Air Pollution XIX (Transactions on Ecology and the Environment) (Wit Transactions on Ecology and the Environment)

Air Pollution XIX

WIT Press publishes leading books in Science and Technology. Visit our website for the current list of titles. www.witpress.com

WITeLibrary Home of the Transactions of the Wessex Institute. Papers presented at Air Pollution XIX are archived in the WIT eLibrary in volume 147 of WIT Transactions on Ecology and the Environment (ISSN 1743-3541). The WIT eLibrary provides the international scientific community with immediate and permanent access to individual papers presented at WIT conferences. http://library.witpress.com

NINETEENTH INTERNATIONAL CONFERENCE ON MODELLING, MONITORING AND MANAGEMENT OF AIR POLLUTION

AIR POLLUTION XIX CONFERENCE CHAIRMEN C. A. Brebbia Wessex Institute of Technology, UK J.W.S. Longhurst University of the West of England, UK

V. Popov Wessex Institute of Technology, UK

INTERNATIONAL SCIENTIFIC ADVISORY COMMITTEE A. Berezin C. Booth C. Borrego M. Jicha F. Patania E. Petrovsky R. San Jose

ORGANISED BY Wessex Institute of Technology, UK

SPONSORED BY WIT Transactions on Ecology and the Environment

WIT Transactions Transactions Editor Carlos Brebbia Wessex Institute of Technology Ashurst Lodge, Ashurst Southampton SO40 7AA, UK Email: [email protected]

Editorial Board B Abersek University of Maribor, Slovenia Y N Abousleiman University of Oklahoma,

G Belingardi Politecnico di Torino, Italy R Belmans Katholieke Universiteit Leuven,

P L Aguilar University of Extremadura, Spain K S Al Jabri Sultan Qaboos University, Oman E Alarcon Universidad Politecnica de Madrid,

C D Bertram The University of New South

USA

Spain

A Aldama IMTA, Mexico C Alessandri Universita di Ferrara, Italy D Almorza Gomar University of Cadiz, Spain

B Alzahabi Kettering University, USA J A C Ambrosio IDMEC, Portugal A M Amer Cairo University, Egypt S A Anagnostopoulos University of Patras, Greece

M Andretta Montecatini, Italy E Angelino A.R.P.A. Lombardia, Italy H Antes Technische Universitat Braunschweig, Germany

M A Atherton South Bank University, UK A G Atkins University of Reading, UK D Aubry Ecole Centrale de Paris, France H Azegami Toyohashi University of Technology, Japan

A F M Azevedo University of Porto, Portugal J Baish Bucknell University, USA J M Baldasano Universitat Politecnica de Catalunya, Spain J G Bartzis Institute of Nuclear Technology, Greece A Bejan Duke University, USA M P Bekakos Democritus University of Thrace, Greece

Belgium

Wales, Australia

D E Beskos University of Patras, Greece S K Bhattacharyya Indian Institute of Technology, India

E Blums Latvian Academy of Sciences, Latvia J Boarder Cartref Consulting Systems, UK B Bobee Institut National de la Recherche Scientifique, Canada

H Boileau ESIGEC, France J J Bommer Imperial College London, UK M Bonnet Ecole Polytechnique, France C A Borrego University of Aveiro, Portugal A R Bretones University of Granada, Spain J A Bryant University of Exeter, UK F-G Buchholz Universitat Gesanthochschule Paderborn, Germany

M B Bush The University of Western Australia, Australia

F Butera Politecnico di Milano, Italy J Byrne University of Portsmouth, UK W Cantwell Liverpool University, UK D J Cartwright Bucknell University, USA P G Carydis National Technical University of Athens, Greece

J J Casares Long Universidad de Santiago de Compostela, Spain

M A Celia Princeton University, USA A Chakrabarti Indian Institute of Science, India

A H-D Cheng University of Mississippi, USA

J Chilton University of Lincoln, UK C-L Chiu University of Pittsburgh, USA H Choi Kangnung National University, Korea A Cieslak Technical University of Lodz, Poland

S Clement Transport System Centre, Australia M W Collins Brunel University, UK J J Connor Massachusetts Institute of Technology, USA

M C Constantinou State University of New York at Buffalo, USA

D E Cormack University of Toronto, Canada M Costantino Royal Bank of Scotland, UK D F Cutler Royal Botanic Gardens, UK W Czyczula Krakow University of Technology, Poland

M da Conceicao Cunha University of Coimbra, Portugal

L Dávid Károly Róbert College, Hungary A Davies University of Hertfordshire, UK M Davis Temple University, USA A B de Almeida Instituto Superior Tecnico, Portugal

E R de Arantes e Oliveira Instituto Superior Tecnico, Portugal L De Biase University of Milan, Italy R de Borst Delft University of Technology, Netherlands G De Mey University of Ghent, Belgium A De Montis Universita di Cagliari, Italy A De Naeyer Universiteit Ghent, Belgium W P De Wilde Vrije Universiteit Brussel, Belgium L Debnath University of Texas-Pan American, USA N J Dedios Mimbela Universidad de Cordoba, Spain G Degrande Katholieke Universiteit Leuven, Belgium S del Giudice University of Udine, Italy G Deplano Universita di Cagliari, Italy I Doltsinis University of Stuttgart, Germany M Domaszewski Universite de Technologie de Belfort-Montbeliard, France J Dominguez University of Seville, Spain K Dorow Pacific Northwest National Laboratory, USA W Dover University College London, UK C Dowlen South Bank University, UK

J P du Plessis University of Stellenbosch, South Africa

R Duffell University of Hertfordshire, UK A Ebel University of Cologne, Germany E E Edoutos Democritus University of Thrace, Greece

G K Egan Monash University, Australia K M Elawadly Alexandria University, Egypt K-H Elmer Universitat Hannover, Germany D Elms University of Canterbury, New Zealand M E M El-Sayed Kettering University, USA D M Elsom Oxford Brookes University, UK F Erdogan Lehigh University, USA F P Escrig University of Seville, Spain D J Evans Nottingham Trent University, UK J W Everett Rowan University, USA M Faghri University of Rhode Island, USA R A Falconer Cardiff University, UK M N Fardis University of Patras, Greece P Fedelinski Silesian Technical University, Poland

H J S Fernando Arizona State University, USA

S Finger Carnegie Mellon University, USA J I Frankel University of Tennessee, USA D M Fraser University of Cape Town, South Africa

M J Fritzler University of Calgary, Canada U Gabbert Otto-von-Guericke Universitat Magdeburg, Germany

G Gambolati Universita di Padova, Italy C J Gantes National Technical University of Athens, Greece

L Gaul Universitat Stuttgart, Germany A Genco University of Palermo, Italy N Georgantzis Universitat Jaume I, Spain P Giudici Universita di Pavia, Italy F Gomez Universidad Politecnica de Valencia, Spain

R Gomez Martin University of Granada, Spain

D Goulias University of Maryland, USA K G Goulias Pennsylvania State University, USA

F Grandori Politecnico di Milano, Italy W E Grant Texas A & M University, USA

S Grilli University of Rhode Island, USA

R H J Grimshaw Loughborough University,

K L Katsifarakis Aristotle University of

D Gross Technische Hochschule Darmstadt,

J T Katsikadelis National Technical

R Grundmann Technische Universitat

E Kausel Massachusetts Institute of

A Gualtierotti IDHEAP, Switzerland R C Gupta National University of Singapore,

H Kawashima The University of Tokyo,

UK

Germany

Dresden, Germany

Singapore J M Hale University of Newcastle, UK K Hameyer Katholieke Universiteit Leuven, Belgium C Hanke Danish Technical University, Denmark K Hayami University of Toyko, Japan Y Hayashi Nagoya University, Japan L Haydock Newage International Limited, UK A H Hendrickx Free University of Brussels, Belgium C Herman John Hopkins University, USA I Hideaki Nagoya University, Japan D A Hills University of Oxford, UK W F Huebner Southwest Research Institute, USA J A C Humphrey Bucknell University, USA M Y Hussaini Florida State University, USA W Hutchinson Edith Cowan University, Australia T H Hyde University of Nottingham, UK M Iguchi Science University of Tokyo, Japan D B Ingham University of Leeds, UK L Int Panis VITO Expertisecentrum IMS, Belgium N Ishikawa National Defence Academy, Japan J Jaafar UiTm, Malaysia W Jager Technical University of Dresden, Germany Y Jaluria Rutgers University, USA C M Jefferson University of the West of England, UK P R Johnston Griffith University, Australia D R H Jones University of Cambridge, UK N Jones University of Liverpool, UK D Kaliampakos National Technical University of Athens, Greece N Kamiya Nagoya University, Japan D L Karabalis University of Patras, Greece M Karlsson Linkoping University, Sweden T Katayama Doshisha University, Japan

Thessaloniki, Greece

University of Athens, Greece Technology, USA Japan

B A Kazimee Washington State University, USA

S Kim University of Wisconsin-Madison, USA D Kirkland Nicholas Grimshaw & Partners Ltd, UK

E Kita Nagoya University, Japan A S Kobayashi University of Washington, USA

T Kobayashi University of Tokyo, Japan D Koga Saga University, Japan S Kotake University of Tokyo, Japan A N Kounadis National Technical University of Athens, Greece

W B Kratzig Ruhr Universitat Bochum, Germany

T Krauthammer Penn State University, USA C-H Lai University of Greenwich, UK M Langseth Norwegian University of Science and Technology, Norway

B S Larsen Technical University of Denmark, Denmark

F Lattarulo Politecnico di Bari, Italy A Lebedev Moscow State University, Russia L J Leon University of Montreal, Canada D Lewis Mississippi State University, USA S lghobashi University of California Irvine, USA

K-C Lin University of New Brunswick, Canada

A A Liolios Democritus University of Thrace, Greece

S Lomov Katholieke Universiteit Leuven, Belgium

J W S Longhurst University of the West of England, UK

G Loo The University of Auckland, New Zealand

J Lourenco Universidade do Minho, Portugal J E Luco University of California at San Diego, USA

H Lui State Seismological Bureau Harbin, China

C J Lumsden University of Toronto, Canada L Lundqvist Division of Transport and

Location Analysis, Sweden T Lyons Murdoch University, Australia Y-W Mai University of Sydney, Australia M Majowiecki University of Bologna, Italy D Malerba Università degli Studi di Bari, Italy G Manara University of Pisa, Italy B N Mandal Indian Statistical Institute, India Ü Mander University of Tartu, Estonia H A Mang Technische Universitat Wien, Austria G D Manolis Aristotle University of Thessaloniki, Greece W J Mansur COPPE/UFRJ, Brazil N Marchettini University of Siena, Italy J D M Marsh Griffith University, Australia J F Martin-Duque Universidad Complutense, Spain T Matsui Nagoya University, Japan G Mattrisch DaimlerChrysler AG, Germany F M Mazzolani University of Naples “Federico II”, Italy K McManis University of New Orleans, USA A C Mendes Universidade de Beira Interior, Portugal R A Meric Research Institute for Basic Sciences, Turkey J Mikielewicz Polish Academy of Sciences, Poland N Milic-Frayling Microsoft Research Ltd, UK R A W Mines University of Liverpool, UK C A Mitchell University of Sydney, Australia K Miura Kajima Corporation, Japan A Miyamoto Yamaguchi University, Japan T Miyoshi Kobe University, Japan G Molinari University of Genoa, Italy T B Moodie University of Alberta, Canada D B Murray Trinity College Dublin, Ireland G Nakhaeizadeh DaimlerChrysler AG, Germany M B Neace Mercer University, USA D Necsulescu University of Ottawa, Canada F Neumann University of Vienna, Austria S-I Nishida Saga University, Japan H Nisitani Kyushu Sangyo University, Japan B Notaros University of Massachusetts, USA

P O’Donoghue University College Dublin, Ireland

R O O’Neill Oak Ridge National Laboratory, USA

M Ohkusu Kyushu University, Japan G Oliveto Universitá di Catania, Italy R Olsen Camp Dresser & McKee Inc., USA E Oñate Universitat Politecnica de Catalunya, Spain

K Onishi Ibaraki University, Japan P H Oosthuizen Queens University, Canada E L Ortiz Imperial College London, UK E Outa Waseda University, Japan A S Papageorgiou Rensselaer Polytechnic Institute, USA

J Park Seoul National University, Korea G Passerini Universita delle Marche, Italy B C Patten University of Georgia, USA G Pelosi University of Florence, Italy G G Penelis Aristotle University of Thessaloniki, Greece

W Perrie Bedford Institute of Oceanography, Canada

R Pietrabissa Politecnico di Milano, Italy H Pina Instituto Superior Tecnico, Portugal M F Platzer Naval Postgraduate School, USA D Poljak University of Split, Croatia V Popov Wessex Institute of Technology, UK H Power University of Nottingham, UK D Prandle Proudman Oceanographic Laboratory, UK

M Predeleanu University Paris VI, France M R I Purvis University of Portsmouth, UK I S Putra Institute of Technology Bandung, Indonesia

Y A Pykh Russian Academy of Sciences, Russia

F Rachidi EMC Group, Switzerland M Rahman Dalhousie University, Canada K R Rajagopal Texas A & M University, USA T Rang Tallinn Technical University, Estonia J Rao Case Western Reserve University, USA A M Reinhorn State University of New York at Buffalo, USA

A D Rey McGill University, Canada D N Riahi University of Illinois at UrbanaChampaign, USA

B Ribas Spanish National Centre for

Environmental Health, Spain K Richter Graz University of Technology, Austria S Rinaldi Politecnico di Milano, Italy F Robuste Universitat Politecnica de Catalunya, Spain J Roddick Flinders University, Australia A C Rodrigues Universidade Nova de Lisboa, Portugal F Rodrigues Poly Institute of Porto, Portugal C W Roeder University of Washington, USA J M Roesset Texas A & M University, USA W Roetzel Universitaet der Bundeswehr Hamburg, Germany V Roje University of Split, Croatia R Rosset Laboratoire d’Aerologie, France J L Rubio Centro de Investigaciones sobre Desertificacion, Spain T J Rudolphi Iowa State University, USA S Russenchuck Magnet Group, Switzerland H Ryssel Fraunhofer Institut Integrierte Schaltungen, Germany S G Saad American University in Cairo, Egypt M Saiidi University of Nevada-Reno, USA R San Jose Technical University of Madrid, Spain F J Sanchez-Sesma Instituto Mexicano del Petroleo, Mexico B Sarler Nova Gorica Polytechnic, Slovenia S A Savidis Technische Universitat Berlin, Germany A Savini Universita de Pavia, Italy G Schmid Ruhr-Universitat Bochum, Germany R Schmidt RWTH Aachen, Germany B Scholtes Universitaet of Kassel, Germany W Schreiber University of Alabama, USA A P S Selvadurai McGill University, Canada J J Sendra University of Seville, Spain J J Sharp Memorial University of Newfoundland, Canada Q Shen Massachusetts Institute of Technology, USA X Shixiong Fudan University, China G C Sih Lehigh University, USA L C Simoes University of Coimbra, Portugal A C Singhal Arizona State University, USA P Skerget University of Maribor, Slovenia

J Sladek Slovak Academy of Sciences, Slovakia

V Sladek Slovak Academy of Sciences, Slovakia

A C M Sousa University of New Brunswick, Canada

H Sozer Illinois Institute of Technology, USA D B Spalding CHAM, UK P D Spanos Rice University, USA T Speck Albert-Ludwigs-Universitaet Freiburg, Germany

C C Spyrakos National Technical University of Athens, Greece

I V Stangeeva St Petersburg University, Russia

J Stasiek Technical University of Gdansk, Poland

G E Swaters University of Alberta, Canada S Syngellakis University of Southampton, UK J Szmyd University of Mining and Metallurgy, Poland

S T Tadano Hokkaido University, Japan H Takemiya Okayama University, Japan I Takewaki Kyoto University, Japan C-L Tan Carleton University, Canada E Taniguchi Kyoto University, Japan S Tanimura Aichi University of Technology, Japan

J L Tassoulas University of Texas at Austin, USA

M A P Taylor University of South Australia, Australia

A Terranova Politecnico di Milano, Italy A G Tijhuis Technische Universiteit Eindhoven, Netherlands

T Tirabassi Institute FISBAT-CNR, Italy S Tkachenko Otto-von-Guericke-University, Germany

N Tosaka Nihon University, Japan T Tran-Cong University of Southern Queensland, Australia

R Tremblay Ecole Polytechnique, Canada I Tsukrov University of New Hampshire, USA R Turra CINECA Interuniversity Computing Centre, Italy

S G Tushinski Moscow State University, Russia

J-L Uso Universitat Jaume I, Spain E Van den Bulck Katholieke Universiteit Leuven, Belgium

D Van den Poel Ghent University, Belgium R van der Heijden Radboud University, Netherlands

R van Duin Delft University of Technology, Netherlands

A Yeh University of Hong Kong, China J Yoon Old Dominion University, USA K Yoshizato Hiroshima University, Japan T X Yu Hong Kong University of Science & Technology, Hong Kong

P Vas University of Aberdeen, UK R Verhoeven Ghent University, Belgium A Viguri Universitat Jaume I, Spain Y Villacampa Esteve Universidad de

M Zador Technical University of Budapest,

F F V Vincent University of Bath, UK S Walker Imperial College, UK G Walters University of Exeter, UK B Weiss University of Vienna, Austria H Westphal University of Magdeburg,

R Zarnic University of Ljubljana, Slovenia G Zharkova Institute of Theoretical and

Alicante, Spain

Germany

J R Whiteman Brunel University, UK Z-Y Yan Peking University, China S Yanniotis Agricultural University of Athens, Greece

Hungary

K Zakrzewski Politechnika Lodzka, Poland M Zamir University of Western Ontario, Canada

Applied Mechanics, Russia

N Zhong Maebashi Institute of Technology, Japan

H G Zimmermann Siemens AG, Germany

Air Pollution XIX

Editors C. A. Brebbia Wessex Institute of Technology, UK J.W.S. Longhurst University of the West of England, UK

V. Popov Wessex Institute of Technology, UK

Editors C. A. Brebbia Wessex Institute of Technology, UK J.W.S. Longhurst University of the West of England, UK V. Popov Wessex Institute of Technology, UK Published by WIT Press Ashurst Lodge, Ashurst, Southampton, SO40 7AA, UK Tel: 44 (0) 238 029 3223; Fax: 44 (0) 238 029 2853 E-Mail: [email protected] http://www.witpress.com For USA, Canada and Mexico WIT Press 25 Bridge Street, Billerica, MA 01821, USA Tel: 978 667 5841; Fax: 978 667 7582 E-Mail: [email protected] http://www.witpress.com British Library Cataloguing-in-Publication Data A Catalogue record for this book is available from the British Library ISBN: 978-1-84564-528-1 eISBN: 978-1-84564-529-8 ISSN: (print) 1746-448X ISSN: (on-line) 1743-3541 The texts of the papers in this volume were set individually by the authors or under their supervision.Only minor corrections to the text may have been carried out by the publisher. No responsibility is assumed by the Publisher, the Editors and Authors for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. The Publisher does not necessarily endorse the ideas held, or views expressed by the Editors or Authors of the material contained in its publications. © WIT Press 2011 Printed in Great Britain by Martins the Printer, UK. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the Publisher.

Preface

This volume contains the peer-reviewed papers accepted for the nineteenth International Conference on Modelling, Monitoring and Management of Air Pollution held on Malta in September 2011. This successful international meeting builds upon the prestigious outcomes of the 18 preceding conferences beginning with Monterrey, Mexico in 1993 and most recently in Kos, Greece in 2010. These meetings have attracted outstanding contributions from leading researchers from around the world. The presented papers have been permanently stored in the WIT eLibrary as Transactions of the Wessex Institute (see http://library.witpress.com). These collected conference papers provide an important record of the development of science and policy pertaining to air pollution. Despite the long history of attempts to manage the consequences of air pollution it remains one of the most challenging problems facing the international community. Air pollution is widespread and growing in importance and has clear and known impacts on health and the environment. The human need for transport, manufactured goods and services brings with it often unintended, but none the less real, impacts on the atmospheric environment at scales from the local to the global. Whilst there are good examples of regulatory successes in minimising such impacts the continuing development of the global economy bring new pressures upon the ability of the atmosphere to process pollutants and to safely remove them. Where the natural processing systems of the atmosphere become overloaded and the systems are unable to process inputs to the atmosphere at the rate they are added then pollution results. This brings risks to human health and the environment. The willingness of governments to move quickly to regulate air pollution is often balanced by concerns over the economic impact of such regulation. This frequently results in a lag between the scientific knowledge about the nature, scale and effect of air pollution and the implementation of appropriate, targeted and timely legislation. Science remains the key to identifying the nature and scale of air pollution impacts and is essential in the formulation of policy relevant information for regulatory

decision-making. Continuous improvements in our knowledge of the fundamental science of air pollution and its application are necessary if we are to properly predict, assess and mitigate the air pollution implications of emissions to the atmosphere. Science must also be able to provide the evidence of improvements to air quality that result from implementation of the mitigation measure or the control regulation. The ability to assess and mitigate using the precautionary principle is a challenge that science must grasp and position itself to convince decision makers that uncertainty does not mean inertia. The outcomes of such activities must be peerreviewed but they must also be translatable into a suitable format to assist policy makers in reaching sustainable decisions and to build public acceptance and understanding of the nature and scale of the air pollution problem. This important volume brings together contributions from scientist from around the world to present recent work on various aspects of the air pollution phenomena. Notable in each of the nineteen conferences in this series has been the opportunity to foster scientific exchange between participants. New collaborations amongst scientists and between scientists and policy makers or regulators have arisen through contacts made in this series and each meeting has provided a further opportunity for identifying new areas of air pollution science demanding collaborative investigation. Contributions in this the nineteenth volume in the series address a broad range of urgent scientific and technical developments in our understanding of the cause, consequence and management of air pollution. Specifically, papers presented at Air Pollution 2011 provide new data or present critical reviews in the fields of modelling, monitoring and management of air pollution, on emission sources, on the effects of air pollution and on the economic costs of air pollution. The Editors wish to thank the authors for their contributions and to acknowledge the assistance of the eminent members of the International Scientific Advisory Committee with the organisation of the conference and in particular for their support in reviewing the submitted papers. The Editors Malta, 2011

Contents Towards a new framework for air quality management in Nigeria A. O. Olowoporoku, J. W. S. Longhurst, J. H. Barnes & C. A. Edokpayi................................................................................................. 1 Section 1: Air pollution modelling Impact of urban planning alternatives on air quality: URBAIR model application C. Borrego, P. Cascão, M. Lopes, J. H. Amorim, R. Tavares, V. Rodrigues, J. Martins, A. I. Miranda & N. Chrysoulakis .................................................... 13 Air quality model for Barcelona J. Lao & O. Teixidó ........................................................................................... 25 A comparison study between near roadway measurements and air pollutant dispersion simulations using an improved line source model R. Briant, C. Seigneur, M. Gadrat & C. Bugajny .............................................. 37 Regional on-line air pollution modelling system in highly complex terrain P. Mlakar, M. Z. Božnar & B. Grašič................................................................ 47 Identification of potential sources and transport pathways of atmospheric PM10 using HYSPLIT and hybrid receptor modelling in Lanzhou, China N. Liu, Y. Yu, J. B. Chen, J. J. He & S. P. Zhao................................................. 59 Performance evaluation of the ADMS-Urban model in predicting PM10 concentrations at the roadside in Chennai, India and Newcastle, UK S. Nagendra, M. Khare, P. Vijay & S. Gulia ..................................................... 71 Coastal influences on pollution transport D. Peake, H. Dacre & J. Methven ..................................................................... 81

Non-parametric nature of ground-level ozone and its dependence on nitrogen oxides (NOx): a view point of vehicular emissions S. Munir, H. Chen & K. Ropkins ....................................................................... 93 Prediction of TSP concentration in a metallurgical city of Brazil using neural networks M. M. C. Lima.................................................................................................. 105 Section 2: Monitoring and measuring The use of mineral magnetic measurements as a particulate matter (PM) proxy for road deposited sediments (RDS): Marylebone Road, London C. A. Booth, C. J. Crosby, D. E. Searle, J. M. Khatib, M. A. Fullen, A. T. Worsley, C. M. Winspear & D. A. Luckhurst .......................................... 117 Elemental carbon as an indicator to monitor the effectiveness of traffic related measures on local air quality M. H. Voogt, A. R. A. Eijk, M. P. Keuken & P. Zandveld ................................ 129 AMEC multigas passive sampler: a green product for cost-effectively monitoring air pollution indoors and outdoors H. Tang, L. Burns, L. Yang & F. Apon ............................................................ 137 Influence of natural and anthropogenic sources on PM10 air concentrations in Spain M. S. Callén, J. M. López & A. M. Mastral ..................................................... 149 Infrared imaging Fourier-transform spectrometer used for standoff gas detection M. Kastek, T. Piątkowski & H. Polakowski ..................................................... 161 POPs in ambient air from MONET network: global and regional trends I. Holoubek, J. Klánová, P. Čupr, P. Kukučka, J. Borůvková, J. Kohoutek, R. Prokeš & R. Kareš....................................................................................... 173 The EC QA/QC programmes for inorganic gas pollutants testing M. Barbiere, A. Borowiak, F. Lagler, M. Gerboles, M. Kapus & C. Belis ...... 185 GIS for data management of environmental surveys, carried out in Biancavilla (CT) superfund experience S. Bellagamba, F. Paglietti, V. Di Molfetta, F. Damiani & P. De Simone ...... 199 BTEX concentrations in the atmosphere of the metropolitan area of Campinas (São Paulo, Brazil) A. C. Ueda & E. Tomaz ................................................................................... 211

The development of an ESEM based counting method for fine dust particles and a philosophy behind the background of particle adsorption on leaves M. Ottelé, W. J. N. Ursem, A. L. A. Fraaij & H. D. van Bohemen .................. 219 Synthesis of metal oxide nanostructure and its characterization as gas pollutant monitoring B. Yuliarto, M. Faizal, M. Iqbal, S. Julia & T. Nugraha ................................. 231 Buildings as sources of mercury to the atmosphere G. F. M. Tan, E. Cairnsa, K. Tharumakulasingam, J. Lu & D. Yap................ 239 Occupational exposure to perchloroethylene in Portuguese dry-cleaning stores S. Viegas .......................................................................................................... 247 Section 3: Air quality management Health impact assessment of PM10 and EC in 1985–2008 in the city of Rotterdam, The Netherlands M. P. Keuken, P. Zandveld, S. van den Elshout, N. Janssen & G. Hoek ......... 257 Assessing air pollution risk potential: case study of the Tohoku district, Japan Y. A. Pykh & I. G. Malkina-Pykh..................................................................... 267 Assessing the potential for local action to achieve EU limit values J. H. Barnes, T. J. Chatterton, E. T. Hayes, J. W. S. Longhurst & A. O. Olowoporoku...................................................................................... 277 A procedure for the evaluation of the historical trend of atmospheric pollution in an urban area F. Murena & M. Urciuolo ............................................................................... 287 Section 4: Aerosols and particles Correlation between the mass of PM2,5 and the chemical composition of acid aerosols in the northwest of the metropolitan zone of Mexico City Y. I. Falcón, E. Martinez & L. Cortes.............................................................. 301

Characteristics of aerosol particle size distributions in urban Lanzhou, north-western China Y. Yu, S. P. Zhao, D. S. Xia, J. J. He, N. Liu & J. B. Chen .............................. 307 Particulates in the atmosphere of Makkah and Mina valley during the Ramadan and Hajj seasons of 2004 and 2005 A. R. Seroji....................................................................................................... 319 Section 5: Emissions studies Effect of biodiesel and alkyl ether on diesel engine emissions and performances D. L. Cursaru, C. Tănăsescu & V. Mărdărescu .............................................. 331 Emissions of selected gas pollutants in the application of the additive EnviroxTM F. Bozek, J. Mares, H. Gavendova & J. Huzlik ............................................... 343 Non-thermal plasma abatement of trichloroethylene with DC corona discharges A. M. Vandenbroucke, A. Vanderstricht, M. T. Nguyen Dinh, J.-M. Giraudon, R. Morent, N. De Geyter, J.-F. Lamonier & C. Leys ............ 353 Monitoring of atmospheric dust deposition by using a magnetic method A. Kapička, E. Petrovský & H. Grison ............................................................ 363 Improving car environmental and operational characteristics using a multifunctional fuel additive E. Magaril........................................................................................................ 373 Section 6: Global and regional Application of methanotrophic biofilters to reduce GHG generated by landfill in Quebec City (Canada) N. Turgeon, Y. Le Bihan, G. Buelna, C. Bourgault, S. Verreault, P. Lessard, J. Nikiema & M. Heitz ..................................................................................... 387 A study of the atmospheric dispersion of an elevated release with plume rise in a rural environment: comparison between field SF6 measurements and computations of Gaussian models (Briggs, Doury and ADMS 4.1) C. Leroy, F. Derkx, O. Connan, P. Roupsard, D. Maro, D. Hébert & M. Rozet....................................................................................................... 399

Ozone pollution during stratosphere-troposphere exchange events over equatorial Africa K. Ture & G. Mengistu Tsidu .......................................................................... 411 Section 7: Economics of air pollution control Environmental tools of atmospheric protection in the Czech Republic O. Malíková & M. Černíková .......................................................................... 423 Environmentally related impacts on financial reporting: the case of pollution permits in Czech legislative conditions J. Horák & O. Malíková .................................................................................. 433 Section 8: Health effects Analysis of lung cancer incidence relating to air pollution levels adjusting for cigarette smoking: a case-control study P. R. Band, H. Jiang & J. M. Zielinski ............................................................ 445 Comparison of fungal contamination between hospitals and companies food units C. Viegas, M. Almeida, C. Ramos, R. Sabino, C. Veríssimo & L. Rosado....... 455 Author Index .................................................................................................. 463

This page intentionally left blank

Air Pollution XIX

1

Towards a new framework for air quality management in Nigeria A. O. Olowoporoku1, J. W. S. Longhurst1, J. H. Barnes1 & C. A. Edokpayi2 1

Air Quality Management Resource Centre, University of the West of England, Bristol, UK 2 Department of Marine Sciences, Faculty of Science, University of Lagos, Nigeria

Abstract Since 1988 the Nigerian Government has introduced environmental legislation aimed at reducing the atmospheric impact of various sources of pollution. Emphasis has often been placed on mitigating pollution from the oil and gas industry. However, various studies indicate significant ambient air pollution from other sources due to vehicular traffic growth in urban areas, increased reliance on petrol and diesel fuelled generators for electricity supply in homes and other public facilities, uncontrolled open incineration of waste and major thermal power stations within the city limits. In this paper, we make the case for the establishment of risk-based air quality management approach based on monitoring, modelling and assessment of these other sources. We outline four important elements that should be considered in order to achieve this recommended approach. These elements are conceptualised within the existing institutional, organisational structures and capacity in Nigeria. Keywords: Nigeria, air quality management, air pollution, air quality standards and objectives, environmental legislation, environmental policy, NESREA, traffic-related emissions.

1 Air pollution as an immediate concern Nigeria has a population of 140 million people, a large percentage of which reside in major cities such as Lagos, Kano, Abuja, Port Harcourt and Kaduna (National Bureau of Statistics [1]). Lagos has been identified as one of the fastest WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110011

2 Air Pollution XIX growing megacities in the world, with the potential of becoming the most populous city in Africa by 2015 (Gandy [2]; Ibem [3]). However, the cost of population growth is not limited to the demand for water, food and energy resources, but also includes the effect of the increased use of such resources on public health and quality of life. Urban population growth implies that the residents of such cities will increase their demand for journeys through vehicular transport means (Chatterton et al. [4]). Traffic-related pollutants, derived from the use of vehicular transport modes such as cars, are associated with effects ranging from poor public health, built and natural environmental degradation and global climate change (Paulley [5]). Studies have shown that the level of air pollution in Nigeria’s major cities is at a level that could lead to respiratory and cardiovascular diseases in vulnerable individuals (Ogunsola et al. [6]; Efe [7]). Without policy and legislative change in air quality management, increasing numbers of Nigerians living and working in its cities and sprawling urban settlements are at risk from poor air quality. Therefore the policy response must include a rigorous, robust and well-informed strategy of reducing the environmental, social and health impacts of air pollution.

Figure 1:

Map of Nigeria showing major cities (US central intelligence agency [8]).

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

3

2 Institutional and legislative context The institutional and legislative frameworks for pollution control in Nigeria have been viewed as inconsistent and too limited to address the scale and nature of urban air pollution (Achi [9]). High population growth, mass migration to unplanned urban developments and under-regulated industrial pollution in large cities present clear and present threats to the environment as well as to the public health of millions (Adegoroye [10]). Legal and regulatory frameworks are weak and in most cases uncertain on the statutory responsibilities and duties of the government with regard to environmental management and protection (Ogunba [11]). The establishment of the Federal Environmental Protection Agency (FEPA) Act in 1988 provided, for the first time, an attempt at coordinating a statutory and institutional response to environmental pollution (Chokor [12]). However, subsequent policies pursued by the government through the Agency were reactive control measures. Most of the policies were directed at regulating pollution from the oil and gas industries without adequate consideration for other sources and their impacts in densely populated areas. (Adegoroye [10]; Ogunba [11]). The emergence of a new democratic government in 1999 brought, among other things, new hopes for environmental management and protection in Nigeria. The new government created a Federal Ministry of the Environment (FMoE) with a more focused agenda of tackling issues of industrial and urban pollution, marine and coastal resources degradation and the growing threat of desertification. The ministry facilitated major reforms in the environmental legislative and institutional framework. In 2007 the National Assembly repealed the FEPA Act and replaced it with the National Environmental Standards and Regulation Enforcement Agency (NESREA) Act (The Federal Government Printer [13]). The new agency, NESREA, was given the primary responsibility for all environmental laws, guidelines, policies and standards. Part II of the NESREA Act provided statutory enforcement powers and functions of the Agency (The Federal Government Printer [13]). This include responsibilities for “compliance monitoring, the environmental regulations and standards on noise, air, land, seas, oceans and other water bodies other than in the oil and gas sector” (The Federal Government Printer [13]). The corporate strategic plan document published by NESREA identified “improved air quality” as one of the major environmental priorities within its corporate vision (NESREA [14]). In December 2010 the agency undertook a consultation process on various National Environmental Regulations including sections on the Control of Vehicular Emissions from Petrol and Diesel Engines. The establishment of NESREA can thus be seen as a progression from the previous laissez-faire approach to air quality management of previous governments.

3 Air pollution from traffic-related and domestic sources Pollutants from industrial sources, especially from the oil and gas sector in Nigeria have been studied extensively. Sources of emissions include flared gases WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

4 Air Pollution XIX in the Niger Delta, fumes from metal-smelting and cement works, fugitive gases from other chemical and allied industries, and charred particulates and sulphur dioxide emissions from the steel industries (Osuji and Avwiri [15]). These pollutants are not usually confined to the emission point sources. For example, pollutants from flared gases have been observed with concentrations beyond recommended exposure limits in residential communities within 60 m range of the emission source (Obanijesu et al. [16]). Existing Environmental Impact Assessment (EIA) legislation and other pollution control policies have been disproportionately focussed on regulating the oil and gas industries (Ogunba [11]). Conspicuously ignored were the emerging problems from traffic growth, unplanned urban settlements and dependence on wood and kerosene for domestic energy. Various studies conducted in Lagos, Abuja, Port Harcourt, Kano, Calabar, and other major cities in Nigeria, attribute significant emissions to transport, domestic and other industrial sources within close proximity of residential areas (Faboye [17]; Iyoha [18]; Magbagbeola [19]; Oluyemi and Asubiojo [20]). A large proportion of the population are increasingly exposed to air pollution due to growth in vehicular transport and consequent congestion in urban areas, increased reliance on petrol and diesel fuelled generators for electricity supply, and uncontrolled open incineration of waste and major thermal power stations within the city limits (Oluyemi and Asubiojo [20]). Pollution from exhaust pipes is often recognisable without measurements, by reduced visibility, adverse smell and eye irritation on most busy roads (Baumbach et al. [21]). In major cities there are high concentrations of PM10, NO2, CO and VOCs with annual mean concentrations many times greater than the WHO or the Nigerian Ministry of Environment acceptable thresholds (Efe [7]; Koku and Osuntogun [22]). A WHO study in 2007 indicated a growing trend in vehicular-derived air pollution in Lagos due to traffic volume comprising of 2-stroke engines motorcycles (which have higher emissions of particulate matter and un-burnt hydrocarbons than other types of engines) and old imported vehicles (Taiwo [23]). An earlier study also indicated high concentrations of aromatic hydrocarbons, CO and PM especially in areas within close proximity of bus stops and industries within and around Lagos (Baumbach et al. [21]). The level of CO concentrations in Lagos has been shown to be higher than those found in oil-producing cities in the Niger Delta (Abam and Unachukwa [24]). These findings highlight the significance of other sources, such transport, to air pollution beyond that of oil and gas operations. The UK National Centre for Atmospheric Sciences conducted an aerial emissions estimate studies in Lagos using the Atmospheric Research BAe146 aircraft (Capes et al. [25]). The results showed that emissions are attributed to the evaporation of fuels, mobile combustion and natural gas activities around the city. However, Nigeria is among the few countries with no effective procedures or framework for managing ambient air quality (Koku and Osuntogun [22]). There are no coordinated or continuous assessments to inform an appropriate policy framework to manage the local air pollution that residents of cities such as Lagos routinely experience (Taiwo [23]). WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

5

4 Pathways to a Nigerian air quality management framework While there are various complex political and economic issues that require urgent attention by the Nigerian Government, the need to meet the challenges of urban air pollution is also important. Unlike water, drug or food quality, the impact of urban air pollution is non-discriminatory, and does not recognise the broad social and economic stratum that separate Nigerians. Everyone breathes the same air, including the most vulnerable groups – the children, the elderly and the sick. Managing such a problem requires a cyclic and continuous process. Figure 2 outlines four elements that will be required to initiate and develop a management framework in Nigeria. These elements are conceptualised within the existing institutional, organisational structures and capacity in Nigeria.

Figure 2:

Key elements for developing an air quality management framework in Nigeria.

4.1 Scientific enquiry and monitoring First, there is a need for a government-led scientific inquiry to identify and analyse both the spatial and temporal components of air pollution problems in Nigeria. Such an enquiry will include systematic collation, evaluation and development of an empirical evidence base for ambient air pollution. Deployment of air quality monitoring stations will be necessary across major cities and potential hotspots such as oil and gas production areas. Monitoring air WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

6 Air Pollution XIX quality concentrations across the nation is pivotal to identifying the nature and scale of the air pollution challenge, its sources and impacts. Understanding the science of air pollution provides the ability to assess and mitigate the challenge through robust and evidence-based policies. The body of knowledge on air pollution in Nigeria can be shared and enhanced through research studies and development of professional fora where collaborations and joint-working can be encouraged. An important output of this will be a national emission inventory providing required resources for subsequent air quality assessment, modelling and management options. 4.2 Standards and objectives The scientific recognition of the geography, scale and consequences of the air pollution problem should lead to the determination of relevant standards and objectives against which ambient air quality in Nigeria can be measured. A body similar to the former UK Expert Panel on Air Quality Standards (EPAQS) could be set up and facilitated by NESREA to provide independent advice on concentrations of air pollution at which no or minimal health effects are likely to occur in Nigeria. Although there are still uncertainties with regards to the science of atmospheric pollution, there is sufficient evidence which links poor air quality to a significant public health risk (COMEAP [26]). Based on the best available epidemiological information, the government needs to establish a set of numerical air quality standards and limit values for individual pollutants with the potential to compromise public health. Pollutant concentrations should be riskassessed in relation the costs and benefits of required actions and expressed as air quality objectives setting out the extent to which the government expects the standards to be achieved within a specified timeframe (Longhurst et al. [27]). 4.3 Legislation and regulation Since air quality standards and objectives are designed to protect public health, there is therefore a need for appropriate air pollution regulations to guarantee these standards and objectives. Proposed legislation on air pollution at the National Assembly should include the introduction of regulations, which are shaped by scientific and expert consensus on the definition of the problem. The legislation should introduce a policy framework requiring routine monitoring, assessment and management of ambient air quality to ensure the achievement and maintenance of these standards and objectives. Statutory powers and duties should be conferred on specific governmental institutions such as NESREA or the Federal Ministry of the Environment with regards to air quality. Such powers should include the prohibition and restriction of certain activities or vehicles, the obtaining of information, the levying of fines and penalties, the hearing of appeals and other criteria (HM Government [28]). 4.4 Management and evaluation An important element of the framework is the implementation of legislative requirements to achieve stated air quality objectives. Since the 1980s, the WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

7

Nigerian Government has introduced reactive legislation and developed institutions aimed at reducing the environmental impact of industrial activities (Chokor [12]). Apart from not being robust enough, subsequent policies emanating from such laws were often impaired by limited technical capacity to implement efficient enforcement and compliance regimes (Adegoroye [10]). For example, there is no specific policy framework for managing or mitigating emissions from light-duty and heavy-duty vehicles and trucks, which are thought to be amongst the most significant contributors to air quality in Nigeria (Taiwo [23]). It is therefore evident that efficient air quality management in Nigeria will rely on suites of proportionate and cost-effective evaluation and management programmes to be undertaken at the local and national level as much as setting standards and regulations. The management framework will need to take account of economic efficiency, practicability, technical feasibility and timescales for achieving legislated air quality objectives. The state government, along with national agencies such as NESREA, will play an important role in setting out and implementing such management procedures. These may include regular reviews and assessments of air quality to identify whether the objectives have been, or will be, achieved at specific geographic locations where public health is, or will be, at risk, by the applicable date. Where applicable, the government should take proactive responsibility for enforcing and implementing appropriate air quality measures that will lead to the achievement of the objectives. This will include source emissions control from both stationary (industries and domestic) and mobile sources (such as transport).

5 Conclusions Establishing an air quality management framework in Nigeria requires the introduction of specific environmental policy reform and legislative changes based on scientific understanding and analysis of the public health risks of air pollution. This paper identifies four important elements that should be considered in order to achieve this. The first element is the development of an empirical evidence base for ambient air pollution through monitoring and analysis of the nature and effect of air pollution problems in Nigeria. The second is the establishment of numerical air quality standards and limit values for individual pollutants with the potential to compromise public health. Third, there is a need for robust legislation and regulations which will guarantee these standards as well as conferring powers and duties on specific governmental institutions such as NESREA and state government agencies with regards to air quality. Last and more importantly, is the introduction of suites of proportionate and cost-effective evaluation and management programmes to be undertaken at the local and national level for achieving the air quality objectives. Significant gains, in terms of quality of life and public health can be achieved if a Nigerian air quality framework is put in place. There are also economic benefits in developing the capacity of Nigerian environmental professional and academic communities to undertake air quality assessment and modelling WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

8 Air Pollution XIX services. The EIA of development projects such as road schemes, commercial and residential developments, industrial developments, airports and mineral extraction, especially in the oil and gas operations will be enhanced by such skills and expertise leading to better environmental outcomes and improved air quality.

References [1] National Bureau of Statistics. Annual Abstracts of Statistics 2009, Federal Republic of Nigeria, 2009. Online http://www nigerianstat.gov.ng/ [Accessed on 10/06/11] [2] Gandy M. Planning, anti-planning and the infrastructure crisis facing metropolitan Lagos. Urban Studies. 43 (2), 371-396, 2006 [3] Ibem E.O. Challenges of disaster vulnerability reduction in Lagos Megacity Area, Nigeria. Disaster Prevention and Management 20(1), 27-40, 2011 [4] Chatterton, T., Coulter, A., Musselwhite, C., Lyons, G. and Clegg, S. Understanding how transport choices are affected by environment and health: views expressed in a study on the use of carbon calculators. Public health. 123(1), 45-49, 2009 [5] Paulley, N. Recent studies on key issues in road pricing. Transport Policy. 9(3), 175-177, 2002 [6] Ogunsola, O J., Oluwole, A F., Asubiojo, O I., Durosinmi, M A., Fatusi, A O., and Ruck, W. Environmental impact of vehicular traffic in Nigeria: health aspects. Science of the Total Environment. 146, 111-116, 1994 [7] Efe, S.I. (2008) Spatial distribution of particulate air pollution in Nigerian cities: implications for human health. Journal of Environmental Health Research, 7(2) online. http://www.cieh.org/jehr/jehr3.aspx?id=14688 [Accessed on 10/06/11] [8] US Central Intelligence Agency. The World Factbook. Online. https://www.cia.gov/library/publications/the-world-factbook/geos/ni html [Accessed on 10/06/11] [9] Achi P. B. U., An update on the Nigerian environment. 3rd International Conference on Quality, Reliability, and Maintenance (QRM 2000) Ed. McNulty GJ Oxford Univ England Consortium Int Activ; Inst Mech Engineers. 2000 [10] Adegoroye, A. The challenges of environmental enforcement in Africa: The Nigerian Experience. The Third International Conference on Environmental Enforcement held in Oaxaca, México, April 25-28, 1994 [11] Ogunba, O.A. EIA systems in Nigeria: evolution, current practice and shortcomings. Environmental Impact Assessment Review. 24, 643–660, 2004 [12] Chokor, B. A. Government policy and environmental-protection in the developing world: the example of Nigeria. Environmental Management. 17 (1) 15-30, 1993

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

9

[13] The Federal Government Printer. National Environmental Standards and Regulations Enforcement Agency (Establishment) Act, 2007. Federal Republic of Nigeria Official Gazette 94(92). 31 July 2007 [14] NESREA. Corporate Strategic Plan 2009-2012: Building Capacity, Enforcing Compliance. A publication of National Environmental Standards and Regulations Enforcement Agency. 2009. Online. http://www.nesrea.org/forms/NESREA%20CSP.pdf [Accessed on 10/06/11] [15] Osuji, L.C., and Avwiri G.O. Flared gases and other pollutants associated with air quality in industrial areas of Nigeria: an overview. 2(10), 1277-89, 2005 [16] Obanijesu, E. O., Adebiyi, F. M., Sonibare, J. A., Okelana, O. A. Air-borne SO2 Pollution Monitoring in the Upstream Petroleum Operation Areas of Niger-Delta. Nigeria. Energy Sources Part A-Recovery Utilization and Environmental Effects. 31 (3), 223-231, 2009 [17] Faboye, O.O., Industrial pollution and waste management. Dimensions of Environmental problems in Nigeria, ed. A. Osuntokun, Davidson Press: Ibadan, pp. 26-35, 1997 [18] Iyoha, M.A., The Environmental effects of oil industry activities on the Nigerian Economy: A theoretical Analysis: Paper presented at National Conference on the management of Nigeria’s petroleum Resources, Department of Economics, Delta State University Nigeria, 2009 [19] Magbagbeola, N. O., The use of Economic Instruments for Industrial pollution Abatement in Nigeria: Application to the Lagos Lagoon. Selected paper, Annual Conferences of the Nigerian Economic Society PortHarcourt, Nigeria, 2001 [20] Oluyemi E.A. and Asubiojo O.I., Ambient air particulate matter in Lagos, Nigeria: A study using receptor modeling with X-ray fluorescence analysis. Bulletin of the Chemical Society of Ethiopia. 15(2), 97-108, 2001 [21] Baumbach, G., Vogt, U., Hein, K.R.G., Oluwole, A.F., Ogunsola, O.J., Olaniyi, H.B., and. Akeredolu, F.A., Air pollution in a large tropical city with a high traffic density - results of measurements in Lagos, Nigeria. The Science of the Total Environment, 169, 25-31, 1995 [22] Koku, C.A., Osuntogun, B.A., Environmental impacts of road transportation in South-Western States of Nigeria. Journal of Applied Sciences. 7 (16), 2536-2360, 2007 [23] Taiwo, O., Carbon Dioxide emission management in Nigerian megacities: the case of Lagos. Presentation at United Nation Environmental Protection. 2009. Online. http://www.unep.org/urban_environment/PDFs/BAQ09_ olukayode.pdf [Accessed on 10/06/11] [24] Abam F.I. and Unachukwu, G.O., Vehicular Emissions and Air Quality Standards in Nigeria. European Journal of Scientific Research. 34 (4), 550560, 2009

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

10 Air Pollution XIX [25] Capes, G., Murphy, J. G., Reeves, C. E., McQuaid, J. B., Hamilton, J. F., Hopkins, J. R., Crosier, J. Williams, P. I., and Coe, H., Secondary Organic Aerosol from biogenic VOCs over West Africa during AMMA. Atmospheric Chemistry and Physics, 9, 3841-3850, 2009 [26] COMEAP. Long-Term Exposure to Air Pollution: Effect on Mortality. Report produced by the Health Protection Agency for the Committee on the Medical Effects of Air Pollutants. 2009. Online. http://comeap.org.uk /images/stories/Documents/Reports/mortality%20report%202009.pdf [Accessed on 10/06/11] [27] Longhurst, J.W.S., Beattie, C.I., Chatterton, T.J., Hayes, E.T., Leksmono, N.S. & Woodfield, N.K., Local Air Quality Management as a risk management process: assessing, managing and remediating the risk of exceeding an air quality objective in Great Britain. Environment International 32, 934-947, 2006 [28] HM Government, Environment Act 1995. The Stationary Office: London, 1995

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Section 1 Air pollution modelling

This page intentionally left blank

Air Pollution XIX

13

Impact of urban planning alternatives on air quality: URBAIR model application C. Borrego1, P. Cascão1, M. Lopes1, J. H. Amorim1, R. Tavares1, V. Rodrigues1, J. Martins1, A. I. Miranda1 & N. Chrysoulakis2 1

CESAM & Department of Environment and Planning, University of Aveiro, Portugal 2 Foundation for Research and Technology, Hellas, Greece

Abstract In the last decades, the study of the urban structure impacts on the quality of life and on the environment became a key issue for urban sustainability. Nowadays the relevance of urban planning for the improvement of the interactions between different land uses and economic activities, and also towards a more sustainable urban metabolism, is consensually accepted. A major interest relies on understanding the role of planning on induced mobility patterns and thereafter on air quality, particularly related with the increasing use of private cars. This is one of the main objectives of BRIDGE, a research project funding by the European Commission under the 7th Framework Programme and focused on “SustainaBle uRban plannIng Decision support accountinG for urban mEtabolism”. In this scope, and to evaluate the impact on air quality due to different city planning alternatives (PA), the urban scale air quality modelling system URBAIR was applied to selected areas in Helsinki (Finland), Athens (Greece) and Gliwice (Poland), to estimate traffic related emissions and induced pollutant concentration of different air pollutants, in a hourly basis for the entire year of 2008. For the Helsinki study case the results suggest that urban traffic and building placement considered on the different PA have an influence on local air quality despite no significant concentration levels. In the Athens case study some PA induce a decrease on traffic flows with an improvement of the air quality over the domain. On the contrary, other leads to an increase of PM10 in selected hotspots. The simulations for the Gliwice study case show minor changes between the baseline and the PA, since the proposed interventions do not imply major changes in traffic flows. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110021

14 Air Pollution XIX URBAIR applications allowed a comparative analysis between current situation and predefined PA in terms of the number of exceedances to air quality thresholds and other parameters established in European legislation. The results provide important information to urban planners and policy makers to choose the best PA according to quality of life standards pursuit by the local authorities. Keywords: sustainability, urban planning, air quality modelling, traffic emissions, integrated air quality system, decision support system.

1 Introduction In the last decades the study of the urban structure impacts on the quality of life and on the environment became a key issue for urban sustainability. Several studies recognize the importance of urban planning for the improvement of the interactions between different land uses and economic activities, and also towards a more sustainable urban metabolism [1]. Urban structure (sprawl or compact) is intimately related with urban fluxes (incoming and outgoing) of material, energy, information, people, etc. [2]. A major interest relies on understanding the role of planning on induced mobility patterns and thereafter on air quality, particularly related with the increasing use of private cars [3]. According to the European Environmental Agency [4] most EU Member States still do not comply with the PM10 limit values (for which the attainment year was 2005 according to Directive 1999/30/EC). Especially in urban areas, the exceedance of the daily mean PM10 limit value is not only a compliance problem but also has important potential adverse effects on human health. The most critical issue for NO2 compliance in European countries is the exceedance of the annual NO2 limit value in urban areas, particularly at measurement stations close to streets [5]. In this context, the current challenge to urban planners and environmental engineers is to reverse the impacts on environment and human health resulting from the problematic cohabitation between intense road traffic and high population densities, as a way to promote a better quality of life to urban populations. Air quality models proves to be an important tool to assess the impact of urban planning alternatives on traffic patterns, on urban air quality allowing the identification and study of hot spots and helping on the definitions of new urban configurations to improve the quality of life for citizens [6–8]. At the same time, the rapid and continuous growth of hardware capabilities opens a vast number of new possibilities to air quality models, especially through the development of online tools, to be implemented in new Decision Support Systems (DSS).

2 Methodology This work presents the development of the Urban Air Quality system (URBAIR) and its implementation, as an on-line tool, into a multi-purpose DSS for sustainable urban planning. In the core of URBAIR system is a second generation Gaussian model, which has been enhanced with a number of WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

15

functionalities, namely the estimation of road traffic emissions. The model provides air quality patterns for a given spatial domain and time period (usually one year, in compliance with the European Legislation (Directive 2008/50/CE) for different air pollutants, namely: particulate matter with aerodynamic diameter smaller than 10 µm (PM10), nitrogen dioxide (NO2), sulphur dioxide (SO2) and carbon monoxide (CO). Because of the capability to simulate the effect of buildings geometry on air pollutants dispersion, URBAIR offers the possibility to assess the impact of urban planning strategies and traffic management scenarios on air quality. 2.1 URBAIR system description URBAIR system integrates a set of pre-processors of urban geometry, meteorological information and air pollutants emission data in a single tool specifically developed to run online in a Decision Support System (DSS) build under a GIS platform. The URBAIR structure is organized into 4 modules as schematically shown in figure 1.

Figure 1:

URBAIR system architecture.

The emission module allows the estimation of road traffic emissions using the code of the Transport Emission Model for Line Sources (TREM) [6], which has been integrated into URBAIR. Because topography and build-up structures characteristics have a significant influence on the dispersion of atmospheric pollutants, in particular in urban areas, transport and dispersion of the emitted air pollutants (gaseous and particles) is modelled applying an improved version of the second generation Gaussian model POLARIS [9], which allows to account for the presence of buildings in the dispersion simulation. In this sense, URBAIR WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

16 Air Pollution XIX requires also the characterization of the spatial variation of terrain surface elevation, buildings 3D coordinates and emission sources location and dimensions, which are usually provided by Geographical Information System maps. The geographic module relies on a Cartesian coordinate system, in which regular and discrete gridded data can be used to characterize and spatially distribute terrain, receptors and sources. Representative terrain-influence heights and ‘projected’ building structures influence are determined following widely used modelling approaches. Topography is specified in the form of terrain heights at receptor locations. The influence of buildings on air pollutants dispersion depends on the orientation of the obstacle relating to the source, the wind direction and the shape of the building. The meteorological pre-processor calculates the parameters needed by the dispersion model, namely the atmospheric turbulence characteristics, mixing heights, friction velocity, Monin-Obukhov length and surface heat flux. The meteorological data needed for this pre-processing stage can be provided by mesoscale meteorological models, or alternatively surface measurements and upper air soundings databases can be used. Meteorological information, geographic and geometric data, and road traffic fluxes constitute the major categories of input data needed by the integrated air quality system URBAIR. The output data includes the estimated emissions from road traffic and pollutant concentration at user-specified receptor points or spatially distributed over a regular grid. The first version of URBAIR was designed for line sources since there are the most important ones in urban environments. New model developments include elevated point sources (such as industrial facilities and combustion activities for residential and services sectors). Different mean averaged concentration values can be defined, depending on the evaluation purposes. 2.2 Study cases description URBAIR system was applied to three European urban areas, selected BRIDGE project case studies, with distinct characteristics namely on dimension and planning attributes: Helsinki, Athens, and Gliwice. With the objective of evaluating the impact on air quality due to different city structure design options, different PA were simulated. The study areas were defined based on detailed information relating the baseline situation and the proposed planning alternatives using ArcGIS maps. Traffic is considered as the main pollutant source in the study areas. Emissions are calculated by the pre-processor TREM using traffic counts provided by each city and average speeds. In URBAIR roads are spatially discretized by defining an adequate number of point sources along each road. Previous sensibility analysis has demonstrated that a spacing of 10 to 15 meters between adjacent point sources guarantees the needed accuracy in the representation of the roads existing in the domain. Meteorological input data, including vertical profiles, were obtained from the WRF mesoscale model simulations over the different case studies domains.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

17

The intervention area in Helsinki is located in the forest area between current housing of Meri-Rastila and Vartiokylä Bay. The planning objectives for this area are: to provide new housing for the growing metropolitan areas; to provide places of work mixed with housing; to deal with demographic polarization; to move towards more owned dwellings and bigger apartments; to improve services; to maintain sufficient and continuous recreation and habitats; and to improve accessibility to nature areas [10]. Three planning alternatives have been proposed with varying combinations of housing density and office space, and differing relative footprints. These alternatives consider three different building configurations with different number of new roads and, consequently, of traffic fluxes. The URBAIR computational domain, with approximately 4000×4000 m2, and a spatial resolution of 100×100 m2, was defined at the centre of the study area. For the current situation (baseline) the urban built-up area was simplified by considering 234 grouped buildings with different configurations both in geometry and heights. PA1 considers a total of 251 grouped buildings, while in PA2 and PA3, 254 and 263 building blocks, respectively, were defined. All the alternatives imply an increase on the number of roads (see Figure 2). The Athens case study is focused on the municipality of Egaleo, which lies in the Western part of Athens. Five main road axes divide the area in four quarters. One of the quarters is an industrial degraded area called Brownfield (Figure 3). The total area of Egaleo is 650 ha. The intervention area is centred at the Brownfield industrial area. The computational domain has an area of approximately 4000×4000 m2, with a spatial resolution of 100×100 m2. Built-up geometry was simplified by grouping the existing buildings in 151 blocks. No simulations were carried out for PA1, because no changes in urban planning or traffic are foreseen. PA2 implies an increase in the number of buildings. Traffic fluxes were assumed as identical to nearby roads in the Egaleo area. PA3 considers the conversion of the intervention area into a green zone. Consequently, a reduction of 90% in traffic in relation to nearby roads was assumed. Gliwice is a city with an old Town in the central part and residential districts around the centre, with a total area of 134 km2 [6]. The alternatives include: PA1) the construction of a sports hall, which will entail an additional load of people in the area; PA2) the construction of a centre for new technologies, a 7storey building incorporating sustainable energy use (e.g. heat energy from solar collectors, energy recovery, etc.); and PA3) the development of both projects considered in PA1 and PA2. The case study will be mainly assessed with regard to the environmental load in the area (particularly from the point of view of emissions and resource use) and the transport and economic implications to the city. The URBAIR computational domain, with 5400×5400 m2 and a spatial resolution of 100×100 m2, was centred at the intervention zone. 92 rearranged building blocks were defined in URBAIR for the baseline situation. PA1 and PA2 considers the construction of only one additional building (the sports hall and the centre for new technologies, respectively), while for PA3 both were WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

18 Air Pollution XIX defined in URBAIR (Figure 5). The most significant change is the increase of traffic flows due to foreseen attraction of public.

3 Air quality results for baseline and planning alternatives In Figure 2, PM10 simulation results for Helsinki on 25th July 2008 are presented for baseline situation and PA1, PA2 and PA3.

Figure 2:

a)

b)

c)

d)

Comparison of 1.5 m high horizontal 24 hour average [PM10] fields in Helsinki domain, on 25th July 2008 for: a) baseline, b) PA1, c) PA2 and d) PA3. Red rectangle indicates the intervention area. (See online for colour version.)

Comparing the results observed in Figure 2 it is possible to conclude that despite the changes on the number of roads and respective traffic fluxes, and also on the number and location of buildings, the different alternatives do not induce significant modifications on the dispersion patterns. However, and according to the simulations, PA2 and PA3 have a higher influence over the [PM10] in the intervention area and, particularly in PA3, in an area located to the north of the new buildings and roads. In general, [PM10] over the domain stay within the limit value established on legislation for 24 hours average (50 µg.m-3), although some hot-spots are visible where concentrations reach values of 90 µg.m-3 for this particular summer day. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

a)

19

b)

c)

Figure 3:

Comparison of 1.5 m high horizontal 24 hour average [PM10] fields in Athens domain, on 22nd September 2008 for: a) baseline, b) PA2, c) PA3.

Figure 3 presents the simulation results for a specific summer day in Athens, for [PM10] levels, one of the most critical pollutants in this area. Analysing the results presented in figure 3, it is clear that PA3 is the one that presents better results in the intervention area regarding [PM10]. Values as high as 130 µg m-3 were obtained for all the situations, with a strong reduction in the intervention area for PA3. Athens is the only city case in which an air quality monitoring station is located within the study area. Figure 4 presents a time series of simulated and measured [PM10] during the year of 2008. Observed air quality levels were acquired at the Aristotelous air quality monitoring station. The simulated values are from a specific cell of the domain which corresponds to the location of the referred air quality station. In general, simulated values reasonably follow the trend of measured concentrations. However, some underestimation tendency was observed. Possible reasons are the lack of information relating background concentrations and local emission point sources, as well as the no consideration of particulate matter resuspension (only direct exhaust emissions were considered). It can be also inferred from the analysis of Figure 4 that both measured and simulated PM10 concentrations show several exceedances to the legislated limit value of 50 µg m-3. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

20 Air Pollution XIX

Figure 4:

Figure 5:

Comparison of measured and simulated [PM10] in the Aristotelous air quality station for the year 2008 (XY coordinates: 2800 m; 2000 m).

a)

b)

c)

d)

Comparison of 1.5 m high horizontal 24 hour average [PM10] fields in Gliwice domain, on 2nd January 2008 for: a) baseline, b) PA1, c) PA2, d) PA3.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

21

In figure 5 the PM10 simulation results are presented for Gliwice study case on 2nd January 2008 for baseline situation and PA1, PA2 and PA3. Comparing the results obtained for the baseline situation and planning alternatives, no major differences in [PM10] are visible, showing that the implementation of the new buildings and the increase in traffic fluxes forecast in the nearby roads do not have a significant impact in [PM10]. In order to have a better understanding on the influence of the different alternatives on air quality, table 2 shows the maximum simulated concentrations of PM10, CO, NO2 and SO2 in Helsinki, Athens and Gliwice during 2008. This value corresponds to the maximum concentration calculated by URBAIR for a height of 1.5 meters above ground. From the analysis of the results shown in Table 1 it is possible to conclude that for Helsinki the planning alternatives do not have an influence in the maximum simulated concentrations despite the construction of new roads. Regarding Athens study case, PA2 will lead, according to the simulations, to an increase of the maximum concentrations for all the pollutants considered, while PA3 supports a decrease of the peak concentration when compared with the baseline situation. In Gliwice, baseline scenario and PA1 present the same results, while PA2 and PA3 have lower maximum values. Table 1:

Maximum simulated concentrations of PM10, CO, NO2 and SO2 at 1.5 meters high for Helsinki, Athens and Gliwice in 2008. Study case

Baseline

Helsinki Athens Gliwice

227 248 37

Helsinki Athens Gliwice

1531 5045 451

Helsinki Athens Gliwice

230 382 58

Helsinki Athens Gliwice

84 236 30

Planning alternative 1 2 PM10 [μg.m-3] 227 227 253 37 42 CO [μg m-3] 1531 1532 5526 451 461 NO2 [μg.m-3] 230 230 388 58 68 SO2 [μg m-3] 84 84 240 30 36

3 227 222 42 1532 4995 461 230 370 68 84 228 36

Another analysis was made in terms of the number of exceedances to the limit value of simulated pollutants during the entire year of 2008 for a specific cell of the domain for each study case. This analysis is presented is Table 2. The selected computational cell in Athens and Gliwice corresponds to the location of

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

22 Air Pollution XIX the air quality station, although in the latter measurements are not available for 2008. In Helsinki, the selected cell corresponds to the centre of the domain. Table 2:

Number of exceedances to PM10, CO, NO2 and SO2 in Helsinki, Athens and Gliwice during 2008.

Planning alternative Compliance with the Directive? 1 2 3 PM10 Limit value: 50 μg.m-3 [24 hours average] with 35 exceedances allowed Helsinki 0 0 0 0 y Athens 122 122 96 n Gliwice 0 0 0 0 y CO Limit value: 10 mg m-3 [8 hours moving averages] Helsinki 0 0 0 0 y Athens 0 0 0 y Gliwice 0 0 0 0 y NO2 Limit value: 200 μg m-3 [1 hour average] with 18 exceedances allowed Helsinki 0 0 0 0 y Athens 5 5 3 y Gliwice 0 0 0 0 y SO2 Limit value: 350 μg m-3 [1 hour average] with 24 exceedances allowed Helsinki 0 0 0 0 y Athens 0 0 0 y Gliwice 0 0 0 0 y

Study case

Baseline

With the analysis based on the selected cell for each study case, only for the pollutant PM10 in Athens study case were found exceedances in terms of the number permitted by the European legislation. For NO2 some exceedances were forecast but within the accomplishing criteria established in legislation. However, if the selected cell was in a different location, the situation could change and more exceedances might be found. Despite the number of exceedances is beyond the allowed number permitted by the legislation, in PA3 there is a reduction on their number for PM10. With the analysis based on the selected cell for each study case, only in Athens study case and for PM10 were found exceedances to the limit value. Despite the number of exceedances is beyond the allowed number, PA3 can potentially lead to an improvement on the local air quality.

4 Conclusions URBAIR applications allowed a comparative analysis between current situations and predefined planning alternatives in terms of the number of exceedances to air quality thresholds and other parameters established in European legislation. In general, it was concluded from the comparisons of simulated concentrations with WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

23

measured data that URBAIR presents some underestimation tendency. Among the reasons for this behaviour the followings issues can be raised: • Background concentrations and local emission point sources were not considered, due to lack of information; • Except for Athens study case, average hourly traffic fluxes were calculated from annual values; • Only exhaust emissions were considered (i.e., the contribution of particles resuspension was not taken into account); • Road traffic emissions were estimated based on vehicles count and average speed. This methodology does not allow accounting for the emissions during traffic jams, which can be relevant, especially in Athens, during the peak hours; • Also the contribution of natural events, which can be relevant in some air pollution episodes, was not considered. Despite the small scale of the considered planning alternatives in terms of project dimension and the area of intervention, the results provide important information to urban planners and policy makers to choose the best planning solution according to quality of life standards pursuit by the local authorities.

Acknowledgements The authors would like to acknowledge the financial support of the BRIDGE Project by the European Commission under the 7th Framework Programme, and the Portuguese Ministry of Science, Technology and Higher Education, through the Foundation for Science and Technology (FCT), for the Post-Doc grant of J. H. Amorim (SFRH/BPD/48121/2008) and for the financial support of project INSPIRAR (PTDC/AAC-AMB/103895/2008), supported in the scope of the Competitiveness Factors Thematic Operational Programme (COMPETE) of the Community Support Framework III and by the European Community Fund FEDER.

References [1] Borrego, C., Lopes, M., Valente, V., Neuparth, N., Martins, P., Amorim, J.H., Costa, A.M., Silva, J., Martins, H., Tavares, R., Nunes, T., Miranda, A.I., Cascão, P. & Ribeiro, I., The importance of urban planning on air quality and human health (Chapter 2). Urban Planning in the 21st Century, eds. D.S. Graber & K.A. Birmingham, Nova Science Publishers Inc., 2009. [2] Martins, H., Miranda, A. & Borrego, C., Atmospheric modelling under urban land use changes: meteorological and air quality consequences. 31st NATO/SPS International Technical Meeting on Air Pollution Modelling and its Application, 27 Sept–1 Oct., Torino, Italy. 2010. [3] Amorim, J.H., Lopes, M., Borrego, C., Tavares, R. & Miranda, A.I., Air quality modelling as a tool for sustainable urban traffic management. Air Pollution XVIII. 21-23 June, Kos, 3-14. Greece. WIT Press. 2010.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

24 Air Pollution XIX [4] European Environment Agency. Towards a resource-efficient transport system. TERM 2009: indicators tracking transport and environment in the European Union. EEA Report, No 2. 2010. [5] ETC/ACC 2009a: European exchange of monitoring information and state of the air quality in 2007. ETC/ACC Technical Paper 2009/3. [6] Borrego, C., Tchepel, O., Costa. A., Amorim, J. & Miranda, A., Emission and dispersion modelling of Lisbon air quality at local scale. Atmospheric Environment, 37, 5197-5205, 2003. [7] Martins, A., Cerqueira, M., Ferreira, F., Borrego, C. & Amorim, J.H., Lisbon air quality – evaluating traffic hot-spots, International Journal of Environment and Pollution - Vol. 39, Issue 3/4, 306-320, 2009. [8] Borrego, C., Tchepel, O., Salmin, L, Amorim, J.H., Costa, A.M. & Janko, J., Integrated modelling of road traffic emissions: application to Lisbon air quality management, Cybernet. Sys.: An International Journal 35 (5-6), 535-548, 2004. [9] Borrego, C., Martins, J.M, Lemos, S. & Guerreiro, C., Second generation Gaussian dispersion model: the POLARIS model. International Journal of Environment and Pollution - Vol. 8, No.3/4/5/6 pp. 789 – 795, 1997. [10] Bridge Newsletter. Issue 2 may 2010. [11] http://www.bridgefp7.eu/images/pdf/211345_001_DM_NKUA_1_0_2nd_ Newsletter.pdf

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

25

Air quality model for Barcelona J. Lao & O. Teixidó Energy & Air Quality Department, Barcelona Regional, Spain

Abstract Some cities and metropolitan areas have a hard time complying with EU regulations regarding certain pollutant concentration levels. In 2008, Barcelona’s monitoring stations reported NO2 levels above the EU limit of 40 µg/m3. This paper shows the process and results of NO2 dispersion analysis in Barcelona using 2008 as the base year, as well as the results of the 2020 forecast. Barcelona City Council has drawn up an air-quality model as part of the “PECQ” (Energy, Climate Change and Air Quality Plan for Barcelona 2011-2020) to help decision makers implement actions aimed at reducing NO2. In the first stage, a real inventory of vehicles was performed, recording over 90,000 vehicle plates and also measuring 42,000 actual emissions via Remote Sensing Devices. We discovered that the vehicles on the road are newer than the city census vehicles. We also found out that real-world vehicle emissions are 16.2% higher than COPERT. We used GIS tools to compile the geographical inventory of emissions inside and outside the city. The base-year results show that 65.6% of NO2 concentration levels come from vehicles, 8.6% from the residential and commercial, 4.8% from industry, including heat and power production close to the city, 2.1% from Barcelona Port, and 0.1% from Barcelona Airport. The local background contribution was calculated as 10.1% and the regional background accounts for 8.6%. The PECQ Plan will run projects from 2011 until 2020 aimed at reducing NOX emissions in various sectors. Improvements in vehicle technology are also expected. The 2020 forecast scenario shows that NO2 concentration levels will drop by 35% to reach EU standards. Keywords: air quality, air pollution modelling, validation, Barcelona, dispersion modelling, NO2, NOX, PECQ, emission inventory, RSD.

1 Introduction The Barcelona PECQ 2011-2020 (acronym in Catalan of Barcelona’s Energy, Climate Change and Air Quality Plan [1]) is an action plan directed by Barcelona WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110031

26 Air Pollution XIX city council and developed by Barcelona Regional (a public company) together with the City Council. The general objectives are: to reduce the increase in energy consumption, to reduce the increase in greenhouse-gas emissions associated with the municipality, and to improve air quality in the city, especially as regards NO2 and PM10, with a specific reduction target of 26% for NOX and 39% for PM10 emissions, in order to achieve European objectives for air-quality levels. The PECQ methodology includes a historical analysis of energy, GHG emissions and air quality in Barcelona city, plus a battery of projects and proposals for the next 10 years. It also contains the expected future scenario, including an in-depth dispersion modelling analysis of the city and its surroundings. The PECQ includes other interesting aspects such as the analysis of social attitudes towards energy consumption, and the effects of the PECQ Action Plan on the local and regional economy. The PECQ development process also included extensive consultation with citizens and stakeholders, from the design stage through to drafting of the Action Plan. This paper will focus on the NO2 air-quality model for the city of Barcelona. We will show the methodology, validation process, results of the base case (2008) and Barcelona’s expected air quality by 2020 according to the various policies and measures adopted under the PECQ Action Plan.

2 Barcelona air quality Like other cities, Barcelona exceeds the annual average NO2 concentration thresholds established by the EU to protect human health. This means that cities, regions and countries must adopt new strategies, on various levels, aimed at improving air quality in metropolitan areas. This includes vehicle manufacturers, legislators, citizens, companies, and so on. For years, Barcelona City Council and other public bodies have been working to improve air quality through various measures involving industry and power plants and by promoting renewable energy. Examples of this include the Solar Thermal Bylaw [2] of 1999 or the application of the Barcelona Energy Improvement Plan (2001-2010) [3]. Given that the main source of pollutants is road transport, Barcelona metropolitan area has made remarkable efforts to achieve a modal split change. This has included promoting the integration of public transport fares and the improvement of public transport networks (bus, metro, trams). Barcelona has also expanded the city’s bicycle network and created “Bicing” – a public bicyclerental service with a very low-cost flat rate. Another measure was to increase the roll-out of parking meters for surface parking, in order to make private transport systems less competitive. Despite all the policies designed to make public transport more attractive and to stimulate the modal split change from private vehicles to the public system, there is still some way to go, since the city does not yet fall within the air-quality standards set by Europe. The EU limit value for annual average nitrogen dioxide concentration in 2009 was 42 µg/m3 (including a tolerance margin of 2 µg/m3 WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

27

© Barcelona Regional, 2010

applicable in 2009). Four out of a total of six measurement stations inside the city exceeded the annual average limit. As shown in Figure 1, NO2 station measurements for Barcelona city and surrounding municipalities exceed EU NO2 limits, meaning that new initiatives are required.

45 µg/m3 El Prat de Llobregat

42 µg/m3 Cornellà de Llobregat

43 µg/m3 L’Hospitalet de Llobregat

Figure 1:

41 µg/m3

62 µg/m3

Barcelona: Sants

Barcelona: Eixample 3

63 µg/m

Barcelona: Gràcia-St Gervasi

46 µg/m3

50 µg/m3

46 µg/m3

Barcelona: Ciutadella

Sant Adrià de Besòs

Badalona

51 µg/m3 Barcelona Poblenou

44 µg/m3 3

40 µg/m

Barcelona: Parc Vall d’Hebrón

Montcada i Reixac

44 µg/m3

Santa Coloma de Gramenet

Annual average NO2 concentration at measurement stations in Barcelona and surroundings (2009).

