Odors In the Food Industry (Integrating Safety and Environmental Knowledge Into Food Studies towards European Sustainable Development)

Odors in the Food Industry ISEKI-FOOD SERIES Series Editor: Kristberg Kristbergsson, University of Iceland Reykjavík, ...

Author: Xavier Nicolay

26 downloads 1181 Views 3MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Odors in the Food Industry

ISEKI-FOOD SERIES Series Editor: Kristberg Kristbergsson, University of Iceland Reykjavík, Iceland

Volume 1

FOOD SAFETY: A Practical and Case Study Approach Edited by Anna McElhatton and Richard Marshall

Volume 2

ODORS IN THE FOOD INDUSTRY

Edited by Xavier Nicolay Volume 3

UTILIZATION OF BY-PRODUCTS AND TREATMENT OF WASTE IN THE FOOD INDUSTRY

Edited by Vasso Oreopoulou and Winfried Russ Volume 4

PREDICTIVE MODELING AND RISK ASSESSMENT

Edited by Rui Costa and Kristberg Kristbergsson Volume 5

EXPERIMENTS IN UNIT OPERATIONS AND PROCESSING OF FOODS

Edited by Maria Margarida Cortez Vieira and Peter Ho Volume 6

CASE STUDIES IN FOOD SAFETY AND ENVIRONMENTAL HEALTH

Edited by Maria Margarida Cortez Vieira and Peter Ho

Odors in the Food Industry Edited by

Xavier Nicolay Institut Meurice Brussels, Belgium

Xavier Nicolay Ingénieur Service de Génie Chimique & Biochimique Instiut Meurice – HELdB 1, avenue Emile Gryzon, bât. 2 1070 Bruxelles Belgium [email protected]

Series Editor Kristberg Kristbergsson Professor of Food Science Dept. Food Science and Human Nutrition University of Iceland Hjardarhaga 2-6 107, Reykjavik Iceland

Library of Congress Control Number: 2006926455 ISBN-10: 0-387-33510-2 ISBN-13: 978-0387-33510-0 e-ISBN-10: 0-387-34124-2 e-ISBN-13: 978-0387-34124-8 ©2006 Springer Science+Business Media, LLC All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. 9 8 7 6 5 4 3 2 1 springer.com

SERIES ACKNOWLEDGMENTS ISEKI Food is a thematic network on food studies, funded by the European Union as project N° 55792-CP-3-00-1-FR-ERASMUS-ETN. It is a part of the EU program in the field of higher education called ERASMUS, which is the higher education action of SOCRATES II program of the EU.

v

SERIES PREFACE The single most important task of food scientists and the food industry as a whole is to ensure the safety of foods supplied to consumers. Recent trends in global food production, distribution, and preparation call for increased emphasis on hygienic practices at all levels and for increased research in food safety in order to ensure a safer global food supply. The ISEKI-Food Series is a collection of books where various aspects of food safety and environmental issues are introduced and reviewed by scientists specializing in the field. In all of the books a special emphasis was placed on including case studies applicable to each specific topic. The books are intended for graduate students and senior level undergraduate students as well as professionals and researchers interested in food safety and environmental issues applicable to food safety. The idea and planning of the books originates from two working groups in the European thematic network “ISEKI-Food,” an acronym for “Integrating Safety and Environmental Knowledge In Food Studies.” Participants in the ISEKI-Food network come from 29 countries in Europe and most of the institutes and universities involved with food science education at the university level are represented. Some international companies and nonteaching institutions have also participated in the program. The ISEKI-Food network is coordinated by Professor Cristina Silva at The Catholic University of Portugal, College of Biotechnology (Escola) in Porto. The program has a website at: http://www. esb.ucp.pt/iseki/. The main objectives of ISEKI-Food have been to improve the harmonization of studies in food science and engineering in Europe and to develop and adapt food science curricula, emphasizing the inclusion of safety and environmental topics. The ISEKI-Food network started on October 1st in 2002, and recently has been approved for funding by the European Union for renewal as ISEKI-Food 2 for another three years. ISEKI has its roots in an EU-funded network formed in 1998 called Food Net where the emphasis was on casting a light on the different food science programs available at the various universities and technical institutions throughout Europe. The work of the ISEKI-Food network was organized into five different working groups with specific task, all aiming to fulfill the main objectives of the network. The first four volumes in the ISEKI-Food book series come from WG2 coordinated by Gerhard Schleining at Boku University in Austria and the undersigned. The main task of the WG2 was to develop and collect materials and methods for teaching safety and environmental topics in the food science and engineering curricula. The first volume is devoted to food safety in general with a practical and case study approach. The book is composed of 14 chapters that were organized into three sections on preservation and protection, benefits and risk of microorganisms, and process safety. All these issues have received high public interest in recent years and will continue to be in the focus of consumers and regulatory personnel for years to come. The second volume in the series is devoted to the control of air vii

viii

Series Preface

pollution and treatment of odors in the food industry. The book is divided into eight chapters devoted to defining the problem, recent advances in analysis, and methods for prevention and treatment of odors. The topic should be of special interest to industry personnel and researchers due to recent and upcoming regulations by the EU on air pollution from food processes. Other countries will likely follow suit with stricter regulations on the level of odors permitted to enter the environment from food-processing operations. The third volume in the series is devoted to utilization and treatment of waste in the food industry. Emphasis is placed on sustainability of food sources and how waste can be turned into by-products rather than pollution or landfills. The book is composed of 15 chapters starting off with an introduction of problems related to the treatment of waste and an introduction to the ISO 14001 standard used for improving and maintaining environmental management systems. The book then continues to describe the treatment and utilization of both liquid and solid waste with case studies from many different food processes. The last book from WG2 is on predictive modeling and risk assessment in food products and processes. Mathematical modeling of heat and mass transfer as well as reaction kinetics is introduced. This is followed by a discussion of the stoichiometry of migration in food packaging, as well as the fate of antibiotics and environmental pollutants in the food chain using mathematical modeling and case study samples for clarification. Volumes five and six come from work in WG5 coordinated by Margarida Vieira at the University of Algarve in Portugal and Roland Verhé at Gent University in Belgium. The main objective of the group was to collect and develop materials for teaching food safety-related topics at the laboratory and pilot plant level using practical experimentation. Volume five is a practical guide to experiments in unit operations and processing of foods. It is composed of 20 concise chapters each describing different food-processing experiments outlining theory, equipment, procedures, applicable calculations, and questions for the students or trainees followed by references. The book is intended to be a practical guide for the teaching of food-processing and engineering principles. The final volume in the ISEKI-Food book series is a collection of case studies in food safety and environmental health. It is intended to be a reference for introducing case studies into traditional lecture-based safety courses as well as being a basis for problem-based learning. The book consists of 13 chapters containing case studies that may be used, individually or in a series, to discuss a range of food safety issues. For convenience the book was divided into three main sections on microbial food safety, chemical residues and contaminants, and a final section on risk assessment and food legislation. The ISEKI-Food book series draws on expertise from close to a hundred universities and research institutions all over Europe. It is the hope of the authors, editors, coordinators, and participants in the ISEKI network that the books will be useful to students and colleagues to further their understanding of food safety and environmental issues.

March 2006

Kristberg Kristbergsson

PREFACE Olfactory nuisances, to the same extent as noise pollution, appear to be a modern social problem that is drawing increasingly more attention and is closely associated with urbanization, industrialization, and overall population density. This issue represents a major concern for decisionmakers in industry as well as for regional or national authorities. Air pollution and odor problems involve the responsibility of industrial plant managers not only in terms of brand image, but also in terms of possible economic consequences due to the implementation of treatment processes, or in the absence of adapted precautions, legal actions from neighboring industries or persons. Authorities also are concerned by this public health problem as they are in charge of implementing regulations and planning the national and regional development which will directly affect the socio economic situation of the region or country. This book was edited within the framework of the ISEKI Thematic Network and was originally intended for students and teachers in food engineering. In this context, its primary aim is to raise future industrial decisionmakers’ awareness in odors and air pollution problems. However, this book also offers to representatives of industry the appropriate tools to better apprehend and manage odors problems, should they ever be confronted with this type of situation whether as a victim or as a liable actor. In particular, this book deals with the diverse questions raised by odors in the food industry and the closely related volatile organic compounds: ranging from perception of the issue to implementation of regulations, from prevention of the problems to their possible treatment, through specific case studies and analysis methods illustrating the different measurement technologies. This exhaustive approach was made possible thanks to the extraordinary extent of the ISEKI Thematic Network and its privileged relationships among multidisciplinary experts from universities in 29 countries across Europe. At this stage I would like to thank all contributors for their efficient and enthusiastic collaboration.

January 2006

Xavier Nicolay

ix

CONTENTS CONTRIBUTORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii 1.

ODOR PROBLEMS IN THE FOOD INDUSTRY . . . . . . . . . . . . . . . . . . . Elefteria Psillakis and Vassilis Gekas

1

2.

ODOR MEASUREMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Elefteria Psillakis

3.

PRECONCENTRATION PRIOR TO GAS CHROMATOGRAPHY . . . . 41 Elefteria Psillakis

4.

THE APPLICATION OF INTELLIGENT SENSOR ARRAY FOR AIR POLLUTION CONTROL IN THE FOOD INDUSTRY . . . . . . . . . . . . . . . 47 Saverio Mannino, Simona Benedetti, Susanna Buratti, and Maria Stella Cosio

5.

ELECTRONIC-NOSE TECHNOLOGY: APPLICATION FOR QUALITY EVALUATION IN THE FISH INDUSTRY . . . . . . . . . . . . . . . . 57 Guðrún Ólafsdóttir and Kristberg Kristbergsson

6.

ODORS PREVENTION IN THE FOOD INDUSTRY . . . . . . . . . . . . . . . . 75 Regina Nabais

7.

ODORS TREATMENT: PHYSICOCHEMICAL TECHNOLOGIES. . . . 105 Regina Nabais

8.

ODORS TREATMENT: BIOLOGICAL TECHNOLOGIES . . . . . . . . . . . 125 Bram Sercu, João Peixoto, Kristof Demeestere, Toon van Elst, and Herman Van Langenhove INDEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

xi

CONTRIBUTORS Simona Benedetti Department of Food Science, Technology and Microbiology, University of Milan, Via Celoria 2, 20133 Milan, Italy Susanna Buratti Department of Food Science, Technology and Microbiology, University of Milan, Via Celoria 2, 20133 Milan, Italy Maria Stella Cosio Department of Food Science, Technology and Microbiology, University of Milan, Via Celoria 2, 20133 Milan, Italy Kristof Demeestere EnVOC Research Group, Department of Organic Chemistry, Faculty of Bioscience Engineering, Ghent University, 9000 Gent, Belgium Toon van Elst Project Research Gent nv, 9030 Gent, Belgium Vassilis Gekas Department of Environmental Engineering, Technical University of Crete, Polytechnioupolis, GR-73100 Chania-Crete, Greece

Kristberg Kristbergsson Department of Food Science and Human Nutrition, University of Iceland, Hjardarhaga 2-6, Reykjavík, Iceland Herman Van Langenhove

EnVOC Research Group, Department of Organic Chemistry, Faculty of Bioscience Engineering, Ghent University, 9000 Gent, Belgium Saverio Mannino Department of Food Science Technology and Microbiology, University of Milan, Via Celoria 2, 20133 Milan, Italy Regina Nabais C.E.R.N.A.S. -Centro de Recursos Naturais, Ambiente e Sociedade, Escola Superior Agrária de Coimbra, Instituto Politécnico de Coimbra. Bencanta, 3040-316 Coimbra, Portugal xiii

xiv

Contributors

Gudrun Olafsdottir Icelandic Fisheries Laboratories, Skulagata 4, 101 Reykjavík, Iceland João Peixoto Department of Biological Engineering, CEB, University of Minho, 4710-057 Braga, Portugal Elefteria Psillakis Department of Environmental Engineering, Technical University of Crete, Polytechnioupolis, GR-73100 Chania-Crete, Greece Bram Sercu EnVOC Research Group, Department of Organic Chemistry, Faculty of Bioscience Engineering, Ghent University, 9000 Gent, Belgium

1 Odor Problems in the Food Industry Elefteria Psillakis and Vassilis Gekas

1. INTRODUCTION With growing population, industrialization, and urbanization, the odor problem has been assuming objectionable proportion. Rapidly growing industrialization has aggravated the problem through odorous industrial operations. Undesirable odors contribute to air quality concerns and affect human lifestyles. On the economic front, loss of property value near odor-causing operations/industries and odorous environment is partly a consequence of offensive odor. Odor is undoubtedly the most complex of all the air pollution problems. Malodors generated by the food industry vary enormously since they can be generated during the production, processing, or even in the waste water treatment areas of the plant. In general, odor can be defined as the “perception of smell” or in scientific terms as the “organoleptic attribute perceptible by the olfactory organ on sniffing certain volatile substances” (ISO 5492:1992). Whether pleasant or unpleasant, odors are induced by inhaling airborne volatile organics or inorganics. Unlike conventional air pollutants, odor has distinctly different characteristics, which, to an extent, can be comparable with noise pollution given that similar to noise, nuisance is the primary effect of odor on people.

2. HUMAN RESPONSE TO ODORS Odor is complex both because of the large number of compounds that contribute to it and because it involves a subjective human response. In general, different people find different odors offensive at different concentrations and this is frequently related to the way different people perceive odors. According to a simple ELEFTERIA PSILLAKIS AND VASSILIS GEKAS ● Department of Environmental Engineering, Technical University of Crete, Polytechnioupolis, GR-73100 Chania-Crete, Greece. e-mail: [email protected], [email protected] 1

2

E. Psillakis and V. Gekas

model (Frechen, 1994) describing human odor perception it is the physiological reception and the psychological interpretation that results in a mental impression of a specific odor. Consequently, although the human olfactory organ is quite sensitive, the response to odor is more related to past memories or cultural experiences. It appears that odors have the ability to evoke memories that have been suppressed for many years and some of them may be associated with sadness or unpleasant situations. Our response to an odor may not be caused by high concentration or intensity but rather to a strong memory or an impression instilled many years ago. Discrimination between “good” and “bad” smells is important since pleasant and unpleasant smells require different behavioral responses (Jacob et al., 2003). Bad smells warn us of danger, poor air quality, “off ” food, poisons, even illness—each of which requires some immediate decision to be made and action to be taken, for example, avoidance or withdrawal. Pleasant smells, on the other hand, do not necessitate immediate actions or decisions. In fact, the biological significance of pleasant smells is not immediately obvious. Recent evidence suggests that malodors activate different areas of the brain from “pleasant” odors (Zald and Pardo, 2000). The human olfactory system consists of the olfactory epithelium, the olfactory bulb, and the olfactory cortex. The olfactory epithelium is an area of approximately 5 cm2 located in the upper nasal cavity, containing between 107 and 108 receptor cells (Lancet, 1991). The receptor cells connect via olfactory neurons to the olfactory bulb, where preprocessing of the electrical outputs from the receptor cells takes place before passing to the olfactory cortex where further processing takes place in the higher-order olfactory structures of the central nervous system (Pearce, 1997). It is said that humans can differentiate about 10000 odors with differing qualities. To date it is not possible to predict an odor sensation due to the chemical structure of an odorant with a view to establishing an odorant classification system (Gostelow et al., 2001). This is due to the fact that substances of similar or dissimilar chemical constitution may have similar odors. Furthermore, the nature and strength of odor may change on dilution and weak odors are not perceived in the presence of strong odors. When odorous compounds with the same strength blend to produce a combination, it is possible that one of them may be unrecognizable. As mentioned earlier, memories are strongly correlated with odors. However, there are several other parameters affecting human perception of odors and these include offensiveness, duration of exposure to odor, frequency of odor occurrence, and tolerance/expectation of the receptor. For example, fatigue from continued exposure to an odor may affect the human sense of smell. This phenomenon is called adaptation. Adaptation may reduce both perceived odor intensity and perceived odor quality (Stuetz et al., 2001). The degree of adaptation resulting from exposure to an odorous air will depend on the odor concentration experienced. The weaker the odor concentration of an air sample, the more adaptation affects perceived strength. The phenomenon of adaptation frequently reveals itself in industrial situations, with workers reporting that an initially repulsive odor eventually seemed less repulsive.

Odor Problems in the Food Industry

3

Other parameters such as age and gender also may contribute to the ability to perceive an odor (Bliss et al., 1996) and to a lesser extent health (e.g., cold, nasal allergy), personality, educational background, training, or hereditary deficiencies in odor sensitivity.

3. ODORS AND HUMAN HEALTH An important area of research is whether odors are simply a nuisance or a legitimate health threat. In fact, very little information is available on the impact of odor on human health. Most studies address the impact of individual chemicals on human health (e.g., the smell of hydrogen sulfide and its lethality to humans). There are several problems to that approach: the toxicity of most chemicals, even those with high production volumes, is not known, and frequently the chemical composition of materials resulting from complex industrial processes is unidentified (Rosenkranz and Cunningham, 2003). It should be mentioned, however, that much of the research to date on the relationships between individual chemicals and human health has been conducted indoors, where gas concentrations are higher than in open conditions. Although there is limited evidence that serious risks to physical health occur downwind, for example, of livestock confinement facilities, some research suggests that odor-causing substances can cause health effects such as eye, nose and throat irritation, headache, and drowsiness, and possibly aggravate allergies, asthma, and bronchitis (Sohn et al., 2003). More frequently, though, strong, unpleasant, or offensive smells can interfere with a person’s enjoyment of life especially if these odors are frequent and/or persistent. Studies indicate that odors may alter a person’s attitude (Schiffman et al., 1995; Radon et al., 2004). However, it is unclear if the attitude altering is a psychological or physiological response to odor. Some people may feel angry and frustrated because of the distasteful smell of the gases rather than being physically affected by the gases. The argument may still be made, however, that in either case the person’s attitude was altered, making the presence of odors a valid health concern. More research is underway, but it is important to recognize that while odors from food industry may not affect public health in general, odors can affect a person’s lifestyle, and in such cases a person’s psychological well-being may be just as important as his or her physical health.

4. ODOR GUIDELINES AND REGULATIONS The regulations and guidelines vary from country to country. It is neither possible nor practical to provide all guidance or regulations of different countries. As such, the aim of this section is to provide some useful information relevant to the idea behind odor nuisance. It also should be mentioned here that the term

4

E. Psillakis and V. Gekas

“guidelines” in the context of this chapter implies not only numerical values but also any kind of guidance given. 4.1. Germany After World War II, the rapid economic development and the densely populated areas were responsible for the early regulatory legislation regarding air pollution in Germany. The legal basis for any requirement with respect to air quality is the German Federal Protection Act for Ambient Air (1974/1990) and the Technical Instruction on Air Quality Control (1986). According to the Federal Protection Act for Ambient Air, odors caused by plants and are treated as an annoyance, although the problem is to find out whether an annoyance has to be considered as significant (relevance of the annoyance). The state of Northrhine-Westfalia is undoubtedly the most densely populated and industrialized area in Germany and for decades now the authorities of this state have been developing and testing a new regulation/directive, according to which a complete system has been designed, beginning with measurement methods of the existing odor load and calculation of the expected odor load, and concluding with ambient air quality requirements expressed as limit values in terms of maximum permitted odor frequency in ambient air in certain areas (Frechen, 2000; Both, 2001). These limit values were developed on the basis of investigations in which the existing odor load measured as odor frequency (Guideline VDI 3940, 1997) and the degree of odor annoyance of residents assessed by questionnaires according to guideline VDI 3883 Part 1 (1997) were correlated. The result was to set a limit impact concentration at 1 odor unit /m3 (see definition in Chapter 2, paragraph 3.2) and then to limit the time percentage during which a higher impact concentration is tolerable (Both, 2001). These time percentages (odor frequencies) were between 10% for residential areas and 15% for industrial areas. In order to assess more correctly the extent of odorous emission to people it is stipulated that an hour may be recognized as exceeding the limit value if the limit value is exceeded during 10% of 1 hour, thus during 6 minutes. In practice, it is assumed to be sufficient to multiply the hourly mean by a factor of 10 (the so-called factor 10 model). Thus an impact concentration of 0.1 odor unit /m3 (hourly average) equals to 1 odor unit /m3 in the sense of the directive. 4.2. The Netherlands The general policy in Netherlands is to keep the population as free as possible from annoyance. The term “annoyance” is translated as the percentage of population perceiving sometimes or even often an annoying odor. In this context in the year 2000, 12% of the population was annoyed by industrial odors (Frechen, 2001). The pig production sector in the Netherlands is considerable in size, relative to both the size of the population and the surface area of the country. Annual production is approximately 30 million pigs, which amounts to 2 pigs per head

5

Odor Problems in the Food Industry

of the population. It is therefore not surprising that odor impact of pig production is a major environmental issue, given the high density and proximity of both residents and pigs. The first guideline on how to take account of environmental odor aspects for licensing as a result of application of the existing the Nuisance Law was first issued in 1971 and revised several times in later years. 4.3. United Kingdom The Environmental Protection Act of 1990 provides the legal framework for avoiding and controlling odor nuisance in the United Kingdom. A comprehensive overview over “The legal context of odor annoyance” in the United Kingdom can be found elsewhere (Salter, 2000). The regulations, however, do not set any general valid emission standards concerning odor and retracts to more general statements concerning the odor nuisance. This is expected to change in the near future after publication of Technical Guidance Note H4 (2002), Integrated Pollution Prevention and Control (IPPC), Horizontal Guidance for Odor by the Environment Agency. 4.4. United States of America The regulations for odor in the United States of America vary from state to state. Depending on the state regulation may concern hydrogen sulfide limits, detection to threshold limits, or use as the main legal basis the nuisance law. An overview on the regulation of odors in the United States is given by Thomas Mahin (2001) and is summarized in Table 1.

Table 1. Examples of Ambient Standards for Odor-Causing Compounds [all agencies listed are state agencies unless otherwise noted; from Mahin (2001)] Location

Compound

Ambient Odor Standard

California Idaho

Hydrogen sulfide Hydrogen sulfide

Minnesota

Hydrogen sulfide

Nebraska New York City North Dakota Pennsylvania

Total reduced sulfur Hydrogen sulfide Hydrogen sulfide Hydrogen sulfide

Texas

Hydrogen sulfide

30 ppbva (1-hr average) 10 ppbv (24 hr average) 30 ppbv (30 min average) 30 ppbv (30 min average)b 50 ppbv (30 min average)c 100 ppb (30 min average) 1 ppbv (for wastewater plants) 50 ppbv (instantaneous, two readings 15 min apart) 100 ppbv (1 hr average) 5 ppbv (24 hr average) 80 ppbv (30 min average) residential/commercial 120 ppbv industrial, vacant or range lands

a

Parts per billion by volume. Not to be exceeded more than two days in a 5-day period. c Not to be exceeded more than two times per year. b

6

E. Psillakis and V. Gekas

4.5. World Health Organization Guideline Values For substances with malodorous properties at concentrations below those where toxic effects occur, guideline values likely to protect the public from odor nuisance were established by the World Health Organization (1987, 2000). These were based on data provided by expert panels and field studies for a limited range of substances as 24-hour average concentrations. They were derived with the aim of providing a basis for protecting the public from the adverse effects of air pollution. For a few of these substances that exhibit malodorous properties at concentrations below that at which toxic effects occur, guideline values have been established for avoidance of substantial annoyance. For example, in the case of hydrogen sulfide, guideline values based on sensory effects or annoyance reactions, for an average time of 30 minutes, were set at 7 µg m−3 in order to protect against substantial annoyance. It is important to note that these guidelines have been established for single chemicals. Mixtures of chemicals can have additive, synergistic, or antagonistic effects.

5. ODORS CHEMISTRY A number of factors affect the emission of compounds from food industry operations. Most of the substances emitted are the products of microbial processes and in most cases it is the microbial environment that will determine which substances are generated and at what rate. This section intends to describe the chemical and biological mechanisms that affect the formation and release of emissions. Most of the information used in this section was taken by the draft report of the US Environmental Protection Agency (2001). 5.1. Ammonia The microbial decomposition of organic nitrogen compounds under both aerobic and anaerobic conditions leads to the production of ammonia as a by-product. The amount of ammonia volatilized from any operation depends on total ammonia concentration, temperature, pH, and storage time (US Environmental Protection Agency, 2001). Furthermore, in solution, the partitioning of ammonia between the ionized (NH4+) and unionized (NH3) species is controlled by pH and temperature. As expected, at lower pH, ammonium is the predominant species, and ammonia volatilization occurs at a lower rate than at higher pH values where the rate of ammonia volatilization is increased. Given that the pH of manures handled as solids ranges between 7.5 and 8.5, ammonia volatilization is enhanced. In the case of manure, other aminocompounds including aliphatic amines (methyl- and ethylamine) also may present but at lower concentrations. It

Odor Problems in the Food Industry

7

should be mentioned here that during the storage of fresh manure, amino acids are most likely to undergo decarboxylation producing putrescine, cadaverine, and ammonia (Zhu, 2000). 5.2. Sulfur Compounds Hydrogen sulfide and other reduced sulfur compounds are produced during anaerobic decomposition and hydrogen sulfide is the predominant reduced sulfur compound emitted from animal feeding operations. In the case of animal manures, there are two primary sources of sulfur (US Environmental Protection Agency, 2001): (1) sulfur amino acids contained in the feed and (2) inorganic sulfur compounds, such as copper sulfate and zinc sulfate, serving as growth stimulants and used as feed additives to supply trace minerals. In general, the magnitude of hydrogen sulfide emissions depends on the liquid phase concentration, temperature, and pH. It is well known that the solubility of hydrogen sulfide in water increases at pH values above 7. Therefore, as pH shifts from alkaline to acidic, the potential for hydrogen sulfide emissions increases (Stumn and Morgan, 1996). As already stated, under anaerobic conditions, any excreted sulfur will be microbially reduced to hydrogen sulfide. Thus, all manures managed as liquids or slurries are potential sources of hydrogen sulfide emissions; as such emissions from confinement facilities with dry manure handling systems and dry manure stockpiles should be negligible if there is adequate exposure to atmospheric oxygen to maintain aerobic conditions. Other sulfur compounds emitted from animal feeding operations include methyl mercaptan, dimethyl sulfide, dimethyl disulfide, and carbonyl sulfide. In general, the very offensive smelling compound methyl mercaptan is a product of sulfur-containing amino acid decomposition, and it can be oxidized to form the unpleasant-smelling compounds dimethyl disulfide or dimethyl sulfide (Zhu, 2000). 5.3. Volatile Organic Compounds During the degradation of organic matter, volatile organic compounds (VOC) are formed as intermediate metabolites. Under aerobic conditions, microbial degradation will lead to the formation of VOC, which are rapidly oxidized to carbon dioxide and water. Under anaerobic conditions, complex organic compounds are degraded microbially to volatile organic acids and other volatile organic compounds, which, however, in turn may be converted to methane and carbon dioxide by methanogenic bacteria (US Environmental Protection Agency, 2001). It should be mentioned here that volatile fatty acids (VFA) represent a large portion of VOC and are responsible for a significant proportion of odor

8

E. Psillakis and V. Gekas

in emission plumes from swine production facilities. Together with CO2, H2, as well as ammonia, they can be produced from the deamination of amino acids that are produced during the process of protein degradation and breakdown of carbohydrates. Typical acids in this group consist of acetic, propionic, butyric, iso-butyric, valeric, iso-valeric, caproic, and capric acids. 5.4. Phenolic Compounds Phenolic compounds such as phenols and p-cresols are produced from the microbial degradation of tyrosine and phenylalanine in the intestinal tract of animals (Ishaque et al., 1985). Indole, skatole, cresol, and 4-ethylphenol appear to be the major components included in this group of compounds. In addition, the metabolism of tryptophan can result in the production of indoleacetate, which is subsequently converted into skatole (3-methylindole) and indole by a different group of bacteria (Zhu, 2000).

6. FOOD INDUSTRY COMPONENTS The major components in the food industry are depicted on Figure 1. As can been seen edible products go through a number of stages before reaching the consumer. It is generally accepted that odor problems in the food industry are more likely to occur during the first two stages. Food service and food retail operations are themselves generally not significant sources of air or odor pollution. No data are available on the portion of total emissions of specific pollutants that can be ascribed directly to the food service and food retail industries. The one exception is emissions from vent hood systems of food service equipment, where in some cases users of certain food service cooking equipment are requested to install pollution control measures. A selection of the food industry activities of the first two stages are given in Table 2. In the food industry large livestock operations, poultry farms, slaughterhouses, food and meat processing industries, and bone mills are among major contributors to odor pollution.

Grower/ Fisherman

Processing

Wholesale

Service/ Retail

Consumption

Figure 1. Major components in the food industry. Adapted from Davies and Konisky (2000).

9

Odor Problems in the Food Industry

Table 2. Selected Food Industry Activities Industry Activities

Description

Livestock & poultry feed operations

• Cattle feedlots • Swine feedlots • Poultry houses • Dairy farms

Animal & meat products preparation

• Meat packing plants • Meat smokehouses • Meat rendering plants • Manure processing • Poultry slaughtering

Fish processing

• Fish canning

Dairy products

• Natural and processed cheese

Preserved fruits and vegetables

• Canned fruits and vegetables • Dehydrated fruits and vegetables • Pickles, sauces, and salad dressings

Grain processing

• Grain elevators & processes • Cereal breakfast food • Pet food • Alfalfa dehydrating • Pasta manufacturing • Bread baking • Corn wet milling

Confectionary products

• Sugar processing • Cane sugar processing • Sugar beet processing • Salted & roasted nuts & seeds • Almond processing • Peanut processing • Vegetable oil processing

7. SELECTED CASE STUDIES OF ODOR PROBLEMS IN THE FOOD INDUSTRY 7.1. Livestock Enterprises (including animal feeding operations and manure processing) Odors and gases emitted are a by-product of the microbial decomposition of manure and other organic matter. The amount and type of emissions are dependent on the amount and type of microbial activity and consequently to the moisture content, temperature, pH, oxygen concentration, and other environmental parameters. Any changes in these parameters will alter odor and gas emissions. For example, as temperature decreases, microbial activity slows down, and as such during the winter months few odors are generated. Emission sources are distributed among livestock buildings, manure storage units, land application of manure, and pastures (Skinner et al., 1997). Gas emissions

10

E. Psillakis and V. Gekas

from buildings and storages are relatively constant and vary with seasonal temperatures. In contrast, land application of manure has the potential to emit large amounts of gases periodically throughout the year. Gas emissions vary significantly with management practices and manure system design. The main odorous compound groups associated with animal feeding operation and wastes are fatty acids, amines, ammonia, aromatics, as well as inorganic and organic sulfur (US Environmental Protection Agency, 2001). The most frequently detected odorous compound from livestock operations is undoubtedly hydrogen sulfide. Another group of malodorous compounds released during animal feeding operations are the volatile organic compounds. Next to the odor problem created by these compounds, the majority of them have an environmental impact, given that some of them participate in atmospheric photochemical reactions, while others play an important role as heattrapping gases. 7.2. Meat Processing Meat smokehouses are used to add flavor, color, and aroma to various meats, including pork, beef, poultry, and fish. The operations typically involved in the production of smoked meat are tempering or drying, smoking, cooking, and finally chilling. In general, the pollutants associated with meat smokehouses include particulate matter (PM), carbon monoxide (CO), volatile organic compounds (VOC), polycyclic aromatic hydrocarbons (PAH), organic acids, acrolein, acetaldehyde, formaldehyde, and nitrogen oxides. Acetic acid (followed by formic, propionic, butyric, and other acids) has been identified as the most prevalent organic acid present in smoke that is the primary source of emissions in the smokehouses. Furthermore, heating zones in continuous smokehouses (and the cooking cycle in batch smokehouses) are not a source of combustion compounds [e.g., polycyclic aromatic hydrocarbons (PAH)]; they are, however, a source of odor including small amounts of VOC as a result of the volatilization of organic compounds contained in the meat or the smoke previously applied to the meat (AP-42, 2004). 7.3. Meat-Rendering Plants Meat-rendering plants process animal by-product materials for the production of tallow, greas, and high-protein meat and bone meal. There are two types of animal-rendering processes and these are (1) the edible rendering plants that process fatty animal tissue into edible fats and proteins and inedible rendering and (2) the inedible rendering plants produce inedible tallow and grease, which are used in livestock and poultry feed, soap, and production of fatty acids. Volatile organic compounds are the primary air pollutants emitted from these rendering operations (AP-42, 2004) and are considered to be an odor nuisance in residential areas in close proximity to rendering plants. The major constituents

Odor Problems in the Food Industry

11

that have been qualitatively identified as potential emissions include organic sulphides, disulfides, C-4 to C-7 aldehydes, trimethylamine, C-4 amines, quinoline, dimethyl pyrazine, other pyrazines, and C-3 to C-6 organic acids. In addition, lesser amounts of C-4 to C-7 alcohols, ketones, aliphatic hydrocarbons, and aromatic compounds are potentially emitted. 7.4. Fish Canning Fish processing includes both the canning of fish for human consumption and the production of fish by-products such as meal and oil. Although smoke and particulates may be a problem, odors are the most objectionable emissions from fish processing plants. The fish in the by-products segment is often in a further state of decomposition, thus causing more odor problems than canning itself. The fish meal driers form part of the fish by-products segment and are the largest odor source. In addition, the odorous gases from reduction cookers consist primarily of hydrogen sulfide (H2S) and trimethylamine [(CH3)3N] but are emitted from this stage in appreciably smaller volumes than from fish meal driers (AP-42, 2004). 7.5. Cane Sugar Processing One of the main environmental problems associated with sugar processing is the management of effluent. All stages in sugar processing and refining produce wastes that collectively are very rich in organic content, high in suspended solids, and also may be colored (AP-42, 2004). The release of untreated wastewaters to surface waters will cause extensive enrichment, stripping oxygen from the water and killing aquatic life. As such, wastewater will need to be treated on-site before released to sewer or surface water. Odor from the waste lagoons and from processing may result in local nuisance complaints. If deemed a nuisance by the authorities, mitigation measures may need to be implemented, though these usually relate to housekeeping measures. 7.6. Baking The oven exhaust from baking industries releases water vapor, CO2, VOC, and various combustion products. The VOCs are primarily ethanol produced by the yeast during the fermentation process. It should be noted, however, that many individuals find the odor of fresh baked bread very desirable.

8. CONCLUSIONS How much odor should a community or individual have to tolerate? This is the primary question that must be answered before any good odor policy can be developed. Everyone would probably agree that if one person smells a facility

12

E. Psillakis and V. Gekas

one day out of the year, that facility should not be declared a nuisance. But several hundred people smelling a facility nearly every day throughout the year would be considered a nuisance. The reality at most sites lies somewhere in the middle. A definition of nuisance must be established that takes into account odor intensity and frequency, meaning how bad and how often. An odor nuisance also may have to consider a relationship between the number of people annoyed. Included in this decision also might be the economic impact of odor control on the food industry.

9. ACRONYMS AND ABBREVIATIONS mve PM PAH VFA VOC

Mestvarkeneenheden or fattener units Particulate matter Polycyclic aromatic hydrocarbons Volatile fatty acids Volatile organic compounds

10. REFERENCES AP-42 (2004) “Compilation of Air Pollutant Emission Factors, AP-42, Fifth Edition, Volume I: Stationary Point and Area Sources.” U.S. Environmental Protection Agency. http://www.epa. gov/ttn/ chief/ap42/index.html Both, R. (2001) “Directive on odor in ambient air: an established system of odor measurement and odor regulation in Germany.” Water Science & Technology 44 (9), 119–126. Bliss, P.J., Schulz, T.J., Senger, T., Kaye, R.B. (1996) “Odor measurement—Factors affecting olfactometry panel performance.” Water Science and Technology, 34 (3–4), 549–556. Davies, T., Konisky, D.M. (2000) “Environmental Implications of the Foodservice and Food Retail Industries.” Discussion Paper 00–11, Resources for the Future, Washington DC. http://www.rff. org [2004. Nov 23] Environmental Protection Agency, Ireland (2001) “Odor Impacts and Odor Emission Control Measures for Intensive Agriculture.” Environmental Research R&D REPORT SERIES No. 14. http://www.epa.ie/pubs/ Federal Protection Act for Ambient Air (1974/1990) Act on the Prevention of Harmful Effects on the Environment Caused by Air Pollution, Noise, Vibration and Similar Phenomena (Federal Emission Control Act, Bundes-Immissionschutzgesetz—BImSchG) Federal Ministry for Environment, Nature Conservation and Reactor Safety, Bonn (BGBl. I p.880). (available in English). Frechen, F.-B. (1994) “Odor emissions from wastewater treatment plants—recent German experiences.” Water Science and Technology, 30 (4), 35–46. Frechen, F.-B. (2000) “Odor measurement policy in Germany.” Water Science and Technology, 41 (6), 17–24. Frechen, F.-B. (2001) “Regulation and policies.” In Stuetz, R.M., Frechen, F.-B. (eds), Odours in Wastewater, Treatment: Measurement, Modeling and Control. London: IWA publishing. Gostelow, P., Parsons, S.A., Stuetz, R.M. (2001) “Odor measurements for sewage treatment works.” Water Research, 35 (3), 579–597. Guideline VDI 3883 Part 1 (1997). Effects and Assessment of Odors. Psychometric Assessment of Odor Annoyance Questionnaires. Dósseldorf (German/English).

Odor Problems in the Food Industry

13

Guideline VDI 3940 (1993). Determination of Odorants in Ambient Air by Field Inspections. Düsseldorf. (German/English). Ishaque, M., Bisaillon, J.G., Beaudet, R., Sylvestre, M. (1985) “Degradation of phenolic compounds by microorganisms indigenous to swine waste.” Agricultural Wastes, 13 (3), 229–235. ISO 5492:1992 (1992) “Sensory analysis—vocabulary” International Organization for Standardization. Jacob, T.J.C., Fraser, C., Wang, L., Walker, V., O’Connor, S. (2003) “Psychological evaluation responses to pleasant and maldor stimulation in human subjects; adaptation, dose response and gender differences.” International Journal of Psychophysiology, 48 (1), 67–80. Lancet, D. “Olfaction—The strong scent to success.” Nature 351 (6324), 275–276. Mahin, T.D. (2001) “Comparison of different approaches used to regulate odors around the world.” Water Science and Technology, 44 (9), 87–102. Pearce T.C. (1997) “Computational parallels between the biological olfactory pathway and its analogue. The Electronic Nose .1. Biological olfaction.” Biosystems, 41 (1), 43–67. Radon K., Peters A., Praml G., Ehrenstein V., Schulze A., Hehl O., Nowak D. (2004) “Livestock odors and quality of life of neighboring residents.” Annals of Agricultural and Environmental Medicine 11 (1), 59–62. Rosenkranz, H.S., Cunningham, A.R. (2003) “Environmental odors and health hazards.” The Science of the Total Environment, 313 (1–3), 15–24. Salter, J. (2000) “The legal context of odor annoyance.” Proceedings of the International Meeting on Odor Measurement and Modeling, Odor 1, Cranfield University. Schiffman, S.S., Sattely Miller, E.A., Suggs, M.S., Graham, B.G. (1995) “The effect of environmental odors, emanating from commercial swine operations on the mood of nearby residents.” Brain Research Bulletin 37 (4), 369–375. Skinner, J.A., Lewis, K.A., Bardon, K.S., Tucker, P., Catt, J.A., Chamber, (1997) “An overview of the environmental impact of agriculture in the U.K.” Journal of Environmental Management, 50 (2), 111–128. Sohn, J.H., Smith, R., Yoong, E., Leis, J., Galvin G. (2003) “Quantification of odors from piggery effluent ponds using an electronic nose and an artificial neural network.” Biosystems Engineering 86 (4), 399–410 Stuetz, R.M., Gostelow, P., Burgess, J.E. (2001) “Odor perception.” In: Stuetz, R.M., Frechen, F.-B. (eds), Odours in Wastewater Treatment: Measurement, Modeling and Control. London, IWA publishing. Stumn, W., Morgan J.J. (1996) Aquatic Chemistry: Chemical Equilibria and Rates in Natural Waters. Third Edition. Environmental Science and Technology, New York, John Wiley & Sons, Inc. Technical Guidance Note H4, Integrated Pollution Prevention and Control (IPPC), Horizontal Guidance for Odor, Part 2 Assessment and Control, Environment Agency, UK, 2002. Available on www.environment-agency.gov.uk Technical Instruction on Air Quality Control (1986) (Erste Allgemeine Verwaltungsvorschrift zum Bundes-Immissionsschutzgesetz – Technische Anleitung zur Reinhaltung der Luft – TA Luft) Federal Ministry for Environment, Nature Conservation and Reactor Safety, Bonn (GMBl. p. 95). (available in English). US Environmental Protection Agency (2001) “Emissions from animal feeding operations, draft.” EPA Contract No. 68-D6-0011, Emission Standards Division, Office of Air Quality Planning and Standards, Research Triangle Park, NC. World Health Organization (1987) “Air quality guidelines for europe.” World Health Organization Regional Publications, European Series No 23. World Health Organization (2000) “Air quality guidelines for Europe. Second edition”. World Health Organization Regional Publications, European Series No 91. Zald, D.H., Pardo, J.V. (2000) “Functional neuroimaging of the olfactory system in humans.” International Journal of Psychophysiology, 36 (2), 1165–1181. Zhu, J. (2000) “A review of microbiology in swine manure odor control.” Agriculture, Ecosystems & Environment, 78 (2), 93–106.

2 Odor Measurement Elefteria Psillakis

1. INTRODUCTION It generally is recognized that for effective odor control measures to be implemented the problem must first be quantified. Odor measurement data can be used to: ●

● ●

●

Predict odor impact in the vicinity of an operation for odor impact assessment purposes. Provide information on the strength and intensity of odors. Identify the causes of an odor problem and quantify the scale of odor emission from a particular source. Measure/evaluate the performance of an odor control technology implemented by a company.

However, odors are temporal and spatially dimensioned and can be considered to be one of the most difficult challenges for scientists to investigate. A person’s response to an odor is highly subjective: different people find different odors offensive and at different concentrations given that physiologically odor recognition is associated with the emotional center of brain. Furthermore, some of the odorous compounds can be detected by the human nose in very low concentrations (e.g., hydrogen sulfide) while others cannot be detected even at very high concentrations (e.g., methane). This is further complicated by the fact that some combinations of compounds may be more odorous than the sum of the individual gases. Determining the impact area of odorous gases also is very difficult. As wind direction and speed change, the odor impact area and intensity change. Depending on specific conditions, odorous gases can travel several meters or several kilometers. Gas transmission and impact area also depend on the specific gas. For these reasons, there is no universally accepted method for the quantification of odors, and odor measurement often has been regarded as an art as opposed to a science (Gostelow et al., 2001). ELEFTERIA PSILLAKIS ● Department of Environmental Engineering, Technical University of Crete, Polytechnioupolis, GR-73100 Chania-Crete, Greece e-mail: [email protected]. 15

16

E. Psillakis

2. SAMPLING The sampling of an environmental parameter, such as odor, is an extremely complex task for environmental engineers and scientists. Prior to preparing a sampling program it is necessary for a field inspection to be carried out in order to determine the potential odor sources at the site to be investigated. Factors to be noted include location and conditions affecting odor emissions (controlled or uncontrolled such as weather), accessibility of odor sampling points, and characteristics of odor emissions (e.g., temperature and humidity) as well as safety requirements due to toxicity. Sampling points as well as frequency, duration, and averaging time of sampling should reflect both spatial and temporal pattern of the facility to be investigated. Increasing the number of samples to be taken will increase the cost of the study. In some cases composite samples across several sampling points are used instead of a large number of discrete samples (Jiang and Kaye, 2001). It therefore is essential to carefully design the sampling program and to coordinate it in advance with the testing program, thus minimizing the time elapsed between sampling and analysis. 2.1. General Considerations for Sampling Odorous Compound In those parts of the sampling equipment coming into contact with the odorant sample, appropriate materials such as PTFE, PET, FEP, Tedlar™, glass, or stainless steel must be used. Materials such as brass and silicone or natural rubber must not be used in sampling lines and fittings. In general, representative odorous samples are collected in the field using special purpose atmospheric sampling bags, whose quality must conform to criteria such as being odor free, nonadsoprtive, leak-free, reasonably robust, and having sufficient volumetric capacity. Commonly, sampling bags are made of Tedlar™, FEP, or the low-cost/ single-use Naoplhan NA™. Every new sampling bag should be tested for their background odor concentration as well as for leakage. Because of their high cost it is a common practice to reuse these bags. Nonetheless, where materials are reused, procedures for cleaning and conditioning must be applied as defined in certified methods. In addition, the interval between sampling and measurement should not exceed 30 hr, during which time samples must be kept in the dark and at a temperature below 25°C (EN 13725,2003). Sampling bags may be filled using either the direct or indirect sampling techniques. According to the direct sampling, the bag is filled under pressure by pumping the air sample into the bag. In indirect sampling, the odor sample is collected using an odor sampling system, having a vacuum pump and a 12volt battery built into a sampling drum as shown in Figure 1. After placing a sampling bag into the sealed sampling vessel, air is pumped out of the sampling drum, creating a vacuum inside the drum. Sample air is drawn into the bag by the pressure difference between the inside and outside of the bag. Because of the

17

Odor Measurement Odour Sample Intake

Tedlar Bag

Vacuum Exhaust

Switch Pump and battery

Figure 1. Arrangement of apparatus for indirect odor sampling system.

risk of contamination the direct method is seldom used and the indirect sampling approach is generally recommended. 2.2. Emission Source Types Typically emissions sources are characterized as point, area, and volume. Another type of emission source is the fugitive one (e.g., leaking valves). However, according to the EN 13725:2003, quantification of such emissions incurs large errors and consequently no known technique can be recommended. 2.2.1. Sampling of Point Sources Typically a point source will be a stack with a known flow rate such as a discharge stack from abattoir or even a vent from a pig shed. Gaseous samples from these sources can be collected using a sampling train consisting of a probe, a delivery pipe, and an optional particulate filter before the collection system (EN 13725:2003) such as the one described in the US EPA Method 0030 (1986). It is important that air velocity, dimensions of the vent, temperature, and humidity are measured before a sample is taken. Appropriate guidance regarding the selection of sampling points and velocity measurement and sampling point location can be found in ISO Method 9096: 1992(E) (Stationary source emissions—Determination of concentration and mass flow rate of particulate material in gas-carrying ducts—Manual gravimetric method), ISO 10780: 1994(E) (Stationary source emissions—Measurement of velocity and volume flow rate of gas streams in ducts), and Australian Standard AS 4323.1 1995 (Stationary Source Emissions, Method 1—Selection of Sampling Positions). Predilution of the sample should be undertaken for samples taken directly from combustion processes where the air temperature and relative humidity

18

E. Psillakis

exceed 50 °C and 90%, respectively, thus preventing condensation in the sample bag as well as reducing the concentration to levels suitable for analysis (Jiang and Kaye, 2001). 2.2.2. Sampling of Area Sources Typically an area source will be a water or solid surface such as the water surface of a slurry storage tank or a cattle feedlot and can be sampled using different methods such as the portable wind tunnels. Area sources can be distinguished in sources with a measurable outward airflow and sources that do emit odor but have no measurable outward flow. 2.2.2a. Sources Without Outward Flow In the early 1970s, Lindvall (1970) introduced an odor emission hood used in the comparison study of odor strengths from different sources. Later, Lockyer (1984) designed a wind tunnel system to measure ammonia loses from pastures. Recently, an improved portable wind tunnel system was developed at the University of New South Wales (Jiang et al., 1995; Bliss et al., 1995) to measure odors emission rates from liquid and solid area sources (sewage and industrial wastewater treatment plants, cattle feedlots, mushroom composting, piggeries, etc). The principle of the wind tunnel system is that controlled air (recently filtered by activated carbon through a series of devices) forms a consistent flow over a defined liquid or solid surface. Convective mass transfer takes place above the surface as odor emission happens in the natural atmosphere. The odor emissions are then mixed with clean air and vented out of the hood. A proportion of the mixture is drawn into a sampling bag via Teflon tubing using the sampling vessel. The air velocity used inside the wind tunnel is 0.3 m/sec, which is the lowest reliable measurable air velocity directly inside the main section of the wind tunnel. The aerodynamic performance of the wind tunnel has been validated based on the wind speeds and height where most complaints occurred. An isometric sketch of a portable wind tunnel system is shown at Figure 2. It should be mentioned here that according to the EN 13725:2003 this method is of limited application when the tunnels use unfiltered atmospheric air as intake, which is not always odor free. An alternative to wind tunnel systems is the isolation chambers, also called flux hoods (Klenbusch, 1986; Gholson et al., 1991). According to this method, during sampling a representative area of the source surface is enclosed by the isolation flux chamber. Then a controlled flow of clean sweep gas (normally nitrogen or odor- and hydrocarbon-free air) is released into the chamber and is used to transport the emission from the surface to be sampled. The performance of isolation chambers largely depends on the configuration of the enclosure and operating procedures (Hwang, 1985). A comparative study on portable wind tunnel system and isolation chamber for the determination of VOCs from real sources (Jiang and Kaye, 1996)

19

Odor Measurement Mixing Chamber Main Section

Floating Tubes Figure 2. Isometric sketch of portable wind tunnel system.

