Chilled Foods: A Comprehensive Guide

Chilled foods

© 2008, Woodhead Publishing Limited

Related titles: Handbook of hygiene control in the food industry (ISBN 978-1-85573-957-4) Complementing the highly successful Hygiene in food processing, this book reviews recent research on improving hygiene in food processing. Part I considers recent research on contamination risks such as biofilms and how they can be assessed. Part II reviews ways of improving the hygienic design of both buildings and equipment, including clean room technology. The final part of the book discusses ways of improving hygiene practice and management. Foodborne pathogens: hazards, risk analysis and control (ISBN 978-1-85573-454-8) As trends in foodborne disease continue to rise, the effective identification and control of pathogens becomes ever more important for the food industry. With its distinguished international team of contributors, Foodborne pathogens provides an authoritative and practical guide to effective control measures and how they can be applied in practice to individual pathogens. Managing frozen foods (ISBN 978-1-85573-412-8) This book examines the key quality factors from raw material selection through processing and storage to retail display. It gives a unique overview of this entire industry and provides frozen food manufacturers, distributors and retailers with a practical guide to best practice in maximising quality. Managing frozen foods serves as an invaluable decision-making tool, providing guidance on the selection of raw materials, freezer technology, packaging materials and retail display equipment. The editor concludes the book with an insight into the future of the industry and examines the opportunities offered by recent developments such as anti-freeze proteins and ultrasonic techniques. Details of these books and a complete list of Woodhead’s titles can be obtained by:

• visiting our website at www.woodheadpublishing.com • contacting Customer Services (e-mail: [email protected]; fax: +44 (0) 1223 893694; tel.: +44 (0) 1223 891358 ext.130; address: Woodhead Publishing Ltd, Abington Hall, Granta Park, Great Abington, Cambridge CB21 6AH, England)

© 2008, Woodhead Publishing Limited

Chilled foods A comprehensive guide Third edition Edited by Martyn Brown

WPTF2005

CRC Press Boca Raton Boston New York Washington, DC

Cambridge England © 2008, Woodhead Publishing Limited

Published by Woodhead Publishing Limited, Abington Hall, Granta Park, Great Abington, Cambridge CB21 6AH, England www.woodheadpublishing.com Published in North America by CRC Press LLC, 6000 Broken Sound Parkway, NW, Suite 300, Boca Raton, FL 33487, USA First published 2008, Woodhead Publishing Limited and CRC Press LLC © 2008, Woodhead Publishing Limited The authors have asserted their moral rights. This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. Reasonable efforts have been made to publish reliable data and information, but the authors and the publishers cannot assume responsibility for the validity of all materials. Neither the authors nor the publishers, nor anyone else associated with this publication, shall be liable for any loss, damage or liability directly or indirectly caused or alleged to be caused by this book. Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, microfilming and recording, or by any information storage or retrieval system, without permission in writing from Woodhead Publishing Limited. The consent of Woodhead Publishing Limited does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific permission must be obtained in writing from Woodhead Publishing Limited for such copying. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation, without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. Library of Congress Cataloging in Publication Data A catalog record for this book is available from the Library of Congress. Woodhead Publishing ISBN 978-1-84569-243-8 (book) Woodhead Publishing ISBN 978-1-84569-488-3 (e-book) CRC Press ISBN 978-1-4200-8775-8 CRC Press order number WP8775 The publishers’ policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp which is processed using acid-free and elementary chlorine-free practices. Furthermore, the publishers ensure that the text paper and cover board used have met acceptable environmental accreditation standards. Typeset by Ann Buchan (Typesetters), Middlesex, England Printed by TJ International Limited, Padstow, Cornwall, England

© 2008, Woodhead Publishing Limited

Contents

Contributor contact details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii 1

Introduction to chilled foods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 M. Brown, mhb Consulting, UK 1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 UK-produced chilled foods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.3 Market and growth drivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.4 Product trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.5 Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1.6 Niche consumers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1.7 Food service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.8 Supply chain management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.9 Material storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 1.10 Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 1.11 Cooking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 1.12 Cooling and chilled storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 1.13 Quality assurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 1.14 New tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 1.15 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

Part I 2

Raw materials and products

Raw material selection: fruit, vegetables and cereals . . . . . . . . . . . . . 25 D. Barney, Bakkavor Ltd, UK and L. Bedford, Campden and Chorleywood Food Research Association, UK (retired) 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.2 Basic principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

© 2008, Woodhead Publishing Limited

vi

Contents 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10

3

4

5

Selection criteria – product quality . . . . . . . . . . . . . . . . . . . . . . . . Selection criteria – supply chain . . . . . . . . . . . . . . . . . . . . . . . . . . Codes of practice and assurance schemes . . . . . . . . . . . . . . . . . . . Raw material reception and handling . . . . . . . . . . . . . . . . . . . . . . Raw material assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Storage conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Future trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

28 32 37 38 38 39 39 41

Raw material selection: dairy ingredients . . . . . . . . . . . . . . . . . . . . . . B. T. O’Kennedy, Moorepark Food Research Centre, Ireland 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Milk composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Milk-based fresh ingredients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 Milk-based dry ingredients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5 Functionality of dairy ingredients . . . . . . . . . . . . . . . . . . . . . . . . . 3.6 Chilled food production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.7 Quality criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.8 Allergen issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.9 Future trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.10 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

42

Raw material selection: meat and poultry . . . . . . . . . . . . . . . . . . . . . . S. James and C. James, Food Refrigeration and Process Engineering Research Centre (FRPERC), UK 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Sources of supply . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Hazards (microbiological and non-microbiological) . . . . . . . . . . 4.4 Influence of live animal on meat quality . . . . . . . . . . . . . . . . . . . . 4.5 Influence of slaughter and processing conditions on meat quality 4.6 Transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.7 Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.9 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

42 42 43 51 54 56 57 58 58 59 61

61 63 63 64 69 75 76 78 79

Raw material selection: fish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 L. Jack, Sea Fish Industry Authority (Seafish), UK and B. Read, formerly of Seafish, UK 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 5.2 The retail sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 5.3 The supply chain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 5.4 Microbiology of seafood and seafood products . . . . . . . . . . . . . 101 5.5 Quality changes in seafood and seafood products . . . . . . . . . . . 106 5.6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

© 2008, Woodhead Publishing Limited

6

7

8

Contents

vii

Non-microbiological factors affecting quality and safety . . . . . . . . . H. M. Brown and M. N. Hall, Campden and Chorleywood Food Research Association, UK 6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Characteristics of chemical reactions . . . . . . . . . . . . . . . . . . . . . 6.3 Chemical reactions of significance in chilled foods . . . . . . . . . . 6.4 Characteristics of biochemical reactions . . . . . . . . . . . . . . . . . . . 6.5 Biochemical reactions of significance in chilled foods . . . . . . . . 6.6 Characteristics of physico-chemical reactions . . . . . . . . . . . . . . 6.7 Physico-chemical reactions of significance in chilled foods . . . 6.8 Non-microbiological safety issues of significance in chilled foods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.9 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.10 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

109

Chilled foods packaging: an introduction . . . . . . . . . . . . . . . . . . . . . D. Dearden, Unilever, UK 7.1 Demands of chilled food packaging . . . . . . . . . . . . . . . . . . . . . . 7.2 Packaging material selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3 Packaging material substrates . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.4 Moisture condensation or fogging . . . . . . . . . . . . . . . . . . . . . . . 7.5 Packing and filling technology . . . . . . . . . . . . . . . . . . . . . . . . . . 7.6 Pack formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.7 Labelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.8 Future trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Modified atmosphere and active packaging of chilled foods . . . . . . B. P. F. Day, Food Science Australia, Australia 8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2 Requirements of chilled food packaging materials . . . . . . . . . . . 8.3 Chilled food packaging materials . . . . . . . . . . . . . . . . . . . . . . . . 8.4 Packaging techniques for chilled food . . . . . . . . . . . . . . . . . . . . 8.5 Modified atmosphere packaging . . . . . . . . . . . . . . . . . . . . . . . . . 8.6 Active packaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.7 Vacuum packaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.8 Future trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.9 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

109 110 110 115 117 122 122 126 132 132 138 138 139 140 145 146 149 151 152 158 158 158 160 162 163 176 182 183 185

Part II Technologies and processes in the supply chain 9

Microbiological hazards and safe design . . . . . . . . . . . . . . . . . . . . . . M. Brown, mhb Consulting, UK 9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3 Raw materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

© 2008, Woodhead Publishing Limited

191 191 196 196

viii

Contents 9.4 9.5 9.6 9.7 9.8 9.9 9.10 9.11 9.12

Chilled foods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Safety and quality control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Manufacturing areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The microbiological hazards . . . . . . . . . . . . . . . . . . . . . . . . . . . . Risk classes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Safe process design – equipment and processes . . . . . . . . . . . . . Safe process design – manufacturing areas . . . . . . . . . . . . . . . . . Safe process design – unit operations for decontaminated products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Control systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

197 202 206 209 211 214 215 223

10 Non-microbiological hazards and safe process design . . . . . . . . . . . R. W. R. Crevel, Safety and Environmental Assurance Centre, Unilever, UK 10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.2 Chilled foods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.3 Definition and principles of food safety . . . . . . . . . . . . . . . . . . . 10.4 Sources of toxicological hazards . . . . . . . . . . . . . . . . . . . . . . . . . 10.5 Public health significance of non-microbiological hazards . . . . 10.6 Assessment and management of risk from nonmicrobiological hazards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.7 Considerations specific to chilled foods . . . . . . . . . . . . . . . . . . . 10.8 Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.9 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

240

9.13 9.14 9.15

228 231 233 235

240 241 241 242 246 249 256 256 256

11 The hygienic design of chilled food plants and equipment . . . . . . . . 262 J. T. Holah, Campden and Chorleywood Food Research Association, UK 11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262 11.2 Segregation of work zones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264 11.3 Barrier 1 – the factory site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265 11.4 Barrier 2 – the factory building . . . . . . . . . . . . . . . . . . . . . . . . . . 266 11.5 Barrier 3 – high risk production area . . . . . . . . . . . . . . . . . . . . . 274 11.6 Barrier 4 – product enclosure . . . . . . . . . . . . . . . . . . . . . . . . . . . 294 11.7 Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296 11.8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302 11.9 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302 12 Cleaning and disinfection of chilled food plants and equipment . . . 304 J. T. Holah, Campden and Chorleywood Food Research Association, UK 12.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304 12.2 Sanitation principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305 © 2008, Woodhead Publishing Limited

Contents

ix

Sanitation chemicals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sanitation methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sanitation procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Evaluation of sanitation effectiveness . . . . . . . . . . . . . . . . . . . . . Sanitation management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

309 318 323 325 330 334 335

13 Operation of plants manufacturing chilled foods . . . . . . . . . . . . . . . M. Brown, mhb Consulting, UK 13.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.2 Supply chain structure and operation . . . . . . . . . . . . . . . . . . . . . 13.3 Building location and layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.4 Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.5 Process stages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.6 Safety measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.7 Systems for controlling and monitoring the supply chain . . . . . 13.8 Stock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.9 Logistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.10 Future trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.11 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

341

12.3 12.4 12.5 12.6 12.7 12.8 12.9

14 Refrigeration, storage and transport of chilled foods . . . . . . . . . . . . S. James and C. James, Food Refrigeration and Process Engineering Research Centre (FRPERC), UK 14.1 Introduction: the cold-chain . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.2 Principles of refrigeration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.3 Chilling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.4 Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.5 Transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.6 Retail display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.7 Specifying refrigeration systems . . . . . . . . . . . . . . . . . . . . . . . . . 14.8 Modelling and simulation to improve cold-chain management . 14.9 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.10 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Temperature monitoring and measurement . . . . . . . . . . . . . . . . . . . J. A. Evans, Food Refrigeration and Process Engineering Research Centre (FRPERC), UK and M. L. Woolfe, Food Standards Agency, UK 15.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15.2 Legislation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15.3 Importance of temperature monitoring . . . . . . . . . . . . . . . . . . . . 15.4 Principles of temperature monitoring . . . . . . . . . . . . . . . . . . . . . 15.5 Temperature monitoring in practice . . . . . . . . . . . . . . . . . . . . . . © 2008, Woodhead Publishing Limited

341 344 352 356 359 366 367 369 371 373 373 375

375 375 378 388 388 391 394 396 399 400 404

404 404 409 410 412

x

Contents 15.6 15.7 15.8 15.9 15.10 15.11

Equipment for temperature monitoring . . . . . . . . . . . . . . . . . . . . Temperature and time–temperature indicators . . . . . . . . . . . . . . Radio frequency identification tags . . . . . . . . . . . . . . . . . . . . . . . Temperature modelling and control . . . . . . . . . . . . . . . . . . . . . . Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

424 434 438 439 440 440

Part III Microbiological hazards 16 Chilled foods microbiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S. J. Walker and G. Betts, Campden and Chorleywood Food Research Association, UK 16.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.2 Why chill? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.3 Classification of growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.4 The impact of microbial growth . . . . . . . . . . . . . . . . . . . . . . . . . 16.5 Factors affecting the microflora of chilled foods . . . . . . . . . . . . 16.6 Spoilage micro-organisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.7 Pathogenic micro-organisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.8 Temperature control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.9 Predictive microbiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.10 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.11 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Predicting the behaviour of micro-organisms in chilled foods . . . . P. McClure and A. Amézquita, Safety and Environmental Assurance Centre, Unilever, UK 17.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.2 Predictive microbiological models: experimental design/ set-up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.3 Availability of predictive microbiology models for chilled foods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.4 Modelling of heating and cooling processes . . . . . . . . . . . . . . . . 17.5 Quantitative microbiological risk assessment . . . . . . . . . . . . . . . 17.6 Recent developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.8 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Conventional and rapid analytical microbiology . . . . . . . . . . . . . . . J. McClure, Safety and Environmental Assurance Centre, Unilever, UK 18.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.2 Sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.3 Analytical microbiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.4 Conventional microbiological techniques . . . . . . . . . . . . . . . . . . © 2008, Woodhead Publishing Limited

445

445 446 446 448 449 454 458 465 465 470 470 477

477 478 491 512 522 526 527 528 545

545 546 548 551

18.5 18.6 18.7 18.8 18.9 Part IV

Contents

xi

Rapid methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Identification and characterisation procedures . . . . . . . . . . . . . . Future trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sources of further information and advice . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

555 564 566 568 568

Safety and quality management

19 Shelf-life of chilled foods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 573 C. M. D. Man, London South Bank University, UK 19.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 573 19.2 Safety of chilled foods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 575 19.3 Product factors affecting shelf-life . . . . . . . . . . . . . . . . . . . . . . . 580 19.4 Intrinsic product properties affecting shelf-life . . . . . . . . . . . . . . 582 19.5 Extrinsic factors affecting shelf-life . . . . . . . . . . . . . . . . . . . . . . 585 19.6 Interaction between intrinsic and . . . . . . . . . . . . . . . . . . . . . . . . . . . extrinsic product factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 588 19.7 Determining product shelf-life . . . . . . . . . . . . . . . . . . . . . . . . . . 588 19.8 Future trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594 19.9 Sources of further information and advice . . . . . . . . . . . . . . . . . 595 19.10 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596 20 Sensory quality and consumer acceptability . . . . . . . . . . . . . . . . . . . D. Kilcast, Consultant, formerly with Leatherhead Food International, UK 20.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.2 Consumer requirements for sensory quality . . . . . . . . . . . . . . . . 20.3 Components of sensory quality . . . . . . . . . . . . . . . . . . . . . . . . . . 20.4 Techniques for quality assessment . . . . . . . . . . . . . . . . . . . . . . . 20.5 Analytical test methods using trained sensory panels . . . . . . . . . 20.6 Hedonic/affective testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.7 Benchmarking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.8 Maintaining sensory quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.9 Taints and off-flavours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.10 Instrumental methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.11 Future trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.12 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Management of product quality and safety . . . . . . . . . . . . . . . . . . . . C. Thomas, Consultant, UK 21.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.2 Application of management and quality procedures in business 21.3 The basic need for a quality system . . . . . . . . . . . . . . . . . . . . . . 21.4 HACCP and the quality system . . . . . . . . . . . . . . . . . . . . . . . . . . © 2008, Woodhead Publishing Limited

599

599 601 601 605 607 612 613 613 615 615 616 618 620 620 623 624 625

xii

Contents 21.5 21.6 21.7 21.8 21.9 21.10 21.11 21.12 21.13

Quality control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Quality assurance and food standards . . . . . . . . . . . . . . . . . . . . . Validation of the HACCP system . . . . . . . . . . . . . . . . . . . . . . . . Verification of the HACCP system . . . . . . . . . . . . . . . . . . . . . . . Traceability as a part of quality management . . . . . . . . . . . . . . . Allergen management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Business quality models and quality system techniques . . . . . . . Useful organisations and websites . . . . . . . . . . . . . . . . . . . . . . . Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

627 628 629 630 631 631 633 635 635

22 Legislation and criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . K. Goodburn, MBE, Chilled Food Association, UK 22.1 What are chilled foods? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.2 Food laws and international trade . . . . . . . . . . . . . . . . . . . . . . . . 22.3 National approaches to legislation . . . . . . . . . . . . . . . . . . . . . . . 22.4 Microbiological criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.5 Self-regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.6 References and bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . .

637

© 2008, Woodhead Publishing Limited

637 637 646 651 653 653

Contributor contact details (* = main contact)

Editor and Chapters 1, 9 and 13 Martyn Brown mhb Consulting 46 High Street Harrold Bedfordshire MK43 7DQ UK Email: [email protected] [email protected]

Chapter 2 David Barney* Bakkavor Ltd West Marsh Road Spalding PE11 2BB UK Email: [email protected] Lynn Bedford formerly of Campden and Chorleywood Food Research Association

Chapter 3 Brendan T. O’Kennedy Moorepark Food Research Centre © 2008, Woodhead Publishing Limited

Fermoy Co. Cork Ireland Email: [email protected]

Chapters 4 and 14 Steve James* and Christian James Food Refrigeration and Process Engineering Research Centre (FRPERC) University of Bristol Churchill Building Langford North Somerset BS40 5DU UK Email: [email protected] [email protected]

Chapter 5 Lorna Jack* Sea Fish Industry Authority (Seafish) 18 Logie Mill Logie Green Road

xiv

Contributor contact details

Edinburgh EH7 4HS UK Email: [email protected] Brigitte Read formerly of Seafish

Chapter 6 Helen M. Brown* and M. N. Hall Campden and Chorleywood Food Research Association Station Road Chipping Campden Gloucestershire GL55 6LD UK Email: [email protected] [email protected]

Chapter 7 David Dearden Unilever, UK Savoury Global SC&T Group Brooke House Manor Royal Crawley West Sussex RH10 9RQ UK Email: [email protected]

Chapter 8 Brian P. F. Day Food Science Australia 671 Sneydes Road (Private Bag 16) Werribee Victoria 3030 Australia Email: [email protected] © 2008, Woodhead Publishing Limited

Chapter 10 Rene Crevel Safety and Environmental Assurance Centre Unilever R&D Colworth Sharnbrook Bedfordshire MK44 1LQ UK Email: [email protected]

Chapters 11 and 12 John T. Holah Campden and Chorleywood Food Research Association Station Road Chipping Campden Gloucestershire GL55 6LD UK Email: [email protected]

Chapter 15 Judith A. Evans* Food Refrigeration and Process Engineering Research Centre (FRPERC) University of Bristol Churchill Building Langford North Somerset BS40 5DU UK Email: [email protected] Mark L. Woolfe Food Labelling, Standards and Consumer Protection Division Food Standards Agency

Contributor contact details Ergon House c/o Nobel House 17 Smith Square London SW1P 3HX UK

xv

MK44 1LQ UK Email: [email protected]

Chapter 19 Chapter 16 Steven J. Walker* and Gail Betts Campden and Chorleywood Food Research Association Station Road Chipping Campden Gloucestershire GL55 6LD UK Email: [email protected] [email protected]

Chapter 17 Peter McClure* and Alejandro Amézquita Safety and Environmental Assurance Centre Unilever R&D Colworth Sharnbrook Bedfordshire MK44 1LQ UK Email: [email protected] alejandro.amezquita@unilever. com

Chapter 18 Jan McClure Safety and Environmental Assurance Centre Unilever R&D Colworth Sharnbrook Bedfordshire © 2008, Woodhead Publishing Limited

C. M. Dominic Man Department of Applied Science Faculty of Engineering, Science and the Built Environment London South Bank University 103 Borough Road London SE1 OAA UK Email: [email protected]

Chapter 20 David Kilcast Consultant in Sensory Quality Wilderness Cottage 25 Monks Walk Reigate Surrey RH2 0SS UK Email: [email protected] (formerly with Leatherhead Food International)

Chapter 21 Christine Thomas The Lodge Southend Goxhill North Lincolnshire DN19 7NE UK Email: [email protected] [email protected]

xvi

Contributor contact details

Chapter 22 Kaarin Goodburn, MBE Chilled Food Association PO Box 6434 Kettering NN15 5XT UK Email: [email protected] kaarin.goodburn@pinebridge. co.uk

© 2008, Woodhead Publishing Limited

Preface

Chilled foods include a very wide range of ingredients, processed using a variety of technologies operating under tight control in areas with high hygiene standards. Since the second edition of this book, there have been changes in the industry and in consumer demands; major changes have also affected the sourcing of raw materials and there has been consolidation of manufacturing and retailing. The range of chilled foods connects with all types of consumer, taking account of different demands for convenience, greater authenticity and healthiness, but foods are still either ready-to-eat or require cooking, and all require chilled storage for safety and stability. The key requirements are still good quality and microbiological safety at the point of consumption. Because of the diversity of materials, products and consumers, and rapid rates of innovation, a simplistic approach to product design and manufacture cannot be used: the whole range of modern techniques and knowledge is needed to come up with good products at competitive prices. This book presents current approaches and a range of knowledge on chilled food design and production, with an emphasis on the microbiological aspects. For chilled foods, the microbiological emphasis is usually on safety, application of the scientific principles of preservation and cautious use of safety factors, which has given the industry a good record in this area. Hygienic facilities and a reliable chill chain, although expensive to run, have remained essentials for the chilled food industry. Economies have been sought by sourcing materials globally and this has introduced new risks in familiar materials. Against this background, the importance of using formal, thorough techniques for identifying safety needs, changes in risk and appropriate controls has increased. However, the shelf-life of many products is still limited by microbiological spoilage and in the eyes of consumers there is little difference between safety and quality defects; both are still unacceptable. Spoilage may show as different signs in different raw materials and products – © 2008, Woodhead Publishing Limited

xviii

Preface

noticeable as changes in smell, taste, colour-loss or texture breakdown caused by either the action of micro-organisms or mechanisms natural in the materials. As environmental pressure mounts for increased shelf-life to reduce the energy burden of frequent deliveries and product wastage, so the prevention of spoilage begins to rank with safety. This pressure is likely to create a demand for spoilage modelling techniques, similar to those already existing for microbiological pathogens, and their use will lead to controls, preservation strategies and management techniques with a wider set of objectives, including the prevention or delay of spoilage. As the diversity of designs and controls for chilled food increases and there are developments in processing and packaging technology, so the need for informed management and regulation increases. In Europe, changes in the food hygiene and microbiological criteria legislation have gone some way towards clarifying and harmonizing principles for product safety. The move from measurement of product quality by testing, to its assurance by validated designs controlled by online or by-line parameter measurements, demands a new approach to quality assurance systems and the approval or rejection of products. This change has to come from improved skills and training, not only in the private sector but also in the public service area, and also a need to accept that microbiological information systems can provide universally accepted bases for judging product safety. Contact with the contributing authors has been a pleasure and all have shared their wide experience and shown their expertise. I have been very fortunate to have had their support in updating this input to the science and technology supporting the chilled food industry. Martyn Brown

© 2008, Woodhead Publishing Limited

1 Introduction to chilled foods M. Brown, mhb Consulting, UK

1.1

Background

The chilled food market has moved on since the last edition of this book; change has come in many areas – in product demands and in the sourcing of ingredients and products, and to a lesser extent in the technology available. The industry has changed from being a developing industry to a mature one and its focus is on consolidation, cost-cutting and innovation. Consumers have also moved on in their expectations, which now not only include convenience and quality, but also environmental and ethical considerations. Products continue to evolve (see Table 1.1), meeting consumers’ changing needs and life-styles. This has caused the market to segment, so that its products now span the three major consumer trends – health, convenience and indulgence or gourmet. It is also segmented in another dimension with products targeting specific consumer groups such as the elderly, single-person families and notably children. Chilled foods have been available since the 1960s and continue to be a success story, currently representing about 10% of all UK retail foods by value. The UK chilled prepared foods market in 2007 had a retail sales value of £9.11bn, a rise of 4.4% on 2006. Cheese was the largest sector of the market in 2007, followed by yoghurts and chilled desserts and ready meals (http://biz.yahoo.com/bw/080218/ 20080218005208.html?.v=1). Chilled recipe dishes and sandwiches are the largest and most innovative sectors in the UK. Recipe dishes and ready meals have grown from an estimated £173 million in 1988 to over £1750 million in 2005, based on an estimated 12 000 different dishes based on recipes originating from all over the world. The sandwich market is estimated at £3bn annually. The European © 2008, Woodhead Publishing Limited

2

Chilled foods

Table 1.1

Development of the chilled food market (CFA, 2006)

1960s

1970s

1980s

Sliced meats Pies

Dressed salads Recipe dishes Dairy desserts Quiches Flans Sandwiches Pizzas Ethnic snacks Pastas Soups

1990s

2000s

Non-dairy desserts Dips Salad dressings Sauces Stocks Prepared fruit Prepared vegetables Leafy salads Sandwich fillings

Meal centres and accompaniments Speciality breads Sushi Luxury meal kits Stir fry kits

chilled prepared food market was worth over €14.62bn in 2005 and meals accounted for about 60% by value (http://www.just-food.com/article.aspx?id= 97917). Sales were expected to increase by almost 4% to €15.15bn during 2006, rising to more than €18bn by 2010 (LFI, 2006). Consumers view the chilled cabinet as providing convenience and quality, and suppliers have managed to generate and maintain this perception through rapidly changing portfolios of products whose branding and presentation reinforce it. The UK chilled foods industry has continually built consumer trust, based on its approach to food safety and quality and continuous innovation leading to a wide, and changing, choice of products. Assurance of the microbiological safety of chilled foods remains a key requirement and it has driven product design and the management of the supply chain. Many areas of technology are used to make safe products and achieve product shelf-life. Apart from their culinary effects, these technologies have three purposes:

• to prevent microbial contamination (e.g. primary packaging; hygienic manufacturing facilities and cleaning procedures)

• to restrict microbial growth (e.g. chilled distribution and storage; intrinsic preservation systems)

• to remove (e.g. by washing) or kill (e.g. by heat treatment or fermentation) micro-organisms. Depending on the type of product, some or all of these objectives have to be achieved within the constraints imposed by the product character and the technology available.

1.2

UK-produced chilled foods

The UK is a major market for chilled foods. Consumer demand has kept it growing, which has in turn supported the expansion of manufacturing facilities, ingredient suppliers and specialist logistics companies providing transport, storage and delivery into large distribution centres and then onward to retail outlets within very short time scales (within 24 h) at accurately controlled temperatures. © 2008, Woodhead Publishing Limited

Introduction to chilled foods

3

In the 1960s, the main chilled foodstuffs were meat and dairy products, such as pre-sliced meat, traditional pies and dairy desserts. Very often these were distributed direct from the manufacturer to retail outlets by fleets of vans. Within twenty years the variety has grown to include ready meals or recipe dishes, prepared salads, dressings, sauces and soups, prepared fruit, speciality breads and sandwiches. Retail outlets have become larger and fewer, and distribution is by specialist logistics networks with efficient and uninterrupted cold chains ensuring the quality, shelf-life and microbiological safety of unpreserved products. The 1990s and 2000s, UK consumers are considered cash-rich and time-poor, with increasing numbers of working meal-providers and people living alone. Shoppers have changed purchasing, dining and social habits. The demand for foods that are still within shelf-life when shopping is done weekly, minimize time spent on the preparation, and offer good quality and some variety accounts for the popularity of chilled recipe dishes (http://www.ukinvest.gov.uk/Trends/14/zhTW.html). There are thought to be several reasons for retailers’ success in meeting this need, including wide choice and frequent changes to product ranges, authenticity or ethnicity of dishes and ingredients, and availability of exotic or difficult recipes that are difficult, or time-consuming, to cook successfully. Because of international travel, dining-out and strong media interest in cookery, the attraction of regional and ethnic food is stronger than ever. A growing willingness by consumers to experiment with novel flavours and ingredients has encouraged the development of new ingredients and regional or speciality products. How products are sold has changed; the retailer own-label brands are prominent because retailers have been successful in developing top quality images for their brands, based on innovation and quality. These now account for 95% of the chilled food market and this provides a rich source of income to food manufacturers who can meet their quality and cost requirements. UK chilled food manufacturing is dominated by high volume producers (typical factory about 7000 m2, producing up to 2 million units per week (http://www.themanufacturer.com/uk/profile) with a wide range of products, pack sizes and presentations. In the UK, Chilled Food association (CFA) member companies employ over 56 000 people (including over 1000 scientists), equivalent to 11% of the total food industry workforce. Even within the manufacturers there is increasing outsourcing of production and support services (e.g. analytical services). Predominantly, outsourcing is done to companies that specialize in the key technologies (e.g. heat treatment and packaging) and/or have sufficient capacity to respond quickly to changes in demand. Currently, there are two main trends in chilled food retailing. Firstly, retailer consolidation and, secondly, changes to the types of outlet provided by retailers, so that many large retailers now span the whole range from superstores down to small express stores, which affects the type of service manufacturers and especially logistics providers have to supply. Factories and systems are technically sophisticated because of the need for lowcost manufacture under hygienic conditions and the demands of customers (e.g. major retailers) and consumers for high quality and a robust approach to food safety, high standards of supply chain management, and quality management systems that enforce specified controls effectively (see CFA, 2006). The chilled © 2008, Woodhead Publishing Limited

4

Chilled foods

prepared foods market includes a number of large food groups, such as Dairy Crest Group plc, Geest plc (Bakkavor Invest), Greencore Group plc, Northern Foods plc, Uniq plc, and Young’s Bluecrest Seafood Ltd (see membership of CFA: http:// www.chilledfood.org/Content/Our_members.asp). There is environmental/sustainability pressure to reduce product packaging (see http://www.igd.com/cir.asp?menuid=149&cirid=2354), but any reductions in pack strength or pack type still need to retain the important protective and preservation functions of primary packaging. This layer has to contain the product without leakage, ensure the shelf-life (e.g. prevent moisture migration or contain a modified atmosphere) and provide an attractive presentation format that is suitable for distribution and storage, re-heating and serving (e.g. by microwave or conventional oven) or multi-use. For safety reasons, packs must remain dimensionally stable and temperature-stable during distribution and pre- and post-heating, and this imposes restrictions on the types and quantity of material that can be used. As most products are distributed on pallets and handled in automated systems, secondary (e.g. sleeves or cartons) and tertiary (boxes or overwrapping) packaging also play important roles in protecting the integrity of the primary pack, especially from crushing or puncturing.

1.3

Market and growth drivers

Dairy (e.g. cheese, yoghurts and chilled desserts) remains the largest sector of the chilled food market in size, followed by chilled meats and chilled composite products (e.g. recipe dishes, pizzas, soups and sandwiches). These products can be ready-to-eat, ready-to-re-heat or ready-to-cook, and include multiple components, e.g. fish, meat, dairy, egg and vegetables. Changes in the requirements of EU hygiene legislation in 2004 (covering manufacturing requirements) from individual Directives for each commodity or component and setting out often different requirements, to a single Hygiene Regulation giving over-arching principles with Appendices for each commodity, has made manufacture of composite products easier. The single Regulation (EC) 852/2004 on the hygiene of foodstuffs has removed existing conflicts and anomalies in manufacturing and layout requirements from the Directives. There are also two supplementary Regulations; EC: 853/2004 laying down specific hygiene rules for food of animal origin and EC: 854/2004 laying down specific rules for the organization of official controls on products of animal origin intended for human consumption) (see also http:// www.chilledfood.org/content/guidance.asp). Originally, market growth in the UK and Europe was driven by quality and convenience, which was interconnected with an increase in the number of households with refrigerators and the emergence of supermarkets, whose shorter, faster lines of supply and rapid turnover allowed for the handling of short shelf-life, prepared foods. Since then, growth has been driven further by domestic technology (e.g. availability of microwave ovens) and the social, economic and demographic changes that have influenced eating habits (e.g. decline of formal © 2008, Woodhead Publishing Limited

Introduction to chilled foods

5

meal times) and consequently the types of product bought. Consumers are thus increasingly drawn towards complete products, such as chilled ready meals, pizzas and soups. Evolution of the market continues and large retailers are examining how they can further develop ‘home meal replacements’ to compete with takeaway outlets and delivery networks serving homes and places of work (http:// www.eatwell.gov.uk/healthydiet/seasonsandcelebrations/howweusedtoeat/ 21stfood/). Demand for some types of product, such as ready meals, may dip as consumers are directed in their thinking and behaviour towards health (LFI, 2006), and healthy eating habits catch on and commodity costs continue to rise (see http:/ /www.foodanddrinkeurope.com/news/ng.asp?id=66405-northern-foods-britvicmcdonalds). Whether they serve the everyday market or provide indulgence, brands have to focus on value, nutrition, health and wellness, and respond to consumer demands and fashion. Looked at from a supply chain perspective, the market and its supply chain can be broken down according to the technologies used. Most products are ready-to-eat and a few require cooking prior to consumption. In the latter group, cooking by the consumer is required to transform a raw material into a meal component (e.g. roast meat or poultry), but also to free the material of infectious pathogens (and spoilage micro-organisms), and for these products it is very important that the design of the product, especially its heating characteristics, ensures it gives both good quality and safety when it is heated according to the instructions. For the ready-to-eat group, in-factory cooking, or decontamination (see CCFRA, 1999 and World Health Organization, 1998), with hygienic handling and packing before or after heating, eliminates (or reduces numbers of) micro-organisms in the product, so any heating by the consumer has only to raise it to a temperature for eating. This means that after manufacture the product has to be free of any food-poisoning micro-organisms that can grow during storage. The most serious hazards are salmonella and Listeria monocytogenes; the latter persists in factories (Holah et al., 2004) and is able to grow at chill temperatures. Such products are generally stored, prepared and served in the manufacturer’s container (e.g. tray). Depending on whether factory cooking is done in-pack or out-of-pack with subsequent hygienic product assembly and primary packaging, different factory layouts, services and hygiene precautions are needed (see CFA, 2006). Depending on consumer use and the type of ingredients, processing and packaging, products may be made in areas that meet basic good manufacturing practice hygiene requirements or in areas with higher levels of hygiene, specifically designed to prevent re-contamination – hygienic or high care or high risk areas (see CFA, 2001). UK multi-component chilled prepared foods do not usually contain preservatives and are minimally heat-processed to retain the quality of ‘freshly cooked’ dishes; this often leads to short shelf-lives, limited either by adverse sensory quality or microbiological changes. This focus on short shelf-life products has been criticised as it is thought to increase waste and lead to increased delivery frequencies. The minimal preservation/short shelf-life approach contrasts with the USA, and some other European countries, where more robust products with longer shelflives are customary and hurdle technology based on intrinsic properties (e.g. acid or © 2008, Woodhead Publishing Limited

6

Chilled foods

reduced water activity) is often used to extend shelf-life and reduce the risks of pathogen growth (see, for example, http://www.foodnavigator-usa.com/news/ ng.asp?id=79872-danisco-preservatives-microgard, which advertises clean-label antimicrobial hurdles (cultured dextrose) for prepared meals and selected meats to replace chemical preservatives such as benzoate). In the USA, the FDA classify foods relying on chill storage alone for safety as ‘potentially hazardous food’, but chilled foods are removed from this category if they contain a preservation system capable of inhibiting pathogens (see http://www.cfsan.fda.gov/~comm/ift4-2.html). Similarly in the UK, the Food Standards Agency (FSA) recommends a restriction on the chilled (at 8 °C or less) shelf-life of vacuum and modified-atmosphere packaged chilled foods to protect consumers against the potentially fatal disease of botulism. Although there have been extensive sales of chilled foods without any incidents of food-borne botulism when the food is correctly stored, model studies have shown that these foods can support the growth of Clostridium botulinum. However, the FSA recommend that ‘for short shelf-life foods where other controlling factors are not identified, it is proposed that storage is maintained at ≤8 °C and a shelf-life of ≤10 days, rather than storage at ≤5 °C and a shelf-life of ≤10 days, or storage at 5 °C– 8 °C and a shelf-life of ≤5 days. The shelf-life should commence once the product is first vacuum or modified-atmosphere packed. The shelf-life clock should not be reset if the product is subject to a further packing under vacuum or modified atmosphere. These recommendations have been approved by the Advisory Committee on the Microbiological Safety of Food and will be included in the revised Agency guidance document for industry on vacuum packed or modified atmosphere packed foods’ (see http://www.food.gov.uk/science/research/research info/ foodborneillness/microriskresearch/b13programme/b13list/b13006/b13006r). Chilled processed meats was by far the biggest category in the global market in 2003, accounting for 57 per cent of the total, with chilled fish and seafood products in second place with 27 per cent. Chilled ready meals accounted for 11 per cent of sales, while chilled noodles, pasta soup and pizza globally accounted for around 1– 2 per cent at most. Developed markets, such as Japan, the USA and the UK, can be characterised by the availability of a wide range of sophisticated and complex chilled food products; chilled processed meats and dressings are the only sectors to have achieved significant growth in emerging markets. The development of markets such as Russia and China is limited by traditional patterns of purchasing raw meat and fish, rather than processed dishes, eating habits that are different to western Europeans, low consumer purchasing power and logistical problems inherent in the size of the countries (poor chilled transportation and lack of commercial and domestic chilled storage) (see http://www.euromonitor.com/ Chilled_food_suited_to_the_modern_market).

1.4

Product trends

Most companies and retailers in the chilled prepared foods market have ongoing programmes of new product development (NPD) covering formulation, process © 2008, Woodhead Publishing Limited

Introduction to chilled foods

7

technology and packaging to engage consumer interest. Not only manufacturers and retailers have NPD programmes, but ingredient, packaging and equipment suppliers have similar programmes to provide platforms for their innovations. Very often developments in these sectors provide opportunities for new products. Few of these programmes focus on breakthrough innovation because consumers are conservative in their food choices; by far the majority of innovation is incremental (e.g. uses known technology and focuses on cost or improvement of existing features to add value) and covers repositioning (e.g. moving cost or quality within an existing range or brand) and line extension (e.g. adding dishes with different ingredients, flavours or identities to an existing range). Branding in areas such as healthy eating (including lower salt and fat with healthier fats and oils), premium, organic and ‘free-from’, regional or ethnic and other sectors has also played an important role in volume growth. Growth of functional foods (e.g. food or foods with dietary components that may provide a health benefit) has driven development of a broad range of products ranging from those with a particular functional ingredient (e.g. stanol-enriched spreads), through to everyday foods such as chilled dairy and supplemented dairy products (e.g. yoghurt fortified with a nutrient or ingredient that would not normally be present to any great extent). This is similar to the addition of folic acid fortified bread (see http://www.ific.org/nutrition/functional/index.cfm). This market may expand further as consumers become more ‘health aware’ and concerned about the relationship between what they eat and health. In future, consumers may look beyond foods that simply maintain their health, and seek foods that optimise health and wellness, besides reducing risks of some diseases (e.g. cardiovascular disease: see http://www.igd.com/cir.asp?menuid=35&cirid=1145). No doubt further development will translate successful recipes from the retail market to the growing food service market, as even in 2004 from UK consumer spend on food, retail (36%) and catering (32%) were nearly equal. There is continuing development of convenience-orientated products and functional packaging for microwave heating and individually wrapped products or meal components for different types of preparation or heating. Technology innovations, such as microwave steam cooking (see http://scentsationaltechnologies. com/news/FoodTechMag0105.pdf), are likely to provide new product types. Given active innovation in microwave heating, the challenge for product developers is to manage the dielectric (microwave heating) properties of products (dependent on salt and water levels in fat and protein ) to give even heating of packed products without hot-spots that lead to runaway heating and burning, and leave other parts uncooked or cold. Similarly, pack shape can be used to improve heating. This is better in circular or oval containers, but for shelf-stacking, presentation and minimum cost transport, rectangular containers are preferred. There are opportunities for active packaging that uses susceptors (a light deposition of a microwave active metal, e.g. aluminium or nickel-based alloy, onto the pack surface) to allow controlled heating by altering absorption or transmission of microwave energy to the product and so alter the evenness of cooking to allow local browning or crisping (for example, see http://www.foodproductdesign.com/ articles/0597CS.html). © 2008, Woodhead Publishing Limited

8

1.5

Chilled foods

Health

Increasing health-consciousness among consumers is another factor influencing consumption patterns in developed markets, as media and government-sponsored initiatives draw consumer attention to health and nutrition issues. Although the chilled market has had a period of steady growth, it needs to take account of these changing pressures and demands as sales of some products have already dipped in the UK market (2006), perhaps because of price rises linked to increased raw material costs (e.g. wheat) or perhaps because of wide publicity on the need for healthy eating. During the last 25 years, rates of obesity and overweight have quadrupled in the UK. In England, 22% of men and 23% of women were classified as clinically obese in 2002, while 43% of men and 34% of women were overweight (http://www.esrcsocietytoday.ac.uk/ESRCInfoCentre/facts/index55.aspx). This has been attributed to lifestyle changes, including diet, and the prepared food industry cannot avoid being linked with this. Hence, these health concerns are directing the market towards products with a healthier image (see http:// www.foodanddrinkeurope.com/news/ng.asp?id=71948-marks-and-spencer-s-eatwell-health) and also drive the diet and health food sectors, giving a market expected to reach £5.3bn by 2009 (http://s3.amazonaws.com/foresight/20.pdf). Given that a healthy diet includes plenty of fruit and vegetables, bread, rice, potatoes, pasta and other starchy foods, some milk and dairy foods, some meat, fish, eggs, beans and other non-dairy sources of protein (http://www.food.gov.uk/ multimedia/pdfs/publication/goodlife.pdf and http://www.food.gov.uk/multimedia/pdfs/bghbooklet.pdf), multi-component prepared foods provide a means to offer consumers optimum dietary combinations and controlled portion sizes. Advice from the FSA on the topic says: ‘Much of the food people eat is in the form of dishes or meals with more than one kind of food in them. For example, pizzas, casseroles, pies, lasagne, spaghetti bolognese and sandwiches are all made with foods from more than one of the five food groups. These are often called combination or composite foods. Many manufactured foods are combination foods. To make healthy choices, people will need to identify the main food items or ingredients in combination foods and think about how these fit with the proportions shown in The Balance of Good Health’. It is an essential challenge for chefs and product developers to apply the product formulation principles to meet this demand and ensure the continued success of the brands they support. UK chilled food manufacturers are aware of recommendations to reduce fat and salt consumption, and already take this into account when developing products and recipes (see http://www.food.gov.uk/healthiereating/salt/saltprogressstatement/).

1.6

Niche consumers

There are trends towards gourmet and ethnic chilled food products, as media attention on food, chefs and cooking, and an increase in international travel makes consumers more willing to experiment with novel flavours and ingredients. © 2008, Woodhead Publishing Limited

Introduction to chilled foods

9

Product segmentation for specific consumer groups (such as the elderly, women and children) also plays an important part in product development and marketing strategies, leading to products formulated and packaged for specific groups. There is debate about whether organic food will become ‘mainstream’ or remain a niche market because of higher prices. Studies into European consumer attitudes towards organic food have found that price was the overriding factor in determining whether consumers bought organic food or not (see http://www.condororganic.org/) and organic food is still less than 10% of the total food spend in Europe.

1.7

Food service

Use of chilled foods by the food service sector is growing, but its materials are still dominated by frozen and ambient ones. The market is worth an estimated £3bn, and is growing at around 4.0% in value and 5.0% in volume as more people eat away from home. Chilled foods, as yet, have only a small impact, probably because the consumer cannot see the origin of the meal served at the table, and the short shelf-life of chilled ingredients and meals presents limitations to the logistics systems of food service companies. Fresh and chilled foods are largely purchased by restaurants (http://goliath.ecnext.com/coms2/summary_0199-2910253_ITM) because food can be prepared at a central location by chefs, and distributed as refrigerated pouches which are merely reheated on the restaurant premises. Food service has very specific demands for recipe formulation, hot–hold stability, packaging and suitability for high-capacity microwave cooking. Cook–chill processes are often used for the bulk preparation of chilled foods for food service and there are many commercial systems on the market based on equipment (such as form–fill–seal machines for packaging hot-fill foods, and blast coolers or cook– chill production systems, e.g. CapKold®; http://randell.difoodservice.com/Links/ CapKold%20RB.pdf) or packaging systems (e.g. Cryovac; http:// www.sealedair.com/eu/en/library/brochure/ready_meals_06.pdf ). Guidelines on optimum times and temperatures in the cook–chill chain are produced by the British Nutrition Foundation and MAFF (http://www.nutrition.org.uk/upload/ 56606CHI.PDF), emphasising the importance of the various process stages and the total time to consumption.

1.8

Supply chain management

The supply chain can be broken down into five key elements: Plan, Source, Make, Deliver and Return (http://www.quality-foundation.co.uk/pdf/SCM_teaser.pdf and http://en.wikipedia.org/wiki/SCOR-model).

1.8.1 Source Ingredient and packaging supply for chilled foods has undergone similar changes © 2008, Woodhead Publishing Limited

10

Chilled foods

to the rest of the industry; there has been consolidation of manufacturing, and sourcing has developed into a global network for many companies. Ingredients may be sourced locally, nationally and globally and increasingly manufacturers concentrate on collaboration, with fewer, larger, technically competent suppliers and there is greater vertical (e.g. agriculture to pre-processing carried out within a single organization) and horizontal (e.g. alternative suppliers) co-ordination of suppliers to optimize the consistency and availability of materials. Irrespective of origin, input materials and products need to have reliable performance based on specifications and similar controls and monitoring. Providing and enforcing specifications are important operational steps in ensuring food safety and quality, and maintaining customer confidence. Manufacturers are likely to have procedures for supplier approval, including on-site auditing and approval of suppliers’ management processes against recognized technical standards, covering hazard analysis critical control point (HACCP) implementation, analytical procedures and, where necessary, good agricultural practice. To assure food safety throughout the supply chain, close partnerships between suppliers and customers are essential, and safety risks are reduced if selected suppliers are consistent in their harvesting, processing and delivery procedures, and hence in the quality of materials they supply. Sourcing specifications and procedures need to ensure that the right material is available at the right time, in the right quantity and at the right cost. Regulations (e.g. EU hygiene regulations or the US Code of Federal Regulations: Title 21) enforce minimum levels of responsibility for food safety on each stage in the supply chain and clearly make manufacturers responsible for the safety of the materials they supply. The EU General Food Law (Regulation 178/2002) requires entrepreneurs to identify and manage risks, and this means that all companies involved in the food supply chain must be able to identify and track all the ingredients and food that they handle (traceability). But, increasingly, suppliers and large retailers drive the food safety/quality process with their own requirements for higher food safety standards, justifiable label claims, lower conversion costs and maintenance of customer confidence. In some cases this restricts market access for suppliers. All this is equally true whether ready-to-eat products are made, where consumer heating plays no part in eliminating pathogens, or where products sold are for cooking. Where cooking is at least partly responsible for safety, there are risks associated with consumer undercooking. The microbiological specifications for raw materials (e.g. maximum levels of pathogens), and the product design, must ensure that decontamination procedures during processing or during home cooking are always effective, and, in factories, subsequent recontamination is prevented. Global supply chains give cost advantages, but also higher safety and quality risks associated with higher transaction costs, and in emerging markets there is a vulnerability to lapses in food safety (e.g. Sudan 1, http://www.food.gov.uk/news/ newsarchive/2005/feb/update and http://europa.eu/rapid/pressReleasesAction. do?reference=IP/05/385&format=HTML&aged=1&language=EN&gui Language=fr). The benefits of long-distance sourcing are, however, limited for short shelf-life materials, as the costs of rapid transport cut into other cost savings, © 2008, Woodhead Publishing Limited

Introduction to chilled foods

11

perhaps arising from lower harvesting or preparation costs. Similar materials from different sources in any supplier base may not respond in a predictable way to processing and can lead to differences in quality or the way processes run. Therefore, to keep supply chains working well, good development work, functional specifications, transparency in procedures, processes, etc. with adequate documentation, records and traceability of materials are needed. For these reasons, market access is sometimes influenced more by quality assurance processes than product quality. Where short shelf-life products are made on a just-in-time (JIT) basis, uncertainties on quality, safety, availability and processability can often restrict the use of materials from suppliers who offer only a cost advantage. Capability building in the area of process control and quality assurance for suppliers in emerging economies is still an issue, even though they are generally receptive to meeting market access requirements, and in some cases are using assistance from overseas regulators and food companies to build this capability and the accompanying capacity and skills for analytical work. From a manufacturing point of view, the importance of reliable sources of supply increases as more manufacturing plants work on a JIT basis, allowing little time for the completion of analytical work before material is needed and no cushion of material to compensate for any that is rejected. Therefore, global sourcing increasingly requires the long-term supply of safe and consistent quality raw materials. In response to this, suppliers operate to specific manufacturer, retailer or trade association (e.g. CFA) codes and there is increasingly rigorous scrutiny of quality assurance certification, including certification by global certification bodies. As the supply of specified materials at the appropriate stage in processing is a critical element in ensuring that final product meets similar hygiene and quality standards irrespective of origin, ISO has proposed a standard for management in the global supply chain (http://www.iso.org/iso/tool_5-04.pdf). This standard (ISO 22000) aims to ensure microbiological food safety by recognizing that food safety hazards may be introduced at any stage of the food chain and adequate controls throughout the food chain are essential, assured by the combined efforts of all the parties in the chain. Sourcing now is not only managed for food safety and adding value, it now needs to take account of ethical considerations, such as sustainability and local or environmental impact, because there is evidence that consumers are negatively affected if non-food safety and environmental aspects are neglected (e.g. sustainability, animal welfare, ethical trading, fair trade and genetically modified organisms). However, as part of the supply chain, it continues to be driven by the need to accommodate cost constraints, short product life-cycles and fast innovation times for new products or technologies. Consumers are increasingly aware of the distance that food has to travel from source to table (food or transport miles) and the energy input that processing and logistics require (carbon footprint). For many food manufacturers, sourcing fresh local ingredients has been a long-term aim, and media and consumer pressure has contributed to moving this up the priorities for supply chain management. In spite of pressure to make chilled foods from locally or nationally sourced raw materials, © 2008, Woodhead Publishing Limited

12

Chilled foods

and therefore reduce the environmental impact of production, there is continued demand for year-round, rather than seasonal, product availability and exotic ingredients (e.g. spices, rice); therefore, overseas sourcing remains essential. The scale of demand and complexity of the global market in foods and ingredients is shown by the presence of a 6500 m2 perishable cargo handling centre at Heathrow with a capacity of 140 000t (see http://www.salvesen.com/case_studyb3a6.html? case_ study_id=25§ion_id=3&sub_section_id=13&article_id=-1).

1.8.2 Make The consumer market demands manufacture of chilled products at low cost with high quality. Against this background, hygiene and safety standards in UK chilled food manufacturers’ premises are amongst the best in the world. Most have high standards of construction and equipment, a full HACCP approach implemented, documented and effective quality management systems, reliable temperature controls with appropriate management of factory hygiene standards, and full traceability of raw materials and products. The CFA ‘Best Practice Guidelines’ outlining this approach form the basis of the European Chilled Food Federation’s operating recommendations for manufacturers, and its principles are part of the British Retail Consortium’s Global Standard for Food (http://www.brc.org.uk/ standards/default.asp). To allow efficient working and rapid dispatch of products, chilled food manufacturers increasingly use computer-based planning, automation and process control systems (e.g. SCADA, MRPII or SAP), many of which incidentally generate quality assurance data. The most extensive systems manage recipes, equipment performance and process parameters (e.g. scaling-up or -down line capacity), operate equipment and correct deviations, monitor performance and collect process data to demonstrate that the specified product quality and safety have been achieved. Advanced systems can provide unified or integrated control over entire processing and logistics chains, and some can base ordering and production scheduling on sales. To minimize waste, production planning systems usually predict customer demand and then manage ingredient sourcing, manufacturing and stock levels accordingly. Management of stock levels of short shelf-life chilled products is critical to providing good service levels and minimizing wastage. There is always a balance between levels of buffer stock needed to cope with fluctuations in demand, and overproduction of stock that may go out of date before delivery because of a mismatch between forecast and actual sales. There is a series of well-defined stages within the manufacture of chilled products and many of the techniques are designed to follow kitchen or chef procedures.

1.9

Material storage

When raw materials and products are in storage, management systems need to focus on maintaining temperature control, storage times and turnover or usage © 2008, Woodhead Publishing Limited

Introduction to chilled foods

13

(e.g. FIFO) to meet specified limits. Monitoring procedures include product and equipment temperature checks, and produce records at the beginning of the traceability trail. Air temperature measurements are often used for monitoring chiller and freezer performance and to control product temperatures, but their reflection of internal product temperature may be affected by how the refrigeration equipment is used, the quantity, type and holding time of products it contains and other operational considerations, such as shift changes. Therefore, careful consideration has to be given to the analysis of records. Ideally, air temperatures should indicate temperatures in the warmest product unit, and to do this it may be necessary to damp the sensor unit, so that it heats and cools at a similar rate to product or to take measurements directly from product. As energy conservation becomes more critical, the importance of good control and accurate, meaningful measurements increases, to ensure that quality and safety are not jeopardised by energy savings. The design of chill stores to meet peak demand can now be based on simulation models to predict maximum capacity and utilization under different production schedules and conditions, and evaluate the impact of changes in batch sizes and production scheduling.

1.10 Processing Of all the supply chain steps, preparation and processing have the greatest variety of activities to be designed, controlled, monitored and recorded. They also are the major users of manual labour and have the major impact on product character, quality and safety. Processing will usually involve several stages, including batching or weighing, mixing, cutting, chopping or slicing, or marinating ingredients to make material with the correct heating, storage and eating properties. If products are heated in-pack, the accuracy of filling exerts a critical effect on heating characteristics. Filling, portioning and pack sealing are critical stages, whether they are done before or after heating. For many of these process stages, prerequisite programmes can be used to control identified hazards and assist in the implementation of food safety and quality management systems. Pre-planning of line operation and run-time needs to minimize quality loss, the creation of food safety hazards, including allergen contamination, and minimize costs arising from down-time. If pre-frozen ingredients are used, maintaining safe temperatures and managing time and hygiene during thawing are essential controls. Use of pre-chilled ingredients to prepare a cold product may assist in maintaining temperature control, but if subsequent process heating stages are involved, then a minimum initial temperature in the materials must be ensured. Hygiene and technical programmes must be communicated to all employees to minimize the potential for mistakes.

1.11 Cooking Cooking, or heat treating, foods is the operational step that usually gives a product © 2008, Woodhead Publishing Limited

14

Chilled foods

its characteristic flavours, textures and colours, and reduces or eliminates microbial contamination. For pumpable products, cooking under specified conditions can be in-line. Pieces, or pieces and liquids, can be loaded as batches into ovens, kettles or fryers. Conveyor or tunnel ovens and fryers may also be used. Cooking has several targets, including inactivation of micro-organisms and enzymes, and development of quality attributes. Cooking conditions to achieve these aims are usually developed as part of the product design, and minimum heat treatments should give good quality and be sufficient to eliminate pathogens from foods sold as ready-to-eat. For technologies using very low heat treatments (e.g. sous vide), microbiological safety relies not only on heating but also on very low contamination levels. Sous vide (and pasteurization) processes (Betts, 1992) do not eliminate bacterial spores and the safety of various product/heat combinations can be established by either challenge or modelling studies (Hyytiä-Trees et al., 2000; Dilaya et al., 2006) that should extend to include the distribution chain. Many types of equipment are available for heating, and innovation continues all the time. Some cookers use direct heating by steam injection or infusion (so that the heating medium becomes part of the product), or use a retained heating medium such as oil (e.g. fryers) or air in contact with the product (e.g. forced convection or high-velocity air ovens). Others heat the product by contact with a heated surface (e.g. continuous heat exchangers or kettles) which is in turn is heated by a heating medium, such as hot water or steam; and others use radiant heat or microwave energy. Heating systems using combinations of methods (e.g. microwaves and steam or hot oil, or convection and steam) are beginning to develop. All these means of heating have different effects on the materials heated, produce different products, and require different controls. The cooking conditions needed to give the product characteristics required, and the product’s heating characteristics, need to be established and validated at the design stage, preferably on a commercial scale, to allow process specifications, parameter values and the HACCP plan to be produced. It is important to recognize that critical limits during heating are likely to be product- and equipment-specific; specifications and work procedures should take care of variations in both. For many large-scale operations, checking the internal temperature of individual product units is not feasible and equipment has to be controlled by process measurements. When this is the case, the scheduled process approach (e.g. heating conditions provide the minimum heat treatment under realistically adverse conditions) should be used. The more uncertainty there is in heating, the greater the cushion of over-processing that has to be built into process conditions, and often this leads to loss of the benefits that a particular technology is claimed to give, such as superior quality or reduced heat processing. In the future, improvements in quality and the availability of products with unique process-derived characteristics will come not only from equipment innovations, but from the better and more effective use of existing equipment; modelling has a big part to play in this. Predictive models for killing rates of micro-organisms (based on D and z values, see http://www.frperc.bris.ac.uk/bugdeath.htm) are used to analyze heating data and optimize cooking conditions, thereby improving quality and yields, and © 2008, Woodhead Publishing Limited

Introduction to chilled foods

15

reducing energy costs with minimum risks of under- or over-processing. Heating rates of packs can be predicted using models for heat transfer to the product and heat penetration within (e.g. CTemp: Centre Temperature Prediction for packaged foods: Tucker et al., 1996). If such models can be incorporated into control systems, they can automatically optimize process conditions in real time, but the specification of materials has to ensure that heating and other characteristics are kept constant.

1.12 Cooling and chilled storage Cooling must rapidly reduce the temperature of food to stop quality change and control growth of micro-organisms surviving heating. During storage, chill temperatures (0–8 °C) must be consistently achieved to ensure quality, safety and shelf-life. Cooling is an energy hungry step and, unless it is done in-line, it can also be labour-intensive. Many different types of cooling and refrigeration units are available, ranging from normal blast coolers or chill rooms with condenser units to high-humidity systems for vegetables and fruits, hydrocooling (where product is in direct contact with chilled water), vacuum cooling and cryogenic cooling systems. All these specialized systems are specific in the products they treat optimally and, in the case of ‘naked’ product being cooled, hygiene and the risks of re-contamination are major considerations (e.g. in spiral systems), to be balanced against increased cooling rates or plant capacity. Again, modelling is beginning to be used to optimize refrigeration systems (Campanone et al., 2002). According to the Carbon Trust, refrigeration (domestic and commercial) represents a significant and growing energy load, currently accounting for 14% of the total electrical energy consumed in the UK. Because of its importance, refrigeration equipment is covered by the Enhanced Capital Allowance (ECA) scheme, which is a key part of the UK Government’s programme to manage climate change, and is designed to encourage businesses to invest in energy-saving equipment, such as refrigerated display cabinets, chillers and refrigeration control systems. (see the ETPL [Energy Technology Product Lists] for refrigeration – http://www.carbontrust.co.uk).

1.12.1 Deliver – logistics and retailing Currently, there is pressure from retailers to extend shelf-life to allow increased efficiency in logistics chains, and this causes manufacturers to examine possibilities (e.g. higher levels of hygiene or more robust processes) to meet this need without sacrificing quality and clean label demands. Chilled foods require temperature-controlled distribution at refrigeration temperatures at or below 8 °C (often around 0 °C), with positive and auditable proof of minimum and maximum temperatures from the point of delivery all the way back to the factory. Time/ temperature data during the entire distribution chain are needed, so that any shelflife and microbiological risks caused by the various stages can be evaluated and © 2008, Woodhead Publishing Limited

16

Chilled foods

corrective actions taken. The detail of the temperature history should indicate the risk of temperature deviations and their potential effects on safety and shelf-life; increasing numbers of measurements increases the costs of collection and analysis. Predictive models for microbial growth can be used to assess microbiological risks (see Koutsoumanis et al., 2005). Efficient land-based distribution systems run distribution vehicles full, the shortest distance in the least time, and deliver products within agreed timescales. To do this, the use of shared-user vehicles or mixed temperature vehicles (zones separated by moveable bulkheads) to optimize efficiency is becoming more common in temperature-controlled distribution networks. To keep logistics chains efficient, supply chain and retailer infrastructure need to be reviewed and operated together, to ensure distances and journeys are minimized. In most cases, the distribution of chilled foods is outsourced to logistics/transport specialists. Their objective is to ensure that the space on delivery vehicles is fully utilized and they will often share vehicles between manufacturers to optimize use of space and reduce the number of journeys. Chilled food storage and distribution are energy expensive and so there is continuing pressure from media and customers to improve efficiency. This pressure is reinforced by increasing energy costs and competition to reduce costs. The impact of food distribution and transport on the environment has been considered in a 2006 DEFRA report by AEA Technology Environment entitled ‘The Validity of Food Miles as an Indicator of Sustainable Development’ (see http://statistics.defra.gov.uk/esg/reports/foodmiles/default.asp). Recognizing the complexities of the food distribution system, the report concluded that although higher levels of vehicle activity lead to higher environmental impacts, it is not just distance that is important but also mode, timing, location and efficiency. The report also concluded that local sourcing does not automatically reduce environmental and social costs, e.g. replacing single trips of large heavy goods vehicles from distribution centres to supermarkets by multiple van deliveries from farmers to shops might possibly worsen congestion of inner-city roads (see http://www.chilledfood.org/Content/Media.asp?id=2971). Generally in the UK and Europe, chilled foods are delivered from manufacturers to retailers’ regional distribution centres (RDCs) and these RDCs then supply product directly to retail stores. RDCs are large, about 65 000 square metres with 100 or more loading bays, 10 000 pallet racking locations and 15 000 order picking locations, operating 24 hours per day, 364 days a year. Many RDCs operate on a wave system where each manufacturer is assigned a timed window for delivery. An RDU may, in turn, deliver to up to 750 stores. For example, in 2005 Tesco had 2365 stores (http://www.bized.co.uk/educators/16-19/business/strategy/activity/ strategic1.htm). Additionally, distribution chains are developing to serve citycentre retail outlets that carry the same ranges of chilled products, and because of their limited size they require more frequent deliveries than large retail outlets.

1.12.2 Display cabinets Maintaining product temperatures below critical values and with a minimum of © 2008, Woodhead Publishing Limited

Introduction to chilled foods

17

temperature fluctuations is the key to maximizing the high-quality display life of chilled foods. Retailers commonly use vertical, multi-deck, open, refrigerated display cabinets to keep products chilled and ensure they are visible from the outside without a physical barrier between the customer and the product. These units may incorporate an integral condensing unit and condensate disposal system or be connected to a remote refrigeration plant. They consume considerable amounts of power and their efficiency depends on their design and location within a store and any heat inputs (such as lighting), compressor and fan efficiency, and their ability to meet load and application requirements (e.g. replenishment frequency and input temperature of products). Cold air from the cabinet falls out of the cabinet causing a considerable demand for its replenishment. This may be reduced by the use of night blinds when there are no customers in the store. Computational fluid dynamic (CFD) modelling has been used to identify the changes that would be required to improve multi-deck display cabinet performance (Foster and Quarini, 2001; Foster et al., 2005).

1.13 Quality assurance End-product testing is of limited value for assuring the safety of short shelf-life products. Chilled food manufacturers focus on real-time ways of assuring quality and safety, and therefore they generally use HACCP systems (see http://www. fsai.ie/publications/haccp/HACCP_TERMINOLOGY.pdf). Most producers have ‘farm to fork’ traceability systems, meeting EU requirements and enabling products to be traced from supplier inputs, through manufacturing (including in-house or supplier analyses), distribution centres, and into retailers or restaurants. It is essential that tracking systems for materials and packaging can be related to particular process conditions, so that problems can be found and assessed for any risks. In the event of a problem, such as contamination, the traceability system must be able to rapidly determine the origin and cause, and transmit information so that other products at risk can be identified and the upstream (producers and suppliers) and downstream (retail outlets) parts of the supply chain and consumers who might be affected can be notified and take corrective actions.

1.14 New tools 1.14.1 New technologies To maintain innovation, dishes and ingredients are drawn from all over the world, with increased emphasis on healthy formulation, ingredients and portioning. New technologies are developing either from a demand for new dishes, where the desired characteristics are known and need to be reproduced on an industrial scale, or from the availability of new equipment or ingredients. Expansion in the demand for healthier and ethnic, regional and gourmet dishes will certainly increase the © 2008, Woodhead Publishing Limited

18

Chilled foods

demand for specialized processing equipment and ingredients to give the flavours and textures associated with particular foods. Because different cooking temperatures, rates of moisture removal, heat transfer and heat penetration give foods their unique characteristics, a great diversity of cooking conditions are used for chilled foods. Utilization of dedicated or specialized equipment is often at variance with supply chain needs for standardized processes and equipment wherever possible. For example, some foods require very high cooking temperatures (e.g. tandoor ovens with temperatures about 400 °C and cooking by radiant heat); other foods require low temperatures for a long time, e.g. sous vide. In sous vide cooking, raw or par-cooked foods such as meat, fish and vegetables are vacuum packaged into laminated plastic pouches or containers, often made from an oxygen-impermeable film, and cooked in water or air at low temperatures (above 60 °C, microbiologically ideally 90 °C), rapidly chilled (within 2 hours to 3 °C) and then reheated for serving or re-packing after a period of chilled storage (up to 21 days at 0–3 °C) (see http://www.techneusa.com/Seal%20Appeal.pdf ). The sous vide heating process, if not adequately controlled, presents significant microbiological hazards (ACMSF, 1992), as microbial spores capable of growing above 3 °C are not inactivated by the low cooking temperatures, which may also provide only limited destruction of vegetative pathogens. Microbiological guidelines are available (Betts, 1992). Many different cooking and preparation technologies are available and well known on the kitchen scale, but the challenge comes from adapting them to largescale (ideally continuous) processes whilst maintaining quality characteristics. For example:

• Marinating may be used to give particular flavours or textures and may be done •

• •

•

industrially by immersion or hot spraying, with or without massaging or tumbling. Large-scale continuous pasta cooking is difficult, as successful operation needs any starch leached from the pasta during cooking to be removed, to stop the pasta becoming sticky. This requires a continuous input of fresh, heated, makeup water or effective treatment of re-cycled water (Korzeniowska et al., 2005). Industrial rice cooking is a complex process as it has to provide different, synchronized stages for washing, soaking, cooking and steaming with, or without, a frying stage. There are many formed or shaped products that require battering, breading, glazing or sauce coating. Frying has been the traditional way of stabilizing or setting these products for chill distribution, but pressure for low-fat products means there is more interest in non-fat or low-fat ways of achieving this, for example by the use of air impingement ovens. Many pastry-based products require specialized baking conditions, for example pizzas. Often ovens are computer controlled, so that the parameters affecting heating can be changed quickly (e.g. air temperature, top or bottom heat). In some ovens, the baking substrate can also be changed between stone, steel plate or band platens to provide speciality products.

© 2008, Woodhead Publishing Limited

Introduction to chilled foods

19

• Stir frying requires high temperature frying with continuous mixing of food particles against the heat exchange surface, quality loss occurs because of physical damage to the pieces, and when the range of residence times in the frying area is wide there are burnt pieces. Hence, prepared mixes are generally sold for home-cooking or cooked in small batches. There have been advances in heating methods, most notably in the use of microwaves, which offer volumetric heating because they can penetrate food materials enabling very rapid heat transfer throughout. This means that the surface of the product does not have to be cooked or heated so that the centre achieves the required heat treatment and hence the volumetric heating is well suited to products with a high volume to surface ratio. Microwave heating equipment now includes batch and tunnel ovens and cylindrical (e.g. pipe-based) in-line heating systems which enable liquid and particulate-containing liquids to be uniformly heated to pasteurizing temperatures on a continuous flow basis (see http://www.industrial microwave.com/foodprocessing.htm). Very often their use is restricted by uneven heating. Pressurized microwave tunnels have been used to heat, hold and cool ready meals in sealed packs and for baking. Microwave hybrid ovens have been used in other industries, but their application in food manufacture is limited (see http://www.microwaveheating.net/anwendungen_en.html). In some cases, product containers use susceptors or specific wave guides that can be installed (in industrial ovens) to redistribute microwave energy and achieve increased uniformity of heating of packs. The most common portion packs are made of microwavetransparent rigid films, such as polypropylene with an ethylene/vinyl alcohol (EVOH) barrier or a polyethylene terephthalate (CPET) film. Robots in the food industry are increasingly used to allow high labour cost economies to compete globally. They generally fall into three categories: pick and place (e.g. loading pieces into trays), material placement (e.g. transporting materials from bulk stores and loading hoppers from large containers) and palletising. They could begin to take a more prominent place if greater cost pressure drives further process automation and replacement of routine jobs currently carried out by manual workers. Robotics holds out the promise of reducing costs and labour requirements, and increasing line capacity but high capital cost and uncertainties over hygiene have made food manufacturers reluctant to introduce the technology to their plants.

1.14.2 Microbiological models Over the past 10–15 years there have been tremendous advances in the predictive modelling of microbial behaviour and this allows estimation of microbial numbers in products with different formulations, under chilled conditions and conditions where the chilled chain fails. Even so, many product developers still follow the approach of counting microbes at different stages of processing and storage to gradually build up assurance of shelf-life and safety. Modelling is a cost- and timeeffective tool, because systematic study of the effects on microbial growth of © 2008, Woodhead Publishing Limited

20

Chilled foods

parameters (e.g. temperature, pH, AW ) over many years has resulted in the general acceptance of limit values for the growth of particular microbes. Even though there have been doubts that such assessments adequately recognize that other factors are important (e.g. storage atmosphere, preservatives and food structure), application of models during the development process may allow the use of less severe limits. Comparison of measured growth rates with predictions from models for the same conditions of pH, temperature and water activity are often sufficiently close to allow evaluation of product formulations and suitable predictive models are available. A joint venture by the Institute of Food Research, the UK FSA and the US Department of Agriculture (USDA) has brought together from each partner, from the literature and other contributing organizations information on how bacteria behave in foods. This database is called Combase and it is available on the internet. It allows easy access to a large volume of organized data on bacterial growth, survival and death under various conditions of temperature, pH, water content and atmosphere. At present it contains around 20 000 growth and survival curves and 8000 records of growth rates (http://wyndmoor.arserrc.gov/combase/). Accompanying Combase, the Food Standards Agency has also launched GrowthPredictor, a package of models enabling predictions to be made about how bacteria may grow in foods under various conditions. Both Combase and GrowthPredictor have been designed to provide those involved in food safety (such as researchers, the food industry and those concerned with public health) with easily accessible information to assist with risk assessment and risk management. ComBase and related software can be found at http://www.combase.cc and GrowthPredictor is accessible at www.ifr.ac.uk/Safety/GrowthPredictor. The USDA Food Safety and Inspection Service and the USDA Agricultural Research Service have produced the Predictive Microbiology Information Portal (PMIP) to assist small food companies in the use of predictive models and food microbiology information. PMIP is a comprehensive website that brings together models, research data and other useful resources that assist in the production of safe foods; for example, predictive models and research data for use in HACCP studies (http:/ /portal.arserrc.gov/).

1.15 References ACMSF (ADVISORY COMMITTEE ON THE MICROBIOLOGICAL SAFETY OF FOOD), 1992. Report

on Vacuum Packaging and Associated Processes. HMSO, London. BETTS G, 1992. The Microbiological Safety of Sous-vide Processing. Technical Manual No.

39. CCFRA (Campden and Chorleywood Food Research Association), Chipping Campden, Gloucestershire, UK. CAMPANONE L A, GINER S A AND MASCHERONI R H, 2002. Generalized model for the simulation of food refrigeration. Development and validation of the predictive numerical method. International Journal of Refrigeration, 25(7), 975–984. CCFRA, 1999. Review of industry practice on fruit and vegetable decontamination, CCFRA Review No. 14, CCFRA, UK CFA, 2001. High Risk Area Best Practice Guidelines. 2nd edition, Chilled Food Association, Kettering, UK. http://www.chilledfood.org/_attachments/Resources/HRBP_2.pdf. © 2008, Woodhead Publishing Limited

Introduction to chilled foods

21

CFA, 2006. CFA Guidelines for Good Hygienic Practice in the Manufacture of Chilled Foods

– 4th edition (The CFA Guidelines), Chilled Food Association, Kettering, UK. DILAYA E, VARGASA J V C, AMICOA S C AND ORDONEZB J C, 2006. Modeling, simulation and

optimization of a beer pasteurization tunnel. Journal of Food Engineering, 77(3), 500– 513. FOSTER A M AND QUARINI G L, 2001. Using advanced modelling techniques to reduce the cold spillage from retail display cabinets into supermarket stores to maintain customer comfort. Journal of Process Mechanical Engineering, Proceedings of the Institution of Mechanical Engineers, 215, Part E, 29–38. FOSTER A M, MADGE M AND EVANS J A, 2005. The use of CFD to improve the performance of a chilled multi-deck retail display cabinet. International Journal of Refrigeration, 28(5), 698–705. HEALTH FOCUS, http://www.foodanddrinkeurope.com/news/ng.asp?id=66405. HOLAH J T, BIRD J AND HALL K E, 2004. The microbial ecology of high-risk, chilled food factories; evidence for persistent Listeria spp. and Escherichia coli strains. Journal of Applied Microbiology, 97(1), 68–77. HYYTIÄ-TREES E, SKYTTÄ E, MOKKILA M, KINNUNEN A, LINDSTRÖM M, LÄHTEENMÄKI L, AHVENAINEN R AND KORKEALA H, 2000. Safety evaluation of sous vide-processed products with respect to nonproteolytic Clostridium botulinum by use of challenge studies and predictive microbiological models. Applied and Environmental Microbiology, 66, 223–229. KORZENIOWSKA J, KORZENIOWSKI D, DEFRANCISCI L AND HOSKINS D R, 2005. Effect of treatment of starchy water on quality of pasta during continuous cooking. Journal of Food Process Engineering, 28(2), 144–153. KOUTSOUMANIS K, TAOUKIS P S AND NYCHAS G J E, 2005. Development of a Safety Monitoring and Assurance System for chilled food products. International Journal of Food Microbiology, 100(1–3), 253–260. LFI, 2006. The European Chilled Prepared Foods Report 3rd Edition, Leatherhead Food International, UK. THE CARBON TRUST, 2006. Fourth International Conference on Predictive Modelling in Foods Refrigeration – Introducing energy saving opportunities for business. The Carbon Trust, Queen’s Printer and Controller of HMSO, UK. TUCKER G S, NORONHA J F AND HEYDON C J E, 1996. Experimental validation of mathematical procedures for the evaluation of thermal processes and process deviations during the sterilization of canned foods. Food and Bioproducts Processing, 74(C3), 140–148. WORLD HEALTH ORGANIZATION, 1998. Surface decontamination of fruits and vegetables eaten raw: A review, WHO/FSF/FOS/98.2. (http://www.who.int/foodsafety/publications/fs_management/surfac_decon/en/index.html).

© 2008, Woodhead Publishing Limited

Part I Raw materials and products

© 2008, Woodhead Publishing Limited

2 Raw material selection: fruit, vegetables and cereals D. Barney, Bakkavor Ltd, UK and L. Bedford, Campden and Chorleywood Food Research Association, UK (retired)

2.1

Introduction

Fruits, vegetables and cereals are present in the vast majority of chilled foods and can comprise up to 100% of the raw materials purchased by some manufacturers of these products. A diverse range of chilled foods are produced and therefore a wide variety of ingredients are used. The complexity of requirements is increased by the different formats in which ingredients might be sourced – for example as primary raw materials (e.g. potatoes or peppers) or in a ‘semi-processed’ form (e.g. diced potatoes or blanched peas) – and the different ways they are treated during processing (see Fig. 2.1). Because of this complexity, it is impossible in this chapter to be specific about raw material selection criteria. Therefore, the following sections attempt to deal with general principles rather than detail and, for the same reason, do not attempt to deal with commercial issues, such as links between price, quality and yield. This lack of detail should in no way be taken as an indication that raw material quality and integrity is of limited importance in the production of high quality chilled foods. As any practitioner knows, the provision of appropriate raw and pre-processed material is essential to the manufacturing process, and without material that is ‘fit for purpose’, chilled foods production would be impossible. In the following sections, the term ‘fit for purpose’ is key; it implies that materials are sourced so that the resultant foods meet all relevant legislative, technical and quality standards. Whilst many of these standards are common to all © 2008, Woodhead Publishing Limited

26

Chilled foods

Fig. 2.1

The prepared produce supply chain.

chilled foods and chilled foods markets, others are specific as demands from particular markets (e.g. for specific claims, such as vitamins, colour, flavour or extended shelf-life).

2.2

Basic principles

In order to manage the sourcing of raw materials for food manufacture, it is necessary to generate a specification for the material required. This specification should describe all of the features required by the manufacturer for the process and product. Whilst some of these features will be determined by the recipe or design of the food, others will be determined technically, often based on the processing equipment available. When selecting a raw material for use in chilled food manufacture, it is necessary to consider three key requirements: (i) The role of the raw material in influencing eating quality and shelf-life. (ii) Process performance during manufacture of the food. (iii) The role of the raw material in determining the safety of the food. In the manufacture of chilled foods, product safety will be determined by proper consideration of the following:

• ingredient hygienic quality (including microbiological, chemical and physical • • • • • •

contaminants) product formulation/characteristics processing parameters intended use of the product storage and distribution conditions manufacturing hygiene shelf-life.

Ingredient hygienic quality is of primary importance; the degree to which raw material can influence food safety will be determined by two key things: (i)

The nature of the raw material itself and the impact that this has on food safety and product integrity.

© 2008, Woodhead Publishing Limited

Raw material selection: fruit, vegetables and cereals

27

• • • • (ii)

Which pathogen(s) and what level of contamination might be expected? Is there a possibility of preformed toxins? Is there a possibility of highly heat-resistant spores? What are reasonable specification levels to apply to minimise risk and meet the product design requirements, and how likely are they to be met by suppliers? The process that will be used to make the product (i.e. what further processing is to be applied – in manufacture or eventual use)?

• Does the process involve an effective heat treatment step appropriate to the shelf-life?

• Are finished products pH controlled? • Is the finished product ready-to-eat, to be eaten following domestic reheating or designed to be cooked? An appreciation of these two issues will permit a risk assessment, or hazard analysis, of the raw material to be carried out. As ever, this involves identifying sources of potential harm, assessing the likelihood that harm will occur and then weighing up the consequences if harm does occur. Certain raw material and process combinations will carry greater risks than others (e.g. potentially pathogen-containing materials for use raw in ready-to-eat products). However, the risk incurred with fruit, vegetable and cereal raw materials can be heavily influenced by location and field production methods. The way in which raw materials are grown/ handled (pre-processed) and delivered will influence their safety. This risk may be controlled by involving quality assurance and development functions in the management of the supply chain.

What is the raw material?

What process will the raw material undergo?

Risk or hazard analysis

Required ‘quality’ performance

Requirements of the supply chain

Specifications

Fig. 2.2

The decision process – raw material specification.

© 2008, Woodhead Publishing Limited

28

Chilled foods

Thus any specification drawn up for product destined for a particular purpose should define two things (see Fig. 2.2): (i) (ii)

2.3

The physical, chemical and microbiological properties of the material required to assure product performance – selection criteria for the application. The ‘technical’ food safety requirements of the raw material and, therefore, the supporting requirements in the supply chain responsible for delivering it – i.e. criteria such as hygiene, handling and storage conditions.

Selection criteria – product quality

Selection criteria for the raw materials chosen for a particular product should include features that are important in determining both its process performance and the ‘organoleptic’ properties of the finished food. Organoleptic refers to any sensory properties of a product involving taste, colour, odour and mouthfeel. These criteria therefore should be considered aspects of raw material ‘quality’. 2.3.1 Material quality When applied to the character of a food, ‘quality’ may refer to: (i) (ii)

the degree or standard of excellence for specified characteristics the suitability for purpose (e.g. processing, shelf-life or cooking/re-heating) or (iii) the consistency of attainment of the specified properties. Because of the wide diversity of chilled foods, and the nature of the processes to which raw materials are subjected, ‘normal’ agricultural and market commodity quality standards will frequently not be the most useful ones and more specific additional attributes (such as retention of crispiness or viscosity after cooking and during storage) may be needed. Many geographical regions have legislative marketing standards that may be helpful in defining target standards for raw materials. For example, EC Marketing Standard No 1148/2001 is enforced in the UK by the Horticultural Marketing Inspection branch of the Rural Payments Agency (see www.agriculture.gov.ie/ crops_ and_plants/marketing_standards/marketing_standards.doc and http://www. food. gov.uk/foodindustry/imports/want_to_import/xfruit_vegx). Traders in fruit and vegetables are required to comply with these standards when grading, labelling and transporting goods. The EC standard prescribes minimum marketing requirements and up to three quality ‘classes’. These may be described briefly as follows: Extra Class – excellent quality and usually refers to specially selected and presented produce. Class I – good quality produce with no important defects. Class II – reasonably good quality, sound but deficient in one or two requirements such as shape, colour, small blemishes and marks. Produce destined for processing operations is not directly covered by these standards. © 2008, Woodhead Publishing Limited

Raw material selection: fruit, vegetables and cereals

29

In many cases, existing legislative standards will not provide sufficient detail or characteristics to properly define the quality required of a raw material. However, the vast majority of a commodity is usually destined for ‘whole’ produce markets, and so the produce used for processing will often be grown, handled and sold in accordance with these standards. In these cases, the classification system will only partially define the raw material quality needed for a particular product application and, if there is not sufficient detail, or key attributes are not covered, then chilled food manufacturers will need to generate their own specifications. In selecting raw materials, first of all it is necessary to consider the factors that contribute to product variability and what makes a raw material suitable for the selected purpose, particularly for use in a chilled food product. Some quality factors are genetically controlled and so varieties may differ in size, shape and other characteristics. But many plant characteristics are additionally influenced by environmental factors during growth and ripening, such as site, climate and seasonal weather patterns. These genetic and environmental aspects interact, contributing to the variability of the produce at harvest.

2.3.2 Agronomic characteristics Traditionally, farmers and growers have selected varieties for their field or ‘agronomic’ performance. They have been concerned to achieve high yields and consistent high quality, these being the major factors affecting profitability, which may be considered in terms of total yield or, more importantly, marketable yield, which is the proportions of a crop meeting specified grading or quality criteria. Variety/cultivar Plant species have recognisable inherited characteristics that distinguish them from other species. Members of a species are generally able to interbreed easily, but much less easily, or not at all, with other species. Within a species, natural variation gives rise to groups of individuals with small but definite differences, which are known as ‘varieties’. When variation is brought about by human intervention, such as plant breeding, then botanists term the products ‘cultivars’. However, in common parlance the term ‘variety’ is used for the man-made products of plant breeding as well. When selecting raw material for particular processes and products, one of the most important criteria under human control is the choice of suitable varieties. There are many different ways in which varieties can differ. Most obvious differences include colour, shape and size. Also there may be differences in field characteristics, such as yield, plant growth habit, or time of ripening and disease resistance. In some cases, sensory characteristics such as flavour and texture may differ. For most crops, a range of varieties can be used for any specific purpose, such as cooking or salad production. A good variety has to meet the requirements of the primary producer, processor, retailer and ultimately the consumer. For example, different varieties of Dutch white cabbage all possess the thick leaf texture and © 2008, Woodhead Publishing Limited

30

Chilled foods

white colour required by the chilled salad producer, but only a few may have resistance to cooking and re-heating. In other cases, the choice of a variety suitable for a specific purpose may be very limited; for instance, an apple variety such as Cox may be specifically selected for a chilled fruit salad because of its skin colour, which will enhance the overall appearance. Shape and size Within crops, there is normally a range of shapes and sizes; for instance, there is a range of types of carrots with different shapes and size of roots. Nantes varieties have cylindrical roots and so are preferred for pre-packing. Berlicum varieties are also cylindrical, but they are generally larger and mature later in the season than Nantes varieties. The other common UK types are the conical Chantenays and Autumn King varieties, which have very large roots tapering to a point. Larger roots of cylindrical or slightly conical varieties are suitable for slicing in chilled food products. A recent development has been the breeding of specific varieties for ‘mini vegetable’ production. The concept is a response to consumer perception that small size equates to high quality and is seen as a way of adding value to vegetable products, such as cauliflower, carrots and peas. The varieties may also require special growing techniques and may be selected according to their ability to produce the correct size for a portion for prepared foods. Cauliflowers are often presented in dishes, ready cut into florets. Varieties differ in the ease with which they can be cut up and the size of florets produced (Bedford and Bond, 1992). For chopping or dicing, large size vegetables are required and the overall yield of prepared product, after wastage, is an important consideration. Large cabbages are specified for processing, such as coleslaw production, and the total yield of ‘cabbage shreds’ is the major concern. This need for large units contrasts with retail sale where smaller varieties are required. Colour and appearance It is often said that the consumer buys by eye, so an attractive colour and appearance, and freedom from defects, is essential for retail sale of whole vegetables. Varieties often vary in colour and this forms the basis of consumer preference; for example, the characteristic colours of different apple varieties make them easily identified. They differ not only in the ground colour (Coxes are yellow and red, Bramleys are green), but also in the markings known as ‘russetting’. Chilled food manufacturers will consider the flesh colour, as well as the skin properties and the ability to resist grey or brown discoloration during storage, when selecting a variety to provide colour to a mixed fruit salad. White cabbage for retail sale needs to have bright colour and fresh appearance, but slight greyness may be acceptable if the cabbage is used for coleslaw, as the colour will be masked by mayonnaise or dressing. Modern commercial carrot varieties are orange and were arrived at by careful selection from a varied ancestral gene pool in which yellow and purple colours were common. More recently, largerooted varieties have been developed for dicing; these have a deeper orange colour, © 2008, Woodhead Publishing Limited

Raw material selection: fruit, vegetables and cereals

31

evenly spread across the core and flesh, giving a good appearance when diced in a product. However, colour may not be so important where the product is chopped or otherwise prepared, for example in a sauce, and under these circumstances other properties may become more important. Flavour and texture Of key importance to the continued development of this sector is the emphasis on quality and safety, as well as value for money. This requires that constant attention is paid to the organoleptic properties of products, which sometimes may be purchased by the consumer after prolonged chilled storage (e.g. up to 42 days). There may also be prolonged storage of vegetable and fruit raw materials in the supply chain before they are made into products. This may be during transport or may be used to compensate for seasonal availability. Preservation techniques may include super-chill or freezing and this can have a detrimental effect on sensory properties. Freeze-induced changes may include loss of colour, overall structure and cellular integrity, loss or migration of water and flavours and some enzymatic or oxidative changes. For frozen materials, storage temperature, minimising temperature fluctuations and the choice of packaging to minimise water loss (e.g. freezer burn), light-induced and oxidative changes, and the rate of thawing prior to manufacture into a product, are critical to product quality. If frozen ingredients rather than fresh ones are used, damage may also be minimised by incorporating cryo-protective compounds into the food (for example by marinating) and some offer protection against the combined deleterious effects of thawing and chilled storage. In products, technological (e.g. provision of barriers) and packaging (e.g. modified atmosphere packaging) solutions have also been adopted by industry to minimise additional adverse effects during product storage. While these are effective, pre-processing itself (e.g. excessive cooking) may also contribute to quality loss. To the consumer, of course, the ultimate requirement is for good eating quality, and food products which have good natural flavour and texture combined with the characteristics of the chosen product. Varieties of many crops, such as apples, have very distinctive flavours and consumers commonly select their own favourites (e.g. Cox, Golden Delicious, Russet). In other crops, many varieties have similar flavours. A consumer would be unlikely to be able to tell the difference between many varieties of, say, iceberg lettuce, but for these crops, other criteria may be used (e.g. texture). On the other hand, flavour has been a particular issue with Brussels sprouts. Varieties differ in levels of bitterness and excess bitterness has caused some varieties to be unacceptable. Flavour and aroma characteristics of different varieties are governed by their chemical constituents and breeders may be able to manipulate these, in order to improve varietal quality. The chemicals responsible for bitterness in Brussels sprouts have been identified as the glucosinolates, sinigrin and progoitrin (Fenwick et al., 1983). In the 1970s and 80s, deep-green coloured varieties appeared on the market, which had been selected for resistance to insect pests. However, taste panel assessments showed these varieties to be bitter and chemical analyses confirmed that they were high in glucosinolates. © 2008, Woodhead Publishing Limited

32

Chilled foods

Some varieties, such as Topline, Rasalon and Lunet, with lower glucosinolate levels, have had consistently good quality over several years (Bedford, 1988). Flavour can vary even within the portion of the crop that is consumed. Both leeks and celery have white tissue at the base of their stems and greener tissue and leaves higher up. There are differences in flavour between the white and green portions (Bedford, 1986). EU regulations dictate the proportions of white and green in celery for sale, and varieties vary in their ability to satisfy this requirement (MAFF, 1996). The texture of a product also contributes to eating quality. This is often related to maturity, with an over-mature product becoming tough. However, there may be variety-linked differences. Potatoes have obvious texture differences, being either waxy or floury. Waxy varieties are firm and will retain their shape after cooking. They are preferred for salad use or for products such as potato scallops. Floury varieties, which soften on cooking, are used for mashed potatoes. Texture is also related to dry matter, usually measured as total solids. Carrots with lower dry matter may be acceptable for uncooked products, but less suitable for cooking and re-heating. Sweetness is an important flavour attribute of most crops but sugar levels can also be important for other reasons. Amounts of reducing sugars in potatoes influence colour after frying and there is a strong variety-related component to this effect (Stanley and Jewell, 1989). Maris Piper is a preferred variety for the chipping trade because it is lower in reducing sugars. Chilled food manufacturers use large volumes of sliced and diced onions to add flavour to their products. Onion varieties differ in pungency. In the UK, the majority of main crop onions are of the Rijnsburger type, which are relatively high in pungency (Bedford, 1984). Some of the varieties grown in Spain are less pungent (mild) and hence ‘Spanish’ onions are used if the onion is to be eaten raw. The American Vidalia type is noted for its sweetness. Flavour variations in tomatoes are related to differences in the amounts of sugars and acids in the fruit. If both are at low levels, the flavour will be bland. Commercial varieties with high levels of both acid and sugar are preferred for production, and these levels are also affected by maturity or ripeness at harvest. Growers consider marketable yield (the proportion of fruit of correct maturity at harvest), average fruit weight, colour, and uniformity of quality measured as °Brix and pH.

2.4

Selection criteria – supply chain

Depending on the nature of a raw material, the processing used, and the type of product, the supply chain requirements will vary. Principle safety concerns relate to the levels (or risk) of physical, chemical and microbiological contamination. For example, if effective controls exist in the processing facility to deal with soil contamination (e.g. by washing or sieving), then this will be less of a concern in the growing, harvesting and handling of that raw material. Equally, if the raw material © 2008, Woodhead Publishing Limited

Raw material selection: fruit, vegetables and cereals

33

is going to be cooked in the manufacturing process, it will not pose the same microbiological risk as a raw material destined for a ‘minimally processed’ or raw application. Consequently, the attention paid to the supply chain and the controls placed upon it should vary according to the application and this should be clearly stated in the product design. In order that an appropriate supply chain specification can be described, a risk assessment or hazard analysis of the raw material supply chain (with reference to the process) should be carried out as part of the HACCP study. As ever, the risk assessment should be an objective process consisting of the following steps: (i) (ii) (iii) (iv)

hazard identification hazard characterisation exposure assessment risk characterisation.

The outputs from this process will determine the realistic hazards, possible controls and the consequence of failure of controls. This will indicate:

• level of traceability required • whether the grower must adhere to any specific agricultural requirements or • • • • •

good agricultural practice what controls over the harvest and handling process are required any special storage/transport requirements whether suitable suppliers and a supply chain can be identified to what extent these requirements are covered by established assurance schemes – and whether these can provide evidence/assurance of supply chain integrity suitable audit and monitoring requirements.

2.4.1 Traceability ‘Traceability’ is the ability to trace and follow food, feed and ingredients through all stages of planting, growing, production, processing and distribution. Consideration should be given to the level of traceability required (the level of supply chain ‘visibility’ required by the processor). An effective traceability system can link a lot or batch with its growing/ production site and any treatment that it has received throughout its growth or after harvest – and so permit rapid access to information regarding the production of that raw material. Where fieldsmen or contractors are used, it may include their identity and training. In a crisis situation, this information can help pin-point the source and nature of a problem. Chilled food manufacturers may wish to specify the level of detail required, where this is above regulatory requirements, and the stages to which the supply chain is ‘traceable’. This will depend on the severity of the risks.

2.4.2 Microbiological risk Pathogenic organisms are widespread in the environment and therefore have the © 2008, Woodhead Publishing Limited

34

Chilled foods

potential to contaminate growing crops. The level to which this represents a risk in foods depends on the nature of the material being grown (and supplied), any process or treatment applied during production of the food product, and the use of the food by a consumer. The following parameters are likely to have an impact on the microbiological ‘fitness for purpose’ of any raw material. Growing location and cropping history Animal or human faeces are the commonest source of contamination with foodborne pathogens. Consequently, consideration should be given to possible faecal contamination of a field, and measures taken to avoid its occurrence. Key questions are:

• Past crop history: Have cattle the grazed land recently and has the land been used for growing non-food crops?

• Proximity to potential sources of contamination: Are crops grown near to stock farms, manure storage facilities or the application of sewage sludge or slurries?

• Does the site have a history of flooding? Use of animal manures and human sewage waste Farmyard manure (FYM) is a valuable source of crop nutrition and organic matter. However, unless thoroughly composted/treated to eliminate them, it can be a source of food-borne pathogens (e.g. salmonella and E. coli). For ‘high risk’ applications (such as salad vegetables), manure heaps should be composted for at least three months prior to use of the material, and the composting process validated for effectiveness. Furthermore, in these cases, any application of FYM should be early enough to ensure the breakdown of organic matter and destruction of pathogens prior to crop establishment. The Chilled Food Association recommends that there is an interval of 12 months between application of any FYM and the establishment of any horticultural crop. The use of untreated (raw) sewage sludge on agricultural land is not permitted. Appropriate care should be taken when permitting the use of treated sludge on any crop destined for chilled food manufacture. Water Water is used in numerous field and glasshouse operations, but notably for irrigation and field applications. Water of poor microbiological quality (e.g. runoff from land used for raising animals) is a potential vector for food-borne pathogens. For ‘high risk’ crop/application combinations, measures should be put in place to limit the possibility for water-borne (faecal) contamination. People and equipment People and equipment harvesting and handling crops have the potential to contaminate raw materials. Depending on the level of risk assessed, the following points may need to be considered:

• training of staff at primary production facilities © 2008, Woodhead Publishing Limited

Raw material selection: fruit, vegetables and cereals

• • • • • • •

35

equipment storage and transport facilities hand-washing and toilet procedures equipment washing and maintenance facilities and procedures protective clothing requirements staff facilities (including toilet and hand-wash provision) separation of staff facilities from production areas staff record keeping.

Allergens The following risks of contamination from the field need to be considered:

• • • •

function of previous cropping (e.g. peanuts in potatoes) function of environment (e.g. nut trees at edge of field) cross-contamination function of people/staff issues (eating in packing houses).

A formal risk assessment or hazard analysis procedures should be used to show the hazards considered. Pesticides Pesticides are used in the production of much of the raw material destined for chilled food manufacture. Whilst this chapter cannot possibly cover the complexity of pesticide legislation, a few general points may be of use. There are essentially two key issues to consider with respect to pesticides used: Approval Essentially, only pesticides approved for a particular crop in a region may be used on that crop. The use of agricultural pesticides in the UK is regulated by a number of pieces of legislation, including the Control of Pesticides Regulations (COPR) and Plant Protection Products Regulations (PPPR). The Pesticide Safety Directorate is responsible for agricultural pesticides. Recent legislation implements a European Directive (91/414/EEC) which regulates ‘plant protection products’ and aims to harmonise the registration of plant protection products across the EU. Global sourcing of ingredients has made it important to establish that similar systems exist in sourcing areas and suppliers should ensure that pesticide approval legislation has been followed wherever the crop has been grown. Maximum residue levels It is illegal to sell produce where pesticide maximum residue levels (MRL) have been exceeded. In the UK, legislation specifies MRLs for more than 28 000 pesticide/commodity combinations covering the more important components of the UK diet. The Pesticide Safety Directorate is responsible for this legislation and for developing it to meet European requirements. MRLs are defined as the maximum concentration of pesticide residue (expressed as milligrams of residue per kilogram of food/animal feeding stuff) likely to occur in or on food as a result of the use of pesticides according to good © 2008, Woodhead Publishing Limited

36

Chilled foods

agricultural practice (GAP), i.e. when the pesticide has been applied in line with the product label recommendations and in keeping with local environmental and other conditions. Where there is an approved use of the compound on a particular crop, the MRL is generally set at a value determined from field trials (i.e. where the crop has been treated with the pesticide and samples of the crop have been analysed to determine residue levels). However, where there is no approved use (or where no residue is expected), a residue level can be set at the limit of determination (LOD). This is an effective ‘zero’ level of residues reflecting the lowest level at which reliable quantitative analysis can be performed. It is important that suppliers are informed of the approved pesticide list and approved procedures by the purchaser, to minimise the chances that a product that has been treated with an approved pesticide in accordance with GAP still fails to meet MRL standards in the country in which it is sold. Suppliers should ensure that pesticide application takes due regard of MRL legislation. Seasonality and crop handling Crop production is greatly influenced by seasonal weather conditions. For optimum crop growth, the right balance of temperature and moisture is required. Not only has this a great influence on crop growth and yield, but it may also affect postharvest quality. In wet conditions, leafy plant material often takes in a lot of water. This results in soft tissue and, if it occurs at harvest, can lead to tissue that bruises easily and has a shorter shelf-life. Once a crop is harvested, its quality cannot be improved. At this stage, the objective must be to maintain the produce in good condition through any short- or long-term storage until it is delivered to the customer. Thus, the ultimate quality and shelf-life of a final product depends not only on growing conditions but also on harvesting and on post-harvest handling and storage. Avoidance of handling damage at this stage, for example during truck transport, is important. Rough handling leads to bruises, which spoil the appearance of produce and can become a focus of infection by spoilage diseases. Bruising can be a major reason for losses of fruit in store, as well as for vegetable crops such as Dutch white cabbage. Maintenance of suitable post-harvest temperature is extremely important to maximise shelf-life, both for produce for immediate use and for that to be stored. Crops continue to respire after harvesting, using up reserves and shortening shelflife through wilting and yellowing. Respiration rate is temperature related and is roughly halved for every 10 °C that the temperature is reduced. The general rule is to remove field heat as quickly as possible after harvest and then to maintain the produce at chill temperature. This is achieved by various methods such as vacuum cooling for lettuce, hydro cooling of carrots and tomatoes, and storage in various types of refrigerated stores. Increasingly, use of the continuous chilled logistics chains aims to retain the produce at the required low temperature during packing and transport to the retailer. Some delicate crops such as lettuce are not suitable for other than very shortterm storage. Others, such as root crops, cabbage and many fruit crops, can be © 2008, Woodhead Publishing Limited

Raw material selection: fruit, vegetables and cereals

37

stored for many months to provide the continuity of supply required by the chilled food manufacturer and customer. Storage may be in field, for example for carrots, which are covered with straw, or in ambient stores for beetroot and other root crops. For longer storage, refrigerated stores are used. Many use moist air cooling, to maintain the desired relative humidity and prevent dehydration, and positive ventilation to draw air through the stored crop. Continuity of supply Produce procurement plans will need to reflect the nature of the material being sourced and the requirement for continuity. For grains and other low moisture content products with a long storage life, it may be possible to use a single source during a 12-month cycle. However, many agricultural raw materials (especially short-term, perishable products) are highly seasonal. Therefore, it is likely that multiple supply sources at different geographical locations are required during the year. This can significantly increase the complexity of supply-chain management and increase the burden on those responsible for supplier assurance. In these cases, there may be significant value in employing the services of a ‘specialist’ producer/ handler who can dovetail seasons and guarantee continuity on the processor’s behalf.

2.5

Codes of practice and assurance schemes

The supply chain requirements for many chilled food applications will be broadly similar, and it is tempting to think that a single, international standard might deliver the required supply chain security for many businesses. However, currently, GAP is defined in many different codes of practice by many different organisations worldwide. The multiplicity of (and variations between) these codes can be confusing for suppliers and purchasing departments. In addition, many food manufacturers and their retail customers will have their own personal requirements which go beyond those of the industry. However, a few schemes are developing an international following and becoming established as benchmarks for GAP. One such scheme is ‘GLOBALGAP’ (formerly known as EUREPGAP, see http:// www.globalgap.org/cms/front_content.php?idcat=2 ). This has established itself as a key player in the global market-place by offering a robust agricultural accreditation scheme in a rapidly growing list of countries – currently more than 80. As the momentum behind GLOBALGAP has developed, existing national or regional farm assurance schemes have been benchmarked against the standard and, where appropriate, recognised as an equivalent. GLOBALGAP comprises of a set of regulations/requirements, together with associated control points and compliance criteria. It is independently audited by accredited certification bodies worldwide. Suppliers achieving GLOBALGAP will be subject to annual inspection. Whilst not necessarily delivering everything required by a chilled foods business, accreditation schemes such as GLOBALGAP are recognised by many © 2008, Woodhead Publishing Limited

38

Chilled foods

retailers internationally and offer a way of providing a level of raw material assurance.

2.6

Raw material reception and handling

At the reception area for production or storage, rapid and efficient unloading is usually desirable – notably with chilled, perishable crops. Depending on the history of the supplier, various types of intake procedure may be used:

• Direct intake of materials from approved suppliers – any monitoring based on audits, history, product risk.

• There may be a raw material testing regime – based on supplier history, risk of

•

• •

contamination of the materials and product usage, certificate of analysis and certificate of conformance covering pesticides and/or microbiological loading. The latter may be used for batch-by-batch approval. There may be quality control checks on incoming raw materials, including sampling and analytical work as part of the intake procedures. Sampling should be based on the definition of a ‘lot’/batch; it should be statistically based according to a sampling plan. Any data on the intake, either from supplier history, certification or quality control testing, should be linked to batch coding and hence to other traceability data forming links with supplier’s data. Allergen issues should be considered.

2.7

Raw material assessment

The aim of raw material assessment on delivery is to ensure that it meets the specifications, including delivery temperatures and type of packaging and the quantity ordered. Often a vegetable raw material (e.g. carrots, bell peppers) will be bought against one specification, but will be used for several products likely to be made using different processes. Unless the performance of materials is well known and not subject to seasonal or origin variations, suitable performance should be confirmed by analytical work (e.g. total solids or moisture content) or by smallscale process studies (e.g. cutting or heating) before the material is taken into fullscale production. If part-processed material is used (e.g. blanched or diced material) then residual enzyme activity or cube size should be confirmed. Increasingly, water consumption for washing and the generation of waste are important issues and the performance of materials in these areas should also be confirmed. If the produce is dirty after harvest, intake procedures should confirm it is clean before intake. The hygiene of the cleaning or washing processes should be established or microbial contaminants will be spread over the produce and will cause rapid deterioration of any damaged items. Where materials are received in palletised packs, then the lots comprising one intake should be noted to ensure traceability is maintained. In some cases with agricultural produce, payment to the © 2008, Woodhead Publishing Limited

Raw material selection: fruit, vegetables and cereals

39

growers will be fixed by the intake assessment. In such cases particular care should be taken to ensure that the characteristics, analytical methods and limit values are agreed and there are clear instructions for the disposal of any rejected material.

2.8

Storage conditions

Irrespective of controlled temperature (e.g. ambient, chill [0 – 3 °C] or frozen) and possibly relative humidity (e.g. 75–95%), storage facilities should be hygienic, free of pests, allow good access to product and protect it from damage (see http:/ /www.fao.org/docrep/V5030E/V5030E00.htm). Storage conditions (e.g. time and temperature) should maintain the required sensory and processing properties of the material; this is done by limiting transpiration (e.g. water loss) and respiration (e.g. metabolism by the living vegetable or fruit). Quality change during storage, including weight loss, is affected by temperature, and the rate of change can be reduced by lowering the temperature to an appropriate level. However, too low a storage temperature will damage stored produce (e.g. by freezing). Once produce is removed from the store for processing, quality change will re-start, maybe at an increased rate (see http://itdg.org/docs/technical_information_service/cold_storage _fruits_vegetables.pdf). The layout and operation of stores should maintain batch integrity for traceability and first-in, first-out (FIFO). The size and positioning of racking, crates or cages should allow free circulation of air. There should also be facilities for returning part-used batches to a store. Stores should have segregated areas (or an electronic system) for identifying rejected material awaiting disposal and product that is not yet approved for use. If different fruits and vegetables are to be stored together, then the compatibility should be confirmed because odour/flavour transfer may cause taint problems and ethylene accumulation may hasten ripening. Where part-processed material is stored in a modified atmosphere (e.g. elevated levels of carbon dioxide) or a preserving brine (e.g. sulphite or ascorbate), then pack integrity should be checked on intake and before use.

2.9

Future trends

2.9.1 Sustainability, food miles and carbon footprint Sustainability (in food and raw material terms) is a complex concept and reflects concern over the long-term environmental impact of agriculture and food supply chains. It covers protection of the environment, responsible use of resources, such as water, and the development and maintenance of economic viability, especially employment, in the regions affected (see http://ec.europa.eu/agriculture/foodqual/ sustain_en.htm). ‘Food miles’ is a term that refers to the distance that food travels from its production to the end-user. It has become widely accepted as a convenient and © 2008, Woodhead Publishing Limited

40

Chilled foods

simple indicator of sustainability, but the question of the sustainability and impact of using the shortest possible supply chains is a continuing issue. If supply chain efficiency is measured in terms of food miles, the following concerns are highlighted because globalisation of the food industry is resulting in increased imports and exports and more food miles being travelled. Globalisation is driven by multiple factors:

• using imports to overcome seasonality and provide constant availability • concentration of the food supply base with fewer, larger suppliers • economies of scale from amalgamating transport requirements by using centralised distribution from large depots

• domination by major supermarket chains and imposition of demanding specifications. Therefore, in practice, procurement teams may wish to favour more local supplies over those from further away. But increased use of locally produced raw materials will, of course, need to consider cost, and availability/seasonality and its impact on the product range offered. However, it is important to note that transport is only one element of sustainability. When considering raw material supply, one should take into account all of the environmental ‘costs’. For example, a DEFRA case study in the UK established that tomatoes grown in Spain (longer transport distance but less energy used in growing) had a lower environmental footprint than tomatoes grown in glasshouses in the UK.

2.9.2 Provenance The provenance of a food ingredient refers to the origin or source from which it comes, and the history of subsequent operations (supply chain). Chilled food markets are seeing an increasing demand for authenticity and traceability of ingredients – with assurance of provenance or origin seen as a key trend. It is possible that selection of raw materials will, more and more, have to consider the story behind their supply. However, any trends in this respect will need to recognise the continuing challenges presented by seasonality and cost, including the costs of any pre-processing – especially if it is done by hand. Procurement teams should consider any limitations imposed by regional identity and the implications of being tied to a restrictive supply market. It is also possible that certain potential ingredients are covered by legislation designed to promote and protect specialist food products. In 1992, the European Union created systems known as PDO (Protected Designation of Origin), PGI (Protected Geographical Indication) and TSG (Traditional Speciality Guaranteed). This legislation has the potential to limit the sourcing options available.

2.9.3 Organics The global market for organic food and drink was worth an estimated £16.7 billion © 2008, Woodhead Publishing Limited

Raw material selection: fruit, vegetables and cereals

41

in 2005. Whilst sales are concentrated in Western Europe and North America, the market for organic food products is growing across the world. This trend means that processors and manufacturers are increasingly entering the organic sector. For example, the number of registered organic processors and importers in the UK had risen by 5% (to 2135) in the 12 months leading up to January 2006. In practice, the term ‘organic’ is one that is strictly defined by law (see EC No 834/2007). It may be used only by producers and manufacturers who are registered with (and approved by) a recognised organic accreditation organisation. In order to become registered, producers must:

• • • •

follow a strict set of guidelines laid down by national and international law segregate effectively organic produce from ‘conventional’ materials keep thorough and accurate records of production processes submit to annual and random inspections.

When manufacturing organic chilled foods, procurement teams will be expected to select raw materials with appropriate and recognised organic certification. Whilst organic standards are overseen by the International Federation of Organic Agriculture Movements, there may be significant national and organisational differences. (Most countries have their own national organic certification body – see http:// ec.europa.eu/agriculture/qual/organic/brochure/abio_en.pdf ). Therefore, it is essential that any raw material certification is accepted locally and that local criteria for organic produce have been met. The simplest way of achieving this is usually to insist on an internationally recognised standard.

2.10 References BEDFORD L V (1984), ‘Dry matter and pungency tests on British grown onions’, J Natn Inst

Agric Bot, (16) 581–91. BEDFORD L V (1986), ‘Sensory quality of white and green portions of drilled and transplanted

leeks’ Processing and Quality Assessment of Vegetables from Trials at Ministry Centres. d, Miscellaneous Autumn and Winter Crops’, CCFRA Technical Memorandum No 413, CCFRA, Chipping Campden, UK. BEDFORD L V (1988), Sensory quality of Brussels sprouts: Survey of varieties 1998, CCFRA Review No 11, CCFRA, Chipping Campden, UK. BEDFORD L V AND BOND S (1992), Quality of fresh market cauliflowers, CCFRA Agrofood Report No 3, CCFRA, Chipping Campden, UK. FENWICK G R, GRIFFITHS N M AND HEANY R K (1983), ‘Bitterness in Brussels sprouts (Brasssica oleracea var. gemmifera): The role of glucosinolates and their breakdown products’, J Sci Food Agric, (34), 73–80. MAFF (1996), EC Quality Standards for Horticultural Produce: Fresh Vegetables, MAFF Publications, London (PB05201). STANLEY R AND JEWELL S (1989), ‘The influence of source and rate of potassium fertiliser on the quality of potatoes for French fry production’, Potato Research (32) 439–46.

© 2008, Woodhead Publishing Limited

3 Raw material selection: dairy ingredients B. T. O’Kennedy, Moorepark Food Research Centre, Ireland

3.1

Introduction

The dairy industry provides a large range of milk-based chilled food products in its own right, as well as fresh and dried ingredients, which can be introduced into chilled foods. These products come mainly from cows, but the milk of other animals is also used for speciality products, e.g. sheep and buffalo. The increased use of dairy ingredients in chilled food production is linked to an increase in the knowledge-base necessary for the production of these modern convenience foods. Many dairy-based chilled foods have a long history (yoghurt, cheeses) and have not apparently changed much over time; their shelf-life and quality changes are well known. However, the increased predictability of dairy product behaviour in less traditional chilled food is mainly as a result of better knowledge of ingredient behaviour in response to changes in pH, ionic strength, temperature and the activity of micro-organisms. In addition, new variants and hybrids of products abound which makes knowledge of the potential interactions between food ingredients a necessary requirement for innovation and the production of good quality chilled foods.

3.2

Milk composition

Milk is a dynamic material and its properties are subject to much variation due to, among other things, cattle breed and age, diet, health, stage of lactation and the environment where the animal lives. However, this white opaque liquid generally © 2008, Woodhead Publishing Limited

Raw material selection: dairy ingredients

43

has a solids level of ~ 13% (w/w). These solids are composed mainly of fat, protein and the milk sugar, lactose. As a mixed colloidal dispersion, milk is extremely stable and can be heated, frozen and dried without any outward sign of destabilisation. In its unprocessed state, milk contains fat globules in the range 1–10 µm in diameter, which on standing form a typical cream layer which may be dispersed by gentle stirring. These globules are stabilised by the natural milk fat globule membrane which is composed of a mixture of proteins, phospholipids, lipoproteins, cerebrocides and sterols. Phospholipids are the most surface-active components (surfactants) and can therefore form a barrier between the fat and the water that the fat globules are suspended in. As their name implies, they contain phosphorous (as a phosphate group). Cerebrocides are glycolipids; as they contain a sugar residue, they have similar behaviour to the phospholipids but do not contain phosphorous. Sterols are highly insoluble in water and associate with phospholipids; the most well-known is cholesterol. The main proteins in bovine milk are casein and whey proteins. Casein exists as discrete spherical particles of 200 nm diameter, generally referred to as casein micelles, and constitutes 2.5–2.8% (w/w) of the milk. The structure of these casein micelles is determined by the interaction between the protein and calcium phosphate. This group of milk proteins is insoluble at the point of net neutral charge (pH 4.6) and this property is of major importance to the structure of some chilled foods. Whey proteins constitute about 0.5% (w/w) of milk and are classified as the proteins that remain in solution after coagulation of the casein through rennet action or by acidification. This group of proteins is heat labile and in certain food applications is intentionally heat-denatured to promote functionality development. Lactose is the major sugar in milk (4.6%, w/w) and is in true solution. It is a reducing disaccharide composed of glucose and galactose, and forms a major source of nutrition for most of the micro-organisms that grow in milk to either ferment it or cause spoilage by acidification or gas production. The main soluble salts comprise potassium, sodium, calcium, magnesium, chloride and phosphate. The organic anion, citrate, is also an integral part of this soluble salt solution. The importance of this soluble salt system in maintaining the integrity and stability of the complex colloidal dispersion cannot be over emphasised. The general composition of milk is outlined in Table 3.1. It is important to note that these are average values and can vary, hence affecting the functionality of the milk. However, it is more important to visualise the various components and their place or function in the complex colloidal dispersion we call milk and the ingredients that can be derived from it.

3.3

Milk-based fresh ingredients

3.3.1 Liquid milk As outlined in Section 3.2, milk is a complex colloidal dispersion which dictates its © 2008, Woodhead Publishing Limited

44

Chilled foods

Table 3.1 Typical levels of the major colloidal and soluble components in bovine whole milk Component

Colloidal

Fat (%) Casein (%) Whey protein (%) Lactose (%) Calcium (mg/100 mL) Magnesium (mg/100 mL) Potassium (mg/100 mL) Sodium (mg/100 mL) Chloride (mg/100 mL) Phosphorous (mg/100 mL) Citrate (mg/100 mL)

Table 3.2

Soluble

4.0–4.4 2.5–2.8 80 3

45 15

0.5–0.6 4.4–4.6 40 8 150 50 100 45 170

Description of pasteurised milk products

Liquid milk type

Comments

Whole milk Homogenised whole milk Standardised whole milk Standardised, homogenised whole milk

Nothing added or removed Homogenised but nothing added or removed Milk standardised to a min. fat content of 3.5% Milk standardised to a min. fat content of 3.5% and homogenised Fat standardised to between 1.5 and 1.8% Fat content of less than 0.1%

Semi-skim milk Skim milk

sensory attributes. The processing of milk prior to packaging and introduction into the chilled chain can determine consumer sensory perception. The whiteness and viscosity of liquid milk is caused by both the size and number of the colloidal entities, which include fat globules and casein micelles. Pasteurisation, UHT heat treatment, homogenisation and fat standardisation of liquid milk are standard processes in the dairy industry and directly affect the whiteness, viscosity, taste and mouthfeel of the liquid milk. A description of the liquid milk types in the chilled cabinets is outlined in Table 3.2.

3.3.2 Yoghurt Yoghurt is a popular chilled dairy product with many variations around the central theme of fermented milk. Yoghurt can also find its way into other products as an ingredient, such as dips, smoothies, desserts and sauces. The fermentation of lactose from milk to lactic acid by lactic acid bacteria reduces the pH from 6.6 to 4.6, resulting in the formation of an acid gel. The main factors governing the formation of acid casein gels as a result of fermentation are casein concentration, pH, and rate of acidification, temperature and ionic strength (Van Vliet et al., 1997). While different strains of Lactobacillus sp. can produce exo-polysaccharides © 2008, Woodhead Publishing Limited

Raw material selection: dairy ingredients

45

which can influence the rheology of the fermented yoghurt, this aspect of structure development will be dealt with elsewhere. Fermentation of unheated milk leads to formation of weak gels on acidification. Heat treatment of milk (e.g. time and temperature) and assurance of the absence of antibiotics are important processing steps during the manufacturing of yoghurt. As the denatured/aggregated whey proteins associate with casein micelles, the stability of the resulting protein particle is reduced and the gelation point is moved to higher pH values (viz. pH 5.4 on heating, pH 5 no heat). The gel properties are significantly stronger when the whey proteins are denatured/aggregated. This is due to a primary aggregation during the heating step and a secondary aggregation during acidification. Syneresis (e.g. separation of water from the solids), or wheying-off, of the yoghurt gel is reduced as a result of the initial milk heat treatment (Dannenberg and Kessler, 1988). This is extremely important for the quality of the yoghurt during storage in the chilled cabinet. It should be noted that milks used in commercial production of yoghurttype products are often supplemented with additional powdered dairy ingredients (e.g. skim milk powder) prior to the heating step, to increase the effective protein content and therefore the final strength of the yoghurt gel. The thermal history of these powders should be in any specification because the producer needs to know the denaturation/aggregation state of the whey proteins in the powder. The formation of acid casein gels has been studied in the presence of a number of different substrates including skim milk (Lucey et al., 1997a) native phosphocasein (Famelart et al., 1996) and sodium caseinate dispersions (Lucey et al., 1997b). While the basic casein composition should be similar in all three substrates, their aggregation states (which will affect both the final gel strength and stability of the fermented product) at neutral pH are different.

3.3.3 Cream Dairy cream is produced for sale, in the chilled food cabinet, with a range of minimum fat contents that are outlined in Table 3.3. Cream also sees application in other chilled food products such as soups, sauces and toppings, and can be produced at various fat contents, depending on the function required. Cream is an oil-in-water emulsion. The milkfat globules in unhomogenised cream have a mean diameter of 3–4 µm. They are stabilised by their natural membrane comprised of phospholipids, lipoproteins, cerebrocides and proteins. The fat itself is composed of a range of triacylglycerols, the distinctive flavour being conveyed by the relatively high concentration of butyric acid. Cream is separated from milk by centrifugal separators which are capable of producing a cream product of 60 to 70% fat and skim milk. Cream is pasteurised between 80 and 95 °C for up to 10 s prior to pumping to the cooling sections of the heat exchanger. These higher temperatures of pasteurisation reduce the increased bacterial load, which occurs as a result of fat globule concentration in the separators, and also controls lipolytic activity. Efficient cooling systems without the introduction of air and significant shear ensure good quality product. There are a number of cream products that have a fat content between 10 and © 2008, Woodhead Publishing Limited

46

Chilled foods Table 3.3

Fat content of various cream products

Coffee cream Whipping cream Single cream Double cream Sour cream Crème fraiche

12% 35% 18% 48% 15–20% 18–35%

20% (Table 3.3). The shelf-life of these creams is limited mainly by the separation of the fat phase during chilled storage to form a cream layer or, under certain circumstances, a cream plug. Homogenisation reduces separation during storage and also varies the viscosity of the product. Cream for whipping, in general should not be homogenised, even at low pressures, as this will affect the functionality. The functionality (whipping ability) of whipping cream is dependent on fat concentration, temperature and the presence of the natural fat globule membrane. Whipping causes partial destabilisation of the cream emulsion. Gradually during whipping, networks of coalesced fat globules build up to surround the air bubbles and form a three-dimensional matrix which provides the foam’s rigidity.

3.3.4 Sour cream Sour cream is produced by the fermentation of cream (18% fat) with selected lactic acid bacteria. The fermentation of lactose to lactic acid reduces the pH of the cream to 4.6, where the casein component in the cream forms a gel. Agitation and pasteurisation of the gel leads to a pumpable product suitable for packaging.

3.3.5 Cheese A large number of cheese varieties exist worldwide, McSweeney (2004) suggesting the number may exceed 1000. Cheeses are generally regarded as chilled foods. Classification of cheese types is difficult and schemes have been based on texture, method of coagulation, and chemical breakdown during ripening or modifications. Cheeses can be made from the milk of many types of animals, e.g. cow, sheep, goat and buffalo. In general, we can view the cheese market as containing extra-hard varieties (Italian Grana types, Pecorino cheeses, Spanish Manchego), Cheddar and related varieties, cheeses with propionic acid fermentation (Emmental, Leerdammer, Gruyere), Gouda and related varieties, pasta-filata types, cheeses ripened under brine (Feta, Domiati), surface mould-ripened cheese (Camembert, Brie), blue cheese (Gorgonzola, Stilton), bacterial surface-ripened cheese, acid-curd cheeses (quarg, cottage), cheeses coagulated by a combination of heat and direct acid addition (ricotta, mascarpone), whey cheese (Mysost), processed cheese (made from natural cheese) and analogue cheese (made from ingredients, dairy or otherwise). The first nine types mentioned above are regarded as ripened cheeses. Chymosin, the major enzyme present in calf rennet, is an endopeptidase, which © 2008, Woodhead Publishing Limited

Raw material selection: dairy ingredients

47

acts very specifically on the stabilising part of the casein micelle. In milk with a pH of 6.7, it cleaves a specific bond of κ-casein, destabilising casein micelles which clot together and form a gel. This is the primary step in most cheese production protocols and therefore the primary purpose of rennet is to concentrate the milk protein/fat globules which are the precursor to cheese. Chymosin acts only on the casein micelles; whey proteins are therefore not present in the curd. To include the whey proteins in the curd, prior heating of the cheese milk is required. The heatinduced denaturation/aggregation of the whey proteins causes them to associate with the casein micelle and be part of the curd. However, this causes impaired clotting of the milk and a weaker curd. It also leads to off-flavour development during ripening. For these reasons, highly heated milk is rarely used for cheese manufacture. However, some cheeses, such as Ricotta and Mysost, are made through heating. Processed and analogue cheese products also undergo extensive heating to produce a plastic mass but these cheeses usually do not contain whey proteins. Manufacture of cheese also requires an acidification step to achieve the texture and taste required. This is achieved using various Lactobacillus species as starter cultures. The basics in ripened cheese production are outlined below. (i) (ii)

Milk is standardised to a set casein:fat ratio. Starter cultures are added to the milk at a set temperature. This sets the rate of acidification of the milk. (iii) Rennet is added after a certain amount of time. This forms the initial gel. (iv) At the appropriate gel strength, the gel is cut and stirred to induce syneresis and release the whey. (v) The whey is pumped out of the vat to leave the pre-cheese. The pH at which whey drainage occurs determines the level of calcium and phosphate in the cheese. The solubility of calcium phosphate increases as the pH decreases. (vi) The curd is allowed to sit, where it develops texture as the pH drops. (vii) The curd is salted (dry salting or by submersion in brine after moulding). (viii) The curd is moulded. (ix) The moulded curd is stored and ripened. The ability for variations around these basic steps explains the large variety of cheeses available in the chill cabinet. In essence, the processing behaviour and texture is intrinsically linked to the interaction between casein, fat, pH, calcium, calcium phosphate and sodium chloride. Superimposed on these physico-chemical interactions is the ability of live starter cultures and their associated enzymes (including residual rennet) to modify the casein matrix, thus influencing the final texture and flavour. Cheddar The main variation of Cheddar cheese manufacture is the ‘cheddaring’ of the drained curds. This process allows time for acidity to develop in the curds (pH decreases from 6.1 to 5.4) and places the curds under some pressure. The curd © 2008, Woodhead Publishing Limited

48

Chilled foods

granules fuse during cheddaring and the curd mass becomes rubbery and pliable. Pasta-filata types At certain pH values (5.2–5.4) the casein/fat calcium/calcium phosphate matrix is capable of stretching at 65–80 °C. This movement of the casein aggregates vis-àvis themselves, at elevated temperatures under shear, is central to the production of pasta-filata type cheeses. Mozzarella (traditionally made from buffalo milk but now more usually from bovine milk) production is central to both the industrial applications in chilled foods such as pizza, and pizza ingredients, supplied to food service outlets. Unsalted milled mozzarella curd is worked mechanically under hot water to obtain its typical structure prior to moulding and cooling. These rectangular blocks are brine salted. Emmenthal Cheeses with a propionic acid fermentation are characterised by the presence of many large ‘eyes’ due to the production of carbon dioxide by propionic acid bacteria. These bacteria metabolise lactate produced by the lactic acid bacteria (also present) from lactose, to propionic acid. While fermentation is initiated at normal temperatures, the variations in the make procedure include a high cooking temperature (54 °C, which denatures most of the rennet), a relatively low level of salt to which the propionic acid bacteria are sensitive, a high pH and a high calcium/casein ratio. This results in an elastic and flexible casein matrix capable of forming ‘eyes’ of trapped carbon dioxide. Gouda This Dutch type cheese is characterised by using citrate-positive cultures (which utilise the citrate in the milk as a food source) and replacement of a portion of the whey with hot water after cutting and stirring. The hot water addition has the effect of cooking the curd and removing some extra lactose, which helps to control acidity development after the curds are moulded. The cultures break down citrate to diacetyl and other volatile flavour compounds. Carbon dioxide is also produced, causing a few small ‘eyes’ in the cheese. Feta Feta types are known as ‘pickled cheeses’ because they are ripened under brine. Fermentation is normal and, after moulding of the curd, the cheese is cut into pieces and salted. Subsequently the pieces are covered in a brine solution (14% NaCl) and ripened at 14–16 °C for a week, until the pH has decreased to 4.5. The pickled cheese is then transferred to rooms at 3–4 °C and stored. At these low pH values and in the presence of brine, lower amounts of calcium are casein bound and the texture is dominated by the ionic strength effect of NaCl on casein–casein interactions. Surface mould-ripened cheese Surface mould-ripened cheeses include Camembert and Brie, soft cheeses © 2008, Woodhead Publishing Limited

Raw material selection: dairy ingredients

49

characterised by the growth of Penicillium camemberti on the cheese surface. The ripening of white-mould cheese is characterised by the extensive breakdown of lactate at the surface of the cheese by the mould, causing an increase in the pH of the surface zone. This in turn creates a pH and lactate gradient from the surface to the centre of the cheese. This sets in motion a movement of some of the components that go to make up the cheese matrix, namely lactate, calcium and phosphate. The casein matrix that goes to make up the structure of the cheese changes from solid to liquid as these multiple reactions progress. Blue cheese Blue cheese varieties are characterised by blue/green veins throughout the cheese caused by the growth of Penicillium roqueforti. Since this mould requires oxygen for growth, the texture of blue cheese must be open to allow the fungal spores to germinate and grow. Lack of pressing in the moulds and physical piercing of the cheeses with needles creates this open structure. Acid curd cheese Acid cheese types are distinguished from yoghurt because their manufacture involves removal of some whey and the absence of heating. A small amount of rennet may be used in certain varieties, such as quarg or cottage cheese, to increase the firmness of the coagulum and minimise losses of casein. Processed cheese products Processed cheese is produced by blending shredded natural cheese of varying maturity with emulsifying salts and other ingredients. The mixture is heated under vacuum with continuous agitation until a homogeneous plastic mass is obtained. Natural shredded cheddar cheese will not form a homogeneous plastic mass on heating. This is due to the inherent negative hydration characteristics of casein matrices in the presence of calcium and calcium phosphate, especially at lower pH values. Emulsifying salts (citrates, phosphates and polyphosphates) act as calcium chelators and adjust the pH of the mixture. The casein matrix hydrates and becomes homogeneous and effectively emulsifies the free fat in the natural cheese. Hence the name ‘emulsifying salts’.

3.3.6 Butter The production of sweet-cream butter or salted butter utilises the phase inversion of the oil-in-water emulsion (cream) to a water-in-oil emulsion (butter). Cooling the cream prior to churning is an important prerequisite in the production of butter. Generally, cream is cooled to 4 °C to allow seeding of the fat crystals to occur, reheated to 18–20 °C for 2 hours, when slow crystallisation of the higher-melting fat components can occur, and finally slowly cooling to the churning temperature (16 °C). Tempered cream is industrially processed in a continuous buttermaker. All Fritz-type continuous buttermaking machines have © 2008, Woodhead Publishing Limited

50

Chilled foods

a number of elements: (i) cream feed pump, (ii) high-speed churning section which ruptures the milk fat globule membranes and forms butter granules, (iii) buttermilk draining, (iv) a working section which smoothes the product, (v) vacuum working which reduces the air content and hence the volume, and (vi) salt, added as a slurry to a final concentration of 1.2–1.6%, all of which is dissolved in the aqueous phase. Salt is incorporated into butter primarily to enhance flavour, but also for microbiological stability. Market butter is packed in foil or tubs which exclude light and are highly resistant to moisture movement. Bulk butter is usually packed in 25 kg cartons after first being wrapped in a suitable film. Uneven colour in butter is generally regarded as a fault and can be classified as streaky, spotted or primrose types (Jebsen, 1994). The ideal body of butter is close, smooth and plastic, so that it readily spreads on bread. The possible body faults include crumbly, sticky, leaky or porous. Lactic butter Lactic butter is produced when fermented sour cream is churned. This usually is achieved using a defined starter cocktail to give a consistent flavour profile. However, this process has a number of disadvantages which include the production of sour buttermilk, which may be difficult to dry, and the pH and flavour may be variable. Alternatively lactic butter can be made in continuous buttermakers by the addition of lactic acid and a starter distillate to sweet cream butter at the point of salt addition. Low-fat dairy spreads Variable-fat dairy spreads, whether of the high, reduced or low fat type, are usually water-in-oil emulsions and appear as plastic solids. Butter and margarine must by law contain a minimum of 80% fat, but spreads conventionally contain 72–80% fat (full fat), 55–60% fat (reduced fat), 39–41% fat (low fat) or 20–30% fat (very low fat), but all are water-in-oil emulsions. The principle ingredients of fat spreads are fat, emulsifier, milk protein, stabiliser, sodium chloride and water, and each of these will affect the emulsion, processing and consumer behaviour of the final product. The level of sodium chloride in the aqueous phase can vary but is usually in the region of 1.5% (w/w). The water-in-oil pre-emulsions of fat spreads are always stabilised by high shear working of the emulsion at low temperatures to a plastic consistency. Before this solidification step, emulsions can become unstable due to either phase separation or phase inversion (Mulder and Walstra, 1974). It is intuitively evident that the likelihood of phase inversion increases as the fraction of added disperse phase is increased. The processing of low-fat spreads comprises two steps, namely preparation of an aqueous phase-in-oil emulsion while stirring, followed by pumping the emulsion through one or two scraped surface coolers in series at a defined agitation rate and at a defined refrigerant temperature (shear at low temperatures, as mentioned above). Patent literature has suggested that the higher the aqueous phase viscosity, the greater stability to inversion (Platt, 1988). Sodium caseinate is often the protein of choice to aid in the stabilisation of the © 2008, Woodhead Publishing Limited

Raw material selection: dairy ingredients

51

water-in-oil emulsion. While the introduction of NaCl into the aqueous phase was initially for organoleptic reasons, the interaction between NaCl and caseinate also has a significant effect on the stability of the emulsion prior to solidification. The viscosity of a caseinate solution is an indicator of the degree of bound water absorbed by the hydrophilic groups, as well as the water trapped inside the aggregated molecules (Korolczuk, 1982). Sodium caseinate contributes to the stability of the water-in-oil emulsion through steric and water binding effects (Keogh, 1992). The same author concluded that NaCl made a significant contribution to the aqueous phase viscosity. While final emulsion stability and its stability to inversion (e.g. becoming an oil-in-water emulsion) may be related to the viscosity of the caseinate-based aqueous phase, the interaction between the level of fat-soluble emulsifier and the aqueous caseinate may also be significant (Barfod et al., 1989). Significant reductions in the level of NaCl in the aqueous phase can lead to inversion problems during processing, and alternative methods of increasing the aqueous phase viscosity may have to be considered, e.g. hydrocolloids, if low-salt spreads are required.

3.4

Milk-based dry ingredients

3.4.1 Whole and skim milk powder Water is the main component of milk (typically 87%) and milk is therefore susceptible to microbial spoilage unless it is preserved; in the case of powder, it is preserved by low water activity (less than 0.2 or 0.3 Aw). The manufacture of many dairy ingredients involves the fractionation of this aqueous colloidal dispersion into its various components and subsequent dehydration. A major family of products in the dairy ingredients portfolio is dried milk, either as whole milk powder (WMP) or skim milk powder (SMP). While WMP is generally standardised to 26% milk fat with separated cream, SMP is produced from milk which has had the cream removed. The skim milk is concentrated by vacuum evaporation to > 45% solids prior to spray drying. The physical properties of the powders (bulk density, particle size, flow properties) are determined by the solids content of the feed to the dryer. The pre-heat treatment of the milk is central to the final functionality of the rehydrated powders, which are generally available as low, medium and high heat powder. The pre-heat treatment can range from 72 °C for 15 s (low-heat powder) to 120 °C for 2 min (high-heat powder) Note that high-heat treatments are performed on the unconcentrated milk (raw milk) prior to evaporation and drying. The major effect of preheating temperature and duration is to alter the level of native whey proteins because they are denatured under severe heating conditions. This results in the formation of a new colloidal particle which is now composed of casein, whey protein and calcium phosphate. These types of dairy products are key ingredients because of their functional, nutritional and organoleptic characteristics in food applications, as well as their storage (they can be stored without refrigeration) and economic versatility. © 2008, Woodhead Publishing Limited

52

Chilled foods

3.4.2 Casein-based ingredients Due to the liquid nature of milk and the physico-chemical differences between the various components, the dairy industry actively fractionates these components in a range of different forms. The solubility differences between casein and the whey proteins, and the inherent pH dependent solubility characteristics of calcium phosphate, make the separation of casein and the production of a range of caseinbased ingredients readily achievable. Caseins are a heterogeneous group of proteins composed of varying levels of αs1, αs2, β and κ-caseins. The interactions of these proteins under the conditions present in milk give casein its distinctive properties. The commercial manufacture of acid casein from skim milk has been reviewed by Mulvihill (1989). It involves acidification, using rapid injection of sufficient mineral acid (HCl) into pasteurised (72 °C for 15 s) skim milk at 30 °C, to alter the pH to 4.4–4.7 (the isoelectric point of the casein). Alternatively, lactic casein can be produced through fermentation of the skim milk, giving a slower coagulation but a more natural process. The pasteurisation temperature/time combination used causes minimal denaturation of whey proteins, preventing their interaction with the casein. Casein, as it exists in fresh skim milk, is in micellar form (particles with diameters in the range 60–300 nm) and, as such, contains large concentrations of colloidal calcium phosphate. At the isoelectric pH, however, colloidal calcium and phosphate becomes soluble and remains with the whey following separation from the casein. Commercially, the casein curd is then washed to eliminate residual lactose, soluble salts and native whey proteins, before being dried. While acid casein itself is functionless, it is the precursor to the production of the highly functional caseinate family. These include sodium, potassium and calcium caseinates, depending on the alkali used to increase the pH of the protein dispersion to a neutral pH (6.5–7.0). These are high protein ingredients, up to 90% protein in dry matter, and have many large applications in the food industry as emulsifiers. Another member of this casein-rich family of ingredients is rennet casein, which is produced using chymosin (rennet) to coagulate the casein micelles at the natural pH of skim milk. This proteolytic enzyme cleaves a crucial fragment of κcasein, which destabilises the casein micelle with the formation of a curd. The curd is washed free of lactose, whey proteins and soluble salts, similar to acid casein production. The major differences between acid and rennet casein are (i) κ-casein is intact in acid casein, (ii) there is a high calcium phosphate content in rennet casein and (iii) acid casein can solubilise on pH adjustment. Rennet casein requires the use of emulsifying salts (citrates, polyphosphates) to initiate functionality, hence their usage in analogue cheese. Phosphocasein, a general term used to describe isolated micellar casein, is generally produced by micro filtration without a precipitation step (Schuck et al., 1994). This results in the elimination of the whey protein, lactose and soluble salt fractions, retaining the micellar casein with its associated colloidal calcium phosphate. Due to the inherent low ionic strength conditions prevailing in phosphocasein dispersions, stability to flocculation at processing temperatures (60–80 °C) is limited (Le Ray et al., 1998). Phosphocasein exhibits physico-chemical and micellar © 2008, Woodhead Publishing Limited

Raw material selection: dairy ingredients

53

behaviour similar to milk in terms of particle size, rennet gelation (Pires et al., 1999) and acid gelation (Famelart et al., 1996). Little is available in the literature on the corresponding properties of casein re-micellised from acid casein. Re-micellisation of acid casein involves pH adjustment to 7.0 with the concomitant reintroduction of the colloidal calcium phosphate. Le Ray et al. (1998) emphasised the determinant role of the aqueous phase, with special reference to salt type and concentration, on the physico-chemical properties of reconstituted casein micelle dispersions in the absence of whey proteins. While stability to heat was a major part of this study, the assay was confined to observations noted after exposure to 95 °C for 30 min. 3.4.3 Whey protein-based ingredients Sweet whey is the liquid by-product of chymosin-coagulated milk, either from cheese production or rennet casein manufacture. It typically has a pH in the region of 6.0–6.4. Acid whey is generally a by-product of cottage cheese or acid casein production and has a pH of about 4.6. The typical compositions of sweet and acid whey are outlined in Table 3.4. The major compositional differences between the different whey streams are the mineral content, lactate concentration and the presence of glycomacropeptide. Acid whey from casein manufacture contains significantly higher levels of calcium, phosphate and chloride (HCl addition). Sweet whey also contains the glycomacropeptide cleaved from κ-casein through the action of chymosin. Cheese whey contains some lactate from the action of the starter cultures. Non-hygroscopic whey powder can be produced through evaporation of the whey to >60% solids, followed by controlled temperature reduction over extended time periods to promote lactose crystallisation, giving a powder protein concentration of ~12% (w/w). The whey protein content of the powders can be increased using ultrafiltration and diafiltration, prior to evaporation and drying. These are Table 3.4

Typical composition of sweet and acid whey

Component

Rennet casein whey

Solids (g/L) pH Protein (g/L) Non-protein nitrogen g/L Lactose (g/L) Lactate (g/L) Ash (g/L) Fat (g/L) Calcium (g/L) Phosphorous (g/L) Magnesium (g/L) Sodium (g/L) Potassium (g/L) Chloride (g/L)

© 2008, Woodhead Publishing Limited

Mineral acid whey

Cheddar cheese whey

66 6.4 6.2

63 4.7 5.8

67 5.9 6.2

0.37 52.3 – 5.0 0.2 0.5 0.46 0.07 0.58 1.45 1.02

0.3 46.9 – 7.9 0.3 1.2 0.63 0.11 0.5 1.4 2.25

0.27 52.4 2.0 5.2 0.2 0.47 0.46 0.08 0.58 1.5 1.0

54

Chilled foods

termed whey protein concentrates (WPC) and can range in protein content from 35 to 80% (w/w). Demineralised whey powders (low mineral content) can be made by ion-exchange and electrodialysis. Whey protein isolates (WPI) are generally produced by ion-exchange and can be up to 92% protein (w/w). The application of these products in chilled foods will depend on the functionality required (e.g. filler, water-binding, viscosity, gelation). Whey powders are used as a source of lactose, but where the more functional whey protein effect is required, the use of WPC may be the choice. Demineralised whey powders are generally used in infant formulations. 3.4.4 Lactose Lactose is a disaccharide composed of D-glucose and D-galactose, and is significantly less sweet than sucrose. It is a reducing sugar and is used in some foods to provide a Maillard reaction-linked browning. It is also routinely added to soup and sauce formulations. It is the major carbohydrate substrate for Lactobacillus spp. during fermentation of milk. Industrially, it is produced from whey which has been concentrated to >60% solids. This forms a supersaturated lactose solution at temperatures >90 °C. On cooling slowly, α-lactose monohydrate crystals are formed which are separated using decanters, washed and dried in fluidised bed driers.

3.5

Functionality of dairy ingredients

3.5.1 Effect of pH, ionic strength, ionic species, temperature Chilled foods, in general, contain a selection of proteins, fat, sugars, salts and water in various proportions. They may also contain other ingredients and additives such as low molecular weight surfactants, stabilisers, colours and flavours. The structure, texture and stability of any chilled food that contains milk proteins will be affected by pH, ionic strength and temperature. Its shelf-life will be directly related to pH and the presence of micro-organisms, highly acid products having a potentially longer shelf-life. This will also be applicable to emulsions stabilised by milk proteins. To put this approach in perspective, casein micelles are unstable to heat and pH reduction when the ionic strength of the aqueous phase is too low (i.e. heated milk protein concentrate solutions will gel at pH 6.5) (Auty et al., 2005). This suggests that the texture of chilled foods from dairy sources or containing dairy ingredients is, to a large extent, dependent on the soluble salt content. Whey protein denaturation and subsequent aggregation are central to texture and structure development in dairy desserts, mousses and yoghurt. In these types of chilled foods, heat-induced whey protein denaturation and aggregation may occur at neutral pH (6.7), prior to fermentation or acidification to their final pH values. In general, the structure of chilled dairy products is determined by the pH and the ionic environment, and these may be modified if fruit and cereal are included in the products. It has been recognised for a long time that the presence of divalent © 2008, Woodhead Publishing Limited

Raw material selection: dairy ingredients

55

cations (calcium) enhances the aggregation of casein at all temperatures and of whey proteins on heat-induced denaturation. Heat-induced denaturation/aggregation of β-lactoglobulin is highly pH-dependent. At low ionic strength, maximal denaturation/aggregation is achieved on heating at pH 5.0 with minima occurring at pH <3.5 and 6.0.

3.5.2 Water-binding Protein hydration is generally considered to be the water that is more-or-less bound or immobilised by a protein (Damodaran, 1997). Creamer (1985) showed that casein micelle hydration was affected by pH, NaCl and rennet treatment. A peak in casein micelle hydration was observed at pH 5.2 and, while the hydration was reduced following the action of rennet, this peak in hydration was still retained. The addition of NaCl was shown to increase casein micelle hydration at all pH values above 4.5, either by displacing calcium or calcium phosphate from the protein matrix with a concomitant increase in the number of ionic groups and a consequent increase in the volume of the matrix or by affecting the ability of the matrix to aggregate. The interaction between NaCl, pH and rennet effects was central to the texture and subsequent functional behaviour of cheese for the consumer (i.e. melting, stretching).

3.5.3 Emulsification Chilled foods can be emulsions with a variable fat content and varying degrees of stability; the degree of stability affects the release of flavours. They can be water continuous, such as yoghurt, or oil-continuous, such as table spreads. The strongly amphipathic nature of proteins, particularly dairy proteins, resulting from their mixture of polar and nonpolar side chains, causes them to be concentrated at interfaces – that is why proteins are good emulsifiers. At fluid/fluid interfaces, it is well established that globular proteins, in general, lose their tertiary structure, existing in extended configurations with hydrophobic side chains orientated towards the nonaqueous phase and hydrophilic side chains directed towards the aqueous phase. The effect of oil droplet size and emulsifier type on the rheology of emulsion gels was studied by McClements et al. (1993). Gel strength increased as the oil droplet size decreased for droplets stabilised by protein, but decreased when the oil droplets were stabilised by a low molecular weight surfactant (Tween). Oil droplets coated with protein can be incorporated into a protein network, reinforcing the structure and so increasing the gel strength, whereas droplets coated with Tween cannot be incorporated into a protein network and disrupt the three-dimensional structure of the gel, hence decreasing the strength. The source of the emulsifier used to stabilise emulsions in chilled foods can have a great effect on texture.

3.5.4 Gelation The formation of three-dimensional matrices during the processing of chilled © 2008, Woodhead Publishing Limited

56

Chilled foods

foods is closely related to the functional attribute that is termed ‘gelation’. This is closely aligned with the process of protein hydration (see Section 3.5.2). Whey proteins are globular proteins and are hence heat labile. They are widely used to enhance the nutritional value of formulated foods and are often cited as being highly functional through their ability to gel on heating (De Wit, 1989; Hermansson, 1975). Thermal induction of gelation is normally the functional property of interest, where large quantities of water are immobilised (i.e. binding of water in meat products). The characteristics of whey protein gels may vary from elastic to curd-like, depending on the preparation technique. Physical properties of thermally-induced whey protein gels depend on the level and type of soluble salt present in solution. Maximum gel firmness for whey protein concentrate was observed at 0.044% (w/w) calcium or 0.46% (w/w) sodium, with higher levels reducing the firmness of the gels (Mulvihill and Kinsella, 1988).

3.5.5 Heat stability Chilled foods of a dairy nature or formulated from added dairy ingredients generally go through a heating step in their preparation or formulation. The heating may be pasteurisation, or cooking, to initiate functional development and give products their character. The formation of reconstituted complex colloidal mixtures of milk proteins, fat and minerals that are stable to subsequent high heating regimes is a complicated sequence of events which is done prior to dehydration or post-reconstitution. The main milk components that make up these colloidal structures are casein, whey proteins, calcium and phosphorous. Since some or all of these components are markedly unstable when they are heated alone, if they are heated together then they may protect one another; this property is central to heat stability. At low concentrations of milk solids (10% solids), most milk is stable to high heating regimes provided the pH is in the normal range (6.55–6.8). However, when milk is concentrated (>20% solids), it has reduced heat stability across the pH range (6.2–7.2).

3.6

Chilled food production

Milk proteins are used extensively in chilled food systems for their functionality. This includes formation of acid gels, rennet gels, emulsion stabilisation, whey protein gelation or stability to extended heating regimes. Formulation or production of chilled foods requires a thorough knowledge of the ingredients and the factors that control their behaviour. The processing variables have to be within specification and finely controlled to achieve consistent quality. Good manufacturing practice entails using processing conditions and procedures that have been proven to deliver consistent quality and safety, based on experience. Therefore, the compositional specification for any ingredient, whether of dairy origin or not, should ensure that critical functional properties are covered. Often the specifications for ingredients cannot be written tightly enough due to insufficient information © 2008, Woodhead Publishing Limited

Raw material selection: dairy ingredients

57

on how the chilled product actually works or is altered by raw material variability. This sometimes leads to friction between ingredient producers and ingredient users. The specification for an ingredient should contain as much functional information as possible that is relevant to the product being manufactured. This could include viscosity on reconstitution to a certain solids concentration, the resulting pH, and an indication of the mineral types and content. Obviously the ingredient should be microbiologically safe with no unwanted enzymatic activity. However, the heat treatment needed to achieve this should not negatively impact on colour or taste. In all probability a selected chilled food is produced from a number of ingredients and is subjected to a number of process stages (heating, homogenisation, fermentation). This blending of different ingredients and subjecting them to a series of sub-processes makes prediction of the outcome potentially variable unless good specifications and process instructions exist. Polysaccharide-based biopolymers (e.g. guar gum, xanthan, carrageenan, alginate, chitosan, starch, pectin) are also incorporated into chilled food products for a variety of reasons. These might include viscosity control, water control or general stabilisation of the food. It has been shown that interactions between polysaccharides and proteins can have detrimental or beneficial effects on food product quality, depending on their nature and the required properties of the food in question (Tolstoguzov, 2003). The use of hydrocolloids can positively influence emulsion creaming kinetics or a negative flocculation/serum separation. Mixtures of biopolymers are often observed to separate into distinct phases, which can be the result of thermodynamic incompatabilty (polymers do not like one another), complex coacervation (positive charges like negative charges, opposites attract) or depletion flocculation (localised concentration of polymer). Bryant and McClements (2000) showed that the thermodynamic incompatibility of xanthan and heatdenatured whey proteins could be utilised to create stronger cold-set whey protein gels on addition of NaCl. They suggested that heat-denatured whey protein and xanthan formed water-in-water emulsions consisting of xanthan-rich regions surrounded by a protein-rich aqueous phase.

3.7

Quality criteria

While traditionally the dairy industry has been good on food quality and safety, food processors and ingredient manufacturers now rely on modern quality management systems to ensure the quality and safety of the products they produce. These generally include good manufacturing practice (GMP), hazard analysis critical control points (HACCP) and quality assurance standards. These quality management systems used by food processors also involve working with suppliers, transporters and retailers. The quality of raw materials is crucial to ensure the safety and quality of the final product. Therefore, a systematic approach is needed from the farm to the consumer in order to avoid contamination of foodstuffs and to identify potential hazards. GMP entails the processing conditions and procedures that have been proven to deliver consistent quality and safety based on experience. © 2008, Woodhead Publishing Limited

58

Chilled foods

HACCP, a relatively recent proactive technique, is focused on identifying potential problems and controlling them during the design and the production process. Adherence to standards outlined by the International Standards Organisation (ISO) and the European Hygiene Regulation (EC 852/2004) ensures that food processing, catering and other food-related industries conform to prescribed and well-documented procedures. Food temperature control is central in the chilled food market and legally products should be stored at a maximum of 8 °C. However, most retailers aim for 4–5 °C. Cold chain adherence was identified as key to food safety assurance in the retail sector. The introduction of new technologies such as active and intelligent packaging (use of time–temperature indicators) should ensure that foods that have been subjected to inappropriate conditions can be identified.

3.8

Allergen issues

Food allergy, or hypersensitivity, refers to an abnormal immunologic reaction in which the body’s immune system produces an allergic antibody, called immunoglobulin E, to usually harmless foods such as milk or egg protein, resulting in symptoms such as wheezing, diarrhoea or vomiting. True food allergies are serious, but a far more common condition is food intolerance. People with food intolerance can often eat small amounts of the offending food without experiencing symptoms. Lactose intolerance is a condition in which a person lacks the enzyme to break down the sugar found in milk. This can result in severe discomfort following ingestion of most dairy products. True allergens are mostly proteinbased and can be derived from a wide variety of foods (nuts, eggs, dairy products). Avoidance of specific foods is paramount once a true food allergy has been diagnosed and vigilance is required because processed food may contain milk proteins as secondary inclusions.

3.9

Future trends

While the traditional range of dairy ingredients used in chilled foods will not change dramatically in the near future, our increased understanding of what they do and how they interact with other ingredients will drive both the nutritional and functional approach to chilled food diversification. The interactive effects of mixed biopolymer combinations will be utilised to achieve both nutritional delivery vehicles and new texture development. These increased understandings of the mechanisms involved in biopolymer interactions should be particularly relevant in formulated chilled foods where ingredient selection is not limited to any particular sector (dairy, plant, egg, etc.). While dairy-based chilled foods will still be an integral and major part of the market, the drive for added-value and nutritional tailoring should spawn a range of alternative chilled foods. Fortification of chilled foods (nutraceuticals) with perceived nutritional benefits will be essential and © 2008, Woodhead Publishing Limited

Raw material selection: dairy ingredients

59

these will include calcium, pre- and probiotics, and plant extracts with clinically proven health attributes. The increased understanding of ingredient behaviour in complex food systems is a prerequisite to advances in processing technology and hence new ingredient production and chilled food formulations.

3.10 References AUTY M A E, O’KENNEDY B T, ALLAN-WOJTAS P AND MULVIHILL D M, 2005. The application

of microscopy and rheology to study the effect of milk salt concentration on the structure of acidified micellar casein systems, Food Hydrocolloids, 19 101–109. BARFOD N M, KROG N AND BUCHEIM W, 1989. Lipid–protein–emulsifier–water interactions in whippable emulsions. In: Kinsella J E and Soucie W G, Food Proteins, Champaign, Illinois, American Oil Chemists Soc., AOCS, 144–158. BRYANT C M AND MCCLEMENTS D J, 2000. Influence of xanthan gum on physical characteristics of heat-denatured whey protein solutions and gels, Food Hydrocolloids, 14 383–390. CREAMER L K, 1985. Water absorption by renneted casein micelles, Milchwissenschaft, 40 589–591. DAMODARAN S, 1997. Food proteins; An overview. In: Damodaran S and Paraf A, Food Proteins and their Applications, New York, Marcel Dekker, Inc., 1–24. DANNENBERG F AND KESSLER H G, 1988. Effect of denaturation of β-lactoglobulin on texture properties of set-style nonfat yoghurt. 1. Syneresis. Milchwissenschaft, 43 632–635. DE WIT J N, 1989. The use of whey protein products. In: Fox P F, Developments of Dairy Chemistry-4, London, Elsevier Applied Science, 323–345. FAMELART M H, LEPESANT F, GAUCHERON F, LE GRAET Y AND SCHUCK P, 1996. pH-induced physiochemical modifications of native phosphocaseinate suspensions: Influence of aqueous phase. Le Lait, 76 445–460. HERMANSSON A M, 1975. Functional properties of proteins for foods – flow properties, Journal of Texture Studies, 5 425–439. JEBSON R S, 1994. BUTTER AND ALLIED PRODUCTS. In: Moran D P J and Rajah K K, Fats in Food Products, London, Chapman and Hall, 69–106. KEOGH M K, 1992. The stability to inversion of a concentrated water-in-oil emulsion, Ph D Thesis, National University of Ireland. KOROLCZUK J, 1982. Hydration and viscosity of casein solutions, Milchwissenschaft, 37 274–276. LE RAY C, MAUBOIS J L, GAUCHERON F, BRULE G, PRONNIER P AND GARNIER F, 1998. Heat stability of reconstituted casein micelle dispersions: Changes induced by salt addition, Lait, 78 373–390. LUCEY J A, TEO C T, MUNRO P A AND SINGH H, 1997a Rheological properties at small (dynamic) and large (yield) deformations of acid gels made from heated milk, Journal of Dairy Research, 64 591–600. LUCEY J A, VAN VLIET T, GROLLE K, GUERTS T AND WALSTRA P, 1997b. Properties of acid casein gels made by acidification with glucono-δ-lactone. 1. Rheological properties, International Dairy Journal, 7 381–388. MCCLEMENTS J, MONAHAN F J AND KINSELLA J E, 1993. Effect of emulsion droplets on the rheology of whey protein isolate gels, Journal of Texture Studies, 24 411. MCSWEENEY P L H, 2004. Biochemistry of cheese ripening: Introduction and overview. In: Fox P F, McSweeney P L H, Cogan T M and Guinee T P, Cheese Chemistry, Physics and Microbiology, General Aspects, 3rd edn., Vol. 1, Amsterdam, Elsevier Academic Press, 347–360. MULDER H AND WALSTRA P, 1974. Isolation of milk fat. In: Mulder H and Walstra P, The Milk Fat Globule, Wageningen, Pudoc, 228–243. © 2008, Woodhead Publishing Limited

60

Chilled foods

MULVIHILL D M, 1989. Caseins and caseinates: Manufacture. In: Fox P F, Developments in

Dairy Chemistry-4, London, Elsevier Applied Science, 97–130. 1988. Gelation of β-lactoglobulin: Effects of sodium chloride and calcium chloride on the rheological and structural properties of gels, Journal of Food Science, 53 231–236. PIRES M S, ORELLANA G A AND GATTI C A, 1999. Rennet coagulation of casein micelles and heated casein micelles: Action of Ca2+ and pH, Food Hydrocolloids, 13 235–238. PLATT B L, 1988. Low fat spread, European Patent No 0 256 712. SCHUCK P, PIOT M, MEJEAN S, FAUQUANT J, BRULE G AND MAUBOIS J L, 1994. Deshydratation des laits enrichis en caseine micellaire par microfiltration; comparaison des proprietes des poudres obtenues avec celles d’une poudre de lait ultra-propre, Lait, 74 47–63. TOLSTOGUZOV V B, 2003. Some thermodynamic considerations to food formulation, Food Hydrocolloids, 17 1–23. VAN VLIET T, LUCEY J A, GROLLE K AND WALSTRA P, 1997. Rearrangements in acid-induced casein gels during and after gel formation. In: Dickinson E and Bergenstahl B, Food Colloids: Proteins, Lipids and Polysaccharides. Cambridge, Royal Society of Chemistry, 335–345. MULVIHILL D M AND KINSELLA J E,

© 2008, Woodhead Publishing Limited

4 Raw material selection: meat and poultry S. James and C. James, Food Refrigeration and Process Engineering Centre (FRPERC), UK

4.1

Introduction

An increasing proportion of red and poultry meat is used as a basic raw material for chilled meat products and ready meals. This meat is primarily muscle but will also contain fat and sometimes skin and bone from the animal. Muscles of freshly killed animals are relaxed, soft, extensible and flexible. At slaughter, the metabolism shifts from an aerobic to an anaerobic state where oxygen is depleted. Glycogen is converted to lactic acid, lowering the muscle pH from about 7 to 5.6. This creates phosphate, ATP declines and the muscle becomes stiff, rigid and contracted – a state called rigor mortis. Proteolysis then begins, which gradually tenderises the muscle. The rate of all these reactions is temperature dependent and is therefore a function of the meat’s temperature history post-slaughter. The quality of meat as an incoming raw material is often judged by its appearance and bacterial condition. Appearance criteria are primarily based on the colour, percentage of visual fat and lean, and the amount of drip exuding from the meat. Bacterial condition is subjectively assessed by the presence or absence of odour or slime. Quantitative tests can also be carried out to determine the total viable counts and the presence of specific pathogens or indicator organisms, but cannot supply an instant assessment. Any unacceptable change in the microbial or appearance criteria will limit the shelf-life of the meat. After cooking, its eating quality is partially judged by its appearance but mainly by its tenderness, flavour and juiciness. Red and poultry meat are very perishable raw materials. If stored under ambient conditions (16 to 30 °C), the storage life of both can be measured in tens of hours © 2008, Woodhead Publishing Limited

62

Chilled foods

to a few days. Under the best conditions of chilled storage (close to the initial freezing point of the meat) the storage life can be extended to over 6 weeks for some red meats. Under the very best commercial practice (strictly hygienic slaughtering and cutting, rapid cooling, vacuum-packing, and storage at superchill temperatures (–1±0.5 °C)), the maximum storage life that can be achieved with red meat is approximately 20 weeks. Freezing will extend the storage life of meat to a number of years. In a perfect world, red meat and poultry meat would be completely free of pathogenic (food poisoning) micro-organisms when produced. However, under current methods of production, pathogen-free meat cannot be guaranteed. While all meat animals carry large numbers of different micro-organisms on their skin/ feathers and in their alimentary tract, only a few types of these bacteria directly affect the safety and quality of the finished meat. In general, the presence of small numbers of pathogens is not a problem because meat is normally cooked before consumption, and this is also the case for many ready meals during production. Adequate cooking will substantially reduce the numbers of, if not completely eliminate, all the vegetative pathogenic organisms present on the meat. Most meat-based food poisoning is associated with inadequate cooking or contamination after cooking, in which case other micro-organisms become important (e.g. Staphylococcus aureus). The purpose of refrigeration is to reduce, or prevent, the growth of pathogens so that they do not reach levels that could cause problems. Few pathogens of major concern grow on meat at temperatures below 7 °C (POST, 1997). Normally, it is the growth of spoilage organisms that has the most important input in limiting the storage life of meat. In general, there is little difference in the microbial spoilage of beef, lamb, pork and other meat derived from mammals (Varnam and Sutherland, 1995). Overall it would also appear that there is little intrinsic difference in the rate of spoilage of different meats. Differences can be accounted for by differences in initial bacteria levels, tissue composition and pH (Blixt and Borch, 2002; Stern et al., 1992). The spoilage bacteria of meats stored in air under chill conditions include species of Pseudomonas, Brochothrix and Acinetobacter/Moraxella. Meat colour is related to species, but within a species can be adversely affected by a variety of factors, including post-mortem handling, chilling, storage and packaging (Church and Wood, 1992; Miller, 2002). The presence of exudate or ‘drip’, which accumulates in the container of prepackaged meat, or in trays or dishes of unwrapped meat, substantially reduces its sales appeal (Malton and James, 1983). Drip loss occurs throughout the cold chain and represents a considerable economic loss to the red meat industry. Poultry meat is far less prone to drip. The potential for drip loss is inherent in fresh meat and is related to the development of rigor mortis in the muscle after slaughter and its effect on pH. It is influenced by many factors. Some of these, including breed, diet and physiological history are inherent in the live animal. Others, such as the rate of chilling, storage temperatures, freezing and thawing, occur during processing. In Australia CSIRO (1988) stated that ‘Toughness is caused by three major © 2008, Woodhead Publishing Limited

Raw material selection: meat and poultry

63

factors – advancing age of the animal, ‘cold shortening’ (the muscle fibre contraction that can occur during chilling), and unfavourable meat acidity (pH).’ There is general agreement on the importance of these factors, with many experts adding cooking as a fourth equally important influence. If we wish to produce red meat and poultry based chilled foods with the optimum qualities, we need to understand the factors that govern the quality of the meat itself.

4.2

Sources of supply

Meat for further processing may be supplied as chilled bone-in or boneless primals or carcasses, as boneless blocks of frozen material, or, increasingly, as minced or diced material. The minced or diced meat may be supplied chilled, or more often nowadays as bags of individually quick frozen (IQF) product. Little appears to be known about the relationship between the frozen storage life of the raw material and the chilled life of the product in which it is used. Meat may also be supplied in a ‘super’ or ‘deep’ chilled state (James and James, 2006). Super-chilling is a conservation method for a meat whereby some of the water in the product is frozen. Normally, the meat temperature is lowered to approximately 1 °C to 3 °C below its initial freezing point. Approximately 30–50% of the water in the product is then in a frozen state.

4.3

Hazards (microbiological and non-microbiological)

The primary hazard associated with raw meat is the presence of food pathogens. Other hazards include, to a lesser extent, food-borne diseases, parasites, and allergens.

4.3.1 Microbiological hazards While the internal musculature of a healthy mammal or bird is essentially sterile after slaughter, all meat animals carry large numbers of various micro-organisms on their skin/feathers and in their alimentary tract. Of these, only a few types of bacteria directly affect the safety and quality of the finished meat. The main microbiological hazards associated with meat have been discussed in detail by McClure (2000) and Nesbakken (2005). Pathogens of particular concern in meats are Campylobacter spp., Salmonella spp., pathogenic serotypes of Escherichia coli, Clostridium perfringens, Clostridium botulinum, Yersinia enterocolitica, Listeria monocytogenes and, to a lesser extent, Staphylococcus aureus and Bacillus cereus. The minimum growth temperatures for some of the major food-borne pathogens associated with red and poultry meat are shown in Table 4.1 (García de Fernando et al., 1995; Mead and Hinton, 1996; Doyle, 2002; Tamplin et al., 2005). Other microbiological hazards associated with meat include diseases such as © 2008, Woodhead Publishing Limited

64

Chilled foods

Table 4.1

Minimum growth temperatures for pathogens associated with red meats Minimun growth temperature (°C)

Campylobacter spp. Clostridium perfringens Clostridium botulinum proteolytic Staphylococcus aureus Pathogenic Escherichia coli strains Escherichia coli O157:H7 Salmonella spp. Bacillus cereus Clostridium botulinum non-proteolytic Listeria monocytogenes Yersinia enterocolitica

30 12 10 7 7 6 to 7 5 5 3 –1 to 0 –2

brucellae, and parasites such as Giardia duodenalis; however, these are regarded as a relatively low risk. Trichinella spiralis, which is responsible for trichinellosis, can be a problem in raw or undercooked pork (but not in cured or frozen pork). A number of pathogens, such as Erysipelothrix rhusiopathiae in pigs, are associated with skin diseases in humans that can be caught through contact with meat during processing.

4.3.2 Meat allergies Food allergies are of increasing concern to the food industry. Meat allergies appear to be rare; they are not in the EU list of most common food allergens, but are reported to be increasing (Orhan and Sekerel, 2003). The incidence of beef allergy, for instance, may be as high as 0.3% in the general population of the UK (Fiocchi et al., 2000). Advice issued by the UK Food Standards Agency states that ‘people with an allergy to meat may react to just one type, such as pork, beef, lamb or chicken, or they may react to a range of types. The most common symptom of meat allergy is dermatitis (an allergic skin reaction)’. Since meat allergies are often specific to a certain type of meat, food processors may want to hold different types of raw meats separately to prevent any possible mixing of ingredients during storage and ensure that suppliers of minced or diced raw meats have established cleaning regimes between the cutting of raw meats to make sure that no mixing takes place.

4.4

Influence of live animal on meat quality

A number of the factors that influence the quality of meat are inherent in the live animal and the way that the animal has been raised.

4.4.1 Species and breed In all meat species, the range of storage lives found in the literature is very large © 2008, Woodhead Publishing Limited

Raw material selection: meat and poultry

65

Table 4.2 Chilled storage life of meat and meat products at various storage temperatures and atmospheres (assuming good initial bacteriological quality and normal pH); d = days, w = weeks, MAP = modified atmosphere packaging Product Beef, carcass Beef, carcass Pork, carcass Pork, carcass Lamb, carcass Chicken, carcass Chicken, carcass Beef, subprimals and joints Pork, subprimals and joints Pork, subprimals and joints Pork, subprimals and joints Lamb, subprimals and joints Lamb, subprimals and joints Chicken, joints Chicken, joints Offal Beef, mince Beef, mince Poultry, mince

Temperature (°C) 4 –1.5 to 0 4 –1.5 to 0 –1.5 to 0 4 –2 to 0 –1.5 to 0 –1.5 to 0 –1.5 to 0 –1.5 to 0 –1.5 to 0 –1.5 to 0 1 1 –1.5 to 0 4 –1.5 1

Packaging Unwrapped Unwrapped Unwrapped Unwrapped Unwrapped Oxygen permeable Oxygen permeable Vacuum Vacuum Oxygen permeable 100% CO2 100% CO2 Vacuum Vacuum Oxygen permeable Unwrapped Vacuum Vacuum MAP

Storage life 10–14 d 3–5 w 8d 3w 3–4 w 1w 2–4 w 11–12 w 4–6 w 1.5 w 12 w 16 w 6–10 w 20–25 d 14–16 d 7d 1–2 w 4.5 w 2w

(Table 4.2) and indicates that factors other than species have a pronounced effect on storage life. Overall, species has little affect on the practical storage life (PSL) of meat. The pigment concentration in meat that governs its colour is certainly influenced by species. Beef and lamb contain substantially more myoglobin than pork and poultry meat, thus accounting for the difference between ‘red’ (beef and lamb) and ‘white’ (pork and poultry) meats. Drip potential also appears to be related to species. In general, beef tends to lose proportionately more drip than pork and lamb. Poultry meat is far less prone to drip. The potential for drip loss is inherent in fresh meat and is related to the development of rigor mortis in the muscle after slaughter and its effect on pH. In pigs, especially, there are large differences in drip loss from meat from different breeds. Taylor (1972) showed that there was a substantial difference, up to 2.5 fold, in drip loss between four different breeds of pig (Table 4.3). Conversely breed does not appear to have an effect on water-holding capacity in chickens (Musa et al., 2006), though meat from different chicken breeds may differ significantly in colour density, pH and tenderness. Although there is a common belief that breed has a major effect on beef meat quality, CSIRO (1992) state that ‘although there are small differences in tenderness due to breed, they are slight and currently of no commercial significance to Australian consumers.’ That said, there are substantial differences in the proportion of acceptable tender and tough meat between Bos indicus and Bos taurus cattle. Bos indicus are tropical and semitropical breeds of cattle, primarily Brahman, © 2008, Woodhead Publishing Limited

66

Chilled foods

Table 4.3 Drip loss after 2 days’ storage at 0 °C, from leg joints of different breeds of pig cooled at different rates Breed Slow cooling Landrace Large White Wessex × Large White Pietrain

0.47 0.73 0.97 1.14

Drip loss (% by wt) Quick cooling 0.24 0.42 0.61 0.62

and Bos taurus are temperate breeds, such as Hereford or Aberdeen Angus. The proportion of acceptable tender meat has been found to decrease from 100% in Hereford Angus crosses, to 96% in Tarentaise, 93% in Pinzgauer, 86% in Brahman and only 80% in Tsahiwal (Koch et al., 1982). Toughness of meat increases as the proportion of Bos indicus increases (Crouse et al., 1989).

4.4.2 Animal-to-animal variation There are little data on any relationship between animal-to-animal variation in the same herd and its effect on chilled storage life. However, it is believed to cause wide variations in frozen storage life. Differences (assessed via a taste panel) can be as great as 50% in lamb (Winger, 1984a; b). The differences would appear to be caused by genetic, seasonal, or nutritional variation between animals, but there is little reported work to confirm this view. Variations have been reported between the fatty acids and ratio of saturated/unsaturated fatty acids in lamb from New Zealand, America and the UK (Crystall and Winger, 1986). Differences related to sex, area and cut were mainly a reflection of fatness, with ewes having a greater percentage of body fat than rams. However, differences between growing areas (countries) were found to produce larger variations between animals than sex differences. A number of other trials have detailed differences between animals. There can also be significant differences in meat texture within a breed. Longissimus dorsi shear force values for double muscled Belgium Blue bulls were significantly higher than those for the same breed with normal conformation (Uytterhaegen et al., 1994). Calpain I levels at 1 hour and 24 hours post mortem were also much lower. The calpain system is one of the major proteolytic systems in meat tenderisation, consisting primarily of three components: µ-calpain, a proteinase that is active at micromolar concentrations of calcium, m-calpain, a proteinase that is active at millimolar concentrations of calcium, and the specific endogenous calpain inhibitor calpastatin. It was suggested that the lower background toughness in the double muscled was compensated for by reduced post-mortem proteolytic tenderisation. Meat from male animals usually contains more pigment than that from female animals. The sex of the animal appears to have little or no influence on tenderness in beef. Huff and Parrish (1993) compared the tenderness of meat from 14-month© 2008, Woodhead Publishing Limited

Raw material selection: meat and poultry

67

old bulls and steers, and cows (55 to 108 months old). No differences were found between the tenderness of bulls and steers. Tenderness decreased with the age of the animal. Hawrysh et al. (1979) reported that beef from bulls may be less tender than that from steers. In chickens, tenderness (values for shear force) has been reported to be higher in males than females (Musa et al., 2006), though not all studies concur (Berri, 2004). The sex of the animal can also have a substantial influence on meat flavour. For example, cooking meat from intact male pigs can produce an obnoxious odour known as ‘boar taint’. Problems can also occur with meat from intact males of other species. However, growing and processing intact males can still be attractive to industry because of their higher rate of growth and lower fat content in comparison with castrated males. 4.4.3 Variations within an animal Reports of variations in the storage life of different cuts of meat from the same animal are scarce and primarily deal with dark and light meat. The difference between dark and light muscles in the same animal is related to the job the muscles carry out. Muscles that do more work contain more myoglobin (Miller, 2002). Both Ristic (1982) and Keskinel et al. (1964) have found that poultry breast meat stores better than thigh meat, for instance. Ristic (1982) states that breast meat will store for 16 months while thigh meat can only be stored only for 12 months, due to its higher fat content. Judge and Aberle (1980) also found that light pork meat stored for a longer time than dark meat. This was thought to be due to either higher quantities of haem pigments in the dark muscle (which may act as major catalysts of lipid oxidation) or higher quantities of phospholipids (which are major contributors to oxidised flavour in cooked meat). There can be large differences in drip loss between different muscles. Taylor (1972) showed that there was a 1.7 to 2.8 fold difference in drip between muscle types in pigs (Table 4.4). Since most of the exudate comes from the cut ends of muscle fibres, small pieces of meat also drip more than large intact carcasses, and the way that different muscles are cut will also have an influence on drip. 4.4.4 Age of animal It is now well established that it is the properties of the connective tissue proteins, Table 4.4 Drip loss after 2 days’ storage at 0 °C from four muscles from two breeds of pig cooled at different rates Breed Pietrain Large White

Cooling rate Quick Slow Quick Slow

© 2008, Woodhead Publishing Limited

Drip (as % muscle weight) SemiSemiAdductor Biceps Combined tendinosus membranous femoris (four muscles) 2.82 3.99 1.69 1.95

4.40 6.47 2.01 3.50

5.52 6.61 2.92 5.07

2.69 4.11 1.04 2.32

3.86 5.30 1.92 3.21

68

Chilled foods

and not the total amount of collagen in meat, that largely determines whether meat is tough or tender (Church and Wood, 1992). As the animal grows older, the number of immature reducible (protein) cross-links decreases. The mature crosslinks result in a toughening of the collagen and this, in turn, can produce tough meat. In beef, increasing connective tissue toughness is probably not commercially significant until an animal is about 4 years old (Husband and Johnson, 1985). Pigment concentration (myoglobin content) also increases with age; for example, veal is brownish pink versus beef from 3-year-old steers that is bright, cherry red (Miller, 2002).

4.4.5 Feeding The way in which an animal is fed can influence its quality and refrigerated storage life. It has been reported that chops from pigs fed on household refuse have half the storage life of those fed on a milk/barley ration (Wismer-Pederson and Sivesgaard, 1957; Palmer et al., 1953). Again, pork from pigs that have been fed materials containing offal have been reported to have half the PSL of pigs that have not been fed this type of diet; this was associated with higher iodine numbers (indicating higher unsaturated fat levels) in the fat (Bogh-Sorensen and Hojmark Jensen, 1981). Conversely, Bailey et al. (1973) did not find any differences in the quality of frozen pork from meal- or swill-fed pigs after 4 and 9 months at –20 °C. Feed with large amounts of highly unsaturated fatty acids tends to produce more unstable meat and fat (i.e. is more prone to rancidity during storage). The type of fatty acid composition of ‘depot fat’ in poultry and its stability have been shown to be directly related to the fatty acid composition of ingested fats (Klose et al., 1953; Kummerow et al., 1948). The feeding of fish oils or highly unsaturated vegetable oils (such as linseed oil) to poultry is known to produce fishy flavours in the meat (Liu et al., 1995). The use of vitamin E supplements in feed is recommended for both beef and turkey. This will ‘result in delayed onset of discoloration in fresh, ground and frozen beef and in suppression of oxidative rancidity, especially in fresh, ground and frozen beef and less so in cooked beef’ (Liu et al., 1995). With turkeys, vitamin E supplements have been shown to improve lipid stability of cooked and uncooked turkey burgers during 6 months of frozen storage at –20 °C (Wen et al., 1996).

4.4.6 Handling prior to and at point of slaughter The way animals are handled and transported before slaughter affects meat quality and its storage life. Increased stress or exhaustion, e.g. from poor transport or diet, can produce PSE (pale soft and exudative) or DFD (dark firm and dry) meat, which is not recommended for storage mainly due to its unattractive nature and appearance. The colour changes observed with PSE and DFD meat are mostly due to structural changes in the muscle. DFD meat has a high ultimate pH and oxygen penetration is low. Consequently, the oxymyoglobin layer is thin, the © 2008, Woodhead Publishing Limited

Raw material selection: meat and poultry

69

purplemyoglobin layer shows through, and the meat appears dark. In PSE meat, the pH falls while the muscle is still warm and partial denaturation of the proteins occurs. An increased amount of light is scattered and part of the pigment is oxidised (to metmyoglobin) so that the meat appears pale. Jeremiah and Wilson (1987) found that the use of PSE pork muscle produced low yields after curing and it was concluded that PSE meat was unsuitable for further processing. PSE meat has been reported to be a growing problem in the poultry industry (particularly regarding turkeys) and is characterised by a rapid post-mortem pH decline (Woelfel and Sams, 2001). Rapid chilling after slaughter can alleviate some of the drip problems in PSE meat. Experiments designed to determine the effect of treatments immediately pre or at the point of slaughter appear to show that they have little effect on meat texture. Exercising pigs before slaughter has been shown to have no effect on texture parameters, i.e. muscle shortening and shear force (Ivensen et al., 1995). The use of different stunning methods (both electrical and carbon dioxide) does not seem to have a significant effect on the quality of pork (Garrido et al., 1994).

4.5

Influence of slaughter and processing conditions on meat quality

The appearance of meat influences the consumer’s willingness to buy it when raw; however, after cooking the eating quality is determined by tenderness, juiciness and flavour, as well as appearance (Aaslyng, 2002). Although some of these quality characteristics are determined by the live animal, how the animal is processed immediately after slaughter has a profound affect on the overall meat quality.

4.5.1 Appearance The red colour (from oxymyoglobin) in meat is more stable at lower temperatures because the rate of oxidation of the pigment (to metmyoglobin) decreases. At low temperatures, the solubility of oxygen is greater and oxygen-consuming reactions are slowed down. There is a greater penetration of oxygen into the meat and the meat is redder than at high temperatures. The rate of oxidation depends on the species, thus the rate of cooling during chilling may have an effect on meat colour. Different chilling methods can certainly affect the appearance of poultry meat. The scalding that the carcasses have received prior to plucking has a marked effect on the final appearance. Carcasses destined for air chilling (Fig. 4.1) can be only ‘soft’ scalded (i.e. at 50 to 53 °C), which retains the outer dermal layer, since higher scalding temperatures remove the outer dermal layer and make the carcasses more susceptible to dehydration and discolouration during air chilling. The problem does not occur with immersion chilling. Hence carcasses chilled using this method can be ‘hard’ scalded at higher temperatures. While air chilling is © 2008, Woodhead Publishing Limited

70

Chilled foods

Fig. 4.1

Continuous air blast chilling of chicken carcasses.

preferred in most of Europe, it is still very much a matter of debate as to the relative merits of the appearance of air or immersion chilled carcasses (James et al., 2006).

4.5.2 Drip Exudate, or ‘drip’, accumulates in the containers of pre-packaged meat, or in trays of unwrapped meat. Not only does this exudate look unattractive, but it also represents an appreciable weight loss to the user when the meat is subsequently removed from its container. Drip can be referred to by a number of different terms including ‘purge loss’, ‘press loss’ and ‘thaw loss’, depending on the method of measurement and when it is measured. Although, the potential for drip loss is predetermined to a large extent by species, breed and conditions before slaughter, the realisation of this potential is influenced by the temperature–time history in the cold chain. In general, beef tends to lose proportionately more drip than pork or lamb. Since most of the exudate comes from the cut ends of muscle fibres, small pieces of meat drip more than large intact carcasses. The protein concentration of drip is about 140 mg ml–1 (about 70% of that of meat itself). The proteins in drip are the intracellular, soluble proteins of the muscle cells. The red colour is due to the protein myoglobin, the main pigment of meat. © 2008, Woodhead Publishing Limited

Raw material selection: meat and poultry

71

Rapid chilling reduces drip loss (Tables 4.3 and 4.4) after subsequent cutting operations. The potential for drip loss is established in the first period of cooling. The temperature range conducive to drip is down to about 30 °C, or perhaps a little lower. It has been recommended that deep muscle temperatures in turkeys should be reduced below 25 °C by 60 min post mortem to reduce drip in this meat, for instance, (Alvarado and Sams, 2002). During chill storage, the rate of drip loss increases with storage temperature, and the amount of drip will increase with storage time. Low storage temperatures will reduce the amount of drip. Excessive drip could have a small effect on the eating quality of meat. Perceived juiciness is one of the important sensory attributes of meat. Dryness is associated with a decrease in the other palatability attributes, especially with lack of flavour and increased toughness (Pearson, 1994). However, moisture losses during cooking are typically an order of magnitude higher than most drip losses during refrigeration. Consequently, small differences in drip loss will have little effect on eating quality.

4.5.3 Texture Chilling can have serious effects on the texture of meat if it is carried out too rapidly when the meat is still in the pre-rigor condition, that is before the meat pH has fallen below about 6.2 (Bendall, 1972). In this state, the muscles contain sufficient amounts of the contractile fuel, adenosine triphosphate (ATP), for forcible shortening to set in as the temperature falls below 11 °C, the most severe effect occurring at about 3 °C. This is the so-called ‘cold-shortening’ phenomenon, first observed by Locker and Hagyard (1963), and its mechanism was described by Jeacocke (1986). The meat ‘sets’ in the shortened state as rigor comes on, and this causes it to become extremely tough when it is subsequently cooked (Marsh and Leet, 1966). If no cooling is applied and the temperature of the meat is above 25 °C at completion of rigor, then another form of shortening, ‘rigor’- or ‘heat-shortening’, will occur (Dransfield, 1994) when cooked. The severity of cold-shortening is highly pH dependent, being much greater at pH 6.8 (i.e. exceptionally rapid chilling) than at pH 6.2 (i.e. at an easily attainable commercial rate of chilling). To allow a safety margin, and taking into account the fact that some carcasses will show high initial pH values in the eye muscle, it is recommended that any part of a beef or lamb carcass should not be chilled below 10 °C until at least 10 hours after slaughter. In pork, cold-shortening occurs if temperatures between 3 and 5 °C are reached before the onset of rigor (normally 3 to 8 hours). This will occur only in rapid pork chilling systems and is less common. Although poultry breast muscles are primary composed of white fibres (Sams, 1999), which are less prone to cold-shortening than the red fibres found in red meats, cold-shortening has been shown to occur (Woods and Richards, 1974; Bilgili et al., 1989). Avoiding cold-shortening in beef through the use of slow chilling rates can lead to problems of ‘bone-taint’ (as explained below). © 2008, Woodhead Publishing Limited

72

Chilled foods

Fig. 4.2

Traditional dry-ageing of beef sides.

Electrical stimulation (ES) of the carcass after slaughter can allow rapid chilling to be carried out without much of the toughening effect of cold-shortening. However, Buts et al. (1986) reported that, in veal, ES followed by moderate cooling affected tenderness in an unpredictable way and could result in tougher meat. ES will hasten rigor and cause tenderisation to start earlier at the prevailing higher temperature. In beef meat from carcasses given high or low voltage stimulation and slow cooling, adequate ageing can be obtained in about half the time of non-stimulated beef. This will therefore reduce the requirement and cost of storage. In poultry processing, ES is used to reduce the toughness of meat that is deboned prior to the normal ageing (or maturation) period (Li et al., 1993; Sams, 2002). When meat is stored at above freezing temperatures it becomes progressively more tender. This process, known as ageing (or alternatively as conditioning or maturation) is traditionally carried out by hanging the carcass for periods of 14 days or longer (in the case of beef) in a controlled environment (Fig 4.2) at between –1 and 5 °C (so called ‘dry ageing’). Alternatively, the carcass may be divided into sub-primals and aged in vacuum packaging (usually referred to as ‘wet ageing’). © 2008, Woodhead Publishing Limited

Raw material selection: meat and poultry Table 4.5

73

Time taken to achieve 50 and 80% ageing at 1 °C for different species

Species Beef Veal Rabbit Lamb Pork Chicken

Time (days) taken to achieve 50% ageing 80% ageing 4.3 4.1 4.1 3.3 1.8 0.1

10.0 9.5 9.5 7.7 4.2 0.3

The rate of ageing differs significantly between animal species (Dransfield, 1986) and necessitates different times for tenderisation. Beef, veal and rabbit age at about the same rate and take about 10 days at 1 °C to achieve 80% of ageing (Table 4.5). Lamb ages slightly faster than beef but more slowly than pork, and chicken is at least 14 times faster than pork. The ultimate tenderness will depend on the initial ‘background’ tenderness of the meat and the tenderisation that has occurred during chilling. The age of the animal is also important. In veal, acceptable tenderness can be obtained after 5 days at 1 °C compared with 10 days for beef. The major increase in tenderness has been shown to occur in less than 14 days in beef. In a study by Martin et al. (1971), in which more than 500 animals were examined, it was concluded that for beef carcasses an ageing period of 6 days was sufficient for a consumer product of satisfactory tenderness. Buchter (1970) also showed that no significant increase in tenderness occurs after 4 to 5 days for calves and 8 to 10 days for young bulls at 4 °C. In the UK, ageing has seen a revival in recent years and some UK supermarkets are currently marketing beef aged for up to 28 days, lamb aged for 14 days and pork aged for 10 days. Although ageing is rapid in poultry meat, deboning before sufficient tenderisation has taken place can result in tough meat. Studies to determine the minimum amount of ageing required before deboning show that at least 2 and possibly 4 hours are required in chicken (Sams, 1999) and at least 6 and possibly 8 hours in turkeys (Fanatico, 2003). The merits of ‘dry’ verses ‘wet’ ageing are ongoing matters of debate. What is clear is that there is greater shrink, weight loss and trim loss associated with dry ageing and hence the process is more expensive than wet ageing. In a study by Parrish et al. (1991) comparing 21-day dry and wet aged loin and rib steaks, although differences could be detected by trained panellists, no differences were detected by consumer panellists. The ageing process can be accelerated by raising the temperature of the meat, and the topic was well studied in the 1940s, 50s and 60s. Ewell (1940) found that the rate of tenderising more than doubled for each 10 °C rise, meat from a 3-yearold steer requiring only 2 days at 23 °C to reach the same tenderness as was reached after 10 days at 0 °C. Sleeth et al. (1958) showed that the tenderness, flavour, © 2008, Woodhead Publishing Limited

74

Chilled foods

aroma and juiciness of beef quarters and ribs aged for 2 to 3 days at 20 °C were comparable to those aged 12 to 14 days at 2 °C. Busch et al. (1967) demonstrated that steaks from excised muscles held at 16 °C for 2 days were more tender than those stored at 2 °C for 13 days. The microbiological hazards of high-temperature ageing were also recognised and several investigators used antibiotics and/or irradiation to control bacterial growth during high-temperature ageing (Deatherage and Reiman, 1946; Wilson et al., 1960). Although high-temperature ageing in conjunction with ultraviolet (UV) radiation was tried, its use has not expanded owing to its high cost and inability to cover all parts of the carcase (Marais, 1968). With irradiation treatment gaining more acceptance in the USA, its use together with modified atmosphere packaging to prevent microbiological growth during accelerated high-temperature ageing has been investigated (Mooha Lee et al., 1995). Irradiated steaks stored for 2 days at 30 °C were more tender than unirradiated controls stored for 14 days at 2 °C. In red meat, there have been some reports that small changes to chilling practices alone may extended the subsequent storage life by up to 50% (Gill, 1987); however, details unfortunately are sketchy. There is little evidence of any relationship between chilling rates and subsequent frozen storage life. However, there is evidence for a relationship between frozen storage life and the length of chilled storage (ageing) prior to freezing. Chilled storage of lamb for one day at 0 °C prior to freezing can reduce the subsequent storage life by as much as 25% when compared to lamb which has undergone accelerated conditioning and only 2 hours of storage at 0 °C (Winger, 1984b). It has been shown that pork that has been held for 7 days prior to freezing deteriorates at a faster rate during subsequent frozen storage than carcasses chilled for 1 and 3 days prior to freezing (Harrison et al., 1956). Ageing for periods greater than 7 days was found by Zeigler et al. (1950) to produce meat with high peroxide and free fatty acid values when stored at –18 or –29 °C. Although shorter ageing times appear to have a beneficial effect on storage life, there is obviously a necessity for it to be coupled with accelerated conditioning to prevent any toughening effects. The consumers’ environment or setting can influence their appreciation of tenderness. In one study, consumers were found to be more critical of the tenderness of beef steaks cooked in the home than those cooked in restaurants (Miller et al., 1995). The Warner-Bratzler force transition level for acceptable steak tenderness was between 4.6 and 5.0 kg in the home and between 4.3 and 5.2 kg in the restaurants.

4.5.4 Microbiology Overall, there appears to be little correlation between carcass chilling rates, or chilling systems, and bacterial numbers after chilling. However, very slow initial chilling rates can result in the development of ‘bone-taint’ in the deep tissue near the bone (James and James, 2002). Bone-taint in beef is usually localised in the region of the hip joint (where the thermal path is longest and cooling slowest) and initial contamination, stored in air at different temperatures.is © 2008, Woodhead Publishing Limited

Raw material selection: meat and poultry

Time (days)

Odour

75

Slime

20 18 16 14 12 10 8 6 4 2 0 20

10

5

0

Storage temperature (°C)

Fig. 4.3

Time (days) for odour or slime to be detected on beef sides, with average

manifested by a ‘typical sewage type odour’ or ‘putrefactive sulphide-type odour’, probably from the growth of Clostridia (though there is debate regarding the organisms that cause this phenomena). This type of bone-taint can be avoided by ensuring that the temperature in the deep leg is reduced below 15 °C within 24 hours of slaughter. Taints in pork products appear to differ from bonetaint in beef, occurring most often after processing into cured hams or gammons. The micro-organisms that usually spoil chilled meat are psychrotrophs, i.e. bacteria capable of growth close to 0 °C. Only a small proportion of the initial microflora on meat will be psychrotrophs; the majority of micro-organisms present are incapable of growth at low temperatures. As storage temperature rises, the number of species capable of growth will increase. The growth rate of microorganisms also accelerates with increasing temperature. In the accepted temperature range for chilled meat, –1.5 to 5 °C, there can be as much as an eight-fold increase in growth rate between the lower and upper temperatures. For any particular treatment, the maximum chilled storage life will be obtained by holding the meat at –1.5 °C. Chilled storage life is halved for each 2 to 3 °C rise in temperature. With beef sides stored at 0 °C, odour (Ingram and Roberts, 1976) and slime caused by the growth of micro-organisms (to 106–108 colony forming units per cm2) will be apparent after approximately 14.5 and 20 days, respectively (Fig 4.3). At 5 °C, the respective times are significantly reduced to 8 and 13 days, respectively. Similar relationships can be seen with other meats.

4.6

Transport

Developments in frozen transport in the 19th century established the international meat market. Developments in temperature control and packaging have established © 2008, Woodhead Publishing Limited

76

Chilled foods

sea transportation of chilled meat from Australia and New Zealand to European and other distant markets, and long-distance road transportation of chilled products throughout Europe, the Middle East and the USA. Air freighting is increasingly being used for high-value perishable meat products such as venison and Wagyu beef. It is particularly important that the meat is at the correct temperature before loading since the refrigeration systems used in most transport containers are not designed to extract heat from the load but to maintain the temperature of the load. In the large containers used for long-distance transportation, meat temperature can be kept within ±0.5 °C of the set point. With this degree of temperature control, transportation times of 8 to 14 weeks can be achieved. While in general it is not advisable to rely on product cooling during transportation, in the Netherlands ‘in-transport cooling’ is used, in certain circumstances, as an integral part of a processing system (which is designed for this purpose) for pork carcasses. This allows the carcasses to be dispatched on the same day they are killed. There are a number of codes of practice, such as Codex Alimentarius (1976; 1999; 2005), that cover transport practices. During transport, there is the potential for growth of pathogenic and spoilage micro-organisms under conditions of inadequate temperature control. Meat should be transported at temperatures that achieve safety and suitability objectives. European regulations allow the transport of red meat at temperatures ≤7 °C, poultry at ≤4 °C and offal ≤3 °C. However, to achieve the required storage life for the long-distance transport of vacuumpackaged meats, meat temperatures have to be maintained close to –1±0.5 °C (without freezing) to slow microbial growth to a minimum. Frozen meat should ideally be maintained at ≤–18 °C, and certainly delivered no warmer than –12 °C. Any frozen meat delivered warmer than –18 °C should be reduced immediately after delivery to –18 °C (unless it is to be used immediately). Transport containers should be pre-cooled before loading. Equipment for continuous monitoring and recording of temperatures should accompany transport vehicles and bulk containers wherever appropriate. Additionally, the conditions of transport should provide adequate protection from exogenous contamination and damage. Loading into vehicles and subsequent unloading should be as fast as practicable and the methods used should minimise product temperature rise.