3 Barcelona urban air-dispersion model Within the PECQ, in order to focus the Action Plan most effectively, it was essential to determine what activities and sectors are responsible for high NO2 concentration levels. A detailed inventory of emissions by sectors has been developed for the base case year 2008 and distributed throughout the territory. This emission inventory was one of the key inputs of the Barcelona AirDispersion Model, a tool that helps decision-makers know what is happening with air quality and what needs to be done in order to improve it. 3.1 Methodology For atmospheric dispersion modelling we used ADMS-Urban [4], developed by CERC in the UK. ADMS-Urban allowed us to calculate NO2 concentration levels based on a Gaussian dispersion model with photochemical reactions and an integrated street canyon model. The entire model is fully integrated in a WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

28 Air Pollution XIX Geographical Information System (GIS) database of emission sources, terrain configurations and other relevant aspects. The main features of the dispersion model used are:  Specific dispersion model for urban and metropolitan areas with resolution down to street level.  Includes a meteorological pre-processing model.  Uses the “FlowStar” module, a processing module designed for hourly flows and turbulence for high-resolution complex plots.  Can use hourly, daily and monthly input profile schedules for each source emission.  Works with the OSPM model, specifically to assess the “Street Canyon” effect resulting from the recirculation of air turbulence among buildings.  Uses the GRS chemistry scheme, a semi-empirical photochemical model which includes the reactions of NO, NO2, O3 and many organic compounds. After gathering the data and in order to map an air-quality model, a highresolution grid was created with up to 150,000 virtual grid points across the territory, plus 50,000 points next to roads using “intelligent gridding” software capability. The result was a mean grid resolution of 35.2 metres in outlying parts of the city and an estimated mean grid resolution of 17.6 metres in the city centre in order to ensure higher accuracy. More than twelve processors were used, working constantly for 30 days. Figure 7 (left) shows the map of NO2 concentration levels after model calibration for 2008. Keep in mind that reality is more complex than the “typical profiles” or mean behaviours introduced in the model. Therefore, unusual traffic jams, fires, construction work, unknown emissions, and other situations can cause deviations between the models and actual data, meaning that model calibration must always be carried out. Actual hourly data from the measurement stations was also compared with modelling results for virtual point detectors. Table 1 shows a comparison of annual mean values and Figure 6 contains a monthly example of hourly comparison. 3.2 Characteristics and emissions of vehicles in Barcelona Since the road transport sector is the main emitter of pollutants, the PECQ established a clear difference with previous studies of vehicle emissions. A realworld characterisation of traffic was developed in order to find out what types of vehicles are driven around the city, as well as their emissions. It is important to mention that previous studies used the vehicle census of the city [5]. We will show below that there is a major “gap” between the traffic on the streets and what is listed in the city census via road tax. An example of this is that the vehicles that use the streets everyday are newer on average than the census vehicles. This is a key point for drawing up effective policies and calculating related NOX reductions. It is also important to have a better idea of actual emissions in order to obtain a better estimate in the air-dispersion model. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

29

Since it is a diffuse sector, traffic emissions must be determined indirectly, based on vehicle km/year, total number of vehicles, average speed, and methodologies based on emission factors (EF) by vehicle characteristics (fuel type, engine size, weight and technology of the vehicle), such as CORINAIR [6]/COPERT [7]. To improve EF methodology, Barcelona City Council conducted an ad-hoc study [8] over 32 days in May and June 2009, setting up 16 roadside points for measuring vehicle exhaust emissions, with a detector system called “RSD” (Remote Sensing Device [9]). This technology can detect pollutant emissions from vehicle exhaust pipes instantly and in a non-intrusive manner, using infrared and ultraviolet light according to the Lambert-Beer law. This means that vehicles do not have to modify their normal driving patterns and thousands of vehicle license plates and exhaust vehicle emission data can be gathered in just a few hours. The emission data shows the actual emissions of the vehicles, unlike other methodologies based on standard emission factors, and the license plate shows the vehicle type. This makes it possible to know the brand, vehicle model, technical characteristics (power, weight, fuel, age, etc.), and city of residence (census). We gathered 90,000 vehicle license plates and emissions data for more than 42,000 vehicles after RSD exhaust data validation. This study revealed some important aspects:  The average age of all vehicles is 5.7 years.  The cars driving around the city are newer (with an avg. age of 5.53 years) than the city vehicle census (9.13 years). This does not mean that the census does not work properly. The difference only shows that older vehicles get driven less than newer vehicles or, in other words, that people who use their car every day tend to have newer vehicles.  Petrol cars are older (7.58 years) than diesel cars (4.43 years) due to a social trend. In Spain, it is typical to buy a diesel car if your annual mileage is very high, so daily car users buy more diesel cars than weekend car users.  The pre-EURO class displays an interesting behaviour pattern. 20% of the cars in the city vehicle census are pre-EURO; by contrast, the EURO class only represents 8% of cars driven daily in the city streets.  The taxi fleet has an average age of 3.4 years, and the average age of trucks is 6.5 years.  The most common fuel used by vehicles in the city is diesel at 55.1%, followed by petrol at 44.1%, biodiesel at 0.6% and, finally, natural gas at 0.3%.  52% of the vehicles come from outside Barcelona city (they are not included in the municipal census). As for cars, 51% come from outside the city. In this group, 41% are petrol cars and 56% are diesel cars, since it is more cost-efficient to own a diesel car if your annual mileage is very high, as diesel is cheaper and more efficient than petrol. This makes sense considering that 62% [10] of daily car trips in Barcelona are made by people that live outside the city and commute in to work (or vice-versa). This is a very important point, since municipal policies focusing on diesel WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

30 Air Pollution XIX cars (emitting more pollutants than petrol cars) included in the municipal census will not be as effective as regional policies. According to EURO class, 39.2% of vehicles are EURO IV, 34.8% are EURO III, 14.4% are EURO II, 5.5% are EURO I and 5.6% are pre-EURO. It should be noted that the number of EURO V vehicles is symbolic – 0.4% – since this classification only applied to buses and trucks in 2009 (not to cars, vans or motorcycles) and the measurements were taken in mid-2009. The segments with the highest percentage of EURO II vehicle and older are petrol vans (LDV) with 44.2%, diesel buses and coaches with 34.7%, and petrol cars with 32.1%. Average age (in years) of vehicles in Barcelona MEAN

5,66

Special vehicles

3,58

BUS (Natural Gas)

4,86

BUS+COACH (Diesel)

6,85

HDV (Diesel)

6,55

MDV (Diesel)

EURO IV 39,2%

7,11

LDV (mean)

EURO V 0,4%

Pre-EURO 5,6%

EURO I 5,5% EURO II 14,4%

5,85

LDV (Diesel)

5,64

LDV (Petrol)

9,14

MOTORB KES (Petrol)

5,54

CARS (mean)

EURO III 34,8%

5,53

CARS (Híbrid)

1,38

CARS (Diesel)

4,43

CARS (Petrol)

7,58 0

Figure 2:

EURO Class distribution of vehicle traffic in Barcelona (2008): 4.439,16 Mveh-km/year

7,41

MDV + HDV (mean)

2

4

6

8

years

10

Average age of different categories of vehicles (left) and EURO class distribution of vehicles driving around city streets. © Barcelona Regional, 2010.

As mentioned above, we gathered actual exhaust pipe emissions from 42,000 vehicles driving around the streets of Barcelona, and we compared actual emissions from the “RSD” system with COPERT methodology. The average result was that RSD measured 16.2% higher NOX emissions than COPERT considering the same number of vehicles in city driving mode, with an average speed of 21.3 km/h and the weather conditions for May/June. RSD actual vehicle emission data with the annual share of vehicular traffic shows that 34.2% of vehicle NOX emissions are from cars (29.3% from diesel cars and 4.9% from petrol cars), followed by vans (LDV) 17.4%, medium and heavy trucks (MDV and HDV) 15.7%, motorcycles and mopeds 12.3% (in Barcelona, 22.8% of all private transport is by motorbike), private buses and coaches 12.2%, and local public buses 8.2%. The average emission factor for total traffic was 1.1297 grams of NOX per kilometre. Figure 3 shows total emissions and emission ratio by vehicle class, and Figure 4 contains the distribution according to mobility, vehicle emission and trips. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

1.600

tn NOx/any

NOx emissions of vehicles in Barcelona by vehicle type (2008). TOTAL: 5.014,72 tn/year [COPERT+RSD] [total mean (♦): 1,1297 g/km]

1.400

g/km

11,0090

31 14 12

9,9671

1.200

8,6364

1.000

10 8

800 6

4,4817

600

4

400 200

0,2711

0,8982

0,6074

0,8311

MOTORB KE (Petrol)

LDV (Petrol)

0,0064

1,3240

2

0

0 CAR (Petrol)

Figure 3:

CAR (Diesel)

CAR (Hybrid)

LDV (Diesel)

HDV (Diesel)

NTERURBAN URBAN BUS BUS (Diesel) (Diesel+NatGas)

NOX emissions and emission ratio by vehicle type for Barcelona. © Barcelona Regional, 2010.

4,9% 6,6%

3,3% 8,2%

5,3%

PM

33,0%

veh-km 36,9%

20,3% 16,1%

5,2%

trips

29,3%

CARS (Diesel)

11,5% 16,7%

25,6%

LDV (Diesel)

0,9%

NTERCITY BUS (Diesel) 10,6% 1,4%

8,3%

% 4,2

22,8%

12,3%

Co lo urs w th ho rizo ntal stripes when the main function of the vehicle is to transpo rt goo ds, not people.

1,0%

12% 11% 11%

10,8%

NOx (sorting criteria)

29,3%

5,0%

MOTORB KES (Petrol) HDV (Diesel)

14,2% 21,1%

16,7%

12,2%

Figure 4:

MDV (Diesel)

URBAN BUS (Diesel+NatGas) MDV (Diesel) CARS (Petrol) LDV (Petrol)

Related share of NOX emissions (sorting criteria), PM10 emissions, mobility (veh-km) and trips. © Barcelona Regional, 2010.

3.3 Barcelona 2008 emissions inventory Database quality is one of the most important aspects of air-quality modelling. All the emission sources input into the air-dispersion model are shown below. The base-case year for the emissions inventory is 2008.  Road transport: we used the COPERT emission model plus the XTRA RSD recorded exhaust emission factors in order to include actual emissions from vehicles. Hourly, daily, weekly and monthly traffic profiles were also used.  Residential and commercial: we used the CORINAIR emission factor to estimate NOX emissions from natural gas and LPG. Hourly and monthly profiles were also implemented for this sector.  Industry and power plants: Actual emissions data for isolated emission sources with continuous environmental control was used. Other industrial emissions were estimated using CORINAIR methodology. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

32 Air Pollution XIX  Port: Land traffic, sea traffic and auxiliary vehicles in Barcelona’s Port were taken into account using the COPERT+XTRA RSD system for the first, and the CORINAIR methodology for the second and third. Monthly profiles were implemented for this sector.  Airport: Barcelona Airport is 12 km from the city centre (outside the city limits). NOX emissions were estimated using CORINAIR methodology and taking into account LTO cycle for airplanes and auxiliary vehicles. Hourly and monthly profiles were used for the airport. All these data provide an inventory of NOX emissions for Barcelona and surroundings covering 1,476 km2, of which Barcelona makes up only 7%, with 102 km2. Air-quality modelling involves studying a larger area, since pollution does not respect municipal boundaries. Thus, total NOX emissions for the entire area covered were 34,186 tonnes in 2008, while NOX emissions for Barcelona city were 10,413 tn/y. As regards Barcelona emissions, vehicle are responsible for 5,015 tn/y (4,299 according to COPERT plus 716 according to RSD), making up 48.2% of the city emissions. The second source of emissions was Barcelona Port, with 3,078 tonnes or 29.5% (2,512.5 for sea traffic plus 565,7 for land activity). The third source was industry and power plants, with 1,394 tn/y or 13.4%. And, finally, the residential and services sectors, with 926 tn/y or 8.9%. Figure 5 shows the georeferenced emissions inventory. NOX (kg/year)

Barcelona City emissions (tn/y) AIRPORT

Model area emissions map

n/a

PORT [ships]

2.512,5 (24,1%) 565,7 (5,4%)

PORT [land vehicles]

1.394,5 (13,4%)

INDUSTRY & POWER PLANTS

925,8 (8,9%)

RESIDENTIAL & COMMERCIAL

715,9 (6,9%)

ROAD TRANSPORT [ RSD]

8.000

7.000

6.000

4.000

3.000

2.000

0

1.000

5.000

4.298,8 (41,3%)

ROAD TRANSPORT [COPERT]

Model area emissions (tn/y) AIRPORT

1.607,5 (4,7%)

PORT [ships]

2.512,5 (7,3%)

PORT [land vehicles] 565,7 (1,7%) 10.012,3 (29,3%)

INDUSTRY & POWER PLANTS 1.462,0 (4,3%)

RESIDENTIAL & COMMERCIAL

2.573,5 (7,5%)

ROAD TRANSPORT [ RSD]

© Barcelona Regional, 2010

25.000

20.000

10.000

5.000

0

Figure 5:

15.000

15.453,2 (45,2%)

ROAD TRANSPORT [COPERT]

NO2 emissions map for 2008 (right) and source apportionment for Barcelona city (top left) and for model area (bottom left).

3.4 Structural inputs for the model Certain structural and meteorological inputs were also required in order to model air quality. This is described briefly below:

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

33

 Meteorological data: actual hourly weather data for 2008 was used, taken from an urban weather station located in the city.  Cartography: the topography and surface features of the area were taken from the database of the Catalonian Institute of Cartography.  Road infrastructure: city network, highway network and bus network for Barcelona city and surrounding municipalities were introduced in the model with AAWT (annual average weekday traffic).  3-D building model: the 3-D Barcelona building model was used to model the street canyon effect. 3.5 Background concentrations Background concentration levels were based on hourly data from a specific background concentration station located on the coast in the north of Catalonia (Cap de Creus, Girona), 140 km from Barcelona. The annual average concentration levels for 2008 were 4.26 µg/m3 for NO2, 0.30 µg/m3 for NO, and 74.4 µg/m3 for O3 [11]. 3.6 Modelling results and validation process By running the ADMS-Urban model for Barcelona with the geo-referenced emissions inventory and the structural inputs of the modelled area, we obtained an annual average NO2 concentration of 14% below actual concentration levels for virtual point receptors representing real measurement stations. Therefore, 5 µg/m3 of NO2 was added in order to calibrate the model – called “local background concentrations” – and the model was re-run. After the calibration process, a very good estimate was obtained for all station measurement points. Table 1 shows the actual concentration levels for 2008 compared with the modelled values. We also obtained a very good model estimate for hourly concentration levels, as can be seen in Figure 6. Table 1: Barcelona monitoring sites Ciutadella Vall d’Hebrón Eixample Gràcia Poblenou Sants Average value

NO2 model results vs. actual results.

Type of location Urban background Urban background High traffic site High traffic site Moderate traffic site Moderate traffic site ---

Actual NO2 (µg/m3) 42.3 36.5 65.4 62.6 47.4 45.3 49.9

Modelled NO2 (µg/m3) 46.2 37.7 63.2 57.9 41.8 50.0 49.5

Model / Actual (%) 109% 103% 97% 93% 88% 110% 99%

The air-dispersion model for Barcelona revealed than 65.6% of the average annual NO2 concentration level comes from traffic, 8.6% from the residential and services sector, 4.8% from industry and power generation, 2.1% from Barcelona Port, and only 0.1% from Barcelona Airport. In addition, 8.6% comes from background pollution and 10.1% comes from “local background concentrations” after the calibration process. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

34 Air Pollution XIX 200 ug NO2/m3

NO2 REAL HOURLY Barcelona_Gràcia (ug/m3)

160

Weekend

NO2 MODELLED Barcelona_Gràcia (ug/m3)

Weekend

Weekend

Weekend

120 80 40 0 200 160

NO2 REAL ROLLING AVERAGE 24h Barcelona_Gràcia (ug/m3)

ug NO2/m3

NO2 MODELLED ROLLING AVERAGE 24h Barcelona_Gràcia (ug/m3)

120 80 40

Figure 6:

31/10/08 (Fr )

30/10/08 (Thu)

28/10/08 (Tue)

29/10/08 (Wed)

26/10/08 (Sun)

27/10/08 (Mon)

2 /10/08 (Fr )

25/10/08 (Sat)

23/10/08 (Thu)

21/10/08 (Tue)

22/10/08 (Wed)

19/10/08 (Sun)

20/10/08 (Mon)

17/10/08 (Fr )

18/10/08 (Sat)

16/10/08 (Thu)

1 /10/08 (Tue)

15/10/08 (Wed)

12/10/08 (Sun)

13/10/08 (Mon)

10/10/08 (Fr )

11/10/08 (Sat)

9/10/08 (Thu)

7/10/08 (Tue)

8/10/08 (Wed)

6/10/08 (Mon)

/10/08 (Sat)

5/10/08 (Sun)

3/10/08 (Fr )

2/10/08 (Thu)

1/10/08 (Wed)

0

Hourly and 24-hour rolling average actual and modelled NO2 concentration levels for Gràcia measurement station in October 2008. © Barcelona Regional, 2010.

4 Future scenario A Trend Scenario was developed to estimate the future air quality of the city, adding up the effects of the PECQ Action Plan. In general, certain foreseeable aspects were taken into consideration, including: future population, mobility and energy consumption considering new urban projects, social behaviour trends, new infrastructures and public transport systems to be developed over coming years, expected GDP growth, urban waste treatment, future power plants, Barcelona Port and Airport expansion plans, as well as technological improvements, especially in the transport sector with new EURO-class vehicles and the expected penetration of alternative-fuel vehicles. A future reduction of NOX emissions of 26.3% is expected with the “PECQ Scenario” by 2020 (based on the “Trend Scenario” plus the “PECQ Action Plan”). This is equivalent to reducing 2,743 tn/year compared with existing technology. To predict future air quality, we have to assess the impact the PECQ Action Plan will have on future cars, with cleaner vehicles expected thanks to the technological improvements and EU regulations. So a reduction of 1,451 tn/y of emissions is expected by 2020, equivalent to an overall reduction of emissions of 19.2% from 2008 emissions, or an annual average reduction rate of 1.8% from 2008 to 2020. We should point out that the Trend Scenario was calculated with an annual reduction rate of 0.5%. Thus, the PECQ Action Plan will help speed up the reduction rate through projects such as: “Urban mobility action plan”, “High-emissions radar”, “Agreements with the business and transport sectors to reduce diesel use in vehicle fleets”, “New more-efficient bus network”, “Increasing low-emissions vehicles in the urban bus fleet”, “Introducing new power sources for transport, including electric and gas vehicles”, etc. The future PECQ scenario fulfils current EU legislation concerning NO2 air pollution with the modelled annual average below 40 µg/m3 at all city monitoring sites. Table 2 shows expected future concentration levels, while Figure 7 shows the concentration maps for 2008 (left) and PECQ-2020 (right). WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Table 2:

35

Expected NO2 concentration levels for future scenario in 2020.

Monitoring site

Type

Ciutadella Vall d’Hebrón Eixample Gràcia Poblenou Sants Average value

Urban background Urban background High traffic site High traffic site Moderate traffic site Moderate traffic site ---

2008 ACTUAL DATA [NO2 µg/m3] 42.3 36.5 65.4 62.6 47.4 45.3 49.9

PECQ-2020 MODEL DATA [NO2 µg/m3] 30.3 24.6 39.9 36.5 29.8 32.7 32.3

© Barcelona Regional, 2010 © Barcelona Regional, 2010 0               10               20     25     30     35      40    45      50               60               70               80                                 100      

NO2  (µg/m3):

Figure 7:

Annual average NO2 concentration map for 2008 (left) and 2020 (right).

5 Discussion and conclusion The Barcelona air-dispersion model was drawn up as part of the Barcelona PECQ 2011-2020 (acronym in Catalan of Barcelona’s Energy, Climate Change and Air Quality Plan). As discussed in the paper, transport is the main culprit for current high NO2 concentrations. This sector is responsible for 48.2% of NOX emissions, and also produces 65.6% of NO2 concentration levels, on an annual average. An important study has been conducted in Barcelona to determine what kinds of vehicles use the city streets every day and their exhaust emissions. The Barcelona air-quality model has been a very important tool for drawing up the PECQ, for understanding local dispersion and detecting main pollutant sectors during the diagnosis stage, and for helping to assess and predict the impact of the measures on the city’s air quality during the policy-making process. With the adoption of the PECQ Action Plan for the year 2020, the Trend Scenario is expected to show an improvement in air quality, with annual NO2 average concentrations below 40 µg/m3 at all city stations.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

36 Air Pollution XIX

Acknowledgements We would like to thank all the institutions that have contributed to the project, particularly to Barcelona Energy Agency and Mobility Depart. of Barcelona City Council, and especially to Barcelona Regional colleagues.

References [1] Barcelona Regional & Energy Agency of Barcelona, PECQ, Pla de l’Energia, el Canvi Climàtic i la Qualitat de l’aire de 2011-2020, March 2011, Barcelona city council, Barcelona. http://www.bcnregional.com/ [2] BOP, Butlletí Oficial de la Província de Barcelona, 14 of march 2006, Num. 62, Pag. 16-22. Barcelona. [3] Barcelona Regional & Barcelona City Council, PMEB. Pla de Millora Energètica de Barcelona. Ed. Ajuntament de Barcelona & Agència d’Energia de Barcelona, 2002, Barcelona. [4] CERC, http://www.cerc.co.uk/environmental-software/ADMS-Urbanmodel html [5] Barcelona Statistics Dept. http://www.bcn.cat/estadistica/catala/dades/ vehicles/index htm [6] EMEP/EEA air pollutant emission inventory guidebook – 2009, http://www.eea.europa.eu/publications/emep-eea-emission-inventoryguidebook-2009 [7] COPERT, http://www.emisia.com/copert/ [8] Barcelona Regional & TechNet S.L., Caracterització del parc mòbil de la ciutat de Barcelona, October 2010, Ajuntament de Barcelona, Barcelona. [9] Technet S.L. http://www.technetsl.es & ESP, http://www.esp-global.com [10] Serveis de Mobilitat Aj, de Barcelona. Dades bàsiques 2008. http://www.bcn.cat/mobilitatl [11] EEA. European Environment Agency. Air quality statistics at reporting stations. http://www.eea.europa.eu/themes/air/airbase.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

37

A comparison study between near roadway measurements and air pollutant dispersion simulations using an improved line source model R. Briant1 , C. Seigneur1 , M. Gadrat2 & C. Bugajny2 1CEREA,

Joint Research Laboratory, ´ Ecole des Ponts ParisTech / EDF R&D, Universit´e Paris-Est, France 2 Centre d’Etude ´ ´ technique de l’Equipement (CETE) Nord Picardie, France

Abstract Gaussian plume models, which are widely used to model atmospheric dispersion, provide an exact analytical solution for line sources, such as roads, only when the wind direction is perpendicular to the road. Some approximations have been developed to provide an analytical formula for a line source when the wind direction is not perpendicular to the road; however, such formulas lead to some error and the solution diverges when the wind direction is parallel to the road. A novel approach that reduces the error in the line source formula when the wind direction is not perpendicular to the road was recently developed. This model, combined with a Romberg integration to account for the road section width, has then been used to simulate NOx concentrations in two large case studies (1371 road sections for the first case study and 100 for the second). NO2 , NO and O3 concentrations are then computed using the photostationary-state approximation. Finally, NO2 concentrations were successfully compared with near-roadway measurements made at various locations in the domain area (224 locations for the first case study and 70 locations for the second). Results obtained with a standard model used for regulatory applications, ADMS, are also presented. Keywords: Gaussian plume model, line source, polyphemus, ADMS. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110041

38 Air Pollution XIX

1 Introduction Atmospheric dispersion models are used to estimate the air quality impacts of road traffic emissions for many purposes, such as attainment of ambient air quality standards, health risk assessment and decision support. It may be used for instance to assess the effect of emission control measures or to help select a new road location. It is thus essential to be able to predict with reasonable accuracy the pollutant concentrations associated with vehicle emissions. To that end, analytical models have been developed to simulate the effect of atmospheric dispersion on pollutant concentrations based on an emission rate from a roadway. In open terrain, Gaussian dispersion models are the most commonly used (e.g., [1–4]). Although the Gaussian dispersion formula provides an exact solution to the atmospheric diffusion equation for the dispersion of a pollutant emitted from a point source given some assumptions on stationarity and homogeneity [5], the Gaussian dispersion formula provides an exact solution for the emissions of a pollutant from a line source only in the case where the wind is perpendicular to the line source [6]. It is, therefore, necessary to develop approximations to model atmospheric dispersion from a line source using a Gaussian formulation. One example of such a formulation is that of [7] which reduces the error in the line source formula of [8] when the wind direction is not perpendicular to the road. Although this model performs well for theoretical cases, it has not been evaluated yet with ambient concentration measurements. Here, we briefly present the model developed in [7] and we combine it with a Romberg integration to simulate the road section width (Section 2). Then in Section 3 we present results of comparison between simulations and measurements. We use this model to simulate NOx concentrations in two large case studies (1371 road sections for the first case study and 100 for the second). NO2 , NO and O3 concentrations are then computed using the photostationarystate approximation and NO2 concentrations are compared with near-roadway measurements made at various locations of the domain area (224 locations for the first case study and 70 locations for the second).

2 Gaussian plume model for line sources The Gaussian formulation of the concentration field for a pollutant emitted from a line source is the result of the integration of the point source solution over the line source: Equation (1) (reflexion terms are neglected for simplicity).


y2

C(x, y, z) = y1

Q exp 2πuσy (s)σz (s)



 −z 2 (y − s)2 − ds 2σz2 (s) 2σy2 (s)

(1)

where C is the pollutant concentration in g.m−3 at location (x, y, z), x is the distance from the source along the wind direction in m, y and z are the crosswind distances from the plume centerline in m, u is the wind velocity in m.s−1 , Q is the emission rate in g.s−1 , y1 and y2 the ordinates of the source extremities, WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

39

Figure 1: Schematic representation of the source (xsource , ysource ) and wind (xwind , ywind ) coordinate systems. The wind angle θ is the angle between the normal to the source and the wind direction.

and σy and σz are the standard deviations representing pollutant dispersion in the cross-wind directions in m, computed here with Briggs’s parameterization. In a perpendicular wind case, both source coordinate system and wind coordinate system are identical (Figure 1). Therefore, the distance of the receptor from the source in the wind direction, needed to compute σy and σz , does not change with the integration variable; so no additional approximation is required. For other wind directions, the dependency of standard deviations on the integration variable makes the integration impossible without approximations. Various approximations can be made [6]; we use here a formulation recently proposed by [8]. The Horst-Venkatram (HV) approximation consists in evaluating the integral by approximating the integrand with its behavior near ywind = 0 (see Figure 1). Solving Equation (1) with the HV approximation leads to Equation (2), which provides the concentration field for all wind directions, except θ = 90◦ . The term ucosθ represents the projection of the wind velocity onto the normal direction to the source. However, when the wind is parallel to the line source (θ = 90◦ ), the term cos θ, on the denominator of the equation, makesEquation (2) diverge.  Q −z 2 × C(x, y, z) = √ exp 2σz2 (deff ) 2 2πu cos θσz (deff ) 

 erf

(y − y1 ) cos θ − x sin θ √ 2σy (d1 )



 − erf

(y − y2 ) cos θ − x sin θ √ 2σy (d2 )

 (2)

This solution to the Gaussian equation for a line source has been shown to lead to small acceptable errors compared to an exact solution [8]; nevertheless, some errors remain due to the approximate nature of the solution, especially when the wind is nearly parallel to the line source. In [7] the error made by Equation (2) was computed and parameterized in order to correct the initial formula. For cases where the wind is parallel to the line source, the use of an WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

40 Air Pollution XIX analytical/discretized line source combination, allows one to minimize the error very effectively. Because this combination is only applied for a small range of wind directions, the increase in the overall computational time is not expected to be significant. The objective of this work was to further improve this solution for the concentration field while retaining a computationally-efficient analytical formulation to the extent possible. It provides some improvement in terms of accuracy over previous formulations of the line source Gaussian plume model without being too demanding in terms of computational resources. In addition to what is presented in [7], the model used here also includes a Romberg integration to simulate the road section width. This model was implemented in the modeling platform Polyphemus [9] which is open source and distributed under GNU GPL (http://cerea.enpc.fr/polyphemus) For simplicity, we refer to this new line source model as Polyphemus hereafter.

3 Comparison to measurements The model presented above is evaluated here with actual concentration measurements made by the French technical study and engineering center CETE Nord Picardie. Here we present results of the comparison of simulation results to two cases studies. Those two case studies includes near-roadway air quality estimations measurements by passive tube (to be distinguished from calibrated methods) along with all necessary data required to conduct simulation with Gaussian dispersion models. 3.1 Case study 1 This first case study includes concentration measurements made in Paris region, France during winter 2007 and summer 2008. The dataset contains: • The coordinates of 1371 road sections divided into 5425 smaller, but straight, sections representing a total of 831 km. • The NOx emission rates associated to each road section computed with European model COPERT 3. • Meteorological data required for a Gaussian model: wind velocity, wind direction, cloud coverage. • The measured concentrations at 224 receptor points, averaged over each overall time period of the measurement campaign (1 month in winter and 1 month in summer). • NO2 and O3 background concentrations computed with the Polyphemus Polair3d model [10]. Meteorological data of the specific time period of the measurement campaign were not available. Therefore, for this preliminary study, a generic meteorology of another year was used instead. Although this induces some uncertainty in the results, the use of values averaged over one-month periods minimize the impact of the meteorology. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

(a) Summer campaign with ‘rural’ option for Polyphemus.

41

(b) Winter campaign with ‘rural’ option for Polyphemus.

(c) Summer campaign with ‘urban’ option for Polyphemus.

(d) Winter campaign with ‘urban’ option for Polyphemus.

Figure 2: NO2 concentrations measured and simulated with Polyphemus and ADMS. (note that ADMS results are annual averages rather than periodspecifics values). The computational time required to simulate a whole month is about 2 to 3 hours with a 2, 4 GHz processor. Moreover, because the meteorological situations are independent, several processors can be used concurrently to decrease the computational burden. Figure 2 shows comparison results for each of the 224 receptor points. Several indicators were computed to estimate the error made by the model: N  (Oi − O)(Mi − M ) • Correlation: r =

i=1 N  i=1

(Oi − O)2

N 

(Mi − M )2

i=1

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

42 Air Pollution XIX

N

1 • RM SE (root mean square error): RM SE =  (Mi − Oi )2 N i=1 N 1  Mi − Oi • M N E (mean normalize error): M N E = N i=1 Oi

1  Mi − Oi N i=1 Oi where Mi and Oi are the modeled and observed values, respectively. Both ‘rural’ and ‘urban’ land category cases were tested with the Polyphemus model and Table 1 summarizes the results. Polyphemus with the option ‘rural’ has the best performance for the summer campaign whereas it is with the option ‘urban’ that performance is the best for the winter campaign (except for the RMSE which is better with the option ‘rural’). In addition, it can be seen that results are better in summer than in winter. Finally, in the last column of Table 1, results, obtain by the CETE Nord Picardie, with the atmospheric dispersion model ADMS [11] are presented. ADMS is a standard Gaussian dispersion model that is widely used for regulatory applications. Polyphemus seems to give better results but it should be noted that results of the ADMS simulation were averaged over a whole year and do not correspond to the time periods of the measurement campaign. That is the reason why the curve of ADMS in Figure 2 is the same in winter and in summer. Nevertheless, this N

• M N B (mean normalize bias): M N B =

Table 1: Performance indicators of Polyphemus and ADMS for the case study 1. Polyphemus

Summer

ADMS1

Rural

Urban

0.73

0.72

0.71

) 10.08

14.77

12.03

MNE

0.30

0.39

0.56

MNB

0.09

−0.36

0.47

Winter

Polyphemus Rural Urban

Correlation RMSE (in µg.m

−3

Correlation RMSE (in µg.m

−3

ADMS

0.65

0.68

0.67

) 12.43

15.12

13.94

MNE

0.48

0.34

0.72

MNB

0.28

−0.15

0.62

1

Note that ADMS results, compared to period-specific values rather than annual average values and, therefore, should be seen as preliminary. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

(a) Summer campaign with ‘rural’ option for Polyphemus.

(c) Summer campaign with ‘urban’ option for Polyphemus.

43

(b) Winter campaign with ‘rural’ option for Polyphemus.

(d) Winter campaign with ‘urban’ option for Polyphemus.

Figure 3: NO2 concentrations measured and simulated with Polyphemus and ADMS. (note that ADMS results are annual averages rather than periodspecifics values)

results provide some preliminary estimates of the model performance on a large case study. 3.2 Case study 2 Measurements were made in the Lille (France) region in 2010 and the dataset contains: • The coordinates of 100 road sections divided into 362 smaller, but straight, sections representing a total of 29.6 km. • The NOx emission rates associated to each road sections computed with the European model COPERT 4. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

44 Air Pollution XIX

Table 2: Performance indicators of Polyphemus and ADMS for case study 2. Polyphemus Rural Urban

Summer Correlation

ADMS2

0.47

0.54

0.51

10.39

9.43

7.54

MNE

0.35

0.32

0.22

MNB

0.32

0.27

0.13

RMSE (in µg.m

−3

)

Polyphemus

Winter Correlation RMSE (in µg.m MNE MNB

−3

)

ADMS

Rural

Urban

0.54

0.59

0.45

6.15

6.54

9.27

0.11

0.12

−0.04 −0.07

0.18 −0.16

1 Note that ADMS results, compared to period-specific values rather than annual average values and, therefore, should be seen as preliminary.

• Meteorological data required for a Gaussian model: wind velocity, wind direction, cloud coverage. • The measured concentrations at 70 receptor points, average over each overall time period of the measurement campaign (1 month in winter and 1 month in summer). • NO2 and O3 background concentrations measured at a fixed urban background measurement station. This case study is much smaller than the previous one and, accordingly, the computational time required for these simulations was much smaller (a few minutes). Nevertheless, this case study presents two advantages over the previous one. First, meteorological data in this case study, match the measurement campaigns time periods. In addition, emission rates were computed with the more recent model COPERT4, instead of COPERT3 in the previous case study. Figure 3 shows comparison results for each of the 70 receptor points. Performance indicators were computed and are summarize in Table 2. ADMS results, obtain by the CETE Nord Picardie, gives better results for the summer campaign except for the correlation, which is better with Polyphemus with the ‘urban’ option. For the winter campaign, Polyphemus is better with both ‘rural’ and ‘urban’ options.

4 Conclusion The Gaussian plume model of Polyphemus for line sources has been presented and evaluated with two case studies. The first case study a large roadway system, but WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

45

meteorological data did not match measurements time periods. The second case study is smaller but with the correct meteorological data. Polyphemus performs well on both cases when confronted to both measurements and to ADMS model results. Ongoing work is now to incorporate this Gaussian model into a 3D Eulerian gridded model to constitute a plume-in-grid model, which would allow us to improve the representation of the impact of roadway traffic in Eulerian models.

References [1] Levitin, J., H¨ark¨onen, J., Kukkonen, J. & Nikmo, J., Evaluation of the caline 4 and car-fmi models against measurements near a major road. Atmos Env, 39, pp. 4439–4452, 2005. [2] Berger, J., Walker, S.E., Denby, B., Berkowicz, R., Fstrøm, P.L., Ketzel, M., H¨ark¨onen, J., Nikmo, J. & Karppinen, A., Evaluation and inter-comparison of open road line source models currently in use in the nordic countries. Boreal Env Res, 15(319–334), 2010. [3] Venkatram, A., Isakov, V., Seila, R. & Baldauf, R., Modeling the impacts of traffic emissions on air toxics concentrations near roadways. Atmos Env, 43, pp. 3191–3199, 2009. [4] Chen, H., Bai, S., Eisinger, D., Niemeier, D. & Claggett, M., Predicting nearroad PM2:5 concentrations: comparative assessment of caline4, cal3qhc, and aermod. Transportation Research Record, Journal of the Transportation Research Board, 2123(26–37), 2009. [5] Csanady, G., Turbulent diffusion in the environment. D Reidel Publishing Company, Dordrecht, The Netherlands, 1973. [6] Yamartino, R., AIR QUALITY MODELING - Theories, Methodologies, Computational Techniques, and Available Databases and Software. Vol IIISpecial Issues. EnviroComp Institute and the Air & Waste Management Association, 2008. [7] Briant, R., Korsakissok, I. & Seigneur, C., An improved line source model for air pollutant dispersion from roadway traffic. Atmos Env, 2010. In press, doi:10.1016/j.atmosenv.2010.11.016. [8] Venkatram, A. & Horst, T., Approximating dispersion from a finite line source. Atmos Env, 40, pp. 2401–2408, 2006. ´ [9] Mallet, V., Qu´elo, D., Sportisse, B., Ahmed de Biasi, M., Debry, E., Korsakissok, I., Wu, L., Roustan, Y., Sartelet, K., Tombette, M. & Foudhil, H., Technical Note: The air quality modeling system Polyphemus. Atmos Chem Phys, 7(20), pp. 5479–5487, 2007. [10] Roustan, Y., Pausader, M. & Seigneur, C., Estimating the effect of onroad vehicle emission controls on future air quality in paris, france. Atmos Environ, 2010. In press, doi:10.1016/j.atmosenv.2010.10.010. [11] McHugh, C., Carruthers, D., Higson, H. & Dyster, S., Comparison of model evaluation methodologies with application to ADMS 3 and U.S. models. Int J Env Pollut, 16(1–6), 2001. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

This page intentionally left blank

Air Pollution XIX

47

Regional on-line air pollution modelling system in highly complex terrain P. Mlakar, M. Z. Božnar & B. Grašič MEIS doo., Slovenia

Abstract In the paper, a national project with the title “Prognostic and diagnostic integrated regional air pollution modelling system” is described. It shows that such a project can significantly contribute to the proper understanding of the air pollution in smaller regions with a very complex topography. It also describes how foreseen scientific problems were solved and the necessary testing, improvements and validation were made. The development of a Lagrangian particle model-based air pollution modelling system that works in an on-line diagnostic mode and covers air pollution from several industrial and other sources in the region over a highly complex terrain, is described. To achieve online efficiency some new methods of obtaining high resolution short-range meteorological fields derived from mesoscale models were developed and the implementation of advanced Lagrangian models, acceleration techniques and novel approaches for whole system integration are presented. The project’s test bed was established as a novel approach to the overall treatment of the scientific – applicative project goal. Keywords: on-line regional air pollution modelling system, Lagrangian particle dispersion model, Zasavje region, test bed, complex terrain, non-hydrostatic numerical weather forecast model, nesting.

1 Introduction In some areas of Slovenia, local inhabitants or environmental associations are strongly against the operation of some industrial plants due to the air pollution they cause. These problems are regulated by the European directive of Integrated Pollution Prevention and Control (IPPC) which requires among others that the industrial source’s influence on the ambient air is modelled once to obtain the WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110051

48 Air Pollution XIX IPPC permit. The future European directives that are now in the preparatory phase also emphasize the use of air pollution models in an on-line mode for informing the local community, as the state-of-the-art science is enabling this already. In this paper, a national project with the title “Prognostic and diagnostic integrated regional air pollution modelling system” is described. It shows that such a project can significantly contribute to the proper understanding of the air pollution in smaller regions with a very complex topography.

2 Zasavje region In Zasavje, air pollution is a very serious problem. Pollutant concentration limit values in the atmosphere are exceeded many times; currently the greatest problem is due to dust particles PM10. Particle PM10 (dust fraction that is regulated by the EU directive) air pollution is one of the most important air pollution problems in Slovenia and also in the EU. Several measurement sites show that the basic norms for daily and yearly PM10 concentrations are not being achieved. Based on this fact, the European commission reminded Slovenia in June 2008 that actions should be made to achieve the EU directives’ requirements for a lower ambient air pollution due to the particles PM10 (especially the new Directive 2008/50/ES on ambient air quality and cleaner air for Europe). Unfortunately the measures taken were not sufficient and therefore in July 2010 the European Commission started suing Slovenia in the Court of the European Union regarding this matter. An air pollution modelling system in the atmosphere provides some answers regarding the causes for pollution and pollution mechanisms; in particular, it gives the answer to the spatial and temporal distribution of pollution. Although measuring stations provide very accurate results, mainly in terms of the highly complex terrain, these data represent only a very small area in the immediate vicinity of the measuring station. Highly complex orography and consequently very complex micrometeorological conditions over the small area of the municipalities of Zasavje represent a considerable scientific challenge for modelling both the meteorological conditions and the spread of pollutants in the atmosphere. This project will contribute towards a new dimension of the general understanding of the problem of the air pollution. From a scientific point of view, one of the important objectives of this project is to demonstrate the correlation (in space and time) between the modelled concentrations and the measured concentrations at the locations of the numerous automatic measuring stations in Zasavje.

3 Regional air pollution modelling system At the European level, the European Commission in cooperation with the European Environment Agency set up a forum FAIRMODE [1], responsible for guidance on good practice and the correct usage of modelling tools for any purpose regarding the regulatory use, that is the use with the aim of modelling the concentration of pollutants in the atmosphere for pollution control purposes. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

49

The regional air pollution modelling system was developed within this project and is fully consistent with all the requirements imposed by FAIRMODE. Among the most important requirements, there are requirements for a previously successful validation of the modelling system in a similarly complex area (size of domain, terrain complexity and most of all a similar complexity of meteorological conditions) [2, 3]. Furthermore, it requires that when using modelling systems in order to assess the impacts of industrial sources, that the cell size in the horizontal direction is at the most 250m. The air pollution modelling system is a mathematical tool which illustrates the mechanism of spreading the pollutants into the atmosphere. The modelling system based on the input data regarding the meteorological conditions and pollutant emissions, calculates the consequences of these emissions as the concentration in the atmosphere in the area of emission sources are taken into consideration. Emission sources that have not (yet) been entered into the system, are not shown by the modelling system (emissions from other industries, traffic, local furnaces, biomass combustion and combustion of wastes in the countryside etc.). The modelling system that allows these calculations for the shown area in Zasavje is composed of several models and uses different input data. General presentation of the model is illustrated on Figure 2. For the most accurate matching of the modelled concentrations with the measured concentrations, high-quality input data are of key importance, especially the qualitative measured meteorological data in the area discussed. For now, the prognosticated data on their own are not yet a sufficient basis for modelling the air pollution spreading over such a complex terrain as Zasavje. In order to achieve the matching results of the models with the actual measured concentrations in space and time it is also necessary to include the local meteorological data measured by appropriate meteorological models. 3.1 Emission In order to obtain a quality modelling system it is necessary to describe the emission as much as possible. This is only possible with industrial sources which emit pollutants through stacks. Part of the emissions cannot be included in the industrial sources as they cannot control the emissions through the stacks. In 2010, we teamed up with four major industrial sources which release the emissions into the atmosphere of Zasavje: the thermal power plant Termoelektrarna Trbovlje, the cement factory Lafarge Cement, the glass product manufacturing company Steklarna Hrastnik and the building material company IGM. They have presented us with a detailed description of the physical features of the dischargers (height, diameter, flow rate etc.) and the amount of pollutants that can be emitted into the air. These data are from the environmental protection permit which must not be exceeded.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

50 Air Pollution XIX 3.2 Measured meteorological data The test bed of the regional air pollution modelling system which was started in 2010 includes the integration of the meteorological data from the national network for monitoring air quality of the Slovenian Environment Agency (Agencija za okolje Repulike Slovenije) and from the environmental measuring system of Termoelektrarna Trbovlje. Measured environmental data from stations are presented in Table 1. Station locations are shown on the right hand side of Figure 1. Table 1:

Figure 1:

Measured environmental data at the automatic stations.

Zasavje region (left - 3D picture of orography, right - 2D picture with station locations).

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Figure 2:

Regional air pollution modelling system.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

51

52 Air Pollution XIX 3.3 Meteorological model In order to perform a diagnostic operation of our 3D air pollution modelling system we need prognostic data for the wind prognosis and temperature profile above the domain (as the only available measurements were performed on the ground), as measurements with the balloon are performed only once per day and those are in Ljubljana, and other advanced measurements (SODAR, RASS etc.) are not available for Zasavje. In the previous year, we chose the WRF Weather Research & Forecast model [4], which consists of two Meso models ARW (Advanced Research WRF), maintained by NCAR [5], and the NMM (Nonhydrostatic Mesoscale Model), maintained by NOAA/NCEP [6]. The model frame (data acquisition, data processing) is common; the difference is only in the dynamic core. Both models were installed in different resolutions and with different numbers of nesting on a four core computer (Dual Core Quattro) with a 64 bit system Open SUSE. Boundary and initial data are provided by a global meteorological model GFS (NCEP centre from America). For our project, the module ARW was chosen, which was intended for research and the module NMM was intended for more routine weather prediction – however, we were interested in more specific meteorological features. The configuration of the ARW model which has been running daily from February 2010 is as follows: - two domains; - duration of forecasting: 2 days and 3 hours; - larger domain (Central Europe): 101 × 101 pixels in a resolution of 12km per hour; - smaller domain (Slovenia with surroundings): 76 × 76 pixels in a resolution of 4km per 30min; - the model starts running at 5:00 UTC; - the simulation runs from 3 to 4hours. An example of cloudiness prediction with the nesting method on two areas of different sizes is shown in Figure 3 with an illustrative presentation of the size of each domain. 3.4 Air pollution dispersion model At first, in order to register the meteorological conditions for every 30 minutes in a continuous mode, meteorological data from the automatic measuring stations in this area are used. As it is also essential to describe the vertical profile of the wind, temperature and relative air humidity, an approximation using a profile calculated by a prognostic meteorological model is used. Measurements with the SODAR would be more appropriate, unfortunately they are not available at the moment. All the data are processed by a meteorological pre-processor SurfPro and a three dimensional mass consistent wind model Swift [7–9] (both products by the company Arianet Srl).

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Figure 3:

Example of forecast with the nesting.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

53

54 Air Pollution XIX The modelling system also implements the data on the terrain altitude and land use, both in the above mentioned resolution. Then, the movement of pollutants from their sources towards the hills and valleys is entered in this three-dimensional meteorological area, resulting in the calculated pollutant concentrations as shown in the Figures. For this part, a numerical Lagrangian particle model Spray [9] by the company Arianet Srl is used. The above-mentioned approach in the current state of science in this area with complex terrain provides the best results. This is evidenced by the articles published in the scientific journals [2, 3]. The examples of the model results are shown in the lower part of Figure 2, on the left hand side is shown one of the modelled ground concentrations and on the right hand side, there is shown a 3D picture of the release of one of the sources.

4 KOOREG - public internet portal All the results from the developed air pollution modelling system for Zasavje are given in real time and are available on-line in the portal site named KOOREG – Air pollution control in the region. The portal site shows a concentration of pollutants SO2, NO2 and PM10 due to the operation of different sources of emissions into the atmosphere of Zasavje (http://www.kvalitetazraka.si). During the first phase of the project, the portal site already shows the concentration of pollutants in the air in real time for each 30 minutes, for the past 2 days. The project specifically authorized the use of their nominal (normal maximum operating) emission values: The thermal power plant Termoelektrarna Trbovlje, the cement plant Lafarge Cement, the glass product manufacturing company Steklarna Hrastnik and the building material company IGM Zagorje. For now, only concentrations in the atmosphere which are the result of emissions from these sources are shown. In the modelling system these values are used as if these facilities are operating 24hours every day of the year, therefore the users of the portal site are asked to check the (non-) operation with the individual participant. In the future, we plan to use the simultaneous and automatic use of the emission values (where this is of interest). Besides the concentration, the portal site also offers a current weather forecast for 48 hours in the future. There are forecasts for precipitation, cloudiness, wind, visibility, air pressure and temperature. The portal is also equipped with a non-technical description of the modelling system and the portal site which is intended for the general public. The description is made in order to offer a concise and as simple and non-technical form as possible so that the end users would be able to become acquainted with the content of the portal site. Figure 4 shows the home page. In the upper menu bar, general descriptions of the portal site, the modelling system and contact information are given.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Figure 4:

55

Public portal KOOREG (http://www kvalitetazraka.si).

This is followed by the first group of results of the modelling system, consisting of a weather forecast. By clicking on one of the group pictures (e.g. temperature) another window opens, providing a detailed review of the predictions for 48 hours with an animation effect, as shown on the left hand side of Figure 5. The second group of results of the modelling system is the average pollutant concentrations for every 30 minutes for the past 24 hours. By clicking on each picture another window opens with a detailed review of the last available concentrations. This is shown in Figure 6. In addition to the last concentration, there is also the criteria description for each pollutant and a legend which allows the user to determine the area with critical situations in a clear and simple way. By clicking on the animation capture another window opens which provide a detailed review of the predictions for the past 24 hours with an animation effect, as shown on the right hand side of Figure 5.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

56 Air Pollution XIX

Figure 5:

Animation of results (left – temperature forecast for 48 hours in the future, right – air pollution ½-hour concentrations for the last 24 hours) (TEST OPERATION).

Figure 6:

Example of ½-hour air pollution concentration of NO2 from TPP Trbovlje.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

57

5 Conclusions The regional air pollution modelling system gives a completely new way of monitoring the air pollution with the main purpose of providing a better living environment in Zasavje. The air pollution modelling system in the atmosphere provides several answers regarding the causes and pollution mechanisms, in particular the answer regarding the spatial and temporal distribution of the pollution. It is important that the modelling system used a set of models which had demonstrated in previous validations [2, 3] a well grounded matching of the modelled and the measured concentrations in space and time on this and a similarly complex terrain in Slovenia. The modelling system is intended for both the residents and the polluters. The portal site KOOREG with the modelling results is supposed to help people in order to think about pollution in a more comprehensive manner. They will also be able to find out in the portal site which of the polluters is responsible for the pollution. As one of these four polluters will not always be responsible for the pollution, eventually the system will try to also include smaller sources of air pollution which do not fall under the IPPC Regulation yet are strongly polluting the atmosphere in their immediate vicinity due to their unfavourable location. In the near future, the portal site will also be updated with the pollution prognosis for one day in advance. The pollution forecast will also be of help to the polluters who will be able to reduce the emissions for the approaching day and thereby help to mitigate the concentration of pollutants in the atmosphere.

Acknowledgements The authors gratefully acknowledge the financing of the project by the Slovenian Research Agency, Project No. L1-2082. The authors would like to acknowledge all the four companies mentioned in this article (Termoelektrarna Trbovlje, Lafarge Cement, Steklarna Hrastnik and IGM Zagorje) for their voluntary participation in this project and their permission to use their nominal operating emission values.

References [1] FAIRMODE., Guidance on the use of models for the European Air Quality Directive, working document of the Forum for Air Quality Modelling in Europe FAIRMODE ETC/ACC report Version 6.2, Editor: Bruce Denby, In: FAIRMODE. Available from: http://fairmode.ew.eea.europa.eu/ fol429189/forums-guidance/model_guidance_document_v6_2.pdf, 08.03.2011, 2010 [2] Grašič B., Božnar M. Z., Mlakar P., Re-evaluation of the Lagrangian particle modelling system on an experimental campaign in complex terrain, Il Nuovo Cimento C, Vol. 30, No. 6, pp. 557-575, 2007 [3] Božnar M., Mlakar P., Grašič B., Air pollution dispersion modelling around Thermal power plant Trbovlje in complex terrain: model verification and WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

58 Air Pollution XIX

[4] [5] [6]

[7] [8] [9]

regulatory planning. V: Borrego, C. (ed.), Miranda, A. I. (ed.). Air pollution modeling and its application XIX : [proceedings of the 29th NATO/OCCMS International Technical Meeting on Air Pollution Modelling and its Application, Aveiro, Portugal, 24-28 September 2007], (NATO science for peace and security series, Series C, Environmental security). Dordrecht: Springer, cop. 2008, pp. 695-696, 2010 WRF, The Weather Research & Forecasting Model, Available from: http://www.wrf-model.org/, 10.05.2011 The National Center for Atmospheric Research (NCAR), NCAR ARW WRF Forecast, Available from: http://www.wrf-model.org/plots/realtime_ main.php, 10.05.2011 National Oceanic and Atmospheric Administration (NOAA)/National Centres for Environmental Prediction (NCEP), NMM Model Analyses and Forecasts, Available from: http://www nco ncep.noaa.gov/pmb/nwprod/ analysis/namer/hiresw/12/model_l.shtml Brusasca G., Tinarelli G., Anfossi D., Particle model simulation of diffusion in low windspeed stable conditions", Atmospheric Environment Vol. 26, pp. 707-723, 1992 Anfossi D., Ferrero E., Brusasca G., Marzorati A., Tinarelli G., A simple way of computing buoyant plume rise in Lagrangian stochastic dispersion models, Atmospheric Environment Vol. 27A, pp. 1443-1451, 1993 Tinarelli G., Anfossi D., Bider M., Ferrero E., Trini Casteli S., A new high performance version of Lagrangian particle dispersion model SPRAY, some case studies., Air pollution modelling and its Applications XIII, S. E. Gryning and E. Batchvarova eds., Kluwer Academic / Plenum Press, New York, pp. 499-507, 2000

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

59

Identification of potential sources and transport pathways of atmospheric PM10 using HYSPLIT and hybrid receptor modelling in Lanzhou, China N. Liu1,2, Y. Yu1, J. B. Chen1, J. J. He1,2 & S. P. Zhao1,2 1

Key Laboratory of Land Surface Process and Climate Change in Cold and Arid Regions, Cold and Arid Regions Environmental and Engineering Research Institute, Chinese Academy of Sciences, Lanzhou, Gansu, China 2 GraduateUniversity of Chinese Academy of Sciences, Beijing, China

Abstract Three-dimensional 4-day backward trajectories arriving at Lanzhou 500m above ground level were calculated every 6 h using HYSPLIT-4 trajectory model for spring (March, April and May) 2001 to 2008. The 8 years were divided into two categories: high dust years (2001, 2002, 2004 and 2006) and low dust years (2003, 2005, 2007 and 2008). Cluster analysis, potential source contribution function (PSCF) model, and concentration-weighted trajectory (CWT) method were used to evaluate the transport pathways and potential source regions affecting PM10 loadings in Lanzhou in spring season. Results indicate that the western and the northwestern pathways, accounting respectively 33% and 19.4% of all trajectories, were major pathways leading to high springtime PM10 loadings in Lanzhou during high dust years. However, the major pathways were the western and the northern pathways in low dust years, accounting for 23.6% and 18%, respectively. There were six potential source regions affecting PM10 concentration in Lanzhou, e.g. the Tarim Basin and the Turpan Basin in Xinjiang, the Qaidam Basin in Qinghai, the Hexi Corridor in Gansu province, the desert and Gobi areas in the middle and west of Inner Mongolia, and the Loess Plateau in the middle of Shaanxi province and eastern Sichuan. Keywords: HYSPLIT, PM10, trajectories, transport pathways, potential sources.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110061

60 Air Pollution XIX

1 Introduction With the rapid economic growth and accelerating urbanization, the primary pollutant in most Chinese urban areas has changed from SO2 and TSP (Total Suspended Particles) to PM10 (aerosol particles with an aerodynamic diameter less than 10μm). PM10, as atmospheric aerosols, affects the Earth’s climate directly by absorbing or scattering solar and terrestrial radiation, and indirectly by altering cloud formation, microphysical properties, and lifetimes, [1, 2]. In East Asia, high PM10 loadings could arise from natural processes, e.g. surface dust, sea spray, and volcanic dust etc., anthropogenic activities, [2, 3], e.g. the combustion of fossil fuels and industrial production activities etc., and secondary aerosol particles. Dust storms occurring in the desert and Gobi desert in Central Asia and the northern regions of China, are not only a major weather phenomenon influencing the springtime climate in East Asia, but also a main source of atmospheric particles in spring in China. The desert areas in China, which occupy approximately 13% of China’s land area, are major sources of Asian dust. A large number of observations and studies have shown that dust storms in Asia could not only increase the atmosphere aerosol concentrations in the local and adjacent areas, but also transport to the eastern parts of China, Korea, Japan and even across the Pacific to North America, [4–6]. Lanzhou, located in the northwest arid and semi-arid regions of China, is one of the most polluted cities in China. The PM10 concentrations in spring are dramatically affected by dust storms, [7]. Most of the previous studies on air pollution in Lanzhou were focused on the temporal variability of pollutants, [8], local transport and dispersion characteristics of pollutants, [9], the impact of dust weather, [10, 11], and control measurements, [12, 13], with little or no studies on the transport pathways and the source regions that affect air quality in Lanzhou. In order to control air quality, pollution sources must be identified, emission estimates made, and effective management strategies developed. Many studies have shown the signification correlation of the spatial and temporal variation of pollutants, including PM10, [7], mercury, [14, 15], pollen, [16], and dust outbreak, [17], with air mass transport pathways. Hybrid receptor modeling, such as potential source contribution function (PSCF) model and concentrationweighted trajectory (CWT) method, has been used successfully for potential source region identification for PM10 and TGM (total gaseous mercury), [14, 18]. This research aims at identifying the transport pathways and the potential source regions that lead to elevated PM10 concentrations in Lanzhou and quantifying the relative contribution of the source regions to the PM10 loadings during spring (March, April and May) of 2001-2008. The results from this research would not only provide some scientific basis for improving the air quality in Lanzhou and the ecological environment of the surrounding regions, but also accumulate some experience for cities with similar situation in the world.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

61

2 Data The PM10 concentration data for Lanzhou in spring 2001 to 2008 were calculated from the air pollution index (API) reports for major Chinese cities, [19]. The process used to calculate PM10 concentrations from the API has been described in Zhang et al. [20]. The dust storm records (daily) were obtained from the Meteorological Administration of Gansu Province for the spring 2001 to 2008. The NCEP/NCAR (National Centers for Environmental Prediction and the National Center for Atmospheric Research) Reanalysis archive data were used as meteorological input data for trajectory calculations. The horizontal resolution of the data are 2.5°×2.5°in latitude and longitude, which are archived four times every day (00, 06, 12, 18 UTC).

3 Methods 3.1 Trajectory clustering analysis Three-dimensional 4-day backward trajectories arriving at Lanzhou (Lat: 36.05N, Lon: 103.85E, 1518m above sea level) 500m above ground level (AGL) were calculated every 6 h (00, 06, 12, 18 UTC) using the National Oceanic and Atmospheric Administration (NOAA) HYSPLIT-4 (Hybrid Single-Particle Lagrangian Integrated Trajectory) model, [21], for the spring 2001 to 2008. The eight years were divided into two categories, i.e. high dust years with more than 6 dust storms in spring (2001, 2002, 2004 and 2006) and low dust years (2003, 2005, 2007 and 2008). The final model outputs were hourly backward trajectory endpoints indicating the geographical location and the height of the air parcel. Ward’s hierarchical clustering method was used for all the eight spring seasons based on the mean angle between all pairs of trajectories, [7, 22]. The major transport pathways leading to the elevated PM10 concentrations in Lanzhou during spring could be obtained by combining the trajectory with the daily PM10 concentration data. 3.2 Potential source contribution function (PSCF) A potential source contribution function (PSCF), [23], was used to identify the potential source regions that affect the PM10 loadings during springtime in Lanzhou. The PSCF values for the grid cells in the study domain were calculated by counting the trajectory segment endpoints terminating within each cell. By defining the number of endpoints that fall in the ijth cell as nij and the number of endpoints that corresponds to a PM10 concentration above an arbitrarily set criterion when arriving at Lanzhou in the same grid cell as mij, the PSCF value for the ijth cell can be defined as: PSCFij  mij / nij . (1) Thus, the PSCF value can be interpreted as a conditional probability describing the potential contributions of a grid cell to the high PM10 loadings in WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

62 Air Pollution XIX Lanzhou. That is, grid cells related to high PSCF values are the potential source regions, and the trajectories passing these cells are the major transport pathways leading to high PM10 loadings during spring in Lanzhou. In this study, the spring time daily-averaged PM10 concentration is used as the criterion for counting mij. To reduce the uncertainty of PSCF resulted from small nij values, an arbitrary weight function Wij is multiplied to the PSCF value to better reflect the uncertainty in the values for cells with small nij. The weight function Wij is defined as follows, [7], 1.00, 40  nij  0.70,10  nij  40 . Wij   0.42,5  nij  10 0.17, n  5 ij 

(2)

The weight function reduced the PSCF value when the total number of endpoints in a particular cell was less than about three times the average value of the endpoints per cell (about 40 in this study), [24]. In this study, geographic areas covered by more than 95% of the backward trajectories were selected as the study domain. For the springtimes in high dust years (2001, 2002, 2004 and 2006), the study domain extends from 55° E to 125° E and from 25° N to 65° N, thus composing 11,200 cells 0.5° × 0.5° in latitude and longitude. The total number of trajectory endpoints located in the study domain is 142,784 so there would be about 13 endpoints per cell on average. That is, it is necessary to reduce the uncertainty of PSCF values by using eqn. (2) when the number of trajectory endpoints nij in a grid cell is less than about 40. 3.3 Concentration-weighted trajectory (CWT) method One limitation of the PSCF method is that grid cells may have similar PSCF values when PM10 concentrations at the receptor site are either only slightly or extremely higher than the average value in spring. The PSCF value can only give the spatial distribution of potential source regions and cannot give information on the relative contribution of different potential source regions. To compensate the limitation, a concentration-weighted trajectory (CWT) method, [18, 25], was used to calculate the trajectory weighted concentration. In the CWT method, each grid cell is assigned a weighted concentration by averaging the sample PM10 concentrations that have associated trajectories crossing the grid cell as follows:

Cij 

1



M

 l 1 ijl

M

C  l 1

l ijl

(3)

where Cij is the average weighted concentration in the ijth cell, l is the index of the trajectory, Cl is the PM10 concentration measured on the arrival of trajectory l, M is the total number of trajectories, and τijl is the time spent in the ijth cell by trajectory l, [18]. The eqn. (2) was also applied to the calculation of CWT to reduce the uncertainties when nij is small. A high Cij value implies that air WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

63

parcels traveling over the ijth cell would be associated with high PM10 concentration at the receptor site Lanzhou.

4 Results and discussion 4.1 PM10 pollution in spring The occurrence of dust storms during springtime 2001 to 2008 in Lanzhou is summarized in table 1. According to the number of dust storm events, years 2001, 2002, 2004 and 2006 are classified as high dust years and years 2003, 2005, 2007 and 2008 are classified as low dust years. Table 1:

Occurrence of dust storms in spring during 2001–2008.

Year Dust events (num)

2001 15

2002 9

2003 6

2004 11

2005 2

2006 12

2007 5

2008 4

The springtime daily-averaged PM10 concentration for Lanzhou is 238.8μg/m3 in high dust years (fig. 1) and there are 138 days with daily PM10 concentrations higher than the average. The daily-averaged PM10 concentration is 157.2μg/m3 in low dust years (fig. 2) and there are 123 days with higher than average PM10 concentrations. The daily-averaged PM10 concentration for the four spring seasons in high dust years is much higher than the national Grade II

Figure 1:

Figure 2:

Daily-averaged PM10 mass concentrations at Lanzhou in spring for the years 2001, 2002, 2004 and 2006.

Same as fig. 1 but for the years 2003, 2005, 2007 and 2008.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

64 Air Pollution XIX standard for daily mean PM10 concentration of 150μg/m3 (GB3095-1996), while it is slightly higher than the Grade II standard in low dust years. The averages in both categories exceed the Grade II standard by 67.7% and 38.0%, respectively. In addition, there are 51 days (accounted for 13.9%) and 13 days (accounted for 3.5%) when the national Grade V PM10 standard of 420μg/m3 were exceeded during the spring of the high and the low dust years, respectively. 4.2 Cluster-mean backward trajectories 4.2.1 Transport pathways of mean backward trajectories Seven clusters (1 to 7) were produced by the clustering algorithm for the high dust years, and the cluster-mean trajectories are shown in fig. 3(a). Six clusters (1 to 3 and 5 to 7) were obtained for the low dust years (fig. 3(b)). The transport routes and the direction of trajectories indicate the geographical areas traveled by air masses before their arrival at the receptor site. The length of the cluster-mean trajectories indicates the transport speed of air masses. The longer is the clustermean trajectory, the faster is the air mass. It is seen from fig. 3 that the western and the northwestern trajectories (cluster 2 and cluster 3) were longer than trajectories from other directions, indicating that air masses from the west and the northwest moved faster than others.

(a)

Figure 3:

(b)

Cluster-mean back-trajectories arriving at Lanzhou in spring for (a) 2001, 2002, 2004 and 2006 (b) 2003, 2005, 2007 and 2008.

In high dust years, the air mass associated with cluster 1 were from western China, including the north Qinghai-Tibet Plateau, and the Gobi desert around the Qinghai and Xinjiang border. These trajectories moved southeasterly over the Qaidam basin and the border areas of Sichuan and Gansu province, and then turned northerly to Lanzhou. The air masses associated with clusters 2 and 3 were from Xinjiang province. The Cluster 2 air masses were from the Tarim Basin in southern Xinjiang; these trajectories passed the Taklimakan desert, and then moved westerly over the Qaidam Basin in northern Qinghai before arriving at Lanzhou. The Cluster 3 air masses were from the Junggar Basin, north of Xinjiang province; these trajectories moved southeasterly into the Gansu province and travelled along the Hexi Corridor to Lanzhou. The air masses associated with clusters 4 and 5 were initially from the desert, semi-desert and Gobi regions of Mongolia. The cluster 4 air masses moved from the southwestern Mongolia and travelled southeasterly over the Badain Jaran Desert WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

65

in the west of Inner Mongolia to Lanzhou; in contrast, the cluster 5 moved from the middle of Mongolia and travelled southerly to the middle of the Inner Mongolia, then passed through the Tengger Desert and finally turned southwesterly to Lanzhou. The air masses associated with clusters 6 and 7 were from the Loess Plateau and the border regions between Shaanxi and Sichuan province, respectively. The six clusters in the low dust years are denoted following the high dust years for easy comparison. In the low dust years, the air masses associated with cluster 1 were from the marginal regions south of the Tarim Basin, these trajectories moved westerly through the neighboring regions of Xinjiang, Qinghai and Tibet and crossed the south of Qinghai province, and then turned northerly to Lanzhou. The air masses associated with cluster 3 in the low dust years were from the Junggar Basin, north of Xinjiang province, same as the cluster 3 in the high dust years, but the part of pathway within the Gansu province was different. The air masses associated with cluster 3 in low dust years moved southeasterly over the desert and Gobi desert regions near the border of the Inner Mongolia and the Gansu province to Lanzhou. The trajectories in cluster 5 in low dust years moved over Inner Mongolia and directed to Lanzhou without passing through the Ningxia province, which is different from the one in high dust years. The air masses associated with cluster 6 traveled more southerly in low dust years than the corresponding one in high dust years. The air masses associated with cluster 7 were from Sichuan province in low dust years, as in high dust years, but located more southerly. 4.2.2 Transport pathways of the polluted trajectories All the backward trajectories were divided into two groups, i.e. the polluted and the clean trajectories, according to the sample PM10 concentration when they arrived at the receptor site. The polluted trajectories are those with PM10 concentrations higher than the average in spring seasons, i.e. 238.8μg/m3 and 157.2μg/m3 for the high and the low dust years, respectively. The number of trajectories, the percentage and the corresponding daily PM10 concentrations in each cluster for all trajectories and for polluted trajectories are summarized in tables 2 (for high dust years) and 3 (for low dust years). In high dust years, the average PM10 concentrations of clusters 2 (287.7μg/m3) and 3 (275.5μg/m3) exceed the spring season average 238.8 μg/m3, with more than 50% of trajectories being polluted ones, which indicates that the air masses associated with these clusters would carry more particulate matters and lead to high PM10 loadings in Lanzhou in high dust years. The clusters 2 and 3 represent the western and the northwestern pathways, which passed through major Asian dust source areas, therefore these two pathways, were major pathways carrying dust to Lanzhou. In comparison, fewer trajectories were assigned to clusters 1, 5, and 6 and the corresponding PM10 concentrations are lower than the spring season average, thus these clusters had less effect on PM10 loadings in Lanzhou. Cluster 7 was not a major transport pathway as no dust sources lie along the route. In high dust years, the pathways represented by

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

66 Air Pollution XIX clusters 2 and 3 were polluted pathways, while clusters 1, 5, 6 and 7 were relatively clean pathways. In low dust years, about 23.6% of trajectories were assigned to cluster 2, of which 44.0% were polluted trajectories. Cluster 2 represents the most important transport pathways with the highest PM10 concentration (177.1μg/m3) among the six clusters. The second important transport pathway was associated with cluster 5, which had the second high PM10 concentration (170.1μg/m3) and a relatively high percent of polluted trajectories (33.2%). The corresponding average PM10 concentrations of the two pathways, i.e. the western and the northern pathways, represented by clusters 2 and 5, exceeded the spring season average of 157.2 μg/m3. These results indicate that air masses associated with clusters 2 and 5 would carry more particulate matters to Lanzhou and lead to high PM10 loadings in low dust years. The third important pathway was associated with cluster 3. Although 47.8% of the trajectories in cluster 1 were polluted, the number of trajectories assigned to the cluster is small, therefore, the pathway represented by cluster 1 was considered less important for PM10 loadings in Lanzhou. The pathways represented by clusters 2, 5 and 3 were polluted pathways, while clusters 1, 6 and 7 were relatively clean pathways in low dust years. Table 2:

Trajectory number and averaged PM10 concentration of each cluster for high dust years 2001, 2002, 2004 and 2006. All trajectories

Polluted trajectories* PM10 PM10 Percent Percent concentration Number concentration Cluster Number of total of total (μg/m3) (μg/m3) 1 87 5.9 218.1±113.4 21 24.1 381.0±111.1 2 486 33.0 287.7±150.3 256 52.7 399.0±121.8 3 286 19.4 275.5±166.1 148 51.7 402.2±132.5 4 190 12.9 213.3±149.5 61 32.1 389.6±136.7 5 147 10.0 179.9±146.3 29 19.7 431.0±136.1 6 117 7.9 150.5±88.0 11 9.4 362.6±68.5 7 159 10.8 184.8±90.9 26 16.4 342.5±85.8 *Trajectories associated with those concentrations higher than the average (238.8 μg/m3).

Table 3:

Same as table 2 but for low dust years 2003, 2005, 2007 and 2008. All trajectories

Polluted trajectories* PM10 PM10 Percent Percent concentration Number concentration Cluster Number of total of total (μg/m3) (μg/m3) 1 161 10.9 171.6±90.2 77 47.8 235.0±91.9 2 348 23.6 177.1±112.5 153 44.0 255.8±129.5 3 255 17.3 148.9±88.6 78 30.6 247.3±98.4 5 265 18.0 170.1±133.2 88 33.2 312.6±142.4 6 162 11.1 126.3±88.3 33 20.4 259.6±111.8 7 281 19.1 137.3±62.3 63 22.4 224.8±72.5 *Trajectories associated with those concentrations higher than the average (157.2 μg/m3). WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

67

4.3 Potential source regions and their relative contribution The distributions of PSCF for high and low dust years are shown in fig. 4. In high dust years (fig.4 (a)), cells with high PSCF values appeared mainly in Xinjiang, Qinghai and Gansu provinces, that is, the potential source regions most likely to have effect on high PM10 concentration in springtime in Lanzhou were located in the Tarim Basin and the eastern Turpan Basin in Xinjiang province, the Qaidam Basin in Qinghai and the Hexi Corridor in Gansu. The air masses from these potential source regions traveled along the pathways represented by clusters 2 and 3 to Lanzhou. In low dust years (fig. 4 (b)), cells related to high PSCF values were located in the Tarim Basin in Xinjiang, the Qaidam Basin in Qinghai, the degradated grassland near the borders of Qinghai, Sichuan and Gansu provinces, and the desert and Gobi desert in central and western Inner Mongolia. The air masses from these potential source regions traveled along pathways represented by clusters 1, 2 and 5 to Lanzhou. The effect of potential source regions on PM10 loadings in Lanzhou is lower in low dust years than that in high dust years. Fig. 5 shows the distribution of weighted trajectory concentrations which gives the information on the relative contribution of potential source regions to PM10 loadings in Lanzhou. The potential source regions that are represented by

Figure 4:

Potential source contribution function map of PM10 in spring during (a) 2001, 2002, 2004 and 2006, (b) 2003, 2005, 2007 and 2008. Darker colors indicate greater potential source.

Figure 5:

Concentration-weighted trajectory method analysis map of Lanzhou in spring during (a) 2001, 2002, 2004 and 2006, (b) 2003, 2005, 2007 and 2008. Darker colors indicate greater influence.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

68 Air Pollution XIX CWT values higher than 250μg/m3 (fig. 5(a)) include the Tarim Basin and the Turpan Basin in Xinjiang, the deserts near the border of Xinjiang and Gansu, the Qaidam Basin in Qinghai, and the Hexi Corridor in Gansu. In low dust years (fig. 5(b)), the potential source regions are mainly located in the southern Taklimakan Desert and the Gobi Desert in eastern Xinjiang, the Gobi Desert in central and western Inner Mongolia and the Qaidam Basin in Qinghai. The contribution of these potential source regions to the PM10 loadings in Lanzhou was 150μg/m3 to 200μg/m3 in low dust years. The PSCF and CWT analyses give somewhat different results for Lanzhou. Compared to the PSCF results (fig. 4(a)), CWT results (fig. 5(a)) give more detailed information on source regions. For example, three more source regions can be seen in the CWT results, i.e. the Gobi Desert in Inner Mongolia, the Losses Plateau, and the western Sichuan and regions between Shaanxi and Gansu. The former two regions were not the major potential source regions as they had less affect on the PM10 loadings in Lanzhou. The third regions correspond to the “southerly sources” in Wang et al. [7]. The distribution of source regions in low dust years was similar to that in high dust years. Compared to the PSCF results (fig. 4(b)), the CWT results (fig. 5(b)) reveal more source regions, e.g. the Tengger Desert, the edge of the Gurbantunggut Desert, the northwestern Gansu and Qinghai, the Gobi Desert in Inner Mongolia and the borders of Sichuan, Gansu and Shaanxi.

5 Conclusion The atmospheric pathways, potential source regions and their relative contribution to the high PM10 loadings in Lanzhou were identified by trajectory clustering techniques, potential source contribution function (PSCF) model, and concentration-weighted trajectory (CWT) method. Results indicate that the western and the northwestern pathways, accounting respectively 33% and 19.4% of all trajectories, were major pathways leading to high springtime PM10 loadings for Lanzhou in high dust years. However, the most important and the second important pathways were the western and the northern pathways in low dust years. In high dust years, the potential source regions were mainly located in the Tarim Basin and the Turpan Basin in Xinjiang, the border regions between Qinghai and Gansu, and the Hexi Corridor. In low dust years, the potential source regions were the Tarim Basin and the Turpan Basin in Xinjiang and the border regions among Xinjiang, Qinghai and Gansu. There were also some moderate potential sources including the northerly source, the southerly source and the Loess Plateau source. The northerly source was mainly distributed in the desert and Gobi desert in Inner Mongolia, the southerly source was in the border regions of Sichuan, Gansu and Shaanxi, and the Loess Plateau source was in the central areas of Shaanxi province. Although the potential source regions are similar in the high and the low dust years, the transport pathways and the relative contribution to PM10 loadings in Lanzhou are different. It should be emphasized that the current study was conducted for spring. More research are needed to provide more detailed scientific basis on improving WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

69

the air quality in Lanzhou, especially the transport pathways and source regions for winter pollutants need to be identified by combining event analysis with high-resolution numerical simulations.

Acknowledgement This research was funded by the Chinese Academy of Sciences through the ‘100 Talent Project’.