(Table 1) reported emission rates that were not in accordance and emphasized the implications of the Henry’s Law constant for mass transfer processes in the case of the isolation chamber. The same report concluded that this was due to the fact that the use of isolation chambers may result in significant underestimations of emission rates for real situations and suggested the use of wind tunnel systems. Nonetheless, the EN 13725:2003 clearly indicates the need for further research and standardization.

Table 1. Comparison of Wind Tunnel and Isolation Chamber (Jiang, and Kaye, 1996) Wind tunnel

Isolation chamber

Aerodynamics

Parallel and even air flow inside the wind tunnel Emissions and air well mixed at the exit

Operating parameters

Temperature and relative humidity are close to ambient condition during sampling

Conditions at the emission surface not known Emissions and sweep gas may not be well mixed Temperature and relative humidity are influenced by solar heating during long sampling duration Sweep gas = 5L/min Equilibrium time = 24 min Sample collection rate = 2 L/min Not for chemicals that have low Henry’s Law constants

Limitations

Air velocity = 0.3 m/s Stabilization time = 5 min Sample collection rate = 20 L/min Sample contamination due to atmospheric air as intake

20

E. Psillakis

2.2.2b. Sources with Outward Flow Although, the use of wind tunnels has been demonstrated for all types of area sources (Jiang and Kaye, 1996), significant limitations may exist in the cases of area sources with outward flow. For example, in some instances the placement of the wind tunnel system may create back pressure, limiting the flow of outward moving air into the wind tunnel, and leading to an underestimation of odor emission rates (Jiang and Kaye, 2001). The ideal situation would be to cover the whole total source (or a large part) with foil. The cover must be left open at some point in order to allow the air to escape as well as to allow sampling using the point source apparatus (EN 13725:2003). 2.2.3. Sampling of Building Sources Typically building sources, such as chicken and pig sheds, have a number of openings. Prior to about 10 years ago, little research was undertaken on the determination of odor emissions from buildings. For building sources, measurements of both odor concentration and air ventilation rate are required. The air ventilation rate from animal housing is dependent on operational conditions (e.g., opening or closure of side flaps or shutters) and ambient wind speed and direction (Jiang and Kaye, 2001). For animal sheds, odor samples are normally taken from several points within a shed. Experience indicates that composite sample may be sufficient to represent a single shed at a particular time. Additional samples can be taken at different times of the day or week or to understand the fluctuation of the odor concentration levels within a day or a week. Similarly, sampling may be carried out for different weeks during the grow-out cycle or for different seasons during a year or longer. 2.2.4. Calculation of Emission Rates Emission rates, required for odor impact assessment, are calculates using the odor concentration measured by olfactometry together with other measured properties of the emission source and the sampling apparatus. For point sources the odor emission rate (OER) is calculated using the odor concentration (OC: ou/m3; see definition in paragraph 3.2) measured by olfactometer and the measured gas flow rate (Q,m3/s) according to the following equation: OER = Q × OC The specific odor emission rate (SOER) may be defined as the quantity (mass) of odor emitted per unit time from a unit surface area. The quantity of odor emitted is not determined directly by olfactometry but is calculated from the concentration of odor (OC: ou/m3) as measured by olfactometry, which is then multiplied by the volume of air passing through the hood per unit time. The volume per unit time is calculated from the measured velocity through the wind

21

Odor Measurement

tunnel, which is then multiplied by the known cross-sectional area of the wind tunnel. Given that A is the area covered by the wind tunnel (m2), then SOER is calculated by the expression: Q × OC SOER = A Finally, for building sources the odor emission rate (OER) is calculated using the odor concentration measured by olfactometer through the door and window openings. The following equation can be applied and the measured gas flow rate (Q, m3/s) according to the following equation: OER = Q × OC where Q is the gas ventilation rate (m3/s) and OC is the odor concentration measured by olfactometer (ou/m3).

3. POINTS TO CONSIDER 3.1. Terms Associated with an Odor Measurement In considering odor measurement, it is important to distinguish between odorants and odors. An odorant is a substance that stimulates a human olfactory system so that an odor is perceived, whereas an odor is an organoleptic attribute perceptible by the olfactory organ on sniffing certain volatile substances (EN 13725:2003). The linkage between odorant properties and odor perception is not clear due to the lack of a comprehensive theory of olfaction. Two broad classes of odor measurement exist as a result: analytical measurements, referring to odorants, and sensory measurements, employing human subjects relating to odors (Gostelow et al., 2001). The terms (also called dimensions) associated with an odor measurement refer to the parameters of an odor that can be measured. There are four generally accepted terms: concentration, intensity, character, and hedonic tone. Odor concentration is the most frequently measured parameter and can be measured analytically or by sensory means. Analytical measurements give the physical concentration for specific odorants, whereas sensory concentration measurements determine the number of dilutions required to reduce an odor to its threshold concentration, which is the lowest concentration at which an odor either can be detected or recognized (Gostelow et al., 2001). Recognition thresholds are typically higher than detection thresholds by a factor of 1.5–10 (Dravnieks and Jarke, 1980). It should be mentioned here that odorant concentration is the only odor dimension that can be measured analytically. The rest of the parameters can be measured only using sensory methods. Odor intensity is the strength of the perceived odor sensation. It is related to the odorant concentration. The odor intensity usually is stated according to a predetermined subjective rating system (i.e., faint, moderate, strong) by subjective

22

E. Psillakis

magnitude estimates, or (i.e., odor A is twice as strong as odor B) or by reference to a specific odorant, whose concentration is adjusted until the reference and test odorant have the same perceived intensity (Gostelow et al., 2001). It is important to recognize the distinction between odor intensity and gas concentration. Odor intensity is a measure of detection sensed by the nose. Odor concentration is the actual concentration of the gas in the air. Nonetheless, the two terms are related and the perceived intensity increases with increasing odor concentration, although the relationship is not linear. The laws that have been proposed to explain intensity–concentration relationships are the Weber–Fechner law and Steven’s law (Sarkar and Hobbs, 2002). Odor character or quality is the property to identify an odor and to differentiate it from another odor of equal intensity. Examples of odor characters for specific odorants are given in Table 2. The hedonic tone is a property of an odor relating to its pleasantness or unpleasantness. When an odor is evaluated for its hedonic tone in the neutral context of an olfactometric presentation, the panelist is exposed to a controlled stimulus in terms of intensity and duration. The degree of pleasantness or unpleasantness is determined by each panelist’s experience and emotional associations. Often, negative values are used to represent unpleasant odors and positive values represent pleasant odors. 3.2. The Unit of Odor Measurement The odor unit is a difficult unit to define because it relates a physiological effect to the stimulus that caused it. Odor concentrations derived by threshold olfactometry are dimensionless. The concentrations may be termed threshold odor numbers (TON) or dilution to threshold (D/T) ratios. Nonetheless, it is becoming increasingly common to envisage odor concentrations as physical concentrations and to express them as odor units per cubic meter (ou/m3). In the United States, the same measure has been expressed as odor units per cubic foot (ou/ft3). Numerically, TON, ou/m3, and ou/ft3 are identical, so care must be taken when the concentration ratios are expressed as a physical concentration. According to the EN 13725:2003, the European odor unit (ouE) is that amount of odorant(s) that when evaporated into 1 m3 of neutral gas at standard conditions elicits a physiological response from a panel (detection threshold) equivalent to that elicited by one european reference odor mass (EROM) evaporated in 1 m3 of neutral gas at standard conditions. One EROM evaporated into 1 m3 of neutral gas at standard conditions is the mass of substance that will elicit the D50, which is 50% incidence of effect in physiological response (detection threshold) assessed by an odor panel in conformity with this standard and has by definition a concentration of 1 ouE/m3. For n-butanol one EROM is 123 mg. Evaporated in 1 m3 of neutral gas, at standard conditions, this produces a concentration of 0,040 mmole/mole (which is equal to a volume fraction of 40 parts per billion). There is one relationship between the ouE for the reference odorant

23

Odor Measurement

Table 2. Odorants Associated with Sewage Treatment Works (Abbott, 1993; Bonnin et al., 1990; Brennan, 1993; Gostelow et al., 2001; Young, 1984) Class

Compound

Formula

Character

Sulfurous

Hydrogen sulfide Dimethyl sulfide Diethyl sulfide Diphenyl sulfide Diallyl sulfide Carbon disulfide Dimethyl disulfide Methyl mercaptan Ethyl mercaptan Propyl mercaptan Butyl mercaptan tButyl mercaptan Allyl mercaptan Crotyl mercaptan Benzyl mercaptan Thiocresol Thiophenol Sulfur dioxide Ammonia Methylamine Dimethylamine Trimethylamine Ethylamine Diamines Pyridine Indole Scatole or Skatole Acetic (ethanoic) Butyric (butanoic) Valeric (pentanoic) Formaldehyde Acetaldehyde Butyraldehyde Isobutyraldehyde Isovaleraldehyde Acetone Butanone

H2S (CH3)2S (C2H5)2S (C6H5)2S (CH2CHCH2)2S CS2 (CH3)2S2 CH3SH C2H5SH C3H7SH C4H9SH (CH3)3CSH CH2CHCH2SH CH3CHCHCH2SH C6H5CH2SH CH3C6H4SH C6H5SH SO2 NH3 CH3NH2 (CH3)2NH (CH3)3N C2H5NH2 NH2(CH2)5NH2 C6H5N C8H6NH C9H8NH CH3COOH C3H7COOH C4H9COOH HCHO CH3CHO C3H7CHO (CH3)2CHCHO (CH3)2CHCH2CHO CH3COCH3 C2H5COCH3

Rotten eggs Decayed vegetables, garlic Nauseating, ether Unpleasant, burnt rubber Garlic Decayed vegetables Putrification Decayed cabbage, garlic Decayed cabbage Unpleasant Unpleasant Unpleasant Garlic Skunk, rancid Unpleasant Skunk, rancid Putrid, nauseating, decay Sharp, pungent, irritating Sharp, pungent Fishy Fishy Fishy, ammoniacal Ammoniacal Decomposing meat Disagreeable, irritating Fecal, nauseating Fecal, nauseating Vinegar Rancid, sweaty Sweaty Acrid, suffocating Fruit, apple Rancid, sweaty Fruit Fruit, apple Fruit, sweet Green apple

Nitrogenous

Acids

Aldehydes / ketones

and that for any mixture of odorants. This relationship is defined only at the D50 physiological response level (detection threshold), where: 1 EROM ≡ 123 mg n-butanol ≡ 1 ouE for the mixture of odorants This linkage is the basis of traceability of odor units for any odorant to that of the reference odorant. It effectively expresses odor concentrations in terms of “n-butanol mass equivalents”. The odor concentration can be assessed only at a

24

E. Psillakis

presented concentration of 1 ouE/m3. As a consequence the odor concentration is expressed as a multiple of one ouEin a cubic meter of neutral gas. The odor concentration, in ouE/m3, can be used in the same manner as mass concentrations (kg/m3) (EN 13725:2003). The stimulus in this case can be a multitude of substances. In that sense the odor unit is very similar to the LD50, as used in toxicology assessments, indicating the dose that causes a lethal effect in 50% of a well-defined test population. The physiological reaction is the unifying reaction that can be caused by a wide range of substances, at an equally wide range of dosages. In odor research the D50 could be described as the dose that 50% of a population that can detect as a sensory stimulus (EN 13725:2003).

4. OVERVIEW OF METHODS There are a number of different methodologies in use for odor analysis. Selection of a particular method will depend upon the purpose of the measurement, the frequency of monitoring, sampling location, type of source emission (e.g., point source), as well as the nature and complexity of the emission. In general, odor can be “measured” using chemical (analytical) or sensory methods. It is generally accepted that these categories do not have clear cutoff points and some assessment methodologies could be considered to fall into more than one. The analytical techniques include chemical analysis and direct reading instrumental analysis. Chemical analysis is the indirect assessment involving the collection of a sample which, when analyzed, will give the concentration of the various chemical species present. This includes substance-specific wet chemistry methods, as well as sample collection followed by analysis by means of instruments such as gas chromatography (GC). Direct reading instrumental analysis provides information on the concentration of specific chemical species or their concentrations relative to each other. This includes among others portable analyzers (such as portable GC), the “electronic nose,” as well as colorimetric tubes. The sensory methods on the other hand relate to human response giving an assessment of the physiological response to a particular mixture—strength, quality, characteristics—which provides information on the likely population response. This is obtained by exposing trained individuals to samples of the odorous air either in the laboratory or in the field.

5. CHEMICAL ANALYSIS 5.1. Substance-Specific Wet Chemistry Techniques Substance-specific wet chemistry techniques are usually undertaken at the source and are considered to be recognized methodologies for assessing compliance with emission limits. These include:

Odor Measurement ●

●

25

Determination of ammonia in air samples by drawing sample gas through a sulfuric acid solution. The ammonia concentration is determined by titration after Kjeldahl distillation. Determination of hydrogen sulfide in air samples by drawing sample gas through an ammoniacal cadmium chloride solution. The hydrogen sulfide concentration is determined by an iodiometric titration of cadmium sulfide.

The detection limits achieved by these techniques are determined by the volume of gas sampled. Typically limits of detection vary between 0.1 and 0.5 mg/ m3. In order to achieve low detection limits, large volumes of gas need to be sampled; but the sampling flow rate through typical sampling equipment is limited to about 2 L/min (IPPC H4-Part 2, 2002). Consequently, long sampling times are required and peaks in concentration will be missed. In addition, the equipment is fragile and set up can be time consuming. 5.2. Gas Chromatography Gas chromatography (GC) is a widely used analytical technique for the separation, identification, and quantification of the components of an odorous air sample. The use of GC allows chromatographic separation of gaseous and liquid mixtures into individual components. In general, when a sample mixture is introduced into the heated injection port of a GC, it vaporizes. The gaseous sample mixes with an inert gas (such as helium) also referred to as carrier gas and passes through a glass or metal tube (column) that contains an absorbent. Because the various components of the sample interact with the absorbent (stationary phase) of column to different degrees, compounds will be released from the tube at different specific times. There is a wide selection of column types from a range of different manufacturers. Nonetheless, in the case of odorous samples, the choice of a column can be difficult as there are several options or combinations necessary to obtain the optimum information about an odor’s components. Eventually the components of the injected sample are separated and exit the column at different times (called “retention times”). These “elution” times are compared to those of known compounds (analytical standards), thus allowing to some degree identification. Once the compounds are separated, they elute from the column and then enter a detector. The detector is capable of creating an electronic signal whenever the presence of a compound is detected, and in most cases this signal is linearly related to the concentration of the target analyte in the sample. A variety of detectors are currently available, including mass spectrometer, flame ionization, and photoionization detectors (Hobbs, 2001). Despite the fact that gas chromatographic analyses result in accurate and reproducible measurements, however, there are some major limitations. First, the chemical concentrations corresponding to the odor detection thresholds

26

E. Psillakis

cannot be determined due to the synergistic olfactory effects of stimuli comprising complex mixtures of gases (Gostelow and Parsons, 2000). In addition, no indication is obtained as to the relevance of individual compounds to the odor of the sample as a whole (Hobbs et al., 1995). Even if individual chemical concentrations and their odor threshold values are known, it is not possible to deduce the overall sample odor threshold or the odor character of the mixture of odorants. In addition, direct calibration for analyzing odors can be challenging because the composition mixture often will be unknown (IPPC H4-Part 2, 2002). Furthermore, in many cases identification remains ambiguous or questionable as a result of the presence of unknown components at very low concentration level (Bockreis and Jager, 1999). This is because analysis of odorous samples by direct injection is in most cases impossible due to the low concentrations of malodorous compounds in these samples (Gostelow et al., 2001; Pillonel et al., 2002). This problem can be overcome if a preconcentration step prior to analysis is introduced and will be presented in more details in Chapter 3. For all these reasons, the characteristics of complex odors cannot be derived reliably from the individual chemical characteristics and chemical concentrations of the odorous compounds present in a gas mixture. Nonetheless, gas chromatography in general may help a process design engineer to select equipment if the type of odor is unknown and may help researchers understand the mechanisms of odor removal. Longer-term samples will average out any peaks, although this may be of secondary importance in source/compound identification. 5.2.1. Mass Spectrometry Mass spectrometers consist of an ion source, a mass-selective analyzer, and an ion detector and use the difference in mass-to-charge ratio of ionized atoms or molecules to separate them from each other. Given that molecules have distinctive fragmentation patterns, mass spectrometry is capable of providing essential structural information to unknown components. Consequently, the power of this technique lies in the production of mass spectra from each of the analytes detected instead of merely an electronic signal that varies with the amount of analyte. It therefore is a powerful tool for the quantitation of analytes and more importantly for determining chemical and structural information about molecules. Nonetheless, the high cost of instrumentation has limited the wide use of this detector as compared to the less expensive GC detectors (Hobbs, 2001). 5.2.2. Gas Chromatography Coupled to Olfactometry An interesting approach that has been used mainly for the determination of odor-active compounds in food is gas chromatography-olfactometry (GC-O). First introduced by Fuller et al. (van Ruth, 2001), GC-O uses the human nose as a detector and has proven to be a valuable method for monitoring the presence of an odorant in the effluent of a gas chromatograph. Initially, the GC

Odor Measurement

27

effluent was sniffed and when an odor was perceived a description was given for each retention time, thus associating odor activity with eluting compounds. The advantages of coupling GC with olfactometry instead of conventional physical detectors is that although the latter provide relevant information on volatile composition, many of them are not as sensitive for odor-active compounds as the human nose (Ferreira et al., 2003). Furthermore, GC-O allows differentiation between those volatile components with a scent from those that do not have one. Early GC-O devices had serious problems of reproducibility caused by the discomfort from sniffing hot, dry effluent gases. In an attempt to minimize the discomfort of sniffing, Dravnieks and O’Donnell (1971) published a GC-O design in 1971, according to which the hot GC effluent was combined with humidified air to reduce nasal dehydration. In general, it is very difficult to judge the sensory relevance of volatiles from a single GC-O run and the method is limited to the screening of odor-active volatile compounds, unless the chemical stimuli and the assessors’ responses are quantified (van Ruth and O’Connor, 2001). It should be kept in mind that in GC-O, single compounds are assessed, and consequently this method does not provide information on their behavior in a mixture, although it may indicate their relevance to odor. 5.3. Direct Reading Instrumental Analysis: Colorimetric Detector Tubes A colorimetric tube is also known as a “length of stain” tube, due to the fact that the concentration of the chemical being tested produces a color stain in the analytical material, proportional to the concentration in the air. These tubes also are called Dräger tubes®, named after the company that introduced them in 1937 (Drägerwerk AG). Colorimetric detector tubes are flame-sealed glass tubes containing a chemical reagent that reacts with a specific compound or group of compounds, causing color change. A sample is collected by attaching the detector tube to a special bellows-type pump that draws a known volume of air with each stroke. If the target chemical is present, the reagent in the tube changes color and the length of the color change typically indicates the measured concentration. Sampling times are generally short and equipment is portable and relatively inexpensive, hence they can be used as quick indicators of poor control to assess a short-term event or for scoping studies. Because they are simple to use, detector tubes have been used in a broad range of applications. Directreading colorimetric detector tubes also were designed for use in testing workplace air and for determining compliance with occupational exposure limits, but they also are used widely to demonstrate compliance with emission limits for specific substances. Today a considerable number of tubes are available from several manufacturers covering a wide range of substances, although it is not possible to differentiate between different chemical species within a generic group. Despite the simplicity of the method, correct use is essential or results can be highly misleading. Further, more interference with other chemical species is possible,

28

E. Psillakis

leading to false results. It should be mentioned here that colorimetric tubes are not suitable for identifying unknown substances, and the correct tube needs to be selected for a particular situation. Although the limits of detection for this technique are typically between 0.2 to 1 ppm, detection limit can vary greatly, depending on the tube used (IPPC H4-Part 2, 2002). According to the manufacturers’ literature, the relative standard deviation of detector tubes varies between 5 and 20%. Unused tubes must be stored correctly (some require refrigeration) and have a finite shelf life, which must be observed. 5.4. Direct Reading Instrumental Analysis: Portable Analyzers A range of different types of portable instruments suitable for odor measurement are currently available. 5.4.1. Portable Gold Leaf Analyzers Portable gold leaf analyzers are frequently used to monitor hydrogen sulfide in the gas phase. This type of detector utilizes the change in resistance of a gold film sensor caused by adsorption of H2S molecules, within an output proportional to the H2S concentration. Eventually the gold leaf becomes saturated and has to be regenerated. A common gold film monitor is the Jerome 631-X H2S analyzer (Able Instruments and Controls Ltd.), which has a reported detection limit of the order of 0.003 ppm and can measure up to 0.005 ppm H2S. Sampling and measurement time depend on the level of sulfides present, but typically it is less than 1 min. 5.4.2. Paper Tape Monitors Paper tape monitors contain a chemically impregnated tape, which when exposed to a gas sample changes color in direct proportion to the amount of gas present. A tape is selected that will react with the gas of interest (IPPC H4-Part 2, 2002). Today a wide range of compounds can be quantified depending on the instrument selected. A particular advantage is that the monitor can be set to sample at regular intervals, exposing an unreacted section of the tape each time, and so leaves a permanent record of the concentration of each sample. However, given that the sampling times can be of the order of minutes rather than seconds, this type of monitor is not recommended when a large number of samples are required in a short period of time. Furthermore, the equipment is generally expensive, and although relatively simple to operate, sufficient information relating to process parameters/activities at the time of measurement must be collected, thus allowing correct interpretation of the concentration data obtained. Some instruments are affected by the presence of moisture, which limits their use for stack monitoring.

Odor Measurement

29

5.4.3. Portable Gas Chromatophy Portable GCs also exist and can be used for “fingerprinting,” i.e., to analyze air samples at the complainant’s location in order to ascertain the identity and concentration of the main odorous components (Santos and Galceran, 2003). However, the cost of the instrument and the expertise required for analysis and subsequent evaluation also limit its use as a “quick check” method for everyday use. Portable GCs can be coupled to a variety of detectors (including MS, FID, PID, etc.) (Santos and Galceran, 2002). It should be mentioned here that these detectors may be used independently as portable instruments for detecting different groups of compounds in air samples. 5.4.4. Flame Ionization Detector Flame ionization detector (FID) is perhaps the most commonly used detector for odorants, yielding analytical methodologies with a large dynamic range and limits of detection in the low nanogram range (Hobbs, 2001). The FID detector is suitable for odorants mostly composed of hydrogen and carbon, and is not recommended for sulfides, where sensitivity was found to be less satisfactory. Inside the FID, fuel (H2) and oxidant (O2 in air) are mixed to create and maintain a flame. The gaseous samples entering are burned and charged particles are formed in that combustion process. This creates a current between the detector’s electrodes, which is measured by the FID. 5.4.5. Photoionization Detector This device uses ultraviolet light as a means of ionizing an analyte. The ions produced by this process are collected by electrodes and the current generated therefore is a measure of the analyte concentration. PID detectors are used to analyze volatile organic compounds as well as ammonia and hydrogen sulfide. Given that the PID responds to compounds with a photoionization potential equal to or less than that of the energy of the source, appropriate setting of the lamp voltage will exclude direct detection of the major compounds in the air including water vapor, methane, and carbon dioxide (Hobbs et al., 1995). PID detectors have a dynamic detecting range of between a few ppb and 10,000 ppm and offer instant indicative measurements from typically rugged, lightweight, handheld instrument. Given that only a small fraction of the analyte molecules are actually ionized in the PID chamber, it is considered to be a nondestructive GC detector. Therefore, when the inlet of the PID detector is connected to a portable GC, its exhaust port can be connected to another detector in series such as an FID or an electron capture detector (ECD). A schematic diagram of a portable GC equipped with in-series detectors [PID, FID, and dry electrolytic conductivity

30

E. Psillakis

Figure 3. Schematic of PID, FID, and dry electrolytic conductivity detector (DELCD) in series with a SRI-8610C portable gas chromatograph. (With permission from Koziel et al., 1999).

(DELCD)] is given in Figure 3 (Koziel et al., 1999). In this way data from different detectors sensitive to different group of analytes can be taken simultaneously providing almost real-time analysis and speciation of a wide range of compounds. It should be mentioned here that the major challenge in this instrumental setup is to make the design of the ionization chamber and the downstream connections to the second detector as low volume as possible so that peaks that have been separated by the GC column do not broaden out before detection. 5.4.6. Fourier Transform Infrared Spectrometry Fourier transform infrared spectrometry (FTIR) is suited to both quantitative and qualitative measurement. In general, infrared (IR) absorption spectroscopy is the measurement of the wavelength and intensity of the absorption of mid-infrared light by a sample. Mid-infrared light (4000–200 cm−1) is energetic enough to excite molecular vibrations to higher energy levels. The wavelength of many IR absorption bands is characteristic of specific types of chemical bonds. Technically, FTIR is well suited to serve as an “optic” nose (van Kempen et al., 2002). FTIR instruments are portable and can be set up to monitor either stack emissions or boundary emissions. In addition, these instruments are adequately sensitive to pick up odorants of interest (Ni and Heber, 2001). Part of the development of these technologies for food industry odor purposes is the identification of which wavelengths (compounds) meaningfully contribute to the instrument’s response.

Odor Measurement

31

6. SENSORY METHODS Sensory measurements employ the human nose as the odor detector. In this sense, measurements are directly related to the properties of odors as experienced by humans and on the relationship between psychological and physical attributes of odor. The problems of complex mixtures, interactions between components, and detectability below the threshold of smell become irrelevant as the “total effect” of the overall odor is measured. There are many factors other than the properties of the odor sample itself that may influence the perception of an odor. Principal among these is the variability in the sense of smell between different observers. This may be overcome to a certain extent by using a panel of several observers, with the result being expressed as a measure of the central tendency of the individual results. For repeatable results, great care must be taken in the presentation of samples to observers. Factors such as the order in which samples are presented, the environment in which the testing takes place, and the flow rate of the carrying gas stream are all important. Sensory evaluation techniques can be divided into two categories (Gostelow et al., 2001): ●

●

Subjective measurements in which the nose is used without any other equipment. Objective measurements that incorporate the nose in conjunction with some form of dilution apparatus (olfactometry).

6.1. Subjective Measurements Subjective evaluation measurements have the advantage of being quick and relatively low cost, given that no special equipment is required. Interpretation of results is difficult and subjective measurements should be handled with caution due to the inherent variation in odor perception even for well-trained personnel. Parameters that may be measured subjectively include odor character, hedonic tone, and intensity. Indeed, for character and hedonic tone, there are no objective techniques available with the possible exception of the electronic nose (Laing et al., 1992; Gostelow et al., 2001). An example of subjective evaluation measurements is the direct scaling technique, which is actually the oldest technique for measuring sensory stimuli. According to this method, the odor panelist assigns a number to the odor-containing sample, relative to the referenced standard. For example, an odor sample might be compared to a sample of butanol, a typical reference standard. This provides a good comparison of odor intensity, but there is no measure of irritation or how objectionable the odor is (Wise et al., 2000).

32

E. Psillakis

6.2. Objective Measurements Objective measurements incorporate the nose in conjunction with an instrument that dilutes the odor sample with odor-free air, usually termed an olfactometer. In general, dilution may be static or dynamic. Static dilution involves the mixing of fixed volumes of odorous and odor-free air, whereas dynamic dilution involves the mixing of known flows. An example of static dilution is the syringe dilution method (ASTM Method D1391-78), which is also the oldest method of vapor dilution. According to this method, a measured volume of odorous air is transferred from a graduated hypodermic syringe to a dilution syringe. Odor-free air is then drawn into the dilution syringe to make the required dilution. The diluted mixture is then expelled into the observer’s nose. Dilutions to threshold are determined by varying the dilution ratios. This method was withdrawn as an official ASTM method in 1986. Dynamic dilution is superior to static dilution as the effects of sample adsorption to the internal surfaces of the instrument are minimized (Dravnieks and Jarke, 1980). An additional advantage of dynamic dilution/olfactometers is that the sample can be delivered to the sniffing port at a constant flow, a factor that has been shown to improve repeatability of results (Schulz and van Harreveld, 1996). 6.3. Dynamic Olfactometry There are two categories of dilution-related measurement techniques. The most common is threshold olfactometry, where the sample is successively diluted until it can just be detected (i.e., the threshold concentration). The concentration is then expressed as the number of dilutions required to achieve the threshold concentration. Another form of dilution-related measurement is when the sample odor is compared to a reference odor and the result is expressed as an equivalent concentration of the reference gas (also termed suprathreshold olfactometry). The sample or reference odor is diluted until the perceived intensity of each stream is the same. In both cases, the use of an olfactometer removes (or at least, reduces) any subjectivity from the measurement. In general, there are two modes in common use for observers to indicate whether an odor can be detected at a particular dilution and these are the yes/no and the forced choice methods (Sneath, 2001). In the yes/no mode, either neutral gas or diluted odor passes from the single port and the panelist is asked to respond yes or no as to whether they can detect an odor. The panelists are aware that in some cases blanks will be presented and the samples may be presented either randomly or in order of increasing concentration. The other method of dynamic olfactometry is the forced-choice method. In this system, there are two or more sniffing ports, one of which delivers the diluted odorous sample and the other(s) odor-free air. The port carrying the odorous sample is chosen randomly during a test. The panelist is “forced” to choose which port contains the odor

Odor Measurement

33

(hence, the name forced-choice method), using a keyboard. They also indicate whether their choice was a guess, whether they had an inkling, or whether they were certain they chose the correct port. Only when the correct port is chosen and the panelist is certain, the choice is considered to be True. 6.4. Factors Affecting Olfactometry Panel Performance There are a number of variables involved in olfactometry that will affect measurements. Fatigue from continued exposure to an odor may affect a panelist’s sense of smell. This phenomenon is called adaptation and may reduce both perceived odor intensity and perceived odor quality (Dravnieks and Jarke, 1980; Gostelow et al., 2001). The degree of adaptation will depend on the odor concentration experienced. In general, the weaker the odor concentration of an air sample, the more does adaptation affect perceived strength. On the other hand, exposure of panelists to strong odors may result in adaptation and affect detection of subsequent weak odors. Consequently, the presentation schedule of odorous samples may influence the results. It should be mentioned that in cases where dilutions occur in a strict order, panelists will begin to expect subsequent samples to be weaker or stronger and may adjust their responses accordingly. Descending order of presentation may result in olfactory fatigue/adaptation and may obstruct detection of a weak odor after exposure to a strong odor. In general, presentation of one dilution series is given either in an ascending or random order of stimuli, thus restraining adaptation (Dravnieks and Jarke, 1980; Gostelow et al., 2001). Another beneficial effect of using an ascending schedule of sample presentation is that it also may minimize the effects of adsorption/desorption of materials commonly leading to sample contamination or alteration. In general, the use of nonreactive, odor-free materials, minimization of internal surface areas, and the provision for flushing or easy replacement of flow lines between samples further reduces problems of adsorption/desorption. It should be mentioned that the supply of odor-free air is important as any residual odor in the dilution air would bias the result. The general requirements for the environment for observations by assessors can be found elsewhere (EN 13725: 2003). When dynamic olfactometry is used, the flow rates at the sniffing ports can have a major influence on the reported odor concentration. It should be pointed out that the airflow is one of the most controversial issues in the effort to reach international consensus on the standardization of odor measurement. Other important elements of the test procedure that may influence results (EN 13725:2003) are the shape/dimensions of the of smelling chamber (olfactometer–nose interface), position of valves of gauge readings on the olfactometer, differences in appearance of sniffing ports where more than one is used, or even the responses of the operator. It therefore is beneficial if panelists are isolated from each other, from the olfactometer controls, and from the operator as shown in Figure 4.

34

E. Psillakis

Figure 4. Representation of a dynamic dilution olfactometer using the ODILE™ of the ODOTEC Inc. (Montreal, Canada).

Sensitivity to odors is variable between different individuals (Bliss et al., 1996). Olfactory responses of individuals vary with age. Increasing age is correlated with decreasing acuity in odor perception. It is generally accepted that only persons between 16 and 60 years of age with a normal sense of smell should be included on an odor panel. Female panelists normally have a greater sensitivity than male panelists from the same age group. It also has been reported that smokers have less sensitivity than non-smokers. Factors such as health (e.g., cold, nasal allergy), personality, education background, and training also may contribute in some degree to the ability to assess an odor. It is expected that different observers will report different odor concentrations for the same sample. This effect is minimized by the use of a panel of several observers and recording the average response. However, a large panel increases the overall cost of measurement and the time taken to determine sample concentrations. Typically, between four (minimum according to EN 13725:2003) and ten observers (Brennan, 1993) are used with eight being common (Bliss et al., 1996). 6.5. Portable Sensory Devices An example of a portable device for field sensory measurements that uses the dilution to threshold approach is the scentometer, originally manufactured by Barnebey-Cheney Company; it consists of a simple, handheld odor dilution device used to measure odor concentration in the field. According to this technique, the person taking measurements breathes through the scentometer. Gases can pass either directly to the nose or pass through an activated carbon filter. The analyst chooses dilution factor by selecting the size of the

Odor Measurement

35

hole passing unfiltered air. The advantages of scentometer include portability, simplicity of use, and ability to give immediate values for odor concentration and intensity. However, it is difficult to use without experiencing odor fatigue with increased exposure to samples. The scentometer method has been superseded largely by dynamic dilution/olfactometry, although it sometimes may be used in field studies as it is specifically intended for this purpose (Ritter, 1989; Sweeten and Miner, 1993). A more recent device, based on the design of a scentometer, is the nasal ranger (St. Croix Sensory, Inc. of Lake Elmo, Minnesota), combining the portability and low cost of a scentometer with the sampling control of more expensive laboratory olfactometers. This device relies on a pressure transducer to ensure that assessors maintain the required inhalation rate for the unit (Newby and McGinley, 2003; Sheffield et al., 2004). Field measurements have some drawbacks, the most important of which is that results may be biased due to the fact that the assessor may anticipate the smell or become desensitized to the odorous air before taking the measurement. These methods also are difficult to verify when it involves only one person. Some researchers have improved the accuracy of scentometry by using respirators to avoid odor desensitization as well as by increasing the number of assessors, thus averaging several measurements for a single observation (Sweeten and Miner, 1993). 6.6. Standardization of Olfactometry In the absence of standardized procedures, reported odor concentration levels simply reflect the experience of the operator, the design of the olfactometer, its operational mode (manual or automatic), its mixing method, the flow rate presented to panelists, and the number of panelists employed. Recognition of these problems has led to the development of standards for olfactometry. Although olfactometry has been used or researched since the 1890s, it was not until 1970 that olfactometry was used for the detection of odorants in environmental samples. The direct consequence was its use in regulatory issues. The first standards providing a general definition of the method and process parameters were the VDI 3881 (odor threshold determination, Germany, 1980) and the ASTM E679–91 (determination of odor and taste threshold by a forced-choice ascending concentration series method of limits, USA, 1991). The main problem with these standards is that they failed to define a required level of accuracy of the results. As a result the differences between laboratories remained considerable and olfactometry was treated as somewhat of a black art, escaping the rigor and discipline of statistics and experimental design (Table 3). The leaders in the rigorous standardization of olfactometry were the Dutch in the 1980s, who introduced the Dutch olfactometry standard

36

E. Psillakis

Table 3. Comparison of the Basic Process Parameters of Different Standard Methods ASTM E679-91 Dynamic dilution Triangular forced-choice method Binary forced-choice method Yes/no method Ascending concentration series Integral odor free dilution air Integral dry dilution air Presentation flow rate (lpm) Presentation face velocity (m/s) Presentation mask diameter (cm)

A&WMA EE-6

●

●

●

●

NVN 2820

VDI 3881

EN13725

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

20 0.09–0.26 4–7

20

20 0.2–0.5 3–5

5–10 0.02–0.05 6–10

NVN2820 (Provisional standard: Air quality. Sensory odor measurement using an olfactometer, Netherlands, 1996). Driven by the need to introduce a strongly quantitative odor regulatory they were the first to work toward the standardization of olfactometry which would improve the consistency within a laboratory (repeatability) as well as between laboratories (reproducibility) and the closeness of agreement of results to a known reference (accuracy) (Schulz and van Harreveld, 1996). This standard was subsequently used as the basis for a European standard for olfactometry EN13725 (air quality-determination of odor concentration by dynamic olfactometry, Europe, 2003, draft on 1999). The European standard has become the official olfactometry odor analysis approach in many countries and dramatically reduced the differences within and between laboratories that previously existed. The Air and Waste Management Association’s EE-6 Odor Committee (Subcommittee on the Standardization of Odor Measurement prepared a document titled “Guidelines for Odor Sampling and Measurement by Dynamic Dilution Olfactometry,” USA, 2002) has forwarded its guidelines to the American Society of Testing Materials (ASTM) as a suggested replacement for ASTM Method E679–91. These guidelines are similar to the European Standard but they do allow quite a bit of flexibility in what olfactometer flow rates can be used. This could potentially be a problem when attempting to compare data and results from different olfactometry laboratories. As can be seen, there clearly is a strong need for compatibility between the odor measurement methods. It is clear that performance-based standardization is far easier to implement and practice than equipment-based approaches. The widespread acceptance of forced-choice dynamic olfactometry as a preferred measurement method is encouraging, although a practical internationally accepted standard for olfactometry will not occur until testing results are anchored to a standard reference material.

Odor Measurement

37

7. ACRONYMS AND ABBREVIATIONS A D50 D/T DELCD EROM ECD FID FTIR GC GC-O LD50 MS OC OER ouE PID ppb ppm ppt Q SOER TON VOC

Area 50% of a population that can detect as a sensory stimulus Dilution to threshold ratio Dry electrolytic conductivity European reference odor mass EROM European reference odor mass Flame ionization detector Fourier transform infrared spectrometry Gas chromatography Gas chromatography-olfactometry Lethal dose for 50% of a well-defined test population Mass spectometer Odor concentration Odor emission rate European odor unit Photoionization detector Parts per billion Parts per million Parts per trillion Gas flow or ventilation rate Specific odor emission rate Threshold odor number Volatile organic compound

8. REFERENCES Abbott, J. (1993) “Enclosed wastewater treatment plants-health and safety considerations.” Foundation for Water Research Report FR/W0001. ASTM Method D1391-78 “Standard method for measurement of odor in atmospheres (dilution method).” American Society for Testing and Materials (ASTM International) Standards, withdrawn 1986. Bockreis, A., Jager, J. (1999) “Odor monitoring by the combination of sensors and neural networks.” Environmental Modeling and Software, 14, 421–426. Bonnin, C., Laborie, A., Paillard, H. (1990) “Odor nuisances created by sludge treatment: problems and solutions.” Water Science and Technology, 22, 65–74. Bliss, P. J., Schulz, T. J., Senger, T., Kaye, R. B. (1996) “Odor measurement—factors affecting olfactometry panel performance.” Water Science and Technology, 34, 549–556. Bliss, P.J., Jiang, J.K., Schultz, T.J. (1995) “The development of a sampling technique for determination of odor emission rate from areal surfaces: II. Mathematical Model.” Journal of Air and Waste Management Association. 45, 989–994. Brennan, B. (1993) “Odor nuisance.” Water and Waste Treatment, 36, 30–33. Dravnieks, A., Jarke, F. (1980) “Odor threshold measurement by dynamic olfactometry: significant operational variables.” Journal of Air Pollution Control Association, 30, 1284–1289.

38

E. Psillakis

Dravnieks, A., O’Donnell, A. (1971) “Principles and some techniques of high-resolution headspace analysis.” Journal of Agricultural and Food Chemistry, 19, 1049–1056. EN 13725(2003). Air Quality-Determination of Odor Concentration by Dynamic Olfactometry. CEN, Brussels. Ferreira, V., Pet’ka, J., Aznar, M., Cacho, J. (2003) “Quantitative gas chromatography–olfactometry. Analytical characteristics of a panel of judges using a simple quantitative scale as gas chromatography detector.” Journal of Chromatography A, 1002, 169–178. Gholson, A.R., Albritton, J.R., Jayanty, R.K.M., Knoll, J.E., Midgett, M.R. (1991) “Evaluation of the flux chamber method for measuring volatile organic emissions from quiescent liquid surfaces.” Environmental Science and Technology, 21, 519–524. Gostelow, P., Parsons, S.A. (2000) “Sewage treatment works odor measurements.” Water Science and Technology, 41(6), 33–40. Gostelow, P., Parsons, S.A., Stuetz, R.M. (2001) “Odor measurements for sewage treatment works.” Water Research, 35, 579–597. Hobbs, P.J., Misselbrook, T.H., Pain, B.F. (1995) “Assessment of odors from livestock wastes by a photoionisation detector, an electronic nose, olfactometry and gas-chromatography-mass spectrometry.” Journal of Agricultural Environmental Research, 60, 137–144. Hobbs, P. (2001) “Odor analysis by gas chromatography.” In. Stuetz, R.M., Frechen, F.-B. (eds), Odours in Wastewater Treatment: Measurement, Modeling and Control. London: IWA publishing. Hwang, S. T. (1985) “Model prediction of volatile emissions: A comparison of several models for predicting emissions for hazardous waste treatment facilities.” Environmental Progress, 4, 141–144. Jiang, J.K., Bliss, P.J., Schultz, T.J. (1995) “The development of a sampling technique for determination of odor emission rate from areal surfaces: I. Aerodynamic performance.” Journal of Air and Waste Management Association, 45, 917–922. Jiang, K., Kaye, R. (1996) “Comparison study on portable wind tunnel system and isolation chamber for determination of VOCs from areal sources.” Water Science and Technology, 34, 583–589. Jiang, J., Kaye, R. (2001) “Sampling techniques for odor measurement.” In Stuetz, R.M., Frechen, F.-B. (eds), Odours in Wastewater Treatment: Measurement, Modeling and Control. London: IWA publishing. Klenbusch, M. R. (1986) “Measurement of gaseous emission rates from land surfaces using an emission isolation flux chamber, user’s guide.” EPA/600/8-86/008; US Environmental Protection Agency. Koziel, J., Jia, M., Khaled, A., Noah, J., Pawliszyn, J. (1999) “Field air analysis with SPME device.” Analytica Chimica Acta, 400, 153–162. IPPC H4-Part2 (2002) Integrated Pollution Prevention and Control (IPPC) Draft: “Horizontal guidance for odor: Part 2—Assessment and control,” Environment Agency. http://www.sepa.org.uk/ pdf/ppc/uktech/ippc_h4_pt2.pdf [2005 Jan 5] Laing, D.G. Doty, R.L., Breipohl, W. (1992) The Human Sense of Smell. Springer-Verlag Berlin Heidelberg. Lindvall. T. (1970) “On sensory evaluation of odorous air pollutant intensities.” Nordisk Hygiejnisk Tidsskrift, suppl. 2, 1–181. Lockyer, D.R (1984) “A system for the measurement in the field of losses of ammonia through volatilization.” Journal of the Science of Food and Agriculture, 35, 837–848. Newby, B.D., McGinley, M.A. (2004) “Ambient odor testing of concentrated animal feeding operations using field and laboratory olfactometers.” Water Science and Technology, 50(4), 109–114. Ni, J. Q., Heber, A. J. (2001) “Sampling and measurement of ammonia concentration at animal facilities— A review.” Proceedings of the American Society of Agricultural Engineers Annual International Meeting, Sacramento, California (Paper Number: 01–4090). Pillonel, L., Bosset, J.O., Tabacchi, R. (2002) “Rapid preconcentration and enrichment techniques for the analysis of food volatile. A review.” Lebensmittel-Wissenschaft und-Technologie, 35(1), 1–14. Ritter, W.F. (1989) “Odor control of livestock wastes: State-of-the-art in North America.” Journal of Agricultural Engineering Research, 42, 51–62.

Odor Measurement

39

Santos, F. J., Galceran, M. T. (2002) “The application of gas chromatography to environmental analysis.” Trends in Analytical Chemistry, 21, 672–685. Santos, F. J., Galceran, M. T. (2003) “Modern developments in gas chromatography–mass spectrometry-based environmental analysis.” Journal of Chromatography A, 1000, 125–151. Sarkar, U., Hobbs, S.E. (2002) “Odor from municipal solid waste (MSW) landfills: A study on the analysis of perception.” Environment International, 27, 655–662. Schulz, T. J., van Harreveld, A. P. (1996) “International moves towards standardization of odor measurement using olfactometry.” Water Science and Technology, 34, 541–547. Sheffield, R., Thompson, M., Dye, B., Parker D. (2004) “Evaluation of field-based odor assessment methods.” Water Environment Federation/A&WMA Odors and Air Emissions Conference. http://www.nasalranger.com/Media.cfm [2005 Jan 5] Sneath, R.W. (2001) “Olfactometry and the CEN standard.” In. Stuetz, R.M., Frechen, F.-B. (eds), Odours in wastewater Treatment: Measurement, Modeling and Control. London: IWA Publishing. Sweeten, J.M., Miner, J.R., (1993) “Odor intensities at cattle feedlots in nuisance litigation.” Bioresource Technology, 45, 177–188. van Kempen, T.A.T.G., Powers, W.J., Sutton, A.L. (2002) “Fourier transform infrared (FTIR) spectroscopy as an optical nose for predicting odor sensation.” Journal of Animal Science, 80, 1524–1527. van Ruth, S.M., O’Connor, C.H. (2001) “Evaluation of three gas chromatography-olfactometry methods: comparison of odor intensity-concentration relationships of eight volatile compounds with sensory headspace data.” Food Chemistry, 74, 341–347. van Ruth, S.M. (2001) “Methods for gas chromatography-olfactometry: a review.” Biomolecular Engineering, 17, 121–128. Wise, P.M, Olsson, M.J., Cain, W.S. (2000) “Quantification of odor quality.” Chemical Senses, 25, 429–443. US EPA Method 0030 (1986) “Volatile organic sampling train.” Test Methods for Evaluating Solid Waste, Physical/Chemical Methods (SW-846) Young, P. J. (1984) “Odours from effluent and waste treatment.” Effluent Water Treatment Journal, 24, 189–195.