4.7

Specifications

Having considered some of the factors that influence the quality of meat, the food producer needs to understand the quality requirements for the raw material to be used in their particular products and with this in mind produce a ‘specification’ against which to check the quality of the incoming raw meat. As with every other ingredient, all deliveries of raw meat need to be recorded and inspected for the purposes of recall and traceability. Raw meat must be inspected on delivery to ensure: © 2008, Woodhead Publishing Limited

Raw material selection: meat and poultry (i) (ii) (iii) (iv)

77

the meat is in appropriate condition for intended use the meat is accompanied by appropriate documentation the transport was hygienic and suitable the appropriate temperature control was observed. Ideally temperatures should be: • frozen meat ≤–18 °C (and certainly no warmer than –12 °C) • chilled meat ≤ +3 °C (and certainly no warmer than 7 °C for red meat, 4 °C for poultry meat, 3 °C for offal) • chilled minced meat ≤ +2 °C

(v) personal hygiene of delivery personnel is appropriate (vi) the meat is within ‘use–by’ dates (vii) the meat is packaged/labelled appropriately. Various aspects of meat quality are covered by legislated standards, such as those produced by the EU. Many contain minimum standards for the legal sale of raw meat and meat products. There are also codes of practice, such as those produced by Codex Alimentarius (1976; 1999; 2003), the UK Food Standards Agency (2006) and the European Chilled Food Federation (2006). However, many companies prefer to issue their own individual specifications, which are often more stringent than the legal standards. These are often influenced by the customer (retailer) to whom the producer is supplying their product. They are often also produced in consultation with their raw material supplier, since it is pointless issuing a specification that cannot be met. Specifications should give definitions of different defects, how they are to be measured, and the tolerances that are acceptable. The following categories are usually included: (i)

Meat type: The type of meat – species, in some cases breed; the composition – on-the-bone, vacuum packaged, etc., primal, sub-primal, cut, diced, minced, etc. For diced meats, the size of dice is often specified. A descriptor may be used, such as ‘lean meat shall have a good pink colour, bright in appearance and free from any dark discolouration’. A carcass classification grade may be specified. (ii) Foreign matter: Material of non-meat origin. This includes stones, soil, wool, glass, plastic, etc. There is usually a nil tolerance for all these items. (iii) Extraneous meat matter: Parts of the meat other than that to be consumed. Meat is often defined as being composed as visual lean and visual fat (e.g. 80% visual lean). A percentage of fat by chemical analysis may also be stated. There is usually a tolerance (often nil) for the presence of bone, cartilage, gristle, skin, hair, etc. (iv) Foreign meat: Meat other than the defined species. Particularly a problem with minced meats where the mincer may not have been changed between mincing. (v) Damage: Bruising or mechanical damage. (vi) Blown packs: In the case of vacuum-packaged raw meat there is often a

© 2008, Woodhead Publishing Limited

78

Chilled foods

tolerance for the number of damaged/blown packs that are allowable, or for pH. (vii) Drip: A level of acceptable drip may be specified. (viii) Microbiology: Meat producers usually provide microbiological criteria within their product specifications. The EU Microbiological Criteria Regulation 2073/2005 establishes microbiological criteria for certain micro-organisms and products, and provides rules to be complied with by food business operators when implementing the general and specific hygiene measures, although producers and suppliers may have derived their own product standards. (ix) Packaging: In the case of packaged raw meat, the packaging should be designed to: (a) protect the organoleptic and other quality characteristics of the product; (b) protect the product against microbiological and other contamination; (c) protect, as far as practicable, against dehydration, heat accumulation by radiation or conduction (sun and exhausts on vehicles), and, where appropriate, leakage; (d) not pass on to the product any odour, taste, colour or other foreign characteristics, throughout the processing (where applicable) and distribution of the product up to the time of final sale. In the EU, the labelling for beef must comply with regulations 1760/2000 and state: product name; net weight; country of origin; packaging code; use by date; batch code; supplier’s name and address. (x) Pre-frozen meat: Previously frozen meat may be delivered in a thawed state. In such a case the meat is required to be delivered at fresh meat temperatures and to be clearly labelled ‘do not refreeze’. The food processor must ensure that the meat, once delivered, is stored and treated appropriately after delivery. Meat cold stores should be operated so as to maintain a product temperature of 3 °C for chilled meat or –18 °C or lower for frozen meat, with a minimum of fluctuation. Excessive product temperature fluctuations either in range or frequency are undesirable. Variations greater than 2 °C in the air temperature should, so far as possible, be avoided. Frequent temperature checks should be carried out, preferably with recording thermometers or devices that will continually monitor storage temperatures. The air velocity in cold stores should be moderate and no higher than necessary to achieve sufficiently uniform temperatures within the store. A system of control stock rotation should be employed in cold stores.

4.8

Conclusions

In the chilled food industry there is an increasing demand for meat of a consistent, guaranteed high eating quality. Specifications should take into account the species, breed, age, feeding regime and handling of the live animal. However, the slaughter, chilling and ageing conditions will have probably the most important influence on subsequent meat quality. © 2008, Woodhead Publishing Limited

Raw material selection: meat and poultry

4.9

79

References

AASLYNG M D (2002), ‘Quality indicators for raw meat’, in Kerry J, Kerry J and Ledward D,

Meat Processing: Improving Quality, Cambridge, Woodhead Publishing, Ch. 8, 158–174. ALVARADO C Z AND SAMS A R (2002), ‘The role of carcass chilling rate in the development

of pale, exudative turkey pectoralis’, Poultry Science, 81, 1365–1370. (1973), ‘The influence of slaughter weight on the stability of pork sides in frozen storage’, Journal of the Science of Food and Agriculture, 24, 1299–1304. BENDALL J R (1972), ‘The influence of rate of chilling on the development of rigor and cold shortening’, Meat Chilling – Why and How?, Langford, Meat Research Institute, 3.1–3.6. BERRI C (2004), ‘Breeding and quality of raw poultry’, in Mead G C, Poultry Meat Processing and Quality, Cambridge, Woodhead Publishing, Ch. 2, 21–37. BILGILI S F, EGBERT W R AND HUFFMAN D L (1989), ‘Research note: Effect of postmortem aging temperature on sarcomere length and tenderness of broiler Pectoralis major’, Poultry Science, 68, 1588–1591. BLIXT Y AND BORCH E (2002), ‘Comparison of shelf-life of vacuum-packed pork and beef’, Meat Science, 60, 371–378. BOGH-SORENSEN L AND HOJMARK JENSEN J (1981), ‘Factors affecting the storage life of frozen meat products’, International Journal of Refrigeration, 4(3), 139–142. BUCHTER L (1970), ‘Development of a standardised procedure for the slaughter of experimental beef animals from the Danish Progeny Station “Egtved” ’, Proceedings of the 16th Meeting European Meat Research Workers, Bulgaria, 45. BUSCH W A, PARRISH F C JR. AND GOLL D E (1967), ‘Molecular properties of post-mortem muscle. 4. Effect of temperature on adenosine triphosphate degradation, isometric tension parameters, and shear resistance of bovine muscle’, Journal of Food Science, 32, 390. BUTS B, CASTEELS M, CLAEYS E AND DEMEYER D (1986), ‘Effects of electrical stimulation, followed by moderate cooling, on meat quality characteristics of veal Longissimus dorsi’, Meat Science, 18, 271–279. CHURCH P N AND WOOD J M (1992), The Manual of Manufacturing Meat Quality, London and New York, Elsevier Applied Science, ISBN 1 85166 628 1. CODEX ALIMENTARIUS (1976), Recommended international code of practice for the processing and handling of quick frozen foods, CAC/RCP 8–1976. CODEX ALIMENTARIUS (1999), Code of hygienic practice for refrigerated packaged foods with extended shelf life, CAC/RCP 46–(1999) CODEX ALIMENTARIUS (2003), Recommended international code of practice general principles of food hygiene, CAC/RCP 1–1969, Rev. 4 (2003). CODEX ALIMENTARIUS (2005), Code of hygienic practice for meat, CAC/RCP 58–2005. CROUSE J D, CUNDIFF L V, KOCH R M, KOOHMARAIE M AND SEIDEMAN S C (1989), ‘Comparisons of bos-indicus and bos-taurus inheritance for carcass beef characteristics and meat palatability’, Journal of Animal Science, 67(10), 2661–2668. CRYSTALL B B AND WINGER R J (1986), Composition of New Zealand Lamb as influenced by geographical area, sex of animal and cut, Meat Industry Research Institute of New Zealand (MIRINZ) Publication No. 842. CSIRO (1988), ‘Tender beef’, Meat Research News Letter, 88/4. CSIRO (1992), Does breed influence the tenderness of beef? CSIRO Newsletter 92/4. DEATHERAGE F E AND REIMAN W (1946), ‘Measurement of beef tenderness and tenderisation of beef by Tenderay process’, Food Research, 11, 525. DOYLE M E (2002), Survival and growth of bacterial pathogens on raw meat during chilling, University of Wisconsin-Madison, Food Research Institute. DRANSFIELD E (1986), ‘Conditioning of meat’, Recent advances and developments in the refrigeration of meat chilling, Meeting of IIR Commission C2, Section 1, Bristol, 61–68. DRANSFIELD E (1994), ‘Optimisation of tenderisation, ageing and tenderness’, Meat Science, 36, 105–121. BAILEY C, CUTTING C L, ENSER M B AND RHODES D N

© 2008, Woodhead Publishing Limited

80

Chilled foods

(2006), Recommendations for the production of prepackaged chilled food, ECFF. EWELL A W (1940), ‘The tenderising of beef’, Refrigeration Engineering, 39, 237–240. FANATICO A (2003), Small scale poultry processing, Fayetteville, ATTRA-National Sustainable Agriculture Information Service. FIOCCHI A, RESTANI P AND RIVA E (2000), ‘Beef allergy in children’, Nutrition, 16(6), 454– 457. FOOD STANDARDS AGENCY (2006), Guide to food hygiene and other regulations for the UK meat industry, FSA. GARCÍA DE FERNANDO G D, NYCHAS G J E, PECK M W AND ORDÓÑEZ J A (1995), ‘Growth/ survival of psychrotrophic pathogen on meat packaged under modified atmospheres’, International Journal of Food Microbiology, 28, 221–231. GARRIDO M D, PEDAUYE J, BANON S, MARQUES F AND LAENCINA J (1994), ‘Pork quality affected by different slaughter conditions and post mortem treatment of the carcasses’, Food Science and Technology, 27(20), 173–176. GILL C O (1987), ‘Prevention of microbial contamination in the lamb processing plant’, in Smulders, F J M, Elimination of Pathogenic Organisms from Meat and Poultry, Amsterdam-New York-Oxford, Elsevier, 203–219. HARRISON D L, HALL J L, MACKINTOSH D L AND VAIL G E (1956), ‘Effect of post mortem chilling on the keeping quality of frozen pork’, Food Technology, 10, 104–108. HAWRYSH Z J, PRICE M A AND BERG R T (1979), ‘The influence of cooking temperature on the eating quality of beef from bulls and steers fed three levels of dietary roughage’, Journal Institute Canadian Science Technology Aliment, 12(2), 72–77. HUFF E J AND PARRISH JR F C (1993), ‘Bovine longissimus muscle tenderness as affected by postmortem ageing time, animal age and sex’, Journal of Food Science, 58(4), 713–716. HUSBAND P M AND JOHNSON B Y (1985), ‘Beef tenderness: The influence of animal age and postmortem treatment’, CSIRO Food Research Quarterly, 45, 1–4. INGRAM M AND ROBERTS T A (1976), ‘The microbiology of the red meat carcass and the slaughterhouse’, Royal Society of Health Journal, 96(6), 270–276. IVENSEN P, HENCKEL P, LARSEN L M, MONLLAO S AND MØLLER A J (1995), ‘Tenderisation of pork as affected by degree of cold-induced shortening’, Meat Science, 40, 171–181. JAMES S J AND JAMES C (2002), Meat Refrigeration, Cambridge, Woodhead Publishing. JAMES S J AND JAMES C (2006), ‘Taking the heat out – Chilling, freezing and tempering of food’, Food Science and Technology, 20(3), 21–23. JAMES C, VINCENT C, DE ANDRADE LIMA T I AND JAMES S J (2006), ‘The primary chilling of poultry carcasses – a review’, International Journal of Refrigeration, 29(6), 847–862. JEACOCKE R E (1986), ‘The mechanism of cold shortening’, Recent advances and developments in the refrigeration of meat chilling, Meeting of IIR Commission C2, Section 4, Bristol, 235–241. JEREMIAH L E AND WILSON R (1987), ‘The effects of PSE/DFD conditions and frozen storage upon the processing yields of pork cuts’, Canadian Institute of Food Science and Technology Journal, 20, 25–30. JUDGE M D AND ABERLE E D (1980), ‘Effect of pre-rigor processing on the oxidative rancidity of ground light and dark porcine muscles’, Journal of Food Science, 45, 1736–1739. KESKINEL A, AYRES J C AND SNYDER H E (1964), ‘Determination of oxidative changes in raw materials by the 2-thiobarbituric acid method’, Food Technology, 101–104. KLOSE A A, HANSON H L, MECCHI E P, ANDERSON J H, STREETER I V AND LINEWEAVER H (1953), ‘Quality and stability of turkeys as a function of dietary fat’, Poultry Science, 32, 83–88. KOCH R M, DIKEMAN M E AND CROUSE J D (1982), ‘Characterization of biological types of cattle (cycle-III). 3. Carcass composition, quality and palatability’, Journal of Animal Science, 54(1), 35–45. KUMMEROW F A, HITE J AND KLOXIN S (1948), ‘Fat rancidity in eviscerated poultry’, Poultry Science, 6, 689–694. EUROPEAN CHILLED FOOD FEDERATION

© 2008, Woodhead Publishing Limited

Raw material selection: meat and poultry LI Y B, SIEBENMORGEN T J AND GRIFFIS C L

81

(1993), ‘Electrical-stimulation in poultry – A review and evaluation’, Poultry Science, 72, 7–22. LIU Q, LANARI M C AND SCHAEFER D M (1995), ‘A review of dietary vitamin E supplementation for improvement of beef quality’, Journal of Animal Science, 73, 3131–3140. LOCKER R H AND HAGYARD C J (1963), ‘Cold shortening in beef muscles’, Journal of the Science of Food and Agriculture, 14, 787. MALTON R AND JAMES S J (1983), ‘Drip loss from wrapped meat on retail display’, Meat Industry, 56(5), 39–41. MARAIS G J K (1968), ‘Aspects of the meat trade in the USA’, Meat Industry, Pretoria, October–December 33. MARSH B B AND LEET N G (1966), ‘Meat tenderness 3’, Journal of Food Science, 31, 450–460. MARTIN A H, FREDEEN H T AND WEISS G M (1971), ‘Tenderness of beef longissimus dorsi muscle from steers, heifers and bulls as influenced by source, post-mortem ageing and carcass characteristics’, Journal of Food Science, 36, 619. MCCLURE P J (2000), ‘Microbiological hazard identification in the meat industry’, in Brown M, HACCP in the Meat Industry, Cambridge, Woodhead Publishing, 7, 157–176. MEAD G C AND HINTON M H (1996), Microbial Control in the Meat Industry: 7. Bacterial Pathogens on Raw Meat and their Properties. Concerted Action CT94–1456. Bristol, University of Bristol Press. MILLER R K (2002), ‘Factors affecting the quality of raw meat’, in Kerry J, Kerry J and Ledward D, Meat Processing: Improving Quality, Cambridge, Woodhead Publishing, 3, 27–63. MILLER E, HOOVER L C, GUERRA A L, HUFFMAN K L, TINNEY K S, RAMSEY C B, BRITTIN H C AND HUFFMAN L M (1995), ‘Consumer acceptability of beef steak tenderness in the home and restaurant’, Journal of Food Science, 60(5), 963–965. MOOHA LEE, SEBRANEK J AND PARRISH JR F C (1995), ‘Accelerated post-mortem ageing of beef utilising electron-beam irradiation and modified atmosphere packaging’, Journal of Food Science, 1(5), 133–136. MUSA H H, CHEN G H, CHENG J H, SHUIEP E S AND BAO W B (2006), ‘Breed and sex effect on meat quality of chicken’, International Journal of Poultry Science, 5, 566–568. NESBAKKEN T (2005), ‘Biological pathogens in animals’, in Sofos J N, Improving the Safety of Fresh Meat, Cambridge, Woodhead Publishing, 1, 3–23. ORHAN F AND SEKEREL B E (2003), ‘Beef allergy: A review of 12 cases’, Allergy, 58(2), 127– 131. PALMER A Z, BRADY, D E, NAUMAN H D AND TUCKER L N (1953), ‘Deterioration in freezing pork as related to fat composition and storage treatments’, Food Technology, 7, 90–95. PARRISH F C, BOLES, J A, RUST R E AND OLSON D G (1991), ‘Dry and wet aging effects on palatability attributes of beef loin and rib steaks from three quality grades’, Journal of Food Science, 56, 601–603. PEARSON A M (1994), ‘Introduction to quality attributes and their measurement in meat, poultry and fish products’, in Pearson A M and Dutson T R, Quality Attributes and their Measurement in Meat, Poultry and Fish Products, Advances in Meat Research Series, Volume 9, Glasgow, Blackie Academic and Professional, 1, 1–33. POOL M F, HANSON H L AND KLOSE A A (1950), ‘Effect of pre-freezing hold time and antioxidant spray on storage stability of frozen eviscerated turkeys’, Poultry Science, 29, 347–350. POST (PARLIAMENTARY OFFICE OF SCIENCE AND TECHNOLOGY) (1997), Safer eating – Microbiological food poisoning and its prevention, London, Parliamentary Office of Science and Technology, ISBN 1 897941 56 0. RISTIC M (1982), ‘Influence of the water cooling of fresh broilers on the shelf life of poultry parts at –15 °C and –21 °C’, Lebensmittel, 15, 113–116. SAMS A R (1999), ‘Meat quality during processing’, Poultry Science, 78, 798–803. SAMS A (2002), ‘Post-mortem electrical stimulation of broilers’, World’s Poultry Science Journal, 58, 147–157. © 2008, Woodhead Publishing Limited

82

Chilled foods

SLEETH R B, KELLEY G G AND BRADY D E (1958), ‘Shrinkage and organoleptic characteristics

of beef aged in controlled environments’, Food Technology, 12, 86. STERN N J, LYON C E, MUSGROVE M T, DICKENS J A AND WILSON R L (1992), ‘Comparison of

spoilage rates in ground turkey and ground-beef’, Journal of Food Protection, 55(7), 518– 521. TAMPLIN L M (2002), ‘Growth of Escherichia coli O157:H7 in raw ground beef stored at 10 °C and the influence of competitive bacterial flora, strain variation, and fat level’, Journal of Food Protection, 65(10), 1535–1540. TAMPLIN L M, PAOLI G, MARMER B S AND PHILIPS J (2005), Models of the behavior of Escherichia coli O157:H7 in raw sterile ground beef stored at 5 to 46 °C, International Journal of Food Microbiology, 100, 335–344. TAYLOR A A (1972), ‘Influence of carcass chilling rate on drip in meat’, Meat Chilling: Why and How? Meat Research Institute Symposium No. 2 (ed. C. L. Cutting), 5.1–5.8. UYTTERHAEGEN L, CLAEYS E, DEMEYER D, LIPPENS M, FIEMS L O, BOUCQUÉ C Y, VAN DE VOORDE G AND BASTIAENS A (1994), ‘Effects of double-muscling on carcass quality, beef tenderness and myofibrillar protein degradation in Belgian Blue White bulls’, Meat Science, 38, 255–267. VARNAM A H AND SUTHERLAND J P (1995), Meat and Meat Products, London, Chapman and Hall, ISBN 0 412 49560 0. WEBB J E, BRUNSEN C C AND YATES J D (1974), ‘Effects of dietary fish meal level on flavour of pre-cooked frozen turkey meat’, Poultry Science, 53, 1399–1404. WEN J, MORRISSEY P A, BUCKLEY D J AND SHEEHY P J A (1996), ‘Oxidative stability and αtocopherol retention in turkey burgers during refrigerated and frozen storage as influenced by dietary α-tocopheryl acetate’, British Poultry Science, 37, 787–795. WILSON G D, BRAWN P D, CHESBRO W R, GINGER B AND WEIR C E (1960), ‘The use of antibiotics and gamma irradiation in the ageing of steaks at high temperatures’, Food Technology, 14, 143. WINGER R J (1984a), ‘Storage life and eating related quality of New Zealand frozen lamb: A compendium of irrepressible longevity’, Thermal processing and quality of foods, Elsevier Applied Science, 541–552. WINGER R J (1984b), ‘Storage life of frozen foods. New approaches to an old problem.’ Food Technology in New Zealand, 19, 75,77,81,84. WISMER-PEDERSON J AND SIVESGAARD A (1957), ‘Some observations on the keeping quality of frozen pork’, Kulde, 5, 54. WOELFEL R L AND SAMS A R (2001), ‘Marination performance of pale broiler breast meat’, Poultry Science, 80, 1519–1522. WOOD D F AND RICHARDS J F (1974), ‘Isometric tension studies on chicken pectoralis major muscle’, Journal of Food Science, 39, 525–529. ZEIGLER P T, MILLER R C AND CHRISTIAN J A (1950), ‘Preservation of meat and meat products in frozen storage’, The Pennsylvania State College, School of Agriculture, State College Bulletin No. 530.

© 2008, Woodhead Publishing Limited

5 Raw material selection: fish L. Jack, Sea Fish Industry Authority (Seafish), UK and B. Read, formerly of Seafish, UK

5.1

Introduction

Seafood is a popular part of the chilled product range and more people are turning to fish as a healthy alternative to meat (see Office for Official Publications of the European Communities, 2000). Average global seafood consumption (2004) is about 16 kg per person per year. Seafood is a wide-ranging term and includes fin and ground fish (e.g. haddock, plaice, cod, tuna), gastropods (periwinkles, seasnails), crustacean shellfish (crab, lobster, shrimp) and bivalve molluscan shellfish (oysters, cockles, clams, mussels). Seafood can be sold in an extensive number of formats. Most fish is sold either as fillets or as a prepared product and there are many chilled products that include a mix of fish and crustaceans. However, cooked crustacean products, breaded seafood, prepared seafood (for example in sauce), salted and smoked seafood, sushi and other products such as surimi, pickled or fermented fish are all popular formats for seafood consumption. Chilled seafood and seafood products may be fresh (iced) or may have been frozen and thawed before processing or sale. The low fat content of many seafood species and the beneficial effects on coronary heart disease of the n-3 polyunsaturated fatty acids found in oily (pelagic) fish are extremely important for health-conscious people, particularly in countries where cardiovascular disease is common. High standards of hygiene and good logistics and quality systems are essential in the industry as fish and shellfish are susceptible to spoilage through the supply chain (Gram and Huss, 2000). While regulatory concerns focus on foodborne illness, poor quality (spoiled or decomposed) products rarely cause illness because they are discarded before consumption. With the exception of scrombroid poison© 2008, Woodhead Publishing Limited

84

Chilled foods

ing, problems generally originate with raw materials from contaminated harvest sites, or from mishandling during or after processing. A very large amount of seafood is traded internationally. About 38% of the global live catch was exported in 2004 (value US$71.5 billion), a 51% increase from 1994. Similarly, global imports rose by 25.4% from 2000 to 2004, reaching about US$75 billion. Developed countries accounted for about 81% of the total value of imports and this fish was sold either chilled or frozen directly to consumers or processed and packaged. The top exporting countries were China (US$6.6 billion), Norway (US$4.1 billion), Thailand (US$4.0 billion), the USA (US$3.9 billion), Denmark (US$3.6 billion) and Canada (US$3.5 billion). The top importing countries were Japan (US$14.6 billion), the USA (US$12 billion), Spain (US$5.2 billion), France (US$4.2 billion), Italy (US$3.9 billion) and the UK (US$2.8 billion). Most of these countries have well-developed markets for chilled fish and products. Imports to Europe in 2004 reached US$28.3 billion (+44% since 2000 and +10% since 2003) with 46% of these imports originating outside the EU. Europe exported products to the value of US$17.1 billion, representing a growth of 51% since 2000. Shrimps were the most important commodity traded in value terms, accounting for 16.5% of internationally traded fish products in 2004. The other main groups of exported species were fin and ground fish (10.2%: e.g. hake, cod, haddock and Alaska pollock), tuna (8.7%) and salmon (8.5%); (http:// www.fao.org/newsroom/common/ecg/1000301/en/enfactsheet2.pdf ; see also http:/ /www.globefish.org/files/AUDUNCroatia2007 FISHTRADEfinal_527.ppt#342,22,CONCLUSIONS). In 2006, the UK seafood market was worth approximately £5.2 billion, with sales being divided between retail at £2.4 billion and foodservice at £2.8 billion. At the other end of the supply chain, seafood worth almost £494 million was landed into the UK by UK vessels in 2006. Although the UK has a long history as a fishing nation, the seafood market is increasingly global, with much of the high-value catch, such as langoustine and monkfish, landed into the UK being exported, while species popular in the UK domestic market, such as cod and tuna are imported. The microflora of fish is usually psychrotrophic, mainly reflecting the microorganisms present in the aquatic environment. Filter feeding bivalve molluscs (e.g. oysters) accumulate and concentrate bacteria and viruses from the environment and hence their preparation and consumption require particular attention to ensure safety. Most seafood is processed by a modern, technologically advanced and complicated industry, with high standards of hygiene. The economic losses due to spoilage are rarely quantified but the US National Research Council estimates that one-fourth of the world’s food supply is lost through microbial activity alone. The UK seafood market differs from other protein markets, as sales through foodservice outlets form a much higher proportion of the total market share. This is due partly to the prevalence of fish and chip shops in the UK and that food’s position as a traditional favourite, but also to the fact that seafood is a popular choice in restaurants. © 2008, Woodhead Publishing Limited

Raw material selection: fish

85

5.1.1 Supply Demand for seafood (all fish and shellfish) continues to grow rapidly, not just in the UK, but across the world. However, the supply is limited, as seafood (with the exception of aquaculture) remains a wild resource and, unlike farmed foodstuffs, production cannot be scaled up to meet increasing demand. This creates a predicament which has led to price rises for wild fish and an increase in aquaculture production – the farming of seafood. Because of consumer demand, the seafood market is truly global, with particular species native to a region being sold around the world for the best price (e.g. tuna from the Indian Ocean, cod from the Atlantic, nephrops (langoustine) from the North Sea). So due to the global nature of the seafood industry, most seafood is not locally caught and therefore it is frozen and defrosted through the supply chain if it is to be sold chilled, to ensure that the consumer gets products that are as fresh-tasting as possible. For prepared products it may be primary processed (e.g. cut or sliced) frozen or thawed before secondary processing (e.g. cooking). The European Union (EU) is by far the world’s biggest importer of fish, seafood and aquaculture products. Import rules for these products are harmonized throughout the EU and rules for import from non-EU countries to the EU are designed to ensure that imports meet the same high standards of hygiene and consumer safety, and, if relevant, animal health (in aquaculture), as products from the EU Member States. European Food Law, which forms the basis for import rules, uses principles based on quality management and process-oriented controls throughout the food chain, from fishing or aquaculture, through processing to the consumer’s table, to ensure safety and quality. Spot checks on the end products alone are not able to provide the required level of safety, quality and transparency to the consumer. Imports of fish and fishery products into the EU are subject to official certification, based on recognition of a competent authority (e.g. veterinary officer) monitoring hygiene, etc. in the non-EU country. Formal recognition by the EU Commission indicates that the competent authority can ensure credible inspection and controls throughout the production chain. This is a prerequisite for a country to be authorized to export to the EU and there is a list of authorized or approved establishments and countries for each type of product. National authorities in nonEU countries must also guarantee that the relevant hygiene and public health requirements are met, including EU-specific requirements on the structure of vessels, landing sites, processing establishments, operational processes including quality systems, and storage. Therefore imports are authorized only from approved vessels and establishments (e.g. processing plants, freezer or factory vessels, cold stores) that have been inspected by the competent authority of the exporting country and found to meet EU requirements. To provide the necessary assurances, the competent authority is obliged to carry out regular inspections and initiate corrective actions, if necessary. Imports of fishery products from non-EU countries must enter the EU via an approved border inspection post under the authority of an official veterinarian. Each consignment is subject to systematic document and identity checks and, as appropriate, physical checks. The frequency of physical checks depends on the © 2008, Woodhead Publishing Limited

86

Chilled foods

risk profile of the product and also on the results of previous checks. Consignments which are found not to meet EU requirements are either destroyed or, under certain conditions, may be re-dispatched within 60 days. Key supply chain controls for fish products, irrespective of origin, are certification of suppliers, provision and enforcement of specifications (which should specify quantities and size of fish), control of time and temperature (especially maintenance of the chilled temperatures by refrigeration or ice, and provision of potable water for processing and ice making), maximum microbial levels, training of operatives, an outline of the production programme to allow sufficient time for cleaning, disinfection and maintenance of equipment, and calibration of key equipment. In recent decades, the international community has increasingly focused on the governance of fisheries. This is due to the growing realization that fish stocks in different parts of the world were being increasingly harvested beyond sustainable levels and the global fishery sector is in economic and social difficulty. The need for limiting the harvesting of fish and thus ending open access in fisheries became widely recognized. The United Nations Law of the Sea Convention (1982), complemented by other related international agreements, established the global framework for the governance of capture fisheries. The Code of Conduct for Responsible Fisheries, adopted by members of the Food and Agriculture Organization (FAO) of the United Nations in 1995, is considered to be the foundation on which to promote sustainable fisheries and aquaculture development for the future (see http://www.fao.org/fishery/ccrf). The EU Common Fisheries Policy was initiated in 1983 as a measure designed to address declining fish stocks through quotas. Quotas are decided upon each December through a review process with representatives from each of the EU states, fishermen and scientists and, in effect, governs the amount of selected species, especially fin and ground fish, landed by EU fishermen. In the UK, we import the majority of the seafood that we eat, whilst we export most of what we catch. The top five species consumed in the UK are cod, haddock, salmon, tuna and prawns, which together account for approximately 70% of total seafood consumption. While much of the haddock is landed into the UK and salmon is farmed here, the vast majority of cod, tuna and prawns are imported. Ensuring that seafood comes from responsible and sustainable sources has become a priority for the retail and foodservice sectors alike. Schemes such as Marine Stewardship Council Accreditation look at the sustainability of stocks and ensure that depleted species are not at risk (see www.msc.org).

5.2

The retail sector

The total UK seafood market is worth £5.2 billion, with approximately half of this value coming from retail sales worth £2.4 billion. Within retail, chilled is the biggest sector, worth £1.13 billion. © 2008, Woodhead Publishing Limited

Raw material selection: fish

87

5.2.1 Chilled vs. frozen All species of seafood, when properly chilled or frozen, will stay fresh for longer periods than if not cooled or preserved effectively. Products brought to market in a well-preserved condition will generally command higher prices, both at wholesale and retail levels, and thus give better returns to producers. There are many factors to be taken into account when deciding whether to chill or freeze fish products for various markets, as both can produce stable products. If chilling is done rapidly and the storage temperature of fresh fish is low, quality is lost slowly as spoilage activity is inhibited. The higher the temperature of storage, the shorter the shelf-life, and roughly for every hour that fish is kept at 15–20 °C, the equivalent of 1 day’s chilled shelf-life is lost. The relationship between the shelflife of fish at 0 °C and at its storage temperature t °C is known as the relative rate of spoilage and is defined as: relative rate of spoilage at t °C = keeping time at 0 °C/ keeping time at t °C (FAO, 1995). Supermarkets have undergone significant improvements to the speed and efficiency of their logistics chains since the early 1990s. This has allowed greater emphasis on shorter shelf-life chilled products than on frozen products, and growth of the chilled sector has also been driven by the consumers’ perception of chilled products as more premium and convenient than their frozen counterparts. Chilled seafood has been central to this change, with retailers focusing on fishmonger counters and increasing the amount of dedicated chiller space for pre-packaged, prepared and deli options, resulting in exceptional growth since 1993. However, perceptions and shopping habits are again starting to change and consumers are being re-educated to the benefits of frozen food, with notable campaigns by BirdsEye and McCain’s in the UK showing the consumer that frozen food can be fresher than chilled and has a longer shelf-life. This has also been supported by the introduction of new frozen seafood products that are both innovative and of premium quality. Awareness of the benefits of frozen food, coupled with price rises in the chilled aisle, has encouraged the consumer to return to frozen seafood purchasing. Chilling is the process of cooling fish or fish products to a temperature (0 °C), approaching that of melting ice; it is done using ice, chilled water, ice slurries (of both seawater and freshwater) or refrigerated seawater. Freezing (cooling fish to storage temperatures well below zero, e.g. –18 to –30 °C) is usually done using mechanical freezing plants and allows the fish to be stored for prolonged periods of time before marketing. The shelf-life of chilled fish as a raw material for further processing is limited by spoilage and sensory and textural changes that are mainly caused by coldgrowing species of bacteria, such as Pseudomonas, Altermonas and Shewanella. Chilled storage slows down these spoilage processes, but cannot stop them; therefore, chilled fish materials and products have shelf-lives of between a few days and a month. Freezing stops many of these deteriorative processes and allows long-term storage (a year or more). It may also open the way for different types of preparation, such as flaking and dicing. Both chilling and freezing prolong shelflife by slowing the reaction rates of enzymes and bacteria, and also other chemical © 2008, Woodhead Publishing Limited

88

Chilled foods

and physical processes that adversely affect quality. A major factor affecting spoilage rates of chilled fish is physical damage, which can burst the guts, contaminating the flesh with bacteria (for this reason fish tends to be gutted at sea before being boxed) and releasing enzymes, causing spoilage of fish flesh. Subsequent spoilage rates are also affected by the initial rates of chilling and storage temperature and also by species-specific intrinsic factors such as fat content and skin characteristics (fragile or tough skin). Although microbial growth is stopped by freezing, autolysis and oxidation still occur slowly (autolytic enzymes are active even at –20 °C and below) and fat rancidity may be accelerated by ‘freezer-burn’ on the surface, caused by sublimation during frozen storage. This provides dry, desiccated environments where oxidative reactions proceed rapidly. On defrosting, growth of microbes surviving freezing resumes, so that spoilage continues in the thawed fish. Freezing will not eliminate any pre-formed toxins such as amines.

5.2.2 Focus on the chilled sector In the chilled seafood sector, growth in product volumes in the UK has been supported by the introduction of tiered (i.e. basic, standard and premium) ranges. Premium ranges such as Tesco Finest, Sainsbury’s Taste the Difference and Asda’s The Best, have put emphasis on the provenance of seafood and the traditional processes behind it, giving a retail platform for niche products. The introduction of tiering has, however, had a negative effect on some products that have become commoditized. Cold water prawns have become victims of this marketing approach, with the consumer purchasing the bottom tier of product as they cannot see a definable quality difference that would entice them to buy a higher priced option. Organic has become the new premium in terms of supermarket tiering and this is no exception when it comes to seafood, with organic seafood fetching, on average, twice the price of non-organic seafood. However, there remains a lot of confusion about organic seafood, with few consumers realizing that the nature of organic certification requires seafood that is farmed and fed organic feed; which may not tie in so easily with organic’s ‘natural’ and ‘environmentally friendly’ connotations. (see www.ams.usda.gov/nosb and www.soilassociation.org). Brands have had limited success in the chilled food aisle in the UK, with ownlabel products having the greater market share. However, the introduction of New Covent Garden and a number of branded ready meals, including Champney’s, has changed this to an extent and challenged the consumers’ perception of premium. Some key brands have been launched into the chilled sector, noticeably the No Catch brand, which was introduced into Tesco and Sainsbury’s. This premium brand marketed its farmed seafood, including the first UK farmed cod, with environmental friendly and sustainability messaging. However, the high price, almost three times that of a standard product in the case of cod, had limited sales and the company went into liquidation in early 2008.. Other premium introductions such as RR Spinks with their Arbroath smokies © 2008, Woodhead Publishing Limited

Raw material selection: fish

89

and smoked fish selections have performed extremely well. At an incrementally higher price point they offer the consumer a heritage-rich brand with wellexecuted provenance messages. The chilled ready meal market has included a number of seafood options, but producers have found it difficult to maintain quality in pre-cooked ready meals and breaded portions. This has limited their main ranges to prawns and a variety of fish pies. However, with consumers and retailers switching focus onto ready-to-cook meals, there is greater opportunity for seafood producers. Ready-to-cook meals allow the producer to retain quality, as well as fitting easily into their existing raw packing production facilities. The benefit of ready-to-cook also means that consumers get a freshly cooked meal, which is important when it comes to the taste and freshness of seafood. The introduction of the deli concept has proved a new opportunity for the chilled seafood processor, with ready-to-eat marinated prawns, glazed chunks of salmon and prepared salads containing fish becoming regular lunch time and light meal purchases for consumers. This sector provides further scope for the seafood processor as the consumer’s drive for health impacts on their purchasing decisions. Deli fish may be prepared by salting or immersion in brine (e.g. bacalao), sold in brine (e.g. shrimps) or salted, smoked, sliced and vacuum packed (e.g. smoked salmon). During salting or curing processes, water is lost from the flesh into the surrounding solution and this lowers the water activity (Aw) of the flesh, preserving the fish and extending its shelf-life. Most spoilage bacteria are inhibited at Aw 0.88, and halophilic bacteria at 0.65. Therefore 6–10% salt /water within the flesh will prevent spoilage under chilled conditions. However, milder, low salt products, e.g. 2–4% NaCl s/w are popular but may allow the growth of psychrotrophic Clostridium botulinum type E if they are incorrectly stored. Shrimps and herrings in brine or in salads may additionally be preserved with acetic, benzoic and/or sorbic acid or citric acid (pH 5.5–5.8). Salted or reduced Aw products eventually spoil with heterofermentative lactic acid bacteria (e.g. Leuconostoc sp.) and occasionally yeasts, giving slime, off odours and gas. Salted fish may also be smoked (e.g. smoked salmon, trout or herrings). Two types of smoking are used. Cold-smoking is done at 30–40 °C. During this process the flesh is not cooked, although proteins may be denatured, and bacteria are not killed and may grow. Products may be vacuum packed and spoil with Gram-negative rods, including Enterobacteriaceae and Vibrionacea. Under some circumstances spoilage may be caused by lactic acid bacteria. Cold-smoked fish have a limited shelf-life of about 1 month at 5 °C. There are high risks that cold-smoked products (e.g. salmon, trout and sea bass) will contain Cl. botulinum and Listeria monocytogenes, or may contain biogenic amines (Flick et al., 2001), but high levels of spoilage micro-organisms (e.g lactic acid bacteria) can inhibit pathogen growth. (Jeppesen and Huss 1993; see also FDA (2001a and http://www.cfsan.fda.gov/~comm/ift2list.html). Hot-smoking is done at higher temperatures, 65 °C or above. The flesh is cooked, enzymes and proteins are denatured and vegetative bacteria destroyed, considerably reducing the bacterial load but having a limited effect on numbers of © 2008, Woodhead Publishing Limited

90

Chilled foods

Cl. botulinum spores (Lindström et al., 2003) In the USA, a heat treatment of 82.2 °C for a minimum of 30 minutes is required for the processing of smoked fish to eliminate all Cl. botulinum type E spores. Commercial hot-smoking processes in Finland reduced numbers of spores of non-proteolytic Cl. botulinum by less than a thousand fold. Therefore, hot-smoked fish should be preserved or stored at or below 3 °C, to ensure product safety. During hot-smoking, the flesh loses moisture and the Aw falls, restricting the growth of micro-organisms. The drying rate has to be carefully controlled to ensure that uniform moisture levels are achieved throughout the fillets and rapid drying does not lead to case-hardening. Smoke is produced from burning wood to give the distinctive smoked taste and contains anti-microbial compounds such as phenols, but its composition has to be carefully controlled to prevent hazards to health from carcinogens being deposited on the fish surface. Typical hot-smoked products are vacuum packed (e.g. salmon and trout) and have a low microbial load (>103 CFU/g). Even after slicing and several months of storage at 5 °C ,Cl. botulinum type E will survive the process, and minimum salt levels (3–4% w/w) are required to prevent toxin formation at 5 °C (Cann and Taylor, 1979). Without vacuum packaging, some bacteria, moulds and yeast will grow on smoked fish, causing spoilage. Cooking during hot-smoking will eliminate vegetative micro-organisms, but there may be recontamination if good hygienic practices are not used during slicing or packaging, e.g. of shelled crustaceans or smoked salmon.

5.2.3 Consumer motivation The consumer is becoming more concerned about health, with 20% of meals now eaten because they are healthy and over half of consumers actively looking to buy a range of healthy products when they are shopping. As the obesity crisis develops, with 65% of men and 56% of women already obese or overweight, there is added pressure to eat healthily, with government and health professionals promoting the links between unhealthy lifestyles and chronic health problems. Health has become one of the consumer’s key motivations for purchasing seafood, and research into omega 3 fats has fuelled this demand. Omega 3 is an essential polyunsaturated fatty acid that has been linked to improved heart health. In addition there is a growing body of evidence that it can protect against cardiovascular disease, optimize brain development in unborn babies and children, prevent neuro-degeneration in old age and help with joint problems. A number of well publicized studies have indicated that increasing omega 3 intake may improve the concentration and performance of children in schools (see http://www.ific.org/ foodinsight/2001/ma/omega3fi201.cfm and http://www.eatwell.gov.uk/ healthydiet/nutritionessentials/fatssugarssalt/fats/. This has led to a vast variety of omega 3 enriched products, including bread, chicken, yoghurt and milk. In humans there are two types of essential fatty acids, omega 3 and omega 6, that cannot be manufactured by the body. Long-chain omega 3 fatty acids are derived from α-linolenic acid while omega 6 is derived from linoleic acid. The human body cannot synthesize omega 3s from simple building blocks, but it can form 20- and © 2008, Woodhead Publishing Limited

Raw material selection: fish

91

22-carbon omega 3s, eicosapentaenoic acid (EPA), and docosahexaenoic acid (DHA) from the 18-carbon α-linolenic acid. As these conversions occur competitively with omega 6 fatty acids, accumulation of EPA and DHA in tissues is more effective when they are obtained directly from food or when competing amounts of omega 6 analogues do not greatly exceed the amounts of omega 3. Omega 3 occurs naturally in oil-rich fish. Although not all seafood is rich in omega 3, it is nutrient rich, naturally low in calories, and consumers see it as a lighter alternative to meat or poultry. Consumers are becoming more familiar with cooking seafood as celebrity chefs continue to demonstrate its versatility and how quick it is to cook, and as more ready-to-cook or prepared products become available. Consumers’ main barriers to eating seafood revolve around its preparation, as many people are squeamish about handling a whole fish. Particular concerns include bones and skin, so convenient, ready-prepared packs of fish allay the majority of their fears. Convenience is a key selling point of seafood, as the majority of consumers no longer have the skills to prepare and de-bone fish themselves. Ready-prepared packs that highlight how quick fish is to cook, compared to other proteins, have become an alluring proposition to the consumer.

5.3

The supply chain

The seafood supply chain in the UK is difficult to measure due to its complexity – a summary of its components is show in Fig. 5.1. Seafood products can take a variety of routes between being landed or imported and finally reaching the consumer via the retail or foodservice sectors. In 2006, UK consumers bought £5.2 billion worth of seafood. The foodservice sector was worth £2.8 billion, while the retail market was worth a total of £2.4 billion. Looking at the retail market in more detail, canned seafood was worth £363 million, frozen seafood £696 million and chilled £1.13 billion. In the chilled sector, the breaded category was worth £860 million, the battered category £6 million, added value products (such as seafood in sauce and ready meals), £274 million, natural £657 million and smoked £224 million. In the frozen sector the categories were as follows: the breaded category was worth £254 million; the battered category £133 million; added value products £137 million; natural £171 million; and smoked £14 million. The difference in retail sales between individual categories in the two sectors indicates that consumers tend to buy battered products as a standby meal option, to keep in the freezer, while smoked seafood is bought chilled and consumed soon after purchase. In the foodservice sector, fish and chip shops contributed £1.1 billion of revenue and restaurants were worth £1.2 billion. Seafood consumed in schools accounted for £0.5 billion, boosted by government healthy eating guidelines. Fish must be served twice a week in Scottish schools (one portion of white fish and one © 2008, Woodhead Publishing Limited

92

Chilled foods

UK Consumers purchase > £5.2 billion of seafood per year

Retail in home £2.4 bn

Catering & institutions out of home £2.8 bn

Frozen/ambient retail £1.06 bn

Fish & chips – £1.1 bn

Chilled retail £1.13 bn

Schools – £0.5 bn

(Independent fishmongers £240 m)

Restaurants £1.2 bn

Wholesale distributor fresh/frozen

Exports (incl processed) £927.8 m

Processing

Fish auctions

UK seafood supply £2.8 bn UK source £787.8 m

Imports £2.0 bn

Imported seafood (incl processed, excl landings by foreign vessels and products not for human consumption) £1.9 bn

Foreign vessels landing into the UK £85.4 m

UK aquaculture £300 m

UK landed seafood by UK vessels £487.8 m (£322.5 m into Scotland)

Landings by UK vessels abroad £109.2 m

Fig. 5.1 UK seafood industry supply chain – estimates for 2006. Source: TNS, HM C&R, MFA and Seafish surveys. No total figure is available for the value of UK aquaculture so estimates from the SSPO have been used.

portion of oil-rich fish), whilst in England and Wales schools must serve white fish once a week and oil-rich fish at least once every 3 weeks. At the other end of the supply chain, UK registered vessels landed seafood worth almost £494 million into the UK last year. Aquaculture also added to this © 2008, Woodhead Publishing Limited

Raw material selection: fish

93

value, although no industry-wide figures for 2006 are currently available. However, initial figures from the Scottish Salmon Producers Organisation (SSPO) indicate that the value of salmon produced in Scottish salmon farms last year reached £300 million. Imported seafood for human consumption was worth £1.9 billion, and foreign vessels landing into the UK contributed a further £108 million, giving a total imports value of £2.0 billion. When combined with UK sources, the total UK seafood supply was £2.8 billion, although seafood worth £928 million was exported. All stages of the seafood supply chain must control material origins and processes according to hazard analysis critical control point principles, in an attempt to minimize the microbiological hazards in the final products (Regulation (EC) 852/2004 on the hygiene of foodstuffs and Office for Official Publications of the European Communities, 2000 – http://ec.europa.eu/publications/booklets/ move/20/txt_en.pdf). The HACCP plans used in all fish processing facilities should be unique for every process and for every product. A detailed study of the supply chain and process flow within a processing area is necessary to identify the hazards and CCPs. However, some general principles can be outlined because seafoods with similar microbial contaminants, and handling, processing and use instructions for consumption can be grouped into microbiological risk categories:

• High. Molluscs, including fresh and frozen mussels, clams, oysters in shell or shucked; usually eaten raw.

• Medium. Lightly preserved products (i.e. NaCl < 6% (s/w), pH > 5.0), including

•

salted, marinated and cold-smoked fish, usually eaten without cooking. Semi-preserved fish products with salt (NaCl > 6% (s/w)) in water phase, or pH < 5.0, with preservatives (sorbate, benzoate added). This group includes salted, hot-smoked and marinated fish, usually eaten without cooking. Low. Heat-processed (pasteurized, cooked or hot-smoked) fish products (including pre-cooked, breaded fillets) and crustacea, may be eaten after reheating or with no additional cooking. Heat sterilized fish packed in sealed containers (e.g. cans), often eaten with no additional cooking. Fish raw materials, fresh and frozen fish and crustaceans and dried or smokedried fish, designed to be eaten after cooking.

For all these products, the level of precautions in the the HACCP plan and especially the CCPs should recognize the level of risk and cover all the process stages minimizing the microbial load, providing decontamination and preventing re-contamination and providing the preservation system (e.g. drying, salt addition or heating, and chilled or frozen distribution and storage). A list of generic HACCP plans can be found at http://seafood.ucdavis.edu/haccp/plans.htm. Examples of generic HACCP plans for whole fish can be found at http://seafood.ucdavis.edu/ haccp/plans/fishwhol.htm; and for hot-smoked fish at http://seagrant.oregon state.edu/sgpubs/onlinepubs/i97001.html. A list of potential hazards and controls can be found at (FDA, 2001b) at Hazards and Controls guide http://www.cfsan. © 2008, Woodhead Publishing Limited

94

Chilled foods

fda.gov/~dms/haccp-2.html. There is also a generic plan for aquaculture products at http://aquanic.org/publicat/state/il-in/ces/linton.pdf .

5.3.1 Assessment of fish (raw material) quality High quality is essential to sell chilled fish and fish products. This usually refers to the appearance and freshness or degree of spoilage of the fish, but may also include factors such as absence of bones or skin. In addition, it should include safety aspects, such as absence of harmful bacteria, parasites and chemicals. Sensory quality means different things to different people and is a term which needs to be well defined to allow effective communication between suppliers, processors and retailers. During the last 50 years, many schemes have been developed for sensory analysis of raw fish. Some are widely used (e.g. the EU grading scheme: EC2406/ 96). Methods for evaluation of (fresh) fish quality can be divided into two types: sensory and instrumental, and include the Quality Index Method (QIM) developed by scientists at Scandinavian fishery institutes, and the Torry system (Shewan et al.,1953), which includes the Torry Taste Panel System and the Torry Freshness Scoring System to identify the quality of fish through sensory evaluation. The first detailed method was that developed by the Torry Research Station (Shewan et al., 1953). Initially, it scored each quality parameter independent of other parameters, and later was modified by grouping characteristic features. Quality was expressed as a score for a broad range of characteristics. The majority of industry assessment remains sensory because of the difficulties of developing reliable and consistent correlations between chemical measures and perceived quality and shelf-life (see Rodríguez-Jérez et al., 2000). In Europe, the EU scheme is widely used for quality assessment. Introduced in Council decision No. 103/76 January 1976, it specifies three quality levels, E (Extra), A, B, where E is the highest quality and below B is the level where fish is discarded as unsuitable for human consumption. The EU freshness grading scheme (now EC2406/96) is based on sensory assessment and is commonly accepted in EU countries. It is supplemented with EC91/492 laying down the health conditions for the production and the placing on the market of live bivalve molluscs and EC 91/493 laying down the health conditions for the production and the placing on the market of fishery products (see http://www.nmfs.noaa.gov/ trade/EUCONTENTS.htm). EC 91/493 is the main text for EU and non-EU countries exporting fish and fishery products and specifies EC standards for handling, processing, storing and transporting fish. There is widespread agreement that the EU grading scheme is too crude and something better is needed (see http://ec.europa.eu/research/success/en/agr/ 0261e.html and http://ressources.ciheam.org/om/pdf/c51/00600284.pdf). It still contains discrepancies as it does not take account of differences between species and uses only general parameters. Additionally, shelf-life predictions cannot be obtained directly from the freshness grades. Because the schemes are complicated, they may not be followed in practice (see http://seafood.ucdavis.edu/pubs/ qualitysafety.doc). The QIM is used for fresh and frozen cod, herring and saithe © 2008, Woodhead Publishing Limited

Raw material selection: fish

95

(see http://www.unuftp.is/Proj04/Alma%20Cardenas%20Bonilla-MexicoPRFn .pdf). QIM is based on the significant sensory parameters for raw fish, using many parameters and a scoring system (see http://www.fao.org/docrep/V7180Ev7180e09 .htm). An internet version of QIM is accessible at http://www.dfu.min.dk/qim/. QIM is a practical rating system in which the fish is inspected and demerit points are recorded. The scores for all the characteristics are then summed to give an overall sensory score, the so-called quality index. QIM gives scores of zero for very fresh fish while increasingly larger totals result as fish deteriorate. A description of the evaluation of each parameter scored is written as a guideline. In the sensory analysis of raw fish, appearance, odour, flavour and texture are evaluated by trained individuals or panels. Variations among individuals (even when trained as part of a panel) in response to materials with the same characteristics can lead to variable responses in an assessment because people differ in their appreciation of colour and sensitivity to chemical stimuli, for example rancidity or cold-storage flavours. Assessment of quality for intake control, on which price may be based, must be as objective and consistent as possible. Therefore it is important to be aware of differences when selecting and training judges for sensory analysis and quality assurance, so that they can accurately and consistently describe the features of any fish being evaluated. Subjective assessment, based on an assessor’s preference for a product, is usually used for market research and product development where consumer reaction needs to be predicted. Sensory test methods used for quality assurance can be divided into two groups: discriminative tests which determine if differences exist between samples (e.g. triangle test) and descriptive tests which describe a material and determine the nature of differences (e.g. profiling different materials). Microbial activity is a major cause of quality change and eventual spoilage of most fresh and lightly preserved seafoods; therefore, total viable count (TVC) or aerobic plate count (APC) are used in seafood standards for the EU, Japan and the USA and in microbiological specifications for purchase contracts. However, it is important to realize that only a small fraction of the micro-organisms present on a raw material or product contribute to spoilage and, consequently, the total microbial load of seafood is often a poor predictor of freshness or remaining shelf-life. Biochemical and chemical indicators have been sought to replace more time-consuming microbiological methods and are becoming available for the evaluation of seafood quality and provision of quantitative standards. It is often difficult to relate these indicators to product quality as assessed by sensory methods, but they can be useful for identifying products of very good or poor quality. Instrumental methods can be correlated with sensory quality, as they are based on specific notable compounds whose concentration increases or decreases with microbial spoilage (e.g. amines, ammonia or trimethylamine, TMA) or autolysis (e.g. aldehydes, ketones, short chain fatty acids resulting from autooxidation.). There is continuing industry pressure to develop rapid, reliable methods for freshness determination and provide prediction of fish spoilage and shelf-life; whatever methods are proposed, they need to be generally accepted as rapid and reliable. The most likely indicators include © 2008, Woodhead Publishing Limited

96

Chilled foods

amine concentration, nucleotides, protein changes and lipid oxidation, along with rapid microbiological and enzymatic methods and histological analysis (see Rodríguez-Jérez et al., 2000; http://ressources.ciheam.org/om/pdf/c5100 600292.pdf).

5.3.2 Product types and processing Chilled (iced), frozen and frozen/thawed raw materials are used to make products in many different formats. The main types are described below. Fillets and portions, packaged or unpackaged These fish may be sold without further processing after filleting, boning, etc. Products are displayed on ice or in refrigerated cabinets and availability and minimization of waste depends heavily on rapid logistics systems giving the maximum display shelf-life. Fish may be fresh, or pre-frozen fish may be thawed during transport and storage in the supply chain. Chilled, packaged fish may be sold in packs containing a modified atmosphere (MA). MA packaging means replacing the air in a pack of fish with a mixture of gases, typically a combination of carbon dioxide, nitrogen and oxygen, to inhibit growth of aerobic spoilage micro-organisms. The proportion of each component gas is fixed when the.pack is sealed and there is no further control during storage but because of microbial metabolism the compostion of the mixture may slowly change (see http:// www.fao.org/wairdocs/tan/x5956e/x5956e00.htm). Wet fish products may exude drip, and packs may contain pads to absorb this. It is prevalent if too high a proportion of carbon dioxide is used in the headspace, shelf-life is too long or fish quality changes. The problem can be minimized by choosing the right gas mixture and by introducing the correct grade of absorbent pad beneath the fish. Layering of products within MA packs should be avoided, as a single fillet or portion is more fully exposed to the action of the gases. Layering is unavoidable when packing sliced smoked fish, so these products do not gain the full benefit of modified or vacuum atmosphere packaging and may require additional preservation factors. MA packs are ineffective if they leak. Packs with faulty seals can be detected by pressing them with the hands; faulty packs will collapse. Packs should be clearly labelled according to existing regulations and should be marked with a sell-by or use-by date. A major market also exists for frozen fillets. Fish raw materials (fresh and frozen raw fish and crustaceans) for further processing Many chilled fish products will be made from further processed raw materials. Hazard analysis of these materials and products is uncomplicated. Fish are caught in the sea or freshwater, handled and processed, and finally distributed with chilling or freezing as the only means of preservation. Unpreserved products are processed and usually cooked before eating; under chilled conditions they have short shelf-lives (< 2 weeks). Fish raw materials may be contaminated with the spoilage and pathogenic bacteria associated with their aquatic environment, such © 2008, Woodhead Publishing Limited

Raw material selection: fish

97

as Cl. botulinum, Vibrio parahaemolyticus, various Vibrio sp., L. monocytogenes and Aeromonas sp. The pathogenic strains require temperatures >1 °C for growth and are competing with the normal spoilage flora, whose growth rate is much higher at low temperatures. Thus, the products are likely to be spoiled before production of toxin or development of high numbers of pathogens. When the products are cooked before consumption, this will completely eliminate the risk from pathogens on the raw materials. Histamine may be produced by histamine-producing bacteria (e.g. Morganella morganii: Klausen and Huss, 1987) and will not be eliminated by cooking or any other heat treatment so there is a risk of histamine poisoning if fish (Scombroidae) have been kept for some time at elevated temperatures (>5 °C). The presence of biotoxins and chemicals in fish depends on fish species, fishing area and season. The biotoxins are heat-stable and the risk of intoxication after consumption (raw or cooked) is high. Heavy contamination and growth of spoilage bacteria will reduce the shelf-life of products and cause commercial problems. Time and temperature control at all times (and in all steps) from catching to distribution is needed to prevent the growth of pathogenic bacteria, histamineproducing bacteria and spoilage bacteria. At t < 1 °C, no growth of pathogenic bacteria takes place. Only small and insignificant amounts of histamine may be formed, and bacterial spoilage is not inhibited but takes place slowly. Time and temperature controls for raw materials are also important for preventing oxidation and chemical spoilage. Exposure of fatty fish to ambient temperatures during handling is sufficient to cause quality loss and initiate chemical spoilage before fish enters the retail chain. Time and temperature recording at specific points in the supply chain, and especially during processing, should preferably be controlled automatically. Process flow must be designed to avoid stops and interruptions, and chill rooms must be supplied with thermometers. Where automatic temperature control is not available, visual inspection (e.g. quantity of ice) and control checks of temperature must be done daily. Sensory assessment (appearance, odour) of the raw material on reception at the factory, or immediately before processing, should be done to ensure that up to this point the material has been under control, and spoiled raw materials cannot enter processing areas. Cleaning, sanitation and factory hygiene procedures should prevent contamination and the effectiveness of control measures should be monitored daily. Packaging and temperature/time controls should be set to control chemical and autolytic spoilage. The packaging methods and materials are an integral part of the product design, providing a key part of the preservation system (e.g. MA or vacuum) and ensuring consumer satisfaction by preventing pack leakage. They are normally specified in the sales contract. Most safety and quality hazards in the production of fish products can be controlled by a routine quality assurance programme using very simple equipment and methods. Only the presence of heat-stable pre-formed biotoxins is potentially an uncontrolled hazard. Lightly preserved fish products Many ready-to-eat preserved fish products have a low salt level (<6% NaCl (w/w) © 2008, Woodhead Publishing Limited

98

Chilled foods

in aqueous phase) and a mild acid (acetic or citric) content (pH > 5.0). Preservatives (sorbate, benzoate, nitrite, smoke) may or may not be added. Such products may be made from raw or cooked material and are normally sold as ready-to-eat. They include salted, marinated and cold-smoked fish with a short shelf-life, even at chill temperatures (FDA, 2001a). Bacterial pathogens,such as L. monocytogenes and salmonella, and biogenic amines are of concern in these products, as the microorganisms occur on the raw materials and are very difficult or impossible to eliminate from the final product by this type of processing and formulation. The numbers and types of micro-organism flora are important determinants of the hygiene of the processing and storage areas.Under conditions of poor hygiene, pathogens and certainly spoilage micro-organisms may contaminate fish factories and high numbers then may be found in any niche where conditions such as temperature and nutrients allow growth or survival. The hygiene of the processing environment is important in the production of microbiologically safe and goodquality, lightly preserved products. Hygienic production with the lowest possible levels of re-contamination is essential, especially for ready-to-eat products such as smoked or pasteurized fish and shellfish. Listeria sp. can be found on both raw and preserved fish; additionally, vacuum packaging (Ben Embarek and Huss, 1993), chill storage, and salt addition may select for pathogenic L. monocytogenes against the naturally occurring microbial flora. In processing factories, Listeria sp. can commonly be found in brines and on the surfaces of machines, slicing equipment, packaging tables and floors and drains. Therefore, cleaning and disinfection should identify and be focused on the most important surfaces and machines to prevent contamination. Performance should be verified regularly with effective and appropriate hygiene-monitoring methods (e.g. swabbing, ATP visualization or visual inspection). Producers of biogenic amines should be included among pathogens and the maximum time–temperature conditions during production, distribution and storage must be specified and controlled to ensure hazardous levels are not generated. For most pathogenic bacteria, chill storage at +5 °C is sufficient, but for some psychrotrophic pathogens, e.g. L. monocytogenes and Cl. botulinum type E (which can multiply and produce toxins at temperatures down to +3 °C), lower temperatures may be needed. Spoilage is retarded by controlling raw material quality, time and temperature conditions during processing and distribution, packaging material composition (oxygen transmission rate of film) and method (degree of vacuum or MA). Heat-treated (pasteurized) fish and shellfish products Pasteurization can make products ready-to-eat by cooking them and eliminating or reducing the numbers of infectious pathogens (e.g. L. monocytogenes, Vibrio vulnificus); heat treatments may also extend shelf-life by destroying spoilage micro-organisms or by setting coatings on battered products. Cooked seafood that is heavily handled may become contaminated or cross-contaminated during processing if there are many manual operations, and poorly maintained or unhygienic equipment, utensils and factory and preliminary processing environments. © 2008, Woodhead Publishing Limited

Raw material selection: fish

99

Crustaceans in particular may be contaminated by Staphylococcus aureus, Salmonella spp., L. monocytogenes, Shigella spp., and other enteric microorganisms. In addition, poor manufacturing practices in warm or tropical areas may result in cross-contamination by indigenous pathogens, especially V. parahaemolyticus, from water or ice, if it is not of potable quality. Clostridium botulinum spores from water, soil or equipment origins may survive in products, depending on the severity and control of the heating process. Heat treatments are typically 70 °C × 2 minutes or equivalent, at the core of the product, and therefore spores remain unaffected. More severe processes (minimum 90 °C × 10 minutes) are used to make products safe for extended refrigerated shelf-life and can eliminate the spores of Cl. botulinum type E (Cann et al.,1966) and non-proteolytic B and F (the types of Cl. botulinum most commonly found in fish). Many fish products receive a heat treatment incidentally to cooking or frying during processing (e.g. cooked and breaded fish fillets, cooked shrimps and crabmeat, cook–chill products and hot-smoked fish). After the heat-treatment these products may be chilled or frozen and further processed or assembled manually or by equipment (e.g. fillers or robotic arms) before being packed and stored/distributed chilled. Some products may be designed to receive additional heating before consumption (cooked and breaded fillets, cook–chill products) or they may be eaten without further heat-treatment (hot-smoked fish, cooked shrimps). In many cases, home heating processes will warm the product but will not pasteurize it. Factory heating processes are generally targeted to eliminate L. monocytogenes. Whilst its heat resisitance is low, it is recognized as the most heat resistant of the infectious pathogens; heat processes equivalent to 70 °C × 2 minutes give safe products. The heat resistance of L. monocytogenes in cod and salmon has been studied by Ben Embarek and Huss (1993) and the results show a significantly higher heat resistance in salmon fillets compared to cod fillets, with D60 being 4.5 minutes in salmon and 1.8 minutes in cod. All ready-to-eat products are very sensitive to re-contamination after heating or pasteurization and this may lead to the presence and the potential for growth of coagulase-positive S. aureus, Enterobacteriaceae and Vibrionaceae during storage. Hence heat-treatment is a critical processing step and microbial hazards may or may not be eliminated depending on the extent of heating. The design principles for heat-treatments should provide sound quality, economic processing and preferably allow use of existing equipment, but also ensure reliable elimination of pathogens (e.g. minimum 70 °C × 2 minutes or equivalent). For a heat process to effectively eliminate harmful micro-organisms, the necessary time/temperature requirements should be based on experimental work with the equipment and process conditions to be used, and demonstrate the lethal effect of the designed heat treatment. Thermal inactivation of non-proteolytic Cl. botulinum type E spores has been investigated in rainbow trout and whitefish media between 75 and 93 °C. Decimal reduction times for the heat-resistant spore fraction in rainbow trout medium were 255, 98, and 4.2 minutes at 75, 85, and 93 °C, respectively, and those in whitefish medium were 55 and 7.1 minutes at 81 and © 2008, Woodhead Publishing Limited

100

Chilled foods

90 °C. The z values were 10.4 C° in trout medium and 10.1 C° in whitefish medium, showing that the food exerts a significant effect on the lethality of heat treatments. Re-contamination and growth after heat treatment are also serious hazards. The likely CCPs during processing of heat-treated fish products are:

• a heat treatment and heat process to effectively eliminate pathogenic bacteria • hygienic and rapid cooling to prevent spore outgrowth and re-contamination • GMP and factory hygiene/sanitation to control re-contamination and possible growth of bacteria after heat treatment

• cooling water or ice microbiological quality (e.g. potable) to prevent recontamination from this source. The criteria and limits to be used for monitoring and verifying CCPs are important and must be specified in detail. Thus, the heating conditions (e.g. water or fryer temperature, residence time or belt speed) necessary to provide the designed minimum internal temperature must be determined by experimentation, and be controlled and monitored during production. Similarly, the requirements to good manufacturing practice, factory hygiene and sanitation procedures must be validated and described in detail. For cleaning and disinfection (sanitation), three operations are involved: preparatory work or scrape down, cleaning, and disinfection. These operations should be linked together so that the final result, a clean factory, is achieved. Water is used for all cleaning and sterilizing agents, intermediate rinses and final rinse of equipment and therefore its chemical and microbiological quality is important for the effectiveness of cleaning and rinsing procedures. Potable water should be used for rinsing. Shellfish and molluscs Shellfish are harvested from the sea bottom in near-shore estuarine water (oysters and mussels) or in the tidal zone (clams and cockles) (see www.seafish.org/sea/ aquaculture.asp). After harvesting, they are sorted, washed and packed for transport to processing or sale. They may be sold either raw or after minimal heating (to open the shells), which is insufficient to reduce microbial contamination. After preliminary processing, shucked meat is washed, packed and sold chilled or frozen. Unfortunately, it is not possible to control the microbiological hazards associated with consumption of raw or lightly heated shellfish as no CCP can be identified for controlling contamination with or eliminating pathogens. However, hazards can be reduced by control of the hygiene and water quality in the growing environment and by high levels of factory hygiene, including control of process water quality. Depuration (rinsing the gills and body cavities of live shellfish) may be used to improve the safety of shellfish (molluscs and oysters) and in a number of countries this is required by regulation. Depuration involves placing the shellfish in tanks with clean circulating, disinfected (e.g. chlorine levels 2–5 ppm free chlorine) seawater. Control of storage and distribution temperatures and times are essential to stop any existing hazards increasing. Prawns and shrimps are generally sold after heat treatment and the major microbiological hazards arise from handling and cooling after heat treatment. © 2008, Woodhead Publishing Limited

Raw material selection: fish

5.4

101

Microbiology of seafood and seafood products

Fish is a perishable commodity and may contain various types of bacteria and other micro-organisms. Some are widely distributed in aquatic environments and occur naturally in tidal, estuarine or fresh waters, depending on temperature, and include both pathogenic (see Rhodes and Kator, 1988) and non-pathogenic or spoilage types. Others are chance contaminants that may be present or persist in water because of specific factors, such as locally higher temperatures, sewage or processing outfalls, or the accumulation of organic debris. Water temperature has a highly selective effect on which bacteria may be present; more cold-tolerant or psychrotrophic species (e.g. Pseudomonads, Cl. botulinum and Listeria) are more common in colder waters, whilst the mesophilic types (Enterobacteriaceae, V. cholerae and V. parahaemolyticus) are found among the natural flora of fish from the coastal and estuarine environments of temperate or warm tropical zones. For limited information on the microbiology of aquaculture see Miettinen and Wirtanen (2005, 2006). Seafood and seafood products that have not been processed to kill bacteria (e.g. by heating) may contain a mixture of these pathogens and spoilage bacteria, but the level of contamination, especially with pathogens, is normally low and it is unlikely that the numbers naturally present in raw seafood are sufficient to cause disease unless there has been growth since harvesting. However, there is an important exception: pathogens may be concentrated by filter feeders (e.g. molluscs) and hygienic, monitored growing conditions and processes (depuration) are used to ensure their microbiological safety. The EU Regulation 2073/2005 for shelled and shucked products of cooked crustaceans and molluscan shellfish contains specific microbiological criteria for Escherichia coli and coagulasepositive staphylococci at the end of the manufacturing process. Salmonella should also be absent from 25 g samples during the shelf-life of these products. There is a requirement that failure to meet these criteria should lead to improvements in production hygiene.

5.4.1 Food poisoning bacteria Although fin and ground fish products have a very good safety record, many different food-poisoning bacteria have been associated with fish and fish products, especially shellfish. Fin and ground fish can contain very low levels of pathogens, but shellfish are more hazardous. Salmonella has been found in both fish and shrimps (D’Aoust, 2000), probably as a result of contact with contaminated water, either in growing areas or during processing. Staphylococcus aureus toxin may be associated with breaded products as it can be formed in hydrated batter mixes and cause consumer illness. The toxin is a concern because it cannot be destroyed by the heating steps that may be used by the processor or the consumer. Listeria monocytogenes has often been associated with raw and smoked fish; its ability to grow at temperatures as low as 3 °C allow it to grow in refrigerated foods. © 2008, Woodhead Publishing Limited

102

Chilled foods

FDA’s draft L. monocytogenes risk assessment indicates that approximately 7% of raw fish are contaminated with from 1 to 103 CFU/g, and approximately 92% are contaminated at less than 1 CFU/g. Less than 1% of raw fish is more heavily contaminated (>103 CFU/g), and none with >106 CFU/g. One of the purposes of cooking products during processing, especially ready-to-eat products, is to eliminate vegetative pathogens or reduce them to an acceptable level. Pathogens such as L. monocytogenes may be introduced to products by raw materials and by processing – before or after a cooking step. For the design of heat treatments, L. monocytogenes is usually selected as the target micro-organism because it is regarded as the most heat-tolerant, food-borne pathogen that does not form spores. Generally, a 6-log reduction (e.g. by minimum 70 °C × 2 minutes or equi-valent) is suitable. Clostridium botulinum has been found in tuna fish, lobster, and smoked and salted fish. Clostridium botulinum type B (the most heat-resistant form of nonproteolytic Cl. botulinum) is the target pathogen. Cooking to eliminate non-proteolytic Cl. botulinum in a soup or sauce to an internal temperature of 90 °C for at least 10 minutes will pasteurize the material. However, these process times may not be sufficient to destroy non-proteolytic Cl. botulinum in materials containing lysozyme (e.g. Dungeness crabmeat), because of its potential to promote pathogen recovery after damage. A range of vibrios has been found in shellfish (e.g. oysters, clams and crabs) and associated with food poisoning (e.g. V. cholerae O1, V. parahaemolyticus and V. vulnificus: Kaysner, 2000). Cholera may be transmitted by shellfish harvested from non-polluted, warm waters since Vibrio cholerae O1 is part of their normal microflora. V. cholerae O1 has also caused sporadic cases from shellfish harvested from polluted coastal waters and consumed raw. Infections with V. parahaemolyticus have been associated with the consumption of raw, improperly cooked, poorly stored (e.g. non-refrigerated) or cooked, re-contaminated fish and shellfish. There is a correlation between the probability of infection and the warmer months of the year. As a principle, consumption of raw, improperly cooked or cooked, recontaminated shellfish offers a high risk of infection. Fish has rarely been associated with the diarrheal type B. cereus food poisoning and the putative pathogen Aeromonas hydrophila is frequently found in fish and shellfish. Since little is known about the virulence mechanisms of A. hydrophila, it is presumed that not all strains are pathogenic, given the ubiquity of the organism.

5.4.2 Viruses Viral disease transmission to humans via consumption of seafood has been known since the 1950s, and human enteric viruses appear to be the major cause of shellfish-associated disease. Viruses implicated include hepatitis type A, Norwalk virus and Astrovirus (Lees, 2000). Viruses are inert outside the living host cell, but they survive, and this means that they do not multiply in seafood. Their presence in seafood is purely as a result of contamination either via infected food handlers © 2008, Woodhead Publishing Limited

Raw material selection: fish

103

or via polluted water. Again, shellfish which are filter-feeders tend to concentrate virus from the water in which they are growing. 5.4.3 Marine biotoxins Marine biotoxins may also be formed and cause seafood-borne diseases. Important toxins are:

• ciguatera, a marine alga found in tropical and sub-tropical fish • paralytic shellfish poison from filter-feeding shellfish (e.g. mussels, clams, cockles and scallops). This is associated with the blooming of dinoflagellates (>106 cells/litre), which may cause a reddish or a yellowish discolouration of the water (see a list of potential hazards and controls at Hazards and Controls guide, http://www.cfsan.fda.gov/~dms/haccp-2.html). Control of these hazards is difficult and their presence, with the possibility of disease, cannot be entirely prevented because cooking, smoking, drying and salting do not destroy them and the appearance of the fish or shellfish flesh does not show whether they are toxic or not. The major preventive measure is inspection and sampling from fishing areas and shellfish beds, and analysis for toxins. To be effective, the monitoring requires reliable sampling plans and efficient means of detection of the toxins. Reliable chemical methods for detection of all toxins are available. Histamine poisoning (also called scombroid fish poisoning because of its frequent association with scombroid fishes – tuna and mackerel) is a chemical intoxication following the ingestion of foods that contain high levels of histamine (Lehane and Olley, 2000). It is a mild disease; its incubation period is very short (few minutes to few hours) and the duration of illness is short (few hours). The most common symptoms are facial flushing with a headache and a tingling, burning sensation in the mouth (Ababouch, 1991). Histamine is formed in fish post mortem by bacterial decarboxylation of the amino acid histidine. The fish frequently involved have a naturally high content of histidine (e.g. Scombridae) but also non-scombroid fish such as Clupeidae and mahi-mahi may be involved. Histamine is produced by some Enterobacteriaceae, Vibrio sp., a few Clostridium and Lactobacillus sp. The most potent histamine producers are among the Enterobacteriaceae (Morganella morganii, Klebsiella pneumoniae and Hafnia alvei), and these microbes can be found on many fish, probably as a result of processing or harvest contamination. However, histamine production is inhibited by chill storage (<5 °C); storage at high temperatures (10– 25 °C) for 24 hours is sufficient for histamine production, because many of the histamine-producing bacteria are mesophilic. Once histamine has been produced in fish, the risk of causing disease is high and it is very heat resistant, so even if the fish is cooked or otherwise heat-treated before consumption, the histamine is not destroyed. 5.4.4 Microbiological spoilage Initial loss of quality of fresh (non-preserved) lean or non-fatty fish species, chilled © 2008, Woodhead Publishing Limited

104

Chilled foods

or not chilled, is caused by autolytic changes, while later spoilage is due mainly to the action of bacteria.The initial microbial flora on fish is diverse, although most often dominated by Gram-negative psychrotrophic bacteria. Fish caught in tropical areas may carry a slightly higher load of Gram-positive organisms and enteric bacteria. During storage, a characteristic flora develops, but only a part of it contributes to spoilage. Spoilage organisms are the specific producers of metabolites responsible for the off-odours and off-flavours associated with spoilage. Microbiological activity can be the cause of spoilage of many preserved fish products stored at temperatures >0 °C and in many cases the bacteria responsible for the spoilage are not known. The addition of small amounts of salt and acid, as in lightly preserved fish products, changes the dominating microflora to consist mainly of Gram-positive bacterial species (lactic acid bacteria, Brochotrix); however, some Enterobacteriaceae and Vibrionaceae may also spoil these products. In products with low levels of preservation, Shewanella putrefaciens can also play a role. Shewanella putrefaciens is typical of the aerobic chill spoilage of many unpreserved fish from temperate waters, and produces TMA, and hydrogen sulphide (H2S) and other volatile sulphides, which give rise to the fishy, sulphidy cabbage like off-odours and flavours. Similar metabolites are formed by Vibrionaceae and Enterobacteriaceae during spoilage at higher temperatures. During storage in CO2-containing modified atmospheres, aerobes are inhibited and a psychrophilic Photobacterium producing large amounts of TMA becomes one of the major spoilage bacteria. Spoilage of some freshwater fish and many fish from tropical waters is characterized by a Pseudomonas type of spoilage during iced, aerobic storage, which is described as fruity, sulphydryl and sickening. Volatile sulphides (e.g. methylmer-captan (CH3SH) and dimethylsulphide ((CH3)2S), ketones, esters and aldehydes (but not hydrogen sulphide) are produced by Pseudomonas. Spoilage or putrefaction proceeds very rapidly once the load of micro-organisms exceeds approximately 107 CFU/g. More strongly preserved fish products such as salted, fermented or hot-smoked products spoil due to the action of micro-organisms. The most numerous microorganisms on these products are Gram-positive, halophilic or halotolerant micrococci, yeasts, spore formers, lactic acid bacteria and moulds. Extremely halophilic, anaerobic Gram-negative rods and halophilic yeasts have also been found and produce off-odours and flavours (sulphidy, fruity). Extreme halophile spoilage bacteria (Halococcus and Halobacterium) can cause pink discolouration of salt, brines and salted fish, as well as the off-odours and flavours normally associated with spoilage (hydrogen sulphide and indole). Some halophilic moulds (Sporendonema, Oospora) may spoil dried fish. They do not produce off odours but their presence causes adverse changes in appearance. Numbers of spoilage bacteria may be about 103–106 CFU/g at harvest and they may grow to much higher levels during storage and processing, eventually causing spoilage. Even with effective chilling, for example on ice, high levels of indigenous bacteria may be found as a result of growth and, if chilling is done slowly or is not effective, potentially hazardous micro-organisms (including © 2008, Woodhead Publishing Limited

Raw material selection: fish

105

pathogenic types) may grow, giving an increased risk of causing illness, especially in raw or in ready-to-eat products. Growth of psychrophilic bacteria (and possible toxin production, e.g. by Cl. botulinum type E) during handling and storage must therefore be prevented. Listeria monocytogenes is an important hazard as it can also grow at chill temperatures (Farber and Peterkin, 2000). Currently, the EU and the FDA in the USA require that L. monocytogenes is absent from ready-to-eat seafoods that are able to support its growth and that may pose a L. monocytogenes risk for public health. For preserved products, maximum numbers at the end of shelf-life are specified. Hence, different criteria are used depending on whether or not the product preservation system can, or cannot, inhibit the growth of Listeria during the shelf-life of the product (see Commission Regulation (EC) No 2073/2005 on microbiological criteria for foodstuffs, but this restriction does not apply to raw products that will be cooked before eating). Whether this requirement is realistic is questionable because many traditional fish processes do not contain a listericidal step, e.g. coldsmoking. For these products, low numbers of L. monocytogenes are tolerable in products, particularly when the organism can be shown to die-off during storage. Control of Enterobacteriaceae The Enterobacteriaceae, (Klebsiella, Salmonella, Shigella, E. coli) are important non-indigenous contaminants, occurring on fish products as the result of contamination from animals, birds, humans or the processing plant, and they may survive in products and processing areas for long times (months) or through processing and sanitation. Major contamination routes are from unhygienic process equipment or from water or water environments, such as ice. Good hygienic design and maintenance of equipment, with hygiene and health education of food handlers are therefore essential in the control of Enterobacteriaceae and Listeria contamination (Miettinen et al., 2001; New Zealand Food Safety Authority – see http:// www.nzfsa.govt.nz/animalproducts/seafood/guidelines/listeria/index.htm). Proper treatment (e.g. chlorination) of water and sanitary disposal of sewage are also essential parts of hygiene control. Factory hygiene, as well as personal hygiene and sanitation, can prevent contamination of products with Enterobacteriaceae and foreign material during processing. The seriousness of the risks depends on local conditions (factory lay-out and design, facilities) and intended use of product (cooked or ready-to-eat). For this reason, a detailed description of the requirements for the HACCP plan and any specifications must be produced in each individual case. Cleaning instructions must specify precisely when to clean and sanitize, how to do it, who is responsible, equipment and what cleaning/sanitizing agents are to be used. The risk of consumers ingesting Enterobacteriaceae can be minimized or eliminated by proper cooking before consumption. The heat resistance of pathogenic types is low, but varies with Aw and with the nature of solutes present. Growth is inhibited by 4–5% (s/w) NaCl (Aw about 0.94) and increased inhibition of growth is seen at low temperature or reduced pH. © 2008, Woodhead Publishing Limited

106

5.5

Chilled foods

Quality changes in seafood and seafood products

Fish tissue is rich in protein and non-protein nitrogen (e.g. amino acids, trimethylamine-oxide (TMAO) and creatinine), but low in carbohydrate giving a high post mortem pH (<6.0). The pelagic or fatty fishes have a high content of lipids, consisting mainly of triglycerides with long-chain fatty acids and phospholipids which are highly unsaturated and this has important consequences for rancid spoilage under aerobic storage conditions. Spoilage is difficult to define in objective terms and may be a matter of personal taste. It is always caused by a combination of microbiological, chemical and autolytic mechanisms. Autolysis causes the early quality changes in fresh fish, but eventually contributes very little to spoilage of chilled fish and fish products, although in ungutted fish there may be rapid development of off-odours and discolourations due to the action of gut enzymes. In frozen fish, the autolytic changes are of great importance where pre-formed TMAO is broken down by autolytic enzymes to dimethylamine (DMA) and formaldehyde (FA). The presence of FA in frozen fish increases denaturation of fish tissue and causes changes in texture and loss of water-binding capacity as pH is altered. Other enzymatic reactions, such as formation of free fatty acids, are also believed to greatly influence the sensory quality of frozen fish, but proceed at a much faster rate at high, sub-zero temperatures. Obvious signs of spoilage in chilled fish are off-odours and flavours, changes in texture, surface slime and discolouration, and gas production in sealed packaging. Spoilage occurs in well-defined phases: firstly there are autolytic changes, caused mainly by enzymes; these initially cause a slight loss of the characteristic odour and flavour which is followed by significant loss of the natural flavour and odour but without off-flavours. Later there are bacteriological changes and fish begins to show signs of spoilage. There are strong off-flavours and stale to unpleasant smells. Texture changes make the flesh become either soft and watery or tough and dry. Eventually the flesh becomes putrid, and inedible. The effect of hygiene in control of spoilage varies with the type of material and microbial contamination which may take place. Efforts to reduce the general contamination during catch handling on board has not significantly delayed the onset of spoilage as only a very small part of the general contamination is made up of specific spoilage bacteria. In contrast, targetted hygiene measures to control contamination of fish and fish products with specific spoilage bacteria greatly influences spoilage rates and shelf-life, and the sites requiring control should be identified by swabbing.

5.5.1 Chemical changes caused by micro-organisms Micro-organisms produce many of the chemicals noted as quality changes in fish. TMA is produced from TMAO by Pseudomonas, Shewanella,Vibrio, Enterobacteriaceae, and Aeromonas. Volatile sulphur compounds are produced by Shewanella and Vibrio; ammonia from amino acids by proteolytic bacteria; aldehydes, ketones and esters by Pseudomonas spp. hypoxanthine from nucleotides © 2008, Woodhead Publishing Limited

Raw material selection: fish

107

by Pseudomonas and Shewanella; and histamine from histadine by Morganella, Enterobacteriaceae, etc. The most important chemical spoilage processes take place in the lipid-rich parts of fish. Oxidative processes include autoxidation involving oxygen and unsaturated lipids. Initially, reactions lead to the formation of hydroperoxides, which are tasteless but can cause brown and yellow discolouration of the fish tissue; their breakdown gives aldehydes and ketones which have a strong rancid flavour. Auto-oxidation may be initiated and catalyzed by heat and light, or organic and inorganic catalysts (e.g. Cu and Fe). Recognized antioxidants (alphatocopherol, ascorbic or citric acid) may reduce the rate of reaction, or in some cases (e.g. frozen products) accelerate it. Chemical spoilage and the development of rancidity can be prevented by rapid catch handling on board and storage of products under anoxic conditions (vacuum or MA packaging), and the use of antioxidants may also be considered.

5.6

References

ABABOUCH L (1991). Histamine food poisoning: An update. Fish Tech News 11, 3–59. BEN EMBAREK P K AND HUSS H H (1993). Heat resistance of Listeria monocytogenes

in vacuum packaged pasteurized fish fillets. International Journal of Food Microbiology, 20, 85–95. CANN D C AND TAYLOR L (1979). The control of botulinum hazard in hot smoked trout and mackerel. Journal of Food Technology, 14, 123–129. CANN D C, WILSON B B, SHEWAN J M AND HOBBS G (1966) Incidence of Clostridium botulinum type E in fish products in the United Kingdom, Nature, 211, 205–206. D’AOUST J-Y (2000). Salmonella. In: Lund B M, Baird-Parker T C and Gould G W (eds). The Microbiological Safety and Quality of Foods. Aspen Publishers, Gaithersberg, Maryland. FAO (1995). Quality and Quality Changes in Fresh Fish, FAO Technical Paper No 348. FARBER J M AND PETERKIN P I (2000). Listeria monocytogenes. In: Lund B M, Baird-Parker T C and Gould G W (eds). The Microbiological Safety and Quality of Foods. Aspen Publishers, Gaithersberg, pp. 1178–1232. FDA (2001a). Processing Parameters Needed to Control Pathogens in Cold Smoked Fish Potential Hazards in Cold-Smoked Fish: Listeria monocytogenes, Clostridium botulinum Type E, Biogenic amines (http://www.cfsan.fda.gov/~comm/ift2list.html). FDA (2001b). Fish and Fisheries Products Hazards and Controls Guidance, 3rd edn, Center for Food Safety and Applied Nutrition, http://www.cfsan.fda.gov/~comm/haccp4o.html and www.cfsan.fda.gov/~comm/haccpsea.html (a list of potential hazards and controls can be found at Hazards and Controls guide http://www.cfsan.fda.gov/~dms/haccp2.html). FLICK G J , ORIA M P AND DOUGLAS L (2001). Potential hazards in cold-smoked fish: Biogenic Amines. Journal of Food Science – Supplement 66, S1088–S1099. GRAM L AND HUSS H H (2000). Fish and shellfish products. In: Lund B M, Baird-Parker T C and Gould G W (eds). The Microbiological Safety and Quality of Foods. Aspen Publishers, Gaithersberg, Maryland. JEPPESEN V F AND HUSS H H (1993). Antagonistic activity of two strains of lactic acid bacteria against Listeria monocytogenes and Yersinia enterocolitica in a model fish product at 5 °C. International Journal of Food Microbiology, 19, 179–186. KAYSNER C A (2000). Vibrio species. In: Lund B M, Baird-Parker T C and Gould G W (eds). The Microbiological Safety and Quality of Foods. Aspen Publishers, Gaithersberg, Maryland, pp. 1336–1362. © 2008, Woodhead Publishing Limited

108

Chilled foods

(1987). Growth and histamine production by Morganella morganii under various temperature conditions. International Journal of Food Microbiology, 5, 147–156. LEES D (2000). Viruses and bivalve shellfish. International Journal of Food Microbiology, 59, 81–116. LEHANE L AND OLLEY J (2000). Histamine fish poisoning revisited. International Journal of Food Microbiology, 58, 1–37. LINDSTRÖM M, NEVAS M, HIELM S, LÄHTEENMÄKI L, PECK M W AND KORKEALA H (2003). Thermal inactivation of nonproteolytic Clostridium botulinum type E spores in model fish media and in vacuum-packaged hot-smoked fish products. Applied and Environmental Microbiology, 69(7), 4029–4036. MIETTINEN H AND WIRTANEN G (2005). Prevalence and location of Listeria monocytogenes in farmed rainbow trout. International Journal of Food Microbiology, 104, 2, 15, 135– 143. MIETTINEN H AND WIRTANEN G (2006). Ecology of Listeria spp. in a fish farm and molecular typing of Listeria monocytogenes from fish farming and processing companies. International Journal of Food Microbiology, 112, 2, 1, 138–146. MIETTINEN H, AARNISALO K, SALO S AND SJOBERG A (2001). Evaluation of surface contamination and the presence of Listeria monocytogenes in fish processing factories. Journal of Food Protection, 64(5), 635–639. OFFICE FOR OFFICIAL PUBLICATIONS OF THE EUROPEAN COMMUNITIES (2000). Healthy food for Europe’s citizens. The European Union and food quality. Europe on the Move series ISBN 92-828-8238-1. RHODES M W AND KATOR H (1988). Survival of Escherichia coli and Salmonella spp. in estuarine environments. Applied and Environmental Microbiology, 54, 2902–2907. RODRÍGUEZ-JÉREZ J J, HERNÁNDEZ-HERRERO M M AND ROIG-SAGUÉS A X (2000). New methods to determine fish freshness in research and industry. In: Global Quality Assessment in Mediterranean Aquaculture . Zaragoza : CIHEAM-IAMZ, p. 63–69 (Cahiers Options Méditerranéennes; v. 51), Workshop of the CIHEAM Networks on Technology of Aquaculture in the Mediterranean (TECAM) and Socio-economic and Legal aspects of Aquaculture in the Mediterranean, 1999/11/29-1999/12/01, Barcelona (Spain). SHEWAN J M, MACINTOSH R G, TUCKER C G AND EHRENBERG A S (1953). The development of a numerical scoring system for the sensory assessment of the spoilage of wet white fish stored in ice. Journal of the Science of Food and Agriculture, 4(6), 283–298. KLAUSEN N K AND HUSS H H

© 2008, Woodhead Publishing Limited

6 Non-microbiological factors affecting quality and safety H. M. Brown and M. N. Hall, Campden and Chorleywood Food Research Association, UK

6.1

Introduction

As the chilled foods market has expanded and become more competitive, so have the demands for diversity, quality and longer shelf-life. Meeting these demands in a responsible, safe and cost-effective manner requires the application of an understanding of the factors that affect product safety and quality. Many problems can be avoided by applying this knowledge to a formalised HACCP approach to identify critical control points relating to quality as well as safety and to make realistic predictions of shelf-life. Considering these issues early in the product development process offers the best chance of providing a product that meets the consumer’s expectations and delivers the desired market opportunities to the company. Food is probably the most chemically complex substance that most people encounter. There are over half a million naturally occurring compounds in fresh plant food and more are formed as a result of processing, cooking and storage. They are responsible for the appearance, flavour, texture and nutritional value of the food (quality), and for its physiological effects when consumed (safety). Non-microbiological factors that affect quality and safety of chilled foods can be broadly divided into chemical, biochemical and physico-chemical factors. Each of these is dependent on properties of the food (e.g. pH, water activity) and the conditions in which the food is held (e.g. temperature, gaseous atmosphere). © 2008, Woodhead Publishing Limited

110

Chilled foods

Attention to the selection of raw materials in order to achieve high quality is paramount, since subsequent processing cannot compensate for poor-quality raw materials, particularly for chilled foods in which the perception of ‘freshness’ is one of the most important criteria for its purchase. The effects of chemical, biochemical and physio-chemical factors are rarely mutually exclusive but these categories provide a convenient framework for discussion. The effects of these factors are not always detrimental and in some instances they are essential for the development of the desired characteristics of a product. In this chapter, some of the characteristics of chemical, biochemical and physico-chemical reactions are described, along with examples that are of significance to chilled foods.

6.2

Characteristics of chemical reactions

Chemical reactions will proceed if reactants are available, if they are in a suitable form and if the activation energy threshold of the reaction is exceeded. The presence of inorganic catalysts reduces the activation energy threshold and causes reactions to proceed that would otherwise not have done so. The reaction rate is dependent on the concentration of the reactants and on the temperature. Increases in temperature speed up the random movement of reactant molecules, increasing the probability of their coming into contact. A general assumption is that for every 10 °C rise in temperature the rate of reaction doubles.

6.3

Chemical reactions of significance in chilled foods

6.3.1 Lipid oxidation Lipid oxidation is one of the major causes of deterioration in the quality of meat and meat products. Cooked meats and poultry rapidly develop a characteristic oxidized flavour, termed ‘warmed-over’ flavour (WOF) by Tims and Watts (1958). The flavour is best described as that associated with reheated meat and has been described as such by sensory assessors during free profiling of precooked meat, reheated after chill storage (Churchill et al., 1988, Lyon, 1987). Further descriptors have been defined for WOF in pork (Byrne et al., 1999a) and chicken meat (Byrne et al., 1999b) and have resulted in the development of sensory vocabularies containing 16 and 18 terms respectively. In cooked meats held at chill storage temperatures, this stale, oxidized flavour becomes apparent within a short time (48 hours) which contrasts with the slower onset of rancidity during frozen storage (weeks) (Pearson and Gray, 1983). Although WOF has generally been recognized as affecting only cooked meat, there is evidence that it develops just as rapidly in raw meat that has been ground and exposed to the air (Greene, 1969, Sato and Hegarty, 1971) and in restructured fresh meat products as a consequence of disruption of the tissue membranes and exposure to oxygen (Gray and Pearson, 1987). Nevertheless, the significance of the development of this flavour to food © 2008, Woodhead Publishing Limited

Non-microbiological factors affecting quality and safety

111

processors has increased with the advent and expansion of markets for cooked chilled ready meals such as ‘TV dinners’, airline catering, and fast food outlets. The consumer expectation in these situations is for ‘freshly prepared’ flavours. The continued development and success of fast food facilities and precooked chilled meals will depend to some extent on the ability of processors to overcome the development of WOF. Lipid oxidation has long been considered to be the primary cause of WOF, supported by studies correlating increases in WOF determined sensorily (Love, 1988) with measurements of the thiobarbituric acid (TBA) number (an indicator of lipid oxidation) (Igene et al., 1979, Igene et al., 1985, Smith et al., 1987), and identification of the volatile compounds extracted from the headspace above meat samples (St Angelo et al., 1987, Ang and Lyon, 1990, Churchill et al., 1990). As with other examples of oxidative rancidity, the process of lipid oxidation results in the formation of many different compounds, some of which are more significant than others to the undesirable odour and flavour associated with rancidity. This gives rise to a less than perfect relationship between measured chemical markers and sensory assessment of rancidity. The reactivity of food lipids is influenced by the degree of unsaturation of constituent fatty acids, their availability and the presence of activators or inhibitors. The composition of fats in meat reflects a number of factors, including the diet of the animal and the type of fat. Lipids are most abundant as either storage depot (adipose) fats or in cell membranes as phospholipids. During cooking, the unsaturated phospholipids, as opposed to the storage triglycerides, are rendered more susceptible to oxidation by disruption and dehydration of cell membranes. The higher degree of unsaturation of fatty acids in the phospholipids contributes to their more rapid rate of oxidation (Igene et al., 1981). The role of phospholipids in the formation of WOF (Igene and Pearson, 1979) and TBA reactive substances (Roozen, 1987, Pikul and Kummerow, 1991) has been demonstrated. Autoxidation of lipids is generally accepted to involve a free radical chain reaction (Fig. 6.1), which is initiated when a labile hydrogen atom is abstracted from a site on the lipid (RH) with the production of lipid radicals (R•) (initiation). Reaction with oxygen yields peroxyl radicals (ROO•) and this is followed by abstraction of another hydrogen from a lipid molecule. A hydroperoxide (ROOH) and another free radical (R•) which is capable of perpetuating the chain reaction are formed (propagation). Decomposition of the hydroperoxides involves further free radical mechanisms and the formation of non-radical products including volatile aroma compounds. Despite much research effort, the mechanism of initiation leading to the formation of the lipid (alkyl or allyl) radical (R•) in meat is still an area of confusion and debate. The involvement of iron has been established (Minotti and Aust, 1987), but beyond this various mechanisms have been suggested but not supported by conclusive evidence (Ashgar et al., 1988). The rate of formation of free radicals is increased by the presence of metal catalysts. In the case of WOF development in cooked meats, both free ferrous ions and haemoproteins, including metmyoglobin in the presence of hydrogen peroxide © 2008, Woodhead Publishing Limited

112

Chilled foods

Fig. 6.1

Free radical chain reaction.

(Asghar et al., 1988), have been shown to have a prooxidant effect. The availability of free iron is known to increase as a result of cooking (Igene et al., 1979) as haemoproteins are broken down and release free iron. The amount released is dependent on the rate of heating and the final temperature, and therefore on the method of heating. Slow heating releases more free iron than fast heating – roasting or braising of meat releases more than does microwave heating (Schricker and Miller, 1983). Procedures for the prevention of WOF were reviewed by Pearson and Gray (1983). The method used is very often restricted by the requirements of the final product. Phenolic antioxidants such as BHT and BHA are of little value in intact meat cuts (Watts, 1961), whereas they may be more suited to comminuted meat products since a more even distribution of the antioxidant can be achieved. Overheating or retorting of meat to produce compounds that have antioxidant activity (Maillard reaction products) may be suitable for canned products but tends to result in a product with characteristics contrary to the ‘fresh’ perception that is a necessary part of many chilled foods. Alternatively, these compounds can be added to meat, but they are then restricted by the same considerations that apply to artificial antioxidants. Reduction of WOF has also been achieved by use of vitamin E. Kerry et al. (1999) demonstrated that addition of alpha-tocopherol to cooked pig meat reduced lipid oxidation and WOF. Difficulties in achieving adequate distribution of the antioxidant in the meat could be overcome by the incorporation of vitamin E supplements to the feed of the animals. Addition of alpha-tocopherol acetate to the diet of rabbits (Lopez-Bote et al., 1997) and broiler chicks (O’Neill et al., 1998) has been shown to be reflected by an increase in the muscle tissue and result in reduced WOF development. Investigations of natural antioxidants present in vegetables have shown some benefits for using extracts from green peppers, onions and potato peelings (Pratt and Watts, 1964) and herbs and spices, particularly rosemary, sage, marjoram (Hermann et al., 1981) and clove (Jayathilakan et al., 1997). Reports of the effectiveness of rosemary oleoresin as an antioxidant in precooked meats are conflicting although Murphy et al. (1998) found rosemary oleoresin and sodium tripolyphosphate to be effective in the prevention of WOF in precooked roast beef slices. Precooked pork balls processed with rosemary stored at 4 °C for 48 hours did not develop oxidized flavours as the controls did (Korczack et al., 1988), whereas restructured beef steaks processed with oleoresin rosemary stored at refrigerated temperatures showed no significant improvement in comparison to the controls (Stoick et al., 1991). The addition of nitrite between 50 and 200 ppm is an effective inhibitor of the © 2008, Woodhead Publishing Limited

Non-microbiological factors affecting quality and safety

Fig. 6.2

113

Nitrosylmyoglobin. Myoglobin with the nitrite ligand.

development of WOF (Sato and Hegarty, 1971, Cho and Rhee, 1997). Nitrite and haemoproteins form nitrosylmyochrome and nitrosylhaemochrome complexes in which the iron is stabilized by the linking of nitric oxide to the porphyrin ring (Fig. 6.2); however, the pink colouration of the meat may be undesirable; causes of pinking in uncured cooked meat are further considered later in this chapter. The effectiveness of pyrophosphate, tripolyphosphate and hexametaphosphate, which chelate metal ions, particularly prooxidative ferrous ions, was demonstrated by Tims and Watts (1958) in pork. It has since been verified for ground beef (Sato and Hegarty, 1971), for restructured beef steaks (Mann et al., 1989), and for battered and breaded chicken (Brotsky, 1976). Phosphates in combination with ascorbic acid may exert a synergistic effect, such that cooked ground pork was protected against lipid oxidation for up to 35 days at 4 °C (Shahidi et al., 1986). An alternative approach is to protect the meat from oxidation. This can be achieved by creating an oxygen barrier, using a sauce or a gravy that can be in place at the time of cooking and during subsequent storage. This principle has been demonstrated by comparing the shelf-life of frozen meat to that of the same meats cooked without gravy coverings (Dalhoff and Jul, 1965). Cooked pork covered with gravy could be stored at –18 °C for more than 100 weeks, whereas pork stored without gravy was unacceptable after 22 weeks. Modified-atmosphere packaging (MAP) to reduce WOF has been applied to precooked turkey and pork products. Although those stored in nitrogen and carbon dioxide atmospheres were less ‘oxidized’ than those in air, vacuum packaging was the most effective (Nolan et al., 1989, Juncher et al., 1998). Shaw (1997) reviewed the potential benefits of the use of MAP for cook–chill ready meals. Protection against oxidation at the time of cooking is also beneficial. Cooking and subsequent storage of chicken breasts in a nitrogen atmosphere reduced the TBA values and sensory scores for WOF intensity as compared with those cooked in air and stored in either nitrogen or air (Fig. 6.3). © 2008, Woodhead Publishing Limited

114

Chilled foods

Fig. 6.3 Effect of cooking and storage atmosphere on WOF development in chicken breasts. (¤—¤) cooked and stored in air; (×—×) cooked in nitrogen, stored in air; (ü—ü) cooked in air, stored in nitrogen; (
—
) cooked and stored in nitrogen.

Autooxidation or oxidative rancidity is by no means confined to meat and meat products. Dairy products and fatty fish are also highly susceptible. Migration of copper into cream on churning can initiate the oxidative sequence of reactions causing rapid flavour impairment. Buttermilk has a high proportion of unsaturated phospholipids, particularly phosphatidylethanolamine, that can bind metal ions in a prooxidative fashion, and the presence of a metal–phospholipid complex at an oil-water interface facilitates lipid hydroperoxide formation. Fish fats contain a high proportion of n-3 polyunsaturated fatty acids, which are vulnerable to oxidation by atmospheric oxygen leading to deteriorative changes. Despite this, rancid flavours only appear to affect the acceptability of fattier species such as trout, sardine, herring and mackerel; and even then, trout and gutted mackerel oxidize at temperatures above 0 °C whereas herring remains relatively unaffected. Castell (1971) has suggested that in fish, oxidized lipids become bound in lipid–protein complexes rather than forming carbonyl compounds associated with rancid flavours. The lipid–protein complexes also contribute to the toughened texture of poorly stored fish. Competing demands for available oxygen from micro-organisms and enzymes, which differ between species, may also influence whether oxygen is available for autooxidation. In trout, reports of lipoxygenase activity in the skin tissue have suggested the potential to initiate lipid oxidation by providing a source of initiating radicals (German and Kinsella, 1985). A © 2008, Woodhead Publishing Limited

Non-microbiological factors affecting quality and safety

115

Table 6.1 Pigment types and causes giving rise to pink colouration in meat products (Brown et al., 1998) Pigment type

Cause of pink discolouration

Oxymyoglobin Nitrosomyoglobin

Low temperature cooking Nitrite contamination directly or from reduced nitrate; nitrogen oxides in ovens Carboxymyoglobin Carbon monoxide in ovens; gamma-irradiation Reduced denatured myoglobin High pH, slow cooking, high salt and availability of reducing agents

complicating factor in the assessment of the significance of oxidation to the quality of fish is that many products distributed chilled have previously been frozen, particularly, for example, herring, to spread seasonal availability. 6.3.2 Pink discolouration in meat products Discolouration in foods is a common problem which can take many forms and be associated with a wide range of chemical reactions: biochemical or enzymic browning is considered later in this chapter. Pink discolouration in cooked meats is a long-standing and all too common problem affecting manufacturing, retailing, food service and domestic sectors and is often interpreted as undercooking. The problem is particularly evident with sliced meats, reformed roast products, pasties and casseroles. Various causes of pinking have been identified and these are indicated in Table 6.1 on the basis of the pigment type thought to be involved. Maga (1994) has reviewed the causes and factors affecting pink discolouration in cooked white meats. Myoglobin is a monomeric globular haem-protein found in all vertebrates which together with haemoglobin give rise to the red colour of meats. The amount of myoglobin varies from species to species, tissue to tissue and is affected by a wide range of environmental factors. As indicated in Table 6.1 myoglobin can be present in several forms, some of which can impart a red or pink residual colour to the meat even after cooking. Recent work has indicated that over 80% of instances of pinking are due to nitrosomyoglobin arising from nitrate contamination and its subsequent bacterial reduction to nitrite (Brown et al., 1998).

6.4

Characteristics of biochemical reactions

Biochemical reactions are catalysed by specialized proteins called enzymes. They are highly specific and efficient catalysts, lowering the activation threshold so that the rate of reaction of thermodynamically possible reactions is dramatically increased. The specificity of enzymes for a particular substrate is indicated in the name, usually by attachment of the suffix ‘-ase’ to the name of the substrate on which it acts: for example, lipase acts on lipids, protease on proteins. The catalytic activity of enzymes is highly dependent on the conformational structure of the © 2008, Woodhead Publishing Limited

116

Chilled foods

protein, and many of the characteristics of enzyme-catalysed reactions result from the influence of the localized environment. Heat, extremes of acidity of alkalinity, and high ionic strength may denature the enzyme, causing impairment or loss of activity. Enzyme inhibitors and activators that bind either reversibly or irreversibly may act by causing changes in conformational structure or acting directly at the active site. The temperature at which denaturation takes place is often a reflection of the environmental conditions that the enzyme naturally operates in. For most enzymes from warm-blooded animals, denaturation begins around 45 °C, and by 55 °C rapid denaturation destroys the catalytic function of the enzyme protein; enzymes from fruit and vegetables are generally denatured at higher temperatures (70– 80 °C); and some microbial enzymes, e.g. lipases and proteases, can withstand temperatures in excess of 100 °C (Cogan, 1977). In the living cell, enzymes catalyse a vast array of reactions that taken together constitute metabolism. In the cellular environment, control and coordination of enzyme activity is achieved by means of feedback mechanisms and compartmentalization. Disruption which occurs at the time of slaughter or harvest may necessitate steps being taken to prevent the subsequent action of enzymes (blanching of vegetables is a good example); or the activity of enzymes may be enhanced if they improve product quality, as in the case of ‘conditioning’ of meats, where protease activity is used to break down muscle fibres to develop full flavour and tenderness. The rate of enzyme-catalysed reactions increases with substrate concentration but only up to a limit (maximal activity) at which the enzyme is saturated with substrate. Further increases in substrate concentration do not increase the rate of reaction. The rate of reaction increases with temperature in the same way as chemical reactions up to an optimum temperature for activity. At temperatures above this, denaturation of the enzyme protein takes place and activity is lost. At chill storage temperatures, the activity of enzymes in most foods is low, but there are notable exceptions. Enzymes in cold-blooded species may be adapted to be active at cold temperatures. In cod, lipase activity at 0 °C shows a marked lag phase before maximal activity is achieved and the rate of activity decreases to 0 °C and increases to a maximum at –4 °C. Enzymes from different sources, although catalysing conversion of the same substrates to the same reaction products, may have different characteristics in terms of rate of reaction, or pH or temperature optima, depending upon their origin. In a chilled pasta salad composed of cooked pasta, onion, red and green peppers, cucumber, sweetcorn, mushrooms and vinaigrette dressing, shelf-life was limited by browning of either the sweetcorn or the mushrooms depending on the holding temperature (Gibbs and Williams, 1990). Holding the salad at storage temperatures between 2 °C and 15 °C showed that the temperature characteristics of the browning reaction, likely to be catalysed by the enzyme polyphenoloxidase, were quite different in the mushrooms and sweetcorn (Fig. 6.4). In mushrooms, the rate of browning reaction appeared to be less temperature-sensitive than was the reaction in sweetcorn, such that at higher temperatures the shelf-life of the salad © 2008, Woodhead Publishing Limited

Non-microbiological factors affecting quality and safety

117

Fig. 6.4 Organoleptic changes in chill-stored pasta salad in vinaigrette. Temperature dependence of the rates of browning of sweetcorn (õ—õ) and mushrooms (¤—¤) (Gibbs and Williams, 1990).

was limited by browning of the sweetcorn, and at lower temperatures by browning of the mushrooms. To prevent such changes or to predict the shelf-life as a function of temperature, the subtleties of the reactions causing the changes in visual appearance need to be known. Enzymes in food may be endogenous, that is they are present naturally in the tissues of the plant or animal that comprises the food. Many hundreds of enzymes fall into this category, though not all will have a significant effect on food quality. Exogenous enzymes in food may be added by the manufacturer to perform a specific function, such as papain for the tenderization of meat, proteases for cheese ripening, or naringinase for the debittering of citrus juices particularly grapefruit juice. Enzymes may be present as a result of ‘contamination’ by migration from one food to another when they are in contact; an example would be the migration of lipases from unblanched peppers in a pizza topping to the cheese where, if the appropriate triacylglycerols are available, lipolysis will result in soapy flavours. Alternatively, there may be ‘contamination’ by extracellular enzymes from microorganisms such as lipases and proteases, where the organism may be destroyed by heat processing but the enzyme which is resistant to the heat treatment remains.

6.5

Biochemical reactions of significance in chilled foods

6.5.1 Enzymic browning In fruits and vegetables, enzymic browning occurs due to damage such as bruising and preparation procedures of cutting, peeling and slicing. The yellowish brown through to black pigments that are formed can appear very rapidly and are unappetizing. In the intact tissue the enzymes responsible, generically referred to as ‘phenolases’, are separated from the substrate. However, when they are brought into contact as a result of damage, naturally occurring phenolic compounds are © 2008, Woodhead Publishing Limited

118

Chilled foods

enzymically oxidized to form yellowish quinone compounds (Vámos-Vigyázó, 1981). A sequence of polymerization reactions follow, giving rise to brown products such as melanins. The extent of browning is dependent on the activity and amount of the polyphenoloxidase in the specific fruit or vegetable and the availability of substrates which may be catechol, tyrosine or dopamine amongst others, but there is always a requirement for oxygen. A number of approaches have been taken to prevent or retard enzymic browning. Reduction of the available oxygen concentration has been achieved via various approaches: vacuum packaging which retarded enzymic browning in potato strips (O’Beirne and Ballantyne, 1987); modified atmosphere packaging, e.g. for shredded lettuce and cut carrots (McLachlan and Stark, 1985); the addition of an oxygen scavenger to the pack, which retarded enzymic browning and textural changes in apricot and peach halves (Bolin and Huxsoll, 1989); and restricting oxygen diffusion into tissues by immersion in water, brine or syrup solutions. In contrast, high levels of oxygen (70–100%) have also been shown to reduce ascorbic acid breakdown, lipid oxidation and enzymic browning in cut lettuce probably as a result of increasing the total antioxidant capacity of the material (Day, 1998). A more direct method to prevent enzymic discolouration is to use enzyme inhibitors, though this may conflict with the ‘fresh’ image of the product or be restricted by legislation. Traditionally, the use of sulphite in the form of metabisulphite dips provided an effective means of preventing enzymic browning in many instances. With restrictions on the use of sulphite, alternatives have been sought. The pH optimum for phenolase activity is generally between pH 5 and 7. Reduction of the pH to less than 4 by the use of edible acids inactivates the enzyme. Citric acid and ascorbic acid dips retard browning by both a reduction in pH and complexation of copper which is essential for the enzyme to function. Levels of 10% ascorbic acid were shown to be effective for potatoes, and 0.5–1% for apples (O’Beirne, 1988). Phenolases from most fruits and vegetables are readily inactivated by heat (Vámos-Vigyázó, 1981) but for salads and preprepared vegetables heat treatment may not be an acceptable option owing to the concomitant changes in colour and texture.

6.5.2 Glycolysis This is a key metabolic pathway of intermediary metabolism found in almost all living organisms. Changes that take place at the time of slaughter and harvest influence the route that substrates metabolized via this pathway subsequently follow. Diversion of the pathway to produce end-products of lactic acid in meat and ethanol in vegetables have marked consequences for the subsequent quality of the food product. Adenosine triphosphate (ATP) is consumed continuously by the living cell to maintain its structure and function. It is produced from the metabolism of glycogen via glycolysis and the Krebs citric acid cycle. At slaughter, the blood supply and therefore replenishment of oxygen to the muscles ceases, but glycolytic activity continues using the stores within muscle cells. Glycogen is metabolized to pyruvate, © 2008, Woodhead Publishing Limited

Non-microbiological factors affecting quality and safety

119

but, under anaerobic conditions, the Krebs citric acid cycle is no longer functional and the pyruvate is reduced by NADH to lactic acid. The supply of NADH is replenished by glycolysis allowing the conversion of glycogen to lactic acid to continue until the glycogen stores are depleted. The breakdown of each glucose unit in muscle glycogen results in the production of two molecules of lactic acid. The accumulation of lactic acid progressively lowers the pH in the muscles, this action finally ceasing when the muscle supply of glycogen is depleted and the pH is about 5.5–5.6. When ATP is no longer generated the muscle fibres go into a state of stiffness known as ‘rigor’. Provided that there is an adequate supply of glycogen at the time of slaughter, the rate and extent of pH fall is dependent on the activity of key enzymes in the glycolytic pathway, competing reactions for adenosine diphosphate (ADP), and the temperature. The lower the temperature the longer the time taken to reach the pH limit, as biochemical reactions are slowed down. The rate of fall and the final pH can have a profound effect on the quality of the meat (Marsh et al., 1987). Lowering of muscle pH leads to protein denaturation and release of a pink proteinaceous fluid called ‘drip’. Reducing the rate at which lactic acid accumulates by rapid chilling of the carcass can dramatically reduce drip loss (Taylor, 1972, Swain et al., 1986); however, rapid chilling to temperatures below 12 °C before anaerobic glycolysis has ceased produces a condition called ‘cold shortening’, resulting in tough meat. Animals that were exhausted at the time of slaughter will have depleted glycogen reserves and produce less lactic acid during the attainment of rigor. Pork that has a pH greater than 6.0–6.2 at rigor is dark, firm, dry meat (DFD), and spoils microbiologically within 3–5 days owing to the high pH. Animals that were stressed at the time of slaughter to such an extent that respiration was anaerobic may attain rigor pH within one hour of slaughter. Pork which falls to pH 5.8 within 45 minutes of slaughter is pale, soft, exudative meat (PSE). It is characterized by excessive drip loss and is pale as a result of membrane leakage and protein denaturation. The shelf-life of such meat is reduced owing to enhanced microbial growth and oxidation of phospholipids.

6.5.3 Proteolysis Activity of proteases can have both beneficial and detrimental effects depending on the situation. Proteases in meat are important in the loss of stiffness that takes place after rigor, known as ‘conditioning’. Traditionally, conditioning is allowed to occur at the slaughterhouse and should be allowed to proceed until the meat is tender and acceptable to the consumer. Ideally this takes 2–3 weeks holding at chill temperatures; but unchilled carcasses lose stiffness sooner, as proteases act faster at higher temperatures. For beef, the conditioning rate increases with temperature up to 45 °C (Q10 2.4), then at a slower rate to 60 °C (Davey and Gilbert, 1976). The role of proteases in conditioning has been reviewed (Goll et al., 1989, Quali and Talmant, 1990). Meat proteases can be classified on the basis of preferred pH for functional activity. Proteases active at acid pH, e.g. the cathepsins, are found in © 2008, Woodhead Publishing Limited

120

Chilled foods

small organelles, lysosomes, located at the periphery of muscle cells. The stability of lysosomes decreases with a fall in pH, allowing leakage of proteases into the cell and eventually extracellular spaces. A protease active at neutral pH and thought to be involved in conditioning is calpain I which requires free calcium ions for activity. In meat, during the onset of rigor, the lack of ATP as an energy source to pump calcium ions out of cells leads to a rise in the levels of free calcium, and conditions suitable for protease activity. The duration of rigor stiffness is dependent on the species, being about one day for beef, half a day for pork and 2–4 hours for chicken. The reasons for these differences are not fully understood. Cathepsin levels are higher in chicken and pork which condition quickly (Etherington et al., 1987), and in beef the myofibrillar structure is more resistant to the action of cathepsin enzymes than it is in chicken (Mikami et al., 1987). More precise details of the proteases responsible and the conditions that control their activity have yet to be fully understood. In cheese making, the addition to the milk of proteases in rennin and the microbial starter culture causes the development of characteristic flavour and texture during ripening. Chymosin, an aspartyl protease in rennin, splits a single peptide bond in κ-casein, a milk protein, which results in clotting. A combination of the action of chymosin and proteases from the starter culture degrade casein to peptides. Many of these peptides can have bitter or sour flavours or no flavour at all, but intracellular proteases from the starter culture break the peptides down further to amino acids and small peptides which have flavour-enhancing properties. Bitter flavours in dairy products may be an adverse effect of protease activity. Peptides that are composed of predominantly non-polar amino acids tend to be bitter. In fermented dairy products, conditions that favour proteolysis and the accumulation of peptide intermediates are likely to have a bitter flavour. In fish, proteases are responsible for the condition known as ‘belly burst’. Heavy feeding prior to capture enhances the concentration and activity of gut enzymes. Unless the fish is gutted or cooled soon after capture, protease activity weakens the gut wall, allowing leakage of the contents to surrounding tissues. Herring and mackerel are notably more susceptible to belly burst; herring can become unsuitable for smoking in one day. In crustacea such as lobster and prawns the process is even more rapid, with gut enzymes attacking the flesh within hours of death. Rapid chilling and processing after catching is required.

6.5.4 Lipolysis The hydrolysis of triacylglycerols at an oil–water interface is catalysed by lipase (Fig. 6.5). The specificity of lipases varies, some being able to attack esters at all three positions in the triacylglycerol whilst others are restricted to positions 1 and 3. The activity of lipases of either endogenous or microbial origin is responsible for changes in functional properties of some dairy products such as a reduction in the skimming properties of skim milk and the churning capacity of cream, but © 2008, Woodhead Publishing Limited

Non-microbiological factors affecting quality and safety

Fig. 6.5

121

Action of lipase on triacylglycerol.

Table 6.2 Free fatty acid concentrations in dairy products and rancid flavour threshold values (Allen, 1989) Product Normal Milk powder Ice cream Butter Cheese Cheddar Brie Blue

Free fatty acid values (meq/g fat) Likely to cause problems

0.3–1.0 0.5–1.2 0.5–1.0

1.5–2.0 1.7–2.1 2.0

1.2 1.2 40.0

2.9 – –

particularly for the soapy and rancid flavours of foodstuffs. Long-chain fatty acids are usually associated with soapy flavours, and short-chain fatty acids with unpleasant rancid flavours; for example, the odour of valeric acid is described as being like ‘sweaty feet’, and hexanoic acid as ‘goat-like’. The flavour threshold of these compounds is generally low, e.g. 14 ppm for hexanoic acid, so even a very little lipolytic activity can have a marked effect on quality. In milk, the release of as little as 1–1.5% of the fatty acids from triacylglycerols can cause it to be unpalatable (Table 6.2). Endogenous milk lipases are most likely to be responsible if lipolysis occurs before the milk has been heat-treated and if the total viable count is less than 106 per ml. Flavour changes due to endogenous lipases in milk are a rare occurrence. Endogenous lipases are denatured by pasteurization, but extracellular microbial lipases released by psychrotrophic bacteria such as Pseudomonas spp are heat-stable, withstanding pasteurization and, in some cases, HTST treatments. As psychrotrophic organisms are able to grow at 2–4 °C, the preferred holding temperature for milk or cream in bulk storage tanks, significant levels of lipase may be reached. Heat-resistant lipases may take weeks to have an effect on product quality and are usually of greater significance to the quality of ambient and long shelf-life products. In cheese-making, hydrolysis due to lipase activity in the rennet may be needed to develop the required flavour (Peppler and Reed, 1987). Almost all strongly flavoured cheeses, such as Stilton, Roquefort, Gorgonzola and Parmesan, depend on free fatty acids for their flavour. With the advent of microbial proteases as rennet substitutes there is a need to add lipases with the appropriate specificity to achieve the precise mixture of fatty acids responsible for the desired flavour. The difficulties © 2008, Woodhead Publishing Limited

122

Chilled foods

associated with achieving the appropriate specificity and amounts of enzyme needed have been demonstrated with Cheddar cheeses. Differences in fatty acid levels that give a normal Cheddar and a rancid Cheddar can be reached despite extremely small differences in the amount of lipase (Law and Wigmore, 1985).

6.6

Characteristics of physico-chemical reactions

Physico-chemical reactions that affect the quality of chilled foods occur as a result of physical changes to the product or the chemical or biochemical reactions that follow. Thus migration of components either by diffusion or osmosis and light absorption by natural or artificial pigments, fall within this category.

6.7

Physico-chemical reactions of significance in chilled foods

6.7.1 Migration In mayonnaise-based salads, such as coleslaw and potato-based salads, the major quality changes observed are sensory changes related to the distribution of oil and water between the mayonnaise and vegetable tissue (Tunaley and Brocklehurst, 1982). In the case of coleslaw, a 13.5% increase in ether-extractable solids from the cabbage and a translucent appearance, indicated the uptake of oil from the mayonnaise by the cabbage within 6 hours of mixing (Tunaley et al., 1985). In the mayonnaise, the change in oil content was reflected by an increase in the polydispersity of the globule size. In addition, migration of water from the cabbage to the mayonnaise, owing to the difference in osmotic potential, caused the mayonnaise to become runny and ‘non-coating’ within the same time frame as the cabbage becoming translucent. Investigations of differences between cabbage varieties with respect to oil absorption have shown that stored Dutch cabbage gave no change in the assessment of ‘creamy-oiliness’ of the mayonnaise, whereas fresh English cabbage gave a significant decrease. Other ingredients with a large difference in osmotic potential with respect to the mayonnaise, such as celery and raisins, may also present problems owing to moisture migration resulting in the formation of pools of water on the surface of the mayonnaise. One of the most widely experienced quality changes involving the migration of water is sogginess in sandwiches. Moisture migration from the filling to the bread can be reduced by the use of fat-based spreads to provide a moisture barrier at the interface (McCarthy and Kauten, 1990). In pastry- and crust-based products such as pies and pizzas, migration of moisture from fillings and toppings to the pastry and crust causes similar problems. The migration of moisture or oils may be accompanied by soluble colours; for example, in pizza toppings where cheese and salami come into contact red streaking of the cheese is seen, and in multilayered trifles migration of colour between layers can detract from the visual appearance © 2008, Woodhead Publishing Limited

Non-microbiological factors affecting quality and safety

123

unless an appropriate strategy for colouring is used. Migration of enzymes from one component to another, for example when sliced unblanched vegetables are placed in contact with dairy products, can lead to flavour, colour or texture problems depending on the enzymes and substrates available (Labuza, 1985).

6.7.2 Evaporation A high volume of chilled foods are sold unwrapped from delicatessen counters, particularly cooked fresh meat, fish, pâtés and cheese. The shelf-life of such products differs markedly from the wrapped equivalent – six hours versus a few days to weeks. The most common cause for this reduction in shelf-life is evaporative losses. These result in a change in appearance, to such an extent that the consumer will select products which have been loaded into the cabinet most recently in preference to those which have been held in the display cabinet. The practical display-life of unwrapped meat products is determined by surface colour changes that may make the product seem unattractive. Changes in appearance are related to weight loss due to evaporation (Table 6.3). The direct cost of evaporative loss from unwrapped foods in chilled display cabinets was estimated to be in excess of £5 million per annum in 1986 (Swain and James, 1986). In stores where the rate of turnover of product is high, the average weight loss will be greater because of the continual exposure of freshly wetted surfaces to the air stream. Weight losses from the surface of unwrapped foods are dependent on the rate of evaporation of moisture from the surface and the rate of diffusion of moisture from within the product. Temperature, relative humidity and air velocity are the most influential factors affecting weight loss. Weight loss during storage of fruit and vegetables is mainly due to transpiration. Most have an equilibrium humidity of 97–98% and will lose water if kept at humidities less than this. For practical reasons, the recommended range for storage humidities is 80–100% (Sharp, 1986). The rate of water loss is dependent on the difference between the water vapour pressure exerted by the produce and the water vapour pressure in the air, and air speed over the product. Loss of as little as 5% moisture by weight causes fruit and vegetables to shrivel or wilt. As the temperature of air increases, the amount of water required to saturate it increases (approximately doubling for each 10 °C rise in temperature). If placed in a sealed container, foods will lose or gain water until the humidity inside the container reaches a value characteristic of that food at that Table 6.3 Evaporative weight loss from, and the corresponding appearance of, sliced beef topside after 6 hours’ display (James, 1985) Evaporative loss (g/cm2)

Change in appearance

Up to 0.01 0.015–0.025 0.025–0.035 0.05 0.05–0.10

Red, attractive and still wet; may lose some brightness Surface becoming drier; still attractive but darker Distinct obvious darkening; becoming dry and leathery Dry, blackening Black

© 2008, Woodhead Publishing Limited

124

Chilled foods

temperature. If the temperature is increased and the water vapour in the atmosphere remains constant, then the humidity of the air will fall. Minimizing temperature fluctuations is crucial for the prevention of moisture loss in this situation.

6.7.3 Chill injury Although low-temperature storage of fruit and vegetables is considered to be the most effective method for preserving the quality of perishable horticultural products, for chill-sensitive crops it may be more harmful than beneficial. Most fruit and vegetables of tropical and subtropical origin are injured by exposure to low but not freezing temperatures (10–15 °C) (Couey, 1982). Some temperate fruit and vegetables are also susceptible to injury, but at lower threshold temperatures (below 5 °C to 10 °C) (Bramlage, 1982). Chill injury is indicated by a range of different symptoms that adversely affect quality. Pitting, a general collapse of the tissue, is induced by dehydration and low temperatures. It is most evident in mangoes, avocados, grapefruit and limes, in which the outermost covering is harder and thicker than that underneath. Surface discolouration is common in fruits with thin soft peels, such as bell peppers, aubergines and tomatoes. Uneven or incomplete ripening is induced in tomatoes, melons and bananas. Most frequently internal breakdown and a weakening of the tissues makes the fruit or vegetable susceptible to decay by post-harvest plant pathogens. Chill injury may occur within a short space of time if temperatures are considerably below the critical level. In some cases, symptoms may only develop and become detectable after removal from cold storage and on holding at warmer temperatures, making it difficult to determine immediately after exposure to low temperatures whether chill injury has occurred. Changes in physical structures occurring at the time of chill injury have been described; however, their association with the development of symptoms of chill injury has not been established in the majority of cases. Changes in membrane lipid structure and composition (Whitaker, 1991), alterations of the cytoskeletal structure of cells and conformational changes in some regulatory enzymes and structural proteins leading to loss of compartmentalization within cells have been reported. Resulting changes in plant physiology include loss of membrane integrity, leakage of solutes, stimulation of ethylene production (Wang and Adams, 1980), and bursts of respiration (Wang, 1982). Approaches to alleviate chill injury are highly dependent on the fruit or vegetable in question (Jackman et al., 1988). The most obvious is to avoid exposure of chill-sensitive fruit and vegetables to low temperatures. However, as already stated, chilling provides a means of reducing respiration rate, evaporation and transpiration and therefore extends storage life. Temperature treatments – such as pre-storage conditioning at temperatures just above the threshold (acclimation) (suitable for cucumber and bananas); intermittent warming during storage (suitable for apples and stone fruits); or holding at ambient temperatures for a short time prior to chill storage – are effective in some cases. Controlled-atmosphere storage has been shown to be beneficial in a limited number of cases, e.g. avocados © 2008, Woodhead Publishing Limited

Non-microbiological factors affecting quality and safety

125

(Spalding and Reeder, 1975), peaches (Anderson, 1982) and okra (Ilker and Morris, 1975), but it is considered to aggravate chill injury by imposing the additional stresses of low oxygen and high carbon dioxide levels on the produce (Wade, 1979). Chemical treatments have been shown to be effective on some fruit and vegetables. On the basis that changes in membrane structure lead to chill injury, treatments leading to an alteration or protection of components of cell membranes have been used. Treatment of tomato seedlings with ethanolamine increased the levels of unsaturated fatty acids incorporated into membrane phospholipids; this reduced damage to cellular components during chilling (Ilker et al., 1976). Free radical scavengers or antioxidants such as ethoxyquin and sodium benzoate, diphenylamine and butylated hydroxytoluene have been shown to be effective on cucumbers, bell peppers (Wang and Baker, 1979) and apples (Huelin and Coggiola, 1970). Coating of fruits in waxes or oils (provided they are approved for food use) prior to chilling is effective by preventing moisture loss and reducing oxygen available for oxidation. Incorporation of the fungicides benomyl or thiabendazole (TBZ) into this type of coating has been shown to have further advantages for peaches and nectarines (Schiffmann-Nadel et al., 1975). The ultimate goal for alleviating chill injury is to select, breed or genetically engineer fruit and vegetable crops to prevent chill sensitivity. Plant breeding and selection have had varying degrees of success. A better understanding of the mechanisms responsible for chill injury should provide the insight required for targeted genetic engineering programmes to overcome this problem, though the varying causes of chill injury are unlikely to be overcome by universal solutions.

6.7.4 Syneresis The weeping or slow spontaneous movement and separation of liquid from a colloidal semi-solid mass is termed syneresis. It occurs as a result of physicochemical changes in carbohydrates or proteins which influence their ability to hold water. As a food ingredient, starch fulfils a number of essential functions – thickens, gels, stabilizes emulsions, controls moisture migration, and influences texture. An inherent limitation of native starches and flour is a lack of stability at low temperatures and at fluctuating temperatures. At low temperatures they become prone to weeping or syneresis. Native starch is a complex carbohydrate composed of the homopolymers, amylose and amylopectin. Amylose is a linear chain molecule composed of 1, 4linked α-D-glucopyranose building blocks. Amylopectin has a backbone structure like amylose but, in addition, 1, 6-linkages give it a branched structure that confers a greater water-holding capacity than amylose. The ratio of amylopectin to amylose therefore alters the properties and texture of a starch. For example, wheat flour, a traditional thickener used in gravies and sauces, provides desirable flavour and opacity, but has no low-temperature stability and chilling results in syneresis. When the starch grains are swollen, the linear amylose molecules tend to leach out © 2008, Woodhead Publishing Limited

126

Chilled foods

into solution and reassociate into aggregates aligned by hydrogen bonding. The reassociated amylose tends to expel water resulting in opacity and syneresis. Cooling or freezing causes the overall structure to shrink, greatly accelerating the rate at which syneresis occurs. Problems of syneresis often occur as a result of improper selection of starch. Incorporation of stabilized waxy maize-based starches into products that are to be chilled resists retrogradation and syneresis. Alternatively, stabilized starches are available that have been modified specifically with monofunctional blocking groups to prevent associations between leached amylose molecules and thereby prevent syneresis. Use of a modified starch in conjunction with a wheat flour provides stability in the final product. Syneresis in milk is known as ‘wheying off’, the point at which the curds and whey separate. It is obviously desirable for cheese making, but not in milk-based products such as yoghurts. Homogenization of milk for yoghurt production decreases syneresis by increasing the hydrophilicity and water-binding capacity by enhancing casein and fat globule membrane interactions and other protein– protein interactions (Tamime and Deeth, 1980). Heat processing for yoghurt manufacture (85 °C for 30 minutes, or 90–95 °C for 5–10 minutes) is unique in dairy processing. It is believed to bring about important changes in the physicochemical structure of the proteins, which minimizes syneresis and results in maximal firmness of the yoghurt coagulum.

6.7.5 Staling The market for sandwiches containing a wide variety of fillings that need chill storage has grown considerably. However, the staling of bread is one of the few reactions that has a negative temperature coefficient (McWeeney, 1968); that is, bread stales more rapidly at reduced temperatures (Meisner, 1953). The term ‘staling’ in relation to bread is used to describe an increase in crumb firmness and crumb-texture hardness, loss of crust crispness and increased toughness, and disappearance of the fresh bread flavour and emergence of a stale bread flavour. Despite extensive research into the mechanism of staling, most researchers are only prepared to agree that firmness changes are attributed to physico-chemical reactions of the starch component, mainly due to its amylopectin fraction, and some include involvement of flour proteins. The shelf-life of commercial bread is considered to be two days (Maga, 1975), which will be reduced by holding at chill temperatures. The use of modified atmosphere packaging, particularly carbon dioxide, is believed to slow the rate of staling of bread (Avital et al., 1990).

6.8

Non-microbiological safety issues of significance in chilled foods

Non-microbiological safety issues associated with chilled foods are rarely a result © 2008, Woodhead Publishing Limited

Non-microbiological factors affecting quality and safety

127

of, or exacerbated by, chilled storage temperature. Some arise as a consequence of the ingredient combinations or minimal processing that subsequent chill storage enables. In most instances, judicious selection of raw materials and a carefully tailored monitoring programme, based on an assessment of the risks posed by individual ingredients and the final product, contributes to the assurance of product safety. If possible, it is always preferable for shelf-life to be limited by changes in quality rather than safety because changes in quality can usually be discerned by the smell, taste or appearance of the product, but such changes cannot be relied upon to indicate when safety limits the shelf-life.

6.8.1 Natural toxicants There is a tendency to associate ‘natural’ with a wholesome and healthy image, yet in some cases there is an awareness that some naturally occurring chemical compounds in food may contribute to human illness. Such an example is greening of potatoes, which is commonly associated with the potential to cause harm. Glycoalkaloids, the group of toxic compounds that can be found in potatoes stored under stress conditions, accumulate just beneath the peel and at eye regions, so peeling reduces potential human exposure. Cooking is not thought to reduce glycoalkaloid concentrations (Bushway and Ponnampalam, 1981). However, as a result of dietary advice to increase the intake of fibre, an increasing number of potato products incorporate or retain the skin, e.g. chilled, filled, baked potatoes and potato skins; such products could present a higher risk. It has been generally agreed that tubers for human consumption should not exceed 20 mg glycoalkaloid per 100 g fresh tuber weight. Monitoring of the levels of glycoalkaloids is advised, particularly in new cultivars and after changes in storage and processing procedures, to ensure that they do not exceed recommended limits. Pulses and grain legumes have long been known to contain highly toxic lectins (haemagglutinins) which agglutinate red blood cells. Haemagglutinins have been detected in a wide range of leguminous seeds including lentils, soyabeans, lima or butter beans, and red kidney beans (Liener, 1974). During the last decade a number of incidents of food poisoning have been associated with red kidney beans, and in one case with butter beans (Bender and Reaidi, 1982, Rodhouse et al., 1990). A tendency to partially cook pulses or to eat them raw, particularly red kidney beans in salads, led to numerous cases of gastrointestinal disturbances. Soaking of beans for at least 8 hours leaches out lectins and boiling in fresh water for at least 10 minutes heat-inactivates any that remain, preventing the possibility of food poisoning. The inclusion of nuts, figs and dates in exotic salads such as hosaf carries the associated risk of contamination by mycotoxins (fungal toxins). Mycotoxins are contaminants rather than natural toxicants, being secondary metabolites of the fungal species e.g. Aspergillus, Penicillium, and Fusarium. These fungal species grow on a wide variety of substrates, most notably cereals and ground nuts and other high carbohydrate seeds (e.g. figs) under environmental conditions ranging from tropical to domestic refrigeration temperatures. Unfortunately, mycotoxin © 2008, Woodhead Publishing Limited

128

Chilled foods

production is associated with storage conditions designed to prevent fungal growth. Mycotoxins are chemically very diverse (they include groups such as aflatoxins, ochratoxins and trichothecenes), ranging in molecular complexity and toxicity (some are extremely toxic and others are carcinogenic). Control of mycotoxin contamination has focused on treatments to prevent mould growth or mycotoxin production during storage (Moss and Frank, 1987), and on the development of improved analytical methods for their detection. Improved quality of raw materials and post harvest treatments, coupled with improved storage and distribution conditions, reduces the incidence of contamination. In keeping with the principles of HACCP awareness of the possibility of mycotoxin contamination should be accompanied by the implementation of a suitable monitoring programme, based on an assessment of the potential risks involved, and written into the raw material specifications.

6.8.2 Phycotoxins Toxic compounds produced by algae (phycotoxins) enter the food chain via seafood, usually either shellfish (shellfish toxins) or finfish (ciguatoxins). The growing awareness of the beneficial dietary effects associated with eating fish and seafood products and the availability of the chill chain to distribute these products has resulted in an increase in their geographic availability and consumption (Przybyla Wilkes, 1991). Importation of seafoods means more exotic forms of phycotoxin are now potentially found on a global scale (Scoging, 1991). Four different forms of shellfish poisoning are recognized: paralytic shellfish poisons (PSP), diarrhetic shellfish poisons (DSP), amnesic shellfish poisons (ASP) and neurotoxic shellfish poisons (NSP). Shellfish, particularly bivalve molluscs, e.g. mussels, clams and oysters, accumulate these toxins and are unharmed by them. Subsequent consumption of shellfish by humans produces immediate and severe effects, depending on the type of toxin involved. Accumulation of toxins by shellfish coincides with high levels of particular algal species in coastal waters, so-called ‘algal blooms’. These result from increased availability of nutrients and light in surface waters associated with seasonal climate and hydrographic changes. In the UK, extensive monitoring is undertaken by the Food Standards Agency during high-risk periods. Coastal waters, shellfish and some crustacea are analysed for PSP toxins. Prohibition orders on the collection of shellfish are put in place when toxins accumulate to levels which are regarded as unsafe for human consumption (West et al., 1985). This is currently believed to be the most effective control method, as the toxins which the shellfish accumulate are reduced, but not eliminated, by cooking (Krogh, 1987) or by holding shellfish in purification tanks. PSP is linked with algal species which occur in waters where ambient temperatures are around 15–17 °C. Initial symptoms, seen within 30 minutes of consumption, are tingling and numbness in the mouth and fingertips which spreads throughout the body, causing impaired muscle coordination and, in severe cases, paralysis. The major toxin is saxitoxin, though 18 other toxic derivatives have been © 2008, Woodhead Publishing Limited

Non-microbiological factors affecting quality and safety

129

identified which are either natural algal toxins or metabolized derivatives found in shellfish. DSP intoxications are common in Japan, but outbreaks have also been recorded in France, Italy and The Netherlands. Symptoms occurring within 30 minutes of consumption are vomiting, abdominal pain and diarrhoea. The major toxic components are okadaic acid and dinophysic toxins found in mussels, clams and scallops. Denaturation of these toxins only occurs after processing at 100 °C for 163 minutes; therefore monitoring and prohibition orders are the only real safeguard. ASP is believed to be caused by a toxic amino acid, domoic acid, produced by a diatom occurring in US, Japanese and Canadian coastal waters. Symptoms include nausea, diarrhoea and confusion/disorientation headaches and, in severe cases, memory loss. NSP intoxications have been mainly associated with the consumption of oysters, clams and other bivalve molluscs in North America. Symptoms occur within 3 hours of consumption and include gastrointestinal disturbances, numbness of the mouth, muscular aches and dizziness. The dinoflagellate responsible for NSP, Ptychodiscus brevis, is notorious for the massive fish kills that occur every 3–4 years off the west coast of Florida. The lack of availability of analytical standards has hampered the development of suitable chemical methods for the determination of these toxins. Most monitoring programmes rely on the use of a mouse bioassay to detect levels of toxicants. Restriction of harvesting of shellfish at those times of the year when algal blooms occur is currently the safest method of prevention. Ciguatera toxins are the largest global public health non-microbial problem associated with seafood. Most incidents occur in the USA, danger areas being the Pacific, Caribbean and Indian Oceans. To date, three incidents have been recorded in the UK (Scoging, 1991). Finfish that harbour the toxin include the barracuda, red snapper, grouper, amberjack, surgeon fish and sea bass. Ciguatoxin is a neuromuscular toxin that affects the membrane potential of neural cells. Symptoms vary widely with the dose ingested but include vomiting, abdominal plain, dizziness, blurred vision, and reversal of the sensations of hot and cold. Onset is usually within a few hours of consumption and the effects can persist for several months. The toxins are heatstable and unaffected by processing methods. The appearance of the fish gives no indication of the toxin. Development of a dipstick immunoassay to detect ciguatoxins has facilitated sampling in the field (Hokama et al., 1989).

6.8.3 Scombroid fish poisoning Scombroid fish poisoning occurs throughout the world, though most incidents are recorded in the USA, Japan and the UK. In the USA, scombrotoxicosis was the cause of 29% of food-poisoning incidents caused by chemical agents between 1973 and 1987 (Hughes and Potter, 1991), and in the UK, 348 suspected incidents were reported between 1976 and 1986 (Bartholomew et al., 1987) with Scombridae and Scomberesocidae families (tuna, mackerel, saury, bonito © 2008, Woodhead Publishing Limited

130

Chilled foods

and seerfish), but incidents have also been associated with non-scombroid fish (sardines, herring, pilchards, anchovies and marlin) (Bartholomew et al., 1987, Morrow et al., 1991). Scombrotoxicosis is characterized by the rapid onset (within a few minutes to 2–3 hours of eating the fish) of symptoms which can include flushing, headache, cardiac palpitations, dizziness, itching, burning of the mouth and throat, rapid and weak pulses, rashes on the face and neck, swelling of the face and tongue, abdominal cramp, nausea, vomiting and diarrhoea. The similarity of these symptoms with those related to food allergy has often resulted in misdiagnosis. Histamine has been considered to be the cause of scombrotoxic poisoning for a number of reasons. Analysis of the fish remaining ‘on the plate’ usually reveals it contains high levels of histamine; metabolites of histamine have been detected in the urine of victims; symptoms resemble those of known histamine responses; and administration of antihistamine drugs reduces the severity of symptoms. Scoging (1998) of the Food Hygiene Laboratory (Public Health Laboratory Service) proposed guidelines with respect to histamine levels and the potential for illness. Histamine is a spoilage product resulting from decarboxylation of the amino acid L-histidine which is abundant in scombroid fish flesh. Formation of histamine requires the enzyme histidine decarboxylase, which is produced by the normal bacterial microflora of fish skin, gut and gills. If fish is stored above 4 °C, these organisms proliferate and levels of histamine in the flesh increase. Prevention of scombroid fish poisoning would therefore appear to be highly dependent on good handling practices – rapid chilling of the catch, and adequate chilling of the fish prior to preparation for eating. However, in medically supervised feeding studies, deliberately spoiled mackerel and mackerel with added histamine, fed to volunteers, failed to reproduce scombrotoxic symptoms (Clifford et al., 1989). These workers suggested that histamine alone is unlikely to be the causative agent. Other amines, such as cadaverine, have been suggested as potentiators or as synergists to histamine (Bjeldanes et al., 1978). Further feeding studies, using mackerel implicated in a scombroid-fish poisoning outbreak which reproduced symptoms in volunteers, showed the potency of the mackerel was not related to the histamine dose (Ijomah et al., 1991), or the content of other amines (cadaverine, putrescine, spermidine, spermine, tyramine), or any relationship between the levels of these amines (Clifford et al., 1991). Vomiting and diarrhoea were abolished by administration of antihistamine drugs. It has been suggested that histamine, released by the human body as a part of the natural defence mechanism, is responsible for the observed symptoms, that dietary histamine has a minor role in scombroid-fish poisoning, and that, as yet, the agent in fish which is responsible for triggering the release of histamine by the body is unidentified.

6.8.4 Allergens Food allergy, as opposed to food intolerance, is an immunological reaction to some component of the food. This component or antigen can stimulate the body to © 2008, Woodhead Publishing Limited

Non-microbiological factors affecting quality and safety

131

release specific immunoglobulin E (IgE) antibodies that give rise to anaphylaxis. Such a reaction can range from a trivial event such as a sneeze to a life-threatening incident. Many food types have been associated with allergic reactions. European labelling legislation currently identifies 14 allergens, including milk, soya, egg, crustacea, molluscs, lupin, cereals containing gluten, sesame, fish, celery, mustard, sulphites, nuts and peanuts. These issues are not specific to chilled foods but with the increase in formulated chilled food products and the severity of the potential hazards associated with the use of peanuts, mention is appropriate here. Studies have indicated that peanut allergy has been reported by 0.5% of the adult population in the UK (Emmett et al., 1999) and that those with sensitivity to peanuts commonly show reaction to other nut types, i.e. hazel nut and Brazil nut (Pumphrey et al., 1999). The allergenicity of peanut residues is heat stable and Ara h 1, a major peanut allergen, has been shown to retain its IgE binding characteristics despite significant structural denaturation (Koppelman et al., 1999). The possible carry over of allergenic material from product to product or production line to production line, therefore, necessitates stringent hygiene practices with associated quality assurance measures. Wherever possible, products containing peanut residues should be prepared and processed in separate areas, away from products that consumers do not expect to contain peanut residues. Risk assessments should be used to identify all potential sources of cross contamination. Where there is potential for cross contamination, product scheduling and appropriate cleaning regimes are essential.

6.8.5 Products of lipid oxidation Lipid oxidation products are of great significance to the sensory properties of food, but, in addition, attention has been given to the health risks that they may pose, and to their role in reduction of nutrient availability via free radical production and destruction of fat-soluble vitamins A and E. Lipid hydroperoxides and their decomposition products may bind and polymerize proteins, and cause damage to membranes and biological components, thus affecting vital cell functions (Halliwell and Gutteridge, 1986, Frankel, 1984). Lipid peroxides and oxidized cholesterol may be involved in tumour promotion and in atherosclerosis. Malonaldehyde, a secondary product of lipid oxidation, has been implicated as a catalyst in the formation of N-nitrosamines and as a mutagen (Pearson et al., 1983, Jurdi-Haldernan et al., 1987, Sanders, 1987). The significance to human health of eating foods which contain high levels of lipid hydroperoxides and their decomposition products is still to be established, particularly as the rate of formation of lipid peroxides in vivo is much greater than that arising from dietary intake. Nevertheless, whilst possible health risks associated with lipid oxidation products remain controversial, high levels of lipid peroxides are undesirable in the diet. Pre-cooked meats have been identified, amongst other products, as an area which requires further research to improve methods for retarding the development of rancidity (Addis and Warner, 1991). © 2008, Woodhead Publishing Limited

132

6.9

Chilled foods

Conclusion

The objective of this chapter has been to illustrate, by example, the way in which many non-microbiological factors interact to influence the quality and safety of chilled foods. The contribution that an understanding of food chemistry can make towards optimization or prevention of these interactions is evident. Further understanding is required for expansion and continued success in the production of safe, high quality chilled foods which achieve the desired shelf-life.

6.10 References ADDIS P B AND WARNER G J (1991) The potential health aspects of lipid oxidation products

in food. In: Aruoma, O I. and Halliwell, B (eds), Free Radicals and Food Additives, Taylor and Francis, London, pp. 77–119. ALLEN J C (1989) Rancidity in dairy products. In: Allen, J C. and Hamilton, R J. (eds) Rancidity in foods 2nd edn, Elsevier Applied Science, London, pp. 199–210. ANDERSON R E (1982) Long-term storage of peaches and nectarines intermittently warmed during controlled atmosphere storage, J. Am. Soc. Hort. Sci., 107(2) 214–16. ANG C Y W AND LYON B G (1990) Evaluations of warmed-over flavor during chill storage of cooked broiler breast, thigh and skin by chemical, instrumental and sensory methods, J. Food Sci., 55(3) 644–8, 673. ASGHAR A A, GRAY J I, BUCKLEY D J, PEARSON A M AND BOOREN A M (1988) Perspectives on warmed-over flavor, Food Technol., 42(6) 102–8. AVITAL Y, MARMHEIM C H AND MILTZ J (1990) Effect of carbon dioxide atmosphere on staling and water relations in bread, J. Food Sci., 55(2) 413–16, 461. BARTHOLOMEW B, BERRY P, RODHOUSE J, GILBERT R AND MURRAY C K (1987) Scombrotoxic fish poisoning in Britain; features of over 250 suspected incidents from 1976 to 1986, Epidemiol Infect., 99 775–82. BENDER A E AND REAIDI G B (1982) Toxicity of kidney beans (Phaseolus vulgaris) with particular reference to lectins, J. Plant Foods, 4(1) 15–22. BJELDANES L F, SCHUTZ D E AND MORRIS M M (1978) On the aetiology of scombroid poisoning: cadaverine potentiation of histamine toxicity in the guinea pig, Food and Cosmetics Toxicology, 16 157–9. BOLIN H R AND HUXSOLL C C (1989) Storage stability of minimally processed fruit, J. Food Proc. Pres, 13(4) 281–92. BRAMLAGE W J (1982) Chilling injury of crops of temperate origin,. HortSci., 17(2) 165–8. BROTSKY E (1976) Automatic injection of chicken parts with polyphosphate, Poultry Sci., 55(2) 653–60. BROWN H M, OSBORN H AND LEDWARD D (1998) Undesirable pink colouration in cooked meat products. RSS No. 80, CCFRA Research Summary Sheets, Campden and Chorleywood Food Research Association, Chipping Campden. Gloucestershire GL55 6LD. BUSHWAY R J AND PONNAMPALAM R (1981) α-Chaconine and α-solanine content of potato products and their stability during several modes of cooking, J. Agric. Food Chem., 29(4) 814–17. BYRNE D V, BAK L S, BREDIE W L P, BERTELSEN G AND MARTENS M (1999a) Development of a sensory vocabulary for warmed-over flavour. I. In porcine meat, J. Sens. Std., 14(1) 47– 65. BYRNE D V, BAK L S, BREDIE W L P, BERTELSEN G AND MARTENS M (1999b) Development of a sensory vocabulary for warmed-over flavour. II. In chicken meat, J. Sens. Std., 14(1) 67– 78. CASTELL C H (1971) Metal-catalysed lipid oxidation and changes of proteins in fish, J. Am. Chem. Soc, 48(11) 645–9. © 2008, Woodhead Publishing Limited

Non-microbiological factors affecting quality and safety

133

CHO S H AND RHEE K S (1997) Lipid oxidation in mutton: species related and warmed-over

flavours, Journal of Food Lipids, 4(4) 283–93. CHURCHILL H M, GRIFFITHS N G AND WILLIAMS B M

(1988) Warmed-over flavour in meat, Campden Food and Drink Research Association Technical Memorandum No. 489, Campden and Chorleywood Food Research Association, Chipping Campden. CHURCHILL H M, GRIFFITHS N G AND WILLIAMS B M (1990) The effect of reheating on warmed-over flavours in chicken and baked potatoes, Campden Food and Drink Research Association Technical Memorandum No. 591, Campden and Chorleywood Food Research Association, Chipping Campden. CLIFFORD M N, WALKER R AND WRIGHT J (1989) Studies with volunteers on the role of histamine in suspected scombrotoxicosis, J. Sci. Food Agric., 47 365–75. CLIFFORD M N, WALKER R, WRIGHT L, MURRAY C K AND HARDY R (1991) Is there a role for amines other than histamines in the aetiology of scombrotoxicosis? Food Additives and Contaminants, 8(5) 641–52. COGAN T M (1977) A review of heat resistant lipases and proteinases and the quality of dairy products, Ir. J. Food Sci. Technol., 1 95–105. COUEY H M (1982) Chilling injury of crops of tropical and subtropical origin, HortSci., 17(2) 162–5. DALHOFF E AND JUL M (1965) In: Penzer, W T. (ed.) Progress in Refrigeration Science and Technology, Vol. 1. AVI Publishing Co., Westport, CT, p. 57. DAVEY C L AND GILBERT K V (1976) The temperature coefficient of beef ageing, J. Sci. Food. Agric., 27(3) 244–50. DAY B P F (1998) Novel MAP – a brand new approach, Food Manufacture, 73(11), 24–6. EMMETT S E, ANGUS F J, FRY J S AND LEE P N (1999) Perceived prevalence of peanut allergy in Great Britain and its association with other atopic conditions and with peanut allergy in other household members, Allergy, 54(4) 380–5. ETHERINGTON D J, TAYLOR M A J AND DRANSFIELD E (1987) Conditioning of meat from different species: relationship between tenderizing and the levels of cathepsin B, cathepsin L, calpain I, calpain II and beta-glucuronidase, Meat Sci., 20(1) 1–18. FRANKEL E N (1984) Recent advances in the chemistry of the rancidity of fats. In: Bailey, A J. (ed.) Recent Advances in the Chemistry of Meat, The Royal Society of Chemistry, Special Publication, No. 47, pp. 87–18. GERMAN J B AND KINSELLA J E (1985) Lipid oxidation in fish tissues. Enzymatic initiation via lipoxygenase, J. Agric. Food Chem., 33(4) 680–3. GIBBS P A AND WILLIAMS A P (1990) Using mathematical models for shelf life prediction. In: Turner, A. (ed.), Food Technology International Europe, Stirling Publications International Ltd. GOLL D E, KLEESE W C AND SZPACENKO A (1989) Skeletal muscle proteases and protein turnover. In: Campion, D R, Hausman, G J and Martin, R J (eds), Animal Growth and Regulation, Plenum, New York, pp. 141–82. GRAY J L AND PEARSON A M (1987) Rancidity and warmed-over flavor. In: Pearson A M and Dutson, T R (eds), Advances in Meat Research. Vol. 3. Restructured Meat and Poultry Products, Van Nostrand Reinhold, New York, pp. 221–69. GREENE B E (1969) Lipid oxidation and pigment changes in raw beef, J. Food Sci., 34(2) 110– 13. HALLIWELL B AND GUTTERIDGE J M C (1986) Free Radicals in Biology and Medicine, Clarendon Press, Oxford. HERMANN K, SCHUTTE M, MULLER H AND BISMER R (1981) Ueber die antioxidative Wirkung von Gerwürzen, Deutsche Lebensmittel-Rundschau, 77 134. HOKAMA Y, HONDA S A A, KOBAYASHI M N, NAKAGAWA L K, ASAHINA A Y AND MIYAHARA J T (1989) Monoclonal antibody (Mab) in detection of ciguatoxin (CTX) and related polyethers by the stick-enzyme immunoassay (S-EIA) in fish tissues associated with ciguatera poisoning. In: Natori, S., Hashimoto, K. and Ueno, Y. (eds), Mycotoxins and Phycotoxins ’88. Elsevier Science Publishers, Amsterdam, pp. 303–10. HUELIN F E AND COGGIOLA I M (1970) Superficial scald, a functional disorder of stored © 2008, Woodhead Publishing Limited

134

Chilled foods

apples. V. Oxidation of alpha-farnesene and its inhibition by diphenylamine, J. Sci. Food Agric., 21(1) 44–8. HUGHES M J AND POTTER M E (1991) Scombroid-fish poisoning: from pathogenesis to prevention, New England J. Med., 324 766–8. IGENE J O AND PEARSON A M (1979) Role of phospholipids and triglycerides in warmed-over flavour development in meat model systems, J. Food Sci., 44(5) 1285–90. IGENE J O, KING J A, PEARSON A M AND GRAY J I (1979) Influence of heme pigments, nitrite, and non-heme iron on development of warmed-over flavor (WOF) in cooked meat, J. Agric. Food Chem., 27(4) 838–42. IGENE J O, PEARSON A M AND GRAY J I (1981) Effects of length of frozen storage, cooking and holding temperatures upon component phospholipids and the fatty acid composition of meat triglycerides and phospholipids, Food Chem., 7(4) 289–303. IGENE J O, YAMAUCHI K, PEARSON A M, GRAY J I AND AUST S D (1985) Evaluation of 2thiobarbituric acid reactive substances (TBRS) in relation to warmed-over flavor (WOF) development in cooked chicken, J. Agric. Food Chem., 33(3) 364–7. IJOMAH P, CLIFFORD M N, WALKER R, WRIGHT L, HARDY R AND MURRAY C K (1991) The importance of endogenous histamine relative to dietary histamine in the aetiology of scombrotoxicosis, Food Additives and Contaminants, 8(4) 531–2. ILKER Y AND MORRIS L L (1975) Alleviation of chilling injury of okra. HortSci., 10 324. ILKER Y, WARING A J, LYONS J M AND BREIDENBACH R W (1976) The cytological responses of tomato-seedling cotyledons to chilling and the influence of membrane modifications upon these responses, Protoplasma, 90 229–52. JACKMAN R L, YADA R Y, MARANGONI A, PARKIN K L AND STANLEY D W (1988) Chilling injury. A review of quality aspects, J. Food Qual., 11(4) 253–78. JAMES S J (1985) Display conditions from the product’s point of view. Institute of Food Research Bristol Laboratory, Teach-in on retail display cabinets. JAYATHILAKAN K, VASUNDHARA T S AND KUMUDAVALLY K V (1997) Effects of spices and Maillard reaction products on rancidity development in precooked refrigerated meat, J. Food Sci. Technol., India, 34(2) 128–31. JUNCHER D, HANSEN T B, ERIKSEN H, SKOVGAARD I M, KNOCHEL S AND BERTELSEN G (1998) Oxidative and sensory changes during bulk and retail storage of hot-filled turkey casserole, Zeit. Lebensm. Untersuch. Forsch. A, 206(6) 378–81. JURDI-HALDERNAN D, MCNEIL J H AND YARED D M (1987) Antioxidant activity of onion and garlic juices in stored cooked ground lamb, J. Food Protect., 50(5) 411–13. KERRY J P, BUCKLEY D J, MORRISSEY P A, O’SULLIVAN K AND LYNCH P B (1999) Endogenous and exogenous alpha-tocopherol supplementation: effects on lipid stability (TBARS) and warmed-over flavour (WOF) in porcine M. longissimus dorsi roasts held in aerobic and vacuum packs, Food Res. Int., 31(3) 211–16. KOPPELMAN S J, BRUIJNZEEL-KOOMEN C A F M, HESSING M AND JONGH, H H J DE (1999) Heatinduced conformational changes in Ara h 1, a major peanut allergen, do not affect its allergenic properties, J. Biol. Chem., 274(8) 4770–4. KORCZACK L, FLACZYK E AND PAZOLA Z (1988) Effects of spices on stability of meat products kept in cold storage, Fleischwirtschaft, 68 64–6. KROGH P (1987) Scientific Report on paralytic shellfish poisons in Europe. Document VI/ 3964/87-EN rev. 1, Commission of the European Communities, Directorate General for Agriculture, VI/B/II.2, Brussels. LABUZA T P (1985) An integrated approach to food chemistry: illustrative cases. In: Fennema, O. R. (ed.) Food Chemistry. Marcel Dekker, New York, pp. 913–38. LAW B A AND WIGMORE A S (1985) Effect of commercial lipolytic enzymes on flavour development in Cheddar cheese, J. Soc. Dairy Technol., 38(3) 86–8. LIENER I E (1974) Phytohaemagglutinins: their nutritional significance, J. Agric. Food Chem., 22(1) 17–22. LOPEZ-BOTE C, REY A, RUIZ J, ISABEL B AND SANZ-ARIAS R (1997) Effects of feeding diets high in monounsaturated fatty acids and alpha tocopherol acetate to rabbits on resulting © 2008, Woodhead Publishing Limited

Non-microbiological factors affecting quality and safety

135

carcass fatty acid profile and lipid oxidation, Animal Sci., 64(1) 177–86. LOVE J (1988) Sensory analysis of warmed-over flavour in meat, Food Technol., 42(6) 140–

3. (1987) Development of chicken flavour descriptive attribute terms aided by multivariate statistics, J. Sensory Studies, 2(1) 55–67. MCCARTHY M J AND KAUTEN R J (1990) Magnetic resonance imaging applications in food research. Trends in Food Science & Technology December, 134–9. MCLACHLAN A AND STARK R (1985) Modified atmosphere packaging of selected prepared vegetables, Campden Food and Drink Research Association Technical Memorandum No. 412, Campden and Chorleywood Food Research Association, Chipping Campden. MCWEENY D J (1968) Reactions in food systems: negative temperature coefficients and other abnormal temperature effects, J. Food Technol., 3(1) 15–30. MAGA J A (1975) Bread staling, CRC Crit. Rev. Food Technol., 5(4) 443–86. MAGA J A (1994) Pink discolouration in cooked white meat. Food Reviews Int., 10(3) 273– 86. MANN T F, REAGAN J O, LILLARD D A, CAMPION D R, LYON C E AND MILLER M F (1989) Effects of phosphate in combination with nitrite or Maillard reaction products upon warmed-over flavour in precooked, restructured beef chuck roasts, J. Food Sci., 54(6) 1431–3, 1437. MARSH B B, RINGKOB T R, RUSSELL R L, SWARTZ D R AND PAGEL L A (1987) Effects of earlypost mortem glycolytic rate on beef tenderness, Meat Sci., 21(4) 241–8. MEISNER J A (1953) Importance of temperature and humidity in the transportation and storage of bread, Baker’s Dig., 27 109. MIKAMI M, WHITING A H, TAYLOR M A L, MACIEWICZ R A AND ETHERINGTON D J (1987) Degradation of myofibrils from rabbit, chicken and beef by cathepsin L and lysosomal lystates, Meat Sci. 21(2) 81–97. MINOTTI G AND AUST S D (1987) The requirement for iron (III) in the initiation of lipid peroxidation by iron (II) and hydrogen peroxide, J. Biol. Chem., 262 1098. MORROW J D, MARGOLIES G R, ROWLAND J AND ROBERTS L J (1991) Evidence that histamine is the causative toxin of scombroid-fish poisoning, New Eng. J. Med., 324 716–20. MOSS M O AND FRANK M (1987) Prevention: effects of biocides and other agents on mycotoxin production. In: Watson, D H. (ed.) Natural Toxicants in Food Progress and Prospects, Ellis Horwood, Chichester, pp. 231–52. MURPHY A, KERRY J P, BUCKLEY J AND GRAY I (1998) The antioxidative properties of rosemary oleoresin and inhibition of off-flavours in precooked roast beef slices, J. Sci. Food. Agric., 77(2) 235–43. NOLAN N L, BOWERS J A AND KROPF D H (1989) Lipid oxidation and sensory analysis of cooked pork and turkey stored under modified atmospheres, J. Food Sci., 54(4) 846–9. O’BEIRNE D (1988) Modified atmosphere packaging of ready-to-use potato strips and apple slices. Abstracts 18th Annual Food Science and Technology Research Conference, Int. J. Food Sci. Technol, 12(1) 94–5. O’BEIRNE D AND BALLANTYNE A (1987) Some effects of modified atmosphere packaging and vacuum packaging in combination with antioxidants on quality and storage life of chilled potato strips, Int. J. Food Sci. Technol, 22(5) 515–23. O’NEILL L M, GALVIN K, MORRISSEY P A AND BUCKLEY D J (1998) Comparison of the effects of dietary olive oil, tallow and vitamin E on the quality of broiler meat and meat products, British Poultry Science, 39(3) 365–71. PEARSON A M AND GRAY J T (1983) Mechanisms responsible for warmed-over flavor in cooked meat. In: Waller, G R. and Feather, M S. (eds) The Maillard Reactions in Foods and Nutrition, American Chemical Society, Washington DC, p. 287. PEARSON A M, GRAY J T, WOLDZAK A M AND HORENSTEIN N A (1983) Safety implications of oxidized lipids in muscle foods, Food Technol., 37(7) 121–9. PEPPLER H J AND REED G (1987) Enzymes in food and feed processing. In: Rehm, H J. and Reed, G. (eds) Biotechnology, VCH, Weinheim. PIKUL J AND KUMMEROW F A (1991) Thiobarbituric acid reactive substance formation as LYON B G

© 2008, Woodhead Publishing Limited

136

Chilled foods

affected by distribution of polyenoic fatty acids in individual phospholipids, J. Agric. Food Chem., 39 451–7. PRATT D E AND WATTS B M (1964) Antioxidant activity of vegetable extracts. I. Flavone aglycones, J. Food Sci., 29 27–33. PRZYBYLA WILKES A (1991) Hooking onto seafood, Food Product Design, August, 51–5. PUMPHREY R S H, WILSON P B, FARAGHER E B AND EDWARDS S R (1999) Specific immunoglobulin E to peanut, Hazel nut and Brazil nut in 731 patients: similar patterns found at all ages, Clinical and Experimental Allergy, 29(9) 1256–9. QUALI A AND TALMANT A (1990) Calpains and calpastatin distribution in bovine, porcine and ovine skeletal muscles, Meat Sci., 28(4) 331–48. RODHOUSE J C, HAUGH C A, ROBERTS D AND GILBERT R J (1990) Red kidney bean poisoning in the UK: an analysis of 50 suspected incidents between 1976 and 1989, Epidemiol. Infect., 105 485–91. ROOZEN J P (1987) Effects of types I, II, and III antioxidants on phospholipid oxidation in a meat model for warmed over flavour, Food Chem., 24 167–85. ST ANGELO, A L, VERCELLOTI J R, LEGENDRE M G, VINNETT C H, KUAN J W, JAMES JR C AND DUPUY H P (1987) Chemical and instrumental analyses of warmed-over flavour in beef, J. Food Sci., 52(5) 1163–8. SANDERS T (1987) Toxicological considerations in oxidative rancidity of animal fats, Food Sci. Technol. Today, 1 162–4. SATO K AND HEGARTY G R (1971) Warmed-over flavor in cooked meats, J. Food Sci., 36(7) 1098–102. SCHIFFMANN-NADEL M, CHALUTZ E, WAKS J AND DAGAN M (1975) Reduction of chilling injury in grapefruit by thiabendazole and benomyl during long-term storage, J. Am. Soc. Hort. Sci., 100 270–2. SCHRICKER B R AND MILLER D D (1983) Effects of cooking and chemical treatment on heme and nonheme iron in meat, J. Food Sci., 48(4) 1340–3, 1349. SCOGING A C (1991) Illness associated with seafood, Communicable Disease Report, 1(11) R117–22. SCOGING A C (1998) Scombrotoxic (histamine) fish poisoning in the United Kingdom: 1987 to 1996, Communicable Disease and Public Health, 1(3) 204–5. SHAHIDI F, RUBIN L J, DIOSADY L L, KASSUM N, FONG J C AND WOOD D F (1986) Effect of sequestering agents on lipid oxidation in cooked meats, Food Chem., 21(2) 145–52. SHARP A K (1986) Humidity: measurement and control during the storage and transport of fruits and vegetables, CSIRO Food Res. Quart., 46(4) 79–85. SHAW R (1997) Cook–chill ready meals: opportunities for MAP, Food Review, 24(4) 23, 25, 27, 29, 31. SMITH D M, SALIH A M AND MORGAN R G (1987) Heat treatment effects on warmed-over flavor in chicken breast meat. J. Food Sci., 52(4) 842–5. SPALDING D H AND REEDER, W F (1975) Low-oxygen high carbon dioxide controlled atmosphere storage for control of anthracnose and chilling injury of avocados, Phytopathology, 65(4) 458–60. STOICK S M, GRAY J I, BOOREN A M AND BUCKLEY D J (1991) Oxidative stability of restructured beef steaks processed with oleoresin rosemary, tertiary butylhydroquinone, and sodium tripolyphosphate, J. Food Sci., 56(3) 596–600. SWAIN M V L AND JAMES S J (1986) Evaporative weight loss from unwrapped meat and food products in chilled display cabinet. In: Meat Chilling ’86. International Institute of Refrigeration Commission C2, Bristol. SWAIN M V L, GIGIEL A J AND JAMES S J (1986) Carbon dioxide chilling of hot boned meat. In: Meat Chilling ’86. International Institute of Refrigeration Commission C2, Bristol pp. 195–202. TAMIME A Y AND DEETH H C (1980) Yoghurt: technology and biochemistry, J. Food Protect., 43(12) 939–77. TAYLOR A A (1972) Influence of carcass chilling rate on drip in meat. In: Proceedings of the © 2008, Woodhead Publishing Limited

Non-microbiological factors affecting quality and safety

137

Meat Research Institute Symposium No. 2, Meat Chilling Why and How? Bristol, pp. 5.1– 5.8. TIMS M J AND WATTS B M (1958) Protection of cooked meats with phosphates, Food Technol., 12 240–3. TUNALEY A AND BROCKLEHURST T F (1982) A study on the shelf life of coleslaw, Chilled Foods, 1(6) 12–13. TUNALEY A, BROWNSEY G AND BROCKLEHURST T F (1985) Changes in mayonnaise-based salads during storage, Lebensm. Wiss. u. Technol., 18(4) 220–4. VÁMOS-VIGYÁZÓ L (1981) Polyphenol-oxidase and peroxidase in fruits and vegetables, CRC Crit. Rev. Food Sci. Nutri., 15(1) 49–127. WADE N L (1979) Physiology of cool-storage disorders of fruit and vegetables. In: Lyons, J. M., Graham, D. and Raison, J K. (eds) Low Temperature Stress in Crop Plants, Academic Press, New York, pp. 81–96. WANG C Y (1982) Physiological and biochemical responses of plants to chilling stress, HortSci., 17(2) 173–86. WANG C Y AND ADAMS D O (1980) Ethylene production by chilled cucumbers (Cucumis sativus L), Plant Physiol., 66(5) 841–3. WANG C Y AND BAKER J E (1979) Effects of two free radical scavengers and intermittent warming on chill injury and polar lipid composition of cucumber and sweet pepper fruits, Plant Cell Physiol., 20 243–51. WATTS B M (1961) The role of lipid oxidation in lean tissues in flavor deterioration of meat and fish. In: Proceedings Flavor Chemistry Symposium, Campbell Soup. Co., Camden, NJ p. 83. WEST P A, WOOD P C AND JACOB M (1985) Control of food poisoning risks associated with shellfish, Journal of the Royal Society of Health, 1 15–21. WHITAKER B D (1991) Changes in lipids of tomato fruit stored at chilling and non-chilling temperatures, Phytochemistry, 30(3) 757–61.

© 2008, Woodhead Publishing Limited

7 Chilled foods packaging: an introduction D. Dearden, Unilever, UK

7.1

Demands of chilled food packaging

Consumers associate chilled foods with greater convenience which offers both superior quality and vitality benefits. These perceived benefits of chilled foods continue to drive growth and product innovation across a wide variety of food categories ranging from very short shelf-life, raw, unprocessed foods (such as meat, fish and vegetables) through to fully prepared extended shelf-life foods, e.g. soups, ready meals, pizza and pasta sauces to name but a few. This broad and diverse range of food products provides significant technological challenges for both packaging material and packaging machinery suppliers. Suppliers of packaging/equipment and food manufacturers need to balance the conflicting needs of the delivery of the optimum packaging materials, pack and packing combinations that will protect, preserve and promote the product throughout the supply chain with material and production cost efficiency. The selected pack format must also ensure that the consumer does not consider the products to be over packaged or the packaging unnecessarily complex. The perceived environmental impact and sustainability of packaging materials and packs will be key purchasing determinants for food manufacturers and consumers, both now and increasingly in the future, particularly as the emphasis on environmental impact on the basis of carbon footprint increases. The days of complex multi component packs will soon be a thing of the past. The need to provide packaging materials, pack sizes and pack formats that are compatible with different markets or retailers and the additional needs of an aging population must not be lost by the technologists, designers and food manu© 2008, Woodhead Publishing Limited

Chilled foods packaging: an introduction

139

facturers. Ease of opening, pack size and the legibility of graphical design are key considerations. This chapter reviews the mainstream packaging materials, pack formats and packing techniques used by the industry and outlines some of the developments that are in process of being, or will be, deployed in the near future.

7.2

Packaging material selection

Chilled food packaging materials must to be able to protect, preserve and promote an extensive range of foods ranging from fresh produce through to prepared meals, some of which will be reheated in-pack using a microwave or conventional oven. In order to adequately define and specify the finished pack – that is the primary, secondary and outer packaging materials and format – a number of characteristics of the product and the onward supply chain and consumer use need to be adequately clarified:

• The dimensions of the food components forming the product, including the tolerance on each.

• Shape of the product – regular or irregular. • Single or multiple compartments. • The bulk density of the product; if the contents are granular in nature – compacted and un-compacted.

• How the product will be preserved in order to maintain optimum appearance and quality throughout its shelf-life, both in store and in the hands of the consumer.

• Does the food need to respire, such as some fruits and vegetables? • Is a protective atmosphere needed in order to maintain food colour to prevent browning or flavour changes such as oxidative rancidity?

• Will vacuum be required in order to preserve the contents of the pack? • Are the contents of the pack light-sensitive and what is likely to be the light intensity exposure it will experience?

• What temperature range will the pack experience during manufacture, distribution, in store and during use by the consumer? This will not only affect substrate choice (e.g. chilled to oven heating temperature range), but also the basis for packaging material toxicological clearance under food contact legislation (Framework Regulation (EC) No 1935/2004 and see http://ec.europa.eu/food/ food/chemicalsafety/foodcontact/spec_dirs_en.htm for specific information).

• Storage temperature chilled or super chilled? • Will the product be cold or hot filled as part of processing and within what temperature range (e.g. 65 to >100 °C)?

• Will the product be in-pack pasteurised and at what temperature and for how long? © 2008, Woodhead Publishing Limited

140

Chilled foods

• Will the pack be required to withstand, and remain rigid during, microwave, fan oven or conventional oven reheating or cooking?

• Will moisture/drip absorbers be required within the pack (e.g. for fresh meat and fish)?

• Are there free oils and/or a high fat content (>20%) of the product – levels can affect pack integrity and especially gas permeability.

• Salt concentration of the product – high concentrations are known to cause delamination within some multilayer structures.

• Certain spices – these have also been known to cause delamination in multilayer •

•

structures and the use of oleo-resins in the product formulation may be a suitable replacement. The need for a high hygiene environment in order to minimise microbiological loading. This will impact upon the organisation of the filling hall, type of filling/ sealing machine and how materials are prepared before moving into the hygienic zone, e.g. delivery format and transit packaging removal. Method of forming, filling and closing – will impact upon the tensile properties of the materials:

• • • • • •

vertical form fill and seal horizontal form fill and seal hand filling and sealing hot or cold filled modified atmosphere vacuum sealing

• Filling, packing and sealing line speeds. • Method of pack closure, e.g. crimped seal, heat or ultrasonic seal, screw top. • Method of consumer opening, e.g. peelability pull tab, clip lid and any need for re-closing.

• Packing machine mechanisms – does the method of packing material handling

• • • • •

or forming require specific material properties – slip level (coefficient of friction), tensile strength (elongation machine or cross direction), stiffness, dead fold, etc. Method of primary pack collation and assembly of formed packs into outer packaging (e.g. for trays or cartons). Modularity of outer packaging (the need to have outer pack footprint matched to pallet dimensions, bulk transit tote and the shelf in the retailer). Reusability of the packaging – home use by the consumer. Environmental impact – carbon footprint, life-cycle analysis and compatibility with recycle schemes (including incineration with energy recovery). Supply chain considerations – stacking height, traceability and warehousing.

7.3

Packaging material substrates

There are a wide number of primary (e.g. in contact with the food) and secondary © 2008, Woodhead Publishing Limited

Chilled foods packaging: an introduction

141

(e.g. over wrapping or covering the primary packaging and presenting the product) packaging materials that can be potentially selected for use in chilled food packs. The choice will depend upon the requirements of the food, packing line, the onward supply chain and consumer end use. The trade off in this equation is pack functionality vs. cost. As a rule of thumb, the heavier the pack (e.g. material type, such as glass and thickness of material) the more costly it will be in terms of both material cost and distribution or transport costs. Clearly, whilst light weighting does address the direct cost issues, it will impact line efficiency, line wastage and wastage in the onward distribution or supply chain as a result of pack damage. These factors are much more difficult to accurately cost and hence it is much easier to drive cost saving or value improvement projects on the basis of material savings rather than on balanced or total business savings. Therefore it is important to get as complete a picture as possible in order to ensure that any economies arising are genuine economies in a total business context. Briefly outlined below are the comparative properties of the major material types in terms of the advantages and disadvantages of each. The information is for illustration only and is not intended as a comprehensive or definitive set of guidelines.

7.3.1 Aluminium Aluminium is a durable, ductile and strong material; it offers a good surface for decoration and it has excellent light, moisture and gas barrier properties. Grade selection in terms of thickness and temper (e.g. hardness and strength) is important for the conversion process used by the packaging material supplier.

• Aluminium foil can be used as a barrier layer (gas or moisture or light) in

• • •

•

combination with paper, board or plastic films (e.g. bricks), although the inclusion of foil as a discrete layer within a laminated structure will also enhance formability (dead fold or cold forming), stiffness and tensile strength. As a thin layer deposited on the surface of films, paper or board by the vacuum metallisation process (e.g. pouches). This can enhance barrier properties and pack appearance. As a standalone or unsupported wrapping or lidding material (e.g. wrapped product or lidded trays). This will require the application of a heat or cold seal coating of some sort. As semi-rigid trays, bowls or pots (with or without heat seal lacquer) (e.g. multicompartment trays). Semi-rigid or formed containers, whilst offering excellent reheat potential in conventional and microwave ovens and bain maries, are susceptible to crush or impact damage unless protected by secondary packaging during distribution. Formed and drawn into two-piece cans or bottles.

Aluminium also offers recyclability, although this will be pack format dependant, foils in laminates being particularly difficult to recover and reuse. In terms of overall life-cycle analysis, the impact caused by mining and smelting of bauxite © 2008, Woodhead Publishing Limited

142

Chilled foods

can tend to colour the picture of aluminium’s environmental credentials if only virgin aluminium is used in packaging materials.

7.3.2 Steel (tinplate or tin-free steel) Steel packaging offers a very similar range of properties compared with those of aluminium in terms of its physical and mechanical properties. Calliper for calliper (e.g. gauge or thickness), steel is superior to aluminium in terms of its strength. It too can easily be formed into shaped or semi-rigid containers. It is perhaps more associated with cans or general line packaging; however, significant strides have been made in recent years by manufacturers to diversify into those markets for semi-rigid packs normally dominated by aluminium. This has been achieved through the development of steel/polymer extrusion lamination materials. Steel cannot be cost-effectively rolled into thin gauge ‘foil’, certainly not to compete with aluminium, and hence its use is limited to heavier weight structures, e.g. semi-rigid metal/polymer trays. Semi-rigid or formed containers offer excellent reheat potential in conventional and microwave ovens and bain maries (although metals have to be avoided in some microwave ovens). Environmentally, steel is recyclable and the impact of iron ore extraction and smelting tends to be much lower than that of aluminium.

7.3.3 Glass Glass offers excellent moisture and gas barrier properties. It can be formed to match a range of closures such as caps, crowns, snap on or sealed film lids. Light protection can be provided by use of chemical additives which colour the glass (opalescent, amber and green), or polymer coating technology can be used, but this is expensive. As an alternative, PET or PVC shrink sleeves can be used to enhance light barrier properties and enhance pack decoration. Glass containers are heavier than alternatives such as steel, aluminium, board and plastics, and hence have an impact in terms of the distribution costs and the environment. Whilst glass is strong in terms of its axial load resistance, it is extremely vulnerable to impact or mechanical damage and hence must be protected by robust transit packaging. Sleeves can be used to increase the strength of glass jars. However, one risk of this solution is the potential for the sleeve to mask transit damage from consumers, i.e. the container may be broken within the supply chain and retail distribution chain and still be held together by the sleeve. The consumer can be put at risk if such an occurrence arises. Glass can be easily recycled, although the certain colours (white or flint) are generally in more demand than coloured glass by the glass packaging manufacturers.

7.3.4 Paper and board Paper and board are very versatile and can be used as primary packaging materials ranging from formed pulp trays (e.g. used for distribution of vegetables) to folding © 2008, Woodhead Publishing Limited

Chilled foods packaging: an introduction

143

cartons (e.g. bricks or bakery trays) to formed ovenable board containers capable of being reheated in both conventional and microwave ovens (e.g. pizza trays), through to paper labels or moisture absorbers or paper. Board can also be used as secondary transit packaging. It can also be used as fitments within packs, e.g. dividers. Paper is predominantly used for labelling and as a printing surface in multilayer laminates. Where paper or board are in direct contact with food, care must be taken to ensure that only ‘virgin’ fibre components are used, to minimise the chances of accidental contamination from recycled board. Recycled board materials have the potential to be contaminated with unsuitable materials such as residues from printing inks. If recycled content is to be used, then its purity of source must be guaranteed, e.g. materials from the board mill itself. There are also a number of coatings and surface treatments available which can further enhance paper’s physical properties and thereby its ubiquity. These include heat seal lacquers, adhesive coatings or coatings to prevent moisture, grease and oil penetration. Paper and board can also be laminated to other structures, such as plastics, to retain an excellent printing surface and enhance its mechanical properties. Paper and board as mono-materials or unlaminated or combined materials offer excellent environmental sustainability. Laminated board structures, such as those used for liquid packaging, offer much more of a challenge to the recycling industry because of the difficulty of separating the components of a laminate.

7.3.5 Plastics Plastics are the most versatile of all packaging materials. They have wide ranging physical properties (such as chemical or physical resistance and stability), many using coextrusions, extrusion lamination or adhesive lamination technology. This capability and flexibility allows the pack structure to be highly designed so that, theoretically, the packaging developer can both define and get the exact combination of material properties and container format required to do the job. However, the packaging industry is sufficiently well developed not to require this degree of creativity for each development and readily offers material combinations that span the range of product needs. In addition to this ‘mix and match’ capability offered by plastics, they can also be deployed in a range of different formats – monolayer films, multilayer films (adhesively laminated, extrusion laminated or coextruded), injection moulded, thermoformed, injection blow moulded, etc., depending on the product requirements, shape, size and consumer end use needed. Table 7.1 is intended to provide a summary of the some of the more commonly used materials and their properties. Properties of these materials can be modified by combining them and/or changing the manufacturing technology used, e.g. orientation. Some of the more basic properties that can be attained by the lamination or coextrusion of combinations of different polymers to produce multilayer plastics are outlined in Table 7.2. The information shown in this table represents some of © 2008, Woodhead Publishing Limited

144

Chilled foods

Table 7.1

Some comparative properties of plastics

Material

Transparency

Polyester (PET) High density polyethylene (HDPE) Polyvinyl chloride (PVC) Polypropylene (PP) Polystyrene PS

Excellent

2

Poor

0.5

Good

3

Poor

0.5

Table 7.2

Moisture and gas barrier properties Moisture Oxygen Carbon (cm3/m2 dioxide (g/50mm2 /24hrs) /24hrs) (cm3/m2 /24hrs)

Excellent

10

Impact strength

75

540

Good

4000

18000

Good

150

380

Fair

3500

7000

Fair

6000

18700

Poor

Differentiation of various plastics structures

© 2008, Woodhead Publishing Limited

•

•

• • • • • •

• • • • • •

• • •

• •

• •

•

Microwaveable

Dual ovenable

Freezing

• • • •

Sterilised

•

• • • • • • • • • • • • • •

Pasteurised

• •

Chilled

Hot fill

• • • • • • • • • • • • • • • • • •

atmosphere

Thermoforming

APET APET/PE APET/EVOH/PE CPET PS/PE PS/EVOH/PE PS/PETG PP/PE PP/EVOH/PE PP/EVOH/PP EPP/EVOH/PE PP/PA/PE PP/PA/PP PET Blend PET Blend/PE PVC PVC/PE PVC/EVOH/PE

Modified

Structures

Applications

•

•

•

•

• • • • • • •

• •

• • • •

• • •

• • •

• • • • • •

Chilled foods packaging: an introduction

145

the more commonly available plastics structures or combinations available and gives typical applications for which they may be used. This list is not intended to be definitive and many other structures are available. Selection will clearly depend upon product preservation needs, manufacturing process requirements, barrier properties, strength, intended consumer usage and cost constraints. All these structures are compatible with a range of packing techniques which will be outlined in greater detail later. However, there are other considerations in the design of chilled packaging.

7.4

Moisture condensation or fogging

Clearly, enclosing high-moisture foods and those that will be respiring (e.g. fresh vegetables) within a sealed pack under chilled conditions, or indeed placing a sealed pack at chilled temperatures, will lead to the phenomenon of fogging. Pack fogging is produced by the condensation of air moisture on cold plastic, causing the formation of tiny droplets on the surface which scatter light and obscure the contents of the pack. Glass materials can also be subject to the same problem. This is because temperatures below the dew point cause the water vapour in the air to condense on the pack surfaces (internally and externally). The low surface tension of polymers (hydrophobic) leads to the formation of droplets on surfaces. The duration of this effect is directly proportional to the thermal conductivity of the polymer or pack. However, plastics that contain an antifog surfactant (hydrophilic) cause droplets to spread evenly and thinly across the pack surface; the resultant thin film of water on the pack surface does not impact upon its transparency and also prevents moisture dripping onto product (which can lead to unwanted quality issues for the pack contents). 7.4.1 Antifogging strategies There is no single universal solution for combating fogging. Potential solutions used to minimise fogging include:

• other polymeric ingredients that can be blended with the plastics and form part

•

of their structure to: • improve the wetting performances of the polymer or • improve the water repellence (this technique is the least used). use of antifogging surface coatings.

The surface tension of water can be halved by adding small levels of conventional surfactants, for example sodium laurylsulfate, or a fluorosurfactant. As they are in contact with food, these active chemicals must be food-grade. Key influences on fogging behaviour of packs are:

• the difference in temperature between the air and the pack • the level of moisture in the air • the surface tension of the polymer. © 2008, Woodhead Publishing Limited

146

Chilled foods

The lower the thermal conductivity of the polymer, the longer the duration of the fogging. Temporary coatings Temporary coatings are not always reliable because they can be easily removed by friction or exposure to liquid during processing, therefore potentially making it necessary to repeat the treatment. Permanent coatings The best antifogging performance is provided by permanent coatings. These systems are often patented and processed by a licensed subcontractor. Permanent coatings achieved by a number of techniques, for example:

• • • • •

polymers including antifog additives water-soluble polymer, such as modified cyclodextrine, containing surfactant water-dispersible polyurethane oligomer water-dispersible silicone oligomer metal oxides deposited by sputtering techniques.

The adhesion of the coating over the polymer substrate is an important and complex issue because it influences seal integrity and strength.

7.5

Packing and filling technology

Once the product preservation requirements have been defined by the product and process developers, the packaging technologist must then select the optimum packaging material(s) compatible with these requirements that will then provide:

• the product with correct shelf-life and maintenance of any preservation system, including modified atmosphere packing (MAP)

• the correct product appearance (colour, visible moisture, pack transparency) and presentation

• required pack size for distribution and sale • correct strength for distribution and storage • brand communication (legal declarations, nutritional information and preparation instructions)

• functionality for the consumer (ease of opening, reclosure, microwaveability, ovenability, home freezing compatibility, etc.). It is considerably easier and quicker to design packaging for extension of an existing range of products, or stretch a design to cover a broader group of product concepts, than it is to take a new product innovation concept through to commercial realisation. The developer can draw on existing know-how and capability from existing products and this removes the need to verify and validate pack performance to the same degree as is required for new developments. © 2008, Woodhead Publishing Limited

Chilled foods packaging: an introduction

147

There are a number of packaging techniques that can be used to help maintain product quality and shelf-life of chilled products. The most commonly used of these is modified atmosphere packaging (MAP); another is vacuum packaging. Vacuum packaging and the gas flushing techniques used during MAP can be used in combinations – back-flushing. MAP is suitable for products that may be crushed if vacuum packaging techniques are used. MAP has been widely used for the last 25 years and is still continuing to develop as a technology. Product shelf-life can usually be extended by between 50% and 500% (product and microbiology dependant) using MAP techniques. This will help minimise product wastage and allow re-stocking and ordering to be managed more flexibly, given that it is not possible to accurately predict daily demand, despite very sophisticated planning techniques and the growing availability of electronic point of sales (EPOS) data. Other benefits of using MAP include:

• increased distribution range of products (and raw materials such as meat)

•

which, in recent years, has been increasingly important with products being brought greater and greater distances to the consumer (however, because of the growing anxiety of the public and the media about climate change, ‘global’ sourcing is perhaps not perceived as such a positive benefit any longer.) MAP can eliminate or reduce the level of preservatives that may be required for certain products to prevent the growth of spoilage bacteria and moulds.

7.5.1 Gases used for modified atmosphere packaging ‘Food grade’ gases are high purity gases, delivered either as liquids in bulk or semibulk containers or as compressed gases in high-pressure cylinders. Before choosing a gas mixture, many factors need to be considered. Product trials and analysis of the functions required from the gas (e.g. colour retention, and prevention of rancidity or microbial growth) are the most effective method of identifying the optimum gas mixture required to provide the maximum benefits. The effects of each gas on food products can be summarised as follows: Carbon dioxide (CO2) Carbon dioxide inhibits the growth of most aerobic bacteria and moulds. The normal rule of thumb is that the greater the level of CO2, then the greater the product shelf-life. However, CO2 is readily absorbed by fats and water; therefore most foods will readily absorb it. Excess levels of CO2 in MAP can cause flavour tainting (e.g. metallic taste), drip loss (because of pH drop) and pack collapse (because of absorption). It is important, therefore, that a balance is struck between the commercially desirable shelf-life of a product and the degree to which any negative effects can be tolerated. Nitrogen (N2) Nitrogen is an inert gas and is used to exclude oxygen. It is also used as a balance © 2008, Woodhead Publishing Limited

148

Chilled foods

gas (filler gas, e.g. 70% N2 and 30% O2) to make up the difference in a gas mixture, to prevent the collapse of packs containing high-moisture and fat-containing foods, which have a tendency to absorb carbon dioxide. Oxygen Oxygen causes oxidative deterioration of foods (e.g. colour loss or rancidity) and is required for the growth of aerobic micro-organisms, which may cause spoilage. Generally, oxygen should be excluded, but there are often good reasons for it to be present in controlled quantities such as:

• to maintain the fresh and natural colour (in red meats, for example, to maintain oxymyoglobin levels)

• to maintain respiration (in fruit and vegetables) and prevent discolouration • to inhibit the growth of anaerobic organisms (such as Clostridia in some types of fish and in vegetables). Specialised gases Argon Argon has the same properties as nitrogen. It is an inert, tasteless, odourless gas that is heavier than nitrogen, hence it can be used in a modified (controlled in a continually flushed pack) atmosphere to replace nitrogen in most applications. Its solubility (twice that of nitrogen) and certain molecular characteristics give it special properties for use with vegetables. Under certain conditions, it slows down metabolic reactions (such as ripening) and reduces respiration. However, there is still a lack of conclusive experimental evidence that the partial or total substitution of nitrogen with argon has commercially beneficial effects in terms of shelf-life extension and quality. The arguments therefore for replacing nitrogen with argon are weak, especially when the additional costs of the gas and associated piping are taken into account. Carbon monoxide Carbon monoxide (CO) is a toxic, colourless, odourless gas that is stable at up to 400 °C with respect to decomposition into carbon and oxygen. Results have shown that the use of carbon monoxide in MAP with high levels of CO2 has resulted in increased shelf-life together with retention of the bright red colour of meat cuts, where it forms a stable red pigment. It is also claimed that carbon monoxide can effectively reduce or inhibit different spoilage and pathogenic bacteria. The use of carbon monoxide in MAP is allowed in certain countries; however it cannot be used within the EU. Ozone Ozone gas (O3) is an unstable form of oxygen which has oxidising and disinfecting properties. It can be delivered safely only up to about 15% concentration in air or oxygen, having a half-life of only 20 min in clean water. One of its major benefits © 2008, Woodhead Publishing Limited

Chilled foods packaging: an introduction

149

is that it will break down to harmless elemental oxygen. Ozone is most effective when in solution or in high humidity atmospheres. Much of the early research was carried out on the disinfection of water, where ozone was shown to be much more effective than chlorine for a broad spectrum of microbial contaminants. The application of ozone gas in MAP, to improve both shelf-life and safety, has been the focus of much research. But its short half-life means that any effect will be limited to the first few minutes after packing, during which time it may disinfect product and packaging surfaces. After this, the ozone breaks down leaving a slightly higher oxygen content. However, too much ozone can cause pack damage or discolouration, and may oxidise the surface of the product, causing the release of nutrients which will encourage growth of the unwanted micro-organisms. In the USA, where ozone is now an approved food additive, most of the products on the market use ozone dissolved in water for washing of equipment and produce (e.g. salad vegetables) to aid in microbial control. Full approval for food disinfection in the EU is awaited.

7.6

Pack formats

As indicated earlier, there a wide range of packaging materials and pack formats that can be used for chilled food packaging. This section will focus on some of the most common packing formats used, and how MAP techniques might be applied during filling and closing. Matching of packaging material properties to the packaging machine mechanisms (e.g. pack presentation, filling and sealing) is key to ensuring that the machine/material interface operates as specified and that down-time and wastage are minimised.

7.6.1 Pack styles and formats There is a wide range of packaging materials for MAP. When choosing materials, the following considerations must be taken into account:

• pack format or style • pouch – pre-made or form-fill-seal from the reel (horizontal or vertical) • flowrap • semi-rigid packs, e.g pre-made tray with heat sealed lids • thermoform fill and seal trays with lids • other packs, such as injection blow moulded bottles • gas transmission barrier properties of the substrate materials. The choice of packaging material will be driven by the product, the pack format, the physical properties of the packaging material, shelf-life, marketing requirements and consumer end-use. The use of MAP techniques will enhance product preservation and appearance, and the choice of packaging material substrate will © 2008, Woodhead Publishing Limited

150

Chilled foods

be driven by their gas and water vapour transmission rates. Materials such as polyester (PET), nylon (PA), polyvinylidene chloride (PVdC) and ethylene vinyl alcohol copolymer (EVOH) provide good gas barrier properties, but in many cases the combination of high humidity and poor water vapour barrier of the layer protecting the EVOH leads to a decrease in gas barrier performance.

7.6.2 Vacuum chambers Vacuum packaging machines use preformed pouches or semi-rigid trays and lids, and utilise a vacuum chamber to extract air. Packs can be manually or automatically placed within the vacuum chamber before evacuation, back-flushing with the desired gas mixture, and heat sealing. The pack is placed in the vacuum chamber, which is then closed, and the programmed level of vacuum is produced in both the vacuum chamber and the pack. The pack is then either sealed under vacuum (vacuum package) or the chamber and therefore the pack is filled with a MAP gas mixture before the sealing operation. This type of machine can be used for smallscale production of vacuum or gas flushed packs.

7.6.3 Vacuum gas injection These machines also use a vacuum which is then back-flushed by a combination of gases. The gas, rather than flooding the whole vacuum chamber, is injected into the open pack via an injection tube or a snorkel. This technique can be used for both bigger catering packs or smaller retail packs, or used to flush big bags containing smaller filled and sealed retail packs, e.g. fresh meat.

7.6.4 Tray fill and sealing A tray sealer uses ready-made trays (ovenable board, injection moulded, thermoformed or aluminium), which are filled with product and then sealed with a flexible top web or lidding material. Air is evacuated from the sealing die and protective gas is added. Then the pack is sealed by the application of heat and pressure through a sealing head. Tray lidding and sealing machines are available from tabletop (manual) versions (≤60 packs/hr) for the small producer through to fully automatic inline versions for larger processors.

7.6.5 Horizontal form-fill-seal – flow wrappers Flow wrappers are capable of making flexible, pillow-pack style pouches or overwrapping semi-rigid trays from a reel of film. Horizontal form-fill-seal machines can also overwrap a pre-filled tray of product. The air from the package is removed by a pulse of gas or continuous gas flushing, but gas mixtures containing levels of O2 > 21% cannot be used due to the use of hot sealing jaws at the end of the machine allowing oxygen ingress. For certain very porous products (e.g. some bakery goods), gas flushing is not capable of reducing the residual O2 © 2008, Woodhead Publishing Limited

Chilled foods packaging: an introduction

151

within the package to low levels as oxygen is retained within the product. In some cases, a gas injection station can be fitted to the machine in-feed so that the product itself is purged with gas immediately prior to packaging.

7.6.6 Vertical form-fill-seal A vertical machine forms a tube; it then fills with product (in most cases dropped from an overhead multi-weigher), purges the part-sealed container with gas and then seals it. At the same time, film is transported vertically downwards. Vertical form-fill-seal machines are predominantly used for packaging foods in powder, granular, shredded or dried form.