References [1] [2]

[3] [4] [5] [6] [7] [8] [9] [10] [11]

Intergovernmental Panel on Climate Change (IPCC). In: Houghton, J.T., et al., (eds), Climate Change 2001, Cambridge University Press, New York, 2001. Kim J., Yoon S.C. & Jefferson A., et al., Aerosol hygroscopic properties during Asian dust, pollution, and biomass burning episodes at Gosan, Korea in April 2001. Atmospheric Environment, 40(8), pp. 1550-1560, 2006. Chen L.L., Carmichael G.R. & Hong M.S., et al., Influence of continental outflow events on the aerosol composition at Cheju island, south Korea. Journal of Geophysical Research, 102, pp. 28551-28574, 1997. Duce R.A., Unni C.K. & Ray B.J., et al., Long-range atmospheric transport of soil dust from Asia to the tropical North Pacific: temporal variability. Science, 209, pp. 1522-1524, 1980. Uematsu M., Duce R.A. & Prospero J.M., et al., Transport of mineral aerosol from Asia over the North Pacific Ocean. Journal of Geophysical Research, 88, pp. 5342-52, 1983. Husur R.B., Tratt D.M. & Schichtel B.A., et al., Asian dust events of April 1998. Journal of Geophysical Research, 106, pp. 18317-18730, 2001. Wang Y.Q., Zhang X.Y. & Arimoto R., The contribution from distant dust sources to the atmospheric particulate matter loading at XiAn, China during spring. Science of the Total Environment, 368, pp. 875-883, 2006. Yu Y., Xia D.S. & Chen L.H., et al., Analysis of particulate pollution characteristics and its causes in Lanzhou, Northwest China. Environmental Science, 31(1), pp. 22-28, 2010. Chu P.C., Chen Y.C. & Lu S.H., Atmospheric effects on winter SO2 pollution in Lanzhou, China. Atmospheric Research, 89, pp. 365–373, 2008. Wang S.G., Wang J.Y. & Zhou Z.J., et al., Regional characteristics of three kinds of dust storm events in China. Atmospheric Environment, 39(3), pp. 509-520, 2005. Wang S.G., Yuan W. & Shang K.Z., The impacts of different kinds of dust events on PM10 pollution in northern China. Atmospheric Environment, 40(40), pp. 7975-7982, 2006.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

70 Air Pollution XIX [12] [13] [14]

[15]

[16] [17] [18]

[19] [20]

[21] [22]

[23] [24] [25]

Zhang L., Chen C.H. & Murlis J., Study on winter air pollution control in Lanzhou, China. Water, Air, and Soil Pollution, 127, pp. 351–372, 2001. Chu P.C., Chen Y.C. & Lu S.H., Afforestation for reduction of NOX concentration in Lanzhou China. Environment International, 34, pp. 688– 697, 2008. Xu X. & Akhtar U.S., Identification of potential regional sources of atmospheric total gaseous mercury in Windsor, Ontario, Canada using hybrid receptor modeling. Atmospheric Chemistry and Physics, 10, pp. 7073-7083, 2010. Ci Z.J., Zhang X.S. & Wang Z.W., et al., Atmospheric gaseous elemental mercury (GEM) over a coastal/rural site downwind of East China: Temporal variation and long-range transport. Atmospheric Environment, 45, pp. 2480-2487, 2011. Efstathiou C., Isukapalli S., & Georgopoulos P., A mechanistic modeling system for estimating large-scale emissions and transport of pollen and co-allergens. Atmospheric Environment, 45, pp. 2260-2276, 2011. Escudero M., Stein A.F. & Draxler R.R., et al., Source apportionment for African dust outbreaks over the Western Mediterranean using the HYSPLIT model. Atmospheric Research, 99, pp. 518-527, 2011. Seibert P., Kromp-Kolb H. & Baltensperger U., et al., Trajectory analysis of aerosol measurements at high alpine sites. In: Borrell P.M., et al., (eds), Transport and transformation of pollutants in the troposphere. Academic Publishing, Den Haag, pp. 689-693, 1994. Ministry of Environmental Protection of the People’s Republic of China (MEP). Beijing, www.zhb.gov.cn/index htm. Zhang X.Y., Gong S.L. & Shen Z.X., et al., Characterization of soil dust aerosol in China and its transport/distribution during 2001 ACE-Asia, 1. Network observations. Journal of Geophysical Research, 108, pp. 4261, 2003. Draxler R.P. & Hess G.D. & An overview of the HYSPLIT-4 modeling system for trajectories, dispersion, and deposition. Australian Meteorological Magazine, 47, pp. 295-308, 1998. Sirois A. & Bottenheim J.W., Use of backward trajectories to interpret the 5-year record of PAN and O3 ambient air concentrations at Kejimkujik National Park, Nova Scotia. Journal of Geophysical Research, 100, pp. 2867-2881, 1995. Ashbaugh L.L., Malm W.C. & Sadeh W.Z., A residence time probability analysis of sulfur concentrations at Grand Canyon National Park. Atmospheric Environment, 19, pp. 1263-1270, 1985. Polissar A.V., Hopke P.K. & Harris J.M., Source regions for atmospheric aerosol measured at Barrow, Alaska. Environmental Science and Technology, 35, pp. 4214-4226, 2001. Hsu Y.K., Holsen T.M. & Hopke P.K., Comparison of hybrid receptor models to locate PCB sources in Chicago, Atmospheric Environment, 37, pp. 545-562, 2003.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

71

Performance evaluation of the ADMS-Urban model in predicting PM10 concentrations at the roadside in Chennai, India and Newcastle, UK S. Nagendra1, M. Khare2, P. Vijay1 & S. Gulia2 1

Department of Civil Engineering, Indian Institute of Technology Madras, Chennai, India 2 Department of Civil Engineering, Indian Institute of Technology Delhi, New Delhi, India

Abstract Recent economic development in many Asian and European countries has shown an increase in vehicle kilometre travel (VKT) in many cities, which has resulted in an increase in the vehicular pollution levels. In particular, particulate matter (PM) concentrations emitted from vehicles are at alarming levels in most of the cities of the world. Therefore, there is growing interest in the formulation of local level air quality management system to tackle vehicular emissions. Many mathematical models have been widely used as tools in urban air quality management. In the present paper, an attempt has been made to evaluate the performance of the ADMS Urban model in predicting roadside PM concentrations at two cities namely Chennai, India and Newcastle, UK during critical winter period of the year 2009. The statistical parameters such as Index of Agreement (IA), Fractional Bias (FB), Normalized Mean Square Error (NMSE), Geometric Mean Bias (MG) and Geometric Mean Variance (VG) have been used to evaluate the ADMS model performance. Results indicated that the roadside PM concentrations predicted by the ADMS model are reasonably accurate for Newcastle than at the roadside in Chennai. Keywords: air quality, particulate matter, model, management, statistical indicator, meteorology.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110081

72 Air Pollution XIX

1 Introduction Air pollution from motor vehicles is one of the most serious and rapidly growing environmental problems in the large cities of the developing world. In many cases, the monitored pollutant concentrations exceed the world health organization (WHO) air quality guidelines or national ambient air quality standards (NAAQS) set in their countries. The recent trends of air pollution in India and UK are showing a low level in most of the criteria pollutants except particulate matter (PM) and oxides of nitrogen concentrations in both the cities. In the United Kingdom, about 80% of the road emissions are generated from particulate matter of which road transport is the dominant share [1]. Vehicular exhaust derived pollution and its effect on human health is now becoming a matter of concern in many urban areas. According to recent epidemiological and toxicological studies, the high concentrations of airborne particles are associated with significant impacts on human health [2]. This holds especially for the fine and ultrafine particle size ranges due to their ability to penetrate deep into the human body. A study conducted by Srimuruganandam and Nagendra [3] showed that the 24-hour average PM10 and PM2.5 concentration are violating of NAAQS as well as world health organisation standards (WHO) during winter and monsoon season and minimus in summer season. Source apportionment studies of Chennai have also revealed the exceedances of particulate matter concentrations above NAAQS in residential areas [4]. Therefore, it becomes necessary for the authorities to assess the quality of air and implement control policies and strategies. In the recent past, many cities have been developed air quality management (AQM) strategies to achieve a specified set of ambient air quality standards (AAQS) or rules. Air quality models plays an important role in formulating air pollution control and management strategies by providing guidelines for better and more efficient air quality planning. Air Quality Models represent essential computational tools for predicting the air quality impacts of emissions from road traffic and also help in testing the accuracy of monitoring equipments once it is validated. In the present work, dispersion models namely ADMS-Urban has been used to simulate the air quality at selected locations in Chennai and Newcastle cities. Further, model sensitivity with respect to traffic and meteorological characteristics has also been studied.

2 Methodology 2.1 Study region Figure 1 shows the details of study region in Chennai city, India. The study area is located in the premises of Indian Institute of Technology Madras (IITM), Sardar Patel Road, Chennai city in India. Traffic flows on SP road is about 0.17 million vehicles per day. Braking is frequent on SP road due to the presence of two busy traffic intersections within a distance of 700 meters. The terrain of the study region is considered as plain and flat terrain (terrain height is 7.6 m from WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

73

N

Location of sampling site at IITM

Figure 1:

Description of study area in Chennai, India.

mean sea level). At this site, there is no local stationary source emission except emissions from road traffic. The PM monitoring instrument was kept at 1.2 m height from ground level and 7 m away from the centerline of SP Road. The project site in Newcastle city is located at 540 58’ 40 N and 10 36’ 49” (Figure 2) which is also one the busiest intersection of Newcastle upon Tyne. This intersection comes under air quality control regions (AQCRs). The whole intersection has been divided into three different roads. The monitoring site is situated near to road i.e. 20 m at city centre. The city centre monitoring station works under urban centre air quality station which is operated under Automatic Urban and Rural Network (AURN), the main air quality compliance network for DEFRA. The monitored data of both cities are used for validation of the model.

Figure 2:

Description of study area in Newcastle city, UK.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

74 Air Pollution XIX 2.2 Traffic characteristics The traffic volume count (TVC) has been monitored continuously for a week at a Sardar Patel road using automatic traffic flow recorder (Video). The traffic flow for Newcastle city has been obtained from SCOOT profile. The traffic flow at Chennai intersection is about 1,70,000 and that in the intersection at Newcastle city is about 25000. The morning peak flow occurs between 8am and 10am and afternoon peak occurs between 5pm and 7pm (Figure 3). The fleet composition in Chennai is dominated by two-wheelers (about 51%) followed by cars (34%)

Figure 3:

Figure 4:

Traffic flow on SP road in Chennai city.

Traffic flow Newcastle city centre road in Newcastle city.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

75

and three-wheelers (6%), respectively. Bus, lorry and van contribute a small percentage on SP road. The traffic pattern in Newcastle is somewhat uniform over weekdays and minimum in weekends as compared to weekdays. The diurnal traffic flow shows that morning peak flow occurs between 7 am and 11 am and evening peak occurs between 5pm and 8pm (Figure 4). The fleet composition of Newcastle is dominated by petrol car about 60% followed by diesel car (30%). HGV, LGV and buses contribute a small percentage. 2.3 PM Emissions PM emission has been estimated using the methodology suggested by Righi et al. [5]. ARAI [6] emission factors for Indian vehicles are used for Chennai site and DfT 2009 is used for Newcastle site [1]. The UK emission factors are speed depended unlike Indian vehicles emission factors. Traffic monitoring data for year 2004 was used to calculate emission rate for the year 2009 (extrapolation) on the basis of business as usual. 2.4 Meteorological data The main input meteorological parameters for ADMS model are wind speed and wind direction plus one of the following parameters, cloud cover, heat flux or reciprocal of Monin-Obukhnov length. The parameter solar radiation will be used only if NOx chemistry has to be used [7]. Sequential hourly meteorological data for the winter period (December 2008-February 2009) were obtained from Laga Systems, Hyderabad. The wind rose for the three months of both the cities is shown in Figures 5 and 6.

Figure 5:

Windrose for Chennai.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

76 Air Pollution XIX

Figure 6:

Windrose for Newcastle.

The windrose diagrams revealed that about 2.5% of winds are calm in nature (<1 m/s) and an average wind speed of 4.64 m/s for Chennai city and 4.82 m/s for Newcastle city. For Newcastle city, the predominant wind direction is towards west direction and for Chennai city, it is North-West. Table 1 summarizes the meteorological characteristics at both cities. Table 1:

Chennai Maximum Minimum Average Standard Deviation Newcastle Maximum Minimum Average Standard Deviation

Summary of meteorological characteristics at Chennai and Newcastle cities. Wind Speed (m/s)

Cloud Cover (tenths)

7.8 0.0 3.66 1.62

10.0 2.0 3.0 2.0

28.9 19.4 25.19 1.38

100.0 49.0 76.4 8.15

1016.0 1004.0 1009.79 1.86

14.0 0.0 4.82 2.37

10.0 2.0 4.0 2.0

10.3 -4.8 3.17 3.29

100.0 58.0 91.63 7.55

1030.0 960.0 1000.47 15.75

Temperature (degree Celsius)

Relative Humidity (percentage)

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Pressure (millibars)

Air Pollution XIX

77

2.5 PM10 sampling The PM is measured at SP road using a real-time GRIMM portable dust monitors with the model number 107 (Grimm Technologies, Inc.). The model 107 measures the particle mass concentrations in terms of PM10, PM2 5 and PM1 with a resolution of 1 µg/m3. The instruments use a light-scattering technology for single-particle counts. The flow rate of instrument is 1.2 L/min. The 47-μm polytetra-fluoro-ethylene (PTFE) filters with 0.2 µm size are used for collecting the dust samples. The measurement of PM mass concentrations has been made from October to December 2009. The 1 minute averaged values of both PM mass and numbers have been later converted into one hour average values. At Civic centre in Newcastle city, PM concentrations are monitored using high volume sampler installed at Civic Centre, Newcastle. The observed daily concentration values under changing traffic and environmental conditions have been used for evaluation of ADMS model performance. The model performance was carried out for critical winter period (December 2008-February 2009). 2.6 Description of ADMS Urban model ADMS – Urban (Atmospheric Dispersion Modeling System) has been developed by Cambridge Environment Research Consultants, United Kingdom. It is an advanced model for calculating concentrations of pollutants emitted both continuously from point, line, volume and area sources, and discretely from point sources. In ADMS-Urban road model sources are treated as line sources. Each road source is decomposed into a maximum of 10 source elements and the concentration of each element is approximated by a crosswind line source of finite length (CERC [7]). The model includes algorithms which take account of the following: effects of main site building, complex terrain, wet deposition, gravitational settling and dry deposition, short term fluctuations in concentration, chemical reactions, radioactive decay and dose plume rise as a function of distance, jets and directional releases, averaging time ranging from very short to annual; condensed plume visibility meteorological preprocessor. 2.7 Statistics for model performance evaluation The statistical descriptors Index of Agreement (IA), Fractional Bias (FB), Normalized Root Mean Square Error (NRMSE), Geometric Mean Bias (MG) and Geometric Mean Variance (VG) have been used to evaluate the performance of ADMS model. IA

1

∑ ∑

|

| |

|

      

FB   2 Pi’‐Oi’ /  Pi’ Oi’

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

(1) (2)

78 Air Pollution XIX

NMSE =    

  ∑

 

(3)

MG= exp ln  –  ln

(4)

VG= exp ln

(5)

ln

In the above equations, Oi is the observed value; Pi is the predicted value by ADMS and is the average of observed values. IA indicates how much the predicted value departs from observed values. It has a theoretical value between 0 and 1, latter indicates perfect agreement. NRMSE is an estimator of the overall deviations between the observed and predicted values. Smaller values of NRMSE indicate better performance and are not biased towards model that over predicts or under predict. FB, a dimensionless number, represents the relative difference between observed and modeled values in a bounded range [7]. The value of FB lies between -2 and +2. MG and VG are measures of dispersion which finds application when values in a set follow a log normal distribution. A perfect model will have both MG and VG equal to 1.0 [8]. However, a model will be deemed acceptable if: NRMSE≤0.5; -0.5≤FB≤0.5; 0.75≤MG≤1.25; 1≤VG≤1.25 [9].

3 Results and discussions 3.1 PM10 mass concentration trend The monitored PM10 concentrations for the study period (December 2008 to February 2009) at Chennai and Newcastle City Centre road are shown in the Figure 7. The 24-hr average pollutant levels for the three months are 130.39 µg/m3 for Chennai city and 16.38 µg/m3 for Newcastle. High PM concentrations are found during calm conditions prevail. Similarly the concentrations are found to be high when the wind is towards the receptor location. The PM10 mass concentration at the SP road is found to be higher than the specified NAAQS limit for the residential area (100 µg/m3). In Figure 7, a clear diurnal and weekly cycles were observed at Chennai study site. In diurnal cycles two peaks were observed corresponding to morning and evening peak traffic flow. The maximum PM concentration was observed between 8AM to 10AM in the morning and in the evening between 5PM to 10PM. The low PM values observed during night time between 11PM to 6AM beacause of lean traffic flow. During afternoon between 12 noon to 3PM the PM concentation is found to be in the lower range.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Figure 7:

79

Variation in monitored PM10 concentrations at Chennai and Newcastle city.

3.2 Performance evaluation of ADMS model Table 2 presents the statistics of ADMS model performance during critical winter period. It is observed that ADMS model predictions are close to observed PM concentrations at Newcastle city. The negative MG values indicate that ADMS model is under predicting the PM concentration at both the site. The IA values of 0.39 and 0.48 indicate that 39% and 48% of the model predictions are error free at Chennai and Newcastle sites, respectively. This is evident from the predicted average and the maximum PM values which are significantly below the observed values. NRMSE values for both sites shows less correlation between observed and predicted values. Table 2:

Summary of ADMS model performance at Chennai and Newcastle city.

Study site

PM10

IA

FB

MG

VG

NRMSE

Chennai

Predicted Observed Predicted Observed

0.39 1.00 0.48 1.00

-1.03 0.00 -1.29 0.00

1 36 1.00 1.27 1.00

1.04 1.00 1.14 1.00

0.19 0.00 0.09 0.00

Newcastle

Average (μg/m3) 41.33 130.39 4.44 16.38

Maximum (μg/m3) 120.11 343.16 28 10 59.00

4 Conclusion Mathematical models have been widely used as a tool in urban air quality management in developed countries. The ADMS model is being widely used in Europe for air quality assessment. However, its application is limited in developing countries such as India due to lack of readily available input data to the test model and time and cost involved in collecting the required data. In the present paper, the performance of Gaussian based dispersion model namely ADMS-Urban in predicting vehicular air pollutant (PM) concentration at road sides in Chennai city and Newcastle city has been evaluated. The model WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

80 Air Pollution XIX performance is evaluated using statistical parameters such as Index of Agreement (IA), Fractional Bias (FB), Normalized Root Mean Square Error (NRMSE), Geometric Mean Bias (MG) and Geometric Mean Variance (VG). Results indicated that ADMS model is able to predict the PM concentrations with reasonable accuracy. The IA for ADMS is found to be 0.39 for Chennai city and 0.48 for Newcastle city.

Acknowledgements The research work is a part of the ongoing UKIERI funded research project titled “Evaluation of Quantitative Dispersion Models for Urban Air Quality Assessment”. We wish to thank the UKIERI, New Delhi and HSBC, Mumbai for funding this research study.

References [1] DfT. Department for Transport, 2010. http://www.dft.gov.uk. [2] DEFRA. Department of Environment, Food and Rural Affairs, 2010. http://www.defra.gov.uk. [3] Srimuruganandam, B. and Nagendra S.M.S. (2011). Characteristics of particulate matter and heterogeneous traffic in the urban area of India. Atmospheric Environment 45 (2011) 3091e3102. [4] CPCB (2010): Status of vehicular pollution control program in India, Program Objective Series/PROBES/136/2010. [5] Righi, S., Lucialli, P., Pollini, E., (2009). Statistical and diagnostic evaluation of the ADMS-Urban model compared with an urban air quality. Atmospheric Environment 43 (2009): 3850-3857. [6] ARAI, (2007). Emission Factor Development for Indian Vehicles. Project Report No.AEF/2006-07/IOCL/Emission Factor Project. Automotive Research Association of India, Pune (Available on www.cpcb nic.in). [7] CERC, (2006), Cambridge Environmental Research Consultants, ADMS – Urban user guide, http://www.cerc.co.uk/environmental-software/ADMSUrban-model html. [8] Mohan, M., Bhati, S., Marrapu, P., 2009. Performance evaluation of AERMOD and ADMS-Urban models in a tropical urban environment. Indian Journal of Air Pollution Control, Vol. IX No. 1, 47-62. [9] Kumar, A., Luo, J., and Bennet, F., G., 1993, Statistical evaluation of Lower Flammability Distance (LFD) using four hazardous release models.Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/prs.680120103.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

81

Coastal influences on pollution transport D. Peake, H. Dacre & J. Methven Department of Meteorology, Reading University, UK

Abstract The ventilation of pollutants from the boundary layer into the free troposphere is an important process in controlling regional air quality. Coastal outflow is the horizontal ventilation of pollutants across a coastline from a layer within the continental boundary layer to above the marine boundary layer. It has been shown using the Met Office Unified Model that the ventilation by coastal outflow occurs with a similar order of magnitude to ventilation by convection, and that it possesses a diurnal cycle induced by the boundary layer height cycle over land. Pollutants with short lifetimes (typically several hours) exhibit the greatest diurnal variability in export by coastal outflow. Ventilation by coastal outflow by pollutants with longer lifetimes are less dependent on the boundary layer height over land and more dependent on the large scale cross-coastal wind strength. A simple model developed to simulate coastal outflow shows that increasing the pollutant lifetime, wind speed or convective boundary layer height increases export of tracer by coastal outflow. Above a threshold windspeed, coastal outflow is reduced due to tracer being exported across a coastline before being able to be mixed to a height greater than the marine boundary layer. Convection slightly decreases export by coastal outflow, although this effect is small. Keywords: coastal outflow, pollution, modelling, coasts, tracer, ventilation, export boundary layer.

1 Introduction The vertical ventilation of pollutants from the boundary layer to the free troposphere through convection and frontal mechanisms has a significant impact on regional air quality [1, 2] and has been extensively studied [3–5]. The prospect of horizontal ventilation across a discontinuity in boundary layer height, such as one present at a coastline, has received relatively little attention [6]. During the WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110081

82 Air Pollution XIX

Figure 1: Schematic of coastal outflow. A logarithmic wind profile is shown on the left. Large circular arrows represent boundary layer scale turbulent eddies, grey shading represents pollution. MBL represents the height of the marine boundary layer away from the coast, MAXCBL represents the maximum convective boundary layer height. Deposition in the marine boundary layer is assumed.

daytime the convective boundary layer (CBL) over the land typically grows to a greater depth than the marine boundary layer (MBL). Thus pollutants advected horizontally across a coastline within the CBL can be transported above the MBL through coastal outflow (Figure 1). This layer of air above the MBL can become decoupled from the surface, and is subject to higher wind speeds and negligible deposition, enabling more efficient long-range transport than within the MBL. Verma [7] and Raman [8] show that pollutants are frequently observed in layers above the top of the MBL. This study uses the Met Office Unified Model with passive tracers to quantify coastal outflow and investigate the meteorological processes that control it.

2 Experiment setup The Met Office Unified Model (UM), version 6.1, has been used in this study. It is a non-hydrostatic fully-compressible model which runs the Navier-Stokes equations forward in time using a semi-implicit and semi-lagrangian numerical scheme across staggered grids in both the horizontal (Arakawa-C grid) and vertical (Charney-Phillips grid). The UM model is documented by [9] and runs the new dynamics regime [10]. The UM is run in Limited Area Mode (LAM) over the east coast of the US, chosen to include a large part of the International Consortium for Atmospheric Research on Transport and Transformation (ICARTT) field campaign domain for observational verification, and is shown in Figure 2. The LAM simulation starts at 0000Z on 13/07/04 and is run for 27 days with a horizontal resolution of 0.11◦ (12km) and has 38 terrain-following model levels in the vertical, of which 10 levels are situated in the lowest 2km above ground level. The model run is initialised WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

83

Figure 2: Map of the limited area domain used in the UM in this study, with latitudes and longitudes included for reference.

using a global analysis from the European Center for Medium-Range Weather Forecasts’ (ECMWF) archive, and its boundaries are forced every six hours with data from ECMWF global re-analyses. Pollution transport is modelled using four passive tracers, two of which have e-folding lifetimes of 3 and 24 hours. The tracers are passive in that they undergo no chemical transformation other than an exponential decay of the specified timescale. The tracers are initialised and emitted (with constant source rate) in the lowest model level over land only. For each lifetime, one tracer is transported by the turbulent mixing parameterisation and advection schemes, the other by turbulent mixing, advection and convection schemes. Although the advective and convective processes are not additive, this will allow the relative importance of convection upon coastal outflow to be ascertained. Any external air mass advected into the domain is assumed to have a tracer mixing ratio of 0kg/kg. The model domain is then split into six regions (boxes) for analysis, the ’boundary’, ’residual’ and ’free tropospheric’ layers over the land and sea respectively, shown in Figure 2. Tracer that has been exported from the continent by coastal outflow is located in box M4 , which extends from the top of the marine boundary layer to the top of the MAXCBL, the maximum height the convective boundary layer reaches during its diurnal cycle. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

84 Air Pollution XIX

Figure 3: Splitting the domain into six regions: boundary (M1 , M2 ), residual (M3 , M4 ) and free tropospheric layers (M5 , M6 ) over land and sea respectively. HCBL and HMBL represent the heights of the convective and marine boundary layers. HMAXCBL represents the maximum height the convective boundary layer reaches through its diurnal cycle. HT OA represents the height at model top. Each box contains a mass of tracer Mx , boxes 1, 3 and 5 are placed over land, boxes 2, 4 and 6 over the sea, with the coast in between. Tracer undergoing coastal outflow is horizontally advected to box M4 .

3 Results 3.1 Diurnal cycle Figure 4 shows a composite of the diurnal cycle of tracer contained within box M4 , i.e. exported through coastal outflow, in the UM model run for the 3- and 24-hour tracers passed through the advection, convection and turbulent mixing schemes. 14 days are averaged together to generate an average diurnal cycle (solid line) and day-to-day variability (grey shading) of coastal outflow. Approximately 5% of the total domain 3-hour tracer (Figure 4a) is exported in box M4 by coastal outflow, with a maximum occuring at 5pm local time (LT) and a minimum at 9am LT. Due to the short lifetime of the tracer, coastal outflow can only occur when the depth of the CBL is greater than that of the MBL, and so the maximum occurs in the late afternoon when the CBL has grown deep enough to enable the export of pollutants at a high enough altitude to be present above the MBL. Consequently the minimum occurs in the late morning when tracer emitted at the surface is trapped by a shallow CBL for the nighttime, leading to little coastal outflow due to tracer decaying q uickly in the residual layer. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

85

Due to its longer lifetime, a greater proportion (12%) of the 24-hour lifetime tracer is exported through coastal outflow (Figure 4b) compared to the 3-hour tracer. Due to its slower decay, coastal outflow can continue to occur during the night as a greater concentration of tracer is present in the residual layer over land. As such, the 24-hour tracer exhibits little diurnal variability and undergoes continuous export by coastal outflow. The percentages of 3- and 24-hour tracer that undergo coastal outflow are of the same order of magnitude as that which is transported above the MAXCBL by resolved ascent and/or convection in the model, with 6% and 15% of the total domain tracer mass respectively located in box M5 . The differences in the export percentage when tracer is included in the convection scheme are negligible (not shown). Convection increases boundary layer ventilation, by decreasing the mass of tracer within boxes M1 and M3 and transporting it to free-tropospheric box M5 . This leads to a reduction in coastal outflow. For the 3- and 24-hour tracers reductions of 0.1% and 0.5% occured, which are small when compared to their magnitudes (5% and 12%). Convection thus has little effect on coastal outflow.

3.2 Day-to-day variability Figure 5 shows the percentage of 3- and 24-hour tracer that is exported by coastal outflow across the whole domain as a time series for the first 17 days of the model run (entire model run time series not shown for clarity reasons). Note that for all timeseries plots, the first two days should be regarded as spin up, when high concentrations of tracer were initialised in the lowest model level over land leading to an unrealistic picture of tracer distribution during this time. The 3hour tracer export exhibits a clear diurnal cycle, whereas the 24-hour tracer does not, consistent with Figure 4. The 24-hour tracer exhibits a strong day-to-day variability that can lead to a near doubling of tracer export at some times (17% on 16th, 17th, 21st July) compared to others (9% on 25th, 26th). The 3-hour tracer possesses less day-to-day variability due to its shorter lifetime. The day-to-day variability could be controlled by the variation in day-to-day maximum convective boundary layer height or by horizontal cross-coastal wind speed. These factors are investigated in Figure 6. Figure 6a shows the domain average zonal wind speed at 850m above ground level, a typical height at which coastal outflow occurs at the coast. When a 24 hour lag is taken into account, there is a strong correlation (r = 0.88) between the 850m wind field and proportion of 24-hour tracer exported by coastal outflow, accounting for 77% of variability in export by coastal outflow. The 90th percentile (chosen to remove outliers) of convective boundary layer height has a maximum that varies by less than 25% from day-to-day (Figure 6b) and accounts for 8% (r = 0.28) the day-to-day variability in modelled coastal outflow. Thus the dominant mechanism for controlling coastal outflow is the large-scale wind speed. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

86 Air Pollution XIX

(a)

(b) Figure 4: Calculated from the UM, panels a and b show the proportion of total domain tracer mass in box M4 (exported by coastal outflow), for the 3and 24-hour life time tracers respectively. The grey shading represents one standard deviation in coastal outflow per hour of day. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

87

Figure 5: Percentage of total domain 3-hour (dashed line) and 24-hour (solid line) tracer exported by coastal outflow from 13/07/04 to 29/07/04. Grey shaded areas represent 6pm–6am. Model spin up occurs during the first two days, hence exported tracer initially starting at zero.

3.3 Simulating variability A 2D ‘box model’ has been constructed to base coastal outflow on as few variables as possible and simulate its observed diurnal cycle and variability. The model tests the sensitivity of coastal outflow to 3 variables: wind speed, maximum boundary layer height and pollutant lifetime. The model consists of six boxes, in which pollutants are assumed to be well mixed, of the same structure as illustrated in Figure 2. The convective boundary layer height is assumed to be sinusoidal with time, the maximum MAXCBL occuring at noon and minimum of 50m at midnight. A wind speed, U ms−1 , is assumed to advect directly from land to sea, the width of the land boxes is prescribed to be L meters, and width of the sea boxes infinitely long to ensure no tracer exits the domain. The proportion of tracer advected from land to sea is U/Ls−1 . Passive tracers with specified lifetimes ranging from minutes to months are emitted in box M1 . There is no resolved ascent, convection or entrainment within this model, therefore no tracer enters boxes M5 or M6 . The model replicate the UM simulated diurnal cycle in coastal outflow (not shown). Figure 7 shows the average percentage of tracer exported each WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

88 Air Pollution XIX

(a)

(b) Figure 6: UM simulated time series of (a) domain average zonal wind speed and (b) 90th percentile (chosen to remove outliers) of domain average boundary layer heights over the land (solid line) and sea (dashed line). Grey shaded areas represent 6pm–6am. Model spin up occurs during the first two days. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

89

Figure 7: For c = M AXCBL/M BL = 2, the dependence of average daily proportion of tracer exported by coastal outflow on tracer lifetime a and wind speed to landwidth ratio U/L. Short lived tracers are represented by ab < 0.1 (dotted line), and long lived tracers are represented by ab > 10 (dash dotted line)

day by coastal outflow and how it varies with tracer lifetime, a, and wind speed to landwidth ratio, b = U/L, for the case of maximum convective boundary layer height equal to twice the marine boundary layer height (c = M AXCBL/M BL = 2). Figure 6 shows that as tracer lifetime is increased, that the proportion of tracer exported by coastal outflow also increases. This occurs as it can reach the coastline from further inland. As the windspeed is increased, the proportion of exported tracer increases until the timescale of advection across the coast is shorter than the diurnal timescale of boundary layer height. Tracer is advected across the coast before it is able to be mixed above the top of the marine boundary layer, leading to an increase in tracer in the MBL and a decrease in the coastal outflow layer. Increasing windspeed leads to a decrease in coastal outflow for tracers that are long lived with respect to the wind speed ab > 10, i.e. when the windspeed U exceeds 10L/a, equivalent to a 1.16ms−1 for L = 10km and a = 86400s. Increasing the boundary layer height ratio c also leads to increasing coastal outflow at all lifetimes and windspeeds, but as already noted this is not the meteorologically dominant factor in controlling coastal outflow. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

90 Air Pollution XIX

4 Conclusions Modelling over the eastern US coast using passive tracers within the UM has shown that the magnitude of ventilation of pollutants by coastal outflow, the export of pollutants from a layer within the continental boundary layer to a layer above the marine boundary layer, occurs on a similar order to ventilation by convection. Due to the diurnal boundary layer height cycle over land, a diurnal cycle is induced in export by coastal outflow. For tracers with modelled 3-hour and 24-hour half lives, 5% and 12% of total tracer emitted over land is exported by coastal outflow. The 3-hour tracer exhibits strong diurnal variability with a maximum in the late afternoon (when the greatest amount of tracer has been mixed above the marine boundary layer top) and a minimum in the early morning (when pollutants have remained trapped near the surface by the shallow nocturnal layer). The 24-hour tracer exhibits little diurnal variation but instead is more variable on a day-to-day timescale, and is strongly correlated with the 850m zonal wind speed (850m being a typical height at which coastal outflow occurs). Based on a simple 2D model, the sensitivity of coastal outflow to wind speed, maximum convective boundary layer and pollutant lifetime has been tested. Increasing any of these variables increases coastal outflow, although it has been shown that the maximum convective boundary layer height varies little from dayto-day and does not explain the variability in coastal outflow for a tracer 24hour lifetime during our simulation. Cross-coastal wind-speed is the dominant mechanism controlling the export of pollutants by coastal outflow. There is a threshold U = 10L/a, where U is the cross-coastal wind speed, L is the width of land and a is the pollutant half life, above which increasing the wind speed leads to a decrease in coastal outflow due to tracer being advected across the coast before having the chance to be mixed to a height above the marine boundary layer. Convection, while active in the UM simulation, has negligible effect on export by coastal outflow.

References [1] Clappier, A., Martilli, A., Grossi, P., Thunis, P., Pasi, F., Krueger, B.C., Calpini, B., Graziani, G. & van den Bergh, H., Effect of sea breeze on air pollution in the greater Athens area. part i: Numerical simulations and field observations. J Appl Meteorol, 39, pp. 546–562, 2000. [2] Liang, Q., L., J., Hudman, R.C., Turquety, D.J., S. Jacob, Avery, M.A., Browell, E.V., Sachse, G.W., Blake, D.R., Brune, W., Ren, X., Cohen, R.C., Dibb, J.E., Fried, A., Fuelberg, H., Porter, M., Heikes, B.G., Huey, G., Singh, H.B. & Wennberg, P.O., Summertime influence of asian pollution in the free troposphere over north America. J Geophys Res, 112(D12S11), 2007. [3] Agusti-Panareda, A., Gray, S.L. & Methven, J., Numerical modeling study of boundary layer ventilation by a cold front over Europe. J Geophys Res, 110(D18304), 2005. [4] Donnell, E.A., Fish, D.J., Dicks, E.M. & Thorpe, A.J., Mechanisms WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

[5] [6]

[7]

[8]

[9] [10]

91

for pollutant transport between the boundary layer and free troposphere. J Geophys Res, 106(D8), 2001. Sinclair, V.A., Gray, S.L. & Belcher, S.E., Boundary-layer ventilation by baroclinic life cycles. Q J R Meteorol Soc, 134, pp. 1409–1424, 2008. Dacre, H.F., Gray, S.L. & Belcher, S.E., A case study of boundary layer ventilation by convection and coastal processes. J Geophys Res, 112(D17106), 2007. Verma, S., Boucher, O., Venkataraman, C., Reddy, M.S., M¨uller, D., Chazette, P. & Crouzille, B., Aerosol lofting from sea breeze during the Indian ocean experiment. J Geophys Res, 111(D07208), 2006. Raman, S., Niyogi, D.d.S., Simpson, M. & Pelon, J., Dynamics of the elevated land plume over the arabian sea and the northern Indian ocean during northeasterly monsoons and during the Indian ocean experiment (indoex). Geophys Res Lett, 29(16), 2002. Cullen, M.J.P., The unified forecast/climate model. Meteor Mag, 122, pp. 81– 94, 1993. Lock, A.P., Brown, A.R., Bush, M.R., Martin, G.M. & Smith, R.N.B., A new boundary layer mixing scheme. part i: Scheme description and single-column model tests. Mon Wea Rev, 128, pp. 3187–3199, 2000.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

This page intentionally left blank

Air Pollution XIX

93

Non-parametric nature of ground-level ozone and its dependence on nitrogen oxides (NOx): a view point of vehicular emissions S. Munir, H. Chen & K. Ropkins Institute for Transport Studies, University of Leeds, UK

Abstract Ground-level ozone has been studied extensively using classic parametric statistics (most commonly conventional linear regression). Very few researchers have considered ozone distributions and even those that do tend to apply parametric techniques. This study assesses ground-level ozone distributions at six locations in the UK and characterises the correlation of nitrogen oxides (NOx) and ozone at a roadside location. The distribution of ozone is investigated, applying Shapiro-Wilk test and graphical presentations. The histograms are right skewed and show maximum frequency at ozone mixing ratios from 0 to 5 ppb (particularly at urban centers and roadsides locations), which is probably caused by high levels of freshly produced NOx associated with road traffic. There is evidence that ground level ozone is not normally distributed (p-values < 0.05). NOx is a dominant sink for ozone at urban and roadside sites due to its ozone scavenging effects. Consistent with literature ozone is negatively correlated with NOx. The negative correlation is stronger at low NOx levels (up to approximately 80 ppb 24 hour mean, Spearman correlation coefficient R is ‘-0.72’) and becomes weaker as NOx levels increase (over 80 ppb R value is ‘-0.53’). When NOx mixing ratios reach approximately 200 ppb or over the correlations become positive. This study investigates how the associations of ozone and NOx vary at different levels of their mixing ratios and suggests that due to the non-normal distribution of ozone, nonparametric statistics should be applied for ozone modelling. Keywords: ground-level ozone, air pollution, nitrogen oxides, nitrogen oxides and ozone, ozone distribution, vehicular emissions and ozone.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110091

94 Air Pollution XIX

1 Introduction Traditionally road traffic has been mainly linked with particulates (e.g. PM10 and PM2 5), carbon monoxides (CO), and nitrogen oxides (NOx), whereas ozone has very rarely been studied in connection with traffic. This is because ozone is not directly emitted by traffic or any other combustion processes, rather, it is a secondary pollutant generated by atmospheric chemistry. Ozone is linked with traffic closely and ozone levels are affected by traffic in two main ways: (i) traffic is the main source of ozone precursors i.e. NOx and volatile organic compounds (VOCs) are emitted by traffic that react in sun light and produce ozone; (ii) freshly produced NO react with ozone and destroy it (NO + O3 NO2 + O2) and that is the main reason that ozone concentrations are generally lower at roadsides and urban areas than in the surrounding rural areas [1]. Ozone is important at local levels, as well as at regional and intercontinental levels. At a given location ozone concentration is the sum of ozone produced by photochemical reactions, ozone brought in by regional transport and ozone descended from the stratosphere; minus ozone destroyed by NOx reactions and dry deposition [2]. Due to these sources and sinks ozone has a typical diurnal and seasonal cycle in the UK. The worrying factor regarding tropospheric ozone pollution is that in spite of the decreasing trends in its precursors, background ozone concentrations have been increasing in the UK, particularly in urban areas [3]. The Air Quality Expert Group (AQEG [3]) has expressed their concerns that ozone levels in urban areas are increasing at faster rates by comparison with surrounding rural areas, which in future may result in urban ozone levels as high as in the surrounding rural areas. If that happens it will increase ozone related health and environmental risks in these highly populated areas. Therefore it is vital to understand uncertainties in ozone predictions and quantify accurately the relationship of ozone with its sources and sinks. Tropospheric ozone has been studied extensively throughout the World using classic parametric statistics (most commonly ordinary least square regression). Very few researchers have considered ozone distributions and even those that do tend to apply parametric techniques. The majority of classical statistical tests are based on the assumption that the data to which the tests are applied exhibit a normal distribution (i.e. bell shape, symmetrical and with a common mean and median). If the parametric tests are applied to non-normal data, they can result in biased or even erroneous results [4]. Therefore, before applying a classical test ,it is vital to check data distributions and if the data are non-normally distributed, robust and non-parametric methods should be applied that are not based on such assumptions. This study intends to undertake a statistical analysis based investigation into ground-level ozone to determine whether ozone data are normally distributed or not. Ozone data distribution is compared spatially (6 monitoring sites) as well as temporally (different months and years). Variations in ozone concentrations are explained in terms of its correlation with NOx.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

95

2 Methodology The study is based on the statistical analysis of ozone and NOx data measured at several air quality monitoring sites in the UK. The sites include a roadside (Kirkstall Road Leeds), a kerbside (Marylebone Road London), 2 urban centres (Nottingham and Leeds centre), 1rural (Harwell) and 1 remote (Strath Vaich) air quality monitoring sites. All the sites, except Kirkstall are part of the UK Automatic Urban and Rural Network (AURN). The Kirkstall site is part of facilities available at Institute for Transport Studies (ITS) University of Leeds for the monitoring of air pollution, traffic and meteorological variables. The Kirkstall site [53°48'31.38"N and 1°35'21.40"W] is located on Kirkstall Road (A65), Leeds. Kirkstall Road runs North-West to South-East through the city of Leeds. At all these sites ozone is measured by ultraviolet absorption analyser and NOx by Chemiluminescent analyser, which are the standard methods for measurement of ozone and NOx in the Europe and UK. The details of AURN sites can be found at reference [5]. Ozone data distributions have been studied using simple graphical methods and statistics tests. The graphical methods used in this study include histograms, scatter diagrams and time variation plots. In addition to graphical presentation, the Shapiro-Wilk test has also been applied to estimate ozone normality. Spearman Rank correlation, which is a non-parametric or distribution free approach, has been applied to estimate the degree of co-variance between ozone and NOx. The statistical language R and the associated ‘openair package’ have been used for performing statistical analysis and making diagrams; see [6] for more details of these software.

3 Results and discussions 3.1 Ozone data distribution Firstly ozone data from the Kirkstall site have been analysed in details and then compared with data from AURN sites. The Kirkstall data analysed are for a 2 years periods (Nov. 2007 to Oct. 2009). Figure 1 shows a histogram of hourly ozone mixing ratios collected at Kirkstall site and shows that ozone data are not normally distributed (p-value for Shapiro-Wilk test is less than 0.01). The histogram shows very high frequency (nearly 2500) at ozone levels 0 to 5 ppb (first column). The frequency when ozone mixing ratio is 40 or over is relatively low. The histogram is skewed towards right. The first bar of the histogram needs investigations to prove that it comes from genuine measurements and is not due to an error or artefact. 3.1.1 Ozone distribution when ozone is less than 5ppb This section explains whether the first bar of the histogram (in Figure 1) is a result of genuine ozone measurements or not. Ozone (ppb) hourly average data from Kirkstall site had 16194 data points (excluding missing values). Out of the total 2462 data points have ozone mixing ratios less than 5ppb and 20 data points WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

96 Air Pollution XIX have ozone concentrations even less than 1 ppb. Firstly, the lower detection limit of the monitor (photometric ozone analyser, model 400E) is checked, which is < 0.6 ppb and hence it is low enough to give an accurate ozone measurements for any mixing ratios higher than 0.6 ppb. Secondly whether these data points are distributed over all 12 months or condensed only in 1 or 2 months. Thirdly to find out if these 2462 data points lie where ozone mixing ratios are expected to be low (e.g. winter months, night hours) or not. If these data points mostly lie in winter months (day or night) or in summer night time then we can say that they are genuine, otherwise they will be considered due to an artefact and discarded.

Figure 1:

Histogram of mean hourly ozone ppb from Kirkstall site Nov. 2007 to Oct. 2009 indicates that the data are not normally distributed (pvalue < 0.010 for Shapiro-Wilk test).

Figure 2 (bottom right) shows that the 2462 data points when ozone < 5 ppb is distributed in all 12 months regardless of winter or summer season. However the majority of these hours come from winter months (Nov. to Feb.). Data from summer months (not shown here) indicate that ozone mixing ratios less than 5 ppb mostly occur during night time hours. Figure 2 (bottom left) shows high frequency over night hours (including early morning and evening) and low during day time (especially 10:00 to 16:00). Figure 2 (top right) shows similar frequencies for most days except Sunday. Ozone levels in the UK are generally higher during the summer and lower during the winter, a trend demonstrated for the Kirkstall dataset later in section 3.2.2. Likewise ozone levels are generally observed to be lower at night and higher in the daytime. The reason for low levels of ozone during the winter and at night is most probably the lack (or reduced level) of solar radiation and lower temperatures which are responsible for reduce photochemical ozone production rates. In addition, dry deposition of ozone during the night can further reduce WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

97

ozone levels. Ozone levels are also linked to traffic activity by the NOx scavenging effect, which are generally low on Sunday, therefore Figure 2 (top right) shows low frequency of low ozone on Sunday as compared to other days. The above explanations clearly indicate that the 2462 data points are genuine measurements. This will become clearer in section 3.1.3, where the distributions of ozone data from different monitoring stations at which different instruments are used are compared.

Figure 2:

Histograms showing the frequency of months, hours and weekdays when ozone (O3) < 5 ppb.

3.1.2 Ozone distribution for different months Shapiro-Wilk test of normality gives p-values less than 0.01 for each month of the year 2008, which reveals that hourly ozone data do not follow a normal distribution in any month of the year at Kirkstall site. Histograms of ozone for each month January to December 2008 are shown in Figure 3, showing how ozone distributions vary in different months. There are 2 main categories of histograms in Figure 3 winter and summer months. In winter months (January, February, September, October, November, December) the highest frequency of ozone is found when the mixing ratios of ozone are 0 to 5 ppb, whereas in the rest of the months (March to Aug) the highest frequency can be observed when ozone mixing ratios are approximately 30 ppb. The winter months can be again subdivided into 2 groups i.e. September, October, November when the distribution is somewhat bimodal (highest frequency of ozone can be observed at 0 to 5 ppb and also at around 30 ppb) and

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

98 Air Pollution XIX January, February, December where the highest frequency of ozone is mainly at 0 to 5 ppb. Although the ozone distribution during the summer months seems closer to normal distribution, statistically it still appears to be non-normal (p values < 0.05). The reason for the difference in ozone distributions during summer and winter months is most probably the difference in the amount of solar radiation and temperature. Solar radiation and temperature are the 2 main meteorological factors responsible for photochemical ozone production and that is why the amount of ozone is mostly lower in winter months. In contrast, during the summer months photochemical ozone production is high which results in high ozone mixing ratios as shown in Figure 3.

Figure 3:

The distribution of ozone during different months of the year, hourly mean ozone data from Kirkstall site.

3.1.3 Ozone distribution at different sites Ozone distributions vary spatially from place to place in the UK depending on the nature of the monitoring site. Roadside and urban centre monitoring sites are generally characterized by high levels of fresh NOx that react with ozone and keep ozone levels low at these sites. In contrast, rural and remote monitoring sites have generally low levels of NOx and high levels of ozone.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

99

Figure 4 shows ozone distributions at 6 different monitoring sites. Ozone distributions at kerbside and roadside monitoring sites (Marylebone and Kirkstall) show high frequencies for low ozone mixing ratios i.e. the first column of the histogram is taller than the rest of the columns, which shows that low ozone mixing ratios occur more frequently than higher ozone mixing ratios. Particularly at the Marylebone site the frequency of higher ozone mixing ratios is very low; the reason probably is that Marylebone site is situated in London at the kerb of very busy road where road traffic exhausts produce a huge amount of fresh NOx [7]. Although the Kirkstall site is a roadside monitoring site, the traffic levels on this road in Leeds is not as high as at Marylebone Road and hence the difference in ozone mixing ratio is clear. In contrast, the Harwell and Strath Vaich monitoring sites have totally different ozone distributions; the histograms almost look like a bell shaped symmetric diagram (statistically it is still non-normal, as p-values < 0.05). The higher frequencies of ozone at rural and remote sites occur at about 30 ppb (60µg m-3). The reason is probably that these sites are far away from busy roads and receive very low fresh NOx inputs. The other two sites Leeds and Nottingham centre (urban centre sites) are intermediate between rural and roadside monitoring sites. These sites although are urban, receive reasonable high levels of fresh NOx which could reduce ozone mixing ratios but still the frequency of high ozone concentrations are higher than the roadside monitoring sites.

Figure 4:

Ozone mean hourly data distribution at different monitoring sites in the UK, where KS, RS, UC, Re and R define the type of monitoring site and stand for kerbside, roadside, urban centre, remote and rural respectively.

After the comparison of different monitoring sites and studying ozone distributions during different months, it can be concluded that ozone data in the UK do not follow a normal distribution and hence non-parametric statistics should be used for its analysis. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

100 Air Pollution XIX 3.2 Ozone and NOx correlation The scatter plot of 24 h average NOx and ozone data from the Kirkstall site Leeds is shown in Figure 5, and exhibits a clear negative correlation between NOx and ozone and as NOx mixing ratios increase ozone mixing ratios decrease. Most probably the negative correlation is due to NO reactions with ozone which destroys ozone molecules and produces NO2: (NO + O3  NO2 + O2). In Figure 5 along the x-axis 3 main segments can be observed. In the first segment (NOx ppb < 80, left circle - solid line) the negative correlation between ozone and NOx seems very strong (Spearman correlation coefficient (R) value is ‘-0.72’) and ozone mixing ratios decrease linearly with increases in NOx mixing ratios. In this area the data points are dense, as on this monitoring site most data points lie in this segment. In the next segment (middle circle - dashed line, 80 < NOx ppb < 220) the negative correlation is still there but not linear. In this section data points are relatively sparse and the negative correlation is weaker (R value ‘-0.53’). In the last segment (right circle – dotted line, NOx ppb > 220) the negative correlation between ozone and NOx disappears and the curve becomes almost totally horizontal. The negative correlation turns into positive correlation (R value ‘+0.30’). The positive correlation between ozone and NOx at atypically high NOx concentrations is most probably due to NO2 oxidation which gives rise to ozone formation. As this correlation is based on 24 h mean ozone and NOx data, the correlation will be further investigated in section 3.2.1 using 1 minute and hourly data from the Kirkstall monitoring site.

Figure 5:

Scatter plot between 24 h mean ozone and NOx data from KS site (01/11/2007 to 31/10/ 2009).

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

101

3.2.1 Correlation between NOx and ozone for 1 minute and hourly data When NOx mixing ratios reach about 200 ppb the correlation between ozone and NOx becomes positive as demonstrated in section 3.2. In this section those 6 days (12–15 December 2007; 11–12 February 2008) when NOx mixing ratios were 365, 304, 250, 198, 271 and 354 ppb, respectively, are further investigated. Figure 6 shows the scatter plot along with their R values for 1 minute data for the 6 days. Over these days ozone and NOx have positive correlation coefficients, except on 13/12/2007 where the correlation is negative despite the fact that NOx mixing ratio is as high as 304 ppb. Figure 6 (top row, middle column) shows that the scatter plot for 13/12/2007 looks somewhat different from the other scatter plots. The main difference is that for some reason there are some higher ozone mixing ratios at the start of the scatter plot. Hourly mean NOx and ozone mixing ratios show more clearly those high ozone mixing ratios points (not shown here). To find out an explanation for these data points, the author investigated the meteorological variables to see if there was some explanation, as only NOx and ozone correlation cannot provide an answer. It was found from the observations of meteorological variable that on 13 December 2007 the sun came out about 09.00 am and was shining until 12.00. The solar radiation triggered photochemical ozone formation and the ozone mixing ratios reached the highest level of the day (7 to 8 ppb); the average ozone on the day was about 3ppb. As a high level of NOx (304 ppb) was present on the day, probably NOx reacted with photochemically formed ozone and brought the ozone level down, explaining the negative correlation on 13 December.

Figure 6:

Scatter plot of ozone vs. NOx for the six days when NOx mixing ratios were nearly as high as 200 ppb or over and NOx is mostly positively correlated with ozone; R stands for Spearman correlation coefficients.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

102 Air Pollution XIX 3.2.2 Temporal variations in NOx and ozone mixing ratios In this section the correlation between ozone and NOx has been investigated using time variation charts (Figure 7). These diagrams depict the association of ozone and NOx showing how their mixing ratios change on average during different hours of the days, days of the week or months of the year. Figure 7 reveals that ozone mixing ratios are normally higher during spring and summer (March, April, May and June) and lower during winter months (January, February, November and December), where as NOx mixing ratios are higher in winter (January, February, November and December) and lower during summer months (May, June, July and August). In winter the high NOx ratios are probably resulted by slow chemical reactions and slower pollutant dilution due to stagnant atmospheric conditions; the opposite happens in summer (better chemistry and dilution). On the other hand ozone is a secondary pollutant and is mostly produced in the atmosphere by photochemical reactions of NOx and VOCs driven by solar radiation (Ultraviolet radiation – UV). Therefore in winter because of low UV radiation and temperature ozone production is minimal (if any at all); and whatever ozone is present is consumed by freshly produced NO (remember UV radiations are required for ozone production but not for its destruction). But in summer high UV radiation and temperature are responsible for the relatively higher levels of ozone.

Figure 7:

Time variation plot of ozone and NOx mixing ratios (ppb) hourly mean data from Kirkstall site Nov 2007 to Oct 2009.

On a weekly basis traffic volume seems to be the dominant factor for controlling NOx and ozone mixing ratios. On Kirkstall Road the volume of road traffic is higher during weekdays and lower during the weekend; as many companies and institutes do not operate at the weekend. Figure 7 clearly shows the lowest NOx and highest ozone levels on Sunday, followed by Saturday. As WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

103

on Sunday low traffic volumes produce less NOx which result in less ozone destruction and hence ozone levels are higher. Higher ozone levels at the weekend are called the ozone weekend effect (OWE) and is studied by several authors (e.g. 8, 9, 10, and 11). The diurnal changes in NOx and ozone levels seem to be linked with both traffic volume and meteorology. Although NOx and ozone are strongly correlated and both of them are strongly linked with the volume of road traffic, the diurnal average trend of ozone seems to be dominated by solar radiation. The highest ozone mixing ratios were observed at 13.00 to 15.00 hour when UV radiation is often at a maximum; and lowest ozone levels at 06.00 to 07.00 hour in the morning due to the overnight dry deposition and NOx scavenging effect. NOx levels are strongly linked with traffic volume and reached a maximum level at 08.00 to 09.00 am when roads traffic activity is at peak. After that NOx levels come down but rise again at about 17.00 to 18.00 hours in the evening, probably due to late afternoon traffic peak hours. It was observed that correlation between NOx and ozone is stronger during winter (R for January was ‘-0.80’) and weaker during summer (R for May ‘-0.40’). NOx is probably the dominant controlling factor for ozone levels in winter when there is not much photochemical ozone production and as solar radiation and temperature increase they weaken the correlation between NOx and ozone. Therefore for ozone prediction in addition to NOx it is essential to quantify the role of meteorology and traffic flow in controlling ozone levels, which is part of our future plan.

4 Conclusion This study investigates ozone distribution and its association with NOx at a roadside monitoring site, where most of the NOx is believed to be emitted by road traffic. The study demonstrates that ozone distribution is not a fixed phenomenon and rather it varies both spatially and temporally. Our data show that ground-level ozone is not normally distributed and hence should be studied by using non-parametric or distribution free statistics. The study also shows that generally ozone is negatively correlated with NOx, although the strength and nature of correlation may vary as NOx level changes. The negative correlation is strongest at NOx levels up to 80 ppb and becomes weaker afterward. The correlation changes to positive when NOx levels go as high as 200 ppb or over. Moreover, the correlation is stronger in winter months and night times; and weaker in summer months and daytimes probably due to solar radiation. Ongoing investigations are intended to explore the associations between ozone and traffic flow using traffic volume, speed and fleet composition characteristics for better understanding the relationship of ozone with road traffic, which may be helpful in accurate ozone prediction. Accurate prediction of ozone episodes may help to pre-warn the public of the potential high levels of ozone and aid policy makers the development of effective mitigation strategies.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

104 Air Pollution XIX

Acknowledgement We gratefully acknowledge Economic and Social Research Council (ESRC) for providing funding for this study, which is a part of my PhD project.

References [1] Air Quality Expert Group (AQEG). Trends in primary nitrogen dioxide in the UK, the fourth report prepared by the air quality expert group 2007. DEFRA Publication London, 2007. [2] Cape, J.N., Surface ozone concentrations and ecosystem health: past trends and a guide to future projections, the Science of the Total Environment, 400, pp. 257-269, 2008. [3] Air Quality Expert Group (AQEG). Ozone in the UK, the fifth report produced by air quality expert group (AQEG), 2009. DEFRA Publication London, 2009. [4] Reimann, C., Filzmoser, P., Garrett, R. and Dutter, R., Statistical data analysis explained: applied environmental statistics with R. John Wiley and Sons, Ltd, 2008. [5] UK automatic urban and rural network. Department for Environment, Food and Rural Affairs. www.aurn.defra.gov.uk. Accessed July 28, 2010. [6] Carslaw, D.C., Ropkins, K., Openair-project: NERC knowledge transfer. www.openair-project.org. Accessed July 28, 2010. [7] Carslaw, D.C and Beevers, S.D., Estimations of road vehicle primary NO2 exhaust emission fractions using monitoring data in London, Atmospheric Environment, 39, pp. 167–177, 2005. [8] Chang, S.C. and Lee, C.T., Ozone variations through vehicle emissions reductions based on air quality monitoring data in Taipei city, Taiwan, from 1994 to 2003, Atmospheric Environment, 40, pp. 3513–3526, 2006. [9] Bronnimann, S. and Neu, U., Weekend weekday differences of near surface ozone concentration in Switzerland for different meteorological conditions, Atmospheric Environment, 31, pp. 1127–1135, 1996. [10] Pont, V. and Fontan, J., Comparison between weekend and weekday ozone concentration in large cities in France, Atmospheric Environment, 35, pp. 1527–1535, 2000. [11] Gao, O.H., Holmen, B.A., and Niemeier, D.A., Non-parametric factorial analysis of daily weight in motion traffic: implication for the ozone “weekend effect” in Southern California, Atmospheric Environment, 39, pp. 1669–1682, 2005.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

105

Prediction of TSP concentration in a metallurgical city of Brazil using neural networks M. M. C. Lima Research and Development Centre, Usiminas, Brazil

Abstract The aim of this study was to predict Total Suspended Particulate concentration (TSP) in the main areas of Ipatinga, a metallurgical city located in Minas Gerais state, southeast of Brazil. Artificial neural networks (ANN) were the modelling tool used. This model is able to predict pollutant concentration just by training the input and output parameters. The input parameters were meteorological such as wind direction, wind speed, rain, and ambient temperature and also seasonal such as, summer and winter. The output parameter used was the historical data of the total suspended particulate concentration taken between 1996 and 2004. In the modelling, the multilayer perceptron (MLP) model was tested. Among the MLP configurations evaluated, the topology 13-7-6 was chosen. The validation of the model was done by comparing the simulated with the observed values. The results of this model were also compared with the industrial source complex short-term dispersion model (ISCST3). The four statistical tools used to evaluate the fitting were mean squared error (MSE), fractional bias (FB), index of agreement (IA) and linear correlation coefficient (R). Comparing the results it was seen that the predicted values were better in some boroughs and were overestimated in others. Besides, the predicted results of the ANN model were better than the ISCST3 dispersion model. Keywords: artificial neural networks modelling, multilayer perceptron, total suspended particulate concentration, prediction, ISCST3 dispersion.

1 Introduction This paper introduces the study of prediction of Total Suspended Particulate concentration (TSP) in Ipatinga city using artificial neural networks. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110101

106 Air Pollution XIX Ipatinga is located in the Vale do Aço Region, in Minas Gerais state of Brazil where the main industrial activity is iron and steel making. In the iron and steel making process, the raw material handling and fuel combustion are the main causes of particulate emission. Depending on the particulate concentration, the air quality can be modified and causes a lot of damage, such as increasing the dust in residential areas, visibility impairment and harmful to health effects (USEPA [1]). Mathematical models are often used to estimate environmental impacts, saving money from air quality monitoring. One of the most important features of models is the ability to predict or simulate future impact scenarios. The dispersion or diffusion models have been traditionally applied to atmospheric mathematical modelling. These models are able to predict the pollutant concentration using mass balance of statistical data (pollutant emission, wind direction and speed, ambient temperature) and introduce a Gaussian mathematical equation as a solution of pollutant dispersion. Mitkiewicz [2] predicted TSP concentration in Ipatinga using industrial source complex shortterm dispersion model (ISCST3). Recently, artificial neural networks model (ANN) has been used in modelling complex problems. ANN, such as the ISCST3 model is able to predict air pollutant concentration just by training a set of input and output variables. It offers a mathematical solution by adjusting the weights in such a way that output will be close to real data. Comparing ANN to ISCST3 models, the first one has the advantage in adapting few variables to evaluate air pollutant dispersion. It can be seen in many papers published about this subject: Linyan and Wang [3], Wal and Janssen [4], Perez and Reyes [5], Viotti et al. [6], Zickus et al. [7], Perez and Reyes [8], Podnar et al. [9], Ordieres et al. [10], Hooyberghs et al. [11]. There are a variety of artificial neural networks models being used in modelling. Among them, the multilayer perceptron (MLP) is the most cited in air dispersion modelling (Gardner and Dorling [12]). The MLP structure consists of processing elements and connections. The processing elements, called neurons, are arranged in layers such as an input layer, one or more occult layers and an output layer (Haykin [13]). They are all interconnected. In this context, this study aimed to develop ANNs using MLP. The input parameters were meteorological data and the output was the TSP concentration data. This model was also compared with the ISCST3 model. Specifically, it intended to: a) predict TSP concentration using ANNs in six air quality monitoring stations distributed in Ipatinga, b) determine the main variables responsible for the measured TSP concentration; c) investigate the behaviour of the created ANN due to different configuration proposals; d) validate the ANN model using the comparison between the simulated and measured values and e) compare simulated results between the ANN and ISCST3 models.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

2

107

Material and methodology

2.1 Sampling TSP concentrations were collected weekly between 1996 and 2004 from six air quality monitoring stations using High Volume sampler (HiVol) distributed in six districts around Ipatinga named: CA, BR, BA, NC, EC and CS. Meteorological variables (wind speed (m/s), wind direction, rain volume (mm), ambient temperature (oC)) were collected hourly during the same period and converted to daily variables. The wind directions evaluated were north (N), south (S), east (E), west (W), northeast (NE), southeast (SE), southwest (SW) and northwest (NW). Further variables were also created as calm hours (wind speed less than 1 m/s) and seasonal cycle (winter and summer) during the same period from the same database to evaluate their effect on the TSP concentrations. The database contained 400 data. 2.2 Artificial neural modelling Mathematical routine was developed using the software Matlab [14]. MLP was the artificial neural network model used. It had been tested lots of models with different configurations. The best topology was defined by the MSE (mean square error) analysis. This statistical analysis is suggested by Zhang et al [15] as being efficient in evaluating the artificial neural networks models. The meteorological and seasonal variables were used in the input layer. The choice for the model type and the input variables were due to a lot of published studies in literature as mentioned above. Two models with thirteen and six input data were evaluated. The thirteen data were: daily eight wind directions frequency, daily mean wind speed, daily mean rain volume, daily mean temperature, daily calm hours and seasonal cycle (winter = 1 and summer = 2) frequencies. The six were obtained using the principal component analysis (PCA). The purpose of this analysis is to obtain a small number of linear combinations which account for most of the variability in the data (Haykin [13]). In this case, six components had been extracted from the thirteen input data and they were evaluated in the modelling as input data. TSP concentrations, measured in the six monitoring sites, were introduced in the output layer. Only one occult layer was considered in the modelling. The number of neurons in the occult layer was changed as suggested by Zhang et al [15] and Kóvacs [16]. First, an exploratory data analysis was made, detecting if there were outliers in the database. After that the data were normalized and separated in “training” and “validation” sets. The training set was equivalent to 80% of the database and the validation set to 20% (Zhang et al [15]). The learning algorithms used during the training test were Levenberg-Marquardt and Backpropagation. The early stopping criterion was applied to stop the training. In the mathematical routine, expected error, initial weight, bias, activation function, learning algorithm, momentum term and the iteration cycles were established as usual. Finally, after comparing the real values to the simulated ones by MSE analysis, the best topology was defined. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

108 Air Pollution XIX 2.3 Performance evaluation The model validation was done by comparing between the real and simulated values. These simulated values were the model results obtained from the validation set. The average percentage error (E) was determined as shown in eqn. (1).

E

1 n  Cr  Ce  n i 1  C r

  * 100  

(1)

where n is the sampling size, Ce , Cr simulated and real values, respectively. The tendency of the simulated results was evaluated by quantile-quantile plot and type I (false negative) and II (false positive) error analyses. In the type I error, the model under predicts the values, when they were supposed to be above a critical value. In this case, the critical value (TSP concentration) was the air quality standard of 80g/m3 applied in Minas Gerais state. In the other hand, in the type II error, the model over predicts the values. So they were supposed to be below a critical value. The cluster analysis was applied to verify the effect of input data in the output data. This multivariate statistical technique aims to classify the observations or variables due to their similarity, applying a distance measure algorithm. 2.4 Comparison between ANN and ISCST3 models The comparison between ANN and ISCST3 models was made. For running ISCST3 model, it was collected the meteorological parameters, topography and air emission sources characterization. The comparison among the two simulated results and real values, registered in the six monitoring sites, was made. The statistical evaluation tools used were linear correlation coefficient (R), mean square error (MSE), mean fractional bias (FB) and mean index of agreement (IA) (Olesen [17]). The linear correlation coefficient is shown in eqn. (2):





1 n i1 Ce  Ce C r  Cr n R

C C e



(2)

r

The mean square error is shown in eqn. (3):

 C MSE  n

i 1

r

 Ce 

2

n

(3)

The mean fractional bias is shown in eqn. (4):

FB 

Ce  C r 1 n  i 1 n 0.5Ce  Cr 

The mean index of agreement of (IA) is shown in eqn. (5):

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

(4)

Air Pollution XIX

Ce  Cr  1 n 1  i 1 n Ce  C r  C r  C r

109

2

IA 



 



2

(5)

where Ce : averaged simulated concentration, Cr : averaged real concentration, n is the sampling size, Ce , Cr simulated and real values, respectively,  Cr : real concentration standard deviation,  Ce : simulated concentration standard deviation.

3 Results The best ANN configurations results considering the MSE analysis are shown in table 1. Table 1: Model Neurons

MLP models results. Alg. MSE

Output neurons

Input Occult Output CA BA BR EC NC CS 1 13 7 6 LM 466.7 + + + + + + 2 13 27 6 BP 636.5 + + + + + + 3 13 14 5 LM 354.7 + + + + + 4 13 19 1 LM 251.5 + 5 13 10 1 LM 385.1 + 6 13 7 1 LM 208.4 + 7 13 7 1 LM 205.7 + 8 13 10 1 LM 911.3 + 9 13 9 1 LM 336.2 + 10 6 4 6 LM 509.1 + + + + + + 11 6 13 1 LM 323.5 + 12 5 8 1 LM 386.4 + 13 6 6 1 LM 220.9 + 14 6 10 1 LM 188.6 + 15 6 4 1 LM 869.5 + 16 6 3 1 LM 382.0 + (Alg.) Algorithm, (LM) Levenberg Marquardt, (BP) Backpropagation, (+) simulated, (-) not simulated. Few ANN models had in the occult layer a half of neurons of the input layer while other ones (models 2 and 11) had more than double of neurons of the input layer as commented by Kóvacs [16]. The results obtained from models 1, 2 and 10, considering 6 neurons in the output layer, had the same order of magnitude. Leaving the NC air monitoring site out of evaluation (models 1 and 3), the MSE was reduced. Models 4 to 9 and 11 to 16 were created to evaluate each result from each air monitoring station. The results were very similar except for NC. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

110 Air Pollution XIX Model 12 had five input data due to PCA results. The results obtained from models 8 and 15 were probably due to other variables that were not introduced in the model, since the used variables were not able to explain entirely the TSP concentrations. Using MSE approach for comparison results between model 1 and 2, it was showed that Levenberg-Marquardt algorithm was better than Backpropagation. On balance, the MSE results among the models were very similar with the same order of magnitude. For this reason model 1 was chosen to describe the subsequent results. Comparing between models 1 and 10, the use of principal components as input data did not altered the MSE results significantly. 3.1 Performance evaluation The average percentage errors for CA, BA, BR, EC, NC, and CS air quality monitoring sites were 35%, 24%, 27%, 31%, 42% and 37%, respectively. BA simulated results showed a good agreement with measured values. NC had the worst results. In regular meteorological conditions, BA is used in suffering particulate emissions from Usiminas more than the other sites, as it is located downwind from it. So the simulated results were better in BA than in the other ones. The result obtained in NC was the worst due to other variables that were not introduced in the model. Mitckiewicz [1] had also got the worst simulated results in NC using ISCST3 modelling. The cluster analysis is shown in fig. 1. Analysing the distance measurement it is possible to identify three groups. Dendrogram Ward's Method, Squared Euclidean 800

Distance

600

400

200

Cluster 3

Cluster 1

Cluster 2

W

SW

S

NW

SE

Rain

seasoncycle

Calm hours

Figure 1:

Temperature

E

windspeed

N

---------------- ---------------- --------------------NE

0

Cluster analysis.

Cluster 1, representing 50% of the output data, was characterized by N, NE, E wind directions, wind speed less than 1m/s, strong raining storms, ambient temperature about 21oC and the presence of two seasonal cycles. In other words, WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

111

50% of TSP concentrations results were grouped due to similar input data mentioned above. Cluster 2 grouped 19% of the output data due to input variables: SE, S, NW, SW, W wind directions, wind speed more than 1,1m/s, weak raining, ambient temperature about 21oC and winter cycle. Finally, cluster 3 (24% of output data) was characterized by N, NE, and E wind directions, wind speed more than 2,1m/s, frequent raining, ambient temperature about 25oC and the presence of summer cycle. The analysis of quantile-quantile plot and type I and II error was made for all air quality monitoring sites, but in this paper, it will be only shown BA and NC results (figs. 2 and 3). The other ones can be seen in more details in Lima, M. M. C. [18]. In fig. 2, according to quantile-quantile plot analysis, the tendency of 52% of the predicted values was to overestimate the real values. They occurred in meteorological conditions from cluster 1. Considering the analysis of false positive and false negative errors, both were verified. But, the type I error was more often and was determined by the variables characterized by cluster 3. One hypothesis that could probably explain is the location of main particulate emission sources from Usiminas in relation to BA air quality monitoring site. Under those meteorological conditions (characterized by cluster 3), BA air monitoring quality site was downwind from them and if there was an emission increase during that period, the model was not able to evaluate it, as they were not introduced in the modelling. It could explain why predicted values was lower than real. 200

negative false error

Cluster 1 Cluster 2

180

Cluster 3 160

TSP real values (g/m3) (BA)

140

120

100

80

60

40

20

positive false error 0 0

20

40

60

80

100

120

140

160

180

200

TSP predicted values (g/m3) (BA)

Figure 2:

Comparison between real and simulated values for BA.

According to fig. 3, quantile-quantile plot analysis showed that a great part of simulated values was overestimated if compared to the real values. They also occurred in meteorological conditions described in cluster 1. The type II error was more common than the other one. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

112 Air Pollution XIX 240

negative false error

Cluster 1 Cluster 2 Cluster 3

TSP real values (g/m3) (NC)

200

160

120

80

40

positive false error 0 0

40

80

120

160

200

240

TSP predicted values (g/m3) (NC)

Figure 3:

Comparison between real and simulated values for NC.

The possible cause was its location is not favourable and exposed to extra contributions. NC site is close to a paved road with heavy traffic. This situation was not modelled. 3.2 Comparison between ANN and ISCST3 model The comparison between ANN and ISCST3 model results is shown is table 2. Table 2:

Statistical analysis results.

Site Statistical analysis

CA BA BR EC NC CS

FB ISCST3 -0.29 -0.06 0.42 -0.61 -0.82 0.29

ANN 0.06 -0.07 0.12 0.11 0.15 0.04

IA ISCST3 0.36 0.53 0.21 0.19 0.53 0.27

ANN 0.66 0.62 0.69 0.70 0.76 0.61

R ISCST3 0.02 0.37 0.02 0.18 0.44 0.08

ANN 0.48 0.47 0.56 0.54 0.65 0.44

MSE ISCST3 1754.86 2944.57 6124.87 1652.21 4003.12 14616.33

ANN 334.79 671.00 173.68 202.97 1027.18 390.40

FB and MSE usually measure bias of a model. R is the correlation between the observed and simulated values and IA shows the degree that the model predictions are error free. The ideal model would result in values of MSE = 0, R = 1, FB = 0 and IA = 1. Analyzing the mean fractional bias results, NC, BR and EC air quality monitoring stations had the worst results in both models. The mean index of agreement (IA) showed the EC and CS results were the worst in ISCST3 and ANN models, respectively. The values of IA and FB obtained in WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

113

ANN were better than the ISCST3 dispersion modelling. The values of R and MSE obtained in ANN were also better than the ISCST3 model. For this reason, ANN model should be considered closer to the ideal model.

4 Conclusions This study showed that ANN can be a powerful data analysis tool to evaluate air pollutant dispersion. Even though, in the modelling, particulate emissions from Usiminas were not introduced as input variable, ANN was able to predict the TSP concentration in Ipatinga atmosphere using meteorological and seasonal cycle data. On balance, the tendency of simulated values was to overestimate the real values. The best modelling results were obtained in BR, BA and CA. BA had the best result and NC, the worst. BA monitoring site is favourable to suffer particulate emissions from Usiminas more than the other sites, as it is located downwind from it in regular meteorological conditions. It could explain why the simulated results were better in BA than in the other ones. The simulated result obtained in NC was the worst due to other variables that were not introduced in the model and its unfavourable location. The type I and II errors results were not representative in the modelling. The type II error only occurred in NC monitoring site and type I error was more common in BA site. Those errors were caused by their location probably. According to statistical analysis of MSE, FB, IA and R, the predicted results from the ANN model were better than the ISCST3 dispersion model.