3 Preconcentration Prior to Gas Chromatography Elefteria Psillakis

1. PRECONCENTRATION TECHNIQUES There are several obstacles impeding complete characterization of odorous samples when using gas chromatography (GC). In many cases identification remains ambiguous or questionable as a result of the presence of unknown components at very low concentration levels. This is due to the fact that direct injection of odorous samples into a GC is in most cases impossible due to the low concentrations of malodorous compounds in these samples. This problem can be overcome if a preconcentration step prior to analysis is introduced. In general there are three possibilities for enriching components in an air sample: absorbing the compounds in a suitable liquid, condensing them at low temperatures (cryotrapping), and adsorbing them on a porous solid material. In the case of liquid absorption, a liquid solution is used to absorb or react with the target volatile compounds. An aliquot of the solvent solution containing the entrapped odorous compounds is then injected into the analytical device for further analyses. There are many drawbacks inherent to liquid absorption. A solvent evaporation step is usually required, leading to large losses of the volatiles (Pillonel et al., 2002). Furthermore, sample contamination due to contact with glassware is possible; therefore, a thorough cleaning of all laboratory equipment coming into direct contact with the sample is required. In cases where impingers are used, the liquid containing the entrapped analytes may be subject to spillage, and in these cases, a liquid trap should be used in order to prevent the solution from getting into the pump. Finally, during analysis, interferences due to solvent impurities also may obstruct quantification. It also should be mentioned here that it is generally accepted that it is easier to precondition a solid phase than to purify a liquid phase.

ELEFTERIA PSILLAKIS ● Department of Environmental Engineering, Technical University of Crete, Polytechnioupolis, GR-73100 Chania, Greece e-mail: [email protected] 41

42

E. Psillakis

In cryogenic trapping, volatile compounds are trapped on an inert surface, such as glass-fiber wool, glass beads, Tenax®, Porapak Q®, or even activated carbon. According to this method, during sampling the trap is most frequently immersed in liquid nitrogen (−196°C) or in liquid argon (−186°C) (Wardencki, 1998). The main advantages of cryogenic trapping versus other trapping techniques are that there are (1) no artifact from thermal desorption, (2) no carryover between runs, and (3) no breakthrough problem (Pillonel et al., 2002). However, the additional equipment needed to handle the cryogen is quite expensive and very sensitive to water. Furthermore, there are another two concerns when using this preconcentration method. First, if the cooling temperature is too low, liquid oxygen also may be trapped, which can readily oxidize organic compounds, thus altering the composition of the sample. To overcome this problem, Peltiercooling devices can be used providing temperature control within the required range for primary and secondary refocusing or concentrating a sample (Hobbs, 2001). Second, when samples are too large, then moisture can condense, forming ice and blocking/restricting the flow of the compounds trapped in the sample. To prevent clogging of the trap, water has to be removed efficiently from the charged carrier gas before entering the trap (Pillonel et al., 2002). More commonly preconcentration is achieved by sampling odorous compounds on a porous material packed in a cartridge or in short columns. In general, adsorbents have proved a successful and relatively inexpensive means of trapping volatile analytes and the associated sorbent tubes are easy to condition and small in size, facilitating collection, transport, and storage. The most commonly used sorbents were recently reviewed by Harper (2000) and the general types are inorganic adsorbents and porous materials based on carbon and organic polymers. Various adsorbents may be used individually or in combination and their selection depends on the compounds to be sampled (e.g., concentration, species and mixtrure) as well as the boiling point involved. The surface area of an adsorbent also has an impact on the amount of a given substance that can be withheld by the medium as well as the surface polarity. Carbon-based adsorbents with a large area are useful to trap very low-boiling compounds, whereas it gets difficult to desorb substances with higher boiling points. Porous polymers with a comparatively small surface area allow adsorption and desorption of lesser volatile components from gaseous samples. The different characteristics of adsorbents show the need to carefully choose the right adsorption material for given mixtures (Peng and Batterman, 2000). For trapping volatile compounds with very different properties, multibed adsorbents can be helpful. Typical combinations include Tenax TA® or graphitized carbon and carbon molecular sieve (Harper, 2000). The weaker sorbent (Tenax TA®) is placed first to trap the heavier molecules and the lighter compounds are retained on the stronger sorbent located in second position. Desorption always takes place in the reverse direction to the adsorption step (Pillonel et al., 2002). Additional sorbent beds may be placed behind the primary bed to guard against breakthrough. Breakthrough is the appearance of sampled molecules in

Preconcentration Prior to Gas Chromatography

43

the outlet stream, either because of saturation within the bed or displacement by another chemical. Sampling is no longer efficient when this occurs, and as breakthrough progresses the sample will be less and less representative of the external environment (Harper, 2000). Another concern when using sorbent tubes is the presence of moisture which inhibits and prejudices the profile of adsorbed species. This can be overcome by using hydrophobic adsorbents, dry purges, and other techniques (Peng and Batterman, 2000). Once target compounds are collected on adsorbent, samples can be transferred to GC for further analysis. This is accomplished by either thermal or liquid desorption. There are several reports where the ability of absorbents followed by solvent or thermal desorption and chromatographic analysis to concentrate volatile organic compounds (VOCs) from air samples has been demonstrated (Hobbs et al., 1995; Rabaud et al., 2003; Peng and Batterman, 2000). Thermal desorption is more popular than solvent extraction due to its speed, ease of desorption, and minimization of artifact formation. It also offers higher sensitivity given that the sample is not diluted as well as higher recoveries for polar and reactive compounds which can pose problems for whole air samples. Nonetheless, sample degradation due to contact with heated transfer line from the desorber to the GC inlet is possible and carryover from previous samples can be minimized or even avoided using short-path desorber systems. Overall, the performance of a sorbent sampling/thermal desorption method depends on many factors, including the target compounds (e.g., concentration, species, and mixture), the method (e.g., sorbent selection, procedures for conditioning, desorption, separation, and analysis of VOCs), and environmental conditions present during sampling (e.g., temperature and humidity) (Peng and Batterman, 2000). A more recent trend in the preconcentration of odorous compounds involves the use of the solid-phase microextraction (SPME) technique. SPME, developed by Arthur and Pawliszyn in 1990, addressed the need for simple and rapid sample preparation. The commercially available SPME unit consists of a short length of narrow diameter fused-silica optical fiber externally coated with a thin film of a polymeric stationary phase located inside a syringelike SPME holder. The coated fiber is exposed to the sample where analytes preferentially partition by adsorption or absorption (depending on the fiber type) from the sample to the stationary phase and are concentrated. There are two main types of SPME sampling: immersion sampling where the fiber is immersed into the aqueous solution and headspace sampling where the fiber is exposed to the headspace above the liquid (or solid) sample (Figure 1). Immersion sampling is widespread in the SPME approach, but for volatile compounds the headspace mode is preferred as it results in faster equilibration times and higher selectivity. After sampling for a well-defined period of time, the fiber is withdrawn and transferred to the heated injection port of a gas chromatograph (GC) or to a modified high-performance liquid chromatography (HPLC) rheodyne valve depending on the target analytes (Eisert and Pawliszyn, 1997).

44

E. Psillakis (a)

(b)

SPME holder

SPME fibre

Sample

Figure 1. Schematic representation of the two SPME sampling modes: (a) headspace and (b) immersion in the case of liquid samples.

SPME rapidly gained wide acceptance incorporating sampling, extraction, concentration, and sample introduction into a single solventless step, improving the method’s detection limits and saving preparation time. Given that the amounts of analyte partitioned to the coating are proportional to their initial concentrations in the matrix, SPME is particularly suited for air sampling, where the effects of partitioning on initial analyte air concentration are negligible. Up to now, SPME has been successfully applied to a wide variety of compounds in the gas, liquid, and solid phase for the extraction of volatile and semivolatile organic compounds from environmental, food, and biological samples (Kataoka et al., 2000; Kataoka, 2002; Psillakis et al., 2003). Recently, the possibility of using SPME as a viable tool to monitor trace quantities of odorous compounds (such as volatile organic compounds and formaldehyde) has been demonstrated (Béné et al., 2001; Davoli et al., 2003; Koziel et al. 1999). It is generally accepted that despite the great technological advances there are some major limitations inherent to gas chromatographic analyses of malodorous air samples. The development of new analytical methodologies presents a major challenge for many research laboratories around the world in that by providing new, simple, and economic means for preconcentating odorous compounds will outpace chromatographic analysis of odorous compounds.

2. ACRONYMS AND ABBREVIATIONS GC HPLC SPME VOCs

Gas chromatography High-performance liquid chromatography Solid-phase microextraction Volatile organic compounds

Preconcentration Prior to Gas Chromatography

45

3. REFERENCES Arthur, C.L., Pawliszyn, J. (1990) “Solid-phase microextraction with thermal desorption using silica optical fibers.” Analytical Chemistry, 62(19), 2145–2148. Béné, A., Fornage, A., Luisier, J.-L., Pichler, P., Villettaz, J.-C. (2001) “A new method for the rapid determination of volatile substances: the SPME-direct method: Part I: Apparatus and working conditions.” Sensors and Actuators B: Chemical, 72(2), 184–187. Davoli, E., Gangai, M.L., Morselli, L., Tonelli D. (2003) “Characterisation of odorants emissions from landfills by SPME and GC/MS.” Chemosphere 51(5), 357–368. Eisert, R., Pawliszyn, J. (1997) “New trends in solid-phase microextraction.” Critical Reviews in Analytical Chemistry 27(2), 103–135. Harper, M. (2000) “Sorbent trapping of volatile organic compounds from air.” Journal of Chromatography A, 885(1–2), 129–151. Hobbs, P., Misselbrook, T. H., Pain, B. F. (1995) “Assessment of odours from livestock wastes by a photoionization detector, an electronic nose, olfactometry and gas chromatography-mass spectrometry.” Journal of Agricultural Engineering Research, 60(2), 137–144. Hobbs, P. (2001) “Odour analysis by gas chromatography.” In: Stuetz, R.M., Frechen F.-B. (eds), Odours in Wastewater Treatment. London: IWA Publishing. Kataoka, H., Lord, H.L., Pawliszyn, J. (2000). “Applications of solid-phase microextraction in food analysis.” Journal of Chromatography A, 880(1–2), 35–62. Kataoka, H. (2002) “Automated sample preparation using in-tube solid-phase microextraction and its application – a review.” Analytical and Bioanalytical Chemistry 373(1–2), 31–45. Koziel, J., Jia, M. Khaled, A., Noah, J., Pawliszyn, J. (1999) “Field air analysis with SPME device.” Analytica Chimica Acta, 400(1–3), 153–162. Peng, C-Y., Batterman, S. (2000) “Performance evaluation of a sorbent tube sampling method using short path thermal desorption of volatile organic compounds.” Journal of Environmental Monitoring, 2(4), 313–324. Pillonel, L., Bosset, J.O., Tabacchi R. (2002) “Rapid preconcentration and enrichment techniques for the analysis of food volatile. A review.” Lebensmittel-Wissenschaft und-Technologie, 35(1), 1–14. Psillakis, E., Mantzavinos, D., Kalogerakis, N. (2003) “Monitoring the sonochemical degradation ofphthalate esters in water using solid-phase microextraction.” Chemosphere, 54(7), 849–857. Rabaud, N.E., Ebeler, S.E., Ashbaugh, L.L., Flocchini, R.G. (2003) “Characterization and quantification of odorous and non-odorous volatile organic compounds near a commercial dairy in California.” Atmospheric Environment, 37(7), 933–940. Wardencki, W. (1998) “Problems with the determination of environmental sulphur compounds by gas chromatography.” Journal of Chromatography A, 793(1), 1–19.

4 The Application of Intelligent Sensor Array for Air Pollution Control in the Food Industry Saverio Mannino, Simona Benedetti, Susanna Buratti, and Maria Stella Cosio

1. INTRODUCTION One of the most significant improvements in the food industry during the next few years is likely to be the development of intelligent sensor array systems that can give useful information not only about the food quality characteristics but also on the environment of the food production. For instance, the sensory array system, often called “electronic nose,” would allow the monitoring of off-odors and taints, often present at very low levels, or the odor quality of a food from the raw material to the final product.

2. HISTORY OF THE INTELLIGENT SENSOR ARRAY The earliest work on the development of an instrument specially to detect odors probably dates back to Moncrieff in 1961. This was really a mechanical nose, but the first electronic nose, based on redox reactions of odorants at an electrode, was reported by Wilkens and Hatman in 1964. In 1965, Buck et al. and Dravieks and Trotter worked on the modulation of conductivity by odorants and modulation of contact potential by odorants, respectively. However, arguably, the concept of an electronic nose as an intelligent chemical array sensor system did not emerge until nearly 20 years later from publications by Persaud and Dodd in 1982 at Warwick University in the United Kingdom. Their purpose SAVERIO MANNINO, SIMONA BENEDETTI, SUSANNA BURATTI, AND MARIA STELLA COSIO ● Department of Food Science, Technology and Microbiology, Univerisity of Milan, Via Celoria 2, 20133 Milan, Italy; e-mail: [email protected]; [email protected]; susanna. [email protected]; [email protected] 47

48

S. Mannino et al.

was to model the current conception of the mammalian olfactory system by demonstrating that a few sensors could discriminate among a larger number of odorants. They constructed an array of three metal oxide gas sensors, which they used to discriminate among 20 odorant substances, including essential oils and pure volatile compounds. In 1980, in North America, a group led by Stetter et al. started to build an analytical portable instrument based on a chemical gas sensor array for the US Coast Guard. This organization needed a portable instrument that would rapidly identify and measure volatile chemical vapors in emergency situations. In Asia, scientists also were beginning to foresee the potential of a sensor array. In 1984, Iwanaga et al. proposed an instrument employing an array of metal oxide semiconductor sensors but suggested using simultaneous equations to compute the relative concentrations of gases in a sample. Then in 1989, a session at a NATO Advanced Workshop on Chemosensory Information Processing was dedicated to the topic of artificial olfaction (Schild, 1990) and the design of an artificial olfactory system was further established (Gardner et al., 1990). Finally, the first conference dedicated to the topic of an electronic nose was held in 1992 (Gardner and Bartlett, 1992). At the beginning of the 1990s, the term “artificial” or “electronic nose” appeared. More extended research began, and applications, especially in food industry, could be tested. Gardner and Bartlett (1993) defined the electronic nose as “an instrument, which comprises an array of electronic chemical sensors with partial specificity and appropriate pattern-recognition system, capable of recognizing simple or complex odors” (p. 212).

3. SENSOR ARRAY COMPARED TO HUMAN NOSE The sensor array is very far from the human nose, and according to Mielle et al. (1995) this analytical system is “obviously electronic but not nose.” In fact the only aspect in common with our odor-sensing organ is its function. There are striking analogies between the human nose and the electronic nose. Comparing the two is instructive. The human nose uses the lungs to bring the odor to the epithelium layer; the electronic nose (E-nose) has a pump. The human nose has mucus, hairs, and membranes to act as filters and concentrators, while the E-nose has an inlet sampling system that provides sample filtration and conditioning to protect the sensors and enhance selectivity. The human epithelium contains the olfactory epithelium, which has millions of sensing cells, selected from 100–200 different genotypes that interact with the odorous molecules in unique ways. The E-nose has a variety of sensors that interact differently with the sample. The human receptors convert the chemical responses to electronic nerve impulses. The unique pattern of nerve impulses is propagated by neurons through a complex network before reaching the higher brain for interpretation. Similarly, the chemical sensors in the E-nose react with the sample and produce

49

The Application of Intelligent Sensor Array

electrical signals. A computer reads the unique pattern of signals and interprets them with some form of “intelligent” pattern classification algorithm.

4. SENSOR ARRAY TECHNOLOGIES Several commercial intelligent gas sensor array instruments are now available on the market that cover a variety of chemical sensor principles, system design, and data analysis techniques. Operationally, an E-nose is a “sensing system” composed of three parts (Fig. 1): a sampling system (a); an array of chemical gas sensors (b) producing an array of signals when confronted with a gas, vapor, or odor; and an appropriate pattern-classification system (c). The ideal sensors to be integrated in an electronic nose should fulfill the following criteria (Bartlett et al., 1993): high sensitivity toward chemical compounds, that is, similar to that of the human nose (down to 10−12 g/ml), low sensitivity toward humidity and temperature; medium selectivity, they must respond to different compounds present in the headspace of the sample; high stability; high reproducibility and reliability; short reaction and recovery time; robust and durable; easy calibration; easily processable data output; and small dimensions. By chemical interaction between odor compounds and the gas sensors the state of the sensors is altered giving rise to electrical signals that are registered by the instrument. In this way the signals from the individual sensors represent a pattern that is unique for the gas mixture measured and is interpreted by multivariate pattern recognition techniques like, for example, the artificial neural network. Samples with similar odors generally give rise to similar sensor response patterns and samples with different odors show differences in their patterns. The sensors of an electronic nose can respond to both odorous and odorless volatile compounds.

SENSOR RESPONSES SINGLE SENSOR CHANGE IN RESISTANCE Figure 1. Basic elements of an electronic nose.

50

S. Mannino et al.

Various kinds of gas sensors are available, but only four technologies are currently used in commercialized electronic noses: metal oxide semiconductors (MOS); metal oxide semiconductor field effect transistors (MOSFET); conducting organic polymers (CP); and piezoelectric crystals [bulk acoustic wave (BAW)]. Others, such as fiberoptic (Dickinson et al., 1996), electrochemical (Baltruschat et al., 1997), and bimetal sensors, are still in developmental stage and may be integrated in the next generation of the electronic noses. In all cases the goal is to create an array of differentially sensitive sensing elements. 4.1. Metal Oxide Semiconductors Metal oxide semiconductor sensors (MOS) were first used commercially in the 1960s as household gas alarms in Japan under the names of Taguchi (the inventor) or Figaro (the company’s name). These sensors rely on changes of conductivity induced by the adsorption of gases and subsequent surface reactions (Kohl, 1992). They consist of a ceramic substrate (round or flat) heated by wire and coated by a metal oxide semiconducting film. The metal oxide coating may be either of the n-type [mainly tin dioxide, zinc oxide, titanium dioxide, or iron (III) oxide], which responds to oxidizing compounds, or of the p-type (mainly cobalt oxide or nickel oxide), which responds to reducing compounds (Mielle, 1996). The film deposition technique divides each sensor type into thin (6–1000 nm) or thick (10–300 µm) film MOS sensors. The first one offers a faster response and significantly higher sensitivity but is much more difficult to manufacture in term of reproducibility. Therefore, commercially available MOS sensors often are based on thick film technologies. Due to the high operating temperature (200– 650 °C), the organic volatiles transferred to the surface of the sensors are totally combusted to carbon dioxide and water, leading to the change in the resistance. MOS sensors are extremely sensitive to ethanol which blinds them to any other volatile compound of interest. 4.2. Metal Oxide Semiconductor Field Effect Transistors Metal oxide semiconductor field effect transistor (MOSFET) sensors rely on a change of electrostatic potential. A MOSFET sensor comprises three layers: a silicon semiconductor, a silicon oxide insulator, and a catalytic metal (usually palladium, platinum, iridium, or rhodium), also called the gate. When polar compounds interact with this metal gate, the electric field, and thus the current flowing through the sensor, are modified. The recorded response corresponds to the change of voltage necessary to keep a constant preset drain current (Lundstrom et al., 1990). The selectivity and sensitivity of MOSFET sensors may be influenced by the operating temperature (50–200°C), the composition of the metal gate, and the microstructure of the catalytic metal. MOSFET sensors have a relatively low sensitivity to moisture and are thought to be very robust.

The Application of Intelligent Sensor Array

51

4.3. Conducting Organic Polymers Conducting organic polymers (CP) sensors like MOS sensors rely on changes of resistance by adsorption of gas. These sensors comprise a substrate (such as a fiberglass or silicon), a pair of gold-plated electrodes, and a conducting organic polymer such as polypyrrol, polyaniline, or polythiophene as a sensing element. The polymer film is deposed by electrochemical deposition between both electrodes previously fixed to a substrate. When a voltage is passed across the electrodes, a current passes through the conducting polymer (Amrani et al., 1995). The addition of volatile compounds to the surface of the sensor alters the electron flow in the system and therefore the resistance of the sensor. In general, CP sensors show good sensitivities especially for polar compounds. However, their low operating temperature (<50 °C) makes them extremely sensitive to moisture. Although such sensors are resistant to poisoning, they have a lifetime of only about 9–18 months. 4.4. Piezoelectric Crystal Sensors Piezoelectric crystal sensors are based on the change of the mass, which may be measured as a change in resonance frequency. These sensors are made of tiny disks, usually quartz, lithium niobate (LiNbO3), or lithium tantalite (LiTaO3), coated with materials such as chromatographic stationary phase, lipids, or any nonvolatile compounds that are chemically and thermally stable (Guilbault and Jordan, 1988). When an alternating electrical potential is applied at room temperature, the crystal vibrates at a very stable frequency, defined by its mechanical properties. Upon exposure to a vapor, the coating adsorbs certain molecules, which increase the mass of the sensing layer, and hence decrease the resonance frequency of the crystal. This change may be monitored and related to the volatile present. The crystal may be made to vibrate in a bulk acoustic wave (BAW) or in a surface acoustic wave (SAW) mode by selecting the appropriate combination of crystal cut and type of electrode configuration. BAW and SAW sensors differ in their structure: BAW are three-dimentional waves traveling through the crystal, while SAW are two-dimentional waves that propagate along the surface of the crystal at a depth of approximately one wavelength. These devices are also called “quartz crystal microbalance” (QCM or QMB) because, similar to a balance, their responses change in proportion to the amount of mass adsorbed. Since piezoelectric sensors may be coated with an unlimited number of materials, they present the best selectivity. However, the coating technology is not yet well controlled, which induces poor batch-to-batch reproducibility.

5. DATA PROCESSING The data processing of the multivariate output data generated by the gas sensor array signals represents another essential part of the electronic nose concept. The statistical techniques used are based on commercial or specially designed

52

S. Mannino et al.

software using pattern recognition routines like principal component analysis (PCA), partial least squares (PLS), linear discriminant analysis (LDA), cluster analysis (CA), fuzzy logic, or artificial neural network (ANN). Principal components analysis (PCA) is a procedure that allows the extraction of useful information from the data and exploration of the data structure, the relationship between objects, the relationship between objects and variables, and the global correlation of the variables. It was used for explorative data analysis as it identifies orthogonal directions of maximum variance in the original data, in decreasing order, and projects the data into a lower-dimensional space formed of a subset of the highest-variance components. The orthogonal directions are linear combinations (principal components PCs) of the original variables and each component explains in turn a part of the total variance of the data; in particular, the first significant component explains the largest percentage of the total variance, the second one, the second largest percentage, and so forth (Beebe et al., 1998). Partial least square (PLS) regression models describe the dependence between two variable blocks, e.g., sensor responses and time variables. Let the X matrix represent the sensor responses and the Y matrix represent time and the X and Y matrices could be approximated to a few orthogonal score vectors, respectively. These components are then rotated in order to get as good a prediction of y variables as possible (Joreskogand and Wold, 1982). Linear discriminant analysis (LDA) is among the most-used classification techniques. The method maximizes the variance between categories and minimizes the variance within categories. This method renders a number of orthogonal linear discriminant functions equal to the number of categories minus one (Meloun et al., 1992). Cluster analysis (CA) performs agglomerative hierarchical clustering of objects based on distance measures of dissimilarity or similarity. The hierarchy of clusters can be represented by a binary tree, called a dendrogram. A final partition, i.e., the cluster assignment of each object, is obtained by cutting the tree at a specified level (Gardner and Bartlett, 1992). The artificial neural networks (ANN) are very sophisticated modeling techniques capable of modeling extremely complex functions (Benedetti et al., 2004). The basic unit of an ANN is the neuron. A neural network consists of a set of interconnected network of neurons (Fig. 2). The input layer has one neuron for each of the sensor signals, while the output layer has one neuron for each of different sample properties that should be predicted. Usually one hidden layer with a variable number of neurons is placed between the input and the output layer. During the ANN training phase, the weights and transfer function parameters are adjusted such that the calculated output value for a set of input value are as close as possible to the unknown true value of the sample properties. The model estimation is more complex than for a linear regression model due to the nonlinearity of the model. Neural networks learn from examples through iteration, without requiring a priori knowledge of the relationship among variables under investigation. The neural network model mostly used was a multilayer perceptron (MLP) that

53

The Application of Intelligent Sensor Array Hidden layer Output layer input e.g. sensor responses

output e.g.conc class

w1

Summation

w2

å

w3

Tranfer-function

Weight multiplication Neuron Figure 2. Architecture of an artificial neural network (ANN).

learns using an algorithm called backpropagation. Since a neural network can arrive at different solutions for the same data, if different values of the initial network weights are provided, the network was trained several times. The goal was to try and find a neural network model for which multiple training approaches the same final mean squared error (MSE).

6. APPLICATIONS OF INTELLIGENT SENSOR ARRAY There is a growing emphasis and consensus that an intelligent sensor array or E-nose is most effective in the quality control of raw and manufactured products, for example, the determination of food freshness and maturity monitoring, shelf-life investigations, authenticity assessments of products, and even microbial pathogen detection and environmental control. Industries including those of cosmetics, cars, packaging, tabacco, environment, and food monitoring all have seen quite heavy investment in recent years. The E-nose has the interesting ability to address analytical problems that have been refractory to traditional analytical approaches: for example, the problem of the volatile compounds from air pollution. The volatile compounds interact in such a way that no single compound or group of compounds is associated with the subjective assessment of odor. The odor of “bad” air cannot be traced to a specific molecule or to a simple list of chemicals by traditional analytical techniques like gas chromatography or mass spectroscopy. In such cases, the determination of odor is performed by a panel of skilled and trained human noses that could find differences in the odor of the samples. The E-nose can learn the fingerprints of air and easily recognize the differences. The molecular basis for this difference is not known exactly, but it is possible to affirm that the pattern responses are different and related to the quality of “good” or “bad” air. The E-noses have been used to assess livestock wastes (Hobbs et al., 1995), the application of cattle slurry to grassland (Misselbrook et al., 1997), airborne pollen levels (Kalman et al., 1997), and jet fuels in soil and water samples (McCarrick et al., 1996). E-noses also are being applied to bioprocess

54

S. Mannino et al.

monitoring. Microrganisms have been employed increasingly in recent years to biodegrade biological waste gases, and because of tightening restrictions on odor emissions the need for reliable gas-monitoring methods has become critical. Researchers in Germany have demonstrated the use of E-nose techniques for monitoring such waste-gas-treatment processes (Homan and Fodisch, in press). Sensor arrays also have been employed in online bioreactor monitoring by observing changes in the volatiles emitted from cultures during the biodegradation process (Bachinger et al., in press). Waste from many sources, such as animal factories and fertilizers, many times is responsible for polluting rivers, streams, and underground water aquifers. The E-nose provides a method for onsite characterization and measuring the fingerprint of polluting samples (Edward, 1999).

7. THE FUTURE There is a plethora of potential application for the electronic nose today. Odor control is of increasing importance in our lives, for examples, in cars, trains, aircraft, inside, and outside of buildings and factories. This environmental application area is particularly important because the E-nose can be trained to recognize hazardous chemicals as well as odors. Furthermore, with respect to the human nose, the E-nose does not fatigue as easily, can be placed in hazardous atmospheres, is less costly, and can travel easily into outer space. It also holds the promise of being much cheaper, smaller, and easier to use and maintain than a mass spectrometer. In conclusion, we hope that what has been written could be a contribution to those in and allied to this field of work.

8. ACRONYMS AND ABBREVIATIONS MOS MOSFET CP BAW SAW QCM QMB PCA PLS LDA CA ANN PCs

Metal oxide semiconductors Metal oxide semiconductor field effect transistors Conducting polymers Bulk acoustic wave Surface acoustic wave Quartz crystal microbalance Quartz microbalance Principal component analysis Partial least square Linear discriminant analysis Cluster analysis Artificial neural network Principal components

The Application of Intelligent Sensor Array

MLP MSE

55

Multilayer perceptron Mean squared error

9. REFERENCES Amrani, M.E.H., Persaud, K.C., Payne, P.A. (1995) “High-frequency measurements of conducting polymers : development of a new technique for sensing volatile chemicals.” Meas. Sci. Technol. 6, 1500–1507. Bachinger T., Martensson P., Mandenius C-F. (in press) Seminars in Food Analysis (vol.2), Chapman & Hall. Baltruschat, H., Kamphauser, I., Oelgeklaus, R., Rose, J., Wahlkamp, M. (1997) “Detection of volatile solvents using potentiodymanic gas sensors,” Analytical Chemistry, 69, 743–748. Bartlett, P.N., Blair, N. and Gardner, J.W. (1993) Electronic nose. Principles, applications and outlook. ASIC 15e Colloque, Montpellier, 1993, 478–486. Beebe K.R., Pell R.J., Seasholtz M.B. (1998) Chemometrics, a practical guide, John Wiley and Sons, New York. Benedetti, S., Mannino, S., Sabatini, A.G., Marcazzan, G.L. (2004) “Electronic nose and neural network use for the classification of honey,” Apidologie 35, 397–402. Buck, T.M., Allen, F.G., Dalton, M. (1965) “Detection of chemical species by surface effects on metals and semiconductors,” in T. Bregman and A. Dravnieks (eds.), Surface Effects in Detection, Spartan Books Inc. Dickinson, T.A., White, J. Kauer, J.S. and Walt, D.R. (1996) “A chemical detecting system based on a cross-reactive optical sensor array,” Nature, 382, 697–700. Dravnieks, A. and Trotter, P.J. (1965) “Polar vapour detection based on thermal modulation of contact potentials”, J. Sci. Instrum., 42, 624. Edward, J. S. (1999) EST Public Report. Web site: www.estcal.com Gardner, J.W. and Bartlett, P.N. (1993) “A brief history of electronic noses”, Sensor and Actuator B, 18, 211–220. Gardner, J.W. and Bartlett, P.N. (1992) Pattern recognition in odour sensing. In Sensors and Sensory systems for an Electronic Nose. Dordrecht: Kluwer Academic Publishers, 212, 161–179. Gardner, J.W., Bartlett, P.N., Dodd, G.H., Shurmer, H.V. (1990) “The design of an artificial olfactory system,” in Schild D. (ed.) Chemosensory Information Processing, NATO ASI Series H: Cell Biology, Vol. 39, Springer, Berlin,1990, 131–173. Guilbault, G., Jordan, J.M. (1988) “Analytical uses of piezoelectric crystal: a review,” Crit. Rev. Anal. Chem. 19, 1–28. Hobbs, P.J., Misselbrook, T.H. and Pain, B.F. (1995) “Assessment of odours from Livestock wastes by a photoionization detector, an electronic nose, olfactometry and gas chromatography-mass spectrometry”, J. Agric. Eng. Res. 60, 137–144. Homan, W.J., Fodisch, H. (in press) Seminars in Food Analysis (vol.2), Chapman & Hall. Iwanaga S., Sato N., Isegami A., Noro T., Arima H., US pat. 4457161, 1984. Joreskogand K.G. and Wold H. (eds.) (1982) Systems under indirect observation, Parts I and II, North Holland, Amsterdam. Kalman, E-L., Winquist, F. and Lundstrom, I. (1997) “A new pollen detection method based on an electronic nose“, Atmos. Environ. 31, 1715–1719. Kohl, D. (1992) “oxidic semiconductor gas sensors”, in Sberveglieri G. (ed.) Gas sensors. Dordrecht: Kluwer Academic Publishers, 43–88. Lundstrom, I., Spetz, A., Winquist, F., Ackelid, U. and Sundgren, H. (1990) “Catalytic metals and field-effect devices useful combination”, Sens. Actuator B, 1, 15–20. McCarrick, C.W., Ohmer, D.T., Gilliland, L.A., Edwards, P.A. and Mayfield, H.Y. (1996) “Fuel Identification by Neural Network Analysis of the Response of Vapor-Sensitive Sensor arrays”, Anal. Chem. 68, 4264–4269.

56

S. Mannino et al.

Meloun M., Militky J., Forina M. (1992) Chemometrics for analytical chemistry. Ellis Horwood, New York. Mielle, P. (1996) “Electronic noses: Towards the objective instrumental characterization of food aroma,” Trends in Food Sci. Technol., 7, 432–438. Mielle, P., Hivert, B. and Mauvais, G. (1995) “Are gas sensors suitable for on-line monitoring and quantification of volatile compounds ?” In Etievant P. and Schreier P. (eds.) Proceedings of Bioflavour 95 Dijon: INRA editions, 81–84. Misselbrook, T.H., Hobbs, P.J. and Persaud, K.C. (1997) “Use of an electronic nose to measure odour concentration following application of cattle slurry to grassland”, J. Agric. Eng. Res. 66, 213–220. Moncrieff, R.W. (1961) “An instrument for measuring and classifying odours”, J. Appl. Physiol.,16, 742. NATO Advanced Research Workshop, Reykjavik, Iceland, August 1991, Gardner J.W. and Bartlett P.N. (eds.), Sensors and Sensory Systems for an Electronic Nose, NATO ASI Series E: Applied Science, Vol. 212, Kluwer, Dordrecht, 1992. Persaud, K. and Dodd, G.H. (1982) “Analysis of discrimination mechanisms of the mammalian olfactory system using a model nose” Nature, 299, 352–355. Schild, D. (ed.). Chemosensory Information Processing, NATO ASI Series H: Cell Biology, Vol. 39, Springer, Berlin, 1990. Stetter J.R., Zaromb S., Findlay M.W. 1984. Sens. Actuator B 6, 269. Wilkens, W.F. and Hatman, A.D. (1964) “An electronic analog for the olfactory processes”, Ann. NY Acad. Sci., 116, 608.

5 Electronic-Nose Technology: Application for Quality Evaluation in the Fish Industry Guðrún Ólafsdóttir and Kristberg Kristbergsson

1. INTRODUCTION The odor of fresh fish is one of the most important quality parameters used to determine whether fish is acceptable for consumption. The composition of volatile compounds in fish contributing to the characteristic odors can be determined and related to quality. Currently there is a need for rapid, automated, and objective tools for process monitoring and quality assurance of perishable food products. The possibility of using electronic nose for rapid quality control of fish therefore is of interest. Knowledge about the composition of the headspace of different fish products during storage is useful to guide the development of electronic noses for quality monitoring of fish products. Complicated postmortem processes in the fish are responsible for the loss of freshness and the onset of spoilage. The spoilage processes are a combination of physical, chemical, biochemical, and microbiological interactions that are species dependent and additionally various extrinsic factors such as handling and different storage conditions will further influence the spoilage pattern and the development of spoilage odors.

2. VOLATILE COMPOUNDS AS INDICATORS OF FISH QUALITY The odor of fish can be classified as species-related fresh fish odor, microbial spoilage odor, oxidized odor, processing odor, and environmentally derived odor, based on the origin of the volatile compounds (Table 1).

GUðRÚN ÓLAFSDÓTTIR AND KRISTBERG KRISTBERGSSON ● Icelandic Fisheries Laboratories, Skúlagötu 4, 101 Reykjavík, Iceland and Department of Food Science, University of Iceland, Hjarðarhaga 2-6, Reykjavík Iceland 57

58

G. Ólafsdóttir and K. Kristbergsson

1000 SULFUR COMPOUNDS

30

CONCENTRATION (PPB)

600

FRESH FISH ALCOHOLS

500

400

SWEET ESTERS

20

AROMATICS

300 10

200

100

SHORT CHAIN ALCOHOLS

5 FRESH FISH CARBONYLS

12

PUTRID

8

STALE

4

SWEET

1

FRESH FLAT

DIENALS

16

20

24

SHORT CHAIN ALCOHOLS CONCENTRATION (PPM)

Because of the nonspecific selectivity and low sensitivity of gas sensors in electronic noses it is necessary to correlate their responses to the main classes of quality indicating volatile compounds of the products. Some compounds present in concentrations below the detection limit of the gas sensors in electronic noses may have an impact on the odor and can function as key odorants due to their low flavor threshold. Compounds contributing to characteristic fresh fish odors like alcohols, carbonyls, and dienals causing oxidation odors are present in very low concentrations (ppb) but have an impact on the overall odor because of their low odor threshold (see Table 1). Microbially formed degradation compounds like alcohols are present in much higher concentrations (ppm) in the headspace of fish but their odor threshold is much higher so their impact on the odor is less (Olafsdottir and Fleurence, 1998). Therefore, it cannot be assumed that the gas sensors are measuring the overall odor, but rather detecting the components that are in the highest concentration in the headspace, which may be of both odorous and nonodorous nature. In some cases the compounds may have little or no effect on the human olfactory system, but they may be detected by the gas sensors. Characteristic odor and the progression of odor in fish during storage have been associated with varying levels of different volatile compounds present in the headspace of fish (Lindsay et al., 1986; Josephson et al., 1986).

28 32 TIME (DAYS)

Figure 1. Generalized changes in concentration of groups of influential aroma compounds and sensory quality stages of refrigerated whitefish (from Josephson et al., 1986)

59

Electronic-Nose Technology

Table 1. Classes of Odors in Fish and Examples of Compounds Contributing to the Odors (from Olafsdottir and Fleurence, 1998) Class of chemical Fish odor species Speciesrelated fresh fish odor

C6-C9 alcohols and carbonyls

Bromophenols

N-cyclic compounds Short chain alcohols Microbial spoilage odor

Short chain carbonyls

Amines

Sulfur compounds

Aromatics

N-cyclic compounds Acids Oxidized odor

Unsaturated aldehydes

Examples of compounds

Aroma description

Odor threshold in water

Hexanal / t-2-hexenal, 1-octen-3-ol, /1octen-3-one 1,5-octadiene-3-ol 1,5-octadiene-3-one 2,6-nonadienal 3,6-nonadienol 2,6-dibromophenol 2,4,6-tribromophenol 2-bromophenol Pyrrolidine piperidine

Green, aldehyde-like Mushroom Heavy earthy, mushrooms Geranium Cucumber Cucumber, melon-like Iodine- and shrimp-like Saltwater fish, brine-like. Sea, marine-like flavor Earthy

4,5ppb / 17ppb a 10ppb/ 0,009ppba

ethanol, propanol, butanol, 3-methyl-1butanol acetone, butanone ethanal, propanal 3-methylbutanal 2-methylbutanal ammonia, TMA DMA histamine, putrecine, cadeverine hydrogen sulfide methyl mercaptan methyl sulfide dimethyl disulfide dimethyl trisulfide

Solvent like

bis-methylthio methane thioesters phenethyl alcohol phenol, p-cresol indole skatole acetic acid, butyric acid isobutyric acid hexanal c4-heptenal 2,4-heptadienal, 2,4,7-decatrienal,

10ppba 0,001ppba 0,001ppba 10ppba 0,0005µg/kgb 0,6µg/kgb

1–100 ppmc

Solvent like Malty Malty Ammoniacal fishy, ammoniacal

0,06ppm d 0,04ppm d 110 ppmc 30 ppmc 0,6 ppmc

Putrid, rotten Sulfury, boiled eggs Rotten, cabbage Cabbage-like Putrid, onion-like Putrid, cabbage and onion-Garlic like

5–40 ppb e 0,05 ppb e 0,9µg/kgf 12 ppb g 0.01ppbg 0,3 µg/kgf

Old roses Phenolic, Pigpen-odors ,horse manure Moth ball or fecal like

Sour, rotten, old socks

2 ppm 300 µg/kgf

34,2ppmc 32,8ppmc

green, planty 4,5ppbf cardboard-like, potato-like 0,04ppbh fishy oxidised flavor burnt, fishy, cod-liver oil-like (Continued )

60

G. Ólafsdóttir and K. Kristbergsson

Table 1. Classes of Odors in Fish and Examples of Compounds Contributing to the Odors (from Olafsdottir and Fleurence, 1998)—Cont’d Processing odors

Environmental odors

2,4-heptadienal and 3,5-octadien-2-one methional 2-methyl-3-furanthiol

ripened anchovies

boiled potato - like odor meaty odor in canned tuna methyl sulfide geosmin petroleum odors 2-methyl-iso-borneol earthy, muddy odors

a

Josephson (1991); b Whitfield et al. (1988); c Kawai T. (1996); d Sheldon et al. (1971); e Fazzalari (1978); f Whitfield and Tindale (1984); g Buttery et al. (1976); h McGill et al. (1974).

Species-related fresh fish odors have been attributed to long-chain alcohols and carbonyl compounds like 1,5-octadien-3-ol and 2,6-nonadienal, respectively, which are oxidatively derived from polyunsaturated fatty acids such as eicosapentaenoic acid 20:5ω3 (Josephson et al., 1984). Spoilage odors develop as a result of microbial activity and oxidative degradation of the fish components. Compounds such as trimethylamine (Oehlenschläger, 1992), short-chain alcohols (Kelleher and Zall, 1983; Ahmed and Matches, 1983), carbonyls, esters, and sulfur compounds like hydrogen sulfide, methylmercaptan, dimethyl disulfide, and dimethyl trisulfide are produced by microbial degradation of fish constituents (Herbert et al.,1975; Kamiya and Ose,1984). Oxidation of fatty acids contributes to the rancid odors of fish with the formation of aldehydes like hexanal, 2,7-heptadienal, and 2,4,7decadienal (McGill et al., 1974). All these compounds are to some degree volatile and may be used to monitor freshness and spoilage of fish. Both single compounds or a combination of compounds representing the different changes occurring during spoilage have been suggested as indicators for freshness and spoilage of fish (Lindsay et al., 1986). Ethanol, 3-methyl-1-butanol, 2-methyl-1-propanol, 3-hydroxy-2-butanone, ethyl acetate, and butanoic acid ethyl ester were the most abundant volatiles in the headspace of haddock stored in ice (Olafsdóttir, 2003). Similar volatile compounds were found in coldsmoked salmon during refrigerated storage (Joffraud et al., 2001; Jörgensen et al., 2001). This is expected since similar profiles of microflora emerge in different food products when subjected to the same conditions despite being heterogeneous initially (Gram et al., 2002). The volatile compounds detected in spoiled cold-smoked salmon were mainly alcohols produced by microbial activity. Some of the volatile compounds produced during spoilage of cold-smoked salmon contributed to the spoilage off-flavor of cold-smoked salmon as confirmed by gas chromatography-olfactometry. These were trimethylamine, 3-methyl butanal, 2-methyl-1-butanol, 3-methyl-1-butanol, 1-penten-3-ol, and 1-propanol (Jörgensen et al., 2001). It is likely that the same set of sensors can be used for monitoring spoilage changes in different fish products because similar volatile compounds emerge

Electronic-Nose Technology

61

in the products. However, because of the complexity of the spoilage processes caused by the diversity of the microflora and the influence of temperature on their spoilage potential, it is likely that models to predict the quality or shelf life of the products based on electronic nose responses will have to be developed for each product and the respective storage conditions.

3. INDUSTRIAL NEEDS FOR RAPID QUALITY EVALUATION OF FISH Although the importance of cultured fish is increasing the fish trade is largely based on wild fish that has been harvested by various catching methods. Consequently, the quality of fish as a raw material for food production is very variable. Therefore, the need for rapid evaluation of fish quality is even greater than for other food commodities where control of the total food chain is better. Chemical measurements of total volatile bases and total microbial counts are used in the industry to verify the freshness of fish according to regulations. Alternative techniques have been developed to monitor changes occurring postmortem in fish. Detection of ATP metabolites and physical measurements evaluating changes in texture, microstructure, electrical properties, and color are well established and newer technologies based on image analysis and spectroscopic methods have shown promising results (Olafsdottir et al., 1997c, 2004; Di Natale, 2003). However, none of these methods have been widely implemented in the industry. This still leaves sensory assessment as the leading method for freshness evaluation in the fish industry to give the best description of the actual consumer preferences. Sensory attributes related to appearance, texture, smell, color, defects, and handling were considered very important in quality control according to a survey on the importance of various quality attributes of fish and detection methods used in the European fish sector (Jørgensen et al., 2003). The results showed general agreement regarding the need for rapid instrumental methods to measure the overall concepts freshness and quality of fish. The need to measure objectively the freshness or quality of perishable food products like fish is of importance both in process management and in the trade of fish. Remote buying via the internet has become more common and an objective method to evaluate the quality would facilitate e-commerce. Additionally, the evaluation of the freshness of various processed fish products is also of interest for the marketability of these products.