7.6.7 Thermoform-fill-seal Packaging material for the base web (thermoformable film) is unwound from the reel. It is heated in the forming die and formed into the base of the pack (tray or bowl). The formed bases are then loaded manually or automatically and filled. The top web of packaging material (lid film) covers the filled bases. The air is evacuated from the sealing die and protective gas is added. Then the pack is sealed by the application of heat and pressure. The web of packs is cut across the machine direction initially. Production of the individual packs has been completed after the longitudinal cutting operation. Whatever the choice of packaging technology that is selected, the combination of packaging materials and packing machines must be tested by the use of robust, adequate and statistically representative trials in order to ensure that the solution not only has the capability to deliver the product shelf-life, but is also capable of meeting the challenges of the onward supply chain and meeting the needs of the consumer such that it functions flawlessly. Product understanding and structured trial design are critical aspects which need to be actively managed through to a conclusion. Adequate planning and allocation of time for representative trials will speed time to market and prevent unwanted issues arising during the critical first few months of new product introduction.

7.7

Labelling

All pre-packed products are labelled, the content of labels being determined by marketing demands for product description and presentation. Legislation controls labelling, presentation and advertising (Directive 2000/13/EC, amended by Directive 2001/101/EC). Legislation makes it obligatory for all ingredients to be indicated on the label. Labelling rules aim to ensure that consumers suffering from food allergies or those who wish to avoid eating certain ingredients for any other reason are informed. The new Directive also establishes a list of ingredients liable to cause allergies or intolerances. Labelling of products should include a Quanti© 2008, Woodhead Publishing Limited

152

Chilled foods

tative Ingredients Declaration to show the relative contribution of all ingredients (see http://ec.europa.eu/food/food/labellingnutrition/resources/fl02_en.pdf) and ingredients should be listed in descending order of weight at the time of their use in the preparation of the food. Generally, labels for chilled products should include the following information: the name of the food, a list of ingredients, the appropriate durability indication, any special storage conditions or conditions of use, the name and address of the manufacturer or seller and instructions for use. In some cases the label may include particulars of the place of origin or specific nutritional claims (see http://www.food.gov.uk/multimedia/pdfs/Fguidnot1.pdf). Two types of durability indication are used:

• ‘Best before’ is appropriate for most foods and indicates the period for which a food can reasonably be expected to retain its optimum condition.

• ‘Use by’ is the required form of date mark for foods that are highly perishable from a microbiological point of view (e.g. most chilled products) and which will have a product life after manufacture of a relatively short period, after which their consumption would present a risk of food poisoning. Use by labelling should include at least the day and month and the storage conditions (e.g. temperature). Requirements for food labelling in the USA are found in the Food, Drug, and Cosmetics Act 21 CFR, Sections 1, 74 and 101 (for a guide see http://www.cfsan. fda.gov/~dms/flg-cfr.html and http://www.cfsan.fda.gov/~dms/flg-1.html), the Food Additives Amendment and the Dietary Supplement Health and Education Act (DSHEA). Nutritional labelling is obligatory, and information must be given in a standard format per serving (as defined by the Food and Drug Administration), or as /100 g or /100 mL.

7.8

Future trends

There are a number of ‘new’ trends that have been covered by the packaging press and packing conferences in recent times which can be summarised under separate but related headings:

• active packaging • smart packaging • nanotechnology • bioactive materials • radio frequency identification – not so much a development of packaging technology in its own right, but an immensely powerful enhancement of packaging technology. This next section, which is a brief exploration of the new, is intended to provide only a flavour of these developments and is not intended to equip the reader with © 2008, Woodhead Publishing Limited

Chilled foods packaging: an introduction

153

full capability. Clearly, further exploration will be required by the reader to fully appreciate how these technologies may be used for the benefit of their businesses.

7.8.1 Active packaging Active packaging is sometimes referred to as interactive packaging. It may be an actual package structure or an adjunct or inclusion, such as a sachet or patch containing a responsive chemical or physical agent. Active packaging systems provide more than just an inert barrier to the outside world: they actually create a new interactive environment inside the package or pouch by removing undesirable components and/or adding desirable ones. Food quality is preserved by modifying the inner atmosphere to very low residual oxygen, thereby retarding the growth of spoilage bacteria and mould, biochemical and enzymatic degradation, and survival of insect larvae, while eliminating the need for food additives such as BHA, BHT and sorbates. Scavenging technology is used to maintain the internal pack environment so that shelf-life and product appearance are maximised. A number of applications are proven and it is widely used within various markets, some of which are applicable to chilled foods:

• oxygen scavenging materials • carbon dioxide production to prevent mould growth • moisture absorption, water vapour removal or humidity control to protect mould sensitive foods

• ethylene removal from horticultural produce such as fruit and vegetables • ethanol release for baked produce – which may be subject to legislative restriction depending upon region. Oxygen scavenging The presence of oxygen in food packages accelerates the spoilage and deterioration of many foods. Oxygen can cause off-flavour development, colour change, nutrient loss and microbial attack. The first uses of oxygen scavenger technology were in the Japanese market and were usually in the format of small sachets containing finely divided iron oxide. This format of the technology has been used in some countries to protect the colour of packaged cured meats from oxidation caused by oxygen in the headspace and to slow down staling and mould growth on baked products, e.g. pizza crusts. Preventing mould growth is one of the most promising applications of oxygen scavenging systems in food packages. Most moulds require oxygen to grow and, in standard packages, it is frequently mould growth which limits the shelf-life of some goods, e.g. packaged cheese and pastry products. Laboratory trials have shown that mould growth on some baked products can be stopped for at least 30 days with active packaging, and significant improvements in the mould-free life of packaged cheese have also been obtained. Another promising application is the use of active packaging to delay oxidation © 2008, Woodhead Publishing Limited

154

Chilled foods

of, and therefore rancidity development in, vegetable oils and oily foods containing unsaturated oils, e.g. olive oil. This approach, of inserting a sachet into the package, is effective but meets with resistance among food packers (due to the specialised equipment needed) and consumers. There is potential for consumers to either open the sachet and be singularly unimpressed with the black or brown powder contained therein or to view the inclusion of the sachet as additional unnecessary packaging. Carbon dioxide release High carbon dioxide levels are desirable in some food packages, because they inhibit surface growth of micro-organisms. Fresh meat, poultry, fish, cheeses and strawberries are foods which can benefit from packaging in a high carbon dioxide atmosphere. However, different foods require varying concentrations of carbon dioxide. Film permeability to carbon dioxide is generally greater than that of oxygen; therefore, carbon dioxide will need to be actively produced in some applications to maintain the desired atmosphere in the pack, or better barrier materials will be needed, which will add to packaging costs. So far the problems associated with diffusion of carbon dioxide through the packaging materials have not been resolved and this remains an important research topic. Humidity control Condensation, or ‘sweating’, is a problem in many kinds of packaged fruit and vegetables. High humidities (greater than 80%) mean there is a very real risk of condensation occurring during chilled transport, storage and display. If the water can be kept away from the produce then its appearance will be preserved. However, when the condensation wets the produce, nutrients leak into the water, encouraging rapid mould growth. Ethylene scavenging A chemical reagent, incorporated into the packaging film, traps the ethylene produced by ripening fruit or vegetables and slows ripening. The reaction is irreversible and only small quantities of the scavenger are required to remove ethylene at the concentrations at which it is produced. An alternative technique is the inclusion in the pack of a small sachet containing an appropriate scavenger, usually potassium permanganate. The sachet material itself is highly permeable to ethylene and diffusion through the sachet is not a serious limitation. Ethanol The antimicrobial activity of ethanol (or common alcohol) is well known and it is used in medical and pharmaceutical applications. Ethanol has been shown to increase the shelf-life of bread and other baked products when sprayed onto product surfaces prior to packaging. It is not usually permitted as a food additive. A novel method of generating ethanol vapour, recently developed in Japan, is through the use of an ethanol releasing system enclosed in a small sachet which is included in a food package. Food grade ethanol is absorbed onto a fine inert © 2008, Woodhead Publishing Limited

Chilled foods packaging: an introduction

155

powder which is enclosed in a sachet that is permeable to water vapour. Moisture is absorbed from the food by the inert powder and ethanol vapour is released and permeates the sachet into the food package headspace. This system is approved in Japan to extend the mould-free shelf-life of various cakes. Other developments The examples given above are only some of the commercial and non-commercial applications of active packaging. This technology is the subject of research in many countries by many companies and rapid developments may be expected. Other systems of active packaging which are either already available or could soon be seen in the market place include:

• sachets containing iron powder and calcium hydroxide which scavenge both oxygen and carbon dioxide

• films containing immobilized microbial inhibitors other than those noted above; other inhibitors being investigated include metal ions and salts of propionic acid

• specially fabricated films to absorb flavours and odours or, conversely, to release them into the package. These may use micro-encapsulation.

7.8.2 Smart packaging A package is made ‘smart’ by its functional attributes that provide benefits for consumers. These may be mechanically, chemically, electrically or electronically driven functions that enhance the usability or effectiveness of the product in some way. Examples are be time–temperature food quality or residual shelf-life labels, self-heating or self-cooling containers for beverages and foods, or milk cartons with electronic displays indicating use-by dates and information about the nutritional qualities and origin of the product in numerous languages. One recent and widely reported development in the USA is that of food quality indicators. These are small disks that are sensitive to volatile amine (TVA) levels in packs of seafood. TVAs increase as the seafood deteriorates. The technology behind the disk is based on a dye locked in a water-repellent material and used as a dot-shaped chemical indicator that changes colour as the seafood product changes in quality. As the gases from the seafood move through the dot and intermingle with the chemicals, a gradual colour change is produced in the disk when a sufficient level of the chemical is present. Clearly there will need to be some education of consumers if this technology is to be of benefit to them. Variations are being developed to monitor the freshness of poultry, meats, carbohydrates, and powdered baby food formula. Longer term, researchers are optimistic that they can manufacture materials to change properties at the same rate as products, depending on external or internal conditions, especially the temperature experienced by the pack. For example, researchers hope to use the changing molecular composition of milk that is beginning to spoil to bring about a reaction with nanoparticles embedded in the packaging, causing the colour of the packaging to change. © 2008, Woodhead Publishing Limited

156

Chilled foods

Nanotechnology enables the alteration of the structure of the indicator and functional materials on the molecular scale, to give these materials desired properties. With different nanostructures, plastics can be given tailored gas/water vapour permeabilities to fit the preservation requirements of chilled produce such as fruit, vegetables and beverages. By adding nanoparticles to packaging substrate materials, other physical properties can be enhanced, such as mechanical and thermal performance, and gas absorption can be reduced or increased. Further properties that can be induced are:

• increased fracture hardness by alumina composites • good tensile strength of packaging films through carbon nanofiber and carbon nanotube composites

• improved barrier, flame resistance, thermal and structural properties of many plastics with nanoclay composites

• improvements in oxygen-barrier properties by the use of nanoparticles in polypropylene. These properties can significantly increase product shelf-life, efficiently preserve flavour and colour, and facilitate transportation and usage. Nanostructured films can effectively prevent contamination of products and ensure food safety. With embedded nanosensors in the packaging, consumers will be able to ‘read’ the food inside. Sensors can alarm consumers before the food spoils or can inform them of the exact nutrition status of the contents. In the long run, nanotechnology is going to change the fabrication of the whole packaging industry. Self-assembly of substrate active components will, in the end, hugely reduce the fabrication costs and infrastructure costs. The introduction of more flexible packaging methods has the potential to provide consumers with fresher and more customised food products (e.g. retention or addition of flavours). Whilst nanotechnology offers huge potential to the industry and the consumer, it is presenting something of a headache to the legislators, who are still struggling to come to terms with the technology, and its potential impact on food safety, occupational safety and environmental safety. Bioactive materials It has recently been reported that a bioactive paper is being developed by researchers at ten universities across Canada. It contains ingredients that can detect and deactivate life-threatening food-, air- and water-borne bacteria and viruses such as E. coli and salmonella. Such a product would be an additional food safety weapon for processors to incorporate in their packaging, helping to prevent recalls and brand damage due to pathogen contamination. The key to developing bioactivepaper products will be the consortium’s ability to merge advances in biochemistry with current paper-production processes. The researchers are also investigating the development of a bioactive ‘ink’, which would allow biologically active chemicals to be printed, coated or impregnated onto or into paper using current processes. © 2008, Woodhead Publishing Limited

Chilled foods packaging: an introduction

157

7.8.3 Radio frequency identification tags This is rapidly developing in the retail industry as it allows traceability and the possibility to change encoded data, such as remaining shelf-life. It offers huge potential to both the retailers and the industries supplying them. It will, in the long run, yield benefits to the consumer as the capability and range of functionality of the tags increases, e.g. time–temperature indicators. The shopping experience – the self-check-out trolley – is not very far away from becoming reality, or even the much publicised internet refrigerator. Currently, the high cost of tags and absence of any agreement on a uniform detector or reading system means that their application is still limited to high margin items such as personal care products or DVDs. Wider deployment will probably be some five to ten years away.

© 2008, Woodhead Publishing Limited

8 Modified atmosphere and active packaging of chilled foods B. P. F. Day, Food Science Australia, Australia

8.1

Introduction

Recently there has been a greatly increased consumer demand for perishable chilled foods which are perceived as being fresh, healthy and convenient. The major food retailers have satisfied this consumer demand by providing an ever increasing range of value-added chilled food products. The wide diversity of chilled foods available is accompanied by a huge range of packaging materials and formats which are used to present attractively packaged foods in retail chill cabinets. This chapter overviews the requirements and types of packaging materials and formats that are commonly utilised for a broad variety of chilled food products. In addition, selected packaging technologies for extending chilled food shelf-life, such as modified atmosphere packaging and active packaging, are described in detail and new developments are highlighted.

8.2

Requirements of chilled food packaging materials

Table 8.1 lists the main requirements for a chilled food package (Turtle, 1988). Depending on the type of food packaged, not all of these requirements will need to be satisfied. The sealed package must contain the food without leaking, be nontoxic and have sufficient mechanical strength to protect the food and itself from the stresses of manufacture, storage, distribution and display. © 2008, Woodhead Publishing Limited

Modified atmosphere and active packaging of chilled foods Table 8.1 • • • • • • • • • • •

159

Main requirements of a chilled food package

Contain the product Have seal integrity Be compatible with food Prevent microbial contamination Be non-toxic Protect from odours and taints Run smoothly on filling lines Prevent dirt contamination Withstand packaging processes Resist insect or rodent infestation Handle distribution stresses

• • • • • • • • • • •

Be cost effective Prevent physical damage Have sales appeal Possess appropriate gas permeability Communicate product information Control moisture loss or gain Show evidence of tampering Protect against light Be easily openable Possess antifog properties Be tolerant to operational temperatures

Certain packs (e.g. those for fresh fruit and vegetables) require a degree of porosity to allow moisture or gaseous exchange to take place between the headspace and surrounding atmosphere, and hence packaging materials used for such applications should possess appropriate permeability properties. Alternatively, most modified atmosphere packs require moisture and gases to be retained within the pack during the shelf-life of the product and hence the packaging materials used should possess appropriate moisture and gas barrier properties. The specific requirements for modified atmosphere packs are described later, in Section 8.4.1. Depending on the type of chilled food product, the packaging material may need to be tolerant of high temperatures experienced during hot filling, in-pack pasteurisation or re-heating prior to consumption. The packaging material, particularly with high-speed continuous factory operations, may need to be compatible with form–fill–seal machines. The pack closure must have seal integrity (i.e. a hermetic seal that is capable of holding a vacuum or retaining gases within a modified atmosphere pack) but at the same time should be easy to open. There may be a need for reclosure during storage after opening in the home. Also, with the increased incidence of malicious contamination, tamper-proof or tamper-evident packaging is desirable. The package is the primary means of displaying the contained chilled food and providing product information and point-of-sale advertising. In addition to the primary packaging in direct contact with the food, many chilled food packs use secondary packaging such as protective sleeves or boxes which carry the marketing and labelling artwork. Clarity and printability are two pertinent features that require consideration in the choice of materials. Packaging materials must also be non-toxic food grade and comply with migration limits on harmful substances (e.g. plasticisers and inks) as mandated by relevant packaging legislation. Finally, the packaging must be cost-effective relative to the contained food. For example, a prepared ready meal retailing at a high price can support a considerably higher packaging cost than a yoghurt dessert selling at a fraction of that price.

© 2008, Woodhead Publishing Limited

160

8.3

Chilled foods

Chilled food packaging materials

Once the requirements (see Table 8.1 and Section 8.2) of a container for a particular chilled food product have been established, the next step is to ascertain which type of packaging material will provide the necessary properties. The answer is almost certain to be more than one type. Table 8.2 compares the main technical advantages of the most commonly used chilled food packaging materials, while the principal types of chilled packaging materials (and their abbreviations) are listed in Table 8.3 (Turtle, 1988). Packaging materials consisting of paper, glass, metal or plastic have their individual advantages and these should be exploited when making a choice. For any particular product, a number of materials can generally be used, either as separate components or in the manufacture of a composite.

8.3.1 Paper-based materials Paper and board are widely used in chilled food packaging. They are easy to decorate attractively and are complementary to all other packaging materials in the form of labels, cartons, trays or outer packaging. They are available with coatings such as wax, silicone and polyvinylidene chloride (PVDC), or as laminates with aluminium foil or flexible plastics. Such coating or lamination imparts heatsealability or improves oxygen (O2), moisture or grease barrier properties. For example, butter is traditionally packed in waxed paper or aluminium laminated paper. Dual-ovenable trays can be made of paperboard that is extrusion-coated with polyethylene terephthalate (PET). They can resist temperatures up to 220 °C and hence are suitable for microwave and conventional oven heating of chilled ready meals. Another application of paperboard in chilled food packaging is in the area of microwave susceptors which enable the browning and crisping of meat and dough products, e.g. pizza and pies, during microwave heating. A typical microwave susceptor is constructed of metallised PET film laminated to paperboard in a specific pattern to promote differential rates of microwave energy absorption.

8.3.2 Glass Glass jars and bottles are the oldest form of high-barrier packaging and have the advantages of good axial strength, product visibility, recyclability and chemical inertness. Until recently, returnable glass bottles were used extensively for pasteurised milk in the UK but these have now been replaced by LDPE bottles and plastic laminated paperboard containers. Aluminium caps and closures make opening simple, whilst tamper-evident features such as pop-up buttons provide an important consumer safety factor. Impact breakage of glass containers is a major disadvantage, but new glass technology and plastics sleeving with polyvinyl chloride (PVC) or expanded polystyrene (EPS) have helped to reduce glass breakage. © 2008, Woodhead Publishing Limited

Modified atmosphere and active packaging of chilled foods Table 8.2

Comparison of chilled food packaging materials

Packaging material

Main technical advantages

Aluminium

• • • • • • • • • • • • • • • • • • • • •

Paper

Semi-rigid plastics

Flexible plastics Glass

Table 8.3 • • • • • • • • • • •

161

Excellent gas and moisture barrier properties Lightweight Container axial strength Withstands internal pressure Great variety of paper grades Ease of decoration Adjunct to all other packaging materials Lightweight Properties variable with type of plastic Choice of container shape In-house manufacture Lightweight Properties variable by combination Very lightweight containers Tailor-made sizing Chemically inert Excellent gas and moisture barrier properties Product visibility Container axial strength Withstands internal vacuum pressure Reuse facility

Chilled food packaging materials

Aluminium foil Cardboard Cellulose Cellulose fibre Glass Natural casings Paper Metallised board Metallised film Steel Plastics: ¤ ABS (acrylonitrile-butadiene-styrene) ¤ APET (amorphous PET) ¤ CA (cellulose acetate) ¤ CPET (crystalline PET) ¤ CPP (cast polypropylene) ¤ EPS (expanded polystyrene) ¤ EVA (ethylene-vinyl acetate)

© 2008, Woodhead Publishing Limited

¤ ¤ ¤ ¤ ¤ ¤ ¤ ¤ ¤ ¤ ¤ ¤ ¤ ¤ ¤ ¤ ¤ ¤

EVOH (ethylene-vinyl alcohol) HDPE (high density polyethylene) HIPS (high impact polystyrene) LDPE (low density polyethylene) LLDPE (linear low density polyethylene) MXDE (modified nylon) OPP (orientated polypropylene) OPS (orientated polystyrene) PA (polyamide–nylon) PC (polycarbonate) PE (polyethylene) PET (polyethylene terephthalate) PETG (modified PET) PP (polypropylene) PS (polystyrene) PVC (polyvinyl chloride) PVDC (polyvinylidene chloride) UPVC (unplasticised polyvinyl chloride)

162

Chilled foods

8.3.3 Metal-based materials Pressed aluminium foil trays have a long history of use for prepared frozen meals and hot take-away food. They are also used for many chilled ready meals. Their temperature stability makes them ideal for conventional oven heating, but precautions should be taken to prevent arcing if used in microwave ovens. Guidelines have been developed for the successful use of foil containers in microwave ovens (Foil Container Bureau, 1991). In some circumstances, aluminium foil enables more uniform heating than microwave-transparent trays (Bows and Richardson, 1990). Aluminium foil or aluminium laminated paper are also used for many dairy products, such as butter, margarine and cheese. Aluminium foil is used in carton board composite containers for chilled fruit juices and dairy beverages. In addition, aluminium or steel aerosol containers are used for chilled creams and processed cheeses.

8.3.4 Plastics Plastics are the materials of choice for the majority of chilled foods. Chilled desserts, ready meals, dairy products, meats, seafood, pasta, poultry, fruit and vegetables are all commonly packed in plastic or plastic-based materials which are often laminated with oxygen- or moisture-barrier materials and/or cardboard for strength or machinability. Semi-rigid plastic containers for chilled foods are predominantly made from polyethylene (PE), polypropylene (PP), polystyrene (PS), PVC, PET and acrylonitrile-butadiene-styrene (ABS). Other plastics such as polycarbonate (PC) are used in small quantities. Containers are available in a wide range of bottle, pot, tray and other shapes and thermoforming, injection moulding and blow moulding techniques give food processors the option of in-house manufacture. Flexible plastics offer the cheapest form of barrier packaging and may be used to pack perishable chilled food under vacuum or modified atmosphere. Multilayer materials are typically made by coextrusion or coating processes, using sandwich layers of PVDC or ethylene-vinyl alcohol (EVOH) to provide an O2 barrier. Alternatively, plastics such as PE or PP may be metallised or laminated with foil to provide very high-barrier materials. The required technical properties and pack size and shape may be matched to a desired specification, thereby ensuring cost effectiveness.

8.4

Packaging techniques for chilled food

Food preservation technology and packaging techniques often go hand-in-hand in ensuring the safety and desired shelf-life of chilled food products. Chilled food products are perishable and their achievable shelf-life is highly dependant on the intrinsic properties of the food or the extrinsic factors of the storage environment. Intrinsic properties of foods include acidity (pH), water activity (aw), nutrient content, occurrence of antimicrobial compounds, redox potential, respiration rate and biological structure. Extrinsic factors include storage temperature, relative © 2008, Woodhead Publishing Limited

Modified atmosphere and active packaging of chilled foods

163

humidity (RH) and the gaseous composition surrounding the food. All of these properties and factors directly influence the quality, microbial safety and achievable shelf-life of chilled food products, as explained in detail in other chapters of this book. Several preservation techniques are available for extending the shelf-life of perishable chilled foods. Obviously, the achievable shelf-life will depend not only on the technique used but on the intrinsic nature of the food, initial microbial load, storage temperature, gaseous environment, packaging material and integrity of the package. For example, mild pasteurisation heat treatments are used for many chilled products (e.g. cook–chill ready meals, fresh pasta, combination food products, milk and other dairy products) to reduce microbial loads and inactivate certain degradative enzymes. Other preservation techniques include the use of additives and preservatives (e.g. acids, salt, sugar, spices, antioxidants, sulphites, nitrates, benzoates, sorbates and propionates). Packaging materials and techniques play a crucial role in ensuring the safety and desired shelf-life of chilled food products. Chilled food packs act as a barrier to external contamination and therefore it is essential that the integrity of the packs is maintained during storage and prior to opening by the consumer. This section specifically describes the techniques of modified atmosphere packaging, active packaging, vacuum packaging and vacuum skin packaging all of which can extend the shelf-life of chilled food products.

8.5

Modified atmosphere packaging

Modified atmosphere packaging (MAP) is an increasingly used food preservation technique that minimally affects the characteristics of fresh food, and hence fits in well with current consumer demands for fresh, healthy, convenient and additivefree food. MAP is now being used for extending the shelf-life, improving the product image and reducing the wastage of a wide range of chilled perishable foods, as well as ambient-stable food products (Day, 1992; Blakistone, 1998; Air Products Plc, 2006). The application of MAP for maintaining the quality and assuring the safety of perishable food items is a complicated subject, but guideline documents are available to help food manufacturers with the implementation of MAP technology (Day, 1992; Betts, 1996; Air Products Plc, 2006). This section provides an overview of MAP technology and highlights the effects of MAP on the spoilage mechanisms of selected major food applications. In addition, summary details of an expert system for MAP are outlined (Day and Wiktorowicz, 1999; Air Products Plc, 2006).

8.5.1 Background information The use of gases for the preservation of food products has been known for over a century but during the last three decades there has been considerable growth in the © 2008, Woodhead Publishing Limited

164

Chilled foods

volumes and applications of MAP for extending the shelf-life of foods. Depending on the food product, shelf-life can be extended by 50–500% by using MAP techniques (Air Products Plc, 2006). Retail sales of fresh, healthy and convenient perishable food products have shown explosive growth during recent years. Such foods demand rationalisation and control throughout the entire chilled distribution chain and major retailers have embraced MAP as a preservation technique to meet their logistical supply requirements. MAP has several related technologies and terminologies: (i)

MAP involves the removal of air from the package and its replacement with a single gas or mixture of gases that is different from the normal composition of air (78.08% nitrogen (N2), 20.96% oxygen (O2), 0.04% carbon dioxide (CO2), variable amounts of water vapour and traces of inert gases). It should be noted that the gaseous atmosphere inside a modified atmosphere (MA) pack changes continuously during storage due to absorption of gases, respiration of certain food products, biochemical reactions and exchange of gases through the MA pack. (ii) The term controlled atmosphere packaging (CAP) is often used synonymously with MAP but this terminology is a misnomer since it is not possible to control the in-pack atmosphere once the pack has been hermetically sealed. (iii) Controlled atmosphere storage (CAS) is similar to MAP in that it involves the storage of food in an atmosphere different from air. However, in CAS the gaseous components are precisely adjusted to specific concentrations throughout the storage and distribution of perishable foods. CAS is used in the bulk warehouse storage of fruit and vegetables and the road or sea-freight transportation of perishable foods. Hypobaric or low pressure storage is a form of CAS where the pressure is accurately controlled along with the temperature and relative humidity. Although hypobaric storage is relatively expensive, it has been used for the bulk storage of soft fruits. (iv) Vacuum packaging involves the simple evacuation of air from within a pack prior to hermetic sealing. Hence, vacuum packaging does not involve the replacement of the evacuated air with a gas mixture, as is the case with MAP. Notwithstanding, vacuum packaging does reduce the partial pressure of atmospheric gases within the vacuum packs and hence is capable of extending the shelf-life of perishable foods. Vacuum packaging is described briefly in Section 8.7. (v) Active packaging refers to the incorporation of certain additives into packaging film or within packaging containers with the aim of maintaining and extending food product shelf-life. Packaging may be termed active when it performs some desired role in food preservation other than providing an inert barrier to external conditions. Active packaging includes additives or ‘freshness enhancers’ that are capable of scavenging or releasing O2 and CO2 and hence the effects of active packaging on food products can be similar to those achieved through MAP. Active packaging is described in Section 8.6. © 2008, Woodhead Publishing Limited

Modified atmosphere and active packaging of chilled foods

165

Table 8.4 Gas-mix guide for the MAP of retail food products (adapted with permission from Air Products Plc, 2006) Food item

CO2 (%)

O2 (%)

N2 (%)

Meat (red) Meat (cured) Meat (cooked) Offal (raw) Poultry (white) Poultry (reddish) Fish (white) Fish (oily) Crustaceans and molluscs Fish (cooked) Pasta (fresh) Ready meals Combination products Cheese (hard) Cheese (soft) Other dairy products Fresh fruit/vegetables Vegetables (cooked) Liquid foods and beverages Carbonated soft drinks Bakery products Dried food products

15–30 30–40 30–40 15–25 30–50 25–35 35–50 35–45 35–50 25–35 25–35 25–35 25–35 100 35–45 – 3–10 25–35 – 100 60–100 –

70–85 – – 75–85 0–20 65–75 25–35 – 25–35 – – – – – – – 2–10 – – – – –

– 60–70 60–70 – 50–70 – 25–40 55–65 25–40 65–75 65–75 65–75 65–75 – 55–65 100 80–95 65–75 100 – 0–40 100

8.5.2 Gases used for MAP The gas mixture used in MAP (see Table 8.4) must be chosen to meet the needs of the specific microbiological and sensory (e.g. taste and colour) characteristics of the individual food product, but for nearly all products some combination of CO2, O2 and N2 will be suitable (Day, 1992; Blakistone, 1998; Day and Wiktorowicz, 1999; Air Products Plc, 2006). For example, Fig. 8.1 illustrates a retail display of MA-packed fresh pasta. Carbon dioxide Carbon dioxide has bacteriostatic and fungistatic properties and will retard the growth of mould and aerobic bacteria. In some types of food products, the intrinsic production of CO2 by respiration and microbial metabolism and its retention by MA packs will also reach inhibitory levels. The combined inhibitory effects on various enzymic and biochemical pathways result in an increase in the lag-phase and generation time of susceptible spoilage micro-organisms. However, CO2 does not retard the growth of all types of micro-organisms. For example, the growth of lactic acid bacteria is improved in the presence of CO2 and a low O2 content. CO2 has little effect on the growth of yeast cells. The inhibitory effect of CO2 is increased at low temperatures because of its enhanced solubility in water to form © 2008, Woodhead Publishing Limited

166

Chilled foods

Fig. 8.1

Retail display of MA packed fresh pasta.

a mild carbonic acid which causes a drop in the food product’s pH. The practical significance of this is that MAP does not eliminate the need for refrigeration. The absorption of CO2 by food products is highly dependent on the water and fat content of the product. Excess CO2 absorption causes a pH drop and can reduce the water-holding capacity of meats, resulting in unsightly drip. In addition, some dairy products can be acid-tainted, and fruit and vegetables can suffer physiological damage (e.g. tissue discolouration) owing to high CO2 levels. If products absorb excess CO2, the total volume inside the package will be reduced, giving a vacuum package look known as pack collapse. Oxygen In MAP, initial O2 levels are normally set as low as possible (typically, 0.3–3.0% residual O2 levels are achievable) to inhibit the growth of aerobic spoilage microorganisms and to reduce the rate of oxidative deterioration of foods. However, there are exceptions; for example, O2 is needed for fruit and vegetable respiration, colour retention in red meats or to avoid anaerobic conditions in white fish MA packs. Nitrogen Nitrogen is effectively an inert gas and has a low solubility in both water and fat. In MAP, N2 is used primarily to displace O2 in order to retard aerobic spoilage and oxidative deterioration. Another role of N2 is to act as a filler gas so as to prevent pack collapse. © 2008, Woodhead Publishing Limited

Modified atmosphere and active packaging of chilled foods

167

Fig. 8.2 The interchangeable meat myoglobin forms and associated colours (adapted with permission from Sørheim et al., 1997).

Carbon monoxide Carbon monoxide (CO) produces a stable, cherry red colour in meat (carboxymyoglobin), due to its strong binding to the muscle pigment deoxymyoglobin (see Fig. 8.2). CO in MAP has been used commercially for retail red meat packaging in Norway since 1985. In 2002, CO in MAP was introduced for industrial master-bag packaging in the USA. A growing interest for commercial use and research on CO has lately been seen in the USA and planned commercialisation of CO technology took place in New Zealand and Australia during 2004. Low concentrations of CO (<0.5%) combined with anaerobic high CO2 atmospheres improve meat colour; inhibit lipid oxidation, bone discolouration and premature browning of cooked meat; extend microbiological shelf-life; reduce the growth of certain pathogenic bacteria; pose no toxic hazard to consumers, and are safe to use in meat packaging plants (Sørheim et al., 1997). The reference of Sørheim et al. contains detailed information on the inhibitory effects of CO on micro-organisms and issues related to the safety of CO use in meat packaging plants and for consumers. Argon Argon (Ar) is classified as a miscellaneous additive and is a permitted gas for food use in the European Union. Air Liquide S.A. (Paris, France) has stimulated commercial interest in the potential MAP applications of using Ar. Spencer (2005) claims that >200 food products in the UK and elsewhere in Europe, including a wide range of ready-to-eat foods, are MA-packed in Ar-containing gas mixtures. Air Liquide’s broad range of patents claim that, in comparison with N2, Ar can more effectively inhibit enzymic activities, microbial growth and degradative chemical reactions in selected perishable foods. Although Ar is chemically inert, Air Liquide’s research has indicated that it does have biochemical effects, probably © 2008, Woodhead Publishing Limited

168

Chilled foods

due to its similar atomic size to molecular O2 and its higher density and solubility in water compared with N2 and O2 (Brody and Thaler, 1996). Hence, Ar is probably more effective at displacing O2 from cellular sites and enzymic O2 receptors, with the consequence that oxidative deterioration reactions are likely to be inhibited (Day, 1998). More independent research is needed to better understand the potential beneficial effects of Ar. Research carried out by Air Products Plc has shown that Ar demonstrates some properties which are beneficial but the argument for replacing N2 with Ar is marginal, especially when the additional costs of the gas and associated piping are taken into account (Air Products Plc, 2006). Other gases Other gases, such as ozone, nitrous oxide, ethylene oxide, helium, neon, propylene oxide, ethanol vapour, hydrogen, sulphur dioxide and chlorine have been used experimentally or on a restricted commercial basis to extend the shelf-life of a number of food products. However, the commercial use of these other gases is severely limited owing to safety concerns, regulatory constraints, negative effects on sensory quality or economic factors (Day, 1992).

8.5.3 The benefits and disadvantages of MAP Any food manufacturer considering the use of MAP should first evaluate the benefits of extended shelf-life under MAP, and then compare these benefits against the additional cost of using this preservation system (Day, 1992; Flores and Matsos, 2005). The benefits of MAP include:

• extended food product shelf-life leading to: • The ability to use centralised packaging and distribution and hence alleviate the need for in-store packaging;

• Better utilisation of labour and equipment by flattening production peaks and allowing longer runs of individual food products;

• Economies of scale by facilitating the purchase of larger quantities of raw materials;

• Longer supply chain distances which help increase market range; • Enhancement of sales appeal owing to the attractive colour and presentation of food products;

• Sealed hygienic packs which prevent drips and odours emanating from the packs during distribution;

• Longer supply chain distances and increased product markets; • No need for artificial preservatives. Against these benefits, the following disadvantages of MAP must be considered:

• The capital cost of MAP machinery and accessories; • Slower pack filling rates; • The cost of gases and specialised MAP materials; © 2008, Woodhead Publishing Limited

Modified atmosphere and active packaging of chilled foods

169

• The cost of gas analysis and seal integrity testing equipment; • Increased pack volume which will affect transport costs and display in retail stores.

8.5.4 MAP materials Specifically with regard to MAP, the main characteristics to consider (Day, 1992; Air Products Plc, 2006) when selecting packaging materials are: Gas permeability In most MAP applications, excluding fresh fruit and vegetables, it is desirable to maintain the atmosphere initially incorporated into the MA pack for as long a period as possible. The correct atmosphere at the start will not serve for long if the packaging material allows its gas composition to change too rapidly. Consequently, packaging materials used with all forms of MA-packed foods (with the exception of fresh fruit and vegetables) should have barrier properties. The permeability of a particular packaging material depends on several factors such as the nature of the gas, the structure and thickness of the material, whether the food product is in direct contact with the material, the temperature and the relative humidity (RH). Although CO2, O2 and N2 permeate at quite different rates, the order CO2 > O2 > N2 is always maintained and the permeability ratios CO2/O2 and O2/N2 are usually in the range 3 to 5. Hence, it is possible to estimate the permeability of a material to CO2 or N2 when only the O2 permeability is known. As a general rule, packaging materials with O2 transmission rates < 100 cm3 m–2 day–1 atm–1 are used in MAP. Packaging materials are usually laminated or coextruded in order to have the necessary barrier properties. Water vapour transmission rate Water vapour transmission rates are quoted in g m–2 day–1 at a given temperature and relative humidity (RH). Similar to gas permeabilities, there is a wide variation between different packaging materials. However, there is no correlation between what is a good barrier to gas and what is a good barrier to water. A further complication is that some materials (e.g. nylons) are moisture-sensitive and their gas permeabilities are dependent on RH. Mechanical properties Packaging materials used for MAP must have sufficient strength to resist puncture, withstand repeated flexing and endure the mechanical stresses encountered during pack forming, filling, handling and distribution. Additionally, if trays are to be thermoformed, the web must draw evenly and not thin excessively on the corners. Poor mechanical properties (e.g. stress cracking and delamination) can lead to pack damage and gas leakage. Sealing reliability It is essential that an integral seal (i.e. one that is hermetic and has a specified © 2008, Woodhead Publishing Limited

170

Chilled foods

strength so that there is no direct path for gas to escape from the pack headspace to the outside atmosphere) is formed in order to maintain the correct atmosphere within a MA pack. Therefore, it is important to select the correct heat-sealable packaging materials and to control the sealing operation. For example, in highspeed form–fill–seal operations, it is important to consider the hot tack of the material. Additionally, there is often a requirement for a peelable seal so that the consumer can gain easy access to the contents. However, the balance between peelability and seal strength must be determined. Transparency For most MA-packed foods, a transparent package is desirable so that the product is clearly visible to the consumer. However, high-moisture foods stored at chilled temperatures have the tendency to create a fog of condensed water droplets on the inside of the package, thereby obscuring the product. Consequently, many MAP films are treated with coatings or additives to impart antifog properties so as to improve visibility. These treatments affect only the wetability of the film and have no effect on the permeability properties of the film. For some MA-packed foods (e.g. green pasta and cured meats), it may be desirable to exclude light in order to reduce undesirable light-induced oxidation reactions. In these cases, light barriers such as colour-printed or metallised films may be used. Type of package The type of package used will depend on whether the product is destined for the retail or the catering trade. Popular options include flexible ‘pillow-packs’, ‘bagin-box’ and semi-rigid tray and lidding film systems. Microwaveability The ability of MAP materials to withstand microwave heating is important, particularly in the case of ready-to-eat food products. For example, the low softening point of PVC makes the popular PVC/LDPE thermoformed trays unsuitable for microwave oven heating as they lose physical strength on heating to >100 °C and present a risk of spilling hot product. Hence, materials with greater heat resistance and structural stability, such as cast polypropylene (CPP), crystalline polyethylene terephthalate (CPET), polypropylene (PP) and polystyrene-high temperature (PSHT) are used for MA-packed food products intended to be heated in a microwave oven. Most sealed MA packs intended for microwave heating need to be punctured prior to heating, although some packs contain pressure release valves (e.g. the Steam Cuisine® range of ready meals from Marks & Spencer, UK) or have been designed with pressure sensitive seals that open upon heating.

8.5.5 Selected MAP food applications MAP has applications in virtually every food product category. In 1995, The MAP Gas Selector expert system was developed by Air Products Plc, in conjunction with the Campden & Chorleywood Food Research Association, who provided the © 2008, Woodhead Publishing Limited

Modified atmosphere and active packaging of chilled foods

171

technical input information. The MAP Gas Selector has recently been updated and is freely available to interested parties via the Air Products Plc FreshlineTM website on http://www.airproducts.com/freshline. The MAP Gas Selector was designed to provide technical advice on the MAP of over 450 individual food items which have been grouped within 17 different food categories. Users can select any food item (in a retail, bulk or primal pack format) and the MAP Gas Selector will provide tabulated information on the recommended MAP gas mixture, labelling requirements, storage temperatures, achievable shelf-lives in air and in MAP, principal spoilage mechanisms, possible food poisoning hazards, typical MAP machines and types of packages, and examples of typical MAP materials. In addition, concise technical advice and help information are provided on the following:

• • • • • • • •

definitions and terminology of MAP related topics abbreviations of commonly used MAP materials line diagrams of the most common types of MAP machinery descriptive paragraphs and available growth data on 23 groups of microorganisms descriptive paragraphs on the principal food spoilage mechanisms of different categories of food products and how MAP inhibits deleterious quality changes recommendations on food hygiene, temperature control, stock rotation and quality assurance tests descriptions of legislation and regulations of specific relevance to MAP the entire text of the Freshline® guide to MAP (Air Products Plc, 2006).

The following text from six of the 17 food categories has been extracted and edited (with permission from Air Products Plc) for the selected MAP food applications described below: Raw red meat products The two principal spoilage mechanisms affecting the shelf-life of raw red meats are microbial growth and oxidation of the red oxymyoglobin pigment. When red meat is kept under proper chilled conditions, the controlling influence on the shelf-life of the product is the rate of oxidation of the red oxymyoglobin pigment to its brown oxidised form, metmyoglobin (see Fig. 8.2). For this reason, high concentrations of O2 are necessary for the MAP of red meats in order to maintain the desirable bright red colour for a longer period. Highly pigmented red meats, such as venison and wild boar, require higher concentrations of O2. Alternatively, as mentioned in Section 8.5.2 , low concentrations of CO (< 0.5%) are currently allowed in the USA, Australia and New Zealand to produce a stable cherry red meat colour (carboxymyoglobin) that is similar to the red oxymyoglobin colour that forms when meat is exposed to O2 in the air, or within high O2 MA packs. However, the use of CO in MAP is controversial and it is currently not a permitted packaging gas within the European Union countries because of perceived safety concerns (although Sørheim et al., 1997 present a compelling scientific argument for its use). Red meats provide an ideal medium for the growth of a wide range of spoilage and food poisoning micro-organisms. It should be noted that raw red meats are © 2008, Woodhead Publishing Limited

172

Chilled foods

Fig. 8.3

Retail display of MA packed red kangaroo meat from Australia.

usually cooked before consumption and thorough heating to pasteurisation temperatures is sufficient to kill the vegetative cells of food poisoning bacteria. Consequently, the risk of food poisoning is greatly minimised by proper cooking. Aerobic spoilage bacteria, such as Pseudomonas species, which are normally predominant on red meats, are inhibited by CO2. Consequently, to create the dual effect of red colour stability and microbial inhibition, initial MAP gas mixtures containing 15–30% CO2 and 70–85% O2 are recommended for extending the chilled shelf-life of red meats from 2–4 days to 5–8 days and even longer. For example, Fig. 8.3 illustrates the retail display of MA-packed red kangaroo meat from Australia. The maintenance of recommended chilled temperatures and good hygiene and handling throughout the butchery, MAP, distribution and retailing chain are also of vital importance in ensuring the microbial safety and extended shelf-life of red meat products. Raw fish and seafood products The principal spoilage mechanisms affecting the quality of fish and seafood are the result of microbial growth and oxidative lipid rancidity. Fish and seafood products are very perishable due to their high aw, neutral pH, and presence of autolytic enzymes which cause the rapid development of undesirable odours and flavours. Fish normally have a particularly heavy microbial load owing to their water origin, method of capture and transport to shore, evisceration (i.e. filleting) and retention of skin in retail portions. Microbial activity causes a breakdown of fish protein, © 2008, Woodhead Publishing Limited

Modified atmosphere and active packaging of chilled foods

173

with the resulting production of undesirable fishy odours such as trimethylamine. Oxidative rancidity of unsaturated fats in oily fish also results in additional offensive odours and flavours. Only the highest quality fish and seafood should be used to benefit from the extended shelf-life advantages of MAP. The achievable shelf-life (typically 4–6 days in MAP compared with only 2–3 days in air) will depend on the species, fat content, initial microbial load, gas mixture and chilled temperature of storage. The maintenance of recommended chilled temperatures and good hygiene and handling practices throughout the entire capture-to-consumption chain is essential for ensuring the safety and extended shelf-life of fish and seafood products. MAP is a very effective technique for delaying microbial spoilage and oxidative rancidity in fish and seafood products. MAP is particularly effective at extending the shelf-life of white fish products. For white fish, crustaceans and molluscs, an initial MAP gas mixture containing 30% O2, 40% CO2 and 30% N2 is recommended. An initial MAP gas mixture containing 40% CO2 and 60% N2 is recommended for oily fish products. The inclusion of CO2 is necessary for inhibiting common aerobic spoilage bacteria, such as Pseudomonas species. However, for retail packs of fish and other seafood, too high a proportion of CO2 in the gas mixture can induce pack collapse and excessive drip, and in cold-eating seafood products such as crab, an acidic, sherbet-like flavour. O2 is necessary to prevent the growth of Clostridium botulinum type E, colour changes and bleaching, and reduce the drip in white fish, crustacean and mollusc MA packs. However, O2 is preferentially excluded from oily fish MA packs so as to inhibit oxidative lipid rancidity. Cooked, cured and processed meat products The principal spoilage mechanisms for cooked, cured and processed meat products are microbial growth, colour changes and oxidative rancidity. In such products, the heating process should kill vegetative bacterial cells, inactivate enzymes and preserve the colour. Problems can arise primarily from under-cooking, post-process contamination and/or poor hygiene and handling practices, which can all lead to unacceptably high microbial levels on cooked, cured and processed meat products. Some raw, uncured, meat products (such as beef burgers and British sausages) contain sulphur dioxide (often added in the form of sodium metabisulphite). This additive (use of which is restricted to products having a minimum of 6% cereal content) is an effective preservative against a wide range of spoilage mechanisms and it also acts as a reducing agent or antioxidant. Cured meat products, whether cooked or not, owe their characteristic pink colour to the use of nitrate which interacts with the myoglobin in the meat to form nitrosylmyoglobin. Although this pigment is fairly stable, it is prone to oxidative bleaching, especially when exposed to light. Cured meat products should therefore be MA packed with the exclusion of O2. Vacuum packaging is still extensively used commercially but MAP is the preferred option for sliced products that require easy slice separation. Meat products containing appreciable levels of unsaturated fat are liable to be spoiled by oxidative rancidity, but MAP with the elimination of O2 will also inhibit this. The addition of nitrate and salt inhibits the growth of most food poisoning © 2008, Woodhead Publishing Limited

174

Chilled foods

bacteria but the food safety importance of good chilled temperature control (0– 3 °C) cannot be over emphasised because microbial inhibition may be compromised in products formulated with reduced levels of salt, nitrate or other preservatives. Therefore, caution must be exercised in assessing the potential effects of any changes in product formulation. Cooked meats without any added preservatives may be at risk from the growth of Clostridium botulinum and Listeria monocytogenes under anaerobic MAP and incorrect chilled storage. Dairy products The principal spoilage mechanisms affecting dairy products are microbial growth and oxidative rancidity. The type of spoilage affecting dairy products will depend on the intrinsic properties of the different products. For example, low aw products such as hard cheeses are generally spoiled by mould growth, whereas higher aw products such as creams and soft cheeses are susceptible to yeast and bacterial spoilage, oxidative rancidity and physical separation. MAP can significantly extend the shelf-life of dairy products. Similar shelflives are achieved for MAP in comparison with vacuum packaging. Hard cheeses are generally packed in CO2, which is very effective at inhibiting mould growth. Soft cheeses are packed in CO2/N2 mixtures, which can also inhibit bacterial spoilage and oxidative rancidity. For soft or grated cheese, 40% CO2, 60% N2 is recommended. MAP is particularly effective for crumbly cheeses such as Lancashire and grated cheese where vacuum packaging would cause undesirable compression. MAP is not recommended for mould-ripened cheeses since CO2/N2 gas mixtures would inhibit desirable mould growth. Creams have a delicate flavour and are adversely affected by CO2-containing atmospheres, which cause acidification of the cream, giving it a sharp rather than a smooth taste. Consequently, N2 is recommended for MAP of creams and cream-containing products. By exclusion of air, N2 is also capable of inhibiting aerobic microbial growth and oxidative rancidity. Aerosol creams use nitrous oxide (N2O) as a propellant, which also inhibits oxidative rancidity. Other dairy products, such as butter and yoghurt, are not usually MA packed but would benefit from packaging under N2. Possible food poisoning hazards associated with dairy products are primarily due to either inadequate pasteurisation or cross-contamination during or after packaging. Consequently, adequate pasteurisation, the maintenance of recommended chill temperatures, and good hygiene and handling throughout are essential for ensuring the safety of dairy products. Bakery and combination products The principal spoilage mechanisms for non-dairy bakery products are mould growth, staling and moisture migration. Yeasts may cause a problem in certain filled or iced products. Since the aw of non-dairy bakery products is generally < 0.96, bacterial growth is inhibited and rarely a problem. However, it is possible that Staphylococcus aureus and Bacillus species may be able to grow in certain products and hence pose a potential food poisoning hazard. Consequently, good hygiene and handling practices must be observed throughout. © 2008, Woodhead Publishing Limited

Modified atmosphere and active packaging of chilled foods

175

The use of MAP can significantly extend the shelf-life of non-dairy bakery products. Since the moulds are aerobic micro-organisms, they are very effectively inhibited by CO2/N2 gas mixtures. Moisture migration from the pack is prevented by using barrier materials for MAP. The gases appear to have little effect on the rate of staling. It should be noted that staling rates are increased at chilled temperatures and hence most cold-eating bakery products are normally stored at ambient temperatures. For hot-eating bakery products, such as pizza bases, the staling process, which is caused by starch retrogradation, is partially reversed during the reheating cycle. Combination products are made up of two or more different food components and include battered and breaded meats, fish, seafood and poultry products; burritos; enchiladas; falafels; filled pancakes and rolls; filled fresh pasta; pies; pizzas; quiche; cook–chill ready meals and sandwiches. Due to the vast differences in the intrinsic properties of these products, and the interactions between separate components in the same food product, only generalisations regarding spoilage mechanisms, possible food poisoning hazards, achievable shelf-lives and recommended MAP gas mixtures can be made. Food manufacturers considering the use of MAP for such products must carry out detailed shelf-life evaluations to determine the optimal gas mixture, spoilage mechanisms, etc. The principal spoilage mechanisms likely to affect combination products are microbial growth and oxidative lipid rancidity. CO2/N2 gas mixtures are recommended to effectively inhibit microbial spoilage and lipid rancidity development and hence to significantly extend shelf-life. However, MAP has no effect on moisture migration between different components of certain combination products (e.g. sandwiches). It should be noted that many combination products are cooked or contain cooked ingredients. Consequently, the possible food poisoning hazards associated with these types of products are primarily due to post-cooking and/or postpackaging contamination. These hazards can be minimised by adequate cooking and maintenance of recommended chilled temperatures (0–3 °C), and good hygiene and handling practices. Fresh produce Unlike other chilled perishable foods that are MA packed, fresh produce (i.e. produce that has not been heat treated) continues to respire after harvesting, and any subsequent packaging must take into account this respiratory activity, which is much greater in magnitude than that contributed by microbial respiration. The depletion of O2 and enrichment of CO2 are natural consequences of the progress of respiration when fresh produce is stored in hermetically sealed packs. Such modification of the atmosphere results in a respiratory rate decrease with a consequent extension of shelf-life (Kader et al., 1989). MAs can passively evolve within hermetically air-sealed packs as a consequence of produce respiration. If a produce item’s respiratory characteristics are properly matched to film permeability values, then a beneficial equilibrium MA (EMA) can be passively established. However, in the MAP of fresh produce, there is a limited ability to regulate © 2008, Woodhead Publishing Limited

176

Chilled foods

passively established MAs within hermetically air-sealed packs. There are many circumstances when it is desirable to rapidly establish the atmosphere within produce packs. By replacing the pack atmosphere with a desired mixture of O2, CO2 and N2, a beneficial EMA may be established more rapidly than a passively generated EMA (Day, 1998). Currently, the key to the successful retail MAP of fresh prepared produce is to use packaging film of correct permeability so as to establish optimal EMAs of typically 3–10% O2 and 3–10% CO2. The EMAs attained are influenced by produce respiration rate (which itself is affected by temperature, produce type, variety, size, maturity and severity of preparation); packaging film permeability; pack volume, surface area and fill weight; and degree of illumination. Consequently, establishment of an optimum EMA for individual produce items is very complex. Furthermore, in many commercial situations, produce is sealed in packaging film of insufficient permeability, resulting in development of undesirable anaerobic conditions (e.g. < 2% O2 and > 20% CO2). Microperforated films, which have very high gas transmission rates and WVTRs, are now used for maintaining aerobic EMAs (e.g. 5–15% O2 and 5–15% CO2) for highly respiring prepared produce items such as broccoli and cauliflower florets, carrot batons, beansprouts, mushrooms and spinach. These microperforated films can also be beneficial in allowing moisture to escape from high aw fresh produce packs, thereby inhibiting microbial rots. However, microperforated films are relatively expensive, permit moisture and odour losses, and may allow for the ingress of micro-organisms into sealed packs during wet handling situations (Day, 1998).

8.6

Active packaging

Active packaging has been variously classified in the literature using a number of differing definitions. Some of these definitions are either so broad that they include many packages that are clearly not active, or so narrow that they exclude important subsets of active packaging (Robertson, 2006). According to previous reviews, active packaging has been classified as a subset of smart packaging and referred to as the incorporation of certain additives into packaging film or within packaging containers with the aim of maintaining and extending product shelf-life (Day, 2001; 2003). However, as pointed out by Robertson (2006), this definition focuses on the additives that make a package active and hence excludes certain categories such as temperature compensating polymeric films for fresh fruit and vegetables. Another definition states that packaging may be termed active when it performs some desired role in food preservation other than providing an inert barrier to external conditions (Rooney, 1995). Robertson (2006) correctly identifies ‘desired’ and ‘inert’ as the key words in this definition since all packaging materials, except glass, are not totally inert and can contribute undesirable components to food or absorb desirable components from food. Consequently, for the purposes of this chapter, active packaging is defined as ‘packaging in which subsidiary constituents © 2008, Woodhead Publishing Limited

Modified atmosphere and active packaging of chilled foods

177

have been deliberately included in, or on, either the packaging material or the package headspace to enhance the performance of the package system’ (Robertson, 2006). The key words here are ‘deliberately’ and ‘enhance’ and implicit in this definition is that performance of the package system includes maintaining the sensory, safety and quality aspects of the food. Hence, active packaging includes components of packaging systems that are capable of scavenging/absorbing O2, CO2, moisture, ethylene and/or flavour/odour taints; releasing CO2, ethanol, antioxidants and/or other preservatives; and/or maintaining temperature control and/or compensating for temperature changes. Pira International Ltd estimated the global value of the total active packaging market in 2005 to be worth $1.558 billion and has forecasted this market in 2010 to be $2.649 billion (Anon., 2005). Table 8.5 lists examples of active packaging systems, some of which may offer extended shelf-life opportunities for new categories of food products (Day, 2003; Rooney, 1995; Brody, 2005; Robertson, 2006). Active packaging has been used with many food products and is being tested with numerous others. Table 8.5 also lists some of the food applications that have benefited from active packaging technology (Day, 2001). The intention of this section is to provide an overview of O2 scavengers and moisture absorbers and to briefly describe the different types of devices, the scientific principles behind them, the principal food applications and some of the food safety and regulatory issues that need to be considered by potential users. More detailed information on other active packaging technologies can be obtained from some of the references listed. It is recommended that readers who are interested in keeping up to date with the latest developments in this field consult with the fortnightly Active Intelligent Pack News bulletin published by Pira International Ltd, UK.

8.6.1 Oxygen scavengers Oxygen can have considerable detrimental effects on foods. O2 scavengers (also referred to as O2 absorbers) can therefore help maintain food product quality by decreasing food metabolism, reducing oxidative rancidity, inhibiting undesirable oxidation of labile pigments and vitamins, controlling enzymic discoloration and inhibiting the growth of aerobic micro-organisms (Day, 2001; Rooney, 1995, 2005). O2 scavengers are becoming increasingly attractive to food manufacturers and retailers and the growth outlook for the global market is bullish. Pira International Ltd estimated the global O2 scavenger market to be 12 billion units in Japan, 500 million in the USA and 300 million in Western Europe in 2001. This market was forecasted to grow to 14.4 billion in Japan, 4.5 billion in the USA and 5.7 billion in Western Europe in 2007 (Anon., 2004). In addition, Pira International Ltd estimated the global value of this market in 2005 to be $588 million and has forecasted this market to be worth $924 million in 2010 (Anon., 2005). O2 scavengers are the most commercially important sub-category of active packaging for food products and the most well known take the form of small sachets containing various iron-based powders plus an assortment of catalysts. © 2008, Woodhead Publishing Limited

178

Selected examples of active packaging systems, mechanisms and food applications

Active packaging system

Mechanisms

Food applications

Oxygen scavengers

• • • • • •

bread, cakes, cooked rice, biscuits, pizza, pasta, cheese, cured meats and fish, coffee, snack foods, dried foods and beverages

Moisture absorbers

• PVA blanket • activated clays and minerals • silica gel

fish, meats, poultry, snack foods, cereals, dried foods, sandwiches, fruit and vegetables

Carbon dioxide scavengers/emitters

• • • •

coffee, fresh meats and fish, nuts and other snack food products, sponge cakes

Ethylene scavengers

• potassium permanganate • activated carbon • activated clays/zeolites

fruit, vegetables and other horticultural products

Preservative releasers

• • • • • •

cereals, meats, fish, bread, cheese, snack foods, fruit and vegetables

© 2008, Woodhead Publishing Limited

iron based metal/acid nylon MXD6 metal (e.g. platinum) catalyst ascorbate/metallic salts enzyme based

iron oxide/calcium hydroxide ferrous carbonate/metal halide calcium oxide/activated charcoal ascorbate/sodium bicarbonate

organic acids silver zeolite spice and herb extracts BHA/BHT antioxidants vitamin E antioxidant chlorine dioxide/sulphur dioxide

Chilled foods

Table 8.5

• encapsulated ethanol

pizza crusts, cakes, bread, biscuits, fish and bakery products

Flavour/odour absorbers

• • • • •

cellulose triacetate acetylated paper citric acid ferrous salt/ascorbate activated carbon/clays/zeolites

fruit juices, fried snack foods, fish, cereals, poultry, dairy products, fruit

Temperature control packaging

• • • • • • • •

non-woven plastics double-walled containers hydrofluorocarbon gas quicklime/water ammonium nitrate/water calcium chloride/water super corroding alloys/salt water potassium permanganate/glycerine

ready meals, meats, fish, poultry, beverages

Temperature compensating films • Side chain crystallisable polymers

fruit, vegetables and other horticultural products

Modified atmosphere and active packaging of chilled foods

Ethanol emitters

179

© 2008, Woodhead Publishing Limited

180

Chilled foods

These chemical systems often react with water supplied by the food to produce a reactive hydrated metallic reducing agent that scavenges O2 within the food package and irreversibly converts it to a stable oxide. The iron powder is separated from the food by keeping it in a small, highly O2 permeable sachet that is labelled ‘Do not eat’ and includes a diagram illustrating this warning. The main advantage of using such O2 scavengers is that they are capable of reducing O2 levels to less than 0.01%, which is much lower that the typical 0.3–3.0% residual O2 levels achievable by MAP. O2 scavengers can be used alone or in combination with MAP. Their use alone eliminates the need for MAP machinery and can increase packaging speeds. However, it is more common commercially to remove most of the atmospheric O2 by MAP and then use a relatively small and inexpensive scavenger to mop up the residual O2 remaining within the food package (Day, 2003; Robertson, 2006). Non-metallic O2 scavengers have also been developed to alleviate the potential for metallic taints being imparted to food products. (The problem of inadvertently setting off in-line metal detectors is also alleviated even though some modern detectors can now be tuned to phase out the scavenger signal whilst retaining high sensitivity for ferrous and non-ferrous metallic contaminants). Non-metallic scavengers include those that use organic reducing agents, such as ascorbic acid, ascorbate salts or catechol (Day, 2003). O2 scavengers were first marketed in Japan in 1976 by the Mitsubishi Gas Chemical Co. Ltd under the trade name Ageless™. Since then, several other Japanese companies, including Toppan Printing Co. Ltd and Toyo Seikan Kaisha Ltd, have entered the market, but Mitsubishi still dominates the O2 scavenger business in Japan (Rooney, 1995; 2005). O2 scavenger technology has been successful in Japan but the acceptance of O2 scavengers in North America and Europe has been relatively slow, although several manufacturers and distributors of O2 scavengers are now established in both these continents (Rooney, 1995; 2005; Brody, 2005). It should be noted that discrete O2 scavenging sachets suffer from the disadvantage of possible accidental ingestion of the contents by the consumer and this has hampered their commercial success, particularly in North America and Europe. However, the development of O2 scavenging adhesive labels that can be adhered to the inside of packages, and the incorporation of O2 scavenging materials into laminated trays and plastic films, have enhanced and will help the commercial acceptance of this technology. For example, Marks & Spencer Ltd were the first UK retailer to use O2 scavenging adhesive labels for a range of sliced cooked and cured meat and poultry products, which are particularly sensitive to deleterious light and O2-induced colour changes. Other UK retailers, distributors and caterers are using these labels for the above food products as well as for coffee, pizzas, speciality bakery goods and dried food ingredients (Hirst, 1998). Other common food applications for O2 scavenger labels and sachets include cakes, breads, biscuits, croissants, fresh pastas, cured fish, tea, powdered milk, dried egg, spices, herbs, confectionery and snack food (Day, 2001). The use of O2 scavengers for beer, wine and other beverages is potentially a © 2008, Woodhead Publishing Limited

Modified atmosphere and active packaging of chilled foods

181

huge market that has only recently begun to be exploited. Iron-based label and sachet scavengers cannot be used for beverages or high aw foods because, when wet, their O2 scavenging capability is rapidly lost. Instead, various non-metallic reagents and organo-metallic compounds which have an affinity for O2 have been incorporated into bottle closures, crowns and caps, or blended into polymer (usually PET) materials so that O2 is scavenged from the bottle headspace and any ingressing O2 is also scavenged. The PureSeal™ O2 scavenging bottle crowns (produced by W.R. Grace Co. Ltd, USA), O2 scavenging plastic (PET) beer bottles (manufactured by Continental PET Technologies, USA), OS2000® cobalt catalysed O2 scavenger films (produced by Cryovac Sealed Air Corporation, USA) and light activated ZerO2® O2 scavenger materials (developed by Food Science Australia, North Ryde, NSW, Australia) are just four of many O2 scavenger developments aimed at the beverage market but which are also applicable to other food applications (Rooney, 1995; 1998; 2005; Scully and Horsham, 2005). More detailed information on the technical requirements of the different types of O2 scavengers can be obtained from the manufacturers and suppliers, as well as from Rooney (1995; 1998; 2005), Labuza and Breene (1989) and Brody (2005).

8.6.2 Moisture absorbers For many chilled food products, a major cause of food spoilage is excess moisture. Soaking up moisture by using various absorbers or desiccants is very effective at maintaining food quality, and extending shelf-life by inhibiting microbial growth and moisture-related degradation of texture and flavour. Several companies manufacture moisture absorbers in the form of sachets, pads, sheets or blankets. For packaged dried food applications, desiccants such as silica gel, calcium oxide and activated clays and minerals are typically contained within Tyvek™ (Dupont Chemicals, Wilmington, Delaware, USA) tear-resistant permeable plastic sachets. For dual-action purposes, these sachets may also contain activated carbon for odour adsorption or iron powder for O2 scavenging (Rooney, 1995). In addition to moisture absorber sachets for humidity control in packaged dried foods, several companies manufacture moisture drip absorbent pads, sheets and blankets for liquid water control in high aw foods such as meats, fish, poultry, fruit and vegetables. Basically these consist of two layers of a microporous non-woven plastic film between which is placed a superabsorbent polymer which is capable of absorbing up to 500 times its own weight in water. Typical superabsorbent polymers include polyacrylate salts, carboxymethyl cellulose (CMC) and starch copolymers which have a very strong affinity for water (Day, 2003; Anon., 2003; Reynolds, 2007). Moisture drip absorber pads are commonly placed under packaged fresh meats, fish and poultry to absorb unsightly tissue drip exudate. Larger sheets and blankets are used for absorption of melted ice from chilled seafood during air freight transportation or for controlling transpiration of horticultural produce (Rooney, 1995). Commercial moisture absorber sheets, blankets and trays include Toppan Sheet™ (Toppan Printing Co. Ltd, Japan), Thermarite™ © 2008, Woodhead Publishing Limited

182

Chilled foods

(Thermarite Pty Ltd, Australia), Luquasorb™ (BASF, Germany) and Fresh-RPax™ (Maxwell Chase Inc., Douglasville, GA, USA). Another approach to the control of excess moisture in high aw foods is to intercept the moisture in the vapour phase. This approach allows food packers or even householders to decrease the water activity on the surface of foods by reducing in-pack RH. This can be done by placing one or more humectants between two layers of water-permeable plastic film. For example, the Japanese company Showa Denko Co. Ltd has developed Pitchit™ film which consists of a layer of humectant carbohydrate and propylene glycol sandwiched between two layers of polyvinyl alcohol (PVA) plastic film. Pitchit™ film is marketed for home use in a roll or single sheet form for wrapping fresh meats, fish and poultry. After wrapping in this film, the surface of the food is dehydrated by osmotic pressure, resulting in microbial inhibition and shelf-life extension of 3–4 days under chilled storage (Rooney, 1995; Labuza and Breene, 1989). Another example of this approach has been applied in the distribution of horticultural produce. Microporous sachets of desiccant inorganic salts such as sodium chloride have been used for the distribution of tomatoes in the USA (Rooney, 1995). Yet another example is an innovative fibreboard box which functions as a humidity buffer on its own without relying on a desiccant insert. It consists of an integral water vapour barrier on the inner surface of the fibreboard, a paper-like material bonded to the barrier which acts as a wick and an unwettable but highly permeable to water vapour layer next to the fruit or vegetables. This multi-layered box, patented by CSIRO Plant Industries, Australia, is able to take up water in the vapour state when the temperature drops and the RH rises. Conversely, when the temperature rises, the multi-layered box can release water vapour back in response to a lowering of the RH (Day, 1993; Scully and Horsham, 2005). Moisture absorbers are the best selling active packaging technology for all applications but O2 scavengers are commercially more valuable for strictly food applications. Pira International Ltd estimated the global value of the moisture absorber market in 2005 to be $722 million ($454 million for desiccants and $268 for moisture drip pads) and has forecasted this market in 2010 will be worth $1286 million ($823 million for desiccants and $463 for moisture drip pads) (Anon., 2005).

8.7

Vacuum packaging

Vacuum packaging involves the simple evacuation of air from within a pack prior to hermetic sealing. Hence, vacuum packaging does not involve the replacement of the evacuated air with a gas mixture, as is the case with MAP. Notwithstanding, vacuum packaging does reduce the partial pressure of atmospheric gases within the vacuum packs and hence is capable of extending the shelf-life of perishable foods. Vacuum packaging is an established technique for packaging chilled foods such as primal red meats, cured meats and cheese. Similar to MAP, vacuum packaging extends the shelf-life of food by removing O2 (and allowing build-up of CO2 if © 2008, Woodhead Publishing Limited

Modified atmosphere and active packaging of chilled foods

183

there is microbial respiration) and thus inhibits the growth of aerobic spoilage micro-organisms and reduces the rate of oxidative deterioration (Yamaguchi, 1990).

8.7.1 Packaging materials In order to maintain a vacuum around the food, high O2 barrier materials and high levels of seal integrity are required. Although the requisite O2 barrier for vacuum packaging depends on the type of food packaged, O2 transmission rates of less than 15 cm3 m–2 day–1 atm–1 are generally required. Also, packaging materials with low WVTRs must be used. Typical vacuum packaging materials consist of coextruded or laminated films such as OPP/EVOH/PE, PA/PE, PET/PE, OPP/PVDC/PE, OPP/PVDC/OPP and PVC/EVOH/PVC (Yamaguchi, 1990).

8.7.2 Vacuum skin packaging Vacuum skin packaging is a technique which was developed to overcome some of the disadvantages of the traditional vacuum pack and MAP (White, 1990). The vacuum skin packaging concept relies upon a highly ductile plastic barrier laminate which is gently draped over a food product, thereby moulding itself to the actual contours of the product to form a second skin. The product’s natural shape, colour and texture are highlighted and, since no mechanical pressure is applied whilst drawing the vacuum, soft or delicate products are not crushed or deformed. Successes of vacuum skin packaging in the UK market include sliced cooked and cured meats, pâté and fish products (e.g. peppered mackerel). Unlike vacuum packaging, vacuum skin packaging and MAP allow pre-sliced meats to be easily separated after pack opening. In vacuum skin packaging, the appearance of the product is enhanced and the wrinkle-free skin prevents product movement, thereby enabling vertical retail display. Also, since the bottom and top web films are sealed from the edge of the pack to the edge of the product, pack integrity is maximised and juice exudation is limited. Finally, vacuum skin packaging saves space in domestic refrigerators compared with MA packs and is ideally suited for freezing since the second skin prevents formation of ice crystals on the product surface, thereby eliminating freezer burn and dehydration (White, 1990).

8.8

Future trends

The following trends are likely to influence the chilled food packaging industry during future years.

8.8.1 Environmental factors The packaging industry is going to face an increasing burden which has been placed on it by packaging waste legislation throughout most developed nations © 2008, Woodhead Publishing Limited

184

Chilled foods

around the world. The need to meet recycling and recovery targets is resulting in short-term packaging solutions which may not be in the best interest of long-term sustainability. The legislation requires more packaging to be designed for reuse or recycling, yet under certain circumstances these may not be the most environmentally friendly solutions. Furthermore, the additional requirement for packaging minimisation may work against design for recycling since the lightest-weight materials may not be the easiest to recycle. Finally, future packaging waste legislation is likely to increase recycling and recovery targets, with the possibility of reuse targets being introduced (Stirling-Roberts, 1999).

8.8.2 Consumer-driven packaging innovation The increased focus on consumer needs provides the packaging industry with opportunities to innovate. Consumers respond well to added-value packaging innovations that improve the functionality and design of packaging, e.g. easy-toopen and resealing devices, easy-to-pour bottles, tamper-evident features, time/ temperature indicating labels and microwave ‘doneness’ indicators. Packaging is increasingly being viewed as a strategic marketing tool. The retail supply chain is becoming more responsive and consumer-driven, and the effect on the packaging industry is that demand is for increasingly smaller quantities of consistent quality packagings delivered against tight schedules. In response to consumer demand, a wider range of packaging formats is now on offer and this range is likely to expand in the future. For example, fresh chilled soups can be bought in glass bottles, plastic tubs, laminated paper board cartons and flexible pouches, each of which has different technical and marketing advantages. Consumers have also responded well to convenient food packages and this trend will undoubtedly continue in the future, e.g. prepared fruits, vegetables and salads which are ready-to-eat or readyto-heat in microwaveable packaging (Stirling-Roberts, 1999).

8.8.3 New materials and technology Driven by both environmental concerns and economics, new lighter-weight packaging materials are being developed throughout Europe and around the world. Examples include the introduction of superior performance plastics using metallocene catalysts, and micro-flute corrugated paper and board materials. Research and development in the fields of edible and biodegradable packaging continues to expand as well as methods of reducing the cost of packaging recycling. Also, more strategic developments in the areas of barcode tagging, active and intelligent packaging, and digital printing will continue to expand in the future (Stirling-Roberts, 1999; Pugh, 1998; Anon., 1998). Regarding new materials, the invention of advanced catalyst technologies, such as metallocene technology, has enabled the design of new plastic resins, many of which allow for thinner, high-performance packaging materials which can be tailored for specific application requirements. Technical advantages claimed for metallocene plastic resins include improved rigidity, clarity and gloss; excellent © 2008, Woodhead Publishing Limited

Modified atmosphere and active packaging of chilled foods

185

heat seal and hot tack strength and puncture resistance; and light-weight or down gauging opportunities that were not previously available with traditional plastic resins (Pugh, 1998; Anon., 1998).

8.8.4 Modified atmosphere packaging MAP is one of the most exciting and innovative areas of the packaging industry. New developments in both packaging materials/machinery and food product applications seem to be opening up at an increasing rate. The MAP market in Europe is substantial and has enjoyed considerable growth in recent years because of the important benefits it provides to food manufacturers, retailers and consumers alike. The success of MA-packed products in the UK is expected to stimulate future growth in other countries around the globe.

8.8.5 Active packaging Active packaging is a technology developing a new trust because of recent advances in packaging, material science and biotechnology, and new consumer demands (Ahvenainen and Hurme, 1997). However, ultimately active packaging must benefit and be accepted by consumers before it is more widely adopted (Lähteenmäki and Arvola, 2003). Also, active packaging must not be driven by technological possibilities but rather by meeting real market needs (Anon., 2006). O2 scavengers and moisture absorbers are by far the most commercially important sub-categories of active packaging and the market has been growing steadily for the last ten years and is predicted to grow even further by 2010 (Anon., 2005). All other active packaging technologies are also predicted to be used more in the future, particularly ethylene scavengers, CO2 scavengers and emitters, moisture absorbers and temperature control packaging. Food safety and regulatory issues in the EU and USA are likely to restrict the use of certain preservative releasers and flavour/odour absorber active packaging technologies (Vermeiren et al., 1999; Brody, 2005). Nevertheless, the use of active packaging is becoming increasingly popular and many new opportunities will open up for utilising this technology in the future.

8.9

References

AHVENAINEN, R. AND HURME, E. (1997). Active and smart packaging for meeting consumer

demands for quality and safety. Food Additives Contaminants, 14, 753–763. (2006). The FreshlineTM guide to modified atmosphere packaging (MAP). Air Products Plc, Basingstoke, Hants., UK. ANON. (1998). Plastics of the future. Packaging News, March edition, 63. ANON. (2003). European expansion for moisture absorber. Active & Intelligent Pack News, 2(2), 6. ANON. (2004). US invasion – it dominates the US, but Multisorb is pushing its oxygen scavengers in Europe. Active & Intelligent Pack News, 2(11), 5. AIR PRODUCTS PLC

© 2008, Woodhead Publishing Limited

186

Chilled foods

ANON.

(2005). Up and active – Pira’s latest market report plots a healthy future for active packaging. Active & Intelligent Pack News, 3(25), 5. ANON. (2006). Active encouragement – what needs to be done to further encourage the uptake of active packaging? Active & Intelligent Pack News, 4(13), 5. BETTS, G.D. (ed.) (1996). Code of Practice for the Manufacture of Vacuum and Modified Atmosphere Packaged Chilled Foods with Particular Regards to the Risks of Botulism. Guideline No. 11, Campden and Chorleywood Food Research Association, Chipping Campden, Glos., UK. BLAKISTONE, B.A. (ed.) (1998). Principles and Applications of Modified Atmosphere Packaging of Foods. 2nd edition. Blackie Academic and Professional, London, UK. BOWS, J.R. AND RICHARDSON, P.S. (1990). Effect of component configuration and packaging material on microwave reheating of a frozen three component meal. International Journal of Food Science and Technology, 25, 538–50. BRODY, A.L. (2005). Commercial uses of active food packaging and modified atmosphere packaging systems. In: Innovations in Food Packaging (ed. Han, J.H.), Elsevier Ltd, London, UK, 457–474. BRODY, A.L. AND THALER, M.C. (1996). Argon and other noble gases to enhance modified atmosphere food processing and packaging. Proceedings of IoPP conference on ‘Advanced technology of packaging’. Chicago, Illinois, USA, 17 November. DAY, B.P.F. (1992). Guidelines for the good manufacturing and handling of MA packed food products. Technical Manual No. 34. Campden and Chorleywood Food Research Association, Chipping Campden, Glos., UK. DAY, B.P.F. (1993). Fruit and vegetables In: Principles and Applications of MAP of Foods (ed. Parry R.T.), Blackie Academic and Professional, New York, USA, 114–133. DAY, B.P.F. (1998). Novel MAP – a brand new approach. Food Manufacture, 73(11), 22–24. DAY, B.P.F. (2001). Active packaging – a fresh approach. brand© – The Journal of Brand Technology, 1(1), 32–41. DAY, B.P.F. (2003). Active packaging. In: Food Packaging Technologies (eds. Coles, R., McDowell, D. and Kirwan, M.), CRC Press, Boca Raton, FL, USA, 282–302. DAY, B.P.F. AND WIKTOROWICZ, R. (1999). MAP goes on-line. Food Manufacture, 74(6), 40– 41. FLORES, J.D. AND MATSOS, K.I. (2005). Introduction to modified atmosphere packaging. In: Innovations in Food Packaging (ed. Han, J.H.), Elsevier Ltd, London, UK, 159–172. FOIL CONTAINER BUREAU (1991). Foil in microwave ovens. Packaging Magazine, 62(684), 24. HIRST, J. (1998). Personal Communication. EMCO Packaging Systems Ltd, Worth, Kent, UK. KADER, A.A., ZAGORY, D. AND KERBEL, E.L. (1989). Modified atmosphere packaging of fruits and vegetables. Critical Reviews in Food Science and Nutrition, 28(1), 1–30. LABUZA, T.P. AND BREENE, W.M. (1989). Applications of active packaging for improvement of shelf-life and nutritional quality of fresh and extended shelf-life foods. Journal of Food Processing and Preservation, 13, 1–69. LÄHTEENMÄKI, L. AND ARVOLA, A. (2003). Testing consumer responses to new packaging concepts. In: Novel Food Packaging Techniques (ed. Ahvenainen, R.), CRC Press, Boca Raton, FL, USA, 550–562. PUGH, M. (1998). A catalyst for change. Packaging Magazine, March 12 edition, 30–31. REYNOLDS, G. (2007). Superabsorbent soaks up packaging problems. www.foodproduction daily-usa.com, 22 January. ROBERTSON, G.L. (ed.) (2006). Food Packaging – Principles and Practice. 2nd edition. CRC Press, Boca Raton, FL, USA. ROONEY, M.L. (ed.) (1995). Active Food Packaging. Chapman and Hall, London, UK. ROONEY, M.L. (1998). Oxygen scavenging plastics for retention of food quality. In: Proceedings of Conference on ‘Advances in Plastics – Materials and Processing Technology for Packaging’ Pira International, Leatherhead, Surrey, UK, 25 February. © 2008, Woodhead Publishing Limited

Modified atmosphere and active packaging of chilled foods ROONEY, M.L.

187

(2005). Introduction to active food packaging technologies. In: Innovations in Food Packaging (ed. Han, J.H.), Elsevier Ltd, London, UK, 63–69. SCULLY, A. AND HORSHAM, M. (2005). Emerging packaging technologies for enhanced food preservation. Food Science and Technology, 20(2), 16–19. SØRHEIM, O., AUNE, T. AND NESBAKKEN, T. (1997). Technological, hygienic and toxicological aspects of carbon monoxide used in modified-atmosphere packaging. Trends in Food Science and Technology, 8, 307–312. SPENCER, K.C. (2005). Modified atmosphere packaging of ready-to-eat foods. In: Innovations in Food Packaging (ed. Han, J.H.), Elsevier Ltd, London, UK, 185–203. STERLING-ROBERTS, A. (1999) Where to next? Packaging News, Dec. edition, 8–9. TURTLE, B.I. (1988) Cost effective food packaging, World Packaging Directory, Cornhill Publications Ltd, London, UK, 67–72. VERMEIREN, L., DEVLIEGHERE, F., VAN BEEST, M., DEKRUIJF, N. AND DEBEVERE, J. (1999). Developments in the active packaging of foods. Trends in Food Science and Technology, 10, 77–86. WHITE, R. (1990) Vacuum skin packaging: a total packaging system. In: Flexible Packaging for Food Products Conference (ed. Monbiot, R.), IBC Technical Services Ltd, Byfleet, Surrey, UK. YAMAGUCHI, N. (1990). Vacuum packaging. In: Food Packaging (ed. Kadoya, T.), Academic Press, London, UK, 279–92.

© 2008, Woodhead Publishing Limited

Part II Technologies and processes in the supply chain

© 2008, Woodhead Publishing Limited

9 Microbiological hazards and safe design M. Brown, mhb Consulting, UK

9.1

Introduction

This section is an introduction to designing chilled food products and the systems that support their supply chain, with an emphasis on microbiology. The design should take the original marketing concept and balance safety, quality and cost risks to come up with a product and process design that is safe, meets customer and consumer expectations and can be manufactured at competitive cost. A product design should cover:

• • • • • •

ingredient quality, and especially maximum anticipated contamination levels product formulation and characteristics process design to be used pack type shelf-life intended use of product.

A process design should start with what marketing would like, and end up with what it is feasible for the supply chain to do. The final design, translated into specifications, should cover:

• type of process (e.g. in-pack pasteurised or heat treated and then assembled and packed)

• compatibility of ingredients with process flows • necessary levels of hygiene in manufacturing and storage facilities • unit operations and types of equipment to be used © 2008, Woodhead Publishing Limited

192

• • • • • • •

Chilled foods

target process and storage conditions dosing or filling, packaging and pack handling equipment envisaged product volumes and line capacity capacity for variability in customer demand tolerable costs controls and monitoring of key performance indicators and product attributes requirements for a logistics chain.

The designs should lead to the selection of suppliers, specification of the process and of manufacturing hygiene standards, selection of manufacturing location and equipment. An underlying design principle in Europe is that chilled storage and limitation of shelf-life can be relied on to prevent the growth of any pathogens (especially spore-formers and Listeria monocytogenes) present in the product after processing (see EC No 2073/2005 on microbiological criteria for foodstuffs). In other parts of the world there is a requirement that intrinsic preservation systems (e.g. reduced pH, Aw or chemical preservatives) provide an additional margin of safety. Details of UK hygienic standards and practices and recommended minimum heat treatments, etc. which should be reflected in designs are given in the Chilled Food Association guidelines (CFA, 2006). Many different ingredients and raw materials are processed to make chilled foods. At harvest or slaughter they may have a wide range of microbes in or on them. The design of product components must ensure that these can be reduced to, and remain at, safe levels until the product is consumed. Some materials carry the micro-organisms that cause their eventual spoilage (e.g. bacilli or lactic acid bacteria) whilst others pick them up during harvesting or processing. Many food poisoning bacteria occur naturally with farm animals and agricultural produce (e.g. Salmonella, Escherichia coli O157 and Campylobacter) and hence can contaminate meat and poultry, milk and vegetable products. The numbers and types present will vary from one ingredient and one source to another, and often product safety at the point of consumption will depend on the presence or numbers of pathogens in the raw material, manufacturing (especially heat processing and hygiene) and consumer use. In order to ensure safe products with a reliable shelflife, the manufacturer must identify which food poisoning and spoilage bacteria are likely to be associated with particular raw materials and products (e.g. by microbiological surveys) and the risks they present. Therefore it is essential to design sourcing and food processing procedures to ensure that food poisoning risks are minimised. This is especially important in the prepared food and cook– chill sectors (especially ready-to-eat), where safety relies on the control of many features of the manufacturing process (ICMSF, 1988; Kennedy, 1997). The appropriate means of control should be specified at the product and process design stage and be implemented in the manufacturing operation as specifications and procedures, often including several stages of the supply chain. The prevention of product re-contamination or cross-contamination after factory heating plays an even more critical role when products are sold as ready-to-eat (see Griffith, 2002) and this will introduce the need for additional controls during © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

193

processing. Specifications and process-stage outcomes (e.g. limits) must be clear. For example, Gill et al. (1997) have suggested that the overall hygienic quality of beef hamburger patties could be improved only if hygienic quality beef (i.e. lowest possible levels of contamination with pathogens) was used for manufacture and there was better management of retail outlets with regard to patty storage and cooking. Good manufacturing practice guides are available for many sectors of the chilled food industry, e.g. IFST Food and drink good manufacturing practice (IFST, 2007); UK CFA guidelines (CFA, 2006); Guidelines on good manufacturing practice for heat processed flexible packaging (CCFRA, 2006). These guides outline design principles and responsibilities in relation to the manufacture of safe products; compliance will ensure that products remain wholesome and safe under the expected conditions of use. In practice, product and process design will always be a compromise between the demands for safety and quality on the one hand, and cost and operational limitations on the other. The skill of the product designer is to balance the competing demands of microbiology and quality to find an acceptable balance. To date, the chilled food industry has been successful in achieving this balance. Heat is the main means of giving products their character, ensuring product safety and eliminating spoilage bacteria. The extent of heating will very often be limited by the need to retain product character, especially colour and texture. Usually, cooking processes (e.g. boiling) will incidentally exceed the minimum heat treatments needed for microbiological purposes, either in-factory or in-home, but minimum processes should always be designed to eliminate specific bacteria (e.g. infectious pathogens, cold-growing spores, or types causing spoilage within the product shelf-life). Good control of heat processing and hygiene in the factory and the home or food service outlet are essential for product safety. Usually, more than one process step contributes to quality and safety, and refrigerated storage is used to retard or prevent the growth of micro-organisms that have survived factory heating or have been added as contaminants. Hence, the safety of chilled foods without inherent preservative properties depends almost exclusively on suitable refrigeration temperatures (0–8 °C) being maintained throughout the supply chain, including, for example, storage and loading of refrigerated vehicles. Where preservatives are used (reduced pH, increased acidity, vacuum packing or combinations of these), chilling will usually increase their effectiveness. The techniques of risk assessment (WHO, 2003; Brown and Stringer, 2002) and hazard analysis (see http://www.rbkc.gov.uk/EnvironmentalServices/foodhygiene andstandards/business6.asp and http://www.rbkc.gov.uk/Environmental Services/ foodhygieneandstandards/fhsha.pdf), either formal or more commonly informal, may be used to guide product designers in achieving safe product designs with an acceptable balance between the sale of products containing raw or non-decontaminated components, the need for heat treatment or decontamination either in the factory or by consumers, and the chances and hazards of pathogen presence in the product at consumption (see http://www.foodrisk.org/index.htm ). Software tools are now available to support risk assessments (see http://www.foodrisk.org/ © 2008, Woodhead Publishing Limited

194

Chilled foods

modeling_simulation_tool.cfm). Higher risks of product contamination with pathogens should always prompt product designers to include effective heat processing or decontamination steps. Successful process designs must consider contaminants likely to be carried by the raw materials, and also the effects of shelf-life and anticipated storage conditions during distribution, retailing or consumer use (CFDRA, 1990) on the eventual numbers in the product (e.g. Listeria monocytogenes). Although the chill chain is an integral part of safety, additional risks may come from consumer mishandling or mis-use during storage or preparation. These should be accommodated within designs (e.g. specifications to minimise contamination levels) where safety and high quality shelf-life rely substantially on correct handling and use after manufacture (e.g. chilled storage and/or cooking). Brackett (1992) has pointed out that many chilled foods contain few, or no, antimicrobial additives to prevent growth of pathogenic micro-organisms and are susceptible to the effects of inadequate refrigeration that may allow pathogen growth. He has also highlighted related issues such as over reliance on shelf-life as a measure of quality and the need to consider the needs of sensitive groups (such as immune-compromised consumers) in the product design. Where the product design relies on consumer heating to free the product of pathogens such as salmonellae, it is important that helpful, accurate and validated cooking instructions are provided by the manufacturer and that use of these instructions results in high product quality. Where products are likely to contain pathogens, handling and storage instructions should indicate how to prevent cross-contamination in the home (Humphrey et al., 2001) or in food service. As global sourcing of raw materials increases, existing risk assessments and hazard analyses will need regular review as new microbiological hazards may be introduced with familiar materials. For example, as sourcing of prawns shifts from the north Atlantic to farmed fresh-water tropical prawns, new pathogens (such as pathogenic vibrios and listeria; Food Standards Australia New Zealand, 2002; Reilly and Twiddy, 1991) may be carried, and therefore suitable treatments must be incorporated into product designs and usage instructions. Chilled foods must be stored at or below specified temperature(s), during manufacture, distribution (including transfers between stages) and storage; these range from –1 ° to +8 °C. Storage at higher temperatures can allow the growth of hazardous micro-organisms and shorten shelf-life. Poorly controlled processing in conjunction with temperature or time abuse during storage will certainly lead to the growth of spoilage micro-organisms and premature loss of quality. Labuza and Bin-Fu (1995) have proposed the use of time/temperature integrators (TTI), e.g. 3M™ MonitorMark™ time/temperature indicator for monitoring temperatures in the distribution chain (see http://www.iifiir.org/en/doc/1009.pdf). However, the data produced by these integrators located on the pack surface may not always be a true reflection of the temperatures within the food, as they indicate temperature exposure, not product temperature or quality. Where good product temperature data exist, they can be used with predictive microbial models to assess the impact of storage conditions on the shelf-life and safety of products. The risks associated © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

195

with any particular products can be investigated at the design stage either by practical trials (such as transit trials and challenge testing) or by the use of mathematical modelling. Heat treatments to optimise quality can be designed using microbial lethality models that interchange time and different temperatures to calculate process lethality (based on D and z values) and, similarly, microbial growth models can be used to predict shelf-life (based on time to a certain level of microbes and possibly the effects of intrinsic preservation factors) and optimise supply chain management. There is growing use of predictive models and other information systems in food microbiology as design tools to assist food safety management by improving access to validated information on hazards, growth kinetics and limits (McMeekin et al., 2006). Existing computer-based predictive microbiology databases from the UK (Food micromodel, FMM) and USA (Pathogen Modeling Program, PMP) applicable to chilled products can be found within the ComBase portal (http:// wyndmoor.arserrc.gov/combase/) which includes the PMP (http://ars.usda.gov/ Services/docs.htm?docid=6786 ). Panisello and Quantick (1998) used FMM to make predictions on the growth of pathogens in response to variations in the product pH and salt content, specifically examining the effect of lowering the pH of paté. Zwietering and Hasting (1997) have taken this concept a stage further and developed a modelling approach to predict the effects of processing on microbial growth during food production, storage and distribution. Their process models were based on mass and energy balances together with simple microbial growth and death kinetics based on meat product and burger processing lines. Used in this way, models can predict the contribution of each individual process stage to the microbial level in a product. Zwietering et al. (1991; 1994a,b) have modelled the impact of temperature and time, and shifts in temperature, during processing on the growth of Lactobacillus plantarum. These predictive models can, in principle, be used for suggesting the temperature conditions needed to control microbial growth during distribution, or indicate the impact of microbial ‘lag’ on safe shelf-life, where temperature fluctuations may be common and could allow growth. Impe et al. (1992) have also built similar models describing the behaviour of bacterial populations during processing in terms of both time and temperature, but have extended their models to cover inactivation at temperatures above the maximum temperature for growth. Adair and Briggs (1993) have proposed the development of expert systems, and Brown et al. (1998) have developed a simple automated system, based on predictive models, to assess the microbiological safety of chilled foods. More complex expert systems could be used to interpret microbiological, processing, formulation and usage data and predict the microbiological safety of a particular design. Betts (1997) has also discussed the practical application of microbial growth models for the determination of shelf-life of chilled foods and has pointed out the usefulness of models in speeding up product development and the importance of validating the output of models in real products. Modelling offers advantages in terms of time and cost, but is still in its infancy (Pin and Baranyi, 1998). Now that ComBase offers a combined database of microbial responses to food environments, supple© 2008, Woodhead Publishing Limited

196

Chilled foods

mented by predictive packages (PMP, ComBase Predictor, Perfringens Predictor, DMFit) it can be used by designers to assess the benefits and limitations of available predictive models. Use of a unified approach based on standardised data sources offers great potential for assuring the safety of product designs and foods in international trade formulated by different groups in different places. It should also prevent the unnecessary repetition of experiments. However, usefulness will be limited by variation in the resistance of the microbial types present in raw material and products, and also in their activities and subtle materials interactions altering growth or survival rates or the production of metabolites recognised by customers as spoilage. There are, not surprisingly, major differences between manufacturers in the degree of time or temperature abuse they design products to withstand and hence the risks they are prepared to accept on behalf of their customers. This can result in major differences in the processes, ingredients and packaging used and the shelf-lives given to apparently similar products.

9.2

Definitions

Definitions are given below to introduce general comments and guidance for the design of supply chains which effectively control microbiological risks. They are separated into the following groups: raw materials; processes, products, and safety and quality control.

9.3

Raw materials

Designs should consider the level of contamination and the types of pathogen or toxin (and spoilage micro-organisms) likely to be present in the raw materials or introduced during processing. All treatments and processes have a maximum capability to destroy these micro-organisms, but usually not toxins. The design should identify if there is a possibility that the maximum realistic or specified levels in an ingredient will exceed this capability. In this case, additional or more severe decontamination treatments or tighter specifications should be incorporated into the design. Data for this discussion on realistic levels can be derived from previous hazard analyses on the same, or similar, materials, microbiological analyses or from generic information on sources and suppliers.

9.3.1 Undecontaminated materials Components that have not been decontaminated to make them free of hazardous bacteria include non-heat-treated materials and heated ones, where heating achieves less than 70 °C for 2 minutes. Therefore any pathogens and spoilage micro-organisms present are likely to end up in the product. If such materials are used, the pathogens carried must be identified and the severity of the risks © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

197

assessed. Accordingly, the producer’s packaging, with supply chain intake and delivery procedures, should be designed to prevent their numbers increasing and ensure that they cannot cross-contaminate other components; for example, those that have already been decontaminated. This may lead to the development of additional controls or critical control points (CCPs), and the layout of storage areas should segregate them. Except with appropriate hygiene controls and separation, uncooked materials should not be handled by personnel who handle finished product, or allowed to enter high care or high risk areas (see below). Failure of controls can lead to the accidental manufacture of products potentially harmful to customers (see below).

9.3.2 Decontaminated materials These materials will usually have been heat treated to reduce their microbial load. If they are intended for direct incorporation into ready-to-eat products, then the heating should meet recognised minimum values. These values are heating to 70 °C for 2 minutes (or equivalent) to cause a substantial reduction (nominally a 6 log reduction) in the numbers of vegetative pathogens (e.g. Listeria monocytogenes, Staphylococcus aureus, salmonellae, pathogenic Escherichia coli) and spoilage micro-organisms present. However, spores and pre-formed toxins may remain. More severe heating to 90 °C for 10 minutes (or equivalent) will additionally reduce numbers of spores of psychrotrophic (cold growing) Clostridium botulinum, again by a nominal 6 log reduction. Which heat treatment is specified will depend on whether the product has a short (<10 days) or long (>10 days) shelf-life, and culinary needs. Suitable precautions must be taken to prevent recontamination during chilling, handling or packaging after any heat treatment. Hence ingredients, or part-processed material designed for hygienic assembly, should be kept in secure primary packaging, which is only removed in a high-hygiene area.

9.4

Chilled foods

Chilled foods rely mainly or completely on chilled storage (originally defined as from –1 ° to +8 °C; Anon., 1982, but see below) for stability (see Table 9.1). They may include foods made from raw, uncooked or heat-treated ingredients. Some may require cooking prior to consumption in order to make them edible and safe, e.g. raw fish and meat products, as it is accepted that such foods may unavoidably contain pathogenic micro-organisms from time to time. Depending on their heat treatment, the remaining microbiological hazards (including post-process contamination) and product usage, minimum levels of hygiene are required in manufacturing and assembly or packing areas (CFA, 2006).

9.4.1 Prepared or ready-to-eat chilled foods Prepared chilled foods contain raw or uncooked ingredients, or mixtures (risk © 2008, Woodhead Publishing Limited

198

Chilled foods

Table 9.1

Microbiological risk classes 1–4

Risk Typical classa shelf-life

Critical hazard

Relative risk

Required minimum heat treatment

Required manufacturing areab MA HA HCA

1

1 week

Infectious pathogens

High

Customer cook (minimum 70 °C, 2 minutes)

ü

(ü)

2

1–2 weeks

Infectious pathogens

Low

Pasteurisation by manufacturer (minimum 70 °C, 2 minutes)

ü

ü

3

<10 days

Infectious pathogens and sporeformers

Low

Pasteurisation by manufacturer (minimum 90 °C, 10 minutes)

ü

4

>10 days

Sporeformers

Low

Pasteurisation by manufacturer (minimum 90 °C, 10 minutes)

ü

ü

ü

ü

a

Class 1: raw chill-stable foods, e.g. meat, fish, etc.; class 2: products made from a mixture of cooked and low-risk raw components; class 3: products cooked or baked and assembled or pirmary packaged in a high-care area; class 4; products cooked in-pack. b MA: manufacturing area; HA: hygienic area; HCA: high-care area.

classes 1 and 2, see Section 9.9 and Table 9.1), e.g. salad vegetable, herb or cheese components. Preparation by the manufacturer means that they are either obviously ready-to-eat (e.g. salads or desserts) or only require re-heating (e.g. prepared meals), rather than full cooking, prior to use. These foods should be designed to be free of hazardous pathogens or hazardous levels of pathogens at the end of their shelf-life, and ingredients should be sourced and processes designed with this objective in view (see EC No 2073/2005) for regulatory criteria. A scheme for the layout of process lines used in their manufacture is given in (Figures 9.1 and 9.2). These foods should be handled in areas designated as either high care (a physically segregated area designed and operated to a high standard of hygiene and used only for the processing of washed or treated materials) or dedicated low risk (meeting good manufacturing practice (GMP) requirements as defined by Codex Alimentarius: (1999)), under the CFA scheme.

9.4.2 Cooked ready-to-eat foods These foods (risk classes 3 and 4, see Section 9.9 and Table 9.1) are made entirely from cooked ingredients and therefore should be freed of infectious pathogens during processing. Factory cooking procedures should be designed to ensure © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

199

Fig. 9.1 Typical flow diagram for the production of chilled foods prepared from only raw components (class 1).

this, and handling procedures after cooking, including cooling, should be designed to prevent recontamination of the product or its components. Often the appearance of such foods makes it obvious to the customer that no heating, or mild re-heat, is all that is required before eating. Heating requirements should be made clear by any instructions. Typical process line layouts are shown in Figures 9.3 and 9.4. 9.4.3 Extended shelf-life foods Refrigerated pasteurised foods of extended durability (REPFEDs) include a wide range of longer shelf-life (42 days or so) in-pack pasteurised foods (Mossel et al., 1987; Notermans et al., 1990) and include sous-vide products and other foods © 2008, Woodhead Publishing Limited

200

Chilled foods

Fig. 9.2

Typical flow diagram for the production of chilled foods prepared from both cooked and raw components (class 2).

pasteurised (90 °C × 10 minutes) or containing a specified preservation system throughout all components (e.g. salt 3.5% (s/w), maximum pH 5.0 or Aw 0.97) or a combination of these factors effective against Clostridium botulinum under chill conditions. To realise the full potential of their formulation and heat treatment, handling and packaging must specifically ensure that contamination with infectious pathogens, and the spore-forming pathogens capable of growing under chilled conditions after heating, is prevented. In some cases, injury of spores by © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

Fig. 9.3

201

Typical flow diagram for the production of pre-cooked chilled meals from cooked components (class 3).

heating may be part of the preservation system. There is still a lack of knowledge on realistic safety boundaries and the risks associated with these products with respect to the most severe hazard, non-proteolytic Clostridium botulinum (Peck, 1997). The determinants and interactions determining the effectiveness of complex combination preservation systems using mild heating, chemical preservatives (such as nitrite, lowered pH and reduced water activity) and chilled storage are not fully understood, although variability may be modelled (Membre et al., 2006). © 2008, Woodhead Publishing Limited

202

Chilled foods

Fig. 9.4

9.5

Typical flow diagram for the production of pre-cooked chilled meals cooked in their own packaging prior to distribution (class 4).

Safety and quality control

9.5.1 Good manufacturing practice GMP covers the principles, procedures and means needed to design and manage the production of safe chilled foods of acceptable quality; good hygienic practice (GHP) describes basic hygiene precautions. Governments (see Anon. 1984, 1989), © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

203

the Codex Alimentarius Committee on Food Hygiene (FAO/WHO) and the food industry, often acting in collaboration with food inspection and control authorities and other groups, have developed GMP and GHP requirements (Jouve et al., 1998). More frequently, hazard analysis critical control point plans have to be based on prerequisites covering hygiene and processing requirements (Wallace and Williams, 2001), which may be formally stated, as in a Canadian food safety enhancement program (see http://www.inspection.gc.ca/english/fssa/polstrat/ haccp/manue/ch2sec1-2e.shtml#2.2) and cover specific aspects of plant and building design and layout, equipment and building maintenance and system calibration, supplier approval and performance monitoring for designated technologies. In principle, they are similar to GMP requirements, but are more demanding for particular technologies to ensure safety (e.g. hygiene levels and systems for high risk and high care areas for chilled food production). Their accurate definition is essential for HACCP and quality assurance (QA) systems based on process data to work successfully. Their identification may reduce the number of CCPs in a HACCP plan because common requirements are covered routinely. GMP codes and the hygiene requirements may sometimes contain important limits for the hygienic manufacture of foods (e.g. pasteurisation or cooling conditions) and these should be applied where relevant. Generally GHP/GMP requirements cover the following: • the hygienic design and construction of food manufacturing premises

• the hygienic design, construction and proper use of machinery • cleaning and disinfection procedures (including pest control) • general hygienic and safety practices in food processing, including: • the microbiological quality of raw materials • the hygienic operation of each process step • the hygiene of personnel and their training in hygiene and the safety of food. A design HACCP should examine, at an early stage in development, material flow, process stages and equipment capabilities against the raw material microbiological inputs and product requirements, to provide the basis for the future HACCP system. It should identify hazards important to the product and thought to be associated with the raw materials, processes and product usage. The design study will not be an end-point, as the effectiveness of the proposed controls, etc. cannot be validated. It should suggest controls for various points in the supply chain; their control limits will be theoretical, but will indicate where the ‘real’ HACCP study should establish controls, limits and monitoring procedures. The design HACCP should cover the whole supply chain, including sourcing, making and delivery of products, and propose the management of hygiene and processing, so that clear requirements for plant, product and process are provided before equipment is ordered or production started. The operational HACCP plan itself must be reviewed when any change is made to the product or process design, equipment, raw material specifications or sources, processing or packaging. © 2008, Woodhead Publishing Limited

204

Chilled foods

The widespread availability of web-based technology allows standard data sets for materials and hazards, and growth or survival kinetics to be used in design HACCP studies (see (http://wyndmoor.arserrc.gov/combase/). In the same way, it has allowed the potential for real time exchange of QA data to shift product approval from end-product inspection and testing to monitoring and use of data from process-based controls at all stages of production with focuses on CCPs, and increases the importance of basing process conditions on clear designs, because an implemented HACCP plan involves:

• identification of realistic (microbiological) hazards and the preservation conditions determining their fate (e.g. elimination, presence, growth or survival)

• identification of specific conditions, including limits and acceptable variability, • •

for the control of hazards and identification of process stage(s) where this is achieved and their noting as CCPs or prerequisites specification of procedures and equipment to monitor and record the performance of HACCP and prerequisite controls specification of limits and the actions required when these are exceeded.

The identification and analysis of hazards at the design stage will provide information to interpret GMP/GHP requirements and then direct training, calibration, etc. for the specific product or process. The Microbiology and Food Safety Committee of the National Food Processors Association (NFPA, 1993) has considered HACCP systems for chilled foods produced at a central location and distributed chilled to retail establishments. They use chicken salad as a model for proposing CCPs and give practical advice on HACCP planning; development of a production flow diagram, hazard identification, establishing critical limits, monitoring requirements; and verification procedures to ensure the HACCP system is working effectively. There are also USDA recommendations and outline HACCP flow diagrams for chilled food processes, cook-in-package and cook-then-package (Snyder, 1992).

9.5.2 Risk and hazard analysis Food producers have always assessed the risks of the products they make using either empirical or experiential approaches; realistic hazards, and the means and reliability of their control (risk assessment), need to be identified at the design stage. By the end of this stage, the information available should allow the following questions to be answered – What should the product be like? What can go wrong with it? How likely is this to happen? What are possible causes? This should lead to a last question – What are the consequences of things going wrong? Later on, when the product has been made for some time, the validity of the original design and design HACCP can be judged. The ability of development and QA departments to easily assess the significance of any changes, such as range extension, supply chain change or improvement, use of different raw material sources or the targeting of new customer groups such as children, will depend on the accuracy of recording these stages. © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

205

As causal links have become established between food-borne illness and the presence (infection) or activities (toxigenisis) of particular food-poisoning microorganisms, so controls have been identified and progressively targeted at specific hazards. Such practical approaches have now developed into formal systems with well defined procedures known as microbiological risk assessments (MRAs). MRAs are described in Microbiological Risk Assessment: An interim report (ACDP, 1996) or by the Codex scheme (Figure 1 in Codex Alimentarius, 1996). Their overall aim is to reduce risk by using a three-stage process: risk assessment – to identify realistic microbiological hazards and characterise them according to severity, and examine the impact of raw material contamination, processing and use on the level of risk; risk management – to fix the process, product and storage criteria to mitigate the risk (see Buchanan, 1995); and lastly risk communication – to convey the information output, decisions taken and the level of risk to the consumer or commercial users. When risk assessment is put together with risk communication and is used to promote sound risk management (Brown et al., 1998), a completed risk analysis is produced (ACDP, 1996).

9.5.3 Stages in a risk assessment Risk assessments are often reactive, being triggered by a change in a process or ingredient, the emergence of a new pathogen, a problem, or a change in public concern. Risk assessments can lead to the review of factory controls, layout, sourcing options, or the revision of use instructions. Clear formulation of a problem (what can go wrong?) is an essential prelude to doing a risk assessment and so the first stage is to identify the new hazard (hazard identification). For example, there may be increased concern over the presence of salmonella in a product and its potential to cause harm, depending on its virulence, incidence and concentration, prevalence in raw materials and the dose-response of consumers. In the first stage of risk assessment, the hazard is identified (hazard identification). Exposure assessment then describes the likely exposure of consumers to it, based on the size of the portion consumed and the impact of prior manufacturing on its level and consumer use on the quantity and infectivity of the agent (i.e. Salmonellae) in the product. For a cooked, ready-to-eat product, exposure will depend on the numbers of salmonella entering the factory heating process, the heating of the product, and hygiene in-factory; these combine to determine the number of pathogens surviving at consumption. If the numbers and heat sensitivity of salmonella and the product’s heat treatment are known, then the chances of survival can be estimated. For many raw chilled foods (e.g. burgers or flash-fried poultry products), microbiological safety cannot be guaranteed by manufacturing processes, but is a joint responsibility with the customer (Notermans et al., 1996). This underlines that the consumer is part of the process for ensuring that the end product is safe, and therefore an assessment of their habits and attitudes towards using the product is an essential part of the risk assessment and should be taken account of by the design. Quantification of the risks of infection of the consumer after product consump© 2008, Woodhead Publishing Limited

206

Chilled foods

tion is known as hazard characterisation. It links sensitivity to infection (i.e. usually making use of expert opinion or knowledge of the dose-response within populations) to the concentration of the agent in the portion. The output of these three stages is a risk characterisation, which describes the risks of Salmonella infection associated for a certain consumer after consumption of a product, sourced and manufactured under the specified conditions. To make effective use of the output of a risk assessment, information and decisions must be accessible to management, staff and suppliers. If, as the result of a study, changes are proposed (e.g. changes to use instructions), then these must be clearly communicated to consumers. Where there is good communication, risk management practices can be readily implemented, consistent standards applied and dangerous changes stopped. Risk assessment has been reviewed (Jaykus, 1996) and applied, to specific problems, listeriosis (Miller et al., 1997), the role of indicators (Rutherford et al., 1995) and links with HACCP (Elliot, 1996). 9.5.4 The precautionary principle (see Regulation (EC) No 178/2002) Generally, the safety aspects of product and process designs are based on sound science. However, from time to time decisions have to be taken when good scientific data are not available; for example, if the prevalence, resistance or severity of a new pathogen is unknown. The precautionary principle has been used to ensure health protection and promote innovation and fair trade in the EU. When there is an unknown or poorly characterised hazard, product designers should use this principle to determine what actions are needed in the context of a lack of knowledge. Their decisions should control any perceived health risks without resorting to excessively restrictive control measures (e.g. very severe processes) and should be proportionate to the severity of the food safety problem. Resulting actions need to be derived in an understandable and justifiable way with any assumptions and uncertainties made clear. Most of all, the reduction in risk must be acceptable to all the parties involved and in case of doubt should be discussed with food safety authorities. For example, various measures have been taken to eliminate E. coli O157: H7 from vegetables (by pasteurisation) and salad crops (by disinfection during washing) or through the application of good agricultural practices (De Roever, 1998), because prevalence of this pathogen is unknown and the illness caused can be severe. Any decisions on suitable controls should be used as a basis for designs with caution and regarded as temporary, awaiting further information that will allow a more reliable risk assessment and specification of appropriate control measures.

9.6

Processes

9.6.1 Cooking Cooking processes should be designed to provide sufficient heat to cause all parts of a food to reach the required sensory quality by combinations of temperature and time, and incidentally cause a significant reduction in the numbers of any micro© 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

207

organisms that may be present. Some cooking stages may deliver considerably more heat than is needed for pasteurisation (for example, those involving prolonged periods of boiling or roasting) and the required sensory effects should be noted at the design stage. If the ‘quality’ process exceeds the microbiological requirements, then a good product is likely to result; if the converse is true, then some redesign may be needed to optimise quality. If the cooking is used for pasteurisation, then subsequent (re)contamination must be prevented. It is important that designs for cooking distinguish between heat treatments (i.e. the conditions within a food) and the process conditions needed to provide the heat treatment. For example, for a given heat treatment, process conditions will vary according to the equipment used and the heating characteristics of the product (e.g. diffusion of heat through the product, its dimensions and the transfer of heat to or through its surface).

9.6.2 Pasteurisation Pasteurisation reduces the numbers of specified micro-organisms in a food or ingredient. A 6-log reduction in numbers of infectious pathogens is achieved by 70 °C for 2 minutes, and is usually considered to be a diligent minimum target heat treatment for cooking short shelf-life products (see Section 9.10.2), although for longer shelf-life products, 90 °C for 10 minutes is a suitable design target. In practice, heating is often more severe than the minimum for safety, being targeted at the more heat-resistant bacteria causing spoilage in the product (see Gaze and Betts, 1992; CFDRA, 1992). The effectiveness of a particular heat treatment may be altered by various factors, including the preservation system in the food. For example, the heat resistance of bacteria and their spores is generally increased by low water activity, but decreased by reduced pH. Gaze and Betts (1992) have produced an overview of types of pasteurisation processes and their microbiological targets, based on published heat resistance and growth data. Pasteurisation values (P-values) specify the effectiveness of pasteurisation treatments and are used to indicate the equivalent heat treatments based on a reference temperature (e.g. 85 °C). Whilst it usually involves heat, pasteurisation may involve the use of ultra high pressure (UHP; Hendrickx et al., 2002).

9.6.3 z-value z-value is an empirical value, quoted in temperature (degrees C or F), and used for calculating the increase or decrease in temperature that is needed to alter by a factor of 10 the rate of inactivation of a particular micro-organism. Quality change does not follow the same kinetics. It assumes that the kinetics of microbial death at constant temperature is exponential (i.e. ‘log-linear’). Although z is fundamental to the calculation of heat process equivalence, it should be used with extreme caution for low temperature pasteurisation processes (below 60 °C), as the death kinetics of many types of microbial cells at these temperatures are not log-linear, especially when heating rates are low. There may be ‘shoulders’ and ‘tails’ on © 2008, Woodhead Publishing Limited

208

Chilled foods

survivor curves (Gould, 1989), ‘heat adaptation’ may occur or there may be an increase in the resistance of micro-organisms during heating (Mackey and Derrick, 1987). Although the validity of the z concept is limited at pasteurisation temperatures, it is commonly used by product designers. If preservatives interact with the heat treatment, or if processes are designed to cause large log reductions (i.e. in excess of 10 000-fold) because poor quality ingredients are expected, then tailing of the survivor curves may become an issue and the impact of the necessary extended heating times should be considered when quality is assessed. Actual or challenge trials should be undertaken to validate safe processes.

9.6.4 Re-heating Customers usually re-heat products according to the manufacturers’ ‘use instructions’, designed to ensure optimum culinary quality. For raw products, heating may be needed to pasteurise them and this implies that it should be reliable throughout the product. To achieve this, products are best designed for conventional heating; if microwave re-heating is required, the possibility of cold-spots should be considered and prolonged heating plus a period of standing (for temperature equilibration) should be specified in the use instructions and its effectiveness validated in a number of different ovens. A suitable method for validation uses alginate beads containing micro-organisms with a known heat resistance (Holyoak et al., 1993). As a rule of thumb, microwave heating should be limited to food that has been freed of infectious pathogens during processing and has remained so until consumption. Uneven heating may be accentuated in multi-component meals and is caused by the different heating characteristics of each component, including saltiness, type of tray, and geometry and layout (Ryynanen and Ohlsson, 1996).

9.6.5 Cooling Cooling reduces the product temperature after factory cooking. Effective cooling is critical to product safety and quality. Target cooling rates and the specification of the hygiene of coolers will depend on whether the primary packaging of the product is done before, or after, cooling and this must be included in the product design. During cooling, time spent in the temperature range 60 down to 5 °C should be as short as possible (no more than 4 hours) to minimise risks of spore germination and outgrowth, and this imposes some limits on the design of products (especially size), pack type and equipment operation. Minimum cooling rates are often specified in legislation; for example, in the EEC Meat and Meat Products Directive 77/99 (now 852/2004), prepared meals must be cooled to below 10 °C within 2 hours of cooking. Cooling of liquid products (e.g. soup) may be done in line, before packing, using heat exchangers. Where this is a possibility, hygiene and chances of damage to particles by pumps should be considered at the design stage, as any contamination at this stage will be packed with the product. Where products are cooled in pack, then pack shape, especially length of the thermal path, and the presence of a headspace and the composition of the packaging material will limit cooling rate. From an equipment © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

209

design perspective, the loading density of packs in trays or other containers for cooling will affect line capacity and also the cooling rate/heat exchange capacity of individual packs. Close packing will reduce cooling rates. The design of chillers, especially their air distribution pattern, air velocity and temperature, and the way that product containers are packed into them, will control cooling rates. Use of very cold heating media (e.g. air well below 0 °C) may freeze pack surfaces and cause damage to textures. For products cooled before packaging, for example in bulk trays, the same limitations apply; product depths exceeding 10–15 cm should not be used because above this thickness, conduction of heat to the surface, rather than removal of heat from the surface, will limit the rate of cooling. For cooling out of the primary packaging, additional hygiene precautions must be taken to prevent contamination. These may include covering trays and ensuring very high standards of cleanliness in chillers and the utilities supplying cold air. Evans et al. (1996) have highlighted the importance of cooling and the mandatory requirements that exist in the UK for cook– chill products (Anon., 1982). These guidelines recommend that 80 mm trays should be chilled to below 10 °C in 2.5 hours, between 10 and 40 mm they should be chilled to below 3 °C in 1.5 hours. Assuming that surface freezing is to be avoided and a simple, single-stage operation used, only a 10-mm deep tray can be chilled within these time limits. Racking or packing systems in chillers should ensure that the flow of cold air over container surfaces is turbulent, so that cooling rates are maximised. Special attention should be paid to hygiene and the control of condensation in chillers, as this is a major potential source of recontamination with Listeria, especially if condensation is re-circulated in aerosols by the air flow. The importance of chilling after preparation, chilled storage and chilled distribution have been discussed by Baird-Parker (1994) as CCPs in the manufacture of raw and cooked chilled foods. Farquhar and Symons (1992) have noted a US code of practice with recommendations for the preparation of safe chilled foods; it covers chilling, chilled storage, pre-distribution storage and handling and temperature management practices.

9.6.6 Chilled storage Chilled storage should be designed to maintain specified temperature in a precooled product. Products or ingredients in containers should enter storage chills at their target temperatures, because the heat exchange capacity and performance of chiller air systems (temperature and air velocity) and the way product is stacked for distribution and storage are normally not designed to allow significant reductions in temperature to be made.

9.7

Manufacturing areas

The design of manufacturing areas is critical to the realisation of product designs. As part of the design process, the hygiene level required at each stage of manufacture should be found. Generally hygiene levels can be reduced once a product is in © 2008, Woodhead Publishing Limited

210

Chilled foods

its primary packaging, the highest levels being required for the handling of heated, unpacked product (see CFA, 2006).

9.7.1 General purpose manufacturing area A general purpose manufacturing area is designed to receive and handle all types of ingredient. Any process intermediates made in this area are usually heat-treated before they are sold as products and should pass into hygienic or high-care areas during or after processing.

9.7.2 Hygienic area A hygienic area is a processing area designed for the storage, handling and preparation of low-risk raw materials and products (Class 1) containing a mixture of heat-treated and undecontaminated ingredients. It should be designed, constructed and laid out for easy cleaning, so that high standards of hygiene can be achieved. Layout and practices should prevent bacteria, such as Listeria, becoming established in it and contaminating products. When it is used for the assembly of final products containing undecontaminated components, such as cheese, it should not be used for the processing or preparation of any other ingredients likely to carry pathogens and likely to increase the risks of contamination. Areas conforming to this minimum standard of hygiene should be used for the post-heat process handling of in-pack pasteurised products (Class 4).

9.7.3 High-care area This is a well-defined, physically separated part of a factory which is designed and operated specifically to prevent the re-contamination of cooked ingredients and products during chilling, storage, assembly and primary packaging. It is an integral part of the factory layout shown in Fig. 9.3 and is used for the preparation of products in Classes 2 and 3. Usually there are specific hygiene requirements covering layout, standards of construction and equipment, the training, hygiene and routing of operatives, engineers and management, and a distinct set of operational procedures (especially covering the intake and exit of food components and packaging material), all designed to minimise the chances of contamination. The usage of re-packed and re-worked materials (e.g. products in defective packs) in high-care areas should be discouraged, and if this is not possible then very strict rules of segregation in time should be enforced.

9.7.4 Air handling All hygienic areas are likely to have a controlled and filtered air supply. Design of the air distribution systems should recognise that air is a significant means of dispersing contaminants and particular attention should be paid to the quality and direction of flow of air within these areas or between areas with different levels of © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

211

hygiene. Air flow should always be from ‘clean’ to ‘dirty’ to minimise the chances of contaminants being carried from raw to decontaminated product. The microbiological or particulate quality of the air should be related to the hygiene category of the area (see manufacturing area, hygienic area and high-care area requirements; CFDRA, 1997).

9.7.5 Cleaning Cleaning should be designed to remove food debris from process equipment, manufacturing and storage areas. Effective cleaning should remove all food debris from work surfaces, machines or an area, so that microbes cannot grow and subsequent production is not contaminated. Effective cleaning cannot be achieved unless equipment is hygienically designed and maintained. In practice, complete removal of food residues is rarely achieved by the techniques used for cleaning open plant (e.g. in slicers and dosing equipment). In factories manufacturing chilled foods, the residues after cleaning may provide growth substrates for the factory microflora, and experience has shown that some modern cleaning techniques and chemicals, when used in chilled areas, may actually select for Listeria for survival. The risks of recontamination by proposed equipment and facilities should be considered by the product design and design HACCP, especially when the product shelf-life is proposed. Cleaning may generate aerosols that contaminate products and equipment with food debris and bacteria, layout designs should minimise the risks, and food and packaging materials should be removed during cleaning. The design HACCP can make a valuable contribution by identifying process stages where hygiene is critical to product quality and safety, and overlaying the process flow diagram on a factory plan to show potential risky areas. After cleaning, sufficient time should be left for the air systems in hygienic and high-care areas to return to their original hygiene levels.

9.8

The microbiological hazards

The microbiological hazards of chilled foods can be roughly classified according to the causes of harm to consumers. Harmful microbes may infect consumers, or they may have pre-formed toxins, or grow in the food, producing toxins that then cause disease after the food is eaten. The micro-organisms of greatest concern are listed in Table 9.2. All realistic microbiological hazards should be identified and then controlled by the product and process design. Realistic, not specified, values should be used in designs; for example, practical storage temperature and shelf-life must be assessed by the design to determine whether a hazard and controls are realistic in a particular product. Infectious pathogens can be hazardous at very low levels of cells, whereas appreciable levels, or growth, of toxigenic micro-organisms are needed in order to cause a hazard. When designing a product or process, it is very risky to assume that © 2008, Woodhead Publishing Limited

212

Chilled foods

Table 9.2 Food-poisoning organisms of major concern and their heat resistance and growth temperature characteristics Minimum growth temperature Low

Medium

High

Heat resistance Low → High Vegetative cells Spores Listeria monocytogenes (INF) Yersini enterocolitica (INF) Vibrio parahaemolyticus (INF) Aeromonas hydrophilia (INF) Salmonella species (INF) Clostridium perfringens (INF) Escherichia coli O157 (INF) Staphylococcus aureus (TOX) Camplylobacter jejuni and coli (INF)

Clostridium botulinum type E, non-proteolytic B&F (TOX) Bacillus cereus (TOX) Bacillus subtillis (TOX) Bacillus licheniformis (TOX) Clostridium botulinum type A and proteolytic B (TOX)

INF = infectious; TOX = toxigenic.

a particular microbiological hazard is unrealistic because it has not so far been detected in a component. Generally, survival of micro-organisms is greater in chilled foods than at ambient or frozen temperatures, but toxin production may be inhibited. The infectious pathogens (see Table 9.2 and Chapter 16) include Salmonella, E. coli O157:H7 and Listeria monocytogenes. They may be present in raw materials such as meat, vegetables and cheese made from unpasteurised milk. All may survive for long periods in chilled products (e.g. Escherichia coli O157:H7 survives for 22 days at 8 °C in crispy salad) if not eliminated by processing. All the infectious pathogens are heat sensitive and will be eliminated by the conditions used for pasteurisation (e.g. 70 °C for 2 minutes or 72 °C for 16.2 seconds). The growth of Salmonella and E. coli O157:H7 in products or in the factory environment may be controlled by refrigeration (i.e. by temperatures below about 10 °C). Escherichia coli O157:H7 has a low infectious dose and causes serious illness, especially in the young and the elderly, as it attaches to the wall of the intestinal tract and causes acute, bloody diarrhoea (hemorrhagic colitis) or haemolytic uraemic syndrome (a kidney disease). Several outbreaks of disease have been linked to chilled food and have usually had a bovine origin (e.g. undercooked ground-beef products), although unpasteurised cider and mayonnaise have been implicated. In the latter case it is thought that improper handling of bulk mayonnaise or cross-contamination with meat juices or meat products was the cause. It has been found that this Escherichia coli is more tolerant of acid environments than other known strains; therefore it may survive in fermented dry sausage and yoghurt. Salmonella enteritidis is a potential hazard in products made from poultry and eggs, whereas the multi-drug resistant Salmonella typhimurium DT 104 can be found in a broad range of foods. Outbreaks in the UK have been linked to poultry, meat and meat products, and unpasteurised milk. Campylobacters may cause intestinal infections leading to fever, diarrhoea and © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

213

sometimes vomiting. Sources include water, milk or meat. Campylobacter jejuni is regularly found on retail raw poultry and outbreaks have been associated with undercooked poultry and the cross-contamination of ready-to-eat materials from kitchen surfaces or via the hands of kitchen staff or work areas. It does not grow below 30 °C and therefore conditions affecting survival are important, since sufficient cells must survive to form an infectious dose. Vibrio cholerae may survive on refrigerated raw or cooked vegetables and cereals if they have been sourced from tropical or warm areas where contamination is endemic. Particularly at risk are sea and other foods harvested from estuarine or inshore waters, waters subject to land-water runoff, or fields irrigated with sewage-contaminated water. Contamination can also occur if the produce itself is cooled, washed or freshened with contaminated water. During preparation, food from this type of origin, which can include raw, pre-cooked and processed molluscs, crustaceans, fish and vegetables, should be handled to minimise the chances of cross-contamination and pasteurised prior to sale. Psychrotrophic pathogens such as Listeria monocytogenes (Walker and Stringer, 1987) can grow at refrigeration temperatures and may readily become established on badly designed or maintained equipment and in the factory environment. They are detected only in low numbers in environmental samples associated with the primary production of food (Fenlon et al., 1996) and are therefore likely to be contaminants arising from manufacturing conditions. Under otherwise optimal conditions, some strains of Listeria monocytogenes have been shown capable of slow growth at temperatures as low as –0.1 °C, Yersinia enterocolitica at –0.9 °C and, Aeromonas hydrophila at –0.1 °C (Walker, 1990). For example, Listeria monocytogenes is able to grow well in components such as chill-stored, prepared vegetables and in many products lacking robust chemical preservations systems – such as chill-stored ready-meals and paté. The most severe toxigenic pathogens are cold-growing non-proteolytic strains of Clostridium botulinum. Their growth in pasteurised foods is a particular risk if processing has eliminated the competing vegetative flora, and their growth could precede spoilage. The proteolytic strains are less hazardous because they are able to grow only at higher temperatures and, unlike the non-proteolytic strains, they normally cause spoilage that makes the product inedible. At chill temperatures (< 8 °C) the growth rate of the non-proteolytic types is slow and so requires controls only in products where the designed chilled shelf-lives exceed about 10 – 14 days or high contamination levels are expected. Graham et al. (1996) suggest that non-proteolytic strains of Clostridium botulinum can grow at chill temperatures and therefore pose a potential hazard in minimally heat-processed chilled foods. They have developed models predicting growth and compared them with published data to demonstrate that they are suitable for use with fish, meat and poultry products. The models cover the combined effects of pH (5.0–7.3), salt concentration (0.1–5.0%) and temperature (4–30 °C), and are based on the growth of non-proteolytic Clostridium botulinum in laboratory media. Fortunately, the spores of the strains able to grow at chill temperatures are relatively heat-sensitive and therefore can be controlled by realistic cooking or pasteurisation processes. The © 2008, Woodhead Publishing Limited

214

Chilled foods

heat process recognised as suitable for these long-life chilled foods is 90 °C × 10 minutes. Although cooking can eliminate the cold-growing types of Clostridia, storage temperature remains the most important control on their growth. The UK Advisory Committee on the Microbiological Safety of Food (1992) have discussed the potential hazards of chilled foods made using vacuum packaging and associated processes, such as sous-vide, and they particularly considered the risks associated with botulism. They highlighted methods to prevent and/or control the risks of botulism, including adequate heating based on the heat resistance of the spores and the restriction of shelf-life (see Smith et al. (1990) for details of the use of HACCP to address this problem). At temperatures above 12–15 °C, mesophilic spore-formers (with more heatresistant spores) can grow, and the processes used in the manufacture of chilled foods certainly do not inactivate their spores. Bacillus cereus is sometimes mentioned as a potential hazard in chilled foods, though evidence of its ability to produce harmful toxins in these foods (with the possible exception of dairy products) is equivocal. Some strains of Bacillus cereus that can grow slowly at temperatures as low as 4 °C (van Netten et al., 1990) may be hazardous, but there is little epidemiological evidence. Staphylococcus aureus is the other toxin producer causing concern in chilled foods, though it is of importance only when the food does not contain a competing microflora and has been cooled slowly and re-contaminated, or substantially temperature abused during re-heating. So it is important that high-care areas and the operational procedures used in them are designed to prevent heat-processed foods becoming contaminated with Staphylococcus aureus.

9.9

Risk classes

Risk classes (1–4) indicate the level of microbiological risk associated with various products. Consumer risk depends on the microbes remaining in the foods after manufacture and any growth during distribution and storage. Hence processing, hygiene and packing requirements should be primarily designed to control the risks of products containing infectious or toxigenic pathogens. Control of spoilage microbes and meeting specified shelf-life should be the secondary consideration, although it may often require the application of more severe heating processes or more stringent conditions of hygiene or preservation than the control of safety. If the microbiology cannot be controlled without prejudicing sensory quality, a commercial decision must be made on the acceptable balance between a controlled loss of quality caused by processing and spoilage in the marketplace. The process and product design should not compromise microbiological safety standards to improve sensory quality. If the required process or preservation conditions cannot ensure safety against the background of realistic consumer usage, then the product design should not be progressed without revision (see Gould, 1992; Walker and Stringer, 1990). There are defined risk classes for chilled food designs (Table 9.1). © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

215

• Class 1 includes prepared chilled foods made entirely from raw ingredients and obviously requiring cooking by the customer.

• Class 2 foods contain mixtures of raw and cooked components, processed or

• •

packaged to ensure a satisfactory shelf-life. They may not so obviously require cooking and may contain infectious pathogens which may (e.g. Listeria monocytogenes) or may not (e.g. Salmonella) be able to grow during chilled storage. The manufacturer can control safety by minimising the levels and incidence of pathogens on incoming materials (e.g. by careful choice of suppliers). Storage and processing procedures should be designed not to introduce additional contaminants or allow numbers to increase. Shelf-life and storage temperature should be designed to ensure that only ‘safe’ numbers of infectious pathogens could be present if foods are stored for the full indicated shelf-life. At present there is no generally accepted estimate of the infectious dose of Listeria and it remains up to individual manufacturers or trade associations to decide on acceptable risks. Because Listeria is able to grow at chilled temperatures (for example during storage, distribution and domestic storage), only its complete absence at the point of manufacture will ensure that ready-to-eat foods are safe, without qualification. Where it is present after manufacture, then the producer is accepting the risk associated with sensitivity of customers to any Listeria monocytogenes that may be found at the point of consumption (see EC Microbiological criteria). Salmonella should be absent, as the infectious dose is very low. Class 3 foods contain only cooked or otherwise decontaminated components and may be assembled in high-care areas. Class 4 foods are cooked by the manufacturer within their primary packaging. If manufactured under well-controlled conditions, these foods will be free of infectious pathogens (such as Listeria and Salmonella) and spoilage microbes, and will have a substantially longer shelf-life (up to 42 days or more, with a severe pasteurisation process) than those containing raw components. This substantial extension of spoilage-free shelf-life has important consequences for the logistics chain, but potential changes in their microbiology during storage should be recognised by the design as limiting safety during storage (see below), and hence which pasteurisation conditions are appropriate during manufacturing.

9.10 Safe process design – equipment and processes The manufacture of chilled foods is a complex process. From a microbiological point of view, processes should be designed to control the presence, growth and activity of the micro-organisms identified as realistic hazards. Some of the unit operations making up a process will be designed to eliminate or reduce numbers of bacteria, some may add preservative factors, and others incidentally provide opportunities for re-contamination or growth. The raw materials, the product design and shelf-life (see Table 9.2), and possibly factory hygiene and layout, will © 2008, Woodhead Publishing Limited

216

Chilled foods

determine the target bacteria at each unit operation. At the very beginning of the supply chain, agricultural produce, farm animals, and their products, can act as reservoirs of food-poisoning bacteria (e.g. Salmonella, Campylobacter and Escherichia coli O157), therefore the design of handling and processing controls should provide reliable means of eliminating those identified as hazards and preventing re-contamination of products. The extent and location of precautions needed for control of a particular hazard and the monitoring measures will be proportional to the length, complexity and scale of the supply chain and should be covered by the design HACCP. For short shelf-life (< 10 days) prepared chilled foods, which are often readyto-eat, the main safety risk is infectious pathogens. Processes should be designed to eliminate them and prevent recontamination. There should be controls for the two main re-contamination routes, via food-handlers and cross-contamination from other foods. Process flows and handling procedures should be designed on the basis that raw foods may contain low numbers of food-poisoning bacteria and therefore effective separation of material flows and personnel is necessary. Prolonging the chilled shelf-life above 10 days or so (maybe to 42 days) introduces an additional hazard, as it allows time for the growth of cold-growing toxigenic bacteria to levels where toxin production is possible (cold-growing strains of Clostridium botulinum), and processes need to be designed to eliminate them. Their spores may survive in foods pasteurised with the processes designed to destroy infectious pathogens. More severe heat treatments (90 °C × 10 minutes) are needed to reduce (6 log) numbers of their heat-sensitive spores. If the preservation system of the food can prevent the outgrowth of these spores, then the heat process needs to be designed only to control spoilage micro-organisms. Many chilled foods are designed with intrinsic preservation systems (e.g. low Aw or reduced pH) and are therefore safe, although they have low heat processes. There is still no general agreement among manufacturers or regulatory authorities on the risks of botulism from unpreserved chill-stored foods. However, there is ample evidence that, in spore-inoculated model systems based on ready meals, growth occurs and toxin is produced at temperatures representing those known to be found under commercial conditions (Notermans et al., 1990). It is essential that, if factory heating has not been done in the primary packaging, unwrapped, heat-treated components intended for long-life foods are chilled, handled and assembled in high-care areas to prevent re-contamination with spores and infectious pathogens. Even if re-contamination is prevented, there is a remaining risk from the survival in the components of heat-resistant bacterial spores (e.g. Bacillus species that can eventually cause a musty form of spoilage) able to grow slowly at chilled storage temperatures.

9.10.1 Equipment Many of the critical attributes of chilled foods are controlled by the technical performance of processing equipment and utilities (e.g. air and cooling agents); hence, the equipment available to make the product should be identified at the © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

217

beginning of a design exercise. Material heating characteristics and the lethality of heating processes used will exert a critical effect on the chances of microbial survival and presence in the product; therefore, the correct design of the heat treatment and an understanding of the variability of control of the heating and cooling stages are most important when the process is designed. Similarly if preservatives, such as curing salts, humectants and acidulants, are dosed and mixed, the variability of their distribution after processing and filling must be known. The effectiveness of packaging machines at producing gas-tight packs (e.g. containing an inhibitory gas mixture such as CO2 and N2) should also be known when a design is turned into operational procedures and specifications. The hygienic design of machines is important as products may be re-contaminated by food residues remaining in a machine or processing area after cleaning, and unless these residues are regularly removed they pose a hazard in finished products; therefore, cleaning requirements should be specified in the design. Many codes of practice state that ‘food processing equipment should be designed to be cleanable and capable of being disinfected’. However, this statement hides the real issues in designing, operating and maintaining food manufacturing equipment, which often centre on finding an acceptable balance between cost, manufacturing efficiency (especially change-over time) and hygienic design (see Chapter 15).

9.10.2 Heat processes – the use of heat to decontaminate products Heating methods Heat is the agent most commonly used to inactivate micro-organisms and cause the flavour, texture and colour changes associated with the product. A product design should specify the required product characteristics, process parameters and the different equipment available for heating. For example, liquid products may be heated directly with culinary steam or hot water in open vessels, or indirectly using vessels with heated jackets. With direct heating, the heating medium condenses, becomes part of the product, and the volume increase from condensation needs to be accounted for in the design. With indirect heating, the product volume is unchanged. Pieces of meat, fish or vegetables may be cooked in atmospheric ovens ‘open’ in trays, or ‘closed’ in moulds or sous-vide packs. Where the packs are sealed, any liquid released as part of the cooking process is retained, whereas with open cooking there are weight losses. Continuous heat exchangers may be used to heat and cool liquids or pumpable ingredients and with suitable pressure controls can achieve product temperatures over 100 °C. Wherever product temperatures above 100 °C are required by a design, closed equipment should be specified. However, if pieces of meat, fish or vegetables are heated to such temperatures by contact heating or deep frying, overpressure control is not needed. In these cases product surface temperatures are likely to exceed 100 °C, but centre temperatures will be much lower, depending on the heating characteristics and initial temperature of the material, the heating time and the heating medium or surface temperature. © 2008, Woodhead Publishing Limited

218

Chilled foods

Packaged products or ingredients may be heated in pressurised vessels (e.g. retorts or pressurised kettles) using heating medium temperatures above 100 °C. For this type of heating the vessel may be ‘overpressured’ to prevent pack distortion. Quite commonly, sealed packs are heated in water baths. Such packs should be vacuumised to prevent pack distortion and ensure good heat transfer. If distortion is accepted, then the heating process should be designed to take account of slower heating rates due to the insulating effect of the expanded headspace during heating, and seal strength should prevent leakage. Control of heating The heating process must be designed to deliver the specified minimum heat treatment. Because of the importance of this, critical control parameters always meeting the designed heat treatment should be defined in a specification, which must be available to the operatives in charge of the process. Training of operatives should enable them to carry out processes reliably and to monitor and record progress as specified in the HACCP study. Control and monitoring of heating is most effectively done by following the time/temperature conditions in the vessel or chamber or sometimes in the product itself, using suitable temperature sensors. For each batch of product, it must be verified that the specified process and hence the target lethality have been achieved. The performance of equipment is not the only factor determining effective heating. Heat transfer to the product surface and movement of heat within the product must also be predictable and are controlled by product (e.g. size, shape, viscosity or presence of any coatings) and also by pack characteristics. These must be noted in the product design. If an oven is used to heat packs, they must consistently have the same size, be made of the same material and be positioned in the same way during heating. Features determining heating of packs or ingredients (such as size) must be set at the design stage and controlled during production, so that specified time/temperature integrals at the coldest or slowest heating points are reliably achieved. Processing is also affected by the characteristics of in-process material, most notably the temperatures of components (e.g. frozen, thawed or warm). Initial temperatures at any stage will dictate heating rates, and so again these variables should be specified in the design and then translated into the process specification. Equipment performance Equipment performance and maintenance can help or hinder the manufacture of products meeting the design. Many pieces of commercial heating equipment, such as ovens and retorts, do not provide uniform heating, and so processes are usually set for the coldest or slowest heating area. This approach does not usually consider the quality penalty of over-processing in the hottest areas. Investigation of equipment features is an essential part of process or product development, and the results should be incorporated into the design to achieve an optimum quality product. In an oven or retort, heat distribution can depend on the position of the inlet for the heating medium, packing density, or positioning of the product, the last two of © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

219

which may create or block channels between product units and upset flow of the heating medium. The change in product centre temperatures over time is used to estimate heating of the product in response to conditions in the equipment (Fraile and Burg, 1998a, b). Therefore, an essential part of designing a heating process is knowledge of the distribution of heat within equipment under the intended conditions of usage, so that processes can be set using parameters to ensure the consistent delivery of a minimum process. Stoforos and Taoukis (1998) have proposed a procedure for process optimisation that relies on the use of a two- or three-component time/temperature integrator for thermal process evaluation. Their proposed procedure can take account of the different z values associated with microbial killing and quality loss to assess the impact of particular combinations of time and temperature. The more variable or non-uniform the delivery of heat during processing, the higher the target heat process needs to be set to ensure that the required minimum is always achieved, and this may cause over-processing and loss of quality when the design is translated into operational procedures. An essential part of developing the design is investigation of the range of heat treatments delivered by the chosen equipment within the predicted range of operating conditions. It is also important that the maximum input level of microbes is controlled, because if the input numbers are exceeded, survivors will be found (possibly by the customer). Cooling Although heat is effective at killing micro-organisms, to ensure safety and quality, the effectiveness of cooling (USDA, 1988) and the hygiene of cooling equipment must be specified equally carefully (James and Bailey, 1990). Even in equipment designed to achieve rapid cooling, the risk of product re-contamination remains from micro-organisms endemic in the chiller, in the heat exchanger, or taken in and then spread by the forced air circulation needed for rapid cooling. The rate of cooling may determine the dormancy of any spores surviving in the product – this will affect their readiness to germinate and grow during product storage. The extent of dormancy is particularly important in shelf-life determination when heat has been used in conjunction with chemical preservation systems, such as salt and nitrite or REPFEDs.

9.10.3 Microbiology of heat processing Packed products are usually heated by hot water or steam, and less commonly by microwave or ohmic heating. Heat treatments are designed to match the shelf-life and usage of the food, and the target micro-organisms carried by the raw materials (see minimum processes suggested for short and long shelf-life products, below). Target bacteria include the pathogenic non spore-forming organisms Salmonella, entero-pathogenic Escherichia coli, Campylobacter, Listeria monocytogenes and Yersinia enterocolitica, and the spore-forming non-proteolytic strains of Clostridium botulinum, type E, some type B and F. As raw materials are sourced globally, other pathogens (such as vibrios) may be introduced and the heat-processing © 2008, Woodhead Publishing Limited

220

Chilled foods

element of the HACCP plan should be reviewed to ensure that adequate heat processes are used. Whilst a heat treatment of 70 °C for 2 minutes at the coldest part of a pack will ensure at least a 106-fold reduction of Listeria monocytogenes, the most heatresistant of the vegetative micro-organisms mentioned above, such a treatment will have no effect on the spores of the psychrotrophic strains of Clostridium botulinum. Consequently, in the UK the low heat treatments of 70 °C for 2 minutes are recommended only for short-shelf-life products or for those in food service operations, for which 3 °C storage is certain to be maintained (Glew, 1990). In The Netherlands, a heat treatment of 72 °C for 2 minutes has been recommended to ensure >108-fold inactivation of Listeria monocytogenes (Mossel and Struijk, 1991). Although long, slow, mild heating, as in the original sous-vide processes, may sometimes be desirable for organoleptic reasons, it is important to remember that slow heating (prolonged period at 45–55 °C) may trigger the so-called ‘heat-shock’ response, which increases the apparent heat resistance of vegetative micro-organisms (Mackey and Bratchell, 1989). Therefore, warm-up or warm-holding times during processing should be short, or increased heat resistance may be found. Although psychrotrophic strains of Clostridium botulinum cannot grow at or below 3 °C, the possibility of their slow growth in long-life products at temperatures just above this demands a more severe thermal process which kills spores. This process should be designed to reduce their numbers more than 106-fold, but there is still debate about the minimum heating required for safety. For instance, Notermans et al. (1990) concluded that there still remained insufficient data on the heat resistance of spores of non-proteolytic Clostridium botulinum to specify adequate minimum processes for conventional sous-vide processing (see below). They found that surviving spores could germinate, grow and form toxin within about 3 weeks at 8 °C. Pre-incubation at 3 °C shortened the subsequent time to toxin at 8 °C. They concluded that if storage below 3.3 °C cannot be guaranteed (as is often the case during retailing and storage in the home), then storage time must be limited. However, it must be said that these products have been on the market for many years, with no recorded microbiological safety problems to date. Such comments refer to pasteurised, vacuum-packed foods that do not have any additional preservation factors and rely only on temperature control during distribution for shelf stability and safety. However, many pasteurised vacuum-packed products do include additional intrinsic preservation factors designed to enhance keepability and safety (CFDRA, 1992), for example salt- and nitrite-containing products such as hams and other cured meat products, acidified pasteurised meat sausages and a wide range of Aw-reduced traditional products, some of which are chill-stable and some even ambient-stable, e.g. the so-called SSPs (‘shelf-stable products’) of Leistner (1985). The processing, safety and shelf-life requirements of these products should not be confused with those of pasteurised, unpreserved chilled foods.

9.10.4 Pasteurisation for short shelf-life (Classes 1 and 2) Short shelf-life products are designed to have a shelf-life of up to 10–14 days. Heat © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

221

treatments in the factory or during consumer heating should cause at least a 6-log reduction in the numbers of infectious pathogens (Salmonella and Listeria), and their handling after heating and packaging in the factory should prevent recontamination. In neutral pH, for high water activity products that do not contain antimicrobial preservatives, combinations of temperatures and times equivalent to 70 °C for 2 minutes are more than adequate for this reduction. But practical experience has shown that longer times in this temperature range are needed for the effective control of more heat-resistant, non-sporing, spoilage bacteria, such as lactic acid bacteria.

9.10.5 Pasteurisation for long shelf-life (Classes 3 and 4) Long shelf-life products, with chilled shelf-lives sufficiently long to allow the outgrowth of psychrotrophic spores, need heat treatments designed to eliminate these spores. The treatments should be designed to give at least a 6-log reduction in the numbers of cold-growing strains of Clostridium botulinum; 90 °C for ten minutes, or an equivalent process, is generally accepted as being sufficient. These processes are not sufficient to eliminate all spores of psychrotrophic Bacilli. In unpreserved products, survivors are able to grow to levels causing spoilage within 3 weeks or so at 7–10 °C, temperatures which are known to occur in the chill distribution chains in many countries (Bogh-Sorensen and Olsson, 1990). These cold-growing Bacillus spores often have D90 values of up to 11 minutes (Michels, 1979).

9.10.6 Microwaves Microwave cooking or re-heating of foods, particularly in the home, has expanded greatly in recent years and so there is an increasing range of chill, ambient, and frozen-stored foods for microwave heating. Furthermore, it is likely that the use of microwaves to cook or pasteurise foods either industrially or domestically will grow in the future. Key issues are the design and preparation of foods that heat predictably with microwave energy. It is known that heating is determined by the dielectric constant of the food and its positioning and thickness within a container (van Remmen et al., 1996). The practical problem is preparing food ingredients and dosing them sufficiently accurately on a commercial scale to ensure uniform and predictable heating from microwave absorption. Where heating relies on microwave application and then a period of holding to allow temperature equilibration, lines are likely to have low capacities. Whether it is microwave or other forms of energy that are used to generate heat, there is no difference in the lethal effect on micro-organisms. There is no welldocumented ‘non-thermal’ additional microbicidal effect from commercial or domestic microwave equipment. There have been continuing concerns about the microbiological safety of foods re-heated in domestic microwave ovens, particularly because of the occurrence of cold-spots; see Sage and Ingham (1998), Tassinari and Landgraf (1997) and Heddleson et al. (1996) for discussions of © 2008, Woodhead Publishing Limited

222

Chilled foods

heating variability and its consequences for microbial survival. There are concerns for foods that have not been fully pre-cooked, may have been contaminated after cooking, or contain raw ingredients. These concerns have been reinforced by demonstration of the presence of Listeria monocytogenes in a wide range of retailed foods, including some suitable for microwave re-heating. For example, a UK Public Health Laboratory Service survey found the organism to be detectable in 25 g samples of 18% of commercially available chill meals tested (Gilbert et al., 1989). This concern led the Ministry of Agriculture, Fisheries and Food (now DEFRA) in the UK to review the issues thoroughly with oven manufacturers, the food industries, retailers and consumers, and to promote recommendations concerning the proper employment of microwave ovens for re-heating, so that effective pasteurisation is achieved. The unexpected survival of micro-organisms in food heated by microwaves has been attributed to enhanced heat resistance (e.g. Listeria monocytogenes; Kerr and Lacey, 1989), but it is now generally accepted that survival resulted from nonuniform heating leading to ‘cold spots’ (Lund et al., 1989; Coote et al., 1991). The generation of cold spots results from non-uniform energy absorption caused by differences in the dielectric properties of ingredients, composition and weight of ingredients, and product and pack geometry. Measurement of the heat-induced inactivation rates of Listeria monocytogenes in a variety of food substrates has shown that a 10-fold reduction in numbers is achieved by heating at 70 °C for 0.14–0.27 minutes (D70 = 0.14–0.27 minutes; Gaze et al., 1989). Consequently, the UK Department of Health recommendation for Listeria-sensitive cook–chill foods to receive a minimum microwave heating throughout achieving 70 °C for 2 minutes, would result in a reduction of more than 106-fold (Anon., 1989). Likewise, processes designed to use microwaves for heating or cooking during the manufacture of short shelf-life chill foods should deliver this minimum amount of heating to all parts of the product or ingredient.

9.10.7 Products cooked in their primary packaging (e.g. sous-vide products and REPFEDs) Although the term ‘sous-vide’ strictly refers to vacuum packing, without any indication of thermal processing, it has nevertheless become an accepted name for pasteurised ingredients or products that are vacuum packed in their primary packaging prior to heat processing (risk class 4). In this system, raw or par-cooked foods, such as meat, fish and vegetables, or meals and meal components, soups and sauces, are vacuum packaged into laminated plastic pouches or containers, often made from an oxygen-impermeable film and cooked at low temperatures (above 60 °C, microbiologically ideally 90 °C), rapidly chilled (within 2 hours to 3 °C) and then processed or re-heated for serving (see http://www.techneusa.com/ Seal%20Appeal.pdf). All have extended chill shelf-lives (up to 3–6 weeks at 0 to 3 °C) and, depending on pack size, are used in manufacture, catering or for retail sale. Although they are heated at relatively low temperatures, heat treatments have to be sufficient for them to remain safe and microbiologically stable during © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

223

prolonged storage below 3 °C (the minimum theoretical growth temperature of the cold-growing types of Clostridium botulinum). The severity of the heat treatment needed depends largely on the microflora of their ingredients (Church and Parsons, 1993). An outline for a HACCP study has been published (Adams, 1991). The most comprehensive early research and evaluation of sous-vide processing was for catering at the Nacka Hospital in Stockholm. Prepared foods were vacuum packed, rapidly chilled, then stored under well-controlled refrigeration for periods of one or two months prior to re-heating for consumption (Livingston, 1985). The technology is now used for catering in a number of European countries and is extending to retail foods. If not adequately controlled, the sous-vide technology presents significant microbiological hazards (ACMSF, 2000), for several reasons. Microbial spores capable of growing above 3 °C are not inactivated by the low cooking temperatures, which may also provide only low levels of destruction of vegetative pathogens. There are also concerns about ensuring the required low storage temperatures (max. 3 °C) throughout long-distance distribution chains and, especially, in the home (see Betts and Gaze, 1995; Juneja and Marmer, 1996; Peck, 1997, for a discussion of the risks of botulism; Hansen and Knochel (1996) and Ben-Embarek and Huss (1993) for the effectiveness of sous-vide processes against Listeria monocytogenes); and Turner et al. (1996) for effectiveness against Bacillus cereus and spoilage bacteria in chicken breast). A proposal for a code of practice is given by Betts (1996).

9.11 Safe process design – manufacturing areas Manufacturing areas and production lines should be designed on the principle of forward flow, so that the chances of cross-contamination, or of products missing a process stage, are minimised.

9.11.1 Raw material and packaging delivery areas Factories should have designated areas for deliveries to prevent cross-contamination, minimise damage to packs and containers and ensure continuity of storage conditions. They may be divided into different areas for the various commodities to be processed, or according to their storage requirements, e.g. frozen, chilled or ambient-stable. Separation may also be governed by legislative requirements. The delivery area should allow the efficient and rapid unloading of vehicles and it should have facilities for the inspection and batch coding or maintaining batch integrity of incoming raw materials. Its organisation should allow the direct removal of these materials to storage areas and there should be facilities for the removal and disposal of secondary packaging such as cardboard boxes. If product is unpacked, clean containers may be required for product handling and storage prior to use. High-risk materials, which may contaminate other ingredients, should be handled in segregated areas that are capable of being effectively cleaned and are © 2008, Woodhead Publishing Limited

224

Chilled foods

not used for the storage of packaging that is removed after delivery. When packed pre-processed ingredients are designed for direct use in high-risk or high-care areas, it may be necessary to disinfect the outer packaging prior to admission. Ideally, these materials should be delivered to areas handling only low-risk ingredients. If facilities for this type of segregated handling do not exist, then products using direct inclusion of ingredients should be excluded from the design portfolio.

9.11.2 Storage areas Depending on the range of risks associated with incoming materials (e.g. carrying pathogens or earth) or their eventual destination in the factory (e.g. GMP or higher hygiene areas), a chilled products factory will require a number of designated storage areas. All these areas should be controlled for time and temperatures (chilled: 0–3 °C, or frozen: below –12 °C), and the layout of the store should allow access to the stored items and effective stock rotation. Batches of stored materials should be labelled so that their use-by dates and approvals for use are clear, and particular deliveries or production batches can be identified. All temperaturecontrolled areas should be fitted with reliable control devices, monitoring systems (to generate a record of conditions in the store) and an alarm system indicating loss of control or failure of services. A low humidity store should be provided for dry ingredients and packaging materials. Storage areas should be designed to accommodate sufficient product to meet factory needs, and services should be able to ensure that the designed conditions (such as 2–4 °C in chillers) are maintained during the working day, during high outside temperatures or at peak demand for cooling. The layout of racking and access to floors, walls and drains should allow cleaning to take place.

9.11.3 Raw material preparation and cooking areas The design of these areas should allow them to receive ingredients from storage areas and convert them into components or ingredients using a variety of techniques; layout should minimise hold-up of material and opportunities for mistakes. Preparation and cooking may take place in a single area if materials are for cookin-pack (Class 4) products, but when prevention of contamination is important (for Class 2 or 3 ready-to-eat products), cooking should be done in a separate area, with hygienic product exit routes to prevent contamination of the cooked product by contact with unprocessed material, airborne particles, dust, aerosols or personnel. Where physical separation cannot be achieved, cooked product should not be handled by personnel or equipment that has been in contact with uncooked material. In some cooking areas, the cooking vessels or ovens form the barrier between ‘dirty’ areas, i.e. those handling uncooked material, and ‘clean’ areas. Air flow should always be from ‘clean’ to ‘dirty’ areas, and the supply of air to steam extraction hoods should be designed to prevent contamination of cooked product; for example, by condensation. Where single-door ovens are used, there is an © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

225

increased risk of cross-contamination as it is not possible to segregate raw and cooked material effectively. Entry/exit areas from single-door oven areas need to be kept clean, not used for storage, with loading and unloading done by separate staff, so that opportunities for product contact are minimised. Areas designed for handling different classes of product should address different microbiological risks. Areas designed for short shelf-life (Class 1) products and products containing components which have not been reliably decontaminated (Class 2), should prevent recontamination of product with infectious pathogens and minimise contamination with spoilage micro-organisms. In areas handling ingredients cooked for long-life products (Class 3), designs should include stringent precautions to prevent re-contamination of heat-treated product with clostridial spores and infectious pathogens. These will include a forward flow layout, physical separation of process stages and control of airflow away from decontaminated product. Typical process routes for short and long shelf-life products are shown in Figures 9.1–9.4. Preparation rooms or areas may be chilled, but it is often impractical to chill cooking areas, and indeed this may increase problems associated with condensation.

9.11.4 Thawing of ingredients and raw materials Prior to use, it may be necessary to thaw frozen ingredients (e.g. meat and fish blocks or IQF vegetable). Processes should be designed to minimise the growth of pathogens, i.e. times should be restricted and the maximum surface temperature of the ingredient should not be within the pathogen’s growth range (i.e. should remain below 10–12 °C). Safe thawing may require special equipment such as microwave tempering units, running-water thawing baths or air thawing units. Designs should exclude thawing of perishable products at ambient temperatures as this can lead to the uncontrolled and unrecognised growth of hazardous micro-organisms. Where frozen ingredients are used for direct heat processing, it is important that the particles or pieces entering the cooking stage are of a uniform size, clumping is prevented and they have controlled minimum temperature, so that cold-spots are not accidentally created.

9.11.5 Hygienic areas Hygienic, not high-care, areas should be used for the assembly or processing of prepared foods made entirely of raw (Class 1), or of mixtures of raw and cooked (Class 2) components. Such areas should be designed and operated to prevent infectious pathogens becoming established and growing in them, by attention to a number of key requirements shown below. Temperature Hygienic areas should ideally be chilled to 10–12 °C or below to limit the potential for growth of Salmonella. However, chilling alone will not limit the growth of Listeria, as these organisms can grow at chill temperatures. A very effective, if © 2008, Woodhead Publishing Limited

226

Chilled foods

expensive, solution is to design these areas to maintain an environmental humidity below 55–65%, which is well below the minimum water activity for the growth of Listeria and also low enough to ensure relatively rapid drying of surfaces. However, if control of humidity is not possible, e.g. because an existing area is being used or has been upgraded, then hygienic controls should be concentrated in areas known to harbour Listeria and these should be effectively cleaned, disinfected and dried on at least a daily basis. To do this effectively is very demanding of time, resources, and training of hygiene operatives and management. Construction The design of hygienic storage and manufacturing areas should promote the effective management of hygiene. Building surfaces should be constructed of materials that are impervious to water, are easy to clean, and allow routine access to services from outside the area. After cleaning and disinfection, surfaces should have fewer than 10 micro-organisms per 9 cm2 and Enterobacteriaceae should not be recoverable. Floors should be sloped so that pools of water do not collect and remain on them. Drains should be designed so that food waste is not retained in them for long periods and the layout design should ensure drain entries are accessible. The direction of flow should ensure that material from areas handling contaminated materials does not flow into or through hygienic areas, and any drain traps used should be capable of being cleaned to the same standards as food processing equipment. All these precautions are designed to minimise the chances of product contamination during processing and to reduce the chances of infectious pathogens being present. Stock control Minimum amounts of material for processing should be kept in hygienic areas or in any associated chillers. Storage periods of materials known to contain infectious pathogens or to support the rapid growth of Listeria, such as prepared vegetables, should be minimised and operational procedures should be designed to minimise three things: (i) the numbers of Listeria or other infectious pathogens brought into the area (ii) the opportunities for their growth in the area, and (iii) the number of environments where they can survive between cleaning. During the time between cleaning and production, production areas may not be chilled Listeria and maybe other pathogens and spoilage bacteria may grow in food residues left on imperfectly cleaned equipment or in drains and chillers. Design and layout, and operating procedures, should minimise the opportunities for contamination of the product by equipment or personnel. Mixed raw and cooked components If the product range includes ready-to-eat products made from mixtures of raw and cooked components, the process design should ensure that any components known to have a high risk of containing pathogens or high numbers of spoilage bacteria © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

227

are excluded from manufacturing and storage areas until they have been effectively decontaminated. Some products, such as hard cheeses, are pre-processed so that they can be used safely in hygienic areas provided surface contamination is controlled. The required hygiene status of ingredients should be made clear at the design stage. It is important to remember that pathogens such as Listeria monocytogenes may sometimes survive processing (e.g. in spray-dried milk powder, and cheddar, cottage and camembert cheese: see listings by Doyle, 1988). Product designs should exclude higher-risk materials, such as prawns and other shellfish from warm waters and untreated herbs and spices, which may carry Salmonella. But they are sometimes used even though from time to time they will contain pathogens, and hence storage conditions and hygiene practices in manufacturing areas need to be designed with this in mind.

9.11.6 High-care areas General High-care areas are designed for the post-cook handling, cooling and assembly of ready-to-eat products made entirely of cooked (or otherwise decontaminated) components (Class 3). These materials have been freed of hazardous bacteria (such as cold-growing strains of Clostridium botulinum) and will be chilled, assembled and packaged, and may be further processed before packaging (e.g. sliced, cut or portioned cooked meat products, such as paté). The aim is to prevent re-contamination with food-poisoning bacteria and minimise re-contamination with spoilage bacteria. For these processes to yield safe products with a reliable long (up to 42 days) shelf-life, effective control of re-contamination after cooking must be consistently achieved. This leads to a number of specific requirements. Physical separation As a minimum, these areas may be designed and operated to prevent contamination by pathogens. Often, a higher standard of hygiene is used to limit product recontamination with spoilage micro-organisms, such as yeasts, moulds and lactic acid bacteria, and to ensure shelf-life is consistently met. These areas should be physically separated from production areas handling contaminated components and only materials (including both foodstuffs and packaging) which have been reliably decontaminated should be admitted. If such areas are not available, then product designs based on the hygienic assembly of heated ingredients should not be proposed. Chillers and cooling Cooling rates and the type of equipment to be used should be part of the process design. An integral part of high-care areas is blast coolers and chillers, designed to hygienically cool components (blast chills: James et al., 1987) or to maintain chill temperatures in previously cooled components. Production scheduling should allow cooling of hot, cooked product to be started as soon as possible after cooking. © 2008, Woodhead Publishing Limited

228

Chilled foods

Cooling rates, based on product size and equipment capability, should be designed to allow rapid cooling and prevent the growth of any surviving spore-forming bacteria. Some product designs will include post-heating assembly of products, therefore chillers will often receive unwrapped product and their hygiene (e.g. air quality and environmental hygiene) is critical. To prevent contamination, the cooling elements of fan-driven evaporator units must be designed and operated to limit the potential for product re-contamination by aerosols. If water is the chosen cooling medium – either as a spray or shower or in a bath for products with hermetically sealed containers – chlorination or some other disinfection procedure should be used to ensure that the product will not be re-contaminated by the cooling water. Stringent cleaning and disinfection systems should be used to ensure the hygiene of circulation or recirculation systems, including heat exchangers. Packs should be dried as soon as possible after cooling and manual handling of wet packs minimised. Air supply The environmental air supply should be filtered to remove particles in the 0.5–50 micron size range and the system should provide control of the air flow, from clean to dirty, by means of a slight overpressure, which prevents the ingress of untreated air. Air supply and heating and ventilation systems should be designed for easy access for inspection and cleaning.

9.12 Safe process design – unit operations for decontaminated products 9.12.1 Working surfaces for manual operations Many of the operations involved in preparation or product assembly will be carried out by hand on tables or other flat surfaces. These surfaces must be hygienic, cleanable and technically efficient, so that operations critical to meeting the product design (e.g. filleting or cutting to selected dimensions) can be done effectively.

9.12.2 Cutting and slicing Many chilled products, such as meats and patés, are prepared or cooked as blocks and sliced after cooking to make consumer packs. Slicers can be potent sources of contamination because they are usually mechanically complex, providing many inaccessible and uncleanable sites that can harbour bacteria. The design shelf-life of these products should take account of the potential for slicing equipment recontaminating products. If shelf-life studies show that shelf-life cannot be met, insight into the routes for re-contamination of product can be gained by auditing the machines for product and debris flow during operation, so that the sources and the risks of re-contamination can be identified and eliminated. © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

229

9.12.3 Transport and transfers From the time that a product is cooked, it can undergo many transfer stages from one production area to another before it is finally put into its primary packaging and contamination is excluded. Layout of process lines should be designed to minimise transfers and opportunities for re-contamination. Containers If products are cooked in open vessels, they are likely to be unloaded into trays or other containers for cooling. These containers should not contaminate product, and their design (i.e. shape, size and loading) should ensure rapid cooling is possible. Generally, stainless steel or aluminium trays are more hygienic than plastic ones, which become more difficult to clean as their surfaces are progressively scratched in use. Plastics also have slower rates of heat conduction than either stainless steel or aluminium, and therefore metal trays are to be preferred. Belts In most complex production lines, transfer or conveyor belts may be used for transporting both unwrapped product or product in intermediate wrappings from one process stage to another. Often the product type will determine the type of belt that has to be used. Belts may be found in tunnels or spiral equipment, such as cookers, ovens and coolers. Although many types of belting material are used, they can be broken down into two broad groups: solid and mesh or link. Fabric conveyors are used only for transport under chill or ambient conditions and should be made of a hygienic material that does not absorb water and has a smooth surface finish. Solid and mesh belts are used for heating or cooling (by conduction) and are usually made of stainless steel, which is hygienic and can be heated or cooled indirectly. Belts can be cleaned in place by in-line spray cleaners, which can provide both cleaning and rinsing and may incorporate a drying stage using an air knife or vacuum system. The hygienic problems of conveyor belts are usually associated with either unhygienic design of drive axles and the beds or frames supporting them, or poor maintenance. Again, before the product design is finalised, the type and hygiene of belts available in the plant should be assessed to ensure design requirements can be met. With all forms of belt, hygienic performance becomes more difficult to achieve if routine engineering maintenance is not carried out correctly and the belt becomes damaged or frayed during use. Specialised belts are often used in product delivery, packaging and collation systems. In such equipment, incorrect setting, or use with fragile products, will increase the quantity of product waste generated, so that even a properly designed and operated cleaning system will not keep the system in a hygienic condition. Product characteristics, such as stickiness and crumbliness, should therefore be considered when transfer systems are designed, so that the generation of debris and, in turn, hygiene problems, are minimised.

9.12.4 Dosing and pumping The realisation of many product designs is limited by the ability to fill or portion © 2008, Woodhead Publishing Limited

230

Chilled foods

them into packs. Most chilled products are sold in weight-controlled packs, often with individual ingredients in a fixed ratio to one another, for example meat and sauce. Where the ingredients are liquid or include small (ca. 5 mm) particles suspended in a liquid, they may be dosed using filling heads or pump systems. If this type of equipment is used for dosing decontaminated materials, then its hygiene and operation are critical to product safety and shelf-life. Fillers may interact with products, for example damaging particulates, and product designs including delicate or soft materials may be limited in the filling equipment available. Dosing and filling systems may be operated at cold (below 8–10 °C), intermediate (20–45 °C) or hot (above 60 °C) temperatures. Products should be designed to flow at the chosen fill temperature. The most hazardous fillers run at intermediate temperatures and allow the growth of food-poisoning bacteria, and unless the food producer is completely confident of the hygiene of the equipment and is prepared for frequent cleaning/disinfection breaks, these temperatures should not be used, even in Class 3 areas. When hot filling is used, control of the minimum filling head temperatures is critical to safety, and prolonged holding may reduce product quality. To ensure minimum temperatures are met, target temperatures for the fill are usually set well above the growth maximum of food-borne pathogens (e.g. 65 °C), as this provides a cushion for cooling during dosing or when there are breaks in production and the flow of product is halted. If packs are filled for in-pack pasteurisation, dosing systems need to deliver an accurately controlled and consistent amount of product to ensure that packs with uniform heating characteristics and headspaces are made. Many pick and place operations are manual and these require careful design to ensure contamination is minimised.

9.12.5 Packaging Most chilled products are sold in a packaged form and hence packaging needs to be defined in the product and process design. Primary packaging has a number of roles, including protecting the product from microbial contamination, providing suitable barriers (e.g. to oxygen and water or retaining a protective atmosphere, see Ahvenainen et al., 1997) and ensuring heating characteristics are predictable; tray packs are often used to provide a container for oven cooking or re-heating by the customer. Packs also need to be chosen for technical reasons, such as running at line speeds on fillers and sealers. There is consumer pressure to reduce quantities of material used in packs, but designs must ensure that the chosen packaging does not have compromised technical properties. Outer, or secondary, packaging plays very important roles in protecting primary packs, displaying the product and providing use instructions; design of the complete packaging system is a key element of the design. Short shelf-life products may be an exception to the general preference for bacteria-tight packs. Some of these packs have a crimped seal and therefore a risk of product contamination unless overwrapped in a sealed container. If packaging films or trays come into direct contact with the products, they must not contaminate them either chemically or microbiologically. When chilled © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

231

products are overwrapped or placed on pallets, there are only limited opportunities for reducing their temperatures, as the surface area to volume ratio is unfavourable to rapid cooling because rates of heat penetration through product and packaging are low. Therefore, it is essential that product in its primary packaging is at the target temperature prior to secondary packaging and palletisation. As there is also pressure to increase the efficiency of logistics systems, pack shape should be designed to maximise weight per unit volume.

9.13 Control systems 9.13.1 Instrumentation and calibration Controls and control systems for equipment should be designed to provide the quality specified by the design. They should ensure equipment works reliably within the specified limits; hence, reliable instrumentation or measurement procedures need to be defined as part of the design, and sensors should be correctly located and calibrated during commissioning to produce meaningful outputs that are used to control process conditions (such as during pasteurisation, chilling or storage) or for monitoring compliance with specifications. Sensors and their associated instruments may be in-line (e.g. oven, heat exchanger or fridge thermometers), at the side of the line (e.g. drained weight apparatus, salt or pH meters), or in the laboratory (e.g. colour measurement or nitrogen determination). Wherever controls are situated, they need to be calibrated and maintained, with sensors kept free of fouling because this may produce erroneous signals. If control is manual, operatives need measurement and recording procedures.

9.13.2 Process monitoring, validation and verification To ensure that a design is realised, there is usually a period of intensive monitoring during and after commissioning of a new product or process. All operations within the supply chain need to be covered to ensure that the whole chain performs within the agreed limits. Data from control systems and product assessments should be used for validation and to produce management and operative information, and trend analyses. Critical sensors, such as thermocouples, should be checked during this period to ensure that there are no errors in temperature readings (Sharp, 1989). As businesses increasingly concentrate on their core activities, vertical integration within the supply chain is becoming uncommon and validation of product designs becomes more complex as it involves a number of different suppliers. Increasing reliance on suppliers further up the supply chain means products may contain pre- or part-processed materials from several suppliers or producers; hence, the performance of each one needs to be validated individually to ensure that the overall design is met. Unsuitable process, hygiene or material features should be uncovered by validation procedures, as unsafe or poor quality products can result if the requirements of the design or HACCP plan are not met or controls © 2008, Woodhead Publishing Limited

232

Chilled foods

are not working effectively. Verification is then a continuing activity, systematically analysing supply chain performance by examining process and product-related data against the design requirements. These data may also be compared with specifications and other technical agreements that form the customer requirements. 9.13.3 Process and sample data As chill foods have relatively short shelf-life and are often distributed immediately after manufacture, results of realistic levels of microbiological product sampling cannot be used for the assurance of safety and the time taken to obtain microbiological results in significantly reduces the time available for sale. Microbiological results should be used only for supplier monitoring, trend analysis of process control and hygiene, and for ‘due diligence’ purposes. Data showing the control of CCPs and prerequisites should be taken at a defined frequency, analysed and kept for a period equal to at least the shelf-life plus the period of use of the product to verify the performance of the supply chain. Process control records may be generated and retained according to the framework proposed in the ISO 9000 documents on quality management. The provision of conformance samples to track long-term changes in quality is difficult because of the short shelf-life of the products. Some manufacturers rely on the descriptions in the product design and others retain frozen samples and accept the quality change caused by freezing. End-of-shelf-life samples can be taken and scored for their sensory or physical attributes against fixed scales or parameters. 9.13.4 Training, operatives, supervisors and managers The process design will suggest the key areas for staff training. Relevant training is essential as staff make an essential contribution to the control of product quality and safety. When a new design with new equipment or procedures is implemented, there should be a period of training or re-training to allow staff to make their full contribution to assuring the safe manufacture of high quality products (Mortimore and Smith, 1998; Engel ,1998). After training, staff should at least understand the critical aspects of hygiene for the product, its composition, presentation and control. Because many manufacturing operations will be done under conditions of ‘high care’ and involve team work, all staff must understand the reasons for specified hygiene standards and procedures. This is critical for ready-to-eat products, and staff and supervisors should have defined responsibilities to ensure that low quality or unsafe products are not made (Griffith, 2002). The role of external HACCP consultants has been identified (White, 1998) as being especially useful to small food companies in helping them put together effective HACCP programmes and training of staff. For such training to be effective, consultants must understand the product and process design. 9.13.5 Process auditing Auditing is the collection of data or information about a process or factory by a visit © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

233

to the premises involved. Inspection and auditing may be used to determine whether suppliers can meet the requirements of the product design and if the HACCP plan is correctly established, implemented, and is suitable for ensuring product safety. The importance of auditing is increased where safety relies on the correct functioning of many units in the supply chain (van Schothorst, 1998; Sperber, 1998). An audit needs to see the line operating under normal production conditions and does more than inspect records and the facilities available – the assessment should extend beyond the HACCP plan and include the way that the product and process designs are realised. Inspection of the design HACCP is an important linking step. Internal or external auditors may do the auditing; a checklist may be used and often a scoring system or noting of non-compliances may be applied if there is a customer–supplier relationship. A more recent development is supplier self-auditing. This involves the development of the supplier–customer relationship on a partnership basis. It starts with a clear statement of the commercial and quality objectives and the resources and materials involved. Next, an evidence package covering specifications and records is agreed between the supplier and customer. Because of the variability of quality parameters and process controls, limits based on the product design and a mechanism for challenge must be agreed to prevent false rejection of product and encourage the realistic management of any risks or non-compliances. Successful operation of such a system relies on the identification and allocation of responsibilities and ownership of technical factors within both organisations, so that the person best placed to know and make reliable decisions can always be identified. This type of audit may have better cost/benefits than the traditional audit visit and produces information for decision-making on a continuing basis if it is carried out by externally accredited auditors; their ability to provide a valid assessment may be limited by reluctance to provide them with access to commercially sensitive data on processing or formulation.

9.14 Conclusions Sales of prepared chilled foods continue to increase, particularly in Europe. There is a constant stream of new products to meet changing consumer requirements for convenience, healthier dishes and ingredients, and less severe processing and preservation. The focus of most new designs is range extension and ingredient substitution. There are also pressures to reduce the energy input and use manufacturing and logistics more efficiently. Supply chains, and especially sourcing of raw materials, are becoming global activities and manufacturers are tending to concentrate on their core activities. Retailers are becoming more dominant in setting manufacturing, and especially quality and safety, requirements and also consolidating their operations. Therefore, basing each product on a sound design that can be critically examined before large-scale production is started is essential to ensuring safety and quality and preventing commercial losses. To help this, increasing numbers of modelling tools are available to assess the likely impact of © 2008, Woodhead Publishing Limited

234

Chilled foods

processing and storage on microbial numbers and the lethality of heat processes. In the UK, there is currently little reliance on preservatives such as sorbate and benzoate to extend shelf-life; elsewhere (e.g. the USA), preservatives are added to reduce the frequency of deliveries by extending shelf-life and provide some protection against pathogen growth in the event of temperature abuse. To minimise their sensory impact, preservatives are used as combinations to give synergistic effects (e.g. reduced pH, water activity (Aw) or chemical or natural preservatives) inhibiting the growth of pathogenic and spoilage micro-organisms. Relatively small changes to two hurdles (e.g. acidity + lowered Aw) may be as effective as large and more noticeable changes to one hurdle. The effectiveness of combination preservation systems is normally established by challenge and shelf-life testing, as modelling cannot accurately predict the effect of combination preservation systems in food. At the same time as consumers seek more ‘naturalness’, assurance of safety remains essential and, as such, is paramount for producers and retailers. It is here that soundly validated designs, developed using modern processing and distribution techniques backed up by HACCP and other control procedures (Mayes, 1992), can more than compensate for milder processing that could allow stability and safety to be lost. Although complex in detail, as indicated above, the basic elements of effective design and safe processing of chilled foods are few. They include:

• the reliable identification and anticipation of hazards • a full description of required product characteristics • development and validation of product and process designs to indicate supply chain, formulation and customer requirements

• selection and approval of a supply chain • development of controls, specifications and procedures • validation that processes can provide predictable inactivation of micro-organisms and hygiene precautions can prevent recontamination

• validation that the logistics chain can meet limits to control quality and limit • •

growth of any micro-organisms remaining in the food (e.g. by refrigeration and stock rotation) choice of supply chains that are energy efficient, minimise losses and ensure traceability provision of validated, clear consumer use instructions that give products meeting expectations.

Quality and cost improvements may come through the availability of new processing techniques, such as ultra high pressure to ‘pressure-pasteurise’ foods, or the use of microwaves to give volumetric heating. UHP is likely to be applicable only to products in which (pressure tolerant) bacterial spores are not a problem, e.g. short shelf-life chilled products, low pH jams and fruit juices (Hoover et al., 1989; Smelt ,1998). Irradiation (Haard, 1992; Grant and Patterson, 1992; McAteer et al., 1995) remains unlikely to meet with consumer approval. © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

235

9.15 References ACDP (Advisory Commitee on Dangerous Pathogens) (1996) Microbiological Risk Assess-

ment: An interim report, HMSO, London. (2000) ACM 500 Advisory Committee on the Microbiological Safety of Food. Clostridium botulinum and the ACMSF advice and industry code of practice on vacuum and modified atmosphere packaged products. (http://www.food.gov.uk/multimedia/ pdfs/adhocB13R0001). ADAIR C AND BRIGGS P A (1993) The concept and application of expert systems in the field of microbiological safety, Journal of Industrial Microbiology, 12, 263–267. ADAMS C E (1991) Applying HACCP to sous vide products, Food Technology, 45(4), 148– 9, 151. AHVENAINEN R, EILAMO M AND HURME E (1997) Detection of improper sealing and quality deterioration of modified atmosphere packed pizzas by a colour indicator, Food Control, 8, 177–184. ANON (1982) Guidelines for the handling of chilled foods, Institute of Food Science and Technology, London, 1–27. ANON (1984) Guidelines on precooked chill foods, HMSO, London. ANON (1986) Health service catering hygiene, Department of Health and Social Security, HMSO, London. ANON (1989) Chilled and Frozen. Guidelines on cook–chill and cook–freeze catering systems, Department of Health, HMSO, London. BAIRD-PARKER A C (1994) Use of HACCP by the chilled food industry, Food Control, 5(3), 167–170. BEN-EMBAREK P K AND HUSS H H (1993) Heat resistance of Listeria monocytogenes in vacuum packaged pasteurised fish fillets, International Journal of Food Microbiology, 20, 85–95. BETTS G D (1996) Code of practice for the manufacture of vacuum and modified atmosphere packaged chilled foods with particular regard to the risks of botulism, Guideline No.11, Campden and Chorleywood Food Research Association. BETTS G (1997) Predicting microbial spoilage, Food Processing, UK, 66, 23–24. BETTS G D AND GAZE J E (1995) Growth and heat resistance of psychrotrophic Clostridium botulinum in relation to ‘sous vide’ products, Food Control, 6, 57–63. BOGH-SORENSEN L AND OLSSON P (1990) The chill chain. In: Gormley, T. R. (ed.) Chilled foods: The state of the art, Elsevier Applied Science, London, 245–267. BRACKETT R E (1992) Microbiological safety of chilled foods: Current issues, Trends in Food Science and Technology, 3, 81–85. BROWN M AND STRINGER M (2002) Microbiological risk assessment in food processing, Woodhead Publishing, Cambridge. ISBN 1 85573 587 7. BROWN M H, DAVIES K W, BILLON C M, ADAIR C AND MCCLURE P J (1998) Quantitative microbiological risk assessment: Principles applied to determining the comparative risk of salmonellosis from chicken products, Journal of Food Protection, 61(11), 1446–1453. BUCHANAN R L (1995) The role of microbiological criteria and risk assessment in HACCP, Food Microbiology, 12, 421–424. CCFRA (1997) Guidelines for Air Quality for the Food Industry, Guideline No. 12. CCFRA (2006) Guidelines on good manufacturing practice for heat processed flexible packaging, CCFRA Guideline No. 50. CFA (2006) Best Practice Guidelines for the Production of Chilled Foods, Chilled Food Association (CFA), The Stationery Office, London. ISBN 9780117022836 CFDRA (1990) Evaluation of the shelf-life for chilled foods, Campden Food and Drink Research Association Technical Manual No. 28. CFDRA (1992) Food pasteurisation treatments: Part 1 Guidelines to the types of food products stabilised by pasteurisation treatments; Part 2 Recommendations for the design ACMSF

© 2008, Woodhead Publishing Limited

236

Chilled foods

of pasteurisation processes, Campden Food and Drink Research Association Technical Manual No. 27. CHANDARY V, JOHNS P, GUPTA R P AND SINHA R P (1990) Lethal effects of pulsed high electric fields on food-borne pathogens, Journal of Dairy Science, 73(Suppl. 1), D92. CHURCH I J AND PARSONS A L (1993) Review: Sous vide cook–chill technology, International Journal of Food Science and Technology, 28, 563–574. CODEX ALIMENTARIUS (1999) Recommended International Code of Practice – General Principles of food hygiene – CAC/RCP 1- 1969 Rev 3, Amd 1999. COOTE P J, HOLYOAK C D AND COLE M B (1991) Thermal inactivation of Listeria monocytogenes during simulated microwave heating, Journal of Applied Bacteriology, 70, 489–494. DE ROEVER C (1998) Microbiological safety evaluations and recommendations on fresh produce, Food Control, 9, 321–347. DOYLE M P (1988) Effect of environmental and processing conditions on Listeria monocytogenes, Food Technology, 42, 169–171. ELLIOT P H (1996) Predictive microbiology and HACCP, Journal of Food Protection, Supplement, 48–53. ENGEL D (1998) Teaching HACCP – Theory and practice from the trainer’s point of view, Food Control, 9(2/3), 137–139. EVANS J, RUSSELL S AND JAMES S (1996) Chilling of recipe dish meals to meet cook–chill guidelines, International Journal of Refrigeration, 19, 79–86. FARQUHAR J AND SYMONS H W (1992) Chilled food handling and merchandising: A code of recommended practices endorsed by many bodies, Dairy, Food and Environmental Sanitation, 12, 210–213. FENLON D R, WILSON J AND DONACHIE W (1996) The incidence and level of Listeria monocytogenes contamination of food sources at primary production and initial processing, Journal of Applied Bacteriology, 81, 641–650. FOOD STANDARDS AUSTRALIA NEW ZEALAND (2002) Listeria monocytogenes in cooked prawns. A Microbiological Survey Report. Technical Report Series No. 26, FSANZ. FRAILE P AND BURG P (1998a) Influence of convection heat transfer on the reheating of a chilled ready cooked dish in an experimental superheated steam cell, Journal of Food Engineering, 33, 263–280. FRAILE P AND BURG P (1998b) Re-heating of a chilled dish of mashed potato in a superheated steam oven, Journal of Food Engineering, 33, 57–80. GAZE J E AND BETTS G D (1992) Food pasteurisation treatments. CFDRA Microbiology Panel Pasteurisation Working Party Technical Manual, No. 27. Campden Food and Drink Res. Association, Chipping Campden, UK. GAZE J E, BROWN G D, GASKELL D E AND BANKS J G (1989) Heat resistance of Listeria monocytogenes in homogenates of chicken, beef steak and carrot, Food Microbiology, 6, 251–259. GILBERT R J, MILLER K L AND ROBERTS D (1989) Listeria monocytogenes in chilled foods, Lancet i, 502–503. GILL C O, RAHN K, SLOAN K AND MCMULLEN L M (1997) Assessment of the hygienic performances of hamburger patty production processes, International Journal of Food Microbiology, 36, 171–178. GLEW G (1990) Precooked chilled foods in catering. In: Zeuthen, P., Cheftel, J. C., Eriksson, C., Gormley, T. R., Linko, P. and Paulus, K. (eds) Processing and quality of foods: Vol. 3. Chilled foods: The revolution in freshness, Elsevier Applied Science, London, 3.31–3.41. GOULD G W (1989) Heat-induced injury and inactivation. In: Gould, G. W. (ed.) Mechanisms of action of food preservation procedures, Elsevier Applied Science, London, 11–42. GOULD G W (1992) Ecosystem approaches to food preservation. In: Ecosystems: Microbes: Food. Soc. Appl. Bact. Symp. Series 20. Suppl. to, Journal of Applied Bacteriology, 72, 585. GRAHAM A F, MASON D R AND PECK M W (1996) Predictive model of the effect of temperature, pH and sodium chloride on growth from spores of non-proteolytic Clostridium botulinum. International Journal of Food Microbiology, 31, 69–85. © 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

237

GRANT I R AND PATTERSON M F (1992) Sensitivity of foodborne pathogens to irradiation in

the components of a chilled ready meal, Food Microbiology, 9, 95–103. GRIFFITH C (2002) Good practices for food handlers and consumers. In: eds Blackburn C. de

W. and McClure P. J. (eds), Foodborne Pathogens, Woodhead Publishing, Cambridge, UK. HAARD N F (1992) Technological aspects of extending prime quality of seafood: A review, Journal of Aquatic Food Product Technology, 1, 9–27. HANSEN T B AND KNOCHEL S (1996) Thermal inactivation of Listeria monocytogenes during rapid and slow heating in sous vide cooked beef. Letters in Applied Microbiology, 22, 425– 428. HEDDLESON R A, DOORES S, ANANTHESWARAN R C AND KUHN G D (1996) Viability loss of Salmonella species, Staphylococcus aureus, and Listeria monocytogenes in complex foods heated by microwave energy, Journal of Food Protection, 59, 813–818. HENDRICKX M E G, KNORR D AND BARBOSA-CÁNOVAS G V (2002) Ultra High Pressure Treatment of Foods, Springer, ISBN-13: 978-0306472787 HOLYOAK C D, TANSEY F S AND COLE M B (1993) An alginate bead technique for determining the safety of microwave cooking, Letters in Applied Microbiology, 16, 62–65. HOOVER D G, METRICK C, PAPINEAU A M, FARKAS D AND KNORR D (1989) Biological effects of high hydrostatic pressure on food micro-organisms, Food Technology, 3, 99–107. HUMPHREY T J, MARTIN K W, SLADER J AND DURHAM K (2001) Campylobacter spp. in the kitchen: Spread and persistence. Journal of Applied Microbiology, 90, Supplement 6, 115–120. ICMSF (1988) Micro-organisms in foods. IV. HACCP in microbiological safety and quality, Blackwell Scientific Publications, London. IFST (2007) Food and drink good manufacturing practice: A guide to its responsible management, 5 Rev Ed edition, Institute of Food Science and Technology, UK, ISBN-13: 978-0905367200 IMPE J F-VAN, NICOLAI B M, MARTENS T, BAERDEMAEKER J-DE AND VANDEWALLE J (1992) Dynamic mathematical model to predict microbial growth and inactivation during food processing, Applied and Environmental Microbiology, 58, 2901–2909. JAMES S J AND BAILEY C (1990) Chilling systems for foods. In: Gormley, T. R. (ed). Chilled Foods: The state of the art, Elsevier Applied Science, London, 1–35. JAMES S J, BURFOOT D AND BAILEY C (1987) The engineering aspects of ready meal production. In: Engineering innovation in the food industry, Institute of Chemical Engineering, Bath, 21–28. JAYKUS L A (1996) The application of quantitative risk assessment to microbial safety risks, Critical Reviews in Microbiology, 22, 279–93. JOUVE J L, STRINGER M F AND BAIRD-PARKER A C (1998) Food Safety Management Tools. Report ILSI Europe Risk Analysis in Microbiology Task Force, 83 Avenue E. Mounier, B-1200 Brussels. JUNEJA V K AND MARMER B S (1996) Growth of Clostridium perfringens from spore inocula in sous-vide turkey products, International Journal of Food Microbiology, 32, 115–123. KENNEDY G (1997) Application of HACCP to cook–chill operations, Food Australia, 49, 65– 69. KERR K G AND LACEY R W (1989) Listeria in cook–chill food, Lancet ii, 37–38. LABUZA T P AND BIN-FU (1995) Use of time/temperature integrators, predictive microbiology, and related technologies for assessing the extent and impact of temperature abuse on meat and poultry products, Journal of Food Safety, 15, 201–227. LEISTNER L (1985) Hurdle technology applied to meat products of the shelf stable product and intermediate moisture types. In: Simatos, D. and Multon, J. L. (eds) Properties of water in foods. Martinus Nijhoff Publishers, Dordrecht, 309–329. LIVINGSTON G E (1985) Extended shelf-life chilled prepared foods, J. Foodserv. Syst., 3, 221–230. LUND B M, KNOX M R AND COLE M B (1989) Destruction of Listeria monocytogenes during microwave cooking, Lancet i, 218. © 2008, Woodhead Publishing Limited

238

Chilled foods

(1989) The heat resistance of Listeria monocytogenes, Letters in Applied Microbiology, 9, 89–94. MACKEY B M AND DERRICK C M (1987) Changes in the heat resistance of Salmonella typhimurium during heating at rising temperatures, Letters in Applied Microbiology, 3, 1316. MAYES T (1992) Simple users’ guide to the hazard analysis critical control point concept for the control of food microbiological safety, Food Control, 3, 14–19. MCATEER N J, GRANT I R, PATTERSON M F, STEVENSON M H AND WEATHERUP S T C (1995) Effect of irradiation and chilled storage on the microbiological and sensory quality of a ready meal, International Journal of Food Science and Technology, 30(6), 757–771. MCMEEKIN T A, BARANYI J, BOWMAN J, DALGAARD P, KIR, M, ROSS T, SCHMID S AND ZWIETERING M H (2006) Information systems in food safety management, International Journal of Food Microbiology, 112, 181–194. MEMBRE J-M, AMÉZQUITA A, BASSETT J, GIAVEDONI P, BLACKBURN C DE W AND GORRIS L G M (2006) A probabilistic modelling approach in thermal inactivation: estimation of post process Bacillus cereus spore prevalence and concentration, Journal of Food Protection, 69(1), 118–129 MICHELS M J M (1979) Determination of heat resistance of cold-tolerant spore formers by means of the ‘Screw-cap tube’ technique. In: Cold tolerant microbes in spoilage and the environment, SAB Technical Series 13, Academic Press. London and New York, 37–50. MILLER A J, WHITING R C AND SMITH J L (1997) Use of risk assessment to reduce listeriosis incidence, Food Technology, 51(4), 100–103. MORTIMORE S E AND SMITH R (1998) Standardized HACCP training: Assurance for food authorities, Food Control, 9(2/3), 141–145. MOSSEL D A A AND STRUIJK C B (1991) Public health implication of refrigerated pasteurised (‘sous-vide’) foods, International Journal of Food Microbiology, 13, 187–206. MOSSEL D A A, VAN NETTEN P AND PERALES I (1987) Human listeriosis transmitted by food in a general medical–microbiological perspective, Journal of Food Protection, 50, 894–895. NFPA (NATIONAL FOOD PROCESSORS ASSOCIATION) (1993) Guidelines for the Development, Production, Distribution and Handling of Refrigerated Foods, NFPA, New York. NOTERMANS S, DUFRENNE J AND LUND B M (1990) Botulism risk of refrigerated, processed foods of extended durability, Journal of Food Protection, 53, 1020–1024. NOTERMANS S, GALLHOFF G, ZWEITERING M H AND MEAD G C (1995) The HACCP concept: Specification of criteria using risk assessment, Food Microbiology, 12, 81–90. NOTERMANS S, MEAD G C AND JOUVE J L (1996) Food products and consumer protection: A conceptual approach and a glossary of terms, International Journal of Food Microbiology, 30, 175–185. PANISELLO P J AND QUANTICK P C (1998) Application of food MicroModel predictive software in the development of Hazard Analysis Critical Control Point (HACCP) systems, Food Microbiology, 15(4), 425–349. PECK M W (1997) Clostridium botulinum and the safety of refrigerated processed foods of extended durability, Trends in Food Science and Technology, 8, 186–192. PIN C AND BARANYI J (1998) Predictive models as means to quantify the interactions of spoilage organisms, International Journal of Food Microbiology, 41, 59–72. REILLY P J A AND TWIDDY D R (1991) Salmonella and Vibrio cholerae in cultured tropical prawns. FAO Fisheries Report (FAO). 0429-9337, No. 470 (Suppl.) RUTHERFORD N, PHILLIPS B, GORSUCH T, MABEY M, LOOKER N AND BOGGIANO R (1995) How indicators can perform for hazard and risk management in risk assessment of food premises, Food Science and Technology Today, 9(1), 19–30. RYYNANEN S AND OHLSSON T (1996) Microwave heating uniformity of ready meals as affected by placement, composition, and geometry, Journal of Food Science, 61, 620–624. SAGE J R AND INGHAM S C (1998) Survival of Escherichia coli O157:H7 after freezing and thawing in ground beef patties, Journal of Food Protection, 61, 1181–1183. SHARP A K (1989) The use of thermocouples to monitor cargo temperatures in refrigerated freight containers and vehicles, CSIRO Food Research Quarterly, 49, 10–18. MACKEY B M AND BRATCHELL N

© 2008, Woodhead Publishing Limited

Microbiological hazards and safe design

239

(1998) Recent advances in the microbiology of high pressure processing, Trends in Food Science and Technology, 9, 152–158. SMITH J P, TOUPIN C, GAGNON B, VOYER R, FISET P P AND SIMPSON M V (1990) A hazard analysis critical control point approach (HACCP) to ensure the microbiological safety of sous vide processed meat/pasta product, Food Microbiology, 7, 177–198. SNYDER O P JR (1992) HACCP – an industry food safety self-control program. XII. Food processes and controls. Dairy, Food and Environmental Sanitation, 12, 820–823. SPERBER W H (1998) Auditing and verification of food safety and HACCP, Food Control, 9, 157–162. STOFOROS N G AND TAOUKIS P S (1998) A theoretical procedure for using multiple response time–temperature integrators for the design and evalua-tion of thermal processes, Food Control, 9, 279–287. TASSINARI A D R AND LANDGRAF M (1997) Effect of microwave heating on survival of Salmonella typhimurium in artificially contaminated ready-to-eat foods, Journal of Food Safety, 17, 239–248. TURNER B E, FOEGEDING P M, LARICK D K AND MURPHY A H (1996) Control of Bacillus cereus spores and spoilage microflora in sous vide chicken breast, Journal of Food Science, 61, 217–219, 234. UK ADVISORY COMMITTEE ON THE MICROBIOLOGICAL SAFETY OF FOOD (1992) Report on vacuum packaging and associated processes ISBN 0-11-321558-4, HMSO, London. USDA (1988) Time–temperature guidance – Cooling heated produce, Food Safety and Inspection Directive, 71, 110–113. VAN NETTEN P, MOSSEL D A A AND VAN DE MOOSDIJK A (1990) Psychrotrophic strains of Bacillus cereus producing enterotoxin, Journal of Applied Bacteriology, 69, 73–79. VAN REMMEN H H J, PONNE C T, NIJHUIS H H, BARTELS P V AND KERKHIF P J A M (1996) Microwave heating distributions in slabs, spheres and cylinders with relation to food processing, Journal of Food Science, 61(6), 1105–1113, 1117. VAN SCHOTHORST M (1998) Introduction to auditing, certification and inspection, Food Control, 9, 127–128. WALKER S J (1990) Growth characteristics of food poisoning organisms at suboptimal temperatures. In: Zeuthen, P., Cheftel, J. C., Eriksson, C., Gormley, T. R., Linko, P. and Paulus, K. (eds) Processing and quality of foods Vol. 3. Chilled foods: The revolution in freshness, Elsevier Applied Science, London, 3.159–162. WALKER S J AND STRINGER M F (1987) Growth of Listeria monocytogenes and Aeromonas hydrophila at chill temperatures, Campden Food Preservation Research Association Tech. Memorandum No. 4652. WALKER S J AND STRINGER M F (1990) Microbiology of chilled foods. In: Gormley, T. R. (ed.) Chilled foods: The state of the art, Elsevier Applied Science, London, 269–304. WALLACE C AND WILLIAMS T (2001) Pre-requisites: A help or a hindrance to HACCP? Food Control, 12(6), 235–240. WHITE L (1998) The role of the HACCP consultant, International Food Hygiene, 9, 29–30. WHO (2003) Hazard Characterization for Pathogens in Food and Water, Microbiological Risk Assessment Series, No. 3, WHO, ISBN 92 4 156237 4. ZWIETERING M H AND HASTING A P M (1997) Modelling the hygienic processing of foods – A global process overview, Food and Bioproducts Processing, 75(C3), 159–167. ZWIETERING M H, KOOS J T-DE, HASENACK B E, WIT J C-DE AND RIET K-VAN’T (1991) Modelling of bacterial growth as a function of temperature, Applied and Environmental Microbiology, 57, 1094–1101. ZWIETERING M H, CUPPERS H G A M, WIT J C-DE AND RIET, K-VAN’T (1994a) Evaluation of data transformations and validation of a model for the effect of temperature on bacterial growth, Applied and Environmental Microbiology, 60, 195–203. ZWIETERING M H, CUPPERS H G A M, WIT J C-DE AND RIET K-VAN’T (1994b) Modelling of bacterial growth with shifts in temperature, Applied and Environmental Microbiology, 60, 204–213. SMELT J P P M

© 2008, Woodhead Publishing Limited

10 Non-microbiological hazards and safe process design R. W. R. Crevel, Unilever Safety and Environmental Assurance Centre, UK

10.1 Introduction While food is often assumed to be safe, it can result in adverse health effects. These may arise for several reasons, including the particular nutritional characteristics of a food, the presence of specific additives or contaminants, residual natural toxicants in the food, and the physical characteristics of the food. This chapter addresses non-microbiological hazards arising from the deliberate or inadvertent presence of substances in the food, which may render it injurious to health. Toxicological hazards from foods can arise at any point during the manufacture of a food product, from raw materials, through to processing and storage. The equipment used to produce foods can also contribute to possible toxicity through contact between the machinery and the product, and in some cases different products manufactured on the same equipment. Finally, the way the prepared food product is stored could give rise to toxicological hazards. The possibility that hazards arise at any point along the food chain is recognised in current approaches to food safety management, which lay heavy emphasis on integrated food safety management systems, as detailed for instance in ISO 22000:2005 or in the European Commission’s White Paper on Food Safety (CEC, 2000). The health consequences that may occur from these hazards range from short-term acute illness, sometimes severe or even fatal, to effects which manifest themselves only after a long period of exposure, up to a lifetime in some instances. The same substance can provoke both types of effect, depending on concentration, as is the © 2008, Woodhead Publishing Limited

Non-microbiological hazards and safe process design

241

case for many heavy metals, for instance. A food product may also prove to be unsafe only for a specific group of consumers. Food-allergic consumers constitute an important example of such a group, which will be discussed in detail, given its public health importance. However, similar examples, although differing in the danger posed to those affected, include individuals with coeliac disease and those suffering from intolerances of metabolic origin, such as lactose intolerance or galactosaemia. In the case of any hazard, a key requirement from a public health perspective is to assess the risk that is posed, which will need consideration not only of the possible adverse health effects and the dose at which they occur, but also of the exposure within the population. Optimising public health outcomes can be fraught with difficulty in the case of nutrients which have a beneficial effect in some sections of the population while posing a risk only in others. This chapter will provide an overview of toxicological considerations relating to foods and how they may be addressed, and then examine whether specific hazards could be exacerbated or attenuated in the context of chilled foods.