References [1] United States Environmental Protection Agency (USEPA). Air Quality Criteria for Particulate Matter – Vol. II – EPA/600/P-99/002a-f, 2004, www.epa.gov/pmresearch/. [2] Mitkiewicz, G. F., Metodologia para avaliação da dispersão atmosférica de poluentes provenientes de um complexo siderúrgico industrial, 2002, Departamento de Engenharia Sanitária e Ambiental. (Dissertação de Mestrado em Meio Ambiente), Escola de Engenharia da UFMG, Belo Horizonte, Brasil. [3] Linyan, S. Wang, Y., A neural network model for environmental predication: case study for China. Computers and Industrial Engineering, China, 31, pp. 879-883, 1995. [4] Wal, J.T., Janssen, L.H.J.M., Analysis of spatial and temporal variations of PM10 concentrations in the Netherlands using Kalman filtering. Atmospheric Environment, 34, pp. 3675-3687, 2000. [5] Perez, P. Reyes, J., Prediction of particulate air pollution using neural techniques. Neural Computing & Applications, Chile, 10, pp. 165-171, 2001. [6] Viotti, P.; Liuti, G.; Genova, P. D., Atmospheric urban pollution: applications of an artificial neural network (ANN) to the city of Perugia. Ecological Modelling, Rome, 143, pp. 27-46, 2002. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

114 Air Pollution XIX [7] Zickus, M., Greig, A.J., Niranjan, M., Comparison of four machine learning methods for predicting PM10 concentrations in Helsinki, Finland. Water, Air, and Soil Pollution, 2, pp. 717-729, 2002. [8] Perez, P., Reyes, J., Prediction of maximum of 24-h average of PM10 concentrations 30 h in advance in Santiago, Chile. Atmospheric Environment, 36, pp. 4555-4561, 2002. [9] Podnar, D., Koracin, D., Panorska, A., Application of artificial neural networks to modelling the transport and dispersion of tracers in complex terrain. Atmospheric Environment, 36, pp. 561-570, 2002. [10] Ordieres, J.B., Vergara, E.P., Capuz, R.S., Salazar, R.E., Neural network prediction model for fine particulate matter (PM2.5) on the US-Mexico border in El Paso (Texas) and Ciudad Juárez (Chihuahua). Environmental Modeling & Software, 20, pp. 547-559, 2005. [11] Hooyberghs, J., Mensink, C., Dumont, G., Fierens, F., Brasseur, O., A neural network forecast for daily average PM10 concentrations in Belgium. Atmospheric Environment, 39, pp. 3279-3289, 2005. [12] Gardner, W. M.; Dorling, R. S., Artificial neural networks (the multilayer perceptron) – a review of applications in the atmospheric sciences. Atmospheric Environment, 32(14/15), pp. 2627-2636, 1998. [13] Haykin, S., Neural networks: a comprehensive foundation, Prentice-Hall Inc.: Canada, pp. 1-842, 1999. [14] Matlab R12, Version 5.1, The language of technical computing: getting started with Matlab, The Mathworks Inc., pp. 1-86, 1997. [15] Zhang, G. Patuwo, B.E. HU, M. Y., Forecasting with artificial neural networks: the state of the art. International journal of forecasting, 14, pp. 35-62, 1998. [16] Kovács, Z.H., Redes neurais artificiais: fundamentos e aplicações, Editora Livraria da Física: São Paulo, pp. 1-174, 2002. [17] Olesen, H., Model validation kit – status and outlook, National Environmental Research Institute: Denmark, 1997. [18] Lima, M. M. C., Estimativa de concentração de material particulado em suspensão na atmosfera por meio da modelagem de redes neurais artificiais (Dissertação de Mestrado em Meio Ambiente), Escola de Engenharia da UFMG, Belo Horizonte, Brasil.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Section 2 Monitoring and measuring

This page intentionally left blank

Air Pollution XIX

117

The use of mineral magnetic measurements as a particulate matter (PM) proxy for road deposited sediments (RDS): Marylebone Road, London C. A. Booth1, C. J. Crosby1, D. E. Searle1, J. M. Khatib1, M. A. Fullen2, A. T. Worsley3, C. M. Winspear1 & D. A. Luckhurst2 1

STech, University of Wolverhampton, UK SAS, University of Wolverhampton, UK 3 NGAS, Edge Hill University, UK 2

Abstract Road deposited sediments (RDS) are a recognised pollution problem and a worrying public health concern of many urban environments. Linkages between the magneto characteristics of RDS and their particle size properties have been explored to determine the extent to which magnetic technologies can be utilised as a proxy for proffering insights to address pollution challenges. Samples (n = 60) were collected (May, 2008) along both sides of a busy urban road (Marylebone Road) in central London, UK. Magnetic concentration parameters (LF, χARM and SIRM) reveal high levels of magnetic material, when compared to previous urban RDS studies. Correlation analysis between the magnetic parameters and textural parameters (LF, χARM, SIRM and PM1.0, PM2.5, PM10) show significantly strong relationships but, unlike earlier studies, the trends display negative correlations. Despite this kinship not adhering to previously identified trends, this does not mean that mineral magnetic measurements cannot be used as a proxy. Moreover, it simply implies that the nature of any trends needs to be established for specific places before it can be reliably applied as a proxy. Keywords: environmental magnetism, particle size, street dust, built environment, epidemiology, public health. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110111

118 Air Pollution XIX

1 Introduction Road deposited sediments (RDS) (sometimes referred to as street dust) can be toxic [1–3] and contribute to the particulate matter (e.g. PM1 0, PM2 5, PM10) loadings of urban sediments. Given the microscopic characteristics of these grains they are easily re-suspended and can become a significant human exposure source. When particles are absorbed through inhalation it can lead to serious health problems, such as cardiovascular disease and respiratory illness [4–6]. In the UK, for instance, it is estimated that 24,000 deaths occur annually due to poor air quality [7]. From an environmental perspective, RDS also cause urban drainage system issues, where urban runoff transfers the mix of sediments and toxic substances to receiving drainage systems and/or watercourses, causing detrimental effects on water quality and the health of the natural environment [8–10]. Other studies have highlighted linkages between road surface runoff and the deleterious influence of inorganic metal toxicants on benthic community structure and function in receiving water bodies [11, 12]. This is because the composition of RDS comprises a variety of both natural materials (quartz, clay and carbonates) and anthropogenic particles [13, 14]. Typically, anthropogenic sediments are derived from industrial and vehicle-generated sources, causing them to contain both metals (metallic fragments and iron oxides) and organic matter [13, 15]. Previous studies [16–18] have demonstrated the distribution of heavy metals within the built environment, in particular their proximity to roadsides and urban catchments. Heavy metals have been found to be associated with road traffic in urban areas and are known to contain particles associated with vehicle wear (such as tyres, body, brake linings), road surface wear, road paint degradation, vehicle fluids, and particulate emissions [19–22]. There is growing awareness of the issues associated with PM pollution [23], particularly within the built environment arena. As such, programmes of PM monitoring are now commonplace in major towns and cities of many countries. For instance, the UK currently has a network of 64 automatic monitoring sites (using gravimetric analysers of PM concentrations). The suite of sites, forming the ‘Automatic Urban and Rural Network’ (AURN), operated by the Department for Environment, Food and Rural Affairs (DEFRA), are positioned at strategic urban locations where data is continually measured and monitored (www.airquality.co.uk), recording hourly and daily measurements of PM10, PM2 5, Nitrogen oxides, Sulphur dioxide and Ozone, amongst others. The European Air Quality Framework Directive (96/62/EC) and the First Air Quality Daughter Directive (1999/30/EC) legislation require PM10 levels not to exceed 50 gm-3 for more than 35-days per year and set the maximum annual mean limit at 40 gm-3. That said, Marylebone Road is one of the most widely publicised air pollution sites in the UK, where PM concentrations are known to regularly exceed regulatory standards [24–27]. Therefore, this venue was selected so as to validate the potential of adopting an alternative technology for monitoring the PM sizes of RDS. Previous RDS studies, elsewhere, have already shown kinships exist between particle sizes and heavy metal content [28–31] but WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

119

the magnetic signature of RDS has only recently been identified to exhibit significant correlations with these characteristics [32, 33]. However, magnetotechniques have not been widely applied to sites where there are known PM issues and, thus, allow new findings, herein, to be compared with other studies. This work aims to demonstrate the extent to which particular magnetic concentration parameters can be used as a particle size proxy for urban RDS and attempt to highlight whether any data associations follow the predictable trends of similar studies.

2 Case study Marylebone Road (Fig. 1) is a major arterial route (A501) for traffic (up to seven lanes, including bus lanes) and pedestrians within the city of Westminster (central London) that forms part of the inner London ring-road and marks the northern limit of the London congestion charging zone. Roadside buildings create an asymmetric street canyon with a height to width ratio of ~0.8 [34]. The road has consistently high daily mean PM10 levels that regularly exceed legislative requirements (e.g. 185 incidents 2002–2004 [34]; 47 (a)

(b)

(c)

(d)

Figure 1: Views of Marylebone Road (May, 2008): (a) Site 9, East facing (Grid reference: 527833 181940); (b) Site 14, East Facing (Grid reference: 527963 181995); (c) Site 29, West facing (Grid reference: 528783 182200); and (d) Site 61, East facing (Grid reference: 526243 180535). WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

120 Air Pollution XIX incidents in 2007; 29 incidents in 2008; and 36 incidents in 2009). These exceedences are thought to be due to congestion and high traffic flows [35], with over 80,000 vehicles per day using the road [34]. This has led to a number of independent studies on PM [25, 34, 36–39], which reveal increases in iron rich dusts [39] and a greater frequency of PM10 exceedences during weekdays when traffic conditions peak [34].

3 Materials and methods 3.1 Sample collection and preparation Street dust was collected from the pavements (sidewalks) at regular spacings along both sides of the road. Typically, 10–50 g dust samples were collected (from ~1 m2) by brushing with a small hand-held fine-bristle brush. Dust was then transferred to clean, pre-labelled, self-seal, airtight plastic bags. In the laboratory, samples were visibly screened to remove macroscopic traces of hair, animal and plant matter [21]. 3.2 Mineral magnetic measurements All samples were subjected to the same preparation and analysis procedure. Samples were dried at room temperature (<40 C), weighed, packed into 10 ml plastic pots and immobilized with clean sponge foam and tape prior to analysis. Initial, low–field, mass–specific, magnetic susceptibility () was measured using a Bartington (Oxford, England) MS2 susceptibility meter. By using a MS2B sensor, low frequency susceptibility was measured (LF). Anhysteretic Remanence Magnetisation (ARM) was induced with a peak alternating field of 100 mT and small steady biasing field of 0.04 mT using a Molspin (Newcastle– upon–Tyne, England) A.F. demagnetiser. The resultant remanence created within the samples was measured using a Molspin 1A magnetometer and the values converted to give the mass specific susceptibility of ARM (ARM). The samples were then demagnetized to remove the induced ARM and exposed to a series of successively larger field sizes up to a maximum ‘saturation’ field of 1000 mT, followed by a series of successively larger fields in the opposite direction (backfields), generated by two Molspin pulse magnetisers (0-100 and 01000 mT). After each ‘forward’ and ‘reverse’ field, sample isothermal remanent magnetisation (IRM) was measured using the magnetometer [40]. 3.3 Laser diffraction measurements All samples were subjected to the same textural preparation and analysis procedure, using sieving (2000 m aperture) followed by laser diffraction analysis. Low Angle Laser Light Scattering (LALLS), using a Malvern (Malvern, England) Mastersizer Long-bed X with a MSX17 sample presentation unit, enabled rapid measurement of particle sizes within the 0.1-2000 m range. Macroscopic traces of organic matter were removed from representative subWIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

121

samples before being dampened by the dropwise addition of a standard chemical solution (40 g/l solution of sodium hexametaphosphate ((NaPO3)6) in distilled water) to help disperse aggregates. To ensure complete disaggregation, each slurry was then subjected to ultrasonic dispersion in a Malvern MSX17 sample presentation unit. For greater precision, the mean of five replicate analyses was measured with a mixed refractive indices presentation setting. A standard range of textural parameters was calculated, including the percentage of sand, silt and clay class sizes and their sub-intervals. The Malvern instrumentation was regularly calibrated using latex beads of known size [40].

4 Results Particle size data (Table 1) indicates samples are dominated by sand (~79%), silt (~19%) and clay (~2%), in respective orders. From a respiratory-health perspective, PM10 grains represent ~7%, PM2 5 ~3% and PM1 0 ~1% of the sediments at pavement level. This is noteworthy because, once suspended, particles <10 m in diameter are able to remain airborne for hours or days and, in some cases, even weeks [41]. Therefore, the presence of PM of these sizes on pavement surfaces indicates either the sediments have not been disturbed recently or they have only just settled-out. Since the pavements normally receive frequent and heavy foot-traffic, it is assumed the time of sampling (0500 – 0900) and weather conditions (warm, dry and still) have permitted sizeable PM accumulations. Mineral magnetic characteristics have been summarised (Table 2). LF is roughly proportional to the concentration of ferrimagnetic minerals within the sample, although in materials with little or no ferrimagnetic component and a relatively large antiferromagnetic component, the latter may dominate the signal. ARM is particularly sensitive to the concentration of magnetic grains of stable Table 1: Summary particle size properties of the RDS: (a) traditional sediment size fractions and (b) respiratory health-related size fractions (n = 60 samples). (a) Sand (63-2000 m)

Mean (%) 78.88

Minimum (%) 45.92

Maximum (%) 91.45

Standard Deviation 7.98

Silt (2-63 m)

19.08

7.27

53.01

7.85

Clay (<2 m)

2.04

0.90

8.50

1.31


Mean (%) 6.57

Minimum (%) 2.84

Maximum (%) 23.99

Standard Deviation 3.88


2.64

1.25

9.82

1.69


1.37

0.66

5.39

1.02

(b)

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

122 Air Pollution XIX single domain size, e.g. ~0.03–0.06 m. SIRM is related to concentrations of all remanence-carrying minerals in the sample, but is also dependent upon the assemblage of mineral types and their magnetic grain size. These data indicate the samples contain moderate to high magnetic concentrations. Compared to previous urban magneto-dust studies, the mean values are sizeably greater than those of Liverpool (23.7 x10-7 m3 kg-1) [42] and Shanghai (29.9 x10-7 m3 kg-1) [43]. Spearman’s rank correlation coefficient values (rs) between the mineral magnetic concentration parameters and particle size parameters have been grouped according to traditional sediment size fractions (Table 3) and respiratory health-related size fractions (Table 4). Significant relationships (p <0.001; n = 50) exist between clay content and all of the magnetic concentration parameters, Table 2: Summary mineral magnetic properties of the RDS (n = 60 samples). Units LF ARM SIRM

10-7m3kg-1 -7

3

10 m kg -5

2

-1

10 Am kg

-1

Mean

Minimum

Maximum

47.88

11.13

87.97

Standard Deviation 16.12

2.18

0.04

11.37

1.53

3779.00

956.00

6161.00

119.80

Table 3: Spearman’s rank correlation coefficients (rs) between mineral magnetic concentration and particle size parameters for the RDS based on traditional sediment size fractions. (a) LF

Clay <2 m -0.453***

Silt 2-63 m 0.010

Sand 63-2000 m 0.076

ARM

-0.384**

0.109

-0.017

SIRM

-0.386**

-0.049

0.135

Note: Significance levels: p <0.05 = *; p <0.01 = **; p <0.001 = ***. Table 4: Spearman’s rank correlation coefficients (rs) between mineral magnetic concentration and particle size parameters for the RDS based on respiratory health-related size fractions (n = 60 samples). (b)




LF

-0.589***

-0.575**

-0.526**

ARM

-0.511**

-0.471**

-0.402**

SIRM

-0.554**

-0.553**

-0.519**

Note: Significance levels: p <0.05 = *; p <0.01 = **; p <0.001 = ***. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

123

which is similar for each of the PM10, PM2 5 and PM1 0 sizes. Therefore, this indicates all the magnetic concentration parameters could potentially be used as a particle size proxy, particularly if the kinship is required with particles < PM10.

5 Discussion Earlier sedimentological works have noted significant correlations exist between magnetic and particle size properties. Oldfield et al. [44] identified anhysteretic remanent magnetisation (ARM) measurements reflect the concentration of finegrained magnetite (<0.1 m) in clay fractions and LF measurements can be used to infer the presence of coarser multi-domain magnetite (>1.0 m) in sands and coarse silts. Clifton et al. [45] found LF was strongly associated with sands and medium silts, ARM was strongly associated with clay and fine silts, and SIRM was strongly associated with very fine to medium silts. Zhang et al. [46] suggested that both percentage frequency-dependent magnetic susceptibility (FD%) and ARM can be used as a proxy for clay content. These studies have illustrated sand to correlate negatively with LF (r = –0.94), ARM (r = –0.96) and SIRM (r = –0.91); silt to correlate positively with LF (r = 0.96), ARM (r = 0.96) and SIRM (r = 0.96); and clay to correlate positively with LF (r = 0.82), ARM (r = 0.94) and SIRM (r = 0.81). When data presented here are compared to earlier investigations, it is apparent that each magnetic parameter also correlates with particle size but the significance is only notable with clay fraction and not the sand and silt fractions. Nevertheless, on first observation, this highlights the potential use of mineral magnetic data as a means of normalizing compositional analytical data (i.e. geochemical) for particle size. However, an important disparity is that the trends observed in this work show negative relationships with the clay fraction, while previous works have always revealed positive trends with the clay fraction. This highlights a sizeable issue for the universal application of mineral magnetic measurements as a particle size proxy. Previous RDS works have also noted significant correlations exist between LF, ARM, SIRM and respiratory health-related particle size fractions. Booth et al. [32] revealed PM10 to correlate with LF (r = 0.69; p <0.001; n = 50); PM2 5 to correlate with LF (r = 0.71; p <0.001; n = 50), ARM (r = 0.30; p <0.05; n = 50) and SIRM (r = 0.33; p <0.05; n = 50); and PM1 0 to correlate with LF (r = 0.66; p <0.001; n = 50), ARM (r = 0.41; p <0.01; n = 50) and SIRM (r = 0.32; p <0.05; n = 50). Similarly, Crosby et al. [33] revealed PM10 to correlate with ARM (r = 0.44; p <0.01; n = 35) and SIRM (r = 0.43; p <0.01; n = 35); PM2 5 to correlate with ARM (r = 0.45; p <0.01; n = 35) and SIRM (r = 0.43; p <0.01; n = 35); and PM1 0 to correlate with ARM (r = 0.42; p <0.01; n = 35) and SIRM (r = 0.40; p <0.05; n = 35). Marylebone Road displays similar correlation significance levels to those of the towns of both Southport [32] and Scunthorpe [33]. However, a notable discrepancy in the correlations is that Marylebone Road again displays significant negative trends; whereas, both of the other places have significant WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

124 Air Pollution XIX positive trends. As such, this represents a potential flaw in the use of mineral magnetic measurements as a universal proxy. However, despite this apparent setback, it does not mean that mineral magnetic measurements cannot be used as a proxy. Moreover, it simply implies that the nature of any trends needs to be established for specific places before it can be reliably applied as a proxy. These differences offer an opportunity to provide speculative reasoning to explain the outcomes; whereby, it is postulated that the differences are due to the RDS being derived from different sources that have varying characteristics and/or are derived from several mixed sources. To support this argument, attention is drawn to the similarities and disparity of the venues already mentioned. For instance, given the size and restrictive (street canyon) nature of Marylebone Road it is proposed that the RDS are chiefly derived from a singular source (i.e. vehicular); whereas, Southport is a seaside resort with no noteworthy industry so it is proposed that the RDS are probably derived from more than one main source (i.e. a mix of wind-blown coastal sediments and vehicular derived dusts); likewise, Scunthorpe is celebrated as an iron and steel town so it is proposed that the RDS are also probably derived from more than one main source (i.e. a mix of industrial emissions and vehicular derived dusts). Verification of this reasoning would require detailed investigation, such as SEM analyses and/or complex sediment source modelling. However, as with most investigations, these findings promote the need for further research on the reliability of using magnetic technologies as a pollution proxy but, likewise, it also offers a provoking avenue to expand the work to provenance studies.

6 Future work This work forms part of a wider investigation that includes other places in the UK (Dumfries, Norwich, Oswestry, Runcorn, Salford, Scunthorpe and Wolverhampton), which is attempting to address the same aims as those posed in this particular work. It is anticipated that it will offer better insights into the reliability of using mineral magnetic measurements as a particulate matter proxy.

7 Conclusions As with previous studies, this work indicates each of the magnetic concentration parameters could be reliably employed as a particle size proxy for urban RDS, where the finest fraction (<10m) is the focus. However, the trends displayed in this work are negative correlations and, since this is unlike other studies, it indicates that the perceived universal relationship does not always exist like previously proposed. Despite this potential discrepancy in its suitability, it does not mean that mineral magnetic measurements cannot be used as a proxy. Moreover, it simply implies that the nature of any trends needs to be established for specific places before it can be reliably applied as a proxy.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

125

Acknowledgements This research forms part of a doctoral investigation supported by the School of Technology at the University of Wolverhampton, for which the second author expresses his gratitude. All authors also thank the School of Applied Sciences at the University of Wolverhampton for unlimited access to analytical facilities. Thanks are also extended to Charlotte Turner for her assistance in preparation of this research article.

References [1] Zhang, M. & Wang, H. (2009) Concentrations and chemical forms of potentially toxic metals in road-deposited sediments from different zones of Hangzhou, China. Journal of Environmental Sciences, 21, 625-631. [2] Shi, G., Chen, Z., Bi, C., Li, Y., Teng, J., Wang, L. & Xu, S. (2010) Comprehensive assessment of toxic metals in urban and suburban street deposited sediments (SDSs) in the biggest metropolitan area of China. Environmental Pollution, 158, 694-703. [3] Faiz, Y., Tufail, M., Tayyeb Javed, M., Chaudhry, M.M. & Siddique, N. (2009) Road dust pollution of Cd, Cu, Ni, Pb and Zn along Islamabad Expressway, Pakistan. Microchemical Journal, 92, 186-192. [4] Brunekreef, B. & Holgate, S.T. (2002) Air pollution and health. The Lancet, 360, 1233-1242. [5] Englert, N. (2004) Fine particulates and human health - a review of epidemiological studies. Toxicology Letters, 149, 235-242. [6] Gulland, A. (2002) Air pollution responsible for 600,000 premature deaths worldwide. British medical Journal, 325, 1380-1381. [7] Department of Health (1998) Quantification of health effects of air pollution in the United Kingdom committee on the medical effects of air pollution. London: The Stationary Office. [8] Pontier, H., Williams, J.B. & May, E. (2004) Progressive changes in water and sediment quality in a wetland system for control of highway runoff. Science of the Total Environment, 319, 215-224. [9] Zanders, J.M. (2005) Road sediment: characterization and implications for the performance of vegetated strips for treating road run-off. Science of the Total Environment, 339, 41-47. [10] Sutherland, R.A., Pearson, D.G. & Ottley, C.J. (2007) Platinum-group elements (Ir, Pd, Pt and Rh) in road deposited sediments in two urban watersheds, Hawaii. Applied Geochemistry, 22, 1485-1501. [11] Maltby, L., Boxall, A.B.A., Forrow, D.M., Calow, P. & Betton, C.I. (1995) The effects of motorway runoff on freshwater ecosystems: 2 Identifying major toxicants. Environmental Toxicology and Chemistry, 14, 1093-1101. [12] Pitt, R., Field, R., Lalor, M. and Brown, M. (1995) Urban stormwater toxic pollutants: assessment, sources, and treatability. Water Environment Research 67, 260-275.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

126 Air Pollution XIX [13] Perry, C. & Taylor, K. (2007) Environmental Sedimentology. Blackwell Publishing, Oxford. [14] Taylor, K.G. & Robertson, D.J. (2009) Electron microbeam analysis of urban road-deposited sediment, Manchester, UK: Improved source discrimination and metal speciation assessment. Applied Geochemistry, 24, 1261-1269. [15] Fergusson, J.E. & Ryan, D.E. (1984) The elemental composition of street dust from large and small urban areas related to city type, source and particle size. Science of the Total Environment, 34, 101-116. [16] Herngren, L., Goonetilleke, A. & Ayoko, G.A. (2006) Analysis of heavy metals in road deposited sediments. Analytica Chimica Acta, 571, 270-278. [17] Rijkenberg, M.J.A. & Depree, C.V. (2010) Heavy metal stabilization in contaminated road-derived sediments. Science of the Total Environment, 408, 1212-1220. [18] Duong, T.T. & Lee, B.K. (2011) Determining contamination level of heavy metals in road dust from busy traffic areas with different characteristics. Journal of Environmental Management, 92, 554-562. [19] De Miguel, E., Llamas, J., Chacon, E., Berg, T., Larssen, S., Royset, O. & Vadset, M. (1997) Origin and patterns of distribution of trace elements in street dust: unleaded petrol and urban lead. Atmospheric Environment, 31, 2733-2740. [20] Sutherland, R.A & Tolosa, C.A. (2000) Multi-element analysis of road deposited sediment in an urban drainage basin, Honolulu, Hawaii. Environmental Pollution, 110, 483-495. [21] Shilton, V.F., Booth, C.A., Giess, P., Mitchell, D.J. & Williams, C.D. (2005) Magnetic properties of urban street dust and its relationship to organic matter content in the West Midlands, U.K. Atmospheric Environment, 39, 3651-3659. [22] Manno, E., Varrica, D. & Dongarra, G. (2006) Metal distribution in road dust samples collected in an urban area close to a petrochemical plant at Gela, Sicily. Atmospheric Environment, 40, 5929-5941. [23] Amato, F., Pandolfi, M., Viana, M., Querol, X., Alastuey, A. & Moreno, T. (2009) Spatial and chemical patterns of PM10 in road dust deposited in urban environment. Atmospheric Environment, 43, 1650-1659. [24] Scaperdas, A. & Colvile, R.N. (1999) Assessing the representativeness of monitoring data from an urban intersection site in central London, UK. Atmospheric Environment, 33, 661-674. [25] Kaur, S., Nieuwenhuijsen, M.J., Colvile, R.N., (2005) Pedestrian exposure to air pollution along a major road in Central London, UK. Atmospheric Environment, 39, 7307-7320. [26] Bigi, A. & Harrison, R.M. (2010) Analysis of the air pollution climate at a central urban background site. Atmospheric Environment, 44, 2004-2012. [27] Wang, H., Colvile, R.N., Pain, C., Aristodemou, E. & ApSimon, H.M. (2011) Understanding peak pedestrian exposures due to traffic emissions within the urban environment. Transportation Research Part D: Transport and Environment, 16, 392-401. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

127

[28] Sutherland, R.A. (2003) Lead in grain size fractions of road deposited sediment. Environmental Pollution, 121, 229-237. [29] Adachi, K & Tainosho, Y. (2005) Single particle characterization of sizefractionated road sediments. Applied Geochemistry, 20, 849-859. [30] Zhao, H., Li, X., Wang, X. & Tian, D. (2010) Grain size distribution of road deposited sediment and its contribution to heavy metal pollution in urban runoff in Beijing, China. Journal of Hazardous Materials, 183, 203210. [31] Fujiwara, F., Rebagliati, R.J., Dawidowski, L., Gómez, D., Polla, G., Pereyra, V. & Smichowski, P. (2011) Spatial and chemical patterns of size fractionated road dust collected in a megacity. Atmospheric Environment, 45, 1497-1505. [32] Booth, C.A., Winspear, C.M., Fullen, M.A., Worsley, A.T., Power, A.L. & Holden, V.J.C. (2007) A pilot investigation into the potential of mineral magnetic measurements as a proxy for urban roadside particulate pollution. In: Air Pollution XV (Editors) C.A. Borrego & C.A Brebbia, WIT Press, 391-400. [33] Crosby, C.J., Booth, C.A., Worsley, A.T., Fullen, M.A., Searle, D.E., Khatib, J.M. & Winspear, C.M. (2009) Application of mineral magnetic concentration measurements as a particle size proxy for urban road deposited sediment. In: Brebbia, C.A. & Popov, V. (Eds), Air Pollution XVII, WIT Press, Southampton, 153-162. [34] Charron, A., Harrison, R.M. & Quincey, P. (2007) What are the sources and conditions responsible for exceedences of the 24 h PM10 limit value (50 μg m−3) at a heavily trafficked London site? Atmospheric Environment, 41, 1960-1975. [35] Air Quality Expert Group (2005) Particulate Matter in the United Kingdom: Summary. DEFRA, London. [36] Thorpe, A.J., Harrison, R.M., Boulter, P.G. & Mccrae, I.S. (2007) Estimation of particle resuspension source strength on a major London Road. Atmospheric Environment, 41, 8007-8020. [37] Chen, T., Gokhale, J., Shofer, S. & Kuschner, W. (2007) Outdoor Air Pollution: Particulate Matter Health Effects. The American Journal of the Medical sciences, 333, 235 - 243. [38] Fuller, G.W. & Green, D. (2006) Evidence for increasing concentrations of primary PM10 in London. Atmospheric Environment, 40, 6134-6145. [39] Harrison, R.M., Jones, A.M. & Lawrence, R.G. (2004) Major component composition of PM10 and PM2 5 from roadside and urban background sites. Atmospheric Environment, 38, 4531-4538. [40] Booth, C.A., Walden, J., Neal, A. & Smith, J.P. (2005) Use of mineral magnetic concentration data as a particle size proxy: a case study using marine, estuarine and fluvial sediments in the Carmarthen Bay area, South Wales, U.K. Science of the Total Environment, 347, 241-253. [41] Harrison, R.M. (2004) Key pollutants – airborne particles. Science of the Total Environment, 334, 3-8.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

128 Air Pollution XIX [42] Xie, S., Dearing, J.A. & Bloemandal, J. (2000) The organic matter content of street dust in Liverpool, UK and its association with dust magnetic properties. Atmospheric Environment, 34, 269-275. [43] Shu, J., Dearing, J.A., Morse, A.P., Yu, L. & Yuan, N. (2001) Determining the source of atmospheric particles in Shanghai, China, from magnetic geochemical properties. Atmospheric Environment, 35, 2615-2625. [44] Oldfield, F., Richardson, N., Appleby, P.G. & Yu, L. (1993) 241Am and 137 Cs activity in fine grained saltmarsh sediments from parts of the N.E. Irish Sea shoreline. Journal of Environmental Radioactivity, 19, 1-24. [45] Clifton, J., McDonald, P., Plater, A. & Oldfield, F. (1999) Derivation of a grain-size proxy to aid the modelling and prediction of radionuclide activity in saltmarshes and mud flats of the Eastern Irish Sea. Estuarine, Coastal and Shelf Science, 48, 511-518. [46] Zhang, W., Yu, L. & Hutchinson, S.M. (2001) Diagenesis of magnetic minerals in the intertidal sediments of the Yangtze Estuary, China, and its environmental significance. Science of the Total Environment, 266, 160-175.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

129

Elemental carbon as an indicator to monitor the effectiveness of traffic related measures on local air quality M. H. Voogt, A. R. A. Eijk, M. P. Keuken & P. Zandveld TNO, Netherlands Applied Research Organisation, The Netherlands

Abstract Even when European standards on air quality are met, health effects occur near busy roads because of the increased exposure to tailpipe emissions. For a proper assessment of these effects, an additional indicator is available: elemental carbon (EC). This is a sensitive indicator for particulate matter from combustion emissions by road traffic. Currently, the concentrations of EC along busy roads are still significantly increased relative to the background, in contrast to PM10 and PM2 5. Therefore, the effectiveness of traffic measures on local air quality and associated health effects can be better assessed with EC than with PM mass. This paper discusses the relevance of EC as an indicator in the assessment of air quality and health. This is illustrated with two pilot studies in the Dutch cities of Arnhem and Helmond where traffic measures were evaluated with EC measurements. Also model calculations of the effect of 80 km/h speed limit zones on Dutch motorways on the health risks of local residents are presented. It is concluded that the assessment of EC through monitoring and/or modelling is a powerful policy instrument to analyse the impact of local traffic measures on air quality and health. Typically, people living close to busy road transport may live 1 to 6 months longer when measures like enhancing traffic flow, speed limiting and environmental zoning are implemented. Keywords: EC, traffic measures, health, measuring, modelling.

1 Introduction Authorities implement traffic measures in order to improve the accessibility and quality of the urban environment. Examples of such actions are: volume measures to reduce inner city traffic e.g. by parking policy and encouraging cycling and WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110121

130 Air Pollution XIX public transport; environmental zoning; speed-limiting measures on urban highways; and traffic flow measures such as “green waves” on urban roads. Model calculations can provide insight into the effectiveness of most of such measures on air quality based on emission factors by vehicle category and speed regime. Regarding PM10 or PM2 5 the effects of traffic measures are relatively small, since traffic emissions contribute only to a small extent to the total concentration. For EC on the other hand, traffic is the major contributor along urban roads. Therefore, effects of traffic measures are more pronounced for EC. EC emission factors can be derived from the emission factors for PM2 5 [1]. In combination with population data, the exposure to EC concentration can then be determined. Also, an assessment can be made of health effects by expressing the concentration differences in gain (or loss) of life years. For local measures aimed at improving the traffic flow, it is quite difficult to calculate the effects on EC concentrations. The actual effect of a "green wave" on the traffic flow at a given location is difficult to predict. Also, since the effects on driving dynamics are not well known, the impact on local exhaust emissions can not be modelled accurately. To gain insight into the effectiveness of local traffic flow measures on air quality, measurement campaigns can be of help. This paper describes two pilot studies where traffic measures were evaluated with EC measurements. In section 2, the study approach is discussed. Section 3 describes the results from two pilot studies based on monitoring in the city of Helmond and Arnhem. In section 4 a modelling exercise regarding the effects of an 80 km/h speed zone at the motorway around Rotterdam is presented. This paper is concluded (section 5) with some insight into the health effects of traffic measures that can be derived from the EC assessment.

2 Study approach of measurement campaigns During the measurement campaigns hourly average concentrations of EC (actually its surrogate black carbon) were measured with MAAP monitors (Multi Angle Absorption Photometer model 5012). Measurements took place both near the road or crossing of interest and at a background location. In addition, real time data on wind direction and speed and traffic intensities by vehicle category were collected. On the basis of the prevailing wind hours were selected during which the monitoring location near the road was actually influenced by traffic emissions. For these hours the contribution of the traffic to the concentration of EC ("traffic contribution") was derived from the difference between the measured concentration at the traffic location and the background location. During part of the measurement campaign the traffic flow measure was in force. This is referred to as the “implementation situation”. The other part is characterized as the “reference situation”. When the average traffic contribution during the implementation situation is less than that during the reference situation, there is a positive effect on air quality on the monitoring location. However, in addition to the measure, there are also other factors that account for changes in the traffic contribution. This concerns the variation in traffic volume and composition and the meteorological conditions. A simple modelling WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

131

approach (linear scaling) was used to correct for these variations. In that way, misinterpretation of the effectiveness of the traffic flow measure on the local air quality is prevented.

3 Monitoring pilot studies 3.1 Helmond: different traffic control schemes at an inner urban junction Commissioned by the SRE, a cooperation of municipalities in the city region of Eindhoven, TNO conducted a measurement campaign to study the impact of two different traffic control schemes at a T-junction in the city of Helmond on the local air quality. The campaign lasted for 2 months and was carried out in the autumn of 2010. During the campaign, both schemes were alternated on a daily basis. One scheme aims at optimizing the traffic flow on a local level: stop times of traffic on all three directions are minimized. The other scheme aims at optimizing the traffic flow on the two directions that are part of the main road through the area: on the specific T-junction local traffic may be delayed. The object of the experiment was to investigate whether differences in traffic contribution to the local air quality between the schemes would occur at all and not to qualify the “best” scheme. The latter can not be assessed by measurements on a single location only. In the vicinity of the T-junction a suitable monitoring location was found (see photo in Figure 1). The location is influenced by traffic at southern wind directions. In the area south of the junction, the background monitoring location was set up.

Figure 1:

Monitoring location at the T-junction in Helmond (map: Google, photo: TNO).

After analysis of the measurement data including the correction for variation in traffic and meteorology, the following was concluded for weekdays between 7 a m. and 21 p m.: the differences in the traffic contribution to the concentration of EC at the monitoring location between the traffic control schemes are in the order of 20%. A first (limited) analysis of queues during peak periods gives confidence that the EC differences are indeed related to differences in the traffic WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

132 Air Pollution XIX flow (or “non-flow”) at the junction. A supplementary analysis of differences in the traffic flow needs to be carried out. 3.2 Arnhem: measures to enhance traffic flow at a busy provincial road In 2010 measures to enhance the traffic flow on the provincial road N325 near Arnhem were implemented. The measures include widening of the road and connecting traffic control installations at the N325. Also, the traffic control installations at the connection of the N325 with the highway A12 were introduced to the network. Commissioned by the province of Gelderland TNO conducted a measurement campaign to study the impact of these measures. At the provincial monitoring location along the N325 (see Figure 2) and in the neighbourhood to the north (background location) EC concentrations were measured.

Monitoring location at the N325 in Arnhem (map: Google, photo: TNO).

EC traffic contribution (ug/m3)

Figure 2:

5 4 3 2 1 0 0

Figure 3:

4

8

12 16 hour on the day

20

Average diurnal traffic contribution to the EC concentration (µg/m3) at the provincial road N325 in Arnhem during the reference period (February-March 2010).

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

133

In February and March 2010 reference measurements were carried out. They are representative of the former situation. The average diurnal concentration of the traffic contribution to the EC concentration is given in Figure 3. This shows the strong relationship between traffic emissions and EC: morning and evening rush hour traffic peaks coincide with peaks in the EC concentration. From January until April/May 2011 EC measurements in the new situation were carried out. A preliminary analysis of the EC measurements yields a decrease of the traffic contribution to the EC concentration in the order of 20 to 30%. In this case, the correction for variation in meteorology (wind speed) plays an important role in the (preliminary) outcome. Actually, measurements of just the traffic contribution to the EC concentration during the reference and implementation period were not that much different. However, the wind speed during the implementation period was significantly lower. The traffic contribution to the EC concentration is inversely proportional to the wind speed. Therefore, had the wind speed been similar to the reference period, the EC concentration contribution during the implementation period would have been significantly lower.

4 Modelling exercise On motorways through urban areas of the Hague, Utrecht, Amsterdam and Rotterdam 80 km/h speed limit zones with strict enforcement are implemented. These zones aim to protect residents along these highways against air pollution and noise. Recently questions about the usefulness and necessity of the 80 km/h zones have been raised. Therefore we investigated its significance for the health of the residents along the 80/h km zone in Rotterdam. With the URBIS model [2], we calculated the difference in EC concentrations with and without the 80 km/h speed limit for the existing 80 km/h zone in Rotterdam and for the whole ring road around Rotterdam. The result of this calculation for 2008 is shown in Figure 4. Figure 4 shows that through the 80 km/h zone the EC concentrations decrease up to a maximum of 0.5 µg/m3 depending on the distance to the highway and the traffic composition. For residents along the current 80 km/h zone in Rotterdam, the risk of premature death decreased by 1-3 months for 2,500 people and with 0-1 month for 6000 people. These health effect calculations are based on relationships between long term exposure to EC and premature mortality derived in epidemiological studies. For EC we used a relative risk of 1.06 expressed per 1 µg/m3 [3]. If the zone would have been implemented over the entire ring road around Rotterdam, the risk of premature death would decrease with 1-3 months for 4000 people and with 0-1 month for 15,000 people. This exercise shows that an 80 km/h zone significantly reduces the health risk for a quarter of the Rotterdam population living less than 500 m from the ring road.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

134 Air Pollution XIX

Figure 4:

The difference in yearly averaged EC concentration (µg/m3) with and without a speed limit of 80 km/h on highways around Rotterdam in 2008 (otherwise speed limits are 100 or 120 km/h).

5 Policy relevance These exercise and pilot studies illustrate that road traffic emissions of EC can be reduced with tens of percents with local measures such as volume reduction, environmental zoning, speed limitation and improvement of the traffic flow. For air quality this means a reduction of EC concentrations in the order of 0.1-1 µg/m3 depending on the effect of the measure and the distance of the location of interest to the traffic source. This decrease in EC yields a decrease in the risk of premature mortality by air pollution of approximately 1-6 months. This is a significant health gain for people living near busy road transport by relatively simple measures. Modelling and/or measuring of changes in EC concentrations provide policy makers with an instrument to assess the impact of traffic measures on air quality and health. Whether or not limit values for particulate matter are met.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

135

Acknowledgements The authors would like to thank the province of North Brabant, the city region of Eindhoven, the city of Helmond, the province of Gelderland and Peek Traffic for their support and contribution to the pilot projects.

References [1] Ntziachristos L and Samaras Z. EMEP/EEA emission inventory guidebook - COPERT4. www.eea.europa.eu/publications/emep-eeaemission-inventory-guidebook-2009/part-b-sectoral-guidance-chapters/1energy/1-a-combustion/1-a-3-b-road-transport.pdf [2] Beelen R., Voogt M., Duyzer J., Zandveld P. and Hoek G. Comparison of the performance of land use regression modelling and dispersion modelling in estimating small-scale variations in long-term air pollution concentrations in a Dutch urban area. Atmospheric Environment 44: 46144621, 2010. [3] Janssen N.A.H., Hoek G., Lawson-Simic M., Fischer P., Bree van L., Brink van H., Keuken M.P., Atkinson R., Brunekreef B. and Cassee F. Black carbon as an additional indicator of the adverse health effects of combustion particles compared to PM10 and PM2.5. Submitted to Environmental Health Perspectives, 2011

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

This page intentionally left blank

Air Pollution XIX

137

AMEC multigas passive sampler: a green product for cost-effectively monitoring air pollution indoors and outdoors H. Tang, L. Burns, L. Yang & F. Apon AMEC Earth and Environmental, Alberta, Canada

Abstract Climate change, sustainable development, and greenhouse gas are several of the hot topics in the world. Saving our limited resources, reducing consumption, and waste are emergent tasks facing the world. As a result, a new generation of passive sampling technology – multi-gas passive sampling system (MGPS) has been developed and reported here. This paper will demonstrate the cost effective unique features of the MGPS compared with many normal passive samplers (NPS). Cross contamination problems have been comprehensively studied and reported in this paper. Keywords: air pollution, passive sampler, air monitoring.

1 Introduction Air pollution indoors and outdoors has become a health issue in the world. Scientific and social interest in monitoring air pollutants indoors and outdoors is increasing. Thus, saving our limited resources, reducing consumption, and reducing waste are emergent tasks facing the world environmental business. Many monitoring technologies for air pollutants have been developed and subsequently improved in the past few decades. Due to its cost effective and more convenient to use, passive sampling technology is becoming more and more popular. In the past decades, many different types of passive samplers have been developed. The first passive sampler in the world was used by Fox [1] in 1873 for monitoring ozone concentrations. Since then, passive samplers have been developed for monitoring air pollution in the ambient environment (including vegetation canopies study), work place, and indoor environment, which include WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110131

138 Air Pollution XIX air pollutants such as SO2 [2, 3], NO2 [4, 5], H2S [6, 7], NOx [8], O3 [9], VOC [10], aldehyde [11] etc. The sampling rate is a key parameter related to the correct measurement of air pollutants using passive samplers. Active samplers have a known sampling rate, which is the pump’s flow rate. The passive sampler’s sampling rates depend on many factors such as temperature, relative humidity, wind direction, wind speed, sampler’s structure, collection media etc. If a passive sampler can be used in all climate conditions, a fixed passive sampling rate obtained from laboratories cannot be used for ambient studies anywhere and anytime in the world. It would be highly unreasonable to expect that a passive sampler’s sampling rate would be the same when temperatures change from -30o C to +30o C and relative humidity change from 90% to 15%. Therefore, the key factor for using passive samplers is how to determine their sampling rates. Tang et al. have reported using equations to address the problems [12–15] Field applications proved that Tang’s approach was practically useful in many cases. Krupa and Legge [16] have summarized the passive samplers into different types, such as badge, diffusion tube with filter absorption or solid absorption, and adsorption cartridge etc. The all season passive sampling system (ASPS) designed by Tang et al. [12, 13, 15] and Tang and Lau [14] (Figure 1) is a mixture of bandage and cartridge which can be reused many times; a rain shelter is designed to hold three passive samplers for triplicate study.

Figure 1:

The all-season passive air sampling system designed by Tang.

So far, in all the badge and cartridge type of passive samplers described above, each one can only be used to collect one air pollutant (such as SO2, NO2, NOx, H2S, O3, NH3 etc) or a group of air pollutants (such as volatile organic compounds (VOC) or aldehyde and ketone). In practice, field studies are requested to monitor several air pollutants in replicated in order to increase confidence level. For example, if 4 air pollutants are monitored at the same time and the same location in triplicate, 3 rain-shelters and 12 passive samplers plus several blanks will be used. This operation is tedious. In order to reduce cost and save our environment, several scientists have studied different ways to address the above problems. Tang et al. have tested a collection medium which can be used to simultaneously sample HF, NO2 and SO2 (Tang 2010). In the Ogawa passive sampler, due to two separate sampling chambers in each side, thus, it can

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

139

be easily installed two different collection media. A study for sampling NO2 and NOx together has been reported [17]. A new generation of passive sampling technology – multigas passive sampling system (MGPS) has been developed and validated. In the MGPS, one to several collection media can be packed into one passive sample body. Compared to the normal passive samplers (NPS), the MGPS is environment friendly, more cost-effective, more convenient to use, more accurate, and more flexible. This paper will report the features of MGPS and the field study results.

2 The multigas passive sampling system 2.1 Principle of passive sampler It is well known that a passive (or diffusive) sampler is a device which is capable of taking samples of gas or vapor pollutants from air at a rate controlled by a physical process such as diffusion through a static air layer or permeation through a membrane. The collected amount of an air pollutant by a passive sampler can be derived from Equation (1). t

Q  CDA  x

(1)

where Q is the amount of the air pollutant collected by the passive sampler, t is the sampling time. From Equation (2), it can be seen that the collection amount is proportional to the collection medium area. Theoretically, if only one quarter of the collection medium is used in the same passive sampler, the passive sampler should only collect one quarter amount of the air pollutant collected by the whole filter. Equation (2) is the principle of the multigas passive sampler. 2.2 Multigas passive sampling system Different from the NPS that use one collection medium for collecting a single or a group of air pollutants such as active charcoal for volatile organic compounds (VOC), the multigas passive sampling system (MGPS) (Figure 2) uses one passive sampler body to pack several different collection media at the same time to collect several air pollutants. In the MGPS, VOC or aldehyde is considered as one (group) air pollutant. The AMEC MGPS passive sampler body with a newly designed insert is shown in Figure 3. The MGPS can be used for both indoor, ambient air quality and personal exposure studies. The insert can be used to install 4 different collection media as the first layer in the passive sampler body. For more pollutants, a second layer or more layers can also be used. In this paper, we only report the first layer study.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

140 Air Pollution XIX

Figure 2:

Figure 3:

AMEC multigas passive sampling system.

AMEC passive samplers for indoor, ambient and personal exposure uses.

The advantages of the MGPS include: -

Environment friendly, which can save materials, chemicals, wastes, etc., cost effective, convenient to use, which can save labor for field jobs and passive sampler management, flexible for clients, in which one to several collection filters can be installed based on clients’ need, more accurate.

Table 1 lists comparisons of sampling 4 air pollutants in triplicate by using MGPS and NPS. 2.3 Collection media preparation and sample analyses The following air pollutants were studied in this report: SO2, NO2, O3, and H2S. The collection media preparations were following References 15, 20, 17, and 23 respectively. All the chemicals used in this study were purified grade (Fisher Scientific, Nepean, CA). WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Table 1:

141

Four air pollutants collected in triplicate by NPS and MGPS.

ITEM No. for NPS No. for MGPS No. Saved % Saved Rain shelter 4 1 3 75 Passive body 12 3 9 75 Diffusion Barrier 12 3 9 75 Collection filter 12 3 9 75 Chemicals* 12 3 9 75 Waste generated** 12 3 9 75 Field installation 12 3 9 75 Average 75 *Comparison of chemicals implies total chemicals used for one collection medium as one unit. In this case, 12 units are used for NPS; only 3 are used for MGPS. **Comparison of waste generated uses extraction volume for one collection medium as one unit. In this case, 12 collection media are used by NPS, which generate 12 unit wastes; but only 3units are generated in MGPS. Analyses of extractions of SO2, NO2, and O3 collection media were used IC. The H2S collection medium extraction was analyzed by a filter fluorometer. 2.4 Field validation The MGPS passive samplers were installed in six locations in Alberta (Figure 4). Duplicate or triplate passive samplers and duplicate field blanks were used. One location was at the Alberta Environmental (ANEV) industrial monitoring site in Edmonton (EIMU) which is equipped with a NOx continuous analyzer (TECO Model 42, Thermo Environmental Instruments Inc., Franklin MA), an SO2 analyzer (TECO 45C), an O3 analyzer (TECO 49), a H2S analyzer (TECO Model 45C), a temperature measurement device (Model 41372 Campbell Scientific Inc., Logan UT), a relative humidity measurement device (Model 41372 Campbell Scientific Inc., Logan UT), and a wind speed monitoring device (Wind Flo 540, Athabasca Research Ltd., Edmonton, AB). The other five locations were at the Parkland Airshed Management Zone (PAMZ) monitoring site near Red Deer (RD), West Central Airshed Society (WACS) monitoring stations near Breton (BT) and Carrot Creek (CC), and ANEV industrial monitoring site in Calgary (CIMU) and Lethbridge. Those stations were equipped with similar devices as in the EIMU. Except in BT and CC, the rain shelters were fastened using an outside bracket so that the passive samplers were at almost the same elevation as the inlet for the air pollutant continuous analyzers (Figure 5). In BT and CC, the rainshelters were installed in the fences about 2 meters above the ground.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

142 Air Pollution XIX

Figure 4: 30 25

Alberta locations for installing MGPS. NO2 Analyzer MGPS Equation

ppb

20 15 10 5 0 1

2

3

4

5

Study ID

Figure 5:

Comparison of O3 concentrations in EIMU.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

143

3 Results and discussion 3.1 Different collection filter size studies and sampling rate determination In order to evaluate Equation (2), filters with 37mm diameter and quarter size of the 37mm filters prepared for sampling SO2 and H2S were packed in the MGPS and installed in EIMU. Results are shown in Table 2. Table 2:

Collection quantities of different size filters in the MGPS.

Filter size 37 mm filter (A) Quarter 1 of A Quarter 2 of A Quarter 3 of A Quarter 4 of A Sum of 4 quarter filter

SO2 µg 1.8 0.5 0.5 0.5 0.6 2.1

H2S ng 129.9 37.2 36.3 34.5 35.1 142.9

The data in Table 2 demonstrate that reducing size of collection filter in the MGPS is applicable. The collection efficiencies inside the MGPS among the four quarter filters are almost the same since the relative standard deviations are only 3 and 4% for SO2 and H2S respectively. The collection quantities of 37 mm filters for both SO2 and H2S are a little bit lower compared to the sum of 4 quarter filters. Besides analytical method deviations, the other reason might be the increase of the quarter filters’ surface area after the 37 mm filters were cut to quarter size. The above results also indicate that the MGPS sampling rate can be calculated by the equations published by Tang before. For example, the ozone sampling rate for ASPS is shown in the following equation [14]. RS = 14.8T1/2 + 0.259 RH + 0.275 WSP – 197

(2)

where RS is the ozone sampling rate, ml/min; T is average temperature of the sampling period, K; RH is average humidity (%); WSP is average wind speed, cm/sec, if WSP>130, then WSP = 130. A quarter of the ASPS sampling rate theoretically can be the pollutant sampling rate in the MGPS. The sampling rate calculation equations for NO2 and H2S have been changed since the collection media for those pollutants were different compared to references 13 and 15. Actually the MGPS sampling rate determinations now are using a new approach which is named as “Integrative Passive Network Data Management” (IPN). The new approach makes the passive results more reasonable and accurate. 3.2 Practical quantitative detection limit Tang et al. have reported practical quantitative determination limit (PQDL) for the all season passive sampling system before. Those PQDL were based on WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

144 Air Pollution XIX laboratory filter blank studies. For example, it was found that the pooled standard deviation was 0.6 µg of nitrate per ozone collection filter based on a 24hour exposure. The practical quantitative detection limit, thus, can be taken as 6 µg per filter (10 times the standard deviation). This is equivalent to exposure of the passive sampler to 3 ppb O3 in the atmosphere for 24 hours. If the exposure period were increased to one month (30 days), the method practical quantitative detection limit for O3 in the atmosphere would be about 0.1 ppb. The collection filter’s area in the MGPS is quarter of the filter in the ASPS. It is easy to obtain the same PQDL through reducing the extract volume to be used in the MGPS filter. For example, the extraction volume for ozone collection filter in the ASPS is 20 ml of DI water. If the volume is reduced to 5 ml in the MGPS, the PQDL for the MGPS will be the same as for the ASPS. 3.3 Interference Numerous reagents have been checked for possible interference in the MGPS. It is found that if carefully choosing collection media and separating the media using different methods, there will be no substantial interferences. A number of studies were conducted in the AMEC laboratory to study the interference among the different collection media. The studies were conducted by using one MGPS passive sampler packed all collection media and only one collection medium (duplicate) packed in one passive sampler. Studies found that several chemicals could cause interferences to other pollutants’ detection. For example, based on properties of chemicals used to collect air pollutants, nitric acid in the H2S paper is expected to cause interference problems since it can generate nitric acid vapor, which can react metal parts in the passive sampler and can be adsorbed by filter papers. In the ozone passive samplers, ozone is reacted with nitrite in the collection paper, and the react product is nitrate. The nitrate concentration in the ozone passive sampler is used to calculate ozone concentration in air. Therefore, nitric acid can directly cause positive interference for ozone collection. Laboratory filter blank studies reflected the theoretical analysis above. Table 3 lists nitric concentrations measured in different ozone collection filter blanks. Table 3: Test No.

Nitric concentrations of different ozone collection filters.

Blank filters*

Stored time (Day)

Nitrate concentration (ug/filter)

Error %

Room Refrigerator (21ºC) (-4ºC) 1 A 14 0.56 0.54 4 2 B 14 1.30 0.55 136 3 C 21 0.65 0.61 6 4 D 21 0.96 0.56 71 5 E 21 1.26 0.55 129 *Blank filters were from different MGPS passive samplers. For detail, refer to the following paragraph. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

145

In Table 1, A blank filters were from MGPS passive samplers packed into only two ozone collection filters. Therefore there were no significant differences of nitric concentrations found between passives stored at room temperature and refrigerator. B blank filters were from MGPS passive samplers packed into two ozone collection filters and two H2S collection filters (coated with silver nitrate with 0.01 N nitric acid) in one sampler. It is clear that nitric acid vapor had adsorbed in the ozone collection filters; but at cooled temperature, the interference was not substantial. C, D, E blank filters were from ozone collection filters packed into MGPS passive samplers together with H2S filters coated with solutions with 0.001 N, 0.004 N, and 0.01 N nitric acid respectively. Although all ozone filters stored at refrigerator had not been affected by H2S filters, nitrate concentrations in the ozone filters stored at room temperature did indeed decrease along with the decrease of nitric acid concentrations in the coating solutions. A new H2S passive sampler without nitric acid is being continuously developed in the AMEC Centre for Passive Sapling Technology. 3.4 Field study results 3.4.1 EIMU studies Many MGPS studies have been conducted in EIMU. For example, several studies results for O3 are listed in Table 4. The study periods, meteorological conditions, calculated sampling rates, MGPS sampling rates, and relative errors in each study are also listed in the tables. Reid [18] conducted a study in 2000 using ASPS for monitoring SO2 concentrations in the Northern Rocky Mountain foothill of BC Canada. He found the SO2 concentrations were substantially different between continuous analyzer results and the ASPS results. But both the SO2 concentration trends kept the same. In this study, we met the same problems. Figure 5 shows comparisons of O3 concentrations obtained by analyzers, calculated from equations, and from the MGPS. It can be found that there are good agreements between results from Table 4: #

Location Day

1 2 3 4

*

EIMU EIMU EIMU EIMU

6 3 14 17

5

EIMU

21

Study results in EIMU for O3 by MGPS. Date

Oct 3-9 Oct 26-29 Dec 3-17 Dec 17Jan 3 Jan 3-23

RH

T ºC

% 61 55 79 82

8 7 -9 -9

79

-8

WSP Cal. RS* Km/h cm/min 10 24.1 11 23.8 8 22.9 7 23.1 7 23.1

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

MGPS Rs cm/min 21.5 19 24 21

Error

22

4.8

% 10.8 20.2 -4.8 9.1

146 Air Pollution XIX analyzers and MGPS, but difference between analyzers and equations although the concentration trends are the same, in which when temperature decreased the concentrations of NO2 increased and the concentrations of O3 decreased. A further discussion for the sampling rate will be continued in the next section. 3.4.2 Comparison of EIMU, CIMU, BT, CC, LG and RD studies In February 2008, the MGPS were installed in EIMU, CIMU, CC, BT, LG and RD. Table 5 summarized the weather conditions, calculated sampling rates using equations, MGPS sampling rates and the relative errors in each parameter. Table 5: ITEM Pass body Diffusion barrier Collection filter Field job Waste generated

Comparison of NPS and MGPS.

NPS 95 95 95 95 95

MGPS 34 34 24 34 24

SAVE % 64 64 75 64 75

ppb

It can be seen that the weather conditions from EIMU to CIMU in the study period had no substantial difference. Therefore, the calculated sampling rates from pollutant to pollutant did not vary a lot. But the concentrations obtained by using equations had relatively large difference compared to the results obtained from analyzers and the MGPS (Figure 6). It is well known that passive sampling rates depend on many factors. In addition to temperature, relative humidity, and wind speed, many other factors such as atmospheric pressure, local terrain, chemicals in the atmosphere etc. also play important roles. Tang et al. simplified the sampling rate calculation through only using temperature, relative humidity, and wind speed, in many cases, it did provide a useful tool for accurately monitoring air quality; in some other cases it might generate large deviation, which has been discussed before. We are conducting more studies to address the problems. 40 35 30 25 20 15 10 5 0

Analyz. MGPS Equa.

LG-H

CIMU-H

RD-H

EIMU-H

LG-O

BT-O

CC-O

CIMU-O

RD-O

EIMU-O

LG-N

BT-N

CC-N

CIMU-N

RD-N

EIMU-N

LG-S

BT-S

CC-S

CIMU-S

RD-S

EIMU-S

Station

Figure 6:

Concentration comparisons of four air pollutants obtained by analyzer, MGPS and equation.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

147

4 Applications The MGPS has been used in many projects across Canada and in the world indoor and outdoor. One project – Southeast Saskatchewan Airshed Association (SESAA) is located in southern Saskatchewan. There are many human being activities in the 38,000 km area including oil and gas industries, power generation, agriculture, transportation etc. Air quality is a big concern by local communities. 30 AMEC passive sampler stations were installed in the area (Figure 7) monitoring SO2, NO2, O3, and H2S. The AMEC multigas passive samplers have been used in the airshed for almost 4 years. Very reasonable results have been obtained.

Figure 7:

AMEC passive stations in the SESAA.

This project also demonstrates the cost effective feature of MGPS (Table 5). Compared to NPS, MGPS saves about 70% of major costs.

5 Conclusions A new generation of passive sampling technology – multigas passive sampling system (MGPS) has been developed. Field studies and applications have proved that the MGPS is environment friendly, more cost effective, more convenient to use, and more accurate. The MGPS is a new useful tool in the air monitoring sector for indoor and ambient atmosphere.

References [1] Fox, C.B.; Ozone and Antozone, J. and A. Churchill, London. 1873. [2] Palmes, E.D.; Gunison, A.F.; Personal monitoring device for gaseous contaminants, Am. Ind. Hyg. Assoc. 34, pp. 78-81, 1973.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

148 Air Pollution XIX [3] Leaderer, B.P.; Koutrakis, P.; Wolfson, J.M.; Sullivan, J.R.; Development and evaluation of a passive sampler to collect nitrous acid and sulphur dioxide, J. Expos. Environ. Epidemiol. 4, 503-511. [4] Gair, A.J.; Penkett, S.A.; Ovola, P.; Development of a Simple Passive Technology for the Determination of Nitrogen Dioxide in Remote Continental Locations, Atmos. Environ. 25, pp. 1927-1939, 1994. [5] Mulik, J.D.; Williams, D.; Development of a Simple Passive Technology for the Determination of Nitrogen Dioxide in Remote Continental Locations, Proceedings of the 1986 EPA/APCA symposium of measurement of toxic air pollutants, Raleigh, NC pp. 61-79, 1986. [6] Kring, E.V.; Damrell, D.J.; Henry, T.J.; Demoor, H.M.; Basilio, A.N.; Simon, C.E.; Laboratory validation and field verification of a new passive colorimetric air monitoring badge for sampling hydrogen-sulfide in air, Am. Ind. Hyg. Assoc. J. 45, pp. 1-9, 1984. [7] McKee, E.S.; Mcconnaaughey, P.W.; Laboratory validation of a passive length-of-stain dosimeter for hydrogen-sulfide, Am. Ind. Hyg. Assoc. J., 47, pp. 475-481, 1986. [8] Ogawa & Company USA Inc. No-NO2 simultaneous sampling protocol, June 1994. [9] Grosjean, D.; Mohamed, W.M.; A Passive Sampler for Atmospheric Ozone, JA & WMA, 42, pp. 169-173, 1992. [10] Bamberger, R.L.; Esposito, G.G.; Jacobs, B.W.; Podolak, G.E.; Mazur, J.F.; New passive sampler for organic vapour, Am. Ind. Hyg. Assoc. J. 39, pp. 701-708, 1978. [11] Brown, V.M.; Crunp, D.R.; Gadiner, D.; Gavin, M.; Assessment of a passive sampler for the determination of aldehydes and ketones in indoor air, Environ. Technol. 15, pp. 679-685, 1994. [12] Tang, H.; Brassard, B.; Brassard R.; Peake, E.; A new passive sampling system for monitoring SO2 in the atmosphere, FACT, 1, pp. 307-315, 1997. [13] Tang, H.; Lau, T.; Brassard B.; Cool, W., A new all-season passive sampling system for monitoring NO2 in air, FACT, 6, 338-345. [14] Tang, H. and Lau, L., A new all season passive sampling system for monitoring ozone in air, Environ. Monit. Assess. 65, 129-137. 1999. [15] Tang, H.; Sandeluk, J,; Lin L,; and Lown W.; A new all-season passive sampling system for sampling H2S in air, The Scientific World, 2, pp. 155168. 2002. [16] Krupa, S.; Legge, A.; Passive sampling of ambient, gaseous air pollutants: an assessment from an ecological perspective, Environmental Pollution, 07, pp. 31-45. 2000, [17] Higuchi, K.; Schaeffer, D.R.; Hirano, K.; Advanced monitoring method for air environment by Ogawa Passive, in Proceedings of Passive sampling workshop and symposium, Reston, VA, April 2007. [18] Reid, P.; The role of passive ambient monitoring for sulphur dioxide in the northern Rocky Mountain foothills, in the proceedings of CPANS conference, Edmonton, AB, Canada, 2001.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

149

Influence of natural and anthropogenic sources on PM10 air concentrations in Spain M. S. Callén, J. M. López & A. M. Mastral Energy and Environment Department, Instituto de Carboquímica (ICB-CSIC), Spain

Abstract Particulate matter samples less than or equal to 10 m (PM10) were collected by using a high-volume air sampler during cold and warm seasons at two different areas in Spain: a rural area which was considered a “non-polluted” area and an urban city, Zaragoza (Spain) in which vehicular traffic and small industries were the potential pollution sources. The PM10 samples were analyzed to determine their organic (polycyclic aromatic hydrocarbons (PAH) by gas chromatographymass spectrometry mass spectrometry) and inorganic (ions: anions and cations, by ionic chromatography and by inductively coupled plasma optical emission spectroscopy (ICP-OES)) composition. Higher PAH and ions concentrations were obtained in the urban area during the cold season when compared to the warm season and these concentrations were always higher than the ones obtained in the rural area. Fuel combustion sources associated with coal, natural gas, vehicular traffic and biomass combustion were the major anthropogenic PM10 pollution sources obtained by principal component analysis (PCA) in the urban area although natural sources associated with marine aerosol were also contributing to this PM10. Cluster analysis corroborated these sources and allowed classifying samples as a function of the meteorological variables, PAH and ion concentrations. Keywords: air pollution, PM10, PAH, anthropogenic sources, anions, cations, PCA, cluster analysis.

1 Introduction Particulate matter is considered one of the main air pollutants, whose origin can be attributed to natural and to anthropogenic sources, at higher proportions. Due WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110141

150 Air Pollution XIX to its different nature, organic and inorganic pollutants can be found. Into the organic pollutants, PAH constitute a special group of pollutants which are formed by two or more fused aromatic rings only containing C and H. The main concern of studying PAH is related to their mutagenic and/or carcinogenic character [1]. Other groups of pollutants which can be included in the PM are inorganic pollutants and ions of inorganic and/or organic nature. Nitrate and sulphate constitute two of the main secondary inorganic components of aerosol and they do not only affect climate change but also predominantly control the aerosol acidification and conductivity via acid rain precursors, mainly sulphur and nitrogen oxides. Therefore, only knowing and characterizing the pollution sources, especially those of anthropogenic character, will be possible to abate PM pollutions and to take further control. This work is based on the characterization of the PM10, in particular PAH and ions in two different sampling sites: one mainly affected by anthropogenic sources (urban area) and another one considered as “non-polluted” area. A comparison of the pollutants in both sampling sites was carried out for both seasons: warm and cold seasons. The study of the main pollution sources was also carried out by using source apportionment techniques, such as principal component analysis and statistical tools based on cluster analysis in the most polluted area corresponding to the urban sampling site.

2 Experimental 2.1 Sampling description The study was performed in the city of Zaragoza (ZGZ), located in the NorthEast of Spain (41º39´49.38´´N; 0º53´16.68´´W), by using a Graseby Andersen high-volume air sampler provided with a PM10 cut off inlet to collect particulate phase in a Teflon-coated, fibre-glass filter (0.6 m pore size; 20.5 cm x 25.5 cm) [2]. Traffic pollution and industrial activities were the main potential pollution sources of this place. The second site was Torrelisa (Huesca) (PIR) (42º27´36´´N; 0º10´48´´E) localized in the Pyrenees Mountain and considered as representative of biogenic sources. Samples were collected during two sampling periods: warm and cold seasons collecting a total of 30 samples for each sampling site during 2008 and 2009. More details regarding the sampling site were given in previous articles [2, 3]. 2.2 Ions analysis 1/8 of the filter was extracted by ultrasonic bath for 30 minutes in 15 mL of Milli-Q water. The analyses of anions (Cl-, NO3-, SO42-, PO43-) were carried out by ion chromatography (Metrohm) with conductivity detection and a Metrosep A Supp 5 anion column. The analyses of cations (Na+, K+, Mg2+, Ca2+) were carried out by ICP-OES (JY 2000 Ultrace Horiba). More details regarding the analytical protocol are shown in a previous article [4].

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

151

The sulphate concentration of marine origin, mSO42- was determined indirectly by considering the Na+ soluble concentration according to the ratio: mSO42-/Na+= 0.25 in weight [5]. The non-sea-salt-sulphate, nmSO42-, generally of anthropogenic origin, was obtained by subtracting this value to the SO42concentration. The identification of compounds was based on the retention time and the quantification was made using calibration curve acquired with the external standards. 2.3 PAH analysis The following PAH (phenanthrene (Phe), anthracene (An), 2+2/4methylphenanthrene (2+2/4MePhe), 9-methylphenanthrene (9MePhe), 1methylphenanthrene (1MePhe), 2,5-/2,7-/4,5-dimethylphenanthrene (Dimephe), fluoranthene (Fth), pyrene (Py), benzaanthracene (BaA), chrysene (Chry), benzobfluoranthene (BbF), benzokfluoranthene (BkF), benzoepyrene (BeP), benzoapyrene (BaP), indeno1,2,3-cdpyrene (IcdP), dibenza,hanthracene (DahA), benzoghiperylene (BghiP) and coronene (Cor) were quantified according to previous publication using gas chromatography mass spectrometry mass spectrometry (GC-MS-MS) [2] with the internal standard method after Soxhlet extraction with dicloromethane. 2.4 Meteorological variables The meteorological data in ZGZ and PIR were provided daily by the Estación Experimental de AULA-DEI (CSIC) and by DGA, respectively. 2.5 Statistical tools The SPSS Version 15.0 statistical package was used as statistical tool: a) to evaluate the correlation between variables by using Pearson correlation coefficients, b) to test for significant differences in seasonal air concentrations by using a parametric test (Student’s t-test of independent samples for each sampling place), c) to run principal component analysis (PCA) in order to find the main pollution sources and multiple linear regression and d) to apply cluster analysis for evaluation of data pattern and classification of data.

3 Results and discussion 3.1 PM10 and PAH concentrations A brief summary of results obtained for each sampling site regarding PM10 and PAH concentrations [2, 6] during the warm and cold seasons for both sampling sites are shown in Table 1. Higher PM10 and PAH concentrations were found during the cold season for both sampling sites with higher concentrations in ZGZ than in PIR. This finding is consistent with the localization of the sampling sites: the rural place was in the WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

152 Air Pollution XIX Table 1:

ClNO3nmSO42SO42Ca2+ K+ Na+ PM10 Total PAH

Average concentration and standard deviation of ions (ng/m3), PM10 (g/m3) and total PAH (ng/m3) for the whole period, for the warm and for the cold periods in PIR and in ZGZ. PIR ZGZ PIR ZGZ PIR ZGZ PIR ZGZ PIR ZGZ PIR ZGZ PIR ZGZ PIR ZGZ PIR ZGZ

Whole period

Warm

Cold

20.97(54.80)

1.06(2.50)

40.87 (73.42)

655.00 (693.20)

123.26(110.43)

1186.73(614.25)

1364.27(1614.63)

122.65(78.76)

2605.90(1443.92)

3026.62(3904.35)

1525.31(988.70)

4527.92(5076.35)

545.46(384.89)

568.69(503.50)

522.23(232.27)

1790.22(1913.07)

1680.18(675.23)

1900.26(2664.43)

594.44 (397.82)

598.46(516.01)

590.42(251.82)

1959.94(1926.57)

1754.77(678.67)

2165.11(2671.64)

217.82(133.87)

175.18(144.89)

260.46(111.37)

830.59(342.16)

882.80(234.54)

778.39(426.21)

142.03(80.26)

88.87(45.16)

195.18(72.53)

241.34(324.70)

103.88(43.74)

378.79(419.51)

199.78(143.53)

126.79(54.88)

272.78(168.42)

678.88(584.63)

298.38(108.86)

1059.39(621.20)

14(8.0)

10(7.4)

19(6.2)

26(16)

21(5.5)

30(21)

0 266(0.235)

0.097(0.117)

0.455(0.176)

5.30(5.98)

2.20(1.45)

8.40(7.18)

mountain with plenty of vegetation and minimum anthropogenic sources whereas the urban area was characterized by high traffic, population and predominance of anthropogenic sources. Meteorological factors related to low solar radiation, low temperature and low photochemical decomposition in addition to contribution of domestic heating during the cold season could explain these higher pollutant concentrations. Regarding the PAH concentrations in PIR, most of the individual PAH concentrations were below the detection limit indicating a lower impact of anthropogenic sources, especially during the warm season. 3.2 Ions analysis Regarding the concentrations of ions in both sampling sites, it is remarkable to say that nitrate ion was found to be the largest component present in the samples at both sampling sites with the following decreasing order NO3->nmSO42>Ca2+>Na+>K+ (Table 1). This trend was only modified during the warm season in the rural area, being nmSO42- the majority ion corroborating the enhanced SO42- concentration due to the increased photochemical activity reported by different authors [7, 8] during the warm season. On the contrary, the winter period with low temperature and high relative humidity favoured the formation of NO3- [9, 10]. For each ion, the concentration was always higher in ZGZ than

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

153

in PIR and it was favoured by the higher impact of anthropogenic activities carried out in ZGZ city. The main sources of NO3- are due to the use of fertilizers in agricultural zones close to the sampling point that could be resuspended and the oxidation of NOx. This last one is the most significant precursor of nitrate, proceeding of coal combustion, biomass burning, industry and traffic emissions, all of them local sources that play a major role near the monitoring site in ZGZ. NO3- represented the 10% in weight of the PM10 and it showed a positive and significant correlation at 99% level with nmSO42-, total PAH, PM10, Ca2+ and K+ indicating that they came from a similar type of emission source, which could be associated with fossil fuel combustion including vehicular emissions. The main possible forms of nitrates could be Ca(NO3)2 and KNO3. In addition to its marine origin, sulphate in ambient air mainly comes from photochemical oxidation of sulphur containing precursors such as SO2, H2S, CS2 where sulphur dioxide is the largest contributor [11]. The precursor of sulphate aerosol may be released from industrial source, coal combustion as well as diesel combustion and oil-fired power plant profiles, which contain higher sulphate concentrations than most of the coal-fired power plant, all of them local pollution sources affecting ZGZ. Other main source is due to mineral aerosols. In ZGZ, nmSO42- represented a 6.6% in weight of the PM10 and it was correlated at 99% level with NO3- and K+ indicating that the main form of non marine sulphate was K2SO4. The average percentage of nmSO42- versus SO42- was 84.3% indicating that its main source had anthropogenic origin. Cl- represented a 3% in weight of the PM10 and it was correlated at 99% level with Na+ showing that sea salt was the most important source of this ion in ZGZ. Regarding cations, it is remarkable to say that Mg2+ was not detected in any sample. K+ represented 0.9% in weight of the PM10 in ZGZ and its main sources, in addition to the marine aerosol and soil resuspension, are anthropogenic sources: forest fires, agricultural or biomass burning [12]. It has been used as a biomass burning tracer in source apportionment [13, 14] and it could also indicate the contribution of waste incineration plants and paper fabrics located in the surroundings of the city. Its positive and significant correlation with PM10, total PAH, NO3- and nmSO42- indicated its possible main anthropogenic origin. Ca2+ (3.7% in weight of the PM10) whose origin is mainly mineral due to soil resuspension [15, 16] and activities related to construction, is the majority cation in both sampling sites except in the cold season where Na+ (3.5% in weight of the PM10) is the predominant. Regarding the seasonal behaviour of ions, it was observed that in ZGZ Cl- and + Na were statistically different at 99% level whereas NO3- and K+ were statistically different at 95% level indicating that sources producing these pollutants followed a different trend in both seasons. In PIR the concentrations of all ions were higher during the cold season with the exception of nmSO42- that was slightly higher during the warm season. NO3(6.8% in weight of the PM10), K+ (0.97% in weight of the PM10) and Na+ (1.4%) were statistically different during the warm and cold seasons. As WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

154 Air Pollution XIX happened in ZGZ, NO3- was statistically correlated at 99% level with total PAH, PM10, Ca2+ (1.4% in weight of the PM10), K+ and also with Na+ so that the main forms of nitrates were Ca(NO3)2, KNO3 and NaNO3. The main difference with regard to ZGZ was related to NO3- and nmSO42- (4.0% in weight of the PM10), which were not correlated indicating that they originated from different secondary sources. The non-marine sulphate (4.0% in weight of the PM10) was mainly correlated to Ca2+ as CaSO4 and it could have a natural origin. The mass ratios of SO42-/NO3- were also used to distinguish the predominance of stationary and/or mobile sources of sulphur and nitrogen in the atmosphere [17, 18]. Whereas in ZGZ this mean ratio was 1.11, indicative of mobile sources, in PIR this ratio was 2.47 indicating that stationary sources were predominant. This seems to be reasonable by considering the affluence of traffic in ZGZ compared to PIR (background rural area). This difference was more remarkable when the ratios were calculated for the two different seasons. In this case, the ratio for PIR during the warm season was 4.16 whereas it was 0.40 in the cold season. In ZGZ these ratios were 1.4 and 0.8 for the warm and cold seasons, respectively. Therefore the influence of stationary sources was reflected in PIR during the warm season and it could be attributed to forest fires and agriculture fires produced at the dry South Mediterranean countries during those dates. 3.3 Principal component analysis Because ZGZ was the sampling point with the highest impact of pollution sources, PCA was applied to samples taken in ZGZ for both sampling periods in order to discern the main pollution sources. Data were auto-scaled before PCA by subtracting the mean concentration of each variable from the observed concentration followed by division of the standard deviation of the concentration of each variable. All factors with eigenvalues over 1 were extracted according to Kaiser-Meyer-Olkin KMO and Bartlett’s test of sphericity and were rotated using the Varimax method. A total of 19 variables were taken and three factors were obtained explaining 88% of the variance. The communalities were also higher than 0.6 for all variables and those variables with low coefficients (<0.6) were not used in interpreting the principal components. Figure 1 shows the loading plot for individual components of PAH and ions in ZGZ and the possible pollution sources. The first factor explained 71.3% of the variance and was associated to most of the PAH: Phe, An, 2+2/4MePhe, Fth, Py, BaA, Chry, IcdP+DahA, BghiP and Cor. Ph, Fth, Py, BaA and Chy are markers of coal combustion [19] with predominance of Ph, Fth and Py [20]. In ZGZ, coal is the most important energy source and is used widely for industrial purposes, especially in the power industry and as fuel in the electric power supply from coal-burning power plants. It is remarkable the presence of three thermal power stations in Aragón. Moreover natural gas is also used as main domestic heating, which is reflected by a high fraction of BaA and Chry [21]. In addition, BghiP [22] and IcdP [23] are tracers of vehicular emissions. Therefore, this factor was related to fossil fuel combustion and vehicular emissions.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

PC2 (10.7%)

Air Pollution XIX

Traffic and light oil burning

Secondary inorg. compounds

1 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0

Phe, An,2+2/4MePhe, Py Natural gas Fth, BbjkF,BeP,BaP and heavy oil IcdP+DahA, BghiP, Cor combustion Cl-, Na+

Marine component

NO3-, nmSO42Coal combustion

0

0,2

0,4

155

0,6

Ca2+ K+ 0,8

1

BaA, Chry

PC1 (71.3%)

Figure 1:

Principal component analyses loading plot for individual components of PAH and ions in ZGZ.