4. APPLICATION OF ELECTRONIC NOSES FOR FISH Electronic noses have been suggested for various applications related to quality evaluation of different food such as monitoring freshness and the onset of spoilage or bioprocesses of food. Many of these applications are based on detecting

62

G. Ólafsdóttir and K. Kristbergsson

volatile compounds produced from the growth of fungi, molds, or microbes or changes occurring in food because of oxidation (Schnürer et al., 1999; Olsson et al., 2002; Keshri et al., 2002; Pradyumna et al., 1998; McEntegart et al., 2000; Boothe and Arnold, 2002). In recent years attempts to use electronic nose technology to track the spoilage processes occurring in fish have been reported in numerous papers. Most of these are feasibility studies, showing the capacity of the electronic nose to discriminate between different spoilage levels or storage time of samples. Electronic nose instruments based on various sensor technologies have been used for fish applications and compared with different reference methods like sensory analysis, chemical measurements of volatile compounds like TVN or TMA, and microbial analysis (Table 2). Different catching methods and handling and storage conditions, in particular the temperature, will influence the spoilage pattern of fish, and consequently information about storage time may give ambiguous information about the freshness stage of the fish (Di Natale et al., 2001). Therefore, a more useful approach is to compare the electronic nose responses with a reference measurement that gives information about the microbial and oxidative processes influencing the freshness quality. Gas sensors that currently are used in electronic noses are nonselective toward individual compounds, but some of them are selective toward certain classes of compounds. Research efforts have focused on developing membranes for selective detection of important quality-indicating compounds like sensors aimed specifically at the detection of compounds contributing to freshness odors of fish (Deng et al., 1996). Many researchers have developed sensors for the selective detection of microbial metabolites like TMA (Storey et al., 1984; Egashira et al., 1990; Saja et al., 1999; Zhao et al., 2002). When monitoring the onset of spoilage of capelin during storage with an array of electrochemical sensors, the NH3 sensor gave the best results to predict TVN value of capelin (Olafsdottir et al., 2000). Ohashi et al (1991) evaluated the performance of a semiconductive trimethylamine gas sensor (In2O3 -MgO) to detect the freshness of cod, sardine, and yellow tail stored at 10 and 27 °C. The same group reported further that even a single gas sensor was quite useful to monitor fish freshness by the selective detection of TMA, DMA, and ammonia (Egashira et al. 1994). Volatiles have been associated with the safety of fish products like histamine which causes scombroid poisoning, a leading cause of finfish-borne illness. Histamine traditionally has been used as an indicator for safety. Selective detection of bioamines has been suggested using an amperometric detection of histamine with a methylamine dehydrogenase polypyrrole-based sensor (Zeng et al., 2000). The development of “smart packaging” has received attention recently. The potential use of sensors based on pH-sensitive films for monitoring spoilage volatiles released into packaged fish headspace were described by Byrne et al. (2003). The sensors were based on cresol red entrapped within a plasticized

NH3 - electrode / sensory

e-Nose 4000

Conducting polymers

Smoked salmon / freshness storage time Cod, sardine, yellow tail/ TMA/ DMA/NH3 detection Semiconductor gas sensor (TiO2 – In2O3-MgO

Metal oxide sensors

Microbial count / sensory

FishNose AlphaMOS Prototype

K value, TVB-N

Sensory/microbial counts

Sensory

AromaScan

e-Nose 4000

Conducting polymers

Salmon fillets

Salmon fillets / freshness, storage time Shrimp / freshness storage time

AromaScan e-Nose 4000 Prototype

TVB-N / sensory Sensory TMA / microbial count / sensory TVB-N / microbial count / sensory Lipid oxidation products Antioxidants/ sensory analyses Microbial count / sensory Microbial count / sensory No reference

TVB-N / sensory

Sensory (QIM), color, texture

Sensory analysis TVB-N Sensory analysis

Reference method

Conducting polymers Conducting polymers Amperometric sensors/ heating filaments Conducting polymers

MOSFET and MOS (Taguchi)

FreshSense

FreshSense

FreshSense LibraNose FreshSense LibraNose EnQbe

Instrument

Tuna Tuna Trout /freshness, storage time

Herring / freshness, storage time Cod roe / ripening stage Redfish / storage time Cod fillets / freshness, storage time Herring fillets

Capelin / freshness, storage time Electrochemical sensors

Metaloxide sensors QMB Electrochemical sensors QMB Metalloporphyrins and polymers coated thickness shear mode resonators: Electrochemical sensors

Haddock /freshness, storage time Cod /freshness, storage time

Cod-fillets/ freshness, storage time Sardine/ freshness, storage time

Sensors

Type of fish/ application

Egashira et al. (1994) Ohashi et al. (1991)

Luzuriaga and Balaban (1999a) Olafsdottir et al. (2005)

Olafsdottir et al. (2002) Olafsdottir et al. (2002) Haugen and Undeland (2003) Du et al. (2001) Newman et al. (1999) Schweizer-Berberich et al. (1994) Luzuriaga and Balaban. (1999b) Du et al. (2002)

Olafsdottir et al. (1997a,b, 2000) Olafsdottir et al. (1998)

Ólafsson et al. (1992) Di Natale et al. (2001) Olafsdottir et al. (2003a) Di Natale et al. (1996) Macagnano et al. (2005)

References

Table 2. Examples of Application of Electronic Noses Based on Various Sensor Technologies to Detect Changes during Storage and Processing of Different Seafood (adapted from Olafsdottir, 2003)

Electronic-Nose Technology

63

64

G. Ólafsdóttir and K. Kristbergsson

cellulose acetate matrix and a simple illumination source The TVB-N levels released from orange roughy and black scabbard, deepwater fish, were monitored by color changes in the sensors.

5. CASE STUDY I: FRESHSENSE — AN ELECTRONIC NOSE PROTOTYPE INSTRUMENT DEVELOPED TO BE SELECTIVE FOR QUALITY INDICATOR COMPOUNDS IN FISH An electronic nose called FreshSense based on electrochemical gas sensors has been developed in our laboratory in cooperation with an Icelandic company (Bodvaki-Maritech). Knowing that odor of fish is one of the most important quality attributes to determine the freshness quality, the aim was to apply the instrument for quality monitoring in the fish industry by rapidly measuring volatile compounds contributing to the spoilage odor of fish. The main classes of compounds that are found in increasing concentrations in the headspace of fish with storage time are short-chain alcohols, aldehydes, ketones, esters, sulfur compounds, and amines (Lindsay et al., 1986) (Figure 2). As discussed before, the information on the identity and quantity of volatile compounds present in the headspace during storage of fish is essential when selecting sensors in an array for quality monitoring of fish. Moreover, it is important to know the sensitivity of the sensors toward key compounds in the sample indicating the quality. Selected standard compounds that are representative of

Ammonia odor, dried fish, old fish odor Amines: ammonia (NH3), trimethylamine (TMA), dimethylamine (DMA) Biogenic amines: histamine, cadeverine, putrecine, tyramine

Sweet-sour, malty odor Alcohols, aldehydes, ketones, esters: ethanol, propanol, butanol, 2-me-1-propanol 3-me-1-butanol, 3-methylbutanal butanone, acetoin ethyl acetate

Putrid odor, onion like, old cabbage, rotten egg Sulfur compounds: hydrogen sulfide (H2S), methylmercaptan (CH3SH), dimethyl disulfide (DMDS), dimethyl trisulfide(DMTS) Acids: acetic acid

Figure 2. Quality indicators of fish. Main classes of compounds contributing to microbial spoilage odor of fish.

65

Electronic-Nose Technology

the main classes of compounds causing spoilage odor have been analyzed by the electronic nose FreshSense (Olafsdottir et al., 1998, 2002). The results showed that the electrochemical gas sensors (Dräger, Germany: CO, SO2; City Technology, Britain: NH3) in the FreshSense had different selectivity and sensitivity toward the compounds selected from these classes. The CO sensor was sensitive toward alcohols, aldehydes, and esters; the NH3 sensor detected amines like TMA and ammonia; and the SO2 sensor detected volatile sulfides. Therefore, the main classes of spoilage indicator compounds present in the headspace can be estimated based on the individual sensor responses. A storage study of haddock fillets kept in cold storage (0–2°C) for 3, 7, 10, and 14 days was performed to obtain more detailed information about the identity and the level of the most abundant volatile compounds present in the headspace during storage of fish (Olafsdottir et al., 2003b). Analyses of the headspace volatiles were conducted using air pump headspace sampling and collection of volatiles by a TENAX preconcentration technique and analysis by gas chromatography (GC-MS). The results were compared to the individual responses of the FreshSense electronic nose sensors and sensory odor evaluation of the fillets. The responses of the electronic nose sensors were comparable to the results of the analysis of volatile compounds obtained by the GC-MS detection (Figure 3). The sum of the GC responses of individual compounds belonging to each class of compounds like the alcohols, aldehydes, and esters corresponded to the CO sensor responses of

1200 CO SO2 Response sensors (nA)

arbitrary units

Alcohols, aldehydes and esters

Amines (TMA)

NH3 800

400

Sulfur compounds

0 (a)

5

10 Days

0

15

0 (b)

5

10

15

Days

Figure 3. Sum of the peak areas of compounds representing the three different classes of compounds detected by GC in the headspace of haddock fillets during storage in ice (a) and responses of the CO, SO2, and NH3 sensors toward haddock fillets during storage in ice (b) (from: Olafsdottir, 2003)

66

G. Ólafsdóttir and K. Kristbergsson

the electronic nose. The sum of the amines and the sulfur compounds detected by GC-MS corresponded to the responses of the NH3 and SO2 sensors, respectively. The response of the CO sensor was the highest and increased early in the spoilage process while the responses of the NH3 sensor and the SO2 sensors increased later in the spoilage process. Because of breakthrough of small polar molecules on the TENAX traps the technique is not suitable for the quantification of compounds like ammonia, hydrogen sulfide, methyl mercaptan, and ethanol that have been known to be present in abundance in the headspace of spoiled fish. However, the electrochemical sensors can detect these compounds and therefore the slopes and shapes of the curves are slightly different. Increasing concentrations of spoilage indicator compounds with storage time resulted in the development of characteristic odors and simultaneously increased responses of the electronic nose sensors were observed. On day 3 the odor of the fresh fillet was very little or neutral and low responses of the sensors were observed. The first spoilage odors of the fillets were sweetlike odors that were contributed by the alcohols which give sweet, solventlike odors in combination with aldehydes giving sweet, rancidlike odors (day 7). The amines contributed to salted fish or stock fish odor and in combination with the sulfur compounds, cheesy and foul odors developed and the fillets became stale on day 10. Finally, esters were analyzed in high levels on day 14 giving characteristic sweet, fruity odors. When these sweet odors were mixed with the foul smell of sulfur compounds and ammonialike stockfish character of the amines the odor of the fillet becomes TMA/ammonialike and sour/putridlike, signaling the overt spoilage. It is of interest to compare the composition of the headspace of different fish products during storage to guide the development of an electronic nose for quality monitoring of fish products. A similar set of sensors can be used for a variety of fish species that are stored and processed by different methods. The FreshSense instrument has been used for freshness monitoring of various species of fish like haddock, capelin, redfish, and cod that were handled and stored under different conditions. A similar trend in the responses of the electronic nose and the development of volatile compounds was observed for the different fish species (Olafsdottir et al., 1997a, b, 1998, 2000, 2002; DiNatale et al., 2001).

6. CASE STUDY II: ELECTRONIC NOSE MEASUREMENTS TO EVALUATE QUALITY OF HADDOCK FILLETS Raw material is often labeled with days from catch; however, because of the effect of various extrinsic and intrinsic factors, the information about days from catch is not always reliable for the determination of quality or freshness of the raw material. Sensory evaluation of fish fillets is more difficult than sensory evaluation of the whole fish. Therefore, instrumental techniques to evaluate the freshness quality of the fillets are of special interest. The most obvious spoilage signs of the fillets are development of spoilage odor, discoloration,

67

Electronic-Nose Technology

and decreased firmness of the flesh. Various instrumental techniques to detect these changes have been developed (Olafsdottir et al., 1997c, 2004) but their implementation in the fish industry has been slow. Several storage studies of haddock stored both as whole fish on ice and as fillets have been performed in our laboratory (Tryggvadottir and Olafsdottir, 2000; Olafsdottir et al., 2003b). The main objective of the studies was to evaluate the possibility of using an electronic nose to monitor spoilage changes in haddock fillets and to study the spoilage pattern of fish caught at different seasons (spring and autumn) using different fishing gear (long line and Danish seine) and storage conditions. The fish was stored whole in ice during the first two experiments (May 1999 and Sep 1999), but in the third one (May 2000) the fish was stored as fillets in Styrofoam boxes in a cold room (0–2°C). Changes of various properties related to freshness quality of haddock stored in ice were monitored by the electronic nose FreshSense and compared to traditional chemical [total volatile bases (TVB-N)], sensory (Torry scheme) and microbial methods [total aerobic viable count (TVC)] and texture measurements (TPA). Measurements were performed on fillets on days 1, 4, 6, 8, 11, 13, and 15 counted from the day of catch. Sensory analysis of the fillets using evaluation of cooked fillets according to the Torry scheme showed that the spoilage pattern was different in the three studies (Figure 4). The limit of shelf life was defined as Torry score = 5.5. Fillets spoil faster than whole fish as expected and seasonal variation influenced the spoilage rate. Recently spawned fish caught by Danish seine and stored as whole fish in ice from the spring (May 1999) had shorter shelf life in ice (10 days) than whole fish caught by long line from the autumn season (Sep 1999) (14–15 days). 10 9 Torry sensory score

8 7 6 5 4

sensory rejection Torry = 5.5

3 2

MAY 00

MAY 99

SEP 99

1 0 0

1

2

3

4

5

6

7

8 9 Days

10 11 12 13 14 15 16

Figure 4. Sensory analysis of haddock fillets from storage studies of whole fish [(May 99 (-■-); Sept 99 (-●-)] and fillets [May 00 (-▲-)].

68

G. Ólafsdóttir and K. Kristbergsson

Fish stored as fillets at 0–2°C (May 2000) spoiled faster than whole fish and had a shelf life of only 8–9 days. The results of the electronic nose measurements and the microbial and chemical analysis were mostly in agreement with the sensory analysis (Figures 5 and 6). The traditional microbial analysis (TVC) and electronic nose measurements (CO sensor) showed a similar overall trend in the storage studies of whole fish (May 1999 and Sept 1999) and fillets (May 2000) (Figure 5) and likewise the results of TVN chemical measurements and the NH3 electronic nose sensor showed similar overall trends (Figure 6). The electronic nose measurements showed that the responses of the sensors were highest for the (May 2000) samples indicating the most rapid spoilage for fish stored as fillets. The results from the sensory analysis, the TVN (total volatile bases) measurements, and the NH3 sensor showed that recently spawned fish from the spring season (May 1999), stored as whole fish, had higher spoilage rate than fish from the autumn season (Sep 1999). Contradictory to these results the TVC and CO measurements indicated that the spoilage rate appeared to be higher in the fall (Sep 1999) than in the spring (May 1999). However, based on the results of sensory analysis it appears that the spoilage potential of the microflora might have been greater in the spring. This suggests that the role of specific spoilage organisms producing offensive odors should be emphasized and the validity of TVC measurement may be questionable as has been suggested by other researchers (Gram et al., 2002). When the end of the shelf life was reached, as indicated by sensory analysis, the responses of the NH3 and SO2 sensors detecting microbially produced amines and sulfur compounds start to increase rapidly in all experiments. The conclusions drawn from these results indicate that spoilage patterns are different depending on season and fillets spoil faster than whole fish, as was to be expected. 900 MAY 00

8 TVC log cfu/g

7 SEP 99

6 5

MAY 99

4 3 2 1 0 0

5

10 Days

15

20

Response CO sensor (nA)

9

800

MAY 00

700 600

SEP 99

500 400 300

MAY 99

200 100 0 0

5

10 Days

15

20

Figure 5. Microbial analysis (TVC) and electronic nose measurements (CO sensor) of haddock fillets from storage studies at 0–2 °C for up to 15 days of whole fish [May 99 (-■-); Sept 99 (-●-)] and fillets [May 00 (-▲-)].

69

Electronic-Nose Technology 70 MAY 00

60 TVB-N mg N/100g

Response NH3 sensor (nA)

100

50 MAY 99

40 30 20

SEP 99

10

MAY 00

90 80 70

MAY 99

60 50 40 30 20

SEP 99

10 0 0

0 0

5

10 Days

15

5

10 Days

20

15

20

Figure 6. Chemical measurements of total volatile bases (TVB-N) and electronic nose measurements (NH3 sensor) of haddock fillets stored at 0–2 °C for up to 15 days, from storage studies of whole fish [May 99 (-■-); Sept 99 (-●-)] and fillets [May 00 (-▲-)].

Principal components analysis is a good way to illustrate the overall results (Figure 7). The first PC explained 60% of the variation in the data set and the spoilage level of the samples increased from left to right along PC1. The second PC explained 16% of the variation in the data and appeared to be discriminating 1.0

PC2

FIRM

1/MAY99

Bi-plot TPA

1/MAY 00

H2S 0.5 RT-Meter 1/SEPT99 Torry 0

4/MAY99

NH3

15/MAY99 MAY 1999

Fillets MAY 2000

TVB13/MAY99 TMA

4/MAY 00

13/MAY00 11/MAY00

8/MAY00 11/MAY99 log TVC logH2S CO

8/MAY99 4/SEP99 6/MAY99 −0.5

SO3

6/MAY 00

15/MAY00 Whole fish

6/SEP99 8/SEP99 11/SEP99

13/SEP99 −1.0 −0.5 all data, X-expl: 60%, 18%

15/SEP99 18/SEP99 Whole fish SEP 1999 PC1

0

0.5

1.0

Figure 7. PCA showing a biplot of sample scores and variable loadings of all data from three storage studies of haddock as whole fish (May 1999 and Sept 1999) and fillets (May 2000) stored at 0–2 °C for up to 15 days. The samples are labeled with storage days and time of the experiment and the variable loadings are labeled as follows: E-nose: CO, NH3, SO2, and H2S; chemical: TVB/TMA; microbial: log H2S, log TVC; texture: Firm /TPA.

70

G. Ólafsdóttir and K. Kristbergsson

samples based on the season. The variable loadings show that the CO sensor and the microbial counts (log H2S and log TVC) are highly correlated and the positioning of the H2S, SO2, and NH3 sensors on the upper half of the plot show the contribution of these sensors to discriminate samples from the spring season in combination with the texture measurement (Firm). Fish spoils faster in the spring because the fish is recently spawned and the texture of the muscle is softer and more gaping than in the fall. Different spoilage patterns are evident depending on the season indicating that days of storage is not a good estimate of freshness quality. The E-nose sensor responses appear to give information on the quality of fish related to sensory scores, microbial growth (TVC), and spoilage indicators like TVB-N.

7. CONCLUSIONS Rapid measurements of the headspace of fish with an electronic nose has the potential to detect the freshness or quality stage of the fish, if the sensor array can detect the respective indicator compounds that truly represent the condition of the fish. The high sensitivity of the electronic nose sensors toward the very volatile degradation compounds, like alcohols, sulfur compounds, and amines signaling overt spoilage suggests that the E-nose may be promising for monitoring off odors caused by fish and animal waste and off odors caused by the processing of fish meal or dried products when the raw material is not of optimal freshness.

8. ACKNOWLEDGMENTS The case studies were part of projects financed by The Icelandic Centre for Research: Accurate predictive models—The effect of temperature fluctuations on microbial growth and metabolites in fish and The European Commission: MUSTEC project FAIR98 4076.

9. RECOMMENDED READING The following papers give a good overview of the application of electronic noses for monitoring food (Garcia-González and Aparicio, 2002; Harper, 2001; Haugen, 2001; Bartlett et al., 1997; Gardner and Bartlett, 1999; Schaller et al., 1998) and Jurs et al. (2000) give an excellent overview of data processing techniques for electronic noses. Information on the website of The NOSE II 2nd Network on Artificial Olfactory Sensing has the aim to stimulate information exchange between scientists, manufacturers, and end users in Europe in order to develop synergy in the sensor

Electronic-Nose Technology

71

community, improve the efficiency of R&D, encourage interdisciplinary research, and promote application of new ideas (http://www.nose-network.org/).

10. REFERENCES Ahmed, A., Matches, J. R. (1983) “Alcohol production by fish spoilage bacteria.” Journal of Food Protein 46, 1055–1059. Bartlett, H.N., Elliott, J.M., Gardner, J.W. (1997). Electronic noses and their application in the food industry, Food Technology 51(12): 44–48. Boothe, A.A.J., Arnold, J.W. (2002). ‘Electronic nose analysis of volatile compounds from poultry meat samples, fresh and after refrigerated storage’, J. Sci. Food Agric.82, 3, 315–322. Buttery, R. G., Guadagni, D.G., Ling, L.C., Seifert, R.M., Lipton, W. (1976). Additional volatile components of cabbage, broccoli and cauliflower. J. Agric. Food Chem., 24 (4), 829–832. Byrne, L.; Lau, K.T.; Diamond, D. (2003). Development of pH sensitive films for monitoring spoilage volatiles released into packaged fish headspace. Irish Journal of Agricultural and Food Research, 42(1) 119–129. Deng, Z., Stone, D.C., Thompson, M. (1996). Selective detection of aroma components by acoustic wave sensors coated with conducting polymer films. Analyst 121,671–679. Di Natale, C., Brunink, J.A.J., Bungaro, F., Davide, F., d’Amico, A., Paolesse, R., Boschi, T., Faccio, M., Ferri, G. (1996), ‘Recognition of fish storage time by a metalloporphyrins-coated QMB sensor array’, Meas. Sci. Technol. 7: 1103–1114. Di Natale, C., Olafsdottir, G., Einarsson, S., Mantini, A., Martinelli, E., Paolesse, R., Falconi, C., D’Amico, A., (2001). Comparison and integration of different electronic noses for the evaluation of freshness of cod fish fillets. Sensors and Actuators B77,572–578. Di Natale, C. (2003), Data Fusion in MUSTEC: Towards the Definition of an Artificial Quality Index. In: Luten J B, Oehlenschlager J and Olafsdottir G (Eds), Quality of Fish from Catch to Consumer: Labeling, Monitoring and Traceability, Wageningen Academic Publishers, The Netherlands, 273–280. Du, W.X., Kim, J., Huang, T.S., Marshall, M.R., Wei, C.I. (2001), ‘Microbiological, sensory and electronic nose evaluations of yellowfin tuna under various storage conditions’, J Food Prot. 64 (12): 2027–2036. Du, W.X., Lin, C.M., Huang, T.S., Kim, J., Marshall, M.R., Wei, C.L. (2002), ‘Potential application of the electronic nose for quality assessment of salmon fillets under various storage conditions’, J Food Sci. 67 (1): 307–313. Egashira, M., Shimizu, Y., Takao, Y. (1994), Fish Freshness detection by semiconductor gas sensors, Olfaction and Taste XI. Proc. Int. Sym. 11, 715–719. Egashira, M., Shimizu, Y., Takao, Y. (1990) Trimethylamine sensor based on semiconductive metaloxides for detection of fish freshness. Sensors and Actuators B-Chem 1 (1–6): 108–112. Fazzalari, F.A. (1978). Compilation of Odor and Taste Threshold Value Data. American Society for Testing and Materials, Philadelphia. Gacrcía-González, D. L. and Aparicio, R. (2002). Sensors: From biosensors to the electronic nose. Grasas y Aceites, 53(1) 96–114. Gardner, J.W., Bartlett, P.N. (1999), Electronic noses. Principles and applications, Oxford University Press. Oxford, England. Gram, L., Ravn, L., Rasch, M., Bruhn, J.B., Christensen, A.B., Givskov, M. (2002) Food spoilage - interactions between food spoilage bacteria. Int. J. Food Microbiology 78 (1–2): 79–97. Harper, W.J. (2001), Strengths and weaknesses of the electronic nose. In: Rouseff and Cadwallader (Eds) Headspace Analysis of Food and Flavors: Theory and Practice. Kluwer Academic / Plenium Publishers, New York pp 59–71. Haugen, J.E. (2001). Electronic noses in food analysis. In: Rouseff and Cadwallader (Eds). Headspace Analysis of Food and Flavors: Theory and Practice. Kluwer Academic / Plenium Publishers, New York pp 43–57.

72

G. Ólafsdóttir and K. Kristbergsson

Haugen, J.E., Undeland, I. (2003), Lipid oxidation in herring fillets (Clupea harengus) during ice storage measured by a commercial hybrid gas-sensor array system, J Agric. Food Chem. 51, 3, 752–759. Herbert, R.A., Ellis, J.R., Shewan, J.M. (1975), Isolation and identification of the volatile sulfides produced during chill-storage of North Sea Cod (Gadus morhua). J.Sci. Food Agric., 26, 1195–1202. Joffraud, J.J., Leroi, F., Roy, C., Berdague, J.L. (2001), Characterisation of volatile compounds produced by bacteria isolated from the spoilage flora of cold-smoked salmon, Int. J. Food Microbiol. 66, 3, 175–184. Jørgensen, B.M., Oehlenschläger, J., Olafsdottir, G., Tryggvadóttir, S.V., Careche, M., Heia, K., Nesvadba, P., Nunes, M.L., Poli, B.M., Di Natale, C., Pérez-Villarreal, B., Ballo, H., Luten, J.B., Smelt, A., Denton, W., Bossier, P., Hattula, T., Åkesson, G. (2003), A study of the attitudes of the European fish sector towards quality monitoring and labeling. In: Luten, J.B., Oehlenschläger, J. and Olafsdottir, G. (Eds) Quality of Fish from Catch to Consumer: Labeling, Monitoring and Traceability, Wageningen Academic Publishers, The Netherlands, 57–71. Jørgensen, L.V., Huss, H.H., Dalgaard, P. (2001). Significance of volatile compounds produced by spoilage bacteria in vacuum-packed cold-smoked salmon (Salmo salar) analyzed by GC-MS and multivariate regression. J. Agric. Food Chem. 49, 2376–2381. Josephson, D.B. (1991). Seafood. In Volatile compounds in foods and beverages, H. Maarse (Ed) Marcel Dekker, New York, 179–202. Josephson, D.B., Lindsay, R.C., Stuiber, D.C. (1984), Variations in the occurrences of enzymically derived volatile aroma compounds in salt- and freshwater fish. J. Agric. Food Chem., 32, 1344–1347. Josephson, D.B., Lindsay, R., Olafsdottir, G. (1986). Measurement of Volatile Aroma Constituents as a Means for Following Sensory Deterioration of Fresh Fish and Fishery Products. In Proceedings of an International Symposium, University of Alaska Sea Grant Program, Anchorage, Alaska, U.S.A., Kramer DE and Liston J (Ed.), Elsevier Science Publishers B.V., Amsterdam pp. 27–47. Jurs, P.C., Bakken, G.A., McClelland, H.E. (2000) Computational Methods for the Analysis of Chemical Sensor Array Data from Volatile Analytes. Chem. Rev. 100 pp. 2649–2678. Kamiya, A., Ose, Y. (1984). Study of the odorous compounds produced by putrefaction of foods. V. Fatty acids, sulfur compounds and amines. Chromat., 292, 383–391. Kawai, T. (1996). Fish flavor. Critical Reviews in Food Science and Nutrition, 36, 3, 257–298. Kelleher, S.D., Zall, R.R. (1983). Ethanol accumulation in muscle tissue as a chemical indicator of fish spoilage. J. Food Biochem., 7, 87–91. Keshri, G., Voysey, P., Magan, N. (2002). Early detection of spoilage moulds in Keshri bread using volatile production patterns and quantitative enzyme assays. J. Appl. Microbiol. 92, 165–172. Lindsay, R.C., Josephson, D.B. and Olafsdottir, G. (1986). Chemical and biochemical indices for assessing the quality of fish packaged in controlled atmospheres. In Proceedings of an International Symposium, University of Alaska Sea Grant Program, Anchorage, Alaska, U.S.A., Kramer, D.E. and Liston, J. (Ed.), Elsevier Science Publishers B.V., Amsterdam pp 221–234. Luzuriaga, D.A., Balaban, M.O. (1999a). Evaluation of odor decomposition in raw and cooked shrimp: Correlation of electronic nose readings, odor sensory evaluation and ammonia levels. In: Hurst, W.J. (Ed), Electronic nose and sensor array based system, Design and Applications. Proceedings of the 5th International Symposium on Olfaction and the Electronic Nose. Technomic Publication Company, Inc, USA, pp.177–184. Luzuriaga, D.A., Balaban, M.O. (1999b). Electronic nose odor evaluation of salmon fillets stored at different temperature. In: Hurst, W.J. (Ed) Electronic nose and sensor array based system, Design and Applications. Proceedings of the 5th International Symposium on Olfaction and the Electronic Nose. Technomic Publication Company, Inc, USA, pp.162–169. Macagnano, A., Careche M., Herrero, A., Paolesse, R., Martinelli, E., Pennazza G., Carmonae, P., D’Amico A., Di Natale, C. (2005), A model to predict fish quality from instrumental features. Sensors and Actuators B 111–112 293–298. McEntegart, C.M., Penrose, W.R., Strathmann, S., Stetter, J.R. (2000). Detection and discrimination of coliform bacteria with gas sensor arrays. Sensors and Actuators B 70, 170–176.

Electronic-Nose Technology

73

McGill, A.S., Hardy, T., Burt, R.J. and Gunstone, F. D. (1974). Hept-cis-4-enal and its contribution to the off-flavor of cold-stored cod. J. Sci. Food Agric., 25, 1477–1481. Newman, D.J., Luzuriaga, D.A., Balaban, M.O. (1999). Odor and Microbiological Evaluation of Raw Tuna: Correlation of Sensory and Electronic Nose Data. In: Hurst, W.J. (Ed), Electronic nose and sensor array based system, Design and Applications. Proceedings of the 5th International Symposium on Olfaction and the Electronic Nose. Technomic Publication Company, Inc, USA, pp.170–176. Oehlenschläger, J. (1992). Evaluation of some well established and some underrated indices for the determination of freshness and/or spoilage of ice stored wet fish’. In: Huss, H.H., Jakobsen, M., Liston, J. (Eds), Quality Assurance in the Fish Industry, , Elsevier Science Publisher, Amsterdam, 339–350. Ohashi, E., Takao, Y., Fujita, T., Shimizu, Y., Egashira, M. (1991). Semiconductive trimethylamine gas sensor for detecting fish freshness, J. Food Sci. 56(5)1275–1278. Olafsdottir, G., Martinsdóttir, E., Jónsson, E.H. (1997a). Rapid gas sensor measurements to predict the freshness of capelin (Mallotus villosus). J. Agric. Food Chem. 45(7) 2654–2659. Olafsdottir, G., Martinsdóttir, E., Jónsson, E.H. (1997b). Gas sensor and GC measurements of volatile compounds in capelin (Mallotus villosus). In Seafood from Producer to Consumer, Integrated Approach to Quality. Eds. Luten, J.B., Börresen, T., Oehlenschläger, J. Amsterdam, Elsevier, pp 507–520. Olafsdottir, G., Martinsdóttir, E., Oehlenschläger, J., Dalgaard, P., Jensen, B., Undeland, I., Mackie, I.M., Henehan, G., Nielsen, J., Nilsen, H. (1997c), Methods to evaluate fish freshness in research and industry. Trends Food Sci. Technol, 8. 258–265. Olafsdottir, G., Högnadóttir, Á., Martinsdóttir, E. (1998), Application of gas sensors to evaluate freshness and spoilage of various sea foods. In: Methods to Determine the Freshness of Fish in Research and Industry, Proceedings of the Final meeting of the Concerted Action “Evaluation of Fish Freshness” AIR3 CT94 2283. Nantes Nov 12–14, 1997. International Institute of Refrigeration, pp100–109. Olafsdottir, G. and Fleurence, J. (1998). Evaluation of fish freshness using volatile compounds Classification of volatile compounds in fish. In Methods to Determine the Freshness of Fish in Research and Industry, Proceedings of the Final meeting of the Concerted Action “Evaluation of Fish Freshness” AIR3 CT94 2283. Nantes Nov 12–14, 1997. International Institute of Refrigeration, 55–69. Olafsdottir, G., Högnadóttir, Á., Martinsdóttir, E. and Jónsdóttir, H. (2000). Application of an Electronic Nose to Predict Total Volatile Bases in Capelin (Mallotus villosus) for Fishmeal Production, J. Agric. Food Chem. 48 ,6, 2353–2359. Olafsdottir, G., Li, X., Lauzon, H. L., Jonsdottir, R. (2002). Precision and application of electronic nose measurements for freshness monitoring of redfish (Sebastes marinus) stored in ice and modified atmosphere bulk storage. JAFP 11, 3/4, 229–249. Olafsdottir, G., Di Natale, C., Macagnano, A. (2003a). Measurements of quality of fish by electronic noses. In: Luten, J.B., Oehlenschläger, J. and Olafsdottir, G. (Eds), Quality of Fish from Catch to Consumer: Labelling, Monitoring and Traceability, Wageningen Academic Publishers, The Netherlands, 225–234. Olafsdottir, G., Jonsdottir, R., Tryggvadottir, S.V., Einarsson, S., Lauzon, H.L., (2003b). Prediction of freshness quality of haddock fillets using electronic nose, texture and TVN measurements First Joint Trans-Atlantic Fisheries Technology Conference (TAFT) 33rd WEFTA Meeting and 48th Atlantic Fisheries Technology Conference, 11–14 June 2003, Reykjavik - Iceland 2003 pp. 148–150. (http://vefur.rf.is/TAFT2003/). Olafsdottir, G. (2003). Developing rapid olfaction arrays for determining fish quality. In Tothill, I.E. (Ed) Rapid and on-line instrumentation for food quality assurance. Woodhead Publishing Ltd, Cambridge, England, pp. 339–360. Olafsdottir, G., Nesvadba, P., Di Natale, C., Careche, M, Oehlenschläger, J., Tryggvadóttir, S.V., Schubring, R., Kroeger, M., Heia, K., Esaiassen, M., Macagnano, A., Jørgensen, B.M. (2004). Multisensor for fish quality determination. Trends in Food Science and Technology 15, 86–93.

74

G. Ólafsdóttir and K. Kristbergsson

Olafsdottir, G., Chanie, E., Westad, F., Jonsdottir, R., Thalman, C., Bazzo, S., Labreche, S., Marcq, P., Lundby, F., Haugen, J.E. (2005). Prediction of Microbial and Sensory Quality of Cold Smoked Atlantic Salmon (Salmo salar) by Electronic Nose. J Food Sci 70 (5) 563–574. Olafsdottir, G. (2005). Volatile compounds as quality indicators of fish during chilled storage: Evaluation of microbial metbolites by an electronic nose. PhD thesis Faculty of Science, University of Iceland, 219p. Olafsdottir, G., Lauzon, H., Martinsdottir, E., Oehlenschläger, J., Kristbérgsson, K. (2006). Evaluation of shelf-life of superchilled cod (Gadus morhua) fillets and influence of temperature fluctuations on microbial and chemical quality indicators. J. Food Sci 71 (2) 97–109. Ólafsson, R., Martinsdóttir, E., Olafsdottir, G., Sigfússon, Þ. I., Gardner, J. W. (1992), Monitoring of fish freshness using tin oxide sensors. In: Gardner, J.W. and Bartlett, P. N. (Eds), Sensors and Sensory Systems for an Electronic Nose, Kluwer Academic Publ. 257–272. Olsson, J., Börjesson, T., Lundstedt, T., Schnürer, J. (2002), Detection and quantification of ochratoxin A and deoxyinvalenol in barley grains by GC-MS and electronic nose. Int. J Food Microb. 72, 203–214. Pradyumna, K., Namadev, Yair Alroy and Vijay Singh (1998) Sniffing out trouble: Use of an electronic nose in bioprocesses, Biotechnol. Prog, 14, 75–78. Saja, R.de, Souto, J., Rodriguez-Mendez, M.L., de Saja, J.A. (1999) Array of lutetium bisphthalocyanine sensors for the detection of trimethylamine. Materials Science & Engineering C-Biomimetic and Supramolecular Systemes 8–9: 565–568. Schaller, E., Bosse, J.O., Escher, F. (1998) ‘Electronic noses’ and their application to food’, LebensmittelWissenschaft-und-Technologie, 31 (4) 305–316. Schnürer, J., Olsson, J. and Börjesson, T. (1999) Fungal volatiles as indicators of food and feeds spoilage. Fungal Genetics and Biology 27, 209–217. Schweizer-Berberich, M., Vahinger, S., Göpel, W. (1994) Characterization of food freshness with sensor array, Sensors and Actuators B, 18–19, 282–290. Sheldon, R.M., Lindsay, R.C., Libbey, L.M., Morgan, M.E. (1971) Chemical nature of malty flavor and aroma produced by streptococcus lactis. Applied Microbiol. 22, 263–266. Storey, r.M., Davis, K., Owen, D., Moore, L. (1984). Rapid approximate estimation of volatiel amines in fish, J. Food Technol. 19,1–10. Tryggvadottir, S.V., Olafsdottir, G., 2000. Multisensor for fish: Questionnaire on quality attributes and control methods -Texture and electronic nose to evaluate fish freshness. Project report for European Commission (MUSTEC CT-98–4076). RF report 04–00, 36p. Whitfield, F.B., Shaw, K.J., Tindale, C.R. (1988). 2,6-dibromophenol: The cause of an iodoform-like off-flavour in some Australian crustacea. J. Sci. Food Agric., 46, 29–42. Whitfield, F.B., Tindale, C.R. (1984). The role of microbial metabolites in food off-flavours. Food Technol. Australia, 36, 204–213. Zeng, K., Tachikawa, H., Zhu, Z.Y., Davidson, V.L. (2000), Amperometric detection of histamine with a methylamine dehydrogenase polypyrrole-based sensor, Analytical Chemistry (10): 2211–2215. Zhao, C.Z., Pan, Y.Z., Ma, L.Z., Tang, Z.N., Zhao, G.L., Wang, L.D. (2002), “Assay of fish freshness using trimethylamine vapor probe based on a sensitive membrane on piezoelectric quartz crystal,” Sensors and Actuators B-CHEM 81 (2–3): 218–222.

6 Odors Prevention in the Food Industry Regina Nabais

1. INTRODUCTION Taking into account the environmental impact, at the local and regional level, the control and elimination of the emission of noxious odors in the food-processing industry is one of that industry’s most difficult problems. The problem normally is exacerbated not only because of the multiple ways in which it can be approached, but also because, in spite of many efforts developed, there still is no universally accepted way to address its solution. The complexity of the issue of off-odors and the necessary multidisciplinary expertise entailing a rigorous approach is the reason this chapter is neither extensive nor intended to represent, even at an introductory level, an odor management guide for the food industry. The scope is to strengthen the necessity to reconcile aims, methods, collaborative attitudes, and efforts; to thoroughly outline strategies and tactics of scientific, technological, political, and legal interventions; and to explore the continued improvement and consistency of effective global and integrated actions that will lessen the general environmental impact, including odor emissions, of the food-processing industry. Whenever possible, the main concepts, principles, and fundamentals of a few mentioned suggestions to odor abatement or control will be mentioned and a list of bibliographic references will be provided.

2. CONCEPTIONS AND MISCONCEPTIONS ABOUT ODORS EMISSION BY THE FOOD INDUSTRY In Europe, the Charter of Fundamental Rights, as proclaimed at the Nice European Council in December 2000 specifically states: “A high level of environmental REGINA NABAIS ● C.E.R.N.A.S. Centro de Recursos Naturais, Ambiente e Sociedade, Escola Superior Agrária de Coimbra, Instituto Politécnico de Coimbra. Bencanta, 3040–316 Coimbra, Portugal. 75

76

R. Nabais

protection and the improvement of the quality of the environment must be integrated into the policies of the Union and ensured in accordance with the principle of sustainable development.” This statement reflects the numerous changes in European politics owing to environmental problems (Mike Holland et al., 2004; SEPA and NIEHS, 2002 a, b). Emissions of foul odors, regardless of their origin, are problems of extreme complexity, assuming an ever-larger role among environmental issues. This is especially the case when they have a direct, negative impact upon the populations near an offending industry, whether they are a simple inconvenience or a harmful treat to neighboring communities. An odor is recognized through sensory perception and subjectively characterized either as agreeable or disagreeable, or even ambivalently perceived, depending upon the specific circumstances of how an individual’s olfactory sense is stimulated. The olfactory sense probably has not received the attention it deserves, considering that it influences the senses of touch and taste, with deep repercussions on psychological behavior, stemming from its possible memorization and affective and cultural interpretations. When the sources of pollution emissions are specific industries, such as food processing, pharmaceutical, or cosmetics, the supposedly marginal questions of production processes such as esthetics and odor become important issues in the production process, because they may be reflected in the price of the products in the marketplace. Concerning odors, even those generally considered agreeable, under certain conditions can become unbearable, depending upon the variability of emission and reception conditions. This particularity is known by the acronym “FIDOL,” which stands for the following: (1) frequency: how often an individual is exposed to odour; (2) intensity: the strength of the odor; (3) duration: the length of a particular odor event; (4) offensiveness/character: the character relates to the “hedonic tone” of the odor (meaning “smells like”), which may be pleasant, neutral, or unpleasant; and (5) location: the type of land, use, and nature of human activities in the neighborhood of an odor source.

3. THE COSTS OF NOT TREATING ODORS Even if the majority of malodors produced by the food processing industry are inoffensive, due to low concentrations or reduced toxicity, the industries that emit them must be prepared to encounter different levels of protest against their activities: 1. People believe that the odor has a negative effect on their well-being and health conditions. 2. Environmental and activist political groups often protest against any activity that may lead to the emission of bad odors.

77

Odors Prevention in the Food Industry

3. The plaintiffs and their lawyers or legal representatives will usually drag the offending industries through interminable legal battles, extremely costly in terms of time, money, and resources. 4. The normative legislation at the local, regional, national, and supranational levels are progressively demanding, regulatory, and restrictive. 5. Other businesses and entities, which complain of the negative or potentially negative impacts on their costumers and their future investments in the area. 6. Some clients may boycott food products produced by factories that emit bad odors. 7. The installation of some potentially economically viable activities (e.g., commercial ventures, tourism, sporting venues, etc.) would be compromised by locations near odor-emitting industries. Some activities of the food-processing industry are potential emissaries of “bad” odors e.g., fish processing, roasting coffee, rendering of fats, extraction of vegetable oils, wineries and breweries, chocolate production, preprocessed foods, even if located some kilometers from inhabited areas, can add significantly to the energy bills of their neighborhoods, due to the fact that they must maintain environmental control systems operating, without the need of their utilization for temperature control. Table 1 synthesizes what is the meaning and consequences for externalities (social costs) that are costs not directly considered or paid by the owners of an economic activity. The creation of good relations and trust with those who live near offending emission producers, combined with an ongoing tightening of regulations, can be decisive factors in the resolution of odor problems. Pushing away the neighbors from the search for solutions environmental problems caused by an industry promotes strong feelings of suspicion and helplessness and gives rise to unbearable resentments, never overcome by developing expensive information campaigns. In fact, effective public relations and communications tactics help producers become aware and strengthen and foster favorable actions by policymakers; otherwise, people tend to exaggerate risks and may be suspicious of arguments limited to technical information. Table 1. Relative Cost for Controlling Industrial Environmental Impacts Enterprise Costs Quality of the prevention or control systems Improper Proper a

Initial investment

Maintenance and operational costs

Depreciation of the enterprise image and final property value

Social Costsa (externalities)

Inexistent or low Medium to high

High Low

High Null

High Low

Concept explained in (Holland et al. (2004)).

78

R. Nabais

One method to enlist the confidence and trust of the local neighborhood residents is to offer them a forum in which to explore their apprehensions and doubts, to offer guided tours of the facility and its production system, and to include them in both exploring the environmental problems and the quest for possible solutions. “Science is losing credibility. Conflicts of interest, biased studies, and secrecy are undermining science’s reputation and its truth-seeking objective affecting public trust” (Collins, 2000). Hence, trust and confidence are the main keys to addressing public perception and understanding. This is not in itself an expensive methodology in monetary terms, but it demands accomplishing the most difficult tasks human beings can achieve: cultural changes. Trust enforces the exchange of information and ideas and increases the likelihood of the community feeling more in control of the solutions for their environmental concerns, instead of promoting overreactions as a result of exaggerating their fears by useless secrecy or hiding the problems.

4. MAJOR INDUSTRIAL FOOD PROCESSES RESPONSIBLE FOR ODOROUS SUBSTANCES EMISSIONS Apart from considering process isolation, capable confinement, and good conveyor systems, immediately followed by efficient onsite treatments (and despite the existence of several exceptions) direct interventions in food process and/or food process control have poor effect on annoying properties of odors. Food process residues, their manipulation, transfer operations, treatment and final disposal systems or facilities are also sources of malodors emissions. A food process residue is usually understood to be an incidental organic material generated by processing agricultural commodities for human or animal consumption. The term includes food residuals, food coproducts, foodprocessing wastes, food-processing sludge, or any other incidental material whose characteristics are derived from processing agricultural and fishery products. Examples include process wastewater from cleaning slaughter areas, rinsing carcasses, or conveying food materials; process wastewater treatment sludge; bone; fruit and vegetable peels; seeds; shells; pits; cheese whey; off-specification food products; hide; and hair and feathers. As any other air pollutant, odor sources may be classified as point source or disperse source; unfortunately, the food industry may present both odor origins. A rough organization of improving food industry odors sources may be seen in Table 2. Thus, odor can be considered as a “primary air pollutant” emitted by some food industrial processes: raw materials handling, transport, storage, cooking, concentration, drying, grilling, fermenting, and roasting processes are examples of sources of odors emissions. If these emissions are potentially annoying to their neighbors, their sources must be rethought, controlled, and treated separately from the fugitive odors of raw materials.

79

Odors Prevention in the Food Industry

Table 2. Examples of Food Processes Susceptible of Odors Emissionsa,b Operations and processess Cooking; drying; evaporation; extraction; fermentation; frying; roasting

Utilities operation and maintenance

Waste handling and disposal

Change, antiquated, inefficient, or improper odor control systems Plan and preview emergency releases

Remediation and recovery of contaminated land Avoid clogged drains, by proper procedures Evaluate emissions from open tanks, lagoons, effluent treatment plants Convenient planning for solid or liquid sludge, slurries and wastes handling, transfer, storage and final disposal sites

Design of heating, combustion, or steam generators Prevention of leaking spills and splashes from vessels, pipes, and connections

Design appropriate local exhaust ventilation discharge points Pumps and flanges sealings Efficient raw materials logistic organization, as transports, storage, and handling Good features for start-up and shut-down processes Design of storage tank “breathing” vents Appropriate installations for end products, by-products, and waste residues Efficient transfer processes, including unloading of vehicles, warehousing locations, and procedures Confinement of skips or bins Adequate design for buildings and storage areas ventilation Appropriate washing, disinfection, cleaning, and maintenance operations a

From IPPC (2003).

On the other side, in general, liquid wastes produced by food industry have high contents of organic pollutants including biochemical oxygen demand, ammonia, suspended solids, oils and greases and are commonly acid, biodegradable, and malodorous (Environmental Protection Agency, 2000). When an unstable organic waste in liquid, slurry, or sludge starts decomposing, gases are liberated and accumulate in fluid solution phase, until the saturation of the liquid is reached; beyond this point, smelly vapors are released. Most of food process treatment systems (ponds, leaching beds, septic tanks, and trickling filters) are poorly built-up, inefficient, and improperly operated. Fortunately, there are several highly efficient technologies available for industrial food manufacturing liquid effluents, but the costs are generally unbearable, considering the narrow marginal profits of these activities. The problem for

80

R. Nabais

small and medium size businesses is to find a technology that is comparatively affordable, with a short return on the investment relative to its life cycle, capable of treating high industrial-strength wastewater, with nonodorous treatment, demanding comparatively little land and flexible enough to be installed as needs arise (Robèrt et al., 2002; Starkey, 1998). On other hand, taking into consideration the temperature dependence of degradation rates, reducing load rates applied to waste treatment features during cold weather periods is a good practice to decrease off-odors emission. Another advised procedure for those kinds of biological waste systems is their continuous loading, at least fed-batch rate additions adapted to temperature patterns, to avoid the biological agents facing variable nutrient concentrations or sudden media composition alterations, which can induce biological nutrient shocks and shifts in metabolic mechanisms to malodorous processes, as those exemplified in Table 3. Whatever may be the origin of malodors emissions, it is common to find in the food industry liquid or gaseous streams containing isolated substances, or in complex mixtures, substances mentioned in the Table 2, Chapter 2, paragraph 3.1. For instance, at rendering plants, among hundreds of detected odorous compounds there may be found acetaldehyde, ethyl mercaptan, ammonia, hydrogen sulfide, butyl amine, indole, butyric acid, methylamine, dimethylamine, methyl mercaptan, dimethyl sulfide, skatole, dimethyl, qinoline and triethylamine. Quinoline is classified as a hazardous air pollutant (UKRA, 2004; US Environmental Protection Agency, 1995).

5. ODOROUS COMPOUNDS AND MIXTURES EMISSIONS IN FOOD INDUSTRY AND THEIR GENERAL AND SPECIFIC PHYSICAL AND CHEMICAL CHARACTERISTICS Odorous substances exhibit common physical and chemical general properties, all contributing to odor perception; these generic properties are their molecular weight (less than 300), high volatility, high vapor pressure, and high solubility (in water needed at the membrane level of the nose and in the lipid components of nervous cells).