10.2 Chilled foods The Institute of Food Science and Technology (IFST) defined chilled foods as ‘perishable foods which, to extend the time during which they remain wholesome, are kept within specified ranges of temperature above –1 °C’ (IFST, 1990). The UK Chilled Foods Association restricted the definition to ‘prepared foods’ which, ‘for reasons of safety and/or quality, are designed to be stored at refrigerated temperatures throughout their entire life’. Chilled foods can thus be designed ready to eat, to be reheated or to be cooked.

10.3 Definition and principles of food safety Recent standards and regulations have defined food safety. For instance, ISO 22000:2005 defines food safety as the concept that food will not cause harm to the consumer when it is prepared and/or eaten according to its intended use. In this definition, it notes that food safety refers specifically to the occurrence of food hazards and does not include adverse effects that may result from nutritional considerations, in other words nutritional imbalances. According to the FAO– WHO (2003), ‘food safety refers to all those hazards, whether chronic or acute, that may make food injurious to the health of the consumer’ and ‘is not negotiable’. The European Union’s Food Law (Regulation (EC) 178/2002) elaborates the concept further. Under the Regulation, food is deemed unsafe if it is either injurious to human health or unfit for human consumption. In line with the EU White Paper on food safety, it adopts a risk-based approach, recognising that safety is not an absolute condition. This concept mirrors the criterion of ‘reasonable certainty of no harm’ used in US law (Food Quality Protection Act, 1996). Thus, © 2008, Woodhead Publishing Limited

242

Chilled foods

in determining whether a food is unsafe, the Regulation requires that two key characteristics are taken into account: the normal conditions of use of the food by food operators as well as by the final consumer, and the information provided about the food, for instance by labelling, but also more generally. The Regulation also provides guidance on how to determine whether any food is injurious to health. This includes consideration of the probable immediate and/ or short-term and/or long-term effects of that food on the health of a person consuming it, but should also include possible effects on subsequent generations; and the probable cumulative toxic effects as well as the particular health sensitivities of a specific category of consumers where the food is intended for that category of consumers. Contemporary approaches to food safety emphasise the need for a comprehensive integrated process (FAO–WHO, 2003; Regulation (EC) 178/2002) both in individual food businesses and along the whole food chain. This approach follows logically from the observation that food hazards may arise at any point along that chain and highlights the need for good communication at all levels, as well as traceability.

10.4 Sources of toxicological hazards As mentioned previously, toxicological hazards in foods may arise at any point within the food chain, starting with the substances that make up the food, i.e. raw materials at different levels of processing, through processing itself to packaging and storage prior to consumption. Toxicological hazards may also arise through substances deliberately added to foods or ingredients, as well as through substances present inadvertently.

10.4.1 Raw materials and ingredients The constituents of food are chemically complex, containing a range of components, some of which can be known toxicants. The hazards inherent in raw materials may be classed into several groups, including hazards inherent in the food itself (e.g. toxic metabolites produced by food crops), hazards arising from the environment in which the foods are grown and/or harvested (e.g. heavy metals in contaminated soils), those arising from substances used to treat food crops (e.g. pesticides, herbicides) or in the husbandry of food animals (e.g. antimicrobial residues) or present on such crops as contaminants (e.g. mycotoxins) (see Table 10.1). Whole foods themselves may intrinsically contain a wide variety of biologically active materials which are capable of producing adverse health effects. These are often metabolites which have evolved to protect the plant against predators. Wellknown examples include alkaloids, such as solanine in Solanaceae species, and cyanogenic glycosides in certain almond species. Potentially toxic substances also occur because of the chemical burden in the soil in which a crop is grown or the feedstock used by food animals. For instance, © 2008, Woodhead Publishing Limited

Non-microbiological hazards and safe process design

243

Table 10.1 Sources of non-microbiological hazards Source

Hazards

Raw materials and ingredients

Mycotoxins, polychlorinated biphenyls, polypolyhalogenated aromatic hydrocarbons, heavy metals, pesticide residues, antibiotic residues, plant toxins Materials with a functional role Additives, colours, processing aids, vitamins, minerals in food products or processes Equipment and processing Maillard reaction products, nitrosamines, PAHs, acrylamide, lipid peroxides, machinery lubricants, sanitising agents Storage Food contact materials Allergenic ingredients Milk, egg, peanut, etc.

low levels of arsenic have been reported in tea leaf samples in China (Yuan et al., 2007). Heavy metals, for instance mercury, are well-known contaminants of many seafood species, the problem being exacerbated by the position of many of these species at the top of the food chain, resulting in accumulation (see Balshaw et al., 2007). Other common contaminants have been reported to occur because the dietary sources of the relevant food animal species include them; in particular, many types of polycyclic aromatic hydrocarbons and polychlorinated biphenyls (Gagnon et al., 2004). Raw materials and ingredients may also contain residues of potentially harmful compounds that have been used to aid their production, such as pesticides in the case of plant food sources and antimicrobial residues in farmed animals. Infestation of crop species with certain micro-organisms can also result in potential chemical toxicity, a good example being the presence of mycotoxins, including aflatoxins, ochratoxins, fumonisins, zearalenone, and trichothecenes from different fungal species. Contamination with mycotoxins can affect crops as diverse as peanuts, maize (corn), pistachio, walnuts, and copra. Aflatoxins are among the most studied mycotoxins, and the relationship between aflatoxin ingestion and primary liver cancer is well established. Almost all plant products can serve as substrates for fungal growth, and subsequently mycotoxin contamination of human food and animal feed. Animal feed contaminated with mycotoxins can result in the carry-over of toxins through milk and meat to consumers.

10.4.2 Materials with a functional role in food products or processes Modern production of foods can also involve the deliberate use of a wide array of materials which come under the broad designation of additives. These substances can also constitute potential risks to health. There are some 2500 chemicals that function as food additives, giving rise to some 5000 trade name products on a worldwide basis (Scotter and Castle, 2004). A recent report by the Nordic Working Group on Food Toxicology and Risk Assessment (NNT, 2002) on behalf of the EU, listed over 300 permitted substances, including synthetic and natural colours, preservatives and antioxidants, emulsifiers, stabilisers and gelling agents, © 2008, Woodhead Publishing Limited

244

Chilled foods

sweeteners and materials with a variety of technological functions. Many nutrients have well-documented adverse effects when recommended intakes are exceeded (e.g. teratogenic effect of vitamin A). In the EU, and many other legislatures, these materials are listed in a positive list, which also specifies conditions under which they may be used and in what quantities. 10.4.3 Equipment and processing Food processing involves a variety of processes, in which thermal treatments of various types and degrees figure prominently. Chemically, these treatments obey conventional chemical rules inasmuch as the formation of particular compounds can be predicted based on the available substrates and the effect of temperature, as well as other factors such as pH and ionic strength, in providing the necessary activation energies and driving the kinetics. These processes thus result in reactions between components of the food, which can produce potentially hazardous substances (as well as other, desirable, compounds). They can also result in the inactivation of some toxicants, such as plant lectins and alkaloids. A number of reaction systems leading to production of compounds with well-documented toxic properties have been extensively studied and well characterised (Table 10.2). These include:

• Formation of polyaromatic hydrocarbons through pyrolysis of carbohydrates during grilling or broiling of meats.

• Maillard reactions, which produce a wide range of compounds, including some • • •

potentially hazardous to health, from the reaction of reducing sugars with amino acids. Nitrosamine formation from reaction of nitrite preservatives with secondary and tertiary amines in cured meat products. Generation of acrylamide from the reaction between asparagine and carbohydrate at the very high temperatures involved in frying and baking. Lipid peroxidation, resulting in the production of reactive peroxides in oils with a high proportion of polyunsaturated fatty acids.

Aside from any consequences of normal equipment operation (e.g. heating), processing can potentially contribute to toxic hazards if materials, e.g. lubricants Table 10.2 Examples of non-microbiological hazards from processing Type of processing Grilling/broiling of meats Nitrite preservation

Reaction

Pyrolysis of carbohydrates Nitrite reaction with secondary and tertiary amines Frying/baking Carbohydrate reaction with asparagine residues Storage of polyunsaturated Oxidation of double bonds fats or exposure to air

© 2008, Woodhead Publishing Limited

Hazards Polyaromatic hydrocarbons Nitrosamines Acrylamide Lipid peroxides

Non-microbiological hazards and safe process design

245

used to ensure its operation become transferred to the food which is being produced. In addition, equipment operation, or perhaps more accurately the ease with which it can be sanitised, can contribute to cross-contact between different products, a situation which can make a product hazardous to particular groups of individuals, such as allergic consumers.

10.4.4 Storage Storage of foods and food products can similarly result in the production or appearance in the food of potentially toxic materials. One of the processes driving this is essentially the same as for processing: over a period of time a variety of chemical reactions will take place in the food, albeit at a much slower rate, with some, of course, not taking place at all because of the lack of adequate activation energy. Enzymes present in foods, particularly those which have undergone little or no heat processing, can catalyse reactions leading to spoilage and production of potentially toxic compounds (e.g. biogenic amines in scombroid fish). Materials may also migrate from the packaging in which food products are stored. For instance, tin concentrations of up to 200mg/kg have been reported in acidic products stored in tin-plated steel cans (Oldring and Nehring, 2007).

10.4.5 Food allergens As with other toxicological hazards, food allergens may arise at any point in the food chain. However, they differ from most other chemical hazards as they pose a risk only to a limited and reasonably well-defined proportion of the population and are harmless to the vast majority at almost any level of intake. Known food allergens are exclusively proteins or polypeptides in foods, which provoke an immune response mediated by IgE antibodies in susceptible individuals. Allergenic proteins from foods belong to a relatively small number of protein families (Breiteneder and Mills, 2005) and tend to share certain molecular characteristics (e.g. thermal stability, resistance to pepsinolysis, possession of intramolecular disulphide bonds), although there is no single set of features which can discriminate between an allergenic and a non-allergenic protein. Small quantities of allergenic foods may become mixed in products at almost any stage in the food chain. This includes commingling through crop rotation or transport, as well as mixing through cross-contact during manufacture, often because of difficulties in adequately sanitising a production line. The extent of cross-contact during manufacture and therefore the magnitude of the risk will depend considerably on the process. In general, dry mix processes will result in products that pose a higher risk than wet processes where extensive cleaning in place (CIP) is possible (and usually necessary for microbiological reasons).

© 2008, Woodhead Publishing Limited

246

Chilled foods

10.5 Public health significance of non-microbiological hazards The toxicological hazards described in the previous section undoubtedly have potential significance for public health. Indeed, in many cases the adverse health effects caused by specific materials have impelled public authorities to draw up food regulations and place limits on specific ingredients and contaminants. In the recent past, episodes such as dioxin contamination of animal feed in Belgium, together with other events, led to a loss of confidence among consumers about the safety of the European food supply. This provoked the European Union to draw up its White Paper on Food Safety (CEC, 2000), proposing a new framework of food law as well as the creation of the European Food Safety Authority. The previous section attempted to describe the range of substances with potentially toxic effects that may be present in foods, both inadvertently and through deliberate addition. However, assessing the impact of these materials on public health remains complex. A key principle, however, remains the well-known statement by Paracelsus, namely ‘the dose makes the poison’. In essence, this means that the inherent properties of any material are insufficient in themselves to pose a risk; instead, this requires consideration of exposure to the hazard. Food itself is often held to be responsible for up to one third of the cases of cancer in the population of industrialised societies, but this figure includes the effects of nutritional as well as toxicological factors. Similarly, the range of noncancer endpoints and the target organs associated with different substances is also extremely wide. Epidemiological findings and animal studies have provided many hypotheses, but the overall adverse health burden from toxicological hazards remains difficult to estimate, even for extremely toxic materials. For substances that have both a beneficial effect, and indeed may be essential nutrients (e.g. vitamins, trace minerals), the calculation becomes even more complex. In this section, several examples of substances in foods that can pose a risk to health will be described to illustrate general principles governing consideration of the public health impact. Polyhalogenated aromatic hydrocarbons (PAHs) form a class of compounds which includes dioxins, and polychlorinated (PCB) and polybrominated biphenyls (PBB), which are ubiquitous in the environment. They are also well-documented toxicants in animal studies with both cancer and non-cancer effects (Wigle and Lanphear, 2005; Bretveld et al., 2006). Numerous studies have examined human exposure, both through accidents and through contamination of the food supply. For instance, van Larebeke et al. (2006) measured PCB levels in Belgian residents in two urban and one rural area and examined correlations with the levels of various tumour-associated proteins (TAPs). They reported rather homogeneous exposure to pollutants in all three populations, but increased levels of TAPs only in the industrial (urban) areas. Thus the study provided, at best, equivocal support for an adverse effect of PCBs at the levels noted. Goldman et al. (2000) found elevated serum levels of PAHs in Californian residents consuming meat and poultry which they had reared and who lived in areas which had been contami© 2008, Woodhead Publishing Limited

Non-microbiological hazards and safe process design

247

nated. They estimated that the increased levels they observed translated into an additional lifetime cancer risk of more than the 1 in a million reference, which is the effective threshold of regulation. Dhooge et al. (2006) studied the effect of dioxin exposure on reproductive parameters in young Flemish men and found an inverse correlation between serum levels of TCDD and semen volume, but not sperm count or morphology. They concluded that TCDD may affect seminal vesicle function. Rignell-Hydbom et al. (2007) reached similar conclusions in a study of Swedish fishermen. In a comprehensive review, a WHO Consultation on dioxin noted that the appropriate metric against which to evaluate dioxin exposure was body burden rather than blood levels (van Leeuwen et al., 2000). They confirmed an excess total cancer risk associated with dioxin body burden, but only for the most highly exposed individuals in industrial cohorts. However, the excess risk was relatively small (RR = 1.4). They also concluded that there was evidence for some neurobehavioural effects. Vitamins are essential nutrients that must often be provided in the diet, either as such, or as precursors. However, many produce adverse health effects if consumed in excessive amounts. Kloosterman et al. (2007) studied this issue in the context of vitamin fortification of foods in the Netherlands. Until 2004, the Netherlands forbade fortification on the grounds of possible adverse effects on public health. However, the policy had to change following a ruling by the European Court of Justice that it constituted a barrier to trade within the European Union. The authors reported the development of a risk assessment model based on dietary intakes of particular foods by different population sections and European upper tolerable intake levels. Based on this model, allowed fortification levels were set at 100 µg/ 100 kcal for folic acid, 3 µg/100 kcal for vitamin D and 0 µg/100 kcal for vitamin A [i.e. no fortfication]. In contrast with contaminants, one of the unique factors that needed to be considered in this assessment was the deliberate intake of these nutrients as supplements in different consumer groups, as well as the consequences of fortifying different products which might all be consumed by the same group (multiple intakes). Another dimension to the vitamin issue is illustrated by the observation that supplementation of the diet with β-carotenoids reduces the frequency of all-type cancers in non-smokers, but increases it in smokers (Touvier et al., 2005), an example of a subpopulation being adversely affected by an otherwise beneficial intervention. Soy isoflavones are an example of a constitutent of a common food which is not an essential nutrient like vitamins, but to which properties have been attributed which are potentially beneficial to some population groups, but might be harmful to others. Specifically, soy isoflavones display oestrogenic activity in vitro and in experimental animal studies (e.g. Diel et al., 2000; Diel et al., 2001). This observation led to the suggestion that they could be beneficial as a substitute for hormone replacement therapy in women undergoing the menopause (Kurzer, 2003; Tice et al., 2003). In contrast, this functional activity was hypothesised to pose a risk of feminisation in infants exposed through soy-based infant formula (Sheehan, 1998), as well as risking exacerbation of disease in women with hormone-sensitive breast cancer (Kurzer, 2003). Exposure to soy infant formula © 2008, Woodhead Publishing Limited

248

Chilled foods

has been extensive over many years (Setchell, 2006). Several reviews concluded that the epidemiological data on soy infant formula showed no evidence of adverse oestrogenic effects. The only well-documented adverse health effect in infants was goitrogenesis, which was remedied early in the use of such formulas through addition of iodine (Doerge and Sheehan, 2002; Borgert et al., 2003; Setchell, 2006). Food additives serve a number of technological functions in foods, including functions relating to food safety (e.g. preservatives). However, these compounds often suffer from unfavourable public perception and their safety is constantly questioned. While considerable information exists on toxicity to experimental animals, gathered through toxicological studies, well-documented reports of adverse health effects in exposed populations are, however, difficult to find for many additives. In practice, reports seem to be confined to a fairly narrow range of materials, which have evoked continuing interest or where a particularly sensitive subpopulation exists which is not protected by the general acceptable daily intake (ADI). In the first category figure BHT and BHA, which have been the subject of scrutiny because of reports of carcinogenesis in rodents. The latest evaluations in man indicate no increased risk, and indeed possibly a decreased risk of cancer in exposed human populations at current levels of use (Williams et al., 1999). Saccharin has been closely examined for the same reason and with the same result (Cordle and Miller, 1984). More recently, aspartame has also been suggested to be carcinogenic based on rodent studies (Soffritti et al., 2006, 2007), but a review of both the existing and new data by regulatory agencies and independent experts failed to support those conclusions (EFSA, 2006; Magnuson et al., 2007). The second category includes substances such as sulphites. Sulphites are well recognised causative agents of bronchial constriction, as well as of certain other symptoms which resemble those of food allergy. In the USA, sulphites accounted for the second highest frequency (after aspartame) of early reports to the FDA’s Adverse Reactions Monitoring System (ARMS) (Tollefson, 1988). Of 887 complaints received, 51% were classed as serious and there were 27 deaths, of which 17 were considered causally linked to sulphites. A distinct population appears to be particularly at risk of suffering adverse reactions to sulphites, namely severe asthmatics who may represent 5–10% of the total number of asthmatics. Based on these observations, although the WHO established an ADI of 42 mg/kg body weight, regulatory authorities worldwide now require sulphites to be declared on the label when they reach or exceed 10 ppm (10 mg/kg) in a food. However, studies show that the amount below which 5% of the sulphite-sensitive population would react is estimated to be as low as 3 mg (expressed as SO2), while the median reactive dose is equivalent to 23 mg SO2 and the amount below which 95% of at risk individuals would react is 172 mg, based on challenges in 236 patients (Corder and Buckley, 1995). It is thus not surprising that reports of reactions to sulphites continue to appear. Possibly also included in this second category are additives which, either alone or in combination, provoke behavioural changes in certain individuals, particularly children. While behavioural changes have long been alleged for many © 2008, Woodhead Publishing Limited

Non-microbiological hazards and safe process design

249

components of foods, particularly certain additives (Feingold, 1975), they have not generally been substantiated when investigated using robust methodology. However, a recent report of a randomised double-blind cross-over placebo-controlled trial of one of two mixes of artificial food colours and additives revealed adverse effects on a global measure of hyperactivity in both 3-year-old and 8/9-year-old children (McCann et al., 2007). As assessment of behavioural effects does not form part of the routine evaluation of toxicological safety; these findings do not indicate that safety studies were misinterpreted, but they do highlight the need to remain alert to the possibility that endpoints that were previously not considered significant may require assessment.

10.5.1 Food allergens Food allergy has been long recognised as a clinical phenomenon, with numerous reports in the 20th century medical literature (Prausnitz and Küstner, 1921; Tuft and Blumstein, 1942; Loveless, 1950). In the last 10–15 years, epidemiological studies have begun to quantify its impact on public health, indicating that a significant minority (2–4%) of the population, and in particular of children (5– 8%), in most countries surveyed suffer from IgE-mediated food allergy (Young et al., 1994; Kanny et al., 2001; Sicherer et al., 2004; Pereira et al., 2005; Rona et al., 2007; Bruijnzeel-Koomen et al., 1995; Venter et al., 2006). However, food allergy, because of its nature, affects the wider household and community, modifying their food-buying habits and generally decreasing their quality of life (Hefle et al., 2007; Gudgeon et al., 2005). Allergic reactions to foods can be lifethreatening and can occur in some individuals in response to exposure to milligram quantities of the relevant food (Taylor et al., 2002). Avoidance of the offending allergenic food, together with the use of rescue medication, is the only therapy. Protection of the food allergic consumer can thus pose difficult problems for food businesses. ILSI-Europe has recently published an excellent overview of the scientific and clinical aspects of food allergy and the issues involved in its management, in the form of a concise monograph (Jackson, 2003).

10.6 Assessment and management of risk from nonmicrobiological hazards Before a risk can be managed, it must first be assessed. Both risk assessment and risk management form part of the process of risk analysis, which the Codex Alimentarius Commission defines as being composed of three components:

• Risk assessment – a scientifically based process consisting of hazard identification, hazard characterisation, exposure assessment and risk characterisation.

• Risk management – the process, distinct from risk assessment, of putting in place measures which minimise the risk, taking into account the interests of all those affected by it. © 2008, Woodhead Publishing Limited

250

Chilled foods

• Risk communication – the interactive exchange of information and opinions throughout the risk analysis process concerning hazards and risks, risk related factors and risk perceptions, among risk assessors, risk managers, consumers, industry, the academic community and other interested parties, including the explanation of risk assessment findings and the basis of risk management decisions.

10.6.1 Risk assessment for chemical hazards in foods Risk assessment is thus critical to the management of risks. As already mentioned, risk assessment is a four-step process (Table 10.3), which must be preceded by problem formulation, which defines the risk management question. Basically, for any substance the broad data requirements will be similar: a comprehensive view of the hazards presented will be needed. Thus a hazard will first need to be identified and then characterised. How this happens will likely differ, depending on the source and nature of the hazard. In many cases, contaminants are likely to be identified as hazards through epidemiological studies or accidental exposure of populations. In those cases, characterisation may well also be incomplete as any studies are likely to focus on the hazard endpoints highlighted by epidemiology or accidental exposure, rather than on providing the comprehensive toxicological profile that is currently required for substances which are intended to be deliberately added to foods, such as additives or nutrients. Similarly comprehensive hazard characterisation data may also be available for certain types of contaminant such as pesticide or antimicrobial residues, if such data had already been required for market approval. A history of safe use may provide some of the data for characterisation of hazards in the case of certain food constituents which are proposed to be used in other than traditional uses (Constable et al., 2007). The legislative approach to food additives in the EU is based on the principle that only additives which have been explicitly authorised may be used. The authorisation may also specify in which foods the materials may be used and in what amounts (Council Directive 89/107/EEC, amended by Directive 1882/2003). The EU requires that ‘any proposed food additive should normally undergo a comprehensive examination for potential toxicological effects before its safety-inuse can be accepted’. While these requirements are flexible, permitting derogations as long as a sound justification is provided, dossiers are normally expected to contain at least the results of core tests, namely metabolism/toxicokinetics, subchronic toxicity (at least 90-day repeat dose studies), genotoxicity, chronic toxicity and carcinogenicity, reproductive and developmental toxicity studies (SCF, 2001). Nutrients, such as vitamins, would also require similarly comprehensive data. Hazard characterisation, if complete, will specify the various hazards inherent in a particular substance and will provide information on the lowest observed adverse effect and no observed adverse effect levels (LOAEL and NOAEL) for each endpoint. In the absence of exposure data these values can be used to derive tolerable daily intakes or ADIs by application of uncertainty factors. Information © 2008, Woodhead Publishing Limited

Non-microbiological hazards and safe process design

251

Table 10.3 Steps of the risk assessment process Step

Risk assessment question

Source of data

Hazard identification

What are the intrinsic hazards of the material? How does toxicity change with dose? What is the NOAEL? Are there particular at-risk groups? How much of this material are people exposed to? Through what sources?

Acute exposure data, epidemiology, animal studies Repeat dose animal studies, epidemiology

Hazard characterisation

Exposure assessment

Risk characterisation

What is the probability of adverse effects? In how many people?

Analytical measurements in products and sources of exposure, measurement of concentrations in tissues, body fluids Analysis of hazard, characterisation and exposure data

on exposure is then required in conjunction with these data to undertake safety or risk assessments. Exposure assessment is the qualitative and/or quantitative evaluation of the likely intake of substances via food, as well as exposure from other sources if relevant. Kroes et al. (2002) describe two different sampling strategies for this purpose: one to obtain a representative picture of chemical levels present in food and one aimed at sampling those products expected to contain higher levels in a cost-effective way. In principle, to assess food consumption, four different types of data can be used. Food supply data are calculated in food balance sheets which are accounts, at a national level, of annual production of food, changes in stocks, imports and exports, and agricultural and industrial use. The result is an estimate of the average food consumption per head of the population. However, it does not take account of food which is not eaten, nor provide information on patterns of consumption. Thus it is of relatively limited value to the risk manager seeking to protect ‘extreme’ consumers, as well as average ones. Household and individual consumption surveys record the amounts of foods and drinks brought into the household or consumed by individuals respectively. They thus help to fill this gap. Finally, individual daily diets can also be examined in more detail as they are consumed using the duplicate portion or duplicate diet approach. It is often feasible and relatively inexpensive to use or construct databases of foods with quite accurate information on content of particular contaminants or residues, based on routinely collected chemical analytical data. However, before using such routinely collected information, it is important to verify how the collection of samples was established, e.g. whether it was by random sampling or by selection due to suspicion. Once exposure data are available, full characterisation of the risk is possible. Using probabilistic methodologies, it also becomes possible to make more complete use of the dose–response information to provide better insights into the public © 2008, Woodhead Publishing Limited

252

Chilled foods

health consequences of specific levels of exposure to a substance. Quantitative methods of risk assessment for chemicals in food have been described in detail in the context of the EU project Food Safety in Europe (Edler et al., 2002; Kroes et al., 2002; van den Brandt et al., 2002). Renwick et al. (2003) published a detailed description of the risk characterisation of chemicals in food and diet, addressing amongst other issues:

• methods of hazard characterisation suitable for different risk assessments • characterisation of the dose–response relationship by mathematical modelling • use of mode or mechanism of action and/or the toxicokinetics for dose– response characterisation

• how to incorporate uncertainty and variability in the resulting output for the further refinements of dose–response characterisation. In the case of contaminants which are present at very low levels, the threshold of toxicological concern (TTC) approach has been proposed as a basis for deciding whether a given exposure is at such a level as to require a detailed toxicological evaluation of the material of concern (Kroes et al., 2000; ILSI-Europe, 2000). The approach is based on the analysis of animal and human toxicological data on a wide range of chemicals and their classification into different categories, based on their NOAELs. Application of the approach to an untested chemical involves allocating it to the correct class according to its structure, and comparing the estimated exposure to the TTC for the relevant class of chemicals. If exposure does not exceed the TTC, the resulting risk is deemed insignificant and further risk management measures unnecessary.

10.6.2 Risk management of chemical hazards in food The same basic principles thus apply to the risk assessment of substances from different sources, although data requirements can differ depending on factors such as exposure and the severity of any effects. However, subsequent management of the risk to ensure that food remains safe will vary according to the nature of the substance, the risk from which is being mitigated. Only limited control is thus possible over the uptake of contaminants by crops, so risk management will focus on specifying upper limits in those crops, at which the risk is considered tolerable. A similar approach may be applicable to substances such as residues from pesticides or antimicrobials used in animal husbandry, although action at an earlier point in the food chain is possible by restricting inputs of those substances. For substances deliberately added to foods, including those that may enter food by migrating from packaging, there is an expectation of no harm. Management is thus determined by the definition of ADIs, which take into account vulnerable subpopulations. As discussed previously, this process can be complex, particularly when nutrients have adverse effects in certain population groups. Furthermore, current ADIs are not immutable, but are periodically reviewed in the light of scientific developments. The NNT review (NNT, 2002) thus identified a need to review the risk assessment for a number of additives © 2008, Woodhead Publishing Limited

Non-microbiological hazards and safe process design

253

used in the EU, highlighting issues such as the replacement of synthetic additives by those from natural sources, the specification of which will differ from the material evaluated toxicologically. Management of risks arising through processing is more complex and requires a detailed consideration of the number of interacting chemistries which need to be understood in order to ensure that a product does not pose a risk to health. The principles above and their application are enshrined in an extensive body of legislation and regulation in many countries and regions and supported by guidance. For instance, in the EU, the General Food Law (Regulation (EC) 178/2002) covers the general principles which must be adhered to in order to assure safe food. However, it is complemented by laws specifically covering substances directly added to foods, as well as those inadvertently present (summarised in Table 10.4). Specifically, food additives are governed by Directive 89/107/EEC. Flavourings constitute a separate category of additive which is Directive 88/388/EEC and Council Decision 88/389/EEC regulate. Food contact materials comprising all materials and articles intended to come into contact with foodstuffs, including packaging materials but also cutlery, dishes, processing machines, containers, etc. come under Regulation 1935/2004. The principle governing regulation of contaminants is that safe levels are determined through scientific assessment, but levels in food should be kept as low as reasonably achievable using good working practices. This principle is elaborated in Directive 315/93/EC, while Regulation 1881/2006 sets maximum permissible levels for a number of materials including several mycotoxins, heavy metals, dioxins, PCBs and PAHs. Separate legislation governs residues from veterinary medicinal products (Regulation 2777/90/EEC) and plant protection products (Regulations 396/2005 and 178/2006).

10.6.3 Risk assessment and management of food allergens Kroes et al. (2002) acknowledge that ‘a particular challenge is the evaluation of food allergens and components causing other forms of intolerances, and how to determine the levels present and actual intakes vs. the limited knowledge of amounts needed for induction or elicitation of a response’. The authors in fact decided to exclude consideration of this issue from their paper. However, food allergens make a significant impact on public health and allergen management is mandated under general food law. Food production is a complex process and generally involves the use of shared equipment at all stages from the transport of raw materials, through manufacturing to packaging. Small amounts of allergenic ingredients which are not part of the recipe are therefore an ever-present possibility in finished products and can pose a risk to allergic consumers, even to the extent of provoking severe reactions. It is now well-established that the reactivity of allergic consumers varies over a considerable range, making assessment of that risk particularly challenging. While identifying the hazard (knowing which foods are allergenic) has proved straightforward, characterising this hazard is still the subject of much discussion. Considerable uncertainty still exists about how many people may react to such © 2008, Woodhead Publishing Limited

254

Chilled foods

Table 10.4 EU legislation on substances added to or inadvertently present in food Type of residue

Applicable document

Additives Flavourings Food contact materials Veterinary medical products

Directive 89/107/EEC Directive 88/388/EEC Framework Regulation 1935/2004 Regulation 2377/90 (establishes procedures for safety evaluation and setting MRLs) Regulation 396/2005 and Annexes I–IV; Regulation 178/ 2006 (crops for which MRLs have been set) Regulation 315/93; Regulation 1881/2006 (levels for specific contaminants in various foods)

Plant protection products Contaminants (general)

small amounts of allergen and how low those amounts are. In particular, it has proved difficult to define amounts of allergen below which no allergic patients react. Important reports from both the EFSA (EFSA, 2004) and the USFDA (US FDA–CFSAN, 2006) have questioned whether adequate data currently exist to set thresholds. To some extent, this situation has happened because clinical food challenge studies are the only means of generating such data and allergic patients at risk of severe reactions to very low quantities of allergen are understandably reluctant to participate in them. In recent years, well-designed clinical studies have begun to define the distribution of the minimum doses of allergen that elicit responses in members of the allergic population. These new data have stimulated the development of new approaches (Bindslev-Jensen et al., 2002; Crevel et al., 2007) to overcome this problem. These approaches use statistical modelling of the population distribution of minimum eliciting doses (thresholds) to characterise the allergenic hazard. Coupled with estimates of exposure to the relevant allergen and knowledge of prevalence, these dose–distribution models can generate quantitative estimates of risk, helping to prioritise allergen management measures (Spanjersberg et al., 2006; Kruizinga et al., 2007). Managing allergen risks parallels essentially management of other chemical contamination risks. Because it can occur at any stage of the food chain, it requires systematic assessment of each of those stages to establish the nature and magnitude of the risk. Traceability and communication along the food chain, as intimated in standards such as ISO 22000:2005 (ISO, 2005) are key prerequisites to a successful allergen management programme, together with the application of HACCP to manufacturing operations. Based on the approach described above to assess the risk from residual allergen, risk managers can make informed decisions about whether instituted procedures, such as sanitation protocols, effectively reduce the risk to the extent intended. In many countries and regions, the special characteristics of food allergy and the risks for food-allergic consumers have led to legislation which supplements general food law. In the EU, the Labelling Directive (2000/13/EC), as amended by Directive 2003/89/EC, governs allergen labelling. The Directive identifies eleven foods or food groups and sulphur dioxide (listed in Annex IIIa) that are found in a wide variety of processed foods and that are known to trigger allergic reactions. © 2008, Woodhead Publishing Limited

Non-microbiological hazards and safe process design

255

These foods or their derivatives must be declared on the label whenever present as ingredients in a pre-packaged product, including previously exempted alcoholic beverages. Subsequent legislation (Directive, 2006/142/EC) added lupin and molluscs to Annex IIIa, while Directive, 2007/68/EC defined derived ingredients exempted following evaluation of dossiers by the EFSA and a conclusion that they did not pose a significant risk to the allergic consumer. The USA also recently implemented new legislation to improve protection for allergic consumers, with the Food Allergen Labeling and Consumer Protection Act, 2004 (FALCPA, 2004) which came into force on 01 January 2006. This Act contains provisions which differ in detail rather than principle from those of EU Directive 2003/89/EC. The Act mandates a shorter list of allergenic food groups, but requires indication of the species on the label in the case of fish, crustacea and tree nuts. It also specifically exempts highly refined oils from allergen labelling. Subsequent guidance on tree nuts describes a much longer list than the EU’s Annex IIIa. FALCPA also provides for a process to obtain exemption from labelling, although at the time of writing no derived ingredients have been exempted. As with the Directive, FALCPA (2004) requires labelling of allergenic ingredients irrespective of the amount ultimately present in the product. Both Directive 2003/89/EC and FALCPA (2004), as well as legislation in other countries such as Australia and Japan, undoubtedly improve the labelling of allergenic foods. However, despite better labelling of allergenic ingredients, the issue of adventitious allergen originating through, for instance, cross-contact during manufacturing is not addressed, even though substantial incorporation of undeclared allergenic constituents can occur (Taylor et al., 2002). A practical way to deal with unintentional allergenic ‘cross-contact’ could be the adoption of an appropriate upper limit for non-ingredient allergenic food components. For example, Switzerland requires the declaration of specified allergenic constituents whenever present in concentrations greater than 0.1% (1000 mg/kg), whether as ingredients or otherwise (Conseil Fédéral Suisse, 2002). However, a threshold of 0.1% might not be considered sufficiently protective of public health, since less than 1 mg of peanut protein has been shown to elicit adverse reactions in allergic subjects (reviewed in Taylor et al., 2002; US FDA–CFSAN, 2006). Obviously, an upper limit for non-ingredient allergenic food components also needs to consider the NOAEL reported for each of the important allergenic foods. Thus, while allergic consumers are protected against undeclared allergenic ingredients, they remain at risk from non-ingredient allergenic components, while the food industry lacks clear guidance from regulatory authorities. This absence of a regulatory threshold also has other consequences, which may reduce the protection of the allergic consumer. Allergenic ingredients present in insignificant quantities must be declared. For example, an ingredient containing refined peanut oil, with almost undetectable protein, could be a component of another ingredient used in very small amounts, such as a flavour. Yet products containing this ingredient must be labelled as containing peanut and allergic consumers who eat them could erroneously conclude that their allergy had resolved when, in fact, the amounts were too small to trigger a reaction. © 2008, Woodhead Publishing Limited

256

Chilled foods

10.7 Considerations specific to chilled foods Chilled foods are characterised by the lower than ambient temperature at which they are held prior to consumption, in order to keep them wholesome longer. As discussed, toxicological risks arise throughout the food chain and, insofar as chilled foods are made using the same ingredients and processes as other foods, they will be subject to the same toxicological risks. The process of chilling takes place at the end of the food manufacturing chain and it is difficult to conceive a change in the chemical risk profile, either from conventional food constituents or contaminants which would arise merely from this stage. It is possible to envisage theoretical scenarios, for instance, where the lower temperature might prevent the breakdown of a contaminant and thus lead to a product with a higher risk profile than its non-chilled equivalent. However, such a situation has not been documented. Other scenarios could be evoked where chemical or allergen contamination might occur during the chilling process, because of other materials or ingredients processed on the same machinery. However, the risk from such situations should normally be assessed as part of the overall process. Absence of reports of incidents from such scenarios suggests that such an assessment does indeed take place and is sufficient to manage any risk.

10.8 Conclusion As the FAO–WHO (2003) guidelines state: food safety is not negotiable. However, the sources from which food and food ingredients are obtained, as well as the processes which they undergo before they reach the consumer, all have the potential to jeopardise safety. Evolution of the principles of food safety has led to the recognition that it can be ensured in a modern food chain only by an integrated approach to the assessment and management of risk, supported by full traceability and communication. These principles are exemplified in the EU’s food safety legislation among others, and mirrored in newly developed international standards such as ISO 22000:2005. In relation to chemical aspects of food safety, this requires a detailed consideration of the toxicological profile of both the inherent constituents of the food or food ingredient and of any constituents that might be present, whether deliberately or otherwise. In keeping with the concept of a risk-based approach, the next step must be to assess the probability of harm, taking into account the toxicological profile and the extent of exposure. Risk assessment for different types of chemical hazard, including the hazards from food allergens, follows similar principles. However, subsequent management differs according to the source and nature of the risk.

10.9 References BALSHAW S, EDWARDS J, DAUGHTRY B, ROSS K. 2007. Mercury in seafood: Mechanisms of

accumulation and consequences for consumer health. Rev Environ Health, 22(2): 91–113. © 2008, Woodhead Publishing Limited

Non-microbiological hazards and safe process design

257

BINDSLEV-JENSEN C, BRIGS D, OSTERBALLE M. 2002. Can we determine a threshold level for

allergenic foods by statistical analysis of published data in the literature? Allergy, Aug; 57(8): 741–746. BORGERT C J, LAKIND J S, WITORSCH R J. 2003. A critical review of methods for comparing estrogenic activity of endogenous and exogenous chemicals in human milk and infant formula. Environ Health Perspect, Jun; 111(8): 1020–36. BREITENEDER H, MILLS E N. 2005. Molecular properties of food allergens. J Allergy Clin Immunol, Jan; 115(1): 14–23. BRETVELD R W, THOMAS C M, SCHEEPERS P T, ZIELHUIS G A, ROELEVELD N. 2006. Pesticide exposure: The hormonal function of the female reproductive system disrupted? Reprod Biol Endocrinol, 4: 30. BRUIJNZEEL-KOOMEN C, ORTOLANI C, AAS K, BINDSLEV-JENSEN C, BJORKSTEN B, MONERETVAUTRIN D, et al. 1995. Adverse reactions to food. European Academy of Allergology and Clinical Immunology Subcommittee. Allergy, 50: 623–635. CODEX ALIMENTARIUS COMMISSION. 2003. Principles for the Risk Analysis of Foods Derived from Modern Biotechnology. CAC/GL44-2003. Codex Alimentarius Commission, Joint FAO/WHO Food Standards Programme, Food and Agriculture Organisation, Rome. http://www.codexalimentarius.net/web/standard_list.do?lang=en COMMISSION OF THE EUROPEAN COMMUNITY. 2000. White paper on food safety. DOC/00/ 1 (COM(1999)719) Brussels, 12 January 2000. COMMISSION REGULATION (EC). 2006. No 1881/2006 of 19 December 2006 setting maximum levels for certain contaminants in foodstuffs. Official Journal of the European Union L364, 20.12.2006, p 5–24. COMMISSION REGULATION. 2006. 178/2006 of 1 February 2006 amending Regulation (EC) No 396/2005 of the European Parliament and of the Council to establish Annex I listing the food and feed products to which maximum levels for pesticide residues apply. Official Journal of the European Union L 29, 2.2.2006 p 3–25. CONSEIL FÉDÉRAL SUISSE. 2002. Ordonnance sur les denrées alimentaires. RO 2002 573. CONSTABLE A, JONAS D, COCKBURN A, DAVI A, EDWARDS G, HEPBURN P, HEROUETGUICHENEY C, KNOWLES M, MOSELEY B, OBERDÖRFER R, SAMUELS F. 2007. History of safe use as applied to the safety assessment of novel foods and foods derived from genetically modified organisms. Food Chem Toxicol, 45: 2513–2525. CORDER E H, BUCKLEY C E. 1995. Aspirin, salicylate, sulfite and tartrazine induced bronchoconstriction. Safe doses and case definition in epidemiological studies. J Clin Epidemiol, Oct; 48(10): 1269–1275. CORDLE F, MILLER S A. 1984. Using Epidemiology to regulate food additives: saccharin casecontrol studies. Public Health Rep, 99: 365–369. COUNCIL DECISION 88/389/EEC. 1988. Of 22 June 1988 on the establishment, by the Commission, of an inventory of the source materials and substances used in the preparation of flavourings. Official Journal of the European Union L184, 15.7.1988, p 67. COUNCIL DIRECTIVE 88/388/EEC. 1988. Of 22 June 1988 on the approximation of the laws of the Member States relating to flavourings for use in foodstuffs and to source materials for their production. Official Journal of the European Union L184, 15.7.1988, p 61. COUNCIL DIRECTIVE 89/107/EEC. 1989. Of 21 December 1988 on the approximation of the laws of the Member States concerning food additives authorized for use in foodstuffs intended for human consumption. Official Journal of the European Union L40, 11.2.1989, p 27. COUNCIL REGULATION (EEC). 1993. No 315/93 of 8 February 1993 laying down Community procedures for contaminants in food. Official Journal of the European Union L 37, 13.2.1993, p 1–3. COUNCIL REGULATION 2777/90/EEC. 1990. Of 26 June 1990 laying down a Community procedure for the establishment of maximum residue limits of veterinary medicinal products in foodstuffs of animal origin. Official Journal of the European Union L L224, 18/08/1990 P. 8. © 2008, Woodhead Publishing Limited

258

Chilled foods

CREVEL R W R, BRIGGS D, HEFLE S L, KNULST A C, TAYLOR S L. 2007. Hazard characterization

in food allergen risk assessment: The application of statistical approaches and the use of clinical data. Food Chem Toxicol, 45(5): 691–701. DHOOGE W, VAN LAREBEKE N, KOPPEN K, NELEN V, SCHOETERS G, VLIETINCK R, KAUFMAN J-M, COMHAIRE F, FOR THE FLEMISH ENVIRONMENT AND HEALTH STUDY GROUP. 2006. Serum dioxin-like activity is associated with reproductive parameters in young men from the general Flemish population. Environ Health Perspect, 114: 1670–1676. DIEL P, SCHULZ T, SMOLNIKAR K, STRUNCK E, VOLLMER G, MICHNA H. 2000. Ability of xenoand phytoestrogens to modulate expression of estrogen-sensitive genes in rat uterus: Estrogenicity profiles and uterotropic activity. J Steroid Biochem Mol Biol, May; 73(1– 2): 1–10. DIEL ET AL. 2001. Phytoestrogens and carcinogenesis-differential effects of genistein in experimental models of normal and malignant rat endometrium. Hum Reprod, 16(5): 997– 1006. DIRECTIVE 2000/13/EC. 2000. Of the European Parliament and of the Council of 20 March 2000 on the approximation of the laws of the Member States relating to the labelling, presentation and advertising of foods. Official Journal of the European Union L109. 06/ 05/2000, p 29–42. DIRECTIVE 2003/89/EC. 2003. Of the European Parliament and the Council of 10 November 2003 amending Directive 2000/13/EC as regards indication of the ingredients present in foodstuffs. Official Journal of the European Union L 308/15. 25/11/2003. DOERGE D R, SHEEHAN D M. 2002. Goitrogenic and estrogenic activity of soy isoflavones. Environ Health Perspec, 110(suppl 3): 349–353. EDLER L, POIRIER K, DOURSON M, KLEINER J, MILESON B, NORDMANN H, RENWICK A, SLOB W, WALTON K, WUERTZEN G. 2002. Mathematical modelling and quantitative methods. Food Chem Toxicol, 40: 238–326. EFSA. 2006. Opinion of the Scientific Panel on Food Additives, Flavourings, Processing Aids and Materials in contact with Food (AFC) on a request from the Commission related to a new long-term carcinogenicity study on aspartame – Question number EFSA-Q-2005122 Adopted on 3 May 2006. EFSA Journal, 356: 1–44. EUROPEAN COMMISSION DIRECTIVE 2005/26/EC. 2005. Of 21 March 2005 establishing a list of food ingredients or substances provisionally excluded from Annex IIIa of Directive 2000/13/EC of the European Parliament and of the Council. Official Journal of the European Union L75/33, 22/03/2005. EUROPEAN COMMISSION DIRECTIVE 2006/142/EC. 2006. Of 22 December 2006 amending Annex IIIa of Directive 2000/13/EC of the European Parliament and of the Council listing the ingredients which must under all circumstances appear on the labelling of foodstuffs. Official Journal of the European Union L368/110, 22/12/2006. EUROPEAN COMMUNITY. 2002. Regulation (EC) 178/2002 of The European Parliament and of The Council of 28 January 2002 laying down the general principles and requirements of food law, establishing the European Food Safety Authority and laying down procedures in matters of food safety. Official Journal of the European Union L31, 1.2.2002, P. 1. EUROPEAN FOOD SAFETY AUTHORITY. 2004. Opinion of the Scientific Panel on Dietitic Products, Nutrition and Allergies on a request from the Commission relating to the evaluation of allergenic foods for labelling purposes. EFSA Journal, 32: 1–197. FAO–WHO. 2003. Assuring food safety and quality. www.who.int/foodsafety/publications/ capacity/en/Englsih_Guidelines_Food_control.pdf FEINGOLD B F. 1975. Hyperkinesis and learning disabilities linked to artificial food flavors and colours. Am J Nurs, 75: 797–803. FOOD ALLERGEN LABELING AND CONSUMER PROTECTION ACT. 2004. 118 STAT. 905. Public Law 108–282, Aug. 2, 2004. FOOD QUALITY PROTECTION ACT (FQPA). 1996. Public Law 104–170, Aug. 3, 1996 110 STAT. 1489. GAGNON F, TREMBLAY T, ROUETTE J, CARTIER J F. 2004. Chemical risks associated with © 2008, Woodhead Publishing Limited

Non-microbiological hazards and safe process design

259

consumption of shellfish harvested on the north shore of the St. Lawrence River’s lower estuary. Environ Health Perspect, Jun; 112(8): 883–888. GOLDMAN L R, HARNLY M, FLATTERY J, PATTERSON D G JR, NEEDHAM L L. 2000. Serum polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans among people eating contaminated home-produced eggs and beef. Environ Health Perspect, Jan; 108(1): 13–19. GUDGEON L A, TREWIN J B, GRIMSHAW K E C, HOURIHANE J O B. 2005. Patients Find Low Dose Threshold Challenges Useful in the Management of their Peanut Allergy. American Academy of Allergy, Asthma and Immunology, San Antonio, USA, March. HEFLE S L, FURLONG T J, NIEMANN L, LEMON-MULE H, SICHERER S, TAYLOR S L. 2007. Consumer attitudes and risks associated with packaged foods having advisory labeling regarding the presence of peanuts. J Allergy Clin Immunol, Jul; 120(1): 171–176. ILSI-EUROPE. 2000. Threshold of Toxicological Concern for Chemical Substances Present in the Diet. Report of a Workshop held in October 1999. ILSI Europe Report Series ISBN 1-57881-101-5. INSTITUTE FOR FOOD SCIENCE AND TECHNOLOGY. 1990. Guidelines for the handling of chilled foods, 2nd edition. IFST, London. ISO 22000. 2005. Food safety management systems – Requirements for any organization in the food chain. International Organisation for Standardisation. http://www.iso.org. JACKSON W F. 2003. Food Allergy. ILSI Europe Concise Monograph Series. ISBN 1-57881160-0. KANNY G, MONERET-VAUTRIN DA, FLABBEE J, BEAUDOUIN E, MORISSET M, THEVENIN F. 2001. Population study of food allergy in France. J Allergy Clin Immunol, 108(1): 133– 140. KLOOSTERMAN J, FRANSEN HP, DE STOPPELAAR J, VERHAGEN H, ROMPELBERG C. 2007. Safe addition of vitamins and minerals to foods: Setting maximum levels for fortification in the Netherlands. Eur J Nutr, Jun; 46(4): 220–229. KROES R, GALLI C, MUNRO I, SCHILTER B, WÜRTZEN G. 2000. Threshold of toxicological concern for chemical substances present in the diet: A practical tool for assessing the need for toxicity testing. Food Chem Toxicol, 38: 255–312. KROES R, MÜLLER D, LAMBE J, LÖWIK M R H, VAN KLAVEREN J, KLEINER J, MASSEY R, MAYER S, URIETA I, VERGER P, VISCONTI A. 2002. Assessment of intake from the diet. Food Chem Toxicol, 40: 327–385. KRUIZINGA A G., BRIGGS D, CREVEL R W R, KNULST A C, VAN DEN BOSCH L M C, HOUBEN G F. 2007. Probabilistic risk assessment model for allergens in food: Sensitivity analysis of the minimum eliciting dose and food consumption. Food Chem Toxicol, doi: 10.1016/ j.fct.2007.09.109. KURZER M S. 2003. Phytoestrogen supplement use by women. J Nutr, 133(6): 1983S–1986S. LOVELESS M H. 1950. Milk allergy: A survey of its incidence; experiments with a masked ingestion test. J Allergy, 21: 489. MAGNUSON B A, BURDOCK G A, DOULL J, KROES R M, MARSH G M, PARIZA M W, SPENCER P S, WADDELL W J, WALKER R AND WILLIAMS G M. 2007. Aspartame: A safety evaluation based on current use levels, regulations, and toxicological and epidemiological studies, Criti Rev Toxicol, 37: 8, 629–727. MCCANN D, BARRETT A, COOPER A, CRUMPLER D, DALEN L, GRIMSHAW K, KITCHIN E, LOK K, PORTEOUS L, PRINCE E, SONUGA-BARKE E, WARNER JO, STEVENSON J. 2007. Food additives and hyperactive behaviour in 3-year-old and 8/9-year-old children in the community: A randomised, double-blinded, placebo-controlled trial. Lancet, 370(9598): 1560–1567. NNT. 2002. Food additives in Europe 2000 – Status of safety assessments of food additives presently permitted in the EU. http://www.foodcomp.dk/foodadd/download/NorFAD.pdf. OLDRING P K T, NEHRING U. 2007. Packaging materials 7. Metal packaging for foodstuffs. ILSI Europe Report Series ISBN 90-78637-06-6. PEREIRA P, VENTER C, GRUNDY J, CLAYTON C B, ARSHAD S H, DEAN T. 2005. Prevalence of © 2008, Woodhead Publishing Limited

260

Chilled foods

sensitization to food allergens, reported adverse reaction to foods, food avoidance, and food hypersensitivity among teenagers. J Allergy Clin Immunol, 116: 884–892. PRAUSNITZ C, KÜSTNER H. 1921. Studien über die Uberempfýndlichkeit (Study of hypersensitivity). Zentralblatt für Bakteriologie Mikrobiologie und Hygiene Abteilung Originale, 86: 160–169. REGULATION (EC). 2005. No 396/2005 of the European Parliament and of the Council of 23 February 2005 on maximum residue levels of pesticides in or on food and feed of plant and animal origin and amending Council Directive 91/414/EEC. Official Journal of the European Union OJ L 70, 16.3.2005, p 1–16. RENWICK A G, BARLOW S M, HERTZ-PICCIOTO I, BOBBIS A R, DYBING E, EDLER L, EISENBRAND G, GREIG J B, KLEINER J, LAMBE J, MÜLLER D J G, SMITH M R, TRITSCHER A, TUIJTELAARS S, VAN DEN BRANDT P, WALKER R, KROES R. 2003. Risk characterization of chemicals in food and diet. Final part of Food Safety in Europe. (FOSIE): Risk assessment of chemicals in food and diet (EC concerted action QLK1-1999-00156). Food Chem Toxicol, 41(9 special issue): 1211–1271. RIGNELL-HYDBOM A, AXMON A, LUNDH T, JÖNSSON B A, TIIDO T, SPANO M. 2007. Dietary exposure to methyl mercury and PCB and the associations with semen parameters among Swedish fishermen. Environ Health, May 8; 6: 14. RONA R J, KEIL T, SUMMERS C, GISLASON D, ZUIDMEER L, SODERGREN E, SIGURDARDOTTIR S T, LINDNER T, GOLDHAHN K, DAHLSTROM J, MCBRIDE D, MADSEN C. 2007. The prevalence of food allergy: A meta-analysis. J Allergy Clin Immunol, 120(3): 638–646. SCIENTIFIC COMMITTEE ON FOOD. 2001. Guidance on Submissions for Food Additive Evaluations by The Scientific Committee on Food. SCF/CS/ADD/GEN/26 Final. 12 July 2001. http://europa.eu.int.comm/food/fs/sc/scf/out98_en.pdf. SCOTTER M J, CASTLE L. 2004. Chemical interactions between additives in foodstuffs: A review. Food Addit Contam, Feb; 21(2): 93–124. SETCHELL K D R. 2006. Assessing risks and benefits of genistein and soy. Environ Health Perspect, 114(6). SHEEHAN D M. 1998. Herbal medicines, phytoestrogens and toxicity: Risk–benefit considerations. Proc Soc Exp Biol Med, 217(3): 379–385. SICHERER S H, MUNOZ-FURLONG A, SAMPSON H A. 2004. Prevalence of seafood allergy in the United States determined by a random telephone survey. J Allergy Clin Immunol, 114(1): 159–65. SOFFRITTI M, BELPOGGI F, DEGLI ESPOSTI D, LAMBERTINI L, TIBALDI E, RIGANO A. 2006. First experimental demonstration of the multipotential carcinogenic effects of aspartame administered in the feed to Sprague-Dawley rats. Environ Health Perspect, 114: 379–385. SOFFRITTI M, BELPOGGI F, TIBALDI E, ESPOSTI D, LAURIOLA M. 2007. Life-span exposure to low doses of aspartame beginning during prenatal life increases cancer effects in rats. Environ Health Perspect, 115: 1293–1297. SPANJERSBERG M Q, KRUIZINGA A G, RENNEN M A, HOUBEN G F. 2007. Risk assessment and food allergy: The probabilistic model applied to allergens. Food Chem Toxicol, Jan; 45(1): 49–54. TAYLOR S L, HEFLE S L, BINDSLEV-JENSEN C, BOCK S A, BURKS A W, CHRISTIE L, HILL D J, HOST A, HOURIHANE J O B, LACK G, METCALFE D D, MONERET-VAUTRIN D A, VADAS P A, RANCE F, SKRYPEC D J, TRAUTMAN T A, MALMEHEDEN YMAN I, ZEIGER R S. 2002. Factors affecting the determination of threshold doses for allergenic foods: How much is too much? J Allergy Clin Immunol, 109: 24–30. TICE J A, ETTINGER B, ENSRUD K, WALLACE R, BLACKWELL T, CUMMINGS S R. 2003. Phytoestrogen supplements for the treatment of hot flashes: The Isoflavone Clover Extract (ICE) Study: A randomized controlled trial. JAMA, Jul 9; 290(2): 207–14. TOLLEFSON L. 1988. Monitoring adverse reactions to food additives in the U.S. Food and Drug Administration. Regulatory Toxicol Pharmacol, 8: 438–446. TOUVIER M, KESSE E, CLAVEL-CHAPELON F, BOUTRON-RUAULT M-C. 2005. Dual association © 2008, Woodhead Publishing Limited

Non-microbiological hazards and safe process design

261

of beta-carotene with risk of tobacco-related cancers in a cohort of French women. J Natl Cancer Inst, 97(18): 1338–1344. TUFT L, BLUMSTEIN G I. 1942. Studies in food allergy. II. Sensitization to fresh fruits: clinical and experimental observations. J Allergy, 13: 574–581. US FDA–CFSAN. 2006. Report prepared by the Threshold Working Group. Approaches to establish thresholds for major food allergens and for gluten in food. http://www.cfsan.fda. gov/~acrobat/alrgn2.pdf. VAN DEN BRANDT P, VOORRIPS L, HERTZ-PICCIOTTO I, SHUKER D, BOEING H, SPEIJERS G, GUITTARD C, KLEINER J, KNOWLES M, WOLK A, GOLDBOHM A. 2002. The contribution of epidemiology to risk assessment of chemicals in food and diet. Food Chem Toxicol, 40: 387–424. VAN LAREBEKE N A, BRACKE M E, NELEN V, KOPPEN G, SCHOETERS G, VAN LOON H, VLIETINCK R. 2006. Differences in tumor-associated protein levels among middle-age Flemish women in association with area of residence and exposure to Pollutants. Environ Health Perspect, 114: 887–892. VAN LEEUWEN F X, FEELEY M, SCHRENK D, LARSEN J C, FARLAND W, YOUNES M. 2000. Dioxins: WHO’s tolerable daily intake (TDI) revisited. Chemosphere, May–Jun; 40(9– 11): 1095–101. VENTER C, PEREIRA B, GRUNDY J, CLAYTON C B, ARSHAD S H, DEAN T.2006. Prevalence of sensitization reported and objectively assessed food hypersensitivity amongst six-yearold children: A population-based study. Pediatr Allergy Immunol, 17: 356–363. WIGLE D T, LANPHEAR B P. 2005. Human health risks from low-level environmental exposures: No apparent safety thresholds. PLoS Med, 2(12): e350. WILLIAMS G M, IATROPOULOS M J, WHYSNER J. 1999. Safety assessment of butylated hydroxyanisole and butylated hydroxytoluene as antioxidant food additives. Food Chem Tox, 37: 1027–1038. YOUNG E, STONEHAM M D, PETRUCKEVITCH A, BARTON J, RONA R. 1994. A population study of food intolerance. Lancet, 343(8906): 1127–1130. YUAN C, GAO E, HE B, JIANG G. 2007. Arsenic species and leaching characters in tea (Camellia sinensis). Food Chem Toxicol, 45(12): 2381–2389.

© 2008, Woodhead Publishing Limited

11 The hygienic design of chilled food plants and equipment J. T. Holah, Campden and Chorleywood Food Research Association, UK

11.1 Introduction The primary concern of chilled food manufacturers is to produce a product that is both wholesome, i.e. it has all the fresh, quality attributes associated with a chilled food, and safe, i.e. free from pathogenic micro-organisms and chemical and foreign body contamination. This is particularly important in this product sector as, due to the nature and method of production, many chilled foods are classified as high risk, ready-to-eat (RTE) products. The schematic diagram shown in Fig. 11.1, which is typical for all food factories, shows that the production of safe, wholesome foods stems from a thorough hazard analysis – indeed this is now a legal requirement (Anon., 2004). The diagram also shows that given specified raw materials, there are four major ‘building blocks’ that govern the way the factory is operated to ensure that the safe, wholesome food goal is realised. Hygienic design dictates the design of the factory infrastructure and includes the factory site and buildings, the process lines and equipment and, until replaced by robots, the operatives! Hygienic practices maintain the integrity of the facility and include cleaning and disinfection, maintenance, personal hygiene and good manufacturing practices (GMP). Process development enables the design of safe, validated processes whilst process control subsequently ensures that each product in each batch on every day meets the process requirements. © 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

Fig. 11.1

263

Schematic stages required to ensure safe, wholesome chilled products.

Risk analysis encompasses identifying the hazards that may affect the quality or safety of the food product, and controlling them at all stages of the process such that the survival of hazardous micro-organisms and product contamination are prevented. In the food industry, this is commonly referred to as hazard analysis critical control point (HACCP). Such hazards are usually described as:

• biological, e.g. bacteria, yeasts, moulds • chemical e.g. cleaning chemicals, lubricating fluids • physical, e.g. glass, insects, pests, metal, dust. A hazard analysis should be undertaken at the earliest opportunity in the process of food production and, if possible, before the design and construction of the processing facility. This allows the design of the production facility to play a major role in hazard elimination or risk reduction. For example, it is possible to identify that glass is a potential hazard and one could eliminate this hazard by designing a glass-free factory. Of the four building blocks illustrated in Fig. 11.1, this chapter deals with hygienic design. For the food factory, hygienic design begins at the level of its siting and construction and is concerned with such factors as the design of the building structure, the selection of surface finishes, the segregation of work areas to control hazards, the flow of raw materials and product, the movement and control of people, the design and installation of the process equipment and the design and installation of services (air, water, steam, electrics, etc.). The ‘hygienic design’ of food operatives, usually referred to as the employment of staff free of pathogen carriage, is not considered. © 2008, Woodhead Publishing Limited

264

Chilled foods

With regard to hygienic building legislation in the EU, Regulation (EC) No 852/ 2004 on the hygiene of foodstuffs details the basic requirements for food premises. Some information, at a very general level, is given on the layout, design and construction of food premises and it covers, for example, the provision of cleaning and sanitary services, water supply and drainage, temperature control and the selection of floor and wall surfaces. Guidance is extended in Regulation (EC) No 853/2004 on laying down specific hygiene rules for food of animal origin to premises preparing meat, bivalve mollusc, fish, milk and egg products. Within both of these documents, however, advice is, at best, concise.

11.2 Segregation of work zones Factories should be constructed as a series of barriers that aim to limit the entrance of contaminants. The number of barriers created will be dependent on the nature of the food product and the product design, and will be established from the HACCP study. Figure 11.2 shows that there are up to four levels of segregation that are typical for food plants: Level 1 represents the siting of the factory, the outer fence and the area up to the factory wall. This level provides barriers against environmental conditions, e.g. prevailing wind and surface water run-off, unauthorised public access and avoidance of pest harbourage areas. Level 2 represents the factory wall and other processes (e.g. UV flytraps) which 2 3

Low risk

1

Goods reception and storage

oven barrier

Product enclosure

4

High risk Product enclosure

4

Product enclosure

4

Primary packing

Preparation

Secondary packing warehouse and dispatch

Administration and amenity area

Fig. 11.2 Schematic layout of a factory site showing ‘barriers’ against contamination. (1) Perimeter fence. (2) Main factory buildings. (3) Walls of high risk area. (4) Product enclosure within high risk.

© 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

265

should separate the factory from the external environment. Whilst it is obvious that the factory cannot be a sealed box, the floor of the factory should ideally be at a different level from the ground outside, and openings should be designed to be pest proof when not in use. Level 3 represents the internal barriers that are used to separate manufacturing processes of different risk, e.g. pre and post-heat treatment. Such separation, which creates zones usually referred to as high care or risk areas, should seek to control the air, people and surfaces (e.g. the floor and drainage systems and the passage of materials and utensils across the barrier). Level 4 represents a product enclosure zone, set within the level 3 high care/risk area. A product enclosure zone could encompass true aseptic filling or ‘ultra clean’ processing and packing areas, such as glove boxes or the use of highly filtered air as a barrier around the process line.

11.3 Barrier 1 – the factory site Attention to the design, construction and maintenance of the site surrounding the factory provides an opportunity to set up the first (outer) of a series of barriers to protect production operations from contamination. The focus on the site barrier has changed over the last 20 years or so from consideration of pests and the protection from environmental conditions, e.g. prevailing wind and surface water run-off, through the control of unwanted access by people, to the threat of bioterrorism. At the site level, a number of steps can be taken to control hazards including:

• The site should be well defined and/or fenced to prevent unauthorised public access and the entrance of domestic/wild animals, etc.

• Measures can be put in place to maintain site security including the use of gate • •

• • •

houses, security patrols and maintenance schedules for barrier fencing or other protection measures. The factory building may often be placed on the highest point of the site to reduce the chance of ground level contamination from flooding. Well-planned and properly maintained landscaping of the grounds can assist in the control of rodents, insects and birds by reducing food supplies, and breeding and harbourage sites. The use of two lines of rodent baits located every 15–21 m along the perimeter boundary fencing and at the foundation walls of the factory, together with a few mouse traps near building entrances is advocated by Imholte (1984). In addition, good landscaping of sites can reduce the amount of dust blown into the factory. Shapton and Shapton (1991) state that there should be a strategy of making the factory site unattractive to birds by denying food and harbourage, otherwise colonies can become established and cause serious problems. Open waterways can attract birds, insects, vermin, etc. and should be enclosed in culverts if possible. Entrances that have to be lit at night should be lit from a distance with the light

© 2008, Woodhead Publishing Limited

266

•

• •

•

• •

• •

Chilled foods

directed to the entrance, rather than lit from directly above. This prevents flying insects being attracted directly to the entrance. Some flying insects require water to support part of their life cycle (e.g. mosquitoes), and experience has shown that where flying insects can occasionally be a problem, all areas where water could collect or stand for prolonged periods of time (waterways, old buckets, tops of drums, etc.) need to be removed or controlled. All waste containers should be lidded to prevent attraction to pests, and any spillages should be cleared up as soon as possible. Waste containers should be regularly emptied. Processes likely to create microbial or dust aerosols, e.g. effluent treatment plants, waste disposal units or any preliminary cleaning operations, should be sited such that prevailing winds do not blow contaminants directly into manufacturing areas. An area of at least 3 m immediately adjacent to buildings should be kept free of vegetation and covered with a deep layer of gravel, stones, paving or roadway, etc. (Katsuyama and Strachan, 1980; Troller, 1983) This practice helps weed control, assists inspection of bait boxes and traps and helps maintain control of the fabric of the factory building. Storage of equipment, utensils, pallets, etc. outside should be avoided wherever possible as they present opportunities for pest harbourage. Wooden pallets stacked next to buildings are also a known fire hazard. Siting of process steps and storage facilities outside (for example silos, water tanks, and packaging stores) should be avoided wherever possible. If not possible, they should be suitably locked off so that people or pests cannot gain unwanted access to food materials. Equipment necessary to connect transport devices to outside storage facilities (e.g. discharge tubing and fittings between tankers and silos) should also be locked away when not in use. Wherever possible loading and unloading operations should be undertaken under cover.

11.4 Barrier 2 – the factory building The building structure is the second and a major barrier, providing protection for raw materials, processing facilities and manufactured products from contamination or deterioration. Protection is both from the environment, including rain, wind, surface runoff, delivery and dispatch vehicles, dust, odours, pests and uninvited people, etc., and internally from microbiological hazards (e.g. raw material cross-contamination), chemical hazards (e.g. cleaning chemicals, lubricants) and physical hazards (e.g. from plantrooms, engineering workshops). Ideally, the factory buildings should be designed and constructed to suit the operations carried out in them and should not place constraints on the process or the equipment layout. © 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

267

With respect to the external environment, whilst it is obvious that the factory cannot be a sealed box, openings to the structure must be controlled. There is also little legislation controlling the siting of food factories and what can be built around them. The responsibility, therefore, rests with the food manufacturer to ensure that any hazards (e.g. micro-organisms from landfill sites or sewage works, or particulates from cement works, or smells from chemical works) are excluded via appropriate barriers. Detailed information on the hygienic design requirements for the construction of the external envelope and internal structure of the factory is not easily found, though the Campden and Chorleywood Food Research Association (CCFRA) has published a suite of design guidelines (Anon., 2002, 2003a,b, 2005). A typical example of a suitable outside wall structure, incorporating new design concepts related to hazard control, is shown in Fig. 11.3. The diagram shows a well-sealed structure that resists pest ingress and is protected from external vehicular damage. The ground floor of the factory is also at a height above the external ground level. By preventing direct access into the factory at ground floor level, the entrance of contamination (mud, soil, foreign bodies, etc.), particularly from vehicular traffic (forklift trucks, raw material delivery, etc.) is restricted. This is now seen as very

Fig. 11.3

Schematic diagram of an external wall construction preventing the ingress of soil (and associated micro-organisms), foreign bodies and pests.

© 2008, Woodhead Publishing Limited

268

Chilled foods

important in preventing microbiological challenges on the high care/risk barrier as it is now generalised that microbiological pathogens of concern can arise from two sources: the intestines of animals, usually entering the factory in raw materials, and the external environment (either directly from the environment, e.g. wind-blown dust, or with raw materials, e.g. soil on vegetables). By restricting soil ingress we may well reduce the challenge of environmental pathogens such as Listeria spp., Bacillus spp and Clostridia spp. Key hygienic design factors include:

• Wherever possible, buildings should be single storey or with varying headroom

•

• •

• • • •

featuring mezzanine floors to allow gravity flow of materials where this is necessary (Imholte, 1984). This prevents any movement of wastes or leaking product moving between floors. Whilst ideally the process line should be straight, this is rarely possible, but there must be no backtracking and, where there are changes in the direction of process flow, there must be adequate physical barriers. Enough chilled storage provision must be made, particularly in high risk, to ensure product safety in the event of line stoppage, etc. Pipework or process lines carrying product, waste or services, etc. that presents a hazard to open product should not pass directly over (above) such product lines. The layout should also consider that provision is made for the space necessary to undertake the process and associated quality control functions, both immediately the factory is commissioned and in the foreseeable future. In addition to process areas, provision may have to be made for a wide range of support activities including raw material storage; packaging storage; water storage; wash-up facilities; plant room; engineering workshop; cleaning stores; microbiology, chemistry and quality control laboratories; test kitchens; pilot plant; changing facilities; restrooms; canteens; medical rooms; observation areas/ viewing galleries; and finished goods dispatch and warehousing. The siting and construction of factory openings should be designed with due consideration for prevailing environmental conditions, particularly wind direction and drainage falls. The floor may be considered as one of the most important parts of a building because it forms the basis of the entire processing operation. It is thus worthy of special consideration and high initial capital investment. The structural floor slab is the most important aspect of the floor and should be soundly constructed with appropriate waterproof membranes and movement joints. Both tiles and synthetic resins are suitable as a surface topping, with tiles being favoured if the predicted use of the processing area (without change) extends for 5 or more years. If change in the processing area is foreseen before this time, an appropriate resin floor of 6–8 mm thickness is appropriate. Advice on tiles can be obtained from the Tile Association (www.tiles.org.uk) and on resin floors from the Resin Flooring Association (www.fefra.org.uk). Ultimately, the skill of the contractors laying the floor is at least as important as the choice of topping

© 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

•

• •

•

•

• •

•

269

and efforts should be made to get references from previous clients or indeed inspect previously laid floors. The floor should be coved where it meets walls or other vertical surfaces such as plinths or columns as this facilitates cleaning. The food industry has had serious problems with slips and trips related to floors, and the Health and Safety Executive (in the UK) have been pro-active in providing the industry with advice. Wherever possible, floors should be specified to be non-slip and, unless the surface structure is heavily embossed with valley to peak heights exceeding 3–4 mm (i.e. when brush bristles cannot penetrate to the bottom of the valley), this should have no effect on cleanability. However, although flooring materials designed to be slip-resistant may have an average peak to valley height measured in millimetres, the size of imperfection that could harbour micro-organisms would be measured in microns (Anon., 2002; Taylor and Holah, 1996; Mettler and Carpentier, 1998). Such micronsized holes are equally likely to be found on ‘smooth’ as ‘rough’ surfaces. Flooring and walling materials manufacturers should also be able to demonstrate that their materials have been tested for absence of water absorption and ease of cleanability so as to meet current EU legislation. Satisfactory drainage can be achieved only if adequate falls to drainage points are provided. Falls should be in the range 1 in 50 to 1 in 80, with the slope being steeper with high water volumes or slip resistant toppings. Falls greater than 1 in 40 may introduce operator safety hazards and also cause problems with wheeled vehicles. The type of drain used depends to a great extent upon the process operation involved. For operations involving a considerable amount of water and solids, channel drains are often the most suitable (Fig. 11.4a). For operations generating volumes of water but with little solids, aperture channel drains are more favourable (Fig. 11.4b). Channel gratings must be easily removable. The fall to drain should not exceed 5 m. The drainage system should flow from high to low risk and, whenever possible, backflow from low risk to high risk areas should be impossible. This is best achieved by having separate low and high risk drains running to a master collection drain with an air-break between each collector and master drain. The drainage system should also be designed such that rodding points are outside high risk areas. If possible, the high risk drains should enter the collection drain at a higher point than the low risk drains, so that if flooding occurs, low risk areas will flood first. The edge of the channel rebate must be properly designed and constructed to protect it, by an angle, from damage (Fig. 11.4), particularly if wheeled vehicles are in use. The author is aware of a number of cases where the seal has been broken between the drain channel and floor structure so as to leave an uncleanable void between the two. When the channel was subsequently walked upon or traversed by wheeled vehicles, a small volume of foul liquid and Listeria was ‘pumped’ to the floor surface. In addition, drainage systems have been observed to act as air distribution channels, allowing contaminated air movement between rooms. This can typically occur when the drains are little used and the water traps dry out.

© 2008, Woodhead Publishing Limited

270

Chilled foods

(a)

(b)

Fig. 11.4

(a) Half-round drainage channel with reinforced rebate for grating and (b) stainless steel aperture channel drain.

Modular insulated panels are now used very widely for non-load-bearing walls. The panels are made of a core of insulating material between 50 and 200 mm thick, sandwiched between steel sheets, which are bonded to both sides of the core. Careful consideration must be given to the choice of filling which must comply with current fire retardation (and insurance) standards. The modules are designed to lock together and allow a silicone sealant to provide a hygienic seal between the units. The modules are best mounted on a concrete upstand or plinth or stainless steel channel filled with concrete (Fig. 11.5). The latter provides useful protection against the possibility of damage from vehicular traffic, particularly forklift trucks. However, it should be appreciated that this arrangement reduces the possibility of relatively easy and inexpensive changes to room layout to meet future production requirements Load-bearing and fire-break walls are often constructed from brick or blockwork. These walls can either be directly painted with an appropriate waterproof coating or rendered with a cement and sand screed to achieve a better surface smoothness for the coating layer. The covering of walls by other materials such as sheets of plastic or stainless steel is not recommended unless the walls are absolutely flat, so as to prevent the harbourage of cleaning water or pests between the two walling materials. Walls and other vulnerable structures such as doors and pillars should be protected by suitable barriers. Barriers may be wall or floor mounted and should be designed to prevent collisions with the wall or other vulnerable structures by the specific types of transport systems, racking, totebins, etc. used within the factory. © 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

Silicone sealant

271

Silicone sealant

Panel base channel

200 mm

200 mm

Resin cove

Coved tile

Screed Slab Resin floor Screed may be removed prior to casting kerb

Fig. 11.5

Tiled floor Optional steel dowels to locate kerb

Modular insulated panel located in U-channel and fixed to a concrete plinth.

Other hygienic design features include:

• The number and size of openings should be kept to a minimum and exterior

• •

•

doors should not open directly into production areas. External doors should always be shut when not in use and, if they have to be opened regularly, should be of a rapid opening and closing design. Doors should be constructed of metal, glass reinforced plastic or plastic, selfclosing, designed to withstand the intended use and misuse, and be suitably protected from vehicular damage where applicable. Plastic strips/curtains are acceptable in interior situations only as they are easily affected by weather. Where necessary, internal or external porches can be provided with one door, usually the external door on an external porch, being solid and the internal door being a flyscreen door; on an internal porch it would be the opposite configuration. Air jets or curtains directed over doorways, designed to maintain temperature differentials when chiller/freezer doors are opened, may have a limited effect on controlling pest access. For many food manufacturers and retailers, glass is seen as the second major food hazard after pathogenic micro-organisms. For this reasons, glass should be avoided as a construction material (windows, inspection mirrors, instrument and clock faces, etc.). If used, e.g. as viewing windows to allow visitor or management observation, a glass register, detailing all types of glass used in the factory, and their location, should be compiled.

© 2008, Woodhead Publishing Limited

272

Chilled foods

• Windows should be glazed with either polycarbonate or laminated glass. Where

• • •

•

opening windows are specifically used for ventilation (particularly in tropical areas), these must be screened and the screens must be designed to withstand misuse or attempts to remove them. Flyscreens should be constructed of stainless steel mesh and be removable for cleaning. The flow of air and drainage should be away from ‘clean’ areas, towards ‘dirty’ ones. The flow of discarded outer packaging materials should not cross, or run counter to, the flow of either unwrapped ingredients or finished products. If a filtered air supply is required to processing areas and the supply will involve ducting, a minimum level of filtration of > 90% of 5 micron particles is required, e.g. G4 or F5 filters (BS EN 779, 1979), to provide both suitably clean air and prevent dust accumulation in the ductwork. Lighting should be a minimum of 500–600 lux for food inspection purposes and should ideally be flush mounted to the ceiling to facilitate bulb changing from outside the food processing area. Alternatively, lighting units should be suspended and plugged in so that in the event of a failure the entire unit can be replaced and the faulty one removed from the processing areas to a designated workshop for maintenance/bulb replacement. Open lighting elements (i.e. those not within enclosed lighting units) should be protected by plastic sleeves.

Within the internal environment, most factories are segregated into food production areas (raw material storage, processing, final product storage and dispatch) and amenities (reception, offices, canteens, training rooms, engineering workshops, boiler houses, etc.). The prime reason for this is to clearly separate the food production processes from the other activities that the manufacturer must perform. This may be to control microbiological or foreign body hazards arising from the amenity functions, but is always undertaken to foster a ‘you are now entering a food processing area’ hygienic mentality in food operatives. Food production areas are typically segregated into raw material intake, raw material storage, processing, packaging and final product warehouse and dispatch. In addition, the flow of ingredients and products is such that, in ideal conditions, raw materials enter at one end of the factory (dirty end) and are dispatched at the opposite end (clean end). Whilst a range of ingredients is brought together for processing, they may need to be stored separately. Storage may be temperature orientated (ambient, chilled or frozen) or ingredient related, and separate stores may be required for fruit and vegetables, meat, fish, dairy and dry ingredients. Other food ingredients such as allergens, and non-ingredients such as packaging, should also be stored separately. Segregation may also extend into the first stages of food processing, where, for example, the production of dry intermediate ingredients, e.g. pastry for pies, is separated from the production of the pie fillings. The degree of segregation for storage and processing of ingredients and intermediates is predominantly controlled by the exclusion of water, particularly in how they are cleaned (Anon., 2001a), viz: © 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

273

• Dry cleaning. This applies to areas where no cleaning liquids are used, only

•

vacuum cleaners, brooms, brushes, etc. Whilst these areas are normally cleaned dry, occasionally they may be fully or partially wet cleaned, when limited amounts of water are used. Wet cleaning. This applies to areas where the entire room or zone is always cleaned wet. The contents (equipment, cable trays, ceilings, walls, etc.), are wet washed without restrictions on the amount of cleaning liquid used.

In addition to segregating dry areas because of a requirement to exclude water, other areas may need to be segregated due to excessive use of water, which can lead to the formation of condensation and the generation of aerosols. Such areas include traywash and other cleaning areas. Ideally, manufacturers who manufacture allergenic and non-allergenic products should do so on separate sites such that there is no chance of cross-contamination from different ingredients. However, this issue has been debated by food manufacturers in both Europe and the USA with the conclusion that it is unlikely to be economically viable to process on separate sites. Segregation of allergenic components will therefore have to be undertaken within the same site. As a preferred alternative to separate factories, it may be possible to segregate the whole process, from goods in through raw material storage and processing to primary packaging, on the same site. If this is not possible, segregation has to be undertaken by time, e.g. by manufacturing non-allergen containing products first and then manufacturing allergen-containing products last. Thorough cleaning and disinfection is then undertaken before the manufacture of the non-allergen containing products is recommenced. If segregation by time is to be considered, a thorough HACCP study should be undertaken to consider all aspects of how the allergen is to be stored, transported, processed and packed, etc. This would include information on any dispersal of the allergen during processing (e.g. from weighing, mixing), the fate of the allergen through the process (will its allergenic attributes remain unchanged), the degree to which the allergen is removed by cleaning, and the effect of any dilution of residues remaining in the subsequent product flow after cleaning. This may be complicated in that other hazards may be seen as more important. For example, some manufacturers prefer to segregate their production on microbiological risk (e.g. in a sandwich operation the manufacturer produces fully-cooked fillings first and then salad fillings, which are more likely to contain spoilage organisms). To a lesser extent and because it is not a safety issue, label declaration issues such as non-organic components in organic foods, GMO components in GMOfree products, vegetarian foods with non-vegetarian components, and ‘non-religion’ processed components in religious based foods (e.g. kosher or halal), have all caused food manufacturers to think about how raw materials are segregated. As for allergenic materials, segregation is usually by time and by the use of separate ingredient stores. Stores containing key components, e.g. meat in a factory producing vegetarian components, are often locked to prevent inadvertent use of © 2008, Woodhead Publishing Limited

274

Chilled foods

these ingredients when not scheduled, and the locking and unlocking of such stores can be recorded in the quality system.

11.5 Barrier 3 – high risk production area The third barrier within a factory segregates an area in which food products are further manipulated or processed following a decontamination treatment. It is, therefore, an area into which a food product is moved after its microbiological content has been reduced. Many names have been adopted for this third level processing area including ‘clean room’ (or ‘salle blanche’ in France) following pharmaceutical terminology, ‘high hygiene’, ‘high care’ or ‘high risk’ area. In some sectors, particularly chilled, RTE foods, manufacturers have also adopted opposing names to describe second barrier areas such as ‘low risk’ or low care’. Much of this terminology is confusing, particularly the concepts of ‘low’ areas which can imply to employees and other people that lower overall standards are acceptable in these areas where, for example, operations concerned with raw material reception, storage and initial preparation are undertaken. In practice, all operations concerned with food production should be carried out to the highest standard. Unsatisfactory practices in so-called low risk areas may, indeed, put greater pressures on the barrier system separating the second and third level processing areas. The Chilled Food Association in the UK (Anon., 2006) have established guidelines to describe the hygiene status of chilled foods and indicate the area status of where they should be processed after any heat treatment. Three levels are described: high risk area (HRA), high care area (HCA) and good manufacturing practice (GMP). Their definitions are: HRA. An area to process components, all of which have been heat treated to ≥ 90 °C for 10 mins or ≥ 70 °C for 2 mins, and in which there is a risk of contamination between heat treatment and pack sealing that may present a food safety hazard. HCA. An area to process components, some of which have been heat treated to ≥ 70 °C for 2 mins, and in which there is a risk of contamination between heat treatment and pack sealing that may present a food safety hazard. GMP. An area to process components, none of which has been heat treated to ≤ 70 °C for 2 mins, and in which there is a risk of contamination prior to pack sealing that may present a food safety hazard. In practice, the definition of HCA has been extended by the author to include an area to further process components, none of which has undergone a pasteurisation treatment but which have been decontaminated. For example, a high care area is appropriate for products that have undergone a wash in chlorinated water, e.g. fruit and salad produce, or fish after low temperature smoking and salting. Many of the requirements for the design of HRA and HCA operations are the same, with the emphasis on preventing contamination in HRA and minimising © 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

275

contamination in HCA operations (Anon., 2006). In considering whether a high risk or high care is required, and therefore what specifications should be met, chilled food manufacturers need to carefully consider their existing and future product ranges, the hazards and risks associated with them, and possible developments in the near future. If budgets allow, it is always cheaper to build to the highest standards from the onset of construction rather than try to retrofit or refurbish at a later stage. The requirements for third barrier level high care/risk segregation for appropriate foodstuffs is now recognised by the major food retailers worldwide and is a requirement in the BRC Global Food Standard (2005) (www.brc.org.uk) and the Global Food Safety Initiative Guidance Document (2004) (www.ciesnet.com/). In general, high care/risk areas should be as small as possible, as their maintenance and control can be very expensive. If there is more than one high care/risk area in a factory, they should be arranged together or linked as much as possible by closed corridors of the same class. This is to ensure that normal working procedures can be carried out with a minimum of different hygienic procedures applying. Some food manufacturers design areas between the second low risk and third barrier high risk level zones and use these as transition areas. These are often termed ‘medium care’ or ‘medium risk’ areas. These areas are not separate areas in their own right as they are freely accessed from low risk without the need for the protective clothing and personnel hygiene barriers, as required at the low/high risk area interface. By restricting activities and access to the medium risk area from the low risk area, however, these areas can be kept relatively ‘clean’ and thus restrict the level of microbiological contamination immediately adjacent to the third level barrier. To simplify the text, the term high risk is used to cover both high care and high risk, unless specified otherwise.

11.5.1 Listeria philosophy In terms of chilled food product safety, the major contamination risk is microbiological, particularly from the pathogen most commonly associated with the potential to grow in chilled foods, Listeria monocytogenes. For many chilled food products, L. monocytogenes could well be associated with the raw materials used and thus may well be found in the low risk area. After the product has been heat processed or decontaminated (e.g. by washing), it is essential that all measures are taken to protect the product from cross-contamination from (predominantly) low risk, L. monocytogenes sources (e.g. materials or the processing environment). Similarly, other pathogens and potential foreign body contamination that would jeopardise the wholesomeness of the finished product could also be found in low risk. A threefold philosophy has been developed by the author to help reduce the incidence of L. monocytogenes in finished product and, at the same time, control other contamination sources. (i)

Provide as many barriers as possible related to the building structure and processing practices to prevent the entry of Listeria into the high risk area

© 2008, Woodhead Publishing Limited

276

Chilled foods

and to minimise locations within high risk in which Listeria could become harboured. (ii) Prevent the growth and spread of any Listeria penetrating these barriers during production. (iii) After production, employ a suitable sanitation system to ensure that all Listeria are removed from high risk prior to production recommencing. The building structure, facilities and practices associated with the high risk production and assembly areas provide the third and inner barrier protecting chilled food manufacturing operations from contamination. This barrier is built up by the use of combinations of a number of separate components or sub-barriers to control contamination that could enter high risk from the following routes:

• • • • • • • • • •

structural defects product entering high risk via a heat process product entering high risk via a decontamination process product entering high risk that has been heat processed/decontaminated off-site but whose outer packaging may need decontaminating on entry to high risk other product transfer packaging materials liquid and solid waste materials food operatives entering high risk air supply utensils, which may have to be passed between low and high risk.

11.5.2 Structure Structurally, creating a third barrier level has been described by Ashford (1986) as creating a box within a box (Fig. 11.6). In other words, the high risk area is sealed on all sides to prevent microbial ingress. Whilst this is an ideal situation, openings are still needed to the box to allow access for people, ingredients and packaging and exit for finished product and wastes. Openings should be as few as possible, as small as possible (to better maintain an internal positive pressure) and should be controlled (and shut if possible) at all times. Similarly, the size of the ‘box’ should be as small as practically possible for a good layout and future expansion, to reduce costs, particularly for any HVAC system, and the perimeter of the box should be inspected frequently to ensure that all joints are fully sealed. The design of the high risk food processing area must allow for the accommodation of five basic requirements, viz:

• • • • •

processed materials and possibly some ingredients processing equipment staff concerned with the operation of such equipment packaging materials finished products.

There is a philosophy which has considerable support, which states that all other © 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

277

Services

Ring mains

Ceiling panels supported off structural framework Recessed lighting

Local drops to specific equipment

Fig. 11.6

õ

õ

Catwalk

Recessed lighting

High care production area (Internal walls not showing)

Basic design concepts – the separation of production from services and maintenance operations.

requirements should be considered as secondary to these five basic requirements and, wherever possible, should be kept out of the high risk processing area. This reduces potential environmental niches for micro-organisms, and aids in cleaning and disinfection, and thus contamination control. These secondary requirements include:

• structural steel framework of the factory • service pipework for water, steam and compressed air; electrical conduits and trunking; artificial lighting units; and ventilation ducts

• compressors, refrigeration units and pumps • maintenance personnel associated with any of these services • ‘furniture’ and computers, etc. associated with office facilities. In addition to being as small as possible, the splitting down of large processing areas into smaller sub-units (e.g. a single twelve-line meat slicing hall into three fully segregated sub-units of four slicing lines) can help in eliminating crosscontamination between lines. Such segregation is also now considered as a method of increasing manufacturing flexibility, for example when some lines need to be shut down for cleaning or maintenance whilst the others need to remain in production.

11.5.3 Heat-treated product Where a product heat treatment forms the barrier between low and high risk (e.g. an oven, fryer or microwave tunnel), two points are critical to facilitate its successful operation:

• All product passing through the heat barrier must receive its desired cooking © 2008, Woodhead Publishing Limited

278

•

Chilled foods

time/temperature combination. This means that the heating device should be performing correctly (e.g. temperature distribution and maintenance are established and controlled and product size has remained constant) and that it should be impossible, or very difficult, for product to pass through the heat treatment without a cook process being initiated and completed. The heating device must be designed, as far as is possible, to form a solid, physical barrier between low and high risk. Where it is not physically possible to form a solid barrier, air spaces around the heating equipment should be minimised and the low/high risk floor junction should be fully sealed to the highest possible height.

The fitting of the heating devices that provide heat treatment within the structure of a building presents two main difficulties. Firstly, the devices have to be designed to load product on the low risk side and unload in high risk. Secondly, the maintenance of good seals between the heating device surfaces, which cycle through expansion and contraction phases, and the barrier structure, which has no thermal expansion, is problematical. Of particular concern are ovens and the author is aware of the following issues:

• Ovens should drain to the low risk area, but some ovens have been designed to

•

•

•

•

drain into high risk. This is unacceptable for the following reason. It may be possible for pathogens present on the surface of product (which is their most likely location if they have been derived from cross-contamination in low risk) to fall to the floor through the melting of the product surface layer (or exudate on overwrapped product) at a temperature that is not lethal to the pathogen. The pathogen could then remain on the floor or in the drain of the oven in such a way that it could survive the cook cycle. On draining, the pathogen would then subsequently drain into high risk. Pathogens have been found at the exit of ovens in a number of food factories and this area should always be a pathogen sampling point within the environmental sampling plan. Problems have occurred with leakage from sumps under the ovens into high risk. Sumps can collect debris and washing fluids from the oven operation which can facilitate the growth of Listeria. Oven floors should be sloped and these areas should be routinely cleaned (from low risk). Where the floor of the oven is cleaned, cleaning should be undertaken in such a way that cleaning solutions do not flow from low to high risk. Ideally, cleaning should be from low risk with the high risk door closed and sealed. If cleaning solutions have to be drained into high risk, or in the case of ovens that have a raining water cooling system, a drain should be installed immediately outside the door in high risk. Heating devices should be designed to load product on the low risk side and unload in high risk. If oven racks of cooked product have to be transferred into high risk for unloading, these racks should be returned to low risk via the ovens, with an appropriate thermal disinfection cycle as appropriate. The design of small batch product blanchers or noodle cookers (i.e. small vessels with water as the cooking medium) does not often allow the equipment

© 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

279

to be sealed into the low/high risk barrier as room has to be created around the blancher to allow product loading and unloading. Condensation is likely to form because of the open nature of these cooking vessels and it is important to extract steam from the area to prevent microbial build-up where water condenses. Condensation forming in the extraction or ventilation system should be drained away hygienically. Any ventilation system should be designed so that the area is ventilated from low risk; ventilation from high risk can draw into high risk large quantities of low risk air. Early installations of open cooking vessels (kettles) as barriers between low and high risk were occasionally constructed with level retaining or bund walls to prevent water movement across the floor and barriers at waist height to prevent the movement of people; whilst innovative in their time, they are now seen as hygiene hazards (Fig. 11.7a). It is virtually impossible to prevent the transfer of contamination, by people, the air and via cleaning, between low and high risk. Such barriers were then improved by fully segregating the floor junction and decanting product into high risk via open hatches (Fig. 11.7b). These are still occasionally found for products that are not suitable for pumping for quality reasons. Later installations had the kettles fully separated from high risk and transferred cooked product (by pumping, gravity, vacuum, etc.) through into high risk via a pipe in the dividing wall (Fig. 11.7c). The first such installations had the kettle output pipe passing through the wall very close to the floor, which led to Listeria contamination problems during unloading as operatives were putting trays directly onto the floor into which the cooked product was unloaded. The kettles need to be positioned in low risk at a height such that the transfer into high risk is well above ground level, usually facilitated by mounting the kettles on a mezzanine (Fig. 11.7d). Pipework connections through the walls should be cleaned from high risk such that potentially contaminated low risk area cleaning fluids do not drain into high risk.

11.5.4 Product decontamination Fresh produce to be processed in a high care area should enter it via a decontamination operation, usually involving a washing process with the clean washwater incorporating a biocide. The use of chlorinated dips , mechanically stirred washing baths or ‘Jacuzzi’ washers are the most common method, though alternative biocides are also used (e.g. bromine, chlorine dioxide, ozone, organic acids, peracetic acid, hydrogen peroxide). In addition, it is now seen as increasingly important, following a suitable risk assessment, to decontaminate the outer packaging of various ingredients on entry into high risk (e.g. product cooked elsewhere and transported to be processed in the high risk area, canned foods and some overwrapped processed ingredients). Where the outer packaging is likely to be contaminated with food materials and/or micro-organisms, decontamination is best done using a washing process incorporating a disinfectant (usually a quaternary ammonium compound). If the packaging is clean and has a surface free of wrinkles, etc., the use of UV light has the advantage that it is dry and thus limits © 2008, Woodhead Publishing Limited

280

Chilled foods

(a)

(b)

Fig. 11.7 (a) Schematic early low risk (white coated worker)/high risk divide around kettles; (b) Improved segregation in that the floor wall junction has been sealed and the opening between low and high risk has been reduced to the size of the hatch through which the product is decanted.

© 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

281

(c)

(d)

Fig. 11.7 (c) Complete segregation in which cooked product is gravity fed or pumped into high risk through pipework. (d) More acceptable schematic arrangement in which cooking equipment is mezzanine mounted so that the cooked product discharges at a height well above floor level. © 2008, Woodhead Publishing Limited

282

Chilled foods

the spread of water into high risk with its associated potential for environmental microbial growth. For companies that also have ovens with low risk entrance and high risk exit doors, it is also possible to transfer product from low to high risk via these ovens using a short steaming cycle that offers surface pasteurisation without ‘cooking’ the ingredients. Decontamination systems have to be designed and installed such that they satisfy three major criteria: (i)

As with heat barriers, decontamination systems need to be installed within the low/high risk barrier to minimise the free space around them. As a very minimum, the gap around the decontamination system should be smaller than the product (e.g. unit pack) to be decontaminated. This ensures that all ingredients in high risk must have passed through the decontamination system and thus must have been decontaminated (it is impossible to visually assess whether the outer surface of an ingredient has been disinfected, in contrast to whether an ingredient has been heat processed). When installing washing systems, as much of the system as possible should be in low risk to minimise water transfer into high risk. Some systems incorporate an airknife for additional drying of decontaminated articles before high risk entry. (ii) Prior to installation, the decontamination process should be established and validated. For a wet process, this will involve the determination of a suitable disinfectant that combines detergency and disinfectant properties and a suitable application temperature, concentration and contact time. Similarly, for UV light, a suitable wavelength, intensity pack orientation and contact time should be determined. The same degree of decontamination should apply to all the product surfaces or, if this is not possible, the process should be established for the surface receiving the least treatment. (iii) After installation, process controls should be established and may include calibrated, automatic disinfectant dosing, fixed speed conveyors, UV light intensity meters, etc. In-process monitoring may include the periodic checking for critical parameters, for example blocked spray nozzles or UV lamp intensity and, from the low risk side, the loading of the transfer conveyor to ensure that product is physically separated such that all product surfaces are exposed.

11.5.5 Other product transfer It is now seen as poor practice to bring outer packaging materials (e.g. corrugated cardboard) into high risk. All ingredients and product packaging must, therefore, be de-boxed and transferred into high risk. Some ingredients, such as bulk liquids that have been heat-treated or are inherently stable (e.g. oils or pasteurised dairy products), are best handled by being pumped across the low/high risk barrier, directly to the point of use, from bulk containers. Dry, stable bulk ingredients (e.g. sugar) can also be transferred into high risk via sealed conveyors linked to silos or pallet sized containers. For non-bulk quantities, it is possible to open ingredients at the low/high risk © 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

283

barrier and decant them through into high risk via a suitable transfer system (e.g. a simple funnel set into the wall), into a receiving container. Transfer systems should, preferably, be closable when not in use and should be designed to be cleaned and disinfected from the high risk side, prior to use as appropriate.

11.5.6 Packaging Packaging materials (film reels, cartons, containers, trays, etc.) are best supplied to site ‘double bagged’. This involves a cardboard outer followed by two plastic bag layers surrounding the packaging materials. The packaging is brought on site, deboxed, inspected for damage and stored double bagged until use in a suitable packaging store. When called for in high risk, the packaging material is brought to the low/high risk barrier, the outer plastic bag removed and the inner bag and packaging enters high risk through a suitable hatch. The second plastic bag keeps the packaging materials covered until they are loaded onto the line or the packaging machine. The hatch, as with all openings in the low/high risk barrier, should be as small as possible and should be closable when not in use. This is to reduce airflow through the hatch and thus reduce the airflow requirements for the air handling systems to maintain high risk positive pressure. For some packaging materials, especially heavy film reels, it may be required to use a conveyor system for moving materials through the hatch. A full-size opening door or, preferably, double door airlock should be used only if the use of a hatch is not technically possible and suitable precautions must be taken to decontaminate an airlock after use. Airlocks should be interlocked so that only one door can be open at any one time.

11.5.7 Liquid and solid wastes On no account should low risk liquid or solid wastes be removed from the factory via high risk and attention is required to the procedures for removing high risk wastes. The handling of liquid wastes from low and high risk is described under the factory building section (Section 11.4). Solid wastes that have fallen on the floor, equipment, etc. through normal production spillages, should be bagged-up or placed in easily cleanable bins on an on-going basis commensurate with good housekeeping practices. It may also be necessary to remove solid waste product from the line at break periods or to facilitate line product changes. Waste bags should leave high risk in such a way that they minimise any potential cross-contamination with processed product and should, preferably, be routed in the reverse direction to the product. For small quantities of bagged waste, existing hatches should be used, e.g. the wrapped product exit hatches or the packaging materials entrance hatch, as additional hatches increase the risk of external contamination and put extra demands on the air handling system. For waste collected in bins, it may be necessary to decant the waste through purpose built, easily cleanable from high risk, waste chutes that deposit directly into waste skips. Waste bins should be colour coded to differentiate them from other food containers and should be used only for waste. © 2008, Woodhead Publishing Limited

284

Chilled foods

11.5.8 Personnel Within the factory building, provision must be made for adequate and suitable staff facilities, including amenities for changing, washing and eating. There should be lockers for storing outdoor clothing in areas that must be separate from those for storing work clothes. Toilets must be provided and must not open directly into food processing areas, all entrances of which must be provided with hand washing facilities arranged in such a way that their ease of use is maximised. With regard to operatives in high risk areas, personnel facilities and requirements must be provided in a way that minimises any potential contamination of high risk operations. The primary sources of potential contamination arise from the operatives themselves and from low risk operations. This necessitates further attention to protective clothing and, in particular, special arrangements and facilities for changing into high risk clothing and entering high risk. Best practice with respect to personnel hygiene is continually developing and has been reviewed by Guzewich and Ross (1999), Taylor and Holah (2000) and Taylor et al. (2000). Factory clothing for high risk areas does not necessarily vary from that used in low risk in terms of style or quality, though it may have to receive higher standards of laundry, especially related to a higher temperature process, sufficient to significantly reduce microbiological levels. Indeed some laundries now operate to the same low/high risk principles as the food industry such that dirty laundry enters ‘low risk’, is loaded into a washing machine that bridges a physical divide, is cleaned and disinfected and exits into ‘high risk’ to be dried and packed. Additional clothing may be worn in high risk, however, to further protect the food being processed from contamination arising from the operative’s body (e.g. gloves, sleeves, masks, whole head coveralls, coats with hoods, boiler suits, etc.). All clothing and footwear used in the high risk area must be colour-coded to distinguish it from that worn in other parts of the factory and to reduce the chance that a breach in the systems would escape early detection. High risk footwear should be captive to high risk, i.e. it should remain within high risk, operatives changing into and out of footwear at the low/high risk boundary. This has arisen because research has shown that boot baths and boot washers are unable to adequately disinfect low risk footwear such that they can be worn in both low and high risk and decontaminated between the two (Taylor et al., 2000). Essentially, they do not remove all organic material from the treads and any pathogens within the organic material remaining are protected from any subsequent disinfectant action. In addition, boot baths and boot washers can both spread contamination via aerosols and water droplets that, in turn, can provide moisture for microbial growth on high risk floors. Bootwashers were, however, shown to be very good at removing large quantities of organic material from boots and are thus a useful tool in low risk areas to both clean boots and help prevent operative slip hazards. The high risk changing room provides the only entry and exit point for personnel working in or visiting the area and is designed and built to both house the necessary activities for personnel hygiene practices and minimise contamination from low risk. In practice, there are some variations in the layout of facilities of © 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

285

high risk changing rooms. This is influenced, for example, by space availability, product throughput and type, which will affect the number of personnel to be accommodated, and whether the changing room is a barrier between low and high risk operatives or between operatives arriving from outside the factory and high risk. Generally, higher construction standards are required for low/high risk barriers than outside/high risk barriers because the level of potential contamination in low risk, both on the operatives hands and in the environment, is likely to be higher (Taylor and Holah, 2000). In each case, the company must evaluate the effectiveness of the changing-room layout and procedure to ensure that the high risk area and products prepared in it are not being put at risk. This is best undertaken by a HACCP approach, so that data are obtained to support or refute any proposals regarding the layout or sequence. Research at the CCFRA has also proposed the following hand hygiene sequence to be used on entry to high risk (Taylor and Holah, 2000). This sequence has been designed to maximise hand cleanliness, minimise hand transient microbiological levels, maximise hand dryness yet at the same time reduce excessive contact with water and chemicals that may both lead to dermatitis issues for the operatives and reduce the potential for water transfer into high risk. (i) (ii) (iii) (iv) (v)

(ix) (x) (xi)

Remove low risk or outside clothing. Remove low risk/outside footwear and place in designated ‘cage’ type compartment. Cross over the low risk/high risk dividing barrier. WASH HANDS. Put on in the following order: (vi) high risk captive footwear; (vii) hair net; put on over ears and covering all hair; (plus beard snood if needed) and hat (if appropriate) (Note: Some RTE food manufactur ers prefer to put on hairnets prior to step (iii)); (viii) overall (completely buttoned up to neck). Check dress and appearance in the mirror provided. Go into the high risk production area and apply an alcohol-based sanitizer to hands. Draw and put on disposable gloves, sleeves and apron, if appropriate.

A basic layout for a changing room is shown in Fig. 11.8 and has been designed to accommodate the above hand hygiene procedure and the following requirements:

• An area at the entrance to store outside or low risk clothing. Lockers should have sloping tops.

• A barrier to divide low and high risk floors. This is a physical barrier such as a

• •

small wall (approximately 60 cm high), that allows floors to be cleaned on either side of the barrier without contamination by splashing, etc. between the two. Open lockers at the barrier to store low risk footwear. A stand on which footwear is displayed/dried, usually soles uppermost.

© 2008, Woodhead Publishing Limited

286 Chilled foods

Fig. 11.8 © 2008, Woodhead Publishing Limited

Schematic layout for a high risk changing room.

The hygienic design of chilled food plants and equipment

287

• An area designed with suitable drainage for bootwashing operations. Research

•

• • • • •

has shown (Taylor et al., 2000) that manual cleaning (preferably during the cleaning shift) and industrial washing machines are satisfactory bootwashing methods. Hand wash basins. Hand wash basins must have automatic or knee/foot operated water supplies, water supplied at a suitable temperature (that encourages hand washing) and a waste extraction system piped directly to drain. It has been shown that hand wash basins positioned at the entrance to high risk, which was the original high risk design concept to allow visual monitoring of hand wash compliance, gives rise to substantial aerosols of Staphylococcal strains (Anon., 2001b) that can potentially contaminate the product. Suitable hand drying equipment, e.g. paper towel dispensers or hot air dryers, and, for paper towels, suitable towel disposal containers. Access for clean factory clothing and storage of soiled clothing. For larger operations this may be via an adjoining laundry room with interconnecting hatches. Interlocked doors are possible such that doors only allow entrance to high risk if a key stage, e.g. hand washing has been undertaken. CCTV cameras (if acceptable) as a potential monitor of hand wash compliance. Alcoholic hand rub dispensers immediately inside the high risk production area.

There may be the requirement to site additional hand wash basins inside the high risk area if the production process is such that frequent hand washing is necessary and these should be clearly separated from sinks for washing utensils. As an alternative to this, Taylor et al. (2000) demonstrated that cleaning hands with alcoholic wipes, which can be done locally at the operative’s work station, is an effective means of hand hygiene.

11.5.9 Air supply The air is an important, potential source of pathogens and the air intake into the high risk area has to be controlled. Air can enter high risk via a purpose built air handling system or can enter into the area from external uncontrolled sources (e.g. low risk production, packing, outside). For high risk areas, the goal of the air handling system is to supply suitably filtered fresh air, at the correct temperature and humidity, at a slight overpressure to prevent the ingress of external air. The capital cost of the air handling systems is one of the major costs associated with the construction of a high risk area and specialist advice should always be sought before embarking on an air handling design and construction project. Following a suitable risk analysis, it may be concluded that the air handling requirements for high care areas may be less stringent, especially related to filtration levels, capacity and degree of overpressure. If there are large movements of air from low risk to high care, positive pressure may be needed to reverse this flow. If airflows are more static, positive pressure in high care may not be required. Once installed, any changes to the construction of the high risk area (e.g. the © 2008, Woodhead Publishing Limited

288

Chilled foods

rearrangement of walls, doors or openings) should be carefully considered as they will have a major impact on the air handling system. To aid the performance of the air handling system, it is also important to control potential sources of aerosols, generated from personnel, production and cleaning activities, in both low and high risk. Air quality standards for the food industry have been reviewed by a CCFRA Working Party and guidelines produced (Anon., 2005). The design of the air handling system should consider the following issues:

• • • • • • • •

degree of filtration of external air overpressure air flow – concerned with operational considerations and operative comfort air movement temperature requirements local cooling and barrier control humidity requirements installation and maintenance.

The main air flows within a high risk area are shown in Fig. 11.9 and a more detailed schematic of the air handling system is shown in Fig. 11.10. Filtration of air is a complex matter and requires a thorough understanding of filter types and installations. The choice of filter will be dictated by the quality of the incoming air, and degree of microbial and particle removal required (BS EN 779, 1979). Filter types are described in detail in the CCFRA guideline document (Anon., 2005). For high risk applications, a series of filters is required (Fig. 11.10) to provide air to the desired standard and these are usually made up of a G4/F5 panel or pocket filter followed by an F9 rigid cell filter. For some high risk operations an H10 or H11 final filter may be desirable, whilst for high care operations an F7 or F8 final filter may be acceptable. A major risk of airborne contamination entering high risk is from low risk processing operations, especially those handling raw produce that is likely to be contaminated with pathogens. The principal role of the air handling system is thus to provide filtered air to high risk with a positive pressure and outward flow into low risk. This means that wherever there is a physical break in the low/high risk barrier, e.g. a hatch, the air flow will be through the opening from high to low risk. Levels of airborne micro-organisms in low risk, depending on the product and processes being undertaken, may be quite high (Holah et al., 1995) and overpressure should prevent the movement of such airborne particles, some of which may contain viable pathogenic micro-organisms, into high risk. To be effective, the pressure differential between low and high risk should be in excess of 5 Pascals. The desired pressure differential will be determined by both the number and size of openings and also the temperature differentials between low and high risk. For example, if the low risk area is at ambient (20 °C) and the high risk areas at 10 °C, hot air from low risk will tend to rise through any openings, e.g. hatches, whilst cold air from high risk will tend to sink through the same opening, causing two way flow. The velocity of air through the opening from © 2008, Woodhead Publishing Limited

Fresh and/or make up air

Exhaust air

Non-return damper

Extract air

Air handling unit

Treatment plant Air-treatment unit

Closed loop recycled air

Ducting or textile ventilation

Local dustfume or vapour extract Return air

Supply air

Air changes/hour (fresh and recirculated)

Controlled space

Return air

Transfer air Leakage Local dustfume or vapour extract

(positive pressure) Process air

Schematic diagram showing the airflows within a high risk or high care production area.

289

Fig. 11.9

© 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

Recirculating air

290 Chilled foods

Fig. 11.10 © 2008, Woodhead Publishing Limited

Schematic diagram of the components of a typical air handling system.

The hygienic design of chilled food plants and equipment

291

high risk may need to be 1.5 m/s or greater to ensure that one way flow is maintained. The requirements for positive pressure in high care processing areas are less stringent and ceiling mounted evaporative chillers together with additional air make-up may be acceptable. In addition to providing a positive over-pressure, the air flow rate and heat exchange capacity of the system must be sufficient to remove the heat load imposed by the processing environment (processes and people) and provide operatives with fresh air. Generally, 5–25 air changes per hour are adequate, though in a high risk area with large hatches/doors that are frequently opened, up to 40 air changes per hour may be required. Air is usually supplied to high risk by either rigid metal ducts to ceiling grills or textile ducts (socks), usually made from polyester or polypropylene to reduce shrinkage. Ceiling grills have the advantage that they are cheap and require little maintenance but have limitations on velocity and flow rate without high noise levels or the potential to cause draughts. With respect to draughts, the maximum air speed close to workers to minimise discomfort through ‘wind-chill’ is 0.3 m/s–1. Air socks have the ability to distribute air at a low, draught-free velocity with minimal ductwork connections, though they require periodic laundering and spare sets are required. A best practice guideline on air flows in high risk areas was published by MAFF in 2001 (Anon., 2001b) following work by CCFRA and Silsoe Research Institute (SRI), centred on the measurement of both air flows and airborne microbiological levels in food factories. Computational fluid dynamics (CFD) models have been developed by SRI to predict air and particle (including micro-organism) movements which can visualise, for example, the influence on airflows of air intakes and air extracts, secondary ventilation systems in, e.g. washroom areas, the number of hatches and doors and their degree of openings and closings. This has led to the redesign of high risk areas, from the computer screen, such that airflow balances and positive pressures have been achieved. The CFD models also allow the prediction of the movement of airborne micro-organisms from known sources of microbial contamination, e.g. operatives. This has allowed the design of air handling systems which provide directional air that moves particles away from the source of contamination, in a direction that does not compromise product safety. As an illustration, Fig. 11.11a shows the predicted air flows in a real factory generated using a CFD software package developed by SRI following air flow measurements. The model was then used to predict the movement of 10 µm particles (similar to shed skin squames) from line operatives (Fig. 11.11b). The predicted tracks indicate that in some cases the airflow is good and moves shed particles away from the product whilst in other cases particles move directly over or along the product conveyors, thus presenting a hygiene risk. Chilled foods manufacturers have traditionally chosen to operate their high risk areas at low temperatures, typically around 10–12 °C, to both restrict the general growth of micro-organisms in the environment and to prevent the growth of some (e.g. Salmonella) but not all (e.g. Listeria) food pathogens. Chilling the area to this temperature is also beneficial in reducing the heat uptake by the product and thus © 2008, Woodhead Publishing Limited

292

Chilled foods

(a)

Hatches H5

H4

H3

H1

Door, D2 Chillers

Entrance doorway, D3 Y

Z

Process lines

X

Low care (despatch) area Doorway, D1

Office Hatch, H6

H5

Hatches, H1–H5

(b)

Door, D2 Chillers

Operators

Entrance doorway, D3 Process lines

Y

X

Office

Z

Low care (despatch) area Doorway, D1

Fig. 11.11 Schematic diagram of: (a) predicted airflows in an actual chilled food factory established from airflow measurements. The length and size of the arrow indicates air speed whilst the orientation shows flow direction. (b) Predicted flow from line operatives. The flow of product down the five lines is in the direction Y to Z. (Courtesy of Silsoe Research Institute.)

© 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

293

maintaining the chill chain. Moreover, chilled food manufacturers have to ensure that their products meet, in the UK, the requirements of the Food Safety (Temperature Control) Regulations 1995 (Anon., 1995) as well as those imposed by their retail customers. In the UK, The Workplace (Health, Safety and Welfare) Regulations (Anon., 1992a) require that the ‘temperature in all workplaces inside buildings shall be reasonable’, which, in the supporting Approved Code of Practice (1992b), is normally taken to be at least 16 °C or at least 13 °C where much of the work involves serious physical effort. To help solve this conflict of product and operative temperature, a Working Group comprising members of the Health and Safety Executive (HSE) and the chilled food industry was established at CCFRA in 1996. The Working Group produced a document Guidance on Achieving Reasonable Working Temperatures and Conditions During Production of Chilled Foods (Brown, 2000) which extends the information provided in HSE Food Sheet No.3 (Rev) Workroom Temperatures in Places Where Food is Handled (Anon., 1999; www.hse.gov.uk/pubns/fis03.pdf). The guidance document (Brown, 2000) states that employers will first need to consider alternative ways of controlling product temperatures to satisfy the Food Safety (Temperature Control) Regulations (Anon., 1995) rather than simply adopting lower workroom temperatures. If the alternative measures are not practical then it may be justified for hygiene reasons for workrooms to be maintained at temperatures lower than 16 °C (or 13 °C). Where such lower temperatures are adopted, employers should be able to demonstrate that they have taken appropriate measures to ensure the thermal comfort of employees. Full guidance on these issues is given in the document. The choice of relative air humidity is a compromise between operative comfort, product quality and environmental drying. A relative humidity of 55–65% is very good for restricting microbial growth in the environment and increases the rate of equipment and environment drying after cleaning operations. Low humidities can, however, cause drying of the product with associated weight and quality loss, especially at higher air velocities. Higher humidities maintain product quality but may give rise to drying and condensation problems that increase the opportunity for microbial survival and growth. A compromise target humidity of 60–70% is often recommended, which is also more optimal for operative comfort. Finally, air handling systems should be properly installed such that they can be easily serviced and cleaned and, as part of the commissioning programme, their performance should be validated for normal use. The ability of the system to perform in other roles should also be established. These could include dumping air directly to waste during cleaning operations, to prevent air contaminated with potentially corrosive cleaning chemicals entering the air handling unit, and recirculating ambient or heated air after cleaning operations to increase environmental drying.

11.5.10 Utensils Wherever possible, any equipment, utensils and tools, etc. used routinely within high risk, should remain in high risk. This should mean that there is provision for © 2008, Woodhead Publishing Limited

294

Chilled foods

storage areas or areas in which utensils can be maintained or cleaned. Typical examples include:

• The requirement for ingredient or product transfer containers (trays, bins, etc.) • •

•

•

•

• •

should be minimised but where these are unavoidable they should remain within high risk and be cleaned and disinfected in a separate wash room area. Similarly, any utensils (e.g. stirrers, spoons, ladles) or other non-fixed equipment (e.g. depositors or hoppers) used for the processing of the product should remain in high risk and be cleaned and disinfected in a separate wash room area. A separate wash room area should be created in which all within-production wet cleaning operations can be undertaken. The room should preferably be sited on an outside wall that facilitates air extraction and air make-up. An outside wall also allows external bulk storage of cleaning chemicals that can be directly dosed through the wall into the ring main system. The room should have its own drainage system that, in very wet operations, may include barrier drains at the entrance and exit to prevent water spread from the area. The wash area should consist of a holding area for equipment, etc. awaiting cleaning, a cleaning area for manual or automatic cleaning (e.g. traywash) as appropriate, and a holding/ drying area where equipment can be stored prior to use. These areas should be as segregated as possible. All cleaning equipment, including hand tools (brushes, squeegees, shovels, etc.) and larger equipment (pressure washers, floor scrubbers and automats, etc.) should remain in high risk and be colour coded to differentiate between high and low risk equipment if necessary. Special provision should be made for the storage of such equipment when not in use. Cleaning chemicals should preferably be piped into high risk via a ring main at the use concentration and temperature (this should be separate from the low risk ring main). If this is not possible, cleaning chemicals should be stored in a purpose-built area. The most commonly used equipment service items and spares, etc., together with the necessary hand tools to undertake the service, should be stored in high risk. For certain operations, e.g. blade sharpening for meat slicers, specific engineering rooms may need to be constructed. Provision should be made in high risk for the storage of utensils that are used on an irregular basis but that are too large to pass through the low/high risk barrier, e.g. stepladders for changing the air distribution socks. Written procedures (and training) should be prepared detailing how and where items that cannot be stored in high risk but are occasionally used there, or new pieces of equipment entering high risk, will be decontaminated. If appropriate, these procedures may also need to detail the decontamination of the surrounding area in which the equipment decontamination took place.

11.6 Barrier 4 – product enclosure The fourth barrier is product enclosure and has the objective of further excluding © 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

295

contamination, particularly from micro-organisms from the surrounding environment. Whilst the fourth barrier approach is essential for the production of aseptic foods, it is also being used for the production of some chilled, RTE foods, particularly sliced meat products. Product enclosure can be undertaken by physical segregation (a box within a box within a box) or by the use of highly filtered, directional air currents. With respect to physical segregation, ‘gloveboxes’ offer the potential to fully enclose product with the ability to operate to aseptic conditions. Gloveboxes for the food industry work in the same way as gloveboxes for the medical, microbiological and pharmaceutical industries; the food is enclosed in a sealed space, totally protected from the outside environment, and manipulated through gloves sealed into an inspection window. They work best if the product is delivered to them in a pasteurised condition, is packed within the box and involves little manual manipulation. The more complicated the product manipulation, the more ingredients to be added, the faster the production line or the shorter the product run, the less flexible gloveboxes become. Operating on a batch basis, pre-disinfected gloveboxes give the potential for a temperature controlled environment with a modified atmosphere if required (e.g. high CO2 or low O2 concentrations), that can be disinfected on-line by gaseous chemicals (e.g. ozone) or UV light. Gloveboxes may also offer some protection in the future to foodstuffs identified by risk assessments as being particularly prone to bioterrorism. Gloveboxes are necessary, of course, only if people are involved in the food production line. If robots undertook product manipulation, there would be less microbiological risk and the whole room could be temperature and atmospherically controlled! Where the use of gloveboxes is impractical, partial enclosure of the product can be achieved in appropriate circumstances by the use of localised, filtered airflows. The high risk air handling system provides control of airborne contamination external to high risk but provides only partial control of aerosols, generated from personnel, production and cleaning activities, in high risk. At best, it is possible to design an air handling system that minimises the spread of contamination generated within high risk from air directly moving over product. Localised airflows are thus designed to:

• provide highly filtered (H11–12) air directly over or surrounding product, and

•

its associated equipment. This could reduce the requirement to chill the whole of the high risk area to 10 °C (13 °C would be acceptable), and reduce the degree of filtration required (down to H8–9). The requirement for positive pressure in low risk is paramount, however, and the number of air changes per hour would remain unchanged. provide a degree of product isolation ranging from partial enclosure in tunnels to chilled conveyor wells, where the flow of the filtered air provides a barrier that resists the penetration of aerosol particles, some of which would contain viable micro-organisms.

An example of such a technology is shown in Fig. 11.12 which shows a schematic diagram of a conveyor that has chilled, filtered air directed over it, sufficient to © 2008, Woodhead Publishing Limited

296

Chilled foods

Food product

Wall of cold-well

Air supply duct

Counts on floor

800 counts

Counts outside trough Counts inside trough Air supply duct Trough

Fig. 11.12 Schematic diagram of: (a) a conveyor belt with cooled, filtered air directed across the product and (b) the reduction in microbial counts (colony forming units, cfu) within the conveyor during operation. The diameter of the circles is directly proportional to the number of cfus recorded. (Courtesy of Silsoe Research Institute.)

maintain the low temperature of the product. When a microbial aerosol was generated around the operational conveyor, microbiological air sampling demonstrated a 1–2 log reduction of micro-organisms within the protected zone as compared to outside the conveyor (Burfoot et al., 2000).

11.7 Equipment The manufacture of a large proportion of chilled foods generally involves some element of batch or assembly operations or both. The equipment used for such operations is predominantly of the open type, that is it cannot be cleaned by recirculation (CIP) procedures, and must be of the highest hygienic design and construction standards. © 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

297

11.7.1 Hygienic equipment design Hygienic equipment design provides three major benefits to food manufacturers: (i)

Quality – Good hygienic design maintains product in the main product flow. This ensures that product is not ‘held-up’ within the equipment where it could deteriorate and affect product quality on rejoining the main product flow or, for example in flavourings manufacture, one batch could not taint a subsequent batch. (ii) Safety – Good hygienic design prevents the contamination of the product with substances that would adversely affect the health of the consumer. Such contamination could be microbiological (e.g. pathogens), chemical (e.g. lubricating fluids, cleaning chemicals) or physical (e.g. glass). (iii) Efficiency – Good hygienic design reduces the time required for an item of equipment to be cleaned. This reduction of cleaning time is significant over the lifetime of the equipment such that hygienically designed equipment which may be initially more expensive (compared to similarly performing, poorly designed equipment), will be more cost effective in the long term. In addition, savings in cleaning time may lead to increased production. Guidance on hygienic design, specifically oriented to equipment manufacturers, is readily available in Europe from the European Hygienic Engineering Design Group (www.ehedg.org) and in the USA from the 3-A Standards (www.3-a.org) and the National Sanitation Foundation (NSF) (www.nsf.org) organisations. In addition to providing guidance for equipment manufacturers on how equipment should be designed, each organisation offers a third party auditing or certification scheme so that equipment suppliers can demonstrate independent compliance with these schemes. Purchasers of food process equipment should also familiarise themselves with guidance from such organisations to ensure that they select equipment that is inherently hygienic and that they have a good knowledge base to install and link together items of equipment in their process lines. In addition, and if possible, product and process development trials can be undertaken with equipment manufacturers and/or discussions can be had with other equipment users to verify hygienic performance. It should be noted that in Europe, hygienic design guidelines tend to be more generic in nature than the more prescriptive requirements American readers may be familiar with. In the EC, the Council Directive on the approximation of the laws of member states relating to machinery (89/392/EEC) was published on 14 June 1989. The Directive includes a short section dealing with hygiene and design requirements which states that machinery intended for the preparation and processing of foods must be designed and constructed so as to avoid health risks and it consists of seven hygiene rules that must be observed. These are concerned with materials in contact with food; surface smoothness; preference for welding or continuous bonding rather than fastenings; design for cleanability and disinfection; good surface drainage; prevention of dead spaces which cannot be cleaned; and design to prevent product contamination by ancillary substances, e.g. lubricants. The Directive requires that all machinery sold within the EU shall meet © 2008, Woodhead Publishing Limited

298

Chilled foods

these basic standards and be marked accordingly to show compliance (the ‘CE’ mark). Subsequent to this Directive, a European Standard EN 1672-2 (Anon., 1997) has been published to further clarify the hygiene rules established in 89/392/EEC. In addition to this, a number of specific standards on bakery, meat, catering, edible oils, vending and dispensing, pasta, bulk milk coolers, cereal processing and dairy equipment are in preparation. The basic hygienic design requirements as presented in EN 1672-2 can be summarised under eleven headings and are described below: (i)

Materials of construction – Materials used for product contact must have adequate strength over a wide temperature range, and a reasonable life, and be non-tainting, corrosion and abrasion resistant, be easily cleaned and capable of being shaped. Stainless steel usually meets all these requirements and there are various grades of stainless steel which are selected for their particular properties to meet operational requirements. For example, Type 316, which contains molybdenum, is used where improved corrosion resistance is necessary. (ii) Surface finish – Product contact surfaces must be finished to a degree of surface smoothness that is enough to enable them to be easily cleaned. Surfaces will deteriorate with age and wear (abrasion) such that cleaning will become more difficult. (iii) Joints – Permanent joints, such as those which are welded, should be smooth and continuous. Dismountable joints, such as screwed pipe couplings, must be crevice-free and provide a smooth continuous surface on the product side. Flanged joints must be located with each other and be sealed with a gasket because, although metal–metal joints can be made leak tight, they may still permit the ingress of micro-organisms. (iv) Fasteners – Exposed screw threads, nuts, bolts, screws and rivets (Fig. 11.13) must be avoided wherever possible in product contact areas. Alternative methods of fastening can be used (Fig. 11.14) where the washer used has a rubber compressible insert to form a bacteria-tight seal. (v) Drainage – All pipelines and equipment surfaces should be self-draining because residual liquids can lead to microbial growth or, in the case of cleaning fluids, result in contamination of product. (vi) Internal angles and corners – These should be well radiused, wherever possible, to facilitate cleaning. (vii) Dead spaces – As well as ensuring that there are no dead spaces in the design of equipment, care must be taken that they are not introduced during installation. (viii) Bearings and shaft seals – Bearings should, wherever possible, be mounted outside the product area to avoid possible contamination of product by lubricants, unless they are edible, or possible failure of the bearings due to the ingress of the product. Shaft seals must be of such design so as to be easily cleaned and, if not product lubricated, then the lubricant must be edible. Where a bearing is within the product area, such as a foot bearing for an © 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

Fig. 11.13

299

Examples of unhygienic fasteners. A = soil trap points, B = metal to metal, C = dead spaces.

Fig. 11.14 Examples of hygienic fasteners. © 2008, Woodhead Publishing Limited

300

Chilled foods

Fig. 11.15 Typical operating switch (a) with inherent crevices. Listeria was harboured between the switch ‘piston’ and body, and when the switched was pressed, Listeria was exuded onto the operatives’ fingers. A hygienic rubber cap (b) prevented the problem and allowed easy cleaning.

agitator shaft in a vessel, it is important that there is a groove completely through the bore of the bush, from top to bottom to permit the passage of cleaning fluid. (ix) Instrumentation – Instruments must be constructed from appropriate materials and if they contain a transmitting fluid, such as in a bourdon tube pressure gauge, then the fluid must be approved for food contact. Many instruments themselves are hygienic but often they are installed unhygienically. (x) Doors, covers and panels – These should be designed so that they prevent the entry of and/or the accumulation of soil. Where appropriate, they should be sloped to an outside edge and should be easily removed to facilitate cleaning. (xi) Controls – Controls, particularly those that are repeatedly touched by food handlers to allow process operation, should be designed to prevent the ingress of contamination and should be easily cleanable (Fig. 11.15). The importance of good hygienic design in chilled foods operations can be illustrated with reference to a sliced meat factory which had slicers whose action was initiated by pressing a control switch identical to that shown in Fig. 11.15. The factory concerned was having problems due to product contamination with Listeria monocytogenes, and was eventually forced to stop production for a few days with a subsequent financial loss in excess of £1million. The problem was finally traced to a source of L. monocytogenes that was being harboured within the body of the slicer switches. At the beginning of production the slicing operative © 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

301

picked up a log of meat, placed it on the slicer and pressed the control switch to start slicing. From this point on, and every time he subsequently repeated this procedure, L. monocytogenes was transferred from his hand to the slicer and, by the middle of the shift, sufficient L. monocytogenes was present on the slicer to be detected in the product. The conclusion to the incident was the purchase of a number of rubber switch covers as shown in Fig. 11.15, for the cost of a few pounds. Today, many items of equipment have touch panels to control their operation though this case study does illustrate that even the smallest dead space or harbourage site is sufficient to retain pathogenic micro-organisms that can subsequently contaminate product. Industrial slicers have been implicated on a number of occasions as causing pathogen contamination problems and must be designed, maintained and operated hygienically (Timperley and Timperley, 1993). It is imperative that technical managers and engineers understand and appreciate the concepts of hygienic design such that only appropriate food production equipment is purchased and installed.

11.7.2 Installation of equipment The potential for well designed and constructed equipment to be operated in a hygienic manner may be easily vitiated by inadequate attention to its location and installation. Timperley (1997), when considering the accessibility of equipment, recommended that it is more effective to consider complete lines instead of individual items of equipment and recommended the following:

• There should be sufficient height from the floor (at least 20 cm) to allow • • •

• • •

adequate access for inspection, cleaning and maintenance of the equipment and for the cleaning of floors. All parts of the equipment should be installed at a sufficient distance from walls, ceilings and adjacent equipment (at least 1 m) to allow easy access for inspection, cleaning and maintenance, especially if lifting is involved. Ancillary equipment, control systems and services connected to the process equipment should be located so as to allow access for maintenance and cleaning. Supporting framework, wall mountings and legs should be kept to a minimum. They should be constructed from tubular or box section material which should be sealed to prevent ingress of water or soil. Angle or channel section material should not be used. Base plates used to support and fix equipment should have smooth, continuous and sloping surfaces to aid drainage. They should be coved at the floor junction. Alternatively, ball feet should be fitted. Pipework and valves should be supported independently of other equipment to reduce the chance of strain and damage to equipment, pipework and joints. Equipment should not be sited above drains. Discharges from equipment, however, should be fed directly into drains to avoid floor flooding. Alternatively, a low wall may be built around the equipment from which water and solids may be drained.

© 2008, Woodhead Publishing Limited

302

Chilled foods

11.8 Conclusion Since the first Listeria infection problems associated with sliced meats in 1988 in the UK, the chilled food industry has been on a steep learning curve to improve its manufacturing infrastructure to help prevent high risk pathogen re-contamination incidents. To a great extent this has been achieved and the level of pathogen incidence or other non-conforming product due to microbiological contamination in RTE chilled products is now much reduced. Emphasis should now be put, therefore, on education. Food plant specialist architects and construction companies should be trained in the principles of hygienic building design and only those contractors who can demonstrate good knowledge in these areas should be selected for factory refurbishment or new-build projects. Similarly, equipment manufacturers should also be trained in the principles of hygienic equipment design and, where a choice of equipment is available, equipment should be selected that has been approved as hygienic or proven to operate hygienically wherever possible. There are now a number of independent third-party approval schemes for building fabrication materials and food processing equipment, and items should be chosen that have met such approval schemes. In essence, suppliers should be challenged to substantiate their hygienic claims. Looking to the future, whilst we have improved our manufacturing infrastructure in terms of hygiene and have many good hygiene practices in place, we are still unclear as to which are the most important routes of product contamination that we should seek to control. Only with this knowledge will we able to design buildings, equipment and practices that will further control contamination of product during manufacture.

11.9 References ANON.

(1992a) The Workplace (Health, Safety and Welfare) Regulations. HMSO ISBN 011-886333-9. ANON. (1992b) Workplace Health, Safety and Welfare. Approved Code of Practice and Guidance on Workplace (Health, Safety and Welfare) Regulations 1992. L24 HSE Books ISBN 0-7176-0413-6. ANON. (1995) Food Safety (Temperature Control) Regulations 1995. HMSO ISBN 0-11053383-6. ANON. (1997) EN 1672-2. Food processing machinery – Basic requirements. Part 2; Hygiene requirements. ISBN 0 580 27957 X. ANON. (1999) Workroom temperatures in places where food is handled. HSE Food Sheet No.3 (Revised). Anon. (2001a) Hygienic design criteria for the safe processing of dry particulate materials. EHEDG Guideline No.22. www.ehedg.org. ANON. (2001b) Best practice guidelines on air flows in high-care and high-risk areas. Published for the UK Ministry of Agriculture, Fisheries and Food by Campden and Chorleywood Food Research Association. ANON. (2002) Guidelines for the design and construction of floors for food production areas (Second edition). Campden and Chorleywood Food Research Association, Guideline No. 40. ANON. (2003a) Guidelines for the hygienic design, construction and layout of food processing factories. Campden and Chorleywood Food Research Association, Guideline No. 39. © 2008, Woodhead Publishing Limited

The hygienic design of chilled food plants and equipment

303

ANON. (2003b) Guidelines for the design and construction of walls, ceilings and services for

food production areas (Second edition). Campden and Chorleywood Food Research Association, Guideline No. 41. ANON. (2004) Regulation (EC) No 852/2004 of the European Parliament and of the Council of Europe of 29 April 2004 on the hygiene of foodstuffs. Official Journal of the European Union, L226/3-21. ANON. (2005) Guidelines on air quality for the food industry. Campden and Chorleywood Food Research Association, Guideline No. 12. ANON. (2006) Best practice guidelines for the production of chilled foods (Fourth edition). Chilled Food Association, PO Box 6434, Kettering, NN15 5XT www.chilledfood.org. ASHFORD M J (1986) Factory design principles in the food processing industry. In: Preparation, Processing and Freezing in Frozen Food Production. The Institution of Mechanical Engineers, London. BS EN 779 (1979) Particulate air filters for general ventilation. Requirements, testing, marking (ISBN 0 580 21267 X), British Standards Institute, UK. BROWN K L (2000) Guidance on achieving reasonable working temperatures and conditions during production of chilled foods. Campden and Chorleywood Food Research Association, Guideline No. 26. BURFOOT D, BROWN K, REAVELL S AND XU Y (2000) Improving food hygiene through localised air flows. Proceedings International Congress on Engineering and Food, Volume 2 April 2000, Puebla, Mexico. Technomic Publishing Co. Inc., Lancaster, Pensylvania, USA, pp 1777–1781. GUZEWICH J AND ROSS P (1999) Evaluation of risks related to microbiological contamination of ready-to-eat food by food preparation workers and the effectiveness of interventions to minimise those risks. Food and Drug Administration, White Paper, Section One, USA. HOLAH J T, HALL K E, HOLDER J, ROGERS S J, TAYLOR J AND BROWN K L (1995) Airborne microorganisms levels in food processing environments. Campden Food and Drink Research Association, R & D Report No. 12. HOLAH J T (1998) Hygienic design: International issues. Dairy, Food and Environmental Sanitation, 18, 212–220. IMHOLTE T J (1984) Engineering for Food Safety and Sanitation. Technical Institute of Food Safety, Crystal, Minnesota. KATSUYARNA A M AND STRACHAN J P (EDS) (1980) Principles of Food Processing Sanitation. The Food Processors Institute, Washington DC. METTLER E AND CARPENTIER B (1998) Variations over time of microbial load and physicochemical properties of floor materials after cleaning in food industry premises. Journal of Food Protection, 61, 57–65. SHAPTON D A AND SHAPTON N F (EDS) (1991) Principles and Practices for the Safe Processing of Foods. Butterworth Heinemann. TAYLOR J, HOLAH J T (1996) A comparative evaluation with respect to bacterial cleanability of a range of wall and floor surface materials used in the food industry. Journal of Applied Bacteriology, 81, 262–267. TAYLOR J H, HOLAH J T (2000) Hand hygiene in the food industry: A review. Review No. 18, Campden and Chorleywood Food Research Association, UK. TAYLOR J H, HOLAH J T, WALKER H AND KAUR M (2000) Hand and footwear hygiene: An Investigation to define best practice. Campden and Chorleywood Food Research Association, R & D Report No. 110. TIMPERLEY A W (1997) Hygienic design of liquid handling equipment for the food industry (Second edition). Campden and Chorleywood Food Research Association, Technical Manual No. 17. TIMPERLEY D A, TIMPERLEY A W (1993). Hygienic design of meat slicing machines. Campden and Chorleywood Food Research Association, Technical Memorandum No. 679. TROLLER J A (1983) Sanitation in Food Processing. Academic Press, New York.

© 2008, Woodhead Publishing Limited

12 Cleaning and disinfection of chilled food plants and equipment J. T. Holah, Campden and Chorleywood Food Research Association, UK

12.1 Introduction Chapter 11 has outlined the concept of ‘hygienic design’ and ‘hygienic practices’ in controlling the safety of chilled food products. This chapter deals with hygiene practices that are used to maintain a hygienic manufacturing infrastructure, specifically those related to cleaning and disinfection. Contamination in chilled food products may arise from four main sources: the constituent raw materials, surfaces (including people), liquids and the air. Control of the raw materials is addressed elsewhere in this book and is the only nonenvironmental contamination route. Food may pick up contamination as it is moved across product contact surfaces (hard surfaces) or if it is touched or comes into contact with people (food handlers) or other animals, e.g. pests (soft surfaces). Liquids include process aids (washing or cooling fluids) or residues of cleaning fluids or rinses left on surfaces. They differ from surface contact in that there is an element of uptake of the liquid into the product. The air acts as both a source of contamination, i.e. from outside the processing area, or as a transport medium, e.g. moving contamination from non-product to product contact surfaces. Provided that the process environment and production equipment have been hygienically designed (Chapter 11), cleaning and disinfection (referred to together as sanitation) are the major day-to-day controls of the hard surface vectors of food product contamination and, if effective, can reduce sources of micro-organisms within the processing environment. In addition, and for the high care/risk area, © 2008, Woodhead Publishing Limited

Cleaning and disinfection of chilled food plants and equipment

305

sanitation practices are the only processes that can control microbial crosscontamination within food processing areas. When undertaken correctly, sanitation programmes have been shown to be cost-effective and easy to manage, and, if diligently applied, can reduce the risk of microbial or foreign body contamination. Given the intrinsic demand for high standards of hygiene in the production of short shelf-life chilled foods, together with pressure from customers, consumers and legislation for ever-increasing hygiene standards, sanitation demands the same degree of attention as any other key process in the manufacture of safe and wholesome chilled foods. This chapter is concerned with the sanitation of ‘hard’ surfaces only – equipment, floors, walls and utensils – as other surfaces, e.g. protective clothing or skin, have been dealt with under personal hygiene (Chapter 11). In this context, surface sanitation is undertaken to: (i)

remove micro-organisms, or material conducive to microbial growth. This reduces the chance of contamination by pathogens and, by reducing spoilage organisms, may extend the shelf-life of some products; (ii) remove materials that could lead to foreign body contamination or could provide food or shelter for pests. This also improves the appearance and quality of product by removing food materials left on lines that may deteriorate and re-enter subsequent production runs; (iii) extend the life of, and prevent damage to equipment and services; provide a safe and clean working environment for employees which boosts morale and productivity; (iv) present a favourable image to customers and the public. On audit, the initial perception of an ‘untidy’ or ‘dirty’ processing area, and hence a ‘poorly managed operation’ is subsequently difficult to overcome.

12.2 Sanitation principles Sanitation is undertaken primarily to remove all undesirable material (food residues, micro-organisms, foreign bodies and cleaning chemicals) from surfaces in an economical manner, to a level at which any residues remaining are of minimal risk to the quality or safety of the product. Such undesirable material, generally referred to as ‘soil’, can be derived from normal production, spillages, line-jams, equipment breakdown, equipment maintenance, packaging or general environmental contamination (dust and dirt). To undertaken an adequate and economic sanitation programme, it is essential to characterise the nature of the soil to be removed. Product residues on surfaces are easily observed and may be characterised by their chemical composition, e.g. carbohydrate, fat, or protein. It is also important to be aware of processing and/or environmental factors, however, as the same product soil may lead to a variety of cleaning problems, depending primarily on moisture levels and temperature. Generally, the higher the product soil temperature (especially if the soil has been baked) and the greater the time period before the © 2008, Woodhead Publishing Limited

306

Chilled foods

sanitation programme is initiated (i.e. the drier the soil becomes), the more difficult the soil is to remove. With respect to micro-organisms on food production surfaces, the food industry has always recognised that they are present (either within the surface soil or attached to the surface) and that their numbers increase during production and are reduced by cleaning and disinfection. Initially, it was considered that microorganisms adhered to the food product contact surfaces could be an important source of potential contamination, leading to serious hygienic problems and economic losses due to food spoilage. For example, pseudomonads and many other Gram negative organisms detected on surfaces are the (spoilage) microorganisms of concern in chilled foods. Since this time, however, we have recognised that food contact surfaces can give rise to biofilms, can support the growth of pathogens, and more recently, that food processing environments will develop their own persistent ‘house’ flora, some of which can be pathogenic. There are a number of factors that have been shown to affect attachment and biofilm formation, such as the level and type of micro-organisms present, surface conditioning layer, substratum nature and roughness, temperature, pH, nutrient availability and time available. Several reviews of biofilm formation in the food industry have been published including Pontefract (1991), Holah and Kearney (1992), Mattila-Sandholm and Wirtanen (1992), Carpentier and Cerf (1993), Zottola and Sasahara (1994), Gibson et al. (1995) and Kumar and Anand (1998). In general, however, biofilm formation may be found on environmental (particularly drains) and some production (e.g. flumes and cooling canals) surfaces and the growth of attached cells through micro-colonies to extensive biofilms is limited only by regular cleaning and disinfection. Indeed, following HACCP principles, if the food processor believes that biofilms are a risk to the safety of the food product, appropriate control steps (e.g. cleaning and disinfection) must be taken, monitored and controlled. Following the first concerns associated with Listeria in chilled foods in the late 1980s, a number of authors have isolated Listeria monocytogenes from a range of food processing surfaces (Walker et al., 1991; Lawrence and Gilmore, 1995 and Destro et al., 1996). Indeed, the assessment of the environment for Listeria spp. is now an essential part of the high care/risk processing area environmental sampling plan. The concept of persistence, i.e. strains of an organism becoming adapted to and surviving in an environment for considerable time periods, is more recent. Evidence for L. monocytogenes strain persistence within factory processing areas has been demonstrated for 8 months (Rørvik et al., 1995), 11 months (McLauchlin et al., 1990), 14 months (Johansson et al., 1999), 1 year (Lawrence and Gilmour, 1995), 17 months (Pourshaban et al., 2000), 2 years (McLauchlin et al., 1991), 3 years (Brett et al., 1998, Holah et al., 2004), 40 months (Unnerstad et al., 1996), 4 years (Nesbakken et al., 1996, Aase et al., 2000; Fonnesbech Vogel et al., 2001), 7 years (Unnerstad et al., 1996, Miettinen et al., 1999) and 10 years (Kathariou, 2003). These authors have, between them, isolated persistent strains from the following processed food products: cheese; mussels, shrimps and raw and smoked fish; fresh, cooked and fermented meats; pâté and pork tongue in aspic; chicken © 2008, Woodhead Publishing Limited

Cleaning and disinfection of chilled food plants and equipment

307

and turkey meat; pesto sauce and ready meals. The nature of strain persistence is unknown but may be due to a number of factors affecting biological/physiological adaptation (surface attachment, biofilm formation, attachment strength, reduced growth rate, quiescence, cleaning and disinfection resistance) to the whole range of environmental conditions typical in chilled food factory environments (low temperature, wide pH range, fluctuating nutrient supply and moisture levels, frequency of cleaning and disinfection, etc.). Lundén et al. (2000, 2003a) have demonstrated that persistent strains may show enhanced surface adherence and increased disinfection resistance at very low disinfectant concentrations, though Holah et al. (2002) found no evidence of resistance of certain persistent strains to disinfectants at their normal in-use concentration. Within the sanitation programme, the cleaning phase can be divided up into three sections, following the pioneering work of Jennings (1965) and interpreted by Koopal (1985), with the addition of a fourth stage to cover disinfection. These are described below: (i)

The wetting and penetration by the cleaning solution of the soil and to the equipment surface. (ii) The reaction of the cleaning solution with both the soil and the surface to facilitate peptisation of organic materials, dissolution of soluble organics and minerals, emulsification of fats and the dispersion and removal from the surface of solid soil components. (iii) The prevention of re-deposition of the dispersed soil back onto the cleaned surface. (iv) Contact of the disinfectant solution with residual micro-organisms to facilitate reaction with cell membranes and/or penetration of the microbial cell to produce a biocidal or biostatic action. Depending on whether the disinfectant contains a surfactant and the disinfectant practice chosen (i.e. with or without rinsing), this may be followed by dispersion of the micro-organisms from the surface. To undertake these four stages, sanitation programmes employ a combination of four major factors as described below. The combination of these four factors varies for different cleaning systems and, generally, if the use of one energy source is restricted, this short-fall may be compensated for by utilising greater inputs from the others. (i) (ii) (iii) (iv)

mechanical or kinetic energy chemical energy temperature or thermal energy time.

Mechanical or kinetic energy is used to physically remove soils and may include scraping, manual brushing and automated scrubbing (physical abrasion) and pressure jet washing (fluid abrasion). Of all four factors, physical abrasion is regarded as the most efficient in terms of energy transfer (Offiler, 1990), and the efficiency of fluid abrasion and the effect of impact pressure has been described by © 2008, Woodhead Publishing Limited

308

Chilled foods

Anon. (1973) and Holah (1991). Mechanical energy has also been demonstrated to be the most efficient for biofilm (Blenkinsopp and Costerton, 1991; Wirtanen and Mattila-Sandholm, 1993, 1994; Mattila-Sandholm and Wirtanen, 1992; Gibson et al., 1999). In cleaning, chemical energy is used to break down soils to render them easier to remove and to suspend them in solution to aid rinsability. At the time of writing, no cleaning chemical has been marketed with the benefit of aiding micro-organism removal. In chemical disinfection, chemicals react with micro-organisms remaining on surfaces after cleaning to reduce their viability. The chemical effects of cleaning and disinfection increase with temperature in a linear relationship and approximately double for every 10 °C rise. For fatty and oily soils, temperatures above their melting point are used to break down and emulsify these deposits and so aid removal. The influence of detergency in cleaning and disinfection has been described by Dunsmore (1981), Shupe et al. (1982), Mabesa et al. (1982), Anderson et al. (1985) and Middlemiss et al. (1985). For cleaning processes using mechanical, chemical and thermal energies, generally the longer the time period employed, the more efficient the process. When extended time periods can be employed in sanitation programmes, e.g. soaktank operations, other energy inputs can be reduced (e.g. reduced detergent concentration, lower temperature or less mechanical brushing). Soiling of surfaces is a natural process which reduces the free energy of the system. To implement a sanitation programme, therefore, energy must be added to the soil to reduce both soil particle to soil particle and soil particle to equipment surface interactions. The mechanics and kinetics of these interactions have been discussed by a number of authors (Jennings, 1965; Schlussler, 1975; Loncin, 1977; Corrieu, 1981; Koopal, 1985; Bergman and Tragardh, 1990), and readers are directed to these articles since they fall beyond the scope of this chapter. Research continues on the theoretical aspects of soil removal to gain a better understanding of the relative importance of both the physical and chemical removal factors and the relative influence of soil to soil and soil to surface bonding mechanisms. A useful review of these factors was undertaken by Fryer and Christian (2005) who have characterised soil removal in terms of deposit thinning (weaker soil to soil bonding) and deposit removal in large ‘chunks’ (weaker soil to surface bonding) for tomato and milk deposits respectively. Suffice to say that it will be many years before soil removal mechanisms can be predicted for all soil types and that practical factory based cleaning trials to optimise cleaning will still predominate. In practical terms, however, it is worth looking at the principles involved in basic soil removal, as they have an influence on the management of sanitation programmes. Soil removal from surfaces decreases such that the log of the mass of soil remaining per unit area is linear with respect to cleaning time (Fig. 12.1a) and thus follows first order reaction (log:linear) kinetics (Jennings, 1965; Schlusser, 1975). This approximation, however, is not valid at the start and end of the cleaning process and, in practice, soil removal is initially faster and ultimately slower (dotted line in Fig. 12.1a) than a first order reaction predicts. The reasons for this are unclear, though initially, non-adhered, gross oil is usually easily removed © 2008, Woodhead Publishing Limited

Cleaning and disinfection of chilled food plants and equipment

309

A

(a) Log soil mass

Soil mass

(b)

B Cleaning time

1 2 3 4 Number of periodic cleans

Fig. 12.1 Soil removal and accumulation. (a) Removal of soil with cleaning time; solid line is theoretical removal, dotted line is cleaning in practice. (b) Build-up of soil (and/or micro-organisms): A, without periodic cleans and B, with periodic cleans. (After Dunsmore et al., 1981).

(Loncin, 1977) whilst ultimately, soil held within surface imperfections, bound to surfaces or otherwise protected from cleaning effects, is more difficult to remove (Holah and Thorpe, 1990). Routine cleaning operations are never 100% efficient, and over a course of multiple soiling/cleaning cycles, soil deposits (potentially including micro-organisms) will be retained. As soil accumulates, cleaning efficiency will decrease and, as shown in plot A, Fig. 12.1b, soil deposits may for a period increase exponentially. The timescale for such soil build-up will differ between processing applications and can range from hours (e.g. heat exchangers) to typically several days or weeks, and in practice can be controlled by the application of a ‘periodic’ clean (Dunsmore et al., 1981). Periodic cleans are employed to return the surface-bound soil accumulation to an acceptable base level (plot B, Fig. 12.1b) and are achieved by increasing cleaning time and/or energy input, e.g. higher temperatures, alternative chemicals (such as a periodic acid de-scale, when alkaline cleaning agents are used routinely) or manual scrubbing. A typical example of a periodic clean is the ‘weekend clean down’ or ‘bottoming’.

12.3 Sanitation chemicals In many instances, management view the costs of cleaning and disinfection as the price of the chemicals purchased, primarily because this is the only ‘invoice’ that they see. In reality, however, sanitation chemicals are likely to represent approximately only 5% of the true costs, with labour and water costs being the most significant. The purchase of a good quality formulated cleaning product, whilst being initially more expensive, will more than cover its costs by increasing both the standard of clean and cleaning efficiency. © 2008, Woodhead Publishing Limited

310

Chilled foods

Table 12.1 Arithmetic and mean log bacterial counts (per swab) on food processing equipment before and after cleaning and after disinfection

Arithmetic mean Log mean No. of observations

Before cleaning

After cleaning

After disinfection

7.41 × 107 4.73 480

4.48 × 105 2.79 698

2.35 × 103 1.23 2004

Within the sanitation programme it has traditionally been recognised that cleaning is responsible for the removal of not only the soil but also the majority of the micro-organisms present. Mrozek (1982) showed a reduction in bacterial numbers on surfaces by up to 3 log orders, whilst Schmidt and Cremling (1981) described reductions of 2–6 log orders. The results of work at the Campden and Chorleywood Food Research Association (CCFRA) on the assessment of well constructed and competently undertaken sanitation programmes on food processing equipment in nine chilled food factories is shown in Table 12.1. The results suggest that both cleaning and disinfection are equally responsible for reducing the levels of adhered micro-organisms and suggest a combined 3 log reduction of micro-organisms. It is important, therefore, not only to purchase quality cleaning chemicals for their soil removal capabilities but also for their potential for microbial removal. Unfortunately no single cleaning agent is able to perform all the functions necessary to facilitate a successful cleaning programme, so a cleaning solution, or detergent, is blended from a range of typical characteristic components: (i) (ii) (iii) (iv) (v)

water surfactants inorganic alkalis inorganic and organic acids sequestering agents.

For the majority of food processing operations it may be necessary, therefore, to employ different cleaning products, for specific operations. This requirement must be balanced by the desire to keep the range of cleaning chemicals on site to a minimum so as to reduce the risk of using the wrong product, to simplify the job of the safety officer and to allow chemical purchase to be based more on the economics of bulk quantities. The range of chemicals and their functions is well documented (Anon., 1991; Elliot, 1980; ICMSF, 1980, 1988; Hayes, 1985; Holah, 1991; Koopal, 1985; Russell et al., 1982) and have changed little since these reviews; an overview of only the principles is given here. Water is the base ingredient of all ‘wet’ cleaning systems and must be of potable quality. Water provides the cheapest readily available transport medium for rinsing and dispersing soils, has dissolving powers to remove ionic-soluble compounds such as salts and sugars, will help solubilise proteins below their coagulation point, emulsify fats at temperatures above their melting point, and, in high pressure cleaning, can be used as an abrasive agent. On its own, however, water is a poor ‘wetting’ agent and cannot dissolve non-ionic compounds. © 2008, Woodhead Publishing Limited

Cleaning and disinfection of chilled food plants and equipment

311

Organic surfactants (surface-active or wetting agents) are amphipolar and are composed of a long non-polar (hydrophobic or lyophilic) chain or tail and a polar (hydrophilic or lyophobic) head. Surfactants are classified as anionic (including the traditional soaps), cationic, or non-ionic, depending on their ionic charge in solution, with anionics and non-ionics being the most common. Amphipolar molecules aid cleaning by reducing the surface tension of water and by emulsification of fats. If a surfactant is added to a drop of water on a surface, the polar heads disrupt the water’s hydrogen bonding and so reduce the surface tension of the water and allow the drop to collapse and ‘wet’ the surface. Increased wettability leads to enhanced penetration into soils and surface irregularities and hence aids cleaning action. Fats and oils are emulsified as the hydrophilic heads of the surfactant molecules dissolve in the water whilst the hydrophobic end dissolves in the fat. If the fat is surface-bound, the forces acting on the fat–water interface are such that the fat particle will form a sphere (to obtain the lowest surface area for its given volume), causing the fat deposit to ‘roll-up’ and detach itself from the surface. Alkalis are useful cleaning agents as they are cheap, break down proteins through the action of hydroxyl ions, saponify fats and, at higher concentrations, may be bactericidal. Strong alkalis, usually sodium hydroxide (or caustic soda), exhibit a high degree of saponification and protein disruption, though they are corrosive and hazardous to operatives. Correspondingly, weak alkalis are less hazardous but also less effective. Alkaline detergents may be chlorinated to aid the removal of proteinaceous deposits, but chlorine at alkaline pH is not an effective biocide. The main disadvantages of alkalis are their potential to precipitate hard water ions, the formation of scums with soaps, and their poor rinsability. It is currently not known whether alkaline cleaning agents modify allergenic protein residues sufficiently to reduce or eliminate their allergenicity. Acids (e.g. nitric) have low detergency, although they are very useful in solubilizing carbonate and mineral scales, including hard water salts and proteinaceous deposits. As with alkalis, the stronger the acid the more effective it is; though, in addition, the more corrosive to plant and operatives. Acids are not used as frequently as alkalis in chilled food operations and tend to be used for periodic cleans. Sequestering agents (sequestrants or chelating agents) are employed to prevent mineral ions precipitating by forming soluble complexes with them. Their primary use is in the control of water hardness ions and they are added to surfactants to aid their dispersion capacity and rinsability. Sequestrants are most commonly based on ethylene diamine tetracetic acid (EDTA), which is expensive. Although cheaper alternatives are available, these are usually polyphosphates which are environmentally unfriendly. A general-purpose food detergent may, therefore, contain a strong alkali to saponify fats, weaker alkali ‘builders’ or ‘bulking’ agents, surfactants to improve wetting, dispersion and rinsability and sequestrants to control hard water ions. In addition, the detergent should ideally be safe, non-tainting, non-corrosive, stable, environmentally friendly and cheap. The choice of cleaning agent will depend on © 2008, Woodhead Publishing Limited

312

Chilled foods

Table 12.2 Solubility characteristics and cleaning procedures recommended for a range of soil types Soil type

Solubility characteristics

Cleaning agent recommended

Sugars, organic acids, salt High protein foods (meat, poultry, fish)

Water-soluble Water-soluble Alkali-soluble Slightly acid-soluble Partly water-soluble Alkali-soluble Water-insoluble Alkaline-soluble Water-insoluble Alkaline-insoluble Acid-soluble

Mildly alkaline detergent Chlorinated alkaline detergent

Starchy foods, tomatoes, fruits Fatty foods (fat, butter, margarine, oils) Heat-precipitated water hardness, milk stone, protein scale

Mildly alkaline detergent Mildly alkaline detergent; if ineffective, use strong alkali Acid cleaner, used on a periodic basis

Modified from Elliot, 1980.

the soil to be removed and on its solubility characteristics, and these are summarised for a range of chilled products in Table 12.2 (modified from Elliot, 1980). Because of the wide range of food soils likely to be encountered and the influence of the food manufacturing site (temperature, humidity, type of equipment, time before cleaning, etc.), there are currently no recognised laboratory methods for assessing the efficacy of cleaning compounds. Food manufacturers have to be satisfied that cleaning chemicals are working appropriately by conducting suitable field trials. Because of their key role in the sanitation process, therefore, food manufacturers should only purchase high quality detergents: cost cutting for less effective chemicals may seriously reduce the efficacy of the sanitation process. Although the majority of the microbial contamination is removed by the cleaning phase of the sanitation programme, there are likely to be sufficient viable micro-organisms remaining on the surface to warrant the application of a disinfectant. The aim of disinfection is therefore to further reduce the surface population of viable micro-organisms, via removal or destruction, and/or to prevent surface microbial growth during the inter-production period. Elevated temperature is the best disinfectant as it penetrates into surfaces, is non-corrosive, is non-selective to microbial types, is easily measured and leaves no residue (Jennings, 1965). However, for open surfaces, the use of hot water or steam is uneconomic, hazardous or impossible, and reliance is therefore placed on chemical biocides. Whilst there are many chemicals with biocidal properties, many common disinfectants are not used in food applications because of safety or taint problems, e.g. phenolics or metal-ion-based products. Other disinfectants are used only to a limited extent in chilled food manufacture and/or for specific purposes, e.g. peracetic acid, biguanides, formaldehyde, glutaraldehyde, organic acids, ozone, chlorine dioxide, bromine and iodine compounds. Of the acceptable chemicals, the most commonly used products for open surface disinfection are: © 2008, Woodhead Publishing Limited

Cleaning and disinfection of chilled food plants and equipment (i) (ii) (iii) (iv)

313

chlorine-releasing components quaternary ammonium compounds quaternary ammonium/amphoteric mixtures alcohols

and for closed surfaces (CIP) (v)

peracetic acid. Chlorine is the cheapest disinfectant and is available as hypochlorite (or occasionally as chlorine gas) or in slow releasing forms (e.g. chloramines, dichlorodimethylhydrantoin). It should not be used with acidic cleaning products because of the hazard of releasing chlorine gas. Quaternary ammonium compounds (Quats or QACs) are amphipolar, cationic detergents, derived from substituted ammonium salts with a chlorine or bromine anion and amphoterics are based on the amino acid glycine, often incorporating an imidazole group. All cleaning agents should be used at the concentrations recommended by the manufacturer because below that level they may have significantly reduced biocidal action, and above that level they will be uneconomic and may pose a health and safety risk. In a (CCFRA) survey undertaken of the UK food industry in 1987, of 145 applications of disinfectants, 52% were chlorine based, 37% were quaternary ammonium compounds and 8% were amphoterics. Of these biocides there were, respectively, 44, 30 and 8 branded products used. In a (CCFRA) European survey of 1993, the most common disinfectants used in the UK and Scandinavian countries were QACs for open surfaces and peracetic acid and chlorine for closed, liquid handling surfaces. The survey also showed that open surfaces were usually cleaned with alkaline detergents which were foamed and then rinsed with medium pressure water (250psi) and closed systems were CIP cleaned with caustic followed by acidic detergents, with a suitable rinse in-between. A survey of the approved disinfectant products in Germany (DVG listed) in 1994 indicated that 36% were QACs, 20% were mixtures of QACs with aldehydes or biguanides, and 10% were amphoterics (Knauer-Kraetzl, 1994). In a (CCFRA) survey undertaken of the UK food industry in 2000, of 117 food premises (incorporating farms, food manufacturers, food hauliers and caterers), 70% used QACs, 49% chlorine releasing agents, 26% alcohol, 10% amphoterics and 7% peracetic acid. From the mid 1990s the synergistic combinations of QACs and amphoterics have been explored in the UK and these compounds are now widely used in chilled food plants. The characteristics of the most commonly used are compared in Table 12.3. The properties of QAC/amphoteric mixes will be similar to their parent compounds, with suggested enhanced micro-organism control. The table essentially confirms that QACs, amphoterics and alcohol are used predominantly for open plant cleaning because they combine sufficient antimicrobial properties yet are ‘friendly’ to operatives and the processing environment. Chlorine has excellent antimicrobial properties though, due to its corrosiveness, it is used sparingly on open surfaces and is followed by thorough rinsing. Peracetic acid is perhaps the most antimicrobial chemical to micro-organisms in suspension, attached to surfaces and © 2008, Woodhead Publishing Limited

314

Chilled foods

Table 12.3 Characteristics of some universal disinfectants Micro-organism control Gram-positive Gram-negative Spores Yeast Developed microbial resistance Inactivation by organic matter water hardness Detergency properties Surface activity Foaming potential Problems with taints stability corrosion safety other chemicals Potential environmental impact Cost per area cleared

Chlorine

QAC

Amphoteric

Alcohol acid

Peracetic

++ ++ + ++ – ++ – – – – +/– +/– + + – ++ –

++ + – ++ + + + ++ ++ ++ – – – – + –/+ ++

++ ++ – ++ + + – + ++ ++ – – – – – –/+ ++

++ ++ – + – ++ – – – – – – – + – – +++

++ ++ ++ ++ – + – – – – +/– +/– – ++ – – +

– no effect (or problem) + effect + + large effect

in biofilms (Holah et al., 1990a), though is more of a hazard to cleaning operatives and, while the most common disinfectant used in closed plant systems, is used only under controlled conditions for open plant. Within the chilled food industry, particularly for mid-shift cleaning and disinfection in high risk areas, alcohol based products are commonly used. This is primarily to restrict the use of water for cleaning during production as a control measure to prevent the growth and spread of any food pathogens that penetrate the high risk area barrier controls. Ethyl alcohol (ethanol) and isopropyl alcohol (isopropanol) have bactericidal and virucidal (but not sporicidal) properties (Hugo and Russell, 1999), though they are active only in the absence of organic matter, i.e. the surfaces need to be wiped clean and then alcohol reapplied. Alcohols are most active in the 60–70% range, and can be formulated into wipe and spray based products. Alcohol products are used on a small, local scale because of their well recognised health and safety issues. The efficacy of disinfectants is generally controlled by five factors: interfering substances (primarily organic matter and hard water ions), pH, temperature, concentration and contact time. To some extent, and particularly for the oxidative biocides, the efficiency of all disinfectants is reduced in the presence of organic matter. Organic material may react chemically with the disinfectant such that it loses its biocidal potency, or spatially such that micro-organisms are protected from its effect. Other interfering substances, e.g. cleaning chemicals, may react with the disinfectant and destroy its antimicrobial properties, and it is therefore essential to remove all soil and chemical residues prior to disinfection. This is © 2008, Woodhead Publishing Limited

Cleaning and disinfection of chilled food plants and equipment

315

particularly important with cationic quaternary ammonium compounds such that, if used, all traces of anionic detergents need to be removed prior to disinfection. Disinfectants should be used only within the pH range specified by the manufacturer. Perhaps the classic example of this is chlorine, which from pH 3–7.5 is predominantly present as HOCl (hypochlorous acid, often termed ‘free chlorine’), which is a very powerful biocide, though the potential for corrosion increases with acidity. Above pH 7.5, however, the majority of the chlorine is present as the dissociated OCl– ion which has about 100 times less biocidal action than HOCl. In general, the higher the temperature the greater the disinfection. For most food manufacturing sites operating at ambient conditions (around 20 °C) or higher, this is not a problem as most disinfectants are formulated (and tested) to ensure performance at this temperature. However, the chilled food industry operates at lower temperatures. Taylor et al. (1999) examined the efficacy of 18 disinfectants at both 10 °C and 20 °C and demonstrated that for some chemicals, particularly quaternary ammonium based products, disinfection was much reduced at 10 °C. They recommended that in chilled production environments only products specifically formulated for low temperature activity should be used. In practice, the relationship between microbial death and disinfectant concentration is not linear but follows a sigmoidal curve. Microbial populations are initially difficult to kill at low concentrations, but as the biocide concentration is increased, a point is reached where the majority of the population is reduced. Beyond this point the micro-organisms become more difficult to kill (through resistance or physical protection) and a proportion may survive regardless of the increase in concentration. It is important, therefore, to use the disinfectant at the concentration recommended by the manufacturer. Concentrations above this recommended level may thus not enhance biocidal effect and will be uneconomic, whilst concentrations below this level may significantly reduce biocidal action. Sufficient contact time between the disinfectant and the micro-organisms is perhaps the most important factor controlling biocidal efficiency. To be effective, disinfectants must find and concentrate at, bind to and transverse microbial cell envelopes before they reach their target site, and begin to undertake the reactions which will subsequently lead to the destruction of the micro-organism (Klemperer, 1982). Sufficient contact time is therefore critical to give good results, and most general-purpose disinfectants are formulated to require at least 5 minutes to reduce bacterial populations by 5 log orders in suspension. This has arisen for two reasons. Firstly, 5 minutes is a reasonable approximation of the time taken for disinfectants to drain off vertical or near vertical food processing surfaces. Secondly, when undertaking disinfectant efficacy tests in the laboratory, a 5-minute contact time is chosen to allow ease of test manipulation and hence timing accuracy. For particularly resistant organisms such as spores or moulds, surfaces should be repeatedly dosed to ensure extended contact times of 15–60 minutes. Ideally, disinfectants should have the widest possible spectrum of activity against micro-organisms, including bacteria, fungi, spores and viruses, and this should be demonstrable by means of standard disinfectant efficacy tests. The range © 2008, Woodhead Publishing Limited

316

Chilled foods

of currently available disinfectant test methods was reviewed by Reybrouck (1998) and fall into two main classes, suspension tests and surface tests. Suspension tests are useful for indicating general disinfectant efficacy and for assessing environmental parameters such as temperature, contact time and interfering matter such as food residues. In reality however, micro-organisms disinfected on food contact surfaces are those that remain after cleaning and are therefore likely to be adhered to the surface. A surface test is thus more appropriate. A number of authors have shown that bacteria attached to various surfaces are generally more resistant to biocides than are organisms in suspension (Dhaliwal et al., 1992; Frank and Koffi, 1990; Holah et al., 1990a; Hugo et al., 1985; Le Chevalier et al., 1988; Lee and Frank, 1991: Ridgeway and Olsen, 1982; Wright et al., 1991; Andrade et al., 1998, Das et al., 1998). In addition, cells growing as a biofilm have been shown to be more resistant (Frank and Koffi, 1990; Lee and Frank, 1991; Ronner and Wong, 1993). The mechanism of resistance in attached and biofilm cells is unclear but may be due to physiological differences such as growth rate, membrane orientation changes due to attachment and the formation of extracellular material which surrounds the cell. Equally, physical properties may have an effect, e.g. protection of the cells by food debris or the material surface structure, or problems in biocide diffusion to the cell/material surface. To counteract such claims of enhanced surface adhered resistance, it can be argued that, in reality, surface tests do not consider the environmental stresses the organisms may encounter in the processing environment prior to disinfection (action of detergents, variations in temperature and pH and mechanical stresses) which may affect susceptibility. Both suspension and surface tests have limitations, however, and research based methods are being developed to investigate the effect of disinfectants against adhered micro-organisms and biofilms in-situ and in real time. Such methods have been reviewed by Holah et al. (1998). In Europe, CEN TC 216 has produced (and is currently still developing) a number of harmonised disinfectant testing standards. The current food industry disinfectant test methods of choice for bactericidal and fungicidal action in suspension are EN 1276 (Anon., 1997a) and EN 1650 (Anon., 1998a), respectively, and food manufacturers should ensure that the disinfectants they use conform to these standards as appropriate. In particular, for chilled processing areas, evidence of performance against EN 1276 at 10 °C should be sought from disinfectant manufacturers. The bactericidal suspension test EN 1276 utilises micro-organisms that are particularly resistant to disinfectants, for example Pseudomonas aeruginosa and Staphylococcus aureus are chosen as they are recognised as one of the most resistant Gram negative and positive species respectively. Pathogens relevant to chilled food manufacture such as Listeria monocytogenes, Salmonella spp. or Escherichia coli are not known to be resistant to disinfectants such that disinfectants that ‘pass’ EN 1276 will most likely be effective against them. In addition, many disinfectant manufacturers have also undertaken additional disinfectant tests using the methodology of EN 1276 with these pathogens as test strains to demonstrate their products’ effectiveness against them. © 2008, Woodhead Publishing Limited

Cleaning and disinfection of chilled food plants and equipment

317

A harmonised surface test, EN 13697 (Anon., 2001), is also available for disinfectant manufacturers to demonstrate efficacy of their products against surface adhered micro-organisms. Because of the limitations of disinfectant efficacy tests, however, food manufacturers should always confirm the efficacy of their cleaning and disinfection programmes by ‘field tests’, either from evidence supplied by the chemical company or from in-house validation trials. CEN TC 216 has also produce disinfectant testing standards for biocidal soaps (EN 1499, Anon., 1997b) and alcoholic hand rubs (EN 1500, Anon., 1997c) and again, chilled food manufacturers should ensure that the hand hygiene products they use conform to these standards (see Lambert, 2001). As well as having demonstrable biocidal properties, disinfectants must also be safe (non-toxic) and should not taint food products. Disinfectants can enter food products accidentally, e.g. from aerial transfer or poor rinsing, or deliberately, e.g. from ‘no rinse status’ disinfectants. The practice of rinsing or not rinsing has been under discussion for many years and has yet to be clarified. The main reason for leaving disinfectants on surfaces is to provide an alleged biocide challenge (this has not been proven) to any subsequent microbial contamination of the surface. It has been argued, however, that the low biocide concentrations remaining on the surface, especially if the biocide is a QAC, may lead to the formation of resistant surface populations (again not proven in practical situations). At a European level there is no legislation requiring disinfectants to be rinsed from surfaces. In other words, disinfectants can be left on surfaces if the food manufacturer is happy that any residues will not affect the wholesomeness or safety of subsequently produced foodstuffs. In terms of the demonstration of non-toxicity, legislation will vary in each country although in Europe, this will be clarified with the implementation of Directive 98/8/EC concerning the placing of biocidal products on the market, which contains requirements for toxicological and metabolic studies. This Directive seeks to produce a list of active biocidal substances that have been assessed for both their toxicological properties and also their inherent antimicrobial properties. Following the establishment of the approved active list, formulated products (the disinfectants sold to the final user) can then be made only by incorporating an approved active ingredient, and the formulated product will then itself be assessed for its toxicological and antimicrobial properties. The effects of this Directive are already being seen, as current formulated products that do not contain biocidal active ingredients that will be supported through the approval process are being removed from the marketplace. This will undoubtedly reduce the choice of disinfectants available to the food industry in the future, though the ones that will be available will have been thoroughly assessed. In the interim, a recognised acceptable industry guideline for disinfectants is a minimum acute oral toxicity (with rats) of 2000 mg/kg bodyweight. Approximately 30% of food taint complaints are thought to be associated with cleaning and disinfectant chemicals and are described by sensory scientists as ‘soapy’, ‘antiseptic’ or ‘disinfectant’ (Holah, 1995). CCFRA has developed two taint tests in which foodstuffs which have and have not been exposed to disinfectant © 2008, Woodhead Publishing Limited

318

Chilled foods

residues are compared by a trained taste panel using the standard triangular taste test (Anon., 1983a). For assessment of aerial transfer, a modification of a packaging material odour transfer test is used (Anon., 1964) in which food products, usually of four types (high moisture e.g. melon, low moisture e.g. biscuit, high fat e.g. cream, high protein e.g. chicken) are held above a disinfectant solution or distilled water for 24 hours. To assess surface transfer, a modification of a food container transfer test is used (Anon., 1983b) in which food products are sandwiched between two sheets of stainless steel and left for 24 hours. Disinfectants can be sprayed onto the stainless steel sheets and drained off, to simulate no rinse status, or can be rinsed off prior to food contact. Control sheets are rinsed in distilled water only. The results of the triangular test involve both a statistical assessment of any flavour differences between the control and disinfectant treated sample, and a description of any flavour changes.

12.4 Sanitation methodology Cleaning and disinfection can be undertaken by hand using simple tools, e.g. brushes or cloths (manual cleaning), though as the area of open surface requiring cleaning and disinfection increases, specialist equipment becomes necessary to dispense chemicals and/or provide mechanical energy. Chemicals may be applied as low pressure mists, foams or gels, whilst mechanical energy is provided by high and low pressure water jets or water or electrically powered scrubbing brushes. These techniques have been well documented (Anon., 1991; Marriott, 1985; Holah, 1991) and as with cleaning chemicals, have changed little since these reviews: this section considers their use in practice. The use of cleaning techniques can perhaps be described schematically following the information detailed in Fig. 12.2 (modified from Offiler, 1990). The figure details the different energy source inputs for a number of cleaning techniques and shows their ability to cope with both low and high (dotted line) levels of soiling. For the manual cleaning of small items, a high degree of mechanical energy can be applied directly where it is needed, and with the use of soak tanks (or clean-out-of-place techniques) contact times can be extended and/or chemical and temperature inputs increased such that all soil types can be tackled. Alternatively, dismantled equipment and production utensils may undergo manual gross soil removal and then be cleaned and disinfected automatically in tray or tunnel washers. As with soak tank operations, high levels of chemical and thermal energy can be used to cope with the majority of soils. Tray washes for high risk chilled production areas should be sited in separated areas with good extraction and air make-up facilities which can be closed off from production areas as they are prone to microbial aerosol production which may lead to aerial product contamination (see Chapter 11). In manual cleaning of larger areas, especially above chest level, for reasons of operator safety, cleaning operatives should use only low temperature water (<60 °C) and low levels of chemical energy. As the surface area requiring © 2008, Woodhead Publishing Limited

Energy input

Light soil

CHEM

CHEM

CHEM

CHEM

CHEM

MECH

MECH

TIME

MECH MECH

MECH

MECH TEMP

TIME

CHEM

CHEM

MECH

TEMP

CHEM

TEMP TEMP

TEMP

TIME

TIME

TEMP

TIME TIME

TEMP

TIME

MECH

TEMP TIME

Manual small items

Manual soak tank

Manual tray wash

Manual large areas

Mist

Foam

Gel

Mechanical scrubbing

Cleaning technique

Relative energy source inputs for a range of cleaning techniques. (Modified from Offiler, 1990.)

319

Fig. 12.2 © 2008, Woodhead Publishing Limited

Cleaning and disinfection of chilled food plants and equipment

Heavy soil

320

Chilled foods

cleaning increases, the manual techniques become uneconomic with respect to time and labour. Labour costs amount to 75% of the total sanitation programme and for most food companies, the cost of extra staff is prohibitive. Only light levels of soiling can be economically cleaned by this method. The main difference between the mist, foam and gel techniques is in their ability to maintain a detergent/soil/surface contact time. For all three techniques, mechanical energy can be varied by the use of high-or low-pressure water rinses, though for open surface cleaning, temperature effects are minimal. Mist spraying is undertaken using small, hand-pumped containers, ‘knapsack’ sprayers or pressure-washing systems at low pressure. Misting will only ‘wet’ vertical smooth surfaces; therefore only small quantities can be applied and these will quickly run off to give a contact time of 5 minutes or less. Because of the nature of the technique to form aerosols that could be an inhalation hazard, only weak chemicals can be applied, and so misting is useful only for light soiling. On cleaned surfaces, however, misting is the most commonly used method for applying disinfectants. Foams can be generated and applied by high-pressure equipment with entrapment of air in the foam or by the addition of compressed air in low-pressure systems. Foams work on the basis of reducing the density of the cleaning agent and forming a layer of bubbles above the surface to be cleaned which then collapses and bathes the surface with fresh detergent contained in the bubble film. The critical element in foam generation is for the bubbles to collapse at the correct rate: too fast and the contact time will be minimal; too slow and the surface will not be wetted with fresh detergent. Gels are thixotropic chemicals which are fluid at high shear rates (i.e. during application) but become thick and gelatinous at at low or zero rates of shear (i.e. after application). Gels are easily applied through high and low pressure systems or from specific portable electric pumped units, and physically adhere to the surface. Gels may be coloured to show coverage. Foams and gels are more viscous than mists; they are not as prone to aerosol formation and thus allow the use of more concentrated detergents, and can remain on vertical surfaces for much longer periods (foams 10–15 minutes, gels 15 minutes to an hour or more). Foams and gels are able to cope with higher levels of soil than misting, although in some cases rinsing of surfaces may require large volumes of water, especially with foams. Foams and gels are well liked by operatives and management as, because of their nature, a more consistent application of chemicals is possible and it is easier to identify areas that have been ‘missed’. Fogging systems have been traditionally used in the chilled food industry to create and disperse a disinfectant aerosol to reduce airborne micro-organisms and to apply disinfectant to difficult-to-reach overhead surfaces. The efficacy of fogging has been reported (Anon., 1998b) and, providing a suitable disinfectant is used, the method is effective at reducing airborne microbial populations by 2–3 log orders in 30–60 minutes. Fogging is most effective using compressed air-driven fogging nozzles producing particles in the 10–20 micron range. For surface disinfection, fogging is effective only if sufficient chemical can be deposited onto © 2008, Woodhead Publishing Limited

Cleaning and disinfection of chilled food plants and equipment

321

Log reduction

8 6 4 2 0 276 cm

204 cm

167 cm

119 cm

10 cm

Height from floor Horizontal

Fig. 12.3

Vertical

Underneath

Comparative log reductions of micro-organisms adhered to surfaces and positioned at various heights and orientations.

the surface. This is illustrated in Fig. 12.3 which shows the log reductions achieved on horizontal, vertical and upturned (underneath) surfaces arranged at five different heights from just below the ceiling (276 cm) to just above the floor (10 cm) within a test room. It can be seen that disinfection is greatest on surfaces closest to the floor and that disinfection is minimal on upturned surfaces close to the ceiling. Commercial fogging units are now available which electrostatically charge the disinfectant during the fogging process. The charged disinfectant droplets have been shown to be more effective at wetting overhead and vertical surfaces. To reduce inhalation risks, sufficient time (45–60 minutes) is required after fogging to allow the settling of disinfectant aerosols before operatives can re-enter the production area. Cleaning chemicals are removed from surfaces by low pressure/high volume (LPHV) hoses operating at mains water pressure (typically <10 bar) or by high pressure/low volume (HPLV) pressure washing systems which require a high pressure pump. High pressure washing systems operate above mains water pressure, typically at between 25–100 bar through a 15° nozzle, and may be mobile units, wall mounted units or centralised ring-mains. Water jets confer high mechanical energy, can be used on a wide range of equipment and environmental surfaces, will penetrate into surface irregularities and are able to mix and apply chemicals. Mechanical scrubbers include traditional floor scrubbers, scrubber/driers (automats) for floors, and water-driven attachments to high pressure systems and electrically operated small-diameter brushes that can be used on floors, walls and other surfaces. Contact time is usually limited with these techniques (though can be increased), but the combination of detergency with high mechanical input allows them to tackle most soil types. These techniques work best when the food processing areas to be cleaned have been designed or refurbished to suit their use (primarily to facilitate ease of access). © 2008, Woodhead Publishing Limited

322

Chilled foods

The hygienic implications of the design and use of all cleaning equipment should be carefully considered and should be similar to those required for other food processing surfaces (Chapter 11). Sanitation equipment should be constructed out of smooth, non-porous, easily cleanable materials such as stainless steel or plastic. Mild steel or other materials subject to corrosion may be used but must be suitably painted or coated, whilst the use of wood is unacceptable. Frameworks should be constructed of tubular or box section material, closed at either end and properly jointed, e.g. welds should be ground and polished and there should be no metal-to-metal joints. Crevices and ledges where soil could collect should be prevented and exposed threads should be covered or dome nuts used. Tanks for holding cleaning chemicals or recovered liquids should be self-draining, have rounded corners and should be easily cleaned. Shrouds around brush heads or hoods and rotary scrubbing heads should be easily detachable to facilitate cleaning. Brushes should have bristles of coloured, impervious material, e.g. nylon, embedded into the head with resin so that no soil trap points are apparent. Alternatively, brushes with the head and bristles moulded as one unit may be used. Cleaning equipment is prone to contamination with Listeria spp. and other pathogenic micro-oganisms and, by the nature of its use, provides an excellent way in which contamination can be transferred from area to area. Cleaning equipment should be specific to high risk and after use, equipment should be thoroughly cleaned and, if appropriate, disinfected and dried. The potential for cleaning equipment to disperse microbial contamination by the formation of aerosols has been reported (Holah et al., 1990b) and it was shown that all cleaning systems tested produced viable bacterial aerosols from test surfaces contaminated with attached biofilms. The degree of contamination impinging on a surface was graded from total coverage to the minimum level thought likely to give food safety concerns if a proportion of the droplets contained viable micro-organisms, and the maximum height and distance travelled by these potentially contaminating droplets are shown in Table 12.4. Assuming an average food contact surface height of 1 m, the results suggest that both the HPLV and LPHV techniques disperse a significant numbers of droplets to this height and should not, therefore, be used during production periods. The other techniques, however, are acceptable for use in clean-as-you-go operations as the chance of contamination to product is low, though care is needed when using floor scrubber/driers (these are useful in that the cleaning fluid is removed from the floor) if product is stored in Table 12.4 Maximum height and distance spread of aerosol (droplet) impingement for a number of cleaning techniques Cleaning technique High pressure/low volume spray lance Low pressure/high volume hose Floor scrubber/drier Manual brushing Manual wiping

© 2008, Woodhead Publishing Limited

Height (cm)

Distance (cm)

309 210 47 24 23

700 350 80 75 45

Cleaning and disinfection of chilled food plants and equipment

323

racks close to the floor. After production, HPLV and LPHV techniques may be safely used (and are likely to be the appropriate choice), but it is required that disinfection of food contact surfaces is the last operation to be performed within the sanitation programme. Subsequent work has shown that reducing water pressure or changing impact angle made little difference to the degree of aerosol spread for HPLV and LPHV systems, dispersal to heights >1 m still being achieved.

12.5 Sanitation procedures Sanitation procedures are concerned with both the stage in which the sanitation programme is implemented and the sequence in which equipment and environmental surfaces are cleaned and disinfected within the processing area. Sanitation programmes are so constructed to be efficient with water and chemicals, to allow selected chemicals to be used under their optimum conditions, to be safe in operation, to be easily managed and to reduce manual labour. In this way, effective sanitation will be achieved economically and with due regard to environmental discharges, and energy and chemical use. The principal stages involved in a typical sanitation programme are: (i)

(ii)

Production. In line with the Listeria control policy as noted in Chapter 11, care should be taken to prevent the growth and spread of pathogenic microorganisms during production periods. This is primarily undertaken by basing production periods on the growth rate of target micro-organisms to minimise their growth, and by excluding or severely limiting the use of water for cleaning purposes and using manual cleaning (brushing, wiping) with alcohol as a solvent. Production staff should be encouraged to operate good housekeeping practices (this is an aid to ensuring acceptable product quality and personnel safety), to clean their work stations prior to break periods (unless cleaning operatives are employed specifically to do this) and to leave their work stations in a hygienic condition. Food debris left in hoppers and on process lines, etc. is wasted product! Sound sanitation practices should be used to clean up major product spillages and re-start production, which may require the use of wet cleaning methods. Preparation. As soon as possible after production, equipment for manual cleaning should be dismantled as far as is practicable or necessary to make all surfaces that micro-organisms could have adhered to during production accessible to the cleaning fluids. All product and unwanted utensils/packaging/equipment should be covered or removed from the area. Dismantled equipment should be stored on racks or tables, not on the floor! Machinery should be switched off, at the machine and at the power source, and electrical and other sensitive systems protected from water/chemical ingress. Preferably, production should not be done in the area being cleaned, but in exceptional circumstances if this is not possible, other lines or areas should

© 2008, Woodhead Publishing Limited

324

Chilled foods

be screened off to prevent transfer of debris as aerosols or splashes by the sanitation process. (iii) Gross soil removal. Where appropriate, all loosely adhered or gross soil should be removed by brushing, scraping, shovelling or vacuum, etc. Wherever possible, soil on floors and walls should be picked up and placed in suitable waste containers rather than washed into drains using hoses. (iv) Pre-rinse. Surfaces should be rinsed with low pressure cold water to remove loosely adhered small debris. Hot water can be used for fatty soils (approximately 60 °C), but too high a temperature (>45 °C) may coagulate proteins. (v) Cleaning. A selection of cleaning chemicals, temperature and mechanical energy is applied to remove adhered soils. (vi) Inter-rinse. Both soil detached by cleaning operations and cleaning chemical residues should be removed from surfaces by rinsing with low-pressure cold water. (vii) Disinfection. Chemical disinfectants (or occasionally heat) are applied to remove and/or reduce the viability of remaining micro-organisms to a level deemed to be of no significant risk. In exceptional circumstances and only when light soiling is to be removed, it may be appropriate to combine stages v–vii by using a chemical with both cleaning and antimicrobial properties (detergent–sanitiser). (viii) Post-rinse. Disinfectant residues should be removed by rinsing away with low-pressure cold water of known potable quality. Some disinfectants, however, are intended to be left on surfaces until the start of subsequent production periods and are thus so formulated to be both surface-active and of low risk, in terms of taint or toxicity, to foodstuffs. (ix) Inter-production cycle conditions. A number of procedures may be undertaken, including the removal of excess water and/or equipment drying, to prevent the growth of micro-organisms on production contact surfaces in the period up until the next production process. Alternatively, the processing area may be evacuated and fogged with a suitable disinfectant. (x) Periodic practices. Periodic practices increase the degree of cleaning for specific equipment or areas to return them to acceptable cleanliness levels. They include weekly acidic cleans, weekend dismantling of equipment, cleaning and disinfection of chillers and sanitation of surfaces, fixtures and fittings above 2 m. A sanitation sequence should be established in a processing area to ensure that the applied sanitation programme is capable of meeting its objectives and that cleaning programmes, both periodic and for areas not cleaned daily, are implemented on a routine basis. In particular, a sanitation sequence determines the order in which the product contact surfaces of equipment and environmental surfaces (walls, floors, drains, etc.) are sanitised, such that once product contact surfaces are disinfected, they should not be re-contaminated. Based on industrial case studies, the following sanitation sequence for high care/risk chilled food production areas has been demonstrated to be useful in © 2008, Woodhead Publishing Limited

Cleaning and disinfection of chilled food plants and equipment

325

controlling the proliferation of undesirable micro-organisms. The sequence must be performed at a ‘room’ level such that all environmental surfaces and equipment in the area are cleaned at the same time. It is not acceptable to clean and disinfect one line and then move onto the next and start the sequence again as this merely spreads contamination around the room. (i) Remove gross soil from production equipment. (ii) Remove gross soil from environmental surfaces. (iii) Rinse down environmental surfaces (usually to a minimum of 2 m in height for walls). (iv) Rinse down equipment and flush to drain. (v) Clean environmental surfaces, usually in the order of drains, walls then floors. (vi) Rinse environmental surfaces. (vii) Clean equipment. (viii) Rinse equipment. (ix) Disinfect equipment and rinse if required. (x) Fog (if required).

12.6 Evaluation of sanitation effectiveness Assessment of the effectiveness of the sanitation programme’s performance is part of day-to-day hygiene testing and, as such, is linked to the factory environmental sampling plan. The control of the environmental routes of contamination is addressed via the development of a thorough risk analysis and management strategy, typically undertaken as part of the factory HACCP study, resulting in the development of the factory environmental sampling plan. The development of environmental sampling plans has been established by a CCFRA industrial working party and is reported in Holah (1998). Environmental sampling is directly linked with both process development and product manufacture and has three distinct phases:

• the HACCP team, and especially the process development function, should

• •

determine whether a process route presents an unacceptable contamination risk and assess whether procedures put in place to control the risk identified are working; QA should provide an assessment of routine hygiene; troubleshooter (usually company microbiologist) to identify why products (or occasionally environmental samples) may have a microbiological count that is out of specification or may contain pathogens.