The second factor explained 10.7% of the variance and was related to BbjkF, BeP, BaP, NO3-, nmSO42- and K+. BeP and BaP are the fingerprints of light oil burning [24] and K+ is a marker of biomass combustion [13, 14]. Nevertheless, BbF, BeP and BaP are also found in the ambient air at the traffic source [25]. Therefore this factor could be associated with industrial emissions including oil and wood burning as well as traffic source. Finally, the last factor associated with Cl- and Na+ explained 6.2% of the variance and was related to marine component. Figure 1 shows the loading plot for individual components of PAH and ions in ZGZ and the possible pollution sources. Afterwards, the absolute principal component scores APCS were used based on the PCA factor scores to derive quantitative estimates of source contributions and source profiles [26]. A 16%, 47% and 4% of the PM10 was explained according to the factors mentioned previously. A 32% of the experimental particulate matter was not explained and the correlation between the experimental and the modelled particulate matter was 0.87 with a slope of 0.99 showing a good correlation between the experimental and the modelled PM10. Figure 2 shows the temporal evolution of each factor along the sampling dates. It is observed that the maximum in PM10 concentrations were obtained during the cold season and corresponded to maximum in the industry+traffic factor. Results were also evaluated by cluster analysis obtaining two different dendograms by using variables (Figure 3) and by using dates (Figure 4). The dendogram created when clustering the variables using Ward’s method (squared Euclidean distance, variables normalized using z-scores) confirmed the presence of three main groups with different sub-divisions. In this way, in the first cluster, two different sub-groups can be considered: on the one hand, Fth, Py, An, 2+2/4Mephe, Phe, PAH of low molecular weight and 3-4 rings which would be attributed to fossil fuel combustion. On the other hand, BaA, Chry, IcdP+DahA, BghiP and Cor which would be mainly associated with natural gas and traffic emissions. A second cluster would be constituted by Cl- and Na+ WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

156 Air Pollution XIX

Figure 2:

Contribution of each factor obtained by PCA/APCS (ng/m3) and PM10 concentration (ng/m3) for each sampling date in ZGZ.

Figure 3:

Cluster dendogram based on PAH and ions using Ward’s method and Euclidean distance in ZGZ (variables normalized using Zscores).

associated with marine component. The third major group would be constituted of, BeP, BaP, BbjkF, NO3-, K+, nmSO42- and Ca2+ including industrial and vehicular emissions corroborating the different pollution sources obtained previously by PCA. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Figure 4:

157

Cluster dendrogram based on dates using Ward’s method and Euclidean distance in ZGZ (Samples taken during consecutive days; Num 1=23/05/2008; Num 15=08/06/2008; Num 16=13/01/2009; Num 30=27/01/2009).

A new dendogram was obtained when the same method was also used to group the different cases as a function of the sampling dates (Figure 4). Two main clusters were distinguished. A first cluster grouping most of the samples: 2, 8, 10-15, 26-29, 22, 24 and 1, 5, 16, 23, 25, 30, 7, 9, 4, 6, 3 and a second cluster corresponding to samples 17-21 (14-18 January 2009), which were associated to the highest PAH concentrations obtained during the cold season. Trying to get the variables, which were statistically different among these clusters, an independent t-test was performed and temperature, solar radiation, relative humidity, total PAH, Cl-, NO3- and K+ were the variables statistically different at 95% level. In this way, typical conditions of winter season: low temperature, low solar radiation and high relative humidity favoured the accumulation of the WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

158 Air Pollution XIX following pollutants: PM10, total PAH, Cl-, NO3- and K+. Inside the first cluster, two subgroups could be distinguished. Samples 2, 8, 10-15 were taken during the warm period and samples 26-29 were collected during the cold period. The variables statistically different between both groups were the ions concentration, the temperature and the solar radiation where the lower temperature and the solar radiation registered for samples 26-29 favoured the high pollutants concentration. The second subgroup inside the first cluster corresponded to samples 1, 5, 16, 23, 25, 30, 7, 9, 4, 6, 3 (first subgroup), 22 and 24 (second subgroup). No variables statistically different between the two subgroups were found although samples 22 and 24, which were collected during the cold period presented meteorological conditions typical of winter season with lower average temperature than the other subgroup. The second main cluster corresponded to samples 17-21 (14-18 January 2009), which were associated to the highest PAH concentrations during the cold season. Inside this group, two different sub-groups could be distinguished. Samples 17-19 and 20-21. By applying a t-test to independent samples, no statistical differences were found among variables, the only difference was that samples 17-19 were taken during weekdays whereas samples 20-21 were taken during the weekend in which San Juan bonfires festivals were celebrated.

Acknowledgements Authors would like to thank Aula Dei-CSIC (R. Gracia) for providing the meteorological data, to the Gobierno de Aragón (DGA) for partial financial support, to the Ministry of Science and Innovation (Spain) through the project CGL2009-14113-C02-01 for funding as well as the Fondo Europeo de Desarrollo Regional (FEDER). J.M. López would also like to thank CSICSpanish Government for his Ramón and Cajal contract.

References [1] Luch, A., (ed). The Carcinogenic Effects of Polycyclic Aromatic Hydrocarbons. Imperial College Press, ISBN 1-86094-417-5, London, 2005. [2] Callén, M.S., López, J.M. & Mastral, A.M., Characterization of PM10bound polycyclic aromatic hydrocarbons in the ambient air of Spanish urban and rural areas. Journal of Environmental Monitoring, 13, pp. 319327, 2011. [3] Callén, M.S., de la Cruz, M. T., López, J.M., Murillo, R., Navarro, M.V. & Mastral, A. M., Long-range atmospheric transport and local pollution sources on PAH concentrations in a South European urban area, Fulfilling of the European Directive. Water, Air and Soil Pollution, 190, pp.1-4, 2008. [4] Callén, M.S., de la Cruz, M.T., López, J.M., Murillo, R., Navarro, M.V. & Mastral, A.M., Comparison of receptor models for source apportionment of the PM10 in Zaragoza (Spain). Chemosphere, 76, pp. 1120-1129, 2009.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

159

[5] Duce, R.A., Arimoto, R., Ray, B.J., Unni, C.K. & Harder, P.J., Atmospheric trace elements at Enewetak Atoll: 1, Concentrations, sources and temporal variability. Journal of Geophysical Research, 88, pp. 53215342, 1983. [6] Callén, M.S., López, J.M. & Mastral, A.M., Seasonal variation of Benzo(a)pyrene in the Spanish airborne PM10. Multivariate linear regression model applied to estimate BaP concentrations. Journal of Hazardous Materials, 180, pp. 648-655, 2010. [7] Khan, Md. F., Hirano, K. & Masunaga, S., Quantifying the sources of hazardous elements of suspended particulate matter aerosol collected in Yokohama, Japan. Atmospheric Environment, 44, pp. 2646-2657, 2010 [8] Husain, L. & Dutkiewicz, V.A., A long-term (1972-1988) study of atmospheric SO4: regional contributions and concentration trends. Atmospheric Environment, 24(A), pp. 1175-1187, 1990. [9] Park, S.S. & Kim, Y.J., Source contributions to fine particulate matter in an urban atmosphere. Chemosphere, 59, pp. 217-226, 2005. [10] Mariani, R.L. & de Mello, W.Z., PM2.-10, PM2.5 and associated watersoluble inorganic species at a coastal urban site in the metropolitan region of Río de Janeiro. Atmospheric Environment, 41, pp. 2887-2892, 2007. [11] Khoder, M.I., Atmospheric conversion of sulphur dioxide to particulate sulphate and nitrogen nitric acid in an urban area. Chemosphere, 49, pp. 675-684, 2002. [12] John, K., Karnae, Crist, K., Kim, M. & Kulkarni, A., Analysis of trace elements and ions in ambient fine particulate matter at three elemental schools in Ohio, Journal of Air Waste Management Association, 57, 394406, 2007. [13] Marmur, A., Mulholland, J.A. & Russell, A.G., Optimized variable source profile approach for source apportionment. Atmospheric Environment, 41, pp. 493-505, 2007. [14] Sheffield, A.E., Gordon, G.E., Currie, L.A. & Riederer, G.E., Organic elemental and isotopic tracers of air pollution sources in Albuquerque, NM. Atmospheric Environment, 28, pp. 1371-1384, 1994 [15] Zhuang, H., Chan, C.K., Fang, M. & Wexler, A.S., Size distribution of particulate sulfate, nitrate and ammonium at a coast site in Hong Kong. Atmospheric Environment, 33, pp. 848-853, 1999. [16] Hedge, P., Pant, P., Naja, M., Dumka, U.C. & Sagar, R., South-Asian dust episode in June 2006: aerosol observations in the central Himalayas. Geophysical Research Letters, 34, L23802, 2007. [17] .Yao, X., Fang, M. & Chan, C.K., The water-soluble ionic composition of PM2,5 in Shanghai and Beijing, China. Atmospheric Environment, 36(26), pp. 4223-4234, 2002. [18] Xiu, G., Zhang, D., Chen, J., Huang, X., Chen, Z., Guo, H. & Pan, J., Characterization of major water-soluble inorganic ions in size-fractionated particulate matters in Shanghai campus ambient air. Atmospheric Environment, 38, pp. 227-236, 2004.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

160 Air Pollution XIX [19] Duval, M.M. & Friedlander, S.K. Source resolution of polycyclic aromatic hydrocarbons in the Los Angeles atmosphere, EPA-600/2-81-161, PB82121336, U.S. EPA, Washington, DC, 1981. [20] Zuo, Q., Duan, Y.H., Yang, Y., Wang, X.J. & Tao, S., Source apportionment of polycyclic aromatic hydrocarbons in surface soil in Tianjin, China. Environmental Pollution, 147, pp. 303-310, 2007. [21] Simcik, M.F., Eisenreich, S.J. & Lioy, P.J., Source apportionment and source/sink relationships of PAHs in the coastal atmosphere of Chicago and Lake Michigan. Atmospheric Environment, 33, pp. 5071-5079, 1999. [22] Boonyatumanond, R., Wattayakom, G., Amano, A., Inouchi, Y. & Takada, H., Reconstruction of pollution history of organic contaminants in the upper Gulf of Thailand by using sediment cores: First report from Tropical Asia Core (TACO) project. Marine Pollution Bulletin, 54, pp. 554-565, 2007. [23] Larsen, R.K. & Baker, J.E., Source apportionment of polycyclic aromatic hydrocarbons in the urban atmosphere: a comparison of three methods. Environmental Science and Technology, 37, pp. 1873-1881, 2003. [24] Bari, M.A., Baumbach, G., Kuch, B. & Scheffknecht, G., Wood smoke as a source of particle-phase organic compounds. Atmospheric Environment, 43, pp. 4722-4732, 2009. [25] Lee, W-J., Wang, Y-F., Lin, T-C., Chen, Y-Y., Lin, W-C., Ku, C-C. & Cheng, J-T., PAH characteristics in the ambient air of traffic source. The Science of the Total Environment, 159, pp. 185-200, 1995. [26] Thurston, G.D. & Spengler, J.D., A quantitative assessment of source contributions to inhalable particulate matter pollution in metropolitan Boston. Atmospheric Environment, 19, pp. 9-25, 1985.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

161

Infrared imaging Fourier-transform spectrometer used for standoff gas detection M. Kastek, T. Piątkowski & H. Polakowski Institute of Optoelectronics, Military University of Technology, Poland

Abstract The article presents the detection of gases using an infrared imaging Fouriertransform spectrometer (IFTS). The Telops Company has developed the IFTS instrument HyperCam, which is offered as short or long wave infrared device. The principle of HyperCam operation and methodology of gases detection has been shown in the paper, as well as theoretical evaluation of gases detection possibility. The calculation of an optical path between an IFTS device, cloud of gases and background has been also discussed. The variation of a signal reaching the IFTS caused by the presence of a gas has been calculated and compared with the reference signal obtained without the presence of a gas in IFTS’s field of view. Verification of theoretical result has been made by laboratory measurements. Some results of the detection of various types of gases have also been included in the paper. Keywords: gas detection, hyperspectral detection, imaging Fourier-transform spectrometer, stand-off detection.

1 Introduction Fourier-transform spectrometers (FTS) are renowned instruments, particularly well-suited to remotely provide excellent estimates of quantitative data. Many authors have presented how they are using conventional (non-imaging) FTS to perform quantification of distant gas emissions. Amongst others, we should mention the important contributions made by Harig and Matz [1] and Hang et al. [2]. His group performed many measurement campaigns for which they obtained excellent results by ensuring proper modeling of the scene and by paying attention to understand (and to take into account) the instrument signature. Other groups developed similar approaches however limiting their study to optically WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110151

162 Air Pollution XIX thin plumes [3, 4] to dedicated instruments performing optical subtraction [5, 6], or introducing Bayesian algorithms to maximize the use of a priori information [7]. On the other hand, detection and quantification activities with an imaging Fourier-transform spectrometer (IFTS) have been first presented by Spisz et al. [8]. The present paper deals with the remote gas identification and quantification from turbulent stack plumes with an IFTS. It presents first the modeling that is required in order to get an appropriate understanding of both the scene and the instrument. Next it covers the methodology of the developed quantification approach. Finally, results are presented to demonstrate the capabilities and the performances of the remote gas quantification by using hyperspectral data obtained from the Telops HyperCam. The latter is described in much detail in prior references [9–12].

2

Instrumentation

The infrared imaging Fourier-transform spectrometer (IFTS) – HyperCam – used on experiment was built by Telops, Inc. This IFTS has 320 x 256 pixel Mercury Cadmium Telluride (MCT) focal plane arrays (FPA) with a 6º x 5º FOV. The FPAs are Sterling cooled to provide good noise figures in a field ready package. Spectral information is obtained using a technique called Fourier Transform Infrared Radiometry (FTIR).

Figure 1:

HyperCam sensor during the measurement.

FTIR is a classical interference based technique applied to gas spectroscopy that uses a Michelson interferometer to mix an incoming signal with itself at several different discrete time delays. The resulting time domain waveform, called an interferogram, is related to the power spectrum of the scene through the Fourier transform. An interferogram for each pixel in an image is created by imaging the output of the interferometer onto a focal plane array and collecting data at each discrete time delay. Advantages of using an FTIR sensor over a grating based (filter) system include higher resolution for equal cost and the absence of misalignment of different color images due to platform motion. However, FTIR does have the disadvantage of producing a slower frame rate than filter based systems because twice as many points are taken for the same WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

163

number of spectral points. For atmospheric tracking of gases however, the HyperCam sensor has sufficient frame rate [8, 11]. Data for this collection can be taken at 0.25–150 cm-1 spectral resolution between 830 cm-1 (12 μm) and 1290 cm-1 (7.75 μm) at a frame rate of 0.2 Hz. In addition to the infrared data, visible imagery was taken using a firewire camera that is boresighted to the IR sensor. Fig. 1 shows a picture of a HyperCam sensor during the measurement. The sensor is controlled by field computer and data is stored on a RAID drive to guarantee integrity of data [12]. The package of software Reveal is prepared for control and registration of data from HyperCam, and for calibration of raw data, and for analysis of these results. Fig. 2 presents the windows of software Reveal Pro for control parameters of HyperCam and registered data. Fig. 3 shows the part of software for data analysis.

Figure 2:

Software Reveal Pro for control HyperCam sensor.

(a)

(b)

Figure 3:

Software for data analysis: registration visible picture (a), analysis of spectral data (b).

3 Model of process detection The physical problem of interest in the present paper concerns imaging Fouriertransform spectrometers taking measurements from a ground-based platform. Nevertheless, the approach described herein directly applies, with some adjustments, to airborne-based measurements. In the typical setup the instrument WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

164 Air Pollution XIX is looking towards a stack releasing gases. The scene, from the instrument pointof-view, is implicitly inhomogeneous. Vertical variations of temperature and gas concentrations are necessarily present both in the released gases and in the atmospheric background. From the small instrument instantaneous field-of-view (IFOV) perspective, the instrument does see a small solid angle in a given direction, thus sensing a narrow pencil of air mass. Within the illustrated disposition, any line of pixels therefore provides a complete section of the plume. Modelling the experimental conditions illustrated in Fig. 3 may be done by considering fundamentally two distinct aspects. The first deals with the whole contribution of the released gases through the atmosphere. And the second is devoted to understanding how the instrument itself does add its own signature into the measurement [11, 12]. Atmosphere Gas Ta 2 Tg  g  g

Background Tb  b IFOV

Figure 4:

Atmosphere Ta 1

IFTS

Scheme of a measuring process and notations accepted for gas analysis in atmosphere.

In order to detect the gas plume, it is essential to begin with an underlying physics-based model for the hyperspectral data. This allows us to mathematically describe the radiance, defined as the flux per unit projected area per unit solid angle, incident at the sensor. This at-sensor radiance, L(λ) is a summation of the contributions of numerous terms including both reflected solar and emitted thermal effects. In the LWIR, emissive terms dominate and reflective effects are generally considered negligible [12]. For ground-based observation of a gaseous plume, the at-sensor radiance can be described as: L  λ L  λ (1) L λ   L  λ where Lg(λ) is the self-emitted radiance from the gas plume at wavenumber λ (cm-1), Lb+g(λ) is the self-emitted radiance from the background transmitted through the plume, and La(λ) is the scattered self-emitted atmospheric effects (also known as upwelled radiance). The self-emitted radiance from the gaseous plume can be expanded such that L  λ

τ λ ε λ Β λ, T  

(2)

where τ1(λ) is the atmospheric transmittance from the gas plume to the sensor, εg(λ) is the effective emissivity of the plume material, Tg is the temperature of the gas plume in Kelvin, and WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX π

B λ, T

λ

165 (3)

λ T

is Planck’s blackbody curve where k is Boltzmann’s constant, c is the speed of light, h is Planck’s constant, and λ is wavenumber (cm-1). Therefore the radiance seen at the sensor due to the gas plume is a function of the gas plume’s emissivity, its temperature, and the atmosphere between it and the sensor. For the background, the radiance seen at the sensor is more complicated. This radiance is defined as  λ τ λ τ λ τ λ ε λ Β λ, T   (4) L where τg(λ) is the transmittance through the gas plume, τ2(λ) is the atmospheric transmittance between the background and the gas plume, and εb(λ) is the emissivity of the background material. In this case, the photons from the background travel through the atmosphere to the gas plume. The photons then travel through the gas plume which has its own transmittance function. The photons that make it through the gas plume travel the final distance to the sensor. Equation (4) can be simplified by assuming the gas plume is optically thin. Under this assumption, the Beer-Lambert Law states the gas plume transmittance is a function of the gas plumes effective emissivity such that 1

τ λ

ε λ

(5)

From equations (2), (4), and (5), (1) and using some simple algebra, (6) can be rewritten such that L λ

τ λ τ λ Β λ, T τ λ ε λ Β λ, T τ λ τ λ ε λ Β λ, T L λ

(6)

To solve (6), we have to know something about the atmospheric transmission, the temperature of the background, the temperature of the gas plume, the emissivity signature of the background, the scattering (or up welled radiance) term, and the emissivity signature of the plume. Unfortunately, we only know the emissivity signature of the gas plume and the ambient air temperature; therefore, we must make some simplifying assumptions to use (6). The first assumption we make is that the atmospheric terms are negligible and/or are the same for all terms because all objects are relatively close to the camera. For the same reason, the second assumption we can make is that the up welled radiance term is also minimal and may be dropped. These assumptions lead to the following simplified model: L λ

τ λ Β λ, T

ε λ Β λ, T

  ε λ Β λ, T

(7)

Equation (7) can be expressed in terms of apparent emissivity if an estimate of can be calculated. To do this, the temperature at each pixel can be , calculated such that λ (8) T λ  

L λ

where L(λ) is the radiance signature for the pixel under test, k is Boltzmann’s WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

166 Air Pollution XIX constant, c is the speed of light, h is Planck’s constant, and λ is wavenumber (cm-1). Using (3) and (8), an estimate for the temperature and Planck’s blackbody curve can be created for every pixel in the scene. Each pixel can be divided by its estimated blackbody curve to obtain an estimate of its apparent emissivity: ε λ

ε λ

B λ,T B λ,T

ε λ

ε λ

(9)

Another assumption we can make is that the factor multiplying plume emissivity can be roughly approximated as a function of the temperatures between the plume and the background: B ,T

 

ε λ

B ,T

f T

T   

10  

These assumptions of course could be removed and more robust atmospheric modeling and compensation implemented. However, we have found that, for the relatively close stand-off distances considered, the current simplified model yields good results. Thus using the assumption in (10), a new apparent emissivity model is described by ε λ

ε λ f T

T

ε λ

(11)

Equation (11) suggests the background can be modeled by one emissivity; however, the background is typically comprised of multiple materials each with their own emissivity. In such a case, the apparent emissivity of the background is ε λ

∑Q β ε λ

(12)

where Q is the number of unique background emissivity signatures, εj is the emittance for the jth background signature and is the amount of jth background material. Using (12), the final apparent emissivity model is ε λ

ε λ β

∑Q β ε λ

(13)

where β

f T

T

(14)

Therefore, each pixel in the scene is processed in the following manner. First, the apparent temperature is estimated using (8). From this estimate, the Planck’s blackbody radiation curve is generated according to (3). The pixel is divided by its blackbody radiation estimate to obtain (13). It is this equation can be used at the detection algorithm of gases. The algorithm of automation gases detection was under development and tests [10, 12]. The influence of emissivity coefficient on measurement results can also be compensated by adopting several methods from pyrometric non-contact temperature measurements. There are many known methods [13–16] for such compensation and an algorithm taking into account real emissivity values should be included in automatic gas detection procedures. In the long infrared wave spectral range (LWIR), for chemicals in the gaseous form, scattering processes are not expected to contribute, so they are neglected. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

167

Note that this assumption would not be valid when aerosols are present. In general the spectral absorption coefficients may be dependent upon light polarization. However atmospheric absorption (and emission) being due to molecules, polarization dependence is not expected. Accordingly, the present description does not include polarization effects. The absorption coefficient is made of a linear combination over the gas compounds of the atmosphere (linear mixture model). These coefficients are obtained from gas reference databases of infrared signatures (often in absolute absorbance’s) as a function of the gas temperature (e.g. HITRAN molecular spectroscopic database [17] and Pacific Northwest National Laboratory (PNNL) Spectral Library [18]). The transmission of the cloud gas  g is computed from the spectral properties of the chemical species using Beers’ Law:

 m (  )  exp  ki (  )Ci d  ,

(15)

where Ci is the average concentration of the chemical compound over the path length d and ki (  ) is the wavelength-dependent absorption coefficient. The sum over index i in Eq. (15) is over all spectrally relevant chemical species. For the computer simulation of transmission atmosphere with gas (e.g. methane) was used the PC MODWIN 3 v.1.0 computer program for simulation of transmission of atmosphere [19, 20]. IFTS technology can be used for the gas quantification but of course the methodology of this is now under the testing. This kind of algorithm used some of filtering methods and method comparison results of measurement with model results [21]. A measurement session was performed in the laboratory and in the open-space conditions to practically verify the possibility of applying IFTS-type spectroradiometer for gas detection. The HyperCam LWIR device was used for the experiment and four different gases were observed. During the laboratory experiments the ambient temperature and humidity were monitored and during open-space tests the wind speed and direction and atmospheric pressure were additionally recorded.

4 Laboratory tests The laboratory tests were performed in order to test the effectiveness of imaging Fourier-transform spectroradiometer as a tool for the measurement of spectral characteristics of single gases or gas mixtures. The real photo of the laboratory test stand used during experiments is depicted in Fig. 5. The measurements on a test stand presented in Fig. 5. were performed in order to determine the minimal gas concentration which can be detected by HyperCam LWIR spectroradiometer. The flow from gas cylinder was controlled by a precise gas pressure regulator and large-area blackbody was used to determine the value of thermal contrast between gas cloud and background. The applied blackbody has a radiative area of 11 x 11 inch and its temperature difference ΔT can be set with respect to ambient or temperature value measured WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

168 Air Pollution XIX

Figure 5:

The laboratory test stand.

by an external sensor. In our case, this external sensor was placed at the output of gas nozzle thus the thermal contrast between gas cloud and background (blackbody area) could be precisely set. The following gases were used during the measurements: NO2, CO2 and propane-butane mixture, all of those having absorption lines in far infrared spectral range. Theoretical absorption spectra for CO2 and NO2 gases calculated by HITRAN software are presented in Fig. 6. Those characteristics were calculated for the same environmental conditions as recorded during laboratory tests, i.e. ambient temperature 294.20 K, and using Middle Latitude Summer atmosphere model from commonly available database of atmospheric data from Modtran software. It can be seen from Fig. 6 plots that calculated spectra fit well inside the measurement range of HyperCam LWIR spectroradiometer and it can be expected that both gases should be detectable using this device. (a)

(b)

Figure 6:

The absorption CO2 (concentration 1%) (a), NO2 (concentration 1%) (b).

The applied gas pressure regulator provided the controlled gas flow, during the measurement session the flow was set at 5 mg/s which corresponds to gas concentration of 1% in the measurement zone observed by HyperCam LWIR. Spectra characteristics of gases were measured for the following values of thermal contrast: ΔT= 1 K, ΔT= 2 K, ΔT=3 K. In order to obtain best possible accuracy the spectral resolution was set at 0,75 cm-1 whereas the frame rate was 0.3 Hz and it resulted from chosen resolution of sub-frame windowing (limited area of an array used during measurements which also reduced the field of view). WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

169

Each measurement was repeated twice in order to avoid accidental errors. The results obtained from HyperCam LWIR spectroradiometer gave finally the spectral characteristics of selected gases for different measurement conditions. In Fig. 7 results of experiments on measurement of characteristics of CO2 and NO2 are presented. (a)

(b)

Figure 7:

The results of measure CO2 (concentration 1%) (a), NO2 (concentration 1%) (b) on laboratory tests.

5 Field experiments During the field tests the following gases was measured: CO2, propane-butane mixture and freon 134 (CH2FCF3 tetrafluoroethane). In each case gas concentrations were tested in order to verify the efficiency of gas measurements in the open space by HyperCam LWIR.

Figure 8:

The theoretical absorption CO2 for different concentration 1% and 5%.

Theoretical absorption spectra for two concentrations of CO2 (1% and 5%) were calculated by HITRAN software are presented in Fig. 8. As expected, higher gas concentration significantly increases the absorption and broadens its spectral range. During the experiment measured absorption characteristics of CO2 for two concentrations were measured too. The results of this measurement gives possibility to detect gases on natural environments used the HyperCam LWIR. During the field measurements the weather conditions were monitored by WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

170 Air Pollution XIX (a)

Figure 9:

(b)

The measure during field test on MUT area: spectroradiometer HyperCam (a), pipe with gas and plate as a background (b).

an automatic weather station. The measurement set-up with HyperCam LWIR spectroradiometer during field tests is shown in Fig. 9. As already mentioned, during the field tests the freon 134 gas was used. This composition it exhibits absorption bands in long wave infrared range, thus it can be detected using HyperCam LWIR. The recordings were made at two distances: 60 m and 10m, for two gas concentrations of 3% and 6%. The results are presented in Figs. 10 and 11. Additionally, some Matlab-based post-processing results are also presented. (a)

Figure 10: (a)

Figure 11:

(b)

The results of measure of freon 134 (concentration 3%) distance 60 m (a), freon 134 results of analysis in Matlab (b). (b)

The results of measure of freon 134 (concentration 6%) distance 10 m (a), freon 134 results of analysis in Matlab (b).

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

171

6 Conclusion After the analysis of result laboratory tests and tests field efficiency of the imaging Fourier-transform spectrometer technology was confirmed. During the experiment measurement of characteristics of gases were made with different conditions (different concentration) and during different environments conditions. For the over side the results of tests confirmed possibility to use this kind of spectroradiometers to automatically detection of gases, after the development method and software for it. The results presented in the paper are supported by realization of the Project is co-financed by the European Regional Development Fund within the framework of the 2. priority axis of the Innovative Economy Operational Programme, 2007–2013, submeasure 2.1. “The development of centres with high research potential” Contract no. POIG.02.01.00-14-095/09.

References [1] [2] [3] [4] [5] [6] [7] [8] [9]

Harig R., Matz G., “Toxic Cloud Imaging by Infrared Spectrometry: A Scanning FTIR System for Identification and Visualization”, Field Analytical Chemistry and Technology, 5(1-2), 75-90 (2001). Harig R., Matz G., Rusch P., “Scanning Infrared Remote Sensing System for Identification, Visualization, and Quantification of Airborne Pollutants”, Proc. of SPIE 4574, 83-94 (2002). Griffin M. K., Kerekes J. P., Farrar K. E., Burke H.-H. K., “Characterization of Gaseous Effluents from Modeling of LWIR Hyperspectral Measurements”, Proc. of SPIE 4381, 360-369 (2001). Burr T., Hengartner N., “Overview of Physical Models and Statistical Approaches for Weak Gaseous Plume Detection using Passive Infrared Hyperspectral Imagery”, Sensors, 6(12), 1721-1750 (2006). Lachance R. L., Thériault J.-M., Lafond C., Villemaire A. J., “Gaseous emanation detection algorithm using a Fourier transform interferometer operating in differential mode”, Proc. of SPIE 3383, 124 (1998). Thériault J.-M., “Passive standoff detection of chemical vapors by differential FTIR radiometry”, Technical Report Defence Research Establishment Valcartier (DREV) TR-2000-156 (2001). Heasler P., Posse C., Hylden J., Anderson K., “Nonlinear Bayesian Algorithms for Gas Plume Detection and Estimation from Hyper-spectral Thermal Image Data”, Sensors 7, 905-920 (2007). Spisz T. S., Murphy P. K., Carter C .C., Carr A. K., Vallières A., Chamberland M., “Field test results of standoff chemical detection using the FIRST”, Proc. of SPIE 6554, 655408 (2007). Farley V., Chamberland M., Lagueux P., Vallières A., Villemaire A., Giroux J., “Chemical agent detection and identification with a hyperspectral imaging infrared sensor”, Proc. of SPIE 6661, 66610L (2007).

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

172 Air Pollution XIX [10] Vallières A., Villemaire A., Chamberland M., Belhumeur L., Farley V., Giroux J., Legault J.-F., “Algorithms for chemical detection, identification and quantification for thermal hyperspectral imagers”, Proc. of SPIE 5995, 59950G (2005). [11] Chamberland M., Belzile C., Farley V., Legault J.-F., Schwantes K., “Advancements in field-portable imaging radiometric spectrometer technology for chemical detection”, Proc. of SPIE 5416, 63-72 (2004). [12] Farley V., Belzile C.; Chamberland M., Legault J.-F., Schwantes K., “Development and testing of a hyper-spectral imaging instrument for field spectroscopy”, Proc. of SPIE 5546, 29-36 (2004). [13] Madura H., Kastek M., Piątkowski T., “Automatic compensation of emissivity in three-wavelength pyrometers”, Infrared Physics & Technology, Volume 51, Issue 1, July 2007, Pages 1-8. [14] Madura H., Kastek M., Sosnowski T., Orżąnowski T., “Pyrometric method of temperature measurement with compensation for solar radiation”, Metrology and Measurement Systems, Volume 17, Issue 1, 2010. [15] Bielecki, Z., Chrzanowski, K., Matyszkiel, R., Pia̧ tkowski, T., Szulim, M., “Infrared pyrometer for temperature measurement of objects of both wavelength- and time-dependent emissivity”, Optica Applicata, Volume 29, Issue 3, 1999, Pages 284-292 . [16] Madura, H. , Piątkowski, T., Powiada, E., “Multispectral precise pyrometer for measurement of seawater surface temperature”, Infrared Physics & Technology, Volume 46, Issue 1-2 SPEC. ISS., December 2004, Pages 6973. [17] Champion J.-P., Chance K., Coudert L. H., et al., “The HITRAN 2008 molecular spectroscopic database”, Journal of Quantitative Spectroscopy and Radiative Transfer 110, 533-572 (2009). [18] Sharpe S. W., Johnson T. J., Sams R. L., Chu P. M., Rhoderick G. C., Johnson P. A., “Gas-Phase Databases for Quantitative Infrared Spectroscopy”, Applied Spectroscopy 58(12), 1452-1461 (2004). [19] Kastek M., Sosnowski T., Orżanowski T., Kopczyński K., Kwaśny M., “Multispectral gas detection method”, WIT Transactions on Ecology and the Environment Volume 123, pp. 227-236 (2009). [20] Włodarski M., Kopczyński K., Kaliszewski M., Kwaśny M., MularczykOliwa M., Kastek M., “Application of advanced optical methods for classification of air contaminants”, WIT Transactions on Ecology and the Environment Volume 123, pp.237-247 (2009). [21] Tremblay P., Savary S., Rolland M., Villemaire A., Chamberland M., Farley V. et al., “Standoff gas identification and quantification from turbulent stack plumes with an imaging Fourier-transform spectrometer”, Proc. of SPIE Vol. 7673, 76730H (2010).

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

173

POPs in ambient air from MONET network – global and regional trends I. Holoubek, J. Klánová, P. Čupr, P. Kukučka, J. Borůvková, J. Kohoutek, R. Prokeš & R. Kareš RECETOX, Masaryk University, Brno, Czech Republic

Abstract The Stockholm Convention (SC) on Persistent Organic Pollutants (POPs) mentioned in Article 16 of the SC that its effectiveness shall be evaluated starting four years after the date of its entry into force. Global Monitoring Plan (GMP) has been developed with an objective of evaluating whether the POPs actually were reduced or eliminated on the global scale. As one of the key matrices for the global monitoring, an ambient air was selected. We are using two approaches for sampling – so called active using the high volume samplers, and passive air samplers (PAS) as new tools for the air quality monitoring. MONET programme (MONnitoring NETwork) is driven by RECETOX as the Regional Centre of the Stockholm Convention for the region of Central and Eastern Europe on the national scale (MONET-CZ, containing 37 sites including 15 backgrounds), and regional scales – the Central, Southern and Eastern Europe and Central Asia (MONET-CEECs), the Pacific Islands (MONET-PIs), the African continent (MONET-AFRICA) and newly the whole of Europe (MONET-Europe). Samples are collected every 28 days; it represents 13 samples from each site every year. Keyword: Stockholm Convention, persistent organic pollutants (POPs), ambient air monitoring.

1 Introduction – Stockholm Convention on POPs and their effectiveness evaluation The Stockholm Convention on Persistent Organic Pollutants (POPs) [1] entered into force on the May 17, 2004 and has currently 173 signatory parties (May 16, 2011). The main objective of the Stockholm Convention (SC) is to protect WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110161

174 Air Pollution XIX human health and the environment from persistent organic pollutants by reducing or eliminating their releases into the environment. A term “persistent organic pollutants” (POPs) represents several classes of organic chemicals including polychlorinated dibenzo-p-dioxins and furans (PCDDs/Fs), polychlorinated biphenyls (PCBs), organochlorine pesticides (OCPs) and other industrial and agricultural chemicals. Due to their wide distribution, ability to accumulate in abiotic matrices such as soils or sediments with high contents of organic matter, bioaccumulate in the biotic tissues, and potential harmful effects such as immunotoxicity, neurotoxicity, developmental toxicity, carcinogenicity, mutagenicity, and endocrine disruption potentials, POPs have remained at the centre of scientific attention for the last few decades. Polycyclic aromatic hydrocarbons (PAHs) are often included in this group of compounds because of their potential for long-range transport even though their physicochemical properties do not suggest the persistency and bioaccumulation potential. According to Article 16 of the Convention, its effectiveness shall be evaluated starting four years after the date of its entry into force, and periodically thereafter at every 6 years [1]. Each effectiveness evaluation should consist of reports and environmental monitoring information pursuant to paragraph 2 of Article 16, and national reports pursuant to Article 15 (reports on the measures taken by the Parties, and the effectiveness of those measures). For the first of these elements, the Guidance of Global monitoring was prepared and its main goal is development and implementation of the arrangements to provide comparable monitoring information on the presence of the chemicals listed in Annexes A, B and C of the Convention in ambient air and human milk, as well as their regional and global environmental transport. The objectives of the POPs Global Monitoring Plan are to evaluate whether the POPs actually were reduced or eliminated as requested in Articles 3 and 5 of the Convention which means that information on environmental levels of the chemicals listed in the annexes should enable detection of trends over time [1]. Therefore focus is upon monitoring of background levels of POPs at locations not influenced by local sources. Reliable identification of trends will require that statistical evaluation is carried out on the design of each national monitoring program contributing to the Global Monitoring Plan, to ensure that it is powerful enough to detect trends in time. In order to meet the objectives of the Global Monitoring Plan (support the preparation of regional reports of comparable information on environmental background levels), the guidance must be provided on how information is to be collected, analyzed, statistically treated, and reported [2].

2 Polyurethane foam BASED passive samplers for sampling of POPs in ambient air - background information As air pollution became an issue of great public health concern and the international conventions and new regulations introduced their demands, a pressing need to obtain more air pollutants including POPs data in a costWIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

175

effective way appeared. Global Monitoring Plan has been prepared for the purpose of the Stockholm Convention with the objective of establishing baseline trends at global background sites [2]. It was the main goal of the first step of this Global monitoring programme – the determination of basic global information in 2009. When developed parties are to conduct source inventories, identify ongoing sources, and provide environmental monitoring evidence that ambient levels of POPs are declining [2,3], developing countries in particular require cost-effective and simple approaches. The experiences from the first report in 2009 and including of newly adopted POPs, will be used for the second assessment in 2015. Based on the long-term work of many international experts, the Guidance for the Global Monitoring Plan [2], was prepared. As one of the key matrices, the ambient air was selected. We have two basic principles for the sampling of ambient air for POPs determination – so called active using the high volume air samplers and passive sampling. Since the high volume air samplers as expensive devices requiring reliable power supply as well as trained operators are not widely available, the air monitoring of POPs has only been conducted at limited number of sites. This was a reason, why in the last years, a lot of various new types of passive air samplers (PAS) as new tools for the air quality monitoring [3-6]. PAS represent a cheap and versatile alternative to the conventional high volume air sampling and they have been currently recommended as one of the methods suitable for the purpose of new long-term monitoring projects. They are capable of being deployed in many locations at the same time, which offers a new option for the large scale monitoring. As it provides information about longterm contamination of selected site, passive air sampling can be used as a screening method for semi-quantitative comparison of different sites with the advantage of low sensitivity to accidental short-time changes in concentration of pollutants. It was demonstrated that passive air samplers using polyurethane foam (PUF) filters are suitable to study vapour-phase air concentrations of POPs, particularly of more volatile compounds [7-10], and they were successfully applied as a tool for POPs monitoring on the global, regional, national levels and also local scale where can provide site- and source-specific fingerprints and they can be used to conduct screening surveys to help to identify the sources [9,10]. This tool based on experiences and results from the long term testing and broad applications was recognized as very useful and effective for the determination of temporal, seasonal and spatial trends on the global, regional and local scales [11]. On the other hand, due to the sensitivity of PAS to local effects, sampling site selection seems to be crucial for the success of such projects since small-scale variability in each region can exceed the continental variability. To develop a monitoring network, the local conditions, sources of contamination and environmental variables must be evaluated very carefully since only detail characterization of potential local effects for every sampling site can assure the successful selection of sites for larger (regional or global) scale monitoring. Performing more detailed local screening studies before designing the final network is advisable. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

176 Air Pollution XIX

3 RECETOX pilot studies Passive air samplers of this design were first introduced in the Czech Republic in 2002 during the European screening campaign performed by Lancaster University and focused on the atmospheric levels of POPs [7,12]. Since 2003, this research topic has been developed in the RECETOX in cooperation with Lancaster University and Environment Canada. RECETOX conceptual approach and contribution to wide application of this method was oriented to the long-term study of effects of environmental variables to applicability of this technique for the long-term monitoring and determination of temporal, seasonal and spatial trends on the global, regional and local scales. Samplers were calibrated against the high volume samplers in the field conditions, and their sensitivity, capacity, robustness, as well as an influence of the various meteorological parameters on their performance were assessed [13]. PAS have been continuously deployed in the regular atmospheric monitoring of POPs in Košetice station since 2003 side by side with active samplers to compare information derived from both techniques. A core network of the Czech Republic was significantly altered based on the evaluation of the results from the pilot study in 2005, This network consists from some urban and suburban areas, vicinity of the industrial sources mentioned in the SC as potential sources of POPs (chemical industry, cement industry, municipal and medical waste incinerators, remediation technologies), and the new set of background sampling sites was included in cooperation with the Czech Hydrometeorological Institute. This set consists of the mountain sites along the Czech borders and it was meant to evaluate the impact of the transboundary transport in this region. The new design was introduced and initiated in January, 2006. Thirteen 28-days samples were collected from each of 50 sampling sites [14]. This sampling period and frequency was/is used in main part of MONET sampling sites and campaigns. This monitoring network – MONET-CZ is still flexible and allows further improvements. At the same time, the backbone of the network allows performance and advanced interpretation of the short term spin-off case studies. The new design of the MONET-CZ sampling sites was introduced and initiated in January, 2006 [14]. This network was further reduced from 50 to 37 sites in January, 2007, freeing a capacity to perform detailed screening studies in 14 regions of the Czech Republic [15]. The feasibility of the long-term application of passive air samplers for the evaluation of persisting influence of the war damages on the atmospheric contamination of the Western Balkan region was assessed in this study. Results of this project were compared to those of previous high volume sampling campaigns (APOPSBAL Project) [16,17]. It was the first international RECETOX sampling campaign which was focused to the collecting an extended number of parallel air samples, to put the data from the short-term high volume sampling events into the right perspective, to gain more information about the spatial and temporal distribution of POPs, and to collect the samples from remote

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

177

places, a passive air sampling (PAS) campaign was organized in Croatia, Serbia, Bosnia, Herzegovina, and Kosovo in 2004. As Central, Southern and Eastern Europe is the region with a lack of data on the atmospheric POP, three screening campaigns were organized between 2006 and 2008 (MONET-CEECs). Sampling sites for the first phase of the MONETCEECs Project, have been selected in cooperation with the local partners in all participating countries [18, 19]. A philosophy was the same as for the model network in the Czech Republic: 5-20 sampling sites were selected per country (according to the size of each country) and they were monitored for 5 months. A background site was included in most countries as a potential candidate of background monitoring for the effectiveness evaluation of the Stockholm Convention. Whenever possible, gradient of other sites (rural, urban, industrial) was developed also to address the range of contamination, possible sources and spatial variations. Soil samples were collected from the air sampling sites as a part of the study. A design of the study was synchronized with the Czech passive air monitoring network (MONET-CZ) which provides continuous data. In addition to the Central and Eastern European region (CEEC), 26 sites from the African continent (MONET-AFRICA) and 21 sites from the Central Asia (former Soviet Union countries as a part of MONET-CEECs) were monitored in 2008 and 3 sites from the Pacific Islands between 2006 and 2007 (MONET-PIs). MONET-AFRICA now continues by the second phase (2010-2012) [20-22]. Previous RECETOX studies [10, 11] confirmed that PAS are sensitive enough to mirror even small-scale differences, which makes them capable of monitoring of spatial, seasonal and temporal variations. Passive samplers can be used for point sources evaluation in the scale of several square kilometres or even less - from the local plants to diffusive emissions from transportations or household incinerators - as well as for evaluation of diffusive emissions from secondary sources. While not being sensitive to short time accidental releases passive air samplers are suitable for measurements of long-term average concentrations at various levels.

4 Methods - sampler, sampling and analysis 4.1 Sampler description Passive air sampling device used in the pilot studies and MONET programme consists of two stainless steel bowls attached to the common axes to form a protective chamber for the polyurethane foam filter [9, 10]. The filter is attached to the same rod and it is sheltered against the wet and dry atmospheric deposition, wind and UV light. Exposure times between four and twelve weeks enable determination of many compounds from the POP group. Average sampling rate was estimated to be 3.5 -7 m3/day which roughly corresponds to 100-200 m3 of the air sampled during four weeks of deployment. Passive air samplers consisting of the polyurethane foam disks (15 cm diameter, 1.5 cm thick, density 0.030 g cm-3, type N 3038; Gumotex Breclav, Czech Republic) housed in the protective chambers were employed in this study. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

178 Air Pollution XIX Sampling chambers were prewashed and solvent-rinsed with acetone prior to installation. All filters were prewashed, cleaned (8 hours extraction in acetone and 8 hours in dichloromethane), wrapped in two layers of aluminum foil, placed into zip-lock polyethylene bags and kept in the freezer prior to deployment. Exposed filters were wrapped in two layers of aluminum foil, labeled, placed into zip-lock polyethylene bags and transported in cooler at 5°C to the laboratory where they were kept in the freezer at –18°C until the analysis. Field blanks were obtained by installing and removing the PUF disks at all sampling sites.

mounting bracket

air circulation

Figure 1:

stainless steel dome

PUF disk

Scheme of the passive air sampler.

4.2 Sample analysis All samples were extracted with dichloromethane in a Büchi System B-811 automatic extractor. One laboratory blank and one reference material were analyzed with each set of ten samples. Surrogate recovery standards (d8naphthalene, d10-phenanthrene, d12-perylene for PAHs analysis, PCB 30 and PCB 185 for PCBs analysis) were spiked on each filter prior to extraction. Terfenyl and PCB 121 were used as internal standards for polyaromatic hydrocarbon (PAH) and polychlorinated biphenyl (PCB)/ organochlorine pesticide (OCP) analyses, respectively. Volume was reduced after extraction under a gentle nitrogen stream at ambient temperature, and fractionation achieved on a silica gel column; a sulphuric acid modified silica gel column was used for PCB/OCP samples. Samples were analyzed using GC-ECD (HP 5890) supplied with a Quadrex fused silica column 5% Ph for PCBs: PCB 28, PCB 52, PCB 101, PCB 118, PCB 153, PCB 138, PCB 180, and OCPs: αhexachlorocyclohexane (HCH), β-HCH, γ-HCH, δ-HCH, 1,1-dichloro-2,2-bis(pchlorfenyl)ethylene (p,p´-DDE), 1,1-dichloro-2,2-bis(p-chlorfenyl)ethan (p,p´DDD), 1,1,1-trichloro-2,2-bis(p-chlorfenyl)ethan (p,p´-DDT), o,p´-DDE, o,p´DDD, o,p´-DDE, hexachlorobenzene (HCB), and pentachlorobenzene (PeCB). 16 US EPA polycyclic aromatic hydrocarbons were determined in all samples WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

179

using GC-MS instrument (HP 6890 - HP 5972) supplied with a J&W Scientific fused silica column DB-5MS [9, 11, 14, 15]. 4.3 Quality assurance / quality control Recoveries were determined for all samples by spiking with the surrogate standards prior to extraction. Amounts were similar to detected quantities of analytes in the samples. Recoveries were higher than 76 % and 71 % for all samples for PCBs and PAHs, respectively. Recovery factors were not applied to any of the data. Recovery of native analytes measured for the reference material varied from 88 to 103 % for PCBs, from 75 to 98 % for OCPs, from 72 to 102 % for PAHs. Laboratory blanks were under the detection limits for selected compounds. Field blanks consisted of pre-extracted PUF disks and they were taken on each sampling site. They were extracted and analyzed in the same way as the samples, and the levels in field blanks never exceeded 3% of quantities detected in samples for PCBs, 1% for OCPs, 3% for PAHs, indicating minimal contamination during the transport, storage and analysis [9, 10, 14, 15].

5 Results The passive samplers have been continuously co-employed with the high volume samplers as a part of the atmospheric monitoring of POPs in Kosetice since 2003. That allowed for the field calibration of PAS not only for the gas-phase chemicals but also for the particle-bound compounds. Five years of data are showing the same seasonal fluctuations as well as temporal trends as the results of the high volume monitoring (Fig. 2). 12000

PAS PAHs

Benzo(gh )perylene D benz(ah)anthracen e Indeno(123cd)pyrene

10000

Benzo(a)pyrene Benzo(k)fluoranthene

8000

ng filter 1

Benzo(b)fluoranthene Chrysene

6000 Benz(a)anthracene Pyrene

4000

Fluoranthene Anthracene

2000

Phenanthrene Fluorene Acenapthene

1.

10 . 24 03 .1 2. 0 17 3 .3 .0 4 9. 6. 04 7. 9. 0 1. 4 12 .0 23 4 .2 .0 18 5 .5 .0 10 5 .8 .0 2. 5 11 .0 25 5 .1 .0 19 6 .4 .0 12 6 .7 .0 4. 6 10 . 27 06 .1 2. 0 21 7 .3 .0 13 7 .6 .0 7 5. 9. 28 07 .1 1. 6. 07 3. 2 28 008 .5 .2 20 008 .8 . 12 200 8 .1 1. 20 08

0

Acenaphtylene Naphthalene

Sampling period

Figure 2:

PAHs in ambient air (ng filter-1), Kosetice observatory, 2003-2008 (passive sampling).

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

180 Air Pollution XIX Example of the results from the first 3 years of the MONET network for PAHs (background sites only) is presented in Fig. 3 (winter months are shown in yellow, summer months in red color). We can see there the typical PAHs seasonal trends showing the winter maxima. But, we can also recognize, the variability of meteorological conditions. Temperatures during strong winter 2005/2006 were much lower than the next winter and, as a reflection of these winter conditions, we can expect higher emissions of air pollutants including PAHs connected with the local heating systems. These higher emissions of PAHs are reflected in higher levels of PAHs in ambient air in all sites during the winter 2005/2006 in comparison with winter 2006/2007 [11].

Figure 3:

Passive air sampling 2006-2008, MONET-CZ, PAHs (ng filter-1).

For DDTs, fall concentrations were consistently the highest. Seasonal maxima of HCHs varied from site to site. Interestingly, HCB concentrations were higher in the winter rather than summer suggesting that combustion sources (seasonal residential heating) were more significant source of HCB than summer evaporation from diffusive secondary sources. No clear temporal trends have been observed yet.

6 Future of MONET In case of the CEEC, it has been recognized that knowledge on Western European POP levels would greatly improve the understanding to CEEC data. Although Western Europe is formally a part of WEOG (Western Europe and Others Group) region, and the rest of Europe is reported under the CEEC, it is desirable to harmonize the monitoring activities in both parts of Europe to gain systematic information on the levels and trends of the atmospheric pollution in this continent. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

181

As the EMEP stations participating in the previous MONET activities seem to be the best candidates for the long-term background monitoring in many CEE countries (Czech Republic, Estonia, Latvia, Moldova, Slovakia, Slovenia), an agreement has been made between RECETOX and EMEP to organize a followup study as a joint activity of these partners. MONET stations from the previous campaigns were complemented by new stations from Western Europe to provide a good geographical coverage (Fig. 4 – MONET-Europe). The goal is to maintain sustainable PAS monitoring at the majority of sites. That would greatly improve the understanding to the sources, fate and transport of POPs in Europe and provide rich information for the modelling databases. At the same time, it would create necessary synergies between the Stockholm Convention and Convention on Long-Range Transboundary Air Pollution.

Figure 4:

European passive air monitoring network.

RECETOX contributes very actively to the development and implementation of the Guidance for Global monitoring plan. For comparison – few years ago, only very limited number of sites with systematic and long-term monitoring of POPs in ambient air have been existed. It was not fully covered Europe (a part of EMEP Stations), Great Lakes region (IADN programme), Japan. Now due to the using of passive samplers, the present situation looks much better also with a significant contribution of data from the MONET networks. In 2009, the Conference of the Parties of the SC decided that levels of POPs in core matrices will be assessed every six years assuming that such period will be sufficient for establishment of the temporal trends. As the Košetice station is the only site worldwide where active and passive samplers have been coemployed for full six years so far, results from this station have an important role in the intercalibration of both techniques, comparison of trends derived from both datasets, and development of future global monitoring programmes [16– 18]. Long-term sustainability of such monitoring programmes is of a great importance for a success of the Global monitoring plan. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

182 Air Pollution XIX

Figure 5:

Global distribution of the sampling sites with on-going air monitoring. MONET sampling sites are coloured blue.

7 Conclusions The paper summarizes results of the ambient air monitoring activities in the Central and Eastern European region (CEEC), Central Asia, Africa and Pacific Islands driven by RECETOX under the common name of the MONET networks. For many of the participating countries these activities generated first data on the atmospheric levels of POPs. This was a reason why the background monitoring was accompanied with the screening of the extent of contamination in the individual countries. To carry these activities beyond the point of the first screening, best candidates for the background monitoring have to be selected in every region, and resources have to be sought to make the program sustainable. Now, the harmonization of the monitoring activities in both parts of Europe to gain systematic information on the levels and trends of the atmospheric pollution over this continent has been started in 2009. The EMEP stations in whole Europe participate in this now MONET phase which is focused on the improvement of the understanding to the sources, fate and transport of POPs in Europe and the providing rich information for the modeling databases.

Acknowledgements This paper was/is supported by the CETOCOEN project (CZ.1.05/2.1.00/01.0001) of the European Structural Funds, the INCHEMBIOL project (MSM 0021622412) of the Ministry of Education of the Czech Republic and the Ministry of Environment of the Czech Republic (SP/1b1/30/07).

References [1]

UNEP Stockholm Convention on Persistent Organic Pollutants. Stockholm Convention on Persistent Organic Pollutants 2001, decisions, information documents, http://www.pops.int WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

[2] [3]

[4]

[5]

[6] [7]

[8]

[9]

[10]

[11]

[12] [13] [14]

183

Draft revised Guidance on the Global monitoring plan for persistent organic pollutants, UNEP, May 2011, UNEP/POPS/COP.5/INF/27 Harner, T., Bartkow, M., Holoubek, I., Klanova, J., Wania, F., Gioia, R., Moeckel, C., Sweetman, A. J. & Jones, K. C., Passive air sampling for persistent organic pollutants: Introductory remarks to the special issue. Environ. Pollut. 144, pp. 361-364, 2006. Peters, A. J., Lane, D. A., Gundel, L. A., Northcott, G. L. & Jones, K. C., A comparison of high volume and diffusion denuder samplers for measuring semivolatile organic compounds in the atmosphere. Environ. Sci. & Technol. 34, pp. 5001-5006, 2000. Harner, T., Farrar, N. J., Shoeib, M., Jones, K. C. & Gobas, F., Characterization of polymer-coated glass as a passive air sampler for persistent organic pollutants. Environ. Sci. & Technol. 37, pp. 2486-2493, 2003. Shahir, U. M., Li, Y. H. & Holsen, T. M., Dry deposition of gas-phase polycyclic aromatic hydrocarbons to greased surrogate surfaces. Aerosol Sci. Technol. 31, pp. 446-455, 1999. Jaward, F. M., Farrar, N. J., Harner, T., Sweetman, A. J. & Jones, K. C., Passive air sampling of polycyclic aromatic hydrocarbons and polychlorinated naphthalenes across Europe. Environ. Toxicol. Chem. 23, pp. 1355-1364, 2004. Harner, T., Shoeib, M., Diamond, M., Stern, G. & Rosenberg, B., Using passive air samplers to assess urban - Rural trends for persistent organic pollutants. 1. Polychlorinated biphenyls and organochlorine pesticides. Environ. Sci. & Technol. 38, pp. 4474-4483, 2004. Klanova, J., Kohoutek, J., Hamplova, L., Urbanova, P. & Holoubek, I., Passive air sampler as a tool for long-term air pollution monitoring: Part 1. Performance assessment for seasonal and spatial variations. Environ. Pollut. 144, pp. 393-405, 2006. Cupr, P., Klanova, J., Bartos, T., Flegrova, Z., Kohoutek, J. & Holoubek, I., Passive air sampler as a tool for long-term air pollution monitoring: Part 2. Air genotoxic potency screening assessment. Environ. Pollut. 144, pp. 406-413, 2006. Klánová, J., Čupr P., Holoubek, I., Borůvková, J., Přibylová, P., Kareš, R., Kohoutek, J., Dvorská, A. & Komprda, J., Towards the Global Monitoring of POPs. Contribution of the MONET Networks. RECETOX MU Brno. RECETOX_TOCOEN REPORTS No. 357. 2009; ISBN 97880-210-4853-9; MASARYK UNIVERSITY 2009. Jaward, F. M., Farrar, N. J., Harner, T., Sweetman, A. J. & Jones, K. C., Passive air sampling of PCBs, PBDEs, and organochlorine pesticides across Europe. Environ. Sci. & Technol.38, pp. 34-41, 2004. Klánová, J., Čupr, P., Kohoutek, J. & Harner, T., Assessing meteorological parameters on the performance of PUF disks passive air samplers for POPs. Environ. Sci. Technol. 42 (2), pp. 550-555, 2008. Klánová, J., Čupr P., Kohoutek, J., Holoubek, I.: Application of Passive Sampler for Monitoring of POPs in Ambient Air. Part I: Model WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

184 Air Pollution XIX

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

monitoring network in the Czech Republic (MONET-CZ), 2006. RECETOX MU Brno. RECETOX-TOCOEN REPORTS No. 318. August 2007. Klanova, J., Cupr, P., Borůvková, J., Kohoutek, J., Kareš, R., Přibylová, P., Prokeš, R., Holoubek, I.: Application of passive sampler for monitoring of POPs in ambient air. IV. Model monitoring network in the Czech Republic (MONET-CZ 2007), 2008. Masaryk Univerzity, Brno, Czech Republic. ISBN 978-80-210-4696-2 Klánová, J., Kohoutek, J., Kostrhounová, R. & Holoubek, I., Are the residents of former Yugoslavia still exposed to elevated PCB levels due to the Balkan wars? Part 1: Air sampling in Croatia, Serbia, Bosnia & Hercegovina. Environ. Int. 33 (6), pp. 719-726, 2007. Klánová, J., Kohoutek, J., Čupr, P. & Holoubek, I. Are the residents of former Yugoslavia still exposed to elevated PCB levels due to the Balkan wars? Part 2: Passive air sampling network. Environ. Int. 33 (6), pp. 727735, 2007. Klánová, J., Čupr P., Holoubek, I.: Application of Passive Sampler for Monitoring of POPs in Ambient Air. Part II: Pilot study for development of the monitoring network in the Central and Eastern Europe (MONET_CEEC), 2006. RECETOX MU Brno. RECETOX-TOCOEN REPORTS No. 319. August 2007. Klanova, J., Cupr, P., Holoubek, I., Borůvková, J., Přibylová, P., Kareš, R., Kohoutek, J.: Application of passive sampler for monitoring of POPs in ambient air. V. Pilot study for development of the monitoring network in the Central and Eastern Europe (MONET-CEEC 2007), 2008. Masarykova Univerzita, Brno, Czech Republic. ISBN 978-80-210-4697-9 Klanova, J., Cupr, P., Holoubek, I., Borůvková, J., Přibylová, P., Kareš, R., Kohoutek, J., Dvorská, A., Tomšej, T., Ocelka, T.: Application of passive sampler for monitoring of POPs in ambient air. VI. Pilot study for development of the monitoring network in the African continent (MONET-AFRICA 2008), 2008. Masarykova Univerzita, Brno, Czech Republic, ISBN 978-80-210-4739-6 Klánová J., Čupr, P., Holoubek, I., Borůvková, J., Kareš, R., Tomšej, T. & Ocelka, T., Monitoring of persistent organic pollutants in Africa. Part 1: Passive air sampling across the continent in 2008. J. Environ. Monit., 11, pp. 1952–1963, 2009. Lammel, G., Dobrovolný, P., Dvorská, A., Chromá, K., Brázdil, R., Holoubek, I. & Hošek, J., Monitoring of persistent organic pollutants in Africa. Part 2: Design of a network to monitor the continental and intercontinental background. J. Environ. Monit., 11, pp. 1964–1972, 2009. Klanova, J., Čupr P., Holoubek, I.: Application of passive sampler for determination of the POPs concentrations in ambient air. Part III: Pilot study in the Southern Pacific, Fiji, 2006. RECETOX MU Brno, RECETOX-TOCOEN REPORTS No. 320. August 2007.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

185

The EC QA/QC programmes for inorganic gas pollutants testing M. Barbiere1, A. Borowiak1, F. Lagler1, M. Gerboles1, M. Kapus2 & C. Belis1 1

European Commission – Joint Research Centre, Institute for Environment and Sustainability, European Reference Laboratory for Air Pollution (ERLAP), Italy 2 Danfoss d.o.o., Slovenia

Abstract The Air Quality Directive (2008/50/EC) asks for the organisation of quality assurance programmes for air quality assessment methods at European level. Since the early 1990s the European Reference Laboratory for Air Pollution (ERLAP) of the EC’s Joint Research Centre (JRC) has carried out Intercomparison Exercises (IE) for air pollution measurements on a regular basis for Member States of the EU. All European National Reference Laboratories (NRLs), joined together in the AQUILA Network, are obliged to participate in IE. More than 45 laboratories and institutes, coming from 35 European countries, have participated in the IE during the last 15 years. The results of the most recent IE which took place from 2005 to 2010 are described. Gas mixtures with some concentrations of CO, SO2, NOx, and O3 were generated and measured by the participants. With the results of the participants’ z’-score, En number, repeatability and reproducibility, outlier through the test of Grubb were evaluated. Keywords: air quality, intercomparison, ozone, sulphur dioxide, nitrogen monoxide, nitrogen dioxide, carbon monoxide.

1 Introduction With the adoption of Directive 2008/50/EC [1] on ambient air quality and cleaner air for Europe, a framework for a harmonized air quality assessment in Europe was set. This Directive specifies, among others, the reference methods WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110171

186 Air Pollution XIX for measuring specific atmospheric pollutants and sets data quality objectives (DQO) for the uncertainty, minimum data capture and time coverage. It establishes limit and target values for sulphur dioxide (SO2), nitrogen dioxide (NO2) and nitrogen oxide (NO), particulate matter, lead, benzene, carbon monoxide (CO) and ozone (O3), not to be exceeded to avoid negative effects on human health and the environment. The European Commission (EC) has supported the development and publication of standard measurement methods for CO [2], SO2 [3] , NO-NO2 [4] and O3 [5] as European standards. Appropriate calibration methods [6–8] have been standardized by the International Organization for Standardization (ISO). As foreseen in the Air Quality Directive the ERLAP organizes interlaboratory comparison exercises (IE) to assess and improve the status of comparability of measurements of National Reference Laboratories (NRL) of the Member States of the European Union. Through the IE ERLAP promotes information and know how exchange among the expert laboratories. Currently, a more systematic approach has been adopted, in accordance with the Network of NRLs for Air Quality (AQUILA) [9], aiming both at providing an alert mechanism for the purposes of implementation of legislation and at supporting the operation of quality schemes by NRLs. The protocol for the organization of IEs was developed by ERLAP in collaboration with the WHO CC and the AQUILA Network, collecting all the experiences of the previous IE for gaseous air pollutants [10]. This evaluation scheme was adopted in December 2008 by the AQUILA Network and WHO CC and is applied to all IEs since then. It contains common criteria to alert on possible performance failures which do not rely solely on the uncertainty claimed by participants. The evaluation scheme implements the z’-score [11] and En method [11] with the uncertainty requirements for calibration gases stated in the European standards [2–5], which are consistent with the DQOs of European Directives. Beside the proficiency of participating laboratories, the repeatability and reproducibility [13] of standardized measurement methods [12–14] are evaluated as well. These group evaluations are useful indicators of trends in measurement quality over different IE.

2 Inter-comparison exercises In this report the results of nine IE that were organized between 2005 and 2010 in three European facilities (Langen (D), Essen (D) and Ispra (I)) are described. In Langen the inter-comparison facility of the German Federal Environment Agency (UBA) Pilot station was used and the IE was carried out under the supervision of the World Health Organization Collaborating Centre for Air Quality Management and Air Pollution Control, Berlin (WHO CC) in collaboration with JRC. In 2007 the JRC organized an IE in Essen at the facility of the North Rhine-Westphalian State Agency for Nature, Environment and Consumer Protection (LANUV) in cooperation with WHO CC [17]. All the others IE took place at the ERLAP laboratory of the JRC in Ispra (IT). In Table 1 the list of IE evaluated is shown. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Table 1:

187

List of IE from 2005 till 2010.

Inter-comparison

Site

N. of participants

June 2005 June 2007 October 2007 April 2008 October 2008_1 October 2008_2 September 2009 October 2009 June 2010

Ispra (IT) Ispra (IT) Essen (DE) Ispra (IT) Ispra (IT) Ispra (IT) Langen (DE) Ispra (IT) Ispra (IT)

11 11 16 9 10 9 8 9 10

The participants were required to participate in the IE with their own measurement instrumentation, data acquisition equipment and working standards to be used for calibrations during the IE. 2.1 The preparation of test mixtures During the IE, gas mixtures were prepared for SO2, CO, O3, NO and NO2 at concentration levels around European Air Quality limit and target. The test mixtures were prepared by the dilution of gases from cylinders containing high concentrations of NO, SO2 or CO using thermal mass flow controllers [8]. O3 was added using an ozone generator and NO2 was produced applying the gas phase titration method [16] in the conditions of excess NO. Several different concentrations steps were generated, each lasting roughly 2 hours. Participants were required to report three half-hour-mean measurements for each concentration level in order to evaluate the repeatability of their measurements. Zero concentration levels were generated for one hour and one half-hour-mean measurement was reported. The sequence program of generated test gases is given in Figure 1. In order to test simultaneous gas mixtures under homogeneous experimental conditions a calibration bench [15] was used to generate the different pollutant mixtures (Figure 1). The calibration bench allows through a dynamic dilution the generation of complex gas mixtures by dilution of high concentration gas cylinders. The system is further equipped with an ozone generator for the implementation of the Gas Phase Titration (GPT) and with a water vapor generator for the preparation of humid gas mixtures. All the functions of the bench are programmable and controlled by computer, so that automated and unattended operation is possible. The gas mixture is supplied to the workbenches. During each IE its reference value is given by the ERLAP laboratory monitor who is connected to one workbench.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

188 Air Pollution XIX

Figure 1:

Table 2: day 1st 2nd 2nd 2nd 2nd 2nd 2nd 2nd 2nd 3rd 3rd 3rd 3rd 3rd 3rd 3rd 3rd 3rd 3rd 3rd 3rd 3rd 4th 4th 4th 4th 4th

Calibration bench scheme.

Example of the sequence program of generated test gases.

start time

duration

12:00 8:00 11:00 12:00 14:00 16:00 18:00 20:00 22:00 0:00 2:00 4:00 6:00 8:00 10:00 12:00 14:00 16:00 < 18:00 20:00 21:00 23:30 1:30 3:30 5:30 7:30 8:30

(h) 6 3 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 1 2:30 2 2 2 2 1 END

operation installation calibration NO-NO2-O3 NO-NO2 NO-NO2 O3 NO-NO2 NO-NO2 O3 NO-NO2 NO-NO2 O3 NO-NO2 NO-NO2 O3 NO-NO2 NO-NO2 O3 calibration CO-SO2 CO-SO2 CO-SO2 CO-SO2 CO-SO2 CO-SO2

zero air

NO

NO2

O3

CO

SO2

ppb

ppb

ppb

ppb

ppm

ppb

500 380

0 120

250 146

0 104

150 90

0 60

50 29 1

0 20 9

15 7 21

0 13 6

86 6 43 2 1

132 47 18 8 75 3

0

120

104

60

20 9

13 6 0

0

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

189

3 The evaluation of the laboratory’s measurement proficiency To evaluate the participants measurement proficiency the methodology described in ISO 13528 [11] was applied. For the IEs organized in Ispra (IT) and in Essen (DE) the measurement results of ERLAP were considered as the reference values according to the AQUILA protocol [10] while in Langen (DE), the results of UBA (DE) were used as the reference values. The proficiency of the participants was assessed by calculating two performance indicators. The first performance indicator (z’-score) evaluates if the difference between the participant measured value and the reference value remains within the limits of a common criterion. The second performance indicator (En-number) tests if the difference between the participant measured value and reference value remains within the limits of a criterion, which is calculated individually for each participant from its declared uncertainty of measurement and the uncertainty of reference value. 3.1 Assigned values The assigned values of tested concentration levels were derived from ERLAP or UBA measurements which were calibrated against their certified reference material gases under strict condition of traceability to international standards. In this perspective the assigned values are reference values as defined in the ISO 13528 [11]. SO2, CO and NO analyzers were calibrated using primary calibration gas mixtures prepared according to the methodology described in the ISO 6144 [7] (UBA) and ISO 6143 [6] (ERLAP). Gas mixtures for the calibration experiment were produced from the reference mixtures by dynamic dilution method using mass flow controllers [8]. All flows were measured with certified devices (Brooks vol-U-meter or Molbox/Molbloc systems). In Ispra since 2008 O3 calibration measurements during IE were carried out using as primary standard the NIST (US National Institute of Standards and Technology) Standard Reference Photometer SN42 (SRP) [18]. Assigned values were validated by comparison to the group statistics (x* and s*) for every parameter and concentration level of the IE. These statistics are calculated from participants, applying the robust method described in ISO 13528 [11]. The validation is taking into account reference laboratory measurement result (X) and its standard uncertainty (uX’) as given in equation (1) [11]:

x  X

1,25  s 

 2

p

2  u X2 '

(1)

where ‘x*’ and ‘s*’ represent robust average and robust standard deviation respectively and ‘p’ is the number of participants.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

190 Air Pollution XIX The homogeneity of test gas mixtures throughout the working bench was evaluated by comparison of measurements at the beginning and at the end of the distribution line. From the relative differences between beginning and end measurements, average and standard deviation (s) were calculated, and the uncertainty of test gas due to lack of homogeneity was calculated as the sum of squares of these average and standard deviation. The upper and lower limits of bias homogeneity was evaluated to be smaller than 0.5% which constitutes the relative standard uncertainty of 0,3% of each concentration level assuming a rectangular distribution of the biases. The standard uncertainties of reference values (uX) were calculated with equation (2). u X2  u X2 '   X  ur hom ogeneity 

2

(2)

3.2 z’-score The z’-score statistic is calculated according to ISO 13528 [11] with equation (3).

z' 

xi  X

 p2  u X2

(3)

where ‘xi’ is a participant’s run average value, ‘X’ is the reference value, ‘σp‘ is the ‘standard deviation for proficiency assessment’ and ‘uX‘ is the standard uncertainty of assigned value. In the European standards [2–5] the uncertainties of calibration gases used in ongoing quality control are prescribed. In fact, it is stated that the maximum permitted expanded uncertainty for calibration gases shall be 5% and that ‘zero gas’ shall not give instrument reading higher than defined limit. The assessment of results in the z’-score evaluation is made according to the following criteria:  |z’|  2 are considered acceptable score.  2 < |z’|  3 are considered warning score.  |z’| > 3 are considered not acceptable score. Scores falling in this range are very unusual and are taken as evidence that an anomaly has occurred that should be investigated and corrected. After more than 15 years of IE the high level of expertise reached by the NRL is confirmed by the high percentage of acceptable results (above 90%). Table 3: IE June 2005 June 2007 October 2007 April 2008 October 2008 1 October 2008 2 September 2009 October 2009 June 2010

Percentage of z’-score results in IE from 2005 till 2010. Acceptable score % 95.5 97.8 93.2 93.8 92.9 97.0 94.3 98.2 97.0

Warning score % 2.3 1.9 4.6 2.1 4.2 3.0 4.7 1.8 3.0

Not Acceptable score % 2.2 0.3 2.2 4.1 2.9 0 1 0 0

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

191

3.3 En - number The normalized deviations [11] (En) were calculated with the following equation:

En 

xi  X U x2i  U X2

(4)

where ‘X’ is the reference value with an expanded uncertainty ‘UX‘ and ‘xi’ is the participant’s average value with an expanded uncertainty ‘UXi’. Results are acceptable when En  1 . 3.4 Discussion about z’-score and En-number For a general assessment of the quality of each result a decision diagram was developed (Figure 2) sorting the results according to the following seven categories. a1 measurement result is completely acceptable a2 measurement result is acceptable (z’-score acceptable and En-number ok) but the reported uncertainty is too high a3 measured value is acceptable (z’-score acceptable) but the reported uncertainty is underestimated (En-number not ok) a4 measurement result is warning (z’-score warning) but due to a high reported uncertainty can be considered valid (En-number ok) a5 measurement result is warning (z’-score warning and En-number not ok) a6 measurement result is not acceptable (z’-score not acceptable) but due to a high reported uncertainty can be considered valid (En-number ok) a7 measurement result is not acceptable (z’-score not acceptable and En-number not ok).

Figure 2:

Diagram for general assessment of proficiency results.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

192 Air Pollution XIX Table 4:

General assessment of proficiency results. “nv”: no values reported. 2007_1

2007_2

2008_1

2008_2

2008_3

2009_1

2009_2

2010

a1

unit

64

55

54

37

27

55

85

80

a2

30

30

14

40

31

27

6

8

a3

4

6

6

14

19

3

8

4

1

1

1

1

1

4

0

1

1

3

1

3

2

1

1

2

a6

0

0

2

1

0

1

0

0

a7

0

2

1

2

0

0

0

0

nv

0

3

21

2

20

9

0

5

a4 a5

%

In Table 4 is presented a summary of z’-score and En Number evaluation of all IE beside 2005 in which this discussion was not carried out. As described above a1 indicates the percentage of acceptable results as measured value and calculated uncertainty. Generally the results have good measured values (a1+a2+a3) but in 2008_2 and 2008_3 the percentage of uncertainty values too high (a2) and too low (a3) are considerably above all the other IE. Beside 2009_2 and 2010 the percentage of a2 category shows a tendency to overestimate the uncertainty. In order to investigate this issue in the future would be interesting asking the participants to provide the method used to calculate the uncertainty. 3.5 Reproducibility and repeatability Reproducibility (R) and Repeatability (r) were determined [13] with equation respectively (5) and (6). In equation (5) p is the number of participants after discarding outliers, si is the standard deviation of the measurements of each participant for each sample, ym is the mean of the measurements of each participant, m is the reference value of each sample and n is the number of repeated measurements.

R  2 .8

s n 1

1  ( y m  m) 2  n p 1 p

r  2 .8

s

2 i

p

p

2 i

(5)

p (6)

p

From Table 5 to Table 14 repeatability and reproducibility of two levels of concentrations for NO, NO2, O3, CO and SO2 are represented. Reproducibility values are quite higher than repeatability and this could be a sign of a possible non homogeneous calibration procedure and reference material used. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

193

Air Pollution XIX

Table 5:

NO (conc. 500ppb) reproducibility and repeatability in all IE.