Table 3. Three Dominant Paths of Anaerobic Biological Substrate Biodegradation Carbohydrates

Alcohols, aldehydes, ketones, organic acids

Lipids Proteins

Long and short chain fatty acids, alcohols, acetate Peptones, peptides, amino acids, volatile fatty acids, sulfides, mercaptans, thiols, phenols, indoles

Methane, carbon dioxide, hydrogen sulfide, ammonia, water and biomass

Odors Prevention in the Food Industry

81

Odorous compounds from most of the food-processing plants are emitted generally in low concentrations and are simultaneously hydrosoluble and biodegradable. Hence, for such pollutants, simultaneously highly soluble and degradable, such as methanol, butanol, tetrahydrofurane, aldehydes, formaldehyde, carbonic acids, and butyric acid, biological treatment technology appears very attractive, as its investment and running costs are often lower by a factor of 2 to 10 than the physical–chemical technologies.

6. INDOOR AND OUTDOOR FOOD PROCESSING FACILITIES: ENVIRONMENTAL QUALITY AND SAFETY, CONCERNING ODOURS EMISSIONS Odor emissions in the indoor environment of food industries, affects the indoor air quality (IAQ). When emissions have negative impacts it is common to observe a decrease in worker productivity and increases in absenteeism; in extreme conditions, malodorous substances may induce serious illness. Beside these already mentioned, serious effects, another caution is crucial in IAQ of a food industry and that is “the taint.” The term taint relates to any foreign smell or taste in goods, e.g., musky, yeasty, catty, furry, disinfectants, cork, cardboard, hearthy (geosmin), musty (methylisoborneol), as a result of adsorption or absorption of volatile compounds, by the commodities materials, throughout the entire production line (from raw materials to the final packed products). The causes for taint occurrences in food products are very diverse, but one of the sources may be poor indoor air quality. Anyhow, occurrences like tainting can result in considerable inconveniences to manufacturers, like lost of production, sales, consumer confidence, damaged brand names, loss of commercial relationships between suppliers, manufacturers, and retailers, insurance disputes, and in extremely expensive and long legal battles. The prevention of taint development is vital to any food process, not only to dignify and preserve product image and stay in business, but also because in some countries and circumstances (e.g., in United States), if the taint is a result of intentional or careless manufacturing and cumulatively implies external commerce depreciation, those legally responsible may be subject to heavy fines and may be sentenced to up to 3 years in prison (US Food and Drug Administration, 2004). The three strategies, in order of effectiveness, for reducing pollutants in the indoor air environment are source control, ventilation, and air cleaning. (1) Source control eliminates individual sources of pollutants or reduces their emissions, and is generally the most effective strategy. (2) Ventilation brings outside air indoors. It can be achieved by opening windows and doors, by turning on local exhaust fans, or, in some situations, by the use of mechanical ventilation systems, with or without heat recovery ventilators (air-to-air heat exchangers). However, there are practical limits to the extent ventilation can be used to reduce airborne pollutants. Costs for heating or cooling incoming air can be significant, and outdoor air

82

R. Nabais

itself may contain undesirable levels of contaminants (Society of Heating, Refrigerating, and Air Conditioning Engineers, 1997). (3) Air cleaning may serve as an adjunct to source control and ventilation. However, the use of air cleaning devices alone cannot assure adequate air quality, particularly where significant sources are present and ventilation is insufficient. The balance of the amount and quality of outdoor air getting in and indoor sources of emissions is of major importance of indoor air quality; this balance is identified by indoor/outdoor ratios, usually in the range of 0.5 to 1.7. Unfortunately, indoor odor consequences are not yet considered when ventilation systems are projected; in fact, they are addressed to provide makeup air, controling ambient temperature, and or relative humidity control. If odor emission is a problem, designing for this objective gives the worker a comfortable environment to work in, minimizes concerns related to worker safety, and improves worker attitudes, thus, increasing productivity. The ventilation system layout for nonworker cover or enclosure areas generally includes an exhaust point(s). Instead, a makeup air access is used and positioned strategically away from the exhaust to achieve the desired sweep direction of airflow motion. To avoid odors releasing through the makeup air inlet, this access must automatically be shut, in the event that positive pressure builds up in the enclosed headspace. In some instances, the makeup air source is not necessary, as other sources of air can replace it. The air contained in the headspace of the drains and sewer commonly provides makeup air; in this case one must be attentive to taint prevention. For worker space areas, the ventilation system layout is site specific, but a few design criteria always need to be considered, including the following aspects: ●

●

●

Position ventilation system exhaust points at or near the source of the odors; avoid fouling worker space areas by alternate exhaust locations. Hydrogen sulfide is heavier than air; hence, its exhaust points must be located near the floor. This gas is commonly a prevalent odor in wastewater treatment systems. As a result, the makeup air supply should be located high, near the ceiling of the enclosure, while the odorous exhaust is low, at the floor level where the source of the odors is generally found. Avoid shortcuts between the makeup air entrances and exhausting system; this is an important consideration to provide indoor air quality to workers. The required internal summer ventilation rate should be at least 10 times greater than the minimum continuous rate. This is the prescriptive approach of the American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE); current ventilation standards are America National Standards Institute (ANSI) and are ANSI/ASHRAE Standard 62–2001 (ASHRAE, 1999) plus ASHRAE BOD approved for publication Addenda and ASHRAE Standard 62.2-2003 (ASHRAE, 2003). In fact, these documents provide techniques to establish minimum ventilation requirements (liter/sec) per person or per surface unit (liter/sec per m2), but the minimum

Odors Prevention in the Food Industry

●

●

●

83

values must be specifically estimated according to specific conditions, such as internal and external temperature, relative humidity, and recommended CO2 (or a specific targeted pollutant) concentration. Anyhow, just to give an order of magnitude, the continuous rate may be 10 liter/sec per person, at 22°C and 40% relative humidity. ASRHRAE standards may be a useful start-up approach for each food industry internal environment, if the gasproblem concentrations within the building must be kept at a known level during the summer in a properly designed facility. It worth reminding that the creation of ASHRAE standards and guidelines are recommendations, determined solely by their usefulness, hence, the conformance to them so far is completely voluntary. The use of forced air supply and exhaust for worker space areas (forced air supply together with exhaustion system) provide the best ventilation coverage of enclosures. Equipping the makeup air supply and exhaust registers with volume dampers to control the rate of air. Volume dampers are useful with field flow balancing the ventilation system. Makeup air flux supply should be somewhat less than exhaust to create the negative air pressure within the enclosure. Air changes are defined two different ways: First, how many times the air in the space circulates through the mechanical equipment. Second, how many times the air in the space is replaced with outside air. Finally, another perspective, ACE (that is, the ratio of the effective ventilation rate at the breathing zone divided by the effective ventilation rate which would occur with perfect mixing while at the same ventilation rate) shall be at least 1.0 to 1.2.

In summary, ventilation rate procedure prescribes: ● ●

●

●

the outdoor air quality for acceptable ventilation; outdoor air treatment when necessary, adapted to ventilation rates for residential, commercial, institutional, vehicular, and industrial spaces; criteria for reduction of outdoor air quantities, when recirculated air is treated by contaminant-removal equipment; criteria for variable ventilation when the air volume in the space can be used as a reservoir to dilute contaminants.

7. MAIN PREVENTIVE TECHNOLOGIES TO CONTROL ODOR EMISSIONS Malodorous emissions are never isolated problems of the food industry environment. Indeed, odorant emissions only reflect environmental mismanagement faults and procedures.

84

R. Nabais

According to the United Nations Environment Program (UNEP) (Pennsylvania State University, 2002; UNEP Working Group, 2004a, b) in what concerns general environmental industrial management and cleaner technologies, there must be a consideration of several issues (multicriteria system), which despite the difficulties in being correctly conceptualized, may result in very positive impacts on odor emissions abatement. Among possible ways are automatic control systems, best practices implementation, energy control, financial planning, features “housekeeping,” improved chain management, inventory management, key performance indicators, preventive maintenance, materials accounting, modify inputs, onsite treatment, primary screening/segregation, process control/improvement, recycling, redesign reuse, wastes management, and storm water Management. The great difference between pollution control and cleaner production is that pollution control is an after-the-event, “reactive actions” approach, whereas cleaner production reflects a proactive, “anticipate and prevent” philosophy. Most importantly, performing good practices in those aspects afterwards can be proudly announced as promotional advertisements for manufactured products or services. Companies therefore are adopting a more systematic approach to environmental management, sometimes through the environmental management system (EMS). EMS is a decision-making structured tool to project an action program, bringing cleaner production into the company’s strategy of development, management, and day-to-day operations; in fact, it provides a continuous globally integrated approach to progressively improve the industrial activities by EMS, depicted in Figure 1. An example of an element of this integrated management model, but applied within a scale of a regional development strategic vision, may be achieved by strategically installing a biorefinary, projected to become a friendly technology, materially and energetically sustained, which enables exploitation of biological raw and natural materials in the form of green or processed, wet and waste wet biomass from a targeted sustainable the regional land utilization, representing the interesting concept of industrial symbiosis (Lifset, 1999; Realff and Abbas, 2003). 7.1. Production Shifting, Reformulation, and or Retrofitting Production shifting, reformulation, and retrofitting, said “frontend” control measures, are best received by industries, because they should be implemented only after favorable or unfavorable economic return is correctly estimated. If careful studies are done, the final results are usually very positive for both perspectives: profits for the company’s stockholders and also important environment gains (Pinchin, 2004). Just as an example, in the case of fisheries there are several tips to consider when processing poor quality fish, such as (1) the temperature in

85

Odors Prevention in the Food Industry

6. Continuous improvement

5. Managing review 1. Environmental principles

2. Planning 3. Implementation and operation 4. Checking and corrective actions

Figure 1. General sketch of an environmental management system.

indirect steam dryers must be controlled to avoid scorching of the fish (overdrying, losing weight) and amplifying odor emission problems; (2) increasing the pressure of the presses will improve the recovery of malodorous constitution solutions; and (3) the size of the holes in the pre-strainer must be increased, because if the fishery processes poor quality raw materials, these tend to produce a slurry that clogs up the outlets of the presses, diminishing its efficiency. 7.2. Storages and Warehousing When handling material that is date sensitive or perishable, which is the case in the food industry, several requirements must be foreseen: ●

●

The storage spaces must be clean and carefully organized. All points of the storage must be easily accessible and offer safe conditions for frequent inspections and surveillance procedures; refrigeration units should be cleaned and maintained on a preset periodic basis. For refrigerated and frozen products, temperature and moisture is critical. Given the effects of temperature on the safety conditions of biological goods and raw materials, alarm devices are necessary to monitor temperatures.

86

R. Nabais ●

●

If refrigeration systems are used, precautions must be taken to avoid drips over the products or raw materials. First-in-first-out rotation (FIFO) principle is the removal of inventory from storage in a systematic way where earlier stock items are used first. This can be accomplished by date coding the inventory according to the date of receipt. Warehousing, mainly, involves three activities (receiving, storage and shipping) that are included in a quality control program. The receiving operation is the foundation for processing finished food products of a designated quality (GDV, 2003 and Tybor et al., 1995). The main recommendations, to improve organization of stocks: (1) minimize queues’ waiting periods and dock time; (2) move refrigerated or frozen items directly into storage; (3) compare the incoming shipments with the order specifications; and (4) date code all incoming shipments directly on the container or pallet load for stock rotation.

Certain supplies or ingredients may require segregation (e.g., acetic acid, sodium hydroxide), even for safety recommendations; as a rule one should always consult material safety data sheets (MSDS). The MSDSs were conceived, developed, and determined by the Occupational Safety and Health Administration (OSHA) in their Hazard Communications Standard 29CFR 1910.1200. This standard is commonly referred to as the “right-to-know law”, requiring companies to provide specific information about ingredients, hazards, and precautions on all hazardous material labels. The MSDS must be developed by the product manufacturer and is mandatorially sent, whenever requested, if not promptly sent with the product. Shipping is the final step in which a food business can have direct control on product quality. Ship items on a first-in–first-out basis and use the same guidelines in shipping that you followed in receiving. If not properly managed, storage items may become ruined and smelly. Figure 2 displays an improper condition of biological residues storage because the bags are not closed. For example, onions have an unpleasant, pungent odor. An increase in odor levels indicates improper storage that promotes self-heating, which is associated with the loss of essential oils, sugar, and vitamins. Damage to onions, caused by setting them down roughly, for example, results in greater respiration intensity. The cleaning and sanitizing operations should address three basic areas: internal facility including floors, walls, ceilings, and ventilation system; exterior facility; and grounds. 7.3. Process Modification A good example of a process modification can be evaluated by the following description. There are two basic types of yeast dough mixing processes used in bakeries: sponge dough and straight dough. For the purpose of estimating emissions, the length of the fermentation time is the critical difference between these

Odors Prevention in the Food Industry

87

Figure 2. Residues improperly maintained in open bags, waiting for reutilization or destruction. Permission by courtesy of Portary Gestão de Resíduos Lda (2004).

two processes. It is during the fermentation process that the Volatile organic compounds (VOC) are produced. The sponge dough process, which is most commonly used by commercial bakeries, produces the largest amount of VOC emissions because the required fermentation time can be 5 hours or more. By cutting spongeous fermentation period, from 4 hours to just 1, through use of a preferment concentrate, the bakeries can lower ethanol emissions by half and more. In a British bakery, two depanners use a vacuum system to remove bread from the baking trays and tins. The vacuum is generated in vacuum chambers by fans driven by standard 15 kW AC motors. With these improved conditions, having a payback period of 1.3 years, technical and economical advantages were registered: (1) energy savings: about 135,000 kWh/year; (2) reduction in CO2 emissions: 67,500 kg/year; (3) damage of bread rolls were minimized; and (4) lower overall maintenance cost was obtained. Also, to control VOC emissions from commercial and industrial boilers, no auxiliary equipment is needed; properly maintaining the burner/boiler package will keep VOC emissions at a minimum. Maintenance services must include keeping the air/fuel ratio at the manufacturer’s specified setting, having the proper air and fuel pressures at the burner, and maintaining the atomizing air pressure on oil burners at the correct levels. An improperly maintained boiler/burner package can result in VOC levels over 100 times the normal levels of specific processes, and the energetic efficient decreases. Another good example in the food industry is deep frying processes. It is important to choose oil with a high smoke point (the temperature at which it starts to smoke). When oil smokes, it begins to decompose and, some experts say many of its unsaturated fatty acid molecules become saturated. Frying oils are exposed to atmospheric oxygen and moisture at elevated temperatures during

88

R. Nabais

frying. The oil gets deteriorated, as the triglycerides in the oil undergo oxidative, hydrolytic, and thermal alterations producing more than 400 various chemical compounds. Breakdown of the glycerol molecules in the fat creates acrolein, an obnoxious-smelling compound that easily inflames. This means that perfectly safe fat, with a flash point of 260 °C, might burst into flames at a dangerously low temperature after having been used, just once in high temperature. Avoid adding salt to food before frying, because salt draws moisture to the food’s surface, which will splatter when the food is added to the hot oil. Salt also lowers the smoke point and breaks down the oil. 7.4. Capturing, Conveying, Confining and Removing Odors The most common odor management control for wastewater systems is a floating cover made from polyethylene (PE) membrane suspended on cables and/or supported by floatation devices, with perforated gas collection pipes, weighted sunken troughs to collect water, sump pumps, and so on. The main function of such cover is to prevent vector migration of vapors or gases into the atmosphere and to collect methane gas to be disposed by flaring or other means to reduce odor pollution. No matter how elaborate installation procedures and the methodologies, efficiency, the puncture of PE membrane, wind, hail, snow, ice vulnerability, explosive hazards due to static electricity and accumulation of volatile gases, vandalism, dependency on electric power supply and reliance on professional maintenance of electric motors and pumps contribute to a substantial initial investment and a high operating cost. However, the best technology in the world or the most costly odor control system can only be effective in reducing odorous emissions if the odors reach the control devices. If the containment and ventilation systems are not able to keep the nuisance smells from escaping, then the money spent on the odor control has been wasted. 7.5. In Situ Compounds Recovering Solving malodorous emissions constantly implies improvements in industrial performance, greater efficiency in resources consumption, and reduction in the levels of waste. At the microlevel, the industrial ecology concept demands close attention to unit processes, facility operations, and industrial behavior and organization. The food industry emits gaseous streams, in many of their processes, such as storage vents, distillation vents, reactor vents, mixers, pan coolers, spray dryers, and granulators. Gaseous compounds may be separated and recovered from the whole gas stream by absorption and adsorption, distillation, extraction and refrigeration

89

Odors Prevention in the Food Industry

or condensation based systems. Of the techniques used in the final separation of gaseous compounds, Table 4 shows comparison of three different vapor recovery systems. Adsorption plays an important role in industry, both in the removal of contaminants from a product and as a direct means of recovery of substances. Its most important application is in the removal of organic contaminants from polluted sources. Another example is the case of citrus oil, produced by pressings citrus peels (to decrease liquid content before the peels can be reused, i.e., pellets production) which is usually in a form of emulsion of water oil; the oil may be recovered by vacuum distillation process or fractional distillation if we desire different grades of oil. With this procedure, the liquid effluents will have a lower biological oxygen demand (BOD) concentration, acquiring better adaptation to be biologically treated, reducing malodors emissions, both from liquid and gaseous sources.

Table 4. Comparing Performances of Refrigeration, Absorption, and Adsorption for in situ Compounds Recoverya Refrigeration

Absorption

Adsorption

Application process

Higher concentration Condensation

Operational temp. Pressure drop Vacuum Operations

Very low temp. Very low Not required Fully automatic/ continuous Low maintenance Power used when required Very low Not required

Low concentration—0.5% Separation and condensation Room and above Low Required Full automatic/ continuous Low maintenance Continuous

Low concentration—0.5% Separation and condensation Room and above High Required Automatic batch process

Low Required for outlet gas

Large Required for outlet gas

Smallest Note required Not required

Small Required Special circumstances

Not required

Not required

Not required

Required

Large Required Required for safe operations Required periodic replacement Not required

Intrinsically safe

Safe

Precautions required

Maintenance Power cost Installation cost Concentration measurement Space required Steam Inert gas Carbon replacement Solvent replacement Safety a

From Mohunta (2004).

High maintenance Continuous

90

R. Nabais

7.6. Managing Onsite Good Technological Practices in All Industrial Process: Preventing and Suppressing Odorous Emissions Whatever may be the industrial manufacturer, good practices always require eliminating unintentional holes in buildings and keeping doors and windows closed, avoiding the storage of odorous materials outside the building and the transport of materials between buildings in open containers, and so on. Containment of highly odorous process gases, i.e., keeping it separate from less odorous streams, may reduce the capital and operating costs of the required abatement system(s). As a general rule all odor control systems should exert a negative pressure locally or within process buildings to prevent odorous air leaking out. The most cost-effective approach to odor control may be to examine the operation and maintenance practices at the processing facility and to introduce prevention pollution techniques such as recovery, product design changes, technology or process modifications, input or material changes, in-process recycling and reuse, and good operating practices, e.g., waste segregation, preventive maintenance, training and awareness programs, and production scheduling (Holland et al., 2004). ●

●

●

Automation of transfer processes and computerized raw material handling and pumping system, totally enclosed, help maintain a cleaner environment (Tybor et al., 1995). Installation of removable insulation on–off valves, pipes, and fittings to reduce losses in the process heat distribution system, with potential energy savings of 2–5% and prevent scale accumulation by ensuring water treatment systems are operating effectively. Scale buildup in boiler feed water tubes inhibits both throughput and heat transfer, with potential gains in boiler efficiency of 10–12%. Whenever heat production increases its efficiency, decreases the combustion operations, and hence potential gaseous emissions also diminish, and in consequence all gaseous pollutants decrease as well.

A comprehensive option for reducing odor was proposed by one rendering industry, involving a process change to replace the existing dryer with a steamheated indirect dryer, improvements to ventilation of the rendering spot, and replacing the scrubber with a biological filter. Whatever may be the alternatives to reduce or solve odor problems, there are not many options outside of the schemes suggested by Figure 3 and Figure 4, as follows: Analysis of the Figure 3 shows that we only have two options to deal with industrial smelly problems, but many times they are complementary to each other: emission and reception. Malodor control may be addressed through the following alternative ways:

91

Odors Prevention in the Food Industry Odor formation process Transfer to air Release to atmosphere Dispersion Frequency, Duration, Intensity of Exposure

Exposure

Detection and perception Time of the day, activity context, relation to source and association with odor

Appraisal by the receptor Other ambient stressors (noise, dust crowding)

Individual perception Annoyance/Nuisance Access to complaint channels, Legal instruments and Expected results COMPLAIN ACTION

Figure 3. Origins of odors complaints. From Van Harreveld (2001).

●

●

●

●

●

Reduction of odor production, extraction of ventilation air with treatment to reduce odor concentration. Using miscellaneous additives that can react or mask odors. Reduction of transfer rate (i.e., for instance pH, temperature emissions control, covering the surface). Reduction of exposed area of slurry, including storage, soiled surfaces, grids, and so on. Variable dilution in the atmosphere through turbulent dispersion (turbulence or stability of the boundary layer, wind direction, wind speed, etc.). Adapt the emissions to the neighbors’ characteristics.

Indeed, it is clear that high waste production level is an indication of process inefficiency and losts of profits, which precisely shows the expenditure of valuable resources to produce a product that cannot be sold: waste. Figure 5 displays the relative distribution of preferences calculated, over 8 fundamental principles, considering 123 different specific clean technologies options, preferred by 48 food industries, located in 21 different countries, resulting from an inventory realized by the International Cleaner Production

92

R. Nabais

Raw materials Extraction and beneficiation

Manufacturing process

Consumers

Residues/Emissions

Recycling

Environment

Retrofitting

Reusing

Environment impact (ExternalitiesSocialCosts)

Figure 4. Cleaner production principles.

Information Clearinghouse (ICPIC), which underlines the high importance of process modification and housekeeping improvements, in what concerns cleaner technologies. Housekeeping involves special training, improving management to avoid leakages, periodic maintenance of equipment, strict control of washing water to reduce wastewater, reuse of cooling water where possible, obtaining guarantees of raw material quality, and improving raw material storage. 7.6.1. Process Modification, Recovery, Reuse, and Recycle Some food processing wastes such as tomato and apple pressings, corn cobs and husks, sweet corn cannery waste, stale bakery products, and spent brewery mash may be used as feed complements to cattle, helping farmer to reduce

93

Odors Prevention in the Food Industry Reuse 13%

Housekeeping 24%

Recycle 13% Material Substitution 2%

Recovery 13%

Product Modification 2%

New technology 9%

Process Modification 24%

Figure 5. Relative frequency of cleaner technologies principles used for pollution abatement in food industry. From International Cleaner Production Information Clearinghouse (2004).

overall feeding costs. Also, feeding these recyclable wastes to animals reduces the volume of solid waste that food processors must dispose of at an approved sanitary landfill or apply as a soil amendment. In conclusion, any material that does not become a part of the final products must be promptly recycled into or off the process, reuse to some other by-product, or ultimately adequately disposed. The costs of odors treatment can be minimized by focusing on the previous options and cleaner production. 7.6.1a. Reduce. The next step is to minimize the use of all materials in the process. This can include reducing errors in batch preparation, optimizing cleaning operations to reduce the volume of water used, and turning off equipment that is not in use. 7.6.1b. Reuse. There are many opportunities to reuse waste products in the food manufacturing industry. This will reduce the demand for raw materials and the cost of treatment and disposal. It may be possible to reuse some clean process waters in the boiler or install drip trays to catch product that eventually may return to the process in next batches. To enable biobased materials production in the food industry, it is only necessary to obey preestablished rules of good waste management. An interesting example in this domain is Cargill Dow’s polylactide (PLA), which is a versatile biopolymer, made from 100% renewable

94

R. Nabais

resources like crop residue (stems, straw, husks, and leaves) from corn or other crops (Vinka et al., 2003). 7.6.1c. Recycle. Are the wastes identified by your assessment really wastes? Can some of these be reclaimed through a simple treatment process, which enables them to be reused on-site. Other by-products that cannot be used on-site may be recycled off-site. In these cases there may be the potential to sell recyclable items and also save by the avoiding disposal costs. 7.6.1d. Treat and Dispose. This option should only be considered after the other options have been exhausted. Generally these options are typically a cost to industry. However, it may be essential to consider this as a part of your overall cleaner production strategy. The costs of treatment can be minimized by focusing on described options, which may be specifically applied to different food industries and services (UNEP, 2004a,b). According to the Ontario Ministry of Agriculture (1995), many of those food wastes are wet and/or decompose readily. Leachates and polluted or contaminated runoffs may easily develop problems with odors. To avoid such problems, farmers should tailor handling and storage arrangements to their situation and the waste or by-product they use. Some general guidelines to minimize odors and water quality problems include: ●

●

●

●

Provide adequate storage facilities (preferably covered) appropriate for the material. Locate storage facilities away from and where possible downwind of neighbors’ residences and buildings. Wherever possible, keep the volume of stored waste to a minimum and feed the material to animals as quickly as possible; this is especially important for wastes that decompose readily or that produce particularly strong odors (e.g., vegetables and fruits in general, such as onions and cabbages). Provide the feed in a trough or receptacle that will allow all the feed to be eaten rather than spread into the ground. Avoid water quality problems through collections; store and properly dispose of leachates or liquor from the stored material; they can be heavily concentrated pollutants, and give rise to smelly emissions.

7.6.2. Enabling Technology Summarizing what has been said, several alternatives should be considered when choosing technological options to environmental integrated management system (United Nations Environment Programme, 2004): (1) Good operating practices (e.g., waste segregation); (2) improved housekeeping (e.g., keep containers covered); (3) material substitution (e.g., water-soluble or no-clean fluxes); (4) product redesign (e.g., design for the environment); (5) product

Odors Prevention in the Food Industry

95

reformulation; (6) in-process recycling; (7) process or equipment modifications; (8) recycling and recovering wastes (e.g., biodiesel production from wastes). An example of a recycling option, for potential smelly residues of edible oils and fats, is to collect and reprocess them to produce biofuel. Previously, it is necessary to remove any water and/or water vapor bubbles present in the waste oil by heating the waste oil at 105 °C. From all the industrial processes two main products result: methyl esters, the chemical name for biodiesel, smelling like French fries, and glycerine (a valuable by-product that can be used in soaps and other products), with the additional advantage of the possibility of easy biodegradation. Smelly waste refuse of vegetable oils, used for frying, is an attractive source of biodiesel, but also is difficult to convert because it contains 2–10% free fatty acids (which are, precisely, the cause of the rancid odors and taste). 7.6.3. Preventive Managing Procedures Cleaner production aims to prevent pollution, reduce the use of energy, water, and material resources, and minimize waste, profitably and without reducing production capacity (Environmental Protection Agency, 2000; ICPIC, 2004; Pennsylvania State University, 2002; Robert et al., 2002). Strict inventory control, e.g., minimizing expired shelf-life materials in storage dependences and avoiding long periods of permanence of the products in the expedition spaces, is good practice. Pollution prevention processes are one of the major procedures of cleaner technologies. Cleaner technology main tasks are: 1. 2. 3. 4.

Energy conservation. Water conservation. Maintenance of the equipment in good repair. The cleaning and sanitizing procedure prevents the build up of dirt and debris. Cleaning and sanitizing must address three basic set points: exterior facility access, pavements, grounds, and connections to sewers; internal facility and utilities, including floors, walls, ceilings, and ventilation system; and drains and equipment.

As illustration, one of the important issues for many food industries is oil and grease residue handling, their removal or treatment systems, and final discharge or disposal devices; in fact, grease tends to solidify downstream, causing frequent blockages and/or backups. They also may cling to wastewater ducts and reduce their flow capacity in the long term. All degreaser equipment only separates oil and grease onto the surface of the bulk effluent and at best relocates it. By accidental overflow onto pavements, in many cases it remains as an oily film that poses a risk for slips and falls. Then, there is the problem with cleaning and disposal. On the other hand, oily waste cannot be dumped down the drain, because haul-off can be a very difficult and expensive consequence.

96

R. Nabais

Interior grease traps should be evaluated for cleaning needs on a weekly basis. Frequency of and proper maintenance is the only sure way to know if the traps are operating properly. Using hot water (120°C or higher) to wash or rinse, the grease stays emulsified, not allowing its separation. Many food treatment facilities require cooling jets that are activated to cool the water before entering the interceptors. Even if the grease amount is not much, as it cools it add and accumulates on the interceptor pipe walls and causes a plug. To really remove oil and grease, gravity separation may be used, provided the oil particles are large enough to float toward the surface and are not emulsified. If this is the case, the emulsion must be first broken by simultaneously decreasing of temperature and pH. Another special recommendation for food industries is to assure proper drains, and if odors develop, to reduce the amount of standing water of the pavements and floors and by changing water-based cleaning process to dry systems. In general, reducing biological solids content (soluble or insoluble) of waste waters produced by any food industry is always a good principle, to achieve potential pollution abatement and the unpleasant odors emissions resulting from their uncontrolled biological degradation. 7.6.4. Planning Rigorous Maintenance Proceedings Inspection and maintenance programs embody the idea that some organizing procedure mechanism must be put into place to ensure that all installations and equipment must be kept in nonpolluting condition. Measuring resource efficiency is the key to making incremental improvements in performance. Examples of environmental performance indicators, for companies, are the amount and the type of energy used, carbon dioxide emissions, water consumption, effluent discharges, waste management, and environmental incidents registers. Avoid leaks from gaskets and seals. Spill, splash discharges, and leak prevention is one good tool to prevent odors emissions, while usually also meaning economical gains (e.g., these interventions must take part in good practice checklist procedures of any preventative detailed maintenance plan). 7.7. Special Measures : Developing Correct Urban Development Plans and Adequate Layout Design of Buffer Areas Proper land use planning avoids allowing incompatible land uses in close proximity and is one of the most important tools in odor management. Many odor problems can be avoided by appropriate placement of new facilities. The lack of consistent methods to deal with odors prevents authorities from establishing policies and standards, which would eliminate, or at least minimize, community odor nuisances [Danish Environmental Protection Agency (DEPA), 2002; New South Wales Department of Environment an Conservation (NSW),

Odors Prevention in the Food Industry

97

2004; UNEP Working Group for Cleaner Production in the Food Industry (UNEPTIE), 2004a,b; Robert et al., 2002]. According to ZERI (Zero Emissions Research Initiative, of United Nations University, 1994), in the future industries will be organized into clusters in a way that waste from one industry becomes input for other industries, forming an integrated system. This simple, genial concept imitates natural ecosystems, where no real wastes exist. Industrial ecology shares with ZERI by anticipating the environmental consequences of production, consumption, and waste management, integrating overall balanced activities for ecoindustrial parks and product life-cycles, “taking a systems perspective on the environmental consequences of production and consumption” (Lifset, 1999; Robèrt et al., 2002). It also is urgent to ensure compatibility of industrial activities with adjacent land uses, adequate separation distances to allow for process upsets, and adopt the principles of waste minimization, cleaner production, and best practice control technology. These objectives-oriented interventions will minimize or eliminate the release of odors from the site, strictly oriented by the specificity of the industrial activity (UNEPTIE, 2004a, b). Avoiding unregulated development zoning serves principally to protect property owners from the negative externalities (social costs) of new developments. The zoning plans are based on the following considerations: ●

●

● ● ●

Designated land use of the site and the surrounding land under the district plan. Confine location of activities within the site and their orientation in relation to prevailing winds. Consider sensitivity uses of the downwind receptors. Respect buffer distances from the site boundaries to sensitive land uses. Use screening, such as by earth bunds, shelter belts, or natural topography, that functions as obstacles for odors transference.

Odor is transported primarily by dispersion and diffusion. Dispersion is the gradual dilution of the odor by mixing it with ambient air, transporting it with the bulk flows assisted by the wind. Once odors are released, they are transferred into the surrounding airspace, where air currents and turbulence disperse and dilute them. Besides its concentration, the noxious effect of odorous gases in the air depends if they are in mixtures, its composition, moisture content, temperature, pH, oxygen concentrations, and weather and environmental conditions (season, wind patterns) wind direction, wind speed, weight of gas and adsorption to dust particles. Ammonia (NH3) is lighter than air; results from anaerobic or aerobic activity; and is soluble in water. Hydrogen sulfide (H2S) is heavier than air; it presents a low odor threshold and a high solubility in water; one serious problem of this

98

R. Nabais

gas is its toxicity increases with higher concentrations, while simultaneously it loses its odor character. Weather patterns, humidity, and temperature largely determine odor transport and detection, all of which can change with the season, the day, or even the time of day. Warm temperatures and high humidity increase the potential for odor nuisances, while cold, dry conditions reduce the potential for nuisance complaints. Odors disperse better under warm, windy conditions compared with still conditions. Odor generally does not disperse much early in the morning or early in the evening, particularly in cool weather. Odorous flow also tends to drain down the valley (katabatic drift) from areas of higher altitude and/or higher pressures to areas of lower altitude and/or lower pressures (New South Wales Department of Environment and Conservation, 2004). Considering layout implementation, effective odor control depends on four factors: 1. 2. 3. 4.

good planning and good industrial management; good relations and communication with neighbors; appropriate control and treatment technology selection; and sensitivity of the region.

New facilities should be carefully sited so that prevailing winds do not carry odors to sensitive neighboring sites. However, no setback distance will work for every site. Setback requirements are site-specific and should be calculated using complete information about local conditions. For example, setback calculations should include the size and type of operation, the facilities covered by the setback (such as lagoons, fields, and houses), as well as the proposed sites and methods for applying wastes to land. Local vegetation, prevailing winds, weather patterns, and neighboring land uses also should be considered in the calculations. An efficient planning option is confining odoros activities to certain districts (by effective zoning regulations). Examples of sensitivity activities are residential dwellings, visitor accommodations (hotels, bed and breakfast, guesthouses), hospitals or nursing homes, schools, churches, holiday and weekend dwellings, shelters, campsites, caravan parks, and sports facilities. Figure 6 gives an idea of how an emission source can have a favorable layout previously roughly settled, by considering its location related to the relative frequency of dominant wind-flow directions. The setback distance, or buffer zones, of implantation of the facilities must considerer not only the frequency and intensity and dominant wind, but also the exposure angle of the facilities (Heber, 1997; Schauberger and Piringer, 1997; Pennsylvania State University, 2002). Several European countries have established such setback models, such as the known Austrian guideline that is based on quantitative odor evaluations and neighbor surveys and is a tool to assess airborne emissions resulting from livestock husbandry and the pollution caused by these emissions.

99

Odors Prevention in the Food Industry N 100

NW

80

NE

60 40 20

W

E

0

SW

SE S

Figure 6. A hypothetical way to draw a wind pattern relative to frequency and regional vicinities borders (….), of a well-located emission site related to wind-orientation dominance, spotted in the center of the winds-rose.

Percentile lines are commonly used to describe the odor concentrations on a map of the surroundings of the plant. Percentile lines (or isoconcentration lines) connect places in a contour plot at which a certain threshold concentration is exceeded, during a certain percentage of the year; this information may be used to indirectly estimate odor concentrations, which will be mentioned in the next chapter. Properly planted and developed, dense vegetable curtains can filter dust particles and reduce or modify wind direction. Fast-growing evergreen trees have good windbreaker properties in buffer zones, around an odorant unit, and help disperse odors. The planting of trees around facilities that emit odorous or VOC compounds is currently being used for odor abatement. Trees will help to a small extent to capture the odorous compounds and particularly the dust. On the other hand, vegetation curtains provide a visual barrier that reduces odor perception for residents and also, under adequate cold weather, may serve as a condenser element for organic vapors. Technically, one must definitely realize that tree barriers are not the solution, but up to a point they contribute to soften the opinions about food industry activities as being dirty, smelly, noxious and undesirable. Besides if they are chosen among pleasant odorant species, their odor may hide or overcome unpleasant smelly emissions. Organic matter decomposes anaerobically in standing water, releasing odors when disturbed or tracked onto public roads. The landscape should be properly graded and maintained so that water does not stand in access roads and around production facilities. An interdisciplinary team should establish these criteria in cooperation with industry, local communities, and the appropriate environment policy authorities.

100

R. Nabais

Using geographic and meteorological data, maps can be drawn up to identify levels of exposure and predict odor annoyance impacts on the residents near those areas. For each of the currently available management options described in this chapter, performance criteria should be developed and data collected so that management improvements can be recommended. In summary, odor management strategies are shifting their approaches, from old systems—complaint driven—to the new systems based on anticipation measures and precautionary inspection systems. 7.8. Using Conventional Plume Dispersion and/or Virtual Chimneys Odors are often low-density gases. Once released into the environment they are transported by wind and diluted and dispersed by atmospheric turbulence. According to theoretic common concepts, odorous gases require special consideration and treatment. Warren Spring Laboratory in the United Kingdom (Williams and Thomson, 1985) suggests the following technique to give rough estimations of the uncorrected chimney height hu (meters), by one of two alternatives: hu = (0.1DQ)0.5 where ●

●

D is the number of dilutions or odor units required to disperse the emission source to the TOC50%, and Q is the volumetric flow rate in m3/sec at 0°C and 760 mm Hg (1 atmosphere standard, or 1.013 bar)

Another equivalent expression is: hu = (0.1Mo/TOC50%)0.5 where ● ●

Mo is the mass emission rate of the odorous gas in g/sec. TOC50%is the result of an odor panel test giving the number of dilutions (odor units) required to disperse the odorous source to levels below the detection threshold odor concentration, for 50% of the odor panel.

It is worth a reminder that these two equations were empirically developed; therefore, they must be applied only respecting the units and the dimensions of all variables. The detection threshold levels normally considered are 5, 10, 20, and 50 OU. The level 5 OU is considered to be the minimum for detection in the open environment, where inhabitants are subject to a wide range of natural odors and consequently these levels correspond to worst-case situations. The 10 and 20 OU are levels that, if occurring frequently enough, may be sufficiently detectable to

Odors Prevention in the Food Industry

101

constitute a nuisance. Finally, investigation of a 50 OU detection threshold gives information about the extent of high levels of impact. It is generally not practical to disperse more than 200 OU/m3 per sec of exhaust flow without correctly designed chimneys. Anyhow, dispersion of odorous emissions via a chimney is not recommended, unless a well-trained odor panel previously evaluates the emission source. Wind is responsible for the rapid horizontal transport of humidity, warm air, pollutants, and odors, while turbulence is responsible for vertical transport. Wind turbulence can be visualized as eddies of different sizes that cause fluctuations in concentration over short time intervals. Internal concentration variations within a plume are created by puffs of air. The turbulence caused by these eddies diffuse the plume. Large eddies cause meandering of the plume near the source, normally in a lateral direction. As the plume is torn by the introduction of pure air puffs, a concentration profile develops within the plume. Rough terrain, valleys, and other topographical features can increase the complexity of airflow patterns. For the usually very low concentration of odorants pollutants in the food processes emissions, and because if odor criterion needed for the analysis is of the order of seconds, then the time-averaged formulas could underestimate a shorter-term peak odor effect. Most odor sampling, however, is also time-averaged to collect sufficient sample for analysis. This makes validation of proposed averaging-time adjustments difficult. Straight standard Gaussian plume model does not accurately simulate atmospheric pollutant transport over rough terrain profile, irregular slopes, and surface singularities, as vegetation or protective fences effects. If advisable, but never if potential health nuisance consequences are previewed, enhanced dispersion is a cost-effective option, which may be achieved by increased exhausting stacks height, increased exhausting gas velocity, or moving the source away from critical reception spots. A promising proprietary technology/device that deserves close observation, both in terms of theoretic principles and technologic instruments, is “extendedstacks” or virtual chimney: a wind-based odor diversion system, simulating the natural transport phenomenon caused by the wind. The odorous gaseous mass is blown vertically through the atmospheric layers while diluting the odors (Finn et al., 2004). This curious proprietary patented system theoretically seems to work beautifully, by the simplicity of the basic theory: when one increases longitudinal transfer velocity in a “flowing wind-polluted tube,” the pressure inside the axe of the “emission wind tube” will decrease, according with Bernoulli equation. The horizontally pointed outer ring of fans blows inward at a high velocity, creating an internal low-pressure zone. The atmospheric air, from the outside environment of “the windy polluted tube,” attempts to equalize this internal underpressure of the windy tube, so the fresh air continuously moves toward the center low-pressure zone. The fresh air mixes with the odorous molecules, diluting them. The diluted air stream is ejected vertically upward, by one vertically pointed fan in the center. This continuous process not only (potentially) causes

102

R. Nabais

massive dilution, and simultaneously “virtually increases a fluid stack height” (to discharge odorous gas stream as far as possible, from the ground level). Enhanced spatial dilution of odorants, as the technique just described, must not be considered to be a “miraculous” solution to any air pollution problem, especially if it is considered to be a nuisance. If pollutant concentrations meet all emissions regulations and the legal compliance requirements are met, but the emissions remain a malodorous nuisance to the surrounding neighborhood, further dilution of the pollutants should be sought.

8. CONCLUSIONS By accomplishing environmental management system integrative principles or at least all achievable pollution prevention procedures, industries and communities may prevent expressions of pollution, including odor disturbances, thus minimizing the need and the costs for control, treatment, and disposal. Indeed, these are still the best options to rethink traditional industrial processes; hence, we must support environmental policies that require creative ecoefficient strategies to address the continuous improvement, sustained development, necessity of organization’s public commitment to environmental compliance and protection, and ultimately, the assurance of our own quality of life.

9. AKNOWLEDGMENTS The author deeply expresses gratitude to: Doctor Xavier Nicolay, of Institut Meurice–UBT, for his strong partnership conscience, team responsibility and consistent coordination, in putting together this “many-hands–multicultures knitted book”; but, specially, for his kind understanding, restless support, caring spirit, endless patience, attentive readings, helpful and valuable advices about this chapter content. However, copyrights remain the author’s responsibility for all the still remaining imprecision or errors.

10. REFERENCES Anonymous (2002) Cleaner production tools: LCA and beyond. Editorial. Journal of Cleaner Production 10 403–406. ASHRAE (2003) HVAC Applications Handbook. (IP). Item 210–8103-03. Publisher: ASHRAE. ISBN: 1–931862–22–2. ASHRAE (1999) Ventilation for Acceptable Indoor Air Quality, Standard 62–1999, Atlanta, GA, American Society of Heating, Refrigeration, and Air Conditioning Engineers. Collins, R. (2000) CSPI PROJECT “Science Conflicted: Restoring Trust in the National Academy of Sciences”. Science Conflicted: Restoring Trust in The National Academy of Sciences”, Integrity in Science Project, Center for Science in the Public Interest. Baltimore sun Journal. Editorial of August, 2000. Danish Environment Protection Agency (DEPA) (2002) “Industrial odour control.” Environmental Guidelines No. 9 2002. Vejledning fra Miljøstyrelsen Environmental Guidelines no. 9, 2002. http://www.mst.dk/udgiv/publications/2002/87-7972-297-0/html/helepubl_eng.htm

Odors Prevention in the Food Industry

103

Environment Agency for England and Wales in collaboration with the Scottish Environment Protection Agency (SEPA) and the Northern Ireland Environment and Heritage Service (NIEHS) (2002a). Technical Guidance. Note IPPC H4. “Integrated Pollution Prevention and Control – IPPC DRAFT. Horizontal Guidance for Odour.” Part 1 – Assessment and Control. www.environment-agency.gov. uk/commondata/105385/ h4_odour_part1.pdf. Environment Agency for England and Wales in collaboration with the Scottish Environment Protection Agency (SEPA) and the Northern Ireland Environment and Heritage Service (NIEHS) (2002b) “Technical Guidance. Note IPPC H4. Integrated Pollution Prevention and Control (IPPC) DRAFT. Horizontal Guidance for Odour Part 2 – Assessment and Control.” www.environment-agency.gov. uk/ commondata/ 105385/ h4_odour_part2.pdf. Environment Protection Authority, Victoria, Australia (2004) “Strategies For Promoting Cleaner Production In The Food Industry.” Rowan Mackenzie and Scott Hamilton. http://www.uneptie. org/pc/cp/library/training/howtoCPC/manual_cdrom/CPlinks/pdfs/Hamilton.pdf. Environmental Protection Agency (2000) “Biosolids and Residuals, Management Fact Sheet, Odor Control in Biosolids Management,” EPA 832-F-00–067, United States, 2000. http://www.epa. gov/owmitnet/mtb/odor_control-biosolids.pdf. Ericsons Naturnomics Company (2004) “Odor Answers. Stack Extenders.” http://www.eric-sons. com/ne4/odoranswers/stackextenders/index.htm. Erwin, T.H., Vinka, Rabago, K.R., Glassner, D.A., Gruber, P.A. (2003) “Applications of life cycle assessment to NatureWorks polylactide (PLA) production. Polymer Degradation and Stability.” 80: 403–419. www.cargilldow.com/corporate/life_cycle/docs/PLA_Article.pdf. 2003. European Commission Directorate-General Jrc. Joint Research Centre. Institute for Prospective Technological Studies Technologies for Sustainable Development. European IPPC Bureau. (2003) “Integrated Pollution Prevention and Control, BREF (Best available techniques REFerence document) on the Slaughterhouses and Animal by-products according to Article 16(2) of Council Directive 96/61/ EC. 2003.” http://www.umweltdaten.de/nfp-bat-e/ slaughterhousesbref-e.pdf. Gesamtverband der Deutschen Versicherungswirtschaft e.V. (GDV, German Insurance Association) (2003) Wild, Y. “Refrigerated containers,” in Chapter 7. Container Handbook. 2003. http://www. containerhandbuch.de/ Heber, A.J. (1997) “Setbacks for sufficient swine odor dispersion and dilution.” Paper presented at the Livestock and Environment Symposium, University of Nebraska Cooperative Extension Service, New World Inn, Columbus, Nebraska, December 10–11, 1997. Modified Dec. 16, 1997. http://pasture.ecn.purdue.edu/~heber/setba.htm. Holland, M., Hunt, A., Hurley, F., Watkiss, P. (2004) Draft final methodology paper for “Service contract for carrying out cost-benefit analysis of air quality related issues, in particular in the clean air for Europe (CAFE) programme.” 5/ 07/ 04 AEAT/ ED51014/ Methodology Paper Issue 3. http://europa.eu.int/comm/environment/air/pdf/cba_methodology_issue3.pdf. Institute for Prospective Technological Studies Technologies for Sustainable Development. European IPPC Bureau. European Commission Directorate-General Jrc. Joint Research Centre. 2003. “Integrated Pollution Prevention and Control, BREF (Best available techniques REFerence document) on the Slaughterhouses and Animal by-products according to Artical 16(2) of Council Directive 96/61/EC.” International Cleaner Production Information Clearinghouse (ICPIC) (2004) Environmental Management Center. Technical Case Studies. Manufacturing. “Manufacture of Food Products and Beverages.” http://www.emcentre.com/unepweb/tec_case/ . Lifset, R. (1999) United Nations University. Zero Emissions Forum. The Industrial Ecology of Renewable Resources. ZEF-EN-1999–2. http://www.unu.edu/zef/publications-e/ZEF-EN-199902-D.pdf Mohunta, D. M. (2004) “Recovery of volatile organic compounds from gaseous streams.” http:// www.ccdcindia.com/index2.php?act=voc&subact=vocpaper. Commercial, Chemical and Development Company. New South Wales Department of Environment and Conservation (NSW) (2004) “Odour control.” http://www.epa.nsw.gov.au/mao/odourcontrol.htm

104

R. Nabais

Ontario Ministry of Agriculture, Food and Rural Affairs and Ontario Ministry of Environment and Energy (1995) “The Guide to Agriculture Land Use and Minimum Distance Separation.” http:// www.gov.on.ca/OMAFRA/english/landuse/facts/guidealu.pdf. Pennsylvania State University in cooperation with the Pennsylvania Department of Agriculture (2002) Odor Management in Agriculture and Food Processing. Odor Management in Pennsylvania. A Reference Manual. Revised printing June, 2004. http://server.age.psu.edu/fac/Elliott/ EntireOdorManual.pdf. Pinchin Environmental Newsletter (2004) Ventilation & Emission Control Group. http://www. pinchin.com/newsletters/air/odourcomplaints.htm. Portary (2004) Recolha e Transporte de Resíduos Industriais Banais Perigosos. http://www.portaryresiduos. pa-net.pt/html/gest.htm. Realff, M.J., Abbas, C. (2003) “Industrial symbiosis. refining the biorefinery.” Journal of Industrial Ecology. Vol. 7, Issues 3–4. Robèrt, K.H. (2002) Schmidt-Bleek, B., Aloisi de Larderel, J., Basile, G., Jansen JL, Kuehr R, Price Thomas P, Suzuki M, Hawken P, Wackernagel M. (2002) “Strategic sustainable development — selection, design and synergies of applied tools.” J Cleaner Prod.; 10(3): 197–214. Schauberger, G., Piringer, M. (1997) Guideline to assess the protection distance to avoid annoyance by odour sensation caused by livestock husbandry. Proceedings of the Fifth International Livestock Environment Symposium, May 29–31. 1997. Society of Heating, Refrigerating and Air-Conditioning Engineers (1997) “Handbook of Fundamentals.” http://www.containerhandbuch.de/chb_e/ wild/index.html. Starkey, R. (1998) Centre for Corporate European Environment Agency. (CCEM). Internet Gateway for Cleaner Production, Pollution Prevention and Sustainable Business. “Environmental management, Tools for SMEs: A handbook.” Environmental Issues Series. Edited for the EEA.1998. http://reports.eea.eu.int/GH-14-98-065-EN-C/en/enviissu10.pdf. Sustainable Agri-Food Production and Consumption Forum (2004) UNEP Publications. United Nations Environment Programme Division of Technology, Industry and Economics. “Industry Sector Guide Lines.” Ref. UNEPTIEb, 2004. Tybor, P.T., Hurst, W.C., Reyknolds, A.E., Schuyler, G.A. (1995) Quality Control. A Model Program for the Food Industry. University of Georgia College of Agricultural & Environmental Sciences. Cooperative Extension Service, Extension Food Science. http://pubs.caes.uga.edu/caespubs/ pubcd/b997-w.html. US Food and Drug Administration (2004) Federal Anti-Tampering Act U.S.C. Title 18 – “Crimes And Criminal Procedure. Part I – Crimes Chapter 65 - Malicious Mischief. § 1365. Tampering with consumer products.” http://www.fda.gov/opacom/laws/fedatact.htm. US Environmental Protection Agency (1995) Technology Transfer Network. Clearinghouse for Inventories & Emission Factors Compilation of Air Pollutant Emission Factors, “AP-42, Fifth Edition, Volume I: Stationary Point and Area Sources.” Research Triangle Park, NC. 1995. UNEP Working Group for Cleaner Production in the Food Industry (2004a) “A Guide to Cleaner Production in the Food Industry.” http://www.geosp.uq.edu.au/emc/cp/Res/guides/GUIDE1. HTM to http://www.geosp.uq.edu.au/emc/cp/Res/guides/GUIDE7.HTM. UNEP Working Group for Cleaner Production in the Food Industry (2004b) “Fact sheets.” http:// www.geosp.uq.edu.au/emc/CP/Res/xfactsheets.htm. The University of Queensland and The United Nations Environment Programme. United Kingdom Renderers Association (UKRA) (2004) http://www.ukra.co.uk/index.htm. United Nations Environment Programme — The UNEP Working Group for Cleaner Production in the Food Industry (2004) “A Guide to Cleaner Production in the Food Industry.” http://www. geosp.uq.edu.au/emc/cp/res/yfood_manual.htm. (access on Wednesday, August 25, 2004). Van Harreveld, A.P. (2001) “From odorant formation to odour nuisance: new definitions for discussing a complex process.” Water Science & Technology. Vol. 44. No 9. pp 9–15 © IWA. Williams, M.L., Thompson, N. (1985) “The effects of weather on odour dispersion from livestock buildings and from fields.” In: Odor Prevention and Control or Organic Sludge and Livestock Farming. V.C. Nielsen, J.H. Voorburg, P. L’Hermite. Elsevier Applied Science Publishers, New York. pp. 227–233.