Within chilled food manufacture, routine hygiene testing is related to the Listeria barrier control philosophy outlined in Chapter 11. It is thus concerned with assessing the performance of the high risk barrier systems in preventing pathogen access during production, the level of pathogen in the processing area during © 2008, Woodhead Publishing Limited

326

Chilled foods

processing and, after production has finished, the performance of the sanitation programme. The performance of the sanitation programme is an important aspect of due diligence and is assessed by monitoring to check sanitation process control, and verifying the sanitation programme success. Monitoring is a planned sequence of observations or measurements to ensure that the control measures within the sanitation programme are operating within specification and are undertaken in a time frame that allows sanitation programme control. Verification is the application of methods in a longer time frame to determine compliance with the sanitation programme’s specification. Monitoring the sanitation programme is via physical, sensory and rapid chemical hygiene testing methods. Microbiological testing procedures are never fast enough to be used for process monitoring. Physical tests are centred on the critical control measures of the sanitation programme’s performance and include, for example, measurement of detergent/disinfectant contact time; rinse water, detergent and disinfectant temperatures; chemical concentrations; surface coverage of applied chemicals; degree of mechanical or kinetic input; cleaning equipment maintenance; and chemical stock rotation. The primary monitoring of any cleaning programme is ‘visual’ cleanliness and involves the assessment of a surface as being free from food debris and other soiling by a person without any sampling aids (other than perhaps a torch). This may involve looking at the surface, feeling the surface for any signs of ‘invisible’ deposits such as grease and oils, and even smelling the equipment. Not all surface materials will necessarily clean to the same degree and not all surfaces will be ‘visible’. Dismantling may be needed to establish that all appropriate surfaces have been visibly cleaned. As well as freedom from soiling, visual inspection should also establish freedom of any other hazards attributed to the cleaning programme. Primarily these are related to cleaning fluids and foreign bodies. Cleaning fluids may be hazards in their own right, particularly in their concentrated form; or when diluted, they may become a nutrient and water source to aid residual microbial growth. Inspection should therefore concentrate on the identification of any undrainable surfaces or other areas in which liquids could be contained. Foreign bodies may arise from cleaning equipment such that rough or sharp equipment surfaces may cause the entrapment of brush bristles or the disintegration of cleaning cloths/pads. Rapid hygiene monitoring methods are methods whose results are generated within a time frame (usually regarded as within approximately 10 minutes) sufficiently quickly to allow process control. Current methodology allows the quantification of micro-organisms (ATP – see below), food soils (ATP, protein) or both (ATP). No technique is presently available which will allow the detection of specific microbial types within this time frame. The most popular and established rapid hygiene monitoring technique is that based on the detection of adenosine triphosphate (ATP) by bioluminescence and is usually referred to as ATP testing. ATP is present in all living organisms, including micro-organisms (microbial ATP), and in a variety of foodstuffs, and may also be © 2008, Woodhead Publishing Limited

Cleaning and disinfection of chilled food plants and equipment

327

present as free ATP (usually referred to together as non-microbial ATP). The bioluminescent detection system is based on the chemistry of the light reaction emitted from the abdomen of the North American firefly Photinus pyralis, in which light is produced by the reaction of luciferin and luciferase in the presence of ATP. For each molecule of ATP present, one photon of light is emitted, which is then detected by a luminometer and recorded as relative light units (RLU). The reaction is very rapid and results are available within seconds of placing the sample to be quantified in the luminometer. The result, the amount of light produced, is also directly related to the level of microbial and non-microbial ATP present in the sample and is often referred to as the ‘hygienic’ status of the sample. ATP detection methodology was first established in the late 1980s, e.g. Bautista et al. (1992), Poulis et al. (1993), Bell et al.(1994), Griffith et al. (1994), Hawronskyj and Holah (1997), and is now a recognised standard method. A number of companies offer ATP hygiene detection kits and they are differentiated primarily by cost, service and data handling packages. It is possible to differentiate between the measurement of microbial and non-microbial ATP but for the vast majority of cases, the measurement of total ATP (microbial and non-microbial) is preferred. As there is more inherent ATP in foodstuffs than in micro-organisms, the measurement of total ATP is a sensitive technique to determine remaining residues. Large quantities of ATP present on a surface after cleaning and disinfection, regardless of their source, is an indication of poor cleaning and thus contamination risk (from microorganisms or materials that may support their growth). Many food processors typically use the rapidity of ATP detection to allow monitoring of the cleaning operation such that if a surface is not cleaned to a predetermined level it can be recleaned prior to production. Similarly, equipment can be certified as being cleaned prior to use in processing environments where kit is quickly re-used or when a manufacturing process has long production runs. Some processors prefer to assess the hygiene level after the completion of both the cleaning and disinfection phases, whilst others monitor after the cleaning phase and go on to the disinfection phase only if the surfaces have been adequately cleaned. Techniques have also been developed which use protein concentrations as markers of surface contamination remaining after cleaning operations. As these are dependent on chemical reactions, they are also rapid but their applicability is perhaps less widespread as they can be used only if protein is a major part of the food product processed. As with the ATP technique, a direct correlation between the amount of protein remaining after a sanitation programme has been completed and the number of micro-organisms remaining as assessed by traditional microbiological techniques is not likely to be achieved. They are cheaper in use than ATP based systems as the end point of the tests is a visible colour change rather than a signal interpreted by an instrument, e.g. light output measured by a luminometer. Protein hygiene test kits were developed later than ATP kits (e.g. Griffith et al., 1997) and are also now seen as established standard methods. Verification of the sanitation programme’s performance is usually undertaken by microbiological methods in the chilled food industry, though ATP levels are © 2008, Woodhead Publishing Limited

328

Chilled foods

also used (especially in low risk). Microbiological sampling is typically for the total number of viable micro-organisms remaining after cleaning and disinfection, i.e. total viable count (TVC) or indicator micro-organisms, as a measurement of the ability of the sanitation programme to both control all micro-organisms and maximise microbial detection. Sampling targeted at specific pathogens or spoilage organisms that are thought to play a major role in the safety or quality of the product, is undertaken to verify the performance of the sanitation programme designed for their control. The use of TVC is applicable to both low and high risk processing areas though detection of specific pathogens should be undertaken in high risk only. Microbiological assessments have also been used to ensure compliance with external microbial standards, as a basis for cleaning operatives’ bonus payments, in hygiene inspection and troubleshooting exercises, and to optimise sanitation procedures. Traditional microbiological techniques appropriate for food factory use involve the removal or sampling of micro-organisms from surfaces, and their culture using standard agar plating methods; these have been reviewed by Holah et al. (2003). Micro-organisms may be sampled via sterile cotton or alginate swabs and sponges (swabbing), after which the micro-organisms are re-suspended by vortex mixing or dissolution into suitable recovery or transport media, or via water rinses for larger enclosed areas (e.g. fillers). Representative dilutions are then incubated in a range of microbial growth media, depending on which micro-organisms are being selected for, and incubated for 24–48 hours. Alternatively, micro-organisms may be sampled directly onto self-prepared or commercial (‘dip slides’) agar contact plates. The choice of sampling site will relate to risk assessment Where there is the potential for micro-organisms remaining after (poor) cleaning and disinfection to contaminate product over a period of time by direct contact, these sources would require sampling much more frequently than other sites. Other sites which may be more contaminated may present lower risks of contaminating product. For example, it is more sensible, and gives more confidence, to sample the points of the equipment that directly contact the product and that are difficult to clean than to sample non direct contact surfaces, e.g. underneath of the equipment framework. In relation to micro-organism numbers, it is difficult to suggest what is an ‘acceptable’ number of micro-organisms remaining on a surface after cleaning and disinfection as this is clearly dependent on the food product, process, ‘risk area’ and degree of sanitation undertaken. There is also no European legislation that describes a microbiologically clean surface. A number of figures have been quoted in the past (as TVC per square decimeter) including 100 (Favero et al., 1984), 540 (Thorpe and Barker, 1987) and 1000 (Timperley and Lawson, 1980) for dairies, canneries and general manufacturing, respectively. The results in Table 12.1 show that in chilled food production, sanitation programmes should achieve levels of around 1000 micro-organisms per swab, which on flat surfaces approximately equates to a square decimeter. Expressing counts arithmetically is always a problem, however, as single counts taken in areas where cleaning has been inadequate (which may be in excess of 108 per swab) produce an artificially high © 2008, Woodhead Publishing Limited

Cleaning and disinfection of chilled food plants and equipment

329

mean count, even over thousands of samples. It is better, therefore, to express counts as log to the base 10, a technique that places less emphasis on a relatively few high counts, and Table 12.1 shows that mean log counts of approximately 1 should be obtained. Because of the difficulty in setting external standards, it is best to set internal standards as a measurement of what can be achieved by a given sanitation programme. A typical approach would be to assess the level of micro-organisms, ATP or protein present on a surface after a series of 10 or so sanitation programmes in which the sanitation programme is carefully controlled (i.e. detergent and disinfectant concentrations are correct, contact times are adhered to, water temperatures are checked, pressure hoses are set to specified pressures, sanitation schedules are followed, etc.). The mean result is the best that can be achieved when the cleaning programme is optimised and undertaken correctly. This is not likely to be repeatable on every occasion and the target may be set as the mean result plus ‘a little bit for day-to-day error’. For example, if a mean ATP value of 100 RLU has been obtained, a target value for subsequent cleans could be set at 150 RLU. In some cases one target may not be appropriate and it may be necessary to set different targets for different areas of the process lines. For example, it may be appropriate to set an ATP target of 150 RLU for conveyor belt surfaces, which can be particularly difficult to clean, and a target of 100 RLU for all other surface material types. In some instances, cleaning may result in ATP or protein levels that do not differ significantly from controls, i.e. residues following cleaning are not detected. In this case these techniques are not appropriate and microbiological standards have to be adopted. As more information is obtained from routine cleaning programme verification, the targets can then be reviewed and the food manufacturer can seek to reduce them as appropriate by changing the manufacturing environment, process, equipment or cleaning programme. A review of the target standard would also be required if any of the variables controlling the target were changed; primarily the food product or process or the sanitation programme The verification of freedom of allergenic residues after sanitation is becoming an issue in parts of the chilled food industry. A number of test kits are now available that can be used or modified for the detection of allergen residues on surfaces, including the detection of peanut, hazelnut, almond, sesame, soya, egg and egg white, milk (casein, β-lactoglobulin), crustacean, wheat gluten and sulfite. Allergen detection kits can be sourced from a number of companies including: Tepnel: www.tepnel.com R-Biopharm: www.r-biopharm.com Neogen: www.neogen.com ELISA Systems: www.elisas.com.au HAVen: www.hallmarkav.com Many food manufacturers have adopted the philosophy of validating their allergen cleaning programme using the appropriate specific allergen test kit and then verifying the performance of the cleaning and disinfection programme on a day-to© 2008, Woodhead Publishing Limited

330

Chilled foods

day basis with an alternative hygiene monitor, e.g. ATP. For this to be an acceptable practice, the validation of the cleaning programme must be undertaken using both the allergen test kit and the hygiene monitor. In this case the validation seeks to demonstrate freedom of allergenic residues after cleaning and a positive relationship between freedom from allergens and freedom from the target for the hygiene monitor (ATP, protein, etc.). At this stage it is not possible to state that freedom from allergenic residues as demonstrated by sample results lower than the detection limit of the test kit equates to freedom from an allergenic effect in the final consumer from product produced from such surfaces. In the absence of any other information, however, allergen cleanliness as determined by results lower than the detection limit of the test kit should be the target of the sanitation programme. As part of the assessment of sanitation programmes, it is worthwhile looking at how the programme is performing over a defined time period (weekly, monthly, quarterly, etc.) as individual sample results are an estimate only of what is happening at one specific time period. This may be to ensure that the programme remains within control, to reduce the variation within the programme or, as should be encouraged, to try to improve the programme’s performance. An assessment of the performance of the programme with time, or trend analysis, can be undertaken simply, by producing a graphical representation of the results on a time basis, or can be undertaken from a statistical perspective using statistical process control (SPC) techniques as described by Harris and Richardson (1996). Generally, graphical representation is the most widely used approach, though SPC techniques should be encouraged for more rigorous assessment of improvement in the programme’s performance.

12.7 Sanitation management Senior management must take full responsibility for the successful operation of the sanitation programme; ultimately, failures in the programme generally reflect poor management. For the majority of chilled food processing operations, the following is a guide to the responsibilities of senior management: (i)

Understand the objectives of cleaning and sanitation for a particular product and the hazards presented by the product design. (ii) Always seek to improve hygiene standards in line with the high risk philosophies adopted in Chapter 11. Hygiene has traditionally not had the same research support as other areas of importance in food manufacture and is thus a new and developing science. It is only relatively recently that new concepts have been developed, based on scientific assessments, and management must be flexible enough to try out and to encourage such concepts when they emerge. (iii) Lead by example by being both always properly attired in food production areas and (occasionally) present in production areas when sanitation is being undertaken (usually in the early hours of the morning!). © 2008, Woodhead Publishing Limited

Cleaning and disinfection of chilled food plants and equipment

331

(iv) Provide the required equipment (including maintenance), the staffing levels and the time to undertake the sanitation programme effectively. Cleaning operatives must be a dedicated labour pool whose priority is sanitation (i.e. not production). Similarly, operatives should not join the cleaning team as an ‘introduction to production’. (v) Management should be capable of giving praise when sanitation is undertaken correctly, as well as discipline when it is not. In companies where bonus systems have been employed based on microbiological assessments of equipment after cleaning, results have indicated that hygiene has generally been improved and bonuses are rarely missed. (vi) Appoint or nominate a manager to be responsible for the day-to-day implementation of the sanitation programme. The manager who assumes responsibility for the sanitation programme must have technical hygiene expertise and has a range of job functions including the following: (i) (ii)

The selection of a suitable chemical supplier. The selection of sanitation chemicals, equipment and methods (see Sections 12.3, 12.4). (iii) The training of cleaning operatives. (iv) The development of cleaning schedules. (v) The implementation of sanitation programme monitoring systems (see Section 12.6). (vi) The representation of hygiene issues to senior management. It is essential to have strong links with a chemical supplier who should be able to do much more than simply supply detergents and disinfectants. They should be chosen on their abilities to undertake site hygiene audits, supply suitable chemical dosing and application equipment, undertake operative training and help with the development of cleaning schedules (by trials) and sanitation monitoring and verification systems. Good chemical companies respond quickly to their customer needs, periodically review their customers’ requirements and visit during sanitation periods to ensure that their products are being used properly and are working satisfactorily. The level of expected service from the chemical supplier should be agreed between the chemical supplier and food manufacturer, and be documented as a specification, as for any other product or service purchased by the food manufacturer. The cleaning manager may also need to visit the chemical supplier’s site to audit their quality systems. Whilst, in theory, systems and/or chemicals could seem appropriate for the required task, every factory, with its water supply, food products, equipment, materials of construction and layout, etc. is unique. All sanitation chemicals, equipment and methods must, therefore, be proven in the processing environment. New products and equipment are always being produced and a good working relationship with hygiene suppliers is beneficial. Only disinfectants that have been approved to the relevant European Standards (Section 12.3) should be used. © 2008, Woodhead Publishing Limited

332

Chilled foods

The cleaning operative’s job is technical and potentially hazardous, and all steps should be undertaken to ensure that sufficient training is given. By the nature of the job, training needs to be comprehensive and should include:

• • • • •

a knowledge of basic food hygiene the importance of maintaining low/high risk barriers during cleaning the implications to product safety/spoilage by poor sanitation practices an understanding of the basic function and use of sanitation chemicals and equipment, and of their sequence of operation a thorough knowledge of the safe handling of chemicals and their application, and the safe use of sanitation equipment.

The components of a management system for a sanitation programme are described in Middleton and Holah (2008) and are shown schematically in Fig. 12.4. The cleaning plan should list all the cleaning and disinfection tasks that need to be undertaken across the food manufacturing site and their frequency (daily, weekly, monthly, etc.). It is primarily used to describe and record all of the cleaning tasks that need to be undertaken each day. The plan incorporates the cleaning of all manufacturing areas including processing equipment and utensils, ancillary equipment, services, cleaning equipment, personal equipment (e.g. boots and chain mail gloves) and the environment. In some cases that plan may include cleaning of outside areas of the site, laundering of air socks and laundering of staff clothing. Cleaning schedules are the written work instructions that detail precisely how the cleaning and disinfection procedures for each task should be undertaken. The cleaning schedule can also be used as the work instruction against which cleaning operatives can be formally trained. In addition, cleaning schedules can be used to describe the cleaning and disinfection process to interested parties (e.g. internal and external auditors) who may be visiting the premises during a time period in which cleaning and disinfection procedures are not taking place. The ‘whole room’ cleaning schedule focuses on two key parameters. The first details all the requirements for the practical management of the cleaning and

Crisis management plan

Cleaning plan

Whole room cleaning schedule

In-production cleaning schedule

Fig. 12.4

End-ofproduction cleaning h d l

Periodic cleaning schedule

Site decontamination schedule

Schematic contents of a documented management system to undertake cleaning and disinfection tasks.

© 2008, Woodhead Publishing Limited

Cleaning and disinfection of chilled food plants and equipment

333

disinfection operation and includes manpower, any specialist engineering support, equipment, chemicals and their dosing, health and safety, room preparation, protection of any food production operations, how cleaned surfaces are protected from recontamination, how the room is prepared for subsequent production and how the cleaning equipment itself is cleaned and maintained. The second details requirements to maximise the removal of micro-organisms from the processing area (rather than simply redistributing them) and to leave the food contact surfaces as free of micro-organisms as possible for subsequent food production. This encompasses both the sanitation sequence (Section 12.5) and detailed instructions on each individual cleaning operative’s task and how they are going to be coordinated with their colleagues to ensure the production area is cleaned as a unit. During production periods, ‘clean as you go’ operations should always be undertaken as they encourage hygienic work practices, help prevent slip and trip accidents and may be motivational to food production operatives. The nature of the cleaning methods used, however, may be very different to those used at the end of production. Consideration should be given to how such cleaning could crosscontaminate other processing lines that may be in production. The use of screening to minimise cross-contamination may be appropriate. Restriction of the use of water is encouraged wherever possible to prevent microbial growth on production surfaces. Any major spillages during production should always be thoroughly cleaned to prevent the growth of micro-organisms on processing surfaces, which may become a significant challenge to the subsequent end-of-production sanitation programme. End-of-production cleaning schedules can be written for all equipment in a room (particularly if the equipment is simple and does not need dismantling), for individual pieces of equipment (particularly if the equipment is complex and requires extensive isolation and dismantling), ancillary equipment, services, items of personnel protective equipment and the processing environment. The following are typical features that may be contained in schedules:

• name of the processing area or item of equipment, etc. to be cleaned • number of cleaning operatives required; for some specific equipment this may • • • • • •

include named individuals who have been specifically trained with respect to the dismantling, safety or reassembly of the equipment list of all personal protective equipment (PPE) equipment required requirement for equipment isolation and dismantling prior to cleaning, which may require additional operatives or specialist engineering support list of cleaning equipment required list of cleaning chemicals to be used and their in-use concentrations, contact times and temperatures detailed work instruction describing the method and scheduling of chemical and rinse application including guidance timings following cleaning and/or disinfection, identification of key inspection points used to monitor programme success.

The need for periodic cleaning practices has been mentioned earlier in the chapter © 2008, Woodhead Publishing Limited

334

Chilled foods

(Section 12.2). Periodic cleaning schedules can be developed to combine periodic cleaning tasks on a range of equipment (typically a ‘weekend’ cleaning schedule) or can be added to the ‘end-of-production’ cleaning schedules if they are detailed for individual items of equipment. A site decontamination plan may be associated with the company crisis management plan, detailing a planned decontamination of the processing areas (usually the high care or high risk zone) following a potential pathogen contamination incident, to allow production to re-start as appropriate. Decontamination of the processing area describes cleaning and disinfection practices beyond end-ofproduction and periodic cleaning and typically involves removing and disposing of all food and packaging materials, further (total) dismantling of routine equipment, dismantling of ancillary equipment (e.g. evaporative condensers, air socks, air supply ductwork) and the wider use of (usually stronger) disinfectants. When new equipment is purchased or processing areas are designed or refurbished, insufficient attention is usually placed on sanitation requirements. Equipment or areas of poor hygienic design will be more expensive to clean (and maintain) and may not be capable of being cleaned to an acceptable standard in the time available. If improperly cleaned, adequate disinfection is impossible and thus contamination will not be controlled. Hygiene management must be strongly represented, thus ensuring that hygiene requirements are considered alongside those of engineering, production and accounts, etc. Three types of sanitation programmes can be implemented by management and each has its advantages and disadvantages: at the end of production, production operatives clean their workstations and then (a) they form a cleaning crew and undertake the sanitation programme; (b) a separate, dedicated cleaning gang complete the sanitation programme; or (c) cleaning and disinfection is undertaken by contract cleaners. Whilst each solution will place different demands on the food manufacturer, the principles as managed above should always be incorporated and the sanitation programme effectively managed.

12.8 Conclusion Within a chilled food factory, cleaning and disinfection can range from the cosmetic, through good housekeeping to maintain food operative health and safety and to control food quality in low risk areas, to the major pathogen and thus food safety control within high risk. It can thus stretch from a relatively low level of management concern to a critical issue in controlling product safety. Due resource should thus be given to its effective management. But is it effective? Viewed in terms of controlling pathogens and spoilage micro-organisms in the final product and thus maintaining shelf-life and product safety, then yes it is. This is confirmed by the very few reported recall or food safety incidences that the chilled food industry enjoys. But does it eliminate all microorganisms within the food processing environment? Clearly, the evidence of persistent strains of pathogens residing in manufacturing areas for many years © 2008, Woodhead Publishing Limited

Cleaning and disinfection of chilled food plants and equipment

335

shows this not to be the case. This is a potential threat as if there were to be a lapse in control of the sanitation system, persistent organisms could quickly multiply in the environment and cause product contamination issues. One goal for the future is thus to further our sanitation control beyond just the food processing surfaces to effectively control the entire processing environment. This is becoming known as ‘wholeroom’ disinfection. It is likely to involve additional cleaning and disinfection methods and may encompass the use of gaseous ozone or hydrogen peroxide, the use of ultraviolet light, possibly in synergistic combination with titanium dioxide in coated surfaces or the wider use of proven antimicrobial surfaces. As the hygiene demands from customers inevitably increase, food manufacturers will look to constantly developing sanitation systems to help them deliver enhanced hygiene control.

12.9 References ANON.

(1964) Assessment of odour from packaging material used for foodstuffs. British Standard 3744. British Standards Institute, London. ANON. (1983a) Sensory analysis of food. Part 5: Triangular test. ISO 4120. International Standards Organisation. ANON. (1983b) Testing of container materials and containers for food products. DIN 10955. Deutsches Institut für Normung e.V., Berlin ANON. (1973) The effectiveness of water blast cleaning in the food industry. Food Technology in New Zealand 8, 15 and 21. ANON. (1991) Sanitation. In: Shapton, D. A. and N. F. (eds) Principles and Practices for the Safe Processing of Foods. Butterworth-Heinemann Ltd, Oxford, pp 117–199. ANON. (1998a) EN 1650:1998 Chemical disinfectants and antiseptics – Quantitative suspension test for the evaluation of fungicidal activity of chemical disinfectants and antiseptics used in food, industrial, domestic, and institutional areas – Test method and requirements (phase 2, step 1). ANON. (1998b) A practical guide to the disinfection of food processing factories and equipment using fogging. Ministry of Agriculture Fisheries and Food, 650, St. Catherine Street, London, SE1 0UD. ANON. (1997a) EN 1276:1997 Chemical disinfectants and antiseptics – Quantitative suspension test for the evaluation of bactericidal activity of chemical disinfectants and antiseptics used in food, industrial, domestic, and institutional areas – Test method and requirements (phase 2, step 1). ANON. (1997b) EN 1499:1977 Chemical disinfectants and antiseptics – Hygienic handwash – Test method and requirements (phase 2, step 2). ANON. (1997c) EN 1500:1997 Chemical disinfectants and antiseptics – Hygienic handrub – Test method and requirements (phase 2, step 2). ANON. (2001) EN 13697:2001 Chemical disinfectants and antiseptics – Quantitative nonporous surface test for the evaluation of bactericidal and/or fungicidal activity of chemical disinfectants and antiseptics used in food, industrial, domestic, and institutional areas – Test method and requirements (phase 2, step 2). AASE B., SUNDHEIM G., LANGSRUD S. AND RØRVICK L.M. (2000) Occurrence of and a possible mechanism for resistance to a quaternary compound in Listeria monocytogenes. International Journal of Food Microbiology, 62, 57–63. ANDERSON M.E., HUFF H.E. AND MARSHALL R.T. (1985) Removal of animal fat from food grade belting as affected by pressure and temperature of sprayed water. Journal of Food Protection, 44, 246–248. © 2008, Woodhead Publishing Limited

336

Chilled foods

ANDRADE N.J., BRIDGEMAN T.A. AND ZOTTOLA E.A.

(1998) Bacteriocidal activity of sanitizers against Enterococcus faecium attached to stainless steel as determined by plate count and impedance methods. Journal of Food Protection, 661, 833–838. BAUTISTA D.A., MCINTYRE L., LALEYE L. AND GRIFITHS M.W. (1992) The application of ATP bioluminescence for the assessment of milk quality and factory hygiene. Journal Rapid Methods Autom Microbiology, 1, 179–193. BELL C., STALLARD P.A., BROWN S.E. AND STANDLEY J.T.E. (1994) ATP-Bioluminescence techniques for assessing the hygienic condition of milk transport tankers. International Dairy Journal, 4, 629–640. BERGMAN B-O. AND TRAGARDH C. (1990) An approach to study and model the hydrodynamic cleaning effect. Journal of Food Process Engineering, 13, 135–154. BLENKINSOPP S.A. AND COSTERTON J.W. (1991) Understanding bacterial biofilms. Trends in Biotechnology, 9, 138–143. BRETT M.S.Y., SHORT P. AND MCLAUCHLIN J. (1998) A small outbreak of listeriosis associated with smoked mussels. International Journal of Food Microbiology, 43, 223–229. CARPENTIER B. AND CERF O. (1993) Biofilms and their consequences, with particular reference to the food industry. Journal of Applied Bacteriology, 75, 499–511. CORRIEU G. (1981) State-of-the-art of cleaning surfaces. In: Proceedings of Fundamentals and Application of Surface Phenomena Associated with Fouling and Cleaning in Food Processing. April 6–9, Lund University, Sweden, pp 90–114. DAS J.R., BHAKOO M., JONES M.V. AND GILBERT P. (1998) Changes in biocide susceptibility of Staphylococcus epidermidis and Escherichia coli cells associated with rapid attachment to plastic surfaces. Journal of Applied Microbiology, 84, 852–858. DESTRO M.T., LEITAO M.F.F. AND FARBER J.M. (1996) Use of molecular typing methods to trace the dissemination of Listeria monocytogenes in a shrimp processing plant. Applied and Environmental Microbiology, 62, 705–711. DHALIWAL D.S., CORDIER J.L. AND COX L.J. (1992) Impedimetric evaluation of the efficacy of disinfectants against biofilms. Letters in Applied Microbiology, 15, 217–221. DUNSMORE D.G. (1981) Bacteriological control of food equipment surfaces by cleaning systems. 1. Detergent effects. Journal of Food Protection, 44, 15–20. DUNSMORE D.G., TWOMEY A. WHITTLESTONE,W.G. AND MORGAN H.W. (1981) Design and performance of systems for cleaning product contact surfaces of food equipment: A review. Journal of Food Production, 44, 220–240. ELLIOT R.P. (1980) Cleaning and sanitation. In: Katsuayama, A. M. (ed.) Principles of food processing sanitation. The Food Processors Institute, USA, pp 91–129. FAVERO M.S., GABIS D.A. AND VESELEY D. (1984) Environmental monitoring procedures. In: Speck, M. L. (ed.) Compendium of Methods for the Microbiological Examination of Foods. American Public Health Association, Washington, DC. FONNESBECH VOGEL B., HUSS H.H., OJENITI B., AHRENS P. AND GRAM L. (2001) Elucidation of Listeria monocytogenes contamination routes in cold-smoked salmon processing plants detected by DNA-based typing methods. Applied and Environmental Microbiology 67, 2586–2595. FRANK J.F. AND KOFFI R.A. (1990) Surface-adherent growth of Listeria monocytogenes is associated with increased resistance to surfactant sanitizers and heat. Journal of Food Protection, 53, 550–554. FRYER P.J. AND CHRISTIAN G.K. (2005) Improving the cleaning of heat exchangers. In: Handbook of Hygiene Control in the Food Industry. Eds. H.L.M. Lelieveld, M.A. Mostert and J.T. Holah, Woodhead Publishing, Cambridge, UK. pp 468–496. GIBSON H., TAYLOR J.H., HALL K.E. AND HOLAH J.T. (1995) Biofilms and their detection in the food industry. CCFRA R&D Report No.1, CCFRA, Chipping Campden, Glos., GL55 6LD, UK. GIBSON H., TAYLOR J.H., HALL K.E. AND HOLAH J.T. (1999) Effectiveness of cleaning techniques used in the food industry in terms of the removal of bacterial biofilms. Journal of Applied Microbiology, 87, 41–48. © 2008, Woodhead Publishing Limited

Cleaning and disinfection of chilled food plants and equipment

337

GRIFFITH J., BLUCHER A., FLERI J. AND FIELDING L. (1994) An evaluation of luminometry as

a technique in food microbiology and a comparison of six commercially available luminometers. Food Science and Technology Today, 8, 209–216. GRIFFITH C.J., DAVIDSON C.A., PETERS A.C. AND FIELDING L.M. (1997) Towards a strategic cleaning assessment programme: Hygiene monitoring and ATP luminometry, an options appraisal. Food Science and Technology Today, 11, 15–24. HARRIS C.S.M. AND RICHARDSON P.S. (1996) Review of principles, techniques and benefits of Statistical Process Control. Review No.4, Campden and Chorleywood Food Research Association, Chipping Campden, UK. HAWRONSKYJ J-M. AND HOLAH J.T. (1997) ATP: A universal hygiene monitor. Trends in Food Science and Technology, 8, 79–84. HAYES P.R. (1985) Cleaning and Disinfection: Methods. In: Food Microbiology and Hygiene, Elsevier Applied Science Publishers, Barking, pp 268–305. HOLAH J.T. (1991) Food surface sanitation. In: Hui, Y. H. (ed.) Encyclopedia of Food Science and Technology. John Wiley and Sons, New York. HOLAH J.T. (1995) Special needs for disinfectants in food-handling establishments. Revue Scientific et Technique Office International des Épizooties, 14, 95–104. HOLAH J.T. (1998) Guidelines for effective environmental microbiological sampling. Guideline No. 20, Campden and Chorleywood Food Research Association, Chipping Campden, UK. HOLAH J.T., BIRD J AND HALL K.E. (2004) The microbial ecology of high risk, chilled food factories; Evidence for persistent Listeria spp. and Escherichia coli strains. Journal of Applied Microbiology, 97, 68–77. HOLAH J.T., BIRD J AND HALL K.E. (2003) Manual of hygiene methods for the food and drink industry. Guideline No. 45, Campden and Chorleywood Food Research Association, Chipping Campden, UK. HOLAH J.T., HIGGS C., ROBINSON S., WORTHINGTON D., AND SPENCELEY H. (1990a) A conductance based surface disinfection test for food hygiene. Letters in Applied Microbiology, 11, 255–259. HOLAH J.T. AND KEARNEY L.R. (1992) Introduction to biofilms in the food industry. In: Biofiilms – Science and Technology, Eds L.F. Melo, T.R. Bott, M. Fletcher, B. Capdeville, pp 35–45. Dordrecht: Kluwer. HOLAH J.T., LAVAUD A., PETERS W. AND DYE K.A. (1998) Future techniques for disinfectant testing. International Biodeterioration and Biodegradation, 41, 273–279. HOLAH J.T., TAYLOR J.H., DAWSON D.J. AND HALL K.E. (2002) Biocide usage in the food industry and the disinfectant resistance of persistent strains of Listeria monocytogenes and Escherichia coli. Journal of Applied Microbiology, Symposium Supplement, 92, 111S– 120S. HOLAH J.T., TIMPERLEY A.W. AND HOLDER J.S. (1990b) The spread of Listeria by cleaning systems. Technical Memorandum No. 590. Campden Food and Drink Research Association, Chipping Campden, UK. HOLAH J.T. AND THORPE R.H. (1990) Cleanability in relation to bacterial retention on unused and abraded domestic sink materials. Journal of Applied Bacteriology, 69, 599–608. HUGO W.B., PALLENT L.J., GRANT D.J.W., DENYER S.P. AND DAVIES A. (1985) Factors contributing to the survival of Pseudomonas cepacia in chlorhexidine. Letters in Applied Microbiology, 2, 37–42. HUGO W.B. AND RUSSELL A.D. (1999) Types of antimicrobial agent. In Principles and Practices of Disinfection, Preservation and Sterilization. Eds A.D. Russell, W.B. Hugo and G.A.J. Ayliffe, Blackwell Sciences, Oxford, pp 5–95. ICMFS (1980) Cleaning, disinfection and hygiene. In: Microbial Ecology of Foods, Volume 1, Factors Affecting Life and Death of Micro-organisms. The International Commission on Microbiological Specifications for Foods, Academic Press, London, pp 232–258. ICMFS (1988) Cleaning and disinfecting. In: Micro-organisms in Foods, Volume 4, Application of the Hazard Analysis Critical Control Point (HACCP) System to Ensure © 2008, Woodhead Publishing Limited

338

Chilled foods

Microbiological Safety and Quality. The International Commission on Microbiological Specifications for Foods, Blackwell Scientific, Oxford, pp 93–116. JENNINGS W.G. (1965) Theory and practice of hard-surface cleaning. Advances in Food Research, 14, 325–459. JOHANSSON T., RANTALA L., PALMU L. AND HONKANEN-BUZALSKI T. (1999) Occurrence and typing of Listeria monocytogenes strains in retail vacuum-packed fish products and in a production plant. International Journal of Food Microbiology, 47, 111–119. KATHARIOU S. (2003) Listeria monocytogenes virulence and pathogenicity, a food safety perspective. Journal of Food Protection, 65, 1811–1829. KLEMPERER R. (1982) Tests for disinfectants: Principles and problems. In: Disinfectants: Their Assessment and Industrial Use. Scientific Symposia Ltd, London. KNAUER -KRAETZL B . (1994) Reinigung und Desinfektion. 12. Fleischwarenforum: Fleischhygiene, BEHR’s Seminar, Darmstadt, 14–15 April 1994. KOOPAL L.K. (1985) Physico-chernical aspects of hard-surface cleaning. Netherlands Milk and Dairy Journal, 39, 127–154. KUMAR C.G. AND ANAND S.K. (1998) Significance of microbial biofilms in the food industry: A review. International Journal of Food Microbiology, 42, 9–27. LAMBERT R.J.W. (2001) Advances in disinfection testing and modelling, Journal of Applied Microbiology, 91, 351–363. LAWRENCE L.M. AND GILMORE A. (1995) Characterisation of Listeria monocytogenes isolated from poultry products and poultry-processing environment by random amplification of polymorphic DNA and multilocus enzyme electrophoresis. Applied and Environmental Microbiology, 61, 2139–2144. LECHEVALIER M.W., CAWTHORN C.D. AND LEE R.G. (1988) Inactivation of biofilm bacteria. Applied and Environmental Microbiology, 44, 972–987. LEE S-L. AND FRANK J.F. (1991) Effect of growth temperature and media on inactivation of Listeria monocytogenes by chlorine. Journal of Food Safety, 11, 65–71. LONCIN M. (1977) Modelling in cleaning, disinfection and rinsing. In: Proceedings of mathematical modelling in food processing, 7–9 September, Lund. Lund Institute of Technology. LUNDÉN J.M., MIETTINEN M.K., AUTIO T.J. AND KORKEALA H.J. (2000) Persistent Listeria monocytogenes strains show enhanced adherence to food contact surface after short contact times. Journal of Food Protection, 63, 1204–1207. LUNDÉN J., AUTIO T., MARKKULA A., HELLSTRÖM S. AND KORKEALA H. (2003a) Adaptive and cross-adaptive responses to persistent and non-persistent Listeria monocytogenes strains to disinfectants. International Journal of Food Microbiology, 82, 265–272. LUNDÉN, J., AUTIO, T., SJÖBERG, A-M. AND KORKEALA, H. (2003b) Persistent and nonpersistent Listeria monocytogenes contamination in meat and poultry processing plants. Journal of Food Protection, 66, 2062–2069. MABESA R. C., CASTILLO M. M., CONTRERAS E. A., BANNAAD L. AND BANDIAN V. (1982) Destruction and removal of micro-organisms from food equipment and utensil surfaces by detergents. The Philippine Journal of Science, 3, 17–22. MARRIOT, N.G. (1985) Sanitation equipment and systems. In: Principles of Food Sanitation, AVI Publishing Co., Westport, CT, pp 117–151. MATTILA-SANDHOLM T. AND WIRTANEN G. (1992) Biofilm formation in the food industry: A review. Food Reviews International, 8, 573–603. MCLAUCHLIN J., GREENWOOD M.H. AND PINI P.N. (1990) The occurrence of Listeria monocytogenes in cheese from a manufacturer associated with a case of listeriosis. International Journal of Food Microbiology 10, 255–262. MCLAUCHLIN, J., HALL S.M., VELANI S.K. AND GILBERT R.J. (1991) Human listeriosis and paté: A possible association. British Medical Journal, 303, 773–775. METTLER E. AND CARPENTIER B. (1998) Variations over time of microbial load and physicochemical properties of floor materials after cleaning in food industry premises. Journal of Food Protection, 61, 57–65. © 2008, Woodhead Publishing Limited

Cleaning and disinfection of chilled food plants and equipment

339

MIDDLEMISS N.E., NUNES C.A., SORENSEN J.E. AND PAQUETTE G. (1985) Effect of a water rinse

and a detergent wash on milkfat and milk protein soils. Journal of Food Protection, 48, 257–260. MIDDLETON K.E. AND HOLAH J.T. (2008) A practical guide to cleaning and disinfection in food factories. Guideline No. 55, Campden and Chorleywood Food Research Association, Chipping Campden, UK. MIETTINEN M.K., BJÖRKROTH K. AND KORKEALA H.J. (1999) Characterisation of Listeria monocytogenes from an ice cream plant by serotyping and pulsed gel electrophoresis. International Journal of Food Microbiology, 46, 187–192. MROZEK H. (1982) Development trends with disinfection in the food industry. DeutscheMolkerei-Zeitung, 12, 348–352. NESBAKKEN T., KAPPERUD G. AND CAUGANT D.A. (1996) Pathways of Listeria monocytogenes contamination in the meat processing industry. International Journal of Food Microbiology, 31, 161–171. OFFILER M.T. (1990) Open plant cleaning: Equipment and methods. In: Proceedings of ‘Hygiene for the 90s’, November 7–8, Campden Food and Drink Research Association, Chipping Campden, Glos., UK, pp 55–63. PONTEFRACT R.D. (1991) Bacterial adherence: Its consequences in food processing. Canadian Institute of Food Science and Technology Journal, 24, 113–117. POULIS J.A., DE PIJPER M., MOSSEL D.A.A. AND DEKKERS P.P.H.A. (1993). Assessment of cleaning and disinfection in the food industry with the rapid ATP-bioluminescence technique combined with tissue fluid contamination test and a conventional microbiological method. International Journal of Food Microbiology, 20, 109–116. POURSHABAN M., GIANFRANCESCHI M., GATTUSO A., MENCONI F. AND AURELI P. (2000) Identification of Listeria monocytogenes contamination sources in two fresh sauce production plants by pulsed-field gel electrophoresis. Food Microbiology, 17, 393–400. REYBROUCK G. (1998) The testing of disinfectants. International Biodeterioration and Biodegradation, 41, 269–272. RIDGEWAY H.F. AND OLSEN B.H. (1982) Chlorine resistance patterns of bacteria from two drinking water distribution systems. Applied and Environmental Microbiolog, 44, 972– 987. RONNER A.B. AND WONG A.C.L. (1993) Biofilm development and sanitizer inactivation of Listeria monocytogenes and Salmonella typhimurium on stainless steel and Buna-N rubber. Journal of Food Protection, 56, 750–758. RØRVIK L.M., CAUGANT D.A. AND YNDESTAD M. (1995) Contamination pattern of Listeria monocytogenes and other Listeria spp. in a salmon slaughterhouse and smoked salmon processing plant. International Journal of Food Microbiology, 25, 19–27. RUSSELL A.D., HUGO W.B. AND AYLIFFE G.A J. (1982) Principles and Practice of Disinfection, Preservation and Sterilization. Blackwell Scientific Publications, London. SCHLUSSLER H.J. (1975) Zur Kinetic von Reinigungsvorgangen an festen Oberflachen. Symposium über Reinigen und Desinfizieren lebensmittelverarbeitender Anlagen. Karlsruhe, 1975. SCHMIDT U. AND CREMLING K. (1981) Cleaning and disinfection processes. IV. Effects of cleaning and other measures on surface bacterial flora. Fleischwirtschaft, 61, 1202–1207. SHUPE W.L., BAILEY J.S., WHITEHEAD W.K. AND THOMPSON J.E. (1982) Cleaning poultry fat from stainless steel flat plates. Transactions of the American Society of Agricultural Engineers, 25, 1446–1449. TAYLOR J.H., ROGERS S.J. AND HOLAH J.T. (1999) A comparison of the bactericidal efficacy of 18 disinfectants used in the food industry against Escherichia coli 0157:H7 and Pseudomonas aeruginosa at 10 °C and 20 °C. Journal of Applied Microbiology, 87, 718– 726. THORPE R.H. AND BARKER P.M. (1987) Hygienic design of liquid handling equipment for the food industry. Technical Manual No. 17. Campden Food and Drink Research Association, Chipping Campden, UK. © 2008, Woodhead Publishing Limited

340

Chilled foods

TIMPERLEY D.A. AND LAWSON G.B. (1980) Test rigs for evaluation of hygiene in plant design.

In: Jowitt, R. (ed.) Hygienic Design and Operation of Food Plant. Ellis Horwood, Chichester. UNNERSTAD H., BANNERMAN E., BILLE J., DANIELSSON-THAM M.-L., WAAK E. AND THAM W. (1996) Prolonged contamination of a dairy with Listeria monocytogenes. Netherlands Milk and Dairy Journal, 50, 493–499. WALKER R.L., JENSEN L.H. KINDE H., ALEXANDER A.V. AND OWENS L.S. (1991) Environmental survey for Listeria species in frozen milk product plants in California. Journal of Food Protection, 54, 178–182. WIRTANEN G. AND MATTILA-SANDHOLM T. (1993) Epifluorescence image analysis and cultivation of foodborne bacteria grown on stainless steel surfaces. Journal of Food Protection, 56, 678–683. WIRTANEN G. AND MATTILA-SANDHOLM T. (1994) Measurement of biofilm of Pediococcus pentosacceus and Pseudomonas fragi on stainless steel surfaces. Colloids and Surfaces B: Biointerfaces, 2, 33–39. WRIGHT J.B., RUSESKA I. AND COSTERTON J.W. (1991) Decreased biocide susceptibility of adherent Legionella pneumophila. Journal of Applied Microbiology, 71, 531–538. ZOTTOLA E.A. AND SASAHARA K.C. (1994) Microbial biofilms in the food processing environment – should they be a concern? International Journal of Food Microbiology, 23, 125–148.

© 2008, Woodhead Publishing Limited

13 Operation of plants manufacturing chilled foods M. Brown, mhb Consulting, UK

13.1 Introduction Successful manufacturers of chilled foods use all the raw materials, process technologies and packaging systems at their disposal to achieve attractive, safe, high quality products at a competitive cost. The basic supply chain elements – plan, source, make and deliver – are used to do this and meet customer and consumer needs for either retail or food service packs. Because chilled foods have a relatively short shelf-life, assurance of product safety and quality has to rely on each stage of the supply chain exerting effective control and monitoring to show that specifications have been met. Information for product release has to be available by the time products are ready for dispatch; basing product release on analysis of finished product, which may take a day or two, will reduce available shelf-life. For the supply chain to work effectively, all contributors need to appreciate that they are directly responsible for meeting specifications at their stage. Depending on the product design, the microbiological quality of the ingredients, any processing and decontamination steps; factory hygiene, and consumer use, finished products can either be:

• ready-to-eat (RTE) and eaten cold, or heated by consumers to warm them to a temperature suitable for eating. This heating is often insufficient to ensure microbiological safety and hence these products have to be free of hazardous micro-organisms, or freed from them during the manufacturing process; or © 2008, Woodhead Publishing Limited

342

Chilled foods

• designed for cooking, where heating by the consumer according to the ‘onpack’ instructions cooks the product and pasteurizes it. In this case, the supply chain needs to ensure that microbiological contamination remains within the levels agreed for the product design. Raw meat products (e.g. minced meat, meat preparations and poultry meat products) intended to be cooked must be clearly labelled by the manufacturer, informing consumers of the need for thorough cooking before consumption. Besides their manufacturing plant and marketing organization, all food-producers have a network of suppliers (e.g. for materials, facilities, equipment and logistics) and customers. This network exerts just as much control over product quality, safety and cost as in-house activities. Good control of the network is facilitated by functional links, or partnerships, exchanging critical information, and coordinating product designs and production schedules. Used well, these links allow quality and cost to be optimized, and new products to be easily produced. Software is increasingly used to support information exchange between remote supply chain operations. It is also used to track demand for products, and then integrate the ordering and supply of raw materials with production planning and logistics (e.g. identifying the quantities, remaining shelf-life and where raw materials, product intermediates and finished products are in the supply chain). Some supply chains are predominantly vertically integrated, so that raw materials move from primary production to the final product at a single location (e.g. poultry products). More commonly, there will be different sources of material from suppliers operating independently at different locations, and finished products will be processed and assembled at another location (e.g. ready meals). Products within a range may also be made in facilities shared with other manufacturers (e.g. co-packing or outsourcing). Complex products, such as ready meals or sandwiches, can be made from ingredients from different origins or specialist suppliers that may be shared with other producers. Products can be distributed to local or national markets by multi-stage chilled distribution systems, which are lengthening to become region-wide (e.g. Europe) as brands and logistics chains grow. There are some regulatory requirements, including microbiological criteria, for chilled ready-to-eat foods products (see 2073/2005 EC, Chapter 1: 1.1 and 1.2 and CFA, 2006a). These products are categorized according to their preservation system and the risks of supporting growth of Listeria monocytogenes. For very mildly preserved products that support rapid growth of L. monocytogenes (e.g. it can grow to numbers exceeding 100 cfu/g during the shelf-life) L. monocytogenes must be undetectable in 25 g of product, at exit from the factory. This requires tight specifications, especially for heating and preservation, and very high hygiene levels during production. Within the hazard analysis and critical control point (HACCP) framework, this requirement creates CCPs for hygiene and decontamination, and a need to monitor the microbiological quality of finished batches (e.g. by microbiological testing) to verify that precautions are working. © 2008, Woodhead Publishing Limited

Operation of plants manufacturing chilled foods

343

For preserved products with a shelf-life greater than 5 days (e.g. pre-packed sliced cooked meat, smoked salmon and soft cheese), the manufacturer must be able to show that with low initial levels of contamination and the designed preservation system, numbers of L. monocytogenes will not exceed 100/g during the product’s shelf-life. After the initial validation has been done, operating limits are fixed for the various stages of production, and these are monitored to show satisfactory performance by raw material suppliers, manufacturing and logistics. The suitability of a product design to limit growth of L. monocytogenes can be validated by challenge testing or predictive modelling of growth. The parameters used in such tests should realistically describe the preservation system (e.g. storage temperatures, pH, aw, concentration of preservatives and packaging system).

13.1.1 The manufacturing process As a starting point, raw materials have to be of a suitable quality to ensure they are within the capabilities of the process and that products will meet customer and consumer expectations after processing, packaging and storage. Manufacturing usually starts with ingredient reception and approval, then batching, preparation and kitchen treatment to give the product its character (e.g. dicing, sautéing or stewing), and sometimes preservation (e.g. salting, acidification or pickling). Next there are decontamination (e.g. heating or pasteurization or washing of ingredients or part-processed product) and cooling steps. Lastly dosing or filling places a controlled amount of product, or a piece such as a pie or pizza, into a tray, pouch or pack which is sealed under conditions of controlled hygiene. Filling and sealing can be done before (e.g. in-pack pasteurized), after (e.g. open assembly of pre-processed ingredients) or as part of (e.g. hot fill) the decontamination step. Sealed packs may have a headspace of air or a specified gas mixture (modified atmosphere of N2, O2 and CO2) or be under vacuum. All these process steps may have manual or automated control systems. Where process operations are manual, all the personnel involved need to understand the importance of process control (e.g. minimum temperatures or separation of processed and unprocessed materials) and high levels of hygiene. Final cooling, handling and distribution of the product are always under chilled conditions (0–8 °C). After storage and distribution, the consumer, or ‘chef’ in a food service operation, prepares the food for consumption. Preparation may include in-pack reheating in a conventional or microwave oven, unpacking and portioning the product and conventional (e.g. simmering, roasting or frying) or microwave heating. Some chilled products may be served and eaten cold without re-heating, and such products must always be in the RTE category. Very simply, all supply chain operations should ensure product quality and safety meet product identity and consumer expectations, and strike the correct balance between quality, safety and conversion cost.

© 2008, Woodhead Publishing Limited

344

Chilled foods

13.2 Supply chain structure and operation 13.2.1 Background Chilled products are made by enterprises with a range of capacities, technical capabilities and knowledge. Although their supply chains may look different, all provide, directly or indirectly, the materials and services used for production and distribution of finished product. A typical supply chain can be visualized as four elements – location and layout, production, stock and logistics – that provide the basic functions:

• Plan – Allocate, manage and monitor the resources needed to meet customer demand for products.

• Source – Choose suppliers and manage the supply of materials to manufacturing.

• Make – Manufacture product, including assurance of quality and meeting specifications and cost targets.

• Deliver – Provide logistics and storage of materials and finished product. • Return – Manage withdrawal of unsafe food from the market, based on traceability (178/2002/EC) and for chilled products, oversee the disposal of obsolete or excess product. These functions manage many different types of hazards (e.g. quality and microbiological) using different controls, procedures and facilities. The stringency of requirements will usually be set by hazards involving the raw materials, consumer use of the product and positioning of the brand. Most stringent ones will apply to unpreserved ready-to-eat products and the least stringent to products that will be cooked prior to consumption; premium brands will often demand higher quality standards. The extent of facilities and skills available to do this are usually related to company size and also willingness to employ outside expertise. Controls in any supply chain will depend on the product design and the hazards from raw materials, processing and factory hygiene, e.g. microbiological, physical or allergens. These should be identified by a formal hazard analysis and controlled by implementing the HACCP plan. In parallel, operational decisions will focus on production scheduling to optimize utilization of equipment and to meet volume demands within the constraints imposed by HACCP, hygiene and maintenance requirements and quality assurance measures required by labelling, customers (who may provide specifications) and legislation. However they are organized, food business operators remain responsible for the safe design and manufacture of their products. In the EU, hygiene legislation (EC 852/2004) imposes principles and some minimum standards, recognizing that food manufacture is a complex network of different facilities and processes using materials with different hazards. Sustainability Increasingly, it is necessary to include not only economic, but also social and environmental considerations in the structure and operation of the supply chain to ensure sustainability (IEMA, 2006; Johr, 2006). Sustainability relies on the © 2008, Woodhead Publishing Limited

Operation of plants manufacturing chilled foods

345

company imposing environmental, social and economic policies to ensure that the needs of the present consumers and customers are met without compromising the ability of the supply chain to meet similar needs in the future. Companies often use a supplier code of conduct to do this. The main sustainability risks for chilled foods are use of non-sustainable raw materials, excessive packaging, unnecessarily energy intensive production and storage conditions, long logistics routes and adverse local impact from company dominance leading to erosion of competition and elimination of local suppliers. For the chilled food industry, a major challenge is sustainable production and distribution of products with a short shelf-life and potentially high rates of wastage. Supply chain management has to encompass all these features, especially as retailers are aware that their reputation is in the public eye and is, in part, derived from consumer perception of the products they sell. Traceability Within complex supply chains, traceability of materials and process conditions can be difficult, but is essential to ensure that consumers are protected, legal requirements are met, (minimum one stage up and one stage down) and potential losses to a business are minimized (see http://europa.eu.int/comm/food/food/foodlaw/traceability/index_en.htm). If there is a problem, accurate identification and isolation of the batches at risk (e.g. materials and product) will minimize losses and speed-up problem solving. Traceability can be made more difficult if products include materials from different batches or locations, part-processed materials or re-work, or if the process period in question included stoppages or breakdowns. Traceability of materials, process conditions and products can rely on automated data logging and control systems (e.g. Siematic – http://www.automation. siemens.co.uk/offsiteBar.asp?Url=http://www.siemens.com/simatic or ABB SattLine – http://www.abb.com/product/seitp334/e3d927acb567331bc12571 e100335b3e.aspx). These provide automatic controls and data capture, and their information can be used to provide insight into process performance at a particular time to facilitate traceability and problem solving. For further functionality they can be interfaced with the overall supply chain management software (e.g. SAP – http://www.sap.com/industries/consumer/businessprocesses/food/demand planning.epx or MRP II {Material Requirement Planning} – http://www.lilly software.com/software_solution/manufacturing/material_requirement_ planning_ MRP.asp). System integration can allow process conditions and material movements to be reconciled with particular process periods or events (e.g. process deviations or breakdowns). Batches of material or individual packs can be identified using bar codes based on the UPC (universal product code – which is a 10 to 12 digit optical, machine readable numeric product or material code). The first digits of the bar code usually represent the manufacturer, the remaining ones the product. radio frequency identification (RFID) labels or tags may also be used to do this. They can automatically provide data for control systems by acting as transponders. RFID labels can contain up to 1 Kbyte of information; some are re-writeable, so that new information can be inserted (e.g. revised expiry data of a product), but their cost is © 2008, Woodhead Publishing Limited

346

Chilled foods

higher than a printed bar code label. They are a big step forward in materials and product identification compared to a bar code, but rapid uptake is likely to be limited by cost and the need to develop industry standards and build infrastructures to collect and interpret information.

13.2.2 Quality management systems Quality management systems operate using two major elements:

• the safety/quality team (e.g. production, quality management/assurance, micro•

biology and other laboratory-based functions, integrated with process and product development, engineering and maintenance and regulatory), data and systems (e.g. automated, manual and laboratory controls, data collection, pre-requisites and HACCP).

Consistent product quality relies on all the staff supporting the quality management system and ensuring that meeting cost and capacity targets does not erode quality. To maximize product shelf-life, assurance of quality has to be based on real-time controls (e.g. process and hygiene monitoring) and pre-approval of ingredients. Any time taken after production for testing (as part of the normal product release procedure) will reduce the available shelf-life. Product testing should be designed for monitoring and trend analysis, and verifying the performance of the factory quality systems including pre-requisites and HACCP (see CFA, 2006b). Recently, an ISO standard for food safety management system (FSMS) has been developed (ISO 2200). It defines generic food safety requirements for all parts of the supply chain and is intended to provide a link between HACCP and the formal ISO quality management systems (ISO 9000: see http://www.iso.org/iso/ en/iso9000-14000/addresources/articles/pdf/tool_5-04.pdf. A quality manual should link the factory systems and operational information (e.g. plan, source, make, deliver and return, see Table 13.1) to business and marketing plans and customer/consumer expectations. It should include: (i) (ii) (iii) (iv)

the organizational structure, responsibilities and authorities of key personnel, flow diagrams of processes, specifications for raw materials and process intermediates, functional descriptions of each software programme controlling and operating equipment (software should be backed-up with validated copies held securely), (v) clear guidance on limits, suitable resources and the infrastructure needed for the product range, (vi) formal procedures for dealing with non-conforming or potentially unsafe materials, process conditions or products to ensure that risks to consumers are minimized and recurrence prevented, (vii) interfacing completely with process, cleaning and maintenance procedures, training and documents (e.g. work instructions, specifications, requirements for records, corrective actions and permits-to-work), and © 2008, Woodhead Publishing Limited

Operation of plants manufacturing chilled foods

347

Table 13.1 Elements of a quality manual Supply chain structure and function

Organizational structure, responsibilities and authorities of key personnel. Key contact information.

Introduction of new or altered products to production

Trial procedures and conditions for hand-over from R&D to production.

Routine production

Flow diagrams of processes. HACCP plans and details of pre-requisites, including monitoring by QA and production. Working documents for raw materials, process intermediates and equipment operation. Functional descriptions of each software program controlling and operating equipment (software should be backed-up with validated copies held securely).

Specifications

Clear guidance on materials and limits, suitable resources and controls and the QA, production and equipment infrastructure needed for the product range. Record of agreement of specifications with suppliers.

Non-conforming or potentially unsafe materials

Formal procedures for dealing with nonconforming or potentially unsafe • materials • process conditions • products • practices • personnel to ensure that risks to consumers are minimized and recurrence prevented. Procedures for malicious threats should be included.

Interfacing of factory systems and procedures

Integration of process, cleaning and maintenance procedures, training and documents (e.g. work instructions, specifications, requirements for records, corrective actions and permits-to-work).

Review

Review procedure for validity and compatibility with the HACCP plan. Means of ensuring it is up-to-date (e.g. triggers for review) and out-of-date documents are withdrawn.

(viii) it should be reviewed for validity to ensure it is up-to-date and that out-ofdate documents are withdrawn. Supply chain performance is most effectively verified by reviewing records from equipment and control systems, laboratory analyses and by auditing. The review should cover product quality, CCPs and pre-requisites. If quality assurance (QA) and process records are easily accessible (e.g. in an ISO 9000 type system), verification can focus on them, rather than on additional testing, and auditing. To be effective, verification requires reviewers and auditors with minimum levels of © 2008, Woodhead Publishing Limited

348

Chilled foods

knowledge about the products, materials and processes. Product failures will significantly increase factory and regulatory costs; for example if replacing or resetting equipment or re-training a workforce is needed to solve the problem. To support management, regulatory visits should include not only enforcement, but also promote continuous improvement. ISO management systems Principles for supply chain and factory systems to assure quality are explained in the ISO 9000 standard Quality Management System. ISO 9000:2005 – covers general quality system requirements, based on management responsibility, management of resources, product realization and ongoing measurement and analysis of performance with accompanying actions or improvements. ISO 9001:2000 – gives requirements for more complex quality systems in companies that design and manufacture products and may also provide a service. It is the standard most often used for approval or certification audits by external assessors. ISO 9004:2000 – a guide to continual improvement. For full ISO 9000 accreditation, external auditing at fixed time intervals is needed and ISO 19011 provides guidelines for auditing quality management and environmental management systems. Within the ISO system there must be procedures for document control, to ensure the issuing of new versions (e.g. specifications, recipes and contracts) and withdrawal of old ones. The compulsory elements of the ISO standards match HACCP requirements very closely: operational procedures, document and record controls, corrective actions, control of non-conforming product and verification of the effectiveness of the whole system by internal auditing. Many companies operate strictly to ISO 9000 because it can help them to be recognized as reliable business partners. With interest in sustainability growing, ISO 14000: Environmental management, has become more important. This ISO standard provides a framework for environmental management, including labelling and product life-cycle management and the supporting audit programme.

13.2.3 Specifications and suppliers Specifications (and contracts) should cover the supply chain and especially the high risk and high value aspects of raw and packaging materials, processes and products (see Table 13.2). High risks include the presence of pathogens and spoilage micro-organisms, deterioration of the material, chemical contaminants, critical conditions (e.g. process, storage and hygiene), key dates (e.g. use by), legal requirements and traceability. High value characteristics include appearance, functional properties, colour and composition. Producers, suppliers and customers should formally agree specifications and contractual conditions before supply starts. © 2008, Woodhead Publishing Limited

Operation of plants manufacturing chilled foods

349

Table 13.2 Contents of a specification

• • • • • • • • • • • •

The supplier’s name and location. General description of the product or material (e.g. diced vegetable). The function of the material (e.g. thickener, etc.). Relevant physical or microbiological parameters (e.g. salmonella absent in 25 g). Use-by date and lot coding to allow identification of batches for traceability into the supplier’s systems. Specific requirements (e.g. process conditions, ingredients or storage). For suppliers with multiple manufacturing sites, the manufacturing site or line may be included. Packaging to be used and unit weight (e.g. moisture-resistant,15 kg). Supply cost. Production, batch or delivery quantities and times, one-off, batch or bulk production. Delivery and storage conditions. Legal and nutritional requirements. Sustainability and environmental requirements.

The most reliable suppliers are ones who:

• are known to meet hygiene and process control requirements, • have a history of supplying material meeting quality, price, and delivery time criteria, and • are able to supply information with each batch showing that specifications are met. They are least likely to generate the problems with the materials they provide. Most companies use QA and production to regularly review supplier performance by auditing and testing; this usually leads to suppliers becoming ‘approved suppliers’. The frequency of testing materials at reception will depend on the supplier’s reliability and quality system (e.g. agreed analytical or process data or certificates of analysis), the risk level and origin of the material and the product to be made. If raw materials are consistently within specification, less testing is required and low level monitoring is used only to verify information provided by the supplier. Hence the use of approved suppliers reduces indirect factory costs because the burden on factory QA and the analytical laboratory is reduced since testing each batch prior to acceptance into production is not required. If a supplier does not have a history of reliability, testing will be required to release material for production and to progress the approval process. Once good performance has been established, levels of testing can be progressively reduced to the level required for monitoring and trend analysis. However, if monitoring shows that materials are out of specification, whether this is shown by the supplier’s data, or not, then batches should not be ‘automatically’ presumed cleared until supplier reliability is re-established. Finished product specifications should represent the product design adapted for the raw materials available, equipment capability, shelf-life and product use. Increasingly they need to take account of healthiness, sustainability and packaging material life-cycles. These additional requirements, in turn, influence the operation of the supply chain and all suppliers now need to accept the wider scope of specifications and their rationale. © 2008, Woodhead Publishing Limited

350

Chilled foods

Writing specifications for similar raw materials from different origins (e.g. prawns from either the North Atlantic or tropical seas), requires extra care because microbiological hazards that are unrealistic for some raw materials, or production areas where controls are in place, may be realistic where such controls are not in place or ineffective (e.g. infectious pathogens may be present after processing or harvest) or growing/harvesting conditions are unknown. Therefore review, possibly leading to a new hazard analysis, including the buying department and research and development (R&D), forms an integral part of developing and maintaining specifications and controls. The scope of specifications needs to ensure that material arrives in a suitable condition; hence logistics suppliers should be covered by specifications which detail temperatures, times, acceptance criteria and hygiene requirements (targets and limits) at key stages of the logistics chain. If material suppliers use third party transport, then separate specifications should make responsibilities and conditions clear. To minimize losses and disputes, responsibilities, authority and conditions for blocking or rejecting product should always be clear in specifications, and logistics suppliers should be able to carry out necessary checks. Microbiological specifications Based on the hazard analysis and product design, specifications for raw materials and products should cover realistic microbiological hazards, (micro-organisms or toxins and harmful metabolites – see Table 13.3). These will depend on the raw material and intended use of end product. Specifications should indicate analytical methods, sampling plans and limits. Specification should only cover realistic pathogens (see Table 13.3), micro-organisms indicating poor hygiene (indicator Table 13.3 Realistic microbiological hazards Bacteria

Parasites

Viruses

Bacillus cereus Brucella spp. Campylobacter spp. Clostridium botulinum (toxigenic) Clostridium perfringens Escherichia coli (enterotoxigenic) Escherichia coli O157:H7 Escherichia coli non-O157 STEC Escherichia coli (other diarrheogenic) Listeria monocytogenes Salmonella (nontyphoidal) Salmonella typhi Shigella spp. Staphylococcus aureus (toxigenic) Streptococcus spp. Vibrio cholerae (toxigenic) Vibrio vulnificus Vibrio other Yersinia enterocolitica

Cryptosporidium parvum Cyclospora cayetanensis Giardia lamblia Toxoplasma gondii Trichinella spiralis

Norovirus Rotavirus Astrovirus

© 2008, Woodhead Publishing Limited

Operation of plants manufacturing chilled foods

351

micro-organisms such as Enterobacteriaceae) and those causing spoilage or challenging the decontamination processes (e.g. cold-growing, heat-resistant spores). Accessible and reliable analytical methods do not exist for all pathogens and those that are not detectable by reliable analytical methods should not be included in specifications, as their presence or level cannot be assessed. However, it is essential to confirm that the product design or processing will prevent such agents reaching consumers (e.g. enterotoxigenic Escherichia coli). For some materials (e.g. fresh-cut vegetables and herbs, intended to be eaten without further decontamination) specifications should extend to the means of minimizing or preventing microbial contamination during growing, harvesting and post harvesting activities (e.g. washing). For materials that will be heated during factory preparation, less stringent specifications and limits may be satisfactory, if the heat treatment can ensure their absence after heating. Sampling plans can be based on either attributes (e.g. sampling to measure an attribute above or below a limit or criterion – present or absent), or variables (e.g. quantifying a continuous variable such as a microbiological count). Plans for variables require more samples, but can be used to estimate the chances of a batch meeting a target value with a predictable certainty (e.g. 90% certain that no more than 5% of the batch is outside the target value). 13.2.4 Hazard analysis and HACCP Hazard analysis and material usage Reliable supply chain operation relies on correct operation against specifications. Raw materials may be classified from low to high risk to define the products, procedures and processes they are suitable for.

• Low risk materials (e.g. previously heat treated or extracted) rarely contain

•

•

pathogens and require minimal levels of analysis, as contamination is not expected. They may be used in products made in good manufacturing practice (GMP) or high care areas (HCAs). Medium risk materials may be contaminated with infectious (e.g. salmonella in pasteurised egg) or toxigenic pathogens (e.g. Staphylococcus aureus) or high levels of spores or chemical or physical contaminants. Risks are most effectively minimized by the use of approved suppliers with implemented HACCP and QA programmes that specifically target these hazards. Regular auditing and review of process records should be used to verify supplier performance and analysis at reception should be done according to a defined sampling plan, until suppliers have been recognized as reliable. They can be used for products made in GMP areas and designed for consumer cooking, or used in hygienic areas after heating or decontamination. High risk materials (e.g. meat, poultry, fish and shellfish) have a history of contamination with hazardous micro-organisms or other contaminants (e.g. heat-resistant spores in dry or milled spices), even if they come from approved suppliers. Therefore, quality assurance and rejection procedures are needed,

© 2008, Woodhead Publishing Limited

352

Chilled foods

and they are usually used in products where a validated heat treatment is part of the product design. HACCP The Hygiene of Foodstuffs Regulation (852/2004) requires business operators to put in place, implement and maintain permanent control and monitoring procedures spanning the supply chain and based on HACCP principles, taking account of the Codex Alimentarius. Therefore they need to identify realistic hazards, controls and corrective actions (e.g. critical control points) for each product or group of products. Critical limits should be used to separate acceptable from unacceptable performance, or quality, at each CCP or key process stage. Monitoring and verification procedures should produce records. When any change is made to the product, process or materials used, the HACCP plan should be reviewed and, if necessary, changed. Regulators view the HACCP system as a means to help food businesses achieve high standards of food safety; not as a method of self-regulation to replace official controls. The EC Regulation ‘Microbiological Criteria for Foodstuffs’ (2073/2005) requires food business operators to use HACCP to ensure that their products meet microbiological criteria by using correct process and materials. Monitoring is required to show that all important process stages meet the criteria and that the supply chain functions hygienically. If process monitoring cannot be done, there will be insufficient data to show that the HACCP plan is effective. Pre-requisite programmes Pre-requisites are the foundations of the HACCP system. Chilled food manufacture requires specialized pre-requisites which are not specific to only one process or product. These provide systems, equipment and facilities to control the hazards (especially microbiological and allergen) associated with particular technologies (e.g. cook–chill) and products (e.g. RTE). Pre-requisites range from the general hygiene and system requirements applicable to all food businesses (Regulation EC 852/2004), often referred to as basic GMP to the more stringent requirements for HCA and high risk areas (HRAs) (see CFA, 2006). Requirements for specific products or lines are then controlled and monitored as CCPs; without prerequisites in place, these cannot assure product safety.

13.3 Building location and layout Within any building, the choice of layout (e.g. GMP, HCA or HRA) and facilities starts with the product concept and the raw materials (e.g. ambient–stable, chilled or frozen) and technology to be used (e.g. cook–chill, in-pack pasteurization, or product assembly after cooking). The facilities and equipment needed are an expensive and long-term commitment for any enterprise, and choices may limit or encourage innovation. Unless there are new buildings, the layout has to be contained within the shape and size of existing buildings. As a minimum, buildings need to provide safe, hygienic and cost effective facilities for manufacturing and storage (see CAC/RCP 1-1969, Rev. 4-2003: http://www.fao.org/docrep/005/ © 2008, Woodhead Publishing Limited

Operation of plants manufacturing chilled foods

353

Y1579E/Y1579E00.HTM and 852/2004: EC). A layout should:

• facilitate movement of personnel and materials, without microbial and allergen

• • •

•

contamination (e.g. forward flow of material, one-way flow of suitable materials into high hygiene areas and mininization of cross-overs to prevent contamination (e.g. between allergenic materials or from uncooked to cooked product), minimize the risks of key process stages being omitted (e.g. product heating), ensure hygiene levels and procedures are compatible with the type of product being made (e.g. cleaning and disinfection, prevention of air-borne contamination and adequate space for hygienic operation and product segregation), site equipment and services (such as distribution pipe-work and heating, ventilation and air conditioning (HVAC) systems) to allow operation, maintenance/repair and calibration with a minimum of disruption and risk of contamination, and ensure segregation of allergen-containing and non-allergenic materials by routing of air handling systems (e.g. when powders are present) and using mixproof valves in lines processing liquids. Products containing allergens may be made at the end of a production schedule, immediately before cleaning, to minimize the chances of carry-over. There is a trend to use dedicated areas and lines for processing allergenic ingredients to prevent accidental contamination.

13.3.1 Manufacturing areas The whole range of chilled foods cannot be made in manufacturing areas that meet only basic GMP requirements. For many types, additional precautions and facilities are needed to ensure product quality, microbiological safety and shelf-life. HCAs or HRAs, separated from the low risk (or GMP) areas, should be used for handling, assembling and packing decontaminated materials. Typical design features of high hygiene areas are segregated changing rooms, air locks for entry and exit, dedicated chills and two-door ovens for entry and exit of materials. There should be separated changing rooms and entry routes for employees working in areas with different levels of hygiene, and hand-washing facilities should be sited at all personnel entrances. GMP areas Storage chills and freezers and batching areas generally are designated as GMP areas. If high or medium risk materials are handled or stored in these areas, there should be effective separation from low risk materials. Some chilled product can be made in GMP areas (e.g. raw products intended for cooking, and raw, preserved, ready-to-eat products, such as salads). As a principle, hygiene during storage and preparation should not increase product safety risks or levels of contamination. If raw materials processed in these areas are not decontaminated before being made into products, they should be low risk (e.g. salad vegetables), and the product preservation system (e.g. vinegar-based) should inhibit microbial growth. If raw products are made, consumer cooking processes (giving heat © 2008, Woodhead Publishing Limited

354

Chilled foods

treatments exceeding 70 °C for 2 minutes) should be specified to eliminate infectious pathogens. Higher hygiene areas Higher hygiene manufacturing areas are separated from GMP areas and can be either HCAs or HRAs. Table 13.4 shows what can be made in each type of area. These areas share common structural features to minimize or eliminate the chances of product contamination from raw materials, food contact surfaces, air or personnel. Very reliable utilities (e.g. steam, cooling media and air supply) are essential for their safe operation, and work procedures should cover personnel and material entry and actions to be taken in the event of utility failure (e.g. temperature rise or loss of air over-pressure). Within them, precautions are specifically designed to prevent Listeria spp. becoming endemic (e.g. walls and floors are sealed or impermeable, to ensure cleaning is effective). Floors are sloped to assist drainage and drying, with drains always flowing away from them. To minimize chances of harbouring contaminants, services (steam, cooling media, compressed air and electricity) should be routed with the minimum of exposed cabling or piping. Where piping is sufficiently cold to cause condensation to form, it should be hygienically insulated with a cleanable material. Microbiological monitoring of these areas may include visual inspection (against a check list) and swabbing of key equipment or areas for Listeria spp., which is a suitable indicator for possible presence of L. monocytogenes. If Listeria spp. are detected, an increased risk is indicated and the type found should be characterized and preventive measures reviewed. The air supply (HVAC) systems provide refreshment air and control temperature in GMP and hygienic areas. Their design (especially capacity) operation, set temperature, and rates of air replacement (> 6 changes/h) should maintain environmental temperatures under envisaged operating conditions. System capacity should provide sufficient overpressure (> 5 Pa) to ensure an outward air flow from hygienic areas during production, and prevent the inward flow of potentially contaminated air (e.g. when doors are opened). Hygiene procedures and maintenance should keep air distribution ducts, filters and heat exchangers operating within specification and as clean and dry as possible to minimize the chances of airborne microbiological contamination. The air flow during and after equipment and area cleaning should prevent contamination and ensure rapid drying. During cleaning in one area, air should not be circulated from it to adjacent areas and it should be exhausted directly to the outside. The HVAC installation should be capable of drying the cleaned area in a reasonable time (e.g. 2 hours). Air refreshment rates will determine the time needed to bring the area back to its design hygiene level after the air system has been turned off. Air can be distributed using either rigid or suspended fabric ducting, but increasingly the latter is used as it is more easily cleaned and disinfected. The air supply should be pre-filtered (to remove insects and dust) and then downstream filtered at higher filtration efficiency (e.g. semi HEPA or H11 – see CCFRA, 2005). System hygiene should be monitored using microbial and particle counts. © 2008, Woodhead Publishing Limited

Operation of plants manufacturing chilled foods

355

Table 13.4 What can be made safely in the types of area Type of area

Type of product

Microbiological Decontamination safety target treatments

GMP

Raw, nonInfectious preserved foods pathogens

Risk level*

Mild pasteurization High by consumer or customer

Typical product chilled shelf-life 1 week

Medium ≤ 2 weeks to low

HCA or Raw preserved Infectious dedicated foods or pathogens GMP mixtures of raw and cooked ingredients

Washing + disinfection or action of preservation system

HCA

Infectious pathogens

Mild pasteurisation V. low

≤ 2 weeks

Pathogenic spore-formers

Severe pasteurization

> 2 weeks

Infectious pathogens

Mild pasteurization Low

≤ 2 weeks

Pathogenic spore-formers

Severe pasteurization

> 2 weeks

HRA

In-pack pasteurized by manufacturer

Ingredients pasteurized by manufacturer product assembly in open area

V. low

Low

*Risk level = chances of product containing hazardous micro-organisms at exit from factory Mild pasteurization = Heat treatment equivalent to 70 °C × 2 minutes Severe pasteurization = Heat treatment equivalent to 90 °C × 10 minutes GMP = Area meeting basic food hygiene standards HCA = High care area designed and operated to minimize contamination HRA = High risk area designed and operated to prevent contamination

Air exposure plates (settle plates) and battery-powered centrifugal samplers are widely used to monitor the microbiological quality of air. It is good practice to use a range of sampling points as well as different times of sampling. High care areas HCAs are necessary to manufacture foods that are designed as RTE and unpreserved (e.g. prepared ready meals). These areas should minimize the risk of contaminating the products, which already contain very low levels of micro-organisms, and they should only handle materials that have been pre-decontaminated (e.g. pasteurized or chemically decontaminated raw materials). To minimize the chance of contamination, materials and products pasteurized in-pack should be unloaded after cooling into areas that have hygiene standards at least equivalent to the poststerilization area of a cannery. High risk areas HRAs are laid out and operated to prevent microbial (re-)contamination, especially © 2008, Woodhead Publishing Limited

356

Chilled foods

with pathogens. They handle unpreserved products composed entirely of pasteurized ingredients, and provide facilities for their manual or automated assembly in an ‘open’ environment. Hence they have stringent precautions against re-contamination, including limited personnel entry from segregated changing rooms. When personnel leave the HRA after work or for a break, work clothing and footware should be removed and remain in the changing room to prevent its contamination.

13.4 Production Marketing will usually determine the product range, and R&D with supplier management will often decide where and how products will be manufactured. As a starting point, it should be established that all stages of the supply chain are capable of producing the product or intermediate materials in the qualities and quantities required. This consideration will lead to the allocation of suppliers, equipment and products to manufacturing plants (e.g. in-house or outsourced) and the accompanying development of onward logistics to distribution centres and customer markets. Chilled food production usually carries high direct and indirect production costs, and control of these affects profitability. Costs can be direct (attributable to a particular product, e.g. production workers manually assembling product) or indirect (not attributable to any particular product, including the infrastructure for manufacturing, e.g. environmental controls, hygiene and quality assurance staff and training). To control production costs, each cost needs to be allocated to its origin (e.g. materials, products, procedures or equipment). To minimize costs, materials must work well with the equipment available. Preventable costs attributable to material variability causing slow machine operation or stoppages should be eliminated; therefore, specifications should include aspects of materials critical to manufacturing. It is often necessary to balance supplier capability and material cost against increased manufacturing costs or loss of capacity or quality. If cost reduction is not done properly, contamination risks may be increased; therefore any proposals for changes should be examined using a hazard analysis procedure. Many chilled factories have frequent product changes and use flexible manufacturing systems that can easily change equipment configuration or pack size. Many factories still rely heavily on manual processing for quick changes. Efficient equipment can make changes (e.g. fill volume) automatically and has easily fitted change parts. The design for manufacturing approach (Ulrich and Eppinger, 2000) can be used to optimize manufacturing operations by matching customer, marketing and manufacturing needs from the beginning of the equipment selection process.

13.4.1 Transfer from development to production Production, QA and R&D departments are jointly responsible for ensuring that the product design is correctly translated into process specifications and work © 2008, Woodhead Publishing Limited

Operation of plants manufacturing chilled foods

357

instructions, taking account of capabilities at each process step and ingredient variability. The suitability of new suppliers, equipment, or materials is most effectively established by development trials. Any constraints imposed by, for example, raw materials, equipment capability and cost/quality targets, should be accounted for during developmen,t and in some cases, extension to a product range may have to be restricted if proposed products are beyond production capability. When the suitability of each process stage has been confirmed, specifications and work instructions based on trial results should be in place before production starts. To ensure a smooth transfer from development to production, scale-up should be done incrementally (e.g. kitchen concept, bench-scale, pilot plant and finally fullscale production), with appropriate testing at each stage to establish that the resulting product meets sensory, performance and regulatory needs.

13.4.2 Production planning The supply chain needs to be able to provide the specified quantity and quality of product for delivery to customers at the planned time. Guided by shelf-life and stock demands (stock is material or product held to cushion against supply failures and variability in demand), factory systems should control the movement of raw materials and process intermediates so that production schedules run without delays; this may be done using software such as SAP or MRP II. The combination of manufacturing capacity and stock levels should be capable of meeting changes in demand (e.g. from changes in the weather), and the costs of higher production capacity (e.g. more equipment) and stock (e.g. finished or part-processed product) have to be balanced. If overproduction for stock is used to compensate for low capacity, materials can deteriorate before they are used and products may become obsolete before they are required for sale. If plant is over-sized, it will provide available, but unused, capacity and may perform unpredictably. If it is undersized, there is an increased risk of outof-stock, which leads to low service levels and can lead to loss of brand value and market share. To decide on the best way for a factory to meet fluctuating demand, assessment of plant capacity under operating conditions is important.

• Total line capacity is based on: • theoretical equipment performance • time available for production (e.g. the shift pattern in the plant – 8, 16 or 24 hours/day for 3, 5 or 7 days/week).

• Total capacity is reduced to available capacity by stoppages: • routine stoppages – e.g. change-overs, cleaning, maintenance or size changes.

• unexpected stoppages – e.g. breakdowns and shortages of materials: • Short production periods between stoppages dramatically reduce available capacity and increase costs. © 2008, Woodhead Publishing Limited

358

Chilled foods

• For maximum available capacity, run lengths should be as long as possible without increasing the risks of out-of-date product. Whenever production procedures (e.g. extending production periods between cleans) or production sequences are altered with a view to increasing available capacity, a hazard analysis should be used to identify any new hazards (e.g. allergen cross-contamination or excessive microbial growth in process equipment) and the HACCP plan reviewed. 13.4.3 Managing production and storage areas The major challenge to successfully managing the hygiene of manufacturing areas is controlling the movement of materials, tools, containers and personnel on a dayto-day basis. The entry of food materials and packaging to each area should be limited to materials that have been pre-treated correctly and the quantities required. Containers of materials and part-processed product should enter hygienic areas only when they have been through an effective cleaning and decontamination process or protective secondary packaging has been removed. For ease of management, tools and equipment should be ‘captive’ in designated areas and cleaned, disinfected and maintained there. Equally well-controlled systems should deal with the removal of materials from these areas. To prevent personnel working in these areas acting as a source of contamination, especially with infectious pathogens, they should be health-screened and trained to respect high standards of and procedures for personal hygiene and provided with the right tools, facilities and time to complete their tasks effectively. They should wear only clothing that is dedicated to the areas where they work. Changing into clean, correctly-designed work clothing should be done in designated changing rooms. Work clothing should be changed when soiled, or at least daily. It should be laundered by laundries with approved practices and hygiene standards. More detail on specific requirements can be found in ‘Best Practice Guidelines for the Production of Chilled Food’ (CFA, 2006b). Gloves should be disposable and used only in conjunction with a hand hygiene regime able to ensure that skin remains healthy. Unless they meet the same standards of hygiene, maintenance staff should work in these areas only when all food materials have been put away and production has stopped. Their work should be completed before the equipment has its final clean and disinfection before production re-starts. Health screening Management procedures should ensure that all staff in contact with food will not act as sources of contamination with infectious pathogens (e.g. salmonella). Staff who handle unwrapped and decontaminated raw materials present the highest risks. Guidance on managing this risk is given (for the UK) in Guidance ‘Food Handlers: Fitness to Work’, issued by the Department of Health (see CFA, 2006b). Food handlers recruited for working in HRAs should have a pre-employment screening using a questionnaire. Routine health screening (e.g. with a questionnaire) should extend to all personnel (e.g. laboratory staff, maintenance personnel, © 2008, Woodhead Publishing Limited

Operation of plants manufacturing chilled foods

by examining how well lines work and identifying the causes of losses (e.g. material, defective product and downtime or efficiency losses)

for easy maintenance

359

for specified maintenance tasks, problem solving, repairs and improvement tasks

Improve equipment effectiveness

Allocate responsibility between operational and maintenance staff

Select equipment Train all staff Promote condition-based maintenance

Analyse failures using root cause analysis build-up a systematic approach to all maintenance acitvities with clear allocation of responsibilities

Fig. 13.1

in relevant maintenance skills with continuous review of performance

use output to move towards zero breakdown maintenance

Aims of TPM.

contractors and visitors) who visit production areas. Where it is suspected that a prospective (or existing) employee may act as a (potential) source of food-borne pathogens, advice should be sought from a medical practitioner about further investigation. Total productive maintenance Hygienic areas run well when a total productive maintenance (TPM) programme for equipment and area maintenance is used. This approach uses small groups of employees working in an area to carry out routine equipment inspection, maintenance and elimination of breakdowns and defects (Nakajima, 1988). Their experience is used to improve reliability and minimize unplanned stoppages. TPM aims to establish good maintenance practice through five goals (Fig. 13.1) based on production staff taking ownership of the facilities and equipment, with maintenance staff providing support.

13.5 Process stages 13.5.1 Raw material reception and storage Procedures at reception need to ensure that incoming raw materials and transport conditions meet specifications. Inspection should at least confirm that quantity, © 2008, Woodhead Publishing Limited

360

Chilled foods

quality and temperature of chilled or frozen materials have not changed since leaving the supplier’s plant and that packs have not been damaged. Quality assurance personnel should be trained to do this but may need specific additional training to assess particular raw materials (e.g. meat and fish). Leaking or damaged containers should be rejected to prevent out-of-specification material reaching processes. Any rejected materials must be clearly marked and blocked in the process management software system, and stored in a segregated designated area until disposal. After reception, raw materials must be handled and prepared to minimize quality loss, cross-contamination and microbial growth. For packaging materials, monitoring is usually restricted to visual controls to confirm quantity and dimensions, and detect damage to containers or pallets.

13.5.2 Handling and preparing ingredients and in-process material Many ingredients of chilled foods are fresh and have a short shelf-life, therefore temperature and time control during storage, handling, transport and production is critical. This requires suitably sized temperature-controlled facilities with provision for temperature/time monitoring and recording. Materials should be coded and used strictly on a FIFO (first-in, first-out) basis. Preparation and batching of non decontaminated materials (e.g. cutting or mixing vegetables), should be done in GMP areas laid out and operated to prevent cross contamination and limit temperature rise (e.g. to maximum 7 °C for chilled materials) and residence time. If longer times or higher temperatures are needed for preparation, any temperature rise should not lead to quality loss or microbiological deterioration. Hygienic conditions should prevent inadvertent contamination if mixtures of materials are prepared (e.g. cut, untreated vegetables for sandwiches or pre-packed salads). If potentially contaminated materials are to be used (e.g. herbs and spices), it is a difficult decision (to be made with marketing and QA) between the use of decontamination processes, which may lose an ‘edge’ of flavour, and taking the risks of introducing contamination by using untreated ones. Blanched vegetables Some vegetables are blanched with water or steam at high temperatures (90 °C– 95 °C) to inactivate the enzymes (usually peroxidises) responsible for quality deterioration (see Marshal et al., 2000). These heat treatments (in excess of 2 min × 70 °C) can ensure inactivation of vegetative micro-organisms, including infectious pathogens, but more severe heat treatments equivalent to 10 minutes at 90 °C (z = 9.2 C°) are needed to inactivate cold growing Clostridium botulinum. If the blanching stage is identified as CCP (e.g. used to eliminate infectious pathogens), it must be validated to ensure that the required minimum heat treatment can be achieved. But unless subsequent handling (e.g. in chilled flumes) is done hygienically, material can be re-contaminated, for example with Listeria spp. Where materials are destined for hygienic areas, then blanching in-pack on closed trays, or in tunnel ovens with integral cooling, is more reliable and hygienic. © 2008, Woodhead Publishing Limited

Operation of plants manufacturing chilled foods

361

Fresh or salad vegetables Where vegetables, herbs and spices are not blanched, other preventive measures are needed to minimize microbial levels in products. These focus on using good agricultural practice (GAP) during growing (especially irrigation with clean water), harvesting and post-harvest handling, based on hygienic equipment practices, and rapid cooling. GAP can help to minimize the incidence of infectious pathogens, such as L. monocytogenes, but cannot prevent their presence (CFA, 2002). If animal waste is used as a fertilizer, contamination risks are increased, and there must be a ‘disinfection’ period (some months) between fertilizer application and harvest, to allow any pathogens present to be destroyed. Evidence that GAP and processing practices can ensure absence of salmonella must be validated through an appropriate sampling plan aimed at building up historical data (not for positive release purposes). If there is evidence that salmonella could be present, an ingredient should not be used in RTE products without decontamination. If vegetables are decontaminated by washing, the microbiological quality of washing water, its circulation, replenishment and agitation, and the level of ‘free’ or active disinfectant (e.g. chlorine or silver) are critical to effectiveness. It will be ineffective irrespective of disinfectant addition rate, if water is not clean. This means that water must be changed regularly and debris removed efficiently. Final rinsing should always be in potable water. When salad vegetables are sold as unpreserved products, the only means of controlling microbiological hazards are supplier selection and restriction of microbial growth by limiting shelf-life and using appropriate packaging and pack atmosphere. Making unpreserved salad products with a shelf-life of > 5 days to meet EU microbiological criteria is difficult because the presence and growth of L. monocytogenes cannot be prevented. Frozen vegetables Frozen raw materials need to be thawed to ensure predictable heating, and hence thawing should be covered in the HACCP plan. Control of thawing times and temperatures is critical to minimizing microbiological growth and quality loss. If these materials are used without further decontamination, then process hygiene and the effectiveness of any prior blanching (e.g. by suppliers), as well as the thawing process, must be assessed to ensure acceptable levels of contamination in the final product. Thawing can be done in chillers, or in special equipment (e.g. microwave tempering units, running-water baths or air thawing units), operated hygienically and according to the supplier’s instructions. Thawing at ambient temperatures should be used only when the risks have been assessed, as it can lead to the growth of pathogens or spoilage micro-organisms and high losses from ‘drip’. Thawed materials should be processed as soon as possible (just in time, JIT) and not be refrozen. Thawed materials should not enter HRAs unless they have been cooked, frozen and thawed in their primary packaging, and surface-decontaminated prior to entry. Rice and pasta Pasta may be used dry, chilled or fresh. Usually it has been heated during © 2008, Woodhead Publishing Limited

362

Chilled foods

manufacture but needs to be cooked by the consumer for optimal texture (about 3 minutes in boiling water). It should be re-hydrated before pasteurization. If precooked pasta is used in salads, then the manufacturing process should re-pasteurize it. Pasta has a reduced water activity (aW about 0.95) which prevents growth of cold growing Bacillus cereus and C. botulinum. Therefore pasteurization treatments of about 2 min × 75 °C can be used to eliminate vegetative micro-organisms and a long shelf-life (>1 month) can be achieved in modified atmosphere packaging. Acidification of prepared salads must account for its pH buffering capacity and the diffusion of acid into the solids from the dressing. Pre-cooked rice is used in many RTE products, and heating during factory cooking is generally sufficient to inactivate cold growing C. botulinum (equivalent to 10 minutes at 90 °C: z = 9.2 C°). The heat process used and process hygiene must be validated for long shelf-life products where C. botulinum and B. cereus are realistic hazards. Pastry, bread and batter These ingredients should be formulated and baked to give good quality at the end of shelf-life, as staling and water migration from moist ingredients will restrict their shelf-life. Baking processes are usually equivalent to pasteurization and will inactivate infectious pathogens. Depending on hygiene levels (GMP, HCA or HRA) during cooling and assembly, re-contamination with spoilage micro-organisms (e.g. yeasts, moulds) and pathogens, (e.g. L. monocytogenes) may occur. Microbial levels at the end of shelf-life will be controlled by storage temperature, but some products use reduced aw, chemical preservatives (such as K-sorbate or Ca-propionate), or modified atmosphere packaging (MAP) to extend shelf-life. Meat, poultry and fish These materials may be sold raw or cooked prior to sale. Raw meats may be contaminated with infectious pathogens (e.g. salmonella and Campylobacter) and fish from tropical water may contain vibrios. Raw meats, poultry and fish, including fried, coated materials (e.g. battered fish), may be handled in GMP areas, as heating may not decontaminate the centre of the product. These materials may also be heated in pack (e.g. in sauces or as pieces, using cook–chill or in-pack heating) and then handled in HCAs; or cooked in ovens or fryers, chilled and then assembled in HRAs. The heating conditions and the maximum piece or particle sizes and initial temperature (e.g. frozen) must be set to ensure effective decontamination. The inclusion of bone in meat products will reduce the rate and predictability of heating. If pre-cooked or fermented meats are used and included in RTE products or taken into hygienic areas, then the procedures used during their manufacture and slicing should be validated for effectiveness.

13.5.3 Heat processing Heating cooks the product and eliminates or reduces numbers of specified micro© 2008, Woodhead Publishing Limited

Operation of plants manufacturing chilled foods

363

organisms. It is important to distinguish between the minimum heat treatment that the food requires to meet its product design and the process conditions needed to ensure that this is consistently done. The process specifications and controls should ensure that cooking and pasteurization conditions are met at the coldest point in a product unit placed at the coldest place in an oven or kettle. Variability of raw materials and controls (e.g. temperature and piece dimensions) has a large impact on the amount of heating needed to ensure that the product meets the designed heat treatment. It is usually necessary to re-establish process conditions if product is transferred from one line, or type of heater, to another. For short shelflife cooked RTE products, a minimum heat treatment equivalent to 70 °C for 2 min is needed to eliminate infectious pathogens, but this will not eliminate spoilage bacteria and yeasts/moulds, and higher heat treatments (e.g. 75 °C for 2 min) are used. Where longer shelf-life is required, 90 °C for 10 min is used to inactivate spores of cold-growing C. botulinum and similarly this may not free the product of cold-growing spoilage bacteria originating from surviving spores. As these grow slowly under chilled conditions, there is a compromise between more heating causing quality loss and the risks of spoilage towards the end of shelf-life. Different types of equipment are used for heating and cooking, depending on whether the material is solid or pumpable and the temperatures required. Nonpumpable foods (such as fillets, slices, etc,) will often be manually handled and cooked on trays in ovens or on belt fryers. Pumpable products (e.g. sauces, stews, etc.) can be heated in jacketed vessels or using continuous heat exchangers. Where these vessels are used for heating and cooling, the flow and heat transfer characteristics of the product (e.g. sauce) at both high and low temperatures need to be considered. Thin liquids will heat more efficiently than thick ones and therefore choice of thickener system and particle loading are important, especially if a thickener is used to keep particles in suspension. This equipment may be heated indirectly (e.g. by a jacket) or directly (by steam injection or infusion into the product stream). Direct steam becomes part of the product and therefore must be of potable quality. Integrated cook, hot-fill–chill systems are widely used for high volume operations. Longer heating times will give greater cooking effects and if the efficiency of heat transfer from the equipment to the product is low (e.g. because of fouling), high surface temperatures will result and may lead to burning. Many types of pack are used are used for chilled food, including trays, highdensity heat-sealed plastic bags, with or without vacuum, and chub packs which are clip-closed and are not hermetically sealed. The latter should be used only where the risk of product re-contamination is accepted. Packaging material, and shape and size (e.g. pack depth or size of particulates), and product initial temperature (e.g. frozen, when additional heat is needed to take the food through the latent heat zone) will dramatically affect process time because the heat transfer and the amount of heating needed to give a specific heat treatment during manufacture or consumer heating will be affected. Packs should be specified to withstand the different temperatures involved in production and consumer heating (e.g. oven or microwave). © 2008, Woodhead Publishing Limited

364

Chilled foods

For products sold raw, cooking by the consumer is critical to ensuring that consumer quality expectations and microbiological safety requirements are met. Consumer use instructions should produce good quality product and be validated to give a minimum heat treatment of 70 °C for 2 min (z = 7.5 C°, lethality calculated excluding temperatures below 65 °C) at the coldest point of the product. Expected performance of domestic equipment should also be taken into account in consumer instructions; for example, microwave ovens can heat products in a nonuniform way and for this reason microwaveable products should be RTE, and preparation instructions should indicate standing time for the diffusion of heat to even out temperatures within the pack. Cooling after heating. After factory heating, products and intermediate materials need to be cooled rapidly enough to prevent quality change and ensure that any micro-organisms surviving the heating process, or introduced during cooling, do not grow. Basic requirements are that cooling to below 10 °C is achieved within 4 hours and the storage temperature should be reached within 12 hours. Rapid chilling to 5–7 °C is required by Corrigendum to Regulation (EC) No 852/2004. Product can be cooled directly in heat exchangers or indirectly in packs or covered trays (for later assembly), using high velocity cold air in air blast chillers (either chambers or tunnels). Where spiral coolers are used, the belt should be kept clean to prevent product contamination by debris dropping from the belt. In cook–chill or hot fill operations, sealed packs are cooled by water sprays or immersion in tanks (static or tumbled) or in linear chillers. If chub packs (which are only clip sealed) are used or in the event of pack leakage, prevention of product contamination and pack soiling by cooling water is essential. Filters plus CIP systems to ensure equipment and water cleanliness are an integral part of design and operation. Product cooling rate will be determined by the supply and temperature of cooling medium, the temperature of the product, loading of the cooler (which influences heat load and circulation of cooling medium) and heat transfer from the pack (e.g. packs with thermal paths longer than about 50 mm and/or a large headspace will have slow cooling rates). If packs are shown to cool slowly, then the safety of the cooling should be validated against the lag time and growth rate of realistic hazards (e.g. L. monocytogenes, B. cereus and C. perfringens, and sporeformers causing spoilage) to ensure that growth does not occur. Fresh meat has specific regulatory requirements for immediate chilling after the post-mortem inspection, to achieve and maintain a constant internal temperature of not more than + 7 °C for carcases and cuts, and + 3 °C for offal (see EC 854/2007). Choice of correct chiller capacity (room volume and heat exchange capacity of evaporator units) and layout (e.g. fixed racks for packs or trays or corridors for trolleys) is critical to the capacity of the factory. Designed capacity is reduced if there is condensation or ice build up on evaporator coils during use. This is accelerated if trays of open, hot product are put into the chiller, and by frequent door opening. Condensation can be minimized by using covered trays, and the effects of door opening minimized by using strip or air curtains in door openings. © 2008, Woodhead Publishing Limited

Operation of plants manufacturing chilled foods

365

Keeping the evaporation units clean (e.g. use of anti-microbial gel to prevent growth) is critical, as condensation on the coils may provide a source of contamination if droplets are blown onto product by the air-circulation fans. To defrost coils, refrigeration has to be shut down. If hot water is used for thawing, it may create moist environments suitable for microbial growth if drying is not done after thawing. Ice removal is best done by using a hot air defrost system to melts the ice on the coils. There are also frost-free refrigeration systems where a continuous spray of propylene glycol is applied to the evaporator coils to prevent condensation freezing. Suppliers of refrigeration, air handling systems and cleaning chemicals can give additional information.

13.5.4 Storage of in-process and part-processed materials Storage conditions (e.g. temperature and maximum time) for raw or part-processed materials must not allow quality change, contamination or microbial growth. Containers should be closed and, if they are opened they should be re-closed or repackaged and batch identification retained. Chillers should not freeze the material, but should ensure that dispatch temperatures can be met after assembly and without further cooling. Re-circulation of cleaned, emptied bins and containers must not provide a source of contamination, and only disinfected containers should be used for product intermediates to be assembled into RTE products.

13.5.5 Filling/product assembly Consumer packs may be filled before or after heating. Filler hygiene and temperature (e.g. hot filling) must be matched to the microbiological safety risks of the product. Hot filling uses temperatures above 70 °C, with a minimum holding time after filling to eliminate any recontamination from the packaging, equipment or environment. Equipment used for hot filling should be pre-heated before filling starts and, as a precaution, fill temperatures for such processes are normally set above the minimum temperature to allow for any cooling during dosing and stoppages. When chilled products are assembled in hygienic areas after heating and cooling, the temperature rise during handling should not exceed the difference between storage temperature and dispatch temperature. To minimize temperature changes, minimum amounts of materials should be removed from chills, and filling (manual or automated) primary and secondary packaging must run smoothly on machines and the amount of product out of the chills awaiting assembly and final cooling should be minimized. There are EU requirements for temperature controls on chilled foods, covering maximum temperatures during production and in retail. The principle is that if foods are likely to support the growth of, and/or toxin production by, pathogenic micro-organisms, they should be held at or below 8 °C or above 63 °C. The emphasis is on food business proprietors to identify hazards and ensure that the correct preventive measures (e.g. temperature control and hygiene) are in place. To © 2008, Woodhead Publishing Limited

366

Chilled foods

ensure product safety, packaging must be robust enough to protect the product during processing, handling and consumer use and during the closed (and open) shelf-life.

13.6 Safety measures 13.6.1 Disposal of waste materials Effective measures need to be in place for collecting and disposing of food debris from production equipment, and storage and handling areas. Waste material should be kept in distinctive, closed containers in separate storage areas until disposal, and collection intervals should prevent the build-up of Listeria spp. After production, the maximum amount of food debris should be removed dry (e.g. by scraping). If this is not possible (e.g. from continuous heat exchangers or CIP systems) then it may be cost effective to have in-plant wastewater treatment facilities to prevent surcharges for high BOD or COD or pH.

13.6.2 Cleaning and disinfection Cleaning and disinfection are large indirect costs in chilled products factories, because they take resources and reduce capacity. All areas and equipment need to be cleaned using validated procedures and there should be effective, rapid hygiene monitoring systems. Some equipment (e.g. kettles) is cleaned in situ or using CIP. Other equipment such as fillers are disassembled for cleaning, and cleaning and disinfection of the demountable parts may be done manually or automatically in tray or tunnel washers. Within any installation where there are significant numbers of tools or change parts, there should be dedicated, separated rooms or areas for cleaning with an adequate supply of cleaning agent, good extraction of steam and sufficient space and racking to allow drying of change parts for equipment.

13.6.3 Safe working environment Food businesses need to ensure that the requirements of food hygiene law are achieved, while maintaining a safe working environment and a ‘reasonable temperature’ for personnel in workrooms. Food hygiene law regulates the temperatures of food, while health and safety law requires a workroom temperature of at least 16 °C (or at least 13 °C if the work involves serious physical effort). Hence, to maintain chilled product temperatures during assembly, product-intermediates food needs to be pre-chilled, and work practices and product flows aimed at minimizing temperature rise. Alternatively, materials can be kept in chilled or insulated enclosures or hoppers, or placed on chilled conveyors or tables, so that the product is chilled rather than the workroom. If this is not practical, a warm working station can be provided within a room where the overall temperature is lower (HSE, 2002). © 2008, Woodhead Publishing Limited

Operation of plants manufacturing chilled foods

367

13.7 Systems for controlling and monitoring the supply chain The intake of materials is assured by supplier approval and monitoring at intake. Equipment and areas may have manual or automated systems (e.g. Siematic or SattLine) to control and monitor key aspects of performance (e.g. temperature, time and pressure). These systems often rely on signals from sensors which may be:

• in-line (e.g. ovens, heat exchangers, chillers) • at the side of the line (e.g. pack seam measurement, drained weight, salt or pH) • in the laboratory (e.g. colour measurement, gas composition, microbiology). Specialized equipment, such as ovens and chillers, will have important additional controls for their utilities (e.g. supply and flow of heat transfer medium or completion of a designated cycle). Process sensors must be located to provide accurate measurements and will need to be regularly cleaned, maintained and calibrated (for key equipment this should be done according to a fixed schedule); otherwise they may produce erroneous signals. Connection of their output to signal processing equipment needs to be validated for accuracy, and cabling needs to be protected to prevent erroneous signals. Where control is manual, operators must be trained to make and respond to measurements and make records. Normally, process records and analytical data are reviewed by both production and QA to ensure that targets are met, and their interpretation should be used by a designated, responsible person for product release. All records should be archived as an aid to problem solving and to meet regulatory requirements (e.g. retained for >2 years). To track equipment performance, data can be compared with specifications (e.g. targets, and upper and lower limits). If this comparison is taken over time, it can be used to visualize trends and differences between materials and shifts and help the workforce to optimize production. This is sometimes called Statistical Process Control (SPC; see Grigg, 1998; Hayes et al., 1997). If patterns of variability are shown, a systematic fault is indicated and, if limits are not exceeded, the cause can be identified and eliminated without stopping production. If limits are exceeded, factory technical staff should stop production and initiate corrective actions to find the cause and prevent recurrence. For faults affecting product safety, corrective actions may be legally specified (e.g. Annex I, Hygiene Regulation: 852/2004) and may include modification of the HACCP plan or withdrawal of product.

13.7.1 Monitoring Monitoring is a planned sequence of measurements used to assess whether the CCPs (and pre-requisites) defined by the HACCP plan are under control (e.g. correct procedures are being followed and criteria met). Its frequency and coverage should detect trends towards unsatisfactory results (e.g. process hygiene criteria are exceeded or temperature targets are not met), it should produce records for use in verification, and it may generate improvement plans. If monitoring © 2008, Woodhead Publishing Limited

368

Chilled foods

shows that all CCPs are consistently within their limits, checking of finished products is not required; if it raises doubts, there should be further experimental or investigative work. Typically it covers:

• compliance of materials with specifications at reception, • performance of process controls (e.g. heating, cooling or thawing temperatures) and automated systems (software),

• flow and segregation of materials, (especially the times, temperatures and routes involved),

• cleaning and disinfection (e.g. contact times, disinfectant strengths and temperatures),

• maintenance of process equipment and processing areas, and • hygiene and training of production personnel, e.g. visual inspection of cleanliness, correct wearing of the factory clothing, respect of hygiene rules and working procedures, hand-washing requirements and medical screening. 13.7.2 Verification Verification reviews routine process, analytical and monitoring data to show dayto-day compliance with specifications, quality requirements and the HACCP plan. It may be supplemented with additional information from sampling, analysis of process control records and specialist auditing (e.g. hygiene) to highlight particular aspects. Company and customer QA departments and regulatory authorities are usually responsible for verification over months or years. And to do this effectively, they need to understand the importance of key process stages (e.g. heat treatment equipment, cooling rates), any preservation hurdles (e.g. pH or aw control), and environmental hygiene. Environmental hygiene data (e.g. temperature, air flow and various pressures) and microbiological testing for hygiene indicator organisms (TVC, Enterobacteriaceae, yeasts and moulds) as well as Listeria spp. can be used to verify that areas provide the required level of hygiene. If hygiene monitoring shows the presence of Listeria, operating, cleaning and disinfection procedures must be reviewed and may need to be changed to increase their effectiveness.

13.7.3 Training All personnel in the supply chain need to be trained so that they are effective in their roles and are empowered to take corrective actions, such as rejecting outof-specification ingredients or products. Management need to determine the skills for each job and arrange employee training, because well-trained and motivated personnel are key enablers of reliable production. Work procedures should form the basis of training and additional training is essential if individuals, or teams, control CCPs or operate complex equipment. Management should actively support and evaluate the effectiveness of training. All permanent and temporary employees need basic hygiene training, plus training on procedures to control hazards and ensure quality at their stage of manufacturing. HACCP © 2008, Woodhead Publishing Limited

Operation of plants manufacturing chilled foods

369

training should support the way HACCP is used in the company, as successful implementation requires the full cooperation and commitment of employees, including senior management. If guides to good hygienic practice are used as an alternative to HACCP, training should make staff familiar with their content, relevance and implementation. For critical or complex equipment, such as pasteurizing ovens or fillers, training may be provided by equipment suppliers. Where in-house training is provided as an alternative, it should always be acceptable to the equipment supplier. Staff should be given regular refresher training and auditing should verify their performance. Changes to equipment, materials, procedures or failures should trigger refresher training. Depending on the change, some aspects may require increased emphasis (e.g. temperature control during cooking, cooling or storage, hygiene procedures or routing of specific materials to prevent recontamination). Training and procedures should ensure that employees use changing areas correctly and are correctly dressed for the areas they enter (e.g. segregated or hygienic areas). Personnel working on the ‘clean’ side of decontamination equipment, such as ovens or baths, must understand the control principles used, monitoring procedures and management of process deviations. Operators in HRAs need specific additional training as they are responsible for the manual or automated handling of pasteurized materials in an open environment and assembling them into RTE products. Training must emphasise the continuing importance of reporting illness or skin infections to management and create a general awareness of precautions that must also be taken by visitors and maintenance personnel, contract workers, etc. visiting hygienic areas.

13.7.4 Analysis of product/processing problems There should be documented procedures for dealing with non conformances or problems involving suppliers or manufacturing and complaints; their primary purpose should be to protect consumers and next to protect the business and brand. Structured techniques (Fig. 13.2.) should be used to identify how and why something happened and to correct the specific causes. Generally, however, events or mistakes have multiple causes and do not just happen; all the causes need to be tracked to provide an effective solution and prevent recurrence. The actions and decisions taken to prevent release of non-conforming product, determine what to do with it and deal with the problem, should be recorded along with an analysis of effectiveness. Some examples of common causes of problems are given in Table 13.5.

13.8 Stock Management of stock levels (e.g. raw material, semi-finished or finished products) at every stage of the supply chain is essential to ensuring that the specified products are made, and the service levels required by customers are met (e.g. products are © 2008, Woodhead Publishing Limited

370

Chilled foods

Fig. 13.2

Structured problem solving.

Table 13.5 Common causes of problems Equipment failure Utilities interruption or failure Out-of-specification raw material or packaging Incorrect response to alarms Mismatch of recipe and process Incorrect decision making by Hygiene failure management Incorrect setting-up, maintenance or alignment Procedural or human error of equipment Product/process design problem, Faults in controls and/or sensors e.g.setting heating process Software failure or corruption Poor training

delivered on time, at the correct temperature and with the correct residual shelflife). Service level reflects the proportion of times when sufficient product is available to meet customer demand; hence the concept of service level applies all along the supply chain. Because of their short shelf-life, chilled products are ideally manufactured JIT (see http://www.tutor2u.net/business/production/justin-time.html). This means that batches of product need to be made and distributed on a frequent basis, and ingredients and packaging materials ordered and delivered as they are needed, with a minimum of stock being held. Minimum stock levels are © 2008, Woodhead Publishing Limited

Operation of plants manufacturing chilled foods

371

based on the quantity of materials needed to replace stock that is being sold or used and, if demand drops, can include materials likely to go out-of-date. Ideally, stock levels should be sufficient to cover the time needed to produce replacement stock after heavy demand or supply chain failure, and manufacturing and warehouse capacity have to be sufficient to cope with this variability. Sometimes stock levels may be driven by minimum order quantities and availability from suppliers. The location of stock (e.g. separate areas) and its storage conditions (e.g. temperature) may be determined by legislation or by customers.

13.9 Logistics Logistics involve transport and storage at all stages of the supply chain (e.g. from delivery of raw materials to the distribution of finished products) using either a company’s own facilities or third-party contractors. Logistics chains for chilled products need to be operated to meet a number of requirements including temperature and time control, product and material segregation, traceability and hygiene on a day-to-day and batch-by-batch basis. Because of short shelf-lives and the demand for high service levels from customers, manufacturing units need effective contact with the logistics team to track where the required materials and products are located. For traceability, the logistics team should be able to identify when batches are divided or enter a new branch of the logistics system and the routes for product return. Very often logistics lines cross different companies (suppliers and customers) and international boundaries; this makes coverage in the HACCP plan both difficult and important.

13.9.1 Product transport and storage All transport should be in chilled vehicles and trailers that comply with food hygiene legislation. These should be insulated and equipped with a chilled air circulating system capable of maintaining a uniform distribution of chilled temperatures (usually at or below 1–5 °C) without freezing product surfaces. Specified product temperatures should be achieved prior to loading, because most trucks and containers have very limited capacity to reduce product temperatures. Temperature monitoring and recording devices and alarms must be present and be regularly tested and calibrated. Products in packs should be palletized or stacked off the floor to allow airflow during transport and storage; loading should not block the distribution of chilled air from evaporator units. If trailers or containers are pre-loaded or uncoupled from tractor units, they must be able to maintain specified temperatures during prolonged stops. If mixed loads or storage facilities are used, special attention should be paid to the risks of flavour uptake or contamination from adjacent products. Vehicles, whether company owned or hired, should be regularly serviced both for road safety and temperature control. Choices on means of transport (e.g. truck or rail + truck) are usually cost-based, © 2008, Woodhead Publishing Limited

372

Chilled foods

but sustainability is starting to be considered. The environmental impact of transporting materials and products should be kept as low as possible, consistent with safe and reliable delivery. Choices are often linked to local availability of transport types and the distances involved, quantities of product and any special quality, safety or regulatory requirements for temperature, packaging or hygiene controls. The number of journeys between manufacturing sites, depots and stores can be greatly reduced by shared use of vehicles. However, product or materials should never be stored or transported with other materials that increase (microbiological) contamination risks. Packs of chilled products are usually transported in cartons (sales units) to protect their primary packaging from damage. Cartons containing up to 25 packs should be placed on pallets in a defined stacking pattern and stabilized with plastic shrink wrap or strapping. Stacking patterns need to ensure stability during handling and may also need to allow circulation of air, if products are not fully cooled before palletization.

13.9.2 Distribution centres Distribution centres (DCs) receive batches of materials and products, and then break them up and recombine them into outbound shipments. Chilled distribution centres typically run at 0–7 °C, and temperatures should be checked daily. Temperature measuring devices should be calibrated as indicated by the manufacturer or as required by law. DCs are often large with a 6 m clear height and 10 000 to 100 000 square meters of floor space containing automated product handling equipment (e.g. sorting systems) and storage racks for batch identification and order-picking. They may serve as short-term warehouses. Dispatch and receiving docks may have floor levelling ramps to compensate for differences in height between truck floors and the warehouse floor, so that forklifts can drive directly onto trucks for loading and unloading. Temperature control during handling can be difficult and pre-cooled vehicles should be loaded directly from chilled areas. To reduce temperature fluctuations and ensure that product maintains its specified temperature during loading, closed docking systems (e.g. forming a closed tunnel between the truck or container and the chilled areas) are used for receipt and dispatch. Docks and strip or air curtains in doorways can be used to protect personnel and product from the weather and ingress of warm air when they are loading or unloading trucks. After distribution, product for retail sale at its retail destination should be unloaded directly into a chilled area and not left in areas where the temperature is not controlled, e.g. on an open docking apron. Raw material and product temperatures should be measured (accuracy ± 0.5 °C) at arrival and dispatch at all stages in the supply chain, to establish that specified temperatures are met. Preferably this is done without damaging packs (e.g. a temperature probe placed between the packs) and it may be sufficient to check the air temperatures in trucks. Actions if temperatures are outside specifications must be addressed in the HACCP plan and contracts. © 2008, Woodhead Publishing Limited

Operation of plants manufacturing chilled foods

373

13.10 Future trends Chilled food manufacture is now an established technology with increasingly uniform ways of working. In the future, facilities are likely to become larger and more highly automated as there is consolidation. Globalization will increase the scope of raw material supply, leading to new hazards and technologies, such as superchill. These will allow sourcing of part-processed materials on a global basis. Demand for increasingly complex products will grow, and the greater reliability of manufacturing and logistics systems will allow manufacturers to place less reliance on intrinsic preservation systems and greater reliance on low temperatures to ensure shelf-life, safety and quality. Health and sustainability concerns will drive product ranges and encourage lower energy usage and probably shorter supply chains. There will be a continuing expansion of the market into the food service sector, leading to the need to process and handle larger packs.

13.11 References (2005). Guidelines on air quality for the food industry, 2nd edition, CCFRA Guideline No. 12. CHILLED FOOD ASSOCIATION (CFA: 2002). Microbiological guidance for produce suppliers to chilled food manufacturers. First edition. Chilled Food Association, PO Box 6434, Kettering NN15 5XT, UK. CHILLED FOOD ASSOCIATION (CFA: 2006a). Microbiological testing and interpretation guidance, 2nd edition. http://www.chilledfood.org/_attachments/Resources/ CFA_Micro_testing_interpretation_2nd_ed.pdf. Chilled Food Association, PO Box 6434, Kettering NN15 5XT, UK. CHILLED FOOD ASSOCIATION (CFA: 2006b). CFA Guidelines for good hygienic practice in the manufacture of chilled foods – 4th edition (The CFA Guidelines). Chilled Food Association, PO Box 6434, Kettering NN15 5XT, UK. CFA/BRC, Guidance on the implementation of the EC Microbiological Criteria Regulation. http://www.chilledfood.org/content/guidance.asp and http://www.chilledfood.org/_attachments/Resources/BRC_CFA_Micro_criteria_guidance_ed_1.2.pdf. CODEX ALIMENTARIUS (2003). Recommended international code of practice. General principles of food hygiene, CAC/RCP 1-1969, rev. 4-2003. EUROPEAN COMMISSION (2002). General Food Law (Regulation No. 178/2002/EC). www.europa.eu.int/eur-lex. EUROPEAN COMMISSION (2005). Commission regulation on microbiological criteria for foodstuffs (2073/2005/EC). EUROPEAN COMMISSION (2004a). Regulation (EC) No. 852/2004 on the Hygiene of foodstuffs. EUROPEAN COMMISSION (2004b). Regulation (EC) No 853/2004 on the Hygiene of food of animal origin. EUROPEAN FEDERATION OF ACCOUNTANTS (2005). Assurance of a sustainable supply chain. http://www.fee.be/publications/default.asp?library_ref=4&content_ref=390 GRIGG N.P (1998). Statistical process control in UK food production: An overview, International Journal of Quality and Reliability Management, 15(2), 223–238. HAYES G.D, SCALLAN A.J., WONG J.H.F (1997). Applying statistical process control to monitor and evaluate the hazard analysis critical control point hygiene data. Food Control, 8(4), 173–176. CAMPDEN AND CHORLEYWOOD FOOD RESEARCH ASSOCIATION

© 2008, Woodhead Publishing Limited

374

Chilled foods

(2002). Workroom temperatures in places where food is handled – HSE Food Information Sheet No 3 (2002) Health and Safety Executive FIS3 (rev1) C50 http://www.hse.gov.uk/food/chilled.htm and http://www.hse.gov.uk/pubns/ fis03.pdf. IEMA (2006). Sustainable supply chain management. Institute of Environmental Management and Assessment. http://www.iema.net/readingroom/show/602/c176. JOHR H (2006). Sustainable Agriculture and the food industry. http://www.nestle.com/NR/ rdonlyres/2C36D7E3-A375-4531-B66E-84F6F6516711/0/WorldFoodPrize2006HJ Nestle.pdf. MARSHALL M.R., KIM J. AND WEI C.I. (2000). Enzymatic browning in fruits, vegetables and seafoods. FAO, Viale delle Terme di Caracalla, 00100 Rome, Italy. (http://www.fao.org/ ag/Ags/agsi/ENZYMEFINAL/Enzymatic%20Browning.html. NAKAJIMA S. (1988). Introduction to total productive maintenance. Cambridge, MA: Productivity Press. ULRICH K.T. AND EPPINGER S.D. (2000). Teaching materials to accompany Chapter 11 Design for manufacturing; Product design and development. 2nd Edition, Irwin McGraw-Hill. (http://ocw.mit.edu/NR/rdonlyres/Sloan-School-of-Management/15-783JProduct-Designand-DevelopmentSpring2002/62BF82DF-E5CA-4A0E-93C9-488A28B97248/0/ 11dfm.pdf). HEALTH AND SAFETY EXECUTIVE

© 2008, Woodhead Publishing Limited

14 Refrigeration, storage and transport of chilled foods S. James and C. James, Food Refrigeration and Process Engineering Research Centre (FRPERC), UK

14.1 Introduction: the cold-chain Chilled foods are cooled to, and held at, a temperature below ambient and above their freezing point to preserve quality. To provide safe food products of high organoleptic quality, attention must be paid to every aspect of the cold-chain, from initial chilling of the raw ingredients to retail display. The cold-chain consists of two distinct types of operation: (i) in processes, such as primary and secondary chilling – the aim is to change the average temperature of the food, (ii) others, such as chilled storage, transport, and retail display – the prime aim is to maintain the temperature of the food. Removing the required amount of heat from a food is a difficult, time and energy consuming operation, but it is critical to the operation of the cold-chain. As a food moves along the cold-chain, it becomes increasingly difficult to control and maintain its temperature. This is because the temperatures of bulk packs of chilled product in large storerooms are far less sensitive to small heat inputs than single consumer packs in open display cases or in a domestic refrigerator.

14.2 Principles of refrigeration A detailed analysis of refrigeration cycles and systems may be found in numerous refrigeration textbooks (Gosney, 1982; Trott, 1989; ASHRAE handbooks). Never© 2008, Woodhead Publishing Limited

376

Chilled foods

theless, a basic understanding of the principles of how a refrigeration system works is useful for all users of refrigeration equipment.

14.2.1 Mechanical refrigeration systems Mechanical refrigeration systems operate using the same basic refrigeration cycle. At its simplest, it utilises four interlinked components: the evaporator, compressor, condenser and the expansion valve. A low-pressure liquid refrigerant is allowed to evaporate to a gas within the ‘evaporator’ coil. This process requires heat, which is extracted (e.g. from the product), thus cooling any medium surrounding the evaporator (e.g. air). The gas from the evaporator is compressed in the ‘compressor’ to a high-pressure hot gas. This high-pressure hot gas is then passed through another coil where it condenses. This process releases heat into any medium surrounding the ‘condenser’ coil. The high-pressure, cold, liquid refrigerant then passes though a valve, the ‘expansion valve’, to a lower pressure area. The now low-pressure liquid then passes back to the evaporator. In a direct expansion system, the evaporator coil is either in direct contact with the food to be refrigerated, or the cooling media (e.g. air or brine) surrounding the food. In a secondary refrigeration system, a liquid is cooled by passing it over the evaporator coil and this cooled liquid is used to cool the cooling media (e.g. air, water, brine) surrounding the food. The energy used by a refrigeration system depends on its design, but generally the larger the temperature difference between the evaporator and condenser, the greater the energy used in the compressor for a given amount of cooling duty.

14.2.2 Cryogenic refrigeration Cryogenic cooling uses refrigerants, such as liquid nitrogen or solid carbon dioxide, in direct contact with the product, with cooling being brought about by boiling off the refrigerant. Due to very low operating temperatures and high surface heat transfer coefficients between product and medium, cooling rates of cryogenic systems are often substantially higher than with other refrigeration systems. However, avoiding surface freezing of the product is a major problem in using cryogens for chilling. Most cryogenic systems use total loss refrigerants, i.e. the refrigerant is released to the atmosphere and not recovered. Due to environmental and economic factors, total loss refrigerants must be both readily available and harmless, which limits the choice to atmospheric air and its components (Heap and Mansfield, 1983). Nitrogen is the main constituent of atmospheric air, and at atmospheric pressure, liquefies at a temperature of –196 °C, giving a refrigerating capacity of 378 kJ kg–1. It is usually supplied and stored at a pressure of 3 to 6 bar, with corresponding boiling points of –185 °C to –177 °C (Heap and Mansfield, 1983; Hoffmanns, 1994). A useful rule of thumb is that 1 ton h–1 of liquid nitrogen is approximately equivalent to 100 kW of mechanical refrigeration. Carbon dioxide gas is present in air at a concentration of 0.03 to 0.05% and, if stored as a pressurised liquid and © 2008, Woodhead Publishing Limited

Refrigeration, storage and transport of chilled foods

377

released into the atmosphere, the liquid changes partly to gas and partly to a frozen solid at –78.5 °C which sublimes directly into gas without going through a liquid phase. Liquid carbon dioxide is generally supplied either at ambient temperature (e.g. 25 °C and 65 bar), giving a refrigerating capacity of 199 kJ kg–1, or at –16 °C and 22 bar, giving a refrigerating capacity of 311 kJ kg–1.

14.2.3 Energy and the environment The dominant types of refrigerant used in the food industry in the last sixty years have belonged to a group of chemicals known as halogenated hydrocarbons. Members of this group, which includes the chlorofluorocarbons (CFCs) and the hydrochlorofluorocarbons (HCFCs), have excellent properties, such as low toxicity, compatibility with lubricants, high stability, good thermodynamic performance and relatively low cost, which make them excellent refrigerants for industrial, commercial and domestic use. However, their high chemical stability leads to environmental problems when they are released and rise into the stratosphere. Scientific evidence clearly shows that emissions of CFCs have been damaging the ozone layer and contributing significantly to global warming. With the removal of CFCs from aerosols, foam blowing and solvents, the largest single application sector in the world is refrigeration, which accounts for almost 30% of total consumption. Until recently R12, R22 and R502 were the three most common refrigerants used in the food industry. R12 and R502 have significant ozone depletion potential (ODP) and global warning potential (GWP) and R22, though of smaller potential, is still dangerous in the long-term. Consequently, as a result of international agreements, pure CFCs (e.g. R12, R502) have been completely banned. Pure HCFCs (mainly R22) are banned in new industrial plant and are soon to be phased out completely. HCFC blends (e.g. R401A, R403B, R408A), pure HFCs (mainly R134a) and HFC blends (e.g. R404A, R407C, R410A), originally introduced as CFC replacements, are covered by F-Gas regulations that limit leak rates. Chemical companies are making large investments in terms of both time and money in developing new refrigerants that have reduced or negligible environmental effects. Other researchers are looking at the many non-CFC alternatives, including ammonia, propane, butane, carbon dioxide, water and air that have been used in the past. Ammonia is the common refrigerant in large industrial food cooling and storage plants. It is a cheap, efficient refrigerant whose pungent odour aids leak detection well before toxic exposure or flammable concentrations are reached. The renewed interest in this refrigerant has led to the development of compact, low charge (i.e. small amounts of ammonia) systems that significantly reduce the possible hazards in the event of leakage. It is expected that ammonia will meet increasing use in large industrial food refrigeration systems. Carbon dioxide is being advocated for retail display cabinets, and hydrocarbons, particularly propane and butane or mixtures of both, for domestic refrigerators. As well as the direct effect of refrigerants on the environment, energy efficiency © 2008, Woodhead Publishing Limited

378

Chilled foods

is increasingly of concern to the food industry since refrigeration is known to constitute up to 50% of the total energy used by the food industry.

14.3 Chilling Chilling is a process of removing heat and can be achieved only by four basic mechanisms: radiation, conduction, convection or evaporation. To achieve substantial rates of heat loss by radiation, large temperature differences are required between the surface of the product and that of the enclosure. Physical contact between the product and the source of refrigeration is required to extract heat by conduction. Plate conduction coolers are used for quick cooling of some packaged products and highly perishable products such as mechanically recovered or hot boned meat. For the majority of foods, the heat lost through evaporation of water from the surface is a minor component of the total heat loss, though it is the major component in vacuum cooling. Most foods are chilled by convection of heat into the air or another refrigerated medium and the rate of heat removal depends on the (i) surface area available for heat flow, (ii) temperature difference between the surface and the medium, (iii) surface heat transfer coefficient. Each combination of product and cooling system can be characterised by a specific surface heat transfer coefficient whose value depends principally on the thermophysical properties and velocity of the medium. Typical values range from 5 W m–2 K–1 for slow moving air to 500 W m–2 K–1 for agitated water. Heat must also be conducted from within the food to its surface before it can be removed. Most foodstuffs are poor conductors of heat and this imposes a severe limitation on attainable chilling times for either large individual items or small items cooled in bulk. For the majority of chilled foods, air systems are used, primarily because of their flexibility and ease of use. However, for certain foods, other systems can offer much faster and more controlled chilling. From a hygiene/HACCP based approach, pre-packing the food prior to chilling may lower the risk of contamination/cross-contamination during the chilling process; however, it will significantly affect the rate of cooling (particularly in systems with low surface heat transfer rates such as low velocity air), and this in turn may allow the growth of any micro-organisms present. Provided the cooling media (air, water, etc.) and refrigeration equipment used is kept sufficiently clean, no one cooling method can be said to be intrinsically more hygienic than any other. It may, however, be said that more rapid cooling systems allow less time for any contamination/cross-contamination to occur than slower cooling systems. 14.3.1 Moving air Air systems range from the most basic in which a fan draws air through a refrigerated coil and blows the cooled air around an insulated room (see Fig. 14.1), © 2008, Woodhead Publishing Limited

Refrigeration, storage and transport of chilled foods Evaporator

379

Reversible fan

False ceiling

Product on trolleys

Fig. 14.1

Example of a chilling tunnel with longitudinal air circulation.

to purpose-built conveyerised blast chilling tunnels or spirals. Relatively low rates of heat transfer are attained from product surfaces in air-cooled systems. The big advantage of air systems are their cost and versatility, especially when there is a requirement to cool a variety of irregularly shaped or individual products. In practice, air distribution is a major problem, often overlooked by the system designer and the operator. The cooling time of the product is reduced as the air speed is increased. An optimum value exists between the decrease in cooling time and the increasing power required to drive the fans to produce higher air speeds. Even when a system has been designed to distribute the air through the product, poor management and/or poor understanding of the requirement of the plant commonly leads to uneven cooling. Products stacked or racked irregularly will leave channels around the stacks that are larger in crosssectional area than those within the stacks, and channels of differing areas through the stacks. Air leaving and returning to the refrigeration coil will take the path of least resistance through the largest gaps, instead of passing evenly through or over the product. Weight loss from exposed unwrapped products can also be a problem. Wrapping or covering the product (when practicable) can stop weight loss, but will also introduce an insulating layer around the product that will substantially reduce heat transfer and increase cooling time. Although greater air velocities will increase the rate of weight loss, its effect in reducing the total cooling time may lead to less weight loss overall. Once the surface of the food has been cooled, then there will be only a small temperature difference between the food and the air. During the final phase of cooling (sometimes referred to as equalisation), very little heat will be transferred from the food and the relative humidity (RH) of the air will be important. In typical mechanical refrigeration processes, the RH varies between © 2008, Woodhead Publishing Limited

380

Chilled foods

80% and 100%. Weight loss can be minimised during this phase by keeping the air velocity to the minimum required to maintain a small temperature difference between the surface and the air, and by keeping the RH as near to 100% as possible. Alternatively, a sacrificial layer of water can be applied to the product surface prior to, or during, chilling to prevent weight loss from the product itself. Batch air coolers Placing warm food items in large refrigerated rooms is the most common method of cooling. Large individual items, such as meat carcasses, tuna, or bunches of bananas, are usually hung from overhead rails. Smaller products are placed either unwrapped, or in cartons, on racks or pallets. Bulk fruits and vegetables are often cooled in large bins. In simple single-stage batch chilling systems, the risk of surface freezing limits the lowest air temperature that can be used. This problem is further complicated if products with different thermal properties and/or sizes are cooled at the same time. For example, two-compartment ready meal consumer packs often contain rice or pasta in one compartment, and a meat, fish or vegetable based product in the second. The components have very different thermal properties and may be filled to different depths. Investigations into the cooling of such products (James, 1990) showed that using air at –10 °C and a high velocity of 5 m s–1 resulted in a cooling time of 34 minutes, but substantial quantities of the product in both compartments were frozen (see Table 14.1). Reducing the air velocity to 0.5 m s–1 more than doubled the cooling time but produced a situation where only a small area of the rice was frozen. With higher air temperatures and longer times the extent of freezing was reduced. The relative effect of changes in air velocity depends on the size of the product being chilled. Increasing the air velocity during the chilling of beef sides substantially reduces chilling times at low air velocities but the effect is smaller at higher velocities (see Table 14.2). Much higher air speeds can be justified with small products, for example individual pork pies (see Fig. 14.2). In a high throughput baking line (>1000 items per hour), the 7% increase in throughput that may be achieved by raising the air velocity from 6 to 10 m s–1 (reducing the cooling time by 10 minutes), could justify the higher capital and the running costs of larger fans. Table 14.1 Effect of air temperature and velocity on cooling time in 30 mm thick twocompartment ready meals Air temperature (°C) –10 –10 –10 –5

Velocity (ms–1)

Time 80 to 4 °C (min)

5.0 1.0 0.5 5.0

34 58 78 42

© 2008, Woodhead Publishing Limited

Minimum temperature Meat (°C) Rice (°C) –4.2 –1.4 –0.2 –2.2

–3.8 –4.8 –2.2 –3.0

Refrigeration, storage and transport of chilled foods

381

Table 14.2 Effect of air velocity on the chilling time (hours) to maximum temperature of 10 °C in a beef side Air velocity (m s–1) * 1.0 2.0

Side weight (kg) 105 140

0.5

0.75

19.5 24.1

18.5 22.8

18.0 21.8

3.0

16.0 19.7

14.8 18.5

* Air temperature 0 °C 70 ¸ ü

Centre temperature (°C)

60


õ

10.0 m s–1




6.0 m s–1

ü

1.0 m s–1

¸

0.5 m s–1

õ ¸ ü

50




40

õ

¸ ü ¸


30

ü

¸




ü

õ ¸ ¸

õ

¸

ü


õ

20

¸ ü
õ

10

¸ ¸

ü
õ

ü
õ


õ

¸ ¸

ü ü
õ

¸ ¸

ü




ü

150

180

ü

¸

¸

¸

¸

ü

¸

¸

0 0

30

60

90

120

210

240

270

300

Time (min)

Fig. 14.2

Effect of air speed on the cooling time of pork pies.

Continuous air coolers Conveying the product through the cooling system often overcomes the problem of uneven air distribution since each item is subjected to the same velocity/time profile. In the simplest continuous air chilling systems, the food is suspended from an overhead line and is moved through a refrigerated room. This process is often used for air chilling of poultry (James et al., 2006a) or in pre-chilling of pork carcasses. In more sophisticated plants, the product is conveyed through a chilling tunnel. The main advantage of this method is that the refrigeration capacity and air conditions can be varied throughout the length of the tunnel. Large capacity evaporators can be installed in the initial stage to cater for the high rates of heat release encountered at the start of a cooling process. Higher air temperatures can be used in the latter stages to avoid surface freezing. © 2008, Woodhead Publishing Limited

382

Chilled foods

Some small cooked products (e.g. pies, cakes) are continuously cooled on racks of trays (8 to 16 high) that are pulled or pushed through a chilling tunnel using a simple mechanical system. However, this process often involves double handling and it is difficult to achieve an even air distribution through the different layers. In larger operations it is more satisfactory to convey the cooked products through a linear tunnel or spiral chiller. Linear tunnels are far simpler constructions than spirals but their use is often ruled out due to constraints on floor area. For example, even if a high packing density of 20 kg m–2 can be achieved, processing 1 tonne per hour of a small product, with a cooling time of 0.5 hour, requires a 1 m wide belt 25 m long. Much lower packing densities (<14 kg m–2) and longer cooling times (1 to 1.5 hours) are often required, which results in belt lengths that are too long to be accommodated in most factories. Belts of this length are most economically used in a spiral configuration where the belt is designed to collapse on its inner edge, enabling it to be wound round a drum in an ascending or descending orientation. Advantages are claimed for either horizontal or vertical air distribution through continuous belt systems. In horizontal systems, a lower pressure drop and smaller temperature difference across the products and coil results in less weight loss from unwrapped products, whilst in vertical systems increased turbulence and higher, and more uniform, heat transfer coefficients lead to reduced chilling times (Everington and Sagoo, 1986) Wet-air/ice-bank cooling One of the principal disadvantages of air cooling systems is their tendency to dehydrate unwrapped products. A way around this problem is to saturate the air with water. Wet-air cooling systems recirculate air over ice cold water so that air leaving the cooler is cold (0 to 1 °C) and virtually saturated with water vapour (100% RH). An ice-bank chiller uses a refrigeration plant with an evaporator (plate or coil) immersed in a tank of water that chills the water to 0 °C. During times of low load and with overnight use of off-peak electricity, a store of ice is built up on the evaporator, which subsequently melts to maintain temperatures during times of high load. As well as the advantage of reduced weight loss, the operation of ice-bank chillers can offer a number of other economic savings over conventional air chillers, in terms of their use of a small refrigeration system operating at peak efficiency and the ability to handle peak loads. The principal disadvantage of wetair and ice-bank coolers is the space required. Wet-air cooling has been used commercially for some 25 to 30 years, principally for the pre-cooling and storage of vegetables and fruit (MacLeod-Smith et al., 1994). Ice-bank refrigeration systems have a proven role in providing chilled water for milk cooling. Other applications remain to be exploited.

14.3.2 Direct contact Contact refrigeration methods are based on heat transfer by contact between products and metal surfaces, which in turn are cooled by either primary or © 2008, Woodhead Publishing Limited

Refrigeration, storage and transport of chilled foods

383

secondary refrigerants. Contact chilling offers several advantages over air cooling, such as much better heat transfer and significant energy savings. Plate heat exchangers Modern plate cooling systems differ little in principle from the first contact freezer patented in 1929 by Clarence Birdseye. Though generally designed for freezing, these systems can be readily modified to operate at higher temperatures for chilling operations. Essentially, product is pressed between horizontal or vertical hollow metal plates containing a circulating refrigerant (see Fig. 14.3). Plate chillers tend to be expensive but have low running costs. Good heat transfer is dependent on product thickness, good contact, and the conductivity of the product. Air spaces in packaging and fouling of the plates can have a significant effect on cooling time. With thin materials a plate chilling system has the potential to halve the cooling time required by an air blast system (see Table 14.3), but is often limited to a maximum thickness of 50 to 70 mm (Ciobanu et al., 1976; Persson and Löndahl, 1993).

Hydraulic ram Freezing plate

Product

Separated plates

Fig. 14.3

Closed plates

Example of a horizontal plate cooler.

Table 14.3 Predicted cooling times (h) from 40 to 2 °C at the centre of meat slabs in various cooling systems Cooling method (operating at –1 °C) Air (still) Air (5 m s–1) Plate Immersion

© 2008, Woodhead Publishing Limited

20 5.0 h 1.2 h 0.7 h 0.4 h

Meat thickness (mm) 40 11.0 h 2.8 h 1.8 h 1.2 h

80 24.0 h 7.4 h 5.5 h 4.4 h

384

Chilled foods

Jacketed heat exchangers Jacketed batch vessel coolers can range in capacity from 100 to 10 000 litres and are often used for batches of liquid foods. The cooling medium may circulate through the jacket of the vessel, through a coil immersed in the liquid, or both. When a direct expansion refrigeration system is used, the refrigeration surface is usually built into the stainless steel lining of the tank. Direct expansion coils are sometimes placed in the base and a reservoir of ice built up before the hot foodstuff is introduced into the vessel. Alternatively, ambient or iced water can be used as a secondary cooling medium. Direct expansion requires only about half the energy of ice water chilling but requires a condensing unit two to four times as large (Veisseyre, 1986). Most vessels are provided with agitators or scrapers to improve the rate of convective heat transfer and prevent temperature stratification or gelling of cooled product on the heat exchange surfaces. In agitated vessels, the design and operation of stirrers is critical if breakdown of delicate solid product is to be avoided. Since large quantities are usually processed at a time, cooling may not be particularly fast in such vessels. Burfoot et al. (1987) found that cooling times from 90 to 20 °C of quite small quantities of liquids (250 litres) in a jacketed vessel with cooling water are over 120 minutes (see Fig. 14.4). Belt heat exchangers Belt systems consist of an endless steel belt (around 1 mm thick), the underside of which is cooled either directly with water, brine or glycol sprays or by sliding over a stationary cold surface. Since only one side of the product is in contact with the cooling surface, relatively thin products are required, such as hamburgers, fish

90

Vacuum

Temperature (°C)

Water 70

50

30

10 0

20

40

60

80

100

120

Time (min)

Fig. 14.4

Cooling of liquid in 250 litre vessel using a water cooling jacket or under vacuum.

© 2008, Woodhead Publishing Limited

Refrigeration, storage and transport of chilled foods

385

fillets; or liquid and semi-liquid products such as purées and sauces. A cooler of this type using a 1 m wide belt, 10 m long, could cool 8 tonnes of 2 cm thick product (e.g. meat) per day from 40 °C to 2 °C. The main advantages of belt systems are: continuous processing; easy continuous cleaning and sanitation; reduced evaporative losses, in comparison with air systems; and, the possibility of operating with several temperature zones. Capital cost is high. Continuous heat exchangers Continuous heat exchangers are often used for the cooling of liquid foods, such as fruit juices and dairy products. Falling film or surface heat exchangers (also known as Baudelot coolers) in which the hot liquid is pumped over the top of a horizontal bank of refrigerated coils and flows down over the cooled surfaces were originally used for liquid foods, such as beer wort and cream (Ciobanu et al., 1976). However, these systems have now largely been replaced by totally enclosed coolers, such as scraped surface, tubular, multi-plate or multi-tube heat exchangers. Multi-plate coolers have: the highest available heat transfer surface; lowest material requirements (material retention/dead space); maximum efficiency (up to 90% heat recovery in counter current mode); and are very flexible in operation and easy to clean. In certain applications for thin liquids such as beer and wine, multitube coolers, which have a much higher resistance to pressure and can use primary refrigerants, have advantages over multi-plate coolers. These types are not suitable for viscous products such as sauces or stews.

14.3.3 Immersion/spray As their names imply, these systems involve dipping product into a cold liquid, or spraying a cold liquid onto the food. This produces high rates of heat transfer due to the intimate contact between product and cooling medium. Both offer several inherent advantages over air cooling in terms of reduced dehydration and coilfrosting problems (Robertson et al., 1976). Clearly, if the food is unwrapped, the liquid has to be ‘food safe’. Cooling using ice or cryogenic substances are essentially immersion or spray processes. Hydrocooling is probably the least expensive method of achieving rapid cooling in small products. The product to be cooled is immersed in, or sprayed with, cool water, either at ambient or near 0 °C. Practical systems vary from simple stirred or unstirred tanks to plants where the product is conveyed through agitated tanks or under banks of sprays. Small weight gains are often recorded during hydrocooling of unwrapped products. Hydrocooling is very effective for chilling fruit and vegetables; however, not all crops can tolerate wetting. Fruit can be cooled from 30 °C to 5 °C in 8 to 45 minutes, depending on the diameter of the fruit, without weight loss (Zerbini, 1990). Commercial hydrocoolers can treat 20– 30 tons/hour. Immersion chilling also has application with larger products. Most turkeys, and some chickens, are cooled by immersion in an ice/water mixture (slush ice) or chilled water. The carcasses can gain weight during the process, with the maximum © 2008, Woodhead Publishing Limited

386

Chilled foods

weight gain controlled by legislation. The carcasses have to be conveyed in the opposite direction to the water flow (counter current) to minimise the possibility of bacteria being transferred from one carcass to another (James et al., 2006a). Overhead conveyers, or Archimedean screws, are used to transport the carcasses through the immersion tank. Immersion chilling investigations have also been carried out with hot jointed pork primals in brine (Brown et al., 1988). Lower fluid temperatures, and consequently faster cooling rates, can be achieved and the joints are vacuum packed to minimise problems of brine uptake and cross contamination. However computer predictions show clearly (see Table 14.3) that the relative reduction in chilling time produced by immersion chilling decreases as product thickness is increased. Spray chilling of red meat carcasses is widely practised in the USA (James and Bailey, 1989) and used for poultry in Europe (Veerkamp, 1986). Spraying with ambient or chilled water is also an effective method of initially cooling cooked products that can withstand wetting, e.g. hams, sausages, chubs. Spray bars are fitted either in the batch cookers or in separate cooling cabinets. Ice Chilling with crushed ice or an ice/water mixture is simple, effective and commonly used for fish cooling. Cooling is more attributable to the contact between the produce and the cold melt water percolating through it (i.e. hydrocooling) than with the ice itself (Geeson, 1989). The individual fish are packed in boxes between layers of crushed ice, which extract heat from the fish and consequently melts. Ice has the advantage of being able to deliver a large amount of refrigeration in a short time, as well as maintaining a very constant temperature (–0.5 to 0 °C where sea water is present). The clearest disadvantage of crushed ice treatment is a considerable labour requirement, although automatic filling systems have been developed. Compressed air, the so-called champagne method, can be used to provide mixing of the fish, water and ice (Jul, 1986). To cut down on the amount of ice required, refrigerated sea water (RSW) and chilled sea water (CSW) systems have been developed for refrigeration of fish at sea. Both of these methods have the disadvantage of no longer guaranteeing the maintenance of 0 °C and of introducing the added expense of temperature control and recording (Jul, 1986). Ice may also used for cooling fruit and vegetables; however, as with immersion/ hydrocooling, ice is only applicable to vegetable and fruit produce that can tolerate wetting and low temperatures. Ice is frequently used by some vegetable producers to cool watercress, broccoli and other brassicas (Geeson, 1989). It is also often used to keep produce cool during transport. Cryogenic Direct spraying of liquid nitrogen onto a food product whilst it is conveyed through an insulated tunnel is one of the most commonly used methods of applying cryogens. Surface freezing can be a problem but an extra refrigeration effect is obtained by pre-cooling the food with the cold gas that results from the vaporisa© 2008, Woodhead Publishing Limited

Refrigeration, storage and transport of chilled foods

387

tion. Forced gas cooling is the only method that can be employed when surface freezing has to be avoided. However, the system tends to be inefficient. Post-harvest liquid nitrogen cooling of grapes has been developed in Brazil (Seger Gomes et al., 1992). In comparison with conventional ‘cooling room’ systems, liquid nitrogen cooling has been shown to significantly reduce cooling times (by between 57 and 76%), thus increasing production rates, whilst consuming significantly less energy. Cooling of solids and solid/liquid mixtures during cutting and mixing is increasingly common to prevent heating of products due to the mechanical movement of the mixing and cutting blades. Liquid nitrogen and liquid, or solid, carbon dioxide are commonly employed for such processes. Liquid nitrogen freezing improves protein breakdown, aiding the granulation of frozen meat and bacon fat (Hoffmanns and Buchmüller, 1994). The refrigerants are introduced directly through valves or spray bars, ensuring controlled distribution and cooling. The gases provide more even and quicker chilling than the ice traditionally used. Cryogenic refrigeration systems generally represent a low initial capital investment, but have high operating costs.

14.3.4 Vacuum Food products having a large surface area-to-volume ratio and an ability to readily release internal water are amenable to vacuum cooling. The products are placed in a vacuum chamber (typically operating at between 530 to 670 N m–2) and the resultant evaporative cooling removes heat from the food. Evaporative cooling is quite significant, the amount of heat released through the evaporation of 1 g of water being equivalent to that released in cooling 548 g of water by 1 °C. Suitable products, such as lettuce, can be vacuum cooled in less than 1 hour. In general terms, a 5 °C reduction in product temperature is achieved for every 1% of water that is evaporated. Since vacuum cooling requires the removal of water from the product, pre-wetting is commonly applied to prevent the removal of water from the tissue of the product. Pre-wetting is also useful in products that do not have a large surface area in proportion to mass. When pre-wetting, a thin film of water must be applied to the product since coarse droplets will evaporate preferentially, causing frost-burn spots on the product. Vacuum cooling is rapid and economical to operate because of low labour costs, but the capital cost of the large vacuum vessels is very high, and this has limited their widespread use. Despite this, increasingly large amounts of lettuce, celery, cauliflower, green peas and sweetcorn are vacuum cooled. Unlike with other systems, the type of packaging has very little effect on the rate of cooling, providing the packaging is not completely closed (Longmore, 1973; Geeson, 1989). Rapid rates of temperature reduction can be achieved in trays of cooked product such as mince, sauces and meat portions when cooled under vacuum. Larger products such as 10 kg cooked turkey carcasses can be cooled from 80 to 10 °C in less than 1 hour, but the rate of pressure reduction has to be carefully controlled if textural quality is to be maintained. © 2008, Woodhead Publishing Limited

388

Chilled foods

14.4 Storage Most unwrapped and wrapped chilled foods are stored in large refrigerated rooms with circulating air. To minimise weight loss and appearance changes associated with desiccation, air movement around unwrapped product should be the minimum required to maintain a constant temperature. With wrapped products, low air velocities are also desirable to minimise energy consumption. However, many storage rooms are designed and constructed with little regard to air distribution and localised velocities over products. Horizontal throw refrigeration coils are often mounted in the free space above the racks or rails of product and no attempt is made to distribute the air around the products. Using a false ceiling or other form of ducting to distribute the air throughout the storage room can substantially reduce variations in air distribution and temperature. Using air socks, a uniform air distribution and temperature can be maintained with localised velocities not exceeding 0.2 m s–1 (Dunne and Harvey, 1986). Jacketed storage rooms, where the cooling is incorporated into the walls, floor and ceiling of the room, produce very good temperature control in the enclosed space, with the minimum of air movement. This type of room is especially suitable for controlled atmosphere storage and for unwrapped product that are very sensitive to air movement or temperature fluctuations.

14.5 Transport Over a million refrigerated road vehicles, 400 000 refrigerated containers and many thousands of other forms of refrigerated transport systems are used to distribute refrigerated foods throughout the world (Gac, 2002). All these transportation systems are expected to maintain the temperature of the food within close limits to ensure its optimum safety and high quality shelf-life. Developments in temperature-controlled transportation systems for chilled products have led to the rapid expansion of the chilled food market. It is particularly important that the food is at the correct temperature before loading since the refrigeration systems used in most transport containers are not designed to extract heat from the load but merely to maintain its temperature. Irrespective of the type of refrigeration equipment used, the product will not be maintained at its desired temperature during transportation unless it is surrounded by air or surfaces at or below the maximum transportation temperature. This is usually achieved by a system that circulates moving air, either forced or by gravity, around the load. Inadequate air distribution is probably the principal cause of product deterioration and loss of shelf-life during transport. If products have been cooled to the correct temperature before loading and do not generate heat, then they have only to be isolated from external heat ingress. Surrounding them with a blanket of cooled air achieves this purpose. Care has to be taken during loading to stop any product touching the inner surfaces of the vehicle because this would allow heat ingress by conduction during transport. In the large containers used for © 2008, Woodhead Publishing Limited

Refrigeration, storage and transport of chilled foods

389

long distance transportation, food temperatures can be kept within ±0.5 °C of the set point. With this degree of temperature control, transportation times of 8 to 14 weeks (for vacuum packed meats stored at –1.5 °C) can be achieved and a sufficient chilled storage-life for retail display retained. Products such as fruits and vegetables that produce heat by respiration, or products that have to be cooled during transit, also require circulation of air through the product. This can be achieved by directing the supply air through ducts to channels at floor level or in the floor itself. In general, it is not advisable to rely on product cooling during transportation.

14.5.1 Sea transport Recent developments in temperature control, packaging and controlled atmospheres have substantially increased the range of foods that can be transported around the world in a chilled condition. Control of the oxygen and carbon dioxide levels in shipboard containers has allowed fruit and vegetables, such as apples, pears, avocados, melons, mangoes, nectarines, blueberries and asparagus, to be shipped (typically 40 days in the container) from Australia and New Zealand to markets in the USA, Europe, Middle East and Japan (Adams, 1988). If the correct varieties are selected and rapidly cooled immediately after harvest, the product arrives in good condition and has a long subsequent shelf-life. With conventional vacuum-packaging it is difficult to achieve a shelf-life in excess of 12 weeks with beef and 8 weeks for lamb (Gill, 1984). However, a shelf-life of up to 23 weeks at –2 °C can be achieved in cuts of lamb individually packed in evacuated bags of linear polyethylene, and then placed in gas-flushed foil laminate bags filled with a volume of CO2 approximately equal to that of the meat (Gill and Penney, 1986). Similar storage lives are currently being achieved with beef primals transported from Australia and South Africa to the EU. Most International Standard Organisation (ISO) containers are either ‘refrigerated’ or ‘insulated’. The refrigerated containers have refrigeration units built into their structure. The units operate electrically, either from an external power supply on board the ship or dock or from a generator on a road vehicle. Insulated containers either utilise the plug type refrigeration units already described or may be connected directly to an air-handling system in a ship’s hold or at the docks. Close temperature control is most easily achieved in containers that are placed in insulated holds and connected to the ship’s refrigeration system. However, suitable refrigeration facilities must be available for any overland sections of the journey. When the containers are fully loaded and the cooled air is forced uniformly through the spaces between cartons, the maximum difference between delivery and return air can be less than 0.8 °C (Heap, 1986). The entire product in a container can be maintained to within ±1.0 °C of the set point. Refrigerated containers are easier to transport overland than the insulated types, but have to be carried on deck when shipped because of problems in operating the refrigeration units within closed holds. On board ship they are therefore subjected © 2008, Woodhead Publishing Limited

390

Chilled foods

to much higher ambient temperatures and consequently larger heat gains, which make it far more difficult to control product temperatures.

14.5.2 Air transport Air-freighting is increasingly being used for high-value perishable products, such as strawberries, asparagus and live lobsters (Sharp, 1988; Stera, 1999). However, foods do not necessarily have to fall into this category to make air transportation viable since it has been shown that ‘the intrinsic value of an item has little to do with whether or not it can benefit from air shipment, the deciding factor is not price but mark-up and profit’ (ASHRAE, 1994). There was a 10 to 12% per year increase in the volume of perishables transported by air during in the 1990s (Stera, 1999). Although air-freighting of foods offers a rapid method of serving distant markets, there are many problems because the product is unprotected by refrigeration for much of its journey. Up to 80% of the total journey time is made up of waiting on the tarmac and transport to and from the airport. During flight, the hold is normally between 15 and 20 °C. Perishable cargo is usually carried in standard containers, sometimes with an insulating lining and/or ice or dry ice but is often unprotected on aircraft pallets (Sharp, 1988). Thus it is important that the product be: (i) transported in insulated containers to reduce heat gain; (ii) pre-cooled and held at the required temperature until loading; (iii) in containers filled to capacity; and (iv) accompanied by a thermograph for each consignment.

14.5.3 Land transport Overland transportation systems range from 12 m refrigerated containers for long distance road or rail movement of bulk chilled or frozen products, to small uninsulated vans supplying food to local retail outlets or even directly to the consumer. Some of the first refrigerated road and rail vehicles for chilled product were cooled by air that was circulated by free or forced systems, over large containers of ice (Ciobanu et al., 1976). Similar systems using solid carbon dioxide as the refrigerant have also been used for cooling of transport vehicles. In a 1970–71 survey of vehicles in the UK used to transfer chilled meat from small abattoirs to shops, almost 70% were unrefrigerated and 20% had no insulation (Cutting and Malton, 1972). However, the majority of current road transport vehicles for chilled foods are now refrigerated, using either mechanical, eutectic plates or liquid nitrogen cooling systems. The rise in supermarket home delivery services, where there are requirements for mixed loads of products that may each require different storage temperatures, is introducing a new complexity to local land delivery (Cairns, 1996). Mechanical units Many types of independent engine and/or electric motor driven mechanical refrigeration units are available for lorries or trailers. One of the most common is a self-contained ‘plug’ unit that mounts in an opening provided in the front wall of the vehicle. The condensing section is on the outside and the evaporator on the © 2008, Woodhead Publishing Limited

Refrigeration, storage and transport of chilled foods

391

inside of the unit, separated by an insulated section that fits into the gap in the wall. Units have one or two compressors, depending upon their capacity, which can be belt driven from the vehicle but are usually driven direct from an auxiliary engine. Total loss refrigerants Many advantages are claimed for liquid nitrogen transport systems, including minimal maintenance requirements, uniform cargo temperatures, silent operation, low capital costs, environmental acceptability, rapid temperature reduction and increased shelf-life due to the modified atmosphere (Smith, 1986). Long hauls can be carried out with containers capable of maintaining a chilled cargo at 3 °C for 50 hours after a single charge of liquid nitrogen. Overall costs are claimed to be comparable with mechanical systems (Smith, 1986). However, trials carried on the distribution of milk showed that the operating costs using liquid nitrogen, per 100 l of milk transported, were 2.2 times that of a mechanically refrigerated transport system (Nieboer, 1988). Fixed costs of the nitrogen system were under half that of the mechanical unit but the use of nitrogen was considered to be viable only at very low throughput. In general, total loss refrigerants such as liquid nitrogen (LN) or solid carbon dioxide (CO2), are increasingly being replaced by mechanical refrigeration in transport.

14.6 Retail display The temperature of individual consumer packs, small individual items and especially thin sliced products responds very quickly to small amounts of added heat. All these products are commonly found in retail display cabinets and marketing constraints require that they have maximum visibility. Maintaining the temperature of products below set limits while they are on open display in a heated store will always be a difficult task. Average temperatures in chill displays can vary considerably from cabinet to cabinet, with inlet and outlet values ranging from –6.7 to +6.0 °C, and –0.3 to +7.8 °C respectively in one survey (Lyons and Drew, 1985). The temperature performance of an individual display cabinet does not depend only on its design. Its position within a store and the way the products are positioned within the display area significantly influences product temperatures. In non-integral (remote) cabinets (i.e. those without built-in refrigeration systems), the design and performance of the store’s central refrigeration system is also critical to effective temperature control. The desired chilled display life for wrapped meat, fish, vegetables and processed foods ranges from a few days to many weeks and is primarily limited by microbiological considerations. Retailers of unwrapped fish, meat and delicatessen products (e.g. sliced meats, paté, cheese and prepared salads) normally require a display life of one working day, which is often restricted by appearance changes caused by desiccation during display rather than spoilage (though consideration of ‘at home shelf-life’ must also be accounted for). © 2008, Woodhead Publishing Limited

392

Chilled foods

14.6.1 Unwrapped products Display cabinets for delicatessen products are available with gravity or forced convection coils and the glass fronts may be nearly vertical or angled up to 20°. Sections through three of the commonest types of delicatessen cabinet are shown in Fig. 14.5. In the gravity cabinet (see Fig. 14.5a), cooled air from the raised rearmounted evaporator coil descends into the display well by natural convection and the warm air rises back to the evaporator. In the forced circulation cabinets (see Figures 14.5b and c), air is drawn through an evaporator coil by a fan and then ducted into the rear of the display, returning to the coil after passing directly over the products (Fig. 14.5b), or forming an air curtain (Fig. 14.5c), via a slot in the front of the cabinet and a duct under the display shelf. Changes in product appearance are normally the criteria that limit the displaylife of unwrapped foods, with the consumer selecting newly loaded product in preference to that displayed for some time. Deterioration in appearance has been related to degree of dehydration in red meat (see Table 14.4) and is likely to similarly occur in other foods. Apart from any relationship to appearance, weight loss is of considerable importance in its own right. A small survey carried out in the 1980s found average relative humidity ranged from 41 to 73% and air velocity from 0.1 to 0.67 m s–1 in delicatessen cabinets. The lowest rate of weight loss was measured in a cabinet of the type shown in Fig. 14.5c, which achieved mean conditions over the products of 0.4 °C, 0.14 m s–1 and 65% RH (James and Swain, 1986). The same study showed that relative humidity was more important than the air temperature or velocity. Reducing the RH from 95 to 40% increased weight loss over a 6 hour display period by a factor of between 14 and 18. In further work, a model developed to predict the rate of weight loss from unwrapped meat under the range of environmental conditions found in chilled retail displays showed that it was governed by the mean value of the conditions (James et al., 1988). Fluctuations in temperature or relative humidity had little effect on weight loss and any apparent effect was caused by changes in the mean conditions. There is a conflict between the need to make the display attractive and convenient to increase sales appeal and the optimum display conditions for the

Fan

(a)

(b)

Fig. 14.5

(c)

Three types of retail display cabinet for unwrapped products.

© 2008, Woodhead Publishing Limited

Coil

Refrigeration, storage and transport of chilled foods

393

Table 14.4 The relationship between evaporative weight loss and the appearance of sliced beef topside after 6 hours on display Evaporative loss (g cm–2)

Change in appearance

Up to 0.01 0.015–0.025 0.025–0.035 0.05 0.05–0.10

Red, attractive and still wet; may lose some brightness Surface becoming drier; still attractive but darker Distinct obvious darkening; becoming dry and leathery Dry, blackening Black

product. High lighting levels increase the heat load and the consequent temperature rise dehumidifies the refrigerated air. The introduction of humidification systems can significantly improve display life (Brown et al., 2005).

14.6.2 Wrapped products To achieve the display life of days to weeks required for wrapped chilled foods, the product should be maintained at a temperature as close to its initial freezing point as possible, to prevent microbial spoilage. In some cases, e.g. particular cheeses, dairy products and tropical fruits, quality problems may limit the minimum temperature that can be used, but for the majority of meat, fish and processed foods the range –1 to 0 °C is desirable. Air movement and relative humidity have little effect on the display life of a wrapped product, but the degree of temperature control can be important, especially with transparent, controlled-atmosphere packs. Large temperature cycles will cause water loss from the product and this water vapour will condense on the inner surface of the pack, consequently reducing consumer appeal. Although cabinets of the type described for delicatessen products can be used for wrapped foods, most are sold from multi-deck cabinets with single or twin air curtain systems (Fig. 14.6). Twin air curtains tend to provide more constant product temperatures but are more expensive. It is important that the front edges of the cabinet shelves do not project through the air curtain since the refrigerated air will then be diverted out of the cabinet. On the other hand, if narrow shelves are used, the curtain may collapse and ambient air can be drawn into the display well. To maintain product temperatures close to 0 °C, the air off the coil must typically be –4 °C, and any ingress of humid air from within the store will quickly cause the coil to ice up. Frequent defrosts are often required, and even in a well maintained unit the cabinet temperature may rise to 10 to 12 °C and the product by at least 3 °C (Brolls, 1986). External factors such as the store ambient temperature, the position of the cabinet, and poor pre-treatment and placement of products substantially affect cabinet performance. Warm and humid ambient air, and loading with insufficiently cooled products can also overload the refrigeration system. Even if the food is at its correct temperature, uneven loading or too much product can disturb the airflow patterns and destroy the insulating layer of cooled air surrounding the product. An in-store survey of 299 pre-packaged meat products © 2008, Woodhead Publishing Limited

394

Chilled foods

Fig. 14.6

Multi-deck display cabinets (open and enclosed) for wrapped products.

in chilled retail displays found product temperatures in the range –8.0 to 14.0 °C, with a mean of 5.3 °C and 18% above 9 °C (Rose, 1986). Other surveys (BøghSørensen, 1980; Malton, 1971) have shown that temperatures of packs from the top of stacks were appreciably higher than those from below, due to radiant heat pick up from store and cabinet lighting. It has also been stated that products in transparent film overwrapped packs can achieve temperatures above that of the surrounding refrigerated air due to radiant heat trapped in the package by the ‘greenhouse’ effect. However, specific investigations have failed to demonstrate this effect (Gill, 1988).

14.7 Specifying refrigeration systems There are three stages in obtaining a chilling system that works: (i)

Determining the process specification, i.e. specifying exactly the condition of the product when it enters the system and what state it is required to be in when it exits, for the throughput capacity needed in tonnes/h, etc. (ii) Drawing up the engineering specification, i.e. turning processing conditions into terms that a refrigeration engineer can understand, independent of the food process. (iii) Procuring and commissioning the total system, including any services or utilities. Poor performance of new chillers can often be traced back to a poor, non-existent, or ambiguous process specification. In older systems, it is often due to a change in use that was not considered in the original specification. The first task in designing a plant is therefore the preparation of a clear specification by the user of how the facility will be used now, and in the foreseeable © 2008, Woodhead Publishing Limited

Refrigeration, storage and transport of chilled foods

395

future. In preparing this specification, the user should consult with all parties concerned: these may include officials enforcing legislation, customers, other departments within the company and engineering consultants or contractors – but the ultimate decisions taken in forming this specification are the user’s alone. The process specification must include, as a minimum, data on the product(s) to be handled, in terms of size, shape and throughput. The maximum capacity must be catered for and the chiller should also be specified to chill adequately and economically at all other throughputs. The range of temperature requirements for each product must also be clearly stated. If it is intended to maximise yield during chilling, it is useful to quantify at an early stage how much extra money can be spent to save a given amount of weight. All the information collected so far, and the decisions taken, will be on existing production. Another question that needs to be asked is, ‘will there be any changes in the use of the chiller in the future?’ Chilling is one operation in a sequence of operations. It influences the whole system and interacts with it. An idea must be obtained of how the system will be loaded, unloaded and cleaned, and these operations must always be intimately involved with those of the rest of the operation. There is often a conflict of interest within the usage of a chiller. In practice, a chiller can often be used as a marshalling yard for sorting orders and as a place for storing product not sold. If it is intended that either of these operations are to take place in the chiller, the design must be made much more flexible in order to cover the conditions needed in a marshalling area or a refrigerated store. In the case of a batch and semi-continuous operation, holding areas will be required at the beginning and end of the process in order to even-out flows of material from adjacent processes. The time available for the process will be, in part, dictated by the space that is available; a slow process will take more space than a fast process, for a given throughput. Other refrigeration loads in addition to that caused by the input of heat from the product also need to be specified. Many of these, such as infiltration through openings, the use of lights, machinery, and people working in the refrigerated space, are all under the control of the user and must be specified so that the heat load given off by them can be incorporated in the final design. Ideally, all the loads should then be summed together on a time basis to produce a load profile. If the refrigeration process is to be incorporated with all other processes within a plant, in order to achieve an economic solution, then the load profile is important. The ambient design conditions must be specified. This means that the temperatures adjacent to the refrigerated equipment and the temperatures of the ambient to which heat will ultimately be rejected. In stand-alone refrigerated processes this will often be the wet and dry bulb temperatures of the outside air. If the process is to be integrated with heat reclamation, then the temperature of the heat sinks must be specified. Finally, the defrost regime should also be specified. There are times in any process where it is critical that coil defrosting and its accompanying temperature rise does not take place, and that the coil is cleared of frost before commencing the specified part of the process. Although it is common practice throughout the food industry to leave much of this specification to refrigeration contractors or engineering specialists, the end © 2008, Woodhead Publishing Limited

396

Chilled foods

user should specify all the above requirements. The refrigeration contractors or engineering specialists are in a position to give good advice on this. However, since all the above are outside their control, the end user, with their knowledge of how well they can control their overall process, should always take the final decision. The aim of drawing up an engineering specification is to turn the user requirements into a specification that any refrigeration engineer can then use to design a system. The first step in this process is iterative. First, a full range of time, temperature and air velocity options must be assembled for each cooling specification covering the complete range of each product. Each must then be evaluated against the user requirements. If they are not a fit, then another option is selected and the process repeated. If there are no more options available there are only two alternatives; either standards must be lowered (recognising in do so that cooling specifications will not be met), or the factory operation must be altered. A full engineering specification will typically include: the environmental conditions within the refrigerated enclosure, air temperature, air velocity and humidity; the way the air will move within the refrigerated enclosure; the size of the equipment; the refrigeration load profile; the ambient design conditions; and the defrost requirements. The final phase of the engineering specification should be drawing up a schedule for testing the engineering specification prior to handing over the equipment. This test will be in engineering and not product terms. The specification produced should be the document that forms the basis for quotations and finally the contract between the user and his contractor, and it must be stated in terms that are objectively measurable once the chiller/freezer is completed. Arguments often ensue between contractors and their clients from an unclear, ambiguous or unenforceable specification. Such lack of clarity is often expensive to all parties and should be avoided.

14.8 Modelling and simulation to improve cold-chain management There are relatively few user-friendly computer programs for modelling heat and mass transfer in foods in the cold-chain presently available to the food industry, and here are a few examples:

• Heatsolv, developed by Dr Paul Nesvadba (http://www.rubislawconsulting.com)

•

predicts temperatures in foods using a finite difference solution of the heat conduction equation for the following shapes: slab, cylinder, sphere and ‘fractal shape’ such as a fish – somewhere between a cylinder and a slab. The food thermal properties may vary with temperature and may include phase change. Food Product Modeller, developed by Mirinz, Food Technology and Research Ltd, New Zealand; available from Dever Associates (http://www.dever.com.au/ fpm/food.htm). This also uses a finite difference solution that ‘enables food industry engineers to design and evaluate chilling processes for almost any food

© 2008, Woodhead Publishing Limited

Refrigeration, storage and transport of chilled foods

•

•

•

397

product – whether fish, dairy, fruit, vegetables, meat or processed foods. It incorporates menu options for meat industry objects such as lamb carcasses, beef sides, and cartoned products.’ ColdTime, developed by TechniSolve Software cc, South Africa (info@ coolit.co.za), calculates cooling times of food products using ‘methods outlined in the ASHRAE 2002 Refrigeration Handbook’. A further program from the same company, ColdRoom (not to be confused with the program of the same name developed by FRPERC) calculates product temperatures and the effects of internal loads, insulation, etc. in a cold room. ColdRoom (developed by FRPERC, University of Bristol (frperc.bris.ac.uk)) allows ‘cold room operators, contractors, and manufacturers to specify and design cold rooms to keep food at optimum temperatures under actual working conditions. Users to rapidly predict the effect of operating conditions and loading patterns on performance and identify how they can avoid unacceptable food temperatures’. Refrigeration Loads Analyser, developed by Mirinz, Food Technology and Research Ltd, New Zealand; available from Dever Associates, http://www.dever. com.au/fpm/food.htm) ‘enables food industry engineers to calculate refrigeration heat loads and evaluate refrigeration process alternative’.

Both of the programs called ColdRoom were developed separately but share similar goals and purposes. ColdRoom by TechniSolve was developed mainly for the designing of cold rooms, while Coldroom by FRPERC was also designed to allow users to predict the effect of different operating regimes. Many of these programs equate food products to simple shapes, such as infinite slabs, infinite cylinders, finite bricks, finite cylinders and spheres, and utilise finite difference techniques. Modelling of complex shapes can be accomplished with finite element programs, such as ANSYS; however, these require time and detailed knowledge to be used effectively and are far from user-friendly. Computational fluid dynamics (CFD) software that has been used by the aerospace, automotive, and chemical industries (e.g. COMSOL (formerly FLUENT), CFX, FIDAP, MARC, PHOENICS, FLOW3D, CFDS, etc.) is also being increasingly used by refrigeration manufacturers to design chilling systems.

14.8.1 Chilling models Heat and mass transfer models for predicting chilling processes have been recently reviewed by Cleland (1990), Pham (2001) and Wang and Sun (2003). Time–temperature charts for infinite slabs, infinite cylinders and spheres can be found in Singh and Heldman (1993) and ASHRAE Fundamentals (2001). These are derived from numerical methods of unsteady state heat conduction and can be used to gain a quick answer for chilling times by assuming simple shapes and constant thermal properties with temperature. To avoid the inaccuracies of manually reading off a chart, the equations used to generate the information on the charts can be put onto a computer. © 2008, Woodhead Publishing Limited

398

Chilled foods

Empirically based diagrams have been published by a number of researchers that allow a first approximation of temperature profiles and of chilling time at any chilling conditions. Some of this data has been supplemented by numerically modelled data. Many of these are for meat carcasses: beef (James and Bailey, 1989), pork (Brown and James, 1992), lamb and mutton (Earle and Fleming, 1967). Similar systems have also been used for the cooling of fish in ice (Jain et al., 2005), and the cooling of trays of ready meals in air (Evans et al., 1996) and immersion systems (Ketteringham and James, 1999). In the last two cases, finite difference methods were used to extrapolate the experimental data to a wider range of products and chilling conditions. Finite difference techniques have also been used to model the immersion chilling of foods (Zorrilla and Rubiolo, 2005a, b). Finite element modelling has been used to study heat and mass transfer in the chilling of chicken in a continuous system (Landfeld and Houska, 2006). A threedimensional explicit finite difference mathematical model was developed to investigate air-impingement cooling of food in the shape of finite slabs (Erdogdu et al., 2005). There has been relatively little use of CFD to model chilling. It has been used to model cooling times of cooked meat joints in an air blast-chilling process (Hu and Sun, 2001) and beef cooling (Trujillo and Pham, 2003), while Mirade et al. (2002) have looked at its use in designing large food chillers. Several models have been developed specifically for vacuum cooling of liquid foods (Burfoot et al., 1989; Houska et al., 1996; Wang and Sun, 2003; Dostal and Petera, 2004) and cooked meat (Wang and Sun, 2002a,b).

14.8.2 Transport models The purpose of refrigerated distribution is to maintain the temperature of the food and this appears to have attracted less attention from modellers than chilling, freezing or thawing (James et al., 2006b). However, there are substantial difficulties in maintaining the temperature of refrigerated foods transported in small-refrigerated vehicles that conduct multi-drop deliveries to retail stores and caterers. The design of the refrigeration system has to allow for extensive differences in load distribution, dependent on different delivery rounds, days of the week and the removal of product during a delivery run (Tso et al., 2002). In the UK, a predictive program called CoolVan has been produced to aid the design and operation of small-refrigerated delivery vehicles (Gigiel, 1998). There are stages in transportation where food is not in a refrigerated environment, e.g. in loading bays, in supermarkets before loading into retail displays, and in domestic transportation from shop to home. Some food transportation models have looked at temperature rises in pallet loads of chilled or frozen food during distribution. They often specifically look at the times during loading, unloading and temporary storage when the pallets may be in a warm ambient that may cause the food temperature to rise (Bennahmias et al., 1997; Stubbs et al., 2004). © 2008, Woodhead Publishing Limited

Refrigeration, storage and transport of chilled foods

399

14.8.3 Storage models In the UK, the FRPERC predictive program ColdRoom has been produced, as part of the LINK project, to aid the design and operation of cold rooms (see Section 14.8). CFD studies have been carried out on airflow in cold stores (Said et al., 1995; Mariotti et al., 1995; Hoang et al., 2000; Nahor et al., 2005) and through cold store doorways (Foster et al., 2002), while Tanner et al. (2002) have developed models for heat and mass transfer in packaged horticultural products during storage.

14.8.4 Retail display models The final stages of the refrigerated food chain, retail display and domestic refrigeration, have also attracted the attention of modellers. Again CFD and other numerical techniques have been used to model air movement in retail display cabinets (Van Ort and Van Gerwen, 1995; Lan et al., 1996; Stribling et al., 1997; Navaz et al., 2002; Cortella, 2002).

14.9 Conclusions In general, after initial chilling, as a chilled product moves along the cold-chain it becomes increasingly difficult to control and maintain its temperature. Temperatures of bulk packs of chilled product in large storerooms are far less sensitive to small heat inputs than single consumer packs in transport or open display cases. If primary and secondary cooling operations are efficiently carried out, then the food will be reduced below its required temperature before it is placed in storage. In this situation the cold-store’s refrigeration system is required only to extract extraneous heat that enters through the walls, door openings, etc. Even when temperature-controlled dispatch bays are used, there is a slight heat pick up during loading. In bulk transportation the resulting temperature rise is small and the vehicle’s refrigeration system rapidly returns the product to the required temperature. Larger problems exist in local multi-drop distribution to individual stores. There is a large heat input every time the doors are opened and product unloaded; small packs rapidly rise in temperature and the vehicle often lacks the refrigeration capacity or time to re-cool the food. Temperature control during retail display is often poor due to the retailers’ need to display as much product as possible in a way that is very assessable to the consumer. Increasing energy costs may be the key factor that persuades retailers to reduce consumer access and hence improve temperature control.

© 2008, Woodhead Publishing Limited

400

Chilled foods

14.10 References ADAMS G R (1988), ‘Controlled atmosphere containers’, Refrigeration for Food and People,

Meeting of IIR Commissions C2, D1, D2/3, E1, Brisbane (Australia), 244–248. (1994), ASHRAE Handbook 1994: Refrigeration: Systems and Applications, American Society of Heating, Refrigeration and Air-Conditioning Engineers, Inc., Atlanta, GA, USA. ASHRAE (2001), ASHRAE Handbook 2001: Fundamentals, American Society of Heating, Refrigeration and Air-Conditioning Engineers, Inc., Atlanta, GA, USA. BENNAHMIAS R, GABORIAU R AND MOUREH J (1997), ‘The insulating cover, a particular logistic means for thermo-sensitive foodstuffs’, International Journal of Refrigeration, 20(5), 359–366. BØGH-SØRENSEN L (1980), ‘Product temperatures in chilled cabinets’, Proceedings 26th European Meeting of Meat Research Workers, Colorado Springs (USA), n.22. BROLLS E K (1986), ‘Factors affecting retail display cases’, Recent advances and developments in the refrigeration of meat chilling, Meeting of IIR Commission C2, Bristol (UK), Section 9, 405–413. BROWN T AND JAMES S J (1992), ‘Process design data for pork chilling’, International Journal of Refrigeration, 15(5), 281–289. BROWN T, CORRY J AND JAMES S J (2005), ‘Humidification of chilled fruit and vegetables on retail display using an ultrasonic fogging system with water/air ozonation’, International Journal of Refrigeration, 27(8), 862–868. BROWN T, GIGIEL A J, SWAIN M V L AND HIGGINS J A (1988), ‘Immersion chilling of hot cut, vacuum packed pork primals’, Meat Science, 22, 173–188. BURFOOT D, HAYDEN R AND BADRAN R (1989), ‘Simulation of a pressure cook/water and vacuum cooled processing system’, in Field R W and Howell J, Process Engineering in the Food Industry: Developments and Opportunities, London, Elsevier Applied Science, 27–41. BURFOOT D, HAYDON R AND BADRAN R (1987), ‘Simulation of a pressure cook/water and vacuum cooled processing system’, Engineering Innovation in the Food Industry, Proceedings of IChemE symposium, Bath (UK), 231–242. CAIRNS S (1996), ‘Delivering alternatives: Success and failures of home delivery services for food shopping’, Transport Policy, 3, 155–176. CIOBANU A, LASCU G, BERCESCU V AND NICULESCU L (1976), Cooling Technology in the Food Industry, Tunbridge Wells, Abacus Press. CLELAND A C (1990), Food Refrigeration Processes – Analysis, Design and Simulation, London, Elsevier Applied Science. CORTELLA G (2002), ‘CFD aided retail cabinets design’, Computers and Electronics in Agriculture, 34, 43–66. CUTTING C L AND MALTON R (1972), ‘Recent observations on UK meat transport’, Proceedings of the Meat Research Institute Symposium No. 2. Meat Chilling – Why and How?, Bristol (UK), 24.1–24.11. DOSTAL M AND PETERA K (2004), ‘Vacuum cooling of liquids: Mathematical model’, Journal of Food Engineering, 61(4), 533–539. DUNNE W A AND HARVEY R D (1986), ‘The use of air permeable ducting in the storage and processing of meat and meat products’, Recent advances and developments in the refrigeration of meat chilling, Meeting of IIR Commission C2, Bristol (UK), Section 4, 227–234. EARLE R L AND FLEMING K A (1967), ‘Cooling and freezing of lamb and mutton carcasses: 1- Cooling and freezing rates in legs’, Food Technology, 21, 79–84. ERDOGDU F, SARKAR A AND SINGH R P (2005), ‘Mathematical modelling of air-impingement cooling of finite slab shaped objects and effect of spatial variation of heat transfer coefficient’, Journal of Food Engineering, 71(3), 287–294. ASHRAE

© 2008, Woodhead Publishing Limited

Refrigeration, storage and transport of chilled foods

401

EVANS J, RUSSELL S AND JAMES S J (1996), ‘Chilling of recipe dish meals to meet cook–chill

guidelines’, International Journal of Refrigeration, 19, 79–86. (1986), ‘Continuous chilling of meat products’, Recent Advances and Developments in the Refrigeration of Meat Chilling, Meeting of IIR Commission C2, Bristol (UK), Section 4, 179–191. FOSTER A M, BARRETT R, JAMES S J AND SWAIN M J (2002), ‘Measurement and prediction of air movement through doorways in refrigerated rooms’, International Journal of Refrigeration, 25, 1102–1109. GAC A (2002), ‘Refrigerated transport: What’s new?’, International Journal of Refrigeration, 25, 501–503. GEESON J D (1989), ‘Cooling and storage of fruits and vegetables’, The Proceedings of the Institute of Refrigeration, 1988–89, 85, 1–10. GIGIEL A (1998), ‘Modelling the thermal response of foods in refrigerated transport’, Refrigerated Transport, Storage and Retail Display, Meeting of IIR Commission D1,D2/ 3 Cambridge (UK), Paris: International Institute of Refrigeration, Section 1, 61–69. GILL C O AND PENNEY N (1986), ‘Packaging of chilled red meats for shipment to remote markets’, Recent Advances and Developments in the Refrigeration of Meat Chilling, Meeting of IIR Commission C2, Bristol (UK), Section 10, 521–525. GILL C O (1984), ‘Longer shelf life for chilled lamb’, 23rd New Zealand Meat Industry Research Conference, Hamilton, NZ. GILL J (1988), ‘The greenhouse effect’, Food, (April) 47, 49, 51. GOSNEY W B (1982), Principles of Refrigeration, Cambridge University Press, UK. HEAP R D AND MANSFIELD J E (1983), ‘The use of total loss refrigerants in transport of foodstuffs’, Australian Refrigeration, Air Conditioning and Heating, 37(2), 23–26. HEAP R D (1986), ‘Container transport of chilled meat’, Recent Advances in the Refrigeration of Chilled Meat, Meeting of IIR Commission C2, Bristol (UK), 505–510. HOANG M L, VERBOVEN P, DE BAERDEMAEKER J AND NICOLAI B M (2000), ‘Analysis of the air flow in a cold store by means of computational fluid dynamics’, International Journal of Refrigeration, 23, 127–140. HOFFMANNS W AND BUCHMÜLLER J (1994), ‘Chilling and freezing with cryogenic gases’, Fleischwirtschaft, 74(8), 845–846. HOFFMANNS W (1994), ‘Chilling, freezing and transport: Refrigeration applications using the cryogenic gases liquid nitrogen and carbonic acid’, Fleischwirtschaft, 72(12), 1309–1311. HOUSKA M, PODLOUCKY S, ZITNY R, GREE R, SESTAK J, DOSTAL M AND BURFOOT D (1996), ‘Mathematical model of the vacuum cooling of liquids’, Journal of Food Engineering, 29(3–4), 339–348. HU Z AND SUN D W (2001), ‘Effect of fluctuation in inlet airflow temperature on CFD simulation of air-blast chilling process’, Journal of Food Engineering, 48, 311–316. JAIN D, ILYAS S M, PATHARE P, PRASAD S AND SINGH H (2005), ‘Development of mathematical model for cooling the fish with ice’, Journal of Food Engineering, 71(3), 324–329. JAMES C, VINCENT C, DE ANDRADE LIMA T I AND JAMES S J (2006a), ‘The primary chilling of poultry carcasses – A review’, International Journal of Refrigeration, 29(6), 847–862. JAMES S J AND BAILEY C (1989), ‘Process design data for beef chilling’, International Journal of Refrigeration, 12, 42–49. JAMES S J AND SWAIN M V L (1986), ‘Retail display conditions for unwrapped chilled foods’, Proceeding of the Institute of Refrigeration, 1985–1986, 82, 1–7. JAMES S J (1990), ‘Cooling of cooked products’, Progress in the Science and Technology of Refrigeration in Food Engineering, Meeting of IIR Commissions B2, C2, D1, D2/3, Dresden (GDR), Section 8, 551–557. JAMES S J, JAMES C AND EVANS J A (2006b), ‘Modelling of food transportation systems – A review’, International Journal of Refrigeration, 29(6), 947–957. JAMES S J, SWAIN M V L AND DAUDIN J D (1988), ‘Mass transfer under retail display conditions’, 34th International Congress Meat Science Technology, Brisbane (Australia), 652–654. EVERINGTON D W AND SAGOO L S

© 2008, Woodhead Publishing Limited

402

Chilled foods

(1986), ‘Chilling and freezing fishery products: Changes in views and usages’, International Journal of Refrigeration, 9, 174–178. KETTERINGHAM L AND JAMES S J (1999), ‘Immersion chilling of trays of cooked products’, Journal of Food Engineering, 40, 256–267. LAN T H, GOTHAM D H T AND COLLINS M W (1996), ‘A numerical simulation of the air flow and heat transfer in a refrigerated food display cabinet’, 2nd European Thermal Sciences and 14th UIT National Heat Transfer Conference, 1139–1146. LANDFELD A AND HOUSKA M (2006), ‘Prediction of heat and mass transfer during passage of the chicken through the chilling tunnel’, Journal of Food Engineering, 72(1), 108–112. LONGMORE A P (1973), ‘The pros and cons of vacuum cooling’, Food Industries of South Africa, May, 6–11. LYONS H AND DREW K (1985), ‘A question of degree’, Food, December, 15–17. MACLEOD-SMITH R I, VAN ESPEN J AND MAGER G (1994), ‘Modern practices in wet air cooling for pre-cooling and storage of fresh produce’, Proceedings of the Institute of Refrigeration, 1993–94, 90, 85–92. MALTON R (1971), ‘Some factors affecting temperature of over-wrapped trays of meat in retailers display cabinets’, Proceedings 17th European Meeting of Meat Research Workers, Bristol (UK), J2. MARIOTTI M, RECH G AND ROMAGNONI P (1995), ‘Numerical study of air distribution in a refrigerated room’, Proceedings of the 19th International Conference of Refrigeration, August 20–25; The Hague, The Netherlands, 98–105. MIRADE P S, KONDJOYAN A AND DAUDIN J D (2002), ‘Three-dimensional CFD calculations for designing large food chillers’, Computers and Electronics in Agriculture, 34(1–3), 67– 88. NAHOR H B, HOANG M L, VERBOVEN P, BAELMANS M AND NICOLAI B M (2005), ‘CFD model of the airflow, heat and mass transfer in cool stores’, International Journal of Refrigeration, 28(3), 368–380. NAVAZ H K, FARAMARZI R, GHARIB M, DABIRI D AND MODARESS D (2002), ‘ The application of advanced methods in analyzing the performance of the air curtain in a refrigerated display case’, Journal of Fluids Engineering – Trans ASME, 124(3), 756–764. NIEBOER H (1988), ‘Distribution of dairy products’, Cold-chains in Economic Perspective, Meeting of IIR Commission C2, Wageningen (The Netherlands), 16.1–16.9. PERSSON P O AND LÖNDAL G (1993), ‘Freezing technology’, in Mallett C P, Frozen Food Technology, London, Chapman and Hall, 20–58. PHAM Q T (2001), ‘Modelling thermal processes: Cooling and freezing’, in Tijskens L M M, Hertog M L A T M, Nicolaï B M, Food Process Modelling, Cambridge, Woodhead Publishing Limited, 15, 312–339. ROBERTSON G H, CIPOLLETTI, J C, FARKAS, D F AND SECOR G E (1976), ‘Methodology for direct contact freezing of vegetables in aqueous freezing media’, Journal of Food Science, 41, 845–851. ROSE S A (1986), ‘Microbiological and temperature observations on pre-packaged ready-toeat meats retailed from chilled display cabinets’, Recent Advances and Developments in the Refrigeration of Meat Chilling, Meeting of IIR Commission C2, Bristol (UK), Section 9, 463–469. SAID M N A, SHAW C Y, ZHANG J S AND CHRISTIANSON L (1995), ‘Computation of room air distribution’, ASHRAE Transactions, 101, 1065–1077. SEGER GOMES T C, FELIX BEIRÁO N A AND TARARES DE ANDRADE W (1992), ‘Comparison between cryogenic and conventional systems for post-harvest grape chilling’, Proposals for the Generation and Use of Refrigeration in the 21st Century, Meeting of IIR Commissions B2, C2 and E2, Buenos Aires (Argentina), Session 1, 332–338. SHARP A K (1988), ‘Air freight of perishable product’, Refrigeration for Food and People, Meeting of IIR Commissions C2, D1, D2/3, E1, Brisbane (Australia), 219–224. SINGH R P AND HELDMAN D R (1993), Introduction to Food Engineering, 2nd Ed. Oxford, Clarendon Press. JUL M

© 2008, Woodhead Publishing Limited

Refrigeration, storage and transport of chilled foods

403

SMITH B K (1986), ‘Liquid nitrogen in-transit refrigeration’, Recent Advances and Develop-

ments in the Refrigeration of Meat Chilling, Proceedings of the conference of IIR Commission C2, Bristol, UK, 383–390. STERA A C (1999), ‘Long distance refrigerated transport into the third millennium’, 20th International Congress of Refrigeration, IIF/IIR Sydney, Australia, Paper 736. STRIBLING D, TASSOU S A AND MARRIOT D (1997), ‘A two dimensional CFD model of a refrigerated display case’, ASHRAE Transactions: Research, 103(1), 88–95. STUBBS D M, PULKO S H AND WILKINSON A J (2004), ‘Wrapping strategies for temperature control of chilled foodstuffs during transport’, Transactions of Instrumental Measurement and Control, 26(1), 69–80. TANNER D J, CLELAND A C, OPARA L U AND ROBERTSON T R (2002), ‘A generalised mathematical modelling methodology for design of horticultural food packages exposed to refrigerated conditions: Part 1, Formulation’, International Journal of Refrigeration, 25(1), 33–42. TROTT A R (1989), Refrigeration and Air-conditioning, London, Butterworths. TRUJILLO F J AND PHAM Q T (2003), ‘Modelling the chilling of the leg, loin and shoulder of beef carcasses using an evolutionary method’, International Journal of Refrigeration, 26(2), 224–231. TSO C P, YU S C M, POH H J, JOLLY P G (2002), ‘Experimental study on the heat and mass transfer characteristics in a refrigerated truck’, International Journal of Refrigeration, 25, 340–350. VAN ORT H AND VAN GERWEN R J M (1995), ‘Air flow optimisation in refrigerated cabinets’, Proceedings of 19th International Congress of Refrigeration, 446–453. VEERKAMP C H (1986), ‘Control of weight loss by evaporative air chilling’, Recent Advances and Developments in the Refrigeration of Meat Chilling, Meeting of IIR Commission C2, Bristol (UK), Section 4, 153–158. VEISSEYRE R (1986), Techniques Laitieres, 2nd Ed., La Maison Rustique, Paris (France). WANG L AND SUN D-W (2002a), ‘Modelling vacuum cooling process of cooked meat – Part 1: Analysis of vacuum cooling system’, International Journal of Refrigeration, 25(7), 854–861. WANG L AND SUN D-W (2002b), ‘Modelling vacuum cooling process of cooked meat – Part 2: Mass and heat transfer of cooked meat under vacuum pressure’, International Journal of Refrigeration, 25(7), 862–871. WANG L AND SUN D-W (2003), ‘Recent developments in numerical modelling of heating and cooling processes in the food industry – A review’, Trends in Food Science and Technology, 14, 408–423. ZERBINI P E (1990), ‘Chilling of top and citrus fruit’, in Zeuthen P, Cheftel, J C, Eriksson C, Gormley T R, Linko P and Paulus K, Processing and Quality of Foods, Volume 3, Chilled Foods: The Revolution in Freshness, London, Elsevier Applied Science Publishers, 3.336–3.355. ZORRILLA S E AND RUBIOLO A C (2005a), ‘Mathematical modelling for immersion chilling and freezing of foods: Part i: Model development’, Journal of Food Engineering, 66(3), 329–338. ZORRILLA S E AND RUBIOLO A C (2005b), ‘Mathematical modelling for immersion chilling and freezing of foods: Part ii: Model solution’, Journal of Food Engineering, 66(3), 339– 351.

© 2008, Woodhead Publishing Limited

15 Temperature monitoring and measurement J. A. Evans, Food Refrigeration and Process Engineering Research Centre, UK and M. L. Woolfe, Food Standards Agency, UK*

15.1 Introduction The safety and quality of chilled foods in the cold chain from harvest or slaughter to consumption relies upon the food being maintained at a sufficiently low temperature to prevent growth of bacterial pathogens and to minimise growth of spoilage microbes. Consequently, temperature measurement and monitoring is an integral part of any food management system as well as being, in many areas of the cold chain, a legislative requirement. Without means to monitor and measure temperatures within the food cold chain, the safety and quality of food can be compromised. Therefore, the accuracy of temperature measurement and the interpretation of data are paramount in the management of the cold chain.

15.2 Legislation 15.2.1 Temperature control regulations EU legislation applied on 1 January 2006 has introduced a raft of new regulations that introduce a risk-based and traceable ‘farm to fork’ approach to food safety. The legislation aims to: *The views expressed in this chapter are those of the author and should not be regarded as a statement of official Government policy.

© 2008, Woodhead Publishing Limited

Temperature monitoring and measurement

405

• modernise, consolidate and simplify previous EU food hygiene legislation; • apply effective and proportionate control through the food chain, from primary production to sale or supply to the final consumer;

• focus controls on what is necessary for public health protection; • clarify that it is the primary responsibility of food business operators to produce food safely. This new legislation requires all food business operators (including farmers and growers) to set up, implement and maintain procedures based on HACCP (hazard analysis critical control point) principles and is structured to be applied in a flexible manner according to the size of the food business. The main regulations of relevance to the cold chain are:

• Regulation (EC) 852/20041 on the hygiene of foodstuffs. • Regulation (EC) 853/20042 laying down specific hygiene rules for food of animal origin.

• Regulation (EC) 854/20043 laying down specific rules for the organisation of official controls on products of animal origin intended for human consumption.

• Directive 2004/41 that repeals the previous EU legislation or, in some cases, amends still existing legislation.

• Directive 2002/99 laying down the animal health rules on products of animal origin for human consumption. The main temperature control requirements for food of animal origin that specify temperature control are covered within Regulation (EC) 853/2004 (Table 15.1). Furthermore, less specific requirements are contained in Regulation (EC) 852/ 2004, Annex II, Chapter I, Paragraph 2(d), Chapter III Paragraph 2(g), Chapter IV Paragraph 7, Chapter V Paragraph (2) and Chapter IX Paragraphs (2), (5), (6) and (7). These state that adequate facilities must be available to maintain foodstuffs at appropriate temperatures and allow those temperatures to be monitored and, where necessary, recorded. The Regulation states that ‘the cold chain is not to be interrupted’ and that foods should be ‘cooled as quickly as possible following the heat-processing stage, or final preparation stage if no heat process is applied’. In addition, ‘thawing of foodstuffs is to be undertaken in such a way as to minimise the risk of growth of pathogenic micro-organisms or the formation of toxins in the foods. During thawing, foods are to be subjected to temperatures that would not result in a risk to health.’ ‘However, limited periods outside temperature control are permitted, to accommodate the practicalities of handling during preparation, transport, storage, display and service of food, provided that it does not result in a risk to health.’ To apply this legislation nationally in the UK, a Statutory Instrument (SI) in England, and equivalent legislation in Scotland, Wales and Northern Ireland, has been drawn up. These cover offences, penalties and powers of entry, revocation of existing implementing legislation, enacting the national measures required or provided for in the EU regulations and any consequential amendments (where the revocation of existing legislation requires changed references in other pieces of © 2008, Woodhead Publishing Limited

Food

Temperature

Primary chilling

Red meat

Immediately chilled post slaughter to ensure a temperature throughout the meat of not more than 3 °C for offal and 7 °C for other meat along a chilling curve that ensures a continuous decrease of the temperature

Storage and transport*

Red meat

3 °C for offal and 7 °C for other meats

Cutting, boning, trimming, slicing, dicing, wrapping and packaging

Red meat

Not more than 3 °C for offal and 7 °C for other meat, by means of an ambient temperature of not more than 12 °C or an alternative system having an equivalent effect

Cutting, boning, trimming, slicing, dicing, wrapping and packaging

Poultry and lagomorphs

Not more than 4 °C by means of an ambient temperature of 12 °C or an alternative system having an equivalent effect

Post production

Minced meat/meat preparations

Chilled to internal temperature of not more than 2 °C for minced meat and 4 °C for meat preparations

Post production

MSM

Chilled to a temperature of not more than 2 °C

Storage, transport

Rendered animal fats and greaves Centres for the collection of raw materials and further transport to processing establishments must be equipped with facilities for the storage of raw materials at a temperature of not more than 7 °C. However, raw materials may be stored and transported without active refrigeration if rendered within 12 hours after the day on which they were obtained. Fish Not more than 3 °C six hours after loading and not more than 0 °C after 16 hours

Post catch © 2008, Woodhead Publishing Limited

Chilled foods

Process

406

Table 15.1 Specific temperature control requirements within regulation 853/2004

Fresh fishery products, thawed unprocessed fishery products, cooked and chilled products from crustaceans and molluscs

Maintained at a temperature approaching that of melting ice

Post collection

Milk

Cooled immediately to not more than 8 °C in the case of daily collection, or not more than 6 °C if collection is not daily

Transport

Milk

Maintained and, on arrival at the establishment of destination, the temperature of the milk must not be more than 10 °C

Food business operators

Milk

Upon acceptance at a processing establishment, milk is quickly cooled to not more than 6 °C and kept at that temperature until processed

Processing

Egg

If processing is not carried out immediately after breaking, liquid egg must be stored either frozen or at a temperature of not more than 4 °C. The storage period before processing at 4 °C must not exceed 48 hours. However, these requirements do not apply to products to be de-sugared, if de-sugaring process is performed as soon as possible

Storage

Egg

Products that have not been stabilised so as to be kept at room temperature must be cooled to not more than 4 °C

*Allowable exceptions for transport of less than 2 hours and if competent authority specifies.

Temperature monitoring and measurement

Storage

407

© 2008, Woodhead Publishing Limited

408

Chilled foods

legislation. Therefore, The Food Hygiene (England) Regulations 2006 (SI 2006/ 14)4 came into force on 11 January 2006 to enforce the EU regulations. An amendment to these regulations, The Food Hygiene (England) (Amendment) Regulations 20075, came into force on 14 February 2007. Part 4, Miscellaneous and supplementary provisions, Section 30 (Schedule 4) of The Food Hygiene (England) Regulations 2006 contains temperature control requirements and applies to food businesses not operating under Regulation 853/ 2004 and any food business that is not carried out on a ship or aircraft. For chilled food, the Regulation states that in any commercial operation, any food likely to support the growth of pathogenic micro-organisms or the formation of toxins should be maintained below 8 °C. If food is kept above 8 °C, the food business operator must have carried out a ‘well-founded scientific assessment of the safety of the food at the specified temperature’. Within The Food Hygiene (England) Regulations 2006, certain tolerances outside of the maximum allowable temperature are given. These include being able to keep food at above 8 °C if the display for sale period is less than 4 hours, if the food is being transferred from or to vehicles or if unavoidable circumstances develop (handling of food, defrosting of refrigerated equipment or breakdown of refrigerating equipment) provided the ‘period was consistent with food safety’.