Parameter

r (ppb)

R (ppb)

IE

Parameter

r (ppb)

R (ppb)

IE

NO

0 006

0 036

2005

NO

0 0048

0 062

2007_1

NO

0 003

0 056

2007_1

NO

0 009

0 086

2007_2

NO

0 004

0 053

2007_2

NO

0 005

0 102

2008_1

NO

0 002

0 022

2008_1

NO

0 011

0 042

2008_2

NO

0 003

0 012

2008_2

NO

0 006

0 133

2008_3

NO

0 0045

0 090

2008_3

NO

0 013

0 118

2009_1

NO

0 005

0 142

2009_1

NO

0 004

0 098

2009_2

NO

0 003

0 048

2009_2

NO

0 005

0 083

2010

NO

0 007

0 065

2010

Table 6:

NO (conc. 50ppb) reproducibility and repeatability in all IE.

NO2 (conc. 20ppb) reproducibility and repeatability in all IE.

Table 7: NO2 (conc. 100ppb) reproducibility and repeatability in all IE.

Table 8:

Parameter

r (ppb)

R (ppb)

IE

Parameter

r (ppb)

R (ppb)

NO2

0 008

0 095

2005

NO2

0 029

0 330

2005

NO2

0 007

0 068

2007_1

NO2

0 016

0 086

2007_1

NO2

0 007

0 090

2007_2

NO2

0 010

0 166

2007_2

NO2

0 006

0 049

2008_1

NO2

0 015

0 099

2008_1

NO2

0 012

0 097

2008_2

NO2

0 025

0 137

2008_2

NO2

0 005

0 043

2008_3

NO2

0 024

0 106

2008_3

NO2

0 009

0 084

2009_1

NO2

0 027

0 061

2009_1

NO2

0 006

0 116

2009_2

NO2

0 008

0 124

2009_2

NO2

0 006

0 108

2010

NO2

0 008

0 143

2010

Table 9: O3 (conc. 120ppb) reproducibility and repeatability in all IE.

Table 10:

IE

O3 (conc. 20ppb) reproducibility and repeatability in all IE.

Parameter

r (ppb)

R (ppb)

IE

Parameter

r (ppb)

R (ppb)

IE

O3

0 014

0 084

2007_1

O3

0 008

0 124

2007_1

O3

0 018

0 027

2007_2

O3

0 014

0 103

2007_2

O3

0 015

0 032

2008_1

O3

0 006

0 052

2008_1

O3

0 007

0 035

2008_2

O3

0 011

0 118

2008_2

O3

0 017

0 037

2008_3

O3

0 006

0 050

2008_3

O3

0 013

0 086

2009_1

O3

0 007

0 140

2009_1

O3

0 013

0 068

2009_2

O3

0 006

0 080

2009_2

O3

0 009

0 077

2010

O3

0 008

0 077

2010

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

194 Air Pollution XIX Table 11: CO (conc. 8ppm) reproducibility and repeatability in all IE.

Table 12:

Parameter

r (ppb)

R (ppb)

IE

Parameter

r (ppb)

R (ppb)

IE

CO

0 004

0 035

2005

CO

0 034

0 145

2005

CO

0 002

0 070

2007_1

CO

0 003

0 150

2007_1

CO

0 004

0 053

2007_2

CO

0 019

0 169

2007_2

CO

0 004

0 049

2008_1

CO

0 005

0 040

2008_1

CO

0 010

0 110

2008_2

CO

0 012

0 305

2008_2

CO

0 001

0 072

2008_3

CO

0 006

0 184

2008_3

CO

0 003

0 073

2009_1

CO

0 007

0 164

2009_1

CO

0 007

0 046

2010

CO

0 003

0 113

2010

Table 13: SO2 (conc. 130ppb) reproducibility and repeatability in all IE. Parameter

r (ppb)

R (ppb)

SO2

0 003

SO2

0 004

SO2 SO2

Table 14:

CO (conc. 2ppm) reproducibility and repeatability in all IE.

SO2 (conc. 20ppb) reproducibility and repeatability in all IE.

IE

Parameter

r (ppb)

R (ppb)

IE

0 077

2005

SO2

0 018

0 100

2005

0 084

2007_1

SO2

0 028

0 184

2007_1

0 004

0 073

2007_2

SO2

0 019

0 098

2007_2

0 003

0 061

2008_1

SO2

0 003

0 050

2008_1

SO2

0 004

0 087

2008_2

SO2

0 015

0 137

2008_2

SO2

0 004

0 131

2008_3

SO2

0 014

0 082

2008_3

SO2

0 008

0 047

2009_1

SO2

0 010

0 026

2009_1

SO2

0 005

0 104

2009_2

SO2

0 015

0 211

2009_2

SO2

0 005

0 960

2010

SO2

0 010

0 132

2010

3.6 Grubbs’ test with outlier and straggler Tests for data consistency and statistical outliers as described in ISO 5725-2 [13] were carried out during the evaluation. Laboratories showing some form of statistical inconsistency were requested to investigate the cause of discrepancies and laboratories were allowed to correct their results in case of identification of exceptional errors. In Table 15 outliers and stragglers are presented for each IE. This table for each failing result is also shown at which level was the anomaly: underestimation (Gmin), overestimation (Gmax) or both. Generally this test didn’t show any relevant situation with a great number of outlier. Per each level of concentration for each pollutant (7) and (8) was used to define if the higher or the lower value was an outlier.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

195

_

( x  x) G max  max s

(7)

_

G min  Table 15:

(8)

Statistical outliers and stragglers according to Grubb’s test for all IE.

IE

Site

Jun-05

Ispra (IT)

Jun-07

Ispra (IT)

Oct-07

Essen (DE)

Apr-08

Ispra (IT)

Oct 08_1

Ispra (IT)

Oct 08_2

Ispra (IT)

Sep-09

Langen

Oct-09

Ispra

Jun-10

( x  x min ) s

Ispra (IT)

Pollutant

Straggler

CO NO2 CO NO2 O3 SO2 NO NO2 O3 CO NO2 O3 NO NO2 O3 all NO O3 SO2 CO NO NO2 O3 SO2

1 2 / 2 1 1 1 1 1 1 1 1 1 2 none 1 1 / 1 2 2 2 1

Failing test level Gmax Gmin/Gmax

Outlier / 2 1 2

Gmax Gmax Gmin Gmax Gmax Gmin Gmin Gmax Gmin Gmax Gmin Gmin Gmax Gmax Gmin Gmin Gmax Gmax

1 / 2 / / / / / / / none 2 1 1 1 / / 3 /

Failing test level Gmin/Gmax Gmax Gmax Gmax Gmin/Gmax

Gmin Gmax Gmin Gmax Gmax

4 Intercomparison exercise as a learning process As discussed in section 3.4, in all IE the high percentages of valid measured values confirmed the general good performance of the laboratories. From the En number results (section 3.4) came out a need to harmonize and to define an estimation procedure of the measurement uncertainty. All results obtained in these exercises were below the DQO of 15% of expanded measurement uncertainty requested by the directive [1]. It must be considered that the IE took place in an ideal situation: ambient temperature under control, constant relative humidity, absence of interferences and all instruments were recently calibrated and maintained. Under routine conditions existing in the networks an increase of measurement uncertainty, repeatability and reproducibility is expected to happen. In order to evaluate the ability of NRL to meet the DQO under field condition, it would be advisable to organize an IE using real samples. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

196 Air Pollution XIX One of the most important results obtained during these intercomparison exercises can be found in the opportunity for all the experts in air quality monitoring to exchange information and technical know-how. The way in which IE were managed gave the chance to experts in young teams, to those in new EU member States and Candidate Countries to get in touch with experienced colleagues.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

[11] [12] [13] [14]

Directive 2008/50/EC of the European Parliament and of the Council of 21 May 2008 on ambient air quality and cleaner air for Europe, L 152, 11.06.2008 EN 14626:2005, Ambient air quality – Standard method for the measurement of the concentration of carbon monoxide by non-dispersive infrared spectroscopy EN 14212:2005, Ambient air quality – Standard method for the measurement of the concentration of sulphur dioxide by ultraviolet fluorescence EN 14211:2005, Ambient air quality – Standard method for the measurement of the concentration of nitrogen dioxide and nitrogen monoxide by chemiluminescence EN 14625:2005, Ambient air quality – Standard method for the measurement of the concentration of ozone by ultraviolet photometry ISO 6143:2001, Gas analysis – Comparison methods for determining and checking the composition of calibration gas mixtures ISO 6144:2003, Gas analysis – Preparation of calibration gas mixtures – Static volumetric method ISO 6145-7:2001, Gas analysis – Preparation of calibration gas mixtures using dynamic volumetric methods – Part 7: Thermal mass-flow controllers http//ies.jrc.ec.europa.eu/aquila-homepage.html AQUILA Position Paper N. 37, (2008) Protocol for intercomparison exercise. Organisation of intercomparison exercises for gaseous air pollution for EU national air quality reference laboratories and laboratories of the WHO EURO region http://ies.jrc.ec.europa.eu/aquilaproject/role-and-tasks-of-national-reference-laboratories html ISO 13528:2005, Statistical methods for use in proficiency testing by interlaboratory comparisons ISO 5725-1:1994, Accuracy (trueness and precision) of measurement methods and results – Part 1: General principles and definitions ISO 5725-2:1994, Accuracy (trueness and precision) of measurement methods and results – Part 2: Basic method for the determination of repeatability and reproducibility of a standard measurement method ISO 5725-6:1994, Accuracy (trueness and precision) of measurement methods and results – Part 6: Use in practice of accuracy values

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

[15] [16] [17] [18]

197

De Saeger E. et al., (1997) European comparison of Nitrogen Dioxide calibration methods, EUR 17661 ISO 15337:2009, Ambient air – Gas phase titration – Calibration of analysers for ozone Kapus M. et al. (2009) The evaluation of the Intercomparison Exercise for SO2, CO, O3, NO and NO2 carried out in October 2007 in Essen. JRC scientific and technical reports. EUR 23788. Viallon J. et al 2009 Metrologia 46 08017. Final report, on-going key comparison BIPM.QM-K1: Ozone at ambient level, comparison with JRC, 2008. doi: 10.1088/0026-1394/46/1A/08017

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

This page intentionally left blank

Air Pollution XIX

199

GIS for data management of environmental surveys, carried out in Biancavilla (CT) superfund experience S. Bellagamba, F. Paglietti, V. Di Molfetta, F. Damiani & P. De Simone INAIL – ex ISPESL, Department for Manufacturing Installations and Human Settlements, Italy

Abstract Biancavilla town has been included by the Italian Environmental Ministry in the Italian Superfund to remediate over the problem of contamination. The contamination is produced by fibrous fluoroedenite, a carcinogen mineral. Since the 1990s many environmental monitoring campaigns have been made in Biancavilla town by private and public Institutions charged to guarantee the safety of human health and environmental protection, such as Italian INAIL-EXISPESL, National Institute of Health (ISS), Regional Environmental Agency (ARPA), Local Health Agency (AUSL), University of Catania, Circum-Etna Railway. The Environmental Ministry charged INAIL-EX-ISPESL to create a specific Geographic Information System (GIS) dedicated to the management and consultation of all data collected in Biancavilla town (Catania Sicily), in order to properly handle the large amount of data, in continuous acquisition. The GIS is a tool based on a continuously evolving repository, i.e. an alphanumeric database and digital map updated in real time, through which managing the information about the different monitoring campaigns. GIS Basic function is to associate a geo-referenced location to a descriptive alphanumeric database and to relate such information to events that have happened in the territory. It is also possible to process all data with a complex query associated to geographic position. This paper discusses the GIS, specifically made to collect all monitoring data in Biancavilla town regarding: excavation activities; street cleaning; building operations; human activities. Using this GIS, it is possible to detect trends and increases in the concentration of dangerous fibres in Biancavilla. With this tool it WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110181

200 Air Pollution XIX is possible to highlight any exceeding the threshold limit value, concerning the concentration of fibrous airborne particulate. The aim of this paper is to present a useful instrument to quickly identify risk situations and to adopt prevention measures of diffusion of hazardous contaminants. Finally the GIS, using some standardized formats, ensures the continuous updating of data and the information exchange among all subjects involved (institutions, experts, citizens). Keywords: geographic information system, risk, fluoroedenite, asbestos.

1 Introduction Since the 1990s numerous environmental monitoring campaigns have been conducted in the Municipality of Biancavilla by leading national bodies and/or competent local agencies set up to protect human health and the environment, including INAIL-DIPIA, ISS, ARPA, AUSL, University of Catania, the Municipal Council and Circum-Etnea Railway, due to the detected widespread presence of a carcinogenic fibre similar to asbestos called “fluoro-edenite”. The Environment Ministry has included the Municipality of Biancavilla in the 57 National Priority Sites requiring remediation. To be able to correctly manage the large amount of data obtained from monitoring, which continues to be gathered, regarding the level of environmental contamination present at the site, INAIL-DIPIA has set up an ad hoc GIS (Geographic Information System) dedicated to the handling and consultation of all data collected in the Municipality of Biancavilla (CT). This GIS makes it possible to associate geo-referencing with monitoring carried out during various activities – e.g. excavation work, street cleaning, building works, etc. – and to gauge in real time any increases in concentrations of harmful fibres in the air and their location. This paper illustrates, as an example, the use of a GIS dedicated to Biancavilla and seeks to highlight the advantages of using such integrated information management systems.

2 The NPL site of Biancavilla One of the best-known sites in Italy’s NPL (National Priority List) is the Municipality of Biancavilla (Catania), a small town of 23,000 inhabitants standing at the foot of Mount Etna, where over the past 20 years there have been about 40 recorded deaths due to pleural mesothelioma, 10 times greater than the national average. The detected pollutant is a fibrous volcanic amphibole called Fluoro-edenite, a bright yellow fibrous mineral made up of very thin (less than 1 micron in diameter) and relatively long fibres (up to 50-80 microns), which cause pathologies similar to asbestos-related diseases. The main source of contamination is the broken stone material present in the area, mined up to the year 2000 in the Monte Calvario quarry. Since then, over the past ten years hundreds of logs and thousands of surveys of airborne WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

201

materials have been carried out to assess the extent of contamination and measure concentration values for these fibres in the air. Analyses have shown that hundreds of samples, both in rocks and airborne, are contaminated by fluoro-edenite, with variable concentrations, often exceeding allowed limits. The Municipality has accordingly been included in the 57 National Priority Sites to be cleaned up. The Environment Ministry receives technical and scientific support from National Scientific Bodies, including INAIL-DIPIA, in managing emergency shutdown and remediation procedures. In greater detail, the following work phases have been monitored: surfacing of non-asphalted roads, disposal of piles of material, rubble and dust in bags after adequate wetting or treatment with encapsulator products, cessation of quarrying activity and creation of an ad hoc dump.

3 Methods While performing the above activities a total of 1,800 samples of airborne materials were taken from 2000 to 2010, with both environmental and personal sampling, performed by Ispesl, the University of Catania, Arpa Sicilia and private Laboratories on behalf of the Municipality. Acquired data have been filed and managed in a dedicated GIS, including details on:  date of sampling,  place of sampling,  sampling volumes,  meteorological conditions,  method of analysis,  value of detected concentration, expressed in f/l,  lower and upper confidence limits,  analytical methods (SEM or PCOM). Collected samples derive from specific sampling techniques, drawn up especially for this NPL Site, for both personal and environment sampling, both indoor and outdoor. The most appropriate instrument for the optimal management of monitoring data has proven to be a computerised database associated with a GIS. The database is used to collect, organise, manage, update and easily consult data, while the GIS allows the entry of data in the local context, in other words on a geographic basis. This GIS can be consulted by means of queries, making it possible to describe the trends over time of pollution/depollution phenomena and to highlight situations at risk. The peculiar nature of the GIS is its topological structuring of data management of information based on the mutual spatial relations of various elements, the possibility of conducting analyses on attributes and the possibility of processing geographic data using mathematical algorithms. Algorithms make it possible to combine different levels of information via simple operators, such as overlaying, or complex operators, such as buffering, making it possible to WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

202 Air Pollution XIX create new information levels, associating data so as to identify relations that would otherwise be difficult to detect. The software used to manage the geographic database is ArcGis 10, produced by Esri, while for the alphanumeric database Microsoft’s Access software is used. In the specific case of the Municipality of Biancavilla the results of samples taken were collected in an ad hoc database, with the construction of interrelated tables and creation of two specific masks, one for the filing of environmental samples, the other for personal samples, to facilitate data entry. A total of 115 environmental sampling points were identified, deriving from monitoring activities performed by INAIL DIPIA, ARPA, the University and private Laboratories on behalf of the Municipality and the newly formed Circum-Etnea Railway. With regard to environmental samplings, about 1,800 records have been entered regarding a period of time of about 10 years (2000-2010).

Figure 1:

Environmental sampling information mask.

With regard to the monitoring of workers and the exposed population, data on 37 subjects subjected to personal sampling were catalogued. There are about 380 stored records for the period 2008-2010, a period of time for which measurements have been standardised. Geo-referenced shapefiles with vectorial layers have been included in the GIS:  sampling points  municipal grid  provincial grid  regional grid  road and rail network obtained from Teleatlas mapping  map of population density obtained with Istat data from the population census (2001)  sensitive receptors (schools, hospitals, etc.) There are also geo-referenced layers in raster format, such as:  high-definition IKONOS satellite images and orthophotography To be entered in the GIS, layers have been geo-referenced in the reference geographic system WGS84 UTM32. The system makes it possible to query the alphanumeric database, view analytical values obtained from samplings on the map and render the various levels of exposure visible in graph form. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Figure 2:

Figure 3:

203

Personal sampling information mask.

Location of environmental sampling points in the GIS of Biancavilla.

4 Discussion The GIS has made it possible to file analytical data on personal and environmental samplings. Personal samplings were conducted on both remedial and territorial rehabilitation site workers and other categories at risk. With regard to remediation operators, two-monthly data for the period 20082009-2010 are given, a period in which remediation activities were realized and for which measurements have been standardised. Personal sampling was conducted by drawing a volume of 480 l at a rate of 2 l/min, with analysis via Phase Contrast Optical Microscopy (PCOM). The values recorded in that period, illustrated in Fig. 4, showed fibre concentration levels of about 0–0.4 f/l in 2008, with a slight increase in more WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

204 Air Pollution XIX recent samplings (1 f/l), but remaining well below the allowed threshold limit (100 f/l). Concentration levels, which were also well below the pre-alarm value of 20 f/l, showed that safety procedures adopted with the specific aim of protecting workers’ health (adoption of specific PPE, immediate removal of piles of contaminated soil present on the sides of dirt roads for replenishment, wetting with water and encapsulators, washing of incoming and outgoing site vehicles, daily washing of equipment, yards and service areas, etc.) had a positive effect. 5,00 4,50

f/l

4,00 3,50 3,00 2,50 2,00 1,50 1,00 0,50 0,00 26/02/200 8

26/06/2008 26/04/2008

Figure 4:

26/08/200 8

26/10/2008

26/02/200

26/12/2008

26/06/2009 26/04/2009

26/10/200

26/08/200 9

26/02/201 26/12/2009

Concentration level of fibres measured from personal sampling (2008-2010) of remedial and territorial rehabilitation site workers.

With regard to the analytical results of personal samplings on other categories at risk, data for 2010 are given, with a data given in graph form on:  housewives  traffic police agents  refuse collectors  building workers. The results of analyses conducted over the course of a full year, corresponding to about 300 samples, are illustrated in Figures 5–8. 50 45 40 35 30 25 20 15 10 5 0 f eb-10

Figure 5:

apr-10

mag-10

giu-10

lug-10

ago-10

set -10

ot t-10

dic-10

gen-11

Concentration level of fibres measured from personal sampling of housewives (year 2010).

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

205

50 45 40 35 30 25 20 15 10 5 0 f eb-10

Figure 6:

apr-10

mag-10

giu-10

lug-10

ago-10

set -10

ot t -10

dic-10

gen-11

Concentration level of fibres measured from personal sampling of building workers (year 2010).

50 45 40 35 30 25 20 15 0 5 0 feb- 0

Figure 7:

apr-10

mag-10

giu- 0

lug-10

ago - 0

set-10

o tt-10

dic-10

gen-11

Concentration level of fibres measured from personal sampling of traffic police agents (year 2010).

50 45 40 35 30 25 20 15 10 5 0 feb-10

Figure 8:

apr-10

mag-10

giu-10

lug-10

ago-10

set-10

ott-10

dic-10

gen-11

Concentration level of fibres measured from personal sampling of refuse collectors (year 2010).

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

206 Air Pollution XIX It should be noted that in Italy current legislation does not establish a reference limit for personal samplings in living environments. Therefore, data on the concentration of asbestos fibres relative to housewives cannot be referenced to a given TLV. As established by the Italian Environmental Ministry in Italian Superfund , in this case it must respect the exposure threshold of 1 f/l in an urban environment, as recommended by the document Air Quality Guidelines for Europe (WHO, 2000). It is necessary nonetheless noted that excess values have been recorded with a maximum of 7.06 in the month of July. With regard to other categories examined (traffic police agents, refuse collectors, and building workers) deemed to be most exposed to this risk, as they perform work activities outdoors, again there is no reference limit for personal samplings in working environments not related to the activities of handling and cleaning up asbestos or items containing asbestos. In the latter case the reference TLV is 100 f/l. Accordingly, and in the same way, in working environments free from direct exposure to asbestos, this reference value is normally used. Below in graph form are the results obtained via PCOM analysis for these three worker categories. As above shown, in any case, the TLV value has been exceeded. With regard to environmental sampling (Fig. 9), monitoring data in which the exposure threshold of 1 f/l in an urban environment, as recommended by the document Air Quality Guidelines for Europe (WHO, 2000), has been exceeded, were analysed. Filed data were processed in graph form, showing 34 cases of excess values for the years 2000, 2008, 2009 and 2010.

Figure 9:

Map of samplings in the municipality of Biancavilla.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

207

Excess values regarding the year 2000 occurred in the spring/summer period, when there were concentration peaks >8 f/l in the month of April. With regard to 2008, excess values were recorded in the spring period, with concentrations >5 f/l in April. In the years 2009 and 2010 there were again excess TLV values in the spring/summer period. For these years fibre concentrations were measured at around 2 f/l. The above highlights a constant increase in risk in the dry period. This conforms to recorded meteorological data, which show in the warmer seasons conditions that facilitate the dispersion in the air of fibres (high winds and dry climate). Overall data collected from 2000 to 2010 in the database are also given in the graph in Fig. 10, showing a gradual reduction in the concentration of fluoroedenite fibres from values in excess of 8 f/l to values on average less than 2 f/l. This reduction is the result of a number of remedial activities (closure of quarrying activity, asphalting of town streets, removal of piles of contaminated soil, encapsulation of contaminated plaster in all public buildings, etc.) carried out in the cited NPL Site. The few times the TLV of 1 f/l has been exceeded have always been seen in conjunction with the moving of contaminated soil. 9

8

7

6

5

4

3

2

1

0

Figure 10:

Results of samplings carried out in the Municipality of Biancavilla (2000 – 2010).

From the map of Fig. 11 three town areas at greatest risk are identified: 1. the area situated in the extreme north and extreme south of the town, both devoid of asphalted surfaces (excess values in year 2000) and still today partially with dirt roads, with soil contaminated by fluoro-edenite;

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

208 Air Pollution XIX 2. 3.

the area around the Monte Calvario quarry where there are outcrops of contaminated rocks exposed to meteorological phenomena and a connected municipal dump for contaminated soils; the area to the south-east of the town, close to the municipal border between Biancavilla and the municipality of Santa Maria di Licodia, where there are worksite areas of the newly formed Circum-Etnea Railway, with trucks and service vehicles entering and leaving contaminated tunnels.

Figure 11:

Exceeding of threshold limits: results of samplings carried out in the Municipality of Biancavilla (2000 – 2010).

5 Conclusions The example of a GIS specially created for the Municipality of Biancavilla has illustrated the advantages of using this system to manage analytical data. In particular, this instrument makes it possible for authorities set up to protect workers and the exposed population to quickly pinpoint critical zones and easily consult data, thanks to implementation procedures and automated analysis. This has enabled competent local supervisory authorities (AUSL and ARPA) and the municipal administration to take suitable precautionary measures for both exposed workers and for urban living environments. This experiment in the sphere of a dedicated GIS has also illustrated the advantages of using these computerised technologies, which have proved to be a valid tool for quickly identifying situations of environmental danger, and consequently an indispensable tool when taking precautionary measures. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

209

References [1] F. Paglietti, S. Bellagamba, S. Malinconico, V. Di Molfetta, P. De Simone, M. Giangrasso “Asbestos presence on the Italian National Territory: Progress Report on Mapping and Remediation Activity” ASTM Johnson Conference: Critical Issues in Monitoring Asbestos, (Vermont, USA), 2008 [2] Gianfagna A., Oberti R., “Fluoro-edenite from Biancavilla (Catania, Sicily, Italy): crystal chemistry of a new amphibole end-member”, (2001). American Mineralogist. vol. 86, pp. 1489-1493.. [3] F. Paglietti, S. Malinconico, F. Damiani, P. De Simone “Natural Asbestos Contamination: Biancavilla’s Case” - ASTM 2008 Johnson Conference: Critical Issues in Monitoring Asbestos, (Vermont, USA) [4] F. Paglietti, V. Di Molfetta, S. Malinconico, M. Giangrasso, S. Bellagamba, F. Damiani. Italian Asbestos Mapping. World Asbestos Conference October 1-3 2009.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

This page intentionally left blank

Air Pollution XIX

211

BTEX concentrations in the atmosphere of the metropolitan area of Campinas (São Paulo, Brazil) A. C. Ueda & E. Tomaz School of Chemical Engineering, State University of Campinas, UNICAMP, Brazil

Abstract Benzene, toluene, ethylbenzene and xylenes (BTEX) are hydrocarbons present in fossil fuels and, consequently, are also present in fuel combustion emissions, as toxic organic volatile compounds. Concentrations of these compounds found in the atmosphere of urban regions are indicative of vehicular pollution. In this work were studied BTEX atmospheric concentrations from five sites of metropolitan area of Campinas (São Paulo, Brazil), which have different characteristics: (1) suburban site; (2) downtown site; (3) urban site A; (4) urban site B; (5) industrial. The most abundant compound found in all sites was toluene (2.4 – 10 μg m-3). B/T ratio was used to study predominance of vehicular emissions and X/E indicates the distance from the site to the sources due to different photochemical reactivity of the compounds. It was possible to observe that sites away from the urban region and located in the predominant wind direction have lower concentration of BTEX as well as lower X/E ratios, indicating that pollution is due to transport emissions. Keywords: atmospheric pollution, BTEX (benzene, toluene, ethylbenzene, xylenes), urban region, vehicular emissions, intensity of traffic, photochemical reaction, X/E ratio, B/T ratio, passive sampling, gas chromatography.

1 Introduction Atmospheric emissions of a variety of gaseous pollutants are of great concern, mainly in urban areas. Some compounds, such as BTEX, are precursors of ground level ozone and are also harmful to human health. Benzene is known for its carcinogenic effect [1] and xylenes’ oxidation products may be toxic or WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110191

212 Air Pollution XIX mutagenic, such as aromatic aldehydes and secondary organic aerosols [2]. BTEX are characteristic of fossil emissions and are reported as indicators of intense vehicular activity, corresponding to 60% of total non-methane VOC [3]. Metropolitan area of Campinas (São Paulo, Brazil) is located northwest of São Paulo city and is formed by 19 cities with a population of 2,797,137 inhabitants, 27,079 km2 and fleet equals to 1,425,125 vehicles. The region is surrounded by important highways and has also an international airport, Viracopos. The industrial sector is also privileged with several companies, such as: pharmaceutical, automotive, textile, food and petrochemical industries. The main objectives of this work were to study BTEX concentrations and proportions, as B/T and X/E ratios, as indicatives of dominant pollution sources. B/T ratio was studied in many cities around the world and observed values between 0.2 and 0.5 were found in sites with vehicular pollution predominance [4]. X/E ratio, instead, is based on different photochemical reactivity of xylenes and ethylbenzene (xylenes are more reactive than ethylbenzene). Therefore, higher ratios give higher xylenes concentrations and the plume is considered a fresh one, i.e., near emission sources [2].

2 Experimental 2.1 Materials Hydrocarbons liquid standards were used to produce calibration curves. Benzene, ethylbenzene and o-xylene were purchased from Merck Schuchardt OHG (purity > 99%), toluene was from Tedia (purity > 99.8%), and m and pxylenes from Acros Organics (purity 99%). Methyl alcohol (Merck KGaA) was used as solvent to construction of the curves. Samples were collected into stainless steel tubes (Supelco) filled with Tenax TA. During sampling period aluminum caps with protection sieve and clips were used to prevent from insects and to fix the tubes, respectively. Sample analysis were performed by thermal desorption (Automatic Thermal Desorption System ATD 400, Perkin Elmer) followed by gas chromatography (AutoSystem XL Gas Chromatograph, Perkin Elmer) with flame ionization detector. Capillary column for gas chromatography was NST-01 (60 m length, 0.25 mm inside diameter and 0.20 μm dimethyl polysiloxane film thickness) purchased from Nano Separation Technologies (NST). 2.2 Methods BTEX were sampled by passive method with period of exposition of ten days. After sampling, tubes were conditioned and analyzed in the same day. Sample analysis begins with two-step thermal desorption, where firstly the compounds are desorbed from the sample tube and are adsorbed into a cold trap during desorption time; then, the second desorption occurs from the trap directly to the column, where compounds will be analyzed. Tables 1 and 2 present the main parameters of thermal desorption and analysis by gas chromatography. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Table 1:

213

Desorption system parameters for thermal desorption and gas chromatography analysis. Parameter Line temperature Oven temperature Desorption time Valve temperature

Table 2:

Value 205°C 300°C 15 min 205°C

Parameter System pressure Inlet Split Outlet Split Desorb Flow

Value 20 psi 45 ml min-1 25 ml min-1 60 ml min-1

Gas chromatography analysis parameters.

Initial Step 1 Step 2

Rate (°C min-1) 2,0 1,5

T (°C) 35 60 80

Hold (min) 13 10 5

Step 3

7,0

100

10

Five different sites within the metropolitan area of Campinas were chosen based on the main wind direction in the region: 1) Suburban site (SU) – sample was taken in a location outside the urban area of Campinas, near an extensive green area; 2) Downtown site (DT) – sample was taken in front of a residential building near Campinas downtown, with great intensity of traffic; 3) Urban site A (UA) – sample taken in Campinas, in a mainly residential neighborhood, far from downtown city and with low intense of traffic; 4) Urban site B (UB) – taken in Paulínia, 10 km far from to Campinas, in a region of high intense of traffic, far from downtown and located in the main wind direction of the region (may receive pollution from Campinas); 5) Industrial site (IN) – sample taken inside an oil refinery area; according to the dispersion study in the local of greater influence of these industrial emissions.

3 Results and discussion Localization of sampling sites is presented in Figure 1 and geographical coordinates (UTM) are presented in Table 3. Table 3: Site SU DT UA UB IN

           

Coordinates from sampling sites.

City Campinas – SP Campinas – SP Campinas – SP Paulínia – SP Paulínia – SP

Coordinates  UTM 300,347.72 E 7,467,775.21 S  288,606.76 E 7,466,289.50 S  290,850.36 E 7,470,140.09 S  277,866.86 E 7,478,325.18 S  280,532.57 E 7,485,671.14 S 

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

214 Air Pollution XIX

Figure 1:

Localization of monitoring sites.

BTEX monitoring occurred between August 2008 and December 2009. Among the five sites monitored the greatest BTEX concentration was found in the industrial site (IN), inside the oil refinery, as expected, once this local represents direct influence of the emissions sources. The lowest concentration was found in site SU, where the intensity of traffic is low and there is no evidence of direct influence of emission sources. In all studied sites the compound found in abundance was toluene, corresponding, at least, to 44% of total BTEX. Data are presented in Table 4. Table 4:

Pollutant Benzene Toluene Ethylbenzene Xylenes

BTEX average concentrations for all monitoring period. SU c

DT c

Site UA c

UB c

IN c

(μg m-3)

(μg m-3)

(μg m-3)

(μg m-3)

(μg m-3)

0.90 2.4 0.73 1.5

1.4 6.5 1.1 3.9

1.2 3.5 0.71 1.9

1.4 5.5 1.1 2.4

2.1 10 1.2 5.0

It was noticed that concentrations observed in the Autumn–Winter months (April to September) were higher than in the Spring–Summer months (October to March). It is due to greater atmospheric stability in cold seasons; and favorable dispersion conditions are found in hot seasons. BTEX concentrations for studied periods are shown in Table 5.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Table 5:

215

Total BTEX average concentrations for all periods, for Spring– Summer months and for Autumn–Winter months. SU c

Period All Spring–Summer Autumn–Winter

DT c

Site UA c

UB c

IN c

(μg m-3)

(μg m-3)

(μg m-3)

(μg m-3)

(μg m-3)

5.5 5.7 5.3

13 11 16

7.3 5.9 9.1

10 7.6 14

19 17 20

Evaluation of B/T and X/E ratios during different periods also show difference of photochemical reactivities of the compounds involved. During Spring-Summer months, the intensive solar radiation enhances the photochemical reactions, reducing toluene and xylenes concentrations, due to their higher photochemical reactivities. Therefore, under these conditions, B/T ratios are high and X/E are low. According to data in Table 6, B/T ratio indicates that emissions in all sites are predominantly from vehicles (values between 0.2 and 0.5). Site SU and UA showed B/T ratio higher than other sites, although this places present the lowest concentrations of BTEX and show no direct influence of emission sources. This indicates that pollution in those sites is indicative of transport. Other sites showed B/T ratios above 0.25, indicating that these locals also present other sources of pollution rather than vehicular, such as fuel and solvent evaporation, although this one is predominant. Based on this parameter, vehicular predominance is also shown in the industrial site (IN). However, once this is an oil refinery, it presents the basic same profile of BTEX than vehicular emissions. Table 6: Period All Spring-Summer Autumn-Winter All Spring-Summer Autumn-Winter

B/T and X/E ratios for studied periods. Ratio B/T X/E

SU 0.38 0.39 0.36 2.05 1.53 2.98

DT 0.22 0.30 0.15 3.55 3.49 3.57

Site UA 0.34 0.48 0.22 2.68 2.28 3.07

UB 0.25 0.38 0.17 2.18 1.95 2.29

IN 0.21 0.29 0.14 4.17 3.75 4.58

X/E ratios are also presented in Table 6 and indicate photochemical aging of the emission plume. As xylenes are more reactive than ethylbenzene, a higher X/E ratio represents higher xylenes concentration and, consequently, closeness to emission sources. Sites SU and UA presented X/E lowest values (between 2.0 and 2.5), which, in comparison to other values, represent locals far from direct source influence and compounds emitted by the same emission sources. Site UA showed higher proximity to sources (X/E = 2.68) than site UB (X/E = 2.18), although the second location presented higher BTEX concentrations. This can be assigned to different sources that influence each site: UB have local sources and WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

216 Air Pollution XIX also receive pollutants by transport from the urban region of Campinas, due to wind direction. On the other hand, sites DT and IN showed greatest X/E ratios, indicating that plume are low photochemically aged, i.e., these locals are nearest to emission sources. Concentrations found in this work were compared to other references [4–7] as show in Table 7. The results shown in this work agree with value found in other cities around the world. Besides, BTEX concentrations found in Campinas are lower than in São Paulo, and more alike other smaller cities in Europe and also less industrialized ones. This can be attributed to less emission sources, as vehicles and industries, and favorable wind regime in the region, which promotes a good dispersion of atmospheric pollutants. Brazilian cities are also favored by ethanol based fuel, which reduces BTEX emissions, in comparison with European and other Latin American cities. BTEX concentrations, in μg m-3, comparison between this work and other references.

Caracas, VE [4]

Santiago, CH [4]

Tarragona, ES Historic Center [7]

Tarragona, ES Oil Refinery [7]

2.1

2.0

2,72

14,2

14,8

14,3

2,2

3,3

T

2.4

6.5

3.5

5.5

10

16

9,53

28,9

29,8

90,0

9,6

8,3

DT

Tarragona, ES Downtown [7]

Fives, FR [6]

1.4

IN

1.2

UB

1.4

UA

0.90

SU

B

Compound

São Paulo, BR [5]

Table 7:

E

0.73

1.1

0.71

1.1

1.2

4.1

1,23

5

6,5

13,9

2,9

7,0

m,p-X

0.94

2.7

1.3

1.6

3.4

7.7

3,36

16,4

25,2

40,2

7,3

4,8

o-X

0.52

1.2

0.65

0.79

1.6

2.7

1,18

5,7

8,9

14,6

1,9

8,2

4 Conclusions The study of BTEX concentrations in urban atmosphere and the relations between these compounds are parameters that help us identifying major sources and proximity to them. Concentrations of major pollutants, as BTEX, in the RMC showed values lower than other Brazilian metropolitan regions, as São Paulo, and other important cities around the world, indicating that there are important contributions by ethanol use as fuel and also by the good dispersion conditions in the region. B/T ratio is also an important indicator of atmospheric pollution from vehicular emissions, which have major importance in urban areas. The evidence of vehicular based pollution may help in the installation of maintenance and prevention programs in order to mitigate emissions. The results obtained for WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

217

RMC showed that all places monitored presented major influence of vehicular pollution. The study of X/E ratios is also important due to its representation of photochemical reaction and proximity to sources. This parameter can be useful in the evaluation of local emission sources and pollution transported from other regions. Industrial site showed that main COV contributions are from sources very close to the monitoring point. On the other hand, suburban site was the most remote location from Campinas urban region, although it presents pollution by transport from other regions upwind, such as São Paulo.

Acknowledgements The authors thank Fundação Araucária and Paraná State Government for financial support.

References [1] Mallorquí, M.R., Marcé-Recasens, R.M., Borrul-Ballarín, F. Determination of volatile organic compounds in urban and industrial air from Tarragona by thermal desorption and gas chromatography-mass spectrometry. Talanta, 72(3), pp. 941-950, 2007. [2] Monod, A., Sive, B.C., Avino, P., Chen, T., Blake, D.R., Rowland, F.S. Monoaromatic compounds in ambient air of various cities: a focus on correlations between the xylenes and ethylbenzene. Atmospheric Environment, 35(1), pp. 135-149, 2001. [3] Lee, S.C., Chiu, M.Y., Ho, K.F., Zou, S.C., Wang, X. Volatile organic compounds (VOCs) in urban atmosphere of Hong Kong. Chemosphere, 48(3), pp. 375-382, 2002. [4] Gee, I.L., Sollars, C.J. Ambient air levels of volatile organic compounds in latin american and asian cities. Chesmophere, 36(11), pp. 2497-2506, 1998. [5] Albuquerque, E.L. Volatile organic compounds in urban atmosphere of metropolitan area of São Paulo (PhD Thesis). School of Chemical Engineering, University of Campinas, Brazil, 2009. [6] Borbon, A., Fontaine H., Locoge, N., Veillerot, M., Galloo, J.C. Developing receptor-oriented methods for non-methane hydrocarbon characterization in urban air – Part I: source identification. Atmospheric Environment, 37(29), pp. 4051-4064, 2003. [7] Ras, M.R., Borrull, F., Marcé, R.M. Sampling and preconcentration techniques for determination of volatile organic compounds in air samples. Trends in Analytical Chemistry, 28(3), pp. 347-360, 2009.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

This page intentionally left blank

Air Pollution XIX

219

The development of an ESEM based counting method for fine dust particles and a philosophy behind the background of particle adsorption on leaves M. Ottelé1, W. J. N. Ursem2, A. L. A. Fraaij1 & H. D. van Bohemen1 1

Faculty of Civil Engineering and Geosciences, Delft University of Technology, The Netherlands 2 Department of Biotechnology, Faculty of Applied Sciences, Delft University of Technology, The Netherlands

Abstract The multi scale benefits of urban greenery (green façades and green roofs) have attracted more and more interest of recent research work. The multi scale benefits of vegetation vary from; mitigation of the urban heat island effect, stimulation of the ecological value and biodiversity, aesthetical reasons and for example air pollution reduction. Air pollution control is at the moment mainly focussed on the reduction of fine particle concentrations. Particulate air pollution is damaging for the human health, it causes cardiovascular and lung diseases. Especially dust particles smaller than 2.5 micrometers are of great interest because they can be deeply inhaled into the respiratory system. To determine the effect of leaves on particle adsorption, micrographs are taken of ivy (Hedera helix) leaves using an Environmental Scanning Electron Microscope (ESEM). The examined leaves are exposed to a simulated rainfall in order to determine a method for particle counting on leaves and to determine the self cleaning effect of adsorbed particles on ivy leaves. The self cleaning effect is considered to be an important factor in the effectiveness of particle adsorption by leaves and the potential for resuspension of particles. Particles on pre- and post-rain leaves were counted via the ESEM micrographs using an image analyzer. Results showed that there is no significant effect on particle loss due to rain in the performed experiment. Our findings suggest that a strong Van der Waals bonding between WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110201

220 Air Pollution XIX particle and leaf surface plays an important role in the retaining process of fine particles on the leaf surface. Keywords: green façade, fine particle accumulation, environmental scanning electron microscope, simulated rainfall, Hedera helix.

1 Introduction Green façade designs offer numerous economic, social and environmental benefits such as greenhouse gas emission reduction, adaptation to climate change, air quality improvements, habitat provision and improved aesthetics. Also sound reductions are possible by the use of vegetation [1]. Despite these benefits, a widespread market penetration of greening technologies over the world remains still in its infancy. Greening the façades of urban buildings or infrastructural projects (i.e. sound barriers, tunnel alignments, etc.) using climbing plants or living wall systems (LWS) modifies the interaction of the building system with the surrounding atmosphere. For example: plants have the ability to dissipate absorbed solar radiation into sensible and latent heat [2]. This is often related to the urban heat island effect, but it has also an effect on the vertical air velocity [3] and is thus (indirectly) related to particle deposition processes. With other words, a specific microclimate will be created around a green building envelope. This microclimate could not only improve the outdoor or indoor climate but will also have an effect on the distribution and accumulation of particles inside the street canyon due to the filtering capacity of climbing plants [4]. The aim of this paper is to classify and to provide a first step in a comprehensive study on the potential use of green façades to improve the local air quality in urban environments. In the presented research, adsorption and removal of particulate matter by vegetation are discussed because of the associated increased morbidity and mortality aspects of inhalation fine atmospheric particles by humans. Especially finer particles (those with an aerodynamic diameter of <10 µm (PM10), pose a long-term threat to the human health, mainly to the human respiratory functions [5]. In general, the smaller the particles, the deeper they penetrate into the respiratory system were it is taken up into the blood. In this way, respiratory and/or cardiovascular disease arise [5]. Research done by Pope et al. [6] in the United States on the life expectancy of humans and the concentration of fine particulate air pollution, shows that there is an increase in life expectancy when there is a decrease of 10 µg per cubic meter in the concentration of fine particles in the ambient air. Literature review shows the importance of altering the amount of fine particles in the air to improve human health. Although this information, less research has been done on the potential impact of lowering the amount of fine particles by vertical greened surfaces and on counting the amount of adsorbed fine particles by leaves. Past research methods focussed on particle mass levels through examine the effluent of washing urban tree leaves [7–10]. In assessing the potential benefit of fine particle adsorption by green façades this paper examines if there is an WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

221

influence of rainfall simulation on particle retention by vegetation. Besides that the paper intended to classify the amount of particles by a counting method based on Environmental Scanning Electron Microscope (ESEM) images. The method of using micrographic images in combination with an image analyzer (Image J) enables to study and identify particle size, origin and amount directly on leaves. Surfaces of vegetation are recognized as a terrestrial sink for atmospheric particulate matter, consisting of particles which are highly variable with respect to origin, chemical and physical properties, elemental composition, and potential biological and environmental impact. Vertical greened surfaces may be especially efficient filters of airborne particles [11] because of their high surface to volume ratio of foliage, abundant petioles and twigs, and due to hairy, wax structure or rough leaf surfaces and shapes. To estimate the filtering effects of façade greening, it is necessary to study the relationships between the retention of particles on the leaf surfaces and the local pollutant concentration. Also the resuspension of already adsorbed particles is an important parameter in the purification process. Processes which have to be taken into account for resuspension of particles are among other things: wind, rainfall and falling of leaves (figure 1).

Figure 1:

Conceptual model particle circulation.

A literature survey undertaken by the authors on the different mechanisms between particles and leaves shows that resuspension of particles is commonly stated but hardly investigated. Once fine particles are adsorbed by the leaf surface it is important to know if they still form a danger for human health as discussed earlier. Two main forces act on small, moving airborne particles: one is the force of gravity and the other is the viscous force exerted by the air through which the particles are moving [12]. If the particles are electrically charged they will also be subject to electrostatic forces, which may alter the dispersal and deposition patterns of charged particles compared with uncharged ones [12]. Some particles may be adsorbed into the tree but most are retained on the plant surface [11]. Once particles are adsorbed to the leaf surface it is important to know the pathway that will bring these particles to the subsoil. Also the particle sizes involved in this process are of interested. In this paper a simulated rainfall experiment was carried out on Hedera helix leaves (common ivy) to get more insight into the percentage of (fine) particles that can be washed WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

222 Air Pollution XIX off or that will remain on the leave surface with the potential to end up (resuspended) in the air again. According to Nowak et al. [13] the resuspension of particles can be up to 50% of the adsorbed particles. The objective of this study aimed in defining the reduction of air pollution and stimulating the ecological performance of vertical green (green façades, sound barriers, etc.) with plants. The interaction or filtering effect between (vertical used) vegetation and air pollution (particulate matter) is not well studied yet. Especially the effect of rainfall on particle retention and resuspension is unclear. The approach in this research is not to measure the mass of particles collected on the leaves but the amount of adsorbed particles. Measuring the mass of particles instead of the size and amount of particles ignores the risk for human health with respect to hearth and lung diseases. Since particle size is correlated with lung diseases (it is important to know how many of those fine or ultra fine particles are adhered on the leaf surface. Counting particles on a specific leaf area seems therefore more suitable for this experiment.

2 Materials and methods Leaves of common ivy (Hedera Helix) were chosen for this experiment. The leaves were collected from a vertical greened fence near Steenbergen (The Netherlands) in the beginning of October 2008, after a period of 6 days without rain. Eleven adult ivy leaves and only entire green leaves were taken randomly from the outside ivy foliage of the greened fence. The collected leaves were sealed and labelled separately in plastic containers to exclude the possibility of contamination after sampling. The sealing procedure was done in a way to keep the leaf surface untouched until the examination in the Microlab of Delft University of Technology. To distinguish the difference between the amount of particulate matter before and after rainfall on the leaves (only the upper side of the leaf is examined in this research), micrographs where taken with an Environmental Scanning Electron Microscope (Philips XL30 ESEM with a tungsten filament) at different magnifications namely 100x, 500x and 5000x. An experimental procure was addressed in order to make micrographs before and after rainfall simulation on the same leaf and on the same spot, valid for each magnification. The micrographs are always taken in the middle of the leaf lefthanded near the central nerve (figure 2, right photograph). When a spot is found, the spot is fixed in the middle of the view and the micrographs are taken at different magnifications. After this session the leaf is exposed to a simulated rainfall. Therefore the leaves were placed in a tripod on a bench in which all rainfall events were performed. Rainfall was simulated with a pressurized system utilizing a rain nozzle designed to project a downward spray. The nozzle was placed about 60 cm above the leave(s) and a uniform spray was attained to simulate the rainfall. The leaves were subjected to a simulated rainfall event of 15 minutes at a rainfall rate of 80 mmh-1 with normal tap water (pH 8.0) measured with a funnel with a diameter of 13 cm. A relative high rainfall rate was chosen for this experiment to make sure that the leaf was fully wetted. After the simulated rainfall event, the leaf was placed WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

223

again in the ESEM chamber with the positioning holder (figure 2, left photograph). Micrographs were taken at the same spot as before the rain simulation. This procedure was repeated for each of the collected leaves. In this system the computer is responsible for beam control, image analysis, data processing and data storage. For image analysis, the backscattered electron signal is used to create a binary image of the leaf surface with the adhered particles.

Figure 2:

Sampling procedure using ESEM analysis.

After collecting microphotographs the particle counting was done automatically with an image analyzer software package called Image J. Particle analysis requires that the image is a “binary” image (i.e. black or white). The biggest issue is to distinguish the particles from the background (threshold). See figure 3 and figure 4 to go from raw to threshold. The automatic threshold function used by Image J was applied, in some cases manual correction of the automatic threshold value was needed. Further information about the program can be found on [14]. Particles that are slightly overlapping in a threshold image must be separated; this was done with the watershed function in Image J. Once the particles have been successfully threshold and watershed, they can be analyzed to obtain information regarding particle size and numbers. In the analysis no boundary to the circularity value was given (i.e. a value of 1.0 indicates a perfect circle), which means that all various shapes of particles were count. Per magnification (100x, 500x, respectively 2500x) the different adsorbed particles were count, also weight factors, respectively 1, 25 and 2500 times, were used to compensate for the loss of counting area (zoom effect). In addition, the cross sectional diameters of each of the particles were calculated by assuming that a calculated area belongs to a certain aerodynamic diameter. The experiment and procedure for counting was repeated and carried out for each individually examined leaf. The presented counting procedure was done by second year’s students from the Faculty of Civil Engineering of Delft University of Technology and the students are instructed how to analyze the micrographs. Each leaf was measured WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

224 Air Pollution XIX (a)

Figure 3: (a)

(b)

(a) ESEM micrograph before rainfall, and (b) after rainfall. (b)

Figure 4:

(a) Threshold micrograph before and (b) after rainfall.

by a new group of students, a group exists out of six students and each individual student in a group analyzes the same photo series before and after simulated rainfall. The groups of students were used to examine if there is an influence on the sensitivity of the used image analyzing procedure. The total of used specimens (different leaves) and thus also the number of groups for this experiment was eleven (each leaf is thus counted by six different persons).The collected data (response variable) from the counting procedure was evaluated by an replicated random block design (ANOVA) to test an eventual influence of the students (independent variable) on the counting procedure and the effect of simulated rainfall (independent variable) on particle loss Each collected data set was filtered (the six outcomes of the counting procedure for each leaf before and after rainfall) by testing for outliers. This filtering procedure was done on the basis of Chauvenet’s criteria assuming that each individual data set follows the normal distribution. The rejected outliers were considered as a “malfunctioning” of the students. The statistical analysis was performed by SPSS 16.0 software package. The level of significance used in the analysis and throughout the paper is alpha 5%. The outcome of this analysis will give more insight into the relevance and interactions between the retention of particles on leaf surfaces and the influence of rainfall on the retention. Also the sensitivity of the counting procedure (with relation to the need of qualified personnel) can be examined. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

225

The statistical outcome enables also HØ hypothesis checking. In this study, we hypothesized that rain changes the amount of adsorbed particulate matter on ivy leaves and that students will not have an effect on the outcome of counting particles.

3 Results 3.1 Outcome of counting particles before and after simulated rainfall The result of counting particles on a specific leaf area (950x1275 μm2) results in a particle distribution (figure 5) for the upper side of the leaf. Particle sizes ≥ 10 μm appeared to be rather rare compared to particles sizes ≤ 10 μm. Table 1:

Outcome of counting fine particles based on ESEM images; before and after simulated rainfall.

Blocks

Treatment after rain (y) average out of six students (n=6) 81396

difference (x-y)

leaf 1

before rain (x) average out of six students (n=6) 96952

leaf 2

78277

64609

13668

leaf 3

34648

27410

7238

leaf 4

33376

28296

5080

leaf 5

86644

87235

-591

leaf 6

130134

51011

79123

leaf 7

29285

21016

8269

leaf 8

75499

52146

23353

leaf 9

199813

160142

39670

leaf 10

185010

164838

20172

leaf 11

48413

144531

-96118

(n=11)

15556

Particles larger than 10 μm will thus not be studied in this paper. Almost all of the peaks were found in the range up to 2.5 μm. The smallest particles that are found with the applied measurement technique are in the range of 0.2 μm. Rainfall did not have a significant effect on the number of particles retained on the leaves. The number of particles (table 1) before the rainfall ranged from ± 29000 to ± 199000 particles per μm2 and was not significantly different from the number of particles (respectively ± 21000 and 160000 particles per μm2) retained on the leaves after the simulated rainfall event (figure 5). For leaf 5 and 11 a negative value was found for the amount of particles after the measurement. This means more particles entered the counting area after the simulated rainfall event.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

226 Air Pollution XIX 200000 180000

Average amount of particles

160000 140000 120000 100000

Before rain After rain

80000 60000 40000 20000 0 1

2

3

4

5

6

7

8

9

10

11

Leaf num ber

Figure 5:

Average amount of particles found on Hedera helix leaves before and after simulated rainfall.

3.2 Evaluation of statistical analysis Throughout this paper differences between samples are indicated to be significant if they were significant at α< 0.05 based on the results of the paired ttest (table 2). The F-test is done on each individual counting procedure. The outcome of the statistical analysis identified that in the presented study no significant differences were found between the treatments. The null hypothesis that rain may not have an effect (P=0.186) is thus rejected. However the hypothesis that students (P=0.000) will not influence the outcome of particle counting is accepted.

4 Discussion In this paper the effect of rain on the cleaning effect of leaves has been studied. Eventual cleaning effect of leaves by rain will lower the chance that particles will be affected by wind loads (resuspension). Rainfall may influence thus (in) directly the capacity and efficiency of the leaf on particle reduction in the ambient air. Simulated rainfall shows that the expectations with respect to particle wash off and the self cleaning effect of Hedera helix leaves are far away from ideal. For leaf 5 and 11 consequently we found more particles in the counting area after the simulated rainfall. This means that particles are moved from the area around the examined spot into the counting area. With other words particles are not always fixed on the leaf surface which can be explained by figure 6, were we see an accumulation of particles especially near the edge or tip of the leaf. The photograph also suggests that there is a particle transfer over the leaf surface. The overall result however of the experiment shows that the simulated rainfall has a minor effect on the self cleaning effect of Hedera helix leaves on particulate matter. This leads us to the next question whether particles retained on the leaf surface after the rainfall have the potential to end up WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

227

(resuspended) into the air again due to wind turbulences. The question that can be raised concerning the outcome of this experiment is how likely this resuspension of particles is after a rainfall? Additional wind experiments concerning particle loss are necessary to detect the effect of air turbulence in regard to this statement.

Figure 6:

Accumulation and clogging of particulate matter on the edge of a leaf.

Most of the particles are still present on the leaf surface after the heavy simulated rainfall, so the particles are really well embedded on the leaf surface. Is there an explanation for the minor particle loss due to the effect of rainfall and that the particles are well embedded? Is there a role for the leave composition (wax layer, hairy, rough surface, etc.), the leaf boundary layer or electrostatic behaviour of particles and leaf surface? The wax layer can be of importance to stick the particles, but is most unlikely to be able to hold on after the rain experiment of a simulated rainfall event of 15 minutes at a rainfall rate of 80 mmh-1 with normal tap water (pH 8.0). Wax structures consist of saturated verylong-chain fatty acids (commonly C20 to C34) and in surface structure not equipped to hold fine dust particles [15–17]. A well understood surface wax structure of the Lotus, Nelumbo nucifera, is even been used for its properties of self cleaning effect, due to the laminar special wax structure [18]. This also explains the observed lacking of large fine dust particles of 10 μm or more on the leaf surface at all. Besides the difficulty of fixation, all large particles of 10 μm or more are washed away at a high relative humidity together with low temperatures, which could easily form a dew point effect on the leaf surface, or in cases of precipitation, or in situations of a dry environment due to wind turbulence. The hairs are lacking on common ivy, so this factor is of no importance concerning the fine dust particle adhesion. The leaf boundary layer is WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

228 Air Pollution XIX extremely related to the wax structure of the surface of the epidermal cells. As mentioned before, the leaf boundary layer and the fixation of fine dust particles of 10 μm or more can been seen in the same way as the given structural analysis of wax surfaces. In conclusion, it is very difficult to hold these large fine dust particles of 10 μm or more on the leaf surface of common ivy based on the leaf composition (wax layer, hairy, rough surface, etc.) or on the leaf boundary layer. If a boundary layer could be part of the adhesion factor after the mentioned rainfall, then it can only remain the adhesion properties of fine dust particles when there is a strong bonding of fixation. The only possible strong bonding on the molecular level which resists the test of a 15 minutes rainfall with a rate of 80 mmh-1 will be a Van der Waals bonding. A Van der Waals bonding can only be induced due to a transition of another extra energy source, like an electrical charged fine dust particle. The main sources of fine dust particles of 10 μm or more are known from sand-, clay- or biological origin. In cases of smaller distributions, particles of less than 10 μm, the main origin is from gas exhaust of combustion and engines. The result of the findings in this research on common ivy shows clearly a dominant persistence of particulate matter smaller than 10 μm in diameter. The samples of common ivy mentioned in this research are all exposed to the influence of traffic. Fine dust particles of combustion and engines consistently exhibits in the soot mode of 20 till 200 nm size with a substantial fraction of 40 till 60 percent of charged particles [19]. This means that more then half of all fine dust particles of combustion or engine origin are charged and these particles will be neutralized as soon as it touches the leaf surface. The common ivy is connected with its root system to the soil, so per definition charged in the same manner as the soil it self or in other words the electrical charge can be seen as grounded [20, 21]. In short, the particulate matter either gains electrons in cases of positive charged fine dust particles or lose electrons in cases of negative charged fine dust particles. Because no electrical charged energy can get away in any other form, as soon as a fine dust particle touches the leaf surface of the common ivy, the electrical charge will be converted into a strong Van der Waals bonding. A Van der Waals bonding is known in physics to be an extreme strong fixation bonding, so this explains and underlines the positioned particulate matter on the leaves of common ivy after the 15 minutes at a rainfall rate of 80 mmh-1.

5 Conclusions Since the research was focused on the effect of rainfall on particle retention on the leaves of common ivy (Hedera helix) we can conclude that the cleaning effect of rainfall for the fine and ultra fine particles is very low. However, the observed phenomenon of the remaining of the particulate matter on the leaf surface brought us to basis principle of physics of a Van der Waals bonding as the only possible explanation of fixating after 15 minutes at a rainfall rate of 80 mmh-1. The electrical charged particles of gas exhaust of a combustion or engine source can be considered as the most important factor of the remaining of the particulate matter of less than 10 μm on the leaf surface of common ivy of our WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

229

research. The fine dust particles of 10 μm or more have not been observed in this study on the leaf surface of all sampled common ivy. The adhesion of these relative large fine dust particles is possible washed or blown away as mentioned and discussed in this article.

Acknowledgements The Microlab of Delft University of Technology, Faculty of Civil Engineering and Geosciences is acknowledged for the permission to make use of the necessary research facilities. The authors would also like to thank the second year students of the year 2008 for collecting the data (analyzing micrographs) needed for this experiment. We also thank Mr. A. Thijssen for his technical support, making the ESEM micrographs and data processing.

References [1] Pal, A.K., Kumar, V. and Saxena, N.C., Noise attenuation by green belts, Journal of Sound and Vibration, 234-1, 149-165 (2000). [2] Stec, W.J., Van Paassen, A.H.A. and Maziarz, A. Modelling the double skin façade with plants, Energy and Buildings, 37, 419-427 (2005). [3] Minke, G.,Witter, G., 1982. Häuser mit grünen pelz. Ein handbuch zur hausbegrünung. [4] Bruse, M., Thönnessen, M. and Radtke, U., Practical and theoretical investigation of the influence of facade greening on the distribution of heavy metals in urban streets. Proceedings International Conference on Urban Climatology & International congress of Biometeorology, Sydney, 8-12 (1999). [5] Pekkanen, J., Timonen, K.L., Tiittanen, P., Vallius, M., Lanki, T., Sinkko, H., 2000. Exposure and Risk Assessment for Fine and Ultrafine Particles in Ambient Air. National Public Health Institute. [6] Pope, A. C., Ezzati, M., Dockery, D.W., 2009. Fine-Particulate Air Pollution and Life Expectancy in the United States. The new England journal of medicine; N Engl J Med 2009; 360:376-86. [7] Bartfelder, F., Köhler, M., 1987. Experimentelle untersuchungen zur function von fassadenbegrünungen, Berlin. [8] Freer-Smith, P.H., Beckett, K.P., Taylor, G., 2004. Deposition velocities to Sorbus aria, Acer campestre, Populus deltoids x trichocarpa ‘Beaupre’, Pinus nigra and x Cupressocyparis leylandii for coarse, fine and ultra-fine particles in the urban environment. Environmental Pollution 133 (2005) 157-167. [9] Maher B, Moore C, Matzka J. Spatial variation in vehicle-derived metal pollution identified by magnetic and elemental analysis of road side tree leaves. Atmospheric Environment 2008:42: 364-373. [10] Thönnessen, M., 2002. Elementdynamik in fassaden begrünendem wilden Wein (Parthenocissus tricuspidata), Kölner Geographischer Arbeiten, Heft 78, Köln. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

230 Air Pollution XIX [11] Powe, N.A., Willis, K.G., Mortality and morbidity benefits of air pollution (SO2 and PM10) adsorption attributable to woodland in Britain, Journal of Environmental Management , 70, 119-128 (2004). [12] McCartney, H.A. et al., Electric charge and the deposition of spores of Barley Mildew Erysiphe Graminis, Atmospheric Environment, 16-5, 1133-1143 (1982). [13] Nowak, D.J., Mcpherson, E.G., Rowntree, R.A., Chicago’s Urban Forest Ecosystem: Results of the Chicago Urban Forest Climate Project, United States Department of Agriculture, General Technical Report NE-186, 1994 [14] http://www.epa.gov/air/airpollutants.html. [15] Koch, K., Ensikat, H.J. The hydrophobic coatings of plant surfaces: Epicuticular wax crystals and their morphologies, crystallinity and molecular self-assembly. Micron, 39-7, 759-772 (2008). [16] Niemietz, A., Wandelt, K., Barthlott, W. and Koch, K. Thermal evaporation of multi-component waxes and thermally activated formation of nanotubules for superhydrophobic surfaces. Progress in Organic Coatings, 66-3, 221-227 (2009). [17] Samuels, L., Kunst, L. and Jetter, R. Sealing plant surfaces: cuticular wax formation by epidermal cells. Ann. Rev. Plant Biol., 59, 683-707 (2008). [18] Müller, F., Michel, W., Schlicht, V., Tietze, A. and Winter, P. SelfCleaning surfaces using the Lotus effect. Handbook for Cleaning/Decontamination of Surfaces, 791-811 (2007). [19] Mariq, M. in Nanoparticles in Medicine and Environment, Marijnissen, J. & Gràdon, L., 19-37 (2010). [20] Becquerel, A.C., Eléments de physique terrestre et de météorologie (1847) [21] Becquerel, A.C., Des forces physico-chimiques et de leur intervention dans la production des phénomènes naturels (1875).

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

231

Synthesis of metal oxide nanostructure and its characterization as gas pollutant monitoring B. Yuliarto, M. Faizal, M. Iqbal, S. Julia & T. Nugraha Processing Materials Laboratory, Department of Engineering Physics, Institut Teknologi Bandung (ITB), Bandung, Indonesia

Abstract The characteristics of metal oxide thin films have been studied for gas pollutant monitoring systems. The metal oxide sensitive layers of ZnO are deposited by sol gel method using a chemical bath deposition technique at moderate temperature. The resulting thin films have been characterized using XRD and SEM to confirm the surface structure of the film. The scanning electron microscopy reveals that ZnO thin films have a nanostructure morphology which is supposed to open a larger area to the gas targeted. The sensor device was prepared using alumina as a substrate and a gold finger electrode on the top of the sensors. Gas pollutant monitoring systems have been developed using the resulting sensor device after the calibrating process. The gas monitoring system can measure and save the data in the system so that the trend of measured pollutant can be analyzed for a certain period of time. The design and electronics structure of the gas pollutant monitoring system in terms of CO emission are described. This monitoring system can be used as a sensing node and can form a dense real-time environmental monitoring network. Keywords: thin film, ZnO, CO, gas pollutants, nanostructure, monitoring, sensors.

1 Introduction In recent times there have been growing concerns about the consequences of air pollution on our atmosphere. This growing awareness is enhanced because atmospheric pollution is not only an environmental problem but also induces serious health hazards on humans. Air pollution emerged around the world in the last decade as a result of explosive industrial growth. After that, efforts were WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110211

232 Air Pollution XIX made to monitor and reduce the emission of pollutants resulting from factories and industrial areas, as well as vehicles. In many countries, there has been a regulation to establish a network of environmental quality monitoring stations, especially in their main cities, to give information and record the data about the environmental quality on a regular basis. However, there have been several limitations of the stations including the high price and lengthy air sampling data. The measurement technique used to detect the pollutants gas is passive samplers or a commercially spectroscopic gas analyzer yield longer and non on time monitoring systems. A gas sensor based on metal oxide should be developed so that the sensor can detect in situ and real time response. Moreover, the nanostructure of the ZnO thin films would give an advantage of having large accessibility for the targeted gases [1–4]. This research proposes the design and fabrications of gas pollutant monitoring systems based on ZnO nanostructure thin films. The using of ZnO is based on others’ research which shows that the thin films of ZnO find many promising applications due to their eccentric properties including a large exciton bonding energy of 60 MeV, non-toxicity, good electrical, optical and piezoelectrical behavior, and low cost [5–7]. ZnO is particularly attractive to researchers to be used as a sensor because ZnO has the desired and required typical properties such as resistivity control over the range 10-3–10 5 Mohm cm, excellent thermal and electrical stability. Here, we present the preparation of ZnO nanostructures thin films on an alumina substrate. The substrates were printed onto a silver finger electrode before the ZnO is deposited. The resulting sensor devices have been constructed as a real time monitoring system. The ZnO surface properties have been studied and the effects on the sensing performance have been measured. The sensitivity of pollutants in terms of CO gas has been measured in different operating temperatures.

2

Experiments

A thin layer was deposited by chemical bath deposition (CBD) on alumina substrates that have been cleaned. The solution was synthesized from zinc nitrate tetrahydrate and the dissolving of urea in the DI water. Zinc nitrate tetrahydrate, Iron (III) nitrate nonahydrate, and DI water + ethanol were dissolved at room temperature and stirred using a magnetic stirrer for 30 minutes. Urea is added to the solution and stirred for 30 minutes to produce a homogeneous solution. Alumina substrate has been cleaned up, immersed and placed in a standing position in a clear solution. Substrates were immersed in clear CBD solution, kept in the furnace at a temperature of 60oC for 24 hours. Substrates that have been deposited are removed from the place of immersion and washed using acetone to stop the growth of crystals. Then the films were rinsed in DI water several times and dried at room temperature for one hour in air atmosphere. A transformation into ZnO was conducted by calcination of the films at 300oC in air atmosphere for 30 minutes with 20oC/minute heat rate [10]. X-ray Diffraction (XRD), and Scanning Electron Microscopy (SEM) were employed to study the WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

233

phase and the surface morphology of the films. The crystalline phase of resulting thin films were analyzed by X-Ray Diffractometer using Philips Analytical XRay. Diffraction patterns obtained with Cu radiation (λ = 1.54060 Å) through generator voltage of 40 kV and current 25 mA. Tool set in step scan mode with 0.02° 2θ step size and 0.5 seconds step time in the range 20o–90o 2θ degree. The resulting thin film is then constructed in the sensor device using alumina substrate and the finger electrode on the bottom of the ZnO thin films. The silver paste of finger electrodes was applied on both ends of the films for making ohmic contacts so that the electrical measurements can be characterized. The sensor device is then connected into the integrated acquisition data system using Arduino uno of micro chip so that the results of sensing measurements can be shown in real time and the data can be transferred into the data base server. The sensing performance of CO pollutants gas is characterized using the line monitoring systems as shown in Fig. 1. The CO concentrations are confirmed using the commercial sensors of Bhacarach Smart Sensor.

Figure 1:

Characterization systems consist of reference room, test room and data acquisition system.

3 Results and discussions Fig. 2 shows the results of XRD measurement on samples of the sensor which is made from alumina substrate. From the XRD result, it is indicated that the zincite crystalline phase was successfully formed and not indicated of another formation crystalline phase. Intensity of a significant peak occurs at certain angles and this is in accordance with the standards of the Joint Committee on Powder Diffraction Standards (JCPDS). The ZnO nanostructure thin films were synthesized on the polycrystalline alumina substrate. Because of its polycrystalline phase and since the lattice WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

0 Figure 2:

20

40 2‐Theta60

(202)

Al2O3 Al2O3 (102) Al2O3 (110) Al2O3 (103) (201)

(002) (100) Al2O3 (101)

Al2O3

Intensity (a.u.)

234 Air Pollution XIX

80

100

XRD patterns of grown ZnO nanoballs on alumina substrate.

mismatch between alumina and ZnO is quite low, lead to the development of fine particle and uniform nanostructure with good adhesion to substrate [12]. Therefore, in XRD investigation, it is found that many patterns appear in high intensity due to the low lattice mismatch hence interaction between substrate surface and condensing species in constructive interferences. And the highest intensity of sample A is laying on (100) plane. The SEM images of the resulting thin films are shown in Figs. 3 and 4. Fig. 3 shows the surface morphology of the ZnO thin film in low magnifications which indicate that the ZnO layers with many nanostructures like forests are formed well. Moreover the high magnification of the surface morphology of ZnO thin films is shown in Fig. 4. The nano ball structure is formed in the thin film indicate that the large surface area formed. This nano ball structure would have good opportunity for the gas access so that the sensor will have good performance.

Figure 3:

The SEM image at low magnifications of ZnO nanostructure thin film deposited on the alumina substrate.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Figure 4:

235

High magnifications SEM images of nanoballs in the range of 250nm.

The ZnO nanostructure sensor is constructed in the gas pollutant monitoring system which can report the result in real time and regularly update the display on the web site. The complete diagram structure of the gas pollutant monitoring systems can be seen in Fig. 5.

Figure 5:

The real time monitoring systems based on ZnO nanostructure thin film.

The sensor of ZnO nanostructure thin film (block diagram #1) detects the presence of the target gas in air (CO) in the certain concentration. In order that the sensor can work optimally, it requires a basic measurement circuit as shown in block diagram of no. 2. This circuit heats the internal heater sensor up to a certain temperature (around 200oC) so that the sensor can measure the target gas at its best performance. The output from these sensors, which is in voltage, is still an analog signal. Before it is converted into a digital signal, the analog signal needs to be adjusted to fit the input range of the microcontroller (Arduino Uno) whose range is between 0–5 volts. These adjustments use the op-amplifier as a voltage amplifier mounted on the basic measurement circuit. The A/D converter utilizing Analog input pin of Arduino Uno (block diagram #3) is used to convert into a digital signal. We set pin A0 and A5 as input of CO yielded from the sensors. After that, through the Arduino software, we compiled into digital signals (ppm value) which then displayed in parallel on the LCD screen (block diagram No. 4) and displayed to the website via DFR Duino ethernet WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

236 Air Pollution XIX shield which is connected to the server by LAN’s cable network type RG45 (block diagram # 5). Fig. 6 shows the sensor characteristics of response dynamics at 600 ppm and 300 ppm of the gas CO. The measurements are performed at 200oC. The clear decreasing at significant levels both on 300 ppm and 600 ppm are happened suddenly after the explosions of the CO gas. The change of resistivity value indicates that the sensor has good sensitivity as well as fast response time both in 300 ppm and 600 ppm of CO gas. When a metal oxide semiconductor of ZnO nanostructure thin film is heated at a certain high temperature in air, oxygen is adsorbed on the semiconductor surface with a negative charge. Then donor electrons in the semiconductor surface are transferred to the adsorbed oxygen, resulting in leaving positive charges in a space charge layer. Thus, surface potential is formed to serve as a potential barrier against electron flow. Inside the sensor, electric current flows through the conjunction parts (grain boundary) of ZnO micro crystals. At grain boundaries, adsorbed oxygen forms a potential barrier which prevents carriers from moving freely. The electrical resistance of the sensor is attributed to this potential barrier. In the presence of a deoxidizing gas, the surface density of the negatively charged oxygen decreases, so the potential barrier in the grain boundary is reduced. The reduced potential barrier decreases sensor resistance. CO ON

CO OFF

30

Response (%)

25 300ppm 600 ppm

20 15 10 5 0 0

Figure 6:

5

10

15

20 25 30 Time (Minutes)

35

40

45

The response dynamics of ZnO nanostructure thin film.

4 Conclusions The synthesis process of chemical bath deposition technique has been performed successfully to grow the ZnO thin film on the alumina structure. The resulting ZnO thin film have the nano ball structure which yield good accessibility for the targeted CO gas adsorbed to the surface of the ZnO sensitive layer. The sensors WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

237

have good sensitivity on the exposure of CO gas at 300 ppm and 600 ppm. The on line and real time measurement systems have been constructed so that the sensors can be performed as the pollutant monitoring system.

References [1] Z.I. Wang, Nanostructure of zinc oxide. Matter. Today, 7(6), pp. 26-33, 2004. [2] D. Calestani, M. Zha, R. Mosca, A. Zappettini, M.C. Carotta, V. Di Natale, L. Zanotti, Growth of ZnO tetrapods for nanostructure-based gas sensors. Sensors and Actuators B, 144, pp. 472-478, 2010. [3] B. Yuliarto, Y. Kumai, S. Inagaki, HS. Zhou, Enhanced benzene selectivity of mesoporous silica SPV sensors by incorporating phenylene groups in the silica framework, Sensors and Actuators B: Chemical, 138 Issue 2, pp. 417421, 2009. [4] M.A. Lim, Y.W. Lee, S.W. Han, I. Park, Novel fabrication method of diverse one-dimensional Pt/ZnO hybrid nanostructures and its sensor application, Nanotechnology, 22, 035601 pp. 8, 2011. [5] C. Klingshirn, ZnO: from basics towards applications, Phys. Status Solidi B, 244, pp. 3027-3073, 2007. [6] D.C. Look, D.C. Reynolds, J.W. Hemsky, R.L. Jones, J.R. Sizelove, Production and annealing of electron irradiation damage in ZnO, Appl. Phys. Lett., 75, pp. 811-813, 1999. [7] F. Tuomisto, K. Saarinen, D.C. Look, Irradiation-induced defects in ZnO studied by positron annihilation spectroscopy, Phys. Status Solidi A, 201, pp. 2219-2224, 2004.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

This page intentionally left blank

Air Pollution XIX

239

Buildings as sources of mercury to the atmosphere G. F. M. Tan1, E. Cairnsa1, K. Tharumakulasingam1, J. Lu1 & D. Yap2 1 2

Department of Chemistry and Biology, Ryerson University, Canada Ontario Ministry of the Environment, Canada

Abstract Atmospheric gaseous elemental mercury (GEM) on the streets in the city core at various heights above ground, in parking lots, and in indoor and outdoor air was measured in Toronto, the largest city in Canada. The GEM in the indoor air ranged from 1.15 – 258.4 (average 12.54 ± 11.26) ng m-3 and was much higher than that in the outdoor air (average 1.89 ± 0.49). The average GEM in underground parking lots ranged from 1.37 – 7.86 ng m-3 and was higher than those observed from the surface parking lots. The GEM values increased with increasing elevation and increasing distance from building walls. All of these findings suggest that mercury used in the indoor environment has been diffused and emitted to the outdoor environment, thus, the buildings serve as sources of mercury of GEM to the urban atmosphere. More studies are needed to estimate the contribution of urban areas to the atmospheric Hg budget and the impact of indoor air on outdoor air quality and human health. Keywords: gaseous elemental mercury, urban atmosphere, source of emission, depth profiling, Toronto.