7 Odors Treatment: Physicochemical Technologies Regina Nabais

7.1. INTRODUCTION Taking into consideration simple economic factors and/or social responsibility, nowadays, every industry tends to analyze its productive organization through total quality management, which includes environmental and safety issues such as avoiding annoyances or nuisances by odor emissions; in this, the food industry is no exception. However, even though many efforts have been developed to prevent and control odorous emissions at the production unit level, these measures usually are not enough to assure good quality solutions for all kinds of odor problems. When this happens, effective treatment or abatement measures must be considered. This chapter’s scope is to describe the main technologies available to reduce and control volatile organic compounds (VOC) and malodorous emissions from industrial processes, especially those of interest for the food industry [Environmental Protection Agency (EPA), 1995]. The main concepts, principles, fundamentals, advantages, and disadvantages of each technology will be generally referred and discussed.Indeed, controlling odors is an important consideration for protecting the environment and the community’s good will, but it is worth remembering that operational improvement usually requires strong competencies, capabilities, and experienced technological assistance. When deciding technological investments, technology isn’t the solution; it only enables people to find good answers for solving problems. Among all technologies, simplicity and adequacy must always be our priority; in accord with the philosophic and scientific principle of Occam’s razor—“Pluralitas non est ponenda sine necessitas”—everything should be made as simple as possible, but not simpler. REGINA NABAIS ● CERNAS, Centro de Recursos Naturais, Ambiente e Sociedade, Escola Superior Agraria de Coimbra, Instituto Politecnico de Coimbra, Bencanta, 3040-316 Coimbra, Portugal 105

106

R. Nabais

7.2. BASIC PRINCIPLES UNDERLYING ODOROUS EMISSIONS TREATMENT AND/ OR ABATEMENT Despite the complexity of the problems of handling off-odors, there are only two ways of management of their emissions: (1) source control, by selecting appropriate industrial processes able to reduce emissions in the source, and (2) treatment of the final resultant emissions. When applied to industrial processing, the goal of cleaner production, is to avoid pollution generation in the first place,which frequently reduces risks, cuts costs, and identifies new business opportunities. Cleaner production aims to reduce waste and inefficiency at the source and also can help to develop the most efficient ways to operate processes, produce better products, and provide new services. Adequate control equipment and effective treatments may be based on capture recovery (nondestructive technologies) or on destruction. Some treatment technologies use combinations of both capture recovery and destruction. Criteria for the selection of the best choice for odor control and treatment must be carefully chosen to assure their effectiveness at sustainable costs. Even so, whatever may be the selected treatment, it will require high initial cost to install, efforts to implement, and careful planning to start up and maintain, and together with the important capital investments will require careful criteria that considers heat conservation/recovery. Large airflow rates with low volatile pollutant concentrations frequently are encountered in food industry processes, considering VOC and odors emissions. Regardless of the selected option to deal with off-odors, rigorous, careful, and strict maintenance programs will need to be designed and implemented, especially if nonthermal emission control devices are chosen. Usually, all techniques need replacement of costly consumables (i.e., adsorbents, reactants, fuel etc.) and require energetic provisions to regenerate subsidiary materials (e.g., adsorbents) and storage of spare parts. Consideration also must be given to the legal requirement of using either the “best available control technology” (the best possible choice of competing processes for reduction of a specific pollutant given economic, chemical, and physical constraints), or the “maximum available control technology” (the process that produces the greatest emissions reduction, regardless of all constraints), or “reasonable available technology” ( a balanced process that considers equal concerns to abatement technology that is more reasonable, economical, and proven in practice), or even the “best available technology not entailing excessive costs” (best at preventing pollution and available in the sense that it is procurable by the operator of the activity concerned, including a balance between environmental benefit and financial costs) [Integrated Pollution Prevention and Control (IPPC), 2003]. However, before one starts to identify and choose any suitable technology for air pollution control, including odor abatement, there are variables and procedures

Odors Treatment: Physicochemical Technologies

107

to consider: nature, concentration, physical, and chemical properties of the pollutants to remove, considering corrosiveness, abrasiveness, toxicity and flammability, temperature, relative humidity, pressure, flow rate, and particles content of the emissions; overall appraisal of the process efficiency, regarding compliance with regulations, legislation, recommendations, and guidelines; feasibility and economic interest of recycling or recovery of the pollutants compared with final disposal costs; power and subsidiary requirements; foreseen initial and operational costs; and finally, planning convenient training programs for the personnel who will operate the system. This chapter presents the most important VOC abatement technologies adaptable to odors treatment, especially in the food industry. All available treatment technologies are effective to removing odorous compounds when correctly applied, but each has specific performances and constraints inherent to the process. In many situations, employing different technologies in a sequential configuration provides synergetic effects, while enhancing the best of each technology. Operational plants confirm that for waste airstreams of food industrial process affected by unpleasant odorant pollutants, such as amines, mercaptans, aldehydes, etc., to be treated at an acceptable cost, enough to obtain sufficiently low pollutant levels of outlet gas, require sequences and selected combinations of more than one of the aforementioned technologies. Examples of common technologies used for industrial proposes including food processes are listed in Table 7.1 [New South Wales Department of Environment and Conservation (NSW), 2004].Because of their low operating costs, if properly operated and maintained biological technologies are usually popular among food industries. These technologies are comprehensively described and discussed in Chapter 8, this volume.

7.3. MAIN TECHNOLOGIC TREATMENT SYSTEMS FOR GASEOUS POLLUTANTS Whenever odor emissions cannot be prevented, containment and often some form of treatment are followed by release to the atmosphere. Reliance is placed on sufficient dispersion taking place before sensitive receptors are reached. Odor control methods may be generally classified as: nondestructive or destructive. Nondestructive methods can be very effective in reducing gaseous emissions from process stream vents, but they do not really destroy the VOC or odorant pollutants. Non-destructive methodologies are advantageous if the recovery substances are valuable to the industry itself and the recovery process is located near the points of generation, e.g., recovery of ethanol in food-grade production factories, at the headspace of fermentation closed devices (fermenters). However, if not carefully conceived, in most cases efforts and resources will be wasted as odor-causing pollutants only trade places, from one medium to another.

108

R. Nabais

Fermentation

Grease trap waste Amines

Best management practices. Cleanliness and avoiding spills requires human effort but is relatively inexpensive.

Amines Aldehydes, amines

Condensers. Condensable liquids removed from the exhaust gas stream, reduces the amount of odorous gas to be treated.

Amines

Aldehydes, amines

Alcohol

Slaughterhouses Ammonia

Animal incineration Aldehydes, amines

Fish meal

Chicken feathers

Amines, aerosols Amines, aerosols

Amines

Amines

Carbon adsorption. Needs regenerating at regular intervals. Can be effective but expensive for large air volumes. Small operations are reasonably inexpensive to deal with.

Aldehydes, amines

Thermal Oxidizer (catalytic). Temperature 400°C. Lower temperature operation than direct method, but the catalyst may be destroyed if not operated and maintained correctly. This is frequently a problem.

Amines

Thermal oxidizer (direct). Temperatures between 600°C and 1000°C with residence time of 0.3 to 1 second. Doubling of residence time may enable afterburner temperature to be reduced 20°C to 100°C. Requires careful design to reduce air volume to a minimum. Capital cost high and cost of running is high for large air volumes. Needs further control if sulfur or chlorine present in exhaust gases.

Aldehydes, amines

Amines

Wet scrubbing. Absorption: moderate to high costs, three stages often required. Needs careful selection of scrubber liquor and usually trials. Not always successful, requires regular maintenance and daily tests of active agent and pH control in some cases.

Amines, aerosols

Coffee

Rendering industry

Boiler

Table 1. Examples of Odor Control Technological Options for Food Process Industries

Destructive methods are those, when properly operated, that are not producers of secondary organic waste—they destroy the pollutants or contaminants by oxidation of pollutants or malodorous compounds. Table 7.2 shows the technologies commonly used to control VOC emissions, which may be easily adapted to food industry odor emissions reduction and control. Table 2. Odor and VOC Control Technologies Destructive technologies

Nondestructive technologies

Biofiltration; thermal oxidation (catalytic, regenerative, recuperative)

Absorption; adsorption; condensation

Odors Treatment: Physicochemical Technologies

109

7.3.1. Absorption In general, absorption is the transference of one substance into or through another of a different state. When applied to air pollution this technology means that gaseous components are transferred from a gas stream to a liquid phase. The liquid phase may be water or aqueous solutions of chemicals that react with the absorbed components. According to the adequacy of their application and maintenance, the efficiency of these systems may be in a range of 70 to over 99% [United States Corps of Engineers (USACE), 2004]. Absorption commonly refers to the intimate mixing of a gaseous exhaust stream with a liquid. The liquid must dissolve or absorb specific constituents of the stream. Wet scrubbers are examples of absorbers, whereby the exhaust stream is routed through a vessel and forced into contact with a circulating liquid solution. Low cost and space requirements make this technology attractive for odor control under certain circumstances; secondary treatment or disposal of the effluent, depending on the gas stream constituents, corrosion, and fouling, may impact maintenance costs. This technology is particularly efficient for the treatment of water-soluble odorous compounds like ammonia, amines, ketones, acetates, etc. The liquid phase can be either an aqueous solution (acidic or alkaline) or a solvent solution, depending on the solubility of the compounds. Chemical oxidation is carried out either in a liquid phase or in a gas phase. The most commonly used oxidizing agents are sodium hypochlorite (NaClO) and hydrogen peroxide (H2O2) although other oxidants can be used (EPA, 1995). Both have advantages and disadvantages: sodium hypochlorite is cheaper but the scrubber effluent contains chlorine and chloride. Hydrogen peroxide has no spent product but is more expensive. There are several types of equipment designs for contact phases. Scrubbers include packed beds, spray towers, tray towers, absorption towers (see Figure 7.1), Venturi scrubber, and static mixers. Most of these configurations also require the installatation of mist eliminators before the treated air is released into the atmosphere. Tables 7.3 and 7.4 list general aspects of air pollutant treatment system, based on absorption technology and the major advantages and disadvantages of this technology, respectively. Contaminants transferred to the liquid phase are removed from the scrubber by continuous or periodic overflow of blowdown. All chemical scrubbers are designed for a large contact area between the gas and liquid phases. Odor control, however, requires some dedicated design features. In particular, the air residence time in the scrubber must be sufficient, typically around 3 seconds. The scrubber and related ductwork must be cleaned to remove gelatinous substances and milk stone deposits, especially if this option is considered by dairy industries.

110

R. Nabais

Table 3. General Characterization of Absorption Performance factors

Malfunction conditions

Temperature pH Gas-liquid contact Adequate liquid flow rate Adequate chemical solutes additions

Inadequate liquid flow Inadequate pH control Inadequate chemical selection or addition Excessive temperature of the liquid Plugged beds or mist eliminators Corrosion

Controls of the systems and usual checklists Temperature gradient between inlet and outlet of the gaseous stream Liquid flow rates and pressures Gaseous effluent opacity

From Crowder (2004).

Table 4. Major Advantages and Disadvantages, of the Absorption Technology Process Asorption scrubbers

Advantages Highly efficient removal of odorous compounds (up to 99.0% depending upon system design and chemicals employed.); Small footprint compared to most equivalent odor control technologies (face velocities can approach 120 meters per minute); Ability to handle very high H2S loadings (upwards to 200 ppm); It even treats fumes on line; Increased relative velocity between scrubbing the fluid and gas stream, increases efficiency for solids.

Disadvantages Requires the use of large quantities of chemicals; Chemicals most commonly used (sodium hydroxide and sodium hypochlorite) are expensive and hazardous to handle; Treatment of multiple compounds generally requires multiple stages and different chemicals; For H2S removal, improper handling of the recycled liquid stream can result in liberation of H2S; High pH operation causes scaling (fouling formation/ precipitation of solids) of the packing in packed tower systems; Scrubbers are complex systems requiring frequent maintenance. Many sensors and moving parts that need replacement or calibration; High pressure drops required; Internal plugging, corrosion, erosion; Increased need for internal inspection; Gas and liquid chemistry control important; Particulate matter is not recycled; High corrosion potential; Liquid waste stream; High energy consumption.

Wet scrubbing using packed towers has proved to be the most reliable means of removing odorous pollutants and odor reduction, handling, for example, a wide variety of sulfides in inlet concentration. Alkaline (NaOH) scrubbing can achieve highly efficient removal of hydrogen sulfide (H2S) but the solution pH must be maintained at a high level to prevent release of the H2S. In most cases an oxidant such as NaOCl is added to prevent this from occurring. Highly volatile sulfides such as dimethyl sulfide (DMS) and dimethyl disulfide (DMDS)

111

Odors Treatment: Physicochemical Technologies

Mist eliminator

Outlet gas

Packing material

Scrubbing liquid distribution

Make up water or solution

Inlet gas

Figure 1. Sketch of an absorption tower; wet scrubber system.

typically require two stages of scrubbing to emsure removal below detectable odor thresholds. Design and operating parameters include scrubber geometric shape, liquid spray or injection locations, gas residence time, gas velocities, gas and liquid temperatures, gas and liquid pressure drop, and liquid–gas flow rate ratio. Air flow rate and or residence time are important parameters to get an initial idea of the required (1) footprint of the equipment, (2) admissible inlet pollutant concentrations, and (3) overall removal efficiency. The order of magnitudes that we may expect are: for air velocity through a chemical wet scrubber washer, 2–3 m/s; Venturi scrubbers operate at 45 to 120 m/s. By comparison, traditional activated carbon systems operate between 0.25 a 0.5 m/s, and a typical biofilter operates much closer to 0.025 m/s. For rendering industries, multistage wet scrubbers using various scrubbing agents are the primary alternative to incinerators for high-intensity odor control. Wet scrubbers also generally are used to control whey dryer emissions in cheese manufacturing. 7.3.2. Adsorption Adsorption is a surface phenomenon where volatile compounds are selectively adsorbed on the surface of solids with large surface-to-volume ratios. There are

112

R. Nabais

two major classes of adsorption processes: chemical adsorption (chemisorption) and physical adsorption (physisorption). Physisorption is an adsorption in which the forces involved are intermolecular forces (weak forces); i.e., no significant change in the electronic orbital patterns of the species involved. In the physisorption, the adsorbed species is chemically identical with those in the fluid phase, so that the chemical nature of the compound is not altered by adsorption and subsequent desorption. Adsorption enthalpy in the physisorption is related to factors such as molecular mass and polarity, typically 5–40 kJ mol-1, while in the chemisorption it is related to the chemical bond strength, and typically 40–800 kJ mol-1. Usually, the kinetics of physisorption is also faster; however, there is no definitive distinction between the two systems, because intermediate cases are common. Because of adsorbent regeneration constraints, chemical adsorption is not very common in industrial air pollution control systems. Furthermore, when desorption is considered, after adsorption processes with higher activation energy, it requires significant activation energy to remove the chemisorbed species from the adsorbent surface. This is a costly operation because it must be performed under very high temperature or high vacuum conditions or by some suitable chemical treatment of the surface; therefore, comparatively chemisorption is a more expensive operation. One may choose between many adsorbent substances, such as activated alumina, activated carbon (the most common), synthetic resins, molecular sieves, peat, polymers, silica gel, zeolites, or simply soil. Larger molecules adsorb better than smaller molecules. Nonpolar molecules adsorb better than polar ones. Nonsoluble or slightly soluble molecules adsorb better than highly soluble molecules. Based on the polarity or solubility of the molecule being adsorbed, pH may have an influence on the extent of adsorption. Alcohols are poorly adsorbed, because they are very soluble and highly polar. Aldehydes are highly polar molecules, but as their molecular weight increases, the polarity decreases, and adsorbability related to alcohols is higher. Adsorption of amines is limited by polarity and solubility. Higher molecular weight organic compounds will generally be more adsorbable because adsorptive attraction increases with the size. Temperature increases the rate of diffusion through the stream phase to the adsorption sites, but since the adsorption process is exothermic, increases in temperature may reduce the degree of adsorption. This temperature effect is negligible in water treatment applications and ambient vapor phase applications. Additionally, if odor is the only environmental problem of the airstream, the malodorous constituents represent only a fraction of the total organic compounds that would be adsorbed by the activated media. Because of the large loading of other organic substances, media such as activated carbon would have to be regenerated in situ or frequently replaced to be regenerated elsewhere. Either way, sorbent regeneration is still a difficult and dirty task.

113

Odors Treatment: Physicochemical Technologies

In regenerative systems, the VOC is removed from the air or process stream and then is recovered by directly heating the adsorbent and using this energy to overcome the intermolecular attractive forces. The resulting vapor stream is condensed, allowing for recovery and reuse of the solvent (Calgon Carbon Corporation, 2002). Continuous and automatic processing of solvent-laden air is achieved using a minimum of two adsorber vessels, one of which is steam stripped while the other is adsorbing solvent onto the carbon bed. Low-pressure steam may be used to strip (remove) adsorbed components from the carbon. Tables 7.5 and 7.6 list general aspects of air pollutant treatment system, based on adsorption technology and the major advantages and disadvantages of this technology: Other considerations in carbon adsorption applicability include moisture content and temperature of the emission stream. As water vapor competes with pollutants for adsorption sites on the carbon bed, relative humidity levels must be kept under 50% for usual applications. Adsorption is a highly efficient technology to remove odorous compounds (up to 99.0% depending on system design and chemicals employed) and has the ability to handle very high H2S loadings (upward to 200 ppm). On the other hand, it sustains face velocities approaching 2m/s and has a very small footprint compared to the closest equivalent odor control technologies. Adsorption on activated carbon is useful for recovery of VOC with intermediate molecular weights. Smaller compounds do not adsorb well and larger compounds cannot be removed during regeneration, which typically is by steam stripping. The pollutants will remain in the bed after each adsorption cycle, ultimately requiring a replacement of the carbon bed. Activated carbon systems have interesting applications in removing VOC from relatively dry airstreams (Figure 7.2).

Table 5. General Characterization of Adsorption Technology Performance factors

Malfunction conditions

Temperature Pressure Organic concentration Contaminant molecular weight Moisture Presence of particulate matter in the system

Infrequent adsorption; Physical deterioration of adsorbent; Plugging of the adsorbent; High temperatures; High organic vapor concentration

From Air Pollution Training Institute (2004).

Controls of the systems and usual checklists Inlet gas temperature and flow rates; Pressure drop across the adsorbent; Regeneration frequency and cycle period

114

R. Nabais

Table 6. Major Advantages and Disadvantages of the Adsorption Technology Process

Advantages

Disadvantages

Simple, passive system design, with few moving parts and few maintenance concerns. Effective removal capacity of organics and low levels of H2S. Safe operation with no hazardous materials. Provides an effective second stage treatment to other technologies. Highly efficient removal of odorous compounds (up to 99.0% depending on system design and chemicals employed); Small footprint compared to most equivalent odor control technologies (face velocities can approach 2m/s); Ability to handle very high H2S loadings (upward to 200 ppm); High removal capacity for H2S; Simple maintenance procedures.

Activated carbon is applied to offgasses containing low VOC concentrations (less than 50 ppm). Is an interesting solution for small gas fluxes. Limited capacity for H2S that translates into frequent change-outs and poor economics; Media change-out is time consuming, labor-intensive, and dirty. Spent media must be landfilled or sent off-site for thermal reactivation. Careless system operation can produce a fire hazard. Reduced capacity for organics removal due to presence of impregnated substance; Caustic regeneration is costly and hazardous, and creates spent caustic disposal problems; Spent carbon cannot be thermally reactivated and must be landfilled; Media change-out is time consuming, labor-intensive, and a very dirty operation; Careless system operation can produce a fire hazard.

Adsorption

Adsorption1 Unimpregnated activated systems (carbons, activated alumina, silica gel) Adsorption 2 Caustically impregnated activated carbons

From Van Deuren et al. (2002); US Army Corps of Engineers (2004).

Polluted gas

Clean gas outlet

Blower

Adsorbent material

Figure 2. Sketch of an activated carbon adsorption system. Adapted from Croll-Reynolds Clean Air Technologies (1996).

115

Odors Treatment: Physicochemical Technologies

7.3.3. Condensation Condensation is a separation process in which gaseous components of a gaseous stream are removed by cooling down the gas stream, causing a changing state phase, converting vapor components to liquid, and then removing the liquid phase. Air-to-air or air-to-liquid heat exchangers are commonly used. The most common equipments are shell and tube heat exchangers (see Figure 7.3). The contaminant removal efficiency of the systems is about 90%, and is higher for higher vapor pressure of the contaminants. There are three types of condensers: 1. Conventional (–18 ºC to 4º C),using brine coolants. Conventional condensers are of two kinds: direct (mixing the condensed gaseous components with the coolant, increasing wastewater treatments or components recovery procedures) or surface condensers. 2. Refrigeration (–100 ºC) ,using common compressed industrial coolants. 3. Cryogenic (–195.5 ºC), using liquefied gases such as nitrogen and carbon dioxide. Condensation is typically combined with particulate removal equipment (i.e., mist eliminators, electrostatic precipitators, etc.) to separate the condensed vapor from the gas stream after which the effluent is reprocessed, reused, or disposed. Odor removal efficiency of condensation systems should be cautiously considered since odors can be associated with gaseous phase constituents that are not readily condensed. Depending upon the nature of the exhaust gas stream, fouling of the heat exchangers and maintenance may become problematic. Finally, since condensate is created or is a valuable product, its disposal or recycling costs must be considered. Tables 7.7 and 7.8 list general aspects of air pollutant treatment system, based on condensation technology and the major advantages and disadvantages of this technology.

Cooling water

Inlet air

Condensate Figure 3. Shell and tube condenser.

Hot water outlet

116

R. Nabais

Table 7. General Characterization of Condensation Technology Performance factors

Malfunction conditions

Controls of the systems and usual checklists

Gaseous components concentration Coolant inlet and outlet temperatures

High outlet temperatures Low coolant flow rates Inadequate heat transfer conditions Buildup of ice-frozen organic compounds

Outlet gas stream temperature Gaseous stream and coolant flow rates Coolant flow rate

From Air Pollution Control Technology, (2004).

Experiments are described in which the carbon content of vapors given off during wort boiling at the Bavarian State Brewery, Weihenstephan, are continuously collected and condensed by means of a plate heat exchanger and the remainder is piped into the furnace of the brewery boiler (Muller, 1990; Muller and Meyer-Pittroff, 1990). Condensation of components of a gaseous stream may be achieved by using indirect cooling of the gas to condense water vapor onto the collecting surfaces. Because condensed water is the medium that carries away the entrained particles there are available in the market entirely self-cleaning devices with low pressure drop. At least, with this technology, if condensed pollutants are removed from the exhaust gas stream, the amount of odorous gas to be treated downstream will be reduced. In biological industries (for example, rendering works) it is usual to place the biofilter after the condenser (NSW, 2004). Condensers are convenient equipment for simultaneously removing contaminants from the gas phase and quenching steam and often are the first stage in an air-cleaning system. Table 8. Major Advantages and Disadvantages of the Condensation Technology Process

Advantages

Disadvantages

Condensation

Ability to cost-effectively recover and reuse condensed components from this process increases the technique’s value; Condensation method will work with any volatile compound, in any concentration, from streams with different flow rates.

The coolant fluid depends on the vapor pressures of the organic contaminants to be isolated; Thus, the identity of the VOC pollutants dictates the operating conditions for the control process; this may be difficult to achieve in gaseous streams composed of complex odors mixtures; The relatively low concentration of solvents in air requires very low temperatures resulting in high energy costs (electrical power), and hence, poor economics; Recovery by cryogenic condensation will condense and freeze water vapor from the airstream requiring complex exchanger thawing techniques.

From Van Deuren et al. (2002).

Odors Treatment: Physicochemical Technologies

117

7.3.4. Oxidation There are a set of technologies based in physicochemical oxidation processes, and despite intrinsic differences of operational systems, they are generally referred to as “incineration.” 7.3.4.1. Incineration Generalities In vapor incineration processes, organic components of gaseous stream are oxidized and form carbon dioxide and water vapor. If organic components of gas stream contain chlorine, fluorine, or sulfur, then hydrochloric acid, hydrofluoric acid, sulfur dioxide, and other compounds may be formed. Like other combustion control devices, thermal incinerators operate on the principle that any gaseous organic compound heated up to a high enough temperature in the presence of sufficient oxygen will be oxidized to carbon dioxide and water. The theoretical combustion temperature for thermal oxidation depends on the properties of the VOC to be combusted. General aspects of air pollutant treatment systems based on incineration (thermal oxidation) are listed in Table 7.9. Odor abatement by incineration in combustion plants is well suited in plants with large consumption of energy (fish meal, meat meal, bone meal factories, etc.); in this case, heating the boilers or cooking devices can enhance the exhaust gas heat, whereas for plants with low energy consumption the operation costs of incineration are often unbearable. In this case the best and cheapest solution is often a combination of ad/absorption.

Table 9. General Characterization of Incineration (Thermal Oxidation Technology)

Performance factors

Malfunction conditions

Controls of the systems and usual checklists

Variations in chamber temperature, residence time, inlet VOC concentration; Vent streams with a low heat content need extra fuel for flame stability and their combustion; Dimension parameters include variable waste flow rate, calorific value, temperature and oxygen content, and residence time; Appropriate turbulence design.

Low combustion temperature; Inadequate residence time; Inadequate mixing of the gases; Generation of additional pollutants in the incinerator; Loss of catalyst activity for catalytic processes; Operating temperatures must be higher than 800 ºC, to avoid carbon monoxide formation; Thermal incinerators can achieve more than 98% destruction efficiency at combustion chamber temperatures ranging from 700 to 1,300º C and residence times of 0.5 to 1.5 seconds.

Outlet gas stream VOC concentration; Inlet and outlet gas stream temperature (thermal oxidizers); Stack effluent opacity; Variations in the gas flow rate.

From Van Deuren et al.( 2002).

118

R. Nabais

7.3.4.2. Thermal Oxidation Thermal oxidation, by definition, is an exothermal reaction, by which a hydrocarbon is converted in carbon dioxide and water vapor, in the presence of oxygen, as follows: Cn H2m + (n + m/2) O2 => n CO2 + m H2O + Heat The number of oxygen atoms converted into n molecules of carbon dioxide and m molecules of water vapor and heat, which is given off in the exothermic reactor. In general, a thermal oxidizer is a specially engineered furnace or chamber where materials are burned, combusted, and/or oxidized. The furnace or chamber operates at relatively high temperatures with a sufficient retention (residence) time and optimum turbulence to destroy combustible material. With thermal incineration odorous chemicals are oxidized into less odorous or non-odorous substances, for instance, carbon dioxide, water and sulfur dioxide. The incineration temperature normally ranges between 500 and 1,200 ºC. Thermal oxidizers also have fast startups; therefore, they are good choices for batch working or shift processes. The air required for the boiler plant can be taken from the exhaust system. This method allows the factory to clean large volumes of air at low costs. Considering combustion heat recovery possibilities, thermal oxidation systems may operate by three modes: direct flame, recuperative, and regenerative. Direct flame systems or flares operate by contact of the waste stream with a flame to achieve oxidation of the VOCs. These systems are the simplest thermal oxidizers and the least expensive to install, but require the greatest amount of auxiliary fuel to maintain the oxidation temperature, thus entailing the highest operating cost. Flares are very useful for destruction of intermittent streams; this is a very common situation in the food industry. Direct fire is essentially a combustion chamber with a burner and the appropriate control system. The exhaust from a direct-fired unit is typically at the combustion temperature with no primary or secondary heat recovery. This is used where heat recovery is not required (e.g., when fuel for the burner is free or very cheap). Flares are used for gas streams with organic vapor contents greater than two- or threefold their lower explosive limits; if the gas stream does not have enough heat content, additional fuel is added to the gaseous stream. Recuperative thermal oxidation systems (see Figure 7.4). In this configuration, an oxidation chamber is coupled to a heat exchanger where combustion heat of exhaust gas is transferred to inlet air. Thermal recovery efficiencies are limited to 40–70% of lower explosive limit to prevent autoignition in the heat exchanger. Supplemental fuel therefore is usually required to maintain a high enough temperature for the desired destruction efficiency. Recuperative systems are more expensive to install than flares, but have lower operating costs. Regenerative thermal oxidizer (RTO) is a developing technique based on the thermal oxidation, by operating with great fuel efficiency. An RTO consists of two or more heat exchangers connected by a common combustion chamber or zone (see Figure 7.5). The heat exchangers consist of beds filled with media,

119

Odors Treatment: Physicochemical Technologies Gas outlet

Heat exchanger

Gas inlet

Burner

Oxidation Chamber Figure 4. Recuperative thermal oxidizer. Adapted from Virtuous Cycle (2004).

which will allow air to pass while serving as a mass to store heat. The media material selection, size, and shape can vary greatly and substantially impact the design and utility efficiency of the RTO. In comparison to both thermal and catalytic oxidation, RTOs have the advantage of the VOC application flexibility and destruction of a thermal oxidizer with better fuel efficiency than a catalytic oxidizer without the risk of poisoning or fouling expensive catalyst.Regenerative thermal oxidation systems are the most expensive thermal oxidizers to build, but savings in auxiliary fuel pays the added initial costs. It is important to reduce the moisture content of any gas stream requiring incineration (above 400ºC) in order to reduce fuel consumption, because Supplement fuel flame

Combustion

Chamber Stack

Bed packing ceramic heat accumulator media

Blower of Waste Air Stream

Blower to return exhausted air from the stack

Treated Air to the stack

Figure 5. Sketch of a regenerative thermal oxidizer.

120

R. Nabais

Table 10. Relative Costs of Incinerating Odorous Vapors Saturated with Water Vapor at Various Temperatures Saturation temperature

Cost ratio

400C 500C 600C 700C 800C 850C

1.00 1.05 1.15 1.36 1.81 2.22

From NSW (2004).

incineration costs increase with the moisture content of the airstream to be incinerated, as shown in Table 7.10. Food and animal products industries are a major area where thermal oxidizers are used to eliminate odors. 7.3.4.3. Catalytic Oxidation “Catalysis” was a term used by Berzelius in 1835 to describe the ability of a chemical element to increase the rate of a chemical reaction between other compounds without it being appreciably consumed by the reaction. Catalysts do this by lowering the amount of chemical energy required for a given reaction to take place. Catalytic oxidation converts VOC into carbon dioxide and water, as do other oxidation processes, but it can be planned to avoid by-products requiring treatments or final disposal. The advantage of this technology, in comparison with noncatalytic oxidation, is a substantial reduction of process temperature (315–455°C); hence catalytic systems represent important energy savings. Typical catalytic oxidizer components include the catalyst housing, blower, burner, heat exchanger, controls, and stack (see Figure 7.6). Catalytic oxidation Treated air

Catalyst Fuel Air stream to treat

Pre heater Blower

Figure 6. Catalytic thermal oxidation. Adapted from (2004).

121

Odors Treatment: Physicochemical Technologies

is well suited to applications with VOC concentrations ranging up to 25% of the lower explosion limit. With proper selection of catalyst, operating conditions, and equipment design, catalytic oxidation can attain VOC conversions of up to 99%. Advantages of this technology are low fuel usage, particularly with the proper choice of heat exchanger, given low operating temperatures, and little formation of partial oxidation products, such as carbon monoxide and aldehydes. Disadvantages include susceptibility to catalyst poisons and the sensitivity of the catalysts to high temperatures. Tables 7.11 and 7.12 list general aspects of air pollutant treatment system, based on catalytic oxidation technology and the major advantages and disadvantages of this technology. Table 11. General Characterization of Catalytic Oxidation Technology Performance factors

Malfunction conditions

Catalytic oxidizers are used to treat gas streams that have organic vapor concentrations under the lower explosive limit

Low combustion temperature; Inadequate residence time; Poor mixture of gaseous compounds; Fouling or plugging; Loss of catalyst activity

Controls of the systems and usual checklists Inlet and outlet gas temperatures; Accumulation of particles, followed by emissions opacity

Table 12. Major Advantages and Disadvantages of the Catalytic Oxidation Technology Advantages

Disadvantages

Lower fuel costs than thermal oxidation; Operates at lower temperatures than thermal oxidation

Higher capital costs than thermal oxidation; Less fuel is required compared with other thermal systems because they operate at lower temperatures: from 315 ºC to 455ºC; Catalytic oxidizers are not effective with high concentrations of solids or liquid particles; Frequent cleaning is desired, so the restoration of catalytic activity can rise to up 90%; Catalyst must be optimized for the contaminant present in the off-gas stream (catalytic conversion is species-dependent); Catalyst must be cleaned regularly (12- to 24-month intervals); Several compounds influence catalysts performances (e.g., fluorine and sulfur); Require safety precautions; Catalysts operate typically for less than 3,000 ppm pollutants concentrations; Catalysts must be replaced periodically; Catalyst material may be hazardous and may require disposal as contaminated waste; Phosphorous, heavy metals (zinc, lead), sulfur compounds, chlorine, bromine, iodine, fluorine, and any particulate can result in shortening the life of the catalyst.

Adapted from Van Deuren et al.(2002).

122

R. Nabais

7.4. REFERENCES Air Pollution Control Technology Verification Center (1999) APC. VOHAP. “Control technologies.” January 28, 1999. Air Pollution Training Institute (APTI) (2004) Air Pollution Control Technology Series Training Tool. Control Technology Series. http://www.epa.gov/air/oaqps/eog/utrain.html Calgon Carbon Corporation (2002)_ ES-OC-TP-005. “Options in odor control. New alternatives and traditional technologies”, June 2002. Croll-Reynolds Clean Air Technologies (1996) “Odor control in edible oil processing. Reprinting of Pollution Engineering”, CRS-11-96. Crowder, J. (2004) “Carbon adsorption control devices”, in: Air Control Techniques, P C. Air Pollution Control Technology APTI, 2004. Control Technology Series. Davis, R. J., Zeiss, R.F. (1998) “Cryogenic condensation: A cost-effective technology for controlling VOC emissions”, Proceedings of the 1998 Annual Meeting of the American Filtration and Separations Society, May 1998, St. Louis, Missouri. Environmental Protection Agency (1979) Snyder, W.H. Technical Report. “Guideline for fluid modelling of atmospheric diffusion”, EPA-450/4-79-016 (Draft) 1979. Environmental Protection Agency. (1995) “Compilation of air pollutant emission factors”, AP-42, Chapter 9: Food and Agricultural Industries. Fifth Edition, Volume I. United States. http:// www.epa.gov/ttn/chief/ap42/ Environmental Protection Agency (2000) Biosolids and residuals, Management fact sheet. “Odor control in biosolids management.” http://www.epa.gov/owmitnet/mtb/odor_control biosolids.pdf. Environmental Protection Agency (1993)“Guideline on air quality models” (revised): Appendix W of 40 CFR Part 51 (supplement b). (AH-FRL-7478-3). RIN 2060-AF01). Revision to the Guideline on Air Quality Models. http://www.weblakes.com/lakeepa6.html#GAQM. Environmental Protection Agency (1992)Technical Support Div. Cox, W.M. Technical Report. “Protocol for determining the best performing model”, PB-93-226082/XAB; EPA-454/R-92/025. Integrated Pollution Prevention and Control (IPPC) (2003) European Commission Directorate-General Jrc. Joint Research Centre. Institute for Prospective Technological Studies Technologies for Sustainable Development. European IPPC Bureau. “BREF (Best available techniques. REFerence document) on the Slaughterhouses and Animal by-products according to Article 16(2) of Council Directive 96/61/EC.” http://europa.eu.int/comm/environment/ippc/brefs/sa_bref_1103.pdf. Julie Van Deuren, Teressa Lloyd, Shobha Chhetry, Raycharn Liou, James Peck. Federal Remediation Technologies Round Table. “Remediation Technologies Screening Matrix and Reference Guide, Version 4.” http://www.frtr.gov/matrix2/section3/3_14.html. Muller, K., Meyer-Pittroff, R. (1990) Master Breweries Association of the Americas. “Condensation and thermal treatment of brew house vapors”, Technical Quarterly. 27(1), 1-4. Muller, K. (1990) Master Breweries Association of the Americas. “Combustion of obnoxious substances contained in brewhouse vapors through the boiler firing system.” Technical Quarterly 27(2), 52-55. New South Wales Department of Environment and Conservation (2004) “Odour control.” http:// www.epa.nsw.gov.au/mao/odourcontrol.htm. New South Wales Environment Protection Authority (2001) “Approved Methods and Guidance, For the Modelling and Assessment of Air Pollutants in New South Wales.” http://www.environment. nsw.gov.au/resources/amgmaap.pdf. Rafson, H.J. (ed) (1998) Odor and VOC Control Handbook. McGraw-Hill, New York. US Department of Energy, Office of Environmental Management and Office of Science and Technology (2001)“Membrane system for the recovery of volatile organic compounds from remediation off-gases. Summary report.” Industry Programs and TRU and Mixed Waste Focus Area. http://apps.em.doe.gov/ost/pubs/itsrs/itsr266.pdf. US Environmental Protection Agency Research Triangle Park. Control Technology Center (1995) Information Transfer and Program Integration Division. Office of Air Quality Planning and Standards. R. A. Zerbonia. J. J. Spivey. S. K. Agarwal, A. S. Damle. C. W. Sanford., “Vapor gas streams, control

Odors Treatment: Physicochemical Technologies

123

technology center, Survey of control technologies for low concentration organic vapor gas streams”, EPA-456/R-95-003.1995. http://home.triad.rr.com/sanfordconsult/low_vo.pdf. US Environmental Protection Agency (2002)Air Pollution Control Cost Manual, Sixth edition, EPA/452/B-02-001. 2002. http://www.epa.gov/ttn/catc/dirl/c_allchs.pdf. US Environmental Protection Agency (1994) Volatile Organic Compound Removal from Air Streams by Membranes Separation. Membrane Technology and Research, Inc. United States Environmental Protection Agency. EPA/540/F-94/503. US EPA Model (2004) Regulatory Models - ISCST3 | ISCLT3 | ISC-PRIME | AERMOD | CALINE3 | CALINE4 | CAL3QHC | FDM | UAM - downloads. Environmental Protection Agency,“40 CFR Part 51(AH-FRL-7478-3)”. RIN 2060-AF01. 04/17/03.http://www.weblakes.com/lakeepa3.html. United States Environmental Protection Agency (1986) Guideline on Air Quality Models (Revised), EPA-450/2-78-027R, July 1986. US Army Center for Health Promotion and Preventive Medicine (2005) Introduction to Air Pollution Control. http://chppm-www.apgea.army.mil/desp/pages/samp_doc/aircontrol_intro/22.2005). US Army Corps of Engineers. Engineering and Design (2004)Adsorption Design Guide. DG 11101-2. Chapter 3.http://www.usace.army.mil/usace-docs/design-guides/all.htm. US Army Corps Of Engineers (2001) Engineering and Design–Air Stripping. CECW-E 1110-1-3. 31. Oct 2001. http://www.usace.army.mil/usace-docs/design-guides/all.htm. Van Deuren, J., Lloyd, T., Chhetry, S., Liou, R., Peck, J. (2002) Federal Remediation Technologies Roundtable, “Remediation technologies screening matrix and reference guide, version 4”, http:// www.frtr.gov/matrix2/section3/3-14.html. Virtuous Cycle (formerly BRM Business) (2004)Major Components of Recuperative Thermal Oxidizer. http://www.virtuouscycle.co.uk.

8 Odors Treatment: Biological Technologies Bram Sercu, João Peixoto, Kristof Demeestere, Toon van Elst, and Herman Van Langenhove

1. INTRODUCTION Physical–chemical waste gas cleaning techniques have proven their efficiency and reliability and will continue to occupy their niche, but several disadvantages remain. Among them are high investment and operation costs and the possible generation of secondary waste streams. With biological waste treatment techniques, reactor engineering is often less complicated and consequently costs are less. In addition, usually no secondary wastes are produced. Biological methods are nonhazardous and benign for the environment. Possible drawbacks are restricted knowledge about the biodegradation processes, limited process control, and relatively slow reaction kinetics. Anyway, the biological methods for the removal of odors and volatile organic compounds (VOCs) from waste gases are cost-effective technologies, when low concentrations (below 1–10 g/m−3) are to be dealt with (Kosteltz et al., 1996). Therefore, decision making can be based merely on economical analysis. Like the treatment of liquid effluents, gaseous streams will be more often considered for biological treatment. For organic compounds, the biological reaction can be described as: CHO + O2 + nutrients

C5 H7 O2 N (cell dry weight) + CO2 + H2O + heat

When heteroatoms are present (e.g., chlorine, sulfur), end-products like HCl or H2SO4 can be formed. For efficient pollutant removal, target pollutants have to be sufficiently biodegradable and bioavailable. A major advantage in the case of odor treatment is that biocatalysts have high affinity for the substrates, which BRAM SERCU, JOÃO PEIXOTO, KRISTOF DEMEESTERE, TOON VAN ELST, AND HERMAN VAN LANGENHOVE ● EnVOC Research Group, Department of Organic Chemistry, Faculty of Bioscience Engineering, Ghent University, 9000 Gent, Belgium. Department of Biological Engineering, CEB, University of Minho, 4710–057 Braga, Portugal. Project Research Gent nv, 9030 Gent, Belgium. 125

126

B. Sercu et al.

allows efficient treatment of low influent concentrations. Biocatalysts also operate at room temperature and they have innocuous final products (e.g., carbon dioxide and water). Provided that you have the right inocula, microorganisms can metabolize almost every compound there is. In general, odors consist of a very complex mixture of volatile organic as well as inorganic compounds. The most relevant compounds regarding odors in the food industry are nitrogencontaining compounds (ammonia, amines, amides, and more complex molecules like indole and scatole), reduced sulfur compounds (hydrogen sulfide and volatile organic sulfur compounds) and VOCs like alcohols, aldehydes, volatile fatty acids, and phenols. Most are easily biodegradable. 2. GENERAL OVERVIEW OF TECHNIQUES Several biological waste gas treatment reactor concepts exist. They can be distinguished according to their filter material (organic or inorganic) and the type of liquid phase (noncontinuous or continuous). Both characteristics influence mass transfer and the presence or type of biofilm. In most cases, pollutants are first transferred from the gas phase to the liquid phase and subsequently to the biofilm (Figure 1). It has been argued, however, that pollutants can directly be transferred from the gas phase to the biofilm when no water film is present, or that fungal mycelia, protruding in the gas phase, can directly take up substrates without a dissolution step (Engesser and Plaggemeier, 2000). Optimal conditions for the organisms (micro)environment should be provided. Temperature, pH, water activity, nutrient availability, oxygen concentration, and osmotic potential are important parameters. In practice, most of these parameters can be controlled in an acceptable range by reactor choice, operation, and control. However, sometimes waste gas characteristics like high or fluctuating temperatures and low oxygen concentrations are more difficult to adjust, leading to limitations in the application area of biological waste gas treatment techniques. Laboratory studies provide much information about the removal of single compounds or simple mixtures, but especially for VOCs, often relatively high

Gas phase Cg

Liquid layer

Biofilm

Support

C1

Cthr Figure 1. Pollutant transfer during biological waste gas cleaning (Waweru et al., 2000).