15.2.2 Enforcement of temperature control legislation To ensure that The Food Hygiene (England) Regulations 2006 (SI 2006/14) and equivalent Scottish, Welsh and Northern Irish legislation are enforced consistently throughout the UK, Food Law Practice Guidance6 and a Food Law Code of Practice7 documents have been issued by the Food Standards Agency for use by environmental health practitioners. During the verification of compliance with the required temperatures, authorised officers should adhere to the following staged approach: Stage 1 – check to determine whether temperature monitoring equipment is in place and whether data on temperatures is logged/recorded, and verify the accuracy of temperature monitoring equipment. Stage 2 – measure between-pack temperature of food (non-destructive temperature checks). Stage 3 – measure the temperature of the product itself (destructive test). If officers are satisfied after Stages 1 or 2, further stages in the enforcement procedure are not carried out. Stage 3 testing must be used to produce evidence for a prosecution. The Guidance puts an emphasis on the accuracy of the equipment used to measure temperatures. It is stated that air temperature monitoring should be used only as a guide to determine how well a refrigerating system functions. When measuring using a between-pack sensor, a tolerance of 2.8 °C is allowed (0.8 °C for instrument accuracy and 2 °C for the limitation of the methodology). If destructive testing is carried out, the Guidance states that the product tested should © 2008, Woodhead Publishing Limited

Temperature monitoring and measurement

409

be taken from the warmest area of the chilled load and that thermochromic (liquid crystal) strips can be used to help identify this position. Thermometers or temperature sensors used for enforcement should be calibrated and certified to an appropriate standard (e.g. NPL) and checked for accuracy in melting ice. They should be pre-cooled prior to insertion between or into products to prevent inaccuracies in measurement. Temperature measurement systems that are used for enforcement purposes should comply with certain requirements:

• should reach 90% of their final reading within 3 minutes; • have an accuracy equal to or less than ±0.5 °C when measuring within the temperature range –20 °C to +30 °C;

• must not change in accuracy by more than ±0.3 °C when operated in temperatures of –20 °C to +30 °C;

• instrument display should be readable to at least 0.1 °C; • system should be robust and shock proof; • the sensor should be robust and easily cleanable and should facilitate good thermal contact with the food;

• the measurement instrument should be powered by a dry cell battery, not mains •

electricity (and should have a means of identifying when the battery needs replacing); the temperature probe should be appropriate for the measurements being made (e.g. robust rigid stem with a sharpened point for insertion into product; flat head for between-pack measurements).

15.3 Importance of temperature monitoring The last few years have seen an enormous advance in computer and communications technology permitting real-time status information on vehicles, chill stores and retail display cabinets. Satellite tracking systems (GPS) can follow a vehicle’s position and give total information about the refrigeration and engine systems to its depot. Retail display cases can also have integrated temperature and humidity control to ensure the full shelf-life of non-pre-packed foods. Thus, temperature measurement can be part of an integrated management, safety and quality system. Reducing the storage temperature of food does not kill micro-organisms, but it retards their growth. Hence keeping raw materials, intermediate and finished products at chill temperatures will play its part in ensuring that the food is safe. Other important areas are the proper hygiene training of operatives, prevention of physical contaminants, suitable fittings and equipment, good cleaning regimes, and pest control. Refrigeration equipment is built to function for long periods without attention; however, there are many events apart from breakdowns that can affect temperature control. The defrost cycles need to be set at the correct frequency and the optimum duration to minimise heat gains to the food. The loading of food into refrigerated systems is also often crucial to its operation and © 2008, Woodhead Publishing Limited

410

Chilled foods

proper air flow. Air temperature monitoring can indicate whether refrigerated equipment is functioning correctly and is being operated correctly, but it may not directly relate to the food temperature. In some circumstances, air temperature monitoring is not possible and product temperature or product simulant temperature is required.

15.4 Principles of temperature monitoring 15.4.1 Choice of system There are an enormous number of different temperature monitoring systems available commercially, from a simple thermometer to a fully computerised system linked to a local refrigeration system or even central control system. The choice of system will depend on the amount of detail the operator requires and its cost. If the monitoring system is to provide detailed information on the operation of a system linked with other reactive management systems, then obviously a more elaborate and complex system is required. This may include a large number of sensors to enable a very complete picture of the temperature distribution within a refrigerated system to be obtained. It may also include other information such as defrost cycles, compressor and expansion valve operation, door openings, and energy consumption, and may be linked to an alarm system (simple or telephone based), stock keeping and batch codes of product. On the other hand, if monitoring is being carried out only to ensure that food is being kept within certain temperatures at a critical control point, then the amount of information that is collected may be reduced.

15.4.2 Which temperature to monitor? When designing a monitoring system, there are certain considerations in the choice of temperatures to be measured in the refrigerated system. These are:

• The choice of whether to monitor air temperatures, product temperatures or • •

simulated product temperatures will depend on the individual system and the way it operates. The sensors should preferably be fixed in a position where they will not be damaged during commercial activity. If manual readings are used, these should be taken from accessible positions. The temperatures chosen should be representative of the refrigerated system and give a picture of its operation, and therefore be linked indirectly with the product temperature.

15.4.3 Air temperature monitoring In terms of regulatory compliance, and as part of a documented system based on HACCP principles, the temperature of the food should be monitored. However, the © 2008, Woodhead Publishing Limited

Temperature monitoring and measurement

411

storage or holding times of chilled food are relatively short, making product temperature monitoring difficult without disruption to normal commercial activity and requiring the intervention of trained operators. It is easier to fix sensors separate from food loads, which are connected to readout systems, where temperatures can be recorded automatically or manually. Most refrigerated systems function by circulating cold air over the system’s evaporator, and then passing this cold air over the food load to remove heat from the food. Movement of air is by mechanical fans or, in some cases, by gravity, which relies on the density of cold air being greater than that of warm air. In the case of mechanical circulation, the air returns to the evaporator after passing over the food, making the returning air the same temperature or warmer than the food it is cooling. Localised heating effects from lighting or other effects may give rise to ‘hot spots’ or uneven temperature distribution, and make a small part of the food load warmer than the return air. In general, the relationship between air temperature and product temperature is best established by examining the difference between the cold air leaving the evaporator and warmer air returning to the evaporator. This gives a measure of the performance of the refrigerated system and its effectiveness in keeping the food cold. This differential is also used as the basis of air temperature monitoring. However, in order to relate the air temperatures to product temperature, it is necessary to carry out a load test. The load test involves examining the differential of air temperatures and comparing it to product temperature over a sufficient period of time to ensure the system is working under normal conditions. With closed systems such as chill stores and vehicles, where the only perturbation derives from defrost cycles, door opening, and changing loads, determination of the relationship between air and product temperature is simpler. The warmest locations in the system have to be determined, and product temperatures followed over a period of time in order to relate them to air temperatures. With open systems such as display cabinets, their operation is more sensitive to environmental conditions and location. Room temperature and humidity variations, perturbation of the air curtain by draughts or customer movement can change the temperature distribution. Under these circumstances, load testing can be more difficult. Commercial cabinet manufacturers perform a load test to check cabinet performance (BS EN ISO 23953:2005 Parts 18 and 29), using a set loading of standardised blocks of a gel (‘Tylose’) under controlled environmental test room conditions where ambient temperature and air flow across the front of the cabinet are controlled. Unless the cabinet is loaded and operated in the same manner as in the test room it is unlikely that the cabinet will operate in the same way as in the test situation. However, the load test may be used as an initial guide to identify the warmest positions within a cabinet.

15.4.4 Alternatives to air temperature monitoring There are some circumstances where air temperature monitoring is not appropriate or needs modification. In closed cabinet systems, such as chill storage cabinets © 2008, Woodhead Publishing Limited

412

Chilled foods

using gravity cooling from an ice-box or backplate, the air temperature inside requires significant time to recover after door openings.10 Thus, periodic readings of air temperatures would have little meaning and bear no relationship to the temperatures of the food being stored. In this case it would be better to monitor either a food sample or a ‘simulated’ food sample. The thermal mass of the sample would make it less sensitive to rapid air temperature changes. Also, it is possible to match the ‘food simulant’ to have a similar cooling factor or similar thermal diffusivity to the food being monitored.11 The use of such monitoring would be essential, for example, where cooling is by conduction such as with cold plate (dole plate) serving units in catering, or where air flows are low velocity (gravity-fed serve-over cabinets). Even where the system is cooled by forced air, there may be situations where variations in air temperatures are large, e.g. small delivery vehicles and cabinet refrigerators. In these cases, air temperature monitoring is difficult to interpret. By increasing the response time or ‘damping’ the sensor or measuring system, the trends in air temperature can be followed, whilst removing the short-term variations. ‘Damping’ can be achieved physically by increasing the thermal mass of the sensor or electronically by alteration of the read-out circuitry.

15.5 Temperature monitoring in practice 15.5.1 Transport, storage and distribution Walk-in chill stores Walk-in stores consist of an insulated store chamber cooled by one or more fanassisted air cooling units, depending on their size. The position of cooling units around the chamber varies, but is usually at ceiling level (Fig. 15.1). Air circulation should be designed to give proper distribution throughout the chamber, and to eliminate any ‘hot spots’ or stratification of air layers. In nearly all cases, air temperature recovery after door openings or defrost is rapid, permitting air temperature to be used as the most convenient means of monitoring. Retention of cold air can be further improved with the use of strip plastic curtains, an air curtain above the door, or a dehumidified air lock minimising the ingress of warm air on door openings. The number of sensors to be used to monitor air temperatures in a chill store will depend on its size and the number of cooling units. Table 15.2 gives an indication of the minimum number of sensors related to the volume of the store, with only one sensor being required in stores less than 500 m3. The sensor should be positioned to gives an indication of the warmest air temperature and hence the warmest food in the store. This warmest location depends on the design of the store, especially the position of the air cooling unit. Figure 15.2 gives an example of air temperatures during the 24-hour operation of a large chill store. The graph shows temperature variations during peak activities of movement of chilled foods in the afternoon and evening compared to quieter © 2008, Woodhead Publishing Limited

Temperature monitoring and measurement

Fig. 15.1

413

Air circulation in a chill store.

Table 15.2 Number of sensors recommended in chill stores Chamber volume above (m3) 500 5000 20 000 50 000 85 000

Number of sensors 2 3 4 5 6

loading activity in the morning. Differences between wall sensors and air return temperatures are very small in this case, and can be affected by their positioning in the store. For chill stores less than 500 m3, the single sensor could be placed in the air return of the cooling unit. In a closed system, such as a store with adequate air distribution, the temperature reading of the air return approximates to the mean temperature of the food load. If there is not good air distribution, then it may be better to put the one sensor in a position more representative of the warmest air temperature. This may be located at the following positions:

• the maximum height of the food load, furthest away from the cooling unit; • at approximately two-thirds the height of the chamber, away from the door and the direct path of the cooling unit;

• two metres above floor level, directly opposite the cooler unit. If the cooling unit is placed above the door, the negative pressure produced by the © 2008, Woodhead Publishing Limited

414

Chilled foods

Fig. 15.2

Air temperature monitoring record of large chill store (40 000 m3).

fan can increase the amount of air drawn into the chamber during door openings. Thus, air return temperature monitoring is not often appropriate in this case. For larger stores, different sensors can be used to indicate the temperatures in different parts of the store. In addition, placing extra sensors in the air outlet and air intakes of one or more of the cooling units gives further information on the performance of the refrigeration system. © 2008, Woodhead Publishing Limited

Temperature monitoring and measurement

Icebox gravity feed

Forced air

C

415

Backplate gravity feed C

C

B B

A

A

B A

(a)

(b)

(c)

Fig. 15.3 Cabinet refrigerators. (a) Forced air refrigerator. (b) Icebox refrigerator. (c) Backplate refrigerator. A: Air off. B: Air return (air-on). C: Load limit or warmest point.

Cabinet refrigerators Cabinet refrigerators are generally free-standing, small-sized units with single or double doors. They are usually cooled by fan-assisted cold air but occasionally can be cooled by gravity-circulated air from an integral icebox or backplate (Fig. 15.3a, b and c). As indicated earlier, air temperature monitoring is not as appropriate to these types of refrigerated systems as it is to walk-in chill stores. Fan-assisted refrigerators will recover relatively quickly after door openings, but a large number of door openings, especially at most active periods of use, will make any temperature readings difficult to interpret. Air temperature monitoring can be more meaningful if a ‘damped’ sensor, with a response time of around 15 minutes, in the air return position is used. Damping can be achieved by using a metal or plastic sheath over the sensor or suspending the sensor in water, oil or glycerol. Figure 15.4 shows the effect of ‘damping’ when the sensor is set at the centre of a plastic tub of water, and readings are compared to air temperatures after door openings. Since cabinets cooled by a back plate or ice box have weak air circulation and long recovery times after door openings, it is more appropriate to monitor their temperatures using food temperatures or, even better, a simulated food temperature. As foods are microbiologically unstable, food temperature monitoring would require using different foods each day, and might lead to wastage. Permanent positioning of a sensor requires a stable food simulant. It is important when choosing a food simulant that it behaves similarly to the food being © 2008, Woodhead Publishing Limited

416

Chilled foods

Door opening for 30 seconds

Temperature (°C)

8

6

4 Sensor placed in plastic tub of water (150 ml) 2

10

20

30

40

Time (minutes)

Fig. 15.4

Effect of ‘damping’ air sensor.

monitored, and that it is robust to different working conditions. It is recommended to determine the cooling factor of the specific package or piece of food and match this with a particular food simulant, or match the thermal diffusivity of the food with that of the simulant.11 Values for cooling factors of different foods and package sizes are also published,11 as well as thermal diffusivities for a range of plastic materials. Regular checks should be made with a food simulant system to ensure that the sensor embedded in it is accurate and functioning properly, and that the simulant is performing as it should. The position of maximum temperature in cabinet refrigerators will vary, depending on the method of air distribution and exact design. However, data collected from 79 cabinets when tested in a test room showed that the maximum product temperature was most usually at the front of the top or bottom shelf 12 (Fig. 15.5).

15.5.2 Chilled transport Distribution of chilled foods is carried out in many different types of vehicle ranging from large 40-foot heavy goods vehicles with independent cooling units, to light goods vehicles relying on insulated containers to maintain the temperature of pre-chilled foods. Pre-chilling to the correct temperature is essential given that most refrigeration units are designed to maintain temperature not cool the load down. © 2008, Woodhead Publishing Limited

Temperature monitoring and measurement

417

50

Percentage of cabinets

45 40 35 30 25 20 15 10 5 0 Top shelf rear

Fig. 15.5

Top shelf Middle shelf Middle shelf Bottom shelf rear front rear front

Bottom shelf front

Position of maximum product temperature in cabinet refrigerators.

Temperature controlled vehicles An independent refrigeration unit, usually powered by diesel, often with an auxiliary electric motor, is used to circulate cold air around the vehicle compartment from an evaporator unit at the front of the vehicle. A trend in many multiple distribution depots is to use vehicles with movable bulkheads so that a vehicle can carry frozen and chilled foods at different temperatures in the same load. Each compartment will have its own evaporator, which can control temperatures independently. The cold air is distributed in different ways within the different vehicles, but the majority have cold air leaving from the top of the cooling unit near the roof, and returning via the base to the front of the vehicle and the return air intake (Fig. 15.6). Correct loading and spacing of the load within the vehicle is crucial to ensure adequate cold air distribution within the compartment. If the load is not spaced correctly, circulation can be restricted and ‘hot spots’ can occur. The maximum length and width of vehicles is set by regulation, and hence the free space available to loads within an insulated chamber places further restrictions on achieving correct loading. Some vehicles are cooled by direct evaporation of liquid nitrogen from a reservoir tank on the vehicle or are cooled by pre-cooled eutectic beams within the vehicle (sometimes in combination with an on-board refrigeration system). These vehicles have the advantage of being much quieter than mechanically refrigerated vehicles, and temperature control can be better. However, an adequate supply of liquid nitrogen or sufficient refrigeration capacity in the eutectic beams is required for the journey, which can limit their range and number of stops. Sensors on temperature readout and single-channel chart records, which have been used for many years on refrigerated vehicles, measure the air return temperature. This returning air should indicate the mean temperature in the load, provided that there is good distribution to all parts of the load. Short circuiting of air may © 2008, Woodhead Publishing Limited

418

Chilled foods Evaporator behind bulkhead

Sensor 2 temperature

Sensor 1 air return to evaporator

Fig. 15.6

Air temperature monitoring of temperature controlled vehicle.

result in colder return air temperatures. In long vehicles, especially those without air duct distribution of cold air in the ceiling of the compartment, the advice is to fit a second sensor placed nearer the rear of the vehicle (Fig. 15.6). The addition of a second sensor is not sufficient to give a full and accurate picture of temperature distribution within the chamber, but it will measure the cold air leaving the evaporator, and may give a better picture of cold air circulation inside the compartment. This second sensor will serve as a check on the functioning of the measuring system, and makes tampering more difficult. It should demonstrate that the evaporator and fan unit are functioning properly and that cold air is reaching the back of the vehicle. It will indicate more easily when the cooling unit has been switched off, or a load added which has been insufficiently cooled. Prevention of freezing of part of the load can also be more easily avoided. Comparison of the normal differential temperatures between the rear sensor and the return air sensor may also indicate poor air distribution within the compartment. Figure 15.7a shows an example of temperature monitoring in a vehicle fitted with two sensors, including the effect of door openings. Figure 15.7b illustrates the care with which air temperature records must be interpreted. The system is operating normally until the chamber is loaded. From this point, the air return sensor gives an acceptable reading apart from slightly longer cycles. However, the compartment sensor at the rear of the vehicle shows a temperature rise indicating that the flow of cold air has been restricted by the load. This causes short-circuiting of cold air from the evaporator and hence the longer cycling period activated by the thermostat. As soon as the driver rearranges the load to restart the air flow to the rear of the vehicle, the temperature drops. This problem would not have been obvious had there been only one sensor on the air return. The frequency of recording for electronic loggers will depend on the length of the journey. Maximum recording intervals are given in EN 12830:1999 (Section 15.6.1). Other information such as defrost cycles, door openings and load identification may also be required. It is important that a driver be aware of any problem © 2008, Woodhead Publishing Limited

Temperature monitoring and measurement

419

Fig. 15.7 (a) Normal air temperature record and (b) poorly-loaded vehicle air temperature record of chilled foods vehicle. (By permission of Cold Chain Instruments.)

occurring with the temperature of the load either through a visible temperature read out or an alarm system. The monitoring of vehicles with movable bulkheads would require more sensors to enable the temperature to be recorded in each separate compartment. This may be achieved in several ways. The easiest would be to monitor the air intake of each cooling unit. Alternatively, more sensors could be fixed to the roof of the chamber to enable compartment temperatures to be monitored, whatever the position of the bulkhead, and in addition to the air return measurements. The use of small temperature loggers, whose position can be changed to suit the bulkhead arrangement, may provide another solution. For vehicles cooled by liquid nitrogen, the sensors have to be positioned in order to account for any temperature gradients occurring in the chamber. Forced circulation should eliminate gradients. If fans are not used, then sensors should be placed above and below the load. Small delivery vehicles Many light goods vehicles delivering chilled foods are fitted with refrigerated units driven from the vehicle engine or transmission. This means that cooling is not possible whilst the vehicle is stationary. Alternatively, the refrigeration units can be powered by the vehicle’s battery allowing continual use of the refrigerating unit during periods when the engine is turned off. The main problem in maintaining good temperature control arises from the number of door openings and the amount of time doors may stay open whilst orders are prepared and delivered. Typical ‘high street’ delivery patterns can result in doors being open for 40% of the working day. This can make temperature control very difficult and also the use of air temperature monitoring inappropriate. Employment of plastic strip curtains © 2008, Woodhead Publishing Limited

420

Chilled foods õ

Air return

ü

Product

¸

Air off

12 õ õ

10 õ

Temperature (°C)

8

õ

õ

õ

õõ õ

õ õ

õ õ

õõ

õõ

õ

õ

õ

õ

õ õ

õ

õ

õ õ

6

4

õ õõ õ õ

ü ü ü ü ü ü ü ü ü ü ü ü ü ü ü üü üü ü ü ü ü ü ü ü ü ü ü ü ü ü

2 ¸

0

¸

¸¸ ¸

¸ ¸

¸¸

¸

¸

¸

¸ ¸

¸

¸¸

¸ ¸ ¸¸

¸

¸

¸ ¸

¸

¸

¸ ¸¸

¸

¸

–2 7:00

8:30

10:00

11:30

13:00

14:30

Time of day

Fig. 15.8

Air temperature record of a small delivery vehicle.

above the door can help to reduce warm air entry whilst doors are open. However, information can be obtained if the air temperature sensors are ‘damped’ by suspending them in a small bottle of liquid such as oil or glycerol. The large fluctuations are removed from the temperature records and the trends in overall chamber temperature followed. An example of this technique of monitoring is shown in Fig. 15.8. Vehicles using eutectic plates or insulated boxes to carry foods normally use a food simulant or actual food to monitor temperature during a journey. The positioning has to be as representative of the load as possible. The temperatures can be read manually, but can also be connected to a chart recorder or logging system. 15.5.3 Display cabinets The majority of chilled foods are displayed in open cabinets. Some sectors use display cabinets with closed doors; for the purposes of monitoring, these can be considered as storage cabinets (see ‘Cabinet refrigerators’ in Section 15.5.1). The open cabinets can be divided into two main groups, multi-deck open cabinets and serve-over cabinets. © 2008, Woodhead Publishing Limited

Temperature monitoring and measurement

Fig. 15.9

421

Air temperature monitoring in retail display cabinets. (a) Multi-deck cabinet, (b) Serve-over cabinet.

Multi-deck cabinets A fan draws air from the front grille of the cabinet, where it is cooled by passing through an evaporator. The cool air emerges from the back of the shelves to cool the food, and from the top grille to form an air curtain in front of the shelves (Fig. 15.9a). Advances have been achieved in the design of cabinets, which include the reduction of heat gain from internal lighting, and the stabilising of the air curtain, by improved design or addition of a second curtain. Thus the ease of monitoring of the temperature in multi-deck cabinets will be determined by their design and operation. In principle, the differential between the air returning from the shelves and the air emerging on the shelves is an indicator of the cabinet’s performance. Positioning of sensors or manual reading of temperature is usually taken from the air entering and leaving the evaporator. If the normal pattern of variation in air © 2008, Woodhead Publishing Limited

422

Chilled foods 60

Percentage of cabinets

50 40 30 20 10 0 Top shelf rear

Fig. 15.10

Top shelf front

Middle shelf rear

Middle shelf front

Bottom shelf rear

Bottom shelf front

Base shelf rear

Base shelf front

Position of maximum product temperature in multi-deck retail display cabinets.

temperatures can be linked to variations in product temperature on the shelves, then air temperature monitoring can be used on a routine basis. If other factors interfere with the air temperature relationship, such as excessive radiant heat absorption, or if a relationship cannot be established between product and air temperatures, then a product or food simulant may have to be used. The position of maximum temperature in multi-deck cabinets will vary depending on the design of the cabinet. Data collected from 87 cabinets when tested in a test room showed that the maximum product temperature in 59% of cabinets was at the front of the base (well) of the cabinet (Fig. 15.10).12 Figure 15.11 shows two different air temperature patterns. The first (Fig. 15.11a) shows regular cyclical changes in air temperature, whilst Fig. 15.11b indicates much steadier conditions except during a defrost cycle. In both cases, establishment of the range of air temperature with the warmest product temperatures would allow effective air temperature monitoring. Serve-over display cabinets There is a wide range of cabinets in this group for displaying meat, fish, delicatessen products, patisserie, cheeses and ready-to-eat products in catering establishments. In most cases, the food is cooled by cold air from a refrigerated unit, but in some older units the food is cooled by contact with a cold plate or well (dole plate) or crushed ice. The effect of radiant heat from lighting or sunlight can be more pronounced with serve-over cabinets and can significantly affect food temperature. Figure 15.9b shows a typical serve-over cabinet for retail sale of delicatessen products, using fan-assisted cold air. Air emerging from the back grille cascades © 2008, Woodhead Publishing Limited

Temperature monitoring and measurement

Fig. 15.11

423

Temperature monitoring records of two different display cabinets. (By permission of the University of Bristol.)

over the food and returns via a front grille. In the case of gravity-fed cabinets, where air enters a back grille and emerges at the base shelf, there is no air return grille. Air velocities are low in serve-over cabinets to reduce dehydration of the displayed products; this also makes air temperature measurement more difficult. Positions of sensors or manual measurements for air temperatures are also shown © 2008, Woodhead Publishing Limited

424

Chilled foods

in Fig. 15.9b. Determining the relationship between air and product temperatures is necessary before air temperatures can be used routinely. In many cases it will be easier to monitor the temperature of the cabinet using food temperatures or a food simulant. The temperature near the front of the cabinet will usually be indicative of the warmest locations, and hence warmest foods, within the cabinet. Air temperature monitoring is not suitable for cabinets cooled by conduction. In this case, direct measurement of food temperatures is appropriate, but should be carried out, as with all determinations of this type, with a clean, well-disinfected probe.

15.6 Equipment for temperature monitoring 15.6.1 Standards Within Europe the following standards exist for thermometers and temperature recording instruments: (i)

BS EN 13485:2001. Thermometers for measuring the air and product temperature for the transport, storage and distribution of chilled, frozen, deep-frozen/quick-frozen food and ice cream – Tests, performance, suitability.13 (ii) EN 12830:1999. Temperature recorders for the transport, storage and distribution of chilled, frozen, deep-frozen/quick-frozen food and ice cream – Tests, performance, and suitability.14 (iii) BS EN 13486:2002. Temperature recorders and thermometers for the transport, storage and distribution of chilled, frozen, deep-frozen/quick-frozen food and ice cream – Periodic verification.15 BS EN 13485:2001 covers specifications for measurement instruments for chilled and frozen foods and methods to verify compliance. The standard specifies lower and higher limit values for temperature (Table 15.3). The accuracy of instrumentation is divided into classes that define errors and resolution for thermometers for air and for internal product temperature measurement (Table 15.4). The response time for thermometers varies depending on the application and is defined as the time required for the indicated value to reach 90% of the applied temperature change. For fixed thermometers for measuring air temperature, the response time must be less than 10 minutes for transport and 20 minutes for storage applications. Portable thermometers must have a maximum response of 3 minutes, as must thermometers for measuring internal product temperatures. The standard defines operation and storage conditions in varied climatic environments for fixed and portable thermometers for air and product temperature measurement (Table 15.5). Instruments can therefore be designated according to BS EN 13485 according to the following classifications:

• application: air, product, or both; • transport (T) or storage (S) use; © 2008, Woodhead Publishing Limited

Temperature monitoring and measurement

425

Table 15.3 Limit values for air and product thermometers Type of application Air (span to be Chilled, frozen greater than 50K) Chilled, frozen, deep-frozen/ quick frozen and ice cream Product (span to be specified by manufacturer)

Lower limit temperature (°C)

Higher limit temperature (°C)

≤–10 ≤–30

≥+20 ≥+15

≤–20

≥+30

Table 15.4 Accuracy classes for thermometers for measuring air and internal product temperatures Class Maximum permissible errors Resolution

0.5 ±0.5 °C ≤0.1 °C

1 ±1 °C ≤0.5 °C

2 ±2 °C ≤1 °C

Class 1 and 2 for air temperature, class 0.5 and 1 for internal product temperature measurement.

• climatic environment (A to E according to Table 15.5); • accuracy class (according to Table 15.4); • measuring range (n.b. this differs from the climatic conditions), e.g. EN 13485, air, T, B, 2, –35 °C +15 °C (suitable for measuring air temperature in transport of between –35 °C and +15 °C). EN 12830:1999 is broadly similar to BS EN 13485:2001 but applies to whole recorder-temperature sensors. Accuracy classes and climatic environments are similar to those specified in BS EN 13485:2001 for air temperature measurement. In addition, recording intervals, durations, response times, accuracies and test methods are specified. Recording intervals for storage are specified as a maximum recording interval of 30 minutes. For transport periods of less than 24 hours, a recording interval of 5 minutes is required, for between 24 hours and 7 days an interval of 15 minutes is required, and for periods greater than 7 days an interval of 60 minutes is required. It is for manufacturers to state the recording capacity of the instrument. Like BS EN 13485:2001, response time for external sensors is 10 minutes for transport and 20 minutes for storage. However, if the recorder has an internal sensor, a response time of 60 minutes is allowed. BS EN 13486:2002 covers frequency and verification of instrument accuracy. It recommends verification of instrument accuracy on a yearly basis with calibration accuracies of ±0.1 °C, ±0.2 °C and ±0.5 °C for class 0.5, 1 and 2 instruments respectively.

© 2008, Woodhead Publishing Limited

426

A Type and use of the apparatus

B

AIR – type of climatic environment C D

Product E

Storage and distribution unit Thermometer type used and method of use Outside refrigerated case, in heated or air conditioned premises with external sensor

Transport vehicle Inside cabin or outside vehicle with external sensor

Storage and Transport distribution unit vehicle Inside refrigeration case with internal or external sensor

Thermometer rated operating conditions Thermometer limiting conditions Thermometer and sensor storage or transport conditions

–30 °C +65 °C –30 °C +70 °C –40 °C +85 °C

–30 °C +30 °C –20 °C +30 °C –20 °C +30 °C –40 °C +50 °C –40 °C +70 °C –30 °C +50 °C –30 °C +50 °C –40 °C +60 °C –40 °C +85 °C –30 °C +70 °C –30 °C +70 °C

© 2008, Woodhead Publishing Limited

+5 °C +40 °C +0 °C +50 °C –20 °C +60 °C

Portable thermometers

Chilled foods

Table 15.5 Operational and storage conditions for thermometers for food according to BS EN 13485:2001

Temperature monitoring and measurement

427

15.6.2 Sensors Accuracy Whatever the system for collecting or recording temperatures, the sensor is the common factor between them. The three principal types of sensors in commercial use are thermocouples, semiconductor (thermistor) and platinum resistance. The choice of sensor will depend on the requirements for accuracy, speed of response, range of temperatures, robustness and cost. Until recently, the majority of general-purpose thermometers and measuring systems used a thermocouple. This is a pair of dissimilar metals joined together at one end, usually by a welded joint. The circuit is completed by a second junction held at a known temperature (often referred to as a ‘cold’ junction). In food applications, where temperatures are relatively close to ambient, two types of thermocouple predominate: Type K thermocouples, which use wires of Chromel (a nickel–chromium alloy) and Alumel (a nickel–aluminium alloy); and Type T thermocouples, which use wires of copper and Constantan (a copper–nickel alloy). The advantages of thermocouples are their low cost, facility to be hand-prepared from reels of wire, and a very wide range of temperature measurement (–184 °C to 1600 °C). Table 15.6 shows the permissible sensor accuracies for the three types of sensors. The difference in instrumental error arises from the fact that the electronic circuitry has to compensate for changes in the reference or ‘cold junction’ (normally ambient temperature). This is measured by a built-in semi-conductor sensor, which automatically compensates for changes in ambient temperature. Measurement errors increase when the ambient temperature varies widely, e.g. moving from a cold to a hot environment. Other errors can be produced by induced voltages from motors or transmitters, moisture, and thermal gradients in other junctions. A move to greater accuracy for measuring and monitoring restricts the use of thermocouple sensors to Type T only, which will normally meet the basic specification for air temperature monitoring (see Section 15.6.1). Thermistor sensors change resistance with temperature, but measure over a narrower range of temperatures (–40 °C to 140 °C) than thermocouples. Their use for measuring food temperatures has increased since the introduction of European measurement standards.13–15 They are rugged, provide good accuracy and repeatability and are not unduly affected by changes in ambient temperatures. Platinum resistance thermometers also have a system accuracy that meets the Table 15.6 Sensor and system accuracies

Sensor accuracy (°C) Instrument accuracy* (°C) System accuracy (°C)

Type K

Type T

Thermistor

Pt resistance

±1.5 ±0.3 ±1.8

±0.5 ±0.3 ±0.8

±0.2 ±0.2 ±0.4

±0.1 ±0.2 ±0.3

*Includes cold junction compensation accuracy.

© 2008, Woodhead Publishing Limited

428

Chilled foods

Table 15.7 Typical response times (seconds) in air and water (from Fairhurst16)

Exposed thermocouple Shrouded thermocouple Exposed thermistor Shrouded thermistor Shrouded platinum resistance

Still air

Forced air

Water*

20 150 45 260 365

5 40 20 50 65

– 6 – 12 15

* Mounted in ‘chisel’ probe in water, time for 20 °C change to 99% level.

European standards. They may be used over a wider range of temperatures (–270 °C to 850 °C). Normally their response time (Table 15.7) is slower, unless they are specially constructed for a fast response. Corrections have to be made for resistance of leads, and a self-heating effect. Their higher cost has restricted their use to applications where high accuracy is required (e.g. calibration). Calibration and periodic verification During manufacture, each sensor and instrument is checked to ensure that it meets specification and achieves an accuracy within tolerances set by each manufacturer and in accordance with European standards.13–15 In many applications of monitoring, different sensors are plugged into an instrumentation system, and they are normally regarded as interchangeable. However, in the case where more precise readings are required, an individual calibration is undertaken of the sensor and instrument together (system). This measures the system’s reading against a range of applied temperatures. The applied temperatures have to be traceable to a national standard (e.g. National Physical Laboratory). The resulting table or graph of the calibration certificate enables the system reading to be corrected to a true reading (within the tolerances of the calibration). It is also recommended that, for temperature measurement to determine compliance with the legislation, the probe and instrument are periodically checked using melting ice. Once a temperature monitoring system is installed, it is essential that periodic checks are carried out to ensure that the equipment is functioning correctly and meets the same specification as when it was purchased, and as described in standards. The frequency of checks depends on the use of the equipment and will consist of routine checks on the functioning of the equipment and those carried out by the manufacturer (or a suitably qualified laboratory). The maximum period recommended is one year for a manufacturer’s check or after a long period of nonuse or operating incident. The equipment is normally checked against another thermometer that has been calibrated against a standard. It is also normal to check the accuracy and functioning of the clock or verify the recording duration. Sensor housing and probes In monitoring applications, the sensor element has to be protected from damage or breakage. This can range from coating with an epoxy resin to embedding in a stainless-steel sheath. If fast response is required, the thermal mass has to be as low © 2008, Woodhead Publishing Limited

Temperature monitoring and measurement

429

as possible. It is also important that sensors which are mounted inside chambers or vehicles to measure air temperatures are protected from damage during the commercial activity of loading or unloading food, but not in such a way as to restrict air flow. Monitoring and measuring food temperatures often requires sensors mounted in hand-held probes. The design of the probe depends on its application. The most common probe is for insertion into foods, and therefore it has a sharpened tip (see Fig. 15.12a). If a non-destructive temperature measurement is required then a probe is required that can be inserted between food packs or cases. Good contact between the packaging and the probe, together with an adequate period to allow readings to settle, are essential to minimise errors in this type of measurement. Examples of probes used to measure between packs and cases are shown in Figures 15.12b and 15.12c. Single readout systems Instrumentation has progressed from the original single readout thermometer, the mercury- or alcohol-in-glass thermometer. The development of dial and stick thermometers with analogue or digital display removed the danger of breakage, but their use can be limited by their low accuracies, especially those based on bimetallic strips. Dial thermometers, which were used to indicate air temperatures in display cabinets, have largely been replaced by digital thermometers. Thermochromic liquid crystals change in orientation and transparency depending on their composition and temperature. When set into strips, they will display the appropriate temperatures printed under them. Their accuracy is limited, but can achieve ±1 °C. More common is the electronic digital readout instrument, which is powered by batteries. The resolution and interval of display temperature will vary with model and type of sensor. Temperatures can be stored and even printed out, and an alarm given if the temperature goes outside a preset limit. Chart recorders Historically, a trace on a moving chart was the only method available to produce a temperature history and record. The use of chart recorders is now less common and they have been overtaken by electronic instruments, but some may still be found in fixed system applications such as cold or chill stores and vehicles. The charts can be circular or mounted on a roll to give a rectangular chart, and a trace is obtained with ink or pressure or heat-sensitive paper. Circular chart recorders have the advantage that the temperature history is visible and abrupt changes apparent, and the chart can be easily stored for future reference. The timescale of the chart is usually over 24 hours, 7 or 31 days, but some long distance marine recorders may operate for 6–8 weeks. The chart clock and electronics can either be battery driven for mobility or driven from the mains for fixed applications. Accuracy of the system varies with sensor, but more modern chart recorders are better than 0.5 °C in the range 0–25 °C. Their limitation is often the resolution on the chart divisions and the thickness of the trace. Fixed system charts can be sophisticated instruments with the possibility of recording 30 or more different channels in different colours and print modes. © 2008, Woodhead Publishing Limited

430

Chilled foods

(a)

(b)

(c)

Fig. 15.12 Hand-held temperature probes. (a) Various air and product temperature probes. (b) A probe for between-pack temperatures. (c) A probe for between-case temperatures.

Chart recorders which are fitted to vehicles (or more often to the trailers) have to be more robust in their construction. They need to withstand the rigours of the road for all types of terrain and weather conditions. Recorders giving two or more traces are available, and event markers (e.g. noting door openings) can be added. © 2008, Woodhead Publishing Limited

Temperature monitoring and measurement

431

Fixed processing systems for chill stores The difficulty in interpretation a large number of different traces and the rapid developments in microelectronics and computer technology have encouraged the replacement of chart recorders by data logging systems. This allows not only the storage of large amounts of data, but also its manipulation and analysis and integration into management systems. In chill store operations where large numbers of temperature measurements are being taken every day throughout the year, it has become increasingly the practice to install computerised systems to handle the data. There may be a digital temperature display on a control unit situated near the refrigerated system, but more often information can be retrieved on a visual display unit situated in a control room. Alarm systems can be integrated with the system when any of the parameters being monitored are outside preset limits. The alarm can be transmitted to maintenance staff inside the system’s locations or to outside premises through telecommunication networks. Vehicle temperature logging systems Several companies have developed dedicated temperature logging systems for vehicle monitoring. These have been designed to withstand the vibration and harsh conditions encountered in transport, as stipulated in European standards. Data are collected over the whole journey from loading to unloading, and alarm signals given if temperatures are outside preset limits. The equipment can either fit inside the vehicle cabin (and is often the same size as a vehicle radio) or outside the vehicle, often next to the refrigeration control unit. In addition, distribution customers increasingly require a record of the temperature history of the food they receive. Systems have been developed which give an instant print-out of temperatures up to the point of delivery, to be attached with the delivery documentation. Other features found are ability for variable logging periods and up to 12 months’ memory, event recorders for defrost and door openings, and multiple channels for multi-compartment monitoring. The retrieval of the information is becoming more sophisticated, with downloading facilities to office PCs via manual collection units or by radio, infrared, or satellite communications. Portable data logging systems The miniaturisation of circuitry has produced some very compact and powerful data logging systems, some of which are small enough to travel with food cases or pallets, and record temperatures during passage through the chill chain. The devices can also be used as fixed systems in stores and vehicles. This is useful where the position of fixed sensors would have to be changed from time to time, e.g. temporary chill stores or movable partitions in multi-compartment vehicles. The choice of system will then depend on its particular application, convenience of its use, and price. An evaluation of two such devices has been reported by Kleer et al.,17 where they were used in large-scale catering systems and were found useful in recording the critical control points of the process. Another type of data logger (‘electronic chicken’) is useful for monitoring © 2008, Woodhead Publishing Limited

432

Chilled foods

display cabinets. The logger is placed on a shelf and then records temperatures from a food simulant contained in the logger, which has the same thermal properties as the food displayed on the shelf. An alarm light is fitted to the logger so that problems can be easily noticed and remedied. The data are downloaded for display and analysis via an infrared remote reader. The specification of commercial systems is changing as micro-electronics improve, and it is certain that miniaturisation of the loggers will continue. Remote sensing devices – non-contact infrared thermometers All objects emit energy at temperatures above absolute zero. This energy increases in intensity, but decreases in wavelength, with temperature. In the temperature range of interest to chilled foods, infrared radiation can be measured to determine the temperature. As temperature increases, the intensity increases and the peak energy moves to shorter wavelengths. Hence, most low temperature commercial infrared thermometers filter a band (8–14 microns) out of the infrared spectrum and measure its intensity. Using such a band reduces the atmospheric (water vapour, carbon dioxide) absorption and ‘distance sensitivity’ of the instrument. Very narrow bands (2.2, 15.2 and 7.9 microns) can be used to give greater accuracy at very high temperatures, but signals are very low, requiring expensive high-gain amplifiers. Not all materials emit the same energy at the same temperatures. The ratio of the energy radiated by a material compared with a perfect radiator or blackbody is known as the ‘emissivity’. Emissivities vary from 0 to 1.0, with most organic substances having a value around 0.95. Different substances vary in the amount of energy they absorb, reflect and transmit. Infrared thermometers often have emissivity compensators that can be set for different values (0.1–1.0) to allow for these differences. The target size is also important. The instrument averages all the temperatures it sees in its field of vision. Unless the object fills all the field of vision, the temperature reading will be an average of the object and its surroundings. Focal distances vary with machine, from close-up to 50 metres. The further away, the more difficult it is to pin-point targets, and laser sighting is a useful feature on many models. There are two main types of remote sensing equipment. In one type, a pistolshaped instrument is pointed at a target and the temperature is read from a digital readout at the back of the instrument. Laser targeting can be built into the pistol to give through-the-lens sighting to locate the target, and long-distance devices are often assisted by optical telescopic sighting. The accuracies claimed for this type of instrument are around ±1 °C. Research carried out by the University of Bristol18 on nine commercially available infrared thermometers revealed that care must be taken in using and interpreting results from these devices. Surface temperatures may be quite different from the internal food temperature. This is a more acute problem with frozen foods where the differential between the surface temperature and interior temperature can be large, especially when the food is being transferred in an ambient temperature higher than –18 °C. The infrared sensor not only measures the © 2008, Woodhead Publishing Limited

Temperature monitoring and measurement

433

Table 15.8 Mean error in °C with standard deviation (in parentheses) of various packaging materials Instrument

Clear MAP

a 0.6 (0.1) b –0.3 (0.0) c 0.7 (0.1) d –3.3 (0.3) e –1.9 (0.6) f 0.8 (0.1) g –0.1 (0.1) h 0.5 (0.4) i –2.2 (0.0) Mod. av. 1.4

Glossy cardboard

Plastic bag

1.7 (0.1) 0.7 (0.0) 0.6 (0.0) –4.5 (0.5) –2.3 (0.1) 0.9 (0.4) –0.5 (0.6) 3.8 (0.3) –1.2 (0.0) 1.3

1.1 (0.6) 0.8 (0.6) 0.5 (0.0) –5.1 (0.4) –2.5 (0.0) 1.0 (0.5) 0.4 (0.6) 6.4 (0.7) –0.8 (0.6) 2.1

Printed laminate foil 6.6 (0.6) 5.3 (0.6) 6.0 (0.1) 7.0 (0.2) 4.1 (0.0) 4.2 (3.0) 6.2 (0.5) 10.4 (0.9) 3.2 (0.0) 5.9

Printed MAP 1.9 (0.5) 0.6 (0.1) 0.4 (0.6) –9.1 (1.0) 1.8 (0.1) 2.9 (0.3) 0.2 (0.2) 6.1 (0.8) –0.9 (0.6) 2.7

Printed Moduli vacuum average pack 1.3 (0.1) 1.4 (0.1) 0.4 (0.0) –7.2 (0.2) –0.6 (0.1) 2.3 (0.1) 0.6 (0.1) 4.0 (1.4) –1.0 (0.6) 2.1

2.2 1.5 1.4 6.0 2.2 2.2 1.3 5.5 1.5

MAP, modified atmosphere packaging

radiation emitted from the surface as a result of its temperature, but also radiation which is reflected from the surroundings of the food, e.g. lighting. Depending on the type of packaging, the reflected radiation can be quite considerable and hence will give an incorrect surface temperature. There was a large difference between the performance of the nine infrared thermometer devices when used in a commercial retail cabinet in a retail outlet. Table 15.8 shows the performance of these using six different packaging materials. Infrared readings were compared with a calibrated thermocouple inserted under the surface of the pack. Out of the five instruments, two (b and g) had errors less than 1 °C, and five less than 2.5 °C, with two further instruments giving unsatisfactory errors. The highest errors of all of the instruments were found with a printed foil pack, which gave the most reflected radiation. It is recommended that angled and brightly lit packs be avoided with an infrared thermometer, and only horizontal or vertical positioned packs in a cabinet be chosen, with the device perpendicular to the top surface. In order to improve the accuracy, lighting must be reduced as much as possible, and distances to take readings must be as short as possible with as consistent a product as the situation allows. If the thermometer is moved from one ambient temperature to another, e.g. room temperature to a chill store, then for best repeatability of measurements it is advisable to allow at least 30 minutes for the instrument to adjust to the new ambient temperature. The thermometer should also be checked regularly against surfaces of known temperature. It is possible to make a relatively cheap black body calibration chamber with black PVC plastic tubing and a copper block. Alternatively, commercial systems are available. The other type of remote sensor is based on infrared video camera-type instruments. Thermal images are displayed on a video display unit, either in colour or in monochrome. A temperature scale on the display gives the temperature that corresponds to the individual colour or shade. Cameras range from low-resolution © 2008, Woodhead Publishing Limited

434

Chilled foods

types, often used for locating victims trapped in collapsed buildings, to very highresolution, sophisticated systems that allow computerised manipulation of the data. Infrared systems have been found to be very useful for industrial control and energy efficiency, with the ability to pick out overheating components and heat losses. They are also being used increasingly for on-line applications in the food industry, either with the hand-held type or with specific models available for this. For example, sealing rollers on microwavable plastic trays can be monitored to ensure evenness of heating, and products emerging from cooking, heating or cooling tunnels can also be screened for evenness of heating or cooling. These are applications where the target is consistent and relative temperatures are more important than accurate temperatures. The readings are instantaneous and the information can be linked directly to control systems. Infrared temperature measurement is unlikely ever to replace electrical temperature measurement for accurate determinations for enforcement of temperature control legislation. However, there are exciting possibilities for their use in routine monitoring and temperature auditing where relative temperatures are very important, but care needs to be exercised in the interpretation of results. Hand-held devices can be used to monitor the surface temperature of cases unloaded from a vehicle on an acceptance or rejection basis, or scan a display cabinet for ‘hot spots’.

15.7 Temperature and time–temperature indicators 15.7.1 Performance of time–temperature indicators Temperature monitoring has been discussed in terms of displaying the temperature readings of the surrounding air or of the food or simulated food itself. However, it is possible to use a physico-chemical mechanism and a resulting colour change to display (a) a current temperature, (b) the crossing of a threshold temperature, or (c) an integration of the temperature and the time of exposure to temperature after activation. Such devices are called temperature indicators (TIs) in the first two cases or time–temperature indicators (TTIs) in the last case. TTIs are based on irreversible mechanical, chemical or enzymatic changes that are expressed as a visible response such as a mechanical deformation, colour development or colour movement. The indicators are normally integrated onto a packaging material that can be attached to the food packaging or the outside of the surrounding or bulk packaging, and can follow the food throughout the chill chain. The type of information than can be provided is one or more of the following:

• • • •

reject or accept on the basis of a colour change temperature abuse above a threshold temperature partial time–temperature history above a threshold temperature full time–temperature history linked to shelf-life.

In order that the devices can be used in commercial situations for monitoring, they © 2008, Woodhead Publishing Limited

Temperature monitoring and measurement

435

should have the following features and be supplied with the following information by the manufacturer:21

• ease of application to food packs; • instructions on the activation of the TTI just before application, including the • •

• • •

temperature at which the device must be stored if it has to be kept at low temperatures (frozen) once manufactured until applied; for TIs, the threshold temperature and its tolerance limits (3 × standard deviation) in ºC, and a response time (inertia) in minutes; for TTIs, the maximum and minimum temperature limits in ºC over which the device will function and the time-to-end-point with the tolerance at sufficient numbers of temperatures throughout the range stated by the manufacturer (above the critical reference temperature in the case of partial history TTIs). The number of temperatures and time-to-end-point combinations must not be less than five; the performance tolerance for the time-to-end-point as in BS 7908: 199919 Category A (up to ±2.5%), Category B (up to ±5%), Category C (up to ±10%), Category D (up to ±20%); for partial history TTIs, the critical reference temperature, i.e. the temperature at which the physico-chemical change is activated to produce an irreversible change in ºC and its tolerance; the storage conditions for these devices so that their performance is unaffected. Also any conditions which could affect the performance apart from temperature, e.g. light.

The devices should be tamper-proof. It is important to appreciate that TTIs are based on physical, chemical or biochemical reactions. Their performance can mimic microbiological changes20–25 or the biochemical and chemical reactions that cause deterioration in the sensory quality of foods. Normally, biochemical reactions change at a faster rate than chemical ones. However, each food will have a different combination of reactions and hence a different rate. In designing an indicator, it may be important that the activation energy of the device is similar to that of the food deteriorations as well as rate of deterioration.26,27 Over 100 patents have been filed on processes that could be used as a basis for indicators. These include changes with temperature based on melting-point temperature, enzyme reaction, polymerisation, electrochemical corrosion, and liquid crystals. The result of the change is usually a colour difference, which can be represented as a static change or a moving band. Available temperature and time– temperature devices have been reviewed.28–30 Many temperature and time– temperature indicators have been launched commercially over the past 10–15 years, but very few have survived. Therefore only a few of the more successful devices will be described. Diffusion-based TTIs 3M (St. Paul, Minn., USA) produce a 3M Monitor Mark® TTI that consists of a © 2008, Woodhead Publishing Limited

436

Chilled foods

paper blotter and track separated by a polyester film layer. The blotter is impregnated with chemicals of specific melting point and a blue dye. If a pre-determined threshold temperature is reached the chemicals melt, and hence the devices are partial history TTIs. The indicator is activated by removing the film, allowing the chemicals and dye to diffuse along the blotter and track. The TTI consists of one or more windows which, as they turn from light grey to dark grey/black, allow the exposure to temperature to be estimated. The diffusion rate increases with temperature above the melt point. Varying tags are produced to correspond to different lengths of time at different melt temperatures (these range from –15 °C to +31 °C). Another diffusion-based TTI produced by 3M is the Freshness Check®. This incorporates a viscoelastic material that migrates into a diffusively light-reflective porous matrix at a temperature dependent rate. The light transmissivity of the porous matrix therefore changes and provides a visual response to temperature changes. Further information on the applicability of 3M TTIs for food and drug use can be found in the literature.24,31 Avery Dennison has also produced TTIs that operate through a diffusion-based reaction that starts when a clear activator label is placed on the base TTI label. The activator label is an acidic substance that diffuses into the base label causing an irreversible reaction (showing as a change from yellow to pink). These TTI are relatively stable prior to use. Enzymatic TTIs Vitsab Checkpoint® (Vitsab AB, Malmö, Sweden) produces a range of TTIs based on an enzymatic release of protons, which changes the colour of a pH indicator from green to yellow. The rate of release is temperature related and the rate can be varied to match the shelf-life and temperature of chilled and frozen foods. The indicator can be stored at room temperature and is activated by pressure which breaks an internal pouch, allowing the components to mix. Each ‘ampoule’ in the TTI contains a different enzyme. One ampoule contains a lipase enzyme and pHindicating dye, and the other a triglyceride substrate. As the pH is lowered via liberation of fatty acids from the triglyceride by the lipase enzyme, a green to yellow colour change occurs that is primarily governed by temperature. The reaction can be controlled by choice of the lipase enzyme and type of triglyceride, to give the required activation energies to match the deterioration of the food. The circular indicator can be printed on flexible or semi-rigid packaging and can be incorporated into, or positioned on, the seal of the packaging. Activation can be post-sealing or during sealing. The TTI is also available mounted on a card, which can be placed either between packs on a pallet or inside a bulk pack. Smolander et al.32 evaluated Vitsab, Fresh-Check and 3M TTIs for monitoring the quality of modified atmosphere packed broiler cuts and found that the performance of the TTIs was closely correlated with spoilage microbes. Polymer-based TTIs Lifelines (Lifelines Technologies, Morris Plains, N.J., USA) have developed several indicators, all of which show a full time–temperature history. The indicators © 2008, Woodhead Publishing Limited

Temperature monitoring and measurement R

R

R

Fresh-Check Indicator

Fresh-Check Indicator

Do not use if center is darker than ring

Do not use if center is darker than ring

Do not use if center is darker than ring

Date of packing

Approximates to ‘sell-by’ date

‘Use-by’ or ‘consume-by’ date

Fig. 15.13

Fresh-Check Indicator

Fresh-Check Indicator

R

Do not use if center is darker than ring

Fresh-Check Indicator

437 R

Do not use if center is darker than ring

Do – not – use

The Lifelines ‘Fresh-Check’ indicator – staged examples.

operate using a solid-state polymerisation reaction in which a colourless acetylenic monomer polymerises, causing the indication area to darken, as a result of accumulated temperature exposure. The indicator is active upon manufacture and therefore requires frozen storage before use. The Lifelines Fresh-Check initially consists of a red label with a dark reference ring surrounding a lighter centre area. When expired, the centre area matches the outer ring. The consumer is advised not to consume the product when the inner ring has become darker than the printed outer one (Fig. 15.13).

15.7.2 Practical use of time–temperature indicators There are some technical difficulties in the use of TTIs compared to other methods of monitoring. The fact that most are applied to the outside of a food pack means that surface temperatures are being used to change the indicators. As long as food packs are in cases, this is probably still a good guide to the food temperature with a tolerance. However, for foods on display, this may give a false indication of shelf-life, owing to radiant heat absorption, unless the effect is eliminated or compensated for. As a means of following the integrity of the chill chain from manufacture up to the point of display, TTIs may have a practical advantage in use over certain other types of monitoring in that they give a simple and individual indication of temperature abuse. The initial potential of TTIs has not, however, been realised to date. Cost, reliability and applicability have all been factors preventing their uptake. The cost of TTIs had been estimated at $0.02 to $0.20 per unit.33 Many of the technical production issues relating to reliablility and reproducibility have been overcome, but issues relating to the relationship between temperature and food quality for varied food types is still current. Similar foods often respond differently to temperature and there is a need for the TTI to mimic the kinetic responses of each particular food to a reasonable level. One way to achieve this is through modelling of the food behaviour.34 A survey of 511 consumers, carried out by the National Consumer Council,35 indicated that almost all respondents (95%) thought that TTIs were a good idea, but only grasped their concept after some explanation, indicating that substantial © 2008, Woodhead Publishing Limited

438

Chilled foods

publicity or an education campaign would be required. Use of TTIs would have to be in conjunction with the durability date, with clear instructions about what to do when the indicator changed colour. The relationship and possible conflict between the indication of the TTI and the durability date on the food was considered a problem. In the retail situation, nearly half those questioned would trust the TTI response if it had not changed but the product was beyond its durability date. If the TTI changed before the end of the durability date when stored at home, the majority of respondents (57%) would use their own judgement in deciding whether a food was safe to eat, with at least 25% putting some of the blame on the food suppliers. However, the value of TTIs was recognised for raising confidence in retail handling, and improving hygiene practices when food is taken home and stored in refrigerators. Concerns over their technical performance (accuracy and reproducibility) were expressed, and the question of whether they could be tampered with or interfered with was also raised. These concerns are shared by the food industry, and have been addressed by the publication of a technical specification for time–temperature indicators.19,36 Reluctance by retailers to use an indicator on retail packs for consumer use is understandable because of the difficulties that are raised. To date, no permanent commercial use of TTIs on retail packs has been adopted in the UK. However, both French and Spanish supermarket chains have used Lifelines Fresh-Check indicators on selected items of chilled food for quite extended trials, but have decided not to use them on a long-term basis. TTIs have found more extensive use in the medical field to ensure that vaccines and medicines are transported and stored correctly. In addition, their use to ensure the integrity of the chill chain up to retail sale by having indicators on outer cartons or pallets as a further check is being examined by the chill and frozen food industry and retailers. The advantages of TTIs over other types of monitoring equipment of giving easy and clear answers to whether temperature abuse has taken place makes them an attractive addition to assuring safety and quality to the consumer.

15.8 Radio frequency identification tags Radio frequency identification (RFID) tags can be attached to products to transit real-time information to the user. They are commonly used in traceability of animals. In its simplest form, the RFID tag contains a small transponder and antenna that communicates with the user antenna via a unique number or alphanumerical sequence. Tags can be embedded in food containers and can be interrogated without exact knowledge of their location or any direct contact. Information on temperature, humidity, product information, etc. can be held by RFID tags that can contain up to 1 Mb of data. Tags can be read only or read/write, and are classified as active or passive. Active tags use battery power and can operate at a distance of up to 50 m from the reader. Passive tags can be read up to 5 m away from the reader and are powered by the reader. RFID frequencies range from low (125 kHz) to © 2008, Woodhead Publishing Limited

Temperature monitoring and measurement

439

UHF (850–900 MHz) and microwave frequencies (2.45 GHz), with the low frequency tags being cheaper, using less power and better able to penetrate nonmetallic objects. The cost of RFID tags varies from around $0.50 to $1.00 but for the technology to be truly competitive it is estimated that tags must cost in the region of $0.01 to $0.05 or less.37

15.9 Temperature modelling and control The use of computer modelling as an aid to predict what is happening in complex systems is well established and has been applied in the refrigerated food sector. It has enabled food temperatures to be predicted if the conditions of use of the refrigerated system are known. A comprehensive review of food models can be obtained from an edition of the International Journal of Refrigeration with special emphasis on data and models on food refrigeration.38

15.9.1 Short-distance delivery vehicles The difficulties in monitoring and maintaining food temperatures in small delivery vehicles with many stops at retail outlets have been examined by the University of Bristol,39 and a commercial computer program (CoolVan40) has been developed to aid the design and operation of these vehicles. The programme examines the changes to the temperature of the air inside the vehicle. The ingress of heat through the insulation from the outside air and solar radiation are taken into account, as is the infiltration of air through the back door when the vehicle is travelling and stationary with the door open. The thermal properties of the insulation of the vehicle and the age of the vehicle enables the reduced heat transfer coefficient of the walls to be predicted and each side of the vehicle can be treated separately. The infiltration of air during door openings is one of the major factors in heat gain. Transparent plastic strip curtains have been recommended as a way of reducing ingress of air, and measurements showed that blowing air from the cooling system directly at the curtains helped to counteract warm air entering the gaps at the top of the curtain. At each stage of the program’s development it was tested against measured data. The program was able to predict the mean temperature of the food inside the vehicle with an accuracy better than 1 °C at any time throughout the journey. However, food temperature within the vehicle actually varied by more than 5 °C at one time, due to the uneven temperature within the vehicle.

15.9.2 Retail display Computer modelling has been developed to examine the way retail cabinets behave and so to improve their design. Computational fluid dynamics (CFD) is a tool that enables changes to be made to the computer model to see which effect produces the best results, before checking this against actual measurements. It has © 2008, Woodhead Publishing Limited

440

Chilled foods

been applied to study design of retail display cabinets41–44 and also the effects of retail cabinets on supermarket environments, in particular the cold air spillage from frozen food cabinets to the aisles.45 The model predicted temperatures at floor level of between 5 °C and 15 °C, whereas measured values ranged from 13 °C to 22 °C. The model is better at showing trends than actually predicting or following actual temperatures. There is potential for computer models to be used to improve the design of all refrigerated equipment in the food chain and improve energy efficiency whilst maintaining food temperatures.

15.10 Sources of information and advice NICHOLAS J V AND WHITE D R,

Traceable Temperatures: An Introduction to Temperature Measurement and Calibration. John Wiley and Sons, 2001. LEIGH J, Temperature Measurement and Control. IET, 1988. JABLONSKI J R, TQM implementation. In: Implementing Total Quality Management: An Overview. Pfeiffer and Co., San Diego, CA, 1991. WEBB N B AND MARSDEN J L, Relationship of the HACCP system to total quality management. In: HACCP in Meat, Poultry and Fish Processing, Pearson A M and Dutson T R (eds),. Advances in Food Research Vol. 10, 1995 Blackie Academic and Professional, Chapman and Hall, Glasgow, 1995. JAMES S J, Controlling food temperature during production, distribution and retail. New Food, 1999, 1(3), pp 35–45. GIGIEL A J, JAMES S J AND EVANS J A, Controlling temperature during distribution and retail. Proceedings of the 3rd Karlsruhe Nutrition Symposium, in ‘European Research towards Safer and Better Food’, 18–20 October, Gaukel V. and Speiss W. E. L. (eds), pp 284–292 1998. International Journal of Refrigeration Issue with Special Emphasis on Data and Models on Food Refrigeration, Volume 29, Issue 6. Nicolaï B. M. and Pham Q. T. (eds), pp 845– 1054, September 2006.

15.11 References 1 REGULATION (EC) No. 852/2004 of the European Parliament and of the Council of 29 April 2004 on the hygiene of foodstuffs. Official Journal of the European Union. 2 REGULATION (EC) No. 853/2004 of the European Parliament and of the Council of 29 April 2004 laying down specific hygiene rules for food of animal origin. Official Journal of the European Union. 3 REGULATION (EC) No. 854/2004 of the European Parliament and of the Council of 29 April 2004 laying down specific rules for the organisation of official controls on products of animal origin intended for human consumption. Official Journal of the European Union. 4 THE FOOD HYGIENE (ENGLAND) REGULATIONS 2006. SI 2006 No. 14. The Stationery Office Limited, London. 5 THE FOOD HYGIENE (ENGLAND) (AMENDMENT) REGULATIONS 2007. SI 2007 No. 56. The Stationery Office Limited, London. 6 FOOD LAW PRACTICE GUIDANCE (England). Food Standards Agency. 7 FOOD LAW CODE OF PRACTICE (England). Food Standards Agency. www.food.gov.uk/ enforcement/foodlaw/foodlawcop/copengland 8 BS EN ISO 23953-1 (2005). Refrigerated Display Cabinets. Part 1: Vocabulary.

© 2008, Woodhead Publishing Limited

Temperature monitoring and measurement

441

9 BS EN ISO 23953-1 (2005). Refrigerated Display Cabinets. Part 2: Classification, requirements and test conditions. 10 JAMES S J, EVANS J A AND STANTON J (1989). Performance of Domestic Refrigerators, Proceedings of 11th International Conference on Home Economics, 13–15 Sept., Middlesex Polytechnic, UK. 11 TUCKER G (1995). Guideline No. 1: Guidelines for the use of thermal simulation systems in the chilled food industry. Campden and Chorleywood Food Research Association. 12 EVANS J A, SCARCELLI S, SWAIN M V L (2007). Temperature and energy performance of refrigerated retail display and commercial catering cabinets under test conditions, International Journal of Refrigeration, 30, 398–408. 13 BS EN 13485 (2001). Thermometers for measuring the air and product temperature for the transport, storage and distribution of chilled, frozen, deep-frozen/quick-frozen food and ice cream – Tests, performance, suitability. ISBN 0 580 39467 0. 14 EN 12830 (1999). Temperature recorders for the transport, storage and distribution of chilled, frozen, deep-frozen/quick-frozen food and ice cream – Tests, performance, suitability. ISBN 0 580 32860 0. 15 BS EN 13486 (2002). Temperature recorders and thermometers for the transport, storage and distribution of chilled, frozen, deep-frozen/quick-frozen food and ice cream – Periodic verification. ISBN 0 580 38909. 16 FAIRHURST D (1990). Temperature monitoring in the cold and chill chain (A one day seminar sponsored by MAFF, 30.01.1990). Food Science Division Report, MAFF, London. 17 KLEER J, PASTARI A, WIEGNER J AND SINELL H (1991). Recording temperature patterns with modern recording systems (original German), Fleischwirtschaft, 71(6), 698–704. 18 JAMES S J AND EVANS J A (1994). The accuracy of non contact temperature measurement of chilled and frozen food, IChemE Food Engineering Symposium, University of Bath, 19–21 Sept. 1994, Publication No. 106, FPERC, University of Bristol. 19 BS 7908 (1999). Packaging–temperature and time–temperature indicator – performance specification and reference testing. British Standards Institution, London. 20 MENDOZA T F, WELT B A, OTWELL S, TEIXEIRA A A, KRISTONSSON H, AND BALABAN M O (2004). Kinetic Parameter Estimation of Time–temperature Integrators Intended for Use with Packaged Fresh Seafood, JFS: Food Microbiology and Safety, 69(3). 21 MOORE C M AND SHELDON B W (2003). Evaluation of time–temperature integrators for tracking poultry product quality throughout the chill chain, Journal of Food Protection, 66(2), February, 287–292. 22 MOORE C M AND SHELDON B W (2003). Use of time–temperature integrators and predictive modeling to evaluate microbiological quality loss in poultry products, Journal of Food Protection, 66(2), February, 280–286. 23 WELT B A., SAGE D S AND BERGER K L (2003). Performance specification of time– temperature integrators designed to protect against botulism in refrigerated fresh foods, Journal of Food Science, 68(1), January, 2–9. 24 SHIMONI E, ANDERSON E M AND LABUZA T P (2001). Reliability of time temperature indicators under temperature abuse, Journal of Food Science, 66(9), 1337–1340. 25 GIANNAKOUROU M C, KOUTSOUMANIS K, NYCHAS G J E, TAOUKIS P S (2005). Field evaluation of the application of time temperature integrators for monitoring fish quality in the chill chain, International Journal of Food Microbiology, 102, 323–336. 26 TAOUKIS P S AND LABUZA T P (1989). Application of time–temperature indicators as shelf-life monitors of food product, Journal of Food Science, 54(4) 783–788. 27 TAOUKIS P S AND LABUZA T P (1989). Reliability of time–temperature indicators as food quality monitors under non-isothermal conditions, Journal of Food Science, 54(4) 789– 792. 28 BALLANTYNE A (1988). An evaluation of time–temperature indicators, Technical Memorandum No. 473, Campden and Chorleywood Food Research Association. © 2008, Woodhead Publishing Limited

442

Chilled foods

29 SELMAN J D AND BALLANTYNE A (1988). Time–temperature indicators: Do they work? Food Manufacture, 63(12), 36–38, 49. 30 SELMAN J D (1990). Time–temperature indicators: How they work, Food Manufacture, 65(8), 30–31 and 33–34. 31 FU B, TAOUKIS P S AND LABUZA T P (1992). Theoretical design of a variable activation energy time–temperature integrator for prediction of food or drug shelf life, Drug Development and Industrial Pharmacy, 18(8), 829–850. 32 SMOLANDER M, ALAKOMI H-L., RITVANEN T, VAINIONPA J AND AHVENAINEN R (2004). Monitoring of the quality of modified atmosphere packaged broiler cuts stored in different temperature conditions. A. Time–temperature indicators as quality-indicating tools, Food Control, 15, 217–229. 33 TAOUKIS P S AND LABUZA T P (2003). Time–temperature indicators (TTIs). In R. Ahvenainen (Ed.), Novel Food Packaging Techniques (pp 103–126). Cambridge, UK: Woodhead Publishing Ltd. 34 TAOUKIS P S (2001). Modelling the use of time–temperature indicators in distribution and stock rotation. In L Tijskens, M Hertog and B Nicolai (Eds.), Food Process Modelling (p 512). Cambridge, UK: Woodhead Publishing Ltd. 35 MINISTRY OF AGRICULTURE, FISHERIES AND FOOD PUBLICATION (1991). Time–temperature indicators: Research into consumer attitudes and behaviour, National Consumer Council. 36 GEORGE R M AND SHAW R (1992). A food industry specification for defining the technical standards and procedures for the evaluation of temperature and time–temperature indicators, Technical Manual No. 35, Campden and Chorleywood Food Research Association. 37 KERRY J P, O’GRADY M N, HOGAN S A (2006). Past, current and potential utilisation of active and intelligent packaging systems for meat and muscle-based products: A review, Meat Science, 74, 113–130. 38 INTERNATIONAL JOURNAL OF REFRIGERATION (2006). Issue with Special Emphasis on Data and Models on Food Refrigeration, Volume 29, Issue 6. Edited by Nicolaï B M and Pham Q T pp 845–1054, September. 39 GIGIEL A J, JAMES S J AND EVANS J A (1998). Controlling Temperature During Distribution and Retail, Proceeding of the 3rd Karlsruhe Nutrition Symposium, European Research towards Safer and Better Food, 18–20 October 1998. Edited by Gaukel V and Speiss W E L, pp 284–92. 40 FRPERC NEWSLETTER (1997). Predicting food temperatures in refrigerated transport, Number 17, May, pp 4–5, University of Bristol. 41 STRIBLING D, TASSOU S A MARRIOT D (1997). A two dimensional CFD model of a refrigerated display case. ASHRAE Transactions: Research Vol. 103, Part 1, pp 88–95. 42 CORTELLA G (2002). CFD aided retail cabinets design, Computers and Electronics in Agriculture, 34, 43–66. 43 NAVAZ H K, FARAMARZI R, GHARIB M, DABIRI D AND MODARESS D (2002). The application of advanced methods in analyzing the performance of the air curtain in a refrigerated display case, J Fluids Eng – Trans ASME, 124(3), 756–764. 44 FOSTER A M, MADGE M AND EVANS J A (2005). The use of CFD to improve the performance of a chilled multi-deck retail display cabinet, International Journal of Refrigeration, 28, 698–705. 45 FOSTER A M AND QUARINI G L (1998). ‘Using advanced modelling techniques to reduce the cold spillage from retail display cabinets into supermarket stores’ ICR/IIR Conference, Refrigerated Transport, Storage and Retail Display, Cambridge, 29 March–1 April.

© 2008, Woodhead Publishing Limited

Part III Microbiological hazards

© 2008, Woodhead Publishing Limited

16 Chilled foods microbiology S. J. Walker and G. Betts, Campden and Chorleywood Food Research Association, UK

16.1 Introduction Chilled foods represent a large and rapidly developing market with an extremely wide range of food types. Traditionally these were simple meat, poultry, fish and dairy products but recent trends have moved towards a greater variety and more complex products (Stringer and Dennis, 2000). As more innovative products are produced, the variety of ingredients have also increased. Many of these ingredients are sourced around the world and relatively little may be known about their microbiological status. The numbers and types of micro-organisms that may be isolated from the full range of chilled foods are very diverse. During the storage of chill products, the microbial flora of the product is not static but affected by many factors, principally the time and temperatures of storage. The spoilage and safety of chilled foods is a complex phenomenon involving physico-chemical, biochemical and biological changes. Often these interact and changes in one affect the rate of change in the others. This review will be concerned only with microbiological issues in relation to chilled foods. With developments in the manufacture and transport of chilled foods, these items may now be rapidly disseminated over a wide geographical area, i.e. different countries and sometimes continents. Therefore should a microbiological issue arise it may be similarly widely spread. Consequently, the microbiological status of chilled foods has become more significant. Greater surveillance both within and between countries will allow such microbiological issues to be more rapidly identified, traced and resolved. © 2008, Woodhead Publishing Limited

446

Chilled foods 10

8 30 °C Log (10) Numbers

20 °C

10 °C

6

4

2

0 0

10

20

30

Time (hr)

Fig. 16.1

Effect of temperature on the growth of micro-organisms.

16.2 Why chill? The effect of reducing temperature is to reduce the rate of food deterioration. This applies not only to the chemical and biochemical changes in foods but also to the activities of micro-organisms. The effect of temperature on microbial growth is shown in Fig. 16.1. As the storage temperature decreases, the lag phase before growth (time before an increase in numbers is apparent) extends and the rate of growth decreases. In addition, as the minimum temperature for growth is approached, the maximum population size attainable often decreases. On a cellular basis, the effect of temperature on growth is a complex issue involving the cell membrane structure, substrate uptake, respiration and other enzyme activities. These have been discussed by Herbert (1989). The range of temperatures over which micro-organisms can grow is extremely wide. Michener and Elliott (1964) reported that a number of micro-organisms, mainly yeasts, were able to grow below 0 °C and a pink yeast isolated from oysters was reported to grow at –34 °C. Therefore, chilling alone cannot be relied upon to prevent all microbial growth. The use of chill temperatures will, however, reduce the rate and extent of microbial growth.

16.3 Classification of growth Microbiologists have attempted to characterise micro-organisms based on their © 2008, Woodhead Publishing Limited

Chilled foods microbiology 447 abilities to grow at various temperatures. Most commonly, the cardinal temperatures for growth (minimum, optimum and maximum growth temperatures) are used. With chilled foods, the factor of most concern is the minimum growth temperature (MGT), which represents the lowest temperature at which growth of a particular micro-organism can occur. If the MGT of a micro-organism is greater than 10 °C, then this micro-organism will not grow during chill storage. Whilst MGT values for micro-organisms have been published, care is needed. If the time period for the investigation reporting this value was too short, or sampling intervals too widely spaced, the resultant value will be erroneous. For example, although an MGT of –0.4 °C has been reported for Listeria monocytogenes, the lag phase before growth was in excess of 15 days (Walker et al., 1990a). Had the study terminated before this time, the reported MGT would have been higher. The MGT is affected by other factors including the pH, salt, preservatives and previous heat treatments. A true estimate of the MGT can be determined only when other factors are optimal for growth. If a micro-organism is stored below its MGT, gradual death may occur, but often the micro-organism will survive and growth will resume should the temperature subsequently be raised. It was noted by Alcock (1984) that the survival of salmonellae was worse at temperatures just below the MGT compared with lower temperatures. Storage at temperatures below the minimum for growth should not be considered to be a lethal process for micro-organisms as in many cases, growth will resume if the temperature is subsequently raised. The optimum growth temperature represents the temperature at which the biochemical processes governing growth of a particular micro-organism are overall operating most efficiently. At this temperature, the lag phase before growth is minimised and the growth rate maximised. As the temperature rises above the optimum, the rate of growth decreases until the maximum growth temperature is reached. In general, the maximum growth temperature is only a few degrees (Celsius) higher than the optimum. With some specialised micro-organisms, isolated from hot springs, the maximum growth temperature may exceed 90 °C (Jay, 1978). At temperatures just above the maximum for growth, cell injury starts to occur. If the temperature is subsequently reduced, then growth may resume, although a period of time may be required to permit cell repair. At higher temperatures, the inactivation of one or more critical enzymes in the microorganism becomes irreversible and cell damage occurs, leading to cell death. Such micro-organisms will not be able to repair and resume growth if temperatures are reduced. The concepts of cell injury and death have been discussed by Gould (1989b). Based on the relative positions of the cardinal temperatures, micro-organisms can be divided into four main groups, viz., psychrophile, psychrotroph, mesophile and thermophile (Table 16.1). With chilled foods, the groups of most concern are the psychrophiles and psychrotrophs. In the past, these terms have been used synonymously, which has led to much confusion. It is now accepted that the term ‘psychrophile’ should only be used for micro-organisms which have a low (i.e. ≤ 20 °C) maximum growth temperature (Eddy, 1960). True psychrophiles are rare © 2008, Woodhead Publishing Limited

448

Chilled foods

Table 16.1 Classification of microbial growth (Jay, 1978; Walker and Stringer, 1990; Morita, 1973) Temperature (°C) Minimum Optimum Maximum

Psychrophile

Psychotroph

Mesophile

Thermophile

< 0–5 12–18 20

< 0–5 20–30 (35)a 35 (40–42)a

(5 to)a 10 30–40 45

(30 to)a 40 55–65 (70 to)a > 80

a

Figures in parentheses are occasionally recorded for micro-organisms assigned to a particular classification.

in food microbiology and generally limited to some micro-organisms from deepsea fish. The major spoilage micro-organisms of chilled foods are psychrotrophic in nature.

16.4 The impact of microbial growth Under suitable conditions, most micro-organisms will grow or multiply. Bacteria multiply by the process of binary fission, i.e. each cell divides to form two daughter cells. Consequently, the bacterial population undergoes an exponential increase in numbers. Under ideal conditions some bacteria may grow and divide every 20 minutes and so one bacterial cell may increase to 16 million cells in 8 hours. Under adverse conditions, e.g. chilled storage, the generation time (doubling time) will be increased. For example with an increased time of two hours, the population obtained after 8 hours would be only 16 cells. Even under ideal conditions, growth does not continue unchecked and is limited by a range of factors including the depletion of nutrients, build-up of toxic by-products, changes to the environmental conditions or a lack of space.

16.4.1 Food spoilage During growth in foods, bacteria will consume nutrients from the food and produce metabolic by-products such as gases or acids. In addition, they may produce a number of enzymes which results in the breakdown of the cell structure or of components (e.g. lipases and proteases). When only a few spoilage micro-organisms are present, the consequences of growth may not be apparent. If however, the micro-organisms have multiplied then the production of gases, acid, off-odours, off-flavours or deterioration in structure in the food may become unacceptable. In addition, the number of micro-organisms may be apparent as a visible colony, production of slime or an increase in the turbidity of liquids. Some of the enzymes produced by spoilage bacteria may remain active, even when a thermal process has destroyed the causative micro-organisms in the food. The relationship between microbial numbers and food spoilage is complex and depends on the number, type and activity of the micro-organisms present, the type of food and the intrinsic and extrinsic conditions. In some cases this is well © 2008, Woodhead Publishing Limited

Chilled foods microbiology 449 understood, e.g. vacuum packed cod (Gram and Huss, 1996). In general, a greater understanding is needed of the relationship between specific spoilage microorganisms in particular foods and the deterioration in sensory quality.

16.4.2 Food-borne pathogens With many human pathogens, the greater the number of cells consumed, the greater the chance of microbial invasion, as the larger number of cells may be able to evade/ swamp the body’s defence mechanism. Higher numbers may also result in a shorter incubation period before the onset of disease. Consequently, control, and preferably inhibition, of growth in foods is essential. However, with some invasive pathogens (e.g. viruses, Campylobacter), the infectious dose is low and growth in the food may not be necessary. Other pathogenic micro-organisms may produce a toxin in the food which results in disease. Preformed toxins are usually produced at high cells densities and so usually growth has occurred. If the toxin is heat stable, it may remain although all micro-organisms have been eliminated from the food. Consequently it is important to control growth at all stages of the chill chain.

16.5 Factors affecting the microflora of chilled foods 16.5.1 Initial microflora With healthy animal and plant tissues, microbial contamination is absent or at a low level except for the exterior surfaces. For example, fresh muscle from healthy animals is usually microbiologically sterile, and aseptically drawn milk from healthy cows contains only a few micro-organisms (mainly streptococci and micrococci) derived from the teat canal. Similarly, the interior of healthy undamaged vegetables does not contain micro-organisms although the exterior may be contaminated with a wide range of micro-organisms of soil origin. During slaughter or harvesting, subsequent processing and packaging, these raw materials become contaminated from a wide range of sites. Typically, these sites include water, air, dust, soil, hides/fleece/feathers, animals, people, equipment and other food materials. Consequently, a large range of micro-organisms can be isolated from foods. Those which are able to grow may potentially give rise to microbial spoilage or public health issues. The hygienic practices of all food operations, from slaughter/harvesting through retail sale to consumer use, will affect the level of microbial contamination of products. In general, the lower the initial level of contamination, the greater the time until microbial spoilage is evident.

16.5.2 Food type The intrinsic properties (e.g. pH, water activity, acidity, natural antimicrobials) of different foods vary greatly. Such factors affect the ability of micro-organisms to grow and the rate of growth and will be discussed in more detail in subsequent © 2008, Woodhead Publishing Limited

450

Chilled foods

sections of this chapter. With different food types, the nutritional status varies although foods are generally not nutritionally limiting for micro-organisms. Foods rich in nutrients (e.g. meat, milk, fish) permit faster growth than those with a lower nutritional status (e.g. vegetables) and so are more prone to spoilage. Slaughter and harvesting practices may affect the intrinsic properties of a food. For example, poor practices in the husbandry and slaughter of pigs may lead to pork being classified as DFD (dark, firm, dry) or PSE (pale, soft, exudative), both of which are more prone to spoilage than ‘normal’ pork. With DFD meat, the pH is higher, so permitting faster growth, whilst nutrient leakage and protein denaturation from PSE meat also allow more rapid microbial proliferation. Even within a single food ingredient or product, variations in the pH, aw and redox potential may occur and so affect the nature and rate of microbial multiplication. The situation may be further complicated in multi-component foods where migration of nutrients and gradients of pH, aw and preservatives may occur. In addition, micro-organisms unable to grow on one ingredient may come into contact with a more favourable environment and so permit growth.

16.5.3 Processing Chill storage The time of storage will affect microbial numbers. Generally, microbial numbers increase with time in chilled foods at neutral pH values, low salt concentrations and the absence of preservatives. However, low pH values or high salt concentrations in foods may cause microbial stasis, injury or even death. At chill temperatures however, the rate of death is often reduced and so the micro-organism may survive for longer periods compared with higher (e.g. ambient) temperatures. In many cases a combination of processing and preservation factors may be used to achieve a safe, high quality product with an acceptable shelf-life. Such combination treatments have been reviewed by Gould (1996). The ability of individual micro-organisms to grow and their rates of growth are affected by temperature. As discussed previously, some micro-organisms (mainly psychrotrophs) are better adapted to growth at chill temperatures. Therefore during chill storage not only will the total number of micro-organisms change, but also the composition of the microflora will alter. For example, with freshly drawn milk, the microflora is dominated by Gram-positive cocci and rods, which may spoil the product by souring if stored at warm temperatures. At chill temperatures, these micro-organisms are largely unable to grow and the microflora rapidly becomes dominated by psychrotrophic Gram-negative rod-shaped bacteria (most commonly Pseudomonas spp.) (Neill, 1974). A similar change in the microflora composition has also been reported for other chill-stored foods (Huis in’t Veld, 1996). Heating As part of their manufacture, many chilled foods undergo a heating process. This © 2008, Woodhead Publishing Limited

Chilled foods microbiology 451 will reduce microbial numbers, generally resulting in a pasteurised rather than a sterilised product, otherwise chill storage would be unnecessary. Food pasteurisation treatments have been reviewed by Gaze (1992). The degree of heat applied will affect the types of micro-organisms able to survive. In general, the Gramnegative rod-shaped bacteria, which proliferate in chilled foods, are sensitive to heat and are readily eliminated. Although these bacteria may be isolated from, and even spoil, heated foods, their presence is usually attributable to post-heating contamination. Some Gram-positive bacteria are tolerant to mild heat and classified as thermoduric (e.g. some Lactobacillus, Streptococcus and Micrococcus species, Jay, 1978). However pasteurisation processes are designed to destroy all vegetative cells. Other bacteria however, produce heat-resistant bodies, called spores, which may survive. The genera of concern are Bacillus and Clostridium species, which include both pathogenic and spoilage strains. Whilst these bacteria are generally out-competed in chilled foods by the Gram-negative rod-shaped bacteria, Bacillus and Clostridium species may grow relatively unhindered in heated foods subsequently stored chilled. Acidification Several types of chilled foods are naturally acidic (e.g. fruit juices) or acidified using either a fermentation process (e.g. yoghurt) or by the direct addition of acids (e.g. coleslaw). As with temperature, micro-organisms have pH limits for growth. The pH optima for most pathogenic bacteria is usually in the range 6.8–7.4 (Jay, 1978), which is similar to the human body pH in which they are adapted to grow. Typical minimum pH values for growth are shown in Table 16.2. The minimum pH for the major spoilage bacteria of meat, poultry and dairy products is approximately 5.0 whereas other microbial types, in particular yeasts and moulds, may grow at pH values of 3.0 or less. Consequently, mildly acidified products may be spoiled by acid-tolerant bacteria (lactic acid bacteria and some Enterobacteriaceae) whilst more acid products are spoiled by yeasts and moulds. Both pH and temperature interact, and the minimum pH for growth at optimal temperatures may be significantly less than that at chill temperatures (George et al., 1988). At pH values below the minimum for growth, some micro-organisms will die rapidly in the food, whilst others may persist for the life of the product. Of particular concern in acid foods is the pathogen E. coli O157: H7, which is more acid tolerant than other pathogens. It may grow at pH values of 4.0 or below and survive for considerable periods at lower pH values (Conner and Kotrola, 1995; Deng et al., 1999). In addition to the pH, the acid type used affects the microbial stability of the foods. The organic acids (lactic, acetic, citric and malic) are more antimicrobial than the inorganic acids (hydrochloric, sulphuric). Care is needed with published literature, as the minimum pH values reported often have used inorganic acids. Therefore the minimum pH for growth in foods is often higher than that quoted, as organic acids are present. Within the organic acids, the order of decreasing antimicrobial efficiency is usually acetic, lactic, citric then malic acid. With the © 2008, Woodhead Publishing Limited

452

Chilled foods

Table 16.2 Typical minimum pH and aw values for growth of microorganisms (Anon.,1991b; Gould, 1989; Mitscherlich and Marth,1984; ACMSF, 1995) Micro-organism Bacillus cereus Campylobacter jejuni Clostridium botulinum (non-proteolytic) Clostridium botulinum (proteolytic) Clostridium perfringens Escherichia coli Escherichia coli O157:H7 Lactobacillus species Pseudomonas species Salmonella species Staphylococcus aureus Many yeasts and moulds Yersinia enterocolitica a

Minimum pH

Minimum aw

4.9 5.3 5.0 4.6 5.0 4.4 3.8–4.2 3–3.5 5.0 4.0 4.0 (4.6)a < 2.0 4.6

0.91 0.985 0.96 0.93 0.93 0.95 0.97 0.95 0.95 0.95 0.86 0.8–0.6 0.95

Minimum pH with toxin production.

organic acids, the undissociated form of the acid is effective against microorganisms and the degree of dissociation is dependent on the pH of the food. Organic acids and their use in food systems have been discussed by Kabara and Eklund (1991). The pH and acid composition does not remain constant during the life of some foods. Changes in pH will affect the types of micro-organisms able to grow and their growth rates. With some foods, fermentation results in a pH decrease during storage whilst in others an increase can be noted. For example, during maturation of mould-ripened cheeses, the pH value of cheese near the surfaces increases owing to proteolytic activity of the mould, and this has been related to the ability of Listeria monocytogenes to grow in these products, but not in the unripened cheeses (Terplan et al., 1987). Reduced aw The aw is a measure of the amount of water available in a food which may be used for microbial growth. As the aw of a food is reduced, the number of microorganisms able to grow and their rate of growth is also reduced (Sperber, 1983) (Table 16.2). The aw of a food may be reduced either by the removal of water (i.e. drying) or by the addition of solutes (e.g. salt or sugar). In response to diet and health issues, many jam and sauce products have reduced their sugar content. Thus the intrinsic preservation system (i.e. low aw) of the product has been compromised and some micro-organisms, mainly yeasts, may now grow. These products generally recommend refrigeration after opening to prevent microbial growth. The aw of a product may interact with other preservation factors, including temperature, to maintain the safety of chilled foods (Glass and Doyle, 1991). In general, yeasts and moulds are more tolerant than bacteria of low aw values in foods (Jay, 1978). As bacterial growth is largely inhibited, yeasts and moulds may then grow and cause the spoilage defects in such products. © 2008, Woodhead Publishing Limited

Chilled foods microbiology 453 Preservatives In order to maintain their microbial stability, many chilled products contain natural or added preservatives, e.g. salt, nitrite, benzoate, sorbate. The presence of these compounds affects the type and rate of product spoilage that may occur. Their applications and mechanisms of action have been reviewed in Russell and Gould (1991). As discussed previously, Pseudomonas species tend to predominate on chilled fresh meat. The addition of curing salts (i.e. sodium chloride and potassium nitrite) to pork meat to form bacon, largely inhibits the growth of these microorganisms and spoilage is caused by other microbial groups (e.g. micrococci, staphylococci, lactic acid bacteria) (Gardner, 1983; Borch et al., 1996). Similarly, the British sausage is largely a fresh meat product, but is preserved by the addition of sulphite. This prevents the growth of the Pseudomonas species, and microbial spoilage will be caused by sulphite-resistant Brochothrix thermosphacta or yeasts (Gardner, 1983). The number and type of micro-organisms able to grow in preservative-containing chilled foods depend on the food type, preservative type, pH of the food, preservative concentration, time of storage and other preservation mechanisms in the food. Overall, yeasts and moulds tend to be more resistant to preservatives compared with bacteria and so may dominate the final spoilage microflora. Recent trends in food processing have tended to reduce or eliminate the use of preservatives. Care is needed with such an approach, as even small changes may compromise the product safety and microbiological stability. Storage atmosphere The use of modified atmospheres, including vacuum packaging, for the storage of chilled foods is increasing. Often these are chosen to maintain sensory characteristics of a product, but many will also inhibit or retard the development of the ‘normal’ spoilage microflora. Pseudomonas species, the major spoilage group in chilled proteinaceous foods, require the presence of oxygen to grow. Therefore, the use of vacuum packaging or modified atmospheres excluding oxygen will prevent the growth of this microbial group. Whilst other micro-organisms can grow in the absence of oxygen, they generally grow more slowly and so the time to microbial spoilage is increased. The spoilage microflora of vacuum-packed meats is usually dominated by lactic acid bacteria or Brochothrix thermosphacta (Borch et al., 1996). In some cases, Enterobacteriaceae or coliforms may cause the spoilage of vacuum-packed and modified-atmosphere-packed (MAP) foods (Gill and Molin, 1991). Most commercial MAP gas mixtures for chilled food usually contain a combination of carbon dioxide, nitrogen and oxygen. The inhibition of bacteria becomes more pronounced as the amount of carbon dioxide increases. The effects of carbon dioxide on microbial growth have been discussed by Gill and Molin (1991). More recently, the use of other gases (including the noble gases) and high levels of oxygen have been used to extend the shelf-life of chilled foods (Day, 2000). Good temperature control is essential to obtain the maximum potential benefits of modified-atmosphere and vacuum packing. Should temperature abuse occur, © 2008, Woodhead Publishing Limited

454

Chilled foods

the rate of spoilage will be similar to that without the atmosphere. It has been suggested that modified atmospheres will inhibit the ‘normal’ spoilage microflora of a food, but the growth of some anaerobic or facultatively anaerobic pathogens (e.g. Clostridium species, Listeria monocytogenes, Yersinia enterocolitica, Salmonella species and Aeromonas hydrophila) will be largely unaffected. Consequently the food may appear satisfactory but contain food-poisoning microorganisms. Of particular concern is the potential for growth of psychrotrophic Clostridium botulinum, which has been addressed by Betts (1996). In general, moulds require oxygen for growth and so are unlikely to create problems in vacuum-packaged or MAP (excluding oxygen) foods. Conversely, many yeasts can grow in the presence or absence of oxygen, although aerobic growth tends to be more efficient, thereby permitting more rapid growth. Combinations Many chill products do not depend on a single preservation system for their microbial stability, but a combination of the factors described above. These can be effective in controlling microbial growth (Gould, 1996). With such foods, care is needed during their manufacture, distribution and sale because inadequate control of one factor may permit rapid growth. Furthermore, the use of two or more systems in combination may select for a particular microbial type (Gould and Jones, 1989). For example, ‘sous-vide’ processing involves the vacuum packaging of foods, followed by a relatively mild heat treatment (pasteurisation). The heat treatment will eliminate vegetative micro-organisms but not spore-forming bacteria. During subsequent chill storage (up to 30 days) in vacuum packaging, anaerobic spore-forming bacteria, including Cl. botulinum, may grow in the absence of other micro-organisms. In order to prevent this happening, the product should be stored below the minimum temperature for growth of Cl. botulinum, the formulation of the product adjusted to prevent growth, or the heat treatment applied increased (Betts, 1992).

16.6 Spoilage micro-organisms Microbiological spoilage of chilled foods may take diverse forms, but all are generally as a consequence of growth which manifests itself in a change in the sensory characteristics. In the simplest form, this may be due to growth per se and often the production of visible growth, and this is common in moulds which produce large often pigmented colonies. Bacteria and yeasts may also produce visible (sometimes pigmented) colonies on foods. Other forms of spoilage include the production of gases, slime (extracellular polysaccharide material), diffusible pigments and enzymes which may produce softening, rotting, off-odours and offflavours from the breakdown of food components. The taints produced by microbial spoilage have been reviewed by Dainty (1996) and Whitfield (1998). Spoilage is usually most rapid in proteinaceous chilled foods such as red meats, poultry, fish, shellfish, milk and some dairy products. These products allow good © 2008, Woodhead Publishing Limited

Chilled foods microbiology 455 microbial growth as they are highly nutritious, have a high moisture content and relatively neutral pH value. In an attempt to reduce the spoilage rates of these foods, they are often modified as discussed previously. For chilled products, these modifications may not entirely prevent microbial growth and spoilage, but do limit the rate and nature of spoilage. In general the micro-organisms responsible for spoilage of a food are those which are best able to grow in the presence of the preservation mechanisms that are operating within that food. Care is needed to distinguish between those micro-organisms present in spoiled food and those responsible for the spoilage defect (often called specific spoilage organisms or SSO) which may be only a fraction of the microflora (Gram and Huss, 1996). Consequently, the relationship between sensory spoilage and microbial numbers is often only poorly correlated. Traditional microbiology is often of limited value for control of spoilage microorganisms as the time taken to get results represents a significant proportion of the shelf-life. Recently more rapid and molecular techniques have become available for the detection of general or specific spoilage organisms (Venkitanarayanen et al., 1997; Gutiérrez et al., 1997). For discussion in this chapter, spoilage microorganisms have been arbitrarily divided into six categories: Gram-negative (oxidase positive) rod-shape bacteria; coliform enterics; Gram-positive spore-forming bacteria; lactic acid bacteria; other bacteria; yeasts and moulds. 1. Gram-negative (oxidase positive) rod-shaped bacteria Overall, this group comprises the most common spoilage micro-organisms of fresh chilled products. The minimum growth temperatures are often 0–3 °C and they grow relatively rapidly at 5–10 °C. Although they may represent only a small proportion of the initial microflora, they rapidly dominate the microflora of fresh proteinaceous chilled stored foods (Huis in’t Veld, 1996; Cousin, 1982; Gill, 1983). Within this general group, the genus Pseudomonas is most common although other genera include Acinetobacter, Aeromonas, Alcaligenes, Alteromonas, Flavobacterium, Moraxella, Shewenella and Vibrio species (Walker and Stringer, 1990). These micro-organisms are common in the environment, particularly in water, and so many easily contaminate foods. Often they may proliferate on inadequately cleaned surfaces of food processing plant or equipment and so contaminate foods. The Gram-negative (oxidase positive) rods may spoil products by the production of diffusible pigments, slime material on surfaces and enzymes which result in food rots, off-flavours and off-odours (Jay, 1978; Cousin, 1982; Gill, 1983). Some of the enzymes produced by Pseudomonas species are extremely heat-resistant and may produce long-term defects (e.g. rancidity or age-gelation) in thermally processed products with extended shelf-lives. Although well adapted to grow at chill temperatures, this group tends to be sensitive to other factors such as the presence of salt or preservatives, lack of oxygen, low (< 5.5) pH and a low (< 0.98) aw. Should these preservation mechanisms be present in a food, the Gram-negative (oxidase positive) rod-shaped bacteria compete less well and other microbial groups may cause spoilage. Vibrio © 2008, Woodhead Publishing Limited

456

Chilled foods

species are unusual as they tolerate relatively high salt levels and so may cause spoilage on chilled stored bacon and other cured products. Photobacterium phosphorum, a very large marine vibrio is the dominant spoilage micro-organism in vacuum packaged cod (Gram and Huss, 1996). Overall, this group is not heatresistant and so is readily removed by mild thermal treatments. Their presence in heat processed foods is usually as a consequence of post-process contamination. 2. Coliform/enteric bacteria This bacterial group also consists of Gram-negative rods, but these may be distinguished from the above group by a negative oxidase reaction. Traditionally, microbiologists have tended to examine for these groups separately, as their sources, significance and factors affecting growth may differ. This group is frequently used as an indicator of inadequate processing or post-process contamination. Compared with the Gram-negative (oxidase positive) rod-shaped bacteria, the coliform-enteric group is generally less well adapted to growth at temperatures of less than 5–10 °C although many may grow at temperatures as low as 0 °C (Ridell and Korkeala, 1997). However, they often dominate the flora at temperatures of 8–15 °C (Huis in’t Veld, 1996; Cousin, 1982). The coliform/enteric group is less sensitive to changes in pH compared with the Gram-negative (oxidase positive) rod-shaped bacteria and so are of more significance in mild acid products. They are however, generally sensitive to low aw, preservatives, salt and thermal treatments (Jay, 1978). The coliform/enteric group do not necessarily require the presence of oxygen for growth. In addition, they have a fermentative metabolism and so may break down carbohydrates to give acids, which may result in souring of milk (Cousin, 1982). In contrast, the metabolism of the Gram-negative (oxidase positive) bacteria is oxidative and fermentation does not occur. Other types of spoilage include the production of pigmented growth, gases, slime, off-odours and off-flavours. Off odours have been described as ‘grassy’, medicinal, unclean and faecal (Walker and Stringer, 1990). Typical spoilage species include Citrobacter, Escherichia, Enterobacter, Hafnia, Klebsiella, Proteus and Serratia (Jay, 1978; Walker, 1988). These micro-organisms are widely disseminated in the environment, including in animals. Poor slaughter and dressing practices may contribute to their presence in foods. 3. Gram-positive spore-forming bacteria This group, of particular significance, can produce heat-resistant bodies (spores) which can survive many thermal processes. Such heating may destroy all vegetative cells, leaving the relatively slow growing spore-formers to dominate the microflora. The minimum growth temperatures are often 0–5 °C, although growth is often slow below 8 °C (Huis in’t Veld, 1996; Cousin, 1982; Coghill and Juffs, 1979). The genera of concern in this group are Bacillus and Clostridium species. Again, these are common in the environment and spores may survive for considerable periods. The most common form of spoilage is the production of large © 2008, Woodhead Publishing Limited

Chilled foods microbiology 457 quantities of gas which may result in pack or product blowing (Cousin, 1982; Walker, 1988). The heat resistance of psychrotrophic strains is considered to be lower than that of mesophilic strains (Reinheimer and Bargagna, 1989), but the former group is of concern in chilled pasteurised foods. 4. Lactic acid bacteria At chill temperatures, lactic acid-producing bacteria grow slowly if at all. Consequently, if they are to cause spoilage, growth of most other bacterial species must be inhibited. This group is more tolerant of low pH than other spoilage bacteria and may multiply at pH values as low as 3.6 (Jay, 1978). The lactic acid bacteria are also more resistant than the previously discussed spoilage bacteria to slight reductions in the aw and some Pediococcus species are salt-tolerant. Lactic acid bacteria usually predominate on vacuum-packed products and in some modifiedatmosphere-stored foods, and may even grow in atmospheres containing 100% carbon dioxide (Gill and Molin, 1991). This bacterial group comprises both rodand coccus-shaped Gram-positive bacteria and typical genera include Carnobacterium, Lactobacillus, Leuconostoc, Pediococcus and Streptococcus species (Borch et al., 1996). Spoilage is generally by the production of acid which results in souring with or without concomitant gas production (Walker and Stringer, 1990). Lactic acid-producing bacteria are deliberately added during the manufacture of some chilled foods (e.g. cheese, yoghurts, some salamis) and are essential for the development of the desired product characteristics. In addition, there is much interest in the potential use of lactic acid bacteria as a novel preservation system, as many produce antimicrobial compounds in addition to acids (Lücke and Earnshaw, 1991). 5. Other bacteria Depending on the food type and preservation system operating, other microorganisms may also cause problems in chilled foods. For example, Brochothrix thermosphacta is a Gram-positive rod-shaped bacterium which is occasionally present on raw meats but does not normally create a spoilage problem. Products preserved with sulphite (e.g. fresh British sausage) may encourage the development of this bacterium (Gardner, 1981). Furthermore, it can grow in atmospheres with a low oxygen level and/or high carbon dioxide concentration and so may cause problems in vacuum-packed or modified-atmosphere-packed meat products. In vacuum-packed sliced meats, this micro-organism produces an objectionable pungent ‘cheesy’ odour. Micrococcus species are Gram-positive cocci which can grow in the presence of high salt concentrations. They tend not to grow well at chill temperatures but can cause souring and slime production on cured meats and in curing brines should temperature abuse occur (Gardner, 1983). Other micro-organisms that may cause spoilage problems in cured meats and/or vacuum-packed meat products are Corynebacterium, Kurthia and Arthobacter species (Gardner, 1983; Gould and Russell, 1991). © 2008, Woodhead Publishing Limited

458

Chilled foods

6. Yeasts and moulds Compared with bacteria, both yeasts and moulds grow more slowly in foods permitting good growth and so are generally out-competed. Therefore this group is seldom responsible for the spoilage of fresh proteinaceous foods. If, however, the conditions in the food are altered to limit bacterial growth, the role of yeasts and moulds may become more significant. Many yeasts can grow at temperatures less than 0 °C (Michener and Elliott, 1964). Furthermore, yeasts and moulds are generally more resistant than bacteria to low pH, reduced aw values and the presence of preservatives (Jay, 1978). Moulds tend to require oxygen for growth whereas many yeasts can grow in the presence or absence of oxygen. Most yeasts and moulds are not heat-resistant and are readily destroyed by a thermal process. The mould genus Byssochlamys however, may produce relatively heat-resistant ascospores (Bayne and Michener, 1979). Freshly collected meat, poultry, fish and dairy products rarely contain yeasts or moulds but they rapidly become contaminated from the environment. In particular, air movements may be an important vector of transmission, especially with mould ascospores. Typical spoilage yeasts include Candida, Debaryomyces, Hansenula, Kluveromyces, Rhodotorula, Saccharomyces, Torula and Zygosaccharomyces species (Walker and Stringer, 1990; Pitt and Hocking, 1985). Moulds that may be isolated from spoiled chilled foods include Aspergillus, Cladosporium, Geotrichum, Mucor, Penicillium, Rhizopus and Thamnidium species (Pitt and Hocking, 1985; Filtenborg et al., 1996). Fungal spoilage may be characterised by the production of highly visible, often pigmented, growth, slime, fermentation of sugars to form acid, gas or alcohol, and the development of off-odours and off-flavours. Odours and flavours have been described as yeasty, fruity, musty, rancid and ammoniacal. As with the lactic acid bacteria, yeasts and moulds are sometimes deliberately added to food products. For example, the development of Penicillium camembertii on the surfaces of Brie and Camembert cheeses is essential for the desired flavour, odour and texture characteristics. This mould growing on other types of cheeses would be described as a spoilage defect.

16.7 Pathogenic micro-organisms Foods may be considered to be microbiologically unsafe owing to the presence of micro-organisms which may invade the body (e.g. Salmonella, Listeria monocytogenes, E. coli O157:H7 and Campylobacter) or those which produce a toxin ingested with a food (e.g. Clostridium botulinum, Staphylococcus aureus and Bacillus cereus). The growth of pathogenic micro-organisms in foods may not necessarily result in spoilage, and so the absence of deleterious sensory changes cannot be relied upon as an indicator of microbial safety. Furthermore, some toxins are resistant to heating and so may remain in a food after viable micro-organisms have been removed. It is therefore essential that an effective programme is used to ensure the safety of foods from production, through processing, storage and distribution to consumption. Within the UK, the trends in food poisoning and the © 2008, Woodhead Publishing Limited

Chilled foods microbiology 459 issues contributing to this have been extensively reviewed by Border and Norton (1997). As discussed previously, storage at chill temperatures cannot prevent all microbial growth, but can prevent the growth of some types and retard the rate of growth in others. As far back as 1936, Prescott and Geer recommended that foods permitting growth of micro-organisms should be stored at less than 10 °C (50 °F) and preferably ca. 4 °C (39 °F) to prevent the growth of pathogens or toxin production. That was sound advice in terms of the food-borne pathogens recognised at that time. The risk of growth by food-borne pathogens is a combination of the minimum growth temperatures, the growth rate at chill temperatures and the time and temperature(s) of storage. The minimum growth temperatures of pathogenic bacteria have been discussed by Walker and Stringer (1990). Whilst the majority of food-borne disease is caused by relatively few bacterial types – mainly Salmonella and Campylobacter (Border and Norton, 1997), the number of bacteria recognised as food-borne pathogens, however, has steadily increased. Whilst this may, in part, reflect a true underlying increase in the incidence, it may also be due to a greater awareness of these micro-organisms and improvements in methodologies. For discussion in this chapter, the pathogenic bacteria of concern for chilled foods can be arbitrarily divided as follows.

16.7.1 Micro-organisms capable of growth at temperatures below 5 °C This group is potentially of greatest concern as they continue to multiply even with ‘good’ refrigeration temperatures. Although growth may continue, temperature control is critical and the growth rate becomes increasingly slow as the temperature is reduced (see Fig. 16.1). In addition, temperature control can interact effectively with other factors to prevent or greatly limit growth. Listeria monocytogenes The bacterium now identified as L. monocytogenes was first recognised as a human pathogen in 1926 (Murray et al.,), but its role in food-borne disease was not apparent until the late 1970s. Reported cases in the UK increased dramatically during the 1980s, and decreased during subsequent years. The symptoms of disease are protean and range from a mild flu-like illness to meningitis, septicaemia, stillbirths and abortions (Ralovich, 1987). In general, the major symptoms of disease are restricted to the pregnant mother, foetus, elderly and immunocompromised. With the last three groups, the mortality level can be high (McLauchlin, 1987). The epidemiology of L. monocytogenes has been discussed by Schuchat et al. (1991). A very wide range of foods including meat, poultry, dairy products, seafoods and vegetables have been reported to be contaminated with L. monocytogenes and have been reviewed by Bell and Kyriakides (1998b). Whilst the total absence of L. monocytogenes from raw meats, poultry and vegetables is difficult to ensure, the bacterium has been isolated from products which have undergone a listericidal thermal process (Lund, 1990). Such isolations are of concern as many of these © 2008, Woodhead Publishing Limited

460

Chilled foods

chilled foods may be consumed without further heating. The presence of L. monocytogenes on cooked foods suggests that post-process contamination may have occurred. Several studies have shown this bacterium has been isolated from a wide range of sites in several types of factory (Cox et al., 1989) and may be spread by some cleaning procedures (Holah et al., 1993). Sites of particular concern include those where water is present. Environmental control of Listeria, particularly in key areas of production (e.g. after cooking) is crucial to the prevention of product contamination. The number of cases of reported listeriosis in England and Wales peaked dramatically between 1986 and 1988 which was associated with contaminated imported pâté. Following public warnings about this, the number of cases declined to the annual rate prior to this (100–150 cases per year) (Border and Norton, 1997). The major concern with L. monocytogenes is its ability to grow at low temperatures, and a minimum growth temperature of -0.4 °C has been reported (Walker et al., 1990a). Temperature control will however, retard the rate of growth (Fig. 16.2). Conversely, temperature abuse during storage of a food can exacerbate problems. Listeria monocytogenes is more resistant than many other vegetative bacteria to some, but not all, of the preservation mechanisms used in food manufacture (e.g. chilling, reduced water activity) and these have been reviewed by Walker (1990). Whilst resistance may be noted to these preservation systems when examined individually, foods are complex and interactions may occur which effectively prevent growth. The use of predictive models (see Section 16.9) for microbiology is an efficient method to identify such interactions. L. monocytogenes is not considered to be a classically heat-resistant bacterium. It is generally accepted that conventional HTST milk pasteurisation (71.7 °C/15 seconds) will eliminate this micro-organism when freely suspended in milk (Bradshaw et al., 1991). In other foods, decimal reduction times of 8–16 seconds have been reported at 70 °C (Gaze et al., 1989). It has been recommended (Anon., 1989) that foods subject to a cook-chill process be heated to a minimum of 70 °C for 2 minutes (or the thermal equivalent) to ensure the effective elimination of this bacterium. Overall, the control of L. monocytogenes in foods and food environments (Holah, 1999) is of concern to food processors. Yersinia enterocolitica Like L. monocytogenes, Y. enterocolitica was first described over 50 years ago (Schliefstein and Coleman, 1939) but largely ignored as an agent of food-borne disease until the 1970s. Outbreaks of disease have implicated chilled foods such as pasteurised milk (Tacket et al., 1984), tofu (Tacket et al., 1985) and chocolate milk (Black et al., 1978). Whilst the reported incidence of Y. enterocolitica in gastrointestinal samples is generally low, it has increased, but as before, this may be due not only to a true underlying increase but also to a greater awareness of this bacterium, recognition of symptoms and improved methodologies. In some countries (e.g. Belgium and the Netherlands) disease by Y. enterocolitica has surpassed that of Shigella and even rivals that of Salmonella (Doyle, 1990). The symptoms of human yersiniosis are protean (Schiemann, 1989). Overall, © 2008, Woodhead Publishing Limited

Chilled foods microbiology 461

Y. enterocolitica L. monocytogenes

Fig. 16.2

Effect of temperature on the generation time of L. monocytogenes and Y. enterocolitica.

acute gastroenteritis is the most common symptom, particularly with children, and is characterised by diarrhoea, abdominal pain, fever and less commonly, vomiting. With adolescents, abdominal pain may be localised in the right iliac fossa area of the body and misdiagnosed as appendicitis. In the outbreak involving chocolate milk, 17/257 (6.6%) of the cases had their appendix removed (Black et al., 1978). The mortality rate from human yersiniosis is low and, other than cases involving appendectomies, the symptoms are generally self-limiting and rarely require treatment (Schiemann, 1989). With adults, secondary symptoms may occur several weeks after the typical gastrointestinal symptoms disappear. Most commonly these are post-infectious polyarthritis and erythema nodosum (Schiemann, 1989). A wide variety of foods have been reported to be contaminated with Y. enterocolitica including many chilled products, e.g. raw and cooked meats, poultry, seafoods, milk, dairy products and vegetables (Greenwood and Hooper, 1989). Care is needed as the isolates responsible for disease generally belong to a few specific bio-serotypes, whilst those from foods and the environment belong to a wide range of bio-serotypes (Gilmour and Walker, 1988; Logue et al., 1996). Therefore the pathogenic significance of food isolates should be ascertained before the food is condemned as a health risk. The bio-serotypes responsible for human disease are frequently isolated from pigs and occasionally pork products (Schiemann, 1989). The minimum reported growth temperature for Y. enterocolitica is –1.3 °C and the bacterium grows relatively well at chill temperatures (Walker et al., 1990b). As © 2008, Woodhead Publishing Limited

462

Chilled foods

with L. monocytogenes, reducing the storage temperature has an increasingly dramatic effect on the growth of Y. enterocolitica (Fig. 16.2). Furthermore, storage at refrigeration temperatures will interact with other preservation factors present in foods to prevent growth of this bacterium. Factors affecting the growth of Y. enterocolitica have been discussed by Walker and Stringer (1990). Y. enterocolitica is a heat-sensitive bacterium and will be readily eliminated from foods by heating (Lovett et al., 1982). It has, however, been reported from cooked meats, seafoods and pasteurised dairy products, which indicates that post-process contamination had occurred. Greater attention is required for the environmental control of Y. enterocolitica in food manufacturing establishments. As yet, relatively little is published on this aspect. Aeromonas hydrophila The role of A. hydrophila as an agent of food-borne disease is still a matter of controversy as no fully documented outbreaks have been reported. This bacterium however, does possess many of the characteristics of other pathogenic bacteria (Cahill, 1990). As with Y. enterocolitica, the number of reported cases of A. hydrophila gastroenteritis in England and Wales has risen during the 1980s (Anon., 1991a). The reasons for this are as described previously. Incidents of foodborne disease implicating A. hydrophila have included oysters and prawns (Todd et al., 1989) – both chilled foods. Within the genus Aeromonas, some of the other major motile species (i.e. A. hydrophila, A. sobria and A. caviae) may be considered to be pathogenic (Stelma, 1989). All three of these species have been isolated from a variety of chilled foods (Abeyta and Wekell, 1988; Fricker and Tompsett, 1989). The minimum reported growth temperature for A. hydrophila is –0.1 to 1.2 °C (four strains tested) and so growth will occur at chill temperatures (Walker and Stringer, 1987). As with the previous psychrotrophic pathogens, temperature control is important and temperature abuse will greatly increase the rate of growth. Relatively little is published about the heat resistance of A. hydrophila but the bacterium is considered to be heat-sensitive and so may be readily eliminated from foods (Palumbo and Buchanan, 1988). The effects of other factors (e.g. pH, salt, preservatives, etc.) on the growth of A. hydrophila have been reviewed by Palumbo and Buchanan (1988). Little has been published on the presence of A. hydrophila in the processing environment, but it is likely that it will be isolated particularly from wet areas. Bacillus cereus The role of B. cereus as a spoilage bacterium of chilled foods is well recognised (Griffiths and Phillips, 1990). Many such strains may grow at temperatures as low as 1 °C (Coghill and Juffs, 1979). This bacterium may also cause food-borne disease but the number of reported cases is generally low (Border and Norton, 1997). The minimum reported growth temperatures of these strains is usually 10– 15 °C (Goepfert et al., 1972; Johnson, 1984) although some isolates from outbreaks which involved vegetable pie, pasteurised milk and cod were able to grow and © 2008, Woodhead Publishing Limited

Chilled foods microbiology 463 produce toxins at 4 °C (van Netten et al., 1990; Jaquette and Beuchat, 1998). In addition, psychrotrophic, presumptively enterotoxigenic strains were frequently isolated from pasteurised milks and some cook–chill meats (van Netten et al., 1990). If temperature abuse of the product occurred, the time until the toxin was detected was reduced by 50% when the temperature was raised from 4 to 7 °C. Bacillus cereus may be of particular significance in foods which have been heated or pasteurised, as the heat treatment may have eliminated other competitor micro-organisms. During subsequent chilled storage, spores which may survive the heat treatment may germinate and grow. Although little published information is available, the heat resistance of psychrotrophic B. cereus (and other related species) is generally lower than that of the mesophilic strains (Reinheimer and Bargagna, 1989). Other Bacillus species (i.e. B. subtilis and B. licheniformis) may also cause human disease (Kramer and Gilbert, 1989). Although psychrotrophic strains of these have been isolated from milk, their association with human disease is at present unclear. Clostridium botulinum Human botulism is caused by the ingestion of a neurotoxin, and, based on the antigenic analysis of this, seven types can be distinguished (named A–G) (Hauschild, 1989). Traditionally, food-borne disease was caused by types A and B. It is now well recognised that types E and F may also cause disease following the ingestion of preformed toxin. The strains responsible for disease can be divided into two main groups. Firstly, types A and some strains of B and F are proteolytic and so often cause putrefaction of foods if substantial growth occurs (Hauschild, 1989). Secondly, types E and others of B and F are non-proteolytic and so the consequences of growth in foods will be less pronounced (Hauschild, 1989). The minimum growth temperature of the mesophilic proteolytic strains is considered to be 10 °C and so these are of limited significance with chilled foods. In 1961, Schmidt et al. reported that type E Cl. botulinum was able to grow and produce toxin in a beef stew after incubation at 3.3 °C for 32 days. It is now recognised that non-proteolytic strains of types B and F are also capable of growth and toxin production at 5 °C or less (Ecklund et al., 1967; Simunovic et al., 1985). Therefore, these non-proteolytic strains may grow in chilled foods. The growth of non-proteolytic Cl. botulinum is of particular concern in ‘sousvide’ processing. This consists of packing foods under vacuum in air-impermeable sealed bags which are then heat processed and stored chilled for extended periods. Whilst the time and temperature of cooking is specific to the food type, it will destroy vegetative microbial cells but may not be sufficient to destroy bacterial spores. These may subsequently germinate and grow in the absence of air during refrigerated storage (Betts, 1992). It should be noted that the heat resistance of the psychrotrophic non-proteolytic strains is considerably lower than that of the mesophilic proteolytic strains (Table 16.3). The risk of botulism from ‘sous-vide’ products can be minimised by the use of an appropriate time–temperature profile during heating, adequately controlled chilled storage and/or alterations in the product formulation to prevent growth © 2008, Woodhead Publishing Limited

464

Chilled foods

Table 16.3 Comparison of proteolytic and non-proteolytic strains of Clostridium botulinum (Betts, 1992; Hauschild, 1989) Proteolytic Minimum temperature Minimum pH Maximum salt Minimum aw D value at 100 °C for spores

10–12 °C 4.6 10% 0.93 25 min

Cl. botulinum Non-proteolytic 3.3–5.0 °C 5.0 5–6.5% 0.95–0.97 < 0.1 min

(Betts, 1992; Betts, 1996). The minimum pH and aw values for growth also differ between the proteolytic and non-proteolytic strains (Table 16.3). Overall, the nonproteolytic strains are less resistant to low pH and aw values (Hauschild, 1989).

16.7.2 Micro-organisms capable of initiating growth at temperatures of 5–10 °C There are a number of other pathogenic bacteria which, although unable to grow at temperatures below 5 °C, may grow if temperature abuse occurs. These include Salmonella species, Escherichia coli and Staphylococcus aureus, with generally accepted minimum growth temperatures of 5.1, 7.1 and 7.7 °C respectively (Alcock, 1987; Angelotti et al., 1961) At temperatures up to 10 °C, the growth rate of these bacteria is generally slow (Matches and Liston, 1968). These bacteria do, however, cause food-borne disease, frequently implicating chilled foods. Psychrotrophic strains of salmonellae have very occasionally been reported and this may be of more concern with regard to the public health issues of chilled foods (d’Aoust, 1991). Several types of E. coli are well recognised as agents of foodborne disease. At present the type of most concern is E. coli 0157:H7 and other verocytotoxigenic E. coli (VTEC) which may produce severe haemorrhagic colitis (Kaper and O’Brien, 1998). Limited growth of some strains may occur at 5–10 °C (Alcock, 1987; Kauppi et al., 1998). This organism has been reviewed by Bell and Kyriakides (1998a). Whilst Staph. aureus may grow at temperatures as low as 7.7 °C, disease is caused by the ingestion of a preformed toxin. The minimum temperature for toxin production is greater than for growth and has been reported to be 14.3 °C (Alcock, 1987). Overall, the bacterial species above do not grow at temperatures below 5 °C, but may survive at these temperatures. Often, pathogens and spoilage bacteria will survive adverse conditions (e.g. low pH or high salt) better at refrigeration temperatures compared with higher temperatures (Faith et al., 1998). Therefore, if the infectious dose of the bacterium is low and/or growth of the pathogen has already occurred (e.g. during slow cooling), growth during chilled storage may not be a prerequisite for disease.

© 2008, Woodhead Publishing Limited

Chilled foods microbiology 465 16.7.3 Micro-organisms capable of initiating growth at temperatures greater than 10 °C These species include mesophilic Cl. botulinum, mesophilic B. cereus and other Bacillus species, Cl. perfringens and Campylobacter species. In general, these will not grow below 10 °C and growth is limited at temperatures between 10 and 15 °C (Walker and Stringer, 1990). Of particular concern in this bacterial group are the Campylobacter species which comprise the most commonly reported cause of gastrointestinal disease in the UK (Border and Norton, 1997). Although many of the reported cases are sporadic, outbreaks have frequently implicated the consumption of raw milk and undercooked chicken (Skirrow, 1990). This bacterial group is unusual as the minimum temperature for growth is 25–30 °C and so it will not grow on most foods. The infectious dose of the micro-organism is very low and so growth may not be necessary for disease to occur (Butzler and Oosterom, 1991). Whilst disease caused by the mesophilic spore-forming bacteria has implicated chilled foods, this is usually as a consequence of poor temperature control during cooling after cooking (Gould and Russell, 1991; Shaw, 1998). These bacteria may grow extremely rapidly during a long slow cooling regime after cooking and then persist during chilled storage.

16.8 Temperature control With chilled foods, good temperature control is essential, not only to maintain the microbiological safety and quality of foods, but also to minimise changes in the biochemical and physical properties of the food. The temperatures of storage of chilled foods may vary greatly during manufacture, distribution, retail sale and in the home. Consequently, during the life of a chilled food, considerable opportunities exist for temperature abuse to occur. The greater the abuse of temperature, then the greater the potential for microbial growth to occur. This may result in a product becoming unsafe and/or a loss in product quality. Temperature control is the key issue with regard to chilled foods and an integral part of the preservation system. In many of the stages in the food chain after primary chilling, the refrigeration equipment is designed to maintain the product temperature. It may not be able rapidly to reduce the temperature of foods that have been abused at higher temperature.

16.9 Predictive microbiology As discussed previously, chill temperatures will not prevent microbial growth completely and additional preservative factors, such as reduced pH and water activity, may be required to extend the time period before significant microbial growth occurs. Traditionally, the effect of combinations of preservation systems on target organisms would have been tested using laboratory studies (often called challenge tests). Whilst challenge tests have an important role, they tend to be © 2008, Woodhead Publishing Limited

466

Chilled foods

expensive, time consuming and results obtained are limited to the specific conditions tested. Should any of these change, the test needs to be repeated. However, the chilled foods market is very dynamic and there is a great demand for continual development of new products (Stringer and Dennis, 2000). These need to be developed and marketed rapidly. Predictive microbiology is a tool which can provide rapid reliable answers concerning the likely growth of specific organisms under defined conditions, including conditions not previously examined. Models can be used to predict the probability of growth, the time until growth occurs or the growth rate of microorganisms. The use of predictive models to describe the microbial kinetics is not new and reference to these techniques can be found in publications dating from the 1920s (Esty and Meyer, 1922). Microbiological modelling has been reviewed by Gould (1989a) and McMeekin et al. (1993). The development of a microbiological model generally uses the following stages:

• careful selection and appropriate preparation of the target micro-organism • inoculation of the target micro-organism into a growth medium (microbiological • • • • •

media or food) with defined characteristics storage of the medium under controlled conditions sampling of the medium for the target micro-organism at relevant intervals construction of a model to describe the target micro-organism’s response validation of the model’s predictions – preferably in food to ensure the predictions are meaningful refinement or further enhancement of the model.

The types of models which have been used vary greatly and include the Arrhenius equation, non-linear Arrhenius (Schoolfield) models, Bélehrádek-type (Ratkowsky or square root) models, polynomial models, mechanistic models (all reviewed by McMeekin et al., 1993) and a dynamic modelling approach (Baranyi and Roberts, 1994).

16.9.1 Food pathogens Over the past decade, there has been considerable work done on predictive modelling of a wide range of pathogenic bacteria, e.g. kinetic growth models have been published for Salmonella (Gibson et al., 1988), L. monocytogenes (Farber et al., 1996), Cl. botulinum (Graham et al., 1996). In order to make such models accessible to food manufacturers, there is a requirement for them to be packaged as user-friendly software. There are systems available for predicting the growth of food pathogens. In the UK, the Food MicroModel system was the largest and most com-prehensive system and was developed from a Ministry of Agriculture, Fisheries and Food sponsored research programme. The data is now incorporated in Growth Predictor (http:// www.ifr.ac.uk/Safety/GrowthPredictor/). There is an extensive range of pathogen models in the system including those shown in Table 16.4. © 2008, Woodhead Publishing Limited

Chilled foods microbiology 467 Table 16.4 Some pathogen models Growth models

Thermal death models

Aeromonas hydrophila Bacillus cereus Clostridium botulinum Clostridium perfringens Escherichia coli O157:H7 Listeria monocytogenes Salmonella Staphylococcus aureus Yersinia enterocolitica

Cl. botulinum E. coli O157:H7 L. monocytogenes Salmonella Y. enterocolitica

Table 16.5 Some models in the Pathogen Modeling Program (USDA) Growth models

Survival models

A. hydrophila B. cereus E. coli O157:H7 Salmonella spp. Shigella flexneri S. aureus Y. enterocolitica

E. coli O157:H7 L. monocytogenes Salmonella S. aureus

The models in the Food MicroModel system were produced from data obtained in laboratory growth media and validated by comparing predictions from the model with data obtained from the literature or obtained from inoculated food studies. Another comprehensive modelling programme has been produced in the USA by the United States Department of Agriculture (USDA). It is called the Pathogen Modeling Program and was designed by Dr Robert L. Buchanan and Dr Richard Whiting. The models in this system include those shown in Table 16.5. This programme is available free of charge and can be obtained from the internet (http://www.arserrc.gov). The models in this programme have been produced from extensive growth data in laboratory media, but have not been validated in foods.

16.9.2 Food spoilage With regard to the modelling of food spoilage organisms, there are few systems available although many individual models have been published. Work in Tasmania has developed Pseudomonas predictor models applicable to milk and raw meats (McMeekin and Ross, 1996). Campden and Chorleywood Food Research Association (CCFRA) has developed a collection of models which can be used to assess spoilage rates or likely stability of foods, including chilled foods. This © 2008, Woodhead Publishing Limited

468

Chilled foods

Table 16.6 Current options for CCFRA Forecast models Model Bacillus spp. Pseudomonasspp. Enterobacteriaceae Yeasts (chilled) Lactic acid bacteria

pH

Salt (% w/v)

Temperature (°C)

4.0–7.0 5.5–7.0 4.0–7.0 2.5–6.3 2.9–5.8

0.5–10.0 0.0–4.0 0.5–10.0 0.5–10.0 0.5–10.0

5–25 0–15 0–30 1–22 2–30

collection of models is called Forecast and is available to potential users via an enquiry service (+44 (0) 1386 842000) which runs the model on behalf of clients after a detailed consultation with respect to their needs. The consultancy aspect of this approach also allows subsequent expert interpretation and consideration of model validation status. Table 16.6 shows the range of models currently available within Forecast. All models within the Forecast system have been produced from data obtained in laboratory media and have been validated in relevant foods using literature data or inoculated challenge test studies. Limited models on spoilage organisms are available in the Food MicroModel programme previously mentioned and these include: Brochothrix thermosphacta, Saccharomyces cerevisiae, Lactobacillus plantarum, Zygosaccharomyces bailii. In addition to bacteria and yeasts, models have also been developed for mould growth (Valík et al., 1999). Furthermore, Membré and Kubaczka (1998) have applied similar models to product degradation (i.e. pectin breakdown) rather than just microbial growth.

16.9.3 Practical application of models Figure 16.3 shows how a model has been put to practical use by comparing predicted values of numbers against predetermined standards for termination of shelf-life. Many other potential applications exist, for example:

• What level of micro-organisms will be present under different temperatures of storage?

• How much salt is needed to restrict microbial numbers to a pre-set level after one week storage at 8 °C?

• What will be the effect of increasing the product pH from 5.0 to 5.4? Several authors have reported deficiencies or inaccuracies in model predictions (Dalgaard and Jørgensen, 1998; Hygtiä et al., 1999) in that they predict faster growth than that observed in foods. However, many of the models, particularly those for pathogens, are designed to be ‘fail safe’ and foods may contain additional antimicrobial factors not present in the model, which may inhibit or prevent the predicted growth. Consequently, it is important to determine that any models used contain the important preservation factors relevant to the study and that the model has been validated in appropriate foods. Most of the models developed have been based on single organisms or groups of organisms in pure cultures and may not therefore © 2008, Woodhead Publishing Limited

Chilled foods microbiology 469

Time (hr)

Fig. 16.3 Graphical representation of predictions made using CCFRA forecast conditions: pH 6.0, salt 3% w/v, temperature of storage 6 °C. The user’s tolerance for enterobactericeae, Pseudomonas spp. and Bacillus spp. are clearly shown in relation to the predicted shelf-life.

take into account any effects of microbial interaction and competition likely to be seen in foods. Pin and Baranyi (1998) have used modelling techniques to examine the interactions between spoilage bacteria. In the wrong hands, the information from predictive models may be misused and may have serious consequences. It is important that the right questions be asked in order to obtain useful information. There are many advantages to the use of predictive models in the development and manufacture of chilled foods. They can help to focus resources during product development to assess the microbiological safety and stability of hundreds of different ingredient combinations before stepping into the development kitchen. Predictive models can be used as decision-making tools to allow productive focusing of effort in process and product development and risk and hazard assessment. They can be of great value in complex HACCP studies if used correctly. They should be followed up with targeted practical trials and challenge tests. Used in this way, predictive models can be powerful tools for industrial food microbiologists. Recently, several workers have proposed the development of predictive models with computational neural networks (Hajmeer et al., 1997) and their incorporation in decision support systems for microbiological quality and safety (Wijtzes et al., 1998). Predictive models also have a role to play in education and training, in that they allow demonstration of microbial behaviour and risk without the need for expensive laboratory exercises. © 2008, Woodhead Publishing Limited

470

Chilled foods

It should be stressed that microbiological models will never completely remove the requirements for microbiological expertise or to conduct microbiological challenge tests and shelf-life studies, but can be very useful for an indication of the safety and stability of chilled products and ingredients.

16.10 Conclusions Chilled foods comprise a diverse and complex group of commodities which contain a large number of ingredients. The composition and number of microorganisms present is affected by the indigenous microflora, micro-organisms contaminating before and after processing, the growth rates and abilities of the micro-organisms, the spoilage abilities of the micro-organisms, the intrinsic properties of the food, the effects of processing and packaging, and the time and temperatures of storage. Consequently, the microbial safety and spoilage of chilled foods is very complex, but certain general principles may be applied. (i)

The microbiological status of all raw materials should be known and only materials of good quality used. (ii) All stages of processing should be defined, monitored and controlled to ensure their correct operation. This is of particular significance in foods which rely on a combination of factors to ensure microbial stability. (iii) The temperatures and times of chill storage should be controlled during all stages, from raw materials through retail sale and preferably to the home. The lower the temperature throughout the process, the slower the rate of growth. (iv) Attention must be given to the hygiene of the entire process to ensure that microbial contamination is minimised. These objectives may be best achieved through the application of a quality system including Hazard Analysis Critical Control Points (HACCP) (Leaper, 1997) which may be powerfully integrated with other systems, including risk analysis (Jouve et al., 1998). The use of appropriate and validated models may greatly help in the decision-making processes of HACCP and risk analysis. Finally, greater education of all involved in food manufacture, distribution and retail sale and better education of the consumer in areas of hygiene and temperature control will be of great benefit.

16.11 References ABEYTA C AND WEKELL M M

(1988) Potential sources of Aeromonas hydrophila. J. Food

Safety 9 11–22. ADVISORY COMMITTEE ON THE MICROBIOLOGICAL SAFETY OF FOOD (ACMSF) (1995) Report

on Verocytotoxin-producing Escherichia coli. HMSO, London. (1984) Growth characteristics of food-poisoning organisms at suboptimal temperatures. II Salmonellae, Campden Food Preservation Research Association Technical Memorandum No 364.

ALCOCK S J

© 2008, Woodhead Publishing Limited

Chilled foods microbiology 471 (1987) Growth characteristics of food-poisoning organisms at suboptimal temperatures, Campden Food Preservation Research Association Technical Memorandum No. 440. ANGELOTTI R, FOTER M J AND LEWIS K H (1961) Time–temperature effects of salmonellae and staphylococci in foods, Am. J. Pub. Health, 36 559–63. ANON (1989) Guidelines for Cook-Chill and Cook-Freeze Catering Systems, HMSO, London. ANON (1991a) The Microbiological Safety of Food, Part II HMSO, London. ANON (1991b) Principles and Practices for the Safe Processing of Foods, ButterworthHeinemann, Oxford. BARANYI J AND ROBERTS T A (1994) A dynamic approach to predicting bacterial growth in food, International Journal of Food Microbiology, 23 277–94. BAYNE H G AND MICHENER H D (1979) Heat resistance of Byssochlamys ascospores, Appl. Environ. Microbiol., 37 449–53. BELL C AND KYRIAKIDES A (1998a) E. coli: A practical approach to the organism and its control in foods, Blackie Academic and Professional, London. BELL C AND KYRIAKIDES A (1998b) Listeria: A practical approach to the organism and its control in foods, Blackie Academic and Professional, London. BETTS G D (1992) The microbiological safety of sous-vide processing, Campden and Chorleywood Food Research Association Technical Manual No. 39. BETTS G D (1996) A code of practice for the manufacture of vacuum and modified atmosphere packaging chilled foods, Campden and Chorleywood Food Research Association, CCFRA Guideline No. 11. BLACK R E, JACKSON R L, TSAI T, MEDVESKY M, SHAYEGANI M, FEELEY J C, MACLEOD K I E, AND WAKELEE A W (1978) Epidemic Yersinia enterocolitica infection due to contaminated chocolate milk, New Engl. J. Med., 298 7679. BORCH E, KANT-MUERMANS M-L, AND BLIXT, Y (1996) Bacterial spoilage of meat and cured meat products. International Journal of Food Microbiology, 33 103–20. BORDER P AND NORTON M (1997) Safer eating: microbiological food poisoning and its prevention. The Parliamentary Office of Science and Technology, London. BRADSHAW J G, PEELER J T AND TWEDT R M (1991) Thermal resistance of Listeria spp. in milk, J. Food Prot., 54 12–4. BUTZLER J P AND OOSTEROM J (1991) Campylobacter: pathogenicity and significance in foods, Int. J. Food Microbiol., 12 1–8. CAHILL M M (1990) Virulence factors in motile Aeromonas species: a review, J. Appl. Bacteriol., 69 1–16. COGHILL D AND JUFFS H S (1979) Incidence of psychrotrophic spore-forming bacteria in pasteurised milk and cream products and effect of temperature on their growth, Australian J. Dairy Technol., 3 150–3. CONNER D E AND KOTROLA J S (1995) Growth and survival of Escherichia coli O157:H7 under acidic conditions, Applied and Environmental Microbiology, 61 382–5. COUSIN M A (1982) Presence and activity of psychrotrophic microorganisms in milk and dairy products: a review, J. Food Prot., 45 172–207. COX L J, KLEISS T, CORDIER J L, CORDELLANA C, KONKEL P, PEDRAZZINI C, BEUMER R AND SIEBENGA A (1989) Listeria spp. in food processing, non-food processing and domestic environments, Food Microbiol, 6 49–61. DAINTY R H (1996) Chemical/biochemical detection of spoilage, International Journal of Food Microbiology, 33 19–34. DALGAARD P AND J ØRGENSEN L V (1998) Predicted and observed growth of Listeria monocytogenes in seafood challenge tests and naturally contaminated cold smoked salmon, International Journal of Food Microbiology, 40 105–15. DAY B P F (2000) Chilled food packaging. In: Stringer, M. F. and Dennis, C. (ed.), Chilled Foods: a comprehensive guide, 2nd edn. Woodhead Publishing Ltd., Cambridge, pp. 137– 50. ALCOCK S J

© 2008, Woodhead Publishing Limited

472

Chilled foods

D’AOUST J Y

(1991) Psychrotrophy and foodborne Salmonella, Int. J. Food Microbiol., 13 207–16. DENG Y, RYU J H AND BEUCHAT L R (1999) Tolerance of acid adopted and non-adopted Escherichia coli O157:H7 cells to reduced pH as affected by type of acidulant. Journal of Applied Microbiology, 86 203–10. DOYLE M P (1990) Pathogenic Escherichia coli, Yersinia enterocolitica and Vibrio parahaemolyticus, The Lancet, 336 1111–15. ECKLUND M W, WIELER D L AND POYSKY F T (1967) Outbreak and toxin production of nonproteolytic type B Clostridium botulinum at 3.3 to 5.6 °C, J. Bacteriol, 93 1461–2. EDDY B P (1960) The use and meaning of the term ‘psychrophilic’, J. Appl. Bacteriol., 23 189–90. ESTY J R AND MEYER K F (1922) The heat resistance of spores of Cl. botulinum and allied anaerobes, J. Infect. Dis., 31 650–63. FAITH N G, WIERZBA R K, IHNOT A M, ROERING A M, LORANG T D, KASPER, C W AND LUCHANSKY J B (1998) Survival of Escherichia coli O157 in full and reduced fat pepperoni after manufacture of sticks, storage of slices at 4 ° or 21 °C under air vacuum and baking of slices on frozen pizza at 135, 191 and 246 °C, Journal of Food Protection, 61 383–9. FARBER J M, CAI Y AND ROSS W H (1996) Predictive modelling of the growth of Listeria monocytogenes in CO2 environments, International Journal of Food Microbiology, 32 133–44. FILTENBORG O, FRISVAD J C AND THRANE U (1996) Moulds in food spoilage. International Journal of Food Microbiology, 33 85–102. FRICKER C R AND TOMPSETT S (1989) Aeromonas spp. in foods: a significant cause of food poisoning, Int. J. Food Microbiol., 9 17–23. GARDNER G A (1981) Brochothrix thermosphacta (Microbacterium thermosphactum) in the spoilage of meats: a review. In: Roberts, T. A. et al. (eds) Psychrotrophic Microorganisms in Spoilage and Pathogenicity, Academic Press, London, pp. 139–73. GARDNER G A (1983) Microbial spoilage of cured meats. In: Roberts, T. A. and Skinner, F. A. (eds) Food Microbiology: Advances and Prospects, Academic Press, London pp. 179– 202. GAZE J E (1992) Food pasteurisation treatments, Campden Food and Drink Research Association Technical Manual No. 27. GAZE J E, BROWN G D, GASKELL D E AND BANKS J G (1989) Heat resistance of Listeria monocytogenes in homogenates of chicken, beef steak and carrot, Food Microbiol., 6 251– 59. GEORGE S M, LUND B M AND BROCKLEHURST T F (1988) The effect of pH and temperature on initiation of growth of Listeria monocytogenes, Letters in Appl. Microbiol., 6 153–6. GIBSON A M, BRATCHELL N AND ROBERTS T A (1988) Predicting microbial growth: growth responses of salmonellae in a laboratory medium as affected by pH, sodium chloride and storage temperature, Int. J. Food Microbiol., 6 155–78. GILL C D AND MOLIN G (1991) Modified atmospheres and vacuum packaging. In: Russell, N. J. and Gould, G. W. (eds), Food Preservatives, Blackie and Son Ltd., Glasgow, pp. 172–99. GILL C O (1983) Meat spoilage and evaluation of the potential storage life of fresh meat, J. Food Prot., 46 444–52. GILMOUR A AND WALKER S J (1988) Isolation and identification of Yersinia enterocolitica and Yersinia enterocolitica-like bacteria, J. Appl. Bacteriol. Suppl., 65 213S–236S. GLASS K A AND DOYLE M P (1991) Relationship between water activity of fresh pasta and toxin production by proteolytic Clostridium botulinum., J. Food Prot., 54 162–5. GOEPFERT J M, SPIRA W M AND KIM H U (1972) Bacillus cereus food poisoning: a review. J. Milk Food Technol. 35 213–27. GOULD G (1989a) Predictive modelling of microbial growth and survival in foods, Food Sci. Technol. Today, 3 89–92. GOULD G W (1989b) Heat-induced injury and inactivation. In: Gould, G. W. (ed.), Mechanisms of Action of Food Preservation Procedures, Elsevier Appl. Sci. London, pp. 11–42. © 2008, Woodhead Publishing Limited

Chilled foods microbiology 473 (1996) Methods for preservation and extension of shelf-life, International Journal of Food Microbiology, 33 51–64. GOULD G W AND JONES M V (1989) Combination and synergistic effects. In: Gould, G W (ed.), Mechanisms of Action of Food Preservation Procedures, Elsevier Appl. Sci., London, pp. 400–21. GOULD G W AND RUSSELL N J (1991) Major food-poisoning and food-spoilage microorganisms. In: Russell, N. J. and Gould, G. W. (eds), Food Preservatives, Blackie and Son Ltd., Glasgow, pp. 1–21. GRAHAM A F, MASON D R AND PECK M W (1996) Predictive model of the effect of temperature, pH and sodium chloride on growth from spores of non-proteolytic Clostridium botulinum, International Journal of Food Microbiology, 31 69–85. GRAM L AND HUSS H H (1996) Microbiological spoilage of fish and fish products, International Journal of Food Microbiology, 53 121–38. GREENWOOD M H AND HOOPER W L (1989) Improved methods for the isolation of Yersinia species from milk and foods, Food Microbiol., 6 99–104. GRIFFITHS M W AND PHILLIPS J D (1990) Incidence, source and some properties of psychrotrophic Bacillus spp. found in raw and pasteurized milk, J. Soc. Dairy Technol, 43 62–6. GUTIÉRREZ R, GARÉIA T, GONZALEZ I, SANZ B, HERNÁNDEZ P E AND MARTIN R (1997) A quantitative PCR-ELISA for the rapid enumeration of bacteria in refrigerated raw milk, Journal of Applied Microbiology, 83 518–23. HAJMEER M N, BASHEER I A AND NAJJAR Y M (1997) Computational neural networks for predictive microbiology II. Application to microbial growth, International Journal of Food Microbiology, 34 51–66. HAUSCHILD A H W (1989) Clostridium botulinum. In: Doyle, M. P. (ed.), Foodborne Bacterial Pathogens, Marcel Dekker, New York, pp. 111–89. HERBERT R A (1989) Microbial growth at low temperatures. In: Gould, G. W. (ed.), Mechanisms of Action of Food Preservation Procedures, Elsevier Appl. Sci., London, pp. 71–96. HOLAH J T (1999) Effective microbiological sampling of food processing environments, Campden and Chorleywood Food Research Association, Guideline No. 20. HOLAH J T, TAYLOR J AND HOLDER J S (1993) The spread of Listeria by cleaning systems, Campden Food and Drink Research Association Technical Memorandum No. 673. HUIS IN’T VELD J H T (1996) Microbial and biochemical spoilage of foods: an overview, International Journal of Food Microbiology 33, 1–18. HYGTIÄ E, HIELM S, MOKKILA M, KINNUNEN A AND KORKEALA H (1999) Predicted and observed growth and toxigenesis by Clostridium botulinum type E in vacuum-packaged fishery product challenge tests, International Journal of Food Microbiology, 47 161–9. JAQUETTE C B AND BEUCHAT L R (1998) Combined effects of pH, nisin and temperature on growth and survival of psychrotrophic Bacillus cereus, Journal of Food Protection, 61 563–70. JAY J M (1978) Modern Food Microbiology, 2nd edn. D. van Nostrand Co., New York. JOHNSON K M (1984) Bacillus cereus foodborne illness – an update, J. Food Prot., 47 145– 53. JOUVE J L, STRINGER M F AND BAIRD-PARKER A C (1998) Food Safety Management Tools, ILSI – Europe, Brussels. KABARA J J AND EKLUND T (1991) Organic acids and their esters. In: Russell N J and Gould G W (eds), Food Preservatives, Blackie and Son Ltd., Glasgow, pp. 44–71. KAPER B AND O’BRIEN A D (1998) Escherichia coli O157:H7 and other Shiga toxinproducing E. coli strains, ASM Press, Washington. KAUPPI K L, O’SULLIVAN D J AND TATINI S R (1998) Influence of nitrogen source on low temperature growth of verocytotoxigenic Escherichia coli, Food Microbiology, 15 355–64. KRAMER J M AND GILBERT R J (1989) Bacillus cereus and other Bacillus species. In: Doyle, M P (ed.), Foodborne Bacterial Pathogens, Marcel Dekker, New York, pp. 21–69. GOULD G W

© 2008, Woodhead Publishing Limited

474

Chilled foods

(1997) HACCP: a practical guide (2nd edn), Campden and Chorleywood Food Research Association Technical Manual, No. 38. LOGUE C M, SHERIDAN G, WAUTERS G, MCDOWELL D A AND BLAIR I S (1996) Yersinia spp and numbers, with particular reference to Y. enterocolitica occurring on Irish meat and meat products, and the influence of alkali treatment on their isolation, International Journal of Food Microbiology, 33 257–74. LOVETT L, BRADSHAW J G AND PEELER J T (1982) Thermal inactivation of Yersinia enterocolitica in milk Appl. Environ. Microbiol., 44 517–19. LÜCKE F K AND EARNSHAW R G (1991) Starter cultures. In: Russell, N J and Gould, G W (eds), Food Preservatives, Blackie and Son Ltd., Glasgow, pp. 215–34. LUND B M (1990) The prevention of foodborne listeriosis, Br. Food J., 92 13–22. MCLAUCHLIN J A (1987) A review: Listeria monocytogenes, recent advances in the taxonomy and epidemiology of listeriosis in humans, J. Appl. Bacteriol., 63, 1–2. MCMEEKIN T A AND ROSS T (1996) Shelf-life prediction: status and future possibilities, International Journal of Food Microbiology, 31 65–84. MCMEEKIN T A, OLLEY J N, ROSS T AND RATKOWSKY D A (1993) Predictive microbiology: theory and application, Research Studies Press, Somerset. MATCHES J R AND LISTON J (1968) Low temperature growth of Salmonella, J. Food Sci., 33 641–5. MEMBRÉ J M AND KUBACZKA M (1998) Degradation of pectin compounds during pasteurized vegetable juice spoilage by Chryseomonas luteola: a predictive microbiology approach, International of Food Microbiology, 42 159–66. MICHENER H D AND ELLIOTT R P (1964) Minimum growth temperatures for food poisoning, fecal indicator and psychrophilic microorganisms, Adv. in Food Res., 13 349–96. MITSCHERLICH E AND MARTH E H (1984) Microbial Survival in the Environment, SpringerVerlag, Berlin. MORITA R Y (1973) Psychrophilic bacteria, Bacteriol. Rev., 39 144–67. MURRAY E G D, WEBB R A AND SWAN M B R (1926) A disease of rabbits characterised by a large mononuclear leucocytosis caused by a hitherto undescribed bacillus Bacterium monocytogenes (n. sp), J. Pathol. Bacteriol, 29 407–39. NEILL S D (1974) A study of the microflora of raw milk stored at low temperature, PhD Thesis, The Queen’s University of Belfast, Northern Ireland. PALUMBO S A AND BUCHANAN R L (1988) Factors affecting growth or survival of Aeromonas hydrophila in foods, J. Food Safety, 9 37–51. PIN C AND BARANYI J (1998) Predictive models as means to quantify the interactions of spoilage organisms, International Journal of Food Microbiology, 41 59–72. PITT J I AND HOCKING A D (1985) Spoilage of fresh and perishable foods. In: Fungi and Food Spoilage, Academic Press, Sydney, pp. 365–82. PRESCOTT S C AND GEER L P (1936) Observations on food poisoning organisms under refrigeration conditions, Refrigeration Engineering, 32 211–2, 282–3. RALOVICH B S (1987) Epidemiology and significance of listeriosis in the European countries. In: Schonberg, A. (ed.), Listeriosis: Joint WHO/ROI Consultation on Prevention and Control, Vet. Med. Hefte, Berlin, pp. 51–5. REINHEIMER J A AND BARGAGNA M L (1989). Response of psychrotrophic strains of Bacillus to different heat treatments, Microbiol-Aliments-Nutr., 5 117–22. RIDELL J AND KORKEALA H (1997) Minimum growth temperatures of Hafnia alvei and other Enterobacteriaceae isolated from refrigerated meat determined with a temperature gradient incubator. International Journal of Food Microbiology 35 287–92. RUSSELL N J AND GOULD G W (1991) Food preservatives, Blackie, Glasgow. SCHIEMANN D A (1989) Yersinia enterocolitica and Yersinia pseudotuberculosis. In: Doyle, M. P. (ed.), Foodborne Bacterial Pathogens, Marcel Dekker, New York, pp. 601–72. SCHLIEFSTEIN J I AND COLEMAN M B (1939) An unidentified microorganism resembling B. lignieri and Past. pseudotuberculosis, and pathogenic for man, New York State J. Med., 39 1749–53. LEAPER S

© 2008, Woodhead Publishing Limited

Chilled foods microbiology 475 SCHMIDT C F, LECHOWICH R V AND FOLINAZZO J F (1961)

Growth and toxin production by Type E Clostridium botulinum below 40 °F, J. Food Sci., 26 626–34. SCHUCHAT A, SWAMINATHAN B AND BROOME C V (1991) Epidemiology of human listeriosis, Clin. Microbiol. Rev., 4 169–83. SHAW R (1998) Identification and prevention of hazards associated with slow cooling of hams and other large cooked meats and meat products, Campden and Chorleywood Food Research Association Review No. 8. SIMUNOVIC J, OBLINGER J L AND ADAMS J P (1985) Potential for growth of non-proteolytic types of Clostridium botulinum in pasteurized and restructured meat products: a review, J. Food Prot., 48 265–76. SKIRROW M B (1990) Campylobacter, The Lancet, 336 921–3. SPERBER W H (1983) Influence of water activity on foodborne bacteria – a review, J. Food Prot., 46 142–50. STELMA G N (1989) Aeromonas hydrophila, In: Doyle, M. P. (ed.) Foodborne Bacterial Pathogens, Marcel Dekker, New York, pp. 1–19. STRINGER M F AND DENNIS C (2000) The market for chilled foods, In Chilled Foods: a comprehensive guide, 2nd edn. Woodhead Publishing, Cambridge. TACKET C O, NAVAIN J P, SATTIN R, LOFGREN J R, KONIGSBERG C, RENDTORFF R C, RAUSA A, DAVIS B R AND COHEN M L (1984) A multistate outbreak of infections caused by Yersinia enterocolitica transmitted by pasteurised milk, J. American Med. Assoc., 51 483–6. TACKET C O, BALLARD L, HARRIS N, ALLARD L, NOLAN C, QUAN T AND COHEN M L (1985) An outbreak of Yersinia enterocolitica infections caused by contaminated tofu, American J. Epidemiol., 121 705–11. TERPLAN G, SCHOEN R, SPRINGMEYER W, DEGLE I AND BECKER H (1987) Investigations on incidence, origin and behaviour of Listeria in cheese. In: Schönberg, A. (ed.), Listeriosis – Joint WHO/ROI Consultation on Prevention and Control, Vet. Med. Hefte, Berlin, pp. 98–105. TODD L S, HARDY J C, STRINGER M F AND BARTHOLOMEW B A (1989) Toxin production by strains of Aeromonas hydrophila grown in laboratory media and prawn purée, Int. J. Food Microbiol., 9 145–56. VALÍK L, BARANYI J AND GÖRNER F (1999) Predicting fungal growth: the effect of water activity on Penecillium roquefortii, International Journal of Food Microbiology, 47 141– 46. VAN NETTEN R, VAN DE MOOSDIJK A, VAN HOENSEL P AND MOSSEL D A A (1990) Psychrotrophic strains of Bacillus cereus producing enterotoxin. J. Appl. Bacteriol., 69 73–9. VENKITANARAYANEN K S, FAUSTMAN C, CRIVELLO J F, KHAN M I, HOAGLAND T A AND BERRY B W (1997) Rapid estimation of spoilage bacterial load in aerobically stored meat by a quantitative polymerase chain reaction, Journal of Applied Microbiology, 82 359–64. WALKER S J (1988) Major spoilage microorganisms in milk and dairy products, J. Soc. Dairy Technol., 41 91–2. WALKER S J AND STRINGER M F (1987) Growth of Listeria monocytogenes and Aeromonas hydrophila at chill temperatures, Campden Food and Drink Research Association Technical Memorandum No 462. WALKER S J AND STRINGER M F (1990) Microbiology of chilled foods. In: Gormley, T. R (ed.), Chilled Foods – The State of the Art. Elsevier Appl. Sci., Barking, pp. 269–304. WALKER S J (1990) Listeria monocytogenes: an emerging pathogen. In: Turner, A. (ed.), Food Technology International Europe, Sterling Publications, London, pp. 237–40. WALKER S J, ARCHER P AND BANKS J G (1990a) Growth of Listeria monocytogenes at refrigeration temperatures, J. Appl. Bacteriol., 68 157–62. WALKER S J, ARCHER P AND BANKS J G (1990b) Growth of Yersinia enterocolitica at chill temperatures in milk and other media, Milchwissenschaft, 45 503–6. © 2008, Woodhead Publishing Limited

476

Chilled foods

(1998) Microbiology of food taints. International Journal of Food Science and Technology, 33 31–51. WIJTZES T, VAN’T RIET K, HUIS IN’T VELD J H J AND ZWIETERING M H (1998) A decision support system for the prediction of microbial food safety and quality, International Journal of Food Microbiology, 42 79–90. WHITFIELD F B

© 2008, Woodhead Publishing Limited

17 Predicting the behaviour of microorganisms in chilled foods Peter McClure and Alejandro Amézquita, Unilever, UK

17.1 Introduction Modelling and predicting the behaviour of micro-organisms in chilled foods is a key step in product and process design that enables manufacturers to produce microbiologically safe and stable products. Models can also be used to estimate the fate of micro-organisms in products during storage, distribution and use. Typical characteristics associated with chilled foods are freshness and quality, and shelf-lives range from a few days up to many weeks. The means of achieving these different shelf-lives often involve use of minimal inactivation/intervention processes, targeting infectious pathogens, in combination with inhibitory factors that delay or prevent growth of other micro-organisms such as toxin-producing organisms, toxico-infectious agents and spoilage organisms. Being able to predict the behaviour or fate of these various groups of micro-organisms is therefore essential if manufacturers are to market their products without harm to consumers and damage to the brand, through loss of quality. The way in which microbial behaviour or fate is predicted is based on observing the effects of various factors on micro-organisms in systems (either laboratory media or foods), usually under well-controlled conditions, and then fitting these data with mathematical functions (models). Previously, safety and stability of foods were determined with ‘challenge’ tests. Although challenge tests are often still used, there has been some development, in the form of mathematical models, in our ability to predict the growth, survival or death of particular groups of micro-organisms, that is not specific to particular © 2008, Woodhead Publishing Limited

478

Chilled foods

foods or only relevant to actual conditions tested. In many cases, the fate of microorganisms in foods is determined by a small number of factors, such as pH and temperature, and it has, therefore, been relatively straightforward to characterise responses in laboratory media and to develop models that predict their fate in foods, provided that the food environment can be adequately simulated. Mathematical modelling is based on simplifying behaviours and interactions. The first step is to identify the dominant factors of the phenomenon (e.g. lag time, growth rate, or death rate) to be described and then to make observations (e.g. enumeration of bacterial numbers), often disregarding the less important factors. The various predictive models that have been developed for application to chilled foods vary in their complexity and also in the ‘end-point’ that is predicted. When deciding on the modelling approach to be used, manufacturers will typically follow a hazard analysis critical control point (HACCP) based approach and consider microbiological hazards that may come from raw materials, transport, handling or processing. Models are often used at early stages in the design of a food, to identify means through which a product developer can control relevant target micro-organisms and set a shelf-life that delivers the required safety and quality characteristics. Validation of the predicted behaviour is a common subsequent step that is carried out to provide additional evidence and confidence that the design is appropriate. In addition to their application in product and process design, predictive models also play a key role in quantitative risk assessment, in the exposure assessment stage. Here, predictive models are used to estimate how certain factors affect the frequency and levels of pathogens in foods at the point of consumption (ICMSF, 2002). Together with baseline surveys of pathogens in raw materials, predictive models have provided valuable sources of information for deriving probable exposure estimates (i.e. the dose of the hazard likely to be presented to the consumer in a portion of food) for pathogenic bacteria (ICMSF, 1998). Risk assessments have been used to compare risks posed by different product types (FAO/WHO, 2004; FDA-CFSAN and USDA-FSIS, 2003) and to identify higher risk products and steps in processing or handling that will have the biggest impact on reducing or increasing risk, using tools such as sensitivity analysis. The purpose of this chapter is to describe the different biological ‘end-points’ that are relevant to the application of predictive models in chilled food manufacture, to refer to key considerations for development of various models, to explain the different mathematical approaches taken and to provide some case studies that demonstrate application.

17.2 Predictive microbiological models: experimental design/ set-up 17.2.1 Important factors in experimental design/set-up Models may be used to estimate survival, growth and die-off of micro-organisms. © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

479

There are a number of considerations that are important in studies intended to generate models; these include choice of micro-organism, preparation of test medium that adequately simulates the product of interest, and recovery, identification and enumeration methods. As a principle, it is best practice to use strains that are most tolerant to the stress conditions being tested but where the response is still representative of the group of organisms identified as the target. In some cases, strains that have been used are atypically resistant or tolerant to the conditions being tested and use of these strains may generate data that lead to an overly conservative model. For example, Salmonella Senftenberg 775W is sometimes used in modelling studies on heating because of its relatively high heat resistance but when this is compared to other salmonellae, it is unusually high. Therefore, other strains should be used, unless there are good reasons to use these atypically resistant strains, e.g. they are representative of the target group of organisms. In modelling studies, cocktails are often used because strains differ in their response to different combinations of factors and to generate a ‘leading edge’ model it is better to include a mixture of strains that may be tolerant of the particular stress conditions being tested. Cocktails are also more likely to contain a range of physiological responses that may not be fully characterised, but may represent the diversity found in naturally occurring microbial populations. Strain-to-strain variability has been shown to be important for survival modelling. Buchanan and Edelson (1996) concluded that pH-dependent and pH-independent stationary-phase acid tolerant phenotypes may exist among enterohaemorrhagic Escherichia coli. Interestingly, they also observed that three of the strains that were pH-independent acid tolerant (no need to induce acid tolerance by pre-exposing to low pH) were associated with large foodborne outbreaks. The same authors also concluded that pre-adaptation of cells is important for survival studies with acid foods. McClure (2001) reported on the influence of pre-adapting strains to an acidic environment by subculturing individual strains of a 15-strain cocktail of enterohaemorrhagic E. coli. Under the conditions tested, there was no advantage conferred to cells that were acid-adapted. Deng et al. (1998) found that acid-adapted cells retain higher viability than unadapted cells in only two of nine foods tested. The recovery method and medium are also important in modelling studies. If naturally contaminated foods are used, e.g. in survival studies, these sometimes employ selective media (selecting for the target group of organisms) that may underestimate the numbers of survivors present if those organisms are stressed. McClure (2001) compared recovery of E. coli O157:H7 on Sorbitol MacConkey agar (SMAC), Sorbitol MacConkey agar with added pyruvate (SMAC-PY) and on TSA after inoculation into the high salt, reduced pH conditions described above. Recovery on SMAC and SMAC-PY was poor compared with TSA. If selective media are used to recover injured or stressed cells, modification to the recovery protocol may be required, as described by McCarthy et al. (1998). Similar effects have been shown for other pathogens, such as Salmonella spp. (Kang and Fung, 2000; Kinsella et al., 2006; Liao and Fett, 2005). In addition to the recovery medium, the composition of diluent used for serial dilutions may also be important © 2008, Woodhead Publishing Limited

480

Chilled foods

(Jordan et al., 1999). Other factors shown to be important in modelling studies include pH and acid type (Deng et al., 1999) and storage temperature. Whereas most adverse conditions, such as reduced pH and aw, tend to increase the death rate of micro-organisms, lower temperatures have the opposite effect, prolonging survival. Although the mechanisms of survival are still relatively poorly understood, death rate influenced by factors such as pH is likely to be related indirectly to energy depletion or ‘metabolic exhaustion’, as termed by Leistner (2000), and lower temperatures probably result in a lower ‘drain’ on the energy resources of microbes. The physiological responses of bacteria to different ‘stresses’ often used in modelling studies are discussed by Gould et al. (1983), McMeekin et al. (2002) and Smelt et al. (2002a). Other important considerations for survival modelling include choice of appropriate independent or explanatory variables and the dependent variable measured. For example, a commonly used independent variable is pH and from a biological perspective, it may be better to use hydrogen ion concentration directly (rather than pH), as used by Takumi et al. (2000), when modelling the effects of stomach pH on survival of E. coli in the human gut. If organic acids are used, then the amount of acid is also important and the relevant independent variable is undissociated acid concentration. For emulsion systems that employ ‘actives’ that partition into the oil phase (such as sorbic acid), there will be less undissociated acid available in the water phase and this is the effective component to use. For thermal processing, temperature, exposure time and the composition of the food are important, as are the numbers and types of micro-organisms present, pH value, amounts and type(s) of preservatives present, water activity (carbohydrate content), fat content, protein content and the expected storage conditions of the product. Other factors such as history of the cells used, growth/sporulation conditions, strains used, etc., are important considerations and can impact on the kinetics of inactivation being followed. Recovery conditions are also important. It is common practice to generate thermal inactivation data in laboratory media, so that these factors can be more easily controlled and comparable results obtained. However, the application of the model will be in food systems and it is therefore important to validate models using data generated in foods, to demonstrate that the models can be used with some degree of confidence.

17.2.2 Modelling considerations for different physiological events Modelling survival Modelling the survival (non-thermal inactivation) of micro-organisms is an important consideration for foods that may contain unwanted contaminants and that do not employ commonly-used intervention processes, such as thermal processing (modelling thermal inactivation is covered later in this chapter). The characteristics of some products (e.g. salads) are adversely affected by heating and these have to rely on alternative intervention techniques. Other products may be multiple-use and contaminated after opening. These include chilled products (such as mayon© 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

481

naise or dips) that use factors such as low pH and/or low water activity (aw) or other components to reduce levels of particular contaminants that may be present on raw materials or introduced during processing or handling. Survival modelling is often focussed on infectious pathogens that must be absent in foods at the point of consumption, such as Salmonella spp. and enterohaemorrhagic E. coli O157:H7, or reduced to low numbers, e.g. Listeria monocytogenes. For multiple-use products, shelf-life can extend for many weeks after opening and it is critical that the same consideration for design is applied to ‘open’ as well as ‘closed’ shelf-life. Chilled products that make use of non-thermal intervention processes include fermented meats, fermented dairy products (e.g. yoghurts and cheeses), dressings or mayonnaises and beverages (e.g. fruit juices). Multiple-use chilled products that should not support growth of pathogens or spoilage organisms during open shelflife include spreads, salad dressings, preserves (e.g. jams) and sauces. In order to determine the potential for micro-organisms to survive in particular foods, during specified processing and storage regimes, it is important to be able to describe the kinetics of survival. It is also important to identify and characterise the food properties that influence survival. This is important from an experimental planning perspective, to identify appropriate intervals, particularly if there are interdependencies between different factors such as pH and organic acid concentration. In some respects, the situation with survival of vegetative pathogens is analogous to the situation some years ago with pathogen growth. Compared to growth modelling, there are relatively few studies that have attempted to derive models predicting the survival of pathogens in response to different environmental conditions. With survival studies, much of the published literature focuses on particular foods, and many of these describe numbers of survivors with time. This survivability is dependent on a number of factors, and, as with growth, some of these are now well known. Unlike growth, however, where the shape of the response (growth curve) is generally the same, i.e. sigmoid, the kinetics of survival are not easily predicted. For example, Fig. 17.1 (data generated in our laboratory) shows the response of a cocktail of E. coli O157:H7 strains to a range of different ‘stress’ conditions. The survival curves showed a range of different shapes, e.g. some conditions showed a steady (linear) decline in numbers (Fig 17.1b), some showed initial rapid decrease with a subsequent tail (Fig 17.1a), others showed an initial shoulder followed by a phase of rapid decrease (Fig. 17.1d), and others showed a biphasic shape (Fig. 17.1c). The curves were ‘fitted’ using two different functions, the log–logistic function (described in Cole et al., 1993) and the nonlinear model of Whiting (1993). This stage of modelling when representative kinetics are being established is often referred to as primary modelling. Even though the kinetics of survival can be variable, it is possible to derive ‘secondary’ mathematical models that can provide relatively good estimates of time to a defined, e.g. 4 or 5 log10, reduction. Performance criteria and limits for successful validation remain to be established (Ross et al., 2000) even though standard statistical measures such as standard errors, sum of squared errors, mean sum of squared errors and root mean squared error give some indication of the ‘goodness of fit’. However, useful models have been derived predicting the fate of © 2008, Woodhead Publishing Limited

(c)

Log (count)

0

20

40

60

80

8 7 6 5 4 3 2 1 0

100

Chilled foods

8 7 6 5 4 3 2 1 0

482

Log (count)

(a)

0

20

Time (days) logcount

Logistic

40

logcount

Whiting

Logistic

40

Fig. 17.1

Logistic

80

100

Whiting

60

80

100

8 7 6 5 4 3 2 1 0 0

20

40

60

Time (days)

Time (days) logcount

100

(d) Log (count)

Log (count)

(b)

20

80

Time (days)

8 7 6 5 4 3 2 1 0 0

60

Whiting

logcount

Logistic

Whiting

Examples of survivor curves: (a) NaCl 19%, pH 5, 20 °C, 0.2% lactic acid and 0 nitrite; (b) NaCl 12%, 20 °C, 0.2% lactic acid and 0 nitrite; (c) NaCl 6%, 25 °C, 1.0% lactic acid and 0 nitrite; (d) NaCl 12%, 20 °C, 1.0% lactic acid and 36 ppm nitrite.

© 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

483

L. monocytogenes in response to lactic acid, pH, NaCl and temperature (Buchanan et al., 1997b). Various approaches have been used and include:

• ‘vitalistic’ approach in a one-step procedure (Skandamis et al., 2002), for • •

describing the effects of pH, refrigeration temperatures and concentration of oregano oil on survival of Salmonella Enteritidis and E. coli O157:H7, neural network modelling to describe the fate of S. Enteritidis in mayonnaise with varying temperature, NaCl and mustard concentration (Xiong et al., 2002), use of the Weibull model to explain the effects of citric and lactic acid and temperature on inactivation of L. monocytogenes and E. coli (Virto et al., 2006).

Recently, Geeraerd et al. (2006) reported on a freeware tool that includes functions that may be used to explain all the known survivor curve shapes shown for vegetative bacterial cells. Modelling growth Modelling growth and minimum growth temperature of micro-organisms able to grow at refrigeration temperatures is an important step in the design of chilled foods. These organisms include infectious agents such as L. monocytogenes, which is tolerated at low levels in many foods in some countries, toxico-infectious agents such as psychrotrophic Bacillus cereus and toxigenic agents such psychrotrophic (non-proteolytic) Clostridium botulinum. Where L. monocytogenes is tolerated in foods, such as in Europe, this is generally up to a level of 100 cfu/g of food, and for psychrotrophic C. botulinum, toxin is produced during exponential growth and when growth occurs, this generally occurs relatively quickly. Therefore, for these two micro-organisms, growth/no growth models are often more appropriate, since manufacturers need to design foods that do not support even small amounts of growth during shelf-life. For organisms that can be tolerated at higher levels, such as B. cereus and spoilage organisms able to grow at chill, such as Pseudomonas spp., yeasts, moulds, lactic acid bacteria and psychrotrophic spoilage clostridia, kinetic models are more relevant to predict time to an n log increase and the onset of spoilage. Growth/no growth or ‘boundary’ models tend to be based on ‘positive’ and ‘negative’ observations, and the type of model used influences the experimental design employed in data generation. Developing growth/no growth models is generally less labour intensive than developing kinetic models since the end-points can be measured more easily and they often make use of automated techniques, e.g. microtitre plates. These models are amongst the most useful in product design since they provide the user with the ability to identify combinations of food characteristics or conditions that clearly delineate areas of inhibition, sometimes referred to as ‘operating windows’. Growth models have been a major area of development in predictive microbiology over the past 20 years. Traditionally, these rely on generation of kinetic data that enable a description of the whole growth curve that includes lag time, exponential phase and stationary phase. The use of these models, as with other models, must take account of the harmful level and it is interesting to note that a © 2008, Woodhead Publishing Limited

484

Chilled foods

number of growth models have been developed for relatively low-infectious dose agents such as Salmonella, E. coli O157:H7 and Shigella flexneri. They are useful, for example, where there are steps in the food chain where opportunities for growth to occur and the numbers of these organisms may well impact on the severity of the intervention process required later to reduce their numbers. For the safety and stability of chilled foods, models should be developed based on relevance of the hazard to a combination of raw material, processing and food characteristics, and focused on those micro-organisms likely to limit the shelf-life or exceed specifications. Other models that rely on some aspect of growth include ‘time to toxin’ models and ‘time to spoilage’ models and these are both based on the population of micro-organisms reaching a critical level. For chilled foods, the most relevant ‘time to toxin’ application is for non-proteolytic C. botulinum (Baker and Genigeorgis, 1990; Genigeorgis et al., 1991) but since the nature of adverse health effects can be very severe, manufacturers must be extremely conservative in using such tools in design of foods and will preferentially base their designs on growth/ no-growth models or growth models where there are sufficiently long lag times to assure safety as a consequence of the preservation system used or heat treatment, or a combination of these. Modelling thermal inactivation Thermal inactivation was probably the first response to be modelled and predictive models of one form or another have been applied for more than 70 years, principally in the sterilisation of foods. The kinetics of inactivation are still the subject of much debate but overall, application of log-linear kinetic models that employ D- and z-values are accepted and applied successfully throughout food processing. Thermal processing remains the foremost method for the preservation of foods. The primary purpose is to destroy all micro-organisms capable of growth during subsequent normal storage of the food (commercial sterility). To determine a microbiologically safe heat process, the concentration of surviving microorganisms must be reduced to a level less than or equal to the maximum ‘harmless’ level. For thermal processes, microbiological data form the basis for the target process (time and temperature) set for a specific product. These data are derived from research that strives to understand and describe microbial death kinetics. This traditionally involves following the inactivation of specific microbial populations exposed to specified lethal temperatures (either constant or predictably changing) under well-controlled conditions, by recovering and enumerating surviving organisms. The principle microbial targets for ‘pasteurisation’ processes that employ relatively mild thermal processing are food-poisoning and spoilage micro-organisms. 70 °C for 2 min is used for the inactivation of vegetative infectious pathogens associated with raw materials and food processing environments. These include micro-organisms such as salmonellae, L. monocytogenes, campylobacters and pathogenic E. coli. Presence of some of these pathogens, even in low numbers, can cause foodborne disease and their control is essential to assure consumer safety. ‘Pasteurisation’ processes allow more heat-resistant micro-organisms, such as © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

485

bacterial sporeformers, to survive and, if these are present in the product they must be controlled through other means, such as refrigeration, presence of preservatives, reduced aw or low pH, or combinations of these. Spoilage micro-organisms must also be controlled through either more severe heating or other preservation processes and it is often the case that measures taken to control these organisms will also control pathogens. At the same time, it is important to appreciate that heat, as well as destroying micro-organisms, may also provide desirable characteristics or damage the nutritional and organoleptic properties of foods. It is therefore important that the specific thermal processes used should give the desired food quality and provide that required for microbiological safety and stability. The key considerations for intervention processes such as thermal inactivation are identification of realistic hazards, agreement on acceptable risk, finding preventive measures, acceptance of effective controls, implementation of agreed controls, monitoring and review of control and pre-agreement of failure management. The effective management system of choice for control and monitoring of processes is the HACCP approach. In order to determine what intervention process is necessary, the performance standard needs to be identified. A common performance standard (or preservation specification) for pasteurised chilled ready-to-eat products marketed in the UK is a 6 log inactivation of vegetative infectious pathogens, such as L. monocytogenes. Since this organism is relatively more heat resistant than other infectious foodborne pathogens, inactivation of this organism will also effect large log reductions (i.e. > 6 log) in numbers of other infectious pathogens. In order to determine what process conditions (e.g. time and temperature of heating) are needed to meet this performance standard or heat treatment, robust and reliable microbiological thermal inactivation data are required. This is one of the key elements involved in designing thermal processes and these data are commonly described by mathematical models. Modelling injury after heating Thermal processes for long shelf-life chilled products are traditionally based on well-defined performance criteria, such as 6D reduction of spores of psychrotrophic strains of C. botulinum. For processes that rely on combinations of heating and preservative factors, i.e. some destruction + injury of surviving micro-organisms, an alternative performance criterion that can be applied is ‘less than a 1 log increase in numbers of C. botulinum within the use-by-date when stored at the recommended storage temperature’ (ICMSF, 2002). To generate models that are able to predict these effects, foods are inoculated with the target organism, heated, cooled and then stored under normal use conditions. The containers are examined at intervals after heating to measure any changes in numbers. Alternatively, they may be examined for organoleptic changes (e.g. pH, colour or texture) if there is no safety risk and spoilage is the concern. Models predicting injury effects are relatively uncommon but offer the potential to provide significant gains in identifying conditions that are less severe but can still be relied upon to assure safety and quality. Mafart (2000) proposed a mathematical approach to take account of © 2008, Woodhead Publishing Limited

486

Chilled foods

damage to heated surviving cells when assessing the effective overall reduction ratio reached after a heat treatment. There are few studies that have derived models of ‘injury’ but some that describe the effects on subsequent ability to grow. An early example of this was provided by Hauschild (1982), where two terms, one for inactivation and one for injury, were used to describe the degree of protection provided by particular preservation systems. This concept was also used by Lund’s group to describe the effects of heat treatment and other factors for controlling C. botulinum (see Lund, 1993, for more details). Smelt et al. (2002b) analysed the effect of sublethal injury on the distribution of lag times of individual cells of Lactobacillus plantarum. They found that the mean and standard deviation of lag times increased when the sublethal treatments became increasingly severe. The influence of various stress conditions on the individual lag time of Listeria monocytogenes was studied by Guillier et al. (2005). The authors ranked the stress conditions as functions of their effect on lag time: the heat-treatment effect was found to be one of the most stressful conditions among the nine ‘stresses’ tested. Palop et al. (1997) reported on the effects of reduced pH and the presence of organic acids in preventing outgrowth of spores of B. coagulans, heated using relatively mild treatments of around 10 s at 100 °C. There are a number of other studies that have shown sublethally heat-injured spores to have an increased sensitivity to antimicrobials (e.g. acid, NaCl) commonly used in foods (Blocher and Busta, 1983; Cook and Pierson, 1983; Faille et al., 1997; Foegeding and Busta, 1981; Roberts et al., 1966). Braithwaite and Perigo (1971) and Bean (1983) carried out extensive studies on the combined effects of heat, pH and water activity on spores from a large number of Bacillus spp. and Clostridium spp., identifying conditions that required less heat treatment than would normally be required to cause a specified reduction in numbers or time to growth. Although these data were not ‘modelled’ in the form of a mathematical representation, they were depicted pictorially with isometric illustrations and isopleths, indicating combinations of conditions that could be used to produce safe and stable products without the need to resort to typical sterilisation processes. As far as we are aware, no such data sets exist specifically for psychrotrophic spore-forming organisms.

17.2.3 Modelling approaches Various mathematical models are been developed and/or applied in food microbiology in the past 30 or so years. There has been extensive discussion over so-called empirical models, which simply aim to describe data with an appropriate mathematical relationship, and mechanistic approaches that provide an interpretation of the response based on some underlying mechanism. This notwithstanding, there are few, if any, mathematical models in food microbiology that are truly mechanistic in nature and based on an understanding of underlying mechanisms. The process followed in model development often proceeds with application of primary models to experimental data, and then fitting of secondary models to the environmental conditions. So-called ‘tertiary’ models refer to software applica© 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

487

tions. This terminology was proposed by Whiting and Buchanan (1993) to distinguish between the different levels of development. The various models that are used in predictive food microbiology are briefly described below. Primary models The purpose of primary models is to determine the magnitude of the physiological response measured, e.g. lag time, maximum specific growth rate, death or inactivation rate, time to specified increase (or decrease) in numbers or concentration of metabolite, under a set of specified conditions. The shape of the response being measured determines the type of model used to fit the data generated. For growth curves, the shape is commonly sigmoidal if a logarithmic transformation of cell numbers is used. Primary models that are most often used to fit growth curves include:

• the Baranyi model (Baranyi et al., 1993), • the Gompertz function (as used by Gibson et al., 1988), • the modified Gompertz function, first described by Zwietering et al. (1990) that was reparameterised, using biologically-relevant terms. Alternative approaches have also been developed that can also take account of subsequent decreases in cell numbers (Whiting and Cygnarowicz-Provost, 1992) or that include terms to describe inactivation (Jones and Walker, 1993; van Impe et al., 1992). In studies that describe fitting of data with different functions, these often report relatively minor differences between the different approaches used. Other forms and more detailed description of primary models are reviewed by Lebert and Lebert (2006), McDonald and Sun (1999), McMeekin et al. (2002) and Swinnen et al. (2004). Microbial thermal inactivation was originally assumed to follow the kinetics of a first order reaction, where the kinetics are described by a log-linear function, first described by Chick (1908) for modelling disinfection. The kinetics in this case are typically described by D-values that describe the time for a 90% reduction in the population. The basis of this principle is that the death rate is constant at any given temperature and independent of the initial number of viable cells present. Thermal death time is defined as the time required to inactivate a given number of organisms at a specified temperature. The most heat-resistant spores of C. botulinum have a D-value at 121.1 °C of about 0.21 min and a 12D process is typically applied in the canning industry for high pH foods (Hersom and Hulland, 1980), using a z-value of 10 C°, taken from the data of Esty and Meyer (1922). However, these data were generated using an end-point method and neutral phosphate buffer and some questions remain about the appropriateness of using this approach (log-linear model) with the original data of Esty and Meyer (McClure et al., 2004). The kinetics of inactivation have often been shown to follow a log-linear shape but many studies have shown deviations from this linearity, with ‘shoulders’ (e.g. rapid die-off of a large fraction of the population) and ‘tails’ (e.g. increased survival of a few cells) being observed in data. Consequently, modelling thermal © 2008, Woodhead Publishing Limited

488

Chilled foods

inactivation has been the subject of much debate (Moats et al., 1971) and a number of alternative approaches (Cerf et al., 1996; Kilsby et al., 2000; Mafart et al., 2002; Peleg and Cole, 1998) have been proposed to take account of non log-linear kinetic data. Tailing can be an artefact of the experimental technique used, but there are numerous studies where meticulous care has been taken to eliminate this as a cause. The novel approaches above assume some distribution of inactivation times within populations of bacterial cells. In addition to these novel approaches, other workers have simply fitted existing sigmoid functions commonly used to fit growth curves. For example, the modified Gompertz (Bhaduri et al., 1991), logistic (Juneja et al., 1999a) and Baranyi model (Whiting, 1993) have all been used to describe thermal inactivation of L. monocytogenes. Secondary models Secondary models describe the response of one or more parameters (e.g. lag time, growth rate, death rate) from a primary model in relation to changes in environmental conditions. Early ‘growth’ studies focussed on development of ‘probability’ models, predicting the probability of toxin formation by C. botulinum in foods with different levels antimicrobial agents (e.g. nitrite, NaCl, sorbate) present. The first attempt to quantify the effects of these various preservation systems (Hauschild, 1982) calculated the probability of outgrowth of spores of C. botulinum and expressed this as the reciprocal of P, where this value is equivalent to the decimal reduction value, e.g. P = 10–6 is interpreted as equivalent to a 6 log10 reduction in numbers. The approach of Hauschild (1982) was used by Lund et al. (1985) to predict the probability of outgrowth of spores of C. botulinum and included a temperature relevant to chilled foods (8 °C) and used spores of psychrotrophic strains. The study did not consider the relative contribution of each factor, nor interactions, but did allow comparison of the effects of each combination of factors. Logistic regression has been applied to toxic/non-toxic data, independent of time, and segmented linear models have also been applied to predict the probability of growth of non-proteolytic type B C. botulinum under conditions relevant to chilled foods (Lund et al., 1990). There have been a number of kinetic secondary models that have been used to describe the effects of various environmental conditions on the response of microorganisms. The most common of these include the Arrhenius, modified Arrhenius, square root, γ-concept (gamma) and polynomial models. The appropriateness of various models has been fiercely debated and each has its own merits. The complexity of these different approaches varies and this has implications for the amount of experimental data needed. Where more factors are considered, more data are required and commonly, more complex secondary models are needed to accommodate the changes in response and interaction of different factors on the dependent variable being modelled. There are key considerations for the models that should be applied to experimental data and these are described in detail in McMeekin et al. (1993). More recent publications (Lebert and Lebert, 2006; McDonald and Sun, 1999; McKellar and Lu, 2004; Swinnen et al., 2004) provide © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

489

brief descriptions of these various functions and refer to more detailed studies that give examples of particular applications. As with growth models, so-called ‘secondary’ or ‘second-level’ models may then be fitted to the parameter(s) describing inactivation at each set of conditions tested. Response surface models are often used for these applications and other techniques include the Bigelow model and modifications to this (Leguerinel et al., 2000; Mafart, 2000). Using a single-equation model for all data (Kilsby et al., 2000) has some advantages over the more common two-stage approach. In many studies investigating thermal inactivation of foodborne bacteria, the data are described using first-order reaction kinetics, i.e. D-values, even though some of these studies clearly show data that are not properly described by log-linear functions. Deterministic vs. probabilistic modelling approaches Deterministic models use inputs (e.g. initial concentration of cells or spores or product temperature) as single values, without considering the variability and uncertainty that may be associated with a specific product and process design (e.g. heat process or product pH). Consequently, the output values from the model are single values as well. In that context, the food industry has found it advantageous to use deterministic modelling approaches because, in most instances, the input and output values for the model are relatively easy to interpret. However, an important drawback is that deterministic models are often made to generate rather conservative outputs, because the model inputs are mostly also chosen conservatively (i.e. worst-case inputs). The approach has been of value for a long time in the food industry as process control is often variable. More recently, models that take account of variability between individual cells have been developed, embedding so-called ‘stochastic’ elements. These models are distinguished from deterministic models that do not include consideration of such random properties. The variability and uncertainty of population dynamics for small spore populations in foods was recently modelled using Bayesian inference methodology, in relation to non-proteolytic C. botulinum (Barker et al., 2005a). Probabilistic modelling approaches have been advocated by several researchers in the area of microbiological food safety (Cassin et al., 1998; Nauta, 2001; van Gerwen and Gorris, 2004; Vose, 1998). These researchers have shown that this approach is helpful in providing insight into the impact of data variability and in articulating data uncertainties. In probabilistic modelling, input parameters are introduced to the model as statistical distributions which are typically analysed using the Monte Carlo simulation technique (Poschet et al., 2003). Each iteration undergone during the Monte Carlo simulation provides a single data point, and multiple iterations build up a distribution of values that constitutes the model output (as opposed to a single value as is the case in deterministic modelling). The resulting probability curve represents, in most cases, a more realistic situation of the likely microbial response to the set of input parameters used in the model. In turn, this model output is more helpful in supporting informed decision making regarding the suitability of product and process design or production controls. © 2008, Woodhead Publishing Limited

490

Chilled foods

Most applications of probabilistic modelling are concerned with predicting the growth of micro-organisms (Ross and McMeekin, 2003). In such cases, Monte Carlo method techniques are oftentimes combined with predictive growth models (Cassin et al., 1998; Poschet et al., 2003) or with process modelling in microbiological risk assessments (Montville and Schaffner, 2005; Nauta, 2001). Our laboratory has been developing applications using probabilistic modelling techniques to assess the efficacy of thermal processes in inactivation of B. cereus in cook–chill foods. These techniques afford real advantages in being able to consider uncertainty and variability, not only in the responses measured but also for other aspects, such as the range of time/temperature combinations achieved in packs during ‘in-pack’ heat treatment (Membré et al., 2006). Other recent applications of probabilistic models include the estimation of shelf-life of pasteurised milk (Schaffner et al., 2004a), determination of microbial contamination rates on plastic cutting boards in use in foodservice kitchens (Schaffner et al., 2004b), and consumer phase (i.e. focused on handling practices in the home) risk assessments for L. monocytogenes in deli meats (Yang et al., 2006) and for S. enterica serovar Enteritidis in egg-containing foods (Mokhtari et al., 2006). 17.2.4 Evaluating the performance of predictive models There are various methods that can be used to assess ‘goodness of fit’ and these have been used to determine whether particular models are acceptable from a statistical viewpoint, and also to compare different models with each other. For assessing goodness of fit to the experimental data, standard indices such as Mean square error and sums of squares of residuals (RSS) (found in standard statistical texts such as Causton, 1987; Draper and Smith, 1981) are used. This aspect of performance is related to the experimental dataset only and is part of the function fitting step (i.e. fitting a function that describes the response being modelled). For ‘validation’ with independent data such as published data (e.g. Blackburn et al., 1997), bias and accuracy factors (Ross, 1996) and estimation of the integral mean of the square differences between different models (Baranyi et al., 1999) have been proposed. A classification of the sources of errors in predictive models is described by Baranyi and Roberts (1995). This step requires careful consideration because there may be good reasons why independent data do not match model predictions. There may be factors or variables related to the independent data (foods are commonly used) that are not accounted for in the experimental data used to generate the (original) model, and sometimes the independent data are outside the range or ‘domain of validity’ of the model – these are important to bear in mind. Deviations from predictions do not necessarily mean that the model is flawed. However, when such deviations occur, great care must be taken in use of models. 17.2.5 Applications in the supply chain For chilled foods, the most relevant models for microbiological safety and quality are: © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

491

• inactivation models for vegetative infectious pathogens; • inactivation models for psychrotrophic spore-forming pathogenic organisms; • growth or boundary models for higher infective dose psychrotrophic infectious pathogens;

• growth or boundary models for psychrotrophic toxigenic or toxico-infectious micro-organisms;

• growth or boundary models for psychrotrophic spoilage micro-organisms; • probabilistic models for both spoilage and pathogenic organisms capable of growing at refrigeration temperatures. The selection of realistic and relevant micro-organisms (and their levels) limits the number of targets that should be considered. Other data that are required in modelling for chilled foods include any preservative factors in the food and temperature distributions during transport and storage, before retail, at retail and post-retail, up to the point of consumption. Temperature data from these various stages during handling storage and distribution are available and are referred to in some modelling studies, e.g. Nauta et al. (2003) and Rosset et al. (2004).

17.3 Availability of predictive microbiology models for chilled foods The considerable advancement of predictive microbiology research and its application since the early 1990s has resulted in an abundant number of models to predict microbial responses and/or growth boundaries in foods. Most efforts have concentrated on modelling the behaviour of pathogenic micro-organisms. With respect to chilled foods, four psychrotrophic pathogens are of particular relevance because of their growth characteristics and potential for harm, and hence require special consideration: Listeria monocytogenes, non-proteolytic Clostridium botulinum, Bacillus cereus and Yersinia enterocolitica (ECFF, 2006; NACMCF, 2005). In the case of chilled foods that include a pasteurisation step as part of the product/process design, Clostridium perfringens should also be considered as a relevant pathogen due to the ability of its spores to germinate and outgrow rapidly during cooling before final refrigerated storage (Labbe and Huang, 1995; USDA, 2005). This section presents a summary of models available in the public domain for the above-mentioned pathogenic micro-organisms, as well as some relevant available models for psychrotrophic spoilage bacteria.