1 Introduction Mercury (Hg) is a persistent and highly toxic element (Sarikaya et al. [1]). It can be emitted from natural and anthropogenic sources, largely as gaseous elemental mercury (GEM) (Bagnato et al. [2]; Song et al. [3]; Goodarzi [4]). The atmosphere receives most of the mercury emitted from these sources (Goodarzi [4]), thus, it is the major pathway of transporting this toxic element. Elemental

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110221

240 Air Pollution XIX Hg has an atmospheric lifetime of around one year (Schroeder and Munthe [5]), therefore, it is considered a true global pollutant. Due to its unique physical and chemical properties, Hg has been widely used in industry (e.g., in electrical equipment and control devices, in the electrolytic preparation of chlorine and alkalis), in agriculture (e.g., as pesticides, fungicides, and bactericides), in dental practices, in pharmaceuticals, and in daily products such as thermometers, barometers, bulbs, batteries, paints, and cosmetic products etc. All the above-listed processes and uses are concentrated in cities. They, therefore, in turn, are sources of mercury to the environment. Limited studies have indeed shown higher atmospheric mercury in urban areas (Witt et al. [6]; St. Denis et al. [7]; Carpi and Chen [8]; Liu et al. [9]) and the concentrations of GEM in urban atmosphere varied with variation of urban structure and height (Song et al. [3]; St. Denis et al. [7]; Carpi and Chen [8]). Our studies showed that local sources which have neither been identified nor been reported in the Canadian National Pollutant Release Inventory (NPRI) might have contributed to the high levels of atmospheric Hg in Toronto (Cheng et al. [10]) and that the higher GEM concentration values were more concentrated in the city core and emissions from vehicles and ground surfaces in the city were not the major sources of GEM to the urban atmosphere (Cairns et al. [11]).

2 Experimental 2.1 Experimental location The experiments were carried out in the City of Toronto (population 2.5 million), Ontario, Canada. The site locations are shown in Fig. 1. According to the NPRI Environment Canada [12], there is no source of Hg (cut-off level of 5 kg) in the city area.

Figure 1:

Experimental locations.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

241

2.2 Instrumental Atmospheric GEM was measured along the major streets and at surface and underground parking lots in the city, using a mercury vapour analyzer (Model 2537A, Tekran Inc., Toronto, Canada). The analyzer has a built-in air pump for air sampling and employs dual gold cartridges, arranged in parallel, for Hg preconcentration, thus, allowing continuous measurements of mercury in the air samples. After every 2.5-minute pre-concentration period, mercury is thermally desorbed from the gold cartridges and determined using cold vapour atomic fluorescence spectrometry (CVAFS). Since the particulate matter in the air stream was removed by a front-end Teflon filter before it entered the analyzer, it is widely accepted that the analyzer measures GEM in the air sample. The analyzer was calibrated automatically through a built-in permeation mercury source every 23 hours. The permeation mercury source was verified by manual injection before and after the field campaign. The internal permeation source provided approximately 1 pg s-1 of gaseous elemental mercury ( Hg0) at 50oC into a zero air stream, whereas the manual calibration was done by injecting a certain volume of air saturated with mercury vapour at a known temperature from a mercury vapour calibration unit (Tekran Inc., model 2505, Toronto). The average detection limit was about 0.1 ng m-3 for GEM. 2.3 Method During the street measurement period, the analyzer was installed on board of a minivan with two inlets: one was hanging over the dashboard glass using a pole that was held on the rooftop of the van and was about 1.8m above ground (i.e., at pedestrian level); the other was attached to a pole that extended about 4m above ground. While driving along the streets, and highways and parking underground, only the 1.8m-inlet was used. A global positioning system (GPS) was used to track the locations of the analyzer. While parked at some of the surface lots, air was sampled from the two inlets in an alternative fashion. A second mercury analyzer, which was a part of the mercury speciation monitoring system, was running on the rooftop of Jorgensen Hall (JOR), Fig.1, which is ~ 59m above ground. Accordingly, while the van was parked around JOR, GEM was measured from three levels: 1.8m, 4m, and 59m above ground. Rooms with different usage types (e.g. research labs and office) were selected for the measurements of GEM in indoor and outdoor air. These rooms are located in the north and east sides of a square-shaped building (Kerr Hall, ~160m long) with a courtyard in the middle. During the experiments, the pump in the analyzer pulled air in from inside and outside (using Teflon tubing through a window) 1, 3, and 6m away from the wall in an alternative order for each room tested.

3 Results and discussion 3.1 GEM at surface parking vs. underground parking The results from the (20) surface parking (at 1.8m) and the (7) underground parking lots are presented in Fig. 2. The average GEM value for the surface sites WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

242 Air Pollution XIX ranged from 1.31 – 3.90 ng m-3 with an overall mean value of 1.89 ± 0.35 ng m-3. The overall mean value agreed well with the overall pedestrian level GEM (1.89 ± 0.62 ng m-3) from driving measurements reported by Cairns et al. [11], suggesting that the movement speed of the analyzer had no significant effect on the measured GEM values. The mean value for the underground sites ranged from 1.37 – 7.86 ng m-3. Values higher than the mean value from surface parking (1.88 ± 0.35 ng m-3) were observed in 4 out of the 7 underground parking lots visited and the average value (3.45 ± 0.99 ng m-3) was statistically significantly higher than that from surface parking lots. The highest GEM value (average 7.86 ng m-3) was observed in the garage at Queen Street and University Ave. This garage was revisited on a different day and the observed average GEM value (1.92 ng m-3) this time was much lower. This suggests that some local sources had contributed to the high GEM value and these sources were not consistent.

Figure 2:

Comparison of atmospheric GEM concentrations measured at surface and underground parking lots in downtown Toronto.

3.2 GEM indoor vs. outdoor Table 1 lists the concentrations of atmospheric GEM observed from both indoor and outdoor. The indoor GEM was measured inside (1) an office in the 1st floor, WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

243

(2) a research lab in the 2nd floor (Lab 1) and (3) a research lab in the 3rd floor (Lab 2). The outdoor GEM was measured on the streets within downtown Toronto. Lab 1 is relatively older and larger than Lab 2 and has chemical storage and laboratory equipments. An average of 17.70 ± 22.44 ng m-3 (ranging from 3.28 – 258.4 ng m-3) GEM was observed in Lab 1. This was approximately 4 times higher than Lab 2 (4.51 ± 3.48 ng m-3, ranging from 1.15 – 25.75 ng m-3) and was comparable to GEM level inside the office (15.42 ± 7.86 ng m-3, ranging from 2.19 – 40.83 ng m-3). The presence of chemicals and reagents containing Hg (e.g., Hg standards, HgCl2, HgBr etc.) in Lab 1 might have contributed to the higher indoor GEM level as compared to Lab 2 and office. In addition, huge spikes (109.5 – 258.4 ng m-3) were observed in Lab 1 when the ventilation unit inside the room was displaced, indicating the possibility of Hg accumulation on the ventilation unit over time. A lower ventilation rate, as observed physically, in Lab 1 compared to Lab 2, could be the other contributing factor to the elevated GEM levels. The office GEM was lower than Lab 1 but higher than Lab 2. The absence of a ventilation unit inside the office might have contributed to this variation, thus, indicating the importance of ventilations in maintaining good air quality. Overall, the average GEM level indoor (12.54 ± 11.26 ng m-3) was approximately 7 times higher than that in outdoor (1.89 ± 0.46 ng m-3). Similar result with higher indoor GEM compared to outdoor was observed by Li et al. [13]. Table 1:

Comparison of atmospheric GEM measured at both indoor and outdoor areas.

Type

Number of Measurements

Range

Mean±Std. Dev.

Outdoor Driving

2147

1.06-8.25

1.89±0.62

Surface parking

722

0.24-7.07

1.89±0.35

Overall

2869

0.24-8.25

1.89±0.49

Office

244

2.19-40.83

15.42±7.86

Lab 1

950

3.28-258.4

17.70±22.44

Lab 2

834

1.15-25.75

4.51±3.48

Overall

2028

1.15-258.4

12.54±11.26

Indoor

3.3 GEM horizontal profile Fig. 3 shows the comparison of the data collected indoor and outdoor at a distance of 1m, 3m, and 6m away from the building wall. The results obtained from 1m, 3m, and 6m away from the building wall have an average of 3.48 (± 1.36) ng m-3, 3.48 (± 1.33) ng m-3, and 3.15 (± 1.10) ng m-3 from the office, respectively; 3.57 (± 2.18) ng m-3, 3.40 (± 2.51) ng m-3, and 3.64 (± 2.93) ng m-3, from Lab 1, respectively ; and 3.17 (± 2.35) ng m-3, 4.16 (± 4.02) ng m-3, and WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

244 Air Pollution XIX 4.94 (± 2.64) ng m-3 from Lab 2, respectively. The GEM concentration values decreased with increasing distance from the building wall outside of the office. For the results from the labs, a trend of decreasing then increasing concentration of GEM was observed the further the sampling location is from the wall. Analysis shows that the results from all the locations (office and labs) are statistically significantly different from one another, with a confidence level of 99%. The tree branches near the labs could have prevented air from well mixing, as the air sampling inlets were supported using tree branches located closed to the buildings, thus, requiring further investigation. All the values observed in the air 6m away from the building walls were higher than the value observed from surface parking, indicating the effect of further dilution.

GEM, ng m‐3

50.00 Office 

40.00

Lab 1

Lab 2

Surface Parking

30.00 20.00 10.00 0.00

Sampling Locations Figure 3:

Comparison of atmospheric GEM concentrations measured indoor and outdoor 1, 3, and 6m away from the building wall and from streets.

3.4 GEM depth profile Fig. 4 shows the statistical analysis of the data collected at 1.8m, 4m, and 59 m from 7 sites around Jorgenson Hall in the city core. The GEM values ranged from 1.25 –1.75 ng m-3 (average 1.44 ± 0.19), 1.30 –1.80 (average 1.48 ± 0.18), and 2.60 – 3.70 (average 3.09 ± 0.47) ng m- 3 for the 1.8m, 4m, and 59m level, respectively. A comparison of the values observed at the three levels showed that the higher the elevation, the higher the GEM concentrations were. The values from the 1.8 and 4m agree well with those obtained from other locations in the city at the same height Cairns et al. [11]. This suggests that buildings serve as sources of GEM to the atmosphere, as if GEM were emitted from the ground surface and vehicles, the GEM concentration should be, due to the dilution effect, lower at higher elevation.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

245

4 3.5 GEM, ng m‐3

3 2.5 2 1.5 1 0.5 0 1.8( N=7)

4 (N=7)

59 (N=7)

Height above ground Figure 4:

Three-Level (1.8, 4 and 59m above ground) depth profiles of GEM around Jorgenson Hall in downtown Toronto, May–July 2009. N=number of sites.

4 Conclusion The results from this study showed that, in Toronto, Canada, (1) the higher the elevation, the higher the atmospheric GEM values were in the street canyon; (2) GEM levels were higher in the underground parking lots than those at the surface parking lots; (3) GEM levels were higher indoors as compared to outdoors. All of these findings suggest that mercury used in the indoor environment has been diffused and emitted to the outdoor environment, thus the buildings serve as sources of mercury and that emission from vehicles and ground surfaces were not the major sources of GEM to the urban atmosphere.

Acknowledgements The Ontario Ministry of Environment (MOE) and Natural Sciences and Engineering Research Council (NSERC) of Canada for financial support.

References [1] Sarikaya, S., Karcioglu, O., Ay, D., Cetin, A., Aktas, C., Serinken, M., Acute mercury poisoning: a case report. BMC Emergency Medicine 10, 7, 2010.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

246 Air Pollution XIX [2] Bagnato, E., Allard, P., Parello, F., Aiuppa, A., Calabrese, S., Hammouya, G., Mercury gas emissions from La Soufriere Volcano, Guadeloupe Island (Lesser Antilles). Chemical Geology 266, 276−282, 2009. [3] Song, X., Cheng, I., Lu, J., Annual atmospheric mercury species in Downtown Toronto, Canada , Journal of Environmental Monitoring 11, 660–669, 2009. [4] Goodarzi, F., Speciation and mass-balance of mercury from pulverized coal fired power plants burning western Canadian subbituminous coals. Journal of Environmental Monitoring 6, 792−798, 2004. [5] Schroeder, W.H., Munthe, J., Atmospheric mercury - an overview. Atmospheric Environment 32 (5), 809−822, 1998. [6] Witt, M.L.I., Meheran, N., Mather, T.A., de Hoog, J.C.M., Pyle, D.M., Aerosol trace metals, particle morphology and total gaseous mercury in the atmosphere of Oxford, UK. Atmospheric Environment 44, 1524−1538, 2010. [7] St. Denis, M., Song, X., Lu, J., Feng, F., Atmospheric gaseous elemental mercury in downtown Toronto. Atmospheric Environment 40, 4016–4024, 2006. [8] Carpi, A., Chen, Y. F., Gaseous elemental mercury fluxes in New York City. Water Air and Soil Pollution 140, 371–379, 2002. [9] Liu, S.L., Nadim, F., Perkins, C., Carley, R.J., Hoag, G.E., Lin, Y.H., Chen, L.T., Atmospheric mercury monitoring survey in Beijing, China. Chemosphere 48, 97−107, 2002. [10] Cheng, I., Lu, J., Song, X., Studies of potential sources that contributed to atmospheric mercury in Toronto, Canada. Atmospheric Environment 43, 6145–6158, 2009. [11] Cairns, E., Tharumakulasingam, K., Athar, M., Yousaf, M., Cheng I., Huang Y., Lu J., Yap D., Source, concentration, and distribution of elemental mercury in the atmosphere in Toronto, Canada, Environmental Pollution, 2011. [12] National Pollutant Release Inventory (NPRI). Environment Canada Website, Canada, [13] Li, J., Yang, Y., Chen, H., Xiao, G., Wei, S., Sourcing contributions of gaseous mercury in indoor and outdoor air in China. Environmental Forensics 11, 154−160, 2010.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

247

Occupational exposure to perchloroethylene in Portuguese dry-cleaning stores S. Viegas Escola Superior de Tecnologia da Saúde de Lisboa, Instituto Politécnico de Lisboa (ESTeSL/IPL), Portugal

Abstract Perchloroethylene (also known as tetrachloroethylene) is a solvent that has been a mainstay of the dry cleaning industry for decades. Since 1995 the International Agency for Research on Cancer considers that dry cleaning entails exposures that are possibly carcinogenic to humans (Group 2B). Meanwhile, the same institution classified perchloroethylene as probably carcinogenic to humans (Group 2A). Some industries have begun using alternative cleaning methods that do not require the use of perchloroethylene. However, in Portugal this solvent is still the most common dry-cleaning agent. An exploratory study was developed that aimed to find the occupational exposure to perchloroethylene in four Portuguese dry-cleaning stores. Activities involving higher exposure and variables that promote exposure were also investigated. Real-time measurements of volatile organic compounds concentrations were performed using portable equipment (MultiRAE, RAE Systems model – calibrated by isobutylene). Considering that perchloroethylene was the only cleaning product used in all the stores studied we deduce that results obtained for volatile organic compound measures correspond to perchloroethylene concentrations. The measurements were performed in the same places in each store and during the same tasks, namely: reception area, dry and washing area; iron area; dry-cleaning machine working; loading and unloading dry-cleaning machine and manual removing of stains with perchloroethylene. Besides measurements, information related with variables that can influence exposure in each store was also collected.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110231

248 Air Pollution XIX Results of volatile organic compounds concentrations range from 0.0 ppm to 558 ppm. Three of the tasks studied obtained concentrations higher than 100 ppm, namely loading and unloading dry-cleaning machine and also manual removing of stains with perchloroethylene. It was possible to identify tasks with priority for the appliance of preventive and protective measures. It seems that exposure to high perchloroethylene concentrations still occurs in Portuguese dry-cleaning stores and lack of ventilation conditions is probably the major cause of this situation. Keywords: perchloroethylene, occupational exposure, dry-cleaning, ventilation conditions, real-time measurements.

1 Introduction Dry cleaning is essentially a waterless process in which clothes are cleaned with an organic solvent rather than with soap and water. Early dry-cleaning solvents such as turpentine, gasoline, or benzene were flammable or explosive. Non flammable dry-cleaning solvents such as carbon tetrachloride, trichloroethylene, and fluorocarbons were introduced in 1900’s. Perchloroethylene (PCE, tetrachloroethylene) introduced in the 1940s, has become the primary drycleaning solvent in use [1, 2]. Nowadays, perchloroethylene (PERC) is widely used in Portugal and in many other countries as a solvent in many commercial dry cleaning stores. As an example, there are over 30.000 commercial dry-cleaning stores and approximately 244.000 dry-cleaning workers in the United States. Approximately 90 percent of these stores use PERC as their primary solvent [3, 4]. Consequently, occupational exposure to PERC has already been demonstrated for dry cleaning workers [2, 5–10]. However, over the past several decades, the dry-cleaning industry has made tremendous progress in reducing workers’ exposure and environmental releases of PERC. In particular, the dry-cleaning machines and solvents used have evolved over time. The development of dry-cleaning machines encompasses five “generations” [4]. The first generation of dry-cleaning machines were transfer machines with separate washers and dryers. Transfer machines, older and less expensive, require manual transfer of solvent laden clothing between the washer and dryer. The transfer activity involves high worker exposure to PERC. In the late 1960s the second generation, non refrigerated, dry-to-dry machines, using a one step process that eliminates clothing transfer, were introduced. In dry-to-dry machines, clothes enter and exit the machine dry. Generally, worker exposure to PERC from dry-to-dry machines is less than exposure from transfer machines. Most second-generation machines are vented dry-to-dry machines that vent residual solvent vapours directly to the atmosphere or through some form of vapour recovery system during the aeration process. The third generation of machines, dry-to-dry machines with refrigerated condensers, was introduced in the late 1970s and early 1980s. These machines are vent less dryto-dry machines that are essentially closed systems, which are only open to the WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

249

atmosphere when the machine door is opened. They promote recirculation of the heated drying air through a vapour recovery system and back to the drying drum and because of that here is no aeration step. This machines generation provides considerable solvent savings over their predecessors. Fourth-generation drycleaning machines are essentially dry-to-dry, no vented machines having controls to reduce residual PERC concentrations in the machine cylinder at the end of the dry cycle. These dry-to-dry machines rely on both a refrigerated condenser and closed-loop carbon absorber that does not vent to the atmosphere to recover PERC vapours during the dry cycle. They are designed to recover residual PERC vapours in the cylinder at the end of the dry cycle. More traditional machines generally emit higher concentrations of PERC into the environment and workers’ breathing zones. Fifth-generation machines have the same features as fourth-generation machines; however, they are also equipped with a monitor inside the machine drum connected to an interlock to ensure that the concentration is below approximately 290 ppm before the loading door can be opened [4]. This can be considering a preventive measure to reduce exposure. PERC vapours are rapidly absorbed through the pulmonary alveolar epithelium and are largely exerted by the same route. Elimination via exhaled breath is biphasic, probably because of storage and release from fatty tissues. Some dermal absorption may occur, but this is generally considered to be minimal. PERC is partially metabolized to trichloroacetic acid, which is exerted in urine. In humans, 80–100% of the absorbed perchloroethylene is exerted as the parent compound by exhalation [2, 11]. Regarding health effects, PERC inhalation can lead to depression of the central nervous system. Symptoms at various concentrations include unconsciousness, dizziness, headache, drowsiness, and visual disturbances. Irritation on the eyes, nose and throat may occur, and gastrointestinal complaints such as nausea and vomiting have been reported. Psycho-physiological effects include fatigue, anorexia, irritability, impaired memory, and confusion. Direct skin contact with liquid PERC may lead to erythema, burns, and vesiculation [8]. In the case of reproductive effects, general mechanisms have been suggested in literature for explaining the possible effect in reproduction, namely: Exposure may cause mutations or other genotoxic damage in the ova or sperm; Exposure may cause injurious effects on parental cells or cell systems, which are essential in the reproductive process (ova, sperm, ovaries, testes, prostate, endometrial tissue, hypothalamus, pituitary, etc ), or it can cause such effects on embryonic or foetal cells, tissues or organs; Exposure may act as an agonist or antagonist of endogenous hormones relevant to reproduction and, finally, exposure may disturb regulatory mechanisms important in reproduction. Each of these mechanisms should be considered as a potential action for PERC to interfere with reproduction [12, 13]. PERC has been shown to cause cancer in laboratory animals. This has raised the concern that it may also cause cancer in humans. Several epidemiological studies have been conducted to determine whether there is an association between occupational exposure to PERC and increased cancer risk. Unfortunately, these studies failed to control for smoking status or other factors WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

250 Air Pollution XIX contributing to cancer such as exposure to petroleum solvents (use some times to remove stains), hepatitis infection, alcohol abuse, and human papilloma virus infection [2, 11, 14] Nevertheless, since 1995 the International Agency for Research on Cancer classified PERC as probably carcinogenic to humans (Group 2A) [13]. The PERC Threshold Limit Value for the Time Weighted Average 8 hour day (TLV-TWA) recommended by the ACGIH is 25 ppm (170 mg/m3) [15]. The same is mention by the Portuguese Norm 1796 (2007) that recommend also 100 ppm as Short Term Exposure Limit (TLV - STEL) [16]. This study was designated to evaluate workers exposure to VOCs in 4 drycleaning stores and also identify the task with higher exposure and, consequently, with priority for preventive measures application. Considering that PERC was the only cleaning product use in all the stores studied we deduce that results obtains with VOCs concentrations measure correspond to PERC concentrations.

2 Materials and methods This study was conducted in four dry-cleaning stores located in Lisbon that used only PERC has dry-cleaning agent. The VOCs measurements were performed in the same places in each store and during the same tasks, namely: reception area, dry and washing area; iron area; dry-cleaning machine working; loading and unloading dry-cleaning machine and also manual removing of stains with PERC. Besides measurements in each store was collected information’s about ventilation resources, drycleaning machine type (generation), number of loadings machine per day and use of individual protection equipment. In this research, real-time measurements of VOCs concentrations were performed using portable direct-reading equipment ((MultiRAE, RAE Systems model – calibrated by isobutylene). The detection technique used in this equipment is Photo Ionization Detection (PID) equipment (10.6 eV lamps). The use of this equipment permitted to identify the worst case scenario concerning to exposure in the tasks studied. The PID equipment was zeroed outside each store in fresh air prior to starting measurements. All measurements were done in the breathing zone of the workers while they were performing their tasks, during 5 to 15 minutes. It was consider the highest concentration obtained in each measurement point/task.

3 Results The dry-cleaning stores were similar considering ventilation conditions. Only one store (D) doesn’t have windows and works with a closed door. This store has also the higher number of loadings per day and it was the only one that had a dry-cleaning machine from 3rd generation. Regarding ventilation conditions, store A was the only one that doesn’t have mechanical general ventilation. Meanwhile, none of the stores have located WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

251

ventilation. All the stores have acclimatization. Concerning manual removing of stains with PERC, only store C doesn’t have this practice. In dry-cleaning stores A and B measurements were performed in the morning, and in dry-cleaning stores C and D in the afternoon. Results of VOCs concentrations range from 0.0 ppm to 586 ppm (Table 2). Table 1:

Characterization of each dry-cleaning store. A

B

C

D

Number of loadings per day

4

4

2

5

Open doors

Yes

Yes

Yes

No

Windows

Yes

No

Yes

No

General ventilation

No

Yes

Yes

Yes

No

No

No

No

No

No

No

No

Climatization sistem

Yes

Yes

Yes

Yes

Machine generation

4th

4 th

4 th

3 rd

Yes

Yes

No

Yes

No

No

No

No

Located ventilation in drycleaning machine door Located ventilation outside in dry-cleaning machine

Manual removing of stains with PERC Individual Protection Equipment use Table 2:

Results of VOCs measurements (ppm). Manual Unloading removing dryof stains cleaning with machine PERC

Stores

Reception area

Dry and washing area

Iron area

Drycleaning machine working

Loading drycleaning machine

A

0.5

-

6.8

2.3

586

88.9

-

B

0.0

0.3

1

0.9

277

46.7

-

C

12.1

26.1

52.7

20.1

-

229

-

D

0.1

0.0

3.8

4.6

90.6

83.4

122

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

252 Air Pollution XIX

4 Discussion In dry-cleaning stores, release of PERC vapours into the environment and subsequent worker exposure to PERC vapours is greatest during loading, unloading and also maintenance of dry-cleaning machine [17]. This is consistent with our results, however, only this last task wasn’t possible to study in this research. Results obtained show various tasks with concentrations higher than 100 ppm (Short Term Exposure Limit) and the majority of tasks have duration’s superior to 15 minutes, namely loading and unloading dry-cleaning machine and manual removing of stains. Additionally, a previous study demonstrated also that machine operators have the greatest PERC exposure probably because they execute the loading and unloading of dry-cleaning machine [2]. In the case of unloading, probably exposures can be reduce if fifth-generation machines were used because they have a monitor inside the machine drum connected to an interlock to ensure that the concentration is below approximately 290 ppm before the machine door can be opened [4]. Other tasks show high VOCs concentrations results, namely iron and manual removing of stains. Probably, results in first task (52.7 ppm) are due to temperature increase that promotes volatilization of PERC residues presents in clothes. In the second task, stains removing, only studied in store D, the short distance from the worker breathing zone to the clothe surface, when applying the PERC, explains the obtained results (122 ppm). Furthermore, it is important to notice that store A was the only one that doesn’t have mechanical general ventilation, and this contributes to the vapour accumulation inside the store and might explain the higher concentration (586 ppm) obtain during dry-cleaning machine loading. This kind of ventilation (general promotes dilution and is very useful in decreasing the contamination in work places [2, 17, 18]. Generally accepted guidelines recommend an air change in the workroom every five minutes with a minimum of 30 cfm of outside air per person. Supply and exhaust systems within the store should move air from a clean area (offices, reception area, etc.) to less clean area (where the dry-cleaning machine is located) [17]. Moreover, none of the studied stores have local exhaust ventilation and this ventilation technique is fundamental to reduce the vapour level reaching the worker’s breathing zone and minimizes also vapour diffusion [17]. Concerning the exposure assessment method applied, it is important to consider that knowing when high exposures occurs permit to identify the tasks and work conditions that promote exposure and where the investments have more impact in the workers safety. Therefore, identifying the task with higher peak concentration is extremely important to perform health risk assessment and to define priorities for the appliance of preventive and protective measures. Additionally, the development of real-time measurements offers the possibility of directly relate performance with exposure [19]. This kind of measurement resource is well describe in a study performed by Viegas et al. [20] that also permitted to identify the tasks that involve higher peak exposure to formaldehyde WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

253

in ten pathology and anatomy laboratories and one formaldehyde-based resins production factory. Moreover, peak exposures are a special concern because high concentration produces a high dose rate into the body and target tissue, which may alter metabolism, overload protective or repair mechanisms, and amplify tissue responses. Therefore, is important to characterize peak exposures, because a peak can produce more and perhaps different health effects than the same administered dose given with less intensity over a longer time period [21–23].

5 Conclusions The results obtained are not detailed in the PERC exposure in this occupational setting; however it brings attention for the exposure characteristics (peak exposure) during some of the tasks. More detailed studies have to be developed with the complementary use of exposure biomarkers. Meanwhile the data obtained can support the adoption of some preventive measures, namely replacement of old machines with fifth-generation machines and improve of ventilation conditions (local and general).

References [1] Solet, D, Robins, TG, Sampaio C, Percloroethylene exposure assessment among dry cleaning workers. Am Ind Hyg Assoc J. 51 (1990) 566–74. [2] Pirsaraei, A, Khavanin, A, Asilian, H, Soleimanian, A, Occupational Exposure to Perchloroethylene in Dry-cleaning Shops in Tehran, Iran. Industrial Health. 47 (2009) 155–159. [3] Environmental Protection Agency. Economic Impact Analysis of Regulatory Controls in the Dry Cleaning Industry. EPA 450/3-91-021. Office of Air Quality Planning and Standards, EPA, Research Triangle Park, NC (1991). [4] Earnest, G, A Control Technology Evaluation of State-of-the-Art, Perchloroethylene Dry-Cleaning Machines. Applied Occupational and Environmental Hygiene. 17 (2002) 352–359. [5] Verberk, M, Scheffers, T, Tetrachloroethylene in exhaled air of residents near dry-cleaning shops. Environ. Res. 21 (1980) 432–437. [6] Lauwerys, R, Herbrand, J, Buchet, A, Gaussin, B, Gaussin, J, Health surveillance of workers exposed to tetrachloroethylene in dry-cleaning shops. Int. Arch. Occup. Environ. Health. 52 (1983) 69–77. [7] Monster, A, Regouin-Peeters, W, Van Schijndel, A Van Der Tuin, J, Biological monitoring of occupational exposure to tetrachloroethene. Scand. J. Work Environ. Health. 9 (1983) 273–281. [8] Materna, M, Occupational exposure to perchloroethylene in the dry cleaning industry. Am. Ind. Hyg. Assoc. J. 46 (1985) 268–273. [9] Droz, P, Guillemin, M, Occupational exposure monitoring using breath analysis. J. Occup. Med. 28 (1986) 593–602.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

254 Air Pollution XIX [10] Thompson, K, Evans, J, Workers’ breath as source of perchloroethylene (PERC) in the home. J. Expos. Anal. Environ. Epidemiol. 3 (1993) 417– 430. [11] ATSDR. Toxicological Profile for Tetrachloethylene. Agency for Toxic Substances and Disease Registry. US. Department of Health and Human Services. 1997. [12] Gulden, J, Zielhuis, G, Reproductive hazards related to perchloroethylene. A review. Int Arch Occup Environ Health 61(1989) 235-242. [13] International Agency for Research on Cancer, Dry cleaning, some chlorinated solvents and industrial chemicals. Lyon: IARC, 1995 [14] Ruder, A, Ward, E, Brown, D, Mortality in Dry-Cleaning Workers: An Update. American Journal of Industrial Medicine. 39 (2001) 121-132. [15] American Conference of Governmental Industrial Hygienists, Threshold limit values for chemical substances and physical agents and biological exposure indices (TLVs and BEIs), ACGIH, Cincinnati, (2006). [16] Instituto Português da Qualidade – NP 1796 : 2007 : segurança e saúde do trabalho : valores limite de exposição profissional a agentes químicos existentes no ar dos locais de trabalho. Caparica : IPQ, 2007. [17] National Institute for Occupational Safety and Health. Control of Exposure to Perchloroethylene in Commercial Drycleaning (Ventilation). Applied Occupational and Environmental Hygiene. 15 (2000) 187-188. [18] Liddament, M, A review of ventilation and the quality of ventilation air. Indoor Air. 10 (2000) 193-199. [19] Walsh, P, Forth, A, Clark, R, Real-time measurement of dust in the workplace using video exposure monitoring: farming to pharmaceuticals. Journal of Physics : conference series. 151 (2009) 012-043. [20] Viegas, S, Ladeira, C, Nunes, C, Malta-Vacas, J, Gomes, M, Brito, M, Mendonça, P, Prista, J, Genotoxic effects in occupational exposure to formaldehyde: A study in anatomy and pathology laboratories and formaldehyde-resin production. Journal of Occupational Medicine and Toxicology. (2010) 5:25. http://www.occup-med.com/content/5/1/25. [21] Smith, T, Studying peak exposure: toxicology and exposure statistics. In: Marklund, S, ed. lit. – Exposure assessment in epidemiology and practice. Stockholm: National Institute for Working Life. (2001) 207-209. [22] Preller, L, Burstyn, I, De Pater, N, Characteristics of peaks of inhalation exposure to organic solvents. The Annals of Occupational Hygiene. 48 (2004) 643–652. [23] Vyskocil, A, Thuot, R, Turcot, A, Peak exposures to styrene in Quebec fibreglass reinforced plastic industry. In: Marklund, S, ed. lit. – Exposure assessment in epidemiology and practice. Stockholm: National Institute for Working Life. (2001) 316-318.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Section 3 Air quality management

This page intentionally left blank

Air Pollution XIX

257

Health impact assessment of PM10 and EC in 1985–2008 in the city of Rotterdam, The Netherlands M. P. Keuken1, P. Zandveld1, S. van den Elshout2, N. Janssen3 & G. Hoek4 1

TNO, Netherlands Applied Research Organisation, The Netherlands DCMR, Rijnmond Environmental Agency, The Netherlands 3 RIVM, Netherlands Environmental Agency, The Netherlands 4 IRAS, Utrecht University, The Netherlands 2

Abstract The health impact assessment (HIA) PM10 and elemental carbon (EC) was investigated in the period 1985–2008 in the city of Rotterdam. The spatial distribution of the concentrations was modelled by the URBIS model. The modelling results for 2008 were validated by PM10 and EC measurements at various locations in Rotterdam. This paper describes the HIA related to improved air quality in the period 1985-2008: at urban background locations 18 µg m-3 PM10 and 2 µg m-3 EC. The gain in life years saved due to long-term exposure to PM10 and EC in this period was, respectively, 13 and 12 months per person. The similar health impacts for PM10 and EC suggests that a reduction of combustion aerosol was important for the reduction in health impact of PM10. It is concluded that EC is a more adequate indicator for HIA of traffic measures than PM10. Keywords: health impact assessment, PM10, EC, historical trend.

1 Introduction Exposure to elevated levels of particulate matter (PM) has been associated with health effects in epidemiological surveys [1]. However, in urban areas the population is exposed in particular to emissions by road traffic related to exhaust emissions, tire/engine wear and re-suspension of road dust. Over the years, a WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110241

258 Air Pollution XIX variety of (inter)national and local measures have been implemented to reduce exhaust emissions by road traffic. Examples are more stringent emission standards and environmental zoning to prevent high emitters near residential areas. Consequently, there is a need to assess the impact of these measures on air quality and health. The mass of PM is less adequate for this purpose. Because even near heavy traffic locations, PM mass is dominated by regional background concentrations. Elemental carbon (EC) is considered a more appropriate indicator for dispersion and exposure of the population to PM exhaust emissions [2]. This paper describes a study is to assess the health impact of PM10 and EC in the city of Rotterdam in the period 1985-2008. Rotterdam has the largest harbour of Europe and consequently, road traffic is intensive with a relatively large contribution of heavy duty vehicles. In sections 2 and 3, the approach of the study is detailed and in section 4, the results are presented and discussed. The conclusions and recommendations are elaborated in section 4. The study was financed by the Ministry of Infrastructure and Environment in the framework of The Netherlands Policy Support Program on Particulate Matter (“BOP”).

2 Study approach The URBIS model has been applied to estimate the spatial distribution of annual average PM10 and EC in the city of Rotterdam for the years 1985, 1995 and 2008. The URBIS model combines a street-canyon and line-source model to compute the contribution of emissions by urban traffic and motorways to air quality [3]. The spatial resolution of the URBIS model is a 10*10 m2 grid up to the housing façade along inner-urban roads and up to 500 m near motorways. The next step in the health impact assessment is to combine the spatial distribution of annual average PM10 and EC with GIS based population distribution data. This provides information on the population exposure to longterm air pollution at house address. Subsequently, concentration-responsefunction for long-term health effects of PM10 and EC are applied to the population exposure maps [4]. Finally, the trend in health effects in the period 1985 - 2008 is evaluated in relation to exposure to PM10 and EC.

3 The modelling and experimental setup 3.1 Modelling The required model input for spatial concentrations of PM10 and EC concerns:  meteorological conditions; The ten-year average meteorological conditions for the period 1995-2004 have been used for all three years 1985, 1995 and 2008 to eliminate the effect of meteorological variation on the health impact assessment;  traffic data; Actual traffic data on the motorways and main urban roads for the years 1985, 1995 and 2008 were available in Rotterdam. The data for 1985 and 1995 concerned aggregated traffic volume on WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX





259

motorways and main inner urban roads. For 2008 more detailed information was available which was used to further detail the 1985 and 1995 traffic data; emission factors; Emission factors for road traffic (e.g. friction and exhaust emissions) for the period 1990-2008 are available for the car fleet in The Netherlands [5]. These emission factors were used to extrapolate emission factors for 1985. For EC, emission factors were derived from an EU database with EC emission factors as a fraction of PM exhaust emissions [6]. These data combined with the information on PM exhaust emission factors in The Netherlands, were used to estimate EC emission factors for 1985, 1995 and 2008; regional and urban background; The regional and urban backgrounds of PM10 in 1985 and 1995 for Rotterdam were extrapolated from the monitoring data in 2008 with an increasing trend of 0.7 µg m-3 PM10 per year for previous years [7]. The regional and urban backgrounds of EC were based on time series of Black Smoke measurements in and near Rotterdam in the period 1985–2008. A factor 10 was applied to convert Black Smoke to EC concentrations. This conversion factor was based on parallel measurements in 2006-2007 of Black Smoke and EC in the city of Rotterdam.

3.2 Measurements In 2006-2007, two-weekly PM samples were collected at traffic, urban and regional locations in and near Rotterdam and analyzed by the Black Smoke method and thermal EC analysis. The results are presented in Figure 1. Rotterdam: 2006-2007 35 30 (10:1) -3

BS index (Pg.m )

25 regional urban traffic

20 15 10 5 0 0,0

1,0

2,0

3,0

EC (Pg.m 3)

Figure 1:

Two-weekly average black smoke index (µg m-3) and thermal EC (µg m-3) at traffic, urban and regional locations in and near Rotterdam in 2006-2007.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

260 Air Pollution XIX Figure 1 illustrates that the Black Smoke index is linear correlated with EC at various locations in and near Rotterdam. The linear regression has a slope 11, an intercept near zero and the regression coefficient (R2) 0.9. Taking into considering also results from other studies, a factor 10 has been derived to estimate EC concentrations from Black Smoke index measurements. At two urban traffic stations and a regional station 15 km south of Rotterdam, annual average Black Smoke data were available for the period 1985-2008 [source: RIVM/DCMR]. The relation presented in Figure 1, has been used to derive from the trend in EC from the Black Smoke measurements in this period. The trend for EC is shown in Figure 2. 3,5

3,0

-3

Annual EC ( g.m )

2,5

2,0

regional traffic 1 traffic 2

1,5

1,0

0,5

Figure 2:

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

0,0

Annual EC (µg m-3) at two traffic locations in Rotterdam and one regional location near Rotterdam in 1985-2008.

Figure 2 shows a decreasing trend in EC concentrations both at the traffic stations and at the regional location. The delta between the traffic and the regional stations especially decreased in the period 1995 – 2003. This reflects the introduction of catalytic convertors in the nineties and emission reduction of combustion aerosol by road traffic. This measure had a relatively larger impact on EC concentrations at the traffic locations than the regional background. Since 2003, no further decreasing trend is detected neither at the regional nor at the traffic locations. This indicates that further reduction of (diesel) emissions by “cleaner” vehicles do not longer balance the growth in traffic volume.

4 Results and discussion 4.1 Air quality of PM10 and EC in 1985, 1995 and 2008 Based on the input presented in section 3.1, the spatial distribution for PM10 and EC has been modelled with the URBIS model for the years 1985, 1995 and 2008 in Rotterdam. In Figures 3A and 3B the results are presented for PM10 and EC in WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

261

1985. In these figures, the motorways are presented around the city centre with extensions to the north, south, east and west. Also, the river “Oude Maas” is presented with harbour areas in the west. As it is difficult to show this data in “black/white” only the data for 1985 are presented.

A

B Figure 3:

A. Annual average concentrations of PM10 (µg m-3) in Rotterdam (1985). B. Annual average concentrations of EC (µg.m-3) in Rotterdam (1985).

Figures 3A and B illustrate that the air quality for PM10 and EC in the city of Rotterdam was elevated near motorways and inner-urban roads with heavy traffic. EC concentrations were a factor two elevated near traffic locations WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

262 Air Pollution XIX compared to the urban background, while PM10 was “only” 20% elevated as PM10 is dominated by regional background concentrations. The results for all years 1985, 1995 and 2008 are presented in Table 1. Table 1:

1985 1995 2008

Annual average PM10 and EC concentrations (µg m-3) at regional, urban and traffic locations in Rotterdam in 1985, 1995 and 2008.

regional 40-45 30-35 20-23

PM10 (µg.m-3) Urban 40-45 35-40 23-27

traffic 45-55 35-45 27-35

regional 2.5-3.5 2.5-3.5 0.5-0.8

EC (µg m-3) Urban 2.5-3.5 2.5-3.5 0.8-1.2

Traffic 3.5-6 3.5-5 1.2-3

The results in Table 1 show that the air quality for both PM10 and EC improved significantly in the period 1985-2008 in Rotterdam. The urban background for PM10 decreased from average 43 µg.m-3 in 1985 to average 25 µg.m-3 in 2008. This is caused for 70% by large-scale emission reduction by industry, energy production and road traffic of precursors (sulfur dioxide, ammonia and nitrogen oxides) for secondary particles [7]. Only 10% reduction of PM10 is related to primary combustion emissions and the remaining 20% by secondary organic aerosols and less water adsorbed to PM10 particles. Table 1 demonstrates that similar to PM10 the air quality for EC also improved significantly in the period 1985-2008. This is attributed to lower regional background concentrations as a result of reduced emissions of soot particles by combustion processes in general (e.g. industry, energy production, shipping and road traffic) and at the urban scale of (diesel) emissions by road traffic in particular. Consequently, the urban background of EC decreased from average 3 µg.m-3 in 1985 to average 1 µg m-3 in 2008. 4.2 Validation of modelled air quality in Rotterdam in 2008 Based on meteorological data for the year 2008, the spatial distribution of PM10 and EC in the city of Rotterdam was modelled by the URBIS model. The modelling results were validated by measurements in 2008 at an urban background and traffic locations in the monitoring network of the environmental protection agency (DCMR) in Rotterdam. The results are presented in Table 2. The uncertainty in monitoring annual concentrations is in the order of 15%, while for modelling, the uncertainty is in the range of 25 to 40% for an urban background and road side location, respectively [8]. Considering these uncertainties, the results in Table 2 show good agreement between modelled and monitoring annual averages for PM10 . For EC, the modelling and monitoring results at the urban background location differ by a factor 2. This may be explained by a.) underestimation of the urban background by a too large conversion factor 11 of Black Smoke to EC concentrations and b.) overestimation of the measured background by the MAAP instrument of EC concentrations due to calibration of MAAP by VDI thermal EC analysis. The WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

263

VDI protocol overestimates EC concentrations due to inadequate correction of artefact EC formation during thermal analysis. Our study underlines that more experimental data on EC is required to further improve modelling of dispersion EC in urban areas. Table 2:

Measurements and modelling results of the annual average PM10 and EC at locations in Rotterdam (2008). PM10 (µg·m-3) model monitoring

Urban background - “Schiedam” Traffic location - “Floreslaan” - “Vasteland” Motorway station - “Ridderkerk”

EC (µg·m-3) model Monitoring

27.9

25.7A

1.0

2.0C

25.6 28.9

27.2A n.a.

1.2 1.4

1.6D 1.7D

30.7

28.3B

2.2

n.a.

A

: gravimetric analysis; : Tapered element oscillating monitor – TEOM corrected with 1.3; C : Multi angle absorption photometer – MAAP; D : Conversion from Black Smoke index; n.a.: not available B

Table 3:

The number of inhabitants in Rotterdam exposed to various levels of annual average PM10 and EC in the period 1985 – 2008.

PM10 (µg m-3) 20-30 30-35 35-40 40-45 45-55 EC (µg m-3) 0.5-1.5 1.5-2.5 2.5-3.5 3.5-4.5 4.5-6.0

Number of inhabitants (#) 1995

2008

515.000 55.000

15.000 550.000 5.000

565.000 5.000 -

535.000 30.000 5.000

560.000 10.000 -

565.000 5.000 -

1985

4.3 Health impact assessment of PM10 and EC in 1985, 1995 and 2008 4.3.1 Population density in Rotterdam in 1985-2008 To investigate the health impact assessment in 1985-2008, we have maintained a constant figure for the population of Rotterdam at 570.000. This number of WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

264 Air Pollution XIX inhabitants per X,Y-coordinate in the URBIS modelling domain has been combined with the air quality for PM10 and EC in the period 1985-2008. The exposure of the population in Rotterdam has been classified to various levels of PM10 and EC and presented in Table 3. 4.3.2 Health effects of PM10 and EC in Rotterdam in 1985-2008 In general, a health impact assessment (HIA) of outdoor air pollution by PM10 and EC is based on four components [9]: 1. 2. 3. 4.

an assessment of the ambient air concentrations of PM10 and EC by monitoring or model-based estimation; a determination of the size of the population exposed to specific concentrations of PM10 and EC; a determination of the health effect of prime interest, including the baseline rate of the health effect being estimated (e.g. the underlying mortality rate in the population in deaths per thousand people); a derivation and application of concentration-response functions from the epidemiological literature that relate ambient concentrations of PM10 and EC to selected health effects.

Population exposure distributions (steps 1 and 2) were taken from Table 3. Health impact calculations were performed for the midpoints of the exposure categories and a rounded value just below the lowest category and above the highest category. Though air pollution has been associated with both mortality and morbidity effects, quantitatively the effects of mortality have been shown to be most important in previous health impact assessments [10]. We therefore focus on quantification of mortality effects. Mortality effects of long-term exposure are substantially larger than mortality effects related to short-term daily exposures [10]. We therefore derived an exposure response function based upon long-term exposure studies. Exposure response functions were selected from a recent review of the evidence for PM2 5 and EC [4]. For PM2 5, we used a relative risk (RR) 1.007 (95% confidence interval 1.002 – 1.011) expressed per 1 µg m-3. For EC we used RR 1.06 (95% confidence interval 1.02 – 1.10) expressed per 1 µg m-3. Note that the RR for EC is a factor 10 higher per mass unit as compared to PM2 5. We assumed that we could apply the PM2 5 exposure response function to the Rotterdam case, even though exposure was characterized as PM10. This may be problematic for the calculation of the health impact for a particular year, but much less so for the calculation of differences between years as most decrease in PM10 is due to a decrease of the fine fraction of PM10 [7]. We have expressed mortality impacts in life years gained or lost estimated with life table calculations [11]. For the calculation we used a population of 500,000 people aged 18 to 64, distributed in age categories comparable to the 2008 Dutch population. We have estimated the effects on this population for a lifetime, as follows: For 1985, 1995 and 2008 we first calculated the life years lost related to the exposure distribution in that year. We then subtracted the life years lost in 1995 and 2008 from the life years lost in 1985 to calculate the gain in life years related to a decrease in concentration. The outcome of the health WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

265

impact assessment is that the decrease in PM10 concentration from 1985 to 2008 results in a gain in life of on average 13 months per person with a range of 7 to 20 months. For EC, a gain of 12 months per person is calculated with a range of 4 to 20 months. The health impact is similar for PM10 and EC. This is explained as follows. The population weighted PM10 concentration dropped from 43 µg m-3 in 1985 to 25 µg m-3 in 2008 which is equivalent to 18 µg.m-3 PM10. EC “only” decreased from 3 to 1 µg/m3 over the same time period. As aforementioned in section 4.1, the decrease in PM10 in the last twenty years in The Netherlands is for 70% due to secondary inorganic aerosol and only for 10% due to primary PM emissions, including combustion aerosol [7]. The similarity in health impact for PM10 and EC suggests that the health impact of PM10 is mainly related to the contribution of combustion aerosol in PM10 and less to the contribution of secondary inorganic aerosol. This demonstrates that measures directed to reduce combustion aerosol (e.g. exhaust emissions of road traffic and (inland) shipping) are more effective to reduce health effects of air quality than reducing PM10 in general.

5 Conclusions and recommendations Our study shows that in Rotterdam in the period 1985-2008 the air quality of PM10 and EC improved significantly both at urban background and near heavy traffic locations. This results in a gain in life on average of 13 months (PM10) or 12 months (EC) per person in Rotterdam. The ten times larger drop in PM10 concentrations as compared to EC results in similar health impact. This is explained by the ten times higher relative risk per µg.m-3 for EC as compared to PM10. This demonstrates that EC is a more sensitive indicator (compared to PM10) to monitor the health effects of traffic measures. It is noted, that EC is likely not causing the health effects but acts as a proxy for the mass of combustion aerosol. Further experimental research is recommended to improve modelling of EC in urban areas (e.g. establish emission factors of EC for free-flowing and congested road traffic) and to validate effects of traffic measures on air quality of EC (e.g. low emission zones and speed limitation).

References [1] Brunekreef B. and Holgate S.T. Air pollution and health. Lancet 360: 12331242, 2002. [2] Schauer J. Evaluation of elemental carbon as a marker for diesel particulate matter. Journal of Exposure Analysis and Environmental Epidemiology 13: 443-453, 2003. [3] Beelen R., Voogt M., Duyzer J., Zandveld P. and Hoek G. Comparison of the performance of land use regression modelling and dispersion modelling in estimating small-scale variations in long-term air pollution

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

266 Air Pollution XIX

[4]

[5] [6]

[7] [8]

[9] [10] [11]

concentrations in a Dutch urban area. Atmospheric Environment 44: 46144621, 2010. Janssen N.A.H., Hoek G., Lawson-Simic M., Fischer P., Bree van L., Brink van H., Keuken M.P., Atkinson R., Brunekreef B. and Cassee F. Black carbon as an additional indicator of the adverse health effects of combustion particles compared to PM10 and PM2.5. Submitted to Environmental Health Perspectives, 2011 PBL Netherlands Emission Factor Database and Annual background concentrations. (http://www mnp nl/bibliotheek/rapporten/500088002.pdf (Table H.1, pp. 109) Ntziachristos L and Samaras Z. EMEP/EEA emission inventory guidebook - COPERT4. www.eea.europa.eu/publications/emep-eeaemission-inventory-guidebook-2009/part-b-sectoral-guidance-chapters/1energy/1-a-combustion/1-a-3-b-road-transport.pdf Hoogerbrugge R., Denier van der Gon H.A.C., van Zanten M.C. and Matthijsen J. Trends in particulate matter. BOP-report 500099014/2010 www.pbl nl/en Denby B., Larssen S., Builtjes P., Keuken M.P., Sokhi R., Moussiopoulos N., Douros J., Borrego C., Costa A.M. and Pregger T. Recommendations on spatial assessment of air quality resulting from the FP6 EU project Air4EU. International Journal of Environmental Pollution Vol. 44 Nos 1-4: 128-138, 2010 Ostro B. Outdoor air pollution: Assessing the environmental burden of disease at national and local levels. Geneva, World Health Organization, 2004 (WHO Environmental Burden of Disease Series No. 5) Künzli N., Kaiser R., Medina S., Studnicka M., Chanel O. and Filliger P. Public-health impact of outdoor and traffic-related air pollution: a European assessment. Lancet 356: 795-801, 2000 Miller B.G. and Hurley J.F. Life table methods for quantitative impact assessments in chronic mortality. Journal Epidemiology Community Health 57: 200-206, 2003

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

267

Assessing air pollution risk potential: case study of the Tohoku district, Japan Y. A. Pykh & I. G. Malkina-Pykh Research Center for Interdisciplinary Environmental Cooperation of Russian Academy of Sciences (INENCO RAS), St. Petersburg, Russia

Abstract The main purpose of the study was to propose the index of air pollution risk potential Kp, to prove its acceptability and to demonstrate the results of its application in a case study of the Tohoku district, Japan. Kp reflects climatic conditions, which are typical for a given area and determine both accumulation and dispersion of pollution in the atmosphere. Important applications of the index Kp are urban planning, industrial location in relation to sensitive areas and air quality management. The results of an index application can be easily interpreted, the number of the elements taken into account is quite small, however, it reflects the basic factors of the air pollution risk potential. Keywords: air pollution, risk potential index, climatological analysis, weather appearance distribution.

1 Introduction The worldwide evidence on airborne particulate matter and its impact on public health consistently shows adverse effects to exposures that are currently experienced by populations in both developed and developing countries. The epidemiological evidence shows impacts following both short-term and longterm exposures. Air pollution is caused by both natural and man-made sources. Major man-made sources of ambient air pollution include industries, automobiles, and power generation. The pollution levels at any place and time represent balance between the rates of emission from their sources and the rate at which they are removed from the atmosphere [1]. The assimilative capacity of the atmosphere determines the dilution and dispersion of the pollutants. From the other hand, air pollutants can WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110251

268 Air Pollution XIX subsequently be deposited on the underlying surface and become a source of soil and water contamination [2]. The main purpose of the study was to analyze the role of climatological factors in the dispersal or diffusion of air pollutants released into the atmosphere of the Tohoku district of Japan and to propose the index of air pollution risk potential Kp in order to assist in industrial zoning as well as city/state or country industrial planning. For air pollution transport with sources such as traffic, power station plumes, industrial plumes or plumes from accidents, most transport interest is within the atmospheric boundary layer (ABL). The ABL is the lower (500 to 1000 m) layer of the earth's atmosphere, which is influenced by the earth's surface shearing and heating effects. Air pollution climatology explains the ability of the atmosphere to dilute or stagnate pollutants over a region at any time [3]. The principal meteorological variables important in the dispersion, transformation and removal of air pollutants, among others, are temperature, horizontal and vertical wind components, water vapor mixing ratio, precipitation, surface flux and boundary layer depth [4]. The majority of these variables change rapidly in the atmospheric boundary layer (ABL). On this basis (ABL) has a significant role in self-purification ability of the atmosphere. ABL controls the vertical extension, concentration and transformation of atmospheric pollution to some extent. Dispersion of pollutants within the ABL is controlled by turbulence which varies strongly with stratification. Maximum mixing depth that is the height in which pollutant can have a vertical movement up to and partially into inversion layer, has a direct relation with ABL [5]. In other words, mixing depth is a critical parameter in determining air pollution concentrations near the ground which represents the depth through which pollutants are vigorously mixed. This parameter is highly important because using such data and the wind speed profiles, the corresponding “ventilation properties” can be calculated [4]. Dry air expanding adiabatically cools at a rate of 9.8°C/km or ~l°C/100m. This is known as “dry adiabatic lapse rate” (DALR) that is the reference lapse rate by which the “ambient lapse rate” is compared. A “neutrally stable” atmosphere occurs when the ambient lapse rate (ALR) is equal to the dry adiabatic lapse rate (DALR); i.e. the rate of cooling with altitude is ~ 1°C/100 m. An “unstable” atmosphere occurs when the ambient lapse rate exceeds the dry adiabatic lapse rate, i.e. the rate of cooling with altitude is > l°C/100m. This steeper temperature gradient encourages greater thermal turbulence. This ambient condition is said to be “unstable” with a “superadiabatic” lapse rate. A “stable” atmosphere occurs when the ambient lapse rate is less than the dry adiabatic lapse rate, i.e. the rate of cooling is < 1°C/100m. The temperature gradient is less steep and thus responsible for less turbulent activity. The ambient condition is said to be “stable” with a “subadiabatic” lapse rate [6]. A “stable inversion” condition is a variant of the stable atmosphere. Here, the temperature increases with altitude. An inversion temperature condition is a very stable condition, forcing air pollutants to remain trapped in the atmosphere for long periods [7]. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

269

Under stable atmospheric conditions the mixing height can drop to less than a hundred meters. This can trap pollutants in a very thin layer near the ground and result in high concentrations. Pollutants trapped above the mixing height also can cause high concentrations when the mixed layer at the ground lifts and suddenly brings the pollutants to the ground. It is obviously important in predicting pollutant dispersion to know the speed and direction of wind. The variation of the horizontal wind speed with height is important in evaluating diffusion of pollutants [8]. When pollutant concentration is low and precipitation is high, the air pollution level is low since the rain will be able to thoroughly wash all the pollutants away. When aerosol concentration is high and precipitation is high, the air pollution is still manageable since the rain will still have enough moisture to wash most of the aerosol away from the atmosphere. It is when aerosol is high and precipitation is low, the condition of air pollution will be the worse. Based on the results of the previous studies [9–11] we propose the index or air pollution risk potential Kp. The index includes both favorable and unfavorable meteorological conditions that determine the air pollution risk. Unfavorable for atmospheric ventilation meteorological conditions are wind velocity of V < 1 m/sec; vertical temperature gradient in the lower (500 m) layer of the atmosphere of γ0 5 ≤ 0.4°/100 m and amount of precipitation of Q ≤ 1 mm. In contrast, favorable meteorological conditions are V > 3 m/sec, γ0 5 ≥ 0.8°/100 m and Q ≥ 3 mm. The associations between γ0 5 and surface temperature differences at 6 and 9 p.m. (ΔT18 ,21 ) , at 9 p m. and daily minimal temperature (ΔT21,min ) , daily maximum and minimal temperatures (ΔTmin,max ) were investigated [10]. The highest correlation was obtained between γ0 5and ΔTmin,max , r = 0.64. Thus, in the proposed index of air pollution risk potential (APRP) the probability P(γ0 5≤0.4°C/100m) was replaced with the probability P( ΔTmin,max ≥9°C), and the probability P(γ0 5≥0.8°C/100m) was replaced with the probability P( ΔTmin,max ≤5°C). Then the equation for calculation of the APRP index (Kp) is looking as follows:

Kp 

P(V  1.0)  P( Tmin,max  9.0)  P(Q  1.0) P(V  3.0)  P( Tmin,max  5.0)  P(Q  3.0)

(1)

The probabilities P in the given equation can be replaced with the number of days with the mentioned values of variables under consideration. Then, the APRP index is looking as follows: Kp 

N w  Nt  Nq MW  M t  M q

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

(2)

270 Air Pollution XIX where Nw is the number of days with calm wind (V≤1 m/s); Nt is the number of days with daily temperature gradient  9ºC; Nq is the number of days with precipitation ≤ 1 mm; Mw is the number of days with wind speed ≥3 m/s; Mt is the number of days with daily temperature gradient  5ºC; Mq is the number of days with precipitation ≥ 3 mm. The values of Kp reflect climatic conditions, which are typical for a given area and which determine both accumulation and dispersion of pollution in the atmosphere. The following index rates were calculated: Kp = 0.3 – 1.0 is a low risk potential, Kp = 1.0 – 2.0 is moderate, Kp = 2.0 – 5.0 is high, and Kp > 5.0 is very high. The higher is the Kp value the worse conditions are for the pollution dispersion and removal. If the value of Kp < 1 then the self-purification process exceeds the accumulation of pollutants in the atmosphere. Index Kp was applied for defining the potential air pollution risk zones in the Tohoku district, Japan. The following section briefly describes the weather variables distribution in the Tohoku region with the special focus on those included into the index of air pollution risk potential.

2 A climatological analysis of the weather variables distribution in the Tohoku district Tohoku is a geographical area of Honshu, largest island of Japan. This region is also called Ou and is even referred as Michinoku. The region occupies about one - fifth of the total area of Japan. It is located on the northeastern part of Honshu. Tohoku comprises of six prefectures - Akita, Aomori, Fukushima, Iwate, Miyagi and Yamagata Prefectures. The topography of the Tohoku district is very complicated. The Ou mountains as the backbone range run through the district north to south, dividing it into the Pacific side and the Japan Sea side. On the Japan Sea side, the Dewa hills and the Echigo mountains run north to south parallel with the backbone range, making may basins between them. On the Pacific side, the Kitakami mountains and Abukuma mountains run parallel with the backbone range. According to the land feature, weather distribution shows complicated pattern in the Tohoku district. Two primary factors influence Tohoku’s climate: a location near the Asian continent and the existence of major oceanic currents. The climate of Tohoku is influenced by the monsoon, especially in winter. The winter monsoon out of Siberia is characterized by high pressure and polar continental air that is greatly modified by the warm Japan Sea in its eastward progress. Frequent snow and cloudy conditions, therefore, are the norm in the windward Tohoku Japan Sea side, while sunny, dry weather prevails on the Pacific side. Dominant weather factor for most of the country during summer is the Ogasawara High, a lower tropospheric anticyclone, associated curved trade winds from the tropics which constitute the summer monsoon. In fact, 70 to 80 percent of the annual precipitation falls in the period between June and September [12].

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

271

Two major ocean currents affect this climatic pattern: the warm Kuroshio Current (Black Current; also known as the Japan Current); and the cold Oyashio (Parent Current; also known as the Okhotsk Current). The Kuroshio Current flows northward on the Pacific side of Japan and warms areas as far north as Tokyo; a small branch, the Tsushima Current, flows up the Sea of Japan side. The Oyashio Current flows southward along the northern Pacific, cooling adjacent coastal areas. The characteristics concerning the weather appearance are described according to five regions. They are as follows [13]: 1) The coastal region along the Japan Sea and 2) coastal region along the Pacific Ocean. The side of the country which faces the Sea of Japan has a climate with much rain and snow, produced when cold, moisture-bearing seasonal winds from the continent are stopped in their advance by the Central Alps and other mountains which run along Japan’s center like a backbone. In winter, the JapanSea coast is exposed to the northwestern monsoon with the mountain in the background. It is, therefore, considered that the distribution of the amount of precipitation in proximity to the Japan Sea coast is not yet complicated by landform but correlates with circulation patterns in the Far East. The Pacific Ocean side of the Tohoku region belongs to the temperate zone and its summers are hot, influenced by seasonal winds from the Pacific. Temperature differences between the Japan Sea and the Pacific sides of the Tohoku district exist not only because of weather conditions, but coastal configuration and the effect of ocean currents as verified by sea surface temperatures as well. Indeed, coastal configuration and ocean currents may be of primary significance in affecting temperatures. 3) The western slopes in Echigo mountains and in the northern part of the Ou mountains. These regions have highest frequency of snowy weather throughout the Tohoku district. As these regions are fully exposed to the W-NW monsoon, snowy weather is frequently seen in every flow-pattern in the winter type pressure pattern. Cloudy weather is not so frequent that the number of fine days in the slopes is larger than in the Japan Sea coast. In these slopes, it snows always when it snows anywhere in the western side of the backbone range, because almost all of the snowing in this side appears as the extension of the snowing in each of the slopes. In this sense, both slopes form the original areas of snowy weather in the Tohoku district. 4) The group of basins in the western side of the backbone range. The appearance frequency of every kind of weather varies with basins, because the topographical conditions of each basin is different each other. The cloudy or snowy weather of the Japan Sea coast tends to spread over the Hirosaki basin and seldom spreads to the basins of Yamagata, Yonezawa and Aizu. Anyway it is frequently snowy and seldom fine in the basins of Shinjo and Obanazawa and in the Aomori region, and in both the basins of Yokote and Yonezawa only in the winter-type pressure pattern. It is less snowy and more fine in both the basins of Yamagata and Aizu. 5) Inland regions on the eastern side of the backbone range. In these regions, on the leeside of the backbone range, snowy or cloudy weather with wide WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

272 Air Pollution XIX extension is rarely seen in mid-winter, because the backbone range interrupts the eastward extension of the weather. They have more of fine weather than the opposite side of the range. Since the shelter-effect by the back born range does not work at the lowered saddles in the back born range, snowy or cloudy weather on the western side extends occasionally eastward through the parts. The regions at such a location are the region around Okunakayama, the regions of Hanamaki and Mizusawa, Sempoku plain the Fukushima basin, and the Koriyama basin. The frequency distribution of snowy weather appearance, which is characterized by the pattern that snowy weather is comparatively frequent in these region when the weather is seen in any place (here Fukushima) leeward of the saddle-like part. There have been several studies of the observed air temperature profiles over mountain slopes compared with nearly free-air temperatures [14, 15]. According to these reports, temperature differences between mountain-air and free-air exhibit considerable variability as related to season, the type of air mass (i.e., wind direction, wind speed), the time of day (i.e., radiative and turbulent heat exchanges), cloud amounts, and the existence of snow-cover. The primary control of the temperature difference between free-air and summit-air appears to be the atmospheric temperature structure related to the lapse rate and adiabatic lapse rate. Most of the mountain-air temperatures exhibit larger variations than the freeair temperatures. The sequence of temperature differences (mountain-air minus free-air) of each mountain can be divided into three groups which have a positive, negative and mixed sense of the temperature difference. When the Ogasawara high extends over Japan, large positive temperature differences are found in the central part of the mountainous region. With respect to the spatial variations of the differences, they appear to clearly change with latitude. Negative or small differences are predominant in the higher latitudes, near the coast of the Japan Sea, and in the lower latitudes, close to the Pacific Ocean. In the central mountainous region, however, the sense of the difference is positive. It appears that heating of the mountain surface plays an important role in this temperature distribution pattern, since this pattern becomes sharper under the conditions of the well-developed Ogasawara high.

3 Calculation of air pollution risk potential index in the Tohoku district Numerous climatological studies of Japan stress monsoonal influences on selfpurification ability of the atmosphere in winter and in summer as well. The most unfavorable for air self-purification are summer climatic conditions with Ogasawara High as dominated weather factor for most of the country and associated stable stratification and curved trade winds from the tropics which constitute the summer monsoon. To calculate the air pollution risk potential index we used daily minimum and maximum surface air temperature, wind velocity and amount of precipitation

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

273

data in January and August at 18 weather stations in the Tohoku district, compiled by the Japan Meteorological Agency (JMA), 1978-1998 [16]. In winter the Japan Sea side of the Tohoku district are characterized with low APRP index values Kp =0.3-1.0. These index values are presented in western, north-western and northern part of the district. The winter monsoon out of Siberia brings frequent snow and cloudy conditions to the windward Tohoku Japan Sea side, while sunny, dry weather prevails on the Pacific side. Since the cyclones developing in the Pacific polar frontal zone pass along the east coast of Asia to the northeast, the polar air out-breaks behind cyclones. In the season, therefore, the current from Siberia is not stationary, but the period of violent blow appears alternately. A cold, dry, continental polar air from Siberia is much modified over the Japan Sea, and becomes warm, moist and unstable after passing the Japan Sea. Islands of Japan lie athwart the air-stream, the Japan Sea side of the Islands is exposed to it, and the Pacific side is sheltered by the relief of mountains. In winter the moderate APRP index is presented on the Pacific coast of Tohoku Kp =1.0-2.0. In winter it snows frequently on the former wind-ward side, but it is usually dry, sunny and fine on the latter leeside with high daily temperature gradients. As it was already marked, high daily temperature gradients result in increasing of stability of stratification of a ground layer of an atmosphere, and consequently, in increasing of potential of air pollution. There are two regions of the Tohoku district with high APRP index in winter Kp >2. These are the areas of Yamagata City and the basin of Abukuma river. Very high values of the APRP index in winter (Kp >5) appear also on the Pacific Ocean side near Miyako City. The temperature in Miyako City and the regions north of this city tend to drop due to effects from the cold current Oyashio which results in a more stable stratification of the atmosphere. Also, this region is located in the basins of Kawaki river, sheltered with Kitakami mountains from NW air masses. Summer weather conditions of the Tohoku region are less favorable for selfpurification of the atmosphere than winter weather because of the influence of Ogasawara High that is characterized with high thermal stability and low wind velocities. In summer the low values of risk potential are shown for the Japan Sea coast Kp =0.3-0.1. Summer season on the Japan Sea coast is characterized with a high amount of precipitation (Q = 100 мм). On the Pacific Ocean side of the district the major part of Miyagi prefecture is characterize by low values of the APRP index in summer. Low APRP index values are also presented in the plain Mutsu that occupies most of the Shimokita Peninsula. It has a cold maritime climate characterized by cool summers and cold winters with heavy precipitations (Q =130 mm) resulted from Ogasawara High influence. The major part of the Pacific Ocean side in summer is characterized with moderate values of the APRP index (Kp =1.0-2.0). These values are the results of the combination of stable stratification of the atmosphere and large amount of precipitations. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

274 Air Pollution XIX

Figure 1:

APRP index values, Tohoku district, Japan.

The central part of the Tohoku district in summer is characterized with high values of the APRP index (Kp > 2.0). This is the result of the combination of WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

275

meteorological variables under study – stable stratification of the atmosphere, low wind velocities (V = 0.2–1.2 m/sec) and low amount of precipitation (Q=80 mm). In summer high values of the APRP index appear on the Pacific Ocean side near Miyako City. Very high values of the APRP index in summer (Kp ≥ 5.0) appear in Wakamatsu basin, first of all, because it is sheltered by the relief of mountains. The combination of winter and summer APRP index values is presented in Figure 1. The regions with stable and unstable air pollution risk potential appear within the Tohoku district. The stable low APRP index values appear on the Japan Sea side of the Tohoku district and the Shimokita Peninsula. The stable moderate APRP index values appear on the northern part of the Pacific Ocean coast of Tohoku. The stable high APRP index values appear in the Yamagata and Koriyama basins. The central and north-eastern parts of the Tohoku district are characterized with unstable APRP index values in winter and summer seasons. The summer values of the APRP index are marked with colors, the winter values are marked with shadings.

4 Conclusion The assimilative capacity of the atmosphere determines the dilution and dispersion of the pollutants. The most important atmospheric conditions are wind speed, amount of precipitation and the vertical temperature characteristics of the local atmosphere. The index of air pollution risk potential was proposed. It is calculated on the base of the unfavorable and unfavorable for air self-purification values of climatic variables. Index Kp was applied for defining the potential air pollution risk zones in the Tohoku region, Japan. Kp reflects climatic conditions, which are typical for a given area and which determine both accumulation and dispersion of pollution in the atmosphere.

References [1] Wark, K. & Warner, C.F. Air Pollution, Its Origin and Control. IEPA Dun Donnellyey Publishers, New York, 1976. [2] Malkina-Pykh, I.G. & Pykh, Yu.A. The Method of Response Function in Ecology. WIT Press, Southampton Boston, 2000. [3] Arya, S.P. Air Pollution Meteorology and Dispersion. Oxford University Press, New York, 1999. [4] Cheremisinoff, N. Handbook of Air Pollution Prevention and Control. Butterworth Heinemann, Elsevier Science, USA, 2002. [5] Deardorff, J. (1980). Progress in understanding entrainment at the top of a mixed layer. Workshop on Planetary Boundary Layer. American meteorological field measurements and their interpretation. Villefranche Sur Mer, France, 4 May, 1988. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

276 Air Pollution XIX [6] Holzworth, C.G. Climatological aspects of the composition and pollution of the atmosphere. W.M.O., Tech. Note, 139, pp. 89-91, 1974. [7] Hosler, C.R, Low-level inversion frequency in the composition and pollution of atmosphere. Mon. Wealth. Rev., 89, pp. 319-339, 1961. [8] Ogawa, Y, Diosey, P.G., Uehara, K. & Ueda, H. (1985) Wind tunnel observation of flow and diffusion under stable stratification. Atmospheric Environment, 19, pp. 65-74, 1985. [9] Linevich, I.L., Sorokina, L.P. Climatic potential of atmosphere selfpurification: experience in different scale estimation. Geography and Natural resources, 4, pp. 160-165, 1992. [10] Lomaya, O.V. The problem of the estimation of air pollution potential // Meteorological aspects of air pollution. Proceedings of the Institute of Geophysics of Gorgian Academy of Sciences, XIV, pp. 14-29, 1979. [11] Monin, A. S. & Obukhov, A.M. Basic laws of turbulent mixing in the ground layer of atmosphere. Trans. Geophys. Inst., Akad. Nauk USSR., 151, pp. 1963-1987, 1954. [12] Suzuki, H. The classification of Japanese climate. Geogr. Rev. Japan, 35, pp. 205-211, 1962. [13] Shitara, H. A climatological analysis of the weather distribution in the Tohoku district in winter. Sci. Rep. Tohoku Univ., 7(15), pp. 35-54, 1966. [14] Makita, H. Mountain air temperature in comparison with upper layer temperature over lowland. Sci. Rep. Tohoku Univ., 7(28), pp. 19-26, 1978. [15] McCutchan, M.H. Comparison temperature and humidity on a mountain slope and in the free air nearby. Mon. Wea. Rev., 3, pp. 836-845, 1983. [16] Japan Meteorological Agency. Meteorological Observations for January – July, 1978-1998.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

277

Assessing the potential for local action to achieve EU limit values J. H. Barnes, T. J. Chatterton, E. T. Hayes, J. W. S. Longhurst & A. O. Olowoporoku Air Quality Management Resource Centre, University of the West of England, Bristol, UK

Abstract Despite 14 years of UK Local Air Quality Management (LAQM), ambient NO2 concentrations have not decreased as expected. Although NO2 concentrations decreased from 1996 to 2002-4, this trend has subsequently levelled off. The UK Government has failed to meet European Union (EU) limit values for NO2 and PM10 and risks incurring fines of ~£300m. The number of local authorities (60%) having declared Air Quality Management Areas (AQMAs), primarily for trafficrelated pollutants (NO2 and PM10), has grown steadily since 2001, and despite the production of local Air Quality Action Plans (AQAPs) there have been no traffic-related AQMA revocations solely on the basis of their implementation. The UK Air Quality Strategies (1997-2007) have focussed on emission reduction technologies to reduce overall pollutant concentrations, whilst LAQM targets specific local hotspots often through air quality measures in Local Transport Plans. The failure of this system to achieve necessary NO2 reductions has been attributed in part to a reliance of national policy on Euro vehicle standards, without significant endeavour to reduce road traffic growth. Locally, deficiencies in funding, interdepartmental communication, political will and public awareness, have been criticised for hindering action plan measures. The UK Government’s localism agenda threatens to reduce the top-down governance of LAQM whilst also introducing the potential for EU fines to be passed to local authorities where limit values are exceeded. At the same time, the UK Government has outlined changes that will put more emphasis on the development of local measures to achieve EU limit values. This paper discusses

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110261

278 Air Pollution XIX the potential for local action to achieve the limit values and concludes that further work is necessary at a national level to assist local authorities in this aim. Keywords: local air quality management, air quality action plans, NOx, NO2, PM10, EU limit values, nitrogen dioxide, particles, air pollution, localism.

1 Introduction Part IV of the UK Environment Act 1995 (HM Government [1]) established a range of roles and responsibilities for both national and local government with respect to air quality management. The Act was a pre-emptive approach to address the 1996 EU Framework Directive (Council Directive 96/62/EC) (European Commission [2]) (subsequently replaced by the Council Directive on Ambient Air Quality and Cleaner Air for Europe (2008/50/EC) (European Commission [3])), the daughter directives of which imposed limit values for specific pollutants to be achieved by Members States. The Environment Act 1995 and the subsequent 1997 National Air Quality Strategy (Department of the Environment [4]) recognised the impact of traffic emissions on ambient air quality, and established national air quality objectives for the seven pollutants of concern (NO2, PM10, SO2, CO, benzene, 1,3-butadiene and lead) which reflected the EU limit values. The Strategy also divided responsibility for managing air quality between central government, which was expected to carry the main burden of the air quality improvements by reducing pollutant concentrations across all relevant locations, and local authorities, whose focus was on tackling residual local pollution hotspots (Longhurst et al. [5]). Little was known at that time about the extent or magnitude of these hotspots, which were assumed to be localised and limited to major urban areas (Chatterton et al. [6]). The experience of local authorities over the last 14 years, however, has found that exceedences of the national air quality objectives (and EU limit values) for NO2 and PM10 are common and widespread, and tend to occur wherever high volume and/or congested traffic and residences coincide. There are currently 244 (60%) local authorities in the UK with AQMAs, primarily for NO2 and PM10 from traffic sources (Defra [7]), where local authorities are required to work towards reducing ambient concentrations of these pollutants to meet the air quality objectives through the production and implementation of AQAPs. However, to date there is no evidence of any trafficrelated AQMAs having been revoked solely on the basis of their implementation (Longhurst et al. [5]). Similarly, national government is also in the process of preparing annual Air Quality Plans for the European Commission in respect of its failure to meet EU limit values for NO2 and PM10 by the specified deadlines. The UK is currently exceeding the NO2 annual mean limit values + Margin of Tolerance (MOT) (48 µg/m3) in 40 out of 43 zones and agglomerations and are intending to submit a time extension notification and Air Quality Plan in September 2011 (Defra [8]). Exceedences of the PM10 limit values + MOT are limited to Greater London and a time extension has recently been awarded to June 2011 (European Commission [9]). Failure to comply with the EU limit

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

279

values may incur significant financial penalties estimated to be in the region of £300 million (ENDS [10]).