Odors Treatment: Biological Technologies

127

influent concentrations are investigated, conditions not prevailing in odorous waste gases. Although the experimental conditions and the influent gas can be controlled carefully in lab studies, it is sometimes difficult to extrapolate results to real situations, especially when the composition of the waste gas is very complex or unknown. Other studies investigate the removal of odors in full-scale installations or use pilot-scale waste gas treatment units receiving waste air from industrial plants. In this case, data with real complex odorous mixtures are obtained. However, only in a few cases the waste gas characteristics in two different plants are much alike. Therefore, before installing a biological waste gas treatment reactor, often a series of pilot-scale experiments is conducted on-site to investigate performance and design criteria. Devinny et al. (1999) recommended the following steps in a protocol for biofilter design and implementation: (1) preliminary investigations of the waste stream (e.g., pollutant concentrations, process flow rate, relative humidity, temperature), (2) literature research and modeling to determine if biofiltration is an appropriate technique, and (3) possibly further bench- and pilot-scale experiments to obtain a final reactor design. Many classifications and denominations of bioreactors have been appearing in the literature. Here we shall classify them in five main groups: biofilters, biotrickling filters, biowashers, and two more recent techniques: the biological plate tower and the membrane bioreactor. Many associations of different reactors are also possible. Stability, adaptability, and low equipment and operational costs are the requests they all must meet. 2.1. Biofilters Biofiltration is often used with a broader meaning, referring to all the waste air biological technologies. Strictly it means the most used technology about this matter: the biofilters. They made their appearance in the 1950s in the deodorization of air from wastewater treatment or composting plants. Nowadays they are aimed not only at air polluted with organic gases, such as VOCs and many other hydrocarbons, but also ammonia or H2S. They are more efficient with low molecular weight gases, with high solubility in water and simpler molecular bonds. Biofiltration has been extensively used, because waste gases in this case generally contain low concentrations of well biodegradable organic and inorganic compounds. Basically, a biofilter is a layer of biologically active media (an organic filter matrix), usually of natural origin. The filter particles are typically soil, compost, peat, wood chips, tree bark, and heather. Granular activated carbon and plastic material are also used. One kind or several combinations of particles have been used. The media must provide a large surface area, nutrients and moisture (around 50% of the media) for the microbial activity, and adsorption/absorption of the odorous molecules. The microflora for the degradation of odors—mainly

128

B. Sercu et al.

bacteria and fungi—is part of the package. There is no continuous water phase. For better results, the addition of nutrients containing nitrogen and phosphorus must be considered, although this will add some cost to the process. The presence of bulking inerts usually calls for the addition of nutrients, mainly with high load regimes (Devinny et al., 1999). Adequate porosity (around 0.50) is essential for low pressure drop (power requirements). To build a conventional open-bed filter (Figure 2), in the early ages of the technique, a hole was excavated in the ground (around 1.0 m deep) and filled up with a bed of the selected media. Nowadays, synthetic material or concrete is used. Perforated piping or other systems are used for gas distribution under the bed. The waste air flow, combined with the void fraction, causes the residence time to be normally between 15 and 60 sec, the time it takes for the odors to be absorbed and metabolized through the filter. Surface loading rates are about 1.2 m3 m−2 min−1 (Devinny et al., 1999). Impermeableness is desirable to avoid liquid leaching. For optimal long-term operation of biofilters, next to controlling the biofilter moisture content, precautions should be made to prevent acidification if sulfur or nitrogen-containing compounds are present. This can be accomplished by buffering, e.g., by adding CaCO3 (Rafson, 1998), or regular replacement of the filter material (every 1 to 5 years, depending on the loading rate). The latter treatment is also needed to remove other accumulated intermediates or end-products, to prevent high pressure drops, and to prevent nutrient limitation if nutrients are not provided during biofilter operation. Indeed, removal efficiencies go from 60% to 100%, depending on the media and the pollutants contaminating the air. Initial performance is very good but as time goes on problems may occur, leading to severe degradation of efficiency. Clogging and channeling are likely to appear. Modular closed systems are commercially available. They minimize the surface for installation because they are stacks of trays that can be set in series or parallel arrangements, or combinations of both. The usual time of operation, with good removal characteristics, for conventional systems (2 to

Clean air

Filter media

Polluted air

Excess water

Figure 2. Schematic of a conventional open-bed biofilter.

129

Odors Treatment: Biological Technologies

4 years, according to Devinny et al., 1999), is extended due to selected media, uniform distribution, and the inclusion of controls for temperature (usually around 37 °C), pH, moisture, and airstream relative humidity (it must be near saturation). 2.2. Biotrickling Filters Biotrickling filters are single unit operation reactors (for both capture and destruction), like biofilters. A packed column is inoculated with microorganisms that attach to the particles. Biofilms grow using nutrients supplied by the contaminated airstream and by a liquid flow that trickles down the packing, continuously or periodically. The liquid moving phase and the inorganic nature of the media particles are the most important differences between biofilters and biotrickling filters (Figure 3). Unlike biofilters that use natural materials, most particles for biotrickling filters are built with plastic, steel, or ceramic material. The simplest of all is the Raschig ring. Many particle designs have been used. The odor is first transferred from the air to the circulating water. Next it must diffuse to the biofilm. Finally the microorganisms oxidize the compounds. In previous studies using a biotrickling filter for VOC removal with Pseudomonas putida as the biodegrading bacteria, several packing materials were checked to try to avoid channeling and clogging to block the process (Peixoto and Mota, 1998). It was proved that it is very difficult to overcome the mentioned problems, even when using 20-mm Raschig rings. The presence of the four phases (gas–vapor, liquid, biological, and solid) involved in the process

Clean air

Packed column

Polluted air

Purge

Fresh water Nutrients

Figure 3. Schematic of a biotrickling filter.

130

B. Sercu et al.

makes things very hard to deal with. Even with a high surface area and porosity, the bacterial growth reduced significantly those parameters in a short time. Ultimately the flow used only one last channel (Figure 4), with evident poor efficiency and higher pressure drop, and the process had to stop. Clogging is a direct result of the bacterial (fungal) growth. Growth means that the microorganisms are metabolizing the pollutants, as they are meant to. Therefore, it does not seem to make sense to try to limit the growth to avoid clogging and channeling. A way to remove the exceeding biomass seems to be the natural answer but very tough to find and a settling tank would be needed after the reactor. Backwashing and high shear stress do not seem to solve the problem. 2.3. Biowashers In biowashers the biomass is suspended in the liquid phase. The waste content from the air is first washed and next oxidized by the suspended microorganisms. There may be one or two separate units for absorption and metabolization. Classically, the airstream is washed in a spray chamber (scrubber; packed bed scrubber) and then the liquid phase is sent to an activated sludge tank, where the pollutants are oxidized (Figure 5). Together, these two units form a bioreactor which is commonly named bioscrubber. Instead of the activated sludge tank, the contaminated liquid may be sent to an airlift reactor (Ritchie and Hill, 1995; Rittmann et al., 2000). This kind of reactor, also named circulating-bed biofilm reactor, is known for its good

Liquid In Air/Vapor Out

Air/Vapor In

Liquid Out

Liquid In

Air/Vapor Out

Air/Vapor In

Liquid Out

Figure 4. Evolution of a biotrickling filter packed with Raschig rings, after colonization by bacteria, due to clogging and chaneling.

131

Odors Treatment: Biological Technologies Clean air

Clean air

Airlift reactor

Washing chamber (scrubber)

Polluted air

Activated sludge tank Air (O2)

Polluted air, fresh water and nutrients

Figure 5. Schematic of biowashers: to the left, a two unit reactor with scrubber and activated sludge; to the right, a single unit airlift reactor.

mass transfer properties. Although the biomass in an airlift reactor is attached to particles, these are still suspended in the circulating liquid phase. In a simpler way, absorption and oxidation may happen in a single unit operation. The airstream is directly bubbled to the liquid phase of the activated sludge tank or the airlift reactor (Figure 5), where regeneration takes place. In both cases, the air is washed and at the same time it supplies oxygen for the aerobic oxidation and mechanical power for the suspension and agitation of the particles/biomass in the bulk water. For this form of operation, the liquid phase never leaves the reactor, only being added to replace losses, due to evaporation and sludge purge. The air (or pure oxygen) for the oxygenation of the tank water is replaced by the polluted airstream. This simple solution, if fit, may be economically interesting. Bioscrubbers are only sporadically used, mainly for removing high concentrations of highly water-soluble compounds. They have been used in the treatment of waste gases from incinerators and foundry industry (amines, phenol, formaldehyde, ammonia). 2.4. Comparison of Technologies Table 1 summarizes the properties of the bioreactors described above and includes the biological plate tower (BPT). When the environmental conditions become toxic or aggressive, the biofilms play an important role in protecting the microorganisms in them. Therefore, attached biomass is less prone to inhibition and destruction. Besides, the growth rate may exceed the dilution rate without washing out the biomass. Biofilters, biotrickling filters, and airlift reactors, all grow attached biofilms.

132

B. Sercu et al.

Table 1. Four entry characterization of biological systems for air pollution control LIQUID PHASE Circulating MICROBIAL

Dispersed

FLORA

Attached

Stationary Spray/ shower

Biowasher Biotrickling filter; BPT

Biofilter

Supplied with water

Part of the packing media

Packing; plates

COLUMN TYPE

MINERAL NUTRIENTS

Compact biofilters reduced significantly the footprint of conventional ones and have shortened the disadvantage related to the need of a huge application area. The absence of a moving liquid phase is its big disadvantage. For the control of the temperature, pH, and dissolved oxygen, and the supply of nutrients, the presence of liquid water is a major advantage. One risk for open-bed biofilters is to get flooded by rain or dried out by the sun. Anyway, when the application area is not a problem, biofilters are still the most-used technology, due to the accumulated knowledge, their low cost, and simplicity to operate. One concern related to openbed biofilters is the release of microorganisms to the surrounding air. Food and fermentation industries need an environment with a controlled amount of microorganisms because of the nature of their processes. Van Groenestijn and Hesselink (1993) refer to emissions up to 104 colony-forming units per cubic meter of treated gas. To minimize the risk to the process, an enclosed biofilter with an induced draft system (vacuum), as described in Devinny et al. (1999), should be preferred. This way only clean air will be drawn into the system if leaks are present. The final ventilation ducting must be positioned accordingly. Biotrickling filters are used because of their ability for process control and are recommended when high concentrations of (acidifying) compounds have to be treated, or when only limited area space is available on site. Biowashers are favored when the pollutants have high solubility in water. Otherwise they are water-consuming and less attractive. If an existing activated sludge tank can receive the waste air from another process, that fact may be very profitable. It is necessary to guarantee that the new pollutants will not interfere with the activated sludge performance. Stripping of dissolved odors, flora inhibition, and filamentous bulking are some of the problems that may appear or get out of control. Contacting times, application area, toxicity of pollutants, acid production during degradation, clogging, chaneling, solubility in water, adhesion to packings, costs of technology, one or several pollutants, selected species or natural consortia, pressure drop, selection of support media, open or closed-beds, and

Odors Treatment: Biological Technologies

133

so on, are variables to be considered. With so many variables, there is still a long way to go before one can be sure of what solution is the best. For the same reason, modeling is also very difficult. For further information about reactor design and modeling, books of Devinny et al. (1999) and Kennes and Veiga (2001) are recommended. 2.5. New Technologies 2.5.1. Biological Plate Tower (BPT) The big mistake in the transposition from physical–chemical to biological reactors is to forget that the presence of biofilms completely changes the behavior and performance of the reactor. Biofilm growth is chaotic and never tridimensionally homogeneous on a random packing. Oriented packing has better results but not yet good enough. A good physical–chemical reactor does not have to be a good biological one and it is indeed a poor option in many situations. Four-phase reactors always bring about hydrodynamic problems. Sloughing, channeling, and clogging always occur. A good efficiency of removal makes it happen faster. To solve these problems and make the process easy to operate steadily for a long time, a new concept of reactor was designed and tested with air polluted with VOCs. Pseudomonas putida was the selected inoculum. The effluent simulation was achieved with the mixing chamber described in Peixoto and Mota (1997). The observation of the growth on the plane surfaces (top liquid distributor, base plate) of a biotrickling filter (Figure 6) suggested a design based on

Figure 6. Liquid distributor showing the growth of biomass on the lower surface. On the top side, that occurrence was even stronger.

134

B. Sercu et al.

horizontal surfaces. Basically, the BPT is a pile of parallel circular plates with a single hole on the border. The plates are placed in such a way that the holes will alternate (180°) from one to the next plate. In this way, a cascade of liquid will go downward, changing direction from plate to plate. The gaseous stream follows the opposite direction, upward. The bacteria attach to their top surface. Figure 7 shows the schematic of the flows and biofilm growth on the plates. The reactor is a four-module (about 28.8 dm3 each) BPT with 20 plates in each module. An individual plate surface area (top face) is about 40 195 mm2. The scratched surfaces of the plates were intended to make the bacterial adhesion easier. Only two or three of the four modules are operated continuously. The other(s) is kept free and ready to replace any one that reaches saturation with biomass. In this way the operation can be kept going virtually forever. The performance is quite stable (the biofilm activity, surface-dependent, is kept approximately constant) and the constant surface contact area makes it easy to model and scale-up the process. The total surface area and the space between plates can be designed for the desired operating time. In theory, the available surface in a BPT is a tenth of the surface in a biotrickling filter, considering the same total volume. The new design proved to ensure a stable operation for longer periods, as well as high VOC removal (92 % removal for inlet toluene concentration of 10 g m−3 and empty bed residence time, EBRT = 108 sec). It has very good hydrodynamic performance and operates continuously without problem. In the long term, the short area is compensated by the steady operation. The disposal of the newly formed biomass is also much easier than in the biotrickling filter. Unlike biofilters whose packing has to be rejected after a certain

Air/Vapor

Liquid

Liquid

Air/Vapor

Air/Vapor

Liquid

Liquid

Figure 7. Simplified schematics of the BPT, with only five plates, to better visualize the directions of both flows and the attached biofilm on the upper surface of the plates.

Odors Treatment: Biological Technologies

135

time of operation, BPT biomass is withdrawn as a water-rich solid phase—the biofilm attached to the plates—and quite easy to handle. When the thickness of the biofilm reaches the maximum value allowable, the set of plates is simply replaced by a clean one and the biomass is dealt with outside the reactor. Sampling the biofilm for analysis is very easy. It does not oblige the operation to be stopped, or severely shaken as it happens with biotrickling filters, and any plate can be sampled. Even operation demands a constant surface of biofilm. Oxygen uptake rate measurements were made to find out if there were great activity differences between different plates and between the surface and inside the biofim. The respiratory activity was similar (about 0.11 mg g−1 s−1, mass of oxygen per mass of volatile solids per time) for the superficial samples of all plates, showing some difference (up to about 20 %) for the lower ones where it was higher. The middle samples had almost zero activity (0.01 mg g−1 s−1 or less) and none of the base samples showed any activity. For the respirometry tests, the carbon source was phenol. The plates at the bottom of each module had thicker biofilms than the upper ones, due to the higher concentration of the carbon source and oxygen in the entrance. The first module, which receives the higher dose, is the one that shows the thickest films, reaching over 15 mm until needing to be replaced (Figure 8). The research on the BPT is still ongoing. Assays to quantify VOC and odor removals are now being planned. In the future, different bacteria, plate shapes, and distances between plates will be tested. The bacterial growth hanging from holes in the plates (sieve trays, similar to Figure 6) also will be investigated. The possibility of using different bacteria in different modules also will be considered.

Figure 8. Photograph of the bottom plates of the first module showing the biofim growth on the BPT plates. The huge biofilm does not endanger the permeability of the system.

136

B. Sercu et al.

2.5.2. Membrane Bioreactor In the membrane bioreactor concept, one side of the membranes is dry and acts as a surface for uptake of pollutants from the air flowing along the membranes, while the other side is kept wet and covered by a biofilm. In Figure 9, a flat membrane bioreactor with a composite membrane is shown, but also other configurations like hollow fiber membranes modules can be applied. Pollutants diffuse through the membrane and are subsequently degraded by the microorganisms in the biofilm or in the recirculating aqueous phase. By continuous recirculation of the aqueous phase, the microbial degradation process can be easily controlled. The main advantages of membrane bioreactors for waste gas treatment include the high specific surface area, the ability to prevent clogging, the good reactor control, the physical separation of gas and biofilm, the low pressure drop, the absence of channeling, and the independent control of gas and liquid phase (De Bo, 2002). Potential disadvantages are the high investment costs, the additional mass transfer resistance caused by the membrane, a decreased biofilm activity as the biofilm ages, and clumping of hollow fiber membranes at high biofilm growth. The reactor concept, although not implemented in practice yet, has potential to eliminate VOCs characterized by poor water solubility, by lack of biodegradability, and by toxicity (Reij et al., 1998). Recently a flat membrane reactor was developed and applied for the degradation of DMS and toluene as single compounds (De Bo et al., 2002, 2003). In this case a composite membrane was used, guaranteeing a

Biofilm

Membrane

Substrate Nutrients Oxygen

pH-buffers

Mineral medium

Waste air

Dense

Porous

Figure 9. Scheme of a flat membrane bioreactor for waste gas treatment (De Bo, 2002).

137

Odors Treatment: Biological Technologies

stable long-term reactor performance because clogging of the porous membrane was prevented. For DMS, an ECmax of 4.8 kg m−3 d−1 was obtained, which was higher than any reported figure for biofilters or biotrickling filters.

3. REMOVAL OF COMMON ODORS Most of the work has been done about removal of ammonia, reduced sulfur compounds (either single or as a mixture), and odorous VOCs, especially with biofiltration. About compounds like amides, indole, scatole, and pyridine no information was found. However, it can be assumed that when bioreactors can operate with high odor reduction efficiency, these compounds are also sufficiently degraded. 3.1. Removal of Ammonia Within the group of odorous nitrogen-containing compounds, ammonia is by far the most investigated and documented compound with respect to its removal from waste gases by biological treatment technologies. Although ammonia has a rather high odor threshold value (3.9 ppmv to 5.8 ppmv) compared to many other odorous (nitrogen) compounds (Weckhuysen et al., 1994; Devos et al., 1990), and consequently dilutes rapidly to below detection downwind from the emission source, its very sharp and unpleasant odor can cause a severe odor nuisance nearby its emission source. Among biotechnological waste gas treatment systems, mainly biofilters have been used to control emissions containing ammonia. The main processes taking place during biofiltration of ammonia are presented in Figure 10.

GAS PHASE

BIOFILM

FILTER MATERIAL

I NH3

II

NH4+

NH3

NH4+

Nitrosomonas III −

NO2 Nitrobacter

NO3− Figure 10. Main processes taking place during ammonia biofiltration (I: absorption; II: adsorption; III: nitrification).

138

B. Sercu et al.

Since ammonia has a low Henry’s law constant (H20 °C = 5.6 × 10−4) (Perry and Green, 1984) and a protonation constant pKa, 20 °C of 9.23 (Weast et al., 1984), in biofilters it is partly retained by adsorption onto the carrier material and by absorption into the water fraction of the carrier material. In this context, Shoda (1991) reports a maximum volumetric NH3 elimination capacity of 15 g m−3 d−1 in a peat biofilter due to these physical–chemical transfer processes. In a compost biofilter, Smet et al. (2000) obtained an NH3 adsorption and absorption capacity, per volume of compost, of 490 g m−3 and 47 g m−3, respectively, at an NH3 inlet concentration of 159 ppmv and a compost moisture content of 40%. Next to these physical–chemical processes, nitrification by the autotrophic bacteria Nitrosomonas and Nitrobacter is generally considered as the main microbiological process for the degradation of NH3 (Terasawa et al., 1986; Van Langenhove et al., 1988; Williams, 1995). More recently, the autotrophic genera Nitrosospira and Nitrospira also are reported to be responsible for nitrification (Schramm et al., 2000; Regan et al., 2002). As a result of these phase transfer and (micro)biological processes, NH3 has been removed efficiently at concentrations up to 50 ppmv. Using a wood bark biofilter, Van Langenhove et al. (1988) obtained removal efficiencies of at least 90% at concentrations between 6 ppmv and 17 ppmv and at NH3 mass loading rates (Bv) up to 58 g m−3 d−1. At similar concentrations (4 ppmv to 16 ppmv), Weckhuysen et al. (1994) observed elimination efficiencies of 83% or higher at NH3 mass loading rates between 6.8 g m−3 d−1 and 27.2 g m−3 d−1. In an inoculated peat biofilter, NH3 elimination capacities (EC) up to 41 g m−3 d−1 are reported at an inlet concentration of 20 ppmv (Hartikainen et al., 1996). More recently, removal efficiencies as high as 99.5% were obtained for 100 days in an inoculated perlite biofilter, at concentrations of 50 ppmv and NH3 loading rates between 8.6 g m−3 d−1 and 21.5 g m−3 d−1 (Joshi et al., 2000). Due to the sensitivity of nitrifying microorganisms, however, biofiltration of waste gases containing high ammonia concentrations (above 50 ppmv) has been reported to be questionable. Don (1985) and Hartikainen et al. (1996) reported the biofilter removal efficiency for NH3 dropped drastically at waste gas concentrations exceeding 35 ppmv to 60 ppmv. However, more recently, Liang et al. (2000) could obtain NH3 removal efficiencies of at least 95% at inlet concentrations between 20 ppmv and 500 ppmv in compost biofilter in which active carbon was added to reduce compaction and channeling, as well as to increase the reactive surface and durability of the biofilter. Similarly, removal efficiencies over 90% were achieved by Kim et al. (2002) at inlet concentrations up to 150 ppmv in a biofilter system packed with small cubes of polyurethane sponge coated with a powder mixture of activated carbon and natural zeolite. Kalingan et al. (2004) obtained complete NH3 removal in a peat biofilter containing inorganic supporting material at a NH3 concentration of 200 ppmv, while Smet et al. (2000) observed no NH3-toxicity in a compost biofilter at concentrations up to 775 ppmv. Besides the ammonia input concentration, the mass loading rate of a biofilter seems to be critical for an efficient performance. In an inoculated activated

Odors Treatment: Biological Technologies

139

carbon biofilter, Yani et al. (1998) found a complete NH3 removal up to Bv = 95 g m−3 d−1, whereas the elimination efficiency decreased at higher loading rates. The highest EC observed by these authors was 220 g m−3 d−1 at Bv = 250 g m−3 d−1 and at an EBRT of 52 s. Smet et al. (2000) and Demeestere et al. (2002) obtained elimination capacities up to 350 g m−3 d−1 in a compost biofilter at EBRT = 131 s and 21 s, respectively. EC peak values of 530 g m−3 d−1 and 1285 g m−3 d−1 were reported at inlet concentrations of 250 ppmv (Bv = 600 g m−3 d−1) and 450 ppmv (Bv = 1329 g m−3 d−1), respectively (Demeestere et al., 2002). According to these authors, the cumulative loading (mass of NH3 per filter material volume, g m−3) is the limiting factor for a NH3 degrading biofilter. Smet et al. (2000) observed a sharp reduction in elimination after a cumulative NH3 removal of 6000 g m−3. Osmotic effects, due to the accumulation of NH4NOx at concentrations (mass of NH4NOx-N per mass of compost) higher than 4 g kg−1, were found to be the reason for the inhibition of NH3 removal (Smet et al., 2000; Demeestere et al., 2002). However, a subsequent loading of the biofilter with a carbon source like methanol could regenerate the biofilter material, due to methylotrophic conversion of NH4+ and NOx− into biomass (Demeestere et al., 2002). In order to achieve optimum NH3 removal in biofilters, the moisture content of the filter material should be between 40% and 60%, the temperature between 30°C and 35°C and the pH between 7 and 8 (Van Lith et al., 1997; Warren et al., 1997). With respect to the latter parameter, acidification of the filter material due to the accumulation of nitrite and/or nitrate can inhibit the long-term stability of a NH3 degrading biofilter, as reported by Heller and Schwager (1996). On the other hand, no acidification was observed by Don (1985) and Smet et al. (2000), who attributed that effect to the establishment of an equilibrium between NH3 absorption increasing the pH and nitrification decreasing the pH. Next to the removal of NH3 in waste gases by biofiltration, bioscrubbers and biotrickling filters also have been used for NH3 degradation. Due to the presence of a recirculating water phase, both these techniques allow to drain off accumulating toxic compounds, to control the pH, and to add nutrients. As an example, Smits et al. (1995) obtained a biological elimination capacity of 96 g m−3 d−1 in a pilot-scale biotrickling filter at superficial gas and liquid velocities of 1300 m h−1 and 2.5 m h−1, respectively. No gas-to-liquid mass transfer limitation was observed under these conditions. Due to its low Henry’s law coefficient, efficient NH3 scrubbing from the gas phase can be obtained. However, it was observed by the authors that up to 70% of the ammonia removed from the waste gas was not nitrified but removed with the drain water. Consequently, although the use of both biotrickling filters and bioscrubbers can be very efficient to remove NH3 from the waste gas, it implicates the subsequent treatment of the NH4+-loaded drain water in a wastewater treatment plant. Another drawback of these technologies is the relative low removal of less water-soluble odorous compounds in bad-smelling waste gases. Although it is shown by some authors (Tang et al., 1996; Chou and Shiu, 1997; Busca and Pistarino, 2003; Chang et al., 2004) that also other odorous

140

B. Sercu et al.

nitrogen-containing compounds than ammonia can be efficiently removed by biological waste gas treatment technologies, fewer experimental data are published so far in that field. For example, as far as we know, there is no information available about the biotechnological removal of nitrogen compounds like amides, indole, scatole, and pyridine. Nevertheless, some reports deal with the removal of gaseous amines by biofiltration. Amines are bad-smelling compounds that often are present in waste gases arising from fish markets, meat treatment industries, and other food industries (Busca and Pistarino, 2003). According to Chou and Shiu (1997), methylamine (MA) can be successfully removed, i.e., hydrolyzed to ammonia and nitrified to nitrate and/or incorporated into microbial biomass, in peat biofilters at mass loading rates up to 160 g m−3 d−1, at a pH between 7.5 and 8.5, and at a moisture content between 55% and 60%. Tang et al. (1996) investigated the removal of triethylamine (TEA, 78 ppmv to 841 ppmv) in a biofilter consisting of a mixture of compost and chaff particles and obtained the highest TEA elimination capacity of 3360 g m−3 d−1 at an inlet concentration of 550 ppmv, above which substrate inhibition occurred. However, a comprehensive picture cannot be drawn for these compounds due to scarceness of the literature available (Busca and Pistarino, 2003). 3.2. Removal of Hydrogen Sulfide Next to ammonia, H2S biofiltration has been studied extensively, because it is one of the most frequently produced odorous compounds in industrial processes like petroleum refining, rendering, wastewater treatment, food processing, and paper and pulp manufacturing (Yang and Allen, 1994a). The bacteria responsible for H2S degradation in biofilters mostly belong to the genera Thiobacillus (e.g., T. thioparus) and Acidithiobacillus (e.g., A. thiooxidans) and can be either neutrophilic or acidophilic. Under optimal conditions, H2S is oxidized to sulfuric acid, but during stress conditions (high loads, oxygen limitation) accumulation of elemental sulfur has been observed. Because H2S is very biodegradable, most investigations report very efficient H2S removal in a wide concentration range. Yang and Allen (1994a), for instance, observed higher than 99.9% removal efficiencies for H2S inlet concentrations ranging from 5 ppmv to 2650 ppmv. However, because sulfuric acid is produced, acidification of the filter material will inevitably occur during the biofiltration process, its rate depending on the buffer capacity of the filter bed and the amount of H2S removed. Degorce-Dumas et al. (1997) found that buffering the packing to a near neutral pH doubled the length of the period during which > 95% H2S removal efficiency was obtained. When the pH decreased below 6.6, the H2S removal efficiency started to decrease, together with the number of nonacidifying thiobacilli. Instead, acidifying thiobacilli became dominant. Therefore, a correlation between the number of nonacidifying thiobacilli and the H2S removal efficiency was suggested. Other authors, however, observed a smaller effect of acidic pH values on the H2S removal efficiency. Yang and Allen (1994a),

Odors Treatment: Biological Technologies

141

for instance, found almost equal H2S removal efficiencies at pH values between 3.2 and 8.8. Only at pH = 1.6, the removal efficiency decreased to 15%. The high H2S removal efficiency at pH = 3.2 was attributed to the abundance of acidophilic sulfur oxidizing bacteria. Also other studies did not report decreased H2S removal efficiencies at pH values as low as 3 (Wada et al., 1986; Cook et al., 1999) or even 1.2 (Yang et al., 1994). During biofiltration, the pH will first decrease at the inlet side of the biofilter, where most of the H2S is oxidized and the low pH front will consequently move to the deeper parts of the biofilter (Cook et al., 1999). In general, it should be sufficient to maintain a pH value higher than 3 for efficient H2S removal. However, it could be useful to maintain neutral pH values to prevent inhibition of the removal of other compounds present in the waste gas, corrosion, and increased filter medium degradation. To increase the pH of the biofilter material, washing can be applied (Yang and Allen, 1994b), although only small pH increases are usually obtained. Smet et al. (1996b) observed that regeneration of an acidified biofilter (pH = 4.7) was not possible by trickling tap water or buffer solution over the bioreactor, because most of the sulfate was leached as the corresponding sulfate salts and not as sulfuric acid. In addition, leaching caused washout of essential microbial elements. Alternatively, the use of more concentrated buffer solutions in combination with a complete mineral medium or mixing with limestone powder was recommended. Next to acidification, the accumulation of elemental sulfur and sulfate in the filter material can potentially inhibit microbial activity. Yang and Allen (1994b) found the highest concentrations of both compounds at the inlet side of the biofilter. Elemental sulfur was present because it was formed as an intermediate during incomplete H2S oxidation after exposure to high H2S concentrations. By adding increasing amounts of sulfate to different biofilters, Yang and Allen (1994a) observed that concentrations (mass of S per mass of compost) exceeding 25 mg g−1 were inhibitory for H2S removal, probably due to toxic effects. This inhibition effect, however, was not confirmed by Jones et al. (2003), for sulfate concentrations up to 100 mg g−1. In general, it is recommended to evaluate the expected H2S loading rate before designing a biofilter. If it is assumed that all sulfur entering a biofilter will ultimately accumulate as sulfate, its cumulative concentration can be calculated to assess the long-term deactivation of a biofilter, e.g., with a threshold of 25 mg g−1. Because H2S is very biodegradable, EBRTs can be rather low, e.g., 15 s (Yang and Allen, 1994a) without affecting the H2S removal efficiency. Possibly other, less biodegradable or water-soluble compounds present in the waste gas will determine the lower limit of the EBRT. Next to organic materials like compost, peat, or wood bark, different alternative carrier materials were described for H2S biofiltration, being rockwool, fuyolite, and ceramics (Kim et al., 1998), a pelletized mixture of pig manure and sawdust (Elías et al., 2000), pellets of agricultural residues (Elías et al., 2000), porous lava inoculated with Thiobacillus thiooxidans (Cho et al., 2000), and microorganisms immobilized in Ca-alginate (Chung et al., 1996a,b, 1997, 1998; Huang et al., 1996; Park et al., 2002).

142

B. Sercu et al.

Next to biofilter applications, more recent articles describe H2S removal with biotrickling filters. Their main advantage is optimal control of pH, nutrients, and accumulation products, although of course treatment costs are higher. At an EBRT between 30 s and 120 s, high H2S removal efficiencies (> 95%) easily can be obtained for H2S concentrations between 200 ppmv and 2000 ppmv (Ruokojärvi et al., 2001; Sercu et al., 2005b). At lower influent concentrations, lower EBRTs can be used at high removal efficiencies. Gabriel and Deshusses (2003) described the retrofitting of existing chemical scrubbers for H2S removal to biotrickling filters, maintaining an EBRT between 1.6 s and 2.2 s. Removal efficiencies > 98% were commonly reached for 30 ppmv inlet concentrations, with decreases to 90% at 60 ppmv peak concentrations. The removal of volatile organic sulfur compounds in the same reactor was lower, however, e.g., 35% ± 5% for carbon disulfide. The authors attributed the residual odor after the biotrickling filter mainly to the persistence of these compounds. Also Wu et al. (2001) obtained > 95% H2S removal efficiency at EBRT = 5 s, at < 6 ppmv influent concentrations in a pilot-scale biotrickling filter. At 20 ppmv influent concentration the removal efficiency decreased to about 89%. 3.3. Removal of Ammonia and Hydrogen Sulfide A number of studies have been performed regarding the simultaneous removal of H2S and NH3, because both can constitute an important part of odorous gas mixtures. Similarly as with the removal of the separate compounds, high removal efficiencies can be obtained during simultaneous dosing of both compounds, at concentration levels usually occurring in odorous mixtures (< 50 ppmv). At higher H2S and NH3 concentrations, inhibition of the NH3 removal can occur. Kim et al. (2002), for instance, obtained higher than 99% and 92% removal efficiencies for H2S and NH3, respectively, in a wood chips biofilter, at influent concentrations of about 50 ppmv (EBRT = 1 min). At concentrations exceeding 200 ppmv, however, H2S inhibited the NH3 removal, which decreased to 30%, but this effect was reversible when the H2S concentration decreased again. By using a granulated activated carbon biofilter, the inhibition during H2S peak loadings decreased due to buffering effects. An important aspect of simultaneous H2S and NH3 biofiltration is that the extent of the pH decrease, caused by production of sulfuric and nitric acids, can decrease, because the accumulation of acidic products can be small due to (NH4)2SO4 formation. At NH3 concentrations equal or higher than H2S (on volumetric basis), Chung et al. (2000), for instance, observed no acidification. Recently, Chung et al. (2004) showed that acidification during H2S and NH3 removal could further be decreased considerably by selection of heterotrophic bacteria (Pseudomonas putida CH11 for H2S and Arthrobacter oxydans CH8 for NH3). Heterotrophs oxidize H2S and NH3 mainly to elemental sulfur and organic nitrogen, causing only very small production of acidic end-products. In the activated carbon biofilter, H2S and NH3 concentrations between 20 ppmv and 120 ppmv could be very efficiently removed during 180 d. A carbon source

Odors Treatment: Biological Technologies

143

had to be supplied every two weeks to support growth of the heterotrophic organisms. A number of researchers used microorganisms immobilized in Ca-alginate to remove mixtures of H2S and NH3, although the performance of these reactors was somewhat lower than with the more traditional biofilters (Chung et al., 2000, 2001a,b). Possible advantages, however, are increased possibilities for pH control and removal of metabolic products (elemental sulfur and (NH4)2SO4), as is also the case with biotrickling filters. 3.4. Removal of Volatile Organic Sulfur Compounds Volatile organic sulfur compounds (VOSCs) include compounds like dimethyl sulfide (DMS), dimethyl disulfide (DMDS), mercaptans, and carbon disulfide. These compounds have been related to odor complaints in some studies. A direct correlation could even be established between the total odor concentration and the concentration of VOSCs in waste gases of rendering plants (Defoer et al., 2002). Van Langenhove et al. (1992) compared a full-scale biotrickling filter and a biofilter for treating rendering emissions. Both techniques removed alkanals very efficiently, but organic sulfur compounds were much less efficiently removed. This was attributed to an insufficient development of microorganisms capable of degrading these compounds. Goodwin et al. (2000) also observed problems removing reduced sulfur compounds with a biofilter at a biosolids composting facility. Increasing the EBRT from 20 s to 32 s improved the removal efficiency somewhat. In contrast, VOCs like methane, formaldehyde, isopentanal, N,Ndimethyl methenamine, and dimethylamine were removed for more than 95% in all cases at average inlet concentrations of 15 ppmv. Different reasons can explain the relation of VOSCs and odor nuisance. First of all, VOSCs combine a very bad smell with very low odor threshold values, for instance 0.1 ppbv to 3.6 ppbv for DMDS and 0.9 ppbv to 8.5 ppbv for methyl mercaptan (Smet et al., 1998). This means that to prevent odor nuisance only very low concentrations can persist in the treated gas stream. Second, compared with H2S, VOSCs are less biodegradable. Degradation rates decrease in the order H2S > MM > DMDS > DMS (Cho et al., 1991; Smet et al., 1998). Therefore, it is recommended to inoculate biofilters to shorten the start-up period and to remove high concentrations of these compounds. Smet et al. (1996a), for instance, increased the maximal DMS elimination capacity from 10 g m−3 d−1 to 680 g m−3 d−1 after inoculation of a compost biofilter with Hyphomicrobium MS3. Also other authors used inocula (e.g., Hyphomicrobium spp., Thiobacillus spp.) to remove VOSCs in biofilters (Cho et al., 1991, 1992; Zhang et al., 1991; Park et al., 1993). However, in full-scale applications, the use of inoculation is not well documented. Smet (1995) reported successful removal of organic sulfur compounds in a full-scale biofilter treating emissions from mushroom composting, after inoculation with a specialized strain. Fifty days after inoculation, the total sulfur removal efficiency (excluding H2S-S) in the inoculated biofilter section had increased to 99% compared with 68% in the noninoculated section. But

144

B. Sercu et al.

even when inoculation is used, in a mixture of reduced sulfur compounds, H2S is preferentially degraded over dimethyl sulfide or other organic sulfur compounds (Cho et al., 1992; Wani et al., 1999; Zhang et al., 1991). This occurs because H2S oxidation yields most energy for the microorganisms (Smet et al., 1998). Therefore, the bioreactor has to be designed large enough to allow H2S degradation at the inlet side of the biofilter and degradation of the remaining VOSCs deeper in the biofilter bed. Finally, when a biofilter is designed properly to remove VOSCs, there is still a change of long-term decrease in removal efficiency because of acidification. Similarly as for H2S, sulfuric acid is formed after complete oxidation of VOSCs. Microorganisms degrading the VOSCs, however, are much more sensitive to low pH values than H2S oxidizing bacteria. Smet et al. (1996b), for instance, observed a decreased DMS elimination capacity when the compost pH decreased below 5. To prevent problems due to acidification, the bioreactor has to be designed large enough, and for high influents loadings pH control should be included. Alternatively, two-stage systems have been proposed, first removing H2S and subsequently VOSCs (Kasakura and Tatsukawa, 1995; Park et al., 1993; Ruokojarvi et al., 2001; Sercu et al., 2005b). Ruokojarvi et al. (2001), for instance, developed a two-stage biotrickling filter for sequential removal of H2S, methyl mercaptan (MM) and DMS. Two bioreactors connected in series were inoculated with enriched activated sludge, the first operating at low pH for H2S removal and the second at neutral pH for DMS removal. MM was removed in both reactors. H2S, DMS and MM elimination capacities (as S) as high as 47.9 g m−3 h−1, 36.6 g m−3 h−1 and 2.8 g m−3 h−1, respectively, were obtained for the entire two-stage biotrickling filter at > 99% removal efficiencies and the reactor showed a good long-term stability. 3.5. Removal of Odorous VOCs Generally, odorous VOCs are biodegradable in biofilters (Van Langenhove et al., 1989b, 1992; Goodwin et al., 2000). In most cases these compounds are not the cause of odor problems when biofilter malfunctioning occurs, and therefore literature data about the removal of low concentrations of these compounds are less available than, for example, ammonia and hydrogen sulfide. The removal of aldehydes, alcohols, and fatty acids is generally very good in biofilters (Kiared et al., 1997; Mohseni and Allen, 2000; Otten et al., 2004; Sheridan et al., 2003; Weckhuysen et al., 1993). For methanol, for instance, it was, found that concentration step changes and periods without methanol loading did not affect its removal efficiency in biofilters (Mohseni and Allen, 1999), which was attributed to the good biodegradability and high water solubility of methanol. In some studies it was shown that nutrient addition could enhance VOC elimination capacities during longer periods, e.g., in the case of butanal (Weckhuysen et al., 1993) or butyric acid (Sheridan et al., 2003). It has been observed that in the case of aldehydes, the corresponding organic acids can accumulate during biofiltration, especially at higher influent concentrations

Odors Treatment: Biological Technologies

145

(Weckhuysen et al., 1993; Sercu et al., 2005a). This can lead to a pH decrease, potentially limiting the removal efficiencies of other compounds in the waste gas. Next to biofilters, biotrickling filters have been used to remove odorous VOCs. Again, high removal efficiencies have been obtained for aldehydes, alcohols, and volatile fatty acids (Chang and Lu, 2003; Chua et al., 2000; Ibrahim et al., 2001; Kirchner et al., 1991), even at low EBRT. Kirchner et al (1987), for instance, showed > 90% removal efficiencies for compounds like aldehydes and alcohols at 5 ppmv to 40 ppmv influent concentrations and 2.4 s EBRT. Ibrahim et al. (2001) found 92% and 95% removal efficiencies for 10 ppmv acetaldehyde and propionaldehyde inlet concentrations, respectively, in a column packed with immobilized activated sludge beads at EBRT = 12.4 s. At higher influent concentrations, the removal of both compounds decreased, however, due to inhibitory effects. For higher influent concentrations, higher EBRT values are needed, as shown by Chang and Lu (2003). They found nearly complete isopropanol removal efficiencies in a biotrickling filter, operated between 20 s and 90 s EBRT time at influent concentrations between 100 ppmv and 500 ppmv. When too high influent loadings are applied, accumulation of compounds or intermediates can occur. Chua et al. (2000) found > 99% removal efficiencies for butyric and valeric acid in a biotrickling filter, at mass loadings between 4.8 g m−3 h−1 and 37.8 g m−3 h−1 (0.05 g m−3 to 0.86 g m−3). However, at loading rates exceeding 32 g m−3 h−1, the maximal biodegradation capacity was reached and accumulation of volatile fatty acids in the liquid phase was observed. 3.6. Removal of Odor Mixtures From the previous sections, it is clear that most of the components present in odorous mixtures can be removed efficiently with biological waste gas cleaning techniques, when properly operated, even at relatively high influent concentrations. Also, in industrial applications treating mixtures of compounds, often high (odor) removal efficiencies can be obtained. Park et al. (2001), for instance, used a biotrickling filter packed with ceramics and inoculated with activated sludge to remove odors at a composting facility. After a 30 d acclimation period, > 95% removal efficiencies were obtained for NH3 and H2S during about 60 d of operation. Also, at a biosolids composting facility, Goodwin et al. (2000) found efficient odor removal with a biofilter (> 95%) after about 3 months of operation at EBRT = 20 s, as determined with olfactometric analyses. Luo (2001) observed > 98% odor reduction with wood bark biofilters treating rendering emissions during a period of 3 years, at EBRT = 6.8 min. Reducing the EBRT to 1.7 min did not affect the odor removal efficiencies during the first 3 months of operation. After 22 months, however, the odor removal efficiency was 99.1% at EBRT = 6.8 min and only 29.7% at EBRT = 1.7 min. This clearly shows that regular filter medium replacement is necessary, especially when lower EBRT values are used. When a complex mixture of odorous compounds has to be treated, removal efficiencies of the single compounds can be smaller than expected. This can be

146

B. Sercu et al.

caused by, e.g., toxic effects of substrates or metabolites. Van Langenhove et al. (1989a) compared the applicability of a tree bark biofilter for removing odors from a vegetable processing industry, mainly emitted during the blanching process. The main odorous compounds identified were sulfides, isothiocyanates, nitriles, and aldehydes. In pilot-scale experiments all compounds were removed with > 95% removal efficiencies at a volumetric loading rate of 200 m3 m−2 h−1. However, a full-scale biofilter, designed according to the results obtained from the pilot-scale studies, had lower removal efficiencies after 6 months of operation (45% to 65% for sulfides). This was found to be caused by the accumulation of isothiocyanates, which was not observed during the short-term pilot-scale experiments. For hexanal, Van Langenhove et al. (1989b) observed 85% removal efficiency in a wood bark biofilter, at 10 ppmv inlet concentration and EBRT = 0.33 min. To simulate emissions from a food processing plant, 40 ppmv SO2 was added to the waste stream, leading to a drastic decrease of the hexanal removal efficiency to 40%. Next to toxic effects, preferential degradation of easily biodegradable compounds can inhibit the removal of other compounds. Smet et al. (1997), for instance, found that isobutanal was preferentially degraded before DMS, in a biofilter inoculated with Hyphomicrobium MS3, when both compounds were simultaneously dosed. This could cause low removal of DMS when a biofilter is designed too small.