17.3.1 Models available for pathogenic micro-organisms relevant to chilled foods The majority of the models presented in this section are available as research papers that are not readily available as software packages or user-friendly spreadsheets. The results and conclusions from these papers may be enough to answer a particular question about the behaviour of a micro-organism under a set © 2008, Woodhead Publishing Limited

492

Chilled foods

of environmental conditions. Although the development of these models may require sophisticated mathematical and/or statistical software packages, the final outcome is oftentimes a relatively straightforward equation or set of equations that the user could implement in any available spreadsheet package. In such cases, it is extremely important that the user verifies that the predictions from the spreadsheet results match the results reported in the original paper. After that, the user can utilise the model to simulate other scenarios, provided the limits of validity of the model and any other constraints and/or assumptions reported in the original paper are understood and followed. This must be done to avoid model misuse and misinterpretation of predictions. Models available for Listeria monocytogenes Of all the psychrotrophic pathogens of concern in the production of chilled foods, most modelling work has been done with L. monocytogenes. In the early days of predictive microbiology, most efforts were spent on population-based deterministic models describing growth and/or inactivation rates of the pathogen as influenced by environmental factors. Growth and/or survival curves were typically fitted with primary models, represented by sigmoidal functions such as the Gompertz, logistic or Baranyi models. The effect of environmental factors on growth kinetics or survival was subsequently modelled using secondary models such as response surface or polynomial models. The main environmental factors included as independent variables in these models were temperature, pH, salt (or aw) and sodium nitrite concentrations (Buchanan and Phillips, 1990; Cole et al., 1990; Hudson, 1994; McClure et al., 1997; Wijtzes et al., 1993). Other factors considered included CO2 concentration (Farber et al., 1996), organic acids (George et al., 1996), lipid concentration (Guerzoni et al., 1994), phenol concentration (Membré et al., 1997), and antioxidants such as BHA and BHT (Yousef et al., 1991). Modelling the thermal inactivation of L. monocytogenes has also been an area of interest in the predictive microbiology field. As mentioned earlier, mild ‘pasteurisation’ processes targeting vegetative infectious pathogens associated with raw materials and food processing environments typically are designed to achieve a minimum of 70 °C for 2 min. It is generally accepted that such treatment will deliver a 6D inactivation of L. monocytogenes cells. Research efforts have consequently focused on determining D- and z-values, as well as trying to determine the effects of heating rate, non-isothermal treatments, and deviations from log-linear inactivation kinetics on the thermal inactivation of this pathogen (Augustin et al., 1998; Bhaduri et al., 1991; Cole et al., 1993; Fernandez et al., 2007; Hassani et al., 2005; Juneja, 2003; Juneja and Eblen, 1999; Linton et al., 1995; Murphy et al., 2003; Schultze et al., 2006; Stephens et al., 1994; Valdramidis et al., 2006). Considering the risk of post-thermal process L. monocytogenes contamination of foods, post-packaging pasteurisation processes are commonly implemented in manufacturing environments dedicated to the production of readyto-eat products. In such cases, the design target should depend on the level (i.e. number of L. monocytogenes cells) of post-process contamination and the limits © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

493

set by regulatory requirements and microbiological criteria. For example, for ready-to-eat meat and poultry products in the United States, the USDA Food Safety and Inspection Service (FSIS) considers a post lethality treatment (PLT) to be effective if it ensures at least a 1D inactivation of L. monocytogenes prior to product release (USDA-FSIS, 2005). More recently, considering the severity of foodborne listeriosis and the stringency of regulatory requirements and microbiological criteria in different countries (e.g. EU microbiological criteria, see 2073/2005 EC, Chapter 1, European-Commission, 2005), the focus of models for L. monocytogenes has changed towards growth/no-growth interface or boundary models. This has allowed food companies to clearly define operating windows when designing new formulations to assure that no growth of L. monocytogenes will occur during the shelf-life of chilled food products. The ability to define product characteristics required to prevent the growth of L. monocytogenes, makes these models useful in identifying critical control points as part of HACCP plans, as well as in establishing control measures conducive to compliance with regulatory requirements. Legan et al. (2004) developed and validated successfully a model that can describe the growth/ no-growth boundary for L. monocytogenes in both cured and uncured ready-to-eat meat products during storage at 4 °C as influenced by the concentration of NaCl (0.8% to 3.6%), moisture (45.5% to 83.5%), potassium lactate (60% wt/wt) syrup (0.25% to 9.25%), sodium diacetate (0.0% to 0.2%), and the presence of curing salts. Mejlholm and Dalgaard (2007) developed a growth boundary model for L. monocytogenes in lightly preserved seafood that included the effect of diacetate, lactate, CO2, smoke components, nitrite, pH, NaCl, temperature, and interactions between all these parameters. Other growth/no-growth models for this pathogen have been successfully developed in other food matrices such as Mexican-style cheese (Bolton and Frank, 1999), vacuum-packed cured ham (Mataragas et al., 2006a) and ground pork (Zuliani et al., 2006). Also, recent modelling efforts with respect to L. monocytogenes have focused on individual cell lag-phase prediction as well as the effect of the previous history on duration of lag. The duration of lag phase of a single cell can be considered a random stochastic event. In our laboratory, we reported significant effects on the distribution of detection times for single cells of L. monocytogenes exposed to 0.5 and 9% NaCl, with increasing stress resulting in increased variability in the response (McClure and Stephens, 1996). It has also been reported by other research groups that the duration of lag phase depends inversely on the size of the inoculum, and this effect can be explained by an increase in the variability of lag time of individual cells when stress factors become more stringent (Augustin et al., 2000). Robinson et al. (2001) reported that the variability of detection times of L. monocytogenes increased when a lower inoculum level was used, or when the salt content increased for a constant inoculum level. Similar results have been reported by other authors for Listeria innocua (Kutalik et al., 2005) and Lactobacillus plantarum (Smelt et al., 2002b). Recently, Francois et al. (2007) reported a model to evaluate the effect of precultural temperature and pH on the individual lag phase of L. monocytogenes at 7 °C, concluding that the lag phase of the pathogen was © 2008, Woodhead Publishing Limited

494

Chilled foods

shorter if the cells were pre-incubated at lower temperatures due to cell adaptation to the cold environment. Models available for non-proteolytic Clostridium botulinum Group I strains of Clostridium botulinum are not reported to grow below 12 °C and are therefore of limited significance to chilled foods, as long as the conditions of the supply chain do not result in temperature-abuse scenarios. A recent outbreak of botulism in the United States, associated with pasteurised carrot juice intended to be stored at chill, has raised questions about the robustness of the preservation system used for these products. From the existing evidence, it appears the outbreak was caused by toxin production from a proteolytic (Group I – type A) C. botulinum strain(s), which implies that there was some temperature abuse of product prior to consumption (Shuler et al., 2006). In light of this outbreak, the US Food and Drug Administration has issued a new guidance for refrigerated carrot juice and other refrigerated low-acid juices, which recommends juice processors subject to the pathogen reduction provisions of the juice HACCP regulation to ‘incorporate validated control measures for all C. botulinum spores into their HACCP plans that will be applied in the processing facility and that will ensure that C. botulinum growth and toxin production will not occur should the juice, as offered for sale by the processor, be kept unrefrigerated in distribution or by consumers’ (FDA, 2007). Examples of validated control measures include acidification of the juice to a pH of 4.6 or below, and a thermal treatment of the juice that ensures destruction of spores of proteolytic C. botulinum. Group II strains, which include non-proteolytic types B and F, and type E (all strains are non-proteolytic), are able to grow at low temperatures (e.g. < 5 °C), and therefore pose a risk for chilled foods. These strains are of particular concern in chilled foods since the organism is able to produce heat-resistant spores that are able to survive low-temperature pasteurisation processes (e.g. 70 °C for 2 min) that are commonly used to destroy vegetative infectious agents such as L. monocytogenes. Consequently, control of non-proteolytic C. botulinum is a primary concern for the safety of chilled foods. It is generally agreed that a temperature of 90 °C for 10 min will deliver a 6D inactivation of non-proteolytic C. botulinum and this is a commonly used performance standard for the heat processing of chilled foods (ECFF, 2006; FDA-CFSAN, 2001). This target time/temperature is based on the heat resistances of the more heat tolerant strains (type B strains) of this pathogen (Gaze and Brown, 1990). Predictive models developed and reported in the public domain for nonproteolytic C. botulinum typically aim at predicting the reduction in spore numbers during thermal inactivation as influenced by environmental factors (Juneja et al., 1995a,b) and/or the probability of growth from a single spore when different spore loads are treated by combinations of time and temperature and then incubated under a range of environmental conditions (Fernandez and Peck, 1997; Lund and Peck, 1994; Peck and Fernandez, 1995; Peck et al., 1995). The outputs from the models are usually lag phase duration and time to toxin production (Baker and Genigeorgis, 1990; Genigeorgis et al., 1991; Jianghong and Genigeorgis, 1993). © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

495

Although these models have attempted to establish the minimum heat treatments and combination of design parameters to assure the safety of chilled foods of extended shelf-life (e.g. > 4 weeks), they must not be taken as a substitute for challenge testing of products prior to release (Hyytia Trees et al., 2000). It is worth mentioning that some heat inactivation models for non-proteolytic C. botulinum have been developed using counts collected on recovery media containing lysozyme. These models are overly conservative because spores that are sub-lethally damaged lose their ability to germinate (due to inactivation of enzymes involved in germination) and can do so only in the presence of lysozyme-containing media. This was thought to have implications for lysozyme-containing foods but has not been shown to be significant since studies using foods containing naturally occurring lysozyme do not show the enhanced survival that would be expected, with very few exceptions (e.g. crab meat). More recent modelling approaches have used Bayesian inference techniques to quantify the variability of spore lag times in relation to contamination of foods with non-proteolytic C. botulinum as part of exposure assessment studies (Barker et al., 2005a,b). Models available for Bacillus cereus Bacillus cereus is reported as one of the psychrotrophic micro-organisms most commonly isolated from pasteurised and sous-vide foods (Carlin et al., 2000; Crielly et al., 1994; Harmon and Kautter, 1991). The psychrotrophy of B. cereus has been a controversial topic since the proposal of psychrotolerant strains as a new species, Bacillus weihenstephanensis (Lechner et al., 1998). This species comprises isolates that grow at 4–7 °C, but not at 42 °C, and are genetically distinct from ‘true’ B. cereus by specific sequences in 16S rDNA (von Stetten et al., 1998) and in the gene of cold shock protein, cspA (Francis et al., 1998). To date, there is no evidence of cases of foodborne illness in humans associated with B. weihenstephanensis strains, although it has been shown that some strains can produce the emetic toxin cereulide at temperatures as low as 8 °C (Thorsen et al., 2006). However, some of the psychrotrophic strains of ‘true’ B. cereus, with minimum growth temperatures of approximately 6–7 °C, have been associated with diarrhoeal foodborne disease cases (Stenfors and Granum, 2001). These strains are of particular concern for chilled foods since the organism is able to produce heat-resistant spores that are able to survive high-temperature pasteurisation processes commonly used in the production of chilled foods of extended shelf-life (e.g. 90 °C for 10 min). As with other pathogenic organisms, several growth models, as influenced by various environmental factors such as pH, aw, and CO2 (Baker and Griffiths, 1993; Olmez and Aran, 2005; Sutherland et al., 1996; Zwietering et al., 1996), as well as growth/no-growth boundary models (Lanciotti et al., 2001) have been reported in the literature for B. cereus. However, the greatest interest has been focused on modelling thermal inactivation of spores, which has been widely documented (Collado et al., 2003; Couvert et al., 1999; Fernandez et al., 1999a,b; Gaillard et al., 1998a,b; Gonzalez et al., 1995, 1997, 1999; Laurent et al., 1999; Leguerinel © 2008, Woodhead Publishing Limited

496

Chilled foods

and Mafart, 2001; Leguerinel et al., 2005; Mazas et al., 1995, 1997, 1998, 1999; Mikolajcik, 1970; Moussa-Boudjemaa et al., 2006; Shehata and Collins, 1972). In general, the heat-resistant characteristics of this micro-organism reported in the scientific literature show variations within the decimal reduction times (D-values) and/or the change in temperature required to change the D-values (z-values). These variations may be caused not only by differences in experimental conditions (e.g. sporulation and recovery conditions, heating media, methodology for data collection of heat inactivation curves), but also by differences among strains. For instance, Choma et al. (2000) reported that 52 strains had D90°C-values between 0.7 and 5.9 min, whereas another study found that nine strains had D90°C-values between 21.0 and 137.0 min (Gilbert et al., 1974). Bradshaw et al. (1975) reported strains isolated from spoiled canned vegetables with D-values at 121.1 °C and 129.6 °C around 2.3 and 0.3 min, respectively. It has been reported that psychrotrophic strains are more sensitive to heat than mesophilic strains (Choma et al., 2000; Dufrenne et al., 1994; Fernandez et al., 1999a). However, among both psychrotrophic and mesophilic strains, significant differences in heat sensitivity of spores are apparent; for instance, D90°C-values reported for the two types have ranged from 2.8 to 100 min and from 4.6 to 200 min, respectively (Dufrenne et al., 1994, 1995). Considering general patterns in heat resistance and relationship to minimum growth temperature, Carlin et al. (2006) reported that spores of emetic toxin-producing strains showed, on average, a higher heat resistance at 90 °C compared to non-emetic toxin-producers. Shehata and Collins (1972) studied the heat resistance of psychrotrophic strains of B. cereus isolated from pasteurised milk and when comparing these to another study that they claim used mesophilic strains, concluded that spores from the strains they used were less resistant to heat. However, this was a rather limited comparison and the other study (Mikolajcik, 1970) did not indicate whether the strains they used were mesophilic or psychrotrophic. Michels and Visser (1976) reported that spores of aerobic psychrotrophic and psychrophilic strains from soil showed comparable heat resistance and this was equal to that of more heat-sensitive mesophilic sporeformers. Recently in our laboratory we have used probabilistic modelling techniques to establish the prevalence and concentration of psychrotrophic and mesophilic B. cereus strains after thermal processing of a chilled food of extended shelf-life (Membré et al., 2006). By combining heat resistance data (D- and z-values) obtained from the scientific literature, ‘in-pack’ temperature distributions obtained from a heat transfer model, and raw material contamination levels obtained from expert opinion, we were able to estimate prevalence rates of 11% and 49% after the heat treatment for psychrotrophic and mesophilic strains, respectively, and a combined concentration of 30 cfu/g (95th percentile) in a low-acid semi-liquid chilled food of extended shelf-life. This information was then used as input for subsequent assessment of shelf-life of the product under realistic temperature conditions in the supply chain (Membré et al., 2005). Other recent applications of probabilistic modelling techniques used to model the behaviour of B. cereus include estimation of shelf-life of seasoned soybean © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

497

sprouts stored at refrigeration conditions (Dong Sun et al., 2006) and exposure assessment of Kimbab (a cooked rice-based Korean product) from retail stores to the point of consumption (Gyung Jin et al., 2007). Modelling approaches for B. cereus in recent years have also focused on estimation of lag time as influenced by the history of the spores (e.g. damage caused by heat treatment) and the environmental conditions of the new environment. For instance, Gaillard et al. (2005) reported that lag times of heated B. cereus spores show complex patterns when plotted against time of heating, but using a heating time delivering a 2-log reduction a linear relationship can be observed. Models available for Yersinia enterocolitica Yersinia enterocolitica is unusual amongst pathogenic enterobacteria due to its ability to grow at low temperatures (e.g. 4 °C). In fact, it has been reported that it is able to grow at temperatures as low as –1.5 °C in vacuum-packed roast beef (Hudson et al., 1994). Early modelling efforts were focused on growth and survival kinetic models as influenced by temperature, pH, NaCl and acidulant concentration (Adams et al., 1991; Bhaduri et al., 1994; Davey, 1994; Hudson, 1993; Jones and Walker, 1993; Little et al., 1992a,b). This organism is an important hazard in refrigerated muscle foods, particularly in pork, beef, fish and poultry. As modified atmosphere packaging is frequently used to extend the shelflife of refrigerated meats, modelling growth of Y. enterocolitica under various combinations of CO2, O2 and N2 has also been of special interest (Pin et al., 2000; Wei et al., 2001). More recently, models have been developed for non-thermal inactivation of this pathogen using pulsed electric fields (Alvarez et al., 2003), high pressure processing (Haiqiang and Hoover, 2003) and organic acid combinations (Virto et al., 2005). Additionally, models evaluating the effect of competitive background flora in fermented foods have been developed with the aim of finding biopreservation alternatives as a control measure for this organism (Janssen et al., 2006; Vereecken et al., 2003). Models available for Clostridium perfringens Clostridium perfringens is a spore-forming foodborne pathogen of particular concern in pasteurised-chilled foods because it is one of the most rapidly growing bacteria, with generation times as short as 7.1 min reported in ground beef at 41 °C (Labbe and Huang, 1995). Therefore, it becomes a particularly relevant hazard during the cooling of products that have undergone a pasteurisation step. A commonly used performance criterion for safe cooling of cooked foodstuffs is that the cooling process must be designed to allow no more than a 1-log increase of C. perfringens (see for instance USDA, 1999). Consequently, most models available in the public domain have attempted to predict the potential growth from spores during cooling of various food products. Researchers from the US Department of Agriculture (USDA) have developed several models for growth of C. perfringens during cooling conditions in laboratory media (Juneja et al., 1999b), cured beef (Juneja et al., 1995b) and cured chicken (Juneja and Marks, 2002). These three models have been implemented © 2008, Woodhead Publishing Limited

498

Chilled foods

into the Pathogen Modeling Program (PMP; discussed later) with the purpose of serving as a tool to be used by meat processors when analysing cooling process deviations (as part of their HACCP plans). However, the accuracy of these models, particularly the uncured beef model (developed in laboratory media) has been questioned by the meat industry in the United States. As a result, other groups of research workers have validated (Smith and Schaffner, 2004) and proposed new models (Smith-Simpson and Schaffner, 2005) in an attempt to address the shortcomings of the existing model. Other models developed by the USDA in cooked ground beef include those reported by Huang (2003a,b) using different dynamic modelling approaches. Another recently developed dynamic model for growth of C. perfringens in uncured beef during cooling was developed and validated by the Institute of Food Research in the United Kingdom (Peck et al., 2004), and has been implemented as a user-friendly software package called Perfringens Predictor (discussed later). Other models available for growth of C. perfringens during cooling include other food matrices such as cured pork (Juneja et al., 2006), pea soup (de Jong et al., 2005) and cooked chilli (Blankenship et al., 1988). When modelling growth during cooling, it is fundamental to integrate physical and biological kinetics in a systematic and effective way. A typical approach followed by many researchers has been to oversimplify the heat transfer problem by assuming that the temperature history of a food product during cooling follows a lumped capacitance behaviour. This means that the internal resistance to heat transfer is neglected (i.e. the temperature within the food is nearly uniform), a phenomenon which only happens in solid metals with very high thermal conductivities. However, during cooling, the temperature inside foods is far from being uniform and not only does it depend on the thermo-physical properties of the material, but also it involves a complex system at the surface which typically combines heat losses due to convection, radiation and evaporation. In order to address this, Amézquita et al. (2005b) integrated a transient heat transfer model (see Amézquita et al., 2005a for details of this model) with a dynamic growth model to predict C. perfringens growth during the cooling of cooked boneless cured ham. The model was validated on three different cooling scenarios, two of them simulating process deviations. Good agreement between predicted and observed values was observed for all tests within each cooling scenario.

17.3.2 Models available for spoilage micro-organisms relevant to chilled foods Modelling of biological kinetics for spoilage micro-organisms poses an important challenge because the behaviour of these organisms during shelf-life of chilled foods may vary as a function of product characteristics and storage conditions. Therefore, models for prediction of shelf-life of chilled foods require the knowledge of growth kinetics of specific spoilage organisms (SSO) in specific spoilage domains; that is, specific food matrices and storage conditions (Dalgaard, 1995). Another type of spoilage modelling approach commonly used in the food industry is based on the relative rate of spoilage (RRS) concept. RRS models allow the © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

499

prediction of shelf-life at different storage temperatures only if the shelf-life of the food product has been determined (e.g. by sensory evaluation) at a single known storage temperature, typically at 0 °C for chilled foods (Dalgaard et al., 2002). RRS models have a wider range of applicability because they do not rely on the kinetics of a SSO in a specific spoilage domain. Many efforts have been spent on spoilage models for raw seafood and raw and cooked meats. In the area of seafood spoilage modelling, researchers from the Danish Institute for Fisheries Research (DIFRES) have developed kinetic models for growth of Shewanella putrefaciens as a function of temperature (0 °C to 10 °C) (Dalgaard, 1995) and Photobacterium phosphoreim as influenced by temperature (0 °C to 15 °C) and CO2 concentration (0% to 100%) (Dalgaard, 1995; Dalgaard et al., 1997). Both of these models have been implemented into a user-friendly software package called Seafood Spoilage Predictor (Dalgaard et al., 2002) that will be discussed later. Other recent models developed by DIFRES researchers have focused on growth of spoilage microflora in cold-smoked salmon and their effect on growth of L. monocytogenes (Gimenez and Dalgaard, 2004). A group of researchers in Greece have also worked extensively on seafood spoilage modelling. This group identified pseudomonads as a good spoilage index in gilt-head seabream (Sparus aurata), by correlating results from the microbiological, organoleptical and chemical analysis conducted on naturally contaminated fish as well as on inoculated sterile fish under isothermal conditions ranging between 0 °C and 15 °C (Koutsoumanis and Nychas, 2000). This model was further developed to be used under non-isothermal conditions simulating realistic fish chilled supply chain environments (Koutsoumanis, 2001). Regarding fresh meat spoilage micro-organisms, models for growth of Brochothrix thermosphacta in laboratory media (Baranyi et al., 1995; McClure et al., 1993) and lactic acid bacteria in vacuum-packed meat (Nicolai et al., 1993) have been reported. In vacuum-packed cured cooked meats, lactic acid bacteria, particularly homofermentative lactobacilli such as Lactobacillus sake or Lactobacillus curvatus, and heterofermentative leuconostocs such as Leuconostoc mesenteroides and Leuconostoc carnosum typically predominate and outgrow competitive flora. Mataragas et al. (2006b) developed and validated a model for spoilage in sliced cooked cured meat products under isothermal and time-varying temperature conditions, demonstrating that lactic acid bacteria were the most prevalent organisms among others evaluated such as B. thermosphacta, Enterobacteriaceae and Pseudomonas spp. Models for spoilage of raw ground cured meat products have also been reported (Aggelis et al., 1998). Growth models for psychrotrophic pseudomonads have also been of interest as different species such as P. fluorescens, P. fragi and P. putida are common causes of spoilage in chilled-stored foods of neutral pH and high aw, especially nonfermented dairy and meat products. Neumeyer et al. (1997a) developed and validated (Neumeyer et al., 1997b) a model to predict the growth of psychrotrophic pseudomonads as influenced by temperature (0 °C to 30 °C) and aw (0.947 to 0.996), which has been implemented in a software package developed by the University of Tasmania called Food Spoilage Predictor© (discussed later). © 2008, Woodhead Publishing Limited

500

Chilled foods

17.3.3 Available predictive microbiology application software Several predictive microbiology software packages are available in the public domain. They represent a powerful tool for the food industry in supporting development of new products, reformulation of existing products, or evaluation of the effect of process deviations. Users must have sufficient expertise to understand the limitations and assumptions behind each model and the conditions under which each model is valid, thereby avoiding misinterpretation of model predictions. Moreover, although these packages may assist processors in reducing the amount of microbiological challenge testing required during design stages, they must not be taken as a complete substitute for testing products before release of a design for writing specifications. Available predictive microbiology application software packages are described below. Pathogen Modeling Program The PMP (http://www.ars.usda.gov/naa/errc/mfsru/pmp) is a stand-alone software package of microbial models and is a research product of the Microbial Food Safety Research Unit (MFS) of the Agricultural Research Service (ARS) of the USDA. The model is available free of charge. The latest version (PMP 7.0) currently contains 38 models for 11 bacterial pathogens, including one model for spoilage flora in chicken (inactivation by irradiation). Table 17.1 presents a summary of the micro-organisms and models included in the PMP 7.0. Most models included in the PMP are isothermal; however, it contains four models that are able to predict the growth of C. botulinum and C. perfringens under timevarying temperature conditions (cooling models). The PMP has become a premier international modelling tool and it is downloaded more than 8000 times each year in over 35 countries. Once downloaded, user-friendly features allow users to easily input food-relevant criteria and then to receive predictions about how pathogenic bacteria react to specific food environments. To further assist food processors in meeting regulatory requirements, references are provided for each model via direct Internet access to PDF files. A drawback of the PMP is the lack of information from validation studies showing the performance of models in specific foods as well as more general facilities to predict the effect of time-varying temperature conditions on growth and inactivation. Growth Predictor and Perfringens Predictor The British counterpart to the PMP, called ‘Food MicroModel’, developed in the early 1990s (McClure et al., 1994), is no longer available. However, data from Food MicroModel have been implemented into a stand-alone software package offered free of charge by the Institute of Food Research (IFR) since 2003 (http:// www.ifr.ac.uk/Safety/GrowthPredictor/) called Growth Predictor. Validation data and the data used to build the models included in Growth Predictor are available via ComBase (discussed later). Growth Predictor v. 1.01 includes 18 models for 13 organisms (17 models for pathogenic organisms and 1 model for growth of B. thermosphacta) and allows predictions to be obtained at constant conditions of temperature, pH, NaCl/aw and, in some models, an additional fourth parameter. All © 2008, Woodhead Publishing Limited

Table 17.1 Summary of microorganisms and models included in the Pathogen Modeling Program version 7.0 (PMP 7.0) Response variables(a)

Independent variables and ranges

Aeromonas hydrophila

Growth (aerobic)

GT, lag, time to n-log increase

Aeromonas hydrophila

Growth (anaerobic)

GT, lag, time to n-log increase

Bacillus cereus (vegetative cells)

Growth (aerobic)

GT, lag, time to n-log increase

Bacillus cereus (vegetative cells)

Growth (anaerobic)

GT, lag, time to n-log increase

Proteolytic Clostridium botulinum (spores) Proteolytic Clostridium botulinum

Growth (during cooling)

Net growth

Probability of growth (time to turbidity) Heat inactivation

Pmax, κ, τ

Non-proteolytic Clostridium botulinum (spores) Non-proteolytic Clostridium botulinum (spores – types E & F) and aerobic competitive flora Clostridium perfringens (vegetative cells)

Probability of growth (time to turbidity) Length of lag phase

Pmax, κ, τ

Temperature (5–42 °C); pH (5.3–7.3); NaCl (0.5–4.5%w/v); NaNO2 (0–150 ppm) Temperature (5–30 °C); pH (5.3–7.3); NaCl (0.5–3.5%w/v); NaNO2 (0–150 ppm) Temperature (5–42 °C); pH (4.5–7.5); NaCl (0.5–5.0%w/v); NaNO2 (0–150 ppm) Temperature (10–42 °C); pH (5.0–9.0); NaCl (0.5–5.0%w/v); NaNO2 (0–150 ppm) Time–temperature history (up to 50 data pairs) Temperature (15–34 °C); pH (5.0–7.2); NaCl (0.0–4.0%w/v) Temperature (70–90 °C); pH (5.0–7.0); NaCl (0.0–3.0%w/v); sodium pyrophosphate (0.0–0.3%w/v) Temperature (5–28 °C); pH (5.0–7.0); NaCl (0.0–4.0%w/v) Temperature (4–30 °C)

Growth (anaerobic)

GT, lag, time to n-log increase

Clostridium perfringens (spores)

Growth (during cooling)

Net growth (in beef broth)

Clostridium perfringens (spores)

Growth (during cooling)

Net growth (in cured beef)

Clostridium perfringens (spores)

Growth (during cooling)

Net growth (in cured chicken)

Non-proteolytic Clostridium botulinum (spores)

© 2008, Woodhead Publishing Limited

Time to n-log reduction

Lag (shelf-life of fresh fish in modified atmospheres)

Temperature (19–37 °C); pH (6.0–6.5); NaCl (1.0–3.0%w/v); sodium pyrophosphate (0.1–0.3%w/v) Time–temperature history (up to 50 data pairs) Time–temperature history (up to 50 data pairs) Time–temperature history (up to 50 data pairs)

501

Physiological event

Predicting the behaviour of micro-organisms in chilled foods

Micro-organism

502

Table 17.1 continued Physiological event

Response variables(a)

Independent variables and ranges

Escherichia coli O157:H7

Growth (aerobic)

GT, lag, time to n-log increase

Escherichia coli O157:H7

Growth (anaerobic)

GT, lag, time to n-log increase

Escherichia coli O157:H7

Heat inactivation

Time to n-log reduction

Escherichia coli O157:H7

Survival

Time to n-log reduction

Escherichia coli O157:H7

Net reduction (in beef tartar)

Listeria monocytogenes

Inactivation by gamma-irradiation Growth (aerobic)

GT, lag, time to n-log increase

Listeria monocytogenes

Growth (anaerobic)

GT, lag, time to n-log increase

Listeria monocytogenes

Heat inactivation

Time to n-log reduction

Listeria monocytogenes

Survival

Time to n-log reduction

Salmonella spp.

Growth (aerobic)

GT, lag, time to n-log increase

Salmonella spp.

Survival

Time to n-log reduction

Salmonella Typhimurium

Growth – previous NaCl (aerobic)

GR, lag, time to n-log increase (in sterile chicken breast)

Temperature (5–42 °C); pH (4.5–8.5); NaCl (0.5–5.0%w/v); NaNO2 (0–150 ppm) Temperature (5–42 °C); pH (4.5–8.5); NaCl (0.5–5.0%w/v); NaNO2 (0–150 ppm) Temperature (55–62.5 °C); pH (4.0–8.0); NaCl (0.0–6.0%w/v); sodium pyrophosphate (0.0–0.3%w/v) Temperature (4–37 °C); pH (3.5–7.0); NaCl (0.5–15.0%w/v); NaNO2 (0–75 ppm); lactic acid (0.0–2.0%w/w) Temperature (–20–10 °C); irradiation dose (0–2 kGy) Temperature (4–37 °C); pH (4.5–8.0); NaCl (0.5–5.0%w/v) or aw (0.97–0.997); NaNO2 (0–150 ppm) Temperature (4–37 °C); pH (4.5–8.0); NaCl (0.5–5.0%w/v) or aw (0.97–0.997); NaNO2 (0–150 ppm) Temperature (55–65 °C); pH (4.0–8.0); NaCl (0.0–6.0%w/v); sodium pyrophosphate (0.0–0.3%w/v) Temperature (4–42 °C); pH (3.2–7.3); NaCl (0.5–19.0%w/v); NaNO2 (0–150 ppm); lactic acid (0.0–2.0%w/w) Temperature (10–30 °C); pH (5.6–6.8); NaCl (0.5–4.5%w/v) Temperature (5–42 °C); pH (3.5–7.2); NaCl (0.5–16.0%w/v); NaNO2 (0–200 ppm) Temperature (10–40 °C); previous growth NaCl (0.5–4.5%w/v)

© 2008, Woodhead Publishing Limited

Chilled foods

Micro-organism

Growth – previous temperature (aerobic)

GR, lag, time to n-log increase (in sterile chicken breast)

Salmonella Typhimurium

Growth – previous pH (aerobic) Inactivation by gamma-irradiation

GR, lag, time to n-log increase

Salmonella Typhimurium Salmonella Typhimurium

Shigella flexneri

Inactivation by gamma-irradiation Inactivation by gamma-irradiation Growth (aerobic)

Shigella flexneri

Growth (anaerobic)

Staphylococcus aureus

Growth (aerobic)

Staphylococcus aureus

Growth (anaerobic)

Staphylococcus aureus

Survival

Yersinia enterocolitica

Growth (aerobic)

Spoilage (normal flora)

Inactivation by gamma-irradiation

Salmonella Typhimurium

Temperature (16–34 °C); previous growth temperature (16–34 °C)

Temperature (15–40 °C); pH (5.2–7.4); previous growth pH (5.7–8.6) Net reduction Temperature (–20–10 °C); irradiation dose (in sterile chicken MDM) (0–3.6 kGy); two models (air and vacuum atmospheres) Net reduction Temperature (–20–10 °C); irradiation dose (in non-sterile chicken MDM) (0–3.6 kGy) Net reduction (in non-sterile Temperature (–20–10 °C); irradiation dose chicken legs) (0–3.6 kGy) GT, lag, time to n-log increase Temperature (10–37 °C); pH (5.0–7.5); NaCl (0.5–5.0%w/v); NaNO2 (0–150 ppm) GT, lag, time to n-log increase Temperature (12–37 °C); pH (5.5–7.5); NaCl (0.5–4.0%w/v); NaNO2 (0–150 ppm) GT, lag, time to n-log increase Temperature (10–42 °C); pH (4.5–9.0); NaCl (0.5–12.5%w/v); NaNO2 (0–150 ppm) GT, lag, time to n-log increase Temperature (12–42 °C); pH (5.3–9.0); NaCl (0.5–16.5%w/v); NaNO2 (0–150 ppm) Time to n-log reduction Temperature (4–37 °C); pH (3.0–7.0); NaCl (0.5–20.0%w/v); NaNO2 (0–200 ppm); lactic acid (0.0–1.0%w/w) GT, lag, time to n-log increase Temperature (5–42 °C); pH (4.5–8.5); NaCl (0.5–5.0%w/v); NaNO2 (0–150 ppm) Net reduction Temperature (–20–10 °C); irradiation dose (in non-sterile chicken MDM) (0–3.6 kGy)

© 2008, Woodhead Publishing Limited

503

(a) GT = generation time; GR = growth rate; Pmax = maximum probability of growth; κ = growth rate of individual samples showing growth; τ = time to ½Pmax (inflection point); MDM = mechanically deboned meat; where there is no indication of a specific food matrix or medium in brackets it means the model was developed in laboratory media (broth).

Predicting the behaviour of micro-organisms in chilled foods

Salmonella Typhimurium

504

Chilled foods

of the primary models in Growth Predictor are based on the model developed by Baranyi and Roberts (1994). This requires the user to clearly understand the concept of initial physiological state (α0) of the cells or spores, which will affect the predictions of lag phase duration. Along with Growth Predictor, users can download Perfringens Predictor, which is a Microsoft® Excel-based model (add-in) for the prediction of growth of C. perfringens during the cooling of meat products that were cooked to maximum temperatures of 70 °C to 95 °C. It is valid only for uncured meats and the sole inputs are time and temperature (dynamic profile). Validation data and the data used to build the models included in Perfringens Predictor are also available through ComBase. ComBase Modelling Toolbox All the models from Growth Predictor as well as Perfringens Predictor are now available as an online tool as part of the ComBase (Combined Database for Predictive Microbiology) initiative (http://www.combase.cc). This initiative is a collaboration between the Food Standards Agency (UK), the Institute of Food Research (UK), the USDA Agricultural Research Service, and the Australian Food Safety Centre of Excellence (Baranyi and Tamplin, 2004). ComBase is a large database of microbiological data, available to the public and accessible via the internet. The database is searchable so that users can look for specific growth or death data for different organisms and conditions. Amongst several different applications, ComBase can be used by industry, academia and government to validate newly created models, to compare the predictions of established models developed in a specific medium (e.g. broth) against data recorded in other matrices, or to identify boundary growth values of specific organisms in a variety of foods. Within ComBase, users can find the ComBase Modelling Toolbox (http:// www.combase.cc/toolbox.html) which includes three main components: (i) ComBase Predictor, (ii) Perfringens Predictor, and (iii) DMFit. The online ComBase Predictor is a modified and expanded version of the stand-alone Growth Predictor. It has a set of 20 growth models, 7 thermal death models and 2 non-thermal survival models (Table 17.2). Furthermore, it includes several user-friendly improved features over the original stand-alone Growth Predictor. For instance, ComBase Predictor includes dynamic temperature versions of all the models from Growth Predictor. In this case, the user simply enters a time–temperature profile (e.g. from a temperature data-logger) into the main interface of ComBase Predictor, and the predictions are presented both in graphic and tabulated form. The predictions can be easily copied into any spreadsheet software for further analyses. Another improvement over the Growth Predictor is the fact that ComBase Predictor can simultaneously produce predictions for up to four micro-organisms, thereby facilitating comparisons amongst several scenarios. The online version of Perfringens Predictor (available as part of the ComBase Modelling Toolbox) is a user-friendly expanded version of the stand-alone Perfringens Predictor. Its main purpose continues to be the prediction of the response of C. perfringens during the cooling process of cooked meats. However,

© 2008, Woodhead Publishing Limited

Table 17.2 Summary of micro-organisms and models included in the ComBase Predictor (all growth models also included in stand-alone Growth Predictor) Physiological event

Response variables(a)

Independent variables and ranges

Aeromonas hydrophila

Growth

µ max, tD

Bacillus cereus

Growth

µ max, tD

Bacillus cereus (spores)

Heat inactivation

kmax, D-value

Bacillus licheniformis

Growth

µ max, tD

Bacillus subtilis

Growth

µ max, tD

Non-proteolytic Clostridium botulinum Growth

µ max, tD

Non-proteolytic Clostridium botulinum Heat inactivation (spores, recovered in presence of lysozyme) Proteolytic Clostridium botulinum Growth

kmax, D-value

Temperature (2–37 °C); pH (4.6–7.5); NaCl (0.0–4.5%) or aw (0.974–1) Temperature (5–34 °C); pH (4.9–7.4); NaCl (0.0–9.4%) or aw (0.94–1); CO2 (0–60%) Temperature (90–100 °C); pH (4.5–7.0); `NaCl (2.5–7.5%) or aw (0.954–0.986) Temperature (13–34 °C); pH (4.0–7.6); NaCl (0.0–13.5%) or aw (0.907–1) Temperature (10–34 °C); pH (4.3–7.8); NaCl (0.0–10.3%) or aw (0.933–1) Temperature (4–30 °C); pH (5.1–7.5); NaCl (0.0–4.5%) or aw (0.974–1) Temperature (80–95 °C); pH (4.1–7.3); NaCl (0.0–5.0%) or aw (0.971–1)

Clostridium perfringens

Growth

µ max, tD

Escherichia coli

Growth

µ max, tD

µ max, tD

Temperature (14–40 °C); pH (4.7–7.2); NaCl (0.0–7.5%) or aw (0.954–1) Temperature (15–52 °C); pH (5.0–8.0); NaCl (0.0–5.0%) or aw (0.971–1) Temperature (10–42 °C); pH (4.5–7.5); NaCl (0.0–6.5%) or aw (0.961–1); CO2 (0–100%)

Predicting the behaviour of micro-organisms in chilled foods

Micro-organism

505

© 2008, Woodhead Publishing Limited

506

continued

Micro-organism

Physiological event

Response variables(a)

Independent variables and ranges

Escherichia coli

Heat inactivation

kmax, D-value

Listeria monocytogenes

Growth

µ max, tD

Listeria monocytogenes/innocua

Growth

µ max, tD

Listeria monocytogenes

Growth

µ max, tD

Listeria monocytogenes

Growth

µ max, tD

Listeria monocytogenes

Heat inactivation

kmax, D-value

Listeria monocytogenes

Survival (non-thermal)

kmax, D-value

Staphylococcus aureus

Growth

µ max, tD

Salmonella

Growth

µ max, tD

Temperature (54.5–64.5 °C); pH (4.2–8.0); NaCl (0.0–8.4%) or aw (0.986–1) Temperature (1–40 °C); pH (4.4–7.5); NaCl (0.0–10.2%) or aw (0.934–1); CO2 (0–100%) Temperature (1–40 °C); pH (4.4–7.5); NaCl (0.0–11.4%) or aw (0.924–1); NaNO2 (0–200 ppm) Temperature (1–40 °C); pH (4.4–7.5); NaCl (0.0–11.4%) or aw (0.924–1); lactic acid (0–20 000 ppm) Temperature (1–40 °C); pH (4.4–7.5); NaCl (0.0–11.4%) or aw (0.924–1); acetic acid (0–10 000 ppm) Temperature (60–68 °C); pH (4.2–7.0); NaCl (0.0–9.0%) or aw (0.943–1) Temperature (0–20 °C); pH (3.5–7.0); NaCl (0.0–25.0%) or aw (0.793–1) Temperature (7.5–30 °C); pH (4.4–7.1); NaCl (0.0–13.5%) or aw (0.907–1) Temperature (7–40 °C); pH (3.9–7.4); NaCl (0.0–4.6%) or aw (0.973–1); CO2 (0–100%)

© 2008, Woodhead Publishing Limited

Chilled foods

Table 17.2

Growth

µ max, tD

Salmonella

Heat inactivation

kmax, D-value

Salmonella

Survival (non-thermal)

kmax, D-value

Shigella flexneri

Growth

µ max, tD

Yersinia enterocolitica

Growth

µ max, tD

Yersinia enterocolitica

Growth

µ max, tD

Yersinia enterocolitica

Heat inactivation

kmax, D-value

Brochothrix thermosphacta

Growth

µ max, tD

Brochothrix thermosphacta

Heat inactivation

kmax, D-value

Pseudomonas spp.

Growth

µ max, tD

(a)

Temperature (7–40 °C); pH (3.9–7.4); NaCl (0.0–4.6%) or aw (0.973–1); NaNO2 (0–200 ppm) Temperature (54.5–65 °C); pH (4.0–7.1); NaCl (0.0–0.6%) or aw (0.997–1) Temperature (0–40 °C); pH (4.3–7.5); NaCl (0.0–26.0%) or aw (0.781–1) Temperature (15–37 °C); pH (5.5–7.5); NaCl (0.0–5.0%) or aw (0.971–1); NaNO2 (0–1000 ppm) Temperature (–1–37 °C); pH (4.4–7.2); NaCl (0.0–7.0%) or aw (0.957–1); CO2 (0–80%) Temperature (–1–37 °C); pH (4.4–7.2); NaCl (0.0–7.0%) or aw (0.957–1); lactic acid (0–10 000 ppm) Temperature (52–60 °C); pH (4.2–7.0); NaCl (0.0–6.5%) or aw (0.961–1) Temperature (0–30 °C); pH (5.5–7.0); NaCl (0.0–8.0%) or aw (0.95–1) Temperature (40–55 °C); pH (5.0–7.0); NaCl (0.0–2.0%) or aw (0.989–1) Temperature (0–20 °C); pH (5.0–7.4); NaCl (0.0–6.5%) or aw (0.961–1)

µ max = maximum growth rate (log10 concentration/h); tD = doubling time; kmax = maximum inactivation rate (log10 concentration/h); D-value = decimal reduction time.

Predicting the behaviour of micro-organisms in chilled foods

Salmonella

507

© 2008, Woodhead Publishing Limited

508

Chilled foods

in the online version, the users can also input the pH of the meat (5.2–8.0), the concentration of salt (0–4%), and indicate whether or not the meat is cured. The cured meat option should be used only if the initial concentration of sodium nitrite is ≥ 100 ppm and the residual concentration is ≥ 10 ppm. DMFit is a tool developed by the Institute of Food Research that allows the user to fit bacterial (growth or survival) curves (i.e. logarithmic cell counts vs time) where a linear phase is preceded and followed by a stationary phase to different primary models, including: (i) the model of Baranyi and Roberts (with three options: full model, model without lag, and model without asymptotic stationary phase), (ii) tri-linear model, (iii) bi-phasic models (with two options: no lag, and no asymptotic stationary phase), and (iv) linear model. The tool displays the fitted parameters for the selected model (maximum growth/death rate, lag time (or shoulder), initial cell count, final cell count, and estimates standard errors on these parameters) as well as the evaluation of fit (adjusted-R2, and standard error of fit). The models included in ComBase Predictor were developed using DMFit. Sym’Previus The Sym’Previus software (http://www.symprevius.org/) is a decision support system developed by the Institut National de Recherche Agronomique (INRA) in France (Leporq et al., 2005). The software includes a database with growth responses of micro-organisms in foods and predictive models for growth, inactivation, survival and growth/no-growth boundaries of various pathogenic bacteria (Table 17.3). The software also includes features for evaluation of heat penetration data in thermally processed foods by allowing the user to load the time-temperature information and providing estimations of the well-known parameters jh (heating lag) and fh (heating rate). The main interface is available in French and in English. The software is available on a commercial basis through contact centres as indicated on the homepage cited above. The software is being modified to include increased functionality incorporating a probabilistic tool to take into account prevalence in food, lag time estimation for stressed and low inocula, reference growth curves in food (challenge tests) and biological variability (multiple strains per species). Seafood Spoilage and Safety Predictor The Seafood Spoilage and Safety Predictor (SSSP) software (http://www.dfu.min. dk/micro/sssp/) was developed by the Danish Institute for Fisheries Research (DIFRES), and it is a considerably expanded version of the Seafood Spoilage Predictor (SSP) software developed in 1999 (Dalgaard et al., 2002). The original SSP included only four seafood spoilage models (one growth model for S. putrefaciens, one growth model for P. phosphoreum, and two product-specific RRS models for shelf-life). The new SSSP v. 2.0 includes: (i) four product-specific RRS models, (ii) three generic RRS models, (iii) four product-specific microbial spoilage models, (iv) a generic model to predict microbial growth and shelf-life, (v) modules to compare predictions from SSSP with users own data of shelf-life or growth of bacteria, and (vi) a model to predict the simultaneous growth of © 2008, Woodhead Publishing Limited

Table 17.3 Summary of micro-organisms and models included in Sym'Previus Physiological event

Response variables(a)

Independent variables

Bacillus cereus Escherichia coli Listeria monocytogenes Salmonella Bacillus cereus Clostridium perfringens Escherichia coli Listeria monocytogenes Salmonella Typhimurium Staphylococcus aureus Bacillus cereus Proteolytic Clostridium botulinum Non-proteolytic Clostridium botulinum Clostridium perfringens Escherichia coli Listeria monocytogenes Salmonella Staphylococcus aureus

Growth Growth Growth Growth Heat inactivation Heat inactivation Heat inactivation Heat inactivation Heat inactivation Heat inactivation Growth/no-growth interface Growth/no-growth interface Growth/no-growth interface Growth/no-growth interface Growth/no-growth interface Growth/no-growth interface Growth/no-growth interface Growth/no-growth interface

µ max µ max µ max µ max D-value D-value D-value D-value D-value D-value Growth boundary Growth boundary Growth boundary Growth boundary Growth boundary Growth boundary Growth boundary Growth boundary

Temperature, pH, aw Temperature, pH, aw Temperature, pH, aw Temperature, pH, aw Temperature, pH, aw Temperature, pH, aw Temperature, pH, aw Temperature, pH, aw Temperature, pH, aw Temperature, pH, aw Temperature, pH, aw Temperature, pH, aw Temperature, pH, aw Temperature, pH, aw Temperature, pH, aw Temperature, pH, aw Temperature, pH, aw Temperature, pH, aw

(a)

µ max = maximum growth rate (log10 concentration/h); D-value = decimal reduction time.

Predicting the behaviour of micro-organisms in chilled foods

Micro-organism

509

© 2008, Woodhead Publishing Limited

510

Chilled foods

L. monocytogenes and spoilage bacteria in cold-smoked salmon. The SSSP software is available free of charge as a download file from the homepage cited above. Opti.Form® Listeria Suppression Model, Opti.Form® Listeria Control Model, and Opti.Form® Listeria Control Model 2007 These three models are available from Purac; information about acquiring a copy of these models is available at http://www.purac.com/meatandpoultry/optiform. html. The Listeria Control Model is a kinetic model developed by Seman et al. (2002) that predicts the growth of L. monocytogenes at 4.4 °C as influenced by concentrations of salt, sodium or potassium lactate, sodium diacetate and moisture. It allows the user to calculate the number of days before the population of L. monocytogenes increases by 2 log. The user can manipulate concentrations of lactate and/or diacetate to restrict the growth of the pathogen to no more than 2 log over the product shelf-life. The Listeria Suppression Model is a growth boundary model developed by Legan et al. (2004) based on the data used for development of the Listeria Control Model. This model is aligned with the USDA-FSIS Interim Final Rule for control of L. monocytogenes for ready-to-eat meat products in the United States (USDA-FSIS, 2003a). This Rule gives processors three alternatives for control of this pathogen affording different regulatory treatment depending on the alternative chosen by the processor. Under alternatives 1 and 2, processors can use antimicrobial compounds, such as lactate and diacetate, for control of L. monocytogenes. The Listeria Suppression Model is able to calculate the level of lactate and/or diacetate needed to prevent growth over the product shelf-life (where growth is defined as a 1-log increase in the population of L. monocytogenes). Both the Listeria Control Model and the Listeria Suppression Model are available on a free CD directly from Purac. The Listeria Control Model 2007 is a new version of the original kinetic growth model, which includes temperature and pH as independent variables, in addition to concentration of salt, lactate, diacetate and moisture (already included in the original model). The model is valid for both cured and uncured meats. The model output consists of 90% and 95% confidence intervals for the predicted growth curve. The output data can be easily exported to a spreadsheet file for further analyses. This model is available as a download from the Purac website after first contacting a Purac sales office to request a user-name and password to a secure webpage. Safety Monitoring and Assurance System The Safety Monitoring and Assurance System (SMAS) is a tool developed by a group of predictive microbiology researchers in Greece (Koutsoumanis et al., 2005) for management of chilled supply chain conditions. SMAS integrates kinetic models for foodborne pathogens, time-varying temperature conditions (simulating realistic supply chain environments) and variation of intrinsic parameters such as pH and aw. The system has been successfully evaluated to estimate microbial safety (i.e. concentration of L. monocytogenes) and stability (concentration of L. sake) of cooked ham at the point of consumption including © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

511

temperature data from distribution (local and export market), retail storage, retail display and domestic refrigeration storage conditions. Food Spoilage Predictor© This program contains a model for the effect of temperature and aw on the growth of psychrotolerant pseudomonads (Neumeyer et al., 1997a). The Food Spoilage Predictor© (FSP) includes facilities to read product temperature profiles, as recorded by data loggers, and thus predict the effect of fluctuating temperatures on growth of the organism. The pseudomonads growth model included in the FSP has been validated in seafood, milk and meat products (Neumeyer et al., 1997b). The FSP is commercially available from data logger manufacturers (e.g. Hastings Data Loggers). Bugdeath Bugdeath 1.0 is a software package developed as part of a project commissioned by the European Union, involving researchers from eight partner institutions from five different countries (Gil et al., 2006; James and Evans, 2006). The software integrates heat transfer equations with microbial inactivation models under constant and time-varying temperature conditions. The program predicts food surface temperature and reduction of microbial population during dry and wet surface pasteurisation treatments. It also includes a database of thermal properties of foods and inactivation kinetic parameters for Salmonella and L. monocytogenes.

17.3.4 Knowledge and model gaps Although many predictive models are available in the public domain, the majority of those are still available only via scientific publications. This poses a considerable barrier for wider application of models that are potentially very useful for the food industry. Although in some cases the implementation of published models in a spreadsheet format by potential users is possible, this can sometimes be limited by the lack of data (used for model development and/or validation) or other necessary modelling details, which are typically not included in the publication due to space limitations. Moreover, small processors may not have sufficient and/or qualified resources to use published models in an effective way. Also, in some cases, the model itself may indeed require software tools more sophisticated than a spreadsheet format in order to be implemented effectively. The development of databases that are available through the Internet, such as ComBase, is an important step forward regarding availability of data. The submission of raw data to such databases anytime a researcher develops a new model that will go into the public domain as a scientific publication should be seen as a good modelling practice. Moreover, details about conditions of the experiment, strains, growth or inactivation media, etc. used in the development of a model should also be reported clearly in order to prevent improper use of the data. Similarly, the presentation of experimental results used to develop a predictive model should, as much as possible, include information about sources of variability and the uncertainty of the results obtained. © 2008, Woodhead Publishing Limited

512

Chilled foods

Development and validation of models in real food matrices continues to be another important gap that needs to be filled. Many of the models available in the public domain have been developed in liquid laboratory media (broth), which, oftentimes, supports the growth of micro-organisms better than actual food matrices. Hence, many of these models are perceived by the food industry as very conservative, to the point that they have limited practical application. Although many efforts have been made in recent years to generate microbiological data in food matrices for model development and validation, there is still a need for more models developed in food materials for a wider range of applications. Effective integration of temperature data under real chilled supply chain conditions with predictive models represents another important gap. Temperature conditions in the chilled supply chain vary amongst different geographical regions and along different stages of the cold chain. As mentioned earlier, temperature data from the various stages of the chilled supply chain are available and have been referred to in some studies (Goldwin et al., 2007; Hong et al., 2006; Membré et al., 2005; Nauta et al., 2003; Rosset et al., 2004). However, it is still common practice to use single temperature values, simulating the worst-case temperature conditions during the supply chain, as inputs to deterministic predictive models for estimation of chilled shelf-life. This limits the design of chilled foods, as predictions obtained in this way are overly conservative. Instead, the definition of temperatures as probability distributions during the different stages of the chilled supply chain (e.g. storage by manufacturer, transport and distribution to retail warehouse, retail display, transport to and storage in consumer fridges) would allow for a more accurate and realistic estimation of shelf-life, if these temperature distributions were integrated effectively with predictive microbiology models. To that extent, more studies dealing with probabilistic models and associated implementation tools are necessary in order to account for the variation of supply chain data. Moreover, the ability to collect temperature data in real time and to use this information in determining the shelf-life of chilled foods remains an important challenge. In order to overcome this limitation, the use of radio frequency identification (RFID) tools (tags and readers) has gained considerable interest in recent years, not only for traceability of products through the supply chain, but also as a tool to collect product information (e.g. temperature) in real-time. Other tools such as time–temperature integrators (TTIs) for monitoring temperature history and shelf-life through the chilled supply chain have been implemented in the food industry for many years (Taoukis and Labuza, 1989). The combination of tools such as TTIs, RFID tags and readers, and quantitative microbiology models will allow food processors to bridge the gap between product traceability and management of food safety and stability.

17.4 Modelling of heating and cooling processes Modelling of heating and cooling processes in the food industry represents a very large area of knowledge and research, and could be the subject of a separate © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

513

textbook. In this section, we attempt to highlight the main modelling principles relating to the design of chilled foods, and the integration of thermo-physical models for heating and cooling of foods with predictive microbiology models.

17.4.1 Modelling principles When modelling heating and cooling processes of food materials, all the relevant dependent variables of interest involved in the fundamental heat, mass and momentum conservation equations seem to obey a generalised conservation principle. If the dependent variable is denoted by the Greek letter φ, the general differential equation can be written as:

∂ ( ρφ ) + ∇ ⋅ ( ρ vφ ) = ∇ ⋅ (Γ∇φ ) + S [17.1] ∂t where t is time, ρ is the density of the material, v is the velocity vector, Γ is the diffusion coefficient, and S is the source term. The quantities Γ and S are specific to a particular meaning of φ. On the left-hand side of Eq. 17.1, the first term is the unsteady or transient term, and the second is the convection term. On the righthand side, the first term is the diffusion term and the second is the source term. The dependent variable φ can stand for a variety of different quantities, such as the temperature or the enthalpy of the material, a velocity component, the mass fraction of a chemical species (e.g. water), or the turbulence kinetic energy. Accordingly, for each of these variables, an appropriate meaning will have to be assigned to the diffusion coefficient (Γ) and the source term (S). In some cases (as will be discussed later), the diffusion fluxes are not governed by the gradient of the relevant variable. Therefore, the use of the diffusion term (Γ∇φ) does not limit the general equation to gradient-driven diffusion processes. In such cases, an appropriate diffusion quantity can be expressed as part of the source term (S). The equations presented below for particular modelling scenarios dealing with heating and cooling processes of food materials are specific forms of Eq. 17.1. In the case of heating and cooling processes for solid foods, the principles are based on heat and moisture exchanges between a solid food body and the (heating or cooling) medium surrounding it. Heat transfer through solid foods is governed by the fundamental law of heat conduction, commonly known as Fourier’s law. The temperature profile within a solid food can then be determined by the solution of the energy equation, which can be written as:

ρc

∂T = ∇ ⋅ ( k ∇T ) + Q''' ∂t

[17.2]

where c and k are the specific heat and thermal conductivity of the food material, T is temperature, and Q''' represents the internal energy generation (typically zero, except for advanced or volumetric heating systems such as microwave or ohmic heating). The other terms have been defined before (see Eq. 17.1). The three food properties present in Eq. 17.2 (i.e. density, specific heat and thermal conductivity) can be combined into a single property called thermal diffusivity which is normally © 2008, Woodhead Publishing Limited

514

Chilled foods

represented by the Greek letter α. The thermal diffusivity of food materials can vary with temperature as well as location within the domain (i.e. spatial distribution). A common approach to consider the effect of temperature and/or product composition on the value of α is to use empirical correlations such as those reported by Choi and Okos (1986). Eq. 17.2 describes the diffusion of energy within the solid food material; however, the heat exchange between the surroundings and the surface of the solid food material depends on the conditions at the boundaries. There are three kinds of boundary conditions for the solution of Eq. 17.2: (i) a prescribed temperature, (ii) a prescribed heat flux, and (iii) a convection flux (via a heat transfer coefficient and the temperature of the surrounding medium). The second term on the right-hand side of Eq. 17.2, representing the internal energy generation rate per unit volume, needs to be considered when modelling advanced or volumetric heating systems. For instance, in the case of ohmic or electrical resistance heating, the internal volumetric heat generation can be described by the following equation:

Q ''' = σ ∇V

2

[Eq. 17.3]

where Q''' is the heat generated by the passage of current at any point in the food, σ is the local electrical conductivity, and ∇V is the voltage gradient. The critical property affecting energy generation is the electrical conductivity (σ), which is a function of both temperature and position (see for instance De Alwis and Fryer, 1990; Palaniappan and Sastry, 1991). The volumetric energy generation rate (Eq. 17.3) depends on the voltage distribution (electrical field strength) within the ohmic heater. The voltage field is determined by the solution of Laplace’s equation (for steady-state case, which is typical) for current:

∇ ⋅ (σ∇V ) = 0

[Eq. 17.4]

In the case of microwave heating systems, the energy generation rate per unit volume (i.e. heat generated by the power deposition of the microwaves), Q''', is related to the electrical field strength, E, and can be described by the following equation:

Q''' = 2π f ε 0ε '' E

2

[Eq. 17.5]

where f is the frequency of microwaves, ε0 is the permittivity of free space (8.8542 × 10–12 Farad/m), ε'' is the dielectric loss factor of the food material (which incorporates all of the energy losses due to dielectric relaxation and ionic conduction), and E is the electric field intensity. In situations where the volumetric heating rate, Q''', is included in the energy equation under transient (i.e. unsteady-state) conditions, the value of E is taken as a time-average value. This can be justified since the variations in electric field intensity are very fast (i.e. of the order of 10–10 s) compared with the typical time scales for thermal diffusion and convection (Datta, 2001). As for mass transfer modelling, the transfer of moisture within a solid food material is generally described by Fick’s law of diffusion, which can be expressed as: © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

∂X w = ∇ ⋅ ( D∇X w ) ∂t

515

[Eq. 17.6]

where Xw is the mass fraction of water and D is the diffusion coefficient. As with the thermo-physical properties of Eq. 17.2, the diffusion coefficient depends on temperature and concentration of solutes in the food material. Temperature dependency is typically modelled using equations based on the Arrhenius law. In some instances where the mechanism of internal moisture transport is not driven by concentration, the validity of Eq. 17.6 does not hold. For example, in microwave heating of solid foods, the amount of water evaporated (or sublimated from ice) varies according to the increased thermal energy absorbed from microwaves. This internal evaporation (or sublimation) can generate very high pressures, depending on the resistance of the food material to the transport of water. This results in pressure-driven (Darcy) flow, which cannot be described by Eq. 17.6. Different approaches to model moisture transport during microwave heating have been studied and are beyond the scope of this chapter (see for instance Ni et al., 1999). Modelling heating and cooling processes of liquid foods is more complex due to the fluid motion. Temperature and flow fields need to be coupled; hence, the principles of conservation of mass, momentum and energy in the fluid need to be considered together. The continuity equation and Navier–Stokes equations are used to describe flow field. The energy equation is used to solve for the temperature field. The actual conditions imposed by the processing equipment are considered as the boundary conditions of the governing equations. Liquid foods are relatively incompressible, so the density (ρ) is essentially constant. Under those conditions the following basic equations are used to describe the flow field: Mass conservation equation (continuity equation)

∇⋅v = 0

[Eq. 17.7]

Momentum (Navier–Stokes equations, three equations in mutually perpendicular directions)

 ∂v  ρ  + v ⋅∇v  = −∇p + µ∇ 2 v + F  ∂t 

[Eq. 17.8]

In Equations 17.7 and 17.8, p is pressure, µ is apparent dynamic viscosity, F is the vector of body forces, and the other terms have been previously defined. The lefthand side of Eq. 17.8 represents the transient and inertial terms. On the right-hand side of this equation, the first term is due to pressure gradients, the second to viscous forces and the third to body forces. As previously explained, the temperature field needs to be solved simultaneously using the energy equation, which can be written as follows:

 ∂T  ρc  + v ⋅∇T  = ∇ ⋅ ( k ∇T ) + Q ''' − p∇ ⋅ v + Φ  ∂t 

[Eq. 17.9]

In Eq. 17.9, the left-hand side represents the transient and convective terms. On the © 2008, Woodhead Publishing Limited

516

Chilled foods

right-hand side of this equation, the first term is due to heat conduction (diffusion) through the fluid, the second term is due to internal energy generation (as illustrated previously for the case of volumetric heating systems), the third term is due to mechanical work, and the fourth term represents viscous dissipation (typically negligible, except for some systems such as scrape-surface heat exchangers due to normal and shear stresses).

17.4.2 Modelling techniques and implementation To solve the equations presented in the previous section, numerical methods are frequently used. Analytical solutions are limited to very simple cases, and are oftentimes of little practical value. Instead, numerical analyses have proven to be very effective in solving complex heating and cooling problems in the food industry. Techniques such as the finite element (FE) and the finite difference (FD) methods are commonly used, particularly for the solution of heating and cooling problems of solid foods. In this case, FE and FD techniques allow for the solution of the energy equation for complex scenarios involving unsteady-state heat transfer, consideration of non-isotropic thermo-physical properties, handling of irregularly shaped foods and time-dependent boundary conditions, and nonuniform initial temperature conditions within the model domain (see for instance Amézquita et al., 2005a; Wang and Sun, 2002). Moreover, FE and FD methods have been successfully applied in modelling heat and mass transfer in volumetric heating systems such as microwave and ohmic heating processes. In this case, the solution of Maxwell’s equations for electromagnetics are coupled with the diffusion equations for heat and mass transfer (De Alwis and Fryer, 1990; Geedipalli et al., 2007; Zhang and Datta, 2000), or alternatively, heat and mass transfer are solved numerically assuming a source term with exponential decay (Lambert’s law) instead of solving the Maxwell’s equations for the electromagnetic field (Campañone and Zaritzky, 2005; Ni et al., 1999). In recent years, the finite control volume method (Patankar, 1980) has gained more interest over FE and FD methods in the solution of coupled flow and energy problems and has become the main computational scheme in commercially available computational fluid dynamics (CFD) software packages. In the past, numerical models for heat, mass and momentum transfer were perceived by the food industry as purely academic exercises with little potential for applicability to real processing scenarios. With the considerable advancement of computational power in recent years, commercially available numerical software packages, especially those with CFD capabilities, have become more popular for food modelling applications in industrial settings. Some commercial CFD solvers commonly used by the food industry include (all websites accessed in June 2007):

• • • •

Fluent® (http://www.fluent.com) ANSYS® CFX® (http://www.ansys.com/cfx) Star-CD (http://www.cd-adapco.com) PHOENICS (http://www.cham.co.uk)

© 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

517

• Flow-3D (http://www.flow3d.com) Other popular software packages based on the FE method used to solve a variety of multiphysics problems in food processing applications include (all websites accessed in June 2007):

• Comsol Multiphysics® (http://www.comsol.com) • ANSYS® MultiphysicsTM (http://www.ansys.com/products/multiphysics.asp) • FIDAP (http://www.fluent.com) Each of the software packages listed above offer different capabilities for different processing scenarios. The choice of any of these packages depends on the balance of licence cost and functionality for the desired application. The potential user must have sufficient knowledge about the fundamental modelling principles and the actual conditions of the process or processes to be modelled in order to make an informed decision. Regarding chilled foods, a few practical examples where CFD solvers have been successfully applied include the optimisation of performance of chilled retail display cabinets (Foster et al., 2005), study of distribution of temperature and air flow in domestic refrigerators (Laguerre et al., 2007), study of airflow patterns and temperature distribution in refrigerated trucks (Moureh and Flick, 2004), calculation of air velocity patterns, heat and mass transfer of industrial cold storage rooms (Nahor et al., 2005), and prediction of heat and moisture transfer during the cooling of beef carcasses (Trujillo and Tuan Pham, 2006). Comprehensive literature reviews of numerical simulation and modelling techniques for the solution of heating and cooling problems in the food industry have been recently reported (Norton and Sun, 2006; Smale et al., 2006; Wang and Sun, 2003).

17.4.3 Modelling of heating and cooling processes to support validation Validation of thermal processes for chilled foods poses a challenge for food manufacturers, especially with respect to inaccessible locations within food products and processing equipment where direct temperature measurements are not feasible (both for in-pack and continuous flow conditions). Validation of such processes is an integral part of safe-by-design and quality-by-design principles. When temperature recording devices cannot be used, food processors must use alternative methods for validation. The most common one is the microbiological method, in which cells or spores of a surrogate micro-organism, with similar temperature-induced death kinetics to the target pathogen, are used either in inoculated package studies (Pflug and Odlaug, 1986) or immobilised into alginate particles (that mimic food pieces) and undergo the thermal process with the food (Brown et al., 1984; Rodrigo et al., 1998; Serp et al., 2002). Enumeration of surviving organisms facilitates calculation of process lethality values (also expressed as F-values or P-values). Another valuable validation tool is the use of enzyme-based TTIs. These have proven to be a feasible and reliable tool for validation of pasteurisation processes, ranging from a few minutes at 70 °C to up to several minutes at 95 °C. These TTIs have employed endo-acting a-amylases © 2008, Woodhead Publishing Limited

518

Chilled foods

from fungal sources, such as Aspergillus oryzae (Raviyan et al., 2003), or from bacteria, e.g. Bacillus species (Tucker, 1999; Tucker et al., 2002; van Loey et al., 1996, 1997). In many cases, however, modelling of heat, mass and momentum transfer needs to be used as an alternative for setting the scheduled process, either for in-pack or continuous flow pasteurisation and/or sterilisation. The next section presents some modelling efforts developed for prediction of temperature and velocity distributions and profiles for both in-pack and continuous flow applications. Most of these efforts have focused on the range of temperatures applicable to commercial sterilisation. However, the modelling principles and assumptions are also relevant to pasteurisation treatments. Models for processing of in-pack and solid foods The fundamental objective of this approach is to calculate the temperature history of the slowest heating zone inside a pack (or a solid food) held under the slowest heating conditions (i.e. the slowest heating location inside the equipment used for thermal processing), to ensure that the target design heat treatment is delivered. Normally, for in-pack heat treatments such as for canned foods or hot-filled glass jars, direct measurement of temperature is possible. This has led to various methods of thermal process evaluation such as the General Method or the Ball formula method(s). Nonetheless, several modelling efforts have been reported in the literature using CFD approaches in order to gain a more thorough understanding of temperature distribution and flow behaviour inside packs where heating within the pack is governed by natural convection (i.e. buoyancy-driven heat transfer), rather than conduction. In this case, the continuity, momentum and energy equations (Equations 17.7 to 17.9) are typically solved by assuming that the density is constant, and that the buoyancy force caused by density differentials is only applied to the momentum equations. This method is known as the Boussinesq approximation, where the density can be expressed as:

ρ = ρ ref 1 − β (T − Tref ) [Eq. 17.10] where β is the thermal expansion coefficient, and ρref and Tref are density and temperature, respectively, at a reference value (initial conditions are generally used as reference values). The first efforts to model natural convection heating in in-pack heat-treated foods were done by Engelnam and Sani (1983), for beer pasteurisation in a bottle, and by Datta and Teixeira (1988), for sterilisation in a cylindrical can. More recently, with the rapidly increasing computational capabilities, other more sophisticated CFD approaches have been applied to study flow and heat transfer in non-Newtonian foods in different pack formats or geometries. Tattiyakul et al. (2001) modelled non-Newtonian flow and heat transfer in an axially rotating can. The authors concluded that for rotating cans, the slowest heating zone in the container was different than for stationary cans, with a strong influence of the rotation speed. Abdul Ghani et al. (2001) modelled natural convection flow and heat transfer in a retort pouch using a three-dimensional transient model. The same © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

519

research group modelled the effect of natural convection on the heating of a solid– liquid mixture (pineapple slices in juice) in a cylindrical can, showing that the slowest heating zone was located below the geometrical centre of the can (about 30–35% of the can height from the bottom), but at a higher distance from the bottom than if the can were filled with liquid only (Abdul Ghani and Farid, 2006). End-over-end rotation in retort processes is a common application for thermal processing of foods (either sterilisation or pasteurisation processes) and successful attempts to model (isothermal) flow alone (Hughes et al., 2003) and coupled flow and heat transfer (James et al., 2006) have been reported. Regarding cooling processes of in-pack heat-treated or solid foods, relevant models deal primarily with prediction of total cooling time under different cooling conditions (e.g. air-chilling vs. water immersion chilling), as well as CFD models for optimisation of airflow patterns in cold storage rooms or display cabinets. Several relevant examples of these applications have been described before (see Section 17.4.2). In recent years, modelling of vacuum cooling of both solid and liquid foods has gained interest, due to the shorter cooling times achieved with this technique compared to conventional cooling systems. By reducing the pressure in the system, the evaporation of water is accelerated and the heat required for evaporation is removed from the food, causing it to cool rapidly. This technique has been applied to cooling of a wide variety of food products, including fruit and vegetables, cooked meats, soups and sauces, and ready meals. A few relevant modelling efforts include those reported by Jin and Xu (2006) and Sun and Wang (2006) for cooked meats, and Dostal and Petera (2004) and Houška et al. (1996) for liquid foods. Models for processing liquid foods in continuous systems In the case of continuous-flow systems, and particularly in the case of aseptic processing of foods containing discrete particles, mathematical modelling gains more importance because the temperature within flowing solid particles cannot be easily measured without significant interference on heating rate by the sensor. In this scenario, the objective of the model is to calculate the temperature history of the coldest location within the slowest heating particle (i.e. the fastest-moving particle), to ensure that a commercially sterile product is obtained. Several modelling approaches have been reported in the literature (Cacace et al., 1994; Chandarana and Gavin, 1989; Lee et al., 1990; Sastry, 1986) but many factors remain to be fully understood. In particular, the quantification of the residence time distribution, especially that of the fastest-moving particles, in a continuously flowing stream of very large numbers of particles, is an important challenge. Other factors, such as coupling the thermal and flow problems, along with the difficulties associated with real food systems such as colliding particles, viscoelasticity and nonlinear systems, limit the applicability and reliability of this approach (Sastry and Cornelius, 2002). For single-phase flow (fluid without particulates), the problem of heating and cooling liquid foods in conventional heat exchangers (e.g. tubular, plate, and scraped surface) is well understood, and successful modelling efforts have been reported (see for instance Ditchfield et al., 2006; Grijspeerdt et © 2008, Woodhead Publishing Limited

520

Chilled foods

al., 2004; Jung and Fryer, 1999; Liao et al., 2000). In this scenario, the difficulties encountered for ensuring accuracy of model predictions are related to foods with complex rheological properties, and conditions that derive from these properties such as slip on the wall, yield stresses and rate of fouling.

17.4.4 Integration of predictive microbiology and process modelling approaches The reliability of predictions from a microbiological model depends on the accuracy of the data describing the environmental conditions used as inputs to the model. During processing of chilled foods, variations in the environmental conditions dictate how thermodynamic and transport phenomena within the food domain will affect the change of product parameters such as temperature, moisture gradients, pH or aw. These parameters will have a direct effect on the response of micro-organisms (e.g. growth and/or death rate). In some cases, data from experimental measurements (e.g. from a temperature data-logger) are available, which facilitates model usage. However, these data may have been collected only at a single location within the food product. In many cases, it may not be possible even to collect data due to the impossibility of locating a sensor at the desired position in the system. This may result in conservative estimates of the microbial response, limiting the possibilities for analysing the effect of different processing conditions on the assurance of microbiological safety and stability of the product. To overcome these shortcomings, the use of models for the solution of the heat, mass and momentum conservation equations (as described in Sections 17.4.1 and 17.4.2) offers a valuable tool to provide quantitative information about the history and the spatial distribution of relevant product/process parameters that can be used as inputs in predictive microbiology models. However, the integration of food engineering and microbiology models is uncommon. Effective integration of these models would support improved design of chilled foods, offering opportunities for process optimisation, and would facilitate assessment of the extent of problems caused by process deviations. In order for this integrated approach to be effective, it must successfully incorporate input from a wide range of disciplines including food microbiology, mathematics and statistics, food processing and engineering, thermodynamics, and physical chemistry. The most basic attempts to integrate physical and microbiological models are found in heat inactivation studies through the calculation of integrated lethality values (F- or P-values) at the slowest heating zone in a system, using the following equation:  T −Tref    z 

F or P = ∫ 10

⋅ dt

[Eq. 17.11]

where z (or z-value) describes the temperature sensitivity of the rate of inactivation of micro-organisms (D-value), and Tref is the reference temperature which can take different values depending on the design target (viz. pasteurisation or commercial sterilisation). However, as explained earlier, it is common in these cases to use a © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

521

temperature history obtained from a single-point measurement (at the slowest heating zone or ‘coldest spot’), or predicted from simplification of a heat transfer model (e.g. assuming lumped capacitance behaviour as explained in ‘Models available for Clostridium perfringens’, in Section 17.3.1) or from an empirical equation. In some food processing applications, this ‘coldest spot’ approach may be sufficient with respect to assessment of the effect of temperature on microbial responses. However it is not applicable to other relevant parameters such as pH or aw (Mafart, 2005). A few good examples of successful integration of food engineering and microbiology predictive models have been reported. For instance, a more robust assessment of the lethality delivered by a commercial sterilisation process is presented by Abdul Ghani et al. (2002). These authors coupled temperature and velocity profile predictions obtained from a natural convection CFD model in a 3D retort pouch to a convection-diffusion-reaction model of the concentration distribution and destruction of spores of Geobacillus stearothermophilus within the container. Bacterial inactivation kinetics were considered in the CFD models as a source term (see Eq. 17.1) defined by a first-order kinetic model based on the Arrhenius equation. Also in the area of thermal inactivation, Membré et al. (2006) used the nodal time–temperature history and spatial distribution from a 3D transient heat conduction model in a retort pouch as one of the inputs into a probabilistic model for prediction of the prevalence and concentration of B. cereus spores after a high-temperature pasteurisation process of a refrigerated food for extended shelf-life. Denys et al. (2005) predicted transient temperature distributions and velocity profiles within shell-eggs undergoing pasteurisation in a water immersion system, and used these predictions as inputs to calculate inactivation profiles (i.e. spatial distributions) of Salmonella Enteritidis. Other examples of heat and mass transfer models coupled with deterministic microbial inactivation kinetics include those reported by Valdramidis et al. (2005) for surface pasteurisation of a model food system using hot air (targeting inactivation of Escherichia coli), and Pradhan et al. (2007) for air–steam impingement cooking of chicken breasts (targeting inactivation of Listeria innocua). Integration of heat transfer models with dynamic microbial growth models has also been reported. Amézquita et al. (2005b) developed and validated an integrated model for predicting growth of C. perfringens during the cooling of cured boneless hams. The authors combined a 2D axisymmetric transient heat conduction model (Amézquita et al., 2005a) with the dynamic growth model of Baranyi and Roberts (1994), and validated the predictions under three different cooling scenarios. The growth of C. perfringens during the cooling of pea soup has also been modelled using an integrated heat transfer and microbial growth modelling approach (de Jong et al., 2005). Dynamic growth of E. coli during conduction cooling has also been reported using an integrated modelling approach (Bellara et al., 2000). Other relevant product parameters, such as moisture gradients, pH and aw, that affect microbial kinetics have been successfully modelled and combined with predictive models for microbial growth. Lebert et al. (2005) predicted the growth © 2008, Woodhead Publishing Limited

522

Chilled foods

of Listeria innocua on the surface of a gelatine gel under drying conditions as a function of temperature, pH and aw using a classic predictive microbiology model. Since the measurement of aw at the surface of a food matrix undergoing drying is experimentally challenging, a combination of a moisture transfer model and a thermodynamic model (as functions of relative humidity and velocity of air surrounding the food) were used to predict values of aw, subsequently used as inputs to the microbial growth model. Wilson et al. (2002) proposed a stoichiometric model to predict local changes in a food environment caused by bacterial conversion of carbon sources into acid metabolites and coupled this model to a Fickian diffusion model for metabolite transport to assess the effect of pH changes on microbial growth. The pH changes from the production of acid metabolites were determined by a physico-chemical model known as the Buffering Theory model (Wilson et al., 2000).

17.5 Quantitative microbiological risk assessment Quantitative microbiological risk assessment (QMRA) is a relatively recent development (see Hathaway, 1997), aimed fundamentally at the protection of consumers, used in decision making on food safety issues and helping responsible authorities to meet public health goals and the food industry to show compliance (EuropeanCommission, 2002; FDA-CFSAN, 2005; FDA-CFSAN and USDA-FSIS, 2003; USDA, 2000). To date, these have mostly focused on risks associated with single hazards (particular pathogens) and broad groups of products, enabling government agencies to identify products of greatest concern to public health and to identify key aspects of their processing/handling that impact most on risk. More recently, QMRA has been applied within the food industry and is fast becoming an indispensable part of product/process design and the evaluation of control measures along the steps of manufacturing and the supply chain. The main objective of QMRA is to quantify risk and/or to determine the effect of process/product interventions on risk. With a number of potential applications, the first step is therefore to clearly determine the purpose of the QMRA; that is, to define what questions will be answered and identify which data and calculations are required. The first draft of a risk ‘profile’ includes a detailed description of the product formulation and the manufacturing process, as well as the control measures implemented in the process (examples can be found at: http://www.nzfsa.govt.nz/ science/risk-profiles/, website accessed in July 2007). It should also include a review of available data in order to identify critical gaps that need to be filled before the QMRA can be conducted. Generally, QMRAs are very data-demanding, and a combination of different sources is usually necessary. For example, data can be generated from experiments (e.g. challenge tests), retrieved from databases (e.g. ComBase) or scientific publications, obtained from models (e.g. temperature distributions from a heat transfer model), or come from expert elicitation. The more relevant data are available, the lower the uncertainty of risk estimates will be, thereby offering © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

523

valuable guidance to risk managers for decision-making. When data or other information are missing, it is necessary to make assumptions during the QMRA process. These assumptions, along with all the data available, need to be clearly documented in order to make the conclusions of the risk assessment as transparent as possible. The well-known QMRA structure established by the Codex Alimentarius Commission (CAC, 1999) comprises four major elements:

• Hazard identification: in which pathogenic agents (micro-organisms or micro-

•

•

•

bial toxins) of concern with the food product are identified, as well as their association with adverse health effects. This part of the QMRA is predominantly qualitative. Sources of information are typically scientific literature, public databases, epidemiological data, relevant government bodies, and expert elicitation. Exposure assessment: which estimates the probability of intake of the pathogenic agent and the cell numbers consumed. For this purpose, exposure assessments must indicate the portion size or the unit used to determine the amount of food to be consumed. The main outcome of exposure assessments is an estimation of the level of the pathogenic agent (with some level of uncertainty) and the likelihood of occurrence in foods at the time of consumption. In processing of chilled foods this includes the initial prevalence and concentration of the pathogenic agent, the effect of any inactivation steps, the response of the micro-organism (growth, survival or inactivation) along the manufacturing process, the frequency of recontamination, the growth during storage and distribution, and the handling practices by the final consumer (e.g. temperatures in the household refrigerator). Hazard characterisation: which describes the health impact (severity and duration of adverse effects) of consuming a specified number of cells of the pathogenic agent on individuals, and is frequently determined by assessing pathogen–host relationships and, ideally, dose–response relationships. It is difficult to obtain dose–response data of highly virulent pathogens because human volunteering feeding studies are not possible. Furthermore, correlating animal responses to the responses of susceptible human populations has considerable uncertainties. It is possible to estimate these relationships by comparing monthly exposure estimates for sensitive populations with monthly observed epidemiological data, as demonstrated by Buchanan et al. (1997a). Risk characterisation: which provides a risk estimate as a result of combining the information and analyses performed in the previous three elements. The risk estimate is an indication of the level of disease resulting from a given exposure. It must include a description of the uncertainties associated with the estimate. The level of confidence of this estimate depends on the assumptions made, and on the variability and uncertainty identified in all the previous steps. It is important to differentiate between variability and uncertainty for subsequent risk management decisions. Separation of uncertainty and variability is an area of growing interest in QMRA, especially as it relates to exposure assessments

© 2008, Woodhead Publishing Limited

524

Chilled foods

where predictive models are used. A risk assessor must understand that predictive microbiology models cannot provide exact predictions of the response of micro-organisms in food products due to natural variability of both microorganisms and foods, as well as lack of knowledge and/or data (i.e. uncertainty) when developing a model. Separation of uncertainty and variability in the context of QMRA has been well documented (Nauta, 2000, 2007). When QMRA is used as part of product/process design of chilled foods, it is not uncommon to focus primarily on exposure assessment (see for instance Nauta, 2001). In this case, the target outcome of the assessment is the reduction of the exposure so that the levels of the pathogenic agent are below those required by regulatory agencies at the point of consumption. In this approach, mitigation and control strategies are designed for the exposure only. The quantitative nature of exposure assessment makes it the most complex part of the QMRA process. Thus, exposure assessments are usually focused on specific food categories and/or processes in order to reduce complexity, and facilitate risk management decisions. It is in the exposure assessment part of the QMRA process that predictive microbiology and process modelling play a fundamental role. For that reason, a properly conducted QMRA requires a multidisciplinary team of people, involving expertise in food safety and microbiology, food processing and engineering, and mathematical/statistical and computing skills.

17.5.1 Resources available There are a number of comprehensive QMRAs for different pathogen–product combinations that have been developed systematically by expert panels commissioned by the Food and Agriculture Organisation (FAO) and the World Health Organisation (WHO). These are available at the WHO website at: http:// www.who.int/foodsafety/micro/en/ (accessed on July 2007). Among the risk assessments available from FAO-WHO, those pathogen–product combinations that are of relevance to chilled foods are:

• Risk assessment of Listeria monocytogenes in ready-to-eat foods: Interpretative Summary, 2004 (Microbiological Risk Assessment Series, No. 4)

• Risk assessment of Listeria monocytogenes in ready-to-eat foods: Technical Report, 2004 (Microbiological Risk Assessment Series, No. 5)

• Risk assessment of Vibrio vulnificus in raw oysters: Interpretive Summary and Technical Report, 2005 (Microbiological Risk Assessment Series, No. 8)

• Risk assessment of choleragenic Vibrio cholerae O1 and O139 in warm water shrimp in international trade: Interpretive Summary and Technical Report, 2005. The website above also includes some guidance documents prepared by FAOWHO for best practice in conducting QMRAs. These documents include guidelines for hazard characterisation for pathogens in food and water, and exposure assessment and risk characterisation of microbiological hazards in food. © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

525

Other comprehensive QMRAs that are relevant for product/process design of chilled foods are those conducted by US government agencies for specific pathogen–product combinations, namely:

• Quantitative assessment of relative risk to public health from foodborne Liste• •

ria monocytogenes among selected categories of ready-to-eat foods (FDA-CFSAN and USDA-FSIS, 2003) Risk assessment for Clostridium perfringens in ready-to-eat and partially cooked meat and poultry products (USDA, 2005) Risk assessment for Listeria monocytogenes in deli meats (USDA-FSIS, 2003b).

Another good source of information about QMRAs for different pathogen– product combinations, foodrisk.org (http://www.foodrisk.org), is provided by the Joint Institute for Food Safety and Applied Nutrition (JIFSAN). This website offers a comprehensive list of available risk assessments, not only for microbiological hazards, but also for chemical hazards and other common contaminants in foods. Similarly, as a result of the European project COST 920 titled ‘Foodborne zoonosis: a co-ordinated food chain approach’, an inventory of QMRA studies in Europe is available at: http://www.cost920.com/00020.html. Regarding stand-alone or web-based software packages for the application of quantitative microbiology models within QMRAs, there are currently no packages available. When stochastic models are used for conducting QMRAs that are reported as scientific publications, risk assessors normally use simulation modelling packages such as @Risk (Palisade Corporation, Ithaca, New York, USA), Crystal Ball® (Decisioneering Inc., Denver, Colorado, USA) or Analytica® (Lumina Decision Systems, Inc., Los Gatos, California, USA). There are, however, a few specific software tools that have been developed to aid in QMRA, such as ‘Risk Ranger’ (Ross and Sumner, 2002) offered free of charge by the Australian Food Safety Centre of Excellence (available at: http://www.foodsafetycentre .com.au/riskranger.htm). Risk Ranger is a risk-ranking decision aid presented in spreadsheet format, which takes the inputs from the user as qualitative statements and/or numerical data on riskcontributing factors, and generates an output consisting of indices of the risk to public health. The user is required to answer eleven questions about the severity of the hazard, the likelihood of a disease-causing dose of the hazard occurring in the food portion and the probability of exposure to the hazard. Another software package that is available free of charge is FARE MicrobialTM (http://www.foodrisk.org/ faremicrobial.htm). This software is more sophisticated than Risk Ranger, and it was developed to perform probabilistic microbiological risk assessment. It was developed by a private company called Exponent at the request of the US Food and Drug Administration (FDA). The software incorporates algorithms developed by the FDA for the L. monocytogenes risk assessment; however, it is said to be capable of performing risk assessments for a wide variety of foodborne pathogens. FARE MicrobialTM consists of two modules: the Contamination and Growth Module, which accepts user-defined models and performs simulations that generate distributions representing contamination levels at the time of consumption, and the Exposure Module, which computes exposure distributions using demographic and © 2008, Woodhead Publishing Limited

526

Chilled foods

food consumption data from the 1994–1996, 1998 Continuing Survey of Food Intakes by Individuals (CSFII).

17.6 Recent developments Public health targets set by responsible authorities have to be interpreted by other parties involved in the provision of food. The World Trade Organization introduced the concept of appropriate level of protection (ALOP) as a public health target and other concepts have been proposed and introduced to translate these targets into meaningful and tangible objectives for the food industry. Amongst these are Food Safety Objectives (FSOs), Performance Objectives (POs) and Performance Criteria (PC) proposed by the International Commission on Microbiological Specifications for Foods (ICMSF) and adopted by the Codex Alimentarius Food Hygiene Committee. The discussion of how FSOs translate to ALOPs is not agreed but there have been recent studies aimed at resolving some of the remaining issues. In a recent study by Rieu et al. (2007), various models were used to relate annual marginal risk to the parameters of the FSO, depending on variability in the survival probability and the concentration of the pathogen. The authors proposed use of a stochastic Monte Carlo simulation and logistic discriminant function to determine which sets of parameters are compatible with ALOP, using risks of listeriosis and salmonellosis in one type of soft cheese. The authors recommended that the definition of FSO should integrate three dimensions: prevalence of contamination, average concentration per contaminated typical serving and dispersion of the concentration among those servings. In a recent application of QMRA, Doménech et al. (2007) reported on an application that aids in prioritisation of safety management measures from a manufacturer’s perspective, to allow better consumer health protection and minimise economic losses due to failures in manufacturing. An example of how an FSO set by competent authorities can guide decisions on safe product and process designs in practical operation has recently been published by our laboratory (Membré et al., 2007). In this case, we compared probabilistic and deterministic approaches to modelling thermal inactivation of salmonellae, and concluded that the probabilistic method allowed for a more transparent, objective and quantifiable means of establishing the stringency of food safety management systems. There are other recent examples of applications that take account of the uncertainty and variability of inputs used in risk assessment. Barker et al. (2005a) described a model for variability of lag time from a spore population. A Bayesian inference model was developed to provide uncertainty distributions for lag times of C. botulinum. It is likely that such approaches will become more common because of the advantages they offer over deterministic approaches even though they are more data hungry and more complex. Biological variability is recognised as a key property of pathogens and spoilage organisms and having the capability to describe this variability in a quantitative manner for risk assessment applications will become important for the future. © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

527

17.7 Conclusions In this chapter, we have described various modelling approaches and applications that have a role in maintaining the microbiological safety and quality of chilled foods. These reflect developments in models that aim to describe biological behaviour (summarised by McMeekin and Ross, 2002) and others that describe transport phenomena in food processing. Of the various types of predictive microbiology models developed, there are several that are particularly relevant to chilled foods. These include heat inactivation models, models for non-thermal inactivation and interface models. With product developers wishing to continually challenge the limits of preservation, driven by consumer demand for fresher, less heavily processed foods, it becomes more critical that food processors are able to control factors such as heat more precisely, and be able to confidently describe the effects of combinations likely to impact on microbial responses. There are relatively few examples of integrated models combining aspects of both process engineering and microbiology, but some have been developed and applied with success (e.g. Amézquita et al., 2005b). It is sometimes the case that data sets and models have been developed independently without anticipating integration with other models, and consequently data gaps may exist that make integration more difficult. The transfer of ‘safety by design’ into a food manufacturing operation requires translation of the output from predictive models into easily interpretable manufacturing conditions and limits, taking into account the various uncertainties and variabilities associated with food manufacturing operations. To this end, being able to properly account for the process variables related to considerations such as heat and mass transfer is critical. Food products are exposed to temperatures and other factors that fluctuate, and in-situ measurement of all of these fluctuations is not possible, so being able to predict these effects becomes more important. We have discussed only a couple of studies dealing with models to predict physico-chemical properties (pH, aw) of foods – these are few in number and often complex. These models are certainly an area that is important for the future but their integration with models predicting biological behaviour or response is still an area of development. Physico-chemical models for homogeneous foods are more straightforward than models for more complex, heterogeneous foods, so these still present a significant challenge. Interaction of the target micro-organism with other micro-organisms in foods is another consideration that adds to the multifaceted nature of predictive microbiology. In reality, this will be dependent on the indigenous flora, which can vary tremendously, depending on how the raw material/food is handled and processed. Such interactions are critical for fermented foods where there is reliance on particular communities of micro-organisms to provide safe and stable products. We conclude that integration of models from different disciplines is of critical importance for food products of the future, where designs will be increasingly based on quantitative microbiological risk assessment. Successful application of these approaches on the factory floor is dependent on a good understanding of © 2008, Woodhead Publishing Limited

528

Chilled foods

process engineering and having models that can accurately predict heat, mass and momentum transfer. This means that the ‘design’ activity can include detailed considerations of ‘in-pack’ or ‘in-line’ processing such that transfer to operation becomes much easier, being able to take account of the particular processing conditions likely to be used. Whilst there are still many aspects that require further investigation and development, models will remain an essential tool to ensure that foods remain safe and stable in the future.

17.8 References (2006) Using computational fluid dynamics to analyze the thermal sterilization of solid-liquid food mixture in cans, Innovative Food Science and Emerging Technologies, 7 55–61. ABDUL GHANI A G, FARID M M AND CHEN X D (2002) Theoretical and experimental investigation of the thermal inactivation of Bacillus stearothermophilus in food pouches, Journal of Food Engineering, 51 221–28. ABDUL GHANI A G, FARID M M, CHEN X D AND RICHARDS P (2001) Thermal sterilization of canned food in a 3-D pouch using computational fluid dynamics, Journal of Food Engineering, 48 147–56. ADAMS M R, LITTLE C L AND EASTER M C (1991) Modelling the effect of pH, acidulant and temperature on the growth rate of Yersinia enterocolitica, Journal of Applied Bacteriology, 71 65–71. AGGELIS G, SAMELIS J AND METAXOPOULOS J (1998) A novel modelling approach for predicting microbial growth in a raw cured meat product stored at 3 °C and 12 °C in air, International Journal of Food Microbiology, 43 39–52. ALVAREZ I, RASO J, SALA F J AND CONDON S (2003) Inactivation of Yersinia enterocolitica by pulsed electric fields, Food Microbiology, 20 691–700. AMÉZQUITA A, WANG L AND WELLER C L (2005a) Finite element modeling and experimental validation of cooling rates of large ready-to-eat meat products in small meat-processing facilities, Transactions of the ASAE, 48 287–303. AMÉZQUITA A, WELLER C L, WANG L, THIPPAREDDI H AND BURSON D E (2005b) Development of an integrated model for heat transfer and dynamic growth of Clostridium perfringens during the cooling of cooked boneless ham, International Journal of Food Microbiology, 101 123–44. AUGUSTIN J C, BROULLAUD-DELATTRE A, ROSSO L AND CARLIER V (2000) Significance of inoculum size in the lag time of Listeria monocytogenes, Applied and Environmental Microbiology, 66 1706–10. AUGUSTIN J C, CARLIER V AND ROZIER J (1998) Mathematical modelling of the heat resistance of Listeria monocytogenes, Journal of Applied Microbiology, 84 185–91. BAKER D A AND GENIGEORGIS C A (1990) Predicting the safe storage of fresh fish under modified atmospheres with respect to Clostridium botulinum toxigenesis by modeling length of the lag phase of growth, Journal of Food Protection, 53 131–40. BAKER J M AND GRIFFITHS M W (1993) Predictive modeling of psychrotrophic Bacillus cereus, Journal of Food Protection, 56 684–88. BARANYI J, PIN C AND ROSS T (1999) Validating and comparing predictive models, International Journal of Food Microbiology, 48 159–66. BARANYI J AND ROBERTS T A (1994) A dynamic approach to predicting bacterial growth in food, International Journal of Food Microbiology, 23 277–94. BARANYI J AND ROBERTS T A (1995) Mathematics of predictive food microbiology, International Journal of Food Microbiology, 26 199–218. ABDUL GHANI A G AND FARID M M

© 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

529

BARANYI J, ROBERTS T A AND MCCLURE P J (1993) A non-autonomous differential equation

to model bacterial growth, International Journal of Food Microbiology, 10 43–59. (1995) Predicting the growth of Brochothrix thermosphacta at changing temperature, International Journal of Food Microbiology, 27 61–75. BARANYI J AND TAMPLIN M L (2004) ComBase: A common database on microbial responses to food environments, Journal of Food Protection, 67 1967–71. BARKER G C, MALAKAR P K AND PECK M W (2005a) Germination and growth from spores: Variability and uncertainty in the assessment of food borne hazards, International Journal of Food Microbiology, 100 67–76. BARKER G C, MALAKAR P K, TORRE M D, STECCHINI M L AND PECK M W (2005b) Probabilistic representation of the exposure of consumers to Clostridium botulinum neurotoxin in a minimally processed potato product, International Journal of Food Microbiology, 100 345–57. BEAN P G (1983) Developments in heat treatment processes for shelf-stable products. In: Roberts, T. A. and Skinner, F. A. (eds.), Food Microbiology: Advances and Prospects, Academic Press, London, England, pp. 97–112. BELLARA S R, MCFARLANE C M, THOMAS C R AND FRYER P J (2000) The growth of Escherichia coli in a food simulant during conduction cooling: Combining engineering and microbiological modelling, Chemical Engineering Science, 55 6085–95. BHADURI S, SMITH P W, PALUMBO S A, TURNER-JONES C O, SMITH J L, MARMER B S, BUCHANAN R L, ZAIKA L L AND WILLIAMS A C (1991) Thermal destruction of Listeria monocytogenes in a liver sausage slurry, Food Microbiology, 8 75–78. BHADURI S, TURNER JONES C O, BUCHANAN R L AND PHILLIPS J G (1994) Response surface model of the effect of pH, sodium chloride and sodium nitrite on growth of Yersinia enterocolitica at low temperatures, International Journal of Food Microbiology, 23 333– 43. BLACKBURN C D W, CURTIS L M, HUMPHESON L, BILLON C AND MCCLURE P J (1997) Development of thermal inactivation models for Salmonella enteritidis and Escherichia coli O157:H7 with temperature, pH and NaCl as controlling factors, International Journal of Food Microbiology, 38 31–44. BLANKENSHIP L C, CRAVEN S E, LEFFLER R G AND CUSTER C (1988) Growth of Clostridium perfringens in cooked chili during cooling, Applied and Environmental Microbiology, 54 1104–08. BLOCHER J C AND BUSTA F F (1983) Bacterial spore resistance to acid, Food Technology, 37 87–99. BOLTON L F AND FRANK J F (1999) Defining the growth/no-growth interface for Listeria monocytogenes in Mexican-style cheese based on salt, pH, and moisture content, Journal of Food Protection, 62 602–09. BRADSHAW J G, PEELER J T AND TWEDT R M (1975) Heat resistance of ileal loop reactive Bacillus cereus strains isolated from commercially canned food, Applied Microbiology, 30 943–45. BRAITHWAITE P J AND PERIGO J A (1971) The influence of pH, water activity and recovery temperature on the resistance and outgrowth of Bacillus spores. In: Barker, A. M., Gould, G. W., and Wolf, J. (eds.), Spore Research, Academic Press, London, England, pp. 289–84. BROWN K L, AYRES C A, GAZE J E AND NEWMAN M E (1984) Thermal destruction of bacterial spores immobilized in food/alginate particles, Food Microbiology, 1 187–98. BUCHANAN R L, DAMERT W G, WHITING R C AND VAN SCHOTHORST M (1997a) Use of epidemiological and food survey data to estimate a purposefully conservative doseresponse relationship for Listeria monocytogenes levels and incidence of listeriosis, Journal of Food Protection, 60 918–22. BUCHANAN R L AND EDELSON S G (1996) Culturing enterohemorrhagic Escherichia coli in the presence and absence of glucose as a simple means of evaluating the acid tolerance of stationary-phase cells, Applied and Environmental Microbiology, 62 4009–13. BARANYI J, ROBINSON T P, KALOTI A AND MACKEY B M

© 2008, Woodhead Publishing Limited

530

Chilled foods

BUCHANAN R L, GOLDEN M H AND WHITING R C

(1997b) Expanded models for the nonthermal inactivation of Listeria monocytogenes, Journal of Applied Microbiology, 82 567–77. BUCHANAN R L AND PHILLIPS J G (1990) Response surface model for predicting the effects of temperature, pH, sodium chloride content, sodium nitrite concentration and atmosphere on the growth of Listeria monocytogenes, Journal of Food Protection, 53 370–76. CAC (1999) Principles and Guidelines for the Conduct of Microbiological Risk Assessment, CAC/GL-30-1999. Available at: http://www.who.int/foodsafety/publications/micro/ cac1999/en/, Accessed June 2007. CACACE D, PALMIERI L, PIRONE G, DIPOLLINA G, MASI P AND CAVELLA S (1994) Biological validation and mathematical modeling of the thermal processing of particulate foods – The influence of heat transfer coefficient determination, Journal of Food Engineering, 23 51– 68. CAMPAÑONE L A AND ZARITZKY N E (2005) Mathematical analysis of microwave heating process, Journal of Food Engineering, 69 359–68. CARLIN F, FRICKER M, PIELAAT A, HEISTERKAMP S, SHAHEEN R, SALKINOJA-SALONEN M, SVENSSON B, NGUYEN-THE C AND EHLING-SHULZ M (2006) Emetic toxin-producing strains of Bacillus cereus show distinct characteristics within the Bacillus cereus group, International Journal of Food Microbiology, 109 132–38. CARLIN F, GUINEBRETIERE M H, CHOMA C, PASQUALININ R, BRACCONIER A AND NGUYEN-THE C (2000) Spore-forming bacteria in commercial cooked, pasteurised and chilled vegetable purées, Food Microbiology, 17 153–65. CASSIN M H, PAOLI G M AND LAMMERDING A M (1998) Simulation modeling for microbial risk assessment, Journal of Food Protection, 61 1560–66. CAUSTON D R (1987) A Biologist’s Advanced Mathematics, Allen and Unwin, London. CERF O, DAVEY K R AND SADOUDI A K (1996) Thermal inactivation of bacteria – A new predictive model for the combined effect of three environmental factors: temperature, pH and water activity, Food Research International, 29 219–26. CHANDARANA D I AND GAVIN A (1989) Establishing the thermal process for foods to be processed aseptically: A theoretical comparison of process development methods, Journal of Food Science, 54 198–204. CHICK H (1908) An investigation of the laws of disinfection, Journal of Hygiene Cambridge, 8 92–158. CHOI Y AND OKOS M R (1986) Effects of temperature and composition on the thermal properties of foods. In: Le Maguer, M. and Jelen, P. (eds.), Food Engineering and Process Applications, Elsevier Science, New York, pp. 93–101. CHOMA C, GUINEBRETIŠRE M H, CARLIN F, SCHMITT P, VELGE P, GRANUM P E AND NGUYEN-THE C (2000) Prevalence, characterization and growth of Bacillus cereus in commercial cooked chilled foods containing vegetables, Journal of Applied Microbiology, 88 617–25. COLE M B, DAVIES K W, MUNRO G, HOLYOAK C D AND KILSBY D C (1993) A vitalistic model to describe the thermal inactivation of Listeria monocytogenes, Journal of Industrial Microbiology, 12 232–39. COLE M B, JONES M V AND HOLYOAK C (1990) The effect of pH, salt concentration and temperature on the survival and growth of Listeria monocytogenes, Journal of Applied Bacteriology, 69 63–72. COLLADO J, FERNANDEZ A, CUNHA L M AND MARTINEZ A (2003) Improved model based on the Weibull distribution to describe the combined effect of pH and temperature on the heat resistance of Bacillus cereus in carrot juice, Journal of Food Protection, 66 978–84. COOK F K AND PIERSON M D (1983) Inhibition of bacterial spores by antimicrobials, Food Technology, 37 115–26. COUVERT O, LEGUERINEL I AND MAFART P (1999) Modelling the overall effect of pH on the apparent heat resistance of Bacillus cereus spores, International Journal of Food Microbiology, 49 57–62. © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

531

CRIELLY E M, LOGAN N A AND ANDERTON A (1994) Studies on the Bacillus flora of milk and

milk products, Journal of Applied Bacteriology, 77 256–63. (1995) Modelling of microbial activity and prediction of shelf life of packed fresh fish, International Journal of Food Microbiology, 19 305–18. DALGAARD P, BUCH P AND SILBERG S (2002) Seafood Spoilage Predictor – Development and distribution of a product specific application software, International Journal of Food Microbiology, 73 343–49. DALGAARD P, MEJLHOLM O AND HUSS H H (1997) Application of an iterative approach for development of a microbial model predicting the shelf-life of packed fish, International Journal of Food Microbiology, 38 169–79. DATTA A K (2001) Fundamentals of heat and moisture transport for microwaveable food product and process development. In: Datta, A. K. and Anantheswaran, R. C. (eds.), Handbook of Microwave Technology for Food Applications, Marcel Dekker, Inc., New York, pp. 115–72. DATTA A K AND TEIXEIRA A A (1988) Numerically predicted transient temperature and velocity profiles during natural convection heating of canned liquid food, Journal of Food Science, 53 191–95. DAVEY K R (1994) Modelling the combined effect of temperature and pH on the rate coefficient for bacterial growth, International Journal of Food Microbiology, 23 295– 303. DE ALWIS A A P AND FRYER P J (1990) A finite-element analysis of heat generation and transfer during ohmic heating of food, Chemical Engineering Science, 45 1547–59. DE JONG A E I, BEUMER R R AND ZWIETERING M H (2005) Modeling growth of Clostridium perfringens in pea soup during cooling, Risk Analysis, 25 61–73. DENG Y, RYU J H AND BEUCHAT L R (1998) Influence of temperature and pH on survival of Escherichia coli O157:H7 in dry foods and growth in reconstituted infant rice cereal, International Journal of Food Microbiology, 45 173–84. DENG Y, RYU J H AND BEUCHAT L R (1999) Tolerance of acid-adapted and non-adapted Escherichia coli O157:H7 cells to reduced pH as affected by type of acidulant, Journal of Applied Microbiology, 86 203–10. DENYS S, PIETERS J G AND DEWETTINCK K (2005) Computational fluid dynamics analysis for process impact assessment during thermal pasteurization of intact eggs, Journal of Food Protection, 68 2005. DITCHFIELD C, TADINI C C, SINGH R K AND TOLEDO R T (2006) Velocity and temperature profiles, heat transfer coefficients and residence time distribution of a temperature dependent Herschel–Bulkley fluid in a tubular heat exchanger, Journal of Food Engineering, 76 632–38. DOMÉNECH E, ESCRICHE I AND MARTORELL S (2007) Quantification of risks to consumers’ health and to company’s incomes due to failures in food safety, Food Control, 18 1419–27. DONG SUN L, KEUM JIN H, IL S, JIN PYO P AND HYUN DONG P (2006) Estimation of shelf life distribution of seasoned soybean sprouts using the probability of Bacillus cereus contamination and growth, Food Science and Biotechnology, 15 773–77. DOSTAL M AND PETERA K (2004) Vacuum cooling of liquids: Mathematical model, Journal of Food Engineering, 61 533–39. DRAPER N R AND SMITH H (1981) Applied Regression Analysis, John Wiley and Sons, New York. DUFRENNE J, BIJWAARD M, TE GIFFEL M, BEUMER R AND NOTERMANS S (1995) Characteristics of some psychrotrophic Bacillus cereus isolates, International Journal of Food Microbiology, 27 175–83. DUFRENNE J, SOENTORO P, TATINI S, DAY T AND NOTERMANS S (1994) Characteristics of Bacillus cereus related to safe food production, International Journal of Food Microbiology, 23 99–109. ECFF (2006) Recommendations for the production of pre-packaged chilled food, Available at: http://www.ecff.net/. Accessed March 2007. DALGAARD P

© 2008, Woodhead Publishing Limited

532

Chilled foods

(1983) Finite-element simulation of an in-package pasteurization process, Numerical Heat Transfer, 6 41–54. ESTY J R AND MEYER K F (1922) The heat resistance of spores of B. botulinus and allied anaerobes, Journal of Infectious Diseases, 31 650–63. EUROPEAN-COMMISSION (2002) Risk assessment of foodborne bacterial pathogens: Quantitative methodology relevant for human exposure assessment, European Commission. Health and Consumer Protection, Directorate General. Available at: http://europa.eu.int/ comm/food/fs/sc/ssc/out252_en.pdf, Accessed June 2007. EUROPEAN-COMMISSION (2005) Commission Regulation (EC) No 2073/2005 of 15 November 2005 on microbiological criteria for foodstuffs, Official Journal of the European Union, 48, L338 1–26. FAILLE C, LEBRET V, GAVINI F AND MAINGONNAT J-F (1997) Injury and lethality of heat treatment of Bacillus cereus spores in buffer and poultry meat, Journal of Food Protection, 60 544–47. FAO/WHO (2004) Risk assessment of Listeria monocytogenes in ready-to-eat foods: Technical report, Available at: http://www.who.int/foodsafety/micro/jemra/assessment/ index.html, Accessed May 2007. FARBER J M, CAI Y AND ROSS W H (1996) Predictive modeling of the growth of Listeria monocytogenes in CO2 environments, International Journal of Food Microbiology, 32 133–44. FDA (2007) Guidance for industry: Refrigerated carrot juice and other refrigerated low-acid juices, Available at: http://www.cfsan.fda.gov/~dms/juicgu15.html, Accessed July 2007. FDA-CFSAN (2001) Fish and Fisheries Products Hazards & Controls Guidance. FDA-CFSAN (2005) Quantitative risk assessment on the public health impact of pathogenic Vibrio parahaemolyticus in raw oysters, Available at: http://www.cfsan.fda.gov/~dms/ vpra-toc.html, Accessed June 2007. FDA-CFSAN AND USDA-FSIS (2003) Quantitative assessment of the relative risk to public health from foodborne Listeria monocytogenes among selected categories of ready-to-eat foods, Available at: http://www.foodsafety.gov/~dms/lmr2-toc.html, Accessed May 2007. FERNANDEZ A, LOPEZ M, BERNARDO A, CONDON S AND RASO J (2007) Modelling thermal inactivation of Listeria monocytogenes in sucrose solutions of various water activities, Food Microbiology, 24 372–79. FERNANDEZ A, OCIO M J, FERNANDEZ P S, RODRIGO M AND MARTINEZ A (1999a) Application of nonlinear regression analysis to the estimation of kinetic parameters for two enterotoxigenic strains of Bacillus cereus spores, Food Microbiology, 16 607–13. FERNANDEZ A, SALMERON C, FERNANDEZ P S AND MARTINEZ A (1999b) Application of a frequency distribution model to describe the thermal inactivation of two strains of Bacillus cereus, Trends in Food Science and Technology, 10 158–62. FERNANDEZ P S AND PECK M W (1997) Predictive model describing the effect of prolonged heating at 70 to 80 °C and incubation at refrigeration temperatures on growth and toxigenesis by nonproteolytic Clostridium botulinum, Journal of Food Protection, 60 1064–71. FOEGEDING P M AND BUSTA F F (1981) Bacterial spore injury – An update, Journal of Food Protection, 44 776–86. FOSTER A M, MADGE M AND EVANS J A (2005) The use of CFD to improve the performance of a chilled multi-deck retail display cabinet, International Journal of Refrigeration, 28 698–705. FRANCIS K P, MAYR R, VON STETTEN F, STEWART G AND SCHERER S (1998) Discrimination of psychrotrophic and mesophilic strains of the Bacillus cereus group by PCR targeting of major cold shock protein genes, Applied and Environmental Microbiology, 64 3525–29. FRANCOIS K, VALERO A, GEERAERD A H, VAN IMPE J F, DEBEVERE J, GARCIA GIMENO R M, ZURERA G AND DEVLIEGHERE F (2007) Effect of preincubation temperature and pH on the individual cell lag phase of Listeria monocytogenes, cultured at refrigeration temperatures, Food Microbiology, 24 32–43. ENGELNAM M S AND SANI R L

© 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods GAILLARD S, LEGUERINEL I AND MAFART P

533

(1998a) Model for combined effects of temperature, pH and water activity on thermal inactivation of Bacillus cereus spores, Journal of Food Science, 63 887–89. GAILLARD S, LEGUERINEL I AND MAFART P (1998b) Modelling combined effects of temperature and pH on the heat resistance of spores of Bacillus cereus, Food Microbiology, 15 625–30. GAILLARD S, LEGUERINEL I, SAVY N AND MAFART P (2005) Quantifying the combined effects of the heating time, the temperature and the recovery medium pH on the regrowth lag time of Bacillus cereus spores after a heat treatment, International Journal of Food Microbiology, 105 53–58. GAZE J E AND BROWN G D (1990) Determination of the heat resistance of a strain of nonproteolytic Clostridium botulinum type B and a strain of type E, heated in cod and carrot homogenate over the temperature range 70 to 92 °C, Technical Memorandum No. 592, Campden Food and Drink Research Association. GEEDIPALLI S S R, RAKESH V AND DATTA A K (2007) Modeling the heating uniformity contributed by a rotating turntable in microwave ovens, Journal of Food Engineering, 82 359–68. GEERAERD A H, VALDRAMIDIS V P AND VAN IMPE J F (2006) GInaFiT, a freeware tool to assess the non-log-linear microbial survivor curves, International Journal of Food Microbiology, 102 95–105. GENIGEORGIS C A, MENG J H AND BAKER D A (1991) Behavior of nonproteolytic Clostridium botulinum type B and E spores in cooked turkey and modeling lag phase and probability of toxigenesis, Journal of Food Science, 56 373–79. GEORGE S M, RICHARDSON L C C AND PECK M W (1996) Predictive models of the effect of temperature, pH and acetic and lactic acids on the growth of Listeria monocytogenes, International Journal of Food Microbiology, 32 73–90. GIBSON A M, BRATCHELL N AND ROBERTS T A (1988) Predicting microbial growth: Growth responses of salmonellae in a laboratory medium as affected by pH, sodium chloride and storage temperature, International Journal of Food Microbiology, 6 155–78. GIL M M, PEREIRA P M, BRANDAO T R S, SILVA C L M, KONDJOYAN A, VALDRAMIDIS V P, GEERAERD A H, VAN IMPE J F AND JAMES S (2006) Integrated approach on heat transfer and inactivation kinetics of microorganisms on the surface of foods during heat treatments – Software development, Journal of Food Engineering, 76 95–103. GILBERT R J, STRINGER M F AND PEACE T C (1974) The survival and growth of Bacillus cereus in boiled and fried rice in relation to outbreaks of food poisoning, Journal of Hygiene, Cambridge, 73 433–44. GIMENEZ B AND DALGAARD P (2004) Modelling and predicting the simultaneous growth of Listeria monocytogenes and spoilage microorganisms in cold-smoked salmon, Journal of Applied Microbiology, 96 96–109. GOLDWIN S L, CHEN F C, CHAMBERS IV E, COPPINGS R AND CHAMBERS D (2007) A comprehensive evaluation of temperatures within home refrigerators, Food Protection Trends, 27 168–73. GONZALEZ I, LOPEZ M, MARTINEZ S, BERNARDO A AND GONZALEZ J (1997) Thermal resistance of Bacillus cereus spores as affected by additives in the recovery medium, Journal of Food Safety, 17 1–12. GONZALEZ I, LOPEZ M, MARTINEZ S, BERNARDO A AND GONZALEZ J (1999) Thermal inactivation of Bacillus cereus spores formed at different temperatures, International Journal of Food Microbiology, 51 81–84. GONZALEZ I, LOPEZ M, MAZAS M, GONZALEZ J AND BERNARDO A (1995) The effect of recovery conditions on the apparent heat-resistance of Bacillus cereus spores, Journal of Applied Bacteriology, 78 548–54. GOULD G W, BROWN M H AND FLETCHER B C (1983) Mechanisms of action of food preservative procedures. In: Roberts, T. A. and Skinner, F. A. (eds.), Food Microbiology: Advances and Prospects, Society for Applied Bacteriology Symposium, Series No. 11. Academic Press, London, pp. 67–84. © 2008, Woodhead Publishing Limited

534

Chilled foods

GRIJSPEERDT K, MORTIER L, DE BLOCK J AND VAN RENTERGHEM R

(2004) Applications of modelling to optimise ultra high temperature milk heat exchangers with respect to fouling, Food Control, 15 117–30. GUERZONI M E, LANCIOTTI R, TORRIANI S AND DELLAGLIO F (1994) Growth modelling of Listeria monocytogenes and Yersinia enterocolitica in food model systems and dairy products, International Journal of Food Microbiology, 24 83–92. GUILLIER L, PARDON P AND AUGUSTIN J-C (2005) Influence of stress on individual lag time distributions of Listeria monocytogenes, Applied and Environmental Microbiology, 71 2940–48. GYUNG JIN B, TODD E C D, CHONG HAE H, DEOG HWAN O AND SANG DO H (2007) Exposure assessment for Bacillus cereus in ready-to-eat Kimbab selling at stores, Food Control, 18 682–88. HAIQIANG C AND HOOVER D G (2003) Pressure inactivation kinetics of Yersinia enterocolitica ATCC 35669, International Journal of Food Microbiology, 87 161–71. HARMON S M AND KAUTTER D A (1991) Incidence and growth potential of Bacillus cereus in ready-to-serve foods, Journal of Food Protection, 54 372–74. HASSANI M, MANAS P, RASO J, CONDÓN S AND PAGAN R (2005) Predicting heat inactivation of Listeria monocytogenes under nonisothermal treatments, Journal of Food Protection, 68 736–43. HATHAWAY S C (1997) Development of food safety risk assessment guidelines for foods of animal origin in international trade, Journal of Food Protection, 60 1432–38. HAUSCHILD A H W (1982) Assessment of botulism hazards from cured meats, Food Technology, 36 95–104. HERSOM A C AND HULLAND E D (1980) Canned Foods: Thermal processing and microbiology, Churchill Livingston, London, England. HONG Y, MOKHTARI A, JAYKUS L A, MORALES R A, CATES S C AND COWEN P (2006) Consumer phase risk assessment for Listeria monocytogenes in deli meats, Risk Analysis, 26 89–103. HOUŠKA M, PODLOUCKÝ R, ZITNÝ R, GRÉE R, ŠESTÁK J, DOSTÁL M AND BURFOOT D (1996) Mathematical model of the vacuum cooling of liquids, Journal of Food Engineering, 39 339–48. HUANG L (2003a) Dynamic computer simulation of Clostridium perfringens growth in cooked ground beef, International Journal of Food Microbiology, 87 217–27. HUANG L (2003b) Estimation of growth of Clostridium perfringens in cooked beef under fluctuating temperature conditions, Food Microbiology, 20 549–59. HUDSON J A (1993) Construction and comparison of response surface kinetic models for the Yersinia enterocolitica type strain and a food isolate under aerobic conditions, International Journal of Food Microbiology, 18 201–09. HUDSON J A (1994) Comparison of response surface models for Listeria monocytogenes strains under aerobic conditions, Food Research International, 27 53–59. HUDSON J A, MOTT S J AND PENNEY N (1994) Growth of Listeria monocytogenes, Aeromonas hydrophila and Yersinia enterocolitica on vacuum and saturated carbon dioxide controlled atmosphere-packaged sliced roast beef, Journal of Food Protection, 57 204–08. HUGHES J P, JONES T E R AND JAMES P W (2003) Numerical simulation and experimental visualization of the isothermal flow of liquid containing a headspace bubble inside a closed cylinder during off-axis rotation, Transactions of the Institution of Chemical Engineers Part C, 81 119–28. HYYTIA TREES E, SKYTTA E, MOKKILA M, KINNUNEN A, LINDSTROM M, LAHTEENMAKI L, AHVENAINEN R AND KORKEALA H (2000) Safety evaluation of sous vide-processed products with respect to nonproteolytic Clostridium botulinum by use of challenge studies and predictive microbiological models, Applied and Environmental Microbiology, 66 223–29. ICMSF (1998) Potential application of risk assessment techniques to microbiological issues related to international trade in food and food products, Journal of Food Protection, 61 1075–86. v

© 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

535

ICMSF (2002) Microorganisms in Foods 7, Kluwer Academic/Plenum, London, UK. JAMES P W, HUGHES J P, JONES T E R AND TUCKER G S (2006) Numerical simulation of non-

isothermal flow in off-axis rotation of a can containing a headspace bubble, Transactions of the Institution of Chemical Engineers Part A, 84 311–18. JAMES S J AND EVANS J A (2006) Predicting the reduction of microbes on the surface of foods during surface pasteurisation – the ‘BUGDEATH’ project, Journal of Food Engineering, 76 1–6. JANSSEN M, GEERAERD A H, LOGIST F, VISSCHER Y D, VEREECKEN K M, DEBEVERE J, DEVLIEGHERE F AND IMPE J F V (2006) Modelling Yersinia enterocolitica inactivation in coculture experiments with Lactobacillus sakei as based on pH and lactic acid profiles, International Journal of Food Microbiology, 111 59–72. JIANGHONG M AND GENIGEORGIS C A (1993) Modeling lag phase of nonproteolytic Clostridium botulinum toxigenesis in cooked turkey and chicken breast as affected by temperature, sodium lactate, sodium chloride and spore inoculum, International Journal of Food Microbiology, 19 109–22 JIN T X AND XU L (2006) Development and validation of moisture movement model for vacuum cooling of cooked meat, Journal of Food Engineering, 75 333–39. JONES J E AND WALKER S J (1993) Advances in modeling microbial growth, Journal of Industrial Microbiology, 12 200–05 JORDAN K N, HALL S AND MCCLURE P J (1999) Osmotic stress on dilution of acid injured Escherichia coli O157:H7, Letters in Applied Microbiology, 28 389–93. JUNEJA V K (2003) Predictive model for the combined effect of temperature, sodium lactate, and sodium diacetate on the heat resistance of Listeria monocytogenes in beef, Journal of Food Protection, 66 804–11. JUNEJA V K AND EBLEN B S (1999) Predictive thermal inactivation model for Listeria monocytogenes with temperature, pH, NaCl and sodium pyrophosphate as controlling factors, Journal of Food Protection, 62 986–93. JUNEJA V K, EBLEN B S, MARMER B S, WILLIAMS A C, PALUMBO S A AND MILLER A J (1995a) Thermal resistance of nonproteolytic type B and type E Clostridium botulinum spores in phosphate buffer and turkey slurry, Journal of Food Protection, 58 758–63. JUNEJA V K, HUANG L AND THIPPAREDDI H H (2006) Predictive model for growth of Clostridium perfringens in cooked cured pork, International Journal of Food Microbiology, 110 85–92 JUNEJA V K AND MARKS H M (2002) Predictive model for growth of Clostridium perfringens during cooling of cooked cured chicken, Food Microbiology, 19 313–27. JUNEJA V K, MARMER B S AND EBLEN B S (1999a) Predictive model for the combined effect of temperature, pH, sodium chloride and sodium pyrophosphate on the heat resistance of Escherichia coli O157:H7, Journal of Food Protection, 19 147–60. JUNEJA V K, MARMER B S, PHILLIPS J G AND MILLER A J (1995b) Influence of the intrinsic properties of food on thermal inactivation of spores of nonproteolytic Clostridium botulinum: Development of a predictive model, Journal of Food Safety, 15 349–64. JUNEJA V K, WHITING R C, MARKS H M AND SNYDER O P (1999b) Predictive model for growth of Clostridium perfringens at temperatures applicable to cooling of cooked meat, Food Microbiology, 16 335–49. JUNG A AND FRYER P J (1999) Optimising the quality of safe food: Computational modelling of a continuous sterilisation process, Chemical Engineering Science, 54 717–30. KANG D H AND FUNG D Y C (2000) Application of thin agar layer method for recovery of injured Salmonella Typhimurium, International Journal of Food Microbiology, 54 127–32. KILSBY D C, DAVIES K W, MCCLURE P J, ADAIR C A AND ANDERSON W A (2000) Bacterial thermal death kinetics based on probability distributions: The heat destruction of Clostridium botulinum and Salmonella Bedford, Journal of Food Protection, 63 1179–203. KINSELLA K J, ROWE T A, BLAIR I S, MCDOWELL D A AND SHERIDAN J J (2006) Survival and recovery of Salmonella enterica serovar Typhimurium DT104 at low temperature and water activity in a broth system, Foodborne Pathogens and Disease, 3 375–83. © 2008, Woodhead Publishing Limited

536

Chilled foods

KOUTSOUMANIS K (2001) Predictive modeling of the shelf life of fish under nonisothermal

conditions, Applied and Environmental Microbiology, 67 1821–29. (2000) Application of a systematic experimental procedure to develop a microbial model for rapid fish shelf life predictions, International Journal of Food Microbiology, 60 171–84. KOUTSOUMANIS K, TAOUKIS P S AND NYCHAS G J E (2005) Development of a Safety Monitoring and Assurance System for chilled food products, International Journal of Food Microbiology, 100 253–60. KUTALIK Z, RAZAZ M, ELFWING A, BALLAGI A AND BARANYI J (2005) Stochastic modelling of individual cell growth using flow chamber microscopy images, International Journal of Food Microbiology, 105 177–90. LABBE R G AND HUANG T H (1995) Generation times and modeling of enterotoxin-positive and enterotoxin-negative strains of Clostridium perfringens in laboratory media and ground beef, Journal of Food Protection, 58 1303–06. LAGUERRE O, BEN-AMARA S, MOUREH J AND FLICK D (2007) Numerical simulation of air flow and heat transfer in domestic refrigerators, Journal of Food Engineering, 81 144–56. LANCIOTTI R, SINIGAGLIA M, GARDINI F, VANNINI L AND GUERZONI M E (2001) Growth/no growth interfaces of Bacillus cereus, Staphylococcus aureus and Salmonella enteritidis in model systems based on water activity, pH, temperature and ethanol concentration, Food Microbiology, 18 659–68 LAURENT Y, ARINO S AND ROSSO L (1999) A quantitative approach for studying the effect of heat treatment conditions on resistance and recovery of Bacillus cereus spores, International Journal of Food Microbiology, 48 149–57. LEBERT I, DUSSAP C G AND LEBERT A (2005) Combined physico-chemical and water transfer modelling to predict bacterial growth during food processes, International Journal of Food Microbiology, 102 305–22. LEBERT I AND LEBERT A (2006) Quantitative prediction of microbial behaviour during food processing using an integrated modelling approach: A review, International Journal of Refrigeration, 29 968–84. LECHNER S, MAYR R, FRANCIS K P, PRÜSS B M, KAPLAN T, WIESSNER-GUNKEL E, STEWART G AND SCHERER S (1998) Bacillus weihenstephanensis sp. nov. is a new psychrotolerant species of the Bacillus cereus group, International Journal of Systematic Bacteriology, 48 1373–82. LEE J H, SINGH R K AND LARKIN J W (1990) Determination of lethality and processing time in a continuous sterilization system containing particulates, Journal of Food Engineering, 11 67–92. LEGAN J D, SEMAN D L, MILKOWSKI A L, HIRSCHEY J A AND VANDEVEN M H (2004) Modeling the growth boundary of Listeria monocytogenes in ready-to-eat cooked meat products as a function of the product salt, moisture, potassium lactate, and sodium diacetate concentrations, Journal of Food Protection, 67 2195–204. LEGUERINEL I, COUVERT O AND MAFART P (2000) Relationship between the apparent heat resistance of Bacillus cereus spores and the pH and NaCl concentration of the recovery medium, International Journal of Food Microbiology, 55 223–27. LEGUERINEL I AND MAFART P (2001) Modelling the influence of pH and organic acid types on thermal inactivation of Bacillus cereus spores, International Journal of Food Microbiology, 63 29–34. LEGUERINEL I, SPEGAGNE I, COUVERT O, GAILLARD S AND MAFART P (2005) Validation of an overall model describing the effect of three environmental actors on the apparent D-value of Bacillus cereus spores, International Journal of Food Microbiology, 100 223–29. LEISTNER L (2000) Basic aspects of food preservation by hurdle technology, International Journal of Food Microbiology, 55 181–86. LEPORQ B, MEMBRÉ J-M, DERVIN C, BUCHE P AND GUYONNET J P (2005) The ‘Sym'Previus’ software, a tool to support decisions to the foodstuff safety, International Journal of Food Microbiology, 100 231–37. KOUTSOUMANIS K AND NYCHAS G J E

© 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

537

LIAO C H AND FETT W F (2005) Resuscitation of acid-injured Salmonella in enrichment broth,

in apple juice and on the surfaces of fresh-cut cucumber and apple, Letters in Applied Microbiology, 41 487–92. LIAO H J, RAO M A AND DATTA A K (2000) Role of thermo-rheological behaviour in simulation of continuous sterilization of a starch dispersion, Food and Bioproducts Processing – Official Journal of the European Federation of Chemical Engineering: Part C, 78 48–56. LINTON R H, CARTER W H, PIERSON M D AND HACKNEY C R (1995) Use of a modified Gompertz equation to model nonlinear survival curves for Listeria monocytogenes Scott A, Journal of Food Protection, 58 946–54. LITTLE C L, ADAMS M R, ANDERSON W A AND COLE M B (1992a) Comparison of a quadratic response surface model and a square root model for predicting the growth rate of Yersinia enterocolitica, Letters in Applied Microbiology, 15 63–68. LITTLE C L, ADAMS M R AND EASTER M C (1992b) The effect of pH, acidulant and temperature on the survival of Yersinia enterocolitica, Letters in Applied Microbiology, 14 148–52. LUND B M (1993) Quantification of factors affecting the probability of development of pathogenic bacteria, in particular Clostridium botulinum, in foods, Journal of Industrial Microbiology, 12 144–55. LUND B M, GRAHAM A F, GEORGE S M AND BROWN D (1990) The combined effect of incubation temperature, pH, sorbic acid on the probability of growth of non-proteolytic, type B Clostridium botulinum, Journal of Applied Bacteriology, 69 481–92. LUND B M AND PECK M W (1994) Heat resistance and recovery of spores of non-proteolytic Clostridium botulinum in relation to refrigerated, processed foods with an extended shelflife, Journal of Applied Bacteriology, Symposium Supplement, 23 115S–28S. LUND B M, WYATT G M AND GRAHAM A F (1985) The combined effect of low temperature and low pH on survival of, and growth and toxin formation from, spores of Clostridium botulinum, Food Microbiology, 4 215–26. MAFART P (2000) Taking injuries of surviving bacteria into account for optimising heat treatments, International Journal of Food Microbiology, 38 25–30. MAFART P (2005) Food engineering and predictive microbiology: On the necessity to combine biological and physical kinetics, International Journal of Food Microbiology, 100 239–51. MAFART P, COUVERT O, GAILLARD S AND LEGUERINEL I (2002) On calculating sterility in thermal preservation methods: Application of the Weibull frequency distribution model, International Journal of Food Microbiology, 72 107–13. MATARAGAS M, DROSINOS E H, SIANA P, SKANDAMIS P AND METAXOPOULOS I (2006a) Determination of the growth limits and kinetic behavior of Listeria monocytogenes in a sliced cooked cured meat product: Validation of the predictive growth model under constant and dynamic temperature storage conditions, Journal of Food Protection, 69 1312–21. MATARAGAS M, DROSINOS E H, VAIDANIS A AND METAXOPOULOS I (2006b) Development of a predictive model for spoilage of cooked cured meat products and its validation under constant and dynamic temperature storage conditions, Journal of Food Science, 71 M157–M67. MAZAS M, GONZALEZ I, LOPEZ M, GONZALEZ J AND SARMIENTO R M (1995) Effects of sporulation media and strain on thermal resistance of Bacillus cereus spores, International Journal of Food Science and Technology, 30 71–78. MAZAS M, LOPEZ M, GONZALEZ I, BERNARDO A AND MARTIN R (1997) Effects of sporulation pH on the heat resistance and the sporulation of Bacillus cereus, Letters in Applied Microbiology, 25 331–34. MAZAS M, LOPEZ M, GONZALEZ I, GONZALEZ J, BERNARDO A AND MARTIN R (1998) Effects of the heating media pH on heat resistance of Bacillus cereus spores, Journal of Food Safety, 18 25–36. MAZAS M, LOPEZ M, MARTINEZ S, BERNARDO A AND MARTIN R (1999) Heat resistance of © 2008, Woodhead Publishing Limited

538

Chilled foods

Bacillus cereus spores: Effect of milk constituents and stabilizing additives, Journal of Food Protection, 62 410–13. MCCARTHY J, HOLBROOK R AND STEPHENS P J (1998) An improved direct plate method for the enumeration of stressed Escherichia coli O157:H7 from food, Journal of Food Protection, 61 1093–97. MCCLURE P J (2001) Predicting the fate of verocytotoxigenic Escherichia coli in foods. In: Duffy, G., Garvey, P., and McDowell, D. A. (eds.), Verocytotoxigenic Escherichia coli, Food and Nutrition Press, Trumbull, Connecticut, USA. MCCLURE P J, BARANYI J, BOOGARD E, KELLY T M AND ROBERTS T A (1993) A predictive model for the combined effect of pH, sodium chloride and storage temperature on the growth of Brochothrix thermosphacta, International Journal of Food Microbiology, 19 161–78. MCCLURE P J, BEAUMONT A L, SUTHERLAND J P AND ROBERTS T A (1997) Predictive modelling of growth of Listeria monocytogenes. The effects on growth of NaCl, pH, storage temperature and NaNO2, International Journal of Food Microbiology, 34 221–32. MCCLURE P J, BLACKBURN C D W, COLE M B, CURTIS P S, JONES J E, LEGAN J D, OGDEN I D, PECK M W, ROBERTS T A, SUTHERLAND J P AND WALKER S J (1994) Modelling growth, survival and death of microorganisms: The UK Food Micromodel approach, International Journal of Food Microbiology, 23 265–75. MCCLURE P J, ROBERTS T A, BARANYI J AND ADAIR C A (2004) Thermal inactivation models – validation and application, Bulletin of the International Dairy Federation, No. 392/2004 111–15. MCCLURE P J AND STEPHENS P (1996) Microbiological methods in safety assessment. In: Ahvenainen, R., Mattila-Sandholm, T., and Ohlsson, T. (eds.), New Shelf-life Technologies and Safety Assessments, Technical Research Centre of Finland, Helsinki, pp. 159–72. MCDONALD K AND SUN D-W (1999) Predictive microbiology for the meat industry: A review, International Journal of Food Microbiology, 52 1–27. MCKELLAR R C AND LU X, eds. (2004) Modeling Microbial Responses in Food. CRC Press, Boca Raton, Florida, USA. MCMEEKIN T A, OLLEY J, RATKOWSKY D A AND ROSS T (2002) Predictive microbiology: Towards the interface and beyond, International Journal of Food Microbiology, 73 395– 407. MCMEEKIN T A, OLLEY J N, ROSS T AND RATKOWSKY D A (1993) Predictive Microbiology: Theory and Application, Research Studies Press, Taunton, UK, 340 p. MCMEEKIN T A AND ROSS T (2002) Predictive microbiology: Providing a knowledge-based framework for change management, International Journal of Food Microbiology, 78 133–53. MEJLHOLM O AND DALGAARD P (2007) Modeling and predicting the growth boundary of Listeria monocytogenes in lightly preserved seafood, Journal of Food Protection, 70 70–84. MEMBRÉ J-M, AMÉZQUITA A, BASSETT J, GIAVEDONI P, BLACKBURN C D W AND GORRIS L G M (2006) A probabilistic modeling approach in thermal inactivation: Estimation of postprocess Bacillus cereus spore prevalence and concentration, Journal of Food Protection, 69 118–29. MEMBRÉ J-M, BASSETT J AND GORRIS L G M (2007) Applying the Food Safety Objective and related standards to thermal inactivation of Salmonella in poultry meat, Journal of Food Protection, 70 2036–44. MEMBRÉ J-M, JOHNSTON M D, BASSETT J, NAAKTGEBOREN G, BLACKBURN C D W AND GORRIS L G M (2005) Application of modelling techniques in the food industry: Determination of shelf-life for chilled foods, Acta Horticulturae (ISHS), 674 407–14. MEMBRÉ J-M, THURETTE J AND CATTEAU M (1997) Modelling the growth, survival and death of Listeria monocytogenes, Journal of Applied Microbiology, 82 345–50. MICHELS M J M AND VISSER F M W (1976) Occurrence and thermoresistance of spores of psychrophilic and psychrotrophic aerobic sporeformers in soli and foods, Journal of Applied Bacteriology, 41 1–11. © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

539

MIKOLAJCIK E M (1970) Thermodestruction of Bacillus spores in milk, Journal of Milk and

Food Technology, 33 61–63. MOATS W A, DABBAH R AND EDWARDS V M (1971) Interpretation of non-logarithmic survivor

curves of heated bacteria, Journal of Food Science, 36 523–26. MOKHTARI A, MOORE C M, YANG H, JAYKUS L A, MORALES R A, CATES S C AND COWEN P

(2006) Consumer-phase Salmonella enterica serovar Enteritidis risk assessment for eggcontaining food products, Risk Analysis, 26 753–68. MONTVILLE R AND SCHAFFNER D (2005) Monte Carlo simulation of pathogen behavior during the sprout production process, Applied and Environmental Microbiology, 71 746–53. MOUREH J AND FLICK D (2004) Airflow pattern and temperature distribution in a typical refrigerated truck configuration loaded with pallets, International Journal of Refrigeration, 27 464–74. MOUSSA-BOUDJEMAA B, GONZALEZ J AND LOPEZ J (2006) Heat resistance of Bacillus cereus spores in carrot extract acidified with different acidulants, Food Control, 17 819–24. MURPHY R Y, DUNCAN L K, DRISCOLL K H, BEARD B L, BERRANG M B AND MARCY J A (2003) Determination of thermal lethality of Listeria monocytogenes in fully cooked chicken breast fillets and strips during postcook in-package pasteurization, Journal of Food Protection, 66 578–83. NACMCF (2005) Considerations for establishing safety-based consume-by date labels for refrigerated ready-to-eat foods, Journal of Food Protection, 68 1761–75. NAHOR N B, HOANGA M L, VERBOVEN P, BAELMANS M AND NICOLAI B M (2005) CFD model of the airflow, heat and mass transfer in cool stores, International Journal of Refrigeration, 28 368-80. NAUTA M J (2000) Separation of uncertainty and variability in quantitative microbial risk assessment models, International Journal of Food Microbiology, 57 9–18. NAUTA M J (2001) A modular process risk model structure for quantitative microbiological risk assessment and its application in an exposure assessment of Bacillus cereus in a REPFED, Available at: http://www.rivm.nl/bibliotheek/rapporten/149106007.html Accessed May 2007. NAUTA M J (2007) Uncertainty and variability in predictive models of micro-organisms in food. In: Brul, S., van Gerwen, S. J. C., and Zwietering, M. H. (eds.), Modelling Microorganisms in Food, Woodhead Publishing Limited, Cambridge, England, pp. 44– 66. NAUTA M J, LITMAN S, BARKER G C AND CARLIN F (2003) A retail and consumer phase model for exposure assessment of Bacillus cereus, International Journal of Food Microbiology, 83 205–18. NEUMEYER K, ROSS T AND MCMEEKIN T A (1997a) Development of a predictive model to describe the effects of temperature and water activity on the growth of spoilage pseudomonads, International Journal of Food Microbiology, 38 45–54. NEUMEYER K, ROSS T, THOMSON G AND MCMEEKIN T A (1997b) Validation of a model describing the effects of temperature and water activity on the growth of psychrotrophic pseudomonads, International Journal of Food Microbiology, 38 55–63. NI H, DATTA A K AND TORRANCE K E (1999) Moisture transport in intensive microwave heating of biomaterials: A multiphase porous media model, International Journal of Heat and Mass Transfer, 42 1501–12. NICOLAI B M, VAN IMPE J F, VERLINDEN B, MARTENS T, VANDEWALLE J AND DE BAERDEMAEKER J (1993) Predictive modelling of surface growth of lactic acid bacteria in vacuum-packed meat, Food Microbiology, 10 229–38. NORTON T AND SUN D-W (2006) Computational Fluid Dynamics (CFD) - an effective and efficient design and analysis tool for the food industry: A review, Trends in Food Science and Technology, 17 600–20. OLMEZ H K AND ARAN N (2005) Modeling the growth kinetics of Bacillus cereus as a function of temperature, pH, sodium lactate and sodium chloride concentrations, International Journal of Food Microbiology, 98 135–43. © 2008, Woodhead Publishing Limited

540

Chilled foods

(1991) Electrical conductivities of selected solid foods during ohmic heating, Journal of Food Process Engineering, 14 221–36. PALOP A, MARCO A, RASO J, SALA F J AND CONDON S (1997) Survival of heated Bacillus coagulans spores in a medium acidified with lactic or citric acid, International Journal of Food Microbiology, 38 25–30. PATANKAR S V (1980) Numerical heat transfer and fluid flow, Hemisphere Publishing Corporation, Washington, D.C. PECK M W, BARANYI J, PLOWMAN J, LEMARC Y AND ALDUS C F (2004) Improved control of Clostridium perfringens, Final report of Food Standards Agency Project B14009. PECK M W AND FERNANDEZ P S (1995) Effect of lysozyme concentration, heating at 90 °C, and then incubation at chilled temperatures on growth from spores of non-proteolytic Clostridium botulinum, Letters in Applied Microbiology, 21 50–54. PECK M W, LUND B M, FAIRBAIRN D A, KASPERSSON A S AND UNDELAND P C (1995) Effect of heat treatment on survival of, and growth from, spores of nonproteolytic Clostridium botulinum at refrigeration temperatures, Applied and Environmental Microbiology, 61 1780–85. PELEG M AND COLE M B (1998) Reinterpretation of microbial survival curves, Critical Reviews in Food Science and Nutrition, 38 353–80. PFLUG I J AND ODLAUG T E (1986) Biological indicators in the pharmaceutical and medical device industry, Journal of Parenteral Science and Technology, 40 242–48. PIN C, BARANYI J AND GARCIA DE FERNANDO G (2000) Predictive model for the growth of Yersinia enterocolitica under modified atmospheres, Journal of Applied Microbiology, 88 521–30. POSCHET F, GEERAERD A H, SCHEERLINCK N, NICOLAI B M AND VAN IMPE J F (2003) Monte Carlo analysis as a tool to incorporate variation on experimental data in predictive microbiology, Food Microbiology, 20 285–95. PRADHAN A K, LI Y, MARCY J A, JOHNSON M G AND TAMPLIN M L (2007) Pathogen kinetics and heat and mass transfer-based predictive model for Listeria innocua in irregularshaped poultry products during thermal processing, Journal of Food Protection, 70 607–15. RAVIYAN P, TANG J, ORELLANA L AND RASCO B (2003) Physicochemical properties of a time– temperature indicator based on immobilization of Aspergillus oryzae α-amylase in polyacrylamide gel as affected by degree of cross-linking agent and salt content, Journal of Food Science, 68 2302–08. RIEU E, DUHEM K, VINDEL E AND SANAA M (2007) Food safety objectives should integrate the variability of the concentration of pathogen, Risk Analysis, 27 373–86. ROBERTS T A, GILBERT R J AND INGRAM M (1966) The effect of sodium chloride on heat resistance and recovery of heated spores of Clostridium perfringens (3679/S2), Journal of Applied Bacteriology, 29 549–55. ROBINSON T P, ABOABA O O, OCIO M J, BARANYI J AND MACKEY B M (2001) The effect of inoculum size on the lag phase of Listeria monocytogenes, International Journal of Food Microbiology, 70 163–73. RODRIGO F, RODRIGO M C AND MARTINEZ A (1998) Evaluation of a new time–temperature integrator in pilot plant conditions, Food Research and Technology, 206 184–88. ROSS T (1996) Indices for performance evaluation of predictive models in food microbiology, Journal of Applied Bacteriology, 81 501–08. ROSS T, DALGAARD P AND TIENUNGOON S (2000) Predictive modelling of the growth and survival of Listeria in fishery products, International Journal of Food Microbiology, 62 231–45. ROSS T AND MCMEEKIN T A (2003) Modeling microbial growth within food safety risk assessments, Risk Analysis, 23 179–97. ROSS T AND SUMNER J (2002) A simple, spreadsheet-based, food safety risk assessment tool, International Journal of Food Microbiology, 77 39–53. ROSSET P, CORNU M, NOEL V, MORELLI V AND POUMEYROL G (2004) Time–temperature PALANIAPPAN S AND SASTRY S K

© 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

541

profiles of chilled ready-to-eat foods in school catering and probabilistic analysis of Listeria monocytogenes growth, International Journal of Food Microbiology, 96 49– 59. SASTRY S K (1986) Mathematical evaluation of process schedules for aseptic processing of low-acid foods containing discrete particles, Journal of Food Science, 51 1323–28. SASTRY S K AND CORNELIUS B D (2002) Aseptic processing of foods containing particulates, John Wiley and Sons, Inc., New York. SCHAFFNER D, SITHOLE S AND MONTVILLE R (2004b) Use of microbial modeling and Monte Carlo simulation to determine microbial performance criteria on plastic cutting boards in use in foodservice kitchens, Food Protection Trends, 24 14–19. SCHAFFNER D W, MCENTIRE J, DUFFY S, MONTVILLE R AND SMITH S (2004a) Monte Carlo simulation of the shelf life of pasteurized milk as affected by temperature and initial concentration of spoilage organisms, Food Protection Trends, 23 1014–21. SCHULTZE K K, LINTON R H, COUSIN M A, LUCHANSKY J B AND TAMPLIN M L (2006) A predictive model to describe the effects of temperature, sodium lactate, and sodium diacetate on the inactivation of a serotype 4b strain of Listeria monocytogenes in a frankfurter slurry, Journal of Food Protection, 69 1552–60. SEMAN D L, BORGER A C, MEYER J D, HALL P A AND MILKOWSKI A L (2002) Modeling the growth of Listeria monocytogenes in cured ready-to-eat processed meat products by manipulation of sodium chloride, sodium diacetate, potassium lactate, and product moisture content, Journal of Food Protection, 65 651–58. SERP D, VON STOCKAR U AND MARISON I W (2002) Immobilized bacterial spores for use as bioindicators in the validation of thermal sterilization processes, Journal of Food Protection, 65 1134–41. SHEHATA T E AND COLLINS E B (1972) Sporulation and heat resistance of psychrophilic strains of Bacillus, Journal of Dairy Science, 55 1405–09. SHULER C, DRENZEK C, LANCE S, GONZALEZ G, MILLER J, TOBIN-D’ANGELO M, GABEL J, BURNETT C, SOBEL J, JULIAO P, MASLANKA S, WIERSMA P AND SHETH A (2006) Botulism associated with commercial carrot juice – Georgia and Florida, September 2006, Morbidity and Mortality Weekly Report, 55 1098–99. SKANDAMIS P N, DAVIES K W, MCCLURE P J, KOUTSOUMANIS K AND TASSOU C (2002) A vitalistic approach for non-thermal inactivation of pathogens in traditional Greek salads, Food Microbiology, 19 405–21. SMALE N J, MOUREH J AND CORTELLA G (2006) A review of numerical models of airflow in refrigerated food applications, International Journal of Refrigeration, 29 911–30. SMELT J P P M, HELLEMONS J C, WOUTERS P C AND VAN GERWEN S J C (2002a) Physiological and mathematical aspects in setting criteria for decontamination of foods by physical means, International Journal of Food Microbiology, 78 57–77. SMELT J P P M, OTTEN G D AND BOS A P (2002b) Modelling the effect of sublethal injury on the distribution of lag times of individual cells of Lactobacillus plantarum, International Journal of Food Microbiology, 73 207–12. SMITH-SIMPSON S AND SCHAFFNER D W (2005) Development of a model to predict growth of Clostridium perfringens in cooked beef during cooling, Journal of Food Protection, 68 336–41. SMITH S AND SCHAFFNER D W (2004) Evaluation of a Clostridium perfringens predictive model developed under isothermal conditions in broth to predict growth in ground beef during cooling, Applied and Environmental Microbiology, 70 2728–33. STENFORS L P AND GRANUM P E (2001) Psychrotolerant species from Bacillus cereus are not necessarily Bacillus weihenstephanensis, FEMS Microbiology Letters, 197 223–28. STEPHENS P J, COLE M B AND JONES M V (1994) Effect of heating rate on the thermal inactivation of Listeria monocytogenes, Journal of Applied Bacteriology, 77 702–08 SUN D-W AND WANG L (2006) Development of mathematical model for vacuum cooling of cooked meats, Journal of Food Engineering, 77 379–85. SUTHERLAND J P, AHERNE A AND BEAUMONT A L (1996) Preparation and validation of a © 2008, Woodhead Publishing Limited

542

Chilled foods

growth model for Bacillus cereus: The effects of temperature, pH, sodium chloride and carbon dioxide, International Journal of Food Microbiology, 30 359–72. SWINNEN I A M, BERNAERTS K, DENS A H, GEERAERD A H AND VAN IMPE J F (2004) Predictive modelling of the microbial lag phase: A review, International Journal of Food Microbiology, 94 137–59. TAKUMI K, DE JONGE K AND HAVELAAR A (2000) Modelling inactivation of Escherichia coli by low pH: Application to passage through the stomach of young and elderly people, Journal of Applied Microbiology, 69 935–43. TAOUKIS P S AND LABUZA T P (1989) Applicability of time–temperature integrators as shelf life monitors of food products, Journal of Food Science, 54 789–91. TATTIYAKUL J, RAO M A AND DATTA A K (2001) Simulation of heat transfer to a canned corn starch dispersion subjected to axial rotation, Chemical Engineering and Processing, 40 391–99. THORSEN L, HANSEN B M, NIELSEN K F, HENDRIKSEN N B, PHIPPS R K AND BUDDE B B (2006) Characterization of emetic Bacillus weihenstephanensis, a new cereulide-producing bacterium, Applied and Environmental Microbiology, 72 5118–21. TRUJILLO F J AND TUAN PHAM Q (2006) A computational fluid dynamic model of the heat and moisture transfer during beef chilling, International Journal of Refrigeration, 29 998– 1009. TUCKER G (1999) A novel validation method: Application of time–temperature integrators to food pasteurization treatments, Transactions of the Institution of Chemical Engineers, 77 223–31. TUCKER G S, LAMBOURNE T, ADAMS J B AND LACH A (2002) Application of a biochemical time–temperature integrator to estimate pasteurisation values in continuous food processes, Innovative Food Science and Emerging Technologies, 3 165–74. USDA (1999) Performance standards for the production of certain meat and poultry products. Final rule, Federal Register, 64 732–49. USDA (2000) Predicting human dose-response relationships from multiple biological models: Issues with Cryptosporidium parvum, USDA Center at Riverside Riverdale, Md. A public meeting sponsored by the Risk Assessment Consortium. Available at: http:// www.foodsafety.gov/~comm/racconf.html, Accessed June 2007. USDA (2005) A risk assessment for Clostridium perfringens in ready-to-eat and partially cooked meat and poultry products, Available at: http://www.fsis.usda.gov/PDF/ CPerfringens_Risk_Assess_Sep2005.pdf, Accessed April 2007. USDA-FSIS (2003a) Control of Listeria monocytogenes in ready-to-eat meat and poultry products; final rule, 9 CFR Part 430. Available at: http://www.fsis.usda.gov/OPPDE/ rdad/FRPubs/97-013F.pdf, Accessed June 2007. USDA-FSIS (2003b) Risk Assessment for Listeria monocytogenes in Deli Meats, Available at: http://www.fsis.usda.gov/OPPDE/rdad/FRPubs/97-013F/ListeriaReport.pdf, Accessed June 2007. USDA-FSIS (2005) Meat and poultry hazards and controls guide, Available at: http:// www.fsis.usda.gov/OPPDE/rdad/FSISDirectives/5100.2/Meat_and_Poultry_Hazards _Controls_Guide_10042005.pdf, Accessed June 2007. VALDRAMIDIS V P, BELAUBRE N, ZUNIGA R, FOSTER A M, HAVET M, GEERAERD A H, SWAIN M J, BERNAERTS K, VAN IMPE J F AND KONDJOYAN A (2005) Development of predictive modelling approaches for surface temperature and associated microbiological inactivation during hot dry air decontamination, International Journal of Food Microbiology, 100 261–74. VALDRAMIDIS V P, GEERAERD A H, GAZE J E, KONDJOYAN A, BOYD A R, SHAW H L AND VAN IMPE J F (2006) Quantitative description of Listeria monocytogenes inactivation kinetics with temperature and water activity as the influencing factors; model prediction and methodological validation on dynamic data, Journal of Food Engineering, 76 79–88. VAN GERWEN S J C AND GORRIS L G M (2004) Application of elements of microbiological risk © 2008, Woodhead Publishing Limited

Predicting the behaviour of micro-organisms in chilled foods

543

assessment in the food industry via a tiered approach, Journal of Food Protection, 67 2033–40. VAN IMPE J F, NICOLAI B M, MARTENS T, DE BAERDEMAEKER J AND VANDEWALLE J (1992) Dynamic mathematical models to predict microbial growth and inactivation during food processing, Applied and Environmental Microbiology, 58 2901–09. VAN LOEY A, HENDRICKX M, LUDIKHUYZE L, WEEMAES C, HAENTJENS T, DE CORDT S AND TOBBACK P (1996) Potential Bacillus subtilis alpha-amylase-based time–temperature integrators to evaluate pasteurization processes, Journal of Food Protection, 59 261–67. VAN LOEY A M, ARTHAWAN A, HENDRICKX M E, HAENTJENS T H AND TOBBACK P P (1997) The development and use of an α-amylase based time-temperature integrator to evaluate inpack pasteurisation processes, Lebensmittel-Wissenschaft und-Technologie, 30 94–100. VEREECKEN K M, DEVLIEGHERE F, BOCKSTAELE A, DEBEVERE J AND IMPE J F V (2003) A model for lactic acid-induced inhibition of Yersinia enterocolitica in mono- and coculture with Lactobacillus sakei, Food Microbiology, 20 701–13 VIRTO R, SANZ D, ALVAREZ I, CONDON S AND RASO J (2005) Inactivation kinetics of Yersinia enterocolitica by citric and lactic acid at different temperatures, International Journal of Food Microbiology, 103 251–57. VIRTO R, SANZ D, ALVAREZ I, CONDON S AND RASO J (2006) Application of the Weibull model to describe inactivation of Listeria monocytogenes and Escherichia coli by citric acid and lactic acid at different temperatures, Journal of the Science of Food and Agriculture, 86 865–70. VON STETTEN F, FRANCIS K P, LECHNER S, NEUHAUS K AND SCHERER S (1998) Rapid discrimination of psychrotolerant and mesophilic strains of the Bacillus cereus group by PCR targeting on 16S rDNA, Journal of Microbiological Methods, 34 99–106. VOSE D J (1998) The application of quantitative risk assessment to microbial food safety, Journal of Food Protection, 61 640–48. WANG L AND SUN D-W (2002) Evaluation of performance of slow air, air blast, and water immersion cooling in the cooked meat industry by the finite element method, Journal of Food Engineering, 51 329–40. WANG L AND SUN D-W (2003) Recent developments in numerical modelling of heating and cooling processes in the food industry – A review, Trends in Food Science and Technology, 14 408–23. WEI Q K, FANG T J AND CHEN W C (2001) Development and validation of growth model for Yersinia enterocolitica in cooked chicken meats packaged under various atmosphere packaging and stored at different temperatures, Journal of Food Protection, 64 987–93. WHITING R C (1993) Modeling bacterial survival in unfavorable environments, Journal of Industrial Microbiology, 12 240–46. WHITING R C AND BUCHANAN R L (1993) A classification of models for predictive microbiology, Food Microbiology, 10 175–77. WHITING R C AND CYGNAROWICZ-PROVOST M (1992) A quantitative model for bacterial growth and decline, Food Microbiology, 9 269–77. WIJTZES T, MCCLURE P J, ZWIETERING M H AND ROBERTS T A (1993) Modelling bacterial growth of Listeria monocytogenes as a function of water activity, pH and temperature, International Journal of Food Microbiology, 18 139–49 WILSON P D G, BROCKLEHURST T F, ARINO S, THUAULT D, JAKOBSEN M, LANGE M, FARKAS J, WIMPENNY J W T AND VAN IMPE J F (2002) Modelling microbial growth in structured foods: Towards a unified approach, International Journal of Food Microbiology, 73 275– 89. WILSON P D G, WILSON D R AND WASPE C R (2000) Weak acids: Dissociation in complex buffering systems and partitioning into oils, Journal of the Science of Food and Agriculture, 80 471–76. XIONG R, XIE G, EDMONDSON A S AND MEULLENET J-F (2002) Neural network modelling of the fate of Salmonella enterica serovar Enteritidis PT4 in home-made mayonnaise prepared with citric acid, Food Control, 13 525–33. © 2008, Woodhead Publishing Limited

544

Chilled foods

YANG H, MOKHTARI A, JAYKUS L A, MORALES R A, CATES S C AND COWEN P (2006) Consumer

phase risk assessment for Listeria monocytogenes in deli meats, Risk Analysis, 26 89–103. YOUSEF A E, GAJEWSKII R J, II AND MARTH E H (1991) Kinetics of growth and inhibition of

Listeria monocytogenes in the presence of antioxidant food additives, Journal of Food Science, 56 10–13. ZHANG H AND DATTA A K (2000) Coupled electromagnetics and heat transfer of microwave oven heating, Journal of Microwave Power and Electromagnetic Energy, 35 71–85 ZULIANI V, LEBERT I, GARRY P, VENDEUVRE J-L, AUGUSTIN J-C AND LEBERT A (2006) Effects of heat-processing regime, pH, water activity and their interactions on the behaviour of Listeria monocytogenes in ground pork. Modelling the boundary of the growth/no-growth areas as a function of pH, water activity and temperature, International Journal of Food Science and Technology, 41 1197–206. ZWIETERING M H, JONGENBURGER I, ROMBOUTS F M AND VAN’T RIET K (1990) Modeling the bacterial growth curve, Applied and Environmental Microbiology, 56 1975–881. ZWIETERING M H, WIT J C D AND NOTERMANS S (1996) Application of predictive microbiology to estimate the number of Bacillus cereus in pasteurised milk at the point of consumption, International Journal of Food Microbiology, 30 55–70.

© 2008, Woodhead Publishing Limited

18 Conventional and rapid analytical microbiology J. McClure, Unilever, UK

18.1 Introduction The analysis of foods for the presence of micro-organisms is standard practice for monitoring that food safety and quality targets have been met and it hence plays a critical role in safe food production and the establishment, verification and validation of HACCP procedures. The food microbiologist is faced with a challenging task, often being required to isolate low numbers of potential pathogens or spoilage micro-organisms from food materials and equipment environments, which are complex in terms of location, and physical and chemical composition. The micro-organisms contained may be injured by processes such as heating or freezing, or product formulation features such as low pH and chemical preservation. The food matrix itself (e.g. a multi-component or fatty structure) and the metabolic state of the micro-organisms (e.g. stress or starvation) can significantly affect recovery and enumeration by any given analytical procedure. Microbiological tests carried out on foods can be described as either quantitative or qualitative, and detect micro-organisms or their products (e.g. toxins). In quantitative testing, specific micro-organisms in a defined sample (e.g. 100 g) or number of samples are counted, whereas in a qualitative test, the presence of specific micro-organisms (or their toxins) is detected, again in a defined sample. The qualitative approach is also known as ‘presence/absence’ testing. Both approaches can be applied to testing for pathogens, spoilage or indicator organisms. The term ‘indicator organisms’ can refer to any group of micro-organisms © 2008, Woodhead Publishing Limited

546

Chilled foods

whose presence or absence can provide information related to the history of the sample and whose presence indicates an increased risk of the presence of pathogens. For example, enterobacteriaceae counts are frequently used as indicators of inadequate processing of food and water, post-process contamination or as an index for the possible presence of enteric pathogens, such as salmonella. Chilled foods and their raw materials are likely to contain a wide variety of micro-organisms, either as part of the normal food flora or as a result of contamination during processing or handling. From a safety point of view, infectious pathogens are of most concern in these foods; these include Salmonella, E. coli O157:H7 and Listeria monocytogenes (Brown, 2000). Commercially, the presence and numbers of spoilage micro-organisms is also of high importance. Pathogens and spoilage micro-organisms may originate from various raw materials and may survive for long periods in chilled products if they are not destroyed during processing, usually by heating. Psychrotrophic pathogenes, such as L. monocytogenes, are problematic from a food safety view, as they have been shown to grow at temperatures as low as –0.1 °C (Walker, 1990) and this means that they may grow in products or become established in manufacturing and storage areas. Many other micro-organisms are considered as potential safety hazards of chilled foods, including cold-growing non-proteolytic Clostridum botulinum and other toxin producers such as Bacillus cereus and Staphylococcus aureus, as the processes used in the manufacture of chilled foods may not inactivate spores or toxins. The combination of consumer demand for higher quality, less preserved and safer foods and the food manufacturers’ and retailers’ desire for efficient and profitable performance has highlighted the need for faster and more accurate analytical methods than ever before. As a result, there have been remarkable advances in recent years in rapid or alternative methods for use in food microbiology. This area of methods is vast in its own right and, together with the broad range of micro-organisms potentially of concern to the chilled foods industry, makes it unfeasible to comment on all available methods in this chapter. The information given here is therefore by no means exhaustive and the aim is to give an overview of the basis of both conventional and rapid methods, together with examples of technologies and systems frequently used today. Suggestions for further reading are given in Section 18.8 together with details of manufacturers of commonly used methods.

18.2 Sampling Sampling is a critical element of food safety management. A sample must be as representative as possible of the material being examined and it must be taken without any contamination occurring. Sample handling, between the time it is taken and the time it is analysed, should prevent microbiological multiplication or death. Wherever possible, appropriate, statistical principles must be applied to the sampling process. © 2008, Woodhead Publishing Limited

Conventional and rapid analytical microbiology

547

18.2.1 Sampling for trend analysis The most efficient approach to sampling for trend analysis is to devise a sampling plan in which results are interpreted from a series of analyses rather than a single result. A HACCP plan is the tool that helps identify points in a process that play a critical role in ensuring product safety, sets appropriate control limits and ensures that they are under control by appropriate monitoring and verification. Sampling of incoming raw materials can be used as part of this control scheme and a statistically based sampling plan can aid in the decision to accept the materials. Environmental sampling plans can also help minimise the possibility of cross-contamination from the environment to the food, and sampling of finished product can verify that the process is under control. Once a baseline is established, a typical plan can then stipulate increased sampling when results indicate a site is significantly above the baseline limit. Statistical analysis of historical data can establish whether a process is capable of the required performance. It is now common for microbiologists to use two or three class sampling plans, in which the number of individual samples to be tested from one batch is specified, together with microbiological limits. More information regarding these types of sampling plans can be found in Anon. (1986).

18.2.2 Sampling for investigation Sampling procedures used during an investigation can be very different from those typically used during sampling for trend analysis. If an investigation is in response to information that a problem already exists, it is probable that the source of the problem needs to be identified. This process can be very lengthy. Targeted sampling and not random sampling should be used and as much information as possible should be collated regarding the food type, process operations, equipment design, information and visual inspection together with microbiological knowledge of the micro-organism(s) of concern. A sampling protocol should be defined and should include:

• lot intergrity (relationship between lots, batch and coding, interpretation of codes, maintenance of lot integrity)

• sample type (use of multiple level samples, role of and source of controls, commercial impact of sampling protocol)

• sample handling (speed and cost of investigation, storage and handling, sample coding protocol). For chilled foods, it is particularly important to consider that most will not be homogeneous mixtures but layers or sections, e.g. a prepared sandwich. It is therefore important that samples of each component of a food should be collected. In these cases, a decision needs to be made if the analysis will be carried out on each component or the product as a whole.

© 2008, Woodhead Publishing Limited

548

Chilled foods

18.3 Analytical microbiology 18.3.1 Considerations The first stage of an analytical microbiological method is taking or obtaining a representative sample. The type of sample will depend on the purpose of the analysis, perhaps investigation of a product failure or routine monitoring. Samples are usually transported under controlled conditions, especially time and temperature, and in containers designed to prevent contamination. On receipt, samples are usually homogenised aseptically in either the primary growth medium to be used or a diluent (e.g. maximum recovery diluent [MRD]), typically in a ratio of 1:10. Even at this initial stage, there are three major considerations: the type of food being tested, the formulation and use of any diluents, and the formulation and use of any primary growth media. Food types As outlined in the introduction, if the food being tested has been processed, it is possible that the micro-organisms present may be in an injured or stressed state. In this state, they are more sensitive to the inhibitory agents in selective media, and direct incubation, for example, at 37 or 41 °C can significantly inhibit their growth and therefore detection (Blackburn and McCarthy, 2000; Stephens and Joynson, 1998). Therefore, analysis of certain food types may require an additional pretreatment or pre-enrichment procedure under non-selective conditions, or a prolonged resuscitation period, before being subjected to selective media. Diluents The use of diluents of unsuitable pH, buffering capacity or osmotic strength can significantly inhibit the recovery and growth of some micro-organisms, particularly those that are sub-lethally injured. Equally, injury may be induced by the use of an unsuitable diluent. Suitable diluents for chilled foods include MRD, buffered peptone water (BPW) and peptone salt diluent (PSD). The formulation and use of diluents is an important consideration when analysing dried (low water activity) and low pH foods. Samples of dried foods should be hydrated gradually to avoid the effect of osmotic shock, which can be lethal to some micro-organisms, and low pH foods often require a diluent with a high buffering capacity to prevent adverse changes in the pH of the homogenate. Samples should not be left in a diluent for a length of time that that could allow growth prior to analysis, unless only qualitative methods are used. Media formulations Many different media are available commercially and their performance and application is reviewed in many suppliers’ handbooks (e.g. Oxoid, bioMérieux and Merck KGaA). Methods designed to detect the total number of viable microorganisms in a sample, such as the total viable count (TVC) or aerobic plate count (APC), generally involve the use of non-selective nutrient media and potential problems related to the resuscitation and growth of injured micro-organisms is not © 2008, Woodhead Publishing Limited

Conventional and rapid analytical microbiology

549

a concern. However, the detection of pathogens in foods requires the use of selective techniques in order to differentiate the target micro-organisms from any other micro-organisms present. A typical conventional pathogen detection method includes: (i) a pre-enrichment stage, (ii) a selective enrichment stage, and (iii) an isolation or detection stage on agar. Many rapid method procedures also require some form of enrichment in a selective medium, to inhibit the growth of competing micro-organisms and allow the target micro-organisms to reach the detection level of the method. The selective process itself often involves preventing or slowing the growth of unwanted types by the use of ‘selective agents’. These agents are added to media, to preferentially, or selectively, allow the growth of the target micro-organisms. Selective agents can include inorganic salts (e.g. sodium azide and lithium chloride), dyes (e.g. crystal violet and malachite green), surface-active agents (e.g. bile and lauryl sulphate), and antibiotics (e.g. polymyxin and amphotericin). Specific culture conditions, such as incubation temperatures or atmospheric conditions (aerobiosis, anaerobiosis and microaerobiosis) can also be used for selection, either solely or in conjunction with selective agents (Stephens, 2003). Components are also used to allow the differentiation of the target species from other species, by using differences in their biochemistry to provide recognisable changes (e.g. pH change or precipitation of some media components). A differential procedure is often employed at the detection stage (e.g. on agar) by incorporating indicator dyes and substrate indicators (e.g. chromogens and fluorogens) into the media. Examples of such dyes are phenol red, neutral red and bromocresol purple, which are added to indicate a change in pH due to growth of micro-organisms. This approach is used in the formulation of Brilliant Green Agar, which is a commonly used medium for the detection of Salmonella spp. Typical colonies appear pink/red, surrounded by a brilliant red zone caused by phenol red indicating a shift in pH. More recently, detection of specific enzymes (e.g. β-galactosidase and β-glucuronidase) has led to the development of a great number of new media (e.g. SMID, bioMérieux and OSCA, Oxoid), particularly for pathogen detection and identification. These media involve the use of chromogens and fluorogens, which are compounds that act as substrates for specific enzymes. The chromogen-based systems incorporate a specific enzyme substrate, such as a sugar or amino acid, linked to a chromophore. The compound remains colourless until utilised, when the chromophore is released resulting in a colour change in the medium. Fluorogens are similar in principle, but after enzyme action release a fluorophore which can be detected using UV light. These types of media have been shown to have greater specificity than those based on changes in pH (Baylis and Patrick, 1999). Numerous chromogenic-based media are now commercially available for a wide range of micro-organisms including Salmonella, E. coli O157, Campylobacter and Listeria. These types of media are becoming increasingly accepted and the formulation of agar Listeria according to Ottaviani and Agosti (ALOA) is now included in the International Standard method for enumeration of L. monocytogenes (Anon., 1997). © 2008, Woodhead Publishing Limited

550

Chilled foods

Diagnostic features The physiological features or virulence factors of specific types can also be exploited as diagnostic or identification tools. Some examples of this include the demonstration of phospholipase activity by inclusion of egg yolk emulsion to distinguish Staphylococcus aureus from the other staphylococci (Baird-Parker, 1962) and lecithin (Chrisope et al., 1976) and coagulase activity by inclusion of rabbit plasma fibrinogen (Beckers et al., 1984). Indeed, for some micro-organisms, the detection of a virulence factor is the optimal approach for meaningful identification. For example, although several serotypes of verotoxigenic E. coli (VTEC) have been associated with outbreaks, most infections have been caused by the serotype O157:H7 (Thomas et al., 1993). Hence in many laboratories, E. coli O157:H7 is the only serotype that is routinely examined for and as a result, the general conception has arisen that this is the only serotype of VTEC of concern (McCarthy and Blackburn, 1999). Detection methods for E. coli O157:H7 have consequently been developed based on the characteristic features of this organism and are usually carried out by either determining its lack of fermentation of sorbitol within 24 hours or by the detection of the ‘O’ somatic antigen. These procedures can result in other verotoxin-producing organisms being undetected. The production of verotoxin is the one common factor and principle virulence determinant in VTEC, and it has been suggested that methods should be directed more towards the detection of the toxins, or genes that code for them, rather than the organism itself.

18.3.2 Novel approaches Novel enrichments The selective enrichment process and the composition of any media used during the process are critical considerations in any microbiological analytical method. Enrichment is used to increase numbers of the target micro-organism, suppress other competing micro-organisms and allow time for typical physiology to develop or be expressed. Ensuring injured cells have the correct conditions and time to resuscitate and reach levels where they will be detected (by whatever means) is often the reason methods can be time consuming. The need to reduce total test time has led to much research and development on novel enrichment formulations, and separation and concentration techniques to either speed up resuscitation or reduce detection limits. One example of a novel enrichment procedure is the universal preenrichment broth (UP) developed by Bailey and Cox (1992). The UP was originally developed for simultaneous detection of Salmonella spp. and Listeria spp. in foods. The medium is highly buffered, low in carbohydrates and allows for the resuscitation and multiplication of sub-lethally heat-injured cells. It does not contain any selective agents and therefore will not necessarily suppress the growth of any competing flora. Investigations into use of UP as a pre-enrichment broth for culturing heat-injured E. coli O157:H7, in addition to Salmonella spp. and Listeria spp. found the broth allowed growth of low levels (≤125 cfu/sample) of each pathogen to at least 104 cfu/ml within 24 hours of incubation at 37 °C (Zhao and © 2008, Woodhead Publishing Limited

Conventional and rapid analytical microbiology

551

Doyle, 2001). This study also emphasised the importance of the incubation time. An incubation period of 6 hours was insufficient to produce cell populations at (or above) typical detection levels of many rapid methods, e.g. the enzyme-linked immunosorbent assay (ELISA). This is an important consideration when using a rapid detection method as some manufacturers advocate this relatively short incubation time in their enrichment protocols. Another example of a novel enrichment is the Simple Pre-enrichment and Rapid Isolation New Technology (SPRINT), manufactured by Oxoid. The SPRINT is a more recent development in enhanced enrichment culture and is available for use with Salmonella spp. The principles of the system are based on traditional techniques, allowing a pre-enrichment in BPW followed by selective enrichment in Rappaport-Vassiliadis (RV), in one 24-hour step. The formulation of the BPW is designed to enhance recovery of injured cells before addition of the selective agent after 6 hours of incubation, by means of timed release capsules. The SPRINT has been shown to significantly improve the rate of detection of low numbers of injured salmonellae in ice cream and milk powder, after 24 hours of enrichment (Baylis et al., 2000). Separation and concentration Recent years have also given rise to considerable interest in separation and concentration techniques, the aim being increase the number of micro-organisms per unit volume. In theory, application of these techniques can shorten a selective enrichment incubation time and remove potentially interfering properties of the food matrix. Different separation and concentration methods include those based on filtration, centrifugation, phase separation, electrophoresis and immunology. An excellent review of the different approaches is given by Betts (1994). Immunomagnetic separation (IMS) also referred to as immunocapture, is probably the most widely used technique and many commercially developed kits are now available (e.g. EHEC IMS, Denka Seiken). Their format consists of magnetic particles, or beads, coated with specific antibodies. The particles can be added to a food homogenate or enrichment broth and, if present, target cells will attach to the particles. A magnetic field is applied to retain the particles and captured cells, allowing any food debris and excess liquid to be discarded. A number of companies produce IMS beads and kits that incorporate an IMS stage for various pathogens including Salmonella, Listeria, Campylobacter and E. coli O157. The technique has proved very successful and is now included in some standard reference methods (Anon., 2001b).

18.4 Conventional microbiological techniques Conventional cultural methods used for detecting micro-organisms in food are well established and can be applied to both quantitative and qualitative testing. These types of methods rely on the growth of micro-organisms in one or more nutrient media, and are designed in such a way that any growth causes a change in © 2008, Woodhead Publishing Limited

552

Chilled foods

the properties of the medium, which can be measured (e.g. by visual assessment of turbidity or colour). They can also include some form of isolation stage (e.g. colonies on agar plates) and, in some cases, subsequent confirmation of isolates (e.g. by biochemical and/or serological tests). Conventional techniques require trained, skilled technicians, preferably operating to some form of quality system such as the United Kingdom Accreditation Scheme (UKAS). It is also very beneficial for a laboratory to participate in an external quality assessment (EQA) or other type of proficiency scheme designed for use with conventional techniques. Participation in an EQA will help ensure the reliability of data generated, allow the laboratory to measure its performance against other laboratories, encourage good performance and raise an interest in quality assurance. Performance testing also aids quality assurance of media, methods, operators and equipment (e.g. balances, incubators, water baths and pipettes).

18.4.1 Conventional quantitative methods The number of viable micro-organisms in a sample can be counted by using the pour plate method or the spread plate method; or the number can be estimated by using the broth-based most probable number (MPN) method. Pour plate method In the pour plate technique, diluent containing the sample (typically 1 ml) is added to a sterile Petri dish. Approximately 12–15 ml of a molten (< 45 °C), tempered agar-based medium is then poured into the dish and the sample and medium are mixed. The agar is then allowed to solidify before being incubated for a specific time and temperature, depending on the type of test. During incubation, microorganisms derived from a single cell (or in some cases a clump of cells) form discrete colonies in or on the medium. These are commonly referred to as colony forming units or cfu. They are counted and the number of micro-organisms in the inoculum can be calculated. Various types of agar media can be used to enumerate different types of micro-organisms, based on the selective and differential approaches described in Section 18.3. Spread plate method In the spread plate method, solidified agar-based media in sterile Petri dishes are prepared in advance and may or may not be dried. Volumes of inoculum (typically 100 µl for standard 90 mm Petri dish) are deposited and spread across the surface of the agar to form a thin film. When the inoculum has absorbed into the medium, the Petri dishes are incubated for specific times and temperatures (as in the pour plate method), depending on the type of test. Microorganisms form discrete colonies on the surface of the agar, which can be counted and they may, or may not, cause characteristic changes to agar surrounding the colonies. Where low numbers are expected, larger Petri dishes can be used to spread up to 1 ml of sample. There are some advantages to using the spread plate method, one being that as © 2008, Woodhead Publishing Limited

Conventional and rapid analytical microbiology

553

colonies are formed on the surface of the agar, they are easier to sub-culture for confirmation tests. Also, as pre-poured solidified agar is used, there is no risk of exposure of the micro-organisms to heat stress, potentially caused by molten agar of too high a temperature. However, the disadvantage is the lower detection limit when compared to the pour plate method and some suggestion that exposure to atmospheric oxygen levels on the surface may reduce recovery of injured cells. MPN method The MPN method produces an estimated count of the number of micro-organisms in a sample and is based on probability statistics. These statistics relate to the probability of an individual tube containing at least one micro-organism at any dilution based on the particular levels in the original material. This estimate is obtained by preparing decimal dilutions of the sample, then transferring subsamples of each dilution into multiple tubes of a broth medium. Typically, three (and sometimes eight) replicate tubes are used, but the number of aliquots and range of dilutions required is dependant on the expected level of contamination in the sample and the precision required from the estimated count. The tubes are incubated, then examined for evidence of growth of micro-organisms, e.g. turbidity, change in the pH, indole or gas production. Reference to probability tables indicates the likely contamination level in the sample. This method is more labour intensive than both the pour and spread plating techniques and the precision is poor unless a high number of replicate tubes are used. However, the MPN method is particularly useful when low counts are expected or when the amount of material is too large for either the pour or spread plating methods. Standard ISO methods are available for the pour plate method (Anon., 2003), the MPN method for enumeration of coliforms (Anon., 2006b) and for psychrotrophic micro-organisms (Anon., 2001a).

18.4.2 Conventional qualitative methods Qualitative methods are also termed presence/absence methods, as their primary aim is to find out if certain micro-organisms are present in the sample or not, and an actual level or count is not required. These methods are generally used to test for pathogens such as Salmonella spp., Listeria spp. and Campylobacter spp. Generally, there are four stages involved in a conventional pathogen presence/absence test which include: (i) (ii) (iii) (iv)

primary or pre-enrichment, selective enrichment, detection/plating and confirmation.

A typical total test time can be days rather than hours. The standard ISO method for detection of Salmonella in foods is an example of this type of protocol (Fig. 18.1). The food homogenate is first incubated in a non-selective primary enrichment (pre-enrichment) medium for approximately 20 hours. This facilitates resuscitation © 2008, Woodhead Publishing Limited

Chilled foods Buffered peptone water (9 parts) + Food sample (1 part, typically 25 g)

Confirmation

Detection/plating

Selective enrichment

Pre-enrichment

554

Fig. 18.1

Incubate at 37°C ± 1°C for 18 h ± 2 h

0.1 ml into 10 ml Rappaport-Vassiliadis broth

1 ml into 10 ml MKTTn broth

Incubate at 41.5°C ± 1°C for 24 h ± 3 h

Incubate at 37°C ± 1°C for 24 h ± 3 h

Plate onto XLD agar and second agar of choice

Plate onto XLD agar and second agar of choice

Incubate at 37°C ± 1°C for 24 h ± 3 h

Incubate at 37°C ± 1°C for 24 h ± 3 h

Select five typical colonies from each plate for confirmation

Select five typical colonies from each plate for confirmation

Microbiology of food and animal feeding stuffs – horizontal method for the detection of Salmonella spp. ISO 6579:2002 (Anon, 2002).

of any injured cells. The following stage involves transfer and incubation into two separate selective enrichment broths (24 hours). Two media are necessary due to the biological diversity of the Salmonella group, the fact they are so closely related to other species of Enterobacteriaceae and the need to maximise the probability of © 2008, Woodhead Publishing Limited

Conventional and rapid analytical microbiology

555

detecting all strains. Isolation on two selective agars (24 or 48 hours) is followed by visual assessment and purification of presumptive colonies on non-selective agars (24 hours) (Anon., 2002).

18.5 Rapid methods As outlined in the introduction, in recent years there have been remarkable advances in the development of rapid methods. Improvements in performance, quality, cost and availability of these methods have led them to being widely adopted and accepted. In many cases, evidence of microbial growth by visual assessment (e.g. colonies) is no longer required, with conductance levels or the presence of certain genes being accepted as alternatives to traditional plates counts. The term ‘rapid methods’ encompasses numerous technologies including those based on microscopy (e.g. phase contrast or FITC-based), measurement of adenosine triphosphate (ATP), monitoring of metabolic activity by electrical measurement, and detection of specific nucleic acids and antibodies. Use of these methods can have many benefits for both laboratories and food producers, because of shorter test times and reduced labour requirements. At this present time, no rapid method has yet been developed that can completely replace conventional methods, and in pathogen detection rapid methods are usually adopted to replace one stage of the total isolation, detection or identification procedure, resulting in conventional and rapid methods being used in combination (e.g. a conventional selective enrichment followed by detection using an immunoassay). In this chapter, examples of rapid methods commonly used by the food microbiologist are given. This is by no means a complete list and suggestions for further reading and sources of information can be found in Section 18.8. The user needs to determine the best rapid method system for use with their food type, facilities and resources, etc. available.

18.5.1 Rapid methods based on conventional techniques Alternative agar plating systems Alternative agar plating systems are designed to be simple, easy to use, disposable versions of conventional agar plates. Most are pre-made by the manufacturer and utilise cold-water setting gels on some sort of solid support, e.g. a film product or thin layer of plastic. One example of such a system that is now commonly used in food microbiology laboratories is Petrifilm™, produced by 3M. This product comprises two plastic films coated with adhesive, powdered culture medium constituents and a dehydrated cold-water-soluble gelling agent. The top layer of film is lifted and 1 ml of inoculum is added to the centre of the bottom film. By lowering the top layer, the inoculum is spread across the surface and is contained within the two layers. © 2008, Woodhead Publishing Limited

556

Chilled foods

A very similar product, Compact Dry, is produced by the company HyServe. Compact Dry also comprises dry media sheets, but on a solid cartridge, into which liquid samples diffuse. In the 3M Petrifilm™ plates, colonies appear coloured; for example, in the TVC plate colonies are red, due to a redox indicator tetrazolium salt. Compact Dry uses the same technology for TVC plates, but incorporates chromogenic substances for differentiation of E. coli and coliforms, where E. coli colonies appear blue and other coliforms appear red. The colouration helps distinguish colonies from food residues, and the presence of a counting grid aids counting and interpretation of results. Alternative agar plates have added advantages, e.g. they are less labour intensive (as no time-consuming media preparation is required) and require reduced incubator space. These types of products are available for total viable counts, enumeration of coliforms, Enterobacteriaceae, environmental Listeria spp (designed for use in processing plants), E. coli, staphylococci, yeasts and moulds. Automated systems – TEMPO As indicated previously, most rapid methods are based on conventional techniques and often replace one or more stages of a procedure, but not the procedure as a whole. However, an automated system, designed initially for quality indicator testing (TVCs and coliforms) in various food types, including solid foods, has now been developed by bioMérieux. Apart from sample preparation (dilution and homogenisation), most other stages, including serial dilutions of the food homogenate, inoculation, detection and interpretation of results, are carried out by the instrument. Testing for indicator micro-organisms can be one of the most laborious, time-consuming activities in the food microbiology laboratory, often requiring multiple dilutions and plate reading. For some companies, it has been reported that this type of testing accounts for up to 80% of their microbiological workload (Betts, personal communication).The TEMPO system is based on the MPN technique. However, it achieves greater sensitivity than the conventional MPN, and a 5 log enumeration range, by expanding the traditional ‘3 tube 3 dilution’ MPN with a ‘16 tube 3 dilution’ method (i.e.16 tubes each containing 250 µl, 16 × 25 µl and 16 × 0.25 µl). The 48-tube system is encased in a card device, approximately 10 cm2. The equipment consists of a TEMPO preparation station and a TEMPO reading station. Data are transferred between the two stations via a wireless communication link. At the preparation station, a test sample is diluted and inoculated into a rehydrated TEMPO culture medium. The sample and medium are automatically associated with a specified TEMPO card using the TEMPO software. The sample is automatically transferred aseptically into the 48well enumeration card, using three different volumes, by a filler/sealer located within the instrument. Inoculated cards are incubated (in a conventional incubator) for a specific time and temperature, depending on the test. Any micro-organisms present in the tubes multiply and change the fluorescent properties. This change is detected by the reader station, using UV light and optical receptors. Depending on the number and location of the positive tubes, the system reports the number of micro-organisms present as the MPN. Data are analysed using on-board MPN © 2008, Woodhead Publishing Limited

Conventional and rapid analytical microbiology

557

tables and the interpreted result is displayed. Results can be manually or automatically validated, with the use of barcodes and wireless technology ensuring traceability throughout the process. Two main advantages of the TEMPO system include time saving and standardisation of the technique. Various validation studies, using a series of different food types and the TEMPO E. coli (EC) and TEMPO total count (TC) cards have been carried out, including those performed by Campden and Chorleywood Food Research Association and AFNOR. Results indicate the performance of TEMPO to be good, having agreement rates with conventional counts above 90% for all food matrices tested (unpublished data). The TEMPO is one of the most automated systems for enumeration of micro-organisms in food and has potential to be developed for detection of specific organisms in the future,

18.5.2 Rapid methods based on metabolic activity The detection of microbial growth by electrical measurement can be based on measurement of the changes caused to the conductivity, or resistance, of a culture medium by microbial growth; this is measured using an alternating electrical current (AC) at specific frequencies. These changes are caused because as microorganisms grow and metabolise, large molecules are broken down to small ones; for example, amines and ammonia from proteins, lactic acid from carbohydrates, polymers reduced to monomers. As a result, ionic changes occur in the medium or food homogenate and its electrical properties alter. Changes in ionic concentration and conductivity can be measured and related to the number of micro-organisms or metabolic activity in the sample. Commercially available systems for detection of microbial growth by electrical measurement were first developed in the 1970s. Presently, there are at least four systems available: Malthus, RABIT, Bactometer and BacTrac. Each system has unique characteristics but all are based on the same principle and are capable of handling multiple samples simultaneously. Each system requires electrodes to be immersed in a growth medium or food homogenate. A typical test protocol involves inoculating a ‘test cell’ (containing media) on the instrument with the food homogenate. The instrument, which also acts as an incubator, takes regular readings indicating when a significant change in conductance has been detected. Each system has similar specifications and a threshold of approximately 106 micro-organisms per ml before any electrical change will be detected. The time taken to reach this ‘threshold of detection’ is known as the detection time and it is this measure that is related to the number of microorganisms initially in the sample. Before such a system is adopted in-house, calibrations must be performed by testing samples using both a conventional plating method and the electrical test, for each product or group of products to be tested. (Although this is generally the case when implementing a rapid method, it is particularly important for electrical methods due to potential interference by the food matrix.) Calibrations must be carried out for every sample type to be tested by the electrical method. As these systems are based on the use of growth media, it is possible, to some extent, to develop protocols using media formulations for © 2008, Woodhead Publishing Limited

558

Chilled foods

enumeration of specific micro-organisms or groups of micro-organisms. Considerable research has been carried out in this area and some examples of published data for specific micro-organisms include Enterobacteriaceae (Cousins and Marlatt, 1990), E. coli (Druggan et al.,1993) and Campylobacter (Bolton and Powell, 1993). Potential problems with electrical methods using direct measurement (i.e. where the electrodes are immersed in the culture medium or food homogenate) include interference from the food matrix itself, changes in the surface chemistry of the probes and the fact that some micro-organisms (e.g. yeasts) produce only small changes in conductance and are thereby very difficult to detect. To address these issues, developments have been made using an indirect conductance measurement approach (Owens et al.,1989; Betts, 1993). Here, the growth medium is contained in a separate compartment to the electrode and the liquid surrounding the electrode is a gas (CO2) absorbent, e.g. potassium hydroxide. The growth medium is inoculated with the sample and, as micro-organisms grow, gas, mainly carbon dioxide, is released and absorbed by the liquid surrounding the electrode, causing a change in conductivity. This design has the potential to increase the number of applications of electrical methods within the food industry.

18.5.3 Immunological methods Immunological methods, also referred to as immunoassays or antibody-based methods, utilise the highly-specific binding reaction between antibodies and the antigens associated with the target micro-organism or their toxins. Antibodies are proteins produced by the white blood cells of animals that have been challenged by a foreign agent, e.g. a molecule or micro-organism. Antibodies attach to specific areas or antigenic sites on the invading agent to neutralise its effects. There are two types of antibody, monoclonal and polyclonal. If the invading agent is a large molecule, such as a protein or micro-organism, with many different antigenic sites, polyclonal antibodies are produced during the immune response of the host. Monoclonal antibodies are very specific and can be produced in vitro using tissue culture techniques, procured for a single antigenic site, using a single white blood cell. Because the attachment is highly specific, immunological tests exploit this to detect specific micro-organisms, proteins or toxins. Antigens may be cellular components such as lipopolysaccharides on the outer cell wall, proteins on flagellae, or a toxin produced by the target microorganism during growth. In order to determine how much binding between antibody and antigen has taken place, a system for visualising or measuring the interaction is required. To achieve this, a detectable ‘label’ is attached to the antibody. Labels can be of various types including fluorescent agents, luminescent chemicals, radioisotopes or enzymes. The widespread use and acceptance of immunological methods has led to a great number of commercially produced kits being available for a range of foodborne bacteria, including those belonging to the genera Salmonella, Listeria, Campylobacter and more specifically, E. coli O157:H7. © 2008, Woodhead Publishing Limited

Conventional and rapid analytical microbiology

559

ELISA Probably the most widely used immunological technique for detection of specific micro-organisms in foods is that based on the enzyme linked immunosorbent assay (ELISA). The technique is designed to identify specific micro-organisms and replaces the detection or isolation stage on agar. The scope of an ELISA can be designed to detect a genus, species or serotype. It is relatively easy to perform, can identify a range of pathogens, can be semi-automated, and gives rapid results. However, a positive result obtained by an ELISA is presumptive and must be confirmed, e.g. by conventional tests. The greatest advantage of this technique therefore is in negative screening (e.g. evidence of absence), and including an ELISA test in a microbiological detection procedure has the potential to greatly increase the number of samples per day that can be screened for the presence of a particular pathogen, assuming there are acceptably low levels of ‘false-negative’ results. In an ELISA, the term ‘enzyme linked’ indicates that the labelling system relies on the activity of an attached enzyme. In most systems, the attached enzyme catalyses the conversion of a colourless substrate to a coloured product. The endpoint of the test can then easily be visualised by eye or by a spectrophotometer, depending on the type of test. Typical enzyme–substrate complexes used in ELISAs include alkaline phosphatase (enzyme) with para-nitrophenyl phosphate (substrate) and horseradish peroxidase (enzyme) with tetramethylbenzidine (substrate); both complexes produce a yellow coloured product. ELISA tests can be made in various formats but the simplest and most commonly used in commercially available kits is the ‘sandwich’ format. The word ‘sandwich’ indicates that the assay uses two antibodies, which trap or sandwich the target antigen. During the procedure, the antigen is first captured then detected. The capture antibody, specific to the target antigen, is attached to the surface of a solid support, e.g. microtitre well. An enriched food sample is added to the well and if the target antigen is present, it will bind to the antibodies. After a washing procedure to remove food debris and unbound material, a second ‘detection antibody’ is added to the well. This antibody has an enzyme label attached. Again, the antibody will bind to the target antigen/antibody complex, creating the ‘antibody sandwich’. More washing procedures are carried out to remove any unbound antibodies, followed by the addition of a colourless substrate which the enzyme converts to a coloured product. Finally, a stop solution is added to prevent any further enzyme activity and any change in colour is measured (Fig. 18.2). The total test time for a sandwich ELISA is typically between 2 and 3 hours. It is possible to use this format in quantitative tests by calibrating the concentration of antigen against colour intensity. In the detection of foodborne pathogens, however, the sandwich ELISA is usually used for qualitative purposes, indicating the presence or absence of a pathogen. All sandwich ELISAs require an enriched food sample (e.g. one where the level of any target micro-organisms present would have exceeded the detection limit) as a starting point. The enrichment procedure needed depends on the food type and the micro-organism being detected. © 2008, Woodhead Publishing Limited

560

Chilled foods

Fig. 18.2

The main steps of a sandwich ELISA.

Numerous commercial kits are available, and manufactured by various companies including TECRA, bioMérieux and Foss Electric. Each kit offers advantages and disadvantages, depending on the individual user requirements and the type of product being analysed. The various formats available include the manual-based dipstick format which incorporates immunocapture (e.g. TECRA UNIQUE™), designed for low sample numbers; the microtitre plate format (e.g. SalmonellaTek, bioMérieux), which can be used with or without an immunocapture stage, designed for high throughput; and fully automated systems such as that manufactured by Foss Electric (EiaFoss). © 2008, Woodhead Publishing Limited

Conventional and rapid analytical microbiology

561

18.5.4 Adenosine triphosphate systems ATP is the energy storage molecule of all living organisms. In any food sample, there are three possible sources of ATP: microbial, somatic (from non-microbial cells), and extra-cellular from either origin. From an analytical point of view only the microbial ATP is of interest, although for monitoring cleaning or hygiene, all sources may be used to indicate its effectiveness. When ATP is combined with an enzyme extracted from the firefly (luciferase) and the substrate luciferin, a reaction takes place that results in the production of light. With the use of sensitive light meters, known as luminometers, it is possible to measure the levels of light emitted. The light output is converted to a digital reading and expressed as relative light units (RLUs). The more ATP present in a sample, the more light is produced, which will result in a higher RLU reading. Many equipment manufacturers go one step further and suggest a correlation between RLUs and microbial numbers. ATP bioluminescence reaction: Luciferase + Mg ++

ATP + O2 + D-luciferin → decarboxyluciferin + CO2 + AMP + PPi + light The amount of ATP present in different microbial cells can vary depending on the species, nutrient level, injury level and stage of growth (Stalker, 1984). Therefore, when methods based on ATP bioluminescence are being implemented, it is important to consider:

• the type of micro-organisms potentially being present; in general, vegetative • •

bacteria contain approximately 1 fg of ATP/cell (Karl, 1980) and spores will contain no ATP (Sharpe et al.,1970) the physiological status of the micro-organisms i.e. whether they could be injured due to nutrition depletion, extreme temperatures or pH change, whereby a resuscitation period may be required whether the cells are in an ATP-free environment or in a complex-rich environment, such as food.

Food testing The use of the firefly bioluminescence assay of ATP for detection of microorganisms in foods was first described by Levin et al. (1964). The main cause for concern when testing food samples by this technique is the ATP level of the food (somatic ATP). Generally, this is much higher than levels found from all the microorganisms present. It is therefore necessary to either physically separate the microbial from the other sources of ATP, or use specific chemical extractants to remove and destroy non-microbial ATP. Although filtration methods have successfully been developed and applied to non-particulate foods such as drinks (Littel and LaRocco, 1986), most systems utilise the chemical extraction approach. This involves lysis of somatic (food) cells, followed by the destruction of the released, and any free, ATP with an apyrase (ATPase). A further reagent is then © 2008, Woodhead Publishing Limited

562

Chilled foods

added to lyse any microbial cells and the released microbial ATP is detected using the luciferase reaction. There are a number of commercially available systems designed for detection of microbial ATP in a variety of food types. Some examples of manufacturers are Lumac, Foss Electric, BioOrbit and Biotrace. Many of these systems are automated, including in some cases any dilution stages required. The systems generally have similar specifications, including a minimum detection limit of 104 bacteria per ml/g and an analysis time of approximately 1 hour. Hygiene monitoring To minimise the risk of contamination in food products it is necessary to know that production lines are effectively cleaned before production commences. In order for a production to be clean, the levels of micro-organisms must be low and product residues removed. Product residues on a surface may provide a source of nutrition for the growth of micro-organisms, which if undetected may affect food safety or quality. Therefore, validated cleaning and disinfection procedures must be in place to ensure that the hygiene of the process line is adequate. The required level of cleanliness is very much dependent on the process and product, e.g. different procedures will be in place and target values met for a raw meat processing line compared to the manufacture of a ready-to eat product. The traditional method for monitoring surface cleanliness is based on swabbing or rinsing and subsequent microbial enumeration. The main advantage of this approach is that it has a low detection limit (1 cfu/9 cm2) and can provide an indication of the types of micro-organisms present. The disadvantage of this type of methodology is that results are not available before 1–3 days. In general, hygiene checks are done to indicate that cleaned equipment will not contaminate food, usually as part of the verification system for HACCP. In some cases, for example in high hygiene areas, there is a need for hygiene monitoring results to be directly available in order to prevent production being started if the area, or equipment, is not clean, and to provide the opportunity for remedial action before processing is started. In addition to the time factor, many food microbiologists now consider total organic soil, which includes food debris as well as micro-organisms, to be very important. Hygiene monitoring using ATP bioluminescence detects the presence of both micro-organisms (if present in high enough numbers) and product residues containing ATP, and thus provides a means by which the cleanliness of production lines can be assessed within minutes. Conventional methods are still essential to investigate cleaning failures, especially at designated critical control points or places where quality defects, such as short shelf-life, are caused. Investigation of the origins of failure, which may be been identified by a rapid screening method such as ATP bioluminescence, may involve cultivation, quantification, identification and tracing sources of contamination, all of which can involve the use of conventional methods. When adopting a new technique for hygiene monitoring, such as ATP bioluminescence, it should not be viewed simply as a replacement for in-house methods. Many companies require an initial rapid screening monitoring system, producing data for trend analysis, with microbiological ‘back-up’. © 2008, Woodhead Publishing Limited

Conventional and rapid analytical microbiology

563

18.5.5 Nucleic acid-based methods Many nucleic acid-based methods, also referred to as molecular methods, were originally developed for research use and are now rapidly being adapted for use in routine analyses. The pharmaceutical and medical industries have long promoted DNA diagnostics (e.g. in drug discovery and response) and the last decade has seen many technologies being adapted to the food industry. Nucleic acid-based methods have often been perceived as being difficult to use and labour intensive, but commercialisation of these types of methods has provided easy-to-use, standardised, powerful tools. Nucleic acid-based methods can be used for both the detection and the identification of specific foodborne micro-organisms. In both applications, hybridisation of complementary strands of nucleic acids is the fundamental stage. Detection methods – polymerase chain reaction Nucleic acid-based detection methods usually require amplification of the target DNA. This can be achieved by the polymerase chain reaction (PCR). This technique has been subjected to extensive research and development and is now the most widely used of the nucleic acid-based methods for detection of microorganisms from food. The technique can be used to amplify, in vitro, a specific region of DNA. The region is determined by short oligonucleotides (usually 20– 30 nucleotides in length), designed such that the sequence matches the ends of the region of DNA of interest. These oligonucleotides are known as primers. Amplification takes place during a number of cycles. In each cycle, the doublestranded DNA template is denatured, by heating, to produce single strands. On cooling, the primers hybridise to the single-stranded DNA and a thermostable enzyme, DNA polymerase, synthesises a complementary strand. The result is newly synthesised, double-stranded, copy DNA. In subsequent cycles, the primers will bind to both the original and the newly synthesised DNA, resulting in an exponential increase in the number of copies. By using this technique, DNA can be amplified approximately a million fold during a typical 2-hour assay. The results of the PCR (also termed PCR product) can be detected using agarose gel electrophoresis, with the amplified DNA appearing as bands of different sizes. The bands can then be compared against standard molecular weight markers. The presence of a band or bands of expected size indicates a positive result (presence of the micro-organism). In theory, a PCR-based system is capable of producing an amplicon of expected size from one micro-organism per gram of food. However, most PCR-based systems have superior performance if food samples are incubated in broth before the PCR is applied; typically, sample sizes of 10–25 g are used. The broth types and incubation times vary depending on the test. Pre-incubation helps reduce any dilute inhibitory substances that may be present in the food, allows the recovery of any injured cells, and increases the possibility of detecting only viable cells. Many other techniques also based on amplification of nucleic acids have been developed. These include random amplified polymorphic dna (RAPD), ligase amplification reaction (LAR), transcript amplification system (TAs) and nucleic acid sequence © 2008, Woodhead Publishing Limited

564

Chilled foods

based amplification (NASBA). However, these techniques are normally applied to specialised research studies and not routine testing. Real-time PCR In recent years, there has been a great deal of development and acceptance of realtime PCR. In real-time PCR, the reaction is continuously monitored using a fluorescent reporter dye that binds specifically to double-stranded DNA. Examples of dyes used include ethidium bromide and SYBR® Green. When bound, the dye emits light upon excitation, resulting in an increase in fluorescence in proportion to the amount of PCR product. These types of dye will bind to any double-stranded DNA in the reaction; therefore, they are most well suited to highly specific assays. Numerous real-time PCR systems are now commercially available, together with kits for detection of specific micro-organisms.

18.6 Identification and characterisation procedures In most situations, micro-organisms isolated from food will need to be identified or characterised. This is particularly so if the micro-organisms are considered to be pathogens. The term ‘characterisation’ rather than ‘identification’ is now used more widely in food microbiology. One reason for this may be because there is less emphasis on nomenclature. This has been influenced by the implementation of nucleic acid-based methods and it is now relatively easy to gain information on the genetic characteristics of micro-organisms. In its simplest terms, an identification method can be described as one in which the result is a taxonomic name, whereas a characterisation method may not necessarily result in naming the micro-organism, but a characteristic property (e.g. ability to produce toxin) will be determined. This level of information is vital for many areas of food microbiology including ecology studies (survival of micro-organisms in the production environment and processes), resistance of micro-organisms to cleaning materials, tracing sources of contamination, challenge studies and identifying ‘high risk’ strains.

18.6.1 Phenotypic methods Traditional identification methods are based on concepts developed at the end of the 19th century. They can involve a plethora of techniques including morphology (cell and colony), cell wall chemistry, fermentation studies, growth temperature tolerance, nutritional requirements, gas requirements, pH tolerance and use of different energy sources. These techniques are often supplemented by serotyping, phage typing, biotyping and antibiotic susceptibility testing. Collectively termed phenotypic methods, they detect characteristics which are expressed by the microorganism. As they rely on an expressed feature, rather than features that may be present but are not expressed at the time of testing, problems can be associated with these types of methods. Some characteristics (or markers) are expressed only in © 2008, Woodhead Publishing Limited

Conventional and rapid analytical microbiology

565

certain environments or culture conditions and many are non-reproducible. Therefore these techniques, although they have been used for many years, may have low discriminatory power and allow subjective interpretation of results. However, many phenotypic methods such as Gram staining and biochemical profiling are used as standard, routine procedures in the food microbiology laboratory and are still the simplest, most cost effective way of quickly determining initial classification of a micro-organism and hence the safety and quality of a food.