4. CASE STUDIES 4.1. Methodology In all case studies mentioned in this paragraph, samples have been taken of the untreated and the treated airflows, in order to determine important parameters. First, the chemical composition of the airflow was revealed using GC-MS analysis. These data are very useful for the determination of the total chemical load going to the bioreactor, as well as for improving the working efficiency of it, being able to indicate the compounds or groups of compounds that are degraded insufficiently. The second type of analyses used is the determination of the total odor concentration, using dynamic olfactometry. These data are used to determine the total odor removal efficiency of the bioreactor, which is the final wanted effect of the use of a bioreactor in case of odor problems. 4.1.1 GC-MS Analysis 4.1.1a. Sampling procedure. The gases were sampled using a method that involved preconcentration on an adsorbent. This preconcentration step was carried out at the sampling location. Tenax TA was used as adsorbent. Tenax TA is a porous polymer based on 2,6-diphenylene oxide. It has been specifically designed for the trapping of volatiles and semivolatiles (SIS, 2000). The collected

Odors Treatment: Biological Technologies

147

waste gases were cooled at about 4 °C before adsorption. This cooling stage was used in order to increase the breakthrough volume and in order to separate excess water vapor. The sampled adsorption tubes were filled with approximately 750 mg Tenax. Sampling rate was 200 ml min−1 and sampled volume varied between 50 ml and 10 L. 4.1.1b. Analysis. The analysis of the VOCs present in the sample was done in different steps, including desorption from the adsorbent, separation by gas chromatographic techniques, quantification by flame ionization detection, and subsequent identification through mass spectrometry. The desorption step consisted of a thermal desorption. A second preconcentration (cryogenic trapping of the VOCs) was necessary in order to achieve good chromatographic separation. The cryogenically concentrated samples were introduced immediately into the GC by rapid heating of the trap. The different compounds were separated in a gas chromatograph (Varian 2700) with a 100% polydimethylsiloxane apolar column (type DB-1, 30 m × 0.53 mm, film thickness 5 µm, J&W Scientific). The mass spectrometer used was a Finnigan MAT 112 S with an electron impact ion source and a magnetic sector analyser. 4.1.2. Olfactometry 4.1.2a. Sampling procedure. The gases were sampled using the static sampling method. In this method, a sample is collected and transferred into a sampling container (bag). Collecting the sample was done with the “lung principle,” where the sample bag is placed in a rigid container and the air is removed from the container using a vacuum pump. The partial vacuum created in the container causes the bag to fill with a volume of sample equal to the volume that was removed from the space around the bag in the rigid container. In some sampling points, where a risk of condensation in the sampling bag existed due to high humidity and high temperatures, a predilution was applied using dry odor-free nitrogen. Sampling materials used were Teflon™ for tubing and disposable sampling bags made of Nalophan™ film. All samples were analyzed by the accredited odor laboratory of PRA OdourNet bv (The Netherlands) within 30 hours after sampling. During transportation, samples were not exposed to direct sunlight. All measurements of the odor concentrations were executed by dynamic olfactometry according to the EN13725 (CEN, 2003). 4.1.2b. Principle of dynamic olfactometry. The odor concentration of a gaseous sample of odorants is determined by presenting a panel of selected and screened human subjects with that sample, varying the concentration by diluting with neutral gas, in order to determine the dilution factor at the 50% detection threshold. At that dilution factor, the odor concentration is 1 ouE m−3 by definition. The odor concentration of the examined sample is then expressed

148

B. Sercu et al.

as a multiple of one European odor unit per cubic meter (ouE m−3) at standard conditions for olfactometry. 4.2. Odor Removal at a Vegetable Oil Extraction Plant 4.2.1. Background (Van Elst and Van Langenhove, 2001) Crushing and extraction plants for vegetable oils often cause considerable emissions of odor, which may cause offense in nearby residential areas. The type of oil seed processed partially determines the amount and type of odor released. Only limited information is found in literature on the composition of these waste gases, on source strength classification, and on possible treatment methods. Lacoste et al. (1996) performed a quantitative study of odorous compounds in gas effluents from three rapeseed crushing plants. Olfactometry was used to determine odor concentrations of gaseous effluents. Chemical analyses revealed the presence of nitriles, aldehydes, and sulfur compounds like mercaptans in conditioning and pressing emissions, while hydrogen sulfide and acetaldehyde were the major odorant compounds in the absorption unit effluents. An oil crushing and extraction plant, situated in an industrial area in the northern part of France, mainly processes soybean, sunflower, and rapeseed to produce vegetable oils. Significantly higher odor emissions occurred when processing rapeseed (colza) compared to those associated with other types of oil seed. Because of the growing number of complaints arising from the surrounding residential area at distances of more than one kilometre, the plant management decided to tackle the odor problem. The process of improving the odor situation in the vicinity of this plant was a process that took several years and included various types of measurements and interim evaluations. The measurements were a combination of chemical analyses, olfactometry, and field panels. Chemical measurements (gas chromatography, combined with mass spectrometry) were mainly used to get a better understanding of the composition of the different waste gas streams on the plant. Different gas streams have been sampled to identify and quantify the VOCs present. In the interpretation, special emphasis was put on compounds with a low odor threshold. Olfactometric measurements were used to determine the total amount of odour present in the waste gas stream. These data were very useful to make a classification of the different sources in order to set priorities for abatement but also to calculate the total odor abatement efficiency of treatment systems. Field panel measurements determined the impact of the total odor emission on the vicinity of the plant. 4.2.2. Identification of the Main Sources The first step in the processing of the seeds consists of a number of physical treatments, like cleaning, crushing, heating up to 60°C, pressing, and cooling. The main odor sources in this treatment are the hot and humid vapors that

149

Odors Treatment: Biological Technologies

arise at the heating stage and during pressing. Also the cooling of the material before entering the extraction unit, which takes place on the open conveyor belt between the crushing and the extraction unit, can be considered as an important source. The residual oil is then extracted from the flakes with hexane in an extraction unit. Hexane and oil are separated in a distillation unit. The residual fraction of the seed is treated in a desolventizer to remove hexane. Before being vented in the atmosphere, the vapors of the extraction process pass through an absorption system with mineral oil to recapture hexane. The emissions of the absorption contain high concentrations of hydrogen sulfide, and thus represent an important odor source. The extracted residue of the seed is dried and cooled and sold as livestock feed. Large amounts of fresh air are used in this process and are emitted to the atmosphere, loaded with odorous components. Olfactometric emission measurements of a selected number of odor sources resulted in the following emissions (see Table 2). The vapors of the conditioners were already incinerated in both steam boilers with an odor removal efficiency higher than 90%. The resulting calculated emissions in European odor units per hour demonstrate the importance of the emission of the absorption unit on one hand (72%) and of the drying-cooling unit on the other hand (18%). Chemical measurements were carried out on the six sampled emission points. Table 3 gives an overview of the compounds per chemical group. The data in Table 3 are the emitted mass flows per hour for five sampling points (conditioners not included). The results of the chemical measurements show an important contribution of mainly organic sulfur compounds and hydrogen sulfide to the total odor concentration, considering their low odor threshold. Though no data were found on the odor threshold of the specific nitriles and 4-isothiocyanato-1-butene, there might be an important influence of these compounds to the global odor concentration. The high hydrocarbon content in absorption, conveyor belt, and drying-cooling are mainly caused by the presence of hexane as extraction solvent (hexane, 2-methylpentane, 3-methylpentane, cyclohexane, methylcyclopentane).

Table 2. Odor Concentrations Measured at Six Points in the Process

Emission point Vapors of conditioners Exhaust steam boiler Presses Absorption unit Conveyor belt from extraction to drying unit Drying/cooling unit

Odor concentration (ouE m−3)

Total flow rate per source (m3 h−1)

Total odor emission (ouEh−1)

Percentage of measured emission (%)

46 × 103 0.9 × 103 425 × 103 5564 × 103

10300 32200 195 400

29 × 106 83 × 106 2225 × 106

0.9 2.7 72.1

515 × 103 8.8 × 103

350 64700

180.106 570 × 106

5.8 18.5

150

B. Sercu et al.

Table 3. Mass Flows in g h−1 for the Different Emission Points

Hydrocarbons Aldehydes Ketones Alcohols Nitriles Organic sulfur compounds 4-isothio-cyanato-1-butene Hydrogen sulfide

Boiler exhaust

Presses

Absorption

Conveyor belt

Dryingcooling

0.27 1.2 0.35 0.35 -

3.1 12 1.5 7.7 17 6.5 0.7 0.89

470 25 1.2 35 1822

395 7.1 1 0.05 151 0.4 0.2 -

3940 60 16 3 3600 15 7 -

4.2.3. Abatement Techniques As a first abatement step, some high concentrated streams limited in volumetric flow were chosen to be incinerated in the existing steam boilers (i.e., waste gases coming from the absorption unit and presses). As a second step, a suitable technique was chosen for the flow coming from the drying-cooling unit. This waste gas stream, relatively low in concentration but high in volumetric flow, was decided to be treated in a biofilter system after doing some pilot tests. A biofilter combined with a scrubber, designed for a flow of 100.000 m3 h−1, was constructed by Monsanto EnviroChem systems and operation started in April 1998. A collecting chamber was installed to receive all waste gas flows. The purpose of this chamber was to create a velocity drop and separate residual dust. The scrubber had three main purposes: capture of small dust particles, humidification of the airstream up to 100%, and cooling up to 37 °C. The biofilter itself is a closed, top-down model. The biofilter material consists of small polystyrene balls surrounded with compost. Inside of the filter are two stages, each subdivided in different compartments, with load measuring cells. These cells measure the weight of the compartment, and depending on the weight, additional water can be sprinkled if dehydration is stated. Figure 11 shows the scrubber, the front side of the biofilter, and the extraction fan and the silencer, both positioned behind the biofilter. Triplicate olfactometric control measurements (June and July 1998) of the ingoing and outgoing odor concentrations, as well as the chemical composition of both flows, confirmed the good odor removal efficiency of the complete system, with a low residual odor concentration (see Table 4). The chemical composition of the waste gas at the outlet of the biofilter only revealed hydrocarbons (hexane, etc.) above the detection limit. Afterward (August 1998), some smaller but concentrated waste gas streams were added to the collecting chamber, which resulted in a complete solution of the odor problem after three years of analyses and investments. Table 4 shows that though the inlet concentration increased over the different measurements, the outlet concentrations were relatively constant.

151

Odors Treatment: Biological Technologies

Figure 11. Scrubber and biofilter (left); fan and silencer (right).

Table 4. Average Odor Concentrations Measured at the Biofilter Date

Concentration at inlet scrubber (ouE m-3)

Concentration at outlet biofilter (ouE m-3)

Removal efficiency (%)

June 1998 July 1998 July 1998 Sept 1998

3 935 11 787 33 914 165 537

622 1 670 1 964 1 277

84.2 85.8 94.2 99.2

A value between 1000 and 2000 (ouE m-3) can be considered as a normal background value for the typical “own smell” of a good working biofilter. In some cases, still lower values are possible (up to 500 ouE m-3). 4.3. Odor Removal at an Animal Rendering Plant Rendering is the transformation of animal by-products into stable products mainly by evaporation of the water and separation of the fat. Fresh animal by-products start to decompose as soon as the animal has been slaughtered into mainly volatile substances through anaerobic processes often initiated by the bacteria of the stomach and intestinal contents. The volatile substances are set free when the raw material is heated and dried (water evaporation). They are found in the water vapor and part of them are condensed with the water to be treated in wastewater treatment plants, whereas others remain in the gaseous phase (noncondensables) depending on the vapor pressure under the given condensation conditions (Oberthür and Vossen, 2001). The odorous substances from rendering originate mainly from the proteins in the animal byproducts through anaerobic decomposition. The main constituents of rendering odors are hydrogen sulfide, ammonia, organic sulfides, aldehydes, organic acids, and other minor compounds, which due to their low odor threshold,

152

B. Sercu et al.

however, might contribute in a characteristic way to the rendering odor (VDI, 1996). Due to the nature of the material processed, animal rendering activities thus result in the emission of volatiles and disgusting odors, causing nuisance in the factory’s neighborhood (De Roo and Van Langenhove, 2000). Several technologies such as thermal or catalytic combustion, stage scrubbers, and biofilters may be used for the elimination of volatiles from waste gases. In the investigated rendering plant, the noncondensable gases are incinerated in the steam boilers, resulting in a highly efficient odor removal. The other odorous waste gases are treated in different biofilters, each preceeded with a scrubber using normal water as scrubbing liquid. The so-called category one material (cadavers, destruction blood, slaughterhouse by-products) is processed in one production line (ca. 300.000 ton year–1). On this line, two large, conventional biofilter units are in use. Biofilter 1 treats the waste air coming from the “clean zone.” General building extraction is used to avoid diffusive emissions, as well as point suction on all process units. Biofilter 1 is divided in two parts (1A and 1B); both airflows could be monitored separately. Biofilter 2 treats the waste air coming from the “unclean zone” (building extraction and point suction on breakers, pasteurization tanks, buffer tanks, etc.). In a separate production line, animal by-products of poultry are processed for use in petfood (ca. 100.000 tonnes year–1). Biofilter 3 treats the air coming from this separate poultry line. In the period between 1998 and 2002 the in- and effluent gas flows have been monitored by olfactometry and GC-MS analyses (unpublished reports Project Research Gent). Table 5 gives an overview of the different chemical compounds found in the waste gas streams. Defoer et al. (2002) showed that the presence and concentration of the organic sulfur compounds is determining for the total odor concentration of the flow. A direct correlation could be established. For this reason, only the total

Table 5. Overview of Different Chemical Compounds in Rendering Air Compound class

Chemicals identified

Hydrocarbons

Pentane, hexane, heptane, octane, nonane, decane, undecane, dodecane, tridecane, 2-methylpentane, 3-methylpentane, benzene, toluene, ethylbenzene, o, m, p-xylene, methylethylcyclohexane, methylcyclopentane Ethanol, 3-methylbutanol Dimethyl sulfide, dimethyl disulfide, dimethyl trisulfide, carbon disulfide Trifluoromethylbenzene, 1-chlorobutane, tetrachloroethylene, dichloromethane 2-Methyl-1,3-dioxolane Furane, 2-methylfurane Acetone, 2-butanone 3-Methylbutanal, 2-methylbutanal, n-hexanal, isobutyraldehyde, benzaldehyde

Alcohols S-Compounds Halogenated VOCs Ethers Furanes Ketones Aldehydes

153

Odors Treatment: Biological Technologies

Table 6. Overview of the Results Odour in Odour out (103 ouE m−3) BF 1A

1998 2000 1998 2000 2000 2002 9/1999 12/1999 2000 2002

BF1B BF2 BF3

638 118 267 48 172 163 1339 128 240 19

VOSCs in

14 1.7 5 3.8 48 4 297 3.5 84 6

1815 322 130 100 569 400 4599 193 931

VOSCs out (µg m−3) 200 12 200 10 438 7 3277 15 531 Not measured

odor concentration (expressed in ouE m-3) and the concentration of volatile organic sulfur compounds (expressed in ouE m-3) are shown in Table 6. This set of data shows that even with high influent odor concentrations, low outlet concentrations can be reached. However, it seems to be difficult to reach a “normal” background value situating between 1000 and 2000 ouE m-3. This can be caused by two factors: ●

●

the influent concentrations are quite variable due to different processes; peak loads can be negative for the efficiency of the biofilter; and the presence of the volatile organic sulfur compounds cause the typical smell of the waste air; as mentioned above VOSCs are less biodegradable, unless the biofilter is inoculated with sulfur-degrading microorganisms.

5. ACRONYMS AND ABBREVIATIONS Bv BPT DMS DMDS EBRT EC MM ppbv ppmv TEA VOC(s) VOSC(s)

Mass loading rate Biological plate tower Dimethyl sulphide Dimethyl disulphide Empty bed residence time Elimination capacity Methyl mercaptan Parts per billion volume Parts per million volume Triethyl amine Volatile organic compound(s) Volatile organic sulphur compounds(s)

154

B. Sercu et al.

6. REFERENCES Busca, G., Pistarino, C. (2003) “Abatement of ammonia and amines from waste gases: a summary.” 16:157. CEN (2003) EN 13725, Air quality — determination of odour concentration by dynamic olfactometry, European Standard, Brussels, Belgium. Chang, C.-T., Chen, B.-Y., Shiu, I.-S., Jeng, F.-T. (2004) Biofiltration of trimethylamine-containing waste gas by entrapped mixed microbial cells, Chemosphere 55:751. Chang, K., Lu, C. (2003) Biofiltration of isopropyl alcohol in a trickle-bed air biofilter, Biodegradation 14:9. Cho, K.-S., Hirai, M., Shoda, M. (1991) Degradation characteristics of hydrogen sulfide, methanethiol, dimethyl sulfide and dimethyl disulfide by Thiobacillus thioparus DW44 isolated from peat biofilter, J. Ferment. Bioeng. 71:384. Cho, K.-S., Hirai, M., Shoda, M. (1992) Enhanced removability of odorous sulphur-containing gases by mixed cultures of purified bacteria from peat biofilters, J. Ferment. Bioeng. 73:219. Cho, K.-S., Ryu, H. W., Lee, N. Y. (2000) Biological deodorization of hydrogen sulfide using porous lava as a carrier of Thiobacillus thiooxidans, J. Biosci. Bioeng. 90:25. Chou, M.-S., Shiu, W.-Z. (1997) Bioconversion of methylamine in biofilters, J. Air Waste Manage. Assoc. 47: 58. Chua, H., Li, X. Z., Yu, P. F. H., Tam, C. Y., Huang, Y. L., Yang, S. T. (2000) Design and performance of a fibrous bed bioreactor for odor treatment, Appl. Biochem. Biotechnol. 84–86:469. Chung, Y.-C., Huang, C., Liu, C. H., Bai, H. (2001a), Biotreatment of hydrogen sulfide- and ammoniacontaining waste gases by fluidized bed bioreactor, J. Air Waste Manage. Assoc. 51:163. Chung, Y.-C., Huang, C., Pan, J. R., Tseng, C.-P. (1998) Comparison or autotrophic and mixotrophic biofilters for H2S removal, J. Environ. Eng. April 1998:362. Chung, Y.-C., Huang, C. P., Tseng, C.-P. (1996a), Biodegradation of hydrogen sulfide by a laboratoryscale immobilized Pseudomonas putida CH11 biofilter, Biotechnol. Prog. 12:773. Chung, Y.-C., Huang, C. P., Tseng, C.-P. (1996b), Operation optimization of Thiobacillus thioparus CH11 biofilter for hydrogen sulfide removal, J. Biotechnol. 52:31. Chung, Y.-C., Huang, C. P., Tseng, C.-P. (1997) Removal of hydrogen sulphide by immobilized Thiobacillus sp. strain CH11 in a biofilter, J. Chem. Technol. Biotechnol. 69:58. Chung, Y.-C., Huang, C., Tseng, C.-P. (2001b), Biological elimination of H2S and NH3 from waste gases by biofilter packed with immobilized heterotrophic bacteria, Chemosphere 43:1043. Chung, Y.-C., Huang, C. P., Tseng, C.-P., Pan, J. R. (2000) Biotreatment of H2S- and NH3-containing waste gases by co-immobilized cells biofilter, Chemosphere 41:329. Chung, Y.-C., Lin, Y.-Y., Tseng, C.-P. (2004) Operational characteristics of effective removal of H2S and NH3 waste gases by activated carbon biofilter, J. Air Waste Manage. Assoc. 54:450. Cook, L. L., Gostomski, P. A., Apel, W. A. (1999) Biofiltration of asphalt emissions: full-scale operation treating off-gases from polymer-modified asphalt production, Environ. Prog. 18 :178. De Bo, I. (2002) Membrane biofiltration of single-compound waste gas streams, PhD thesis, Ghent University, Belgium. De Bo, I., Heyman, J., Vincke, J., Verstraete, W., Van Langenhove, H. (2003) Dimethyl sulfide removal from synthetic waste gas using a flat poly(dimethylsiloxane)-coated composite membrane bioreactor, Environ. Sci. Technol. 37:4228. De Bo, I., Jacobs, P., Van Langenhove, H. (2002) Removal of toluene and trichloroethylene from waste air in a membrane bioreactor, Environ. Sci. Pollution Res. Special Issue 3:28. Defoer, N., De Bo, I., Van Langenhove, H., Dewulf, J., Van Elst, T. (2002) Gas chromatographymass spectrometry as a tool for estimating odour concentrations of biofilter effluents at aerobic composting and rendering plants, J. Chromatogr. A 970:259. Degorce-Dumas, J. R., Kowal, S., Le Cloirec, P. (1997) Microbiological oxidation of hydrogen sulphide in a biofilter, Can. J. Microbiol. 43:264.

Odors Treatment: Biological Technologies

155

Demeestere, K., Van Langenhove, H., Smet, E. (2002) Regeneration of a compost biofilter degrading high loads of ammonia by addition of gaseous methanol, J. Air Waste Manage. Assoc. 52:796. De Roo, K., Van Langenhove, H. (2000) Emissions from animal rendering, in: Proceedings of the 7th symposium on analytical and environmental problems, Z. Galbacs, ed., Szeged, Hungary, October 2, 2000, p. 38. Devinny, J. S., Deshusses, M. A., Webster, T. S. (1999) Biofiltration for Air Pollution Control. CRC Press LLC, Boca Raton. Devos, M., Patte, F., Rouault, J., Laffort, P., Van Gembert, L. J. (1990) Standardized Human Olfactory Thresholds, Oxford University Press, Oxford, 165 p. Don, J. A. (1985) The rapid development of biofiltration for the purification of diversified waste gas streams, VDI Berichte 561:63. Elías, A., Barona, A., Ríos, F. J., Arreguy, A., Munguira, M., Peñas, J., Sanz, J. L. (2000) Application of biofiltration to the degradation of hydrogen sulfide in gas effluents, Biodegradation 11:423. Engesser, K.-H., Plaggemeier, T. (2000) Microbiological aspects of biological waste gas purification, in: Biotechnology, 2nd edition, Volume 11c. Environmental Processes III, Solid waste and waste gas treatment, preparation of drinking water, J. Klein, J., and J. Winter, eds., Wiley-VCH, Weinheim, pp. 275–302. Gabriel, D., Deshusses, M. A. (2003) Retrofitting existing chemical scrubbers to biotrickling filters for H2S emission control, Proc. Natl. Acad. Sci. U. S. A. 100:6308. Goodwin, J. P., Amenta, S. A., Delo, R. C., Del Vecchio, M., Pinnette, J. R., Pytlar, T. S. (2000) Odor control advances at cocomposting facility, Biocycle January 2000:68. Hartikainen, T., Ruuskanen, J., Vanhatalo, M., Martikainen, P. J. (1996) Removal of ammonia from air by a peat biofilter, Environ. Technol. 17:45. Heller, W., Schwager, H. (1996) Luftreinigung durch Kompost-Biofiltration, Agrarforschung 3:520. Huang, C. P., Chung, Y. C., Hsu, B. M. (1996) Hydrogen sulfide removal by immobilized autotrophic and heterotrophic bacteria in the bioreactors, Biotechnology Techniques 10:595. Ibrahim, M. A., Mizuno, H., Yasuda, Y., Fukunaga, K., Nakao, K. (2001) Removal of mixtures of acetaldehyde and propionaldehyde from waste gas in packed column with immobilized activated sludge gel beads, Biochem. Eng. J. 8:9. Jones, K., Martinez, A., Rizwan, M., Boswell, J. (2003) Evaluation of sulfur toxicity and media capacity for H2S removal in biofilters packed with both natural and commercial media, in: Proceedings of the 96th annual Air & Waste Management Association Conference, June 22–26, San Diego, CA. Joshi, J. A., Hogan, J. A., Cowan, R. M., Strom, P. F., Finstein, M. S. (2000) Biological removal of gaseous ammonia in biofilters: space travel and earth-based applications, J. Air Waste Manage. Assoc. 50:1647. Kalingan, A. E., Liao, C. M., Chen, J. W., Chen, S. C. (2004) Microbial degradation of livestockgenerated ammonia using biofilters at typical ambient temperatures, J. Environ. Sci. Health Part B-Pestic. Contam. Agric. Wastes 39:85. Kasakura, T., Tatsukawa, K. (1995) On the scent of a good idea for odour removal, Water Quality International 2:24. Kennes, C., Veiga, M.C. (2001) Bioreactors for Waste Gas Treatment, Kluwer Academic Publishers, Dordrecht. Kiared, K., Wu, G., Beerli, M., Rothenbühler, M., Heitz, M. (1997) Application of biofiltration to the control of VOC emissions, Environ. Technol. 18:55. Kim, N. J., Hirai, M., Shoda, M. (1998) Comparison of organic and inorganic carriers in removal of hydrogen sulfide in biofilters, Environ. Technol. 19:1233. Kim, H., Xie, Q., Kim, Y. J., Chung, J. S. (2002) Biofiltration of ammonia gas with sponge cubes coated with mixtures of activated carbon and zeolite, Environ. Technol. 23:839. Kirchner, K., Hauk, G., Rehm, H. J. (1987) Exhaust gas purification using immobilized monocultures (biocatalysts), Appl. Microbiol. Biotechnol. 26:579. Kirchner, K., Gossen, C. A., Rehm, H. J. (1991) Purification of exhaust air containing organic pollutants in a trickle-bed bioreactor, Appl. Microbiol. Biotechnol. 35:396.

156

B. Sercu et al.

Kosteltz, A. M., Finkelstein, A., Sears, G. (1996) What are the ‘real opportunities’ in biological gas cleaning for North America?, in: Proceedings of the 89th Annual Meeting & Exhibition of Air & Waste Management Association, A&WMA, Pittsburgh, PA; 96-RA87B.02. Lacoste, F., Bosque, F., Brenne, E. (1996) Monitoring atmospheric emissions from industrial colza trituration plants, Oleagineux Corps Gras Lipides 3:369. Liang, Y. K., Quan, X., Chen, J. W., Chung, J. S., Sung, J. Y., Chen, S., Xue, D. M., Zhao, Y. Z. (2000) Long-term results of ammonia removal and transformation by biofiltration, J. Hazard. Mater. 80:259. Luo, J. (2001) A pilot-scale study on biofilters for controlling animal rendering process odours, Wat. Sci. Technol. 44:277. Mohseni, M., Allen, D. G. (1999) Transient performance of biofilters treating mixtures of hydrophilic and hydrophobic volatile organic compounds, J. Air Waste Manage. Assoc. 49:1434. Mohseni, M., Allen, D. G. (2000) Biofiltration of mixtures of hydrophilic and hydrophobic volatile organic compounds”, Chem. Eng. Sci. 55:1545. Oberthür, R., Vossen, F. (2001) Intrinsic odour emission from rendering and the odour history of a rendering plant, in: Proceedings of the 1st IWA International Conference on Odour and VOCs: Measurement, Regulation and Control Techniques, The University of NSW, Sydney, Australia, 25 – 29th March, pp. 131–138. Otten, L., Afzal, M. T., Mainville, D. M. (2004) Biofiltration of odours: laboratory studies using butyric acid, Adv. Environ. Res. 8:397. Park, S.-J., Hirai, M., Shoda, M. (1993) Treatment of exhaust gases from a night soil treatment plant by a combined deodorization system of activated carbon fabric reactor and peat biofilter inoculated with Thiobacillus thioparus DW44, J. Ferment. Bioeng. 76:423. Park, S.-J., Nam, S.-I., Choi, E.-S. (2001) Removal of odor emitted from composting facilities using a porous ceramic biofilter, Wat. Sci. Technol. 44:301. Park, D. H., Cha, J. M., Ryu, H. W., Lee, G. W., Yu, E. Y., Rhee, J. I., Park, J. J., Kim, S. W., Lee, I. W., Joe, Y. I., Ryu, Y. W., Hur, B. K., Park, J. K., Park, K. (2002) Hydrogen sulfide removal utilizing immobilized Thiobacillus sp IW with Ca-alginate bead, Biochem. Eng. J. 11:167. Peixoto, J., Mota, M. (1997) Volatile organic compounds evaporation chamber for the simulation of gas effluents in laboratory research, Biotechnology Techniques 11:1. Peixoto, J., Mota, M. (1998) Biodegradation of toluene in a trickling filter, Bioproc. Eng. 19:393. Perry, R. H., Green, D. (1984) Perry’s Chemical Engineers’ Handbook, 6th ed., McGraw-Hill International Editions. Chemical Engineering Series, New York, 2234 p. Rafson, H. J. (1998) Odor and VOC Control Handbook . McGraw-Hill, New York. Regan, J. M., Harrington, G. W., Noguera, D. R. (2002) Ammonia- and nitrite-oxidizing bacterial communities in a pilot-scale chloraminated drinking water distribution system, Appl. Environ. Microbiol. 68:73. Reij, M. W., Keurentjes, J. T. F., Hartmans, S. (1998) Membrane bioreactors for waste gas treatment, J. Biotechnol. 59:155. Ritchie, B. J., Hill, G. A. (1995) Biodegradation of phenol polluted air using an external loop airlift bioreactor, J. Chem. Technol. Biotechnol. 62:339. Rittmann, B. E., Schwarz, A. O., Sáez P. B. (2000) Biofilms applied to hazardous waste treatment, in: Biofilms II, J. Bryers, ed., John Wiley & Sons, Inc., New York, pp. 207–234. Ruokojärvi, A., Ruuskanen, J., Martikainen, P. J., Olkkonen, M. (2001) Oxidation of gas mixtures containing dimethyl sulfide, hydrogen sulfide, and methanethiol using a two-stage biotrickling filter, J. Air Waste Manage. Assoc. 51:11. Schramm, A., De Beer, D., Gieseke, A., Amann, R. (2000) Microenvironments and distribution of nitrifying bacteria in a membrane-bound biofilm, Environ. Microbiol. 2:680. Sercu, B., Demeestere, K., Baillieul, H., Verstraete, W., Van Langenhove, H. (2005a), Degradation of isobutanal at high loading rates in a compost biofilter, J. Air Waste Manage. Assoc. 55:1217. Sercu, B., Núñez, D., Van Langenhove, H., Aroca, G., Verstraete, W. (2005b), Operational and microbiological aspects of a two-stage biotrickling filter removing hydrogen sulfide and dimethyl sulfide, Biotechnol. Bioeng. 90:259.

Odors Treatment: Biological Technologies

157

Sheridan, B. A., Curran, T. P., Dodd, V. A. (2003) Biofiltration of n-butyric acid for the control of odour, Bioresource Technol. 89:199. Shoda, M. (1991) Methods for the biological treatment of exhaust gases, in: Biological Degradation of Wastes, A. M. Martin, ed., Elsevier Science Publishers, New York, pp. 31–46. SIS, (2000) Scientific Instruments Services: adsorbent resins, general listing and description of seven resins; http://www.sisweb.com/index/referenc/resin2.htm Smet, E. (1995) Evaluatie werking biofilter – De Kleijn (Maaseik), Unpublished Report, Ghent University. Smet, E., Chasaya, G., Van Langenhove, H., and Verstraete, W. (1996a) The effect of inoculation and the type of carrier material used on the biofiltration of methyl sulphides, Appl. Microbiol. Biotechnol. 45:293. Smet, E., Lens, P., and Van Langenhove, H. (1998) Treatment of waste gases contaminated with odorous sulphur compounds, Crit. Rev. Environ. Sci. Technol. 28:89. Smet, E., Van Langenhove, H., and Verstraete, W. (1996b) Long-term stability of a biofilter treating dimethyl sulphide, Appl. Microbiol. Biotechnol. 46:191. Smet, E., Van Langenhove, H., Maes, K. (2000) Abatement of high concentrated ammonia loaded waste gases in compost biofilters, Water Air Soil Poll. 119:177. Smet, E., Van Langenhove, H., and Verstraete, W. (1997) Isobutyraldehyde as a competitor of the dimethyl sulfide degrading activity in biofilters, Biodegradation 8:53. Smits, M. C. J., Hoek, A. P., Osinga, B., Ottengraf, S. P. P., and Wijngaard, M. H. (1995) Removal of ammonia with odour from composting facility waste gas with a bio-trickling filter, Meded. Fac. Landbouww. Gent 60:2313. Tang, H.-M., Hwang, S.-J., and Hwang, S.-C. (1996) Waste gas treatment in biofilters, J. Air Waste Manage. Assoc. 46:349. Terasawa, M., Hirai, M., and Kubota, H. (1986) Soil deodorization systems, Biocycle 27:28. Van Elst, T., and Van Langenhove, H. (2001) Case study: odour emission at a vegetable oil plant in France, in: Proceedings of the 1st IWA International Conference on Odour and VOCs: Measurement, Regulation and Control Techniques, The University of NSW, Sydney, Australia, 25–29th March, pp. 91–398. Van Groenestijn, J. W., and Hesselink, P. G. M. (1993) Biotechniques for air pollution control, Biodegradation 4:283. Van Langenhove, H., Bendinger, B., Oberthür, R., and Schamp, W. (1992) Organic sulfur compounds: persistant odourants in the biological treatment of complex waste gases, in: Biotechniques for Air Pollution Abatement and Odour Control Policies, A. J. Dragt, and J. van Dam, Eds., Elsevier, Amsterdam, pp. 309–313. Van Langenhove, H., Cornelis, C., Schamp, N., and Wuyts, E. (1989a) Biofiltration in a vegetables processing industry, Meded. Fac. Landbouww. Gent 54:1471. Van Langenhove, H., Lootens, A., and Schamp, N. (1988) Elimination of ammonia from pigsty ventilation air by wood bark biofiltration, Meded. Fac. Landbouww. Gent 53:1963. Van Langenhove, H., Lootens, A., and Schamp, N. (1989b) Inhibitory effects of SO2 on biofiltration of aldehydes, Water Air Soil Poll. 47:81. Van Lith, C., Leson, G., and Michelsen, R. (1997) Evaluating design options for biofilters, J. Air Waste Manage. Assoc. 47:37. VDI (1996) Emissionsminderung, Anlagen zur Verwertung und Beseitigung von Tierkörperteilen und tierischen Erzeugnissen, Düsseldorf, Verein Deutscher Ingenieure Wada, A., Shoda, M., Kubota, H., Kobayashi, T., Katayama-Fujimura, J., and Kuraishi, H. (1986) Characteristics of H2S oxidizing bacteria inhabiting a peat biofilter, J. Ferment. Technol. 64: 161. Wani, A. H., Lau, A. K., and Branion, R. M. R. (1999) Biofiltration control of pulping odors – hydrogen sulfide: performance, macrokinetics and coexistence effects of organo-sulfur species, J. Chem. Technol. Biotechnol. 74:9. Warren, J., Loerh, C., and Fellow A. S. C. E. (1997) Biofiltration: fundamentals, design and operation principles and applications, J. Environ. Eng. 123:538. Waweru, M., Herrygers, V., Van Langenhove, W., and Verstraete, W. (2000) Process engineering of biological waste gas purification, in: Biotechnology, 2nd edition, Volume 11c. Environmental Pro-

158

B. Sercu et al.

cesses III, Solid Waste and Waste Gas Treatment, Preparation of Drinking Water, J. Klein, and J. Winter, eds., Wiley-VCH, Weinheim, pp. 260–273. Weast, R. C., Astle, M. J., and Beyer, W. H. (1984) CRC. Handbook of Chemistry and Physics, 64th ed., CRC Press, Inc., Boca Raton, Florida, 2303 p. Weckhuysen, B., Vriens, L., and Verachtert, H. (1993) The effect of nutrient supplementation on the biofiltration removal of butanal in contaminated air, Appl. Microbiol. Biotechnol. 39:395. Weckhuysen, B., Vriens, L., and Verachtert, H. (1994) Biotreatment of ammonia-containing and butanal-containing waste gases, Appl. Microbiol. Biotechnol. 42:147. Williams, T. O., 1995, Odors and VOC emissions control methods, Biocycle 36:49. Wu, L., Loo, Y.-Y., and Koe, L. C. C. (2001) A pilot study of a biotricking filter for the treatment of odorous sewage air, Wat. Sci. Technol. 44:295. Yang, Y., and Allen, E. R. (1994a) Biofiltration control of hydrogen sulfide 1. Design and operational parameters, J. Air Waste Manage. Assoc. 44:863. Yang, Y., and Allen, E. R. (1994b) Biofiltration control of hydrogen sulfide 2. Kinetics, biofilter performance, and maintenance, J. Air Waste Manage. Assoc. 44:1315. Yang, Y., Togna, A. P., and Skladany, G. (1994) Treatment of CS2 and H2S vapors by biofiltration, in: Proceedings of the 49th Purdue Industrial Waste Conference, Purdue University, pp. 449–456. Yani, M., Hiral, M., and Shoda, M. (1998) Ammonia gas removal characteristics using biofilter with activated carbon fiber as a carrier, Environ. Technol. 19:709. Zhang, L., Hirai, M., and Shoda, M. (1991) Removal characteristics of dimethyl sulfide, methanethiol and hydrogen sulfide by Hyphomicrobium sp. I55 isolated from peat biofilter, J. Ferment. Bioeng. 72:392.

Index

Page numbers followed by f and t indicate figures and tables, respectively.

Absorption, 30, 41, 43, 81, 88, 89t, 108t, 109, 110t, 111f, 117, 127, 130–131, 138–139, 148–150 Activated carbon, 34, 42, 111–113, 114t, 127, 138, 142 Adaptation, 2, 33 Adsorbent, 42–43, 112–113, 114f, 146–147 Adsorbing, 41, 113 Adsorption, 28, 32–33, 42–43, 81, 89, 108, 111–113, 114f, 146–147 Advertisements, 84 Air pollution, 47–54, 102, 106, 109, 112, 132t Alcohols, 11, 58, 59t, 60, 64–66, 70, 80, 112, 126, 144–145, 150t, 152t Aldehydes, 11, 23t, 59t, 60, 64–66, 80t, 81, 107, 108t, 112, 121, 126, 144–146, 148, 150t, 151, 152t Amines, 6, 10–11, 23t, 59t, 62, 64–66, 68, 70, 107, 108t, 109, 112, 126, 140 Ammonia, 6–8, 10, 18, 23t, 25, 29, 62, 65–66, 79–80, 97, 109, 127, 137–140, 142–143 Ammonia-like stockfish odor, 64f Animal feeding, 7, 9–10 Animal rendering, 10, 151–153 Area sources, 18, 20 Artificial neural network, 51–54 Biodegradation, 54, 80t, 95, 125, 145 Biofilm, 126f, 136f–137f Biofilter, 128f, 129–130, 137f

Biofiltration, 108t, 127, 137–142 Biomass, 80t, 84, 130–131, 133f, 134–135, 139–140 Bioscrubber, 130–131, 139 Biotrickling filter, 129–130 Biowasher, 127, 130–132 Biological plate tower (BPT), 127, 131, 133–135 Brewery, 92, 116 Buffer areas, 96–100 Building Sources, 20–21 Bulk acoustic wave, 50–51 Cane sugar processing, 9t, 11 Capelin, 62, 63t, 66 Capture-recovery, 106 Case study, 64–70 Catalytic metal, 50 Catalytic oxidation, 119–121 Catalytic oxidizer, 119–121 Catalytic combustion, 152 Chemical analysis, 24, 68 Chemical compounds, 49, 88, 152 Chemical gas sensors, 49, 64–65 Cleaner production, 84, 92f, 93–95, 97, 106 Clogging, 42, 128–130, 132–133, 136–137 Cluster analysis, 51–52 Cod, 59t, 62, 63t, 66 Cod roe, 63t Colorimetric Detector Tubes, 27–28 Condensation, 18, 88–89, 108t, 115–116, 147, 151 159

160 Conducting polymer sensors (CP), 51 Conducting polymers, 51 Cryotrapping, 41 Deodorization, 127 Destructive methods, 107–108 Development plans, 96–100 Dilution to threshold (D/T) ratios, 22, 34 Direct scaling technique, 31 Dispersion, 91, 97, 101–102, 107 Dynamic olfactometry, 32–33, 36, 146–147 Electrochemical gas sensors, 64–65 Electronic nose, 49f, 50–51, 57–70 Emission rates, 18–21 Environmental control, 53, 77 Environmental management systems (EMS), 84 Esters, 58f, 60, 64–66, 95 European reference odor mass (EROM), 22–23 Externalities, 77, 92f, 97 F.I.D.O.L., 76 Fatigue, 2, 33, 35, 54 FIFO, 86 Fingerprints, 53 Fish, 9t, 57–70 Fish Canning, 9, 11 Fish fillets, 66–70 Flame ionization detector (FID), 29, 30f Food industry, 1–12, 47–54, 75–102 Fourier transform infrared spectrometry (FTIR), 30 Fresh fish odor, 59t Freshness, 53, 57, 60–62, 63t, 64, 66–67, 70 Front-end control, 84 Frying, 79t, 87–88, 95 Gas chromatography (GC), 25–27, 41–44, 60, 65, 148 Gas chromatography-mass spectrometry (GC-MS), 65–66, 146–147 Gas chromatography- olfactometry (GCO), 26–27

Index

Herring, 63t High-performance liquid chromatography (HPLC), 43 Human health, 3 Human olfactory system, 2, 21, 58 Human response, 1–3, 24 Hydrogen sulfide, 64f Incineration, 108t, 117–120 Indoor air quality, 81–82 Indoors, 3, 81 Industry activities, 8, 9t, 99 Isolation chamber, 18–19 Ketones, 11, 23t, 64, 80t, 109, 150t, 152t Linear discriminant analysis, 51–52 Manure processing, 9–10 Mass spectrometry, 26, 147–148 Meat rendering plants, 9t, 10–11 Membrane bioreactor, 127, 136–137 Metal oxide semiconductor field effect transistors, 50 Metal oxide semiconductors, 50 Metaloxide sensors (MOS), 50–51, 63t Microbial counts, 61, 63t, 70 Micro-organisms, 126, 129–132, 136, 138, 141, 143–144, 153 Multivariate analysis, 49, 51 Nitrification, 135f, 138, 139–140 Non destructive methods, 107 Nuisance, 1, 3–6, 10–12 Objective measurements, 31–32 Odor character, 22, 26, 31, 98 Odor concentration, 2, 16, 20–24, 33–35, 91, 99, 100, 143, 146–150, 151t, 152–153 Odor control, 12, 15, 54, 79t, 88, 90, 98, 106–107, 109, 111, 113 Odor guidelines, 3–6 Odor intensity, 2, 12, 21–22, 31, 33 Odor mixtures, 145–146 Odor removal, 26, 115, 135, 145–146, 148–153 Odor sampling, 16, 17f, 36, 101 Odor sources, 16, 78, 148–149 Odor unit, 4, 22–24, 100, 148–149

161

Index

Off-odor, 47, 75, 80, 106 Olfactometer, 20–21, 32–33, 34f, 35–36 Olfactometry, 20, 22, 26–27, 32–33, 35–36, 60, 146–148, 152 Olfactometry Panel, 33–34 Organic sulfur compounds, 7, 126, 142–144, 149, 150t, 152–153 Organoleptic, 1, 21 Outdoors, 81–83 Oxidation, 58, 60, 62, 63t, 108–109, 117–121, 131, 141, 144 Paper tape monitors, 28 Partial least square, 51–52 Pattern recognition, 48–49, 52 pH sensitive films, 62–64 Phenolic compounds, 8 Photoionization detector (PID), 25, 29–30 Point sources, 17–18, 20 Portable analyzers, 24, 28–30 Portable gold leaf analyzers, 28 Preconcentration, 41–44, 65, 146–147 Pre-concentration Prevention, 75–102 Preventive odor emission control technologies, 83–102 Preventive managing procedures, 95–96 Primary pollutants, 10, 78 Principal component analysis (PCA), 51–52, 69f Properties of odors, 31, 78 Quality, 57–70, 81–83 Quality indicators, 64f Quartz microbalance, 51, 63t Quartz microbalance sensors (QMB), 51 Rapid evaluation of fish quality, 61 Recuperative thermal oxidation systems, 118, 119f Recycling, 84, 90, 92f, 95, 107, 115 Reformulation, 84–85, 94–95 Regeneration, 112–113, 114t, 131, 141 Regenerative thermal oxidizer (RTO), 118–120 Regulations, 3–6, 61, 77, 102, 107 Rendering, 10–11, 80, 90, 111, 140, 143, 151–153

Retrofitting, 92f Reuse (of solvent), 113 RTO, 118–120 Salmon, 60, 63t Scentometer, 34–35 Scrubber, 90, 108t, 109, 110t, 111, 130–131, 139, 142, 150, 151f, 152 Seasonal variation, 67 Sensing system, 49 Sensor array, 47–54 Sensory analysis, 62, 63t, 67–68 Sensory methods, 24, 31–37 Shifting, 84–85 Shrimp, 63t Solid-phase microextraction (SPME), 43–44 Sorbent bed, 42 Sorbent sampling, 43 Sour and putrid odors, 23t, 58f, 59t, 64f, 66 Specific odor emission rate, 20–21 Spoilage, 57, 59t, 60–62, 64–70 Storage studies, 67–68, 69f Subjective measurements, 31 Sulfur compounds, 7, 58f, 64f–65f, 66, 68, 70, 126, 137, 142–144, 149 Sulphur oxidation, 141, 144 Surface acoustic wave, 51 Sweet- and fruity odors, 66 Tainting, 81 Texture measurements, 67 Threshold odor numbers (Ton), 22 TMA (trimethylamine), 11, 23t, 59t, 60, 62, 63t, 64f, 65–66, 69f Total volatile basic nitogen (TVB-N), 63t, 64, 67, 69f, 70 Trapping techniques, 42 Trout, 63t Tuna, 60t, 63t Vegetable oil extraction, 148–151 Ventilation, 20–21, 81–83, 86, 88, 90–91, 95, 132 Volatile compounds, 27, 41–43, 48–49, 51, 53, 57–62, 64–66, 81, 111 Volatile fatty acids (VFA), 7–8, 126, 145

162 Volatile organic compound (VOC), 7–8, 10–12, 29, 43–44, 87, 107, 108t, 116–117, 119–121, 125, 135 Volatile organic sulphur compound (VOSC), 143–144, 153

Index

Volatiles, 27, 50, 54, 60, 62, 65, 146, 152 Waste gas, 125–127, 131, 136–141, 143, 145, 147–148, 150, 152 Wind tunnel system, 18–20

Food Safety: A Practical and Case Study Approach (Integrating Safety and Environmental Knowledge Into Food Studies towards European Sustainable Development)

Experiments in Unit Operations and Processing of Foods (Integrating Safety and Environmental Knowledge Into Food Studies towards European Sustainable Development)

Odors In the Food Industry (Integrating Safety and Environmental Knowledge Into Food Studies towards European Sustainable Development)

Food Safety: A Practical and Case Study Approach (Integrating Safety and Environmental Knowledge Into Food Studies towards European Sustainable Development)

Experiments in Unit Operations and Processing of Foods (Integrating Safety and Environmental Knowledge Into Food Studies towards European Sustainable Development)

Predictive Modeling and Risk Assessment (Integrating Safety and Environmental Knowledge Into Food Studies towards European Sustainable Development)

Food Safety in the International Hospitality Industry

Oils and Fats in the Food Industry (Food Industry Briefing)

Case Studies in Food Microbiology for Food Safety and Quality

Food Safety

Food Safety Culture: Creating a Behavior-Based Food Safety Management System (Food Microbiology and Food Safety)

Sustainable Agriculture and Food

Safe Food: The Politics of Food Safety, Updated and Expanded (California Studies in Food and Culture)

Safe Food, The Politics Of Food Safety, Updated And Expanded (California Studies In Food And Culture)

Environmental Regulation And Food Safety: Studies of Protection And Protectionism

Fermentation and Food Safety

Safe Food, The Politics Of Food Safety, Updated And Expanded (California Studies In Food And Culture)

Bioterrorism and Food Safety

Fermentation and Food Safety

Fermentation and Food Safety

Auditing in the food Industry

Foodborne Parasites (Food Microbiology and Food Safety)

Safe Food: The Politics of Food Safety

Food Safety Control in the Poultry Industry (Woodhead Publishing in Food Science and Technology)

Food and Environmental Analysis

Benchmarking in Food and Farming (Gower Sustainable Food Chains Series)

Advances in Food Protection: Focus on Food Safety and Defense

PCR Methods in Foods (Food Microbiology and Food Safety)

Interdisciplinary Food Safety Research

Food Safety Hazard Guidebook

Food Safety in Shrimp Processing

Food Safety and International Competitiveness:

Occupational Safety and Health Simplified for the Food Manufacturing Industry

Odors In the Food Industry (Integrating Safety and Environmental Knowledge Into Food Studies towards European Sustainable Development)

Food Safety: A Practical and Case Study Approach (Integrating Safety and Environmental Knowledge Into Food Studies towards European Sustainable Development)

Experiments in Unit Operations and Processing of Foods (Integrating Safety and Environmental Knowledge Into Food Studies towards European Sustainable Development)

Predictive Modeling and Risk Assessment (Integrating Safety and Environmental Knowledge Into Food Studies towards European Sustainable Development)

Food Safety in the International Hospitality Industry

Oils and Fats in the Food Industry (Food Industry Briefing)

Case Studies in Food Microbiology for Food Safety and Quality

Food Safety

Food Safety Culture: Creating a Behavior-Based Food Safety Management System (Food Microbiology and Food Safety)

Sustainable Agriculture and Food

Safe Food: The Politics of Food Safety, Updated and Expanded (California Studies in Food and Culture)

Safe Food, The Politics Of Food Safety, Updated And Expanded (California Studies In Food And Culture)

Environmental Regulation And Food Safety: Studies of Protection And Protectionism

Fermentation and Food Safety

Safe Food, The Politics Of Food Safety, Updated And Expanded (California Studies In Food And Culture)

Bioterrorism and Food Safety

Fermentation and Food Safety

Fermentation and Food Safety

Auditing in the food Industry

Foodborne Parasites (Food Microbiology and Food Safety)

Safe Food: The Politics of Food Safety

Food Safety Control in the Poultry Industry (Woodhead Publishing in Food Science and Technology)

Food and Environmental Analysis

Benchmarking in Food and Farming (Gower Sustainable Food Chains Series)

Advances in Food Protection: Focus on Food Safety and Defense

PCR Methods in Foods (Food Microbiology and Food Safety)

Interdisciplinary Food Safety Research

Food Safety Hazard Guidebook

Food Safety in Shrimp Processing

Food Safety and International Competitiveness:

Occupational Safety and Health Simplified for the Food Manufacturing Industry

Recommend Documents