18.6.2 Genotypic methods Recent advances in molecular sequencing of the prokaryotic genome have led to the development of new, universally acceptable DNA-based methods for the identification of micro-organisms to sub-species level. This level of identification is often referred to as subtyping or ‘strain-level characterisation’. Characterisation techniques are powerful tools for the food microbiologist and provide valuable information in cases where a species level identification is not specific enough and it is necessary to know if strains are the same or different, for example when tracing sources of contamination and establishing if an isolate from a finished product originated from a raw material, the production environment or inadequately cleaned equipment. Discriminatory subtyping methods have also been successfully applied to epidemiological investigations, identification of new or emergent strains, identification of routes of transmission and improving the understanding of the epidemiology of foodborne disease (Swaminathan et al., 2001). Developments in characterisation methods have most notably utilised the highly conserved regions of the prokaryotic genome, namely those encoding the ribosomal 16s rRNA sub-unit. The 16s rRNA gene is often referred to as the ‘universal marker’, being genetically stable and having a high information content. Until relatively recently, one disadvantage of genetic characterisation methods was that they required highly skilled technicians to carry them out and lacked standardisation and consistency. Recent developments have resulted in automated systems, simplifying operator training, eliminating subjective interpretation of results and minimising errors due to technique. DNA sequencing Comparative DNA sequencing has become a more widely used alternative method for subtyping of micro-organisms and since the introduction of automated sequencers, the technique has become more rapid, accurate, reproducible, cost effective and suitable for use in routine laboratories. The MicroSeq® Microbial Identification System (manufactured by Applied Biosystems) was developed specifically for automated microbial DNA sequencing. The system combines all the instrumentation, reagents, sequence libraries, and software required, with results available in 5 hours. The main stages include extraction of genomic DNA from a pure colony followed by amplification of the target portion of the ribosomal gene. The PCR product then goes through a cycle sequencing process and the fluorescently labelled product is analysed using an © 2008, Woodhead Publishing Limited

566

Chilled foods

automated fluorescent DNA sequencer. Finally, the resulting sequence is compared with a library of known sequences. DNA sequence data provides an unambiguous chromosomal comparison of isolates. However, care must be taken when interpreting the data. Any identification system is only as good as the database of information to which the unknown is compared. Typically, sequence data are first aligned and compared. Alignment often involves the use of a tool known as the basic local alignment search tool (BLAST). This tool breaks the sequence to be searched into smaller segments of arbitrary length, then locally aligns them to the same segments of the unknowns. This process continues until the alignment level drops below a defined threshold level when alignment orders and scores are reported. The BLAST algorithm is very fast and generally works well; however, it does not provide accurate distance measurements. This can be achieved using a ‘pairwise search’ where sequences are aligned and compared at every base position. This algorithm provides the most accurate distance measurement but is very slow. After alignment and comparison, a phylogenic tree can then be generated and data interpreted. Ribotyping Another example of a strain-level characterisation technique is ribotyping. Ribotyping has been used extensively in epidemiological investigations of many foodborne pathogens and is based on the analysis of restriction fragment length polymorphisms (RFLPs) of ribosomal RNA genes. The main disadvantage of conventional ribotyping is the multiple steps involved, making the technique very labour intensive and time consuming. However, the technique has been made more accessible by the introduction of a fully automated system, the RiboPrinter®, manufactured by DuPont Qualicon. The RiboPrinter® requires a single colony as inoculum and, during an 8-hour period, performs a restriction digest (using the standard EcoRI or other restriction enzymes) of the chromosomal DNA, separates the restriction fragments by gel electrophoresis and simultaneously blots the DNA fragments to a membrane which is used for Southern blot analysis. Restriction digest fragments are hybridised to a bacterial probe that is based on the conserved regions of the genes for the ribosomal DNA operon. The test result is a DNA fingerprint (RiboPrint™) which is strain specific. Each fingerprint is stored in a database so it can be accessed for future comparisons and identifications. The discriminatory power of ribotyping is limited and as different organisms have different ribosomal gene copies, the suitability of the method is to some extent organism specific. For many organisms, ribotyping has most value when used in conjunction with other subtyping methods, such as DNA sequencing or pulsed field gel electrophoresis (PFGE).

18.7 Future trends Historically, microbiological analysis of foods has relied heavily on conventional testing methods, isolating and enumerating micro-organisms on specialised media, © 2008, Woodhead Publishing Limited

Conventional and rapid analytical microbiology

567

yielding results after several days and repeated culture enrichment steps. Developments in immunological methods, and more recently in molecular-based methods, have lead to rapid, specific, high-throughput, automated methods designed for use in the routine laboratory. Implementation and standardisation of these types of methods will allow the food industry to make timely assessments on the microbiological safety of its products and improve knowledge of the role of some micro-organisms in pathogenesis. DNA microarrays One example of a molecular-based method that has become increasingly applied and used within the food industry is the DNA microarray. These arrays are extremely powerful tools for identifying DNA sequences, comparing different genomes of isolates, monitoring gene expression and screening for virulence determinants. They consist of thousands of unique DNA sequences, each attached at a known location onto a small solid surface, typically a glass slide. Complimentary, labelled mRNA or DNA can bind to the fixed sequences to produce a pattern indicative of nucleic acid sequences. These sequences can be qualitatively and quantitatively analysed by a computer. In a standard array, the ‘probe’ refers to known DNA that is used to identify the unknown ‘target’ sequence. DNA microarrays have revolutionised the understanding of microbial genomes and molecular mechanisms underlying cell functions. Over the past few years, microarray technology has been use to explore transcriptional profiles and genome differences for a variety of micro-organisms, greatly facilitating the understanding of microbial metabolism. With the increasing availability of microbial genomes, DNA arrays are becoming a common tool in many areas of microbial research, including physiology, epidemiology, ecology and phylogeny, and have become a recognised assay method in species identification, providing information related to the presence of antibiotic markers and pathogenicity regions. Systems designed for use in a routine microbiology laboratory have recently become commercially available, with one example being the ‘IdentiBac’ system. The IdentiBac is a range of bacterial genotyping kits based on the Array Tube platform technology from Clondiag Chip. The systems are simple, rapid and can be used in high throughput or initial screening procedures. Currently, two products are available: IdentiBac AMR (which detects antimicrobial resistance genes in Gram-negative bacteria) and IdentiBac EC (which detects virulence genes in E. coli).The IdentiBac EC array tube contains probes for 60 different genes present in E. coli and allows identification of virulence genes associated with a range of pathotypes including shigatoxigenic and enterohaemorrhagic. Results from an E. coli isolate are available within 24 hours.

© 2008, Woodhead Publishing Limited

568

Chilled foods

18.8 Sources of further information and advice Useful publications BAYLIS C L (2006) Catalogue of Rapid Methods (5th Edition), Campden and Chorleywood

Food Research Association. (1960) The use and meaning of the term ‘psychrophilic’, Journal of Applied Bacteriology, 23, 189–190. SPENCER R C, WRIGHT E P AND NEWSOM S W B (eds) (1994) Rapid Methods and Automation in Microbiology and Immunology, Intercept Ltd., Andover.

EDDY B P

Useful websites www.identibac.com www.AOAC.org www.bioMérieux.com www.oxoid.com www.appliedbiosystems.com www.denka-seiken.co.jp

18.9 References ANON. (1986) Microorganisms in Foods 2. Sampling for microbiological analysis: princi-

ples and specific applications, ICMSF. Blackwell Scientific, Oxford. (1997) ISO 11290-1:1997 Microbiology of food and animal feeding stuffs – Horizontal method for the detection and enumeration of Listeria monocytogenes – Part 1: Detection method. Geneva: International Organisation for Standardisation. ANON. (2001a) ISO 17410:2001 Microbiology of food and animal feeding stuffs – Horizontal method for the enumeration of psychrotrophic microorganisms. Geneva: International Organisation for Standardisation. ANON. (2001b) ISO 16654:2001 Microbiology of food and animal feeding stuffs – Horizontal method for the detection of Escherichia coli O157. Geneva: International Organisation for Standardisation. ANON. (2002) ISO 6579:2002 Microbiology of food and animal feeding stuffs – Horizontal method for the detection of Salmonella spp. Geneva: International Organisation for Standardisation. ANON. (2003) ISO 4833:2003 Microbiology of food and animal feeding stuffs – Horizontal method for the enumeration of micro-organisms – Colony-count technique at 30 degrees C. Geneva: International Organisation for Standardisation. ANON. (2006a) ISO 4831:2006 Microbiology of food and animal feeding stuffs – Horizontal method for the detection and enumeration of coliforms – Most probable number technique. Geneva: International Organisation for Standardisation. ANON. (2006b) ISO 4831:2006 Microbiology of food and animal feeding stuffs – Horizontal method for the detection and enumeration of coliforms – Most probable number technique. Geneva: International Organisation for Standardisation. BAILEY J S AND COX N A (1992) Universal pre-enrichment broth for the simultaneous detection of Salmonella and Listeria in foods, Journal of Food Protection, 55(4) 256–259. BAIRD-PARKER A C (1962) An improved diagnostic and selective medium for isolating coagulase positive Staphylococci, Journal of Applied Bacteriology, 25 12–19. BAYLIS C L AND PATRICK M (1999) Comparison of a range of chromogenic media for enumeration of total coliforms and E. coli in foods. Leatherhead International Technical Notes No. 135–99. BAYLIS C L, MACPHEE S AND BETTS R P (2000) Comparison of methods for the recovery and ANON.

© 2008, Woodhead Publishing Limited

Conventional and rapid analytical microbiology

569

detection of low levels of Salmonella in ice cream and milk powder, Letters in Applied Microbiology, 30(4) 320–324. BECKERS H J, VAN LEUSDEN F M, BINDSCHEDLER O AND GUERRAZ D (1984) Evaluation of a pour-plate system with a rabbit plasma-bovine fibrinogen agar for the enumeration of Staphylococcus aureus in food, Canadian Journal of Microbiology, 30(4) 470–74. BETTS R P (1993) Rapid electrical methods for the detection and enumeration of food spoilage yeasts, International Biodeterioration and Biodegradation, 32 19–32. BETTS R P (1994) The separation and rapid detection of micro-organisms. In: Rapid Methods and Automation in Microbiology and Immunology, Spencer R. C., Wright E. P. and Newsom S.W.B. (eds) pp. 107–120, Intercept Press, Hampshire, England. BLACKBURN C D AND MCCARTHY J D (2000) Modifications to methods for the enumeration and detection of injured Escherichia coli O157:H7 in foods, International Journal of Food Microbiology, 55(1–3) 285–90. BOLTON F J AND POWELL S J (1993) Rapid methods for the detection of Salmonella and Campylobacter in meat products, European Food and Drinks Review, Autumn 73–81. BROWN M H (2000) Microbiological hazards and safe process design. In: Stringer M. and Dennis C., Chilled Foods, Woodhead Publishing Ltd., Cambridge, UK, 287–339. CHRISOPE G L, FOX C W AND MARSHALL R T (1976) Lecithin agar for detection of microbial phospholipases. Applied and Environmental Microbiology, 31(5) 784–6. COUSINS D L AND MARLATT F (1990) An evaluation of a conductance method for the enumeration of enterobacteriaceae in milk, Journal of Food Protection, 53 568–70. DRUGGAN P, FORSYTHE S J AND SILLEY P (1993) Indirect impedance for microbial screening in the food and beverages industries. In: New Technologies in Food and Beverage Microbiology (eds R G Kroll and A Gilmour) SAB Technical Series 31, London, Blackwell. KARL D M (1980) Cellular nucleotide measurements and applications in microbial ecology, Microbiological Review, 44 739–96. LEVIN G V, GLENDENNING J R, CHAPELLE E W, HEIM A H AND ROCECK E (1964) A rapid method for the detection of micro-organisms by ATP assay: Its possible application in cancer and virus studies, Bioscience, 14 37–8. LITTEL K J AND LAROCCO K A (1986) ATP screening method for presumptive detection of microbiologically contaminated carbonated beverages, Journal of Food Science, 51 474– 6. MCCARTHY J AND BLACKBURN D C (1999) The detection of verotoxins produced by E. coli using the Ridascreen Verotoxin, Premier EHEC and VTEC-RPLA assay kits, Seventeenth International Conference of the International Committee on Food Microbiology and Hygiene (ICFMH) proceedings, Veldhoven, The Netherlands, 13–17 September 1999. OTTAVIANI F, OTTAVIANI M AND AGOSTI M (1997) Differential agar medium for Listeria monocytogenes, Quimper Froid Symposium Prodeedings, P6 ADRIA Quimper, 16–18 June. OWENS J D, THOMAS D S, THOMPSON P.S AND TIMMERMAN J W (1989) Indirect conductimetry: A novel approach to the conductimetric enumeration of microbial populations, Letters in Applied Microbiology, 9 245–50. SHARPE A N, WOODROW M N AND JACKSON A K (1970) Adenosine triphosphate levels in foods contaminated by bacteria, Journal of Applied Bacteriology, 33 758–67. STALKER R M (1984) Bioluminescent ATP analysis for the rapid enumeration of bacteria – principles, practice, potential areas of inaccuracy and prospects for the Food Industry. Campden Food and Drink Research Association Technical Bulletin 57, Campden and Chorleywood Food Research Association, UK. STEPHENS P J (2003) Culture methods. In: McMeekin T. A., Detecting Pathogens in Food, Woodhead Publishing Ltd., Cambridge, UK, 123–146. STEPHENS P J AND JOYNSON J A (1998) Direct inoculation into media containing bile salts and antibiotics is unsuitable for the detection of acid/salt stressed Escherichia coli O157:H7, Letters in Applied Miocrobiology, 27 147–51. © 2008, Woodhead Publishing Limited

570

Chilled foods

SWAMINATHAN B, BARRETT T, HUNTER S, TAUXE R AND THE CDC TASKFORCE

(2001) ‘PulseNet: The molecular subtyping network for foodborne bacterial disease surveillance, United States’, Emerging Infectious Diseases, 7(3) 382–9. THOMAS A, CHART, H, CHEASTY T, SMITH H R, FROST, J A AND ROWE B (1993) Verotoxinproducing Escherichia coli, particularly serogroup O157, associated with human infections in the United Kingdom, Epidemiology and Infection, 110 591–600. WALKER S J (1990) Growth characteristics of food poisoning organisms at suboptimal temperatures. In: Zeuthen, P., Cheftel, J. C., Eriksson, C., Gormley, T.R., Linko, P. and Paulus, K. (eds) Processing and quality of foods Vol. 3. Chilled foods: The revolution in freshness, Elsevier Applied Science, London. 3 159–162. ZHAO T AND DOYLE M P (2001) Evaluation of universal pre-enrichment broth for growth of heat-injured pathogens, Journal of Food Protection, 64 (11) 1751–1755.

© 2008, Woodhead Publishing Limited

Part IV Safety and quality management

© 2008, Woodhead Publishing Limited

19 Shelf-life of chilled foods C. M. D. Man, London South Bank University, UK

19.1 Introduction All foods deteriorate during storage due to their perishable nature. Chilled foods in particular have relatively short shelf-lives, even when stored under refrigerated conditions. In the UK, chilled foods include a very wide variety of processed and prepared food and drinks ranging from raw prepared meats and fish and similar meal components, prepared fruit and vegetables, ready-to-eat salads, ready-to-eat dairy products, delicatessen meats, yellow fats and fat spreads, heat-and-eat snacks such as pizzas and savoury pastry products, ready meals, chilled soups and sauces, fruit juices and smoothies, to complete meals prepared for cooking (Anon., 2005). Until recently, the legal equivalent of the ‘shelf-life of food’ in the UK (and indeed within the EU) was the appropriate ‘minimum durability’ required by the Food Labelling Regulations 1996 (SI, 1996/1499), which have been amended by a number of subsequent regulations. This particular food labelling requirement applies to all food which is ready for delivery to the ultimate consumer or to a catering establishment, subject to certain exceptions. The requirement is fulfilled by a food business operator through the correct use of one of the following:

• In the case of a food which, from the microbiological point of view, is highly •

perishable and in consequence likely after a short period to constitute an immediate danger to health, a ‘use by’ date. In the case of a food other than one specified above, an indication of minimum durability, a ‘best before’ date.

In addition, the ‘best before’ date and the ‘use by’ date must be followed by any © 2008, Woodhead Publishing Limited

574

Chilled foods

special storage conditions which need to be observed, such as ‘keep refrigerated at 0 °C to +5 °C’ or ‘keep in a cool, dry place’. Generally, storage conditions are important because in the EU’s original food labelling Directive 79/112/EEC (EEC, 1979), which the Food Labelling Regulations (HMSO, 1996) implement, the date of minimum durability is defined as the date until which the foodstuff retains its specific properties when properly stored. No food manufacturer can reasonably be held responsible for its food that does not live up to its declared shelf-life if it has been stored under conditions for which it is not intended. Recently, the term ‘shelflife’ has appeared in EU/UK food law (European Commission, 2005). The Commission Regulation (EC) 2073/2005 on microbiological criteria for foodstuffs, which came into force in the UK in January 2006, defines ‘shelf-life’ in Article 2 as either the period corresponding to the period preceding the ‘use by’ or the minimum durability date, as defined respectively in Articles 9 and 10 of Directive 2000/13/EC, the most recent food labelling Directive. Therefore, all food business operators have a legal obligation to determine, assign and maintain shelf-lives of their products, as this makes legal as well as commercial sense. What is less obvious from the labelling regulations is which foods should carry a ‘use by’ and which foods, a ‘best before’ date. In fact, it is the food manufacturer and food processor’s responsibility to decide which minimum durability indication is appropriate, either a ‘use by’ or a ‘best before’ date; after all, they are supposed to know their own products. Useful guidance, however, is available and it has been suggested that the following food groups, essentially all chilled foods, are likely to require a ‘use by’ date (Crawford, 1998):

• • • • • • •

dairy products, e.g. fresh cream filled desserts cooked products, e.g. ready-to-heat meat dishes smoked or cured ready-to-eat meat or fish, e.g. hams, smoked salmon fillets prepared ready-to-eat foods, e.g. sandwiches, vegetable salads such as coleslaw uncooked or partly cooked savoury pastry and dough products, e.g. pizzas, sausage rolls raw ready-to-cook products, e.g. uncooked products comprising or containing meat, poultry or fish, with or without raw prepared vegetables vacuum or modified atmosphere packs, e.g. raw ready-to-cook duck breast packed in modified atmosphere.

A comprehensive and perhaps more useful definition of the shelf-life of food has also been available for some time (IFST, 1993): It is the period of time during which the food product will (i) (ii)

remain safe; be certain to retain its desired sensory, chemical, physical, microbiological and functional characteristics; (iii) where appropriate, comply with any label declaration of nutrition data, when stored under the recommended conditions. Clearly, safety and quality are the two main aspects of food shelf-life, and safety must always take priority over quality, as food safety is both a fundamental and © 2008, Woodhead Publishing Limited

Shelf-life of chilled foods 575 legal requirement. The aim of this chapter is to review these important aspects of shelf-life as they apply to chilled foods.

19.2 Safety of chilled foods In the UK, the Food Safety Act (HMSO, 1990) prohibits the sale of food that:

• • • • •

has been rendered injurious to health is unfit is so contaminated it would be unreasonable to expect it to be eaten is not of the nature or substance or quality demanded is falsely or misleadingly labelled.

It is well known that food safety hazards can be microbiological, chemical or physical in nature. While physical hazards like foreign matters that are sharp or can choke, and chemical hazards such as illegal food additives or agrochemical residues, pose serious food safety risks to all food categories, many of the known microbiological hazards that may be found in chilled foods tend to be specific to them as the prevailing environmental conditions select for the types of organisms that can survive and grow at refrigerated temperatures. Table 19.1 lists various pathogenic micro-organisms which may be associated with chilled foods, and with which chilled food manufacturers should be conversant. Today the most effective way to assure the safety of food, and to control all major types of food safety hazards, is to use the internationally recognised system based on the Hazard Analysis and Critical Control Points (HACCP) principles, as detailed in Article 5 of the EU Regulation (EC) No 852/2004 on the hygiene of foodstuffs. These principles consist of the following (European Commission, 2004): (i) (ii)

(iii)

(iv) (v) (vi) (vii)

identifying any hazards that must be prevented, eliminated or reduced to acceptable levels; identifying the critical control points (CCPs) at the step or steps at which control is essential to prevent or eliminate a hazard or to reduce it to acceptable levels; establishing critical limits at CCPs which separate acceptability from unacceptability for the prevention, elimination or reduction of identified hazards; establishing and implementing effective monitoring procedures at CCPs; establishing corrective actions when monitoring indicates that a CCP is not under control; establishing procedures, which shall be carried out regularly, to verify that the measures outlined in (i) to (v) above are working effectively; and establishing documents and records commensurate with the nature and size of the food business to demonstrate the effective application of the measures outlined in (i) to (vi) above.

© 2008, Woodhead Publishing Limited

576

Chilled foods

Table 19.1 Minimum growth conditions for micro-organisms which may be associated with chilled foods (Betts et al., 2004; Voysey, 2007) Micro-organism

Minimum growth Minimum temp. (°C) pH for growth

Salmonella Staphylococcus aureus Bacillus cereus (spores/heat resistant) Clostridium botulinum (proteolytic A, B, F) Clostridium botulinum (non-proteolytic B, E, F) Listeria monocytogenes Escherichia coli Escherichia coli O157:H7 Clostridium perfringens Vibrio cholerae Vibrio parahaemolyticus Yersinia enterocolitica Aeromonas hydrophila

Minimum aw for growth

Anaerobic growth

4* 5.2* (10 for toxin) 4

3.8 4.0 (4.5 for toxin) 4.9

0.92 – 0.95 Yes 0.83 Yes (0.9 for toxin) 0.93 – 0.95 Yes

10

4.6

0.94

Yes

3* –0.4 7–8 6.5 12 10 5 –1.3 –0.1

4.7* 4.3 4.4 4.5 4.5 5 4.9 4.4 < 4.5

0.97 0.92 0.95 0.935 0.93 – 0.95 0.97 0.94 0.96 0.97

Yes Yes Yes Yes Yes Yes Yes Yes Yes

*when more than one figure is available, the lower one is given.

Earlier, Article 4 of the Regulation requires food business operators to adopt as appropriate a number of specific hygiene measures, which include compliance with microbiological criteria for foodstuffs, compliance with temperature control requirements for foodstuffs, maintenance of the cold chains, and sampling and analysis, all of which are applicable to chilled foods. Besides defining ‘shelf-life’ for the first time, Commission Regulation (EC) No. 2073/2005 on microbiological criteria for foodstuffs also establishes two types of microbiological criteria: food safety criteria and process hygiene criteria (FSA, 2005). A food safety criterion is one that defines the acceptability of a product or a batch of foodstuff, applicable to products placed on the market. Food safety criteria given in the Regulation should therefore be used to assess the safety of a food product or batch of products within the framework of an effective HACCP system. These criteria will assist chilled food manufacturers with validating and verifying their food safety management systems while they are being developed and implemented. Some of the microbiological (food safety) criteria set out in Annex I of the Regulation which are relevant to chilled foods are given in Table 19.2. When the results of testing against these criteria are unsatisfactory, the food business operators are required to take the measures laid down in Article 7 of the Regulation, which include withdrawing unsafe food from the market, and maybe product recall. Food products which have been found to be unsafe, or whose safety has been called into question, effectively have lost their shelf-life. The importance of food safety in relation to an acceptable and meaningful shelflife cannot be overemphasised. © 2008, Woodhead Publishing Limited

Table 19.2 Some food safety criteria that are relevant to chilled foods, and applicable to products placed on the market during their shelf life (taken from Regulation (EC) No 2073/2005 – Annex 1, Chapter 1 and CFA, 2005) Sampling plan+ Limits n C m M

Analytical reference method

Examples of chilled foods

1.1

Listeria monocytogenes Ready-to-eat foods intended for infants and ready-to-eat foods for special medical purposes

Ready-to-eat baby foods 10 Ready-to-eat foods intended for infants less than 12 months old Ready-to-eat dietary food for special medical purposes for infants less than 6 months old

0

Absence in 25 g

EN/ISO 11290-1

1.2

Listeria monocytogenes Ready-to-eat foods able to support the growth of L. monocytogenes, other than those in 1.1

Chilled ready-to-eat products with more than 5 days’ life Pre-packed delicatessen products Pre-packed sliced cooked meat Smoked salmon Pate Soft cheese

5

0

100 cfu/g

EN/ISO 11290-2

5

0

*Absence in 25 g EN/ISO 11290-1

Yoghurt Hard cheese Products with a pH less than 4.4, e.g. coleslaw Products with shelf-life less than 5 days, e.g. sandwiches

5

0

100 cfu/g

1.3

Listeria monocytogenes Ready-to-eat foods unable to support the growth of L. monocytogenes, other than those in 1.1

© 2008, Woodhead Publishing Limited

EN/ISO 11290-2

Shelf-life of chilled foods 577

Criterion Micro-organism and food category

578

Table 19.2

continued Analytical reference method

Examples of chilled foods

1.4

Salmonella Minced meat and meat preparations intended to be eaten raw

Steak tartare

5

0

Absence in 25 g

EN/ISO 6579

1.8

Salmonella Meat products intended to be eaten raw, excluding products where the manufacturing process or the composition of the product will eliminate the salmonella risk

Salami Parma ham Cold smoked duck

5

0

Absence in 25 g

EN/ISO 6579

1.11

Salmonella Cheeses, butter and cream made from raw milk or milk that has undergone a lower heat treatment than pasteurisation

Roquefort, Brie de Meaux

5

0

Absence in 25 g

EN/ISO 6579

1.15

Salmonella Ready-to-eat foods containing raw egg, excluding products where the manufacturing process or the composition of the product will eliminate the salmonella risk

Mayonnaise and meringues made with unpasteurised egg

5

0

Absence in 25 g

EN/ISO 6579

1.16

Salmonella Cooked crustaceans and molluscan shellfish

Mussels, prawns, shrimp, lobster, crab 5

0

Absence in 25 g

EN/ISO 6579

© 2008, Woodhead Publishing Limited

Chilled foods

Sampling plan+ Limits n C m M

Criterion Micro-organism and food category

Salmonella Sprouted seeds (ready-to-eat)

Cress, salad/ready-to-eat bean sprouts 5

0

Absence in 25 g

EN/ISO 6579

1.19

Salmonella Pre-cut fruit and vegetables (ready-to-eat)

Ready-to-eat undressed vegetable 5 salads, green salads, mixed cut lettuce salads; prepared ready-to-eat mixed fruit salads, exotic fruit salads, prepared ready-to-eat fruit

0

Absence in 25 g

EN/ISO 6579

1.20

Salmonella Unpasteurised fruit and vegetable juices (ready-to-eat)

Freshly squeezed unpasteurised fruit juices, mixed fruit juices; smoothies; vegetable juices

5

0

Absence in 25 g

EN/ISO 6579

1.21

Staphylococcal enterotoxins Cheeses, milk powder and whey powder, as referred to in the coagulase-positive staphylococci criteria in Chapter 2.2 of Annex I of the Regulation

Cheeses, excluding processed cheese and non-fermented cheese

5

0

Not detected in 25 g

European screening method of the CRL for milk

1.24

E. coli Live bivalve molluscs and live echinoderms, tunicates and gastropods

Oysters, clams, sea urchins, winkles and welks

1

0

230 MPN/100 g ISO TS 16649-3 of flesh and intravalvular liquid

1.25

Histamine Fishery products from fish species associated with a high amount of histine

Tuna, mackerel, sardines, mahi

9

2

100 mg/kg

*applies to food before it has left the immediate control of the food business operator who has produced it. + all, with the exception of that for histamine, are essentially two-class attribute sampling plans.

© 2008, Woodhead Publishing Limited

200 mg/kg HPLC

Shelf-life of chilled foods 579

1.18

580

Chilled foods

A process hygiene criterion on the other hand is a criterion that indicates the acceptable functioning of the production process. It is not applicable to products placed on the market, and should not be used to assess the safety of a food product or batch of products. It sets an indicative contamination value, again within the framework of an effective HACCP system, above which corrective actions are required in order to maintain the hygiene of the process in compliance with food law. For example, the Regulation stipulates the enumeration of E. coli as an indicator for the level of hygiene in the manufacture of ready-to-eat pre-cut fruit and vegetables, and requires improvements in production hygiene and selection of raw materials should the limit of 1000 cfu/g (M) be exceeded (sampling plan: n = 5, c = 2).

19.3 Product factors affecting shelf-life In order to determine and assign a shelf-life that is acceptable to themselves and their customers, chilled food manufacturers must have a satisfactory understanding of how their products deteriorate and spoil during storage, as well as an adequate knowledge of the main factors that influence these changes. In principle, and in practice, a food can deteriorate microbiologically, chemically, biochemically, physically and organoleptically. Indeed, all these deteriorative changes can take place at the same time, albeit at different rates. In the end, it is the most damaging change, be it microbiological or organoleptical, which will ultimately decide how and when a chilled food becomes unacceptable, because of either safety or quality issues. Knowing the relevant mechanisms and kinetics of food deterioration and spoilage, and the important factors that influence them, is a prerequisite to a successful determination and assurance of the shelf-life of food. Over the years, much knowledge about food deterioration and spoilage has been accumulated, and a number of well-known mechanisms can be used to explain the loss of shelf-life in many foods (Man, 2004):

• moisture and/or vapour transfer leading to gain or loss • physical transfer of substances such as oxygen, odours or flavours other than moisture and/or water vapour

• light-induced changes, i.e. changes caused and/or initiated by exposure to daylight or artificial light

• chemical or biochemical changes • microbiological changes • other mechanisms or changes such as a loss of an important functional property. Table 19.3 shows some examples of food deterioration and spoilage applicable to chilled foods. By definition, chilled foods have the shortest shelf-lives of processed foods, and consumers do generally perceive them as ‘fresh’. Nevertheless, knowing the factors that can influence the main deteriorative and spoilage mechanism in a chilled food does help the manufacturer in preventing premature product failure, in mitigating the adverse effects of the change, and in assigning a realistic © 2008, Woodhead Publishing Limited

Shelf-life of chilled foods 581 Table 19.3 Some examples of food deterioration and spoilage in chilled foods Product

Quality description

Cut lettuce, apple Discolouration (brown cut edges) Waldorf salad Discolouration of mayonnaise Cut vegetables Bioyoghurt Sandwiches

Limpness (loss of turgidity) Loss of function (bioactivity) Bread staleness

Fresh meat Bad/objectionable odour Freshly squeezed Separation, bad/objectionable orange juice odour and taste

Mechanism Enzymatic browning Transfer of brown colour from walnut into mayonnaise Moisture/water vapour loss Death of bioactive culture Moisture re-distribution (starch retrogradation) Microbial spoilage Microbial spoilage (yeast fermentation)

and acceptable shelf-life. Many different factors are known to affect product shelflife and they are briefly described below.

19.3.1 Raw materials In general, the quality of a finished product is largely a reflection of the quality of its raw materials, and chilled foods are no exception. Not all the quality characteristics and parameters of a raw material will have an impact on product shelf-life but those that do will need to be identified and controlled, and their effect on shelf-life established. Produce offers a good example: some produce, e.g. cabbage and apples, which are used in many chilled foods such as coleslaw and fruit pies respectively, are often laid down in controlled atmosphere storage so as to maintain a continuity of supply outside their seasons. Freshly harvested cabbage tends to have a low yeast count, whereas cabbage from cold storage has a higher yeast count (Betts and Everis, 2000). Use of the latter results in coleslaw with a markedly reduced shelf-life for parts of the year, owing to the higher starting levels of yeast introduced via the raw material.

19.3.2 Product composition and formulation The composition and formulation of a food product can be a most important shelflife determining factor in many chilled foods. Butter and margarine are water-in-oil emulsions, and by law have a high fat content (minimum 80%) that limits the growth of most micro-organisms, including pathogens and spoilage organisms (Delamarre and Batt, 1999). Although rare today, they are still prone to oxidation. The development of fat spreads with a reduced fat content, ranging from 59% through 38% (light), to 23% (diet), has seen a shift from oxidation as a potentially predominant deterioration mechanism to others, such as emulsion instability and microbiological spoilage. Consequently, emulsifiers (e.g. mono- and diglycerides and lecithin) and preservatives (e.g. potassium sorbate) are normally added to increase the shelf-lives of these products. In products where there is a real risk of © 2008, Woodhead Publishing Limited

582

Chilled foods

oxidation, vitamin E is usually added in preference to the less consumer-friendly artificial anti-oxidants. 19.3.3 Food structure Many solid and semisolid food products (e.g. sausages, mayonnaise, margarine) do not have a truly homogeneous and uniform structure. As a result, the chemical and physical conditions relevant to microbial growth and/or chemical and biochemical changes can vary with position in the food microstructure. Electron microscopy studies (Katsaras and Leistner, 1991) have revealed that the natural flora and the added starter cultures in fermented sausages are not evenly distributed, but are arrested in tiny cavities of the product; the ripening flora only grow in nests. These nests are 100–5000 µm apart, and large volumes of the sausage must be influenced by enzymes and metabolites (e.g. nitrate reductase, catalase (enzymes), lactic acid, bacteriocins (metabolites)) accumulated in such nests or cavities. The distribution and activity of these microbes in the microstructure of the sausage have a major influence on the ripening process, and hence the microbiological safety and quality of the product. Indeed, food structure is a most likely reason why in some foods real-life observations differ significantly from predictions obtained from predictive microbiological models based in liquid culture (Dens and Van Impe, 2001; Wilson et al., 2002). 19.3.4 Product assembly Product assembly or make-up may be viewed as the macrostructure of the food product. In complex and multi-component products, contact between components often results in migration of moisture, enzymes, colours, flavourings or oil from one component to another, as long as a concentration gradient exists for the substance concerned. When this happens, shelf-life can be affected if it leads to significant deterioration in quality. For instance, in fruit pies, migration of moisture from the filling to the pastry leads to a gradual loss of the desired texture, unless an effective barrier exists between the two components. And in multilayered trifles, the visual appearance is impaired by the migration of colouring components from one layer to another, in particular the sponge layer. Similarly, combining components of different microbiological status such as diced cheese or ham and coleslaw can result in a shorter shelf-life for the end product. The inclusion of a protein component in the form of diced cheese or ham creates additional interfaces with the mayonnaise and the raw vegetables in coleslaw, which allow the migration and re-distribution of acetic acid, enzymes, colour, moisture and so on, thus limiting the shelf-life of the product (Brocklehurst, 1994).

19.4 Intrinsic product properties affecting shelf-life 19.4.1 pH The pH value of a food product varies according to its composition, formulation © 2008, Woodhead Publishing Limited

Shelf-life of chilled foods 583 and processing (e.g. fermentation processes), and it needs to be controlled where acidity has a major influence on the safety and quality of the product. The pH of a system is related to the concentration of hydrogen ions which, in the case of food, come from ‘acid’ ingredients, either indigenous or added, which dissociate in water, releasing them in the process. It is well known that the growth of microorganisms becomes more inhibited as the pH of the environment in which they live decreases. Minimum pH values for growth for some pathogenic micro-organisms that may be associated with chilled foods are given in Table 19.1. The inside of a bacterial cell is at a pH close to neutral and needs to be maintained at this level for the organism to grow, develop and reproduce. In acidic and therefore hostile environments, micro-organisms have to use an extra amount of energy to maintain their internal state of homeostasis and, in consequence, their growth is slowed or prevented where energy sources are limited; they may not necessarily die, although they may be injured by intracellular acidification, particularly at chill temperatures. Certain organic and inorganic acid ingredients (Table 19.4) have specific antimicrobial effects of their own, besides their pH-lowering property. In the case of the organic acids, their preserving effect is attributed to their undissociated forms, which can freely enter the cell and thereby reduce its internal pH. Charged molecules from the dissociated forms (e.g. protons and anions) are unable to cross the bacterial membranes and enter the cell. The antimicrobial effectiveness of weak organic acids generally increases in the order acetic, propionic, sorbic and benzoic, and is dependent on the concentration of undissociated acid. The proportion of the acid that is in the undissociated form is determined by its dissociation constant, the pKa, the pH value at which half of the acid is in its undissociated form, and the pH of the food. Therefore, the effectiveness of weak organic acids as preservatives is increased in acidic foods and decreased in foods with neutral or alkaline pH values. pH also affects many chemical and biochemical changes in food, such as enzymatic and non-enzymatic browning, degradation of aspartame, and the shade of some colours. These, in turn, can have an impact on shelf-life.

19.4.2 Water activity Water activity is a term used to describe the amount of free or unbound water within a system that is available and can be used by micro-organisms. The level of available water can be lowered by physical means such as dehydration or concentration, but more usually in chilled foods, by the addition of solutes known commonly as humectants such as salt and sugar, which form chemical bonds with water and prevent it from being used. Micro-organisms, like all living organisms, require water to survive. Any reduction in water activity in their environment will cause micro-organisms to dehydrate and reduce the diffusion of metabolites to and from cells, and will ultimately inhibit or prevent their growth. Water activity (aw) is defined as the ratio of the partial pressure of water in the atmosphere in equilibrium with the substrate (e.g. a food) to that of the atmosphere in equilibrium © 2008, Woodhead Publishing Limited

584

Chilled foods

Table 19.4 Common organic and inorganic antimicrobial acids (Gould, 1996) Acids

Examples of foods in which used

Weak organic acid and ester preservatives Propionate Sorbate Benzoate Benzoate esters (parabens)

Bread, cakes, cheeses, grain Cheeses, syrups, cakes, dressings Pickles, soft drinks, dressings Marinated fish products

Organic acid acidulants Lactic, citric, malic, acetic, etc.

Low pH sauces, mayonnaise, dressings, salads, drinks, yoghurts, fruit juices and concentrates

Inorganic acid preservatives Sulphite Nitrite

Fruit pieces, dried fruit, wine, meat sausages Cured meat products

Mineral acid acidulants Phosphoric, hydrochloric

Soft drinks

with pure water at the same temperature, and is expressed on a scale of 0 to 1 where 1 is for pure water. There is no reported microbial growth below an aw of 0.6 and most bacteria (except Staphylococcus aureus) do not generally grow below an aw of 0.9. The minimum aw for growth for some pathogenic micro-organisms which may be associated with chilled foods are given in Table 19.1. In general, the water activity of a product can be considered in three main categories (Betts et al., 2004):

• Low aw (< 0.85). The significant spoilage organisms are yeasts and moulds; most pathogenic organisms should not grow at these aw levels.

• Intermediate aw (0.85–0.92). Within this range, aerobic spore-forming bacte-

•

ria, e.g. Bacillus licheniformis, may be able to grow, in addition to yeasts and moulds. Most pathogenic bacteria should be inhibited, with the exception of Staphylococcus aureus, which may be able to grow and produce toxin at aw close to or lower than 0.90. Temperature control below 8 °C should be sufficient to prevent this from occurring. High aw (> 0.92). Most pathogenic bacteria should be able to grow at this aw and above such that it is no longer a major antimicrobial factor in its own right.

Besides microbial growth, water activity values have been widely used to indicate the stability of foods with respect to the potential for chemical and biochemical changes, and physical transfer such as moisture migration. The latter is particularly relevant to multi-component or layered chilled foods.

19.4.3 Redox potential (Eh) Oxidation-reduction (redox) reactions involve a transfer of electrons between atoms or molecules, and as an atom or compound loses electrons, it is oxidised. © 2008, Woodhead Publishing Limited

Shelf-life of chilled foods 585 Oxidation of course takes places when a compound reacts with oxygen. The tendency of a medium to accept or surrender electrons, to oxidise or to reduce, is termed its redox potential (Eh). The redox potential that can be measured in food is the result of several factors (Adams and Moss, 2000):

• • • • • •

redox couples present ratio of oxidant and reductant pH poising capacity, i.e. the capacity to resist a change in the food’s redox potential availability of oxygen microbial activity.

An oxidising or reducing environment results, depending on the overall balance in the redox potential of the food. Also, as redox potential is a function of oxygen availability, it is closely linked with the extrinsic factor of storage atmosphere; the exclusion of air in vacuum packaging or canning reduces the Eh. In general, microbial growth in a food reduces its Eh, which exerts an important selective effect on the microflora of a food such that, for instance, aerobes, e.g. Bacillus spp., are capable of growth at full oxygen tensions, and obligate anaerobes, e.g. Clostridium perfringens, cannot survive in the presence of oxygen. Leistner, who pioneered the concept of ‘hurdle technology’, identified redox potential, besides competitive flora, as one of the major preservation ‘hurdles’ in assuring the safety and consistent quality of fermented (raw) meat sausages (Leistner, 2000).

19.5 Extrinsic factors affecting shelf-life 19.5.1 Hygiene Good hygiene, an integral part of good manufacturing practice (GMP), is fundamental to the manufacture of safe and wholesome food products. Hygienic design and operation requirements distinguish food from non-food manufacture, and can be viewed as part of the prerequisite programmes, fundamental to the effective application of HACCP principles. Poor hygiene leads to contamination, which may be physical, chemical or microbiological in nature, and which can have a major impact on the safety and quality of foods. Microbiologically, poor hygiene control may result in high levels of unwanted micro-organisms being introduced, which may have adverse effects on the safety and stability of the product and hence its shelf-life. Research has shown that poor slicing hygiene caused the spoilage (and reduced shelf-life) of some chilled vacuum-packed cured meats (Holley, 1997) and that the in-house flora had a definite impact on the microbiological quality and shelf-life of cold-smoked salmon (Hansen et al., 1998). While the Commission Regulation (EC) No. 2073/2005 on microbiological criteria for foodstuffs does not impose a general requirement for increased end product microbiological testing or positive release, it does highlight and re-enforce the following basic hygiene requirements: © 2008, Woodhead Publishing Limited

586

Chilled foods

• monitoring of processing areas and equipment for Listeria monocytogenes in •

•

the manufacture of ready-to-eat foods, which may pose a Listeria monocytogenes risk for public health (Article 5); monitoring of processing areas and equipment for Enterobacteriaceae in the manufacture of dried infant formulae or dried foods for special medical purposes intended for infants below 6 months which pose an Enterobacter sakazakii risk (Article 5); analysing trends in the test results (against process hygiene criteria), and taking appropriate actions without undue delay to remedy the situation in order to prevent the occurrence of microbiological risks when a trend towards unsatisfactory results is observed (Article 9).

19.5.2 Processing Processing covers a wide range of operations, which may be applied to food for a variety of purposes. It can exert a considerable effect on the microflora, physical, chemical, biochemical, nutritional and sensory properties of a food product, and hence its shelf-life. Milk is a classic example. Fresh milk that has been pasteurised usually has a few days’ shelf-life at refrigeration temperature. The same material that has been pasteurised and undergone microfiltration can last for a few weeks at chill temperature. And, if it has been given an ultra-high temperature (UHT) treatment and packaged aseptically, it can have a shelf-life of months under ambient condition. Whatever the type of processing, it must not be considered in isolation. This is particularly true with chilled foods as the processing employed (e.g. pasteurisation) is unlikely to provide adequate preservation on its own, given the current trend towards mild processing technology. As heat treatment is widely used in food manufacture and processing, it is worth summarising the three main categories of heat treatment used to stabilise foods (Betts, 2006; CCFRA, 2006):

• Mild pasteurisation to inactivate vegetative micro-organisms. Typically, this is

•

•

a process of 70 °C for 2 minutes or equivalent (z value of 7.5C°). This is primarily aimed at achieving a 6 log reduction in Listeria monocytogenes and other vegetative pathogens; it is also sufficient to inactivate most Enterobacteriaceae, Pseudomonas and yeasts that could spoil chilled foods. Severe pasteurisation to inactivate psychrotrophic or acid-tolerant spore-formers. Typically, this is a process of 90 °C for 10 minutes or equivalent (z value of 9C°) that may be given to chilled food products that are vacuum packed or modified atmosphere packed and have a shelf-life of greater than 10 days. The process is designed to achieve a 6 log reduction of psychrotrophic strains of Clostridium botulinum; it will also inactivate vegetative spoilage organisms. A similar process of 95° C for 5 minutes or 95° C for 10 minutes or equivalent (z value of 8.3C°) may be given to acidic ambient stable products, designed to inactivate acid-tolerant spore-formers that could grow and spoil the product if present after heat treatment. Sterilisation to achieve commercial sterility in low-acid canned foods. This well-known heat treatment equivalent to 121.1 °C for at least 3 minutes (Fo3)

© 2008, Woodhead Publishing Limited

Shelf-life of chilled foods 587 based on a z value of 10C° is given to low-acid canned foods to achieve a 12 log reduction of mesophilic Clostridium botulinum. The process will also inactivate all vegetative micro-organisms and the majority of spore-forming organisms capable of causing spoilage in temperate zones. Chilled foods are not usually given such a harsh treatment, which could result in inferior organoleptic quality.

19.5.3 Packaging materials and systems Packaging and packaging systems are integral parts of modern food processing and preservation. Packaging materials protect and preserve by virtue of their many different properties. Primary packaging protects food against physical damage and in many cases attack by pests, and prevents contamination during transport, storage and distribution. Suitable packaging materials offer a barrier against light, gaseous exchange and/or moisture vapour transfer, protecting the food from many of the deteriorative changes that can be shelf-life limiting. In many cases, packaging has become part of a food preservation system such as in aseptic processing and packaging, and modified atmosphere packaging (MAP). Modified atmosphere packaging extends the shelf-life of fresh red meat where the bright red colour of the meat is maintained by a high level of oxygen, e.g. 70%; excluding oxygen from airpacked chilled foods such as fresh pasta products, many of which also have a reduced aw, delays spoilage by Pseudomonas spp. and extends their shelf-lives significantly. The safety, with respect to Clostridium botulinum, of chilled foods that have been mildly pasteurised in hermetically sealed packages or heated and packed without recontamination (i.e. sous vide foods) has been evaluated by a European Chilled Food Federation ‘Botulinum Working Party’. It was concluded that for such foods, safety can be assured by a minimum heat process and strict limitation of chill shelf-life (less than about 5 days) or, for longer life products (more than about 5 days), by storage below 3 °C, by heat treatment sufficient to deliver a 6 log reduction in numbers of spores of psychrotrophic strains of Clostridium botulinum and storage below 10 °C, or by intrinsic preservation factors shown to be effective in modelling or inoculated pack/challenge tests (Gould, 1999).

19.5.4 Storage, distribution and retail display Conditions, which include temperature, humidity and lighting, experienced by chilled foods during their storage, distribution and retail display can greatly influence their shelf-lives. The current temperature control regulations in the UK stipulate a maximum of 8 °C during distribution and retail display of chilled foods. Compliance with temperature control requirements for foodstuffs, and maintenance of the cold chains, are two of the specific legal requirements contained in the EU Regulation (EC) No 852/2004 on the hygiene of foodstuffs. For chilled foods, many of which are high-risk products, controlled storage at the correct temperature and maintenance of the cold chain are therefore essential in the assurance of their microbiological safety and stability. It has been shown that temperature abuse © 2008, Woodhead Publishing Limited

588

Chilled foods

during storage could initiate yeast growth in fruit yoghurt resulting in excessive gas formation, off flavours and discolouration (Viljoen et al., 2003) whereas under proper storage conditions, this product can be expected to last up to 30 days at chill temperatures.

19.5.5 Consumer handling and use Consumer handling and use, over which the manufacturers often have little control apart from the provision of clear storage and/or use instructions on the label, can be an important shelf-life influencing factor. The little information that is available about chilled foods in the UK (Evans et al., 1991; Evans, 1998) seems to suggest that shopping and carrying-home patterns, as well as their duration and conditions, are very variable. Furthermore, in a separate piece of research, some consumers were found to use potentially unsafe practices such as transporting and storing chilled foods at the wrong temperature, holding cooked food at ambient temperature for prolonged periods, and using inadequate re-heating (Worsfold and Griffith, 1997). In the long run, widespread and sustained education of the consumers is probably the only effective remedy, which in the UK appears to have been taken seriously by the Food Standards Agency (see www.food.gov.uk/).

19.6 Interaction between intrinsic factors and extrinsic product factors In recent years, trends towards the use of fewer preservatives, lower levels of salt and, in general, milder forms of heat treatment have meant that many chilled food products do not rely exclusively on a single factor for preservation but instead depend on the interaction of a number of preservation factors such as salt, pasteurisation, chill temperature and so on, working collectively and synergistically to assure microbiological safety and stability, and give the product its character. This approach to food preservation, as mentioned earlier, is referred to as hurdle technology (Leistner, 2000). In practice, it is therefore important that all relevant hurdles are correct and preservation factors meet their target levels consistently in production. If any of the hurdles is incorrect or the target levels not met, then food safety and quality may be compromised. Thus, all should be considered in the HACCP plan, and if necessary, controlled by appropriate CCPs.

19.7 Determining product shelf-life The most common and direct way of determining shelf-life is to conduct storage trials of the product in question under conditions that mimic those it is likely to experience during storage, distribution, retail display and consumer use. For most chilled foods, this direct approach to shelf-life determination is the most appropriate approach in practice. As food safety and consistent quality are the two main © 2008, Woodhead Publishing Limited

Shelf-life of chilled foods 589 aspects of shelf-life, and both have to be designed into a food product, it is hardly surprising that shelf-life determination features prominently during the development of a new product. At least four types of shelf-life determinations can be distinguished for chilled foods, each serving slightly different purposes (IFST, 1993):

• Initial shelf-life study. This is normally conducted during the concept product

•

•

•

development stage when neither the actual production process nor the product or packaging format has been finalised. Safety (especially microbiological) of the product has either been evaluated or is evaluated alongside this study, based on HACCP principles. This initial study provides an indication of the probable mechanism by which the product is likely to deteriorate and spoil, and the main factors that affect it. It is essential to stress that concept product samples must not be offered for taste tests until their safety has been established. Preliminary shelf-life determination. This is the first detailed determination. It is normally carried out during the latter part of the kitchen/pilot development stage or when successful plant/factory trials have been completed. Information and data obtained are used to assign a provisional shelf-life, which may be included in the draft product, process and packaging specifications. Confirmatory shelf-life determination. This is normally carried out towards the end of the product development process, using product samples made under factory conditions and to a set of provisional specifications. By then, a complete HACCP study will also have been completed and a plan validated. Information and data obtained are intended to confirm or revise the provisional shelf-life previously established, and to help finalise the provisional specifications in preparation for product launch. Routine shelf-life determination. This is carried out in support of on-going production post product launch. It should include verification of the HACCP plan, assessment of the variability of product shelf-life and, if required, a revision of the assigned shelf-life and/or the production process. In certain types of products such as fresh fruits and vegetables, because of their variable nature, routine shelf-life determination is usually an integral part of the daily packing operations. Here, shelf-life test results are used to forewarn packers and retailers of potential quality problems, inform management regarding any shelf-life adjustment, and reveal temporal patterns in quality that can be used to trigger a change in the source of supply (Aked, 2000).

Comprehensive information and guidance on the evaluation of product shelf-life of chilled foods are available (Betts et al., 2004) and will not be covered in detail in this chapter. The following aspects of direct storage trials deserve some consideration.

19.7.1 Objective of the storage trial The objective of the storage trial is to determine how the trial should be designed, planned and undertaken, and how the results should be interpreted. The same © 2008, Woodhead Publishing Limited

590

Chilled foods

chilled food product destined for both retail sale and for food service from a delicatessen counter, where portions of the product are expected to be sold over a period of time and during which the container is open for most of the day, would require two different experimental designs to reflect the two applications. Likewise, the design of the storage trial should take into account whether the food is intended to be consumed in one sitting or for multiple uses. Shelf-life studies, including a review of the HACCP plan, are recommended in the following circumstances (CFA, 2006):

• • • • • • • •

new product development, modification or extension new process development or modification new packaging development range extensions any change of ingredient/s or packaging to an existing product shelf-life extensions on existing products change of production site change or movement of production equipment that could influence the site plan.

19.7.2 Storage conditions These are governed by the existing temperature control regulations applicable to chilled foods (more often by the capabilities of the logistics and retail chains, which operate at lower temperatures), and by the conditions the product being studied is likely to encounter during distribution, storage and display. A baseline storage protocol, which is intended to be representative of manufacture and distribution of most chilled foods in the UK, has been recommended (Betts et al., 2004), and is given in Table 19.5. Individual chilled food manufacturers can of course modify this protocol according to their own requirements. There is, however, little point in adopting a protocol that subjects chilled foods to excessive and unrealistic abuse in respect of temperature, or time, or both, as such conditions will invariably lead to premature deterioration and spoilage, which are unlikely to be characteristic of their actual behaviour when the products are stored properly under the recommended storage conditions. Chilled food manufacturers, however, may wish to distinguish and employ optimum, typical (average) and worse-case Table 19.5 Recommended storage protocol for chilled foods (Betts et al., 2004) Manufacturing stage

Storage temperature

Storage time

Under commercial control In-house storage at manufacturer Distribution vehicles and storage depot Retail display

5 °C or 7 °C 5 °C or 7 °C 5 °C or 7 °C

To be defined by manufacturer and/or retailer

Outside commercial control Consumer purchase Consumer storage

22 °C 7 °C

2 hours Remainder of shelf-life

© 2008, Woodhead Publishing Limited

Shelf-life of chilled foods 591 storage conditions provided they have the confidence about these conditions, and have the capabilities to simulate them.

19.7.3 The end of shelf-life The main task of a shelf-life study is to find out as accurately as possible, under specified storage conditions, the point in time at which the product has become either unsafe or unacceptable to the target consumers, and if the product meets its shelf-life objectives. Regarding microbiological safety and stability, the following should be of help in fixing an end-point for chilled foods: (i)

relevant food legislation, e.g. Commission Regulation (EC) 2073/2005 on microbiological criteria for foodstuffs; (ii) guidelines for the microbiological quality of some ready-to-eat foods (Gilbert et al., 2000) given by enforcement authorities or agencies in support of their work, e.g. those given by the UK Health Protection Agency (previously the UK Public Health Laboratory Service); (iii) guides on microbiological criteria for foods provided by independent professional bodies such as the UK Institute of Food Science and Technology (IFST, 1999); (iv) guides on microbiological criteria for foods produced by independent food research associations such as Campden and Chorleywood Food Research Association (Voysey, 2007); (v) current industrial best practice as published in the primary literature, which suggests probiotic functional foods and drinks should contain at least 107 live and active bacteria per g or ml for their functional claims to be maintained over the shelf-life period (Knorr, 1998; Holzapfel et al., 1998; Shortt, 1999; Birollo et al., 2000); (vi) predictive models (e.g. ComBase) as outlined in Section 19.7.5. Non-microbiological criteria that are used to set shelf-life are much more productspecific. In an ideal situation, these criteria are either contained in the original marketing brief or can be developed from it. Crucially, the criteria, be they physical, chemical or sensory, need to be correlated to the quality attributes that are critical to product acceptability, and hence quality (as opposed to safe) shelf-life and, where appropriate, they should be agreed between the manufacturer and its customer. Once product safety has been assured, sensory evaluation is the most popular means by which the end of shelf-life is determined. This is important because safe food does not necessarily mean organoleptically acceptable food to the consumer. In the absence of a trained panel of sensory assessors, consumer tests give a useful measure of liking that can be used more directly to estimate shelf-life (Kilcast, 2000). The most common procedure is to ask consumers representative of the target population to scale acceptability on a nine-point hedonic scale, anchored from ‘like extremely’ to ‘dislike extremely’. A minimum of 50 consumers should be recruited for the test, although lower numbers (32–40) have been used. Some examples of sensory characteristics which may change © 2008, Woodhead Publishing Limited

592

Chilled foods

Table 19.6 Examples of sensory characteristics which may change during storage (adopted from Betts et al., 2004) Descriptor Appearance Browning Discoloration Colour loss Darkening Thinning Sogginess

Sensory characteristics Cut surfaces turning brown, e.g. apple, lettuce Development of different colour, e.g. greying of cooked potato, pinking of cooked poultry meat;over-cooked products (nonenzymatic browning) Disappearance of the usual colour, e.g. loss/fading of red colour of cooked cured meat Colour becoming darker, e.g. tomato products Reduction in consistency of sauces, gravies, etc. Appearance of having absorbed moisture or liquid, e.g. pastry products, sponge cakes Loss of turgidity, product looking tired, e.g. vegetable crudités

Limpness Odour/flavour Flavour loss Loss of characteristic/typical flavour, e.g. fruit and vegetables Sourness/cheesiness Acid, lactic flavour of old milk Mouldy Odour/flavour associated with mould growth, e.g. mouldy sandwiches Alcoholic Resembling wine, found in flavoured milk products and some modified atmosphere packs; fermenting yoghurts Ammoniacal Pseudomonas on spoiling meat and poultry products Cardboard/stale Odour/flavour associated with old/rancid fat products ‘Off’ notes Associated with specific organisms and/or specific foods, e.g. putrid, rotting meat dishes Texture Firmness High resistance to deformation, e.g. carrot Crispness Tendency to yield suddenly, e.g. celery Crunchiness Making a crunching noise when bitten, e.g. apple Chewiness Tough and fibrous texture requiring much chewing, e.g. tough meat, citrus fruit with tough membranes Sogginess Associated with uptake of moisture/liquid, e.g. pastry Dryness Lack of moisture, e.g. stale old bread

during storage, and their corresponding descriptors, are given in Table 19.6. Another more powerful and informative technique for shelf-life assessment is quantitative descriptive analysis (QDA), which requires the use of a small panel of highly trained (6–12) assessors, and the use of training samples that illustrate quality changes that occur during storage. The technique involves three main steps: development of a standardised vocabulary; quantification of selected sensory characteristics that can be used to describe critical quality changes during storage; and analysis of the results using parametric statistical methods. The results may be presented as spider diagrams aiding interpretation (Lewis and Dale, 1994). These requirements together with the need for a dedicated sensory evaluation facility often preclude small chilled foods manufacturers from employing QDA to study shelf-life. Given that many of the raw materials used in food processing and manufacture © 2008, Woodhead Publishing Limited

Shelf-life of chilled foods 593 are biological in origin, and microbial responses are naturally variable even under a defined set of conditions, it is vitally important that shelf-life storage trials are replicated a number of times to provide sufficient confidence in the assigned shelflife. And as a precaution, it is prudent to build into the assigned shelf-life a generous margin of safety, which can always be revised in light of further experience. Relevant information, for instance obtained via customer complaints and/or a consumer ‘help line’, may be used to inform a management decision should it be necessary to revise product shelf-life.

19.7.4 Challenge testing As pointed out earlier, food safety is best assured by applying HACCP principles, which is also a legal requirement within the European Union. There are times, however, when it is necessary to find out if the product is likely to remain microbiologically safe and stable during its shelf-life should it become contaminated with undesirable micro-organisms (either pathogenic or spoilage) that may cause the product to be unsafe or unstable, or be exposed to higher temperatures than designed. This can be determined using microbiological challenge testing where product is deliberately inoculated with relevant pathogenic or spoilage organisms and evaluated for their potential to survive or grow within the product under normal (and abnormal) storage conditions. Results can allow the risk of food poisoning or microbial spoilage due to the organism(s) used to be evaluated if contamination or temperature abuse did occur. Clearly, challenge testing is nonroutine and requires specialist skills and expertise, which many small chilled foods manufacturers may not have. In this case, the use of an external accredited laboratory that has the necessary skills, experience and facilities could be a costeffective option. Further information on challenge testing is available (Rose, 1987; Notermans et al., 1993; Notermans and in’t Veld, 1994).

19.7.5 Predictive models for estimating microbiological shelf-life A lot has been published about predictive microbiology for almost two decades. In the first book on the subject written by McMeekin and co-authors, ‘predictive microbiology’ was defined as a quantitative science that enables users to evaluate objectively the effect of processing, distribution and storage operations on the microbiological safety and quality of foods (McMeekin et al., 1993). Since then, predictive microbiology and its applications have seen spectacular growth and development, principally in the UK and the US, but also in other countries like Australia. Today, an internet-based, common database called ‘ComBase’ (www.combase.cc), hosted by the Institute of Food Research (IFR), Norwich, UK, is publicly and freely available for research and training/education purposes, for food microbiologists, manufacturers, risk assessors, enforcement officers, and teachers and students of food-related courses. The database consists of data sets from two major software packages, namely the Food MicroModel (FMM), and the Pathogen Modeling Program (PMP) developed by a UK’s former Ministry of © 2008, Woodhead Publishing Limited

594

Chilled foods

Agriculture, Fisheries and Food (MAFF) initiative and the Eastern Regional Research Center (ERRC) of the US Department of Agriculture – Agricultural Research Service (USDA-ARS) respectively, as well as data from other members of the ComBase consortium and those compiled from the scientific literature. ComBase represents a major advancement in the evolution of predictive modelling and its widespread application, as the data contained in it have been standardised and organised in a manner that permits efficient access and retrieval (Baranyi and Tamplin, 2004). As has been possible for some time, predictive models can be used to predict the growth rates of bacteria under various conditions, and obtain data on heat inactivation, survival and time to toxin production (where applicable), which can be applied to all stages of the manufacturing process. For instance:

• • • • • • • •

new product, process and packaging development product reformulation microbial hazard identification estimating microbiological shelf-life (safety and stability) setting microbiological specifications trouble shooting alternative to challenge testing aid to quantitative microbiological risk assessment.

As an illustration, the effect of changes in storage temperature on growth of Yersinia enterocolitica can be seen in Fig.19.1, which shows predicted growth of the organism at three different temperatures in a broth culture at pH 6.5, containing 2% salt with an initial inoculum level of 103 cfu/ml. The predictions were obtained using the PMP7 Release 1 (PMP, 2005: www.arserrc.gov/mfs/pathogen.htm). It remains a fact, however, that the knowledge, skills and experience of a food microbiologist are invaluable in interpreting the information and results from predictive microbiology.

19.8 Future trends In the UK, the chilled foods sector has undoubtedly been a success story. The total UK chilled prepared foods market has been estimated to be worth in excess of £8bn at retail selling prices by the end of 2004 (Anon., 2005). The sector has performed much better than the food sector as a whole for a number of years, and this is expected to continue. Shelf-life is about safety and consistent quality. Almost by definition, chilled foods have relatively short shelf-lives, and with this, there is a general perception by consumers that they are ‘fresh’, ‘wholesome’ and ‘healthy’. As has been explained in this chapter, and discussed elsewhere in this book, many chilled foods are high-risk products, and food safety is of paramount importance to the chilled foods sector as it is to the rest of the industry. Recent consumer-driven trends towards fewer artificial preservatives, lower levels of salt, milder forms of heat processing, and fortification of foods with minerals and vitamins have created a greater technological challenge for chilled foods manufacturers, not least in the © 2008, Woodhead Publishing Limited

Shelf-life of chilled foods 595 Yersinia enterocolitica in broth culture – effect of temperature on growth

12 10

Log cfu/ml

8

7 °C 6 °C

6

5 °C

4 2 0 0

5

10

15

20

25

Days

Temperature °C

Days to 106 cfu/ml

7

5.5

6

6.5

5

8

Fig. 19.1 Predictions from Pathogen Modeling Program (PMP7 Release 1, 2005) (broth culture – pH: 6.5, salt: 2.0%, initial level of inoculum: 1000 cfu/ml).

assurance of food safety. To ensure continued success, more effort and investment, not only in research and development but also in training and education, will be needed by all concerned: suppliers, manufacturers, caterers, retailers, trade associations, professional bodies, enforcement agencies, etc. to maintain and raise the standard of safety of chilled foods, and to continue to provide consumers with enjoyable products of safe, consistent and acceptable shelf-lives.

19.9 Sources of further information and advice There are numerous papers and articles on the shelf-life of chilled foods and related areas in the primary literature. The following are a number of publications on shelflife which are particularly useful: ANON.

(2006) Extension of Product Shelf-life for the Food Processor. A strategic report compiled for the Food Processing Faraday by the Scientific and Technical Information Section, Leatherhead Food International, Leatherhead, UK.

© 2008, Woodhead Publishing Limited

596

Chilled foods

BETTS, G D, BROWN, H M AND EVERIS, L K (eds) (2004) Evaluation of Product Shelf-life for

Chilled Foods. Guideline No. 46, Campden and Chorleywood Food Research Association, Chipping Campden, UK. ESKIN, N A M AND ROBINSON, D S (eds) (2000) Food Shelf Life Stability: Chemical, Biochemical and Microbiological Changes. CRC Press, Boca Raton, USA. IFST (1993) Shelf Life of Foods – Guidelines for Its Determination and Prediction. Institute of Food Science and Technology, London, UK. KILCAST, D AND SUBRAMANIAM, P (eds) (2000) The Stability and Shelf-life of Food. Woodhead Publishing Limited, Cambridge, UK. MAN, D (2002) Shelf Life. Food Industry Briefing Series, Blackwell Science, Oxford, UK. MAN, C M D AND JONES, A A (eds) (2000) Shelf-life Evaluation of Foods, second edition, Aspen Publishers, Gaithersburg, Maryland, USA. STEELE, R (ed) (2004) Understanding and Measuring the Shelf-life of Food. Woodhead Publishing Limited, Cambridge, UK.

19.10 References ADAMS, M R AND MOSS, M O (2000) Food Microbiology. Second Edition. The Royal Society

of Chemistry, Cambridge, UK. AKED, J (2000) Fruits and Vegetables. In: The Stability and Shelf-life of Food. Kilcast, D and

Subramaniam, P (eds.), pp 249–278. Woodhead Publishing, Cambridge, UK. ANON. (2005), Chilled Foods, Key Note Limited, Middlesex UK. BARANYI, J AND TAMPLIN, M (2004) ComBase: A Common Database

on Microbial Responses to Food Environments. Journal of Food Protection, 67(9), 1967–1971. BETTS, G D (2006) Determining the stability and shelf-life of foods. In: Food Spoilage Microorganisms. de W Blackburn, C (ed.), pp 119–143. Woodhead Publishing, Cambridge, UK. BETTS, G D, BROWN, H M AND EVERIS, L K (eds) (2004) Evaluation of Product Shelf-life for Chilled Foods. Guideline No. 46, Campden and Chorleywood Food Research Association, Chipping Campden, UK. BETTS, G AND EVERIS, L (2000) Shelf-life determination and challenge testing. In: Chilled Foods – A Comprehensive Guide, 2nd edn., Stringer, M and Dennis, C (eds), pp 259–285. Woodhead Publishing, Cambridge, UK. BIROLLO, G A, REINHEIMER, J A AND VINDEROLA, C G (2000) Viability of lactic acid microflora in different types of yoghurt. Food Research International, 33, 799–805. BROCKLEHURST, T F (1994) Delicatessen salads and chilled prepared fruit and vegetable products. In: Shelf Life Evaluation of Foods, Man, C M D and Jones, A A (eds), pp 87– 126. Blackie Academic and Professional, London, UK. CCFRA (2006) Pasteurisation: A Food Industry Practical Guide. Guideline No. 51, 2nd edn., Campden and Chorleywood Food Research Association, Chipping Campden, UK. CHILLED FOOD ASSOCIATION (CFA) (2006) Best Practice Guidelines for the Production of Chilled Foods, 4th edn., TSO, Norwich, UK. CHILLED FOOD ASSOCIATION (CFA) (2005) Guidance on the Practical Implementation of the EC Regulation on Microbiological Criteria for Foodstuffs. Chilled Food Association, Kettering, UK. CRAWFORD, C (1998) The New QUID Regulations. Chandos Publishing, Oxford, UK. DELAMARRE, S AND BATT, C A (1999) The microbiology and historical safety of margarine. Food Microbiology, 16, 327–333. DENS, E J AND VAN IMPE, J F (2001) On the need for another type of predictive model in structured foods. International Journal of Food Microbiology, 64, 247–260. EEC (1979) Council Directive 79/112/EEC of 18 December 1978 on approximation of the laws of the Member States relating to the labelling, presentation and advertising of © 2008, Woodhead Publishing Limited

Shelf-life of chilled foods 597 foodstuffs, Official Journal of the European Communities (L33) of 8 Febraury 1979, pp. 1–14. EUROPEAN COMMISSION (2004) Regulation (EC) No 852/2004 of the European Parliament and of the Council on the hygiene of foodstuffs. Official Journal of the European Union, 25th June 2004, L 226/3–L 226/21. EUROPEAN COMMISSION (2005) Regulation (EC) No 2073/2005 on microbiological criteria for foodstuffs. Official Journal of the European Union, 22nd December 2005, L 338/1– L338/26. EVANS, J A (1998) Consumer perceptions and practice in the handling of chilled foods. In: Sous Vide and Cook–Chill Processing for the Food Industry, Ghazala, S (ed), pp 312–360, Aspen Publishers, Gaithersburg, Maryland, USA. EVANS, J A, STANTON, J I, RUSSELL, S L AND JAMES, S J (1991) Consumer Handling of Chilled Foods: A Survey of Time and Temperature Conditions. MAFF Publications, London, UK. FSA (2005) General Guidance for Food Business Operators. EC Regulation No. 2073/2005 on Microbiological Criteria for Foodstuffs. Food Standards Agency, UK. (www.food.gov.uk/) GILBERT, R J, DE LOUVOIS, J, DONOVAN, T, LITTLE, C, NYE, K, RIBEIRO, C D, RICHARDS, J, ROBERTS, D AND BOLTON, F J (2000) Guidelines for the microbiological quality of some ready-to-eat foods sampled at the point of sale. Communicable Disease and Public Health, 3(3), 163–167. GOULD, G W (1996) Methods of preservation and extension of shelf life. International Journal of Food Microbiology, 33, 51–64. GOULD, G W (1999) Sous vide foods: Conclusions of an ECFF Botulinum Working Party. Food Control, 10, 47–51. HANSEN, L T, RØNTVED, S D AND HUSS, H H (1998) Microbiological quality and shelf life of cold-smoked salmon from three different processing plants. Food Microbiology, 15, 137– 150. HMSO (1990) Food Safety Act. Her Majesty’s Stationery Office, London, UK. HMSO (1996) The Food Labelling Regulations (SI 1996/1499). Her Majesty’s Stationery Office, London, UK. HOLLEY, R A (1997) Impact of slicing hygiene upon shelf life and distribution of spoilage bacteria in vacuum packaged cured meats. Food Microbiology, 14, 201–211. HOLZAPFEL, W H, HABERER, P, SNEL, J, SCHILLINGER, U AND HUIS IN’T VELD, J H J (1998) Overview of gut flora and probiotics. International Journal of Food Microbiology, 41, 85–101. IFST (1993) Shelf Life of Foods – Guidelines for its Determination and Prediction. Institute of Food Science and Technology (UK), London. IFST (1999) Development and Use of Microbiological Criteria for Foods. Institute of Food Science and Technology (UK), London. KATSARAS, K AND LEISTNER, L (1991) Distribution and development of bacterial colonies in fermented sausages. Biofouling, 5, 115–124. KILCAST, D (2000) Sensory evaluation methods for shelf-life assessment. In: The Stability and Shelf-life of Food, Kilcast, D and Subramaniam, P (eds), pp 79–105. Woodhead Publishing Limited, Cambridge, UK. KNORR, D (1998) Technology aspects related to micro-organisms in functional foods. Trends in Food Science and Technology, 9, 295–306. LEISTNER, L (2000) Minimally processed, ready-to-eat, and ambient-stable meat products. In: Shelf-life Evaluation of Foods, 2nd, Man, C M D and Jones, A A (eds), pp 242–263. Aspen Publishers, Inc. Gaithersburg, Maryland, USA. LEWIS, M AND DALE, R H (1994) Chilled yoghurt and other dairy desserts. In: Shelf Life Evaluation of Foods, Man, C M D and Jones, A A (eds), pp 127–155. Blackie Academic and Professional, London, UK. MAN, C M D (2004) Shelf-life testing. In: Understanding and Measuring the Shelf-life of Food, Steele, R (ed), pp 340–356. Woodhead Publishing, Cambridge, UK. © 2008, Woodhead Publishing Limited

598

Chilled foods

MCMEEKIN, T A, OLLEY, J N, ROSS, T AND RATKOWSKY, D A (1993) Predictive Microbiology:

Theory and Application. Research Studies Press, Somerset. (1994) Microbiological challenge testing for ensuring safety of food products. International Journal of Food Microbiology, 24, 33–39. NOTERMANS, S, IN’T VELD, P, WIJTZES, T AND MEAD, G C (1993) A user’s guide to microbial challenge testing for ensuring the safety and stability of food products. Food Microbiology, 10, 145–157. ROSE, S A (1987) Guidelines for Microbiological Challenge Testing. Technical Manual No. 20. Campden Food and Drink Research Association, Chipping Campden, Gloucestershire. SHORTT, C (1999) The probiotic century: Historical and current perspective. Trends in Food Science and Technology, 10, 411–417. VILJOEN, B C, LOURENS-HATTINGH, A, IKALAFENG, B AND PETER, G (2003) Temperature abuse initiating yeast growth in yoghurt. Food Research International, 36, 193–197. VOYSEY, P A (2007) Establishment and Use of Microbiological Criteria (Standards, Specifications and Guidelines) for Foods. Guideline No. 52, Campden and Chorleywood Food Research Association, Chipping Campden, UK WILSON, P D G, BROCKLEHURST, T F, ARINO, S, THUAULT, D, JAKOBSEN, M, LANGE, M, FARKAS, J, WIMPENNY, J W T AND VAN IMPE J F (2002) Modelling microbial growth in structured foods: Towards a unified approach. International Journal of Food Microbiology, 73, 275–289. WORSFOLD, D AND GRIFFITH, C (1997) Food safety behaviour in the home. British Food Journal, 99(3), 97–104. NOTERMANS, S AND IN’T VELD, P

© 2008, Woodhead Publishing Limited

20 Sensory quality and consumer acceptability D. Kilcast, Consultant, formerly with Leatherhead Food International, UK

20.1 Introduction The chilled foods market is highly diverse, ranging from fruit and vegetables, meat and dairy products through to complex products such as ready meals, and each part of this broad category has its own characteristics that defines consumer demand. High sensory quality, however, is a common factor that increasingly drives success across the entire category. In value terms, the total western European market for chilled prepared foods was worth over EUR14.62bn in 2005 (LFI, 2006) (Table 20.1), and sales are believed to be in excess of 3 million tonnes. This compares with up to EUR90bn for the frozen foods market in western Europe, with volume sales worth of 12 million tonnes. It should be noted that both of these figures refer to sales via foodservice, as well as retail channels. Value sales of chilled foods are particularly high compared with many frozen or ambient equivalents, reflecting the much higher prices that these products typically command. In 2006, the western European market for chilled prepared foods was expected to reach a value worth more than EUR15.15bn (Table 20.2). This represents an increase of almost 4% compared with 2005. By 2010, value sales are forecast to reach more than EUR18.15bn, up by almost a quarter (24%) from present levels. One of the major drivers in the growth of this market is that many varieties of chilled prepared foods (especially in sectors such as ready meals and pizza) are viewed as being of a higher quality than their frozen equivalents, and consumers © 2008, Woodhead Publishing Limited

600

Chilled foods

Table 20.1 Western European chilled prepared foods market by sector, 2005 (EURm) (Source: Leatherhead Food International) Ready Pizza meals UK France Germany Italy Spain Ireland Belgium Netherlands Total

2200 511 40 190 92 58 88 235 3414

452 278 12 36 298 5 86 24 1191

Coated foods

Pasta

631 405 195 140 63 12 95 Na 1541

202 286 194 547 53 14 40 16 1352

Soups Salads and sauces 246 152 25 83 70 16 5 18 615

1285 495 720 295 18 45 Na Na 2858

Bakery

Total

2935 574 25 Na 44 73 Na Na 3651

7951 2701 1211 1291 638 223 314 293 14622

Na = Not applicable

Table 20.2 Forecast sales of chilled prepared foods in Western Europe, 2006–2010 (EURm) (Source: Leatherhead Food International) Sector Ready meals Pizza Coated foods Pasta products Soups Sauces Prepared salads Sandwiches Savoury pastry products Total

2006

2010

% per annum change, 2006–2010

3607 1223 1513 1350 323 327 3006 2044 1763 15156

4390 1374 1827 1539 445 393 3690 2525 1968 18151

5.0 3.0 4.8 3.3 8.3 4.7 5.3 5.4 2.8 4.6

are prepared to pay more as a result. This is especially true of the retail-led UK market, which has generated intensive competition between the major retailers, and has in turn driven the need to optimise sensory quality to maintain competitive advantage. When considering sensory quality issues, it must be remembered that many factors other than sensory characteristics can influence consumer purchase decisions. For some years, psychology researchers have been developing models to understand consumer behaviour (Shepherd and Sparks, 1994). There are many possible circumstances under which non-sensory factors such as price and nutritional image can have dominant effects. Although the sensory characteristics of foods are central to continued purchase of foods, care should be taken not to overlook these extrinsic factors. The sensory evaluation of food is frequently defined by the term ‘tasting’, but this term is clearly inadequate to describe all the perceptual processes involved in eating food. When we eat food, we perceive a whole range of different character© 2008, Woodhead Publishing Limited

Sensory quality and consumer acceptability

601

istics relating to the appearance, flavour and texture of the food. It is now recognised that all these stimuli are integrated in the brain, and cannot be considered in isolation (Rolls, 2005). Physiological differences between individuals result in a range of responses to these stimuli, and we must expect these disparate responses to be encountered within a given consumer population. Further, differences in ethnic and cultural backgrounds and in experiences of foods will again broaden the response of consumers to foods. In using sensory methods, we must be prepared not only to encounter and work within this wide response, but also to interpret data generated by sensory measurements in the context of the target consumer population. There is often a temptation to interpret measured sensory changes in terms of perceived quality, but unless consumer requirements are well understood, this can be a risky process. In general, we dislike extremes, preferring intermediate levels of a sensory characteristic, leading to an inverted-U relationship, and simple linear relationships are not often seen within a typical consumer population, although a range of relationships can be seen in segmented populations. Assessment of sensory quality can therefore be approached in one of two ways: from measurement of sensory characteristics, or from measurement of consumer liking.

20.2 Consumer requirements for sensory quality Consumers in general do not analyse the characteristics of the food they are eating. Whilst a specific flavour or textural characteristic might hold some key importance, either positive or negative in direction, consumers subconsciously synthesise all the sensory inputs into a decision in some form of ‘I like’ …… ‘I don’t like’ scale. In an increasingly affluent population in the western world, the quality levels demanded by consumers are steadily rising, even with the increased price premium that this entails. One aspect of changing consumer demands is that heavily processed foods are increasingly being perceived as unhealthy products of an unscrupulous food industry, and even though the convenience aspects of dried, canned and frozen foods are recognised, there has been a steady consumer demand for less processed foods that are perceived as both more ‘fresh’ and more ‘natural’, even though the understanding of these concepts is often uncertain. This changing pattern has been reflected in upheavals in the food market; for example in the sale in 2006 by Unilever of the Birds Eye frozen foods business.

20.3 Components of sensory quality Human beings employ a range of senses in perceiving food quality (Table 20.3). The discussion below summarises these senses briefly. Fuller descriptions can be found in the following references: appearance in Hutchings (1999); odour, taste and flavour in Taylor and Roberts (2004); and texture in Bourne (2002) and Kilcast (2004). © 2008, Woodhead Publishing Limited

602

Chilled foods Table 20.3 The human senses Sense

Perception

Vision Gustation Olfaction Chemical/trigeminal Touch Hearing

Appearance Taste Odour/aroma Irritant Texture Texture

20.3.1 Appearance The visual senses are of particular importance in generating an initial impression of food quality that often precedes the input from the remaining senses. Indeed, if the appearance of the food creates a negative impact, then the other senses might not come into play at all. The visual sense is often equated only with colour, but provides input on many more appearance attributes that can influence food choice, for example size, shape, surface gloss, and clarity. In particular, the visual senses can provide an early, and strong, expectation of the flavour and textural properties of foods. This can occur even if the food is packaged, when photographic images and product descriptions can generate an expectation of quality that must be fulfilled.

20.3.2 Odour The odour response is complex, with odours being detected as volatiles entering the nasal cavity, either directly via the nose or indirectly through the retronasal path via the mouth. The odorants are sensed by the olfactory epithelium, which is located in the roof of the nasal cavity. Some 150–200 odour qualities have been recognised, and there is a very wide range (ca 1012) between the weakest and the strongest stimulants (Meilgaard et al., 1999). The odour receptors are easily saturated, and specific anosmia (blindness to specific odours) is common. It is thought that the wide range of possible odour responses contributes to variety in flavour perception. Both taste and odour stimuli can be detected only if they are released effectively from the food matrix during the course of mastication.

20.3.3 Taste Taste (gustation) is strictly defined as the response by the tongue to soluble, involatile materials. These have classically been defined as four primary basic taste sensations: salt, sweet, sour and bitter, now extended to include the savoury response, umami. This list is sometimes extended to include sensations such as metallic and astringency. However, the metallic response is imprecise, and can arise from both metal ions and rancidity products, and the astringency response arises from loss of lubrication from reaction of salivary proteins with phenolic © 2008, Woodhead Publishing Limited

Sensory quality and consumer acceptability

603

materials in the food. The taste receptors are organised groups of cells, known as taste buds, located within specialised structures called papillae. These are located on the tip, sides and rear upper surface of the tongue, and also on other soft oral surfaces. Maximum sweetness sensitivity lies primarily on the tip of the tongue; salt and sour on the sides of the tongue; and bitter on the rear of the tongue. Taste stimuli are characterised by the relatively narrow range between the weakest and the strongest stimulants (ca 104), and are strongly influenced by factors such as temperature and pH (Meilgaard et al., 2006).

20.3.4 Chemical sense The chemical sense corresponds to a pain response arising from stimulation of the trigeminal nerve. This is produced by chemical irritants such as ginger and capsaicin (from chilli), both of which give a heat response, and chemicals such as menthol and sorbitol, which give a cooling response. With the exception of capsaicin, these stimulants are characterised by high thresholds. The combined effect of the taste, odour and chemical responses gives rise to the sensation generally perceived as flavour, although these terms are often used loosely.

20.3.5 Flavour Flavour is not a single entity, but a complex response arising from the aroma, taste and chemical responses (Fig. 20.1). The taste involatiles provide the base to the flavour, the aroma volatiles the variety, and the chemical stimuli the excitement. Each of these classes of stimuli are released from the food in different ways, and are perceived through different sensory mechanisms. Successful flavour balance requires a careful harmonisation of the physical processes that release these components.

Chemical: Excitement

Smell: Variety

Taste: Base notes

Fig. 20.1 © 2008, Woodhead Publishing Limited

Flavour as a composite sensation.

604

Chilled foods

Fig. 20.2

Interactive nature of the senses.

20.3.6 Texture Texture is perceived by the sense of touch, and comprises two components: somesthesis, a tactile, surface response from skin, and kinesthesis (or proprioception), which is a deep response from muscles and tendons. For many foods, visual stimuli will generate an expectation of textural properties. The touch stimuli themselves can arise from tactile manipulation of the food with the hands and fingers, either directly or through the intermediary of utensils such as a knife or spoon. Oral contact with food can occur through the lips, tongue, palate and teeth, all of which provide textural information (Kilcast, 1999). The descriptions given above, whilst appropriate for the individual sensing modalities, fail to take into account their interactive nature, shown schematically in Fig. 20.2. These interactions have been extensively reviewed by Cardello (1996). Colour, which is obviously an important appearance characteristic, can be shown to have an influence on flavour perception. For example, Dubose et al. (1980) found significant increases in perceived flavour intensity in beverages with increasing colour intensity. Textural properties of foods have substantial effects on the percep-tion of flavour, and sound emission from crisp and crunchy foods has been shown to be of great importance in the perception of their texture (e.g. Vickers, 1991). The importance of the interaction between the texture of foods and their perceived flavour can be clearly seen if the time course of events during food consumption is considered. As already indicated, strong expectations of the flavour and texture characteristics can be generated before the food is introduced into the mouth. As food enters the mouth, and is either bitten or manipulated between tongue and palate, catastrophic changes occur to the structure of the food © 2008, Woodhead Publishing Limited

Sensory quality and consumer acceptability

605

that strongly influence the way in which tastants and odorants are released from the food. These processes can result in important effects on perceived flavour, and can produce substantial changes in flavour and texture quality if changes to food structure have already occurred on storage.

20.4 Techniques for quality assessment The complex nature of food quality perception creates many difficulties for the sensory analyst, whose primary task is to use human subjects as an instrument to measure the sensory quality of foods. The factors that should be considered in assessing the performance of human subjects in this way are accuracy, precision and validity (Piggott, 1995). Sensory measurements are a direct measure of human response, and have an inherently higher validity than instrumental measures, which are nonetheless of value as a complement to sensory data in shelf-life assessment. In measuring human responses, low precision must be expected, but variation can be reduced by careful selection of a range of human subjects who can produce a response with lower variability, and by extensive training. Improving accuracy (giving the correct answer without systematic error or bias) can be achieved by recognising the various sources of physiological and psychological biases that can influence human subjects (Meilgaard et al., 2006). The effect of physiological differences between individuals can be reduced, but not completely eliminated, by careful selection procedures. Psychological factors can introduce systematic biases that might not be recognised. These include those arising from unwanted interaction between panellists, and those from more subtle sources. These can be greatly reduced by choice of sensory test procedure and by careful experimental design and operation of sensory test procedures. In developing and implementing a high-quality sensory evaluation system, a number of inter-related requirements can be defined; these are discussed below, and more detailed discussions can be found in standard texts (e.g. Piggott, 1988; Meilgaard et al., 2006; Muñoz et al., 1992; Stone and Sidel, 1993; Lawless and Heymann, 1998). The requirements are shown schematically in Fig. 20.3. 20.4.1 Clearly defined objectives Clear objectives are central to the establishment of any system that will be sufficiently accurate to measure the required sensory characteristics with the required precision and that will be cost-effective. Problems commonly seen in industry include: underestimation of panellist requirements, including enforced changes in personnel; ambiguity in the type of sensory information to be generated; and absence of guidelines on the interpretation of storage changes in terms of consumer response. 20.4.2 Sensory testing environment A suitable environment is essential for generating high-quality sensory data with © 2008, Woodhead Publishing Limited

606

Chilled foods

Environment

OBJECTIVE

Subjects

Validation Tests

Fig. 20.3

Requirements for sensory analysis.

minimal bias. The environment is important not only in providing standardised working conditions for the assessors, but also in providing a work area for sample preparation and for data analysis. Detailed advice is given in a number of publications (e.g. Stone and Sidel, 1993; ISO 8589, 1988). The three main components of a sensory evaluation environment are:

• A preparation area of adequate size and appropriately equipped • A testing environment, adjacent to, but separated from, the preparation area • Individual booths to eliminate assessor interaction. 20.4.3 Selection of suitable test procedures Many sensory test methodologies are available, but they fall into two main classes, shown schematically in Fig. 20.4:

• Analytical tests. These tests are used to measure sensory characteristics of products by providing answers to the questions: • Is there a difference? • What is the nature of the difference(s)? • How big is(are) the difference(s)?

• Hedonic/affective tests. These tests are used to measure consumer response to sensory characteristics of the products by providing answers to the questions: • Which product is preferred? • How much is it liked? The two classes comprise tests that satisfy completely different objectives, and which are subject to different operating principles. Analytical tests use human © 2008, Woodhead Publishing Limited

Sensory quality and consumer acceptability

607

Sensory testing procedures

Analytical tests

Difference tests

Hedonic tests

Quantitative tests

Preference Acceptability

Paired comparison

Simple descriptive

Duo–trio

Profiling

Triangle

Time–intensity

Fig. 20.4

Classification of sensory test procedures.

subjects as a form of instrument to measure properties of the food. Hedonic tests measure the response of consumer populations to the food in terms of likes or dislikes. Different psychological processes are used for each type of test, and in general there is no simple linear relationship between the two types of data, with the inverted U-shape relationship being typical. Of great practical importance, the type and numbers of subjects used for the analytical and hedonic tests are quite different.

20.4.4 Selection and training of suitable test subjects The subjects to be used are defined by the objective of the test and by the consequential choice of test. The numbers of subjects to be used depends on the level of expertise and training of the assessors: in general, the lower the level of expertise, the higher the numbers required.

20.5 Analytical test methods using trained sensory panels Both discriminative and descriptive tests use small panels of assessors chosen for their abilities to carry out the tests. Guidelines for establishing such assessors are given in ISO 8586-1 (1993). A general scheme for establishing a panel requires the following steps:

• Recruitment. Panellists can be recruited from within the company, or dedicated •

part-time panellists can be recruited from the local population (company employees should not be compelled to participate). Screening. These preliminary tests are used to establish that sensory impairment is absent, to establish sensitivity to appropriate stimuli and to evaluate the ability to verbalise and communicate responses. Selection of suitable panellists

© 2008, Woodhead Publishing Limited

608

•

•

Chilled foods

is usually made on the basis of a good performance across the entire range of tests, rather than excellence in some and poor response to others. If the panel is to be used for a specific purpose, then the tests relevant to that purpose can be weighted appropriately. Training. In the initial stages, training is limited to the basic principles and operations, following which further selection can be made. More closely targeted training can then be carried out using the products of interest and aimed towards the specific tests to be used in practice. Monitoring. Close monitoring of panel performance is essential, and any drift that is identified must be corrected by retraining procedures.

20.5.1 Discrimination tests Discrimination tests are perceived as one of the easiest classes of sensory testing to apply in an industrial environment, and are consequently heavily used. The tests can be used in two ways: to determine whether there is an overall difference between two samples, or to determine whether one sample has more or less of a specific attribute than another. However, there are inherent limitations to such tests, for example the restricted information content and the difficulty in determining whether the absence of a difference can be interpreted as the samples being the same. Consequently, such tests are often overused in circumstances in which alternatives, such as quantitative descriptive methods, would be superior. In industry, difference tests are almost universally used to ascertain whether two samples are different, not to ascertain whether two samples are the same. However, current revisions of ISO standards advise sensory analysts on how to use the tests for the latter purpose, although there is controversy over their use. Paired comparison test In the most common form of this test (less commonly referred to as the 2-AFC, alternative forced choice, test), two coded samples are presented either sequentially or simultaneously in a balanced presentation order (i.e. AB and BA). There are two variations in the test. In the directional difference variant, the panellists are asked to choose the sample with the greater or lesser amount of a specified characteristic. The panellists are usually instructed to make a choice (forced-choice procedure), even if they have to make a guess. In the directional form, it is important that the panellists clearly comprehend the nature of the attribute of interest. Duo-trio test In the most common variant of the duo–trio test, the panellists are presented with a sample that is identified as a reference followed by two coded samples, one of which is the same as the reference and the other different. These coded samples are presented in a balanced presentation order, and the panellists are asked to identify which sample is the same as the reference. The duo–trio test is particularly useful when testing foods that are difficult to prepare in identical portions. Testing such heterogeneous foods using tests which rely on identical portions can give rise to © 2008, Woodhead Publishing Limited

Sensory quality and consumer acceptability

609

difficulties, but in the duo-trio test there are no inherent difficulties in asking the question: Which sample is most similar to the reference? Triangle test Three coded samples are presented to the panellists, two of which are identical, using all possible sample permutations. The panellists are asked to select the odd sample in a fixed-choice procedure. The increased number of samples compared with a paired comparison test can result in problems with flavour carry-over when using strongly flavoured samples, making identification of the odd sample more difficult. Difficulties can also be encountered in ensuring presentation of identical samples of some foods. Difference from control test Of particular value when a control is available, the panellists are presented with an identified control and a range of test samples. They are asked to rate the samples on suitable scales anchored by the points ‘not different from control’ to ‘very different from control’. The test results are usually analysed as scaled data.

20.5.2 Quantitative descriptive tests The major advantages of discrimination tests are their relative simplicity to set up and operate, and their high sensitivity. However, they have two important limitations. Firstly, only two sample treatments are compared together. Secondly, the information content of discrimination tests is limited, even when operated in an extended format, incorporating a range of questions. More informative tests can produce more quantitative data that can be subjected to a wider range of statistical treatments. Scaling procedures Quantification of sensory data is needed in many applications, and the recording of perceived intensity of attributes or liking requires some form of scaling procedure. These procedures should be distinguished from quality grading systems, which are used to sort products into classes defined by a combination of sensory characteristics. Such systems are not open to quantitative numerical analysis. Scaling procedures are mainly used to generate numeric data that can be manipulated and analysed statistically. Before this can be carried out, however, thought must be given to how the scales used are seen and interpreted by the assessors, and how this may influence the type of analysis that can be safely applied. The different types of scale used are described below:

• Category scales use a defined number of boxes or categories (often 5, 7 or 9, •

although other numbers are often used). The scale ends are defined by verbal anchors, and intermediate scale points are often given verbal descriptions. Graphic scales (line scales) consist of a horizontal or vertical line with a minimum number of verbal anchors, usually at the ends. Other anchors can be

© 2008, Woodhead Publishing Limited

610

• •

• •

Chilled foods

used, for example to define a central point, or to denote the position of a reference sample. Unipolar scales have a zero at one end, and are most commonly used in profiling, especially for flavour attributes. Bipolar scales have opposite attributes at either end. Definition of the central point can often give rise to logical difficulties, as can ensuring that the extreme anchors are true opposites. This can be a particular problem for textural attributes, for example when using soft………..hard type scales. Bipolar scales are frequently used for consumer acceptability testing, especially using the like extremely………….dislike extremely format. Hedonic scales are use to measure consumer liking or acceptability. Category scales are usually used. Relative to ideal scales are a type of hedonic scale which measures deviation from a personal ideal point.

In practice, establishing a trained sensory panel can often proceed from a category scale with a small number of scale points (e.g. 5), through a category scale with more points (e.g. 9) to a line scale. Sensory analysts should be aware of difficulties that panellists have in using scales, and careful training is needed to ensure that scales are unambiguous and can measure the intended response. Simple descriptive procedures Scaling may often be needed in order to quantify a single, well-defined attribute. However, it should be established that there is no ambiguity in the attribute of interest. This is particularly relevant during product development or modification, when the assumption that a process or ingredient modification will change only a single attribute is frequently violated. Such changes are especially common when textural changes are a consequence of process or ingredient modifications. If it is suspected that several attributes might be of interest, then the profiling procedures described in the subsequent sections should be considered. Quantitative descriptive analysis Quantitative descriptive analysis (QDA) is a total system covering sample selection, panellist screening, vocabulary development, testing and data analysis (Stone and Sidel, 1993). Variants of the original QDA procedures are probably used more than any other profiling procedure. The QDA technique uses small numbers of highly trained panellists. Typically, 6 to 12 people are screened for sensory acuity and trained to perform the descriptive task. Three major steps are required: development of a standardised vocabulary; quantification of selected sensory characteristics; and analysis of the results by parametric statistics. Vocabulary development Development of the vocabulary is a group process for creating a complete list of descriptors for the products under study. Panellists freely describe the flavour, appearance, odour, mouthfeel, texture and aftertaste characteristics of different © 2008, Woodhead Publishing Limited

Sensory quality and consumer acceptability

611

samples. No hedonic (good or balanced), general (full or typical) or intensitybased (strong or weak) terms are permitted. Terminology should be consistent from product to product and tied to reference materials. The references decrease panellist variability, reduce the amount of time necessary to train sensory panellists, and allow calibration of the panel in the use of intensity scales. References should be simple, reproducible and clear to the assessors, and illustrate only a single sensory descriptor. They can be single chemical substances or finished products, and are made available during both the training and the testing phase, at various concentrations or intensity. The attributes are collected and compiled into a master list. This individual preliminary evaluation of the samples may be revised during an open discussion to eliminate any redundant or synonymous descriptors. New terms might be added and physical references proposed. The panel leader condenses and formats the information into a proposal for standardised vocabulary. This vocabulary is then modified and improved in several interactive sessions. Multivariate statistical methods (e.g. factor analysis) are sometimes used to reduce the number of descriptors. Finally, definitions for the attributes are agreed. Intensity measurement Once panellists reach agreement on the vocabulary, further training is performed. The number of training sessions is dependent on the subjects’ performance, product and attribute difficulties and the time allowed for QDA testing. Panel training increases panellist sensitivity and memory, and helps panellists to make valid, reliable judgements independent of personal preferences. Once the training sessions have established satisfactory panel performance, and removal of ambiguities and misunderstandings, the test samples can be evaluated. This is usually carried out in replicated (commonly three) sessions, using experimental designs that minimise biases. The Spectrum™ method This method provides a tool with which to design a descriptive procedure for a given product category (Meilgaard et al., 1999). The method resembles QDA in many respects; for example the panel must be trained to fully define all product sensory attributes, to rate the intensity of each and to include other relevant characterising aspects such as change over time, difference in the order of appearance of attributes, and integrated total aroma and/or flavour impact. Panellists develop their lists of descriptors by first evaluating a broad array of products that define the product category. The process includes using references to determine the best choice of term and to best define that term so that it is understood in the same way by all panellists. Words such as vanilla, chocolate or orange must describe an authentic vanilla, chocolate, and orange character for which clear references are supplied. All terms from all panellists are then compiled into a list that is comprehensive yet not overlapping. The Spectrum™ method is based on an extensive use of reference points. The choice of scaling technique may depend on the available facilities for computer © 2008, Woodhead Publishing Limited

612

Chilled foods

manipulation of data and on the need for sophisticated data analysis. Whatever the scale chosen, it must have at least two, or preferably three or five, reference points distributed across the range. Free choice profiling Free choice profiling (FCP) is a very different concept, which removes the need to generate a compromise consensus vocabulary, and can also be used in consumer research (Jack and Piggott, 1992; Jack et al., 1993) . Assessors are allowed to develop their own individual vocabularies to describe sensory perceptions and to use these to score sets of samples. As a consequence of removing the need to agree vocabularies, free choice profiling requires little training – only instruction in the use of the chosen scale. Assessors merely have to be objective, capable of using line scales, and able to use their developed vocabulary consistently. Thus, assessors can be still regarded as representing naïve consumers. Characteristics being judged can be restricted by the panel leader, but the number of descriptors produced is limited only by the perceptual and descriptive skills of the assessors. A range of sensory characteristics such as appearance, flavour, aroma or texture can be examined. One particular advantage of the technique for shelf-life assessment is that new attributes that develop on storage can readily be incorporated into the profile. Disadvantages include the need to use a complex statistical analysis technique (generalised Procrustes analysis), and the absence of any agreed terminology.

20.6 Hedonic/affective testing Consumer tests are used to assess responses to sensory quality in terms of preference or levels of acceptability or liking (hedonic tests). They have a strictly limited role in probing reasons for likes and dislikes, for correlation with sensory profile data is the preferred route, using techniques such as preference mapping (Greenhoff and MacFie, 1994). Many factors contribute to the validity of consumer tests, but the central issue is the choice of the appropriate respondents. Subjects (respondents) for hedonic tests are chosen to represent the target consumer population, and to reflect any inhomogeneity in that population. Consequently, they need to be used in sufficient numbers to give statistical confidence that they are representative, and they must be given the opportunity to behave as they would in a real consumption situation. In particular, they must not be selected on the basis of sensory ability and must not be given any training. For the early stages of concept development, qualitative studies using focus groups with small numbers of respondents can be used, but the data generated should be treated carefully and conclusions must not be generalised. The most common procedure is to ask consumers representative of the target population to scale acceptability on a 9-point category scale, anchored from like extremely to dislike extremely. A minimum of 50 consumers should be used, and preferably 100, although lower numbers (32–40) can be used for exploratory work. © 2008, Woodhead Publishing Limited

Sensory quality and consumer acceptability

613

Suitable experimental designs should be used, and appropriate statistical analysis. Other information on individual modalities (appearance, odour, flavour and texture) can also be obtained, together with attribute intensity information, but it is preferable to keep such tests simple and to focus on overall acceptability. The most common procedure for operating the tests is to recruit consumers from a convenient high street or mall location and to carry out the tests in a convenient hall. Alternatively, a mobile test laboratory can be used to increase the degree of control. If there is a substantial element of consumer input in the preparation of the product, for example chilled ready meads that might be reheated in a microwave or in a conventional oven, then products might be placed in the respondent’s home for testing. Whilst this presents a real consumption environment, there is little control over the actual actions of the respondents, giving reduced test sensitivity unless high numbers of respondents are tested.

20.7 Benchmarking The term ‘benchmarking’ can have several meanings, but in the context of sensory quality it usually refers to the measurement of some aspect of sensory quality of a product against that of competitive equivalents. Depending on exact objectives, either profiling or consumer acceptability testing can be used, but increasingly a hybrid method is used that comprises elements of both. Typically, a benchmarking panel will comprise 7–50 panellists who are screened on the basis of sensory acuity and descriptive ability, but are otherwise given minimum training. Consequently, they can be seen as consumers, but not drawn from a target population in statistically representative numbers, and are used to give outline measures of consumer response. It is important to realise that such panels are not true consumer panels, and that such testing is not used as a substitute for formal consumer testing. However, they are useful for screening purposes in circumstances in which the use of profiling or consumer testing is not practical. For example, they are often used by retailers who can have several thousand product lines to assess, or manufacturers producing a wide range of products that can undergo regular changes. Such methods can be particularly useful in the catering or foodservice sector.

20.8 Maintaining sensory quality Sensory methods for use in routine assessment of sensory quality need to work within the limitations imposed by a busy production environment. Consequently the formal sensory methods used for research purposes usually need to be adapted and simplified, but it is important to retain as much objectivity as possible.

20.8.1 Sensory quality management Sensory profiling can be used for quality assessments, and gives the most complete © 2008, Woodhead Publishing Limited

614

Chilled foods

quality information (Muñoz et al., 1992). However, the methods are highly resource hungry, and accessible to relatively few companies. Most commonly, quality rating methods are used in which a small number of key attributes are preselected (often through knowledge of key consumer likes and dislikes), and which are quantified on simple rating scales. Another common method is to rate quality against a reference, either on an overall quality basis or on individual quality characteristics. The simplest method is to classify quality against specification on a binary basis: in-spec or out-of-spec. An essential requirement is to ensure that personnel used are trained adequately in the methods to be used, however simple. Increasing use is being made within the retail sector in the UK of detailed written sensory specifications that form a contract between supplier and retailer. Associated with these specifications is a simple but formalised assessment of sensory quality that is transparent to both parties, and that utilises the principles of best sensory practice. Although there is no generally agreed procedure across the retail sector, assessments are usually carried out on the basis of an extension of the in-out method to incorporate intermediate points, for example a three-category scale using ‘accept’, ‘review’ and ‘reject’, sometimes visualised as a traffic light system (green, amber, red). Good sensory practice is inherent in most of these schemes; for example all assessors must have been selected using screening systems described previously, and must have undergone training to recognise the various quality levels.

20.8.2 Sensory shelf-life assessment For microbiologically stable foods and drink products, shelf-life is limited by changes in physicochemical characteristics that are perceived as deterioration in sensory quality. The most common definition of shelf-life, as defined by the IFST in the UK (IFST, 1993), is: ‘Shelf life is the time during which the food product will: • remain safe; • be certain to retain desired sensory, chemical, physical and microbiological characteristics; • comply with any label declaration of nutritional data when stored under the recommended conditions.’ In practice, there are difficulties in using this definition to convert measured sensory characteristics to shelf-life, and a recent American standard (ASTM, 2005) had attempted to clarify this with the following definition: ‘Sensory shelf-life is the time period during which the product’s sensory characteristics and performance are as intended by the manufacturer. The product is consumable or usable during this period, providing the enduser with the intended sensory characteristics, performance, and benefits. After this period, however, the product has characteristics or attributes that are not as intended, or it does not perform the same functions as fresh products or those selected before the end of shelf-life’. © 2008, Woodhead Publishing Limited

Sensory quality and consumer acceptability

615

In each definition there is an implication that shelf-life characteristics can be measured by both trained sensory panels (to give analytical indices) and consumers (to measure hedonic response). Both types of measurement are valid, and have been described in detail by Kilcast (2000). In assessing the shelf-life of chilled foods, however, it must be recognised that the primary shelf-life limiting factor is likely to be microbiological growth, and any attempt to assess sensory changes must be designed to protect sensory assessors against any microbiological risks. This is especially important if attempts are made to accelerated changes in order to develop predictive shelf-life models. Such methods are of dubious relevance unless carried out with great care (Kilcast and Subramaniam, 2000), and if the microbial environment is changed, then microbiological safety problems can ensue.

20.9 Taints and off-flavours Taints can be defined as unpleasant odours or flavours imparted to food through external sources, and should be distinguished from off-flavours, which are also unpleasant odours or flavours, but which are imparted to food through internal deteriorative change. Although the two types can render food equally unpleasant, this distinction is of great assistance in identifying the cause of problems. The difficulties in identifying taint problems result from a number of sources. Firstly, consumer descriptions of taint are, with a few exceptions, notoriously unreliable, partly from a lack of any training in analytical descriptive methods but mainly from unfamiliarity with the chemical species responsible for taint. One notable exception is taint resulting from chlorophenol contamination, which is reliably described as antiseptic, TCP or medicinal, this reliability being a consequence of consumer familiarity with products characterised by these sensations. Secondly, the extremely low concentrations that can give rise to taint present immense difficulties for the analyst who tries to identify the chemical nature of the taint. Thirdly, taint can occur at all stages of the food manufacture and supply chain, and from many different sources at each stage. Consequently, the detective work needed to identify the cause of taint-oriented consumer complaints can be quite different for taints and for off-flavours. As a direct consequence of the low concentrations of some chemicals that cause taint, sensory panels are the main tools for establishing that foods, ingredients and food contact materials are free from taint. All panels used for quality assessment should be screened to ensure that at least some individuals are sensitive to the common taints, principally the halophenol class and their close relatives, the haloanisole class (Kilcast, 2003).

20.10 Instrumental methods Most quality assessment systems recognise the importance of sensory assessment © 2008, Woodhead Publishing Limited

616

Chilled foods

of any product prior to delivery to the retail sector and then to consumers. This frequently runs counter to modern business practices that aim to reduce dependence on human intervention and instead to place more reliance on instrumentation and automation to generate the relevant quality data. Consequently there are requirements for simple instrumental measurements of quality that can be used in place of sensory panels, and which can cover all the sensory modalities: appearance, aroma, flavour, and texture. Sophisticated instrumentation can be used carry out specific quality measurements with high precision, but it is highly unlikely that instruments will ever be able to measure the wide range of responses that the human senses are capable of, and at best they should be seen as a useful adjunct to sensory measurement. Examples of methods that are currently widely used in industry are as follows:

• Colour: Colour is difficult to assess accurately using sensory panels, being

•

• •

strongly influenced by factors such as lighting, contrasting objects and scaling procedures. Colour measuring instrumentation can be used to give parameters that are often readily correlated with perceived colour. The methods are sometimes incorporated into more flexible image analysis systems. Aroma: Many methods are available for measuring aroma volatiles, based primarily on gas chromatography and mass spectrometry, but are usually found to be too slow and expensive to use routinely. Much publicity has been given to ‘electronic noses’, which are in reality sensitive gas sensing arrays, and which do not operate in the same way as the human nose. Operating on the basis of pattern recognition, they do not identify specific chemical species, but can be trained to recognise standard and non-standard volatile patterns. Applications are not yet widespread in the food industry. Taste: Analogous to ‘electronic noses’, ‘electronic tongues’ are now available. Operating on similar principles, they are not yet used for routine testing in the food industry, but are finding applications in the pharmaceutical sector. Texture: Perhaps the most widespread type of instrumental method in use in the food industry, the methods are used for the full range of solid, semi-solid and liquid foods (Kilcast, 2001).

20.11 Future trends 20.11.1 Health and nutrition issues Most food producers are now being faced by increasing pressures to reduce levels of salt, fat and sugar in foods. Currently, the average daily intake of salt in the UK population is around 9 g/day, against a target set in 1994 in the report by the Committee on Medical Aspects of Food and Nutrition Policy (COMA) of 6 g/day. Concerns have been increasing that salt levels in the diets of children are particularly high. Consequently, food manufacturers and retailers in the UK are coming under intense pressure to reduce salt levels in processed foods, which contribute, on average, 75% of dietary salt. The target is to reduce the salt level in processed © 2008, Woodhead Publishing Limited

Sensory quality and consumer acceptability

617

foods by 30% in 3 years (Kilcast and Angus, 2007). Salt has a number of technological functions in food, and these have been described as sensory, process and microbiological (Man, 2007). Consequently, reducing salt whilst maintaining these functional characteristics is complex. Even in microbiologically stable foods in which salt has no process role, the options for reducing salt whilst maintaining acceptable sensory quality are limited (Kilcast and den Ridder, 2007). Sugar (sucrose) also has analogous sensory, structural and microbiological functions in food, but in comparison with salt many sugar reduction strategies have been devised and can be used to deliver high quality products (Kilcast et al., 2008). Fat, whilst having no preservative function, also has substantial structural and sensory functions, one of which is to act as a flavour release modifier. Fat reduction technologies are available, and some flavour companies are able to supply flavour compositions that can be tailored to provide required flavour at a range of fat concentrations. Even when suitable strategies are available for reducing salt, sugar and fat on an individual basis, more difficulties will be encountered in delivering good sensory quality in the many products that contain combinations of two or all three of these important ingredients.

20.11.2 Clean labels Driven largely by the retail sector, there is an increasing trend for companies to eliminate the use of additives that need to be declared as such on labels. This commonly involves more extensive use of natural ingredients, including replacement of synthetic flavours and colours with their natural equivalents. Such moves can have important consequences on microbiological stability and physicochemical stability, and ultimately on shelf-life. As maintaining microbiological stability is taken to be the priority, then shelf-life will potentially be determined by changes in sensory characteristics. The desire for clean labels can also impose limits on the strategies for salt, sugar and fat reduction described above.

20.11.3 Ethical issues The use of sensory panels in the food and drinks industry has too often disregarded the issues of the use of human subjects for experimental purposes. Good sensory practices have always frowned upon compulsory use of company staff on taste panels, but attendance continues to remain compulsory in many companies. Several aspects are now underlying the need for more careful use of human subjects in a manufacturing environment. Firstly, increasing allergy and intolerance problems require considerable care in selecting assessors. Secondly, if new functional ingredients are to be used, care needs to be taken when intensive assessments are to be made on products and ingredients if high intakes are delivered that might result in adverse effects. This has resulted in guidelines for the assessment of novel ingredients and processes, initially originating from the UK and now adopted by the EU (ACNFP). Thirdly, it is now increasingly recognised © 2008, Woodhead Publishing Limited

618

Chilled foods

that companies must adopt procedures designed to protect the health of any human subjects ingesting foods. These can be rigorous in nature, but a suitable starting point is the Ethical Policy published by the Institute of Food Science and Technology in the UK (IFST, 2006).

20.12 References ASTM E2454-05 (2005). Standard Guide for Sensory Evaluation Methods to Determine the

Sensory Shelf Life of Consumer Products, ASTM, USA. (2002). Food Texture and Viscosity: Concept and Measurement. Academic Press, USA. CARDELLO A V (1996). The role of the human senses in food acceptance. In Food Choice, Acceptance and Consumption, ed. H. L. Meiselman and H. J. H. MacFie, Blackie A&P, Glasgow, 1–82. DUBOSE C N , CARDELLO A V AND MALLER O (1980). Effects of colorants and flavorants on identification, perceived flavor intensity and hedonic quality of fruit flavored beverages and cake. Journal of Food Science, 45, 1393–1415. GREENHOFF K AND MACFIE H J H (1994). Preference mapping in practice. In Measurement of Food Preferences, ed. H. J. H. MacFie and D. M. H. Thomson, Blackie A&P, 137–166. HUTCHINGS J F (1999). Food Colour and Appearance, 2nd edition, Aspen, USA. IFST (1993). Shelf Life of Foods – Guidelines for its Determination and Prediction. IFST, London. IFST (2006). Policy Statement on Ethical and Professional Practices for the Sensory Analysis of Foods. IFST, London. ISO 8589 (1988). Guide to design of test rooms for sensory analysis of food, International Organisation for Standardisation. ISO 8586-1 (1993). Assessors for sensory analysis. Part 1. Guide to the selection, training and monitoring of selected assessors, International Organisation for Standardisation. JACK F R AND PIGGOTT J R (1992). Free Choice Profiling in consumer research. Food Quality and Preference, 3, 129–134. JACK F R, PIGGOTT J R AND PATERSON A (1993). Discrimination of texture and appearance in cheddar cheese using consumer free-choice profiling. Journal of Sensory Studies, 8, 167–176. KILCAST D (1999). Sensory techniques to study food texture. In Food Texture, Perception and Measurement, ed. A. Rosenthal, Aspen Publishers Inc, USA, 30–64. KILCAST D (2000). Sensory evaluation methods for shelf-life assessment. In The Stability and Shelf-life of Food, ed. D. Kilcast and P. Subramaniam, Woodhead Publishing, Cambridge, UK, 79–106. KILCAST D (2001). Modern methods of texture measurement. In Instrumentation and Sensors for the Food Industry, 2nd edition, eds. E. Kress-Rogers and C.J.B. Brimelow, Woodhead Publishing, Cambridge, UK, 518–549. KILCAST D (2003). Sensory analytical methods in detecting taints and off-flavour in food. In Taints and Off-flavours in Food, ed. B. Baigrie, Woodhead Publishing, Cambridge, UK, 5–30. KILCAST D (2004). Texture in Food. Volume 2: Solid Foods, Woodhead Publishing/CRC Press, Boca Raton, USA. KILCAST D AND ANGUS F (2007). Reducing Salt in Foods, Woodhead Publishing/CRC Press, Boca Raton, USA. KILCAST D AND DEN RIDDER C (2007). Sensory issues in reducing salt in food products. In Reducing Salt in Foods, eds. D. Kilcast and F. Angus, Woodhead Publishing/CRC Press, Boca Raton, USA, pp 201–220. BOURNE M C

© 2008, Woodhead Publishing Limited

Sensory quality and consumer acceptability

619

(2000). The Stability and Shelf-life of Food. Woodhead Publishing, Cambridge, UK. KILCAST D, DEN RIDDER C AND NARAIN C (2008). Challenges to reducing sugar in food. In Sweetness and Sweeteners: Biology, Chemistry and Pschophysics. ACS Symposium Series 979, 481–491. LAWLESS H T AND HEYMANN H (1998). Sensory Evaluation of Food. Principles and Practices. Chapman and Hall, London. LFI (2006). The European Chilled Prepared Foods Market. Leatherhead Food International, UK. MAN C M D (2007). Technological functions of salt in food products. In Reducing Salt in Foods, eds. D. Kilcast and F. Angus, Woodhead Publishing/CRC Press, Boca Raton, USA, pp 157–173. MEILGAARD M, CIVILLE G V AND CARR B T (1999). Sensory Evaluation Techniques, 3rd edition, CRC Press, Boca Raton, USA. MEILGAARD M, CIVILLE G V AND CARR B T (2006). Sensory Evaluation Techniques, 4th edition, CRC Press, Boca Raton, USA. MUÑOZ A M, CIVILLE G V AND CARR B T (1992). Sensory Evaluation in Quality Control. Van Nostrand Reinhold, New York. PIGGOTT J R (1988). Sensory Analysis of Food, 2nd edition, Elsevier Applied Science. PIGGOTT J R (1995). Design questions in sensory and consumer science. Food Quality and Preference, 6(4), 217–220. ROLLS, E T (2005). Taste, olfactory, and food texture processing in the brain, and the control of food intake. Physiology and Behavior, 85, 45–56. SHEPHERD R AND SPARKS P (1994). Modelling food choice. In Measurement of Food Preferences, eds. H. J. H. MacFie and D. M. H. Thomson, Blackie A&P, 202–226. STONE H AND SIDEL J L (1993). Sensory Evaluation Practices. Academic Press. TAYLOR A J AND ROBERTS D D (2004). Flavour Perception, Blackwell. KILCAST D AND SUBRAMANIAM P

© 2008, Woodhead Publishing Limited

21 Management of product quality and safety C. Thomas, Consultant, UK

21.1 Introduction The management of quality is an extensive topic that has troubled businesses over generations of development. However, today’s chilled food sector has stronger foundations on which to base its approaches than ever before. Many proven approaches and techniques are currently available and, as with all good things, these will continue to develop. The object of this chapter is to give some insights and guidance when entering this realm. Today, pressures on chilled food businesses have become considerable; they arise from a variety of sources including the media, customers, the end consumers and the legislature. Companies are operating in a demanding market place, where consumer expectations are high and public confidence is paramount. Over the last twenty years, the chilled food industry has evolved enormously, with chilled products becoming desirable additions to the shopping basket. Ranges available are now vast, including complex ready-meals, soups, fish, salads and a host of other innovative products. Standards throughout the manufacturing and retailing sectors have risen sharply largely due to the focus on food safety. During the growth of the industry considerable technical and scientific activity added to the picture. The need for impeccable food safety drove chilled manufacturers to invest in improved factory designs and the logistics required to ensure microbiological, physical and chemical food safety. As for the legislation, probably one of the most significant events in the food © 2008, Woodhead Publishing Limited

Management of product quality and safety

621

sector as a whole was the introduction of the Food Safety Act 1990, and the reaction to it of the larger businesses in chilled foods at the time. This Act drew together previously different pieces of legislation and assembled them in one place whilst changing the emphasis under the law regarding how a business could be convicted of an offence. The Act embraced a move away from the idea of an absolute offence, to the concept of due diligence. This had a considerable effect on the UK policy makers in companies, who recognized for the first time that they had an opportunity to shine when challenged, even when faced with the apparent errors or mistakes made by their businesses. The opportunity was to be found innocent and diligent in the public forum and under the wider scrutiny of media. This opportunity still exists today. Some organizations set themselves the goal to never be found guilty when challenged by an environmental health office complaint. This, and the need to develop an excellent public perception of the sector, probably became the single biggest driver in the adoption of an array of current food safety and product quality systems. Through this stage of the industry’s development, companies in the chilled food sector particularly invested a large amount of money to ensure their strength using the growing array of food safety and quality systems. For those aspiring to lead the UK industry, or to trade with the vast UK market place, the requirement to harness the concept of due diligence led to a hunger for a much more systemized approach to quality than had existed before. There was a new need to prove actions, and to show the strength and depth of the forethought and planning that went into safe manufacturing. Later, with the increasing development and influence of Europe, came a battery of EU regulations and the further development of regulatory affairs expertise in corporate businesses; these being used to track and examine the implications of legislation and aim to influence legislation where possible. Amongst the EU Regulations was (EC) No 852/2004 of the European Parliament and of the Council of the 29th April 2004 on hygiene of foodstuffs, which has introduced the concept of appropriate level of protection (ALOP) through manufacture and food handling as a central feature. This recognizes that the hazard analysis and critical control point (HACCP) system is not a panacea, but is the best way to produce ALOP. Additionally, the Commission Regulation (EC) No 2073/2005 of the 15th November on microbiological criteria for foodstuffs removed much of the national interpretation available previously around this subject. Current quality thinking widely embraces the concept of the systemized approach; systemization now has been universally adopted. Success of the systemized approach has endured, bringing with it a constant striving for best practice, and a raft of authoritative guidelines. These include the British Retail Consortium (BRC) standards for manufacturers supplying retailer branded goods, the Guidelines for good hygienic practice in the manufacturing of chilled foods (CFA, 1997; CODEX, 1997a), and the HACCP system and guidelines for its application, to name but a few. Smaller businesses have benefited from a halo effect. In the UK much help became available in the form of training and support from various independent bodies, and from local authorities who introduced home authority principles to environmental health officers. European guidance documents, such as the © 2008, Woodhead Publishing Limited

622

Chilled foods

implementation of procedures based on HACCP principles, European Commission Health and Consumer Protection Directorate-General, Brussels 16 November 2005, introduced some clarity for many businesses. It is now not uncommon for a food business to use a quality model as a platform for defining the quality of its whole business enterprise (for example total quality management (TQM)). The use of whole business quality modeling will be briefly discussed later in this chapter. The success of business modeling has, in part, been due to the co-operation of companies in management and quality discussion groups and the innovative approaches by management teams in individual companies who recognize the need for stronger leadership. The implementation of a whole business model is a complex scenario requiring excellent co-ordination of strategic and policy actions. Every aspect of business activity is embraced in the whole business quality model. The ideal is to produce an enterprise that operates as close to perfection as possible. However lofty the quality goals for the whole range of business functions, the need for excellence in food safety and product and process quality still sits at the core of chilled food businesses. This is the non-negotiable element; no chilled food business has any value if it cannot make safe products consistently. To this end many good management techniques have been designed into the food systems themselves; for example, the use of a well-constructed HACCP to achieve ALOP and then the management of HACCP by the use of the verification, validation and review steps in the hazard analysis critical control points plan. These ensure, if properly used, that the system is maintained in an up to date and relevant manner. Other management requirements fall into the need to clearly define those elements that are required for competent manufacturing and logistics. Some of these may be very specific to individual businesses. However, there are a set of general requirements that are widely accepted as the elements of good manufacturing practices (GMPs) throughout the food industry. Reference to documents from the International Organization for Standardization, particularly the ISO 22000 food standard and guidance document, is recommended. It is now widely accepted that for a good quality system to be in place, all GMPs should be defined, maintained and managed inside a comprehensive quality system. Commentators on the subject of quality estimate that there are over twenty schemes worldwide relating to food safety and the supply chain, all of which vary in important ways; ISO 22000 brings a level of harmonization to this. Reference to CODEX (1997b) Recommended International Code of Practice: General Principles of Food Hygiene, CAC/RCP 1-1969, Ref 3, 1997, Codex Alimentarius Commission, FOA/ WHO, Rome, is valuable here. Organizations at differing stages of development may choose different appropriate ways of achieving their quality goal. The application of quality control (QC), or quality assurance (QA) or indeed any of the business modeling approaches may be attempted. This now comes down to what is perceived to be appropriate and practical for the needs of the business in its regulatory and commercial setting. Objectively, there are many threads shared in common by all of the successful © 2008, Woodhead Publishing Limited

Management of product quality and safety

623

approaches, such as attempts to produce full staff involvement, and to measure improvement. Focused key performance indicators are now widespread, allowing collected results to be analyzed and used as a driver to focus future actions and produce further progress in target areas. Businesses that invest heavily in the achievement of their quality goals now have to look at a range of issues, particularly those associated with the rise of computer technology, which potentially change the way data is marshalled and communication is achieved in a company. Alongside this, for the success of quality systems of ever increasing size, there is a need to produce leaders throughout the company structure, not only at the top. The use of leadership skills and team working skills are widely considered to be paramount to the success of most approaches. Leadership, with its motivational quality, is adding to the mix of skills required in business, and indeed the term is replacing the term management in many organizations. This move marks a change in emphasis regarding how a business deals with its people in the working situation. To aid and stimulate people involved in the development of the quality system, which can be a culture changing process, companies may choose to form or join best practice groups. These, along with other industry forum groups, can help stimulate new ideas and help the speed of improvement by enabling managers and business leaders to embrace the experiences of others. Additionally, involvement may aid the maintenance of momentum, in what can be an extended and exacting process. The following may help to act as a springboard for those wanting to achieve and maintain high standards of quality management in their own food businesses.

21.2 Application of management and quality procedures in business When the topic of procedures and their value is approached, it is common to be challenged with thoughts of bureaucracy and expense. It is therefore worthwhile considering why any sort of procedure or procedural management should be employed in food businesses, let alone form the backbone of modern reputable chilled food manufacturing organizations. The chilled food manufacturing industry produces a vast range of extended shelf-life products by the application of food science and technology, and by the use of skilled personnel. This includes chefs and their support teams who must understand the raw materials and their effect on food safety and durability, and operational and storage personnel who must handle food in a hygienic and timely manner. Products can now be safely and conveniently distributed through a variety of routes to the mass market. This development has come about by using a multidisciplinary approach, joining sciences such as microbiology to technologies such as temperature and cook control through to good operational practices defined using verified methods. The orchestration of all this under one umbrella is the quality and product management system. © 2008, Woodhead Publishing Limited

624

Chilled foods

The potential for widespread public harm or dissatisfaction from getting it wrong in manufacturing is enormous. High numbers of products are produced through individual industrial processes every hour, and without good procedures in place it is possible for a product that has been poorly produced to be distributed widely before the problem is recognized or corrected. Clearly this must not happen in order to maintain reputation. Most members of the general public can think of a food safety issue they have heard about on TV, some of these having caused serious illness or even death to unfortunate individuals. For the manufacturers or retailers involved, the damaged to reputation is difficult to redress; loss of a valuable brand and even legal action can ensue. In extreme cases it has meant the total loss of the business. It is, however, a testament to businesses using systemized and proceduralized approaches that these instances are rare. If the product design is correct and the sourcing of raw materials is properly controlled, then the correct use of well designed manufacturing and distribution procedures ensures safety. Short life products need efficient swift handling, hence procedures must be designed to fulfill this need. So what exactly is a procedure? As with all things that make a difference to people doing a practical job of work, having a framework of actions to abide by, which are known to bring reliable results, is enormously helpful. Procedures can be seen as a manifestation of this. Individual procedures can be written on a wide range of topics, such as anything from crisis management to metal detection. Clearly this makes them a large and important area. Procedural definitions are commonly found in the manuals that accompany any well formed quality system. ISO 22000 or the BRC guidance document and certification process indicates a complete range of areas where proceduralization is required. If a problem is encountered, nothing can be surer than that the question will be asked, ‘Was a validated procedure followed?’ Many people have found to their cost that if the answer is no, then adverse consequences can ensue. The best procedures are recognized as the application of industry best practice. When applied correctly, the flexibility of the manufacturing operation is maintained, and improvements and team working are encouraged. Procedures ensure the same lessons do not have to be learned over and over, but on the other hand procedures themselves can be changed, to enable the development and improvement of the system of which they are a part. Good management and leadership principles dictate that a blame culture must not be produced, so a learning opportunity is therefore encountered both for individuals and for businesses when things go wrong with a procedure. It is evident, however, that where a prescriptive procedure meets an intuitive and innovative culture, careful leadership is required to ensure the integrity of both.

21.3 The basic need for a quality system Most people agree that to have any quality system, there first needs to be commitment at the highest level of the business. Once this is in place, the system owners and © 2008, Woodhead Publishing Limited

Management of product quality and safety

625

stakeholders have to be properly trained and motivated. Commitment comes from many motivations. These include upward or inward pressures from staff, from customers, from consumers via complaints and from local authorities. Often, the senior business leaders decide that the business must be seen to be the best at what it does to survive and grow commercially in a competitive market place. Today, all significant chilled food manufacturing companies need to be seen to be strong and competent during a variety of external audits, particularly those that look at food safety, quality and the strength of the quality system in place. Audits such as these come from commercially important customers and regulatory sources, or from third party certification bodies commissioned for the assessment, such as DNV, Lloyds register (LRQA) and SGS, to name but a few. The audits are undertaken by trained assessors against relevant documentation such as the BRC, the European Food Safety Inspection service (EFSIS) guidelines or ISO 22000, ISO 9000:2000 or on behalf of the retailers or manufacturers themselves, who may have their own audit documents and guidelines. Mostly these are aimed at assessing the food safety and quality status of a business, leading to a grade which relates to the security of doing business with a specific manufacturing company as seen during the audit process. Additionally, all UK food businesses must satisfy their local area authority regarding environmental health and trading standards issues; in Europe there is widespread use of people specially trained to vet these roles. Whilst very small businesses may have the simplest systems of all, demonstrable procedures are still of great value, both in terms of consistency of operation and in proving commitment to good practices when challenged. This means that it is wise to produce a suitable level of documentary evidence during day-to-day operation. Small businesses still need to prove they have good hygiene standards, for instance, which should be as defined by a simple but formal quality system and should be obvious on inspection. To achieve auditability it is recognized that quality systems need to be well designed: this in itself is a major task. There is said to be genius in simplicity, and it is true that the best systems are kept as simple as possible. The advent of the use of computer technology has aided the handling of monitoring data produced by the quality system in operation, but it is inevitable that some systems become large. Computer technology can make trend analysis and control of product release easy, but can also have downsides if data are not properly communicated to those who need them. In all cases it is important to ensure the data capture is consistent, and that data are properly used to help systems improvement. Some data, such as those needed for traceability, need to be easy to access in case of a problem. Quality systems give definition to the core processes in any specific business, and by doing this have a role in maintaining both the business integrity and its marketability.

21.4 HACCP and the quality system The growth and practical development of the HACCP system has been paramount © 2008, Woodhead Publishing Limited

626

Chilled foods

in aiding the food safety process in all styles of food manufacturing and handling businesses. From its early NASA days with the space programme, through to its practical development as a food safety tool in mainstream manufacturing as laid out in HACCP: A Practical Guide (1994), Campden Technical Manual 38, Issue 2, and subsequent issues, HACCP has been seen as a very powerful tool. It is now recognized internationally. HACCP, however, cannot exist in isolation; it requires prerequisites in the form of good manufacturing practices such as hygiene, pest control and suitable factory design. In other words commitment to overall quality must be present. HACCP itself, however, must form a specific and leading part of any food manufacturer’s quality system. The skills and knowledge base required to produce a good hazard analysis stretch across several specialisms in food science and technology. For example, knowledge of raw materials and the make-up of the product is vital, this knowledge often residing with the chefs or technical team. Also, a knowledge of microbiology for potential pathogen identification and control may reside with the laboratory or external authorities, dependent on the size and type of the business involved. Knowing the end use of the material or product is important to ensure that hazards and controls are included for end users. For example, whether the product is for the consumption of babies or vulnerable groups, such as those with allergies, should be asked. If that is the case, questions focused towards identifying specific hazards for those groups should be added to the picture. General allergy control aimed at safeguarding the general population must always be included: this area is covered later in the chapter. Specific members of the team who understand process engineering and good operational practice should be present during the conducting of the hazard analysis. Hazards have many sources, such as foreign bodies from equipment and hazards produced out of excessively long processing times, so it is important to be able to identify these and understand the controls fully for an effective HACCP system to be produced. There is also a specific need to have a team member with HACCP systems skills, in order to draw these individual pieces of knowledge together, and to steer the HACCP team methodically to a cohesive HACCP plan produced against the HACCP principles (CODEX, 1997a). In any specific organization there may be a skills or knowledge gap. This can happen particularly where a company does not employ a microbiologist or any another key skill needed in the production of a HACCP plan. In these circumstances it is important to recognize that the gap cannot be ignored and help should be sought. Maybe recourse to an external authoritative food organization would fill the gap. The overall point is that a HACCP plan cannot be complete or effective without a full and complete consideration of all the known or potentially predictable hazards that may be encountered whilst processing, handling or using a specific product. Quality systems come in many forms, all of which have a role to play in modern food businesses. In fact there are very few quality systems designed inhouse by individual businesses that are not hybrids of several approaches. © 2008, Woodhead Publishing Limited

Management of product quality and safety

627

21.5 Quality control QC is often considered the oldest approach. It is also one of the more durable approaches, often being incorporated at some point in the newer larger models which are more leadership oriented. QC is basically based on testing. In QC, the process operatives produce the product and the quality of the product is independently tested for its compliance to specification or other food standard. This is done by an inspection team or laboratory who then record the results and pass or fail the product against the predetermined criteria. In the better versions of quality control, these results are communicated back to the process operatives in order that errors can be understood, addressed and not reproduced. The use of independent online tests and the careful location of the quality testing room can minimize the delays in gaining test result data. However, delay is a perceived weakness of QC and some results may be significantly retrospective. If, for instance, a positive release on microbiological data is required, then the time span for this is governed by the laboratory’s ability to complete an approved test regimen. In some cases this may be days, meaning the product has to be securely held before release. The use of rapid methodology in laboratories has done much to lessen the delay. Testing has always remained a vital part of any food process, whether it is a laboratory verification check at end of process, or a measure of product quality to match, for example, a colour requirement. The way that quality control is applied is a matter for the individual business and will usually be designed to specifically match the process and product type. Much needs to be measured and analyzed during any food manufacturing process. Aspects such as temperature of the food, sizes of particles included and appearance of the product, through to moisture content, and nutritional composition, all form part of the array of potential QC checks required to prove adherence to a specification. Measurement when done away from the production line can be considered independent in nature, therefore giving QC its perceived strength. External arbiters can point to objective measures against a known standard; this makes QC a very good measure of the success of production. Its very independence however, can also be seen as its greatest weakness. People producing the food products on the production line have the most influence in making a success or failure of the manufacturing process. It is therefore considered essential in terms of modern management practicethat these people are involved in their own monitoring and improvement processes, rather than having decisions imposed after an otherwise invisible set of actions away from their influence. This brings us to the concept of quality assurance.

© 2008, Woodhead Publishing Limited

628

Chilled foods

21.6 Quality assurance and food standards In a QA system, much of the inspection and testing of the food during the process is done on the manufacturing line by the operatives producing the foodstuffs. Specific tools and training need to be provided for those on line in order to ensure a robust system. This approach places ownership for the success or failure of the process with the staff on which the process itself relies. QA relies on effective training of the operatives and full involvement of the operatives in their own areas of responsibility. QA systems also rely on consistency of action, therefore making systemized procedures necessary for repeatability and success. Whilst QA inevitability gives better job satisfaction to operatives, introducing the system into businesses that until then have used QC, needs to be carefully done. Operatives may not react as expected, and many business managers have found that operatives often argue that it is extra work and fear blame and finger pointing if failures occur. However, approached positively with good leadership, the programme can be introduced in a trouble-free manner using full consultation of the staff involved. This usually means the use of a range of motivational techniques and close involvement of staff in the introduction, the result being satisfaction for all parties without any personnel resistance. A formal set of procedures written and incorporated into the body of a QA system will, when followed, ensure the required consistency of the process. If well designed, these same procedures will, if followed, produce a good quality safe product achieving the standard required. To maximize this, rather like with the HACCP plan, a range of skills needs to be used in the design and writing of the whole quality system. Supermarkets and industrial customers of manufacturers tend to worry about the low levels of knowledge present on the process line, so correct training and appropriate corrective actions regarding what to do if things go wrong is now not only expected, but is mandatory in many supply situations. Specific emphasis is placed on the HACCP corrective actions here as these relate to food safety. However, branded houses also require very consistent general product quality to protect their brand images, which can have cost millions of pounds to establish. It is therefore necessary to take care to focus clearly on the food safety aspects and, once having secured these, then decide on a limited number of quality elements that are essential to the products’ commercial success and which should then be implementated into the QA system. The BRC represents the majority of the major supermarket chains. The BRC standards have been developed following many years of discussion and cooperation between major food retailing businesses. Their combined experience has been used to pull together the elements that must be addressed by their suppliers. A supplier covered by this quality system has to embody specific food standards in its manufacturing site that can be assessed by auditing. These standards are now getting extensive recognition. They are also used as the basis for external third party accreditation schemes. These schemes were originally meant to lower the cost of effective auditing, this being achieved by lessening the number of audits required by using a scheme whereby a company, having © 2008, Woodhead Publishing Limited

Management of product quality and safety

629

undergone a BRC audit, and having been graded, could be accepted for supply without having to be revisited by every supermarket on the business’s books. Various commentators may argue regarding the success of this goal. However, this has not detracted from uptake of the BRC standard, which embraces HACCP as a part of its basis for company food safety and product quality systems. It is necessary for a quality system that is designed to satisfy BRC requirements to define practical standards of operation for a particular manufacturing premises, its processes and products. The incorporation of HACCP is an important factor as there is a legal requirement for the full implementation and documentation of this system. Using the skill and knowledge available at large, is a good way to ensure the incorporation of best practice throughout the quality system, and to fill in the gaps in knowledge which can be present in businesses even of large size. Often, the trick is to question the technologies and product design and, based on this, find the right people to ask questions of to fill in gaps in knowledge, particularly if these relate to technologies available and to specific product design issues. This may mean making full use of the food research organizations, such as the Campden and Chorleywood or Leatherhead research associations. The major supermarkets have good technical teams, so product safety, systems and process audits in one form or another are embodied in the terms of supply. It is vital for manufacturers to be perceived amongst the best available to ensure commercial acceptability to most of the supermarkets. Once produced by the HACCP team, using the principles of HACCP itself, the system will demand full verification and validation for food safety to be assured. Producing a good HACCP system is a systematic process which is laid down in Codex and Campden documents, but here we will focus on validation and verification, which are important management tools embedded in the system itself. The next two sections deal with these areas.

21.7 Validation of the HACCP system Validation relies on ‘obtaining evidence that the elements of the HACCP plan are effective’ (FOA/WHO Codex Alimentarius Commission). In practical terms, this means taking the HACCP plan as laid down and looking for evidence in the records, or by additional sampling, to show the effectiveness of the plan. The decisions made regarding what is a realistic hazard in any HACCP plan depend on up-to-date scientific knowledge, and in-depth knowledge of material sourcing and the food production processes and factors surrounding them. The choice of prerequisites and critical control points is the way safety is assured ultimately. Many people involved with HACCP plan validation will look at two main considerations, namely the expertise present in the HACCP team that undertook the hazard analysis itself and the plan design, followed by the proper use of this expertise in the implementation of the plan when all stages are completed. It is not © 2008, Woodhead Publishing Limited

630

Chilled foods

unusual to come across extra scientific evidence that has been imported from external sources when examining a well-constructed HACCP plan. This is usually the scientific validations of the assumptions made during the work of the team, e.g. does a particular specified temperature really kill a named micro-organism in the time allotted in a particular process? Scientific validation of this is recommended, due to dangers associated with accepting without question generalizations regarding effective kill temperatures when associated with a production process. Here, it should be pointed out that heat treatments for lethality of a micro-organism are well understood. Predictive models can be useful in helping to understand lethality over a range of factors. However, process conditions that are often less well characterized may require additional testing and sampling to ensure safety. In some situations, either protective properties of the foods in question or other factors in the process influence effective kill. The same is true of generalizations made around other types of control. Scientific validations of this kind are best commissioned through laboratories or authorities that can demonstrate that their test methods conform to a recognized national or international standard. In the UK, UKAS fulfills this role. These institutions are likely to be abreast of new scientific information and methodology. By its nature, the validation step in any HACCP should come before the implementation of the plan in practice. This allows any improvements that have been highlighted during validation to be incorporated into the work. Proof of validation, like all systemized proofs, should be a fully documented process as this strengthens the plan and allows external examiners to easily judge the level of competence applied to the HACCP work done.

21.8 Verification of the HACCP system Verification is the process of ensuring that the HACCP plan is effective in operation and has been fully implemented. The verification process is systematic and forms the foundation of most internal HACCP auditing. Information has to be gathered from the operating HACCP and assessed against the system that is described in the HACCP plan. This ensures that the HACCP is doing what it was planned to do. Any changes to the manufacturing process must be effectively captured and corresponding changes must be reflected in the HACCP plan. Records and documents covering the complete HACCP system should be available during HACCP verification; these should be examined for correctness in the way they are filled out, for formatting and dating, and completeness against the plan. It is important to verify the methods that monitor critical control points (CCPs). To verify procedures surrounding CCPs, it is valuable to observe these in action. For example, if metal detection is a CCP, then the procedure by which the metal detector is set up and checked must be carried out correctly each time. Therefore, observing a check correctly carried out to written procedure by a named operator whom one can go on to prove to be trained to do the check, strengthens the © 2008, Woodhead Publishing Limited

Management of product quality and safety

631

verification process. (It will have been pre-decided during the validation step that the metal detection step itself is a valid way of controlling the metal hazards associated with the product. The implementation of the step is assessed during verification.) To strengthen the evidence gathering and ensure that nothing is missed out, many people involved in verification employ a checklist; most third-party auditors use this useful tool. These checklists can be divided into the HACCP principles and detailed with things to look for under each principle. This approach is recommended.

21.9 Traceability as a part of quality management The EU General Food Law Regulation 178/2002 defines ‘traceability’ as ‘the ability to trace and follow a food, feed, food-producing animal or substance, through all stages of production, processing and distribution.’ This regulation makes traceability compulsory for all food and feed businesses. Because of this, the traceability of foods is now an essential part of any well-constructed quality system. Issues such as BSE in Britain and dioxin levels in milk from Dutch farms in 2004 have heightened public and food industry awareness of the need for robust systems. Traceability must allow for the ingredients to be traced to source and products to be tracked to receiving customers in the event of a food issue or crisis. All traceability systems are reliant on capturing and recording correct information, and must be designed well to ensure effectiveness and maintenance over time. Traceability should be possible in both directions. A product sitting on a supermarket shelf should be coded in a way that allows the contents of that specific product to be traced back to the manufacturer and, in turn, to the ingredient suppliers, and then to the farm, field or other processor if appropriate. The alternative direction is to be able to identify the fate of an ingredient; that is, to identify into which products it was incorporated during manufacture, and then where all of these product are located. Traceability systems must tie up each link in the supply chain clearly and robustly. This includes farms, fields, processors, storage operations, distributors, and retailers. The ways of capturing and recording this data will vary considerably. Bar coding or manual paper systems based on unique codes and numbering are common approaches. Sector specific legislation applies to certain categories of food product – fruit, vegetables, beef, fish, honey, and olive oil. This is so that consumers can make informed choices regarding the origins of the products. For more information, a visit to the europa website, http://ec.europa.eu/food/index_en.htm, may prove useful.

21.10 Allergen management According to the US Department of Heath and Human Services (July 2004), food © 2008, Woodhead Publishing Limited

632

Chilled foods

allergy affects up to 8% of children under the age of 3 years, and 2% of adults. Authorities in other parts of the world have put these figures even higher. What is agreed is that the problem is a significant issue for a substantial number of people, making good food labeling essential for the safety of people affected. Allergic reactions can occur in sensitive individuals to a range of foodstuffs. These reactions vary in severity from mild discomfort to life-threatening anaphylactic shock. Nuts and peanuts are often quoted as having caused anaphylaxis even when present in extremely small amounts if sensitized individuals are exposed to them. Allergic responses are triggered by the body’s own immune system, the immune system being an important part of the bodily defense mechanism against infection. For reasons not fully understood, the immune system of an individual may produce an immunoglobulin IgE against a certain foodstuff; this attaches to the mast cells found in body tissues in common sites of allergic reactions, typically the gut, skin, and respiratory tract. The mast cells trigger a series of biochemical responses that include the release of histamine. Typical allergic responses are gastrointestinal upset of various degrees of severity, eczema and asthma; there are, however, many more. People who are affected in this way are termed hypersensitive. Codex general standard for the labeling of prepackaged foods – (Codex, 1991) highlights the following foods as being known to cause hypersensitivity. These foods or ingredients must always be declared:

• cereals containing gluten, e.g. wheat, rye, barley, oats, spelt or their hybridized • • • • • • •

strains and products of these crustacea and products of these eggs and egg products fish and fish products peanuts, soybeans and products of these milk and milk products (lactose included) tree nuts and nut products sulphite in concentrations of 10 mg/kg or more.

To allow consumers to have good information, these allergens must be effectively handled and managed in order that correct labeling can take place. This is a food safety issue to those sensitized. There are three basic levels of control required:

• Identification of all sources of the allergen, including those hidden in raw • •

materials. Full specifications are required, including details of any allergens handled on the supplier’s factory or plant. Separation and segregation of allergenic materials or products containing allergens. Definition of handling practices to ensure food safety including the transporting stage.

Manufacturing and storage premises can offer many opportunities for crosscontamination of ingredients if handling practices are poor. The design and construction of manufacturing and storage areas can aid the segregation of © 2008, Woodhead Publishing Limited

Management of product quality and safety

633

allergens and allergen-containing products. For example, a nut-free area where no cross-contamination is possible can be designed. Here, physical barriers and management systems must be in place to ensure nut-containing materials cannot enter the area and that product can be removed and transported without contacting the nut allergens. Whilst physical barriers and exclusion of potentially allergenic food stuffs are the most desirable way to ensure clear labeling, some manufacturers may have closed production lines which are positioned in a site that elsewhere handles the allergen in question. Here, careful segregation and failsafe management must be present to safeguard the systems that allow correct labeling. If a doubt exists as to the potential presence of an allergen, the foodstuff must be labeled appropriately in order that an affected individual can avoid it. On supermarket shelves it is common to see a label reading ‘This food may contain nuts’. Quantities of the allergen, which is a protein component of the food, needed to cause an allergic response are difficult to define. They depend on the individual, and is subject to work by immunologists and other members of the medical profession. However, certain allergens such as peanuts are known to be more likely to cause the most serious sorts of response. A key factor in the control of allergens is the training of the management and manufacturing teams, and also of store staff regarding the significance of allergens to the sensitized sector of the population. Most people will consider these foods to be good, wholesome and harmless, and may not understand why they are so important in a food safety context.

21.11 Business quality models and quality system techniques There is a confusing array of different approaches to business quality modeling. All start at strategic level and aim to produce excellence or near perfection through clear goal setting and measures of progress. In fact it is not surprising that businesses find it difficult to choose between approaches when selecting a route forward. Models to chose from are many, and all have their own specific characteristics. Tried and tested models including the European Business Excellence model, which the standard ISO9000:2000 is based around, TQM which originated in Japan, and approaches like the data driven Six Sigma all have their own specific strengths. The value of the model to any organization depends on the preferred style of execution, as each model emphasizes different approaches to the quality issue. To make any of these models work, however, commitment and resourcing at the highest level in the business is necessary. The ramifications run into all areas of the business. It should be emphasized that the current popular quality models are well defined in their own right, and it is recommended that full understanding of the principle behind the chosen model is achieved before any attempt at implementation is commenced, otherwise its practical implementation will be seriously impaired. Here we can only briefly introduce a couple of the models commonly encountered © 2008, Woodhead Publishing Limited

634

Chilled foods

in the fast moving consumer goods industry, which includes many food and drink manufacturers. A good source of further information is the Chartered Quality Institute (www.thecqi.org). TQM has been around in the western world since the 1980s when it arrived from Japan. It is based on continuous improvement and the model embodies five pillars of support to this goal. These are generally known as the five steps – sort, straighten, shine, standardize and sustain. Each of these steps is, in itself, an improvement tool. However, when all five are used together in fully implemented TQM, they can bring great benefits. To take an example from the sort step of TQM, this uses practical principles to achieve its goal. For instance, a chosen process would be literally sorted out, making sure only items immediately necessary for use were available. These items would be those that are in constant use in a process. Then, going forward, the sort step keeps to a minimum what is needed in the next stage of the operation. This brings about the well known just in time (JIT) principle. Many equally useful techniques are embodied in the other five steps of TQM. Overall, if TQM is well implemented, it brings about improved communication and staff commitment, and indeed these are foundations of the model. Most business quality models are built on simple principles but may be complex to implement, so when considering their worth it is valuable to review other companies who have embarked on the route. In the case of TQM, which has been around for long enough to be evaluated, a considerable number of implementing companies report serious improvements to their performance by its use. The Six Sigma model differs in its approach from TQM by using statistical principles at its core in a measurement-based strategy. The basic idea is to eliminate defects and strive for near perfection. Six Sigma relates to six standard deviation from the specification mean. The Six Sigma model use two main substrategies. One of these is specifically for new processes and the other is for current processes that need improvement. Project and process managers involved with Six Sigma require training to the Six Sigma green belt level, whilst those staff expected to lead, consult and facilitate within the Six Sigma project require training to black belt level. These levels are defined within the model and are understood between the different companies undertaking Six Sigma modeling. Taking a general look at the experiences of the food industry shows that successful quality modeling approaches have some common threads for example, a clear goal, good leadership and good motivation, excellent training and support to ensure that the end goals are set and met. Team working and use of knowledge from a variety of sources strengthens all quality approaches. The use of best practice groups allows benchmarking against other companies, whereas focus groups are examples of ways to stimulate ideas and encourage the raising of standards. These all fall into a pool of generally useful techniques. In whatever way motivation is produced and maintained, it is important to have widespread ownership and clarity of action at all levels in an organization to prevent a loss of direction when introducing a complex model. This requires a strategic level commitment to good clear communication and to the principle of staff involvement © 2008, Woodhead Publishing Limited

Management of product quality and safety

635

and empowerment. To achieve this the main board must be fully behind the introduction of the approach. Most quality models are results of oriented approaches and demand progress to be measurable. Performance measures are often linked to operational targets and goals. Trend analysis and statistical techniques may be employed, and the specific techniques vary, dependent on their purpose. However, as part of the good communication aspect of the models, it is necessary to ensure that information gathered is used effectively and is made available to those that require it. Once quality modeling is embarked upon nothing is surer than that cultural, behavioral and performance changes will follow.

21.12 Useful organizations and websites British Retail Consortium, 21 Dartmouth Street, London SW1H 9BP The Chartered Quality Institute, 12 Grosvenor Crescent, London SW1X 7EE europa website http://ec.europa.eu/food/index_en.htm Leatherhead Food Research Association, Randalls Road, Leatherhead, Surrey DNV www.dnv.co.uk Lloyds register www.Lrqa.co.uk SGS www.training.uk.sgs.com

21.13 Bibliography BRC,

Global standard – Food Issue 5, British Retail Consortium, 21 Dartmouth Street, London. SW1H 9BP. CAC/RCP 1-1969, Ref 3, 1997, Codex Alimentarius Commission, FOA/WHO, Rome. CFA (1997) Guidelines for good hygienic practice in the manufacture of chilled foods, ISBN. 1 901978 00 3. CODEX (1997a) Hazard Analysis and Critical Control (HACCP) System and Guidelines for its Application, CAC/RCP 1-1969, Rev.3, 1997 Codex Alimentarius Commission, FAO/ WHO, Rome. CODEX (1997b) Recommended international code of practice: General principles of food hygiene CODEX (1991) General standard for the labelling of prepackaged foods – Codex stan 1-1985 (rev 1-1991). COMMISSION REGULATION (EC) No 2073/2005 of the 15th November on microbiological criteria for foodstuffs. (EC) NO 852/2004 of the European Parliament and of the council of 29th April 2004 on hygiene of foodstuffs. EU GENERAL FOOD LAW REGULATION 178/2002. FOA/WHO Codex Alimentarius Commission. HACCP system and guidelines for its application. Food and drink manufacture – Good manufacturing practice: A guide to its responsible management. HACCP: A Practical Guide (1994) Campden technical manual 38, issue 2. HACCP: A Practical Guide, 3rd edition, Campden and Chorleywood Food Research Association, Chipping Campden, GL 55 6LD. HACCP principles. European Commission Health and Consumer Protection DirectorateGeneral, Brussels 16 November 2005. © 2008, Woodhead Publishing Limited

636

Chilled foods

HMSO (1990) Food Safety Act, ISBN 0-10-541690-8 HMSO London. ILSI EUROPE REPORT SERIES – validation and verification of HACCP

2001. ILSI Europe, Avenue E Mournier 83, Box 6, B-1200 Brussels Belgium ISBN1-57881-060-4. ISO 22000. 2005, Food safety management systems. REGULATION (EC) No 2073/2005 of 15 November 2005 microbiological criteria for foodstuffs. REGULATION (EC) 852/2004 hygiene of foodstuffs.

© 2008, Woodhead Publishing Limited

22 Legislation and criteria K. Goodburn, MBE, Chilled Food Association, UK

22.1 What are chilled foods? Chilled foods are prepared, generally multicomponent foods relying on chilled storage to achieve their shelf-lives. They are either ready to eat (RTE), ready to reheat (RTRH), ready to cook (RTC) or ready to wash (RTW). Chilled foods have been available in the UK since the 1960s and the market has developed particularly rapidly in the last 25 years. For example, the UK market for chilled ready meals has grown from an estimated £173 million in 1988 (Chilled Food Association (CFA) estimate) to £1968 million in 2007 (TNS). The total UK chilled prepared food market was an estimated £8230 million in 2007 (CFA/TNS: www.chilledfood.org) and continues to evolve, reflecting consumers’ changing needs and lifestyles. The range and nature of chilled foods available in each country, be it within the EU or globally, differs greatly. Table 22.1 shows the scope and simplified chronology of the development of chilled prepared foods in the UK.

22.2 Food laws and international trade There are few regulations that specifically apply to chilled foods. Regulations, as opposed to Directives, apply to all member states and are binding as soon as they are adopted and published in the Official Journal of the European Union. A Regulation is not, unlike a Directive, required to be transposed into law through national legislation therefore, not having the flexibility of a Directive. Regulations © 2008, Woodhead Publishing Limited

638

Chilled foods

Table 22.1

The chronology of chilled foods (UK)

1960s

1970s

Sliced meats Pies

Dressed salads Ready meals Dairy desserts Recipe dishes Quiches Flans Sandwiches Pizzas Ethnic snacks Pastas Soups

1980s

1990s

2000s

Non-dairy desserts Meal centres and Dips accompaniments Salad dressings Chilled speciality Sauces breads Stocks Sushi Prepared fruit Ready to cook Prepared vegetables meals Prepared leafy salads Luxury meal kits Sandwich fillings Stir fry kits

have become the favoured legislative tool in the EU, particularly on issues relating to food hygiene and safety and consumer protection. However, a key feature in the development of chilled foods, particularly in Europe, has been the industry’s approach to food safety and quality, which is hazard analysis and critical control point (HACCP) focussed and founded on self regulation. 22.2.1 CODEX – role and relevance In international food law, the two most important regulatory bodies are the World Trade Organisation (WTO) and the Codex Alimentarius Commission, often referred to as ‘CODEX’. CODEX standards and codes relate to internationally traded goods. CODEX was formed in 1962 to facilitate the development of trade in foodstuffs. It is run under the aegis of the United Nations (UN) Food and Agricultural Organisation (FAO) and World Health Organisation (WHO) and has developed standards, recommendations and guidelines on food safety assurance to facilitate fair trade. Its members are national governments and the EC is represented at its meetings on behalf of the EU as a whole. EU legislation relates to the WTO and is, where relevant, in line with CODEX requirements. In January 1995, when the WTO updated and replaced the General Agreement on Tariffs and Trade, the Agreement was supplemented by several more detailed agreements including the Agreement on Sanitary and Phytosanitary (SPS) Measures and the Agreement on Technical Barriers to Trade. CODEX standards are recognised as the basic standards upon which national food safety measures of SPS member countries should be based, and are therefore particularly relevant in the resolution of trade disputes brought to the WTO’s disputes panel. SPS members are required by the Agreement to accept the SPS measures of other members as being equivalent, even if these measures differ from their own, if the exporting member objectively demonstrates to the importing member that its measures achieve the importing member’s appropriate level of SPS protection. The General Principles set out in Recommended International Code of Practice – General Principles of Food Hygiene are the foundation for food hygiene © 2008, Woodhead Publishing Limited

Legislation and criteria

639

assurance approaches for primary producers, manufacturers, processors, foodservice operators, retailers and consumers. The document is designed to be used in conjunction with specific codes of hygienic practice, where appropriate, and CODEX guidelines on microbiological criteria. The General Principles document follows the food chain from primary production through to final consumption, highlighting the key hygiene controls at each stage. It recommends and sets out a HACCP-based approach wherever possible to enhance food safety as described in the Annex ‘Hazard Analysis and Critical Control Point (HACCP) System and Guidelines for its Application’. This is the authoritative statement of HACCP principles. Over recent years in particular, specific standards, codes and guidelines have been developed which directly relate to or impinge on the production of chilled foods. The Code of Hygienic Practice for Refrigerated Packaged Foods with Extended Shelf-Life (CAC/RCP 46, 1999: www.codexalimentarius.net/download/standards/347/CXP_046e.pdf) is applicable to low-acid (pH > 4.6), high water activity (aw > 0.92), heat treated, chilled foods with a shelf-life of more than 5 days. It sets out HACCP-based recommendations for the processing, packaging, storage and distribution of these foods, aiming to prevent growth of pathogenic microorganisms. The Code defines key terms used in chilled foods manufacture including ‘high risk area’. The Code excludes those food products for which there is already a specific Codex Alimentarius Code of Practice (see Codex website for examples, http://search.fao.org/opensearch). The Code is designed to be used in conjunction with the Codex General Principles of Food Hygiene (Alinorm 97/13A). The Code does not stipulate particular thermal treatments, shelf-lives or storage regimes, but states that: ‘It is the responsibility of the manufacturer to ensure that the product is safe throughout its shelf-life, taking into consideration the potential for temperature abuse. This may warrant the use of hurdles to microbial growth in addition to refrigeration. When using the hurdle concept (e.g. combinations of preservation factors) for product development, even where refrigeration is the sole hurdle, the effect of the hurdle(s) on product safety and shelf-life should be considered thoroughly. Predictive microbiological models may be used to estimate both the effectiveness of preservation conditions and the effects of modifying product composition and varying handling/storage conditions on safety. Unless scientific evidence previously exists, challenge studies should be conducted to confirm the effectiveness of the chosen hurdle(s) against the pathogen(s) of concern.’ Product shelf-life is stated to depend on a number of factors, such as:

• product formulation (e.g. decreased pH, decreased aw, other hurdles) • scheduled heat or other preservation treatments • cooling methods applied to product © 2008, Woodhead Publishing Limited

640

Chilled foods

• type of packaging (e.g. hermetically sealed or not, modified atmosphere packaging (MAP)

• storage temperature • other hurdles. CODEX Guidelines on the Application of General Principles of Food Hygiene to the Control of Listeria monocytogenes in Ready-to-Eat Foods provide advice to governments on a framework for the control of L. monocytogenes in RTE foods. The primary purpose of the guidelines is to minimise the likelihood of illness arising from this cause. At the time of writing, work is ongoing in a CODEX working group on the development of Microbiological Criteria for Ready to Eat Foods, which is intended to be used within the context of the L. monocytogenes control guidelines (CAC/GL 61, 2007) and is specifically linked to Section 5.2.3 (Microbiological criteria and other specifications) of that document. For foods sold through foodservice, the Code of Hygienic Practice for Precooked and Cooked Foods in Mass Catering (CAC/RCP 39-1993) applies to foods with shelf-lives of no more than 5 days.

22.2.2 Agreement on Transportation of Perishable Foods The United Nations Economic Commission for Europe (UNECE) Inland Transport Committee, established in 1947, is a mechanism for intergovernmental action to facilitate international transport while improving its safety and environmental performance. The main outputs of UNECE work are 56 international agreements and conventions providing the international legal and technical framework for the development of international road, rail, inland waterway and combined transport in the UNECE region. These international legal instruments, some of which are applied also by countries outside the UNECE region, address a range of transport issues under the responsibility of governments that impact on international transport. This includes uniform rules and regulations aimed at ensuring a high level of efficiency, safety and environmental protection in transport. These agreements and conventions are legally binding for the states who become contracting parties to them. The UNECE Agreement on the international carriage of perishable foodstuffs and on the special equipment to be used for such carriage (ATP) was first developed in 1971 by a UNECE Working Party on the Transport of Perishable Foodstuffs and came into force in 1976. ATP aims to improve the conditions of preservation of the quality of perishable (chilled, deep- and quick-frozen) foodstuffs during their carriage, particularly in international trade, and promote the expansion of such trade. It sets out:

• equipment performance requirements (refrigerated, insulated or thermal equipment)

• procedures for sampling and measuring temperature of chilled, frozen and quick-frozen perishable foodstuffs

• allowable tolerances in temperature measurement © 2008, Woodhead Publishing Limited

Legislation and criteria

641

Table 22.2 ATP chilled perishable food temperature conditions (2007) Chilled food type

Maximum temperature

Raw milk1 Red meat2 and large game (other than red offal) III. Meat products,3 pasteurised milk, fresh dairy products (yoghurt, kefir, cream and fresh cheese4), ready cooked foodstuffs (meat, fish, vegetables), ready to eat prepared raw vegetables and vegetable products5 and fish products3 not listed below IV. Game (other than large game), poultry2 and rabbits V. Red offal2 VI. Minced meat2

+ 6 °C + 7 °C

I. II.

VII. Untreated fish, molluscs and crustaceans6

Either at + 6 °C or at temperature indicated on the label and/or on the transport documents

+ 4 °C + 3 °C Either at +2 °C or at temperature indicated on the label and/or on the transport documents On melting ice or at temperature of melting ice

Notes 1 When milk is collected from the farm for immediate processing, the temperature may rise during carriage to + 10 °C. 2 Any preparations thereof. 3 Except for products fully treated by salting, smoking, drying or sterilisation. 4 ‘Fresh cheese’ means a non-ripened (non-matured) cheese which is ready for consumption shortly after manufacturing and which has a limited conservation period. 5 Raw vegetables which have been diced, sliced or otherwise size reduced, but excluding those which have been only washed, peeled or simply cut in half. 6 Except for live fish, live molluscs and live crustaceans.

• specific requirements for the selection of equipment and temperature conditions to be observed for the carriage of chilled foodstuffs, providing a common international approach (see Table 22.2).

22.2.3 European Union There is no chilled food-specific EU legislation, provisions being incorporated into general labelling and HACCP-based requirements for food business operators (FBOs). The General Food Law Regulation (178/2002/EC) is the overarching European legislation applicable to FBOs. A FBO is defined in that Regulation as the natural or legal persons responsible for ensuring that the requirements of food law are met within the food business under their control. The General Food Law stipulates common requirements for FBOs, including traceability of food and feed products, general responsibilities, withdrawal of unsafe food or feed from the market and notification to the competent authorities. © 2008, Woodhead Publishing Limited

642

Chilled foods

Another term commonly used but not defined in law is brand owner, being an FBO packaging a product under its own name or mark. EU hygiene legislation The 1993 Council Directive 93/43/EEC on the hygiene of foodstuffs (the Food Hygiene Directive, FHD) for the first time introduced concepts from HACCP and risk assessment into European horizontal (generally applicable) food law. It required food sector businesses to identify ‘any step in their activities which is critical to ensuring food safety and ensure that adequate safety procedures are identified, implemented, maintained and reviewed’ on the basis of five of the seven principles of HACCP. This was a step away from the previous approach of legislation prescribing detailed statutory requirements relating to specific practices, although certain core requirements (e.g. general requirements for food premises) were given in the ten appendices to the Directive. The FHD specifically referred in Article 5 to the production of voluntary guides to good hygiene practice for food businesses as a guide to compliance with Article 3 of the Directive, which related to food operations being carried out in a ‘hygienic way’ and to HACCP principles be used in doing so. Article 5 also referred directly to the CODEX recommended International Code of Practice, General Principles of Food Hygiene, stating that guides shall, where appropriate, have regard to it. However, there are relatively few such guides since industry generally wishes to retain ownership of its standards and the ability to update them without additional bureaucracy created by some member states’ governments’ requirements in relation to the mechanism of their development. For this reason, there are more industry guidelines/guidance documents than guides. The FHD opened the way for further requirements on microbiological and temperature control criteria for certain classes of foodstuffs. However the second of these has not yet been covered by European rules. In the White Paper on Food Safety of January 2000, the EC announced a series of initiatives to improve and complete the legislation on food and feed controls. The FHD and 17 ‘vertical’ Directives relating to specific protein product and raw material groups such as meat of various species, eggs, and fishery products, were consequently replaced in 2004 through a consolidation, harmonisation and intended simplification of all food hygiene legislation (the ‘Hygiene Package’). The Hygiene Package, which came into force in January 2006, comprises Regulations 852/2004, 853/2004 and 183/2005, the Regulation on microbiological criteria for foodstuffs (2073/2005) and the Regulation on official feed and food controls (882/2004). A key aspect of the legislation is that all food and feed operators, from farmers and processors to retailers and caterers, have primary responsibility for ensuring that food put on the EU market meets the required safety standards. The requirements also apply to imported products. FBOs have to apply compulsory self-checking programmes and follow the HACCP principles. Regulation 882/2004 on official controls aims to ensure the verification of compliance with feed and food law, animal health and animal welfare rules. The © 2008, Woodhead Publishing Limited

Legislation and criteria

643

key elements of the Regulation, which became applicable on 1 January 2006, are:

• to ensure that official controls on feed and food are carried out regularly, on a risk basis and with appropriate frequency;

• to establish a clear EU framework for a control system, systematically setting

• • •

• • • •

out the rules to be respected with the aim of greater harmonisation and the integration of controls across the entire food and feed chain under the ‘farm to fork’ principle; to allow a competent authority to delegate specific tasks to official control bodies; to provide regular training for competent authority staff; to establish appropriate control methods and techniques such as monitoring, surveillance, verification, audit, inspection, sampling and analysis (in compliance with relevant Community rules, or with internationally recognised rules or protocols; to improve efficiency of the Commission’s inspection services by way of a more transparent, strategic and integrated approach; to establish a community and national reference laboratories network; to require that the competent authorities prepare a single integrated multiannual national control plan to ensure the effective implementation of the Regulation; to organise and develop a Community food safety training strategy ‘Better Training for Safer Food’ to ensure a more harmonised approach.

The EC’s guidance on official controls (882/2004) and the microbiological criteria regulations were published in November 2006 (http://ec.europa.eu/food/food/ biosafety/salmonella/microbio_en.htm) and the EC has a legal obligation to formally review the operation of the hygiene legislation by 20 May 2009. Good practice guides at either EU or national level are allowed under 852/2004 to assist FBOs to implement self-enforcement programmes and assure food hygiene. A register of national guides is available at: http://ec.europa.eu/food/ food/biosafety/hygienelegislation/register_national_guides_en.pdf. However, this listing does not include a myriad of new and established guidance documents that industry has chosen not to officially submit to the authorities (see Section 22.5). EU Regulation 2073/2005 (as amended) on microbiological criteria for foodstuffs The purpose of this Regulation, which came into force in January 2006, is to protect public health and provide harmonised reference points for FBOs. It relates to the General Food Law (178/2002/EC) that obliges FBOs to withdraw or recall unsafe food from the market. It states that safety is ensured by a preventative approach such as employing good hygiene practices and use of HACCP principles. According to the European Commission’s strategy document which preceded the Regulation (European Commission, 2005): ‘The Regulations do not bring any new obligations or new administrative requirements for food businesses and do not © 2008, Woodhead Publishing Limited

644

Chilled foods

cause additional costs for food businesses’. The scope of the Regulation is food production and retail including catering, but it does not cover all foods. For example, agriculture is excluded apart from sprouted seeds and shellfish. The Regulation sets:

• Food Safety Criteria, which require action under 178/2002 (General Food Law) in terms of reporting exceedances to the authorities and carrying out recalls.

• Process Hygiene Criteria, which are quality-related, requiring no action under 178/2002, but corrective action needing to be carried out in terms of the process/ raw material chain to ensure that the opportunity for occurrence of future exceedances is minimised. The Regulation generally relates to finished manufactured foods and not to ingredients or raw materials used to manufacture that food. However, FBOs producing/supplying raw materials may be affected by the Regulation through the application of criteria and corrective actions required by their customers’ food safety management plans. With the exception of raw minced meat, meat preparations, certain meat products intended to be cooked and mechanically separated meat, the Regulation does not specify the frequency of sampling/testing, and it is for the FBO to decide the appropriate level of sampling/testing to help validate and verify their food safety management systems, e.g. HACCP. An integral part of HACCP is the appropriate use of microbiological criteria in validation and verification. The Microbiological Criteria for Foodstuffs Regulation, which came into force in January 2006, is linked to new hygiene legislation that came into force at the same time (852/2004, 853/2004) and General Food Law Regulation 178/2002, since Food Safety Criteria relate to the definition of ‘unsafe’. It aims to harmonise a large number of disparate criteria used in different member states, set a common approach to their interpretation, and require standard action if the criteria are breached. It is important to note that breaching the criteria does not contravene the law, but an FBO not taking the action specified in the Regulation does. For example, if a Process Hygiene Criterion is exceeded, the FBO should find and address the cause; and if a Food Safety Criterion is exceeded, reporting to the authorities and conditional recall are required. Enforcers have to satisfy themselves that the FBO’s HACCP system is working. FBOs are required to provide evidence that the necessary food safety management procedures are in place to ensure all relevant criteria are met. The Regulation sets Food Safety Criteria and Process Hygiene Criteria. Food Safety Criteria have been set for Escherichia coli, Enterobacter sakazakii, Listeria monocytogenes, Salmonella spp, Staphylococcus enterotoxin and histamine. Process Hygiene Criteria have been set for Salmonella, aerobic colony count, E. coli, Enterobacteriaceae, Staphylococcal enterotoxin and Bacillus cereus. The following are minimum legal requirements:

• There is an obligation within 178/2002/EC to withdraw or recall unsafe food from the market. © 2008, Woodhead Publishing Limited

Legislation and criteria

645

• If the Food Safety Criteria given in Annex I, Chapter 1 of the microbiological

• •

•

•

•

• • •

criteria for foodstuffs Regulation are not met, the product is deemed as ‘unsatisfactory’. Under 178/2002/EC there is an obligation on the brand owner to withdraw unsafe food from the market and to notify the competent authority. Further action may include product recall by the brand owner. This would be agreed on a case-by-case basis in consultation with the competent authority. FBOs must analyse trends in the test results. When a trend towards unsatisfactory results is found, or if the Process Hygiene Criteria given in Annex I, Chapter 2 of the Regulation are exceeded, then appropriate investigation and corrective action is required. Corrective action is to include the actions specified in the Annex I of the Regulation (see pages 17–52 of this guidance) together with other corrective actions defined in the FBO’s HACCP-based procedures and other actions necessary to protect the health of consumers. In addition, the FBO shall take measures to find the cause of the unsatisfactory results in order to prevent the recurrence of the unacceptable microbiological contamination. Those measures may include modifications to the HACCP-based procedures or other food hygiene control measures in place. Environmental monitoring can form part of the investigatory action undertaken to prevent a reoccurrence of an exceedance. No notification is required in the case of a Process Hygiene Criterion being exceeded. A food management preventative approach, such as employing good hygiene practices and a system based on HACCP principles must be in place. Food testing against the appropriate criteria should be undertaken, if appropriate, when verifying or validating HACCP. The Regulation stipulates a minimum frequency of sampling and testing for carcases, minced meat, meat preparations, meat products covered in Annex I of the Regulation that are intended to be eaten cooked, and mechanically separated meat. Where there are no frequencies defined by the Regulation (all except carcases, minced meat, meat preparations, meat products covered in Annex I of the Regulation that are intended to be eaten cooked and mechanically separated meat), sampling and testing frequencies should be determined by the FBO based on HACCP principles. Evidence to support the sampling/testing regime should be held on file and made available on request to the competent authority. Raw minced meat, raw meat preparations and raw poultry meat products intended to be cooked must be clearly labelled by the manufacturer informing the consumer of the need for thorough cooking prior to consumption. If official control testing (for example by a port health authority) is conducted on a product from outside the EU, the Food Safety Criteria will be used as a minimum requirement. Alternative methods to those prescribed in the Regulation may be used as long as those methods provide equivalent results validated against the reference method given in Annex I of the Regulation. Alternative methods must be:

© 2008, Woodhead Publishing Limited

646

Chilled foods a. validated against the reference method, and if a commercial kit, certified by a third party using an internationally accepted protocol, i.e. ISO 16140 or a similar protocol or b. validated by an internationally accepted protocol and authorised by the competent authority.

• Environmental monitoring is one important tool when investigating why hygiene criteria and/or food safety criteria are not met.

• Processing areas and equipment used in the manufacture of ready-to-eat foods must be monitored for L. monocytogenes.

• Processing areas and equipment used in the manufacture of dried infant formula and related products must be monitored for Enterobacteriaceae. Where not set in the Regulation, member states may keep national microbiological criteria providing they are scientifically justified and do not pose barriers to intraCommunity trade. Member states had to scientifically justify to the EC by 1 January 2006 any rules they wished to keep in place. National rules rejected by the EC cannot be used for imports, but can still be used for internal trade management. The Regulation on Microbiological Criteria for Foodstuffs is subject to future review in order to take account of developments in microbiology and food safety.

22.3 National approaches to legislation There is little chilled food-specific legislation applied on a national basis. In many cases emphasis is on temperature control (see Chapter 15), the hygiene measures required if a food is ready to eat, shelf-life marking (see Chapter 7), shelf-life limitation, shelf-life assessment (see Chapter 19) or on any particular microbiological hazard.

22.3.1 UK EU General Food Law, hygiene and labelling requirements apply in the UK, with emphasis being on industry self regulation (see Section 22.5), based on the CFA’s Best Practice Guidelines for the Production of Chilled Foods (CFA, 2006) and retailers’ own codes that are applied for the majority of production by volume and value.

22.3.2 France In France, legislation has permitted the use of heat treatments that deliver less than a 6-log non-proteolytic C. botulinum process, which is the basis of European industry recommendations (CFA, 2006 and ECFF, 2006). An Arrêté of 26 June 1974 and related memoranda of 1988 and 1992 apply to the production of chilled prepared meals and their allowed shelf-life. However, their scope now covers only © 2008, Woodhead Publishing Limited

Legislation and criteria

647

meals made from products of animal origin, other than fish and meat products (i.e. milk, eggs). The 1974 Arrêté limited shelf-life of all pre-cooked chilled products to 6 days. The ‘Note de Service du 31 Mai 1988’ (modified in 1992) established changes in manufacturing protocols to extend shelf-life up to 42 days, according to the type of products (sous vide or not), the pasteurisation value (Pv) and the attained core temperature. The 1988 legislation enabled ready meals reaching a centre temperature of 65 °C and a process equivalent of 70 °C for 100 min (comparable to 90 °C for 1 min, or 0.6 log inactivation of spores of non-proteolytic C. botulinum) to have a shelf-life of up to 21 days, although the precise shelf-life was to be determined by the manufacturer. Products reaching a centre temperature of 70 °C and receiving equivalent lethality of 1000 min (i.e. equivalent to 90 °C for 10 min) could have a shelf-life of 42 days. For ready-to-eat meals made from meat and fish products, the shelf-life is set by the manufacturer using a protocol developed by industry with AFNOR, the French standards organisation.

22.3.3 The Netherlands TNO (1994) issued a code developed with industry in relation to cook–chill products (excluding vacuum packed (VP) sliced meats) with a shelf-life of 11–42 days. The code requires a total 6D heat process of 90 °C/10 mins (equivalent) including products that are duo pasteurised (cooked, filled, sealed, cooked). Final product storage temperatures of 0–3 °C are referred to for post-production storage on-site and a maximum of 6 weeks post-production is allowed, including a maximum of 3 weeks to consumption in the commercial and domestic chill chain (0–5 °C) and 1 week maximum consumer storage (0–5 °C).

22.3.4 North America USA The primary legal food safety instrument in the USA is the Food Code. The Code aims to safeguard public health and ensure that food is unadulterated and honestly presented when offered to the consumer. It applies to food offered at retail and in foodservice. This Code is ‘offered for adoption by local, state, and federal governmental jurisdictions for administration by the various departments, agencies, bureaus, divisions, and other units within each jurisdiction that have been delegated compliance responsibilities for food service, retail food stores, or food vending operations’. Alternative approaches that offer an equivalent level of public health protection to ensure that food at retail and foodservice is safe are recognised. The 2005 Code states that a food establishment may package food using a cook– chill or sous vide process without obtaining a variance (from legislated prerequisites) if it implements a HACCP plan and the food is: © 2008, Woodhead Publishing Limited

648

Chilled foods

• prepared and consumed on the premises, or prepared and consumed off the • • • •

premises but within the same business entity with no distribution or sale of the bagged product to another business entity or the consumer cooked to heat all parts of the food to a specified temperature/time protected from contamination after cooking placed in a package or bag with an oxygen barrier before cooking, or placed in a package or bag immediately after cooking and before reaching a temperature below 57 °C cooled to 5 °C in the package or bag and then cooled to 1 °C or less within 48 hours of reaching 5 °C, and: • held at 1 °C and consumed or discarded within 30 days after the date of preparation, or • if removed from a storage unit that maintains food at 1 °C, held at 5 °C or less for no more than 72 hours before consumption.

The emphasis is therefore on low temperature storage and the control of Listeria monocytogenes. Heat processes stipulated by the FDA are significantly milder than those used by UK industry, where 6D processes are applied with respect to Listeria monocytogenes (for shelf-lives no more than 10 days) or non-proteolytic Clostridium botulinum (for shelf-lives more than 10 days). The 2005 Code requires, with exceptions, raw animal foods such as eggs, fish, meat, poultry, and foods containing these raw animal ingredients to be cooked throughout to a temperature and one of the time–temperature combinations for ratites and injected meats if they are comminuted (Table 22.3). A heat treatment of 74 °C or above for 15 seconds is required for the cooking of poultry, baluts, wild game animals, stuffed fish, stuffed meat, stuffed pasta, stuffed poultry, stuffed ratites, or stuffing containing fish, meat, poultry, or ratites. In the case of whole meat roasts including beef, corned beef, lamb, pork, and cured pork roasts such as ham, it is necessary to heat all parts of the food to one of the time–temperature combinations shown in Table 22.4. The 2005 Food Code requires cooked potentially hazardous food to be cooled:

• within 2 hours from 57 °C to 21 °C; and • within a total of 6 hours from 57 °C to 5 °C or less, or to 7 °C or less. Except for infant formula and some baby food, product dating is not required by Federal regulations. There is no uniform or universally accepted system used for food dating in the USA. Although dating of some foods is required by more than 20 states, there are areas of the country where much of the food supply has some type of open date and other areas where almost no food is dated (USDA, 2001). Based on the results of the FDA’s 2001 Listeria risk assessment and the recom– mendations from the 2004 Conference for Food Protection, the FDA re-evaluated date marking provisions and focused its recommendations for date marking in the 2005 Food Code on high-risk foods with respect to Listeria contamination. It is notable that the 2005 Food Code exempts ‘deli salads’ (e.g. ham, chicken, egg, © 2008, Woodhead Publishing Limited

Legislation and criteria

649

Table 22.3 Heat treatments included in the 2005 US Food Code for comminuted ratite and injected meats Minimum Temperature ( °C)

Time

63 66 68 70

3 minutes 1 minute 15 seconds <1 second (instantaneous)

Table 22.4 Heat treatments included in the 2005 Food Code for whole meat roasts Temperature (°C)

Time in minutes1

Temperature (°C)

Time in seconds1

112 89 56 36 28 18 12 8

63.9 65.0 66.1 67.2 68.3 69.4 70.0

134 85 54 34 22 14 02

54.4 55.0 56.1 57.2 57.8 58.9 60.0 61.1 1

Holding time may include post-oven heat rise To reach 70.0 °C

2

seafood, pasta, potato, and macaroni) prepared and packaged in a food processing plant since ‘Scientific data support the exemption of these products because deli salads prepared and packaged by a food processing plant contain sufficient acidity and preservatives to prevent the growth of Listeria monocytogenes.’ Whether this statement truly applies to all deli salads manufactured commercially in the USA is not known, however. Canada The Canadian Code of Recommended Manufacturing Practice for Pasteurized/ Modified Atmosphere Packed/Refrigerated Food (Agriculture Canada, 1990) emphasises low temperature storage of these foods, i.e.:

• care should be taken to prevent the product temperature from rising above 10 °C during cold filling

• product should be maintained at temperatures of less than 4 °C (or above 65 °C) • refrigerated products to be used as ingredients or prepared foods should be held at temperatures below 4 °C at all times

• the processed final product must be kept refrigerated (–1 to +4 °C) at all times • whenever chilled food is received with the product temperature of +7 °C or higher the manufacturer shall be notified immediately and ‘special handling instructions requested’ © 2008, Woodhead Publishing Limited

650

Chilled foods

• chilled food storage facilities/retail display cases must be capable of maintaining product temperature between –1 and +4 °C. Health Canada’s 1992 Guidelines for the Production, Distribution, Retailing and Use of Refrigerated Pre-packaged Foods with Extended Shelf life note that ‘under Division 27 of the Food and Drug Regulations, refrigeration is defined as ‘exposure to a temperature of 4 °C or less’’. However, provincial regulatory provisions for refrigeration range from no stipulated transportation temperature requirements (five provinces) to <5 °C (two provinces), and regarding storage from voluntary 4 °C (one province) to up to 5 °C. With respect to shelf-life, if this exceeds 10 days, ‘the processor should, on request, make available appropriate data to the agency bearing responsibility for food safety, showing that the food in question can safely be marketed for the intended shelf-life. Such information should include results of microbiological challenge tests involving food poisoning or appropriate non-pathogenic organisms placed in the food and incubated under conditions of temperature abuse. Shelf-life tests should also be performed to determine when spoilage occurs relative to the growth of potential pathogens.’ The document states that federal/government agencies’ microbiological surveillance programmes ‘should be instituted to determine the presence of food borne pathogens in those products for which the shelf-life has not been suitably validated by the processor. Priority for monitoring should be given to new extended shelf-life foods (e.g. processed by sous vide technology) rather than products such as cured meats which generally have a long established record of safety.’ The Food Institute of Canada, in its undated Canadian Code of Recommended Handling Practices for Chilled Food, reiterates –1 °C to +4 °C as the storage temperature for chilled foods but also states that:

• ‘Processed products intended for chilled distribution and sale should be de•

signed with additional hurdles to inhibit food spoilage and poisoning, e.g. pH less than 4.5, reduced water activity, vacuum or modified atmosphere packaging. Whenever chilled food is received with the products at a temperature of +7 °C or warmer, the warehouse/receiver shall immediately notify the manufacturer and request instruction for special handling. These instructions may consist of any available method for effectively lowering the temperature such as lowtemperature rooms with air circulation and proper use of dunnage or separators in stacking.’

In its 12/1/91 Decision on the use of ‘best before’ dates/durable life, the Canadian Food Inspection Agency provides the following information indicating no particular shelf-life limitation for MAP meat: ‘Question: When meat is pre-packed in individual portions at a manufacturing plant and shipped in retail stores in outer containers that have been flushed with a modified atmosphere (i.e. CO2) designed to extend shelflife, is the meat manufacturing plant required to label the individual portions of meat with a ‘best before’ date (durable life date)? © 2008, Woodhead Publishing Limited

Legislation and criteria

651

Answer: No, although Section B.01.007 of the Food and Drug Regulations requires date marking on food products that have a durable life of less than 90 days, administratively, these products are not required to be so labelled by the manufacturer. The durable life of these products is largely dependent upon when the outer shipping container is opened and the pre-packaged product is exposed to air. Consequently, a best before date established by the manufacturer could be potentially misleading to the consumer. These products are date labelled by the retailer.’

22.3.5 Australia/New Zealand The Australian Quarantine Inspection Service, AQIS (1992), set a 10-day limit for MAP/VP foods held at ≤ 5 °C (unless frozen), or >5 days if storage is at 3 °C. For storage at >5 °C to 10 °C, a 6D process for non-proteolytic C. botulinum is required. Extended shelf-life is possible by storage at 3 °C. If the product is at >10 °C, it should be discarded. The Guidance is no longer in print, however. AFIC/ASI/RWTAA (1999) reiterates legal temperature requirements (0–4 °C but never more than 5 °C), including MAP products. AIFST/AFIC/ACCC (2000) guidelines are applicable to a variety of chilled food production systems including cook–chill, sous vide and MAP. The emphasis is on low temperature (0–3 °C) storage and the selection of heat process ‘according to the recipe’. Temperatures of 70 °C/2 min or 90 °C/11 min equivalent are referred to. The guidance does not give specific shelf-life limitations but states ‘long shelf-life with correct practice can be achieved. It has been found that some products have a shelf-life of up to 45 days.’ The guidelines appear to be designed to apply to the catering sector rather than retail and are not believed to be complied with in the retail sector.

22.4 Microbiological criteria It is of key importance to be aware that the safety of food is neither guaranteed nor controlled by microbiological testing. The following factors all contribute to the variable microbiological population both within and between products within a batch of chilled foods:

• • • • •

They are not homogeneous; many are multicomponent. They may not be heat processed. Those that are heat processed could have significant post process handling. They may be ready-to-eat, ready-to-reheat or ready-to-cook. They are generally of short shelf-life.

Conclusions are therefore difficult to draw from isolated results and the overlying trend must be considered. Due to the short life of chilled products, testing results will often not be © 2008, Woodhead Publishing Limited

652

Chilled foods

available within the shelf-life. Positive release testing cannot therefore be carried out for short life chilled products. It is important to note that positive release is neither the intent nor a requirement of the EU Microbiological Criteria for Food Regulation (2073/2005, as amended). In addition, it is not commercially viable to test sufficient products for positive release in order to have statistically reliable results. The CODEX ‘Revised Principles for the Establishment and Application of Microbiological Criteria for Foods’ states: ‘Microbiological criteria should be based on scientific analysis and advice, and where sufficient data are available, on a risk analysis appropriate to the foodstuff and its use.’ These criteria may be relevant to the examination of foods, including raw materials and ingredients of unknown or uncertain origin, or when no other means of verifying the efficacy of HACCP-based systems and good hygienic practices are available. Microbiological criteria may also be used to determine that processes are consistent with the general principles of food hygiene. Microbiological criteria are not normally suitable for monitoring critical limits as defined in the HACCP system. Microbiological testing must be used to validate and monitor processes, verify CCPs identified through HACCP, and provide for due diligence. Indeed, the criteria are not obligatory where a food business is confident of its HACCP systems. Microbiological testing of end products should not be relied upon for anything other than due diligence purposes. Microbiological criteria are essentially of three types, defined by CODEX in 1981 and revised in 1993 to cover the working definitions below: (i)

Standard – This is a microbiological criterion contained in a law or regulation where compliance is mandatory. As well as being an offence, products not complying with the Standards are rejected as unfit for intended use. (ii) Guideline – This is a criterion applied at any stage of food processing which indicates the microbiological condition of the sample. Significant deviations from the norm may indicate the need for attention before control is lost. Investigative action is required to identify and rectify the cause. The UK Public Health Laboratory Service (now the Health Protection Agency) in 2000 published Guidelines on ready-to-eat foods to aid food examiners and enforcement officers; these are under review at the time of writing. (iii) Specification – This is a criterion applied to a purchase agreement and may include pathogens, toxins, spoilage or indicator organisms. Non-conforming products require investigation to determine the cause. Microbiological criteria are intended to give some degree of assurance that food is safe and of suitable quality, and that it will remain so to the end of its shelf-life provided it is handled appropriately. The EC Regulation on Microbiological Criteria for Foodstuffs (see subsection in Section 22.2.3) requires FBOs to use the criteria given in the Regulation when carrying out validation and verification checks as part of food safety management systems based on HACCP principles. It is vital to recognise that the safety of food is neither guaranteed nor controlled by microbiological testing. Microbiological testing can be used to validate and © 2008, Woodhead Publishing Limited

Legislation and criteria

653

monitor processes and verify CCPs identified through HACCP. Microbiological testing of final product alone cannot be relied upon to demonstrate product safety and may be insufficient to demonstrate due diligence.

22.5 Self-regulation The limitations of end-product testing mean that chilled food manufacturers must focus on HACCP implementation to assure food quality and safety. Raw materials, for example, are bought to agreed specifications through approved/audited suppliers. Suppliers are required to provide appropriate documentation with analyses carried out by competent laboratories and full traceability. Given the high risk nature of much chilled food in that it may be cooked and ready to eat, particular attention is required to be paid to hygiene in its production. Industry guidance provides detail omitted from HACCP-based legislation and is used as the basis of self-regulation, particularly in the EU. Two key industry guidance documents focusing on chilled food production are the CFA’s Best Practice Guidelines for the Production of Chilled Foods, published in January 2006 to assist the chilled industry to build on HACCP from new product development concepts to final production. The Guidelines set out prerequisites including the key principles of factory and process layout and design, such as area segregation, and identify how to implement HACCP in a chilled food manufacturing business. The core elements of these Guidelines have been taken up in the second of these documents, the European Chilled Food Federa-tion’s 2006 Recommendations for the Production of Pre-packaged Chilled Food. The principle of audited self-regulation was established in the late 1980s in the chilled food manufacturing sector in the UK and has become widely taken up internationally – for example, by the British Retail Consortium and Global Food Safety Initiative. Following a dramatic increase in third party auditing in the UK in the mid 1990s, standards were established for the performance and administration systems applicable to auditors, accreditation and certification bodies (e.g. ISO/IEC 17011, ISO/IEC Guide 65:1966 EN45011). The potential for application of third party auditing as part of formal food law enforcement remains, but if it is to be taken up, robust mechanisms must be in place to assure common high standards of audit.

22.6 References and bibliography AFIC/ASI/RWTAA (1999), The Australian Cold Chain Food Safety Programs 1999. Australian

Food and Grocery Council, Australian Supermarket Institute, Refrigerated Warehouse and Transport Association of Australia. www.afgc.org.au (accessed 26/3/06). AFNOR (2003) XP V 01-003 Lignes directrices pour l’élaboration d’un protocole de validation de la durée de vie microbioloque, AFNOR, France. AGRICULTURE CANADA (1990) Canadian Code of Recommended Manufacturing Practices for Pasteurized/Modified Atmosphere Packed/Refrigerated Food. March 1990. Agrifood Safety Division, Agriculture Canada. © 2008, Woodhead Publishing Limited

654

Chilled foods

AIFST/ACCC (2000) Guidelines for Chilled Food Production Systems including Food Safety

Programs. Australian Institute of Food Science and Technology/Australian Cook Chill Council Inc. www.aifst.asn.au/templates/aifst.aspx?edit=false&pageID=395 (accessed 7/1/06). AQIS (1992), Code of practice for heat-treated refrigerated foods packaged for extended shelf life. Australian Quarantine Inspection Service. CFA (2006), Best Practice Guidelines for the Production of Chilled Food, The Stationery Office, Norwich, UK, ISBN-13 978-1-901798-11-1, http://www.tsoshop.co.uk/book store.asp?DI=561965. CFA (2006), Guidance on the practical implementation of the EC Regulation on Microbiological Criteria for Foodstuffs (edition 1.2), ISBN-13 978-1-901798-13-5, www.chilled food.org . CFA (2006), Microbiological Testing and Interpretation Guidance, 2nd edition, ISBN-13 978-1-901798-14-2, www.chilledfood.org. CODEX (1995), Guidelines on the Application of the Principles of Risk Assessment and Risk Management to Food Hygiene Including Strategies for their Application. CX/FH 95/8. CODEX (1997), General Principles of Food Hygiene, CODEX, Alinorm 97/13A, http:// www.codexalimentarius.net/download/standards/23/cxp_001e.pdf (accessed 8/2/08). CODEX (1997), Principles for the establishment and application of microbiological criteria for food, CAC/GL 21-1997, http://www.codexalimentarius.net/download/standards/394/ CXG_021e.pdf (accessed 8/2/08). CODEX (1999), Recommended International Code of Practice – General Principles of Food Hygiene (CAC/RCP 1-1969, Rev. 3 Amd, http://www.codexalimentarius.net/download/ standards/23/cxp_001e.pdf (accessed 8/2/08). CODEX (1999), Code of Hygienic Practice for Refrigerated Packaged Foods with Extended Shelf-Life, CAC/RCP 46, www.codexalimentarius.net/download/standards/347/ CXP_046e.pdf (accessed 8/2/08). CODEX (2007), Guidelines on the Application of General Principles of Food Hygiene to the Control of Listeria monocytogenes in Ready-to-Eat Foods, CAC/GL 61-2007, http:// www.codexalimentarius.net/download/standards/10740/cxg_061e.pdf (accessed 8/2/08). ECFF (2006), Recommendations for the Production of Pre-packaged Chilled Food, www.ecff.net. EUROPEAN COMMISSION (2002), Regulation (EC) No 178/2002 of the European Parliament and of the Council laying down the general principles and requirements of food law, establishing the European Food Safety Authority and laying down procedures in matters of food safety, http://eur-lex.europa.eu/pri/en/oj/dat/2002/l_031/l_03120020201en0001 0024.pdf (accessed 8/2/08). EUROPEAN COMMISSION (2004), Regulation (EC) No 852/2004 of the European Parliament and of the Council of 29 April on the hygiene of foodstuffs, http://eur-lex.europa.eu/ LexUriServ/LexUriServ.do?uri=OJ:L:2004:139:0001:0054:EN:PDF (accessed 8/2/08). EUROPEAN COMMISSION (2004), Regulation (EC) No 853/2004 of the European Parliament and of the Council of 29 April laying down specific hygiene rules for food of animal origin, http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2004:139:0055: 0205:EN:PDF (accessed 8/2/08). EUROPEAN COMMISSION (2004), Regulation 854/2004 laying down specific rules for the organisation of official controls on products of animal origin intended for human consumption, European Commission, http://eur-lex.europa.eu/LexUriServ/LexUriServ. do?uri=OJ:L:2004:139:0206:0320:EN:PDF (accessed 8/2/08). EUROPEAN COMMISSION (2004), Regulation (EC) No 882/2004 of the European Parliament and of the Council of 29 April 2004 on official controls performed to ensure the verification of compliance with feed and food law, animal health and animal welfare, http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2004:165:0001: 0141:EN:PDF (accessed 8/2/08). EUROPEAN COMMISSION (2004), Guidance on the implementation of Articles 11, 12, 16, 17, © 2008, Woodhead Publishing Limited

Legislation and criteria

655

18, 19 and 20 of Regulation (EC) No 178/2002 on General Food Law, http://ec.europa.eu/ food/food/foodlaw/guidance/guidance_rev_7_en.pdf (accessed 8/2/08). EUROPEAN COMMISSION (2005), Discussion paper on strategy for setting microbiological criteria for foodstuffs in Community legislation, European Commission, SANCO/ 1252/ 2001 Rev.11, http://ec.europa.eu/food/food/biosafety/salmonella/discussion_paper_en.pdf (accessed 8/2/08). EUROPEAN COMMISSION (2005), Guidance document on Implementation of certain provisions of Regulation (EC) No 852/2004 on the hygiene of Foodstuffs, http://ec.europa.eu/ food/food/biosafety/hygienelegislation/guidance_doc_852-2004_en.pdf (last accessed 8/2/08). EUROPEAN COMMISSION (2005), Regulation (EC) No 2073/2005 of 15 November 2005 on microbiological criteria for foodstuffs, http://eur-lex.europa.eu/LexUriServ/ LexUriServ.do?uri=CELEX:32005R2073:EN:NOT (accessed 8/2/08). EUROPEAN COMMISSION (2006), Guidance document on certain key questions related to import requirements and the new rules on food hygiene and on official food controls, http:/ /ec.europa.eu/food/international/trade/interpretation_imports.pdf (accessed 8/2/08). EUROPEAN COMMISSION (2007), Regulation (EC) No 1441/2007 of 5 December 2007 amending Regulation (EC) No 2073/2005 on microbiological criteria for foodstuffs, http:/ /eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2007:322:0012:0029:EN:PDF (accessed 8/2/08). FOOD STANDARDS AGENCY (2006), Guidance for food business operators on microbiological criteria for foodstuffs, Food Standards Agency, http://www.food.gov.uk/multimedia/ pdfs/ecregguidmicrobiolcriteria.pdf (accessed 8/2/08). PHLS (2000), Guidelines for the microbiological quality of some ready-to-eat foods sampled at the point of sale, Commun. Dis. Public Health, 2000; 3: 163-7. http://www.hpa.org.uk/ cdph/issues/CDPHvol3/No3/guides_micro.pdf (accessed 8/2/08). UNECE (2007), ATP Handbook, http://www.unece.org/trans/main/wp11/wp11fdoc/ATP2007e.pdf (accessed 8/2/08). USDA (2001) Food Product Dating. http://www.fsis.usda.gov/Fact_Sheets/Food_Product_ Dating/index.asp (accessed 8/2/08). US PUBLIC HEALTH SERVICE AND FOOD AND DRUGS ADMINISTRATION (2005), Food Code, http://www.cfsan.fda.gov/~dms/fc05-toc.html (accessed 8/2/08).

© 2008, Woodhead Publishing Limited