2 Why are limit values being breached? 2.1 Emission factors/Euro standards The national strategy to reduce concentrations of NOx and NO2 has, to date, focused predominantly on emission reduction strategies, primarily relying on the integration of Euro standard vehicles into the national fleet, often in advance of formal compliance dates. However, recent work suggests that this reliance may have been overoptimistic (Carslaw et al. [11]). Although trends in ambient NOx and NO2 in the UK decreased from ~1996 to 2002-4, from 2004 to 2009 these trends have levelled off. Furthermore, whilst total NOx emissions may have decreased, NOx emitted as primary NO2 from motor vehicles has increased from 5-7% to 15-16% over the same period (Carslaw et al. [11]). Data from roadside remote sensing detectors has suggested that, under typical urban-type driving conditions, emissions are higher than those recorded in the National Atmospheric Emissions Inventory (NAEI). In particular, NOx emissions for the later Euro standard diesel and light duty vehicle classes have not fallen as predicted and instead have remained relatively stable over the last 15 years (Carslaw et al. [11]). Also, the slight decrease in NOx from petrol emissions has, to some extent, been offset by a significant increase in the number of diesel vehicles on the road (partly due to government policies such as adjustments to vehicle excise duty that have encouraged ‘dieselisation’ for climate change reasons). There may also be further underestimations regarding the rate of vehicle renewal, meaning that there are potentially older, dirtier vehicles on the road than previously anticipated. These flawed NAEI emission factors have also been used by the Department for Food and Rural Affairs (Defra) to derive future year projection factors for roadside NO2, which have in turn been used by local authorities to predict when local roadside NO2 concentrations are likely to fall below air quality objective levels, and by developers undertaking air quality assessments to indicate the likely future impact of their developments. Although the validity of the emission factors had already been called into question as forecast concentrations failed to reflect monitoring trends (Institute for Air Quality Management [12]), this revelation is likely to have repercussions for local air quality where AQAPs may fall short of expectations and developments that may not otherwise have been approved on air quality grounds have been granted planning permission. 2.2 Lack of interdepartmental responsibility/political will One of the main criticisms voiced in the recent UK Environmental Audit Committee report on air quality was the lack of interdepartmental communication in central government regarding the importance of air quality (Environmental Audit Committee [13]). Despite the clear cost-benefit case and WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

280 Air Pollution XIX obvious need for transport policy to play its part in reducing the effects of traffic emissions, air pollution does not appear to be a significant political priority. This is illustrated by the Transport Minister’s response to the UK’s failure to meet the EU limit value for NO2, in which he is reported to be querying with Defra and the EU the validity of the limit value which he claims has “perverse side effects” and will require “disproportionate effort” to achieve (Air Quality Bulletin [14]). The fact that ‘improving air quality’ has been demoted from a “shared priority” with ‘tackling congestion’, delivering accessibility’ and ‘ensuring safer roads’ in the second round of Local Transport Plans (LTP) to a secondary consideration in Round 3, again conveys a lack of political recognition. (Though evidence from Round 2 strongly suggested that “shared priority did not mean equal priority” (Olowoporoku et al. [15]).) Even where central policy guidance has included air quality (e.g. Planning Policy Statement (PPS) 23 (Office of the Deputy Prime Minister [16]), which established air quality as a material planning concern, and Public Service Agreement (PSA) 28 (HM Treasury [17]), which gave the Department for Transport and Defra joint ownership of the air quality indicators for NO2 and PM10), air quality still fails to receive adequate recognition. A lack of recognition of the importance of air quality at a national level is reflected at the local level in terms of political priorities. Environmental Health Officers (EHOs) struggle to raise awareness of local air quality and the lack of perception of the relative importance of air quality restricts their ability to negotiate action plan measures and air quality assessments with transport and planning departments, particularly where there are conflicting priorities regarding economic development promoted by central government departments (Carmichael and Lambert [18]). At a local level there may also be the limitations of local politics with the potential for vested interests and short-termism and a lack of political leadership and courage to pursue more novel and sustainable approaches to economic development. 2.3 Lack of funding Air quality management, including monitoring and modelling, is an expensive operation, which is significantly underfunded due in part to its low political status. Limited air quality grants are available annually from Defra but are always oversubscribed – £2m has been made available for 2011/12 to assist the 244 local authorities with one or more AQMAs to implement their AQAP measures. In previous years the Defra grant has been ring-fenced for air quality purposes, however this “limitation” has been removed in 2011 to allow local authorities more flexibility to manage their own priorities, given the substantial cuts in public funding that local authorities are currently facing. For those local authorities that are able to link their AQAPs with LTPs, measures to improve local air quality can also be funded through this means. However, the reductions in LTP funding too, together with the deprioritisation of air quality in Round 3 of the LTP process, will inevitably mean difficulties in ensuring air quality improvements are properly considered. There are also mechanisms to obtain funding from developers to offset the air quality impacts of developments through the planning process under section 106 WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

281

agreements. These funds are typically acquired at the planning approval stage, however, political pressures to drive economic development will likely restrict any potential obstruction that may be seen to limit inward investment. 2.4 Scientific complexities 2.4.1 Health Part of the reason for the lack of political support for air quality issues stems from a lack of understanding. Despite the publication of reports from the Committee on the Medical Effects of Air Pollution (COMEAP) expounding the health impacts of human exposure to pollutants, it may be difficult for nonexperts to realise the significance of “200,000 premature deaths” or an “average two years life lost” (COMEAP [19]). Added to that, these figures are often couched in terms of uncertainty as the nature of air pollution means that, unlike road traffic accidents for example, it is not identifiable as a direct physiological cause of death, but as a contributory factor in reducing people’s overall lifeexpectancy, along with many other contributory, and potentially inter-related, lifestyle factors such as diet and exercise patterns. Such reports are also usually pollutant specific, whereas public exposure is not to a single pollutant exclusively but to a range of interacting and potentially exacerbating pollutants. It is also unclear whether the separately reported health effects of short-term and long-term exposure to pollutants are additive or exclusive, what the specific or cumulative effects of fine and ultrafine particles are, or the effects of pollutants on morbidity as well as mortality (COMEAP [19]). 2.4.2 Monitoring and modelling The assessment of local air quality depends on the accurate monitoring and modelling of pollutant concentrations. Unfortunately the physicochemical properties of air pollution are complex and significant scientific uncertainties remain. In order to be comparable with the health-based air quality objectives and limit values monitoring methods should represent public exposure to the pollutants in question. The technical difficulties of measuring pollutants over specified averaging periods, whilst accounting for the effects of meteorology, technology and human error, means that there is inevitably a degree of uncertainty in the monitoring results. Indeed the most common method of monitoring NO2 in the UK, using passive diffusion tubes, is subject to ±25% uncertainty (Defra [20]). Although precision and accuracy are controlled to some extent through the use of quality assurance and quality control procedures and techniques, there is still an acceptance that there will be a certain level of uncertainty inherent in the monitoring results. Dispersion modelling too has inherent error, as a simplification of reality with limited inputs and various assumptions made in the absence of complete and accurate data. Models are also often verified and adjusted against monitoring data and so are therefore only as accurate as the monitoring data against which the results have been assessed. This uncertainty in assessment data is difficult to convey to non-air quality specialists, e.g. developers, land use and transport planners and even the general WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

282 Air Pollution XIX public, which can hamper political acceptance of air quality problems and the integration of air quality into LTPs and local planning guidance.

3 Where next for local air quality management? 3.1 National agenda There has been a move to reduce the bureaucratic burden on local authorities under the Government’s ‘Freedoms and Flexibilities’ agenda since the publication of the Local Government Act 2000 (section 6) (HM Government [21]). However, the current UK Government is making this a political imperative though the ‘localism’ agenda. The Localism Bill (HM Government [22]), introduced to the UK Parliament in December 2010, included a clause in Part 2 enabling the transfer of responsibility for EU penalties to local authorities, however the precise detail is subject to Parliamentary approval. This potential for local authorities to find themselves financially liable for breaches of the EU limit values has been described as “unfair” by local authorities whose legal remit was to develop AQAPs “in pursuit” of achieving the national air quality objectives and with no direct responsibility to the EU (Local Government Association [23]). While Defra may be keen to share responsibility for meeting the EU limit values with local authorities, they also maintain that they will be available to support them. Although there are not yet any firm guidelines on how the LAQM framework is likely to change, some of the possibilities have been outlined (Air Quality Bulletin [24]), including:     

Consolidating EU and national air quality objectives. Sharing information on compliance assessment with local authorities. Including local AQAP measures in national Air Quality Plans. Continuing local screening for hotspots. Introducing proportionate screening and reporting.

3.2 Local implications Local authorities in the UK are currently under significant pressure to ‘do more with less’ as they face an estimated £6.5bn public funding cut over the next two years (Local Government Association [25]). This will have significant consequences for air quality management, which is already marginalised by other financial and political priorities. The removal of ring-fencing from air quality grants will further reduce the ability of EHOs to safeguard future monitoring and to provide resources for the implementation of AQAPs. As shrinking budgets force redundancies in local government, EHOs will also find themselves stretched to cover additional duties and having to deal with new priorities. One of the implications of the current UK Government’s move towards localism is that public health will be devolved from regional Public Health Authorities to upper-level councils. For unitary or metropolitan councils this will WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

283

put LAQM and public health within the same local government body, which could benefit communication between EHOs and public health officers. For those local authorities that operate in a two-tier municipality (i.e. district and county councils) public health will reside with the county council, whilst environmental health responsibilities will continue to sit at a lower level, district council. This will create the same inter-organisational divide between air quality and public health as exists between air quality and transport planning, where district level EHOs have experienced difficulties in raising the profile of air quality (Olowoporoku et al. [15]). What is not clear is what local authorities will choose to prioritise under localism. Communities and neighbourhoods are likely to be given greater rights and remit to influence local politics and to drive the local political agenda, however if this is the case the implications for air quality are not promising. At present, outside of London, the public could be forgiven for assuming air quality is good, given the absence of the dense smogs of 60 years ago. Public admission of air quality as a “problem” may also be clouded by a reluctance to admit complicity, as a nation of motorists, or ownership of the solution in the absence of alternative modes of transport or requirements to travel. There are also legal responsibilities incumbent on local authorities under the Environment Act 1995, namely to declare AQMAs where air quality objectives are being breached, and subsequently to prepare Further Assessments and AQAPs to investigate and work towards meeting the objectives. Unless the Act is amended by Parliament, local authorities will still be required to meet these obligations. The decentralisation of power and devolution to local authorities could be seen as potentially damaging for the future of LAQM, which until now has maintained its, albeit limited, profile to a large degree due to the statutory annual reporting requirements. While the focus on diagnosis and reporting has itself been criticised as detracting from the more challenging need to manage air quality (In-House Policy Consultants [26]), there is a risk that if the statutory responsibility is removed, and local authorities are left to devise their own priorities, then air quality will come second to other more tangible or visible local needs. On the other hand, it may (optimistically) be argued that without the conflicting priorities of disconnected central government departments, local authorities may be able to manage air quality as part of a more holistic, locally sustainable approach to development.

4 Conclusions When first devised, the role of LAQM was undervalued by government as a supplementary role in achieving the EU limit values, and has suffered a lack of political support because of this early misconception. Over the following 14 years however, local authorities have risen to the challenge of LAQM and have excelled at diagnosing air quality problems. Their ability to successfully devise and implement AQAPs subsequently, however, has been constrained by other political priorities and a seeming lack of appreciation of the significance of air WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

284 Air Pollution XIX quality issues by departments outside of environmental protection, both at a local and a national level. In addition to identifying the difficulties that UK local authorities have faced to date in managing local air quality, this paper has indicated that the forthcoming localism agenda is unlikely to improve this situation and may have the potential to reduce local authorities’ ability to meet national air quality objectives and, therefore, EU limit values. At the same time, the failure of the UK Government to achieve EU limit values has finally brought to Defra’s attention the value of local AQAPs. What remains to be seen is how the government propose to ensure that they can rely on effective local action in the event of devolved power and moreover how they can best assist and support local authorities to bring these plans to fruition.

References [1] HM Government, Environment Act 1995, The Stationery Office, London, 1995. [2] European Commission, Council Directive 96/62/EC of 27 September 1996 on ambient air quality assessment and management. 21/11/1996. Official Journal L 296, 55–63, 1996. [3] European Commission, Council Directive 2008/50/EC of 21May 2008 on ambient air quality and cleaner air for Europe. 11/6/2008. Official Journal L 152, 1–44, 2008. [4] Department of the Environment, The United Kingdom National Air Quality Strategy, 1997. [5] Longhurst, J.W.S., Irwin, J.G., Chatterton, T., Hayes, E.T. & Leksmono, N.S., The Development of Effects Based Air Quality Management Regimes, Atmospheric Environment, 43(1) 64-78, 2009. [6] Chatterton, T.J., Longhurst, J.W.S., Leksmono, N.S., Hayes, E.T. & Symons, J.K., Ten years of Local Air Quality Management experience in the UK: An analysis of the process, Clean Air and Environmental Quality, 41(4): 26-31, 2007. [7] Defra, List of Local Authorities with AQMAs, http://aqma.defra.gov.uk/ list.php [Accessed 31/5/11]. [8] Defra, Air Pollution in the UK 2009- Edition B, 2010. [9] European Commission, C(2011) 1592 final, Commission Decision of 11.3.2011 on the notification by the United Kingdom of Great Britain and Northern Ireland of an exemption from the obligation to apply the daily limit value for PM10 in zones UK0001 and UK(GIB) 2011. [10] ENDS Report, Government seeks time to meet air quality limits, 409, p. 46, February 2009. [11] Carslaw, D., Beevers, S. Westmoreland, E. Williams, M. Tate, J. Murrells, T. Stedman, J. Li, Y., Grice, S., Kent, A. and I. Tsagatakis (2011). Trends in NOx and NO2 emissions and ambient measurements in the UK. Version: 3rd March 2011. Draft for Comment.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

285

[12] Institute for Air Quality Management, Report on the Meeting on Future NO2 Concentrations – 1 July 2010, http://www.iaqm.co.uk/text/resources/ report_on_no2.pdf [Accessed 1/6/11]. [13] Environmental Audit Committee, Air Quality, Fifth Report of Session 2009-10, Vol. 1, HC 229-I, The Stationery Office, London, 22 March 2010. [14] Air Quality Bulletin, Government lobbies on NO2, 60, p.1, April 2011. [15] Olowoporoku, A.O., Hayes, E.T., Leksmono, N. S., Longhurst, J.W.S. and Parkhurst, G., A longitudinal study of the links between Local Air Quality Management and Local Transport planning policy processes in England. Journal of Environment Planning and Management, 53(3), 385-403, 2010. [16] Office of the Deputy Prime Minister, Planning Policy Statement 23: Planning and Pollution Control. Annex 1: Pollution Control, Air and Water Quality, 2004. [17] HM Treasury, PSA Delivery Agreement 28: Secure a healthy natural environment for today and the future, 2007. [18] Carmichael, L. and Lambert, C., Governance, knowledge and sustainability: the implementation of EU directives on air quality in Southampton, Local Environment, 16: 2, 181-191, 2011. [19] COMEAP, The Mortality Effects of Long-Term Exposure to Particulate Air Pollution in the United Kingdom, Health Protection Agency, ISBN 978-085951-685-3, 2010. [20] Defra, Part IV of the Environment Act 1995 [Environment (Northern Ireland) Order 2002 Part III] Local Air Quality Management Technical Guidance LAQM.TG(09), 2009. [21] HM Government, Local Government Act 2000, The Stationery Office, London, 2000. [22] HM Government, Localism Bill, December 2010. [23] Local Government Association, Localism Bill – EU Fines, LGA Briefing, April 2011. [24] Air Quality Bulletin, Defra hints on future changes, 60, p.5, April 2011. [25] Local Government Association, Funding settlement ‘will lead to cuts in services’, First Online, December 2010, http://www.lga.gov.uk/lga/core/ page.do?pageId=15662390 [Accessed 1/6/2011]. [26] In-House Policy Consultants, Review of Local Air Quality Management, A report to Defra and the devolved administrations, March 2010.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

This page intentionally left blank

Air Pollution XIX

287

A procedure for the evaluation of the historical trend of atmospheric pollution in an urban area F. Murena1 & M. Urciuolo2 1

Chemical Engineering Department, University of Naples “Federico II”, Italy 2 Combustion Research Institute CNR, Naples, Italy

Abstract Air pollution in large urban areas is still a sanitary emergency even though significant improvements have been reached in fuel quality and vehicular emissions. Many air pollutants when inhaled are toxic to humans and can cause threats to the respiratory system and chest congestion and other health problems. The correlation between the concentration of atmospheric pollutants and health impact is not straightforward due to the complexity of atmospheric mixtures and the antagonistic or synergic effects of pollutants. At the same time public authorities need tools for the evaluation of their policies for the mitigation of air pollution. A possible approach to both these problems is the evaluation of historical trends of air pollution in urban areas to be compared with sanitary data or for the evaluation of the effectiveness of environmental policies. In this paper a procedure to evaluate historical trends of air pollution has been developed using air pollution indexes. Data collected by the air monitoring quality network in the urban area of Naples from 2001 to 2007 have been analysed to evaluate three different daily air quality indexes. The different indexes were compared to check the correlation among them. Some correlation exists among the indexes considered but the distribution in the defined risk classes was different. Therefore, the three indexes considered cannot be assumed as equivalent. Only correlation with sanitary data can help to select which index is the most apt. A statistical analysis was then developed to obtain an historical trend of air pollution in Naples. Even though there were some differences, all indexes show that from 2001 to 2007 the air quality in Naples has, on average, improved. Keywords: atmospheric pollution, index, urban area, trend, statistics. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110271

288 Air Pollution XIX

1 Introduction To characterize the quality of the air at a given location, government agencies elaborate data from air quality monitoring stations to calculate numerical indexes named air pollution indexes (API) or air quality indexes (AQI). The aim of each index is to synthesise local atmospheric pollution, due to concentration of several pollutants expressed with different units, in a single number. In this way risks associated with local atmospheric pollution and behaviours to be followed to minimize the impact on human health can be effectively communicated to population. Several procedures are reported in literature to calculate these indexes. In the USA the Air Quality Index (AQI) of Environmental Protection Agency (EPA) (EPA [1]) is the standard. In Europe there is not a single index adopted by all the countries, but the single nations or the local authorities adopt their own index. Several indexes are proposed and reported in literature: Albergamo et al. [2]; Cheng et al. [3]; Cogliani [4]; Guzzi [5]; Kassomenos et al. [6]; Murena [7]. In this paper three different indexes have been considered: CAQI (Common air quality index); APHI (Air Pollution Health Index) and AQHI (Air Quality Health Index). All are daily indexes. It means that their value would be representative of the atmospheric pollution in a solar day. The Common Air Quality Index (CAQI) was developed in the CiteAir project (http://www.citeair.eu). It belongs to the category of Air Pollution Indexes and is based on limit values of the European legislation EC [8–10]. In the same category belong, with some peculiarities, other indexes like: Air quality Index (AQI) of the US Environmental Protection Agency (www.epa.gov) [10]; ATMO adopted in France (http://www.airparif.asso fr) and UK-AQI adopted in UK (http://www.airquality.co.uk). The CAQI is a daily index with values in the range 0-100 subdivided in 5 classes. CAQI procedure defines two different categories of stations: traffic and city background. For the first category mandatory pollutants are NO2 and PM10, for the second one O3 is added. For this study the two categories have been merged. Table 1:

Pollutants and calculation grid adopted for calculation of CAQI index. All concentrations are in g/m3. Daily representative values of concentration are: NO2 and O3 maximum hourly value; CO 8 maximum hours moving average and PM10 daily average.

0-50 51-100 101-200

PM10 1 hour 0-25 26-50 51-90

PM10 24 hours 0-12 13-25 26-50

76-100

201-400

91-180

51-100

> 100

> 400

> 180

> 100

Class of risk Very Low Medium

Index value 0 -25 26-50 51-75

High Very

NO2

O3

CO

0-60 61-120 121180 181240 > 240

0-5000 5001-7500 750110000 1000120000 >20000

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

289

The calculation grid reported in Table 1 has been adopted to evaluate the CAQI index. Values of CAQI index are calculated for each pollutant by linear interpolation from the calculation grid. The daily index of the station is the maximum of the daily indexes calculated for the air pollutants. The Air Pollution Health Index (APHI) was proposed by Cairncross et al. [11] and developed starting from an analysis of the mortality risk associated at each atmospheric pollutant. The index is calculated by summing the contribution of each pollutant by the formula:

APHI   ai Ci

(1)

where ai is a coefficient and Ci is a value “representative” of the concentration of the pollutant “i” in the 24 hours considered. Values of ai and Ci to be adopted are reported in Table 2. The two values of PM and ozone that can be calculated from Table 2 are in alternative: only the highest must be considered in Eq. (1). Differently from CAQI, APHI has not a ceiling value. Table 2:

Daily representative values of concentration (Ci) and coefficients (ai) for the calculation of APHI (representative values are all in g/m3 but CO in mg/m3).

Pollutant

Ci

PM10 PM2.5 SO2 O3 O3 NO2 CO

Daily average Daily average Daily average max of 8 h mobile average max of hourly average max of hourly average max of 8 h mobile average

ai 0.048 0.10 0.026 0.033 0.03 0.02 0.25

The classes of risks are four as reported in Table 3. The Air Quality Health Index, was developed by Stieb et al. [12], and is based on an analysis of mortality data collected in Canada. Table 3: Class of risk Low Medium High Very high

Classes of risk of APHI and AQHI. APHI 1-3 4-6 7-9 ≥10

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

AQHI 0-3 4-6 7-10 ≥10

290 Air Pollution XIX It is calculated by the following formula: 10

AQHIPM10

11.7

100 e0.000871·NO2 ‐1 e0.000537·O3 ‐1 e0.000297·PM10 ‐1

(2)

if data on PM10 are available, or by the formula: 10

AQHIPM2.5

10.4

100 e0.000871·NO2 ‐1 e0.000537·O3 ‐1 e0.000487·PM2.5 ‐1

(3)

if data on PM2.5 are available. Concentrations are in ppb for NO2 and O3 and in g/m3 for PM10 and PM2.5. Daily representative values of concentration are: the maximum 3 h mobile average for NO2 and O3 and the 24 h average for PM. AQHI, as APHI, does not have a ceiling value and four classes of risk (Table 3). Both APHI and AQHI deriving from sanitary data belong to the category of “health indexes”.

2 Methodology The three pollution indexes considered have been adopted to evaluate air pollution in Naples from 2001 to 2007. The air quality network in the urban area of Naples is composed of nine fixed monitoring stations. The location of fixed stations is reported in Fig. 1, while a scheme of pollutants monitored by each station is reported in Table 4.

8 1

6 2 4

7

9

3

5

Figure 1:

Map of Naples with locations of air quality monitoring stations.

The location criteria and the techniques adopted to monitor air pollutants respect the European directives [9, 10]. The monitoring efficiency was generally high: in many cases above 90%. The modifications occurred from 2001 to 2007 about the monitoring facilities of the air quality network as reported in Table 4.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Table 4:

291

Analytical facilities of air quality network of Naples.

NO2 PTS PM10 PM2.5 # CO 1 Oss Astronomico X X X1 X1 1 2 Santobono X X X4 3 Policlinico X1 X X5 2 4 Fuorigrotta X X X 5 Vanvitelli X X X X4 3 6 Cavour X X 7 Ferrovia X X X 8 Nuovo Pellegrini X 9 Via Argine X X1 1 2 working in 03-07; station working only in 02-03; 3station 06-07; 4working 03-05 and 07; 5working 02-05 and 07.

O3 X

SO2 X X5 X5

X X X1 working only

For each of the three indexes considered the procedure to evaluate the daily pollution index in a station is fixed. But in this case we want to evaluate a daily pollution index for a whole urban area not for a single station. Therefore, a specific procedure was defined: i) daily representative values were calculated and validated for each pollutant and each station; ii) daily representative value for each pollutant was calculated and validated for the whole urban area; iii) daily indexes for the whole urban area were calculated and validated on the basis of daily representative values of each pollutant calculated at point ii). The daily representative values, required by each calculation procedure, are validated if at least 75% of data mandatory for their evaluation were present. Table 5 shows in detail the validation criteria adopted. Table 5:

Validation criteria for calculation of representative values.

Representative values Daily maximum of hourly average

Minimum number of data for validation

Daily average Daily maximum of 8 h mobile average

18 hourly average 18 values of 8 h mobile average (6 hourly average to validate each 8 hour mobile average)

Daily maximum of 3 h mobile average

18 values of 3 h mobile average (2 hourly average to validate each 3 h mobile average)

18 hourly average

Once representative values were calculated for each pollutant and for each station then a representative value of the urban area was calculated for each pollutant as: average or maximum of representative values of all stations. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

292 Air Pollution XIX In this paper only results corresponding to the first case are reported. Therefore the representative value of concentration of each pollutant for the whole urban area was calculated as average value of the representative values calculated for all stations. The representative value for the whole urban area was assumed valid even in the case of a single representative value validated for all the stations. Once obtained the representative values of concentration for the whole urban area the calculation of pollution index is straightforward: the procedure defined for each index, and reported before, has to be followed. Urban area daily pollution indexes have been validated only if representative concentration for all mandatory pollutants were validated for the urban area. To conform the different procedures we have assumed as mandatory pollutants: NO2, PM10 and O3. We have not considered data of PM2.5 because it was measured only at #5 station and not in all the years (Table 4).

3 Results Comparison of results of the tree indexes selected is not straightforward, essentially because they are different for the scale adopted and for a presence of a ceiling value (CAQI) or its absence (APHI and AQHI). For CAQI if the value of 100 is exceeded the notation > 100 is adopted. For this reason a first comparison was tested evaluating the frequency of classes of risk for each index reported in Tables 1 and 3. To conform the three indexes the 5 classes of risk of CAQI were reduced to 4 putting together the first two classes associated with the minimum risk. Results are reported in Fig. 2.

80 70 60 50 40 30 20 10

AQHI APHI

0 Low

Figure 2:

Medium

CAQI High

Very high

Frequency distribution in risk classes of the air pollution indexes from 2001 to 2007.

There is a significant difference among the indexes. CAQI is characterised by a high frequency of days with a low (24%) and medium (58%) class of risk, while the most frequent class for APHI and AQHI is the “high” risk class 75% WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

293

and 65% respectively. “Very high” risk class occurs rarely with a frequency of 0.7% for CAQI, 7.7% for APHI and 3.6% for AQHI. Correlation between the indexes was then studied. As an example results obtained for 2007 are reported in Fig. 3. 2007 10 2

R = 0.4366

APHI

08 06 0.4 02 00 00

0.2

0.4

06

08

10

06

08

10

06

0.8

CAQI 10 2

R = 0 5871

AQHI

08 06 0.4 02 00 00

0.2

0.4 CAQI

10 2

R = 0 8422

AQHI

08 06 0.4 02 00 0.0

0.2

0.4

1.0

APHI

Figure 3:

Correlations between indexes for 2007. Values relative to maximum are reported for each index.

The highest correlation was observed between APHI and AQHI (average value R2 = 0.81). The lowest between CAQI and APHI (average value R2 = 0.40). Determination coefficients obtained are reported in Table 6. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

294 Air Pollution XIX Table 6:

Correlation analysis: determination coefficient (R2) between indexes.

2001 2002 2003 2004 2005 2006 2007 Average

CAQI / APHI 0.61 0.63 0.42 0.44 0.10 0.16 0.44 0.40

CAQI / AQHI 0.46 0.67 0.32 0.61 0.46 0.50 0.59 0.52

APHI / AQHI 0.89 0.87 0.78 0.81 0.77 0.74 0.84 0.81

A statistical multiple comparison procedure of average yearly index was carried out to analyse the historical trend. On the basis of daily urban index a yearly average urban index was calculated and the trend from 2001 to 2007 of the yearly average index was studied. In Fig. 4 box and whisker plots are reported for the tree indexes showing the trend of yearly average values. Result of multiple range test are summarised in Table 7. X’s in the same column identify years with means of the air quality index that have not statistically significant differences at the 95.0% confidence level. Table 7:

Results of multiple range test. APHI

CAQI 2001 2002 2003 2004 2005 2006 2007

X X X

X

X X

X X

X X

X X X X X

AQHI X

X X X

X X X X

Also in this case the three index show different results. Some common results can be, however, highlighted. The yearly average of 2001 has a statistically significant difference with average of other years in all the three cases (it is the most polluted year). For all the three indexes 2005, 2006 and 2007 are the three less polluted years. More in detail the less polluted years are: 2005, 2006 and 2007 for CAQI (the difference is not statistically significant); 2007 for APHI and 2005 for AQHI. 2003 is the second highest polluted year for all the three indexes (together with 2004 for APHI or 2002 for AQHI). 2002 is the second (APHI and AQHI) or the third (CAQI) highest polluted year. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

295

Therefore, it can be concluded that apart from the differences in the pollution indexes considered a decreasing trend for air pollution in Naples from 2001 to 2007 can be observed. 2001 2002 2003 2004 2005 2006 2007 0

20

40

60

80

100

120

APH1 2001 APHI 2002 APHI 2003 APHI 2004 APHI 2005 APHI 2006 APHI 2007 0

4

8

12

16

AQHI 2001 AQHI 2002 AQHI 2003 AQHI 2004 AQHI 2005 AQHI 2006 AQHI 2007 0

Figure 4:

3

6

9

12

15

Box and whisker plot for the three indexes showing the trend from 2001 to 2007.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

296 Air Pollution XIX

4 Conclusions The procedure developed allows the evaluation of historical trend of air pollution in urban areas. This goal was obtained using air pollution indexes. The final result depends on the pollution index adopted. In the case of Naples data from 2001 to 2007 show a certain decrease of air pollution. APHI and AQHI indexes are more similar and correlated. CAQI gives more different results. The final decision on the most apt index to be adopted can be taken only after that a analysis of correlation with sanitary data has been carried out. In such analysis it is necessary to take in count that any sanitary data are not exclusively determined by the air pollution levels, but also by many other social, demographical and infrastructural factors.

References [1] Environmental Protection Agency, 1999. Guideline for reporting of daily air quality – air quality index (AQI). EPA-454/R-99-010. Office of Air Quality Planning and Standards, Research Triangle Park, NC 27711. [2] Albergamo, C., Cagnetti, P., Mammarella, M.C., Tondo, A., 1996. Indice giornaliero della qualità dell’aria:IGQA. ENEA Casaccia Italy. [3] Cheng, W., Kuo,Y., Lin,P., Chang, K., Chen, Y., Lin, T., Huang, R., 2004. Revised air quality index derived from an entropy function. Atmospheric Environment 38, 383-391. [4] Cogliani, E., 2001. Air pollution forecast in cities by an air pollution index highly correlated with meteorological variables. Atmospheric Environment 35, 2871-2877. [5] Guzzi, D., 2004. Inquinamento atmosferico in aree urbane: Indici di qualità dell’aria. Chemical Engineering Thesis University of Naples “Federico II” Italy. [6] Kassomenos, P., Skouloudis, A.N., Lykoudis, S., Flocas, H.A., 1999. “Airquality indicators” for uniform indexing of atmospheric pollution over large metropolitan areas. Atmospheric Environment 33, 1861-1879. [7] Murena, F. Measuring air quality over large urban areas: development and application of an air pollution index at the urban area of Naples. Atmospheric Environment 2004, 38, 6195-6202. [8] European Community, 1999. Council Directive 1999/30/EC of 22 April 1999 relating to limit values for sulphur dioxide, nitrogen dioxide and oxides of nitrogen, particulate matter and lead in ambient air. Official Journal L 163, 29/06/1999 P. 41-60. [9] European Community, 2000. Directive 2000/69/EC of the European Parliament and of the Council of 16 November 2000 relating to limit values for benzene and carbon monoxide in ambient air. Official Journal L 313, 13/12/2000 P. 0012 – 0021. [10] European Community, 2002. Directive 2002/3/EC of the European Parliament and of the Council of 12 February 2002 relating to ozone in ambient air. Official Journal L 67, 9/3/2002. P. 14-30. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

297

[11] Cairncross, E. K., John, J., Zunckel M., 2007. A novel air pollution index based on the relative risk of daily mortality associated with short-term exposure to common air pollutants. Atmospheric Environment 41, 84428454. [12] Stieb D.M., Burnett R. T., Smith-Doiron, M., Brion, O., Shin H.H., and Economou V. 2008. “A New Multipollutant, No-Threshold Air Quality Health Index Based on Short-Term Associations Observed in Daily TimeSeries Analyses”. Journal of the Air & Waste Management Association 58: 435-450.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

This page intentionally left blank

Section 4 Aerosols and particles

This page intentionally left blank

Air Pollution XIX

301

Correlation between the mass of PM2,5 and the chemical composition of acid aerosols in the northwest of the metropolitan zone of Mexico City Y. I. Falcón, E. Martinez & L. Cortes Departamento de Energía, Universidad Autónoma Metropolitana-Azcapotzalco, Mexico

Abstract Mexico City has been considered one of the most polluted areas worldwide. Since it has a population of over 20 million, some 3 million vehicles and over 4000 industries, ambient air monitoring and respiratory and cardiovascular diseases surveillance is extremely important in order to evaluate particles’ effects on human health. Acid aerosol sampling was performed with an annular denuder system and particle sampling was carried out with a low volume sampler. Sampling was 24 hours from Monday to Tuesday and Thursday to Friday. Particle and acid aerosol sampling was simultaneous. Both samplers were located in the air monitoring cabin at the Metropolitan Autonomous University, Azcapotzalco campus. The monitoring campaign lasted from January to July, 2004. During January and February sulfur dioxide concentrations were higher compared to sulfates concentrations but in June and July, sulfates concentrations were higher than concentrations of sulfur dioxide. In January and February, nitrates concentrations were the highest and nitric acid concentrations were the lowest during the whole campaign. In the correlation between Chemical Species / Mass: Nitrates / Mass; Sulfates / Mass, the highest rate was for Sulfates / Mass. Regarding mass, there were two periods of uniform concentration. In the first one, from January to March, mass concentrations were higher than they were in the period April to July. During the June–July period, meteorological conditions (high relative humidity and low wind speed) favored sulfates formation while in January and February relative humidity favored nitrates formation. Results of this correlation show that the dominant chemical species in collected particles WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110281

302 Air Pollution XIX were sulfates. Registered low wind speed from January to March (an average of 7.8 kph) did not favor particle dispersion so the highest mass concentrations were collected during this period. Keywords: PM2.5, acid aerosols.

1 Introduction Mexico City is located at 19°03' North latitude and 99°22' West longitude and at 2200 m above sea level [1]. With over 20 million inhabitants, some 3 million vehicles and over 4000 industrial facilities, it is considered one of the most polluted cities in the world. In order to evaluate the effect of respirable particles on health it is essential to monitor air pollutants concentrations and to carry out a survey on respiratory and cardiovascular diseases [2]. An epidemiologic study was performed in the southwest area of Mexico City during the 1993–1995 period, in which fine particle concentrations were correlated with mortality. It was detected that an increase of 10 g/m3 in PM2,5 concentrations increased mortality by 1.4% [3]. Emissions of pollutants generated by vehicles deteriorate air quality and that is one of the main reasons, at a worldwide level, to improve fuel quality. In Mexico, it is considered to have reductions up to an 88% in the sulfur content of the Pemex Premium gasoline, and between 84 and 93% of the Pemex Magna gasoline which are distributed in the Metropolitan Areas (Valley of Mexico, Guadalajara and Monterrey), and 92 to 96% in the remainder of the country. The sulfur reduction in the diesel fuel will be up to 98.5%. Presently, Pemex Premium gasoline has a sulfur content of between 250 and 300 parts per million (ppm) although the Mexican Official Standard establishes an average of 30 ppm and a maximum of 80 ppm. Diesel fuel must have one of the most important reductions, from 500 to only 15 ppm [4], because it is the main particles emitter in the metropolitan areas. Recent studies indicate that the ionic species (SO4-2, NO-3, NH+4) contribute to particle formation in many areas. They also indicate that SO4-2, NO-3, NH+4 are the main water soluble ionic species in the PM2 5. These three species constitute over 30% of the PM2 5 mass in Hong Kong, and in some Korean and Swiss cities [5]. The present study was performed within the UAM-Azcapotzalco facilities, which are located in Northwest Mexico City in a neighborhood with a mixed (industrial and residential) land use.

2 Methodology Acid aerosols (gases and particles) sampling was performed with a denuder system formed by a selective filters head (Teflon and nylon) for the PM2,5 particles collection and diffusion separator tubes for the gaseous molecules collection. Gas molecules rapidly diffuse towards the separator tube walls while

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

303

the fine particles are not affected in their travel through the separator tube and are finally captured on the filters [6]. Respirable PM2,5 particles were collected with a portable low volume sampler, operated with a rechargeable battery. Particles separation was performed using a head with two impactors (PM10 and PM2,5) where particles were separated by size and collected on a 47 mm glass fiber filter with a 2 µm porosity. Particles mass was gravimetrically determined. Both sampling devices (the portable low volume sampler and the denuder) were collocated at the sampling site for 24-hour periods.

3 Results

0

1,80

35

1,60 1, 0

Concentration (g/m3)

30

1,20 25 1,00 20 0,80 15 0,60 10

0, 0

08-09/jul/200

05-06/jul/200

01-02/jul/200

28-29/jun/200

2 -25/jun/200

21-22/jun/200

17-18/jun/200

1 -15/jun/200

10-11/jun/200

07-08/jun/200

03-0 /jun/200

31-01/jun/200

27-28/may/200

2 -25/may/200

01-02/apr/200

20-21/may/200

29-30/mar/200

17-18/may/200

25-26/mar/200

22-23/mar/200

18-19/mar/200

15-16/mar/200

11-12/mar/200

26-27/feb/200

08-09/mar/200

23-2 /feb/200

01-02/mar/200

19-20/feb/200

16-17/feb/200

12-13/feb/200

29-30/jan/200

09-10/feb/200

26-27/jan/2007

02-03/feb/200

0,00 22-23/jan/200

0,20

0 19-20/jan/ 200

5

Solar Radiation UVA (W/cm2)

The monitoring campaign included the 2004 January–July period. Data on relative humidity and solar radiation were also collected since weather conditions directly affect acid aerosol formation. Wind speed information was also gathered because it is related to pollutant dispersion. Gaseous species in the acid aerosols were sulfur dioxide and nitric acid, whether sulfates and nitrates are present as particles. Correlations were performed between chemical species collected with the denuder (particle fraction) and the aerosol total mass collected with the low-vol sampler.

Sampling Date Sulfates

Figure 1:

Sulfur Dioxide

Solar Radiation

Ultraviolet radiation (UVR) effect on the formation of SO4-2 from SO2

4 Discussions and conclusions 4.1 Sulfates and sulfur dioxide During January and February, sulfur dioxide concentrations were higher than sulfate concentrations. ZMCM meteorological conditions favored pollutants dispersion. Wind speed in February was higher than it was on January. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

304 Air Pollution XIX In Figure 1 it can be seen than sulfate concentrations were not detected during February. In March, meteorological conditions followed the previous month’s trend and sulfur dioxide concentrations were higher than sulfates concentrations. Solar radiation during March was higher than that registered during January and February. This favored sulfates formation so its concentration could be detected. During April and May solar radiation was higher than it was during the previous months and sulfate concentrations increased perceptibly. Some days they were even higher than sulfur dioxide concentrations which can be an indication of photochemical activity during these months. Sulfate concentrations during June and July were higher than sulfur dioxide concentrations. This seems to indicate that high relative humidity and low wind speed favored sulfates formation (Figure 1). 4.2 Nitrates and nitric acid During January and February the registered average relative humidity was the lowest of the whole campaign (39,5%) which favored nitrates formation. As a consequence, nitrates concentrations were higher than nitric acid concentrations. Actually, nitrates concentrations were the highest and nitric acid, the lowest, of the whole campaign. During the remainder of the campaign, relative humidity didn’t show a direct influence on the nitric acid formation (Figure 2). 100 90

4

80

3,5

70

3

60

2,5

50

2

40

1,5

30

1

20

0,5

10

08-09/jul/200

05-06/jul/200

01-02/jul/200

28-29/jun/200

2 -25/jun/200

21-22/jun/200

17-18/jun/200

1 -15/jun/200

10-11/jun/200

07-08/jun/200

03-0 /jun/200

31-01/jun/200

27-28/may/200

2 -25/may/200

01-02/apr/200

20-21/may/200

29-30/mar/200

17-18/may/200

25-26/mar/200

22-23/mar/200

18-19/mar/200

15-16/mar/200

11-12/mar/200

26-27/feb/200

08-09/mar/200

23-2 /feb/200

01-02/mar/200

19-20/feb/200

16-17/feb/200

12-13/feb/200

29-30/jan/200

09-10/feb/200

26-27/jan/2007

02-03/feb/200

22-23/jan/200

0 19-20/jan/ 200

0

R H (%)

Concentration (g/m3)

5 4,5

Sampling Date Nitric Ac d

Figure 2:

Nitrates

RH

Relative humidity (RH) effect on the formation of HNO3 from NO3-

4.3 Chemical species / mass correlation When chemical species / mass correlation results are compared it is clearly seen that the nitrates / mass correlations values are the highest, so the predominant chemical species in the collected particles is sulphate (Figure 3). WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

305

0,25

1,40

1,20 0,20

0,15

0,80

0,60

0,10

Nitrates/Mass

Sulfates/Mass

1,00

0,40 0,05 0,20

Figure 3:

05-06/jul/200

01-02/jul/200

28-29/jun/200

2 -25/jun/200

21-22/jun/200

17-18/jun/200

1 -15/jun/200

10-11/jun/200

07-08/jun/200

31-01/jun/200

27-28/may/200

2 -25/may/200

01-02/apr/200

20-21/may/200

29-30/mar/200

Sampling Date Su fates/Mass

17-18/may/200

25-26/mar/200

22-23/mar/200

18-19/mar/200

15-16/mar/200

11-12/mar/200

26-27/feb/200

08-09/mar/200

01-02/mar/200

23-2 /feb/200

19-20/feb/200

16-17/feb/200

12-13/feb/200

29-30/jan/200

09-10/feb/200

02-03/feb/200

0,00 19-20/jan/ 200

0,00

Nitrates/Mass

Correlation (chemical species/mass).

Sulfur content in fuels which are used in Mexico is very high so flue gases contain a high percentage of sulfur dioxide. This can be seen in the obtained results. 4.4 Mass For mass, there were two uniform concentration periods. From January to March, collected mass was higher, which can be due to the average wind speed, of 7,8 kph. During the next period, from April to June, mass concentration decreased. Since the average wind speed was 8,6 kph it was deduced that wind speed directly affects particles dispersion which can be observed in PM2,5 particles mass during the whole monitoring campaign (Figure 4). 100,00

16

90,00

14 12

70,00 10

60,00

8

50,00 40,00

6

Wind speed (kph)

RH (%) & PM2 5 (g/m3)

80,00

30,00 4 20,00 2

10,00

08-09/jul/200

05-06/jul/200

01-02/jul/200

28-29/jun/200

2 -25/jun/200

21-22/jun/200

17-18/jun/200

1 -15/jun/200

10-11/jun/200

07-08/jun/200

31-01/jun/200

27-28/may/200

2 -25/may/200

20-21/may/200

01-02/apr/200

17-18/may/200

29-30 mar/200

25-26 mar/200

22-23 mar/200

18-19 mar/200

15-16 mar/200

11-12 mar/200

08-09 mar/200

01-02 mar/200

26-27/feb/200

23-2 /feb/200

19-20/feb/200

16-17/feb/200

12-13/feb/200

09-10/feb/200

02-03/feb/200

29-30/jan/200

0 19-20/jan/ 200

0,00

Sampling Date Mass

Figure 4:

RH

Wind speed

Relative humidity (RH) and wind speed effect on the PM2,5 mass concentration.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

306 Air Pollution XIX The 65 µg/m3 value (24 hrs average) that has been considered in the PM2,5 Mexican Standard project was never exceeded during the whole monitoring campaign.

References [1] http://www.inegi.gob.mx/geo/default.asp?c=124&e=09 [2] Contaminación Atmosférica y sus Consecuencias en la Salud, Servicio de Salud del Ambiente Región Metropolitana, SESMA, Chile 2001 [3] Borja-Aburto Victor H, Castillejos Margarita, et al, 1998, Mortality and Ambient Fine Particles in Southwest Mexico City, 1993-1995, Environmental Health Perspectives, 106-12, 849-855 [4] http://www.ref.pemex.com/octanaje/o66/o.htm [5] Yi-Chyun Hsu, Mei-Hsiu Lai, et al, 2008, Characteristics of Water-Soluble Ionic Species in Fine (PM2 5) and Coarse Particulate Matter (PM10-2 5) in Kaohsiung, Southern Taiwan, Journal of the Air & Waste Management Association, 1579-1589 [6] Denuder Tutorial, Air Pollution Training Institute, Environmental Protection Agency, U.S.A.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

307

Characteristics of aerosol particle size distributions in urban Lanzhou, north-western China Y. Yu1, S. P. Zhao1,2, D. S. Xia1, J. J. He1,2, N. Liu1,2 & J. B. Chen1 1

Key Laboratory of Land Surface Process and Climate Change in Cold and Arid Regions, Cold & Arid Regions Environmental & Engineering Research Institute, CAS, Lanzhou, Gansu, China 2 Graduate School of the Chinese Academy of Sciences, Beijing, China

Abstract Continuous particle size data (0.5–20μm) were collected using aerodynamic particle sizer (TSI 3321) at an urban site in Lanzhou, north western China from 1st August to 31st October. Variations of particle concentrations and general characteristics of particle size distributions were analyzed. The hourly averaged particle number, surface area and volume concentration are 108.1±92.2cm-3,282.9±267.9μm2cm-3 and 92.2±127.3μg3m-3, respectively. Fine particles (0.5–2.5μm) accounted for 98.7%, 73.8% and 37.5% of the total particle number, surface area and volume concentrations in 0.5–20μm, respectively. The size distribution of number concentrations is unimodal with a maximum at 0.54–0.58μm, while that of the surface area and volume concentrations are bimodal. The main peak of surface area concentration appears near 0.63–0.67μm with a secondary peak at 3.79–4.07μm, and the main peak of volume concentration is at 4.7–5.1μm with a secondary peak near 0.67–0.72μm. K-means cluster analysis was used to group the particle volume size distributions into 7 clusters by their dominant mode and average concentration. Particle volume size distributions observed during dust storms and under dry and clear weather conditions were characterized by a single coarse mode, while particle volume size distributions affected by fog, smog and traffic related emissions were bimodal with peaks at accumulation mode and coarse mode, respectively. Keywords: air pollution, size distribution, aerosols particles, volume concentration, meteorological condition. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110291

308 Air Pollution XIX

1 Introduction Atmospheric aerosol particles play an important role in climate change, [1]. They also have a determining effect on visibility, [2] and human health, [3]. Several studies have indicated the dependence of the direct and indirect climate effects, the health effect and environmental effect of the atmospheric aerosols on particle properties, such as size distributions, [4, 5]. Knowledge of the properties and the factors affecting particle size distributions is thus important not only for understanding aerosol formation mechanisms and improving the parameterisation in climate models, but also for identifying the sources. In recent years, many studies have been done on urban particle size distributions. However, most of the studies were carried out in the European and North American cities, [6, 7]. In China, most of the studies were conducted in central eastern or coastal areas, [8 –10]. Studies in north western China were mainly concentrated on particulate mass, with little or no information on size distributions, [13], especially for urban areas. Lanzhou, the capital of Gansu Province, has an urban population of about 2.6 million and a total population of 3.6 million in 2010, [11] and is one of the most polluted cities in China, [12]. Like many cities in northern China, particulate matter is one of the most formidable air quality and public health issues in Lanzhou. Actions toward reducing emissions of aerosol particles have being put into effect, but the levels of particulate pollution remain high. Yu [13] showed that annual mean mass concentration of PM10 in urban Lanzhou has decreased from 236μgm-3 in 2001 to 127μgm-3 in 2007, but still exceed the national Grade II standard for annual mean PM10 concentration (100μgm-3) by 27%. There have been several studies about particulate pollution in Lanzhou: Wei et al. [14] analyzed the PM2 5 and PM10 in four Chinese cities including Lanzhou, with emphasis on the elements’ size distributions; Wang et al. [15] reported the mass concentrations of PM2 5 and PM10 and their annual and diurnal variations; Wang et al. [16] reported particle mass size distributions for PM10 under different weather conditions, but with rather coarse resolution. Previous measurement of atmospheric particle size distributions in Lanzhou urban area have been scarce, and generally been conducted only for a short period, until now the knowledge of the aerosol particle size distributions in Lanzhou urban area is very limited. In order to better understand the particulate pollution in Lanzhou and establish effective control strategies, continuous measurements of particle size distributions in the range of 0.5–20μm are reported in this paper, with the intention of understanding the characteristics of atmospheric particle size distributions and their affecting factors.

2 Method 2.1 Sampling site Lanzhou (36.1oN, 103.8oE) is situated at the intersection of Qinghai–Tibet Plateau, the Inner Mongolian Plateau and the Loess Plateau. It is surrounded by WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

309

mountains and hills that rise to 500–600m. Fig. 1 shows the topography of Lanzhou and the location of the sampling site. The sampling site was on the roof of an academic building of the Cold & Arid Regions Environmental & Engineering Institute (CAREEI), located in the eastern part of Lanzhou urban area. The inlet of sampling system was about 32 m above the ground level. The site is a primarily residential and commercial area without obvious industrial sources. There are two major roads with heavy traffic, i.e. Donggang West Road and Tianshui road, at the south and the west of the sampling site, with a distance of about 40 m and 300 m from the sampling site, respectively.

Figure 1:

The topographic map of Lanzhou and location of the sampling site.

2.2 Measurement Aerosol size distribution between 0.5 and 20μm in particle aerodynamic diameter was continuously measured during 1st August and 31 October using an aerodynamic particle sizer (APS, TSI model 3321). The flow rate of APS is 1L min-1 with a sheath air flow rate of 4Lmin-1. The time resolution is 5min per scan. Flow rate checks were made once a week with a bubble flow meter. Meteorological data were obtained from an automatic meteorological station co-located with the sampling site. Weather observations from a weather station 1.7km from the site were also used. The observation period covered the end of summer and the autumn of Lanzhou. East and northeast wind prevailed during the observation period. 2.3 Data processing Hourly average size distributions were calculated from the original data. To be included in the subsequent analysis, each average value must consist of at least half hour’s data. After data screening, a total of 2082 hourly particle size distributions were left. Cluster analysis was used to reduce the number of hourly volume size distributions into several groups with similar characteristics. The Kmeans clustering routine available in MATLAB© was used in this study. The Kmeans clustering routine split the existing multidimensional data into predefined WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

310 Air Pollution XIX number of subgroups, i.e. clusters, which are as different as possible from each other, but as homogeneous as possible within themselves, by iteratively minimizing the sum of squared Euclidean distances from each member to its cluster centroid. The K-means clustering method has been used in various studies and been justified as a preferred technique for particle size distribution data analysis, [17]. In this study we limited the analysis to include 7 clusters. Different number of clusters was tested and it is found that increasing the number of clusters above seven did not provide more information, while decreasing the number of clusters below seven lost information on some well characterized size distributions.

3 Results and discussion The aerosol particle size distributions were integrated to calculate particle number, surface area and volume concentrations in different size ranges. In this study, particles in the diameter from 0.5 to 20μm were divided into three subranges: 0.5-1.0μm, 1.0-2.5μm, and 2.5-10μm. The total particle concentrations refer to particle concentrations within 0.5-20μm. The particle surface area and volume concentrations were calculated from the measured number size distributions with an assumption of spherical particles. 3.1 Particle concentrations Table 1 summarizes the statistics for hourly averaged particle number, surface area, and volume concentrations for different size ranges covering the observation period. The total particle number, surface area, and volume concentrations are 108.1±92.2cm-3, 282.9±267.9μm2cm-3, and 92.2±127.3μm3cm-3, respectively. The total number and surface area concentrations are dominated by fine particles (0.5-2.5μm), which account for 98.7% and 73.8% of the corresponding total concentrations. 62.5% of the total volume concentration was contributed by coarse particles (2.5-20μm). The particle number concentrations in size ranges 0.5-1.0μm, 1.0-2.5μm, and 2.510μm are 98.1cm-3, 8.5cm-3, and 1.5cm-3, respectively, and 90.8% of the particle numbers are in 0.5-1.0μm. When aerosol particle concentrations in this study are compared with other studies in Chinese urban areas, it is noticed that the average number concentration in size range 0.5-1.0μm in Lanzhou is close to the values observed in spring of Guangzhou, [18] and in summer of Ji’nan urban areas, [19], while the number concentration in size range 2.5-20μm is higher than that observed in Guangzhou and lower than that observed in winter of Beijing (3.4 cm-1), [20] and in winter and summer of Shenyang (4.4 cm-1), [21] urban areas. In addition to differences in local sources, the influence of surrounding arid and semi-arid areas may lead to a background with relatively higher coarse particles in Lanzhou.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Table 1:

311

Statistics of hourly averaged particle concentration for different size ranges during 1st August-31st October 2010. Size range (μm)

Mean±Std

Max.

Min.

0.5-1.0 1.0-2.5 2.5-10 0.5-20 0.5-1.0 1.0-2.5 2.5-10 0.5-20 0.5-1.0 1.0-2.5 2.5-10 0.5-20

98 1±86.6 8.5±10.1 1.5±2.7 108.1±92.2 148 9±135.8 59 9±77.0 73.4±134.1 282 9±267.9 18 2±16.9 16.4±22.1 56.0±101.9 92.2±127.3

586.8 97.2 32.2 627.4 907.0 786.7 1672.2 2621.5 111.3 230.2 1331.3 1646.8

4.0 0.1 <0.1 4.1 6.1 0.6 0.1 7.0 0.7 0.1 <0.1 1.0

Number concentration (/cm3) Surface area concentration (μm2 /cm3) Volume concentration (μm3 /cm3)

Percentage of the total concentration (%) 90.8 7.9 1.4 100.0 52.6 21.2 26.0 100.0 19.7 17.8 60.8 100.0

3.2 Average particle size distributions The average particle number, surface area and volume size distributions during the observation period are shown in fig. 2. The maximum of the number size distribution appear at 0.54-0.58μm. The surface area and volume size distributions contain two predominant modes at 0.63-0.67μm and 3.79-4.07μm, and at 4.70-5.05μm and 0.67-0.72μm, respectively. Fine particles with diameter below 2.5μm, accounting for 98.7% of the total particle numbers, were the main contributor to the total particle number concentrations, while coarse particles were the main contributor to the total volume concentrations. The coarse mode in our study (at 4.70-5.05μm) appeared at a larger particle size than that for urbaninfluenced aerosols (at 3-4μm) studied by Morawska et al. [22], indicating the possible effect of the arid and semi-arid background, [23]. 1600 Number Conc. Surface area Conc. Volume Conc.

,m2 cm 3,m3 cm 3 3

cm

(dN,dS,dV)/dlogDp

1400 1200 1000 800 600 400 200 0

1

10 Dp / m

Figure 2:

Size distribution of mean particle number, surface area and volume concentrations during the observation period.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

312 Air Pollution XIX 3.3 Clusters of volume size distributions The volume size distribution data were grouped into seven distinctly separated clusters, which are referred to as A to G and shown in fig. 3. The statistical description of the seven clusters and the corresponding meteorological conditions are summarized in table 2. 1600

200 B

A 1200

150

800

100

400

50

0

0

120

C

100

D

30

80 dV/dlogDp / m3 cm-3

40

60

20

40

10

20 0

0

350

E

300

250

F

200

250 200

150

150

100

100

50

50 0

0

500

G

1

10 Dp /m

400 300 200 100 0

1

10 Dp /m

Figure 3:

Average volume size distributions for clusters A to G.

Clusters A, B and G show a similar shaped coarse mode at 4.70-5.42μm, with clusters B and G having another small peak at the accumulation mode. Cluster A contains only 61 particle volume size distributions (2.9%). The total particle (0.5-20μm) volume, number and surface area concentrations of cluster A are the highest among the seven clusters, with coarse particles contributing 78% to its total volume concentration. Cluster A was exclusively observed in August and experienced the highest wind speed and temperature among the seven clusters. Lanzhou was affected by an exceptional dust storm during 12-13 August 2010 (dust storms generally occur during springtime, i.e. March, April WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

313

and March) and most of the size distributions in cluster A were observed during the dust storm. Thus cluster A represents the volume size distributions affect by dust storms. A total of 17.9% and 8.3% of the volume size distributions belong to clusters B and G, respectively. Fig. 4 shows the diurnal variation of the hourly ratio of the number counts of a cluster to the hourly total count of all the clusters. There is no preferred time in a day for the occurrence of cluster B. Cluster B was related to the driest meteorological conditions among the seven clusters and had relatively high wind speed and temperature. The total particle volume, number and surface area concentrations of cluster B are the lowest among the three clusters dominated by coarse mode. Cluster B represents the volume size distributions observed under dry, clear and good dispersion conditions. The hourly ratio of the number counts of cluster G (fig. 4) shows that the cluster occurred most frequently from midnight till noon, corresponding well to the diurnal variation of wind speed, i.e. low wind speed during night and morning hours and high wind speed in the afternoon. The total particle volume, number and surface area concentrations of cluster G is the second highest among the seven clusters. The above analysis indicate that cluster G represents the volume size distributions observed in general polluted conditions when wind speed was low and the atmosphere was stable. Table 2:

Volume, surface area and number concentrations associated with each cluster, and their corresponding meteorological conditions.

Clusters

A

B

C

D

E

F

G

Frequency of occurrence (%) Mean Volume (μm3/cm3) Mean surface area (μm2/cm3) Mean Number (cm-3) Temperature (oC)

2.9

17.9

21.5

40.1

2.8

6.4

8.3

714.6

97.2

61.8

30.1

196.6

113.7

220.5

1591.6

250.4

348.3

131.7

1013.6

639.8

522.0

1088

186.6

339.6

123.5

840.3

556.5

403.3

22.2

21.6

14.1

18.8

8.7

10.1

20.8

2.16

2.0

1.5

2.1

1.3

1.4

1.6

66.6

48.6

58.0

52.2

65.8

61.5

60.2

Wind speed (m/s) Relative humidity (%)

Cluster D is the most frequently observed cluster type, with 40.1% of size distributions belong to it. The cluster is characterized by two modes with comparable amplitude: an accumulation mode at 0.63-0.67μm and a coarse mode at 4.07-4.37μm. Cluster D had the lowest total volume concentration (30.1μg3cm-3) as well as the lowest total surface area and number concentrations. The average wind speed related to cluster D was the second highest among the seven clusters (table 2), indicating good dispersion condition. Cluster D also showed a strong dependency on the occurrence of rain, with 54.5% of size distributions been observed on days with precipitation. The hourly ratio of the number counts of cluster D (fig. 4, middle) has high values during the afternoon WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

314 Air Pollution XIX

70

Cluster A Cluster B Cluster G

Percentage (%)

60 50 40 30 20 10 0 Hour (Local Time)

70

Cluster D

Percentage (%)

60 50 40 30 20 10 Hour (Local Time)

70

Cluster C Cluster E Cluster F

Percentage (%)

60 50 40 30 20 10 0

0

4

8

12

16

20

23

Hour (Local Time)

Figure 4:

Hourly ratio of the number counts of a cluster (A-G) to the hourly 7

total count of all the clusters calculated as Ni/  N i given in %. i 1

and evening when the frequency of rain events is high and the wind speed is relatively high. Thus cluster D, which could be regarded as a clean background for the observation site, represents the volume size distributions observed under clean and good dispersion conditions. Clusters C, E and F have bimodal distributions, with a major mode at 0.630.67μm and a secondary mode at 3.79-4.07μm (clusters C and F) or 5.05-5.43μm (cluster E). The three clusters are dominated by fine particles, with fine particles accounting for 68%, 66% and 72% of the average total volume concentrations of WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

315

clusters C, E and F, respectively. Cluster C is second frequent cluster type, with 21.5% of the size distributions belong to it. Clusters E and F occurred much less frequently compared to cluster C, accounting for 2.8% and 6.4% of the total size distributions, respectively. Of the three clusters, cluster C has the lowest mean volume concentration (the second lowest among the seven clusters), while cluster E has the highest mean volume concentration. It is inferred from fig. 4 that particle volume size distributions within cluster C were observed most frequently in the early morning when calm conditions prevailed, i.e. local emissions tended to be trapped in the atmospheric boundary layer. There is also a bump of high values at around 15:00 LT, which matches with the time when wind mostly came from the south, southeast and southwest at the observation site, indicating the influence of traffic-related emissions. Thus cluster C represents the volume size distributions affected by local sources, mainly trafficrelated emissions (e.g. combustion and resuspended dust). A close inspection of the weather observation data at a weather station about 1.7 km east of the observation site reveal that the particle volume size distributions within clusters E and F were associated with the occurrence of fog and smog, respectively. Both clusters E and F were related to the lowest average wind speed and temperature and a relatively high relative humidity when compared to other clusters, indicating cold and stable weather conditions.

4 Conclusions Aerosol particle size distributions were measured from 1st August to 31st October 2010 at an urban site in Lanzhou, north-western China with APS. The measurements covered the size range from 0.5 to 20µm. The characteristics of the particle concentrations were investigated by dividing the particles into three sub-ranges: 0.5-1.0μm, 1.0-2.5μm, and 2.5-10μm. During the observation period, the total particle number, surface area and volume concentrations (0.5-20µm) were 108.1±92.2cm-3, 282.9±267.9μm2cm-3 and 92.2±127.3μg3m-3, respectively. The particle number size distributions are unimodal, while the corresponding surface area and volume concentrations are bimodal. The total number and surface area concentrations were dominated by fine particles (0.5-2.5μm), while coarse particles (2.5-20μm) were the main contributor to the total volume concentration. Compared with urban areas near coast, Lanzhou has higher coarse particles due to its arid and semiarid environment. K-means clustering analysis resulted in seven well separated clusters of the volume size distributions that were different in their dominant mode and average concentrations. The responsible processes for the different clusters were inferred according to the cluster characteristics and relationship to concurrently measured meteorological conditions. The meteorology was found to exert a high influence on the shape of particle volume size distributions. The cluster containing volume size distributions affected by dust storms had a single coarse mode and the highest volume concentration. Under low wind speed and stable conditions, e.g. early morning or night, the volume size distributions of the cluster were dominated by coarse mode and the volume concentration is the second highest. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

316 Air Pollution XIX The size distributions in the cleanest cluster were mostly observed on days with precipitation and had two modes, at 0.63-0.67μm and 4.07-4.37μm, respectively, with comparable amplitude. Bimodal clusters dominated by accumulation mode were generally observed on days with fog or smog. This study showed the size resolved characteristics of aerosol particles in an urban area, north-western China and the possible effect of meteorological conditions on volume size distributions. However, in order to characterize the size distribution and interactions between meteorology and aerosol physics, a full spectral covering nano-particle and a long-term measurement are needed.

Acknowledgements This research is funded by the Chinese Academy of Sciences through the ‘100 Talent Project’ and the National Natural Science Foundation of China (41071125).

References [1] [2]

[3] [4]

[5] [6] [7]

[8] [9]

Haywood, J. M. & Boucher, O., Estimates of the direct and indirect radiative forcing due to tropospheric aerosols: A review. Reviews of Geophysics, 38, pp. 513–543, 2000. See, S. W., Balasubramanian, R. & Wang, W., A study of the physical, chemical, and optical properties of ambient aerosol particles in Southeast Asia during hazy and nonhazy days. Journal of Geophysical Research, 111, D10S08, doi: 10.1029/2005JD0061, 2006. Dockery, D.W. & Pope, C.A., Acute respiratory effects of particulate air pollution. Annual Review of Public Health, 15, pp. 107–132, 1994. Dusek, U., Frank, G. P., Hildebrandt, L., Curtius, J., Schneider, J., Walter, S., Chand, D., Drewnick, F., Hings, S., Jung, D., Borrmann, S. & Andreae, M. O., Size matters more than chemistry for cloud-nucleating ability of aerosol particles. Science, 312, pp. 1375–1378, 2006. Nel, A., Air pollution-related illness: effect of particles. Science, 308, pp. 804–806, 2005. Hussein, T., Puustinen, A., Aalto, P. P., Mäkelä, J. M., Hämeri, K. & Kulmala, M., Urban aerosol number size distributions. Atmospheric Chemistry & Physics, 4, pp. 391–411, 2004. Tunved, P., Nilsson, E. D., Hansson, H.-C. & Ström, J., Aerosol characteristics of air masses in northern Europe: Influences of location, transport, sinks, and sources. Journal of geophysical research, 110, D07201, doi: 10.1029/2004JD005085, 2005. Wu, Z.J., Hu, M., Lin, P., Liu, S., Wehner, B. & Wiedensohler, A., Particle number size distribution in the urban atmosphere of Beijing, China. Atmospheric Environment, 42, pp. 7967–7980, 2008. Yue, D. L., Hu, M., Wu, Z. J., Guo, S., Wen, M. T., Nowak, A.,Wehner, B., Wiedensohler, A., Takegawa, N., Kondo, Y., Wang, X. S., Li, Y. P., Zeng, L. M. & Zhang, Y. H., Variation of particle number size WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

[10]

[11] [12] [13] [14]

[15]

[16] [17] [18]

[19]

[20] [21]

317

distributions and chemical compositions at the urban and downwind regional sites in the Pearl River Delta during summertime pollution episodes. Atmospheric Chemistry & Physics, 10, pp. 9431–9439, 2010. Gao, J., Wang, T., Zhou, A.H., Wu, W.S. & Wang, W.X., Measurement of aerosol number size distributions in the Yangtze River delta in China: Formation and growth of particles under polluted conditions. Atmospheric Environment, 43, pp. 829–836, 2009. 2010 China’s Sixth National population census data bulletin for Lanzhou, http://www.lzstats.gov.cn/newsshow.asp?id=2760 Zhang, L., Chen, C.H. & Murlis, J., Study on winter air pollution control in Lanzhou, China. Water, Air, and Soil Pollution, 127, pp. 351–372, 2001. Yu, Y., Xia, D.S., Chen, L.H., Liu, N., Chen, J.B. & Gao, Y.H., Analysis of particulate pollution characteristics and its causes in Lanzhou, Northwest China. Environmental Science, 31(1), pp. 22–28, 2010. Wei, F., Teng, E., Wu, G., Hu, W., Wilson, W. E., Chapman, R. S., Pau, J. C. & Zhang, J., Ambient concentrations and elemental compositions of PM10 and PM2 5 in four Chinese cities. Environmental Science & Technology, 33(23), pp. 4188–4193, 1999. Wang, SG, Feng, X.Y., Zeng, X.Q., Ma, Y.X. & Shang, K.Z., A study on variations of concentrations of particulate matter with different sizes in Lanzhou, China. Atmospheric Environment, 43(17), pp. 2823–2828, 2009. Wang, X., Xi, X.X., Guo, Z.L., Liu, Z.G. & Chen, X.L., Analysis of the sand-stormy weather in Lanzhou during the spring of 2002. Journal of Lanzhou University, 42(3), pp. 44–47, 2006. Beddows, D.C .S, Dall’osto, M. & Harrison, R.M., Cluster Analysis of Rural, Urban, and Curbside Atmospheric Particle Size Data. Environmental Science & Technology, 43, pp. 4694–4700, 2009. Zhang, T., Tao, J., Wang, B. G. & Zhang, R.J., Research on size distribution of particles and its impact on visibility in urban Guangzhou during spring. Journal of the Graduate School of the Chinese Academy of Sciences, 27( 3), pp. 331–337, 2010. Gao, J., Wang, J., Cheng, S.H., Yang, L.X., Wang, W.X. & Wang, T., Studies on characteristics of size distribution and formation mechanism of fine particle matters in Summer of Ji’nan. Journal of the Graduate School of the Chinese Academy of Sciences, 24(5), pp. 680–687, 2007. Chen, Y.Z., Chai, F.H. & Wei, Q., Study on size distribution of aerosol composition in winter in Beijing. Journal of Safety and Environment, 6(1), pp. 80–84, 2006. Yan, W.L., Zhou, D.P., Wang, Y.F., Yang, J. & Li, Z.H., Concentrations and Size Distributions of Inhalable Particles in Summer and Winter in Shenyang. Journal of Applied Meteorological Science, 19(4), pp. 435– 443, 2008.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

318 Air Pollution XIX [22] [23]

Morawska, L., Thomas, S., Jamriska, M. & Johnson, G., The modality of particle size distributions of environmental aerosols. Atmospheric Environment, 33, pp. 4401–4411, 1999. Zhang, X.Y., Gong, S.L., Shen, Z.X., Mei, F.M., Xi X.X., Liu, L.C., Zhou, Z.J., Wang, D., Wang, Y.Q. & Cheng, Y., Characterization of soil dust aerosol in China and its transport/distribution during 2001 ACEAsia, 1. Network observations. Journal of Geophysical Research, 108, pp. 4261, doi: 10.1029/2002JD002633, 2003.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

319

Particulates in the atmosphere of Makkah and Mina valley during the Ramadan and Hajj seasons of 2004 and 2005 A. R. Seroji The Custodian of the Tow Holy Mosques Institute of Hajj Research, Umm Al-Qura University, Saudi Arabia

Abstract This work has been devoted to study TSP, PM10 and PM2 5 in the atmosphere of Makkah and the Mina valley during the Ramadan and Hajj periods, 1424 and 1425 H. On the occasion of Hajj, about 2.5 million persons gather in Makkah and move to Mina valley (4 km2), 7 km outside east of Makkah. Pilgrims spend 3 nights in the valley. Congested traffic and the high rates of emissions in such a valley of small area coupled with severe weather conditions, make the area ideal for the accumulation of air pollutants. The present investigation shows that the diurnal cycle of PM10 in air coincides with the pattern of traffic movements. Particulate matters (PM10) daily concentrations in the atmosphere of the Mina valley ranged between 191–262µg/m3 during the presence of the pilgrims in Mina compared to the European standard of 50µg/m3. These concentrations represent 34%–40% of TSP. These high PM10 concentrations are due to the massive transportation movements at Mina valley. Moreover, TSP concentrations reached 665µg/m3 in the Makkah atmosphere during the last ten days of Ramadan compared to the Saudi standard of 340µg/m3. Chemical analysis of PM10 indicated high levels of sulphates, ammonium, nitrates and chlorides. For example, the concentrations of nitrates and sulphates of PM10 were about 4.9% and 6.1% respectively, compared to 2.1% of nitrates and 2.7% of sulphates in TSP. Health dangers that might be encountered by pilgrims due to these pollutants were estimated. It is recommended to set a well planned air quality management program to protect the air of Makkah. Keywords: Makkah, Mina, particulates, PM10, PM2.5, pilgrims, transportation.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110301

320 Air Pollution XIX

1 Introduction Makkah, Saudi Arabia is the holy city for Muslims. Mina valley lies just outside Makkah, 7 km to the east of the city centre and the holy mosque (Al-Masjid AlHaram). Towards the Muslim pilgrimage period every year, Muslims start to gather in Makkah at the beginning of Zulhijah (the last month of the Muslims calendar year). Pilgrims (more than 2.5 millon) move on the 8th of Zulhijah to Mina Valley. They stay in the valley for one night before they proceed to Arafat (about 12 km to the south) on the morning of the 9th of Zulhijah, “Day of AlWakffah”. On the early morning of 10th of Zulhijah, all pilgrims return to Mina on their way back to Makkah. This time they spend 3 days “Tashreek days” and sleep in tents at Mina which has an area of 4 km2, and each pilgrim has less than 2 m2 to live on. Sources of air pollution are mainly fuel combustion with reference to auto exhaust and the movements on the bottom of the valley and the surrounding mountains. Means of transportation are mainly automobiles. The central area of Makkah is characterized by a very dense population, high buildings, narrow streets, and congested traffic flow. Congestion and high rates of pollutant emissions in such valleys of small areas coupled with predominant weather condition of high temperature, lack of rainfall, prevailing one wind direction, low wind speeds and the potentiality of thermal inversions make the area an ideal situation for the accumulation of air pollutants. The problem is of concern when scavenging mechanisms fail to dilute pollutants, pilgrims may face an acute health problem (Nasralla [1]). This study has discussed the concentration levels of particulate in the atmosphere of Makkah and Mina valley during the Ramadan and Hajj seasons in 2004 and 2005. The assessments of health risks for pilgrims are also presented.

2 Methodology The air pollution monitoring program was conducted in Makkah and Mina valley to monitor air pollutants (CO, SO2, NOx, O3 and particulates) and weather elements during pilgrimage periods of 2004 and 2005 (1424 and 1425H). Particulate matters were also measured in Makkah during the month of Ramadan (October 2004). Total suspended particulates, PM10 and PM2 5 were sampled using high volume samplers, PM10 sampler and PM2 5 sampler (Staplex Co.) respectively during pilgrimage period in Mina valley. Measurements were carried out at two levels. The first level is 4 m above the ground surface and the second level is 1.6 m at the entrance of one of the pilgrims’ tents. Sampling period started during both years of the study on 5th to 15th of Zulhijah in 1424 and 1425H (27th Jan.–6th Feb. 2004, and 14th–25th Jan. 2005). Total suspended particulates were sampled in three locations at Makkah city. These sites are (a) the central area adjacent to the holy mosque, (b) city centre about 1km distance from the holy mosque and (c) Alazizia residential district at a distance of 6 km west of the central area of Makkah and about 2 km east of Mina valley. PM10 was also continuously monitored at the central area of Makkah adjacent to the holy mosque using ambient particulate monitor (beta ray absorption). TSP, WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

321

PM10 and PM2 5 were sampled on quartz filters during pilgrimage period in Mina Valley. TSP in Makkah locations were sampled using glass fiber filters. Ions in half of each sample were extracted using deionised water and analysed for chlorides (titrimetry), ammonium and nitrates (colouremetry) and sulphates (turbidimetry) according to the methods described by Nasralla [2]. Particulate samplers were periodically calibrated using air flow calibration kit. Quality assurance/Quality control program for chemical analysis include the extraction of clean filter using deionised water. This extraction was used as a blank and solutions of known concentrations of analysed ions were used for standardization of the analysis procedures.

3 Results and discussions 3.1 Particulate concentrations during Ramadan period

TSP, microgram / m3

Fig. 1 shows the recorded TSP concentrations in the atmospheres of the three monitored locations in Makkah during the month of Ramadan 1424H (Oct/ Nov 2004). About 2 million people visited Makkah to pray in the holy mosque during nights of Ramadan. They spend 1–4 nights in the city. The number of the visitors increased gradually reaching its peak during the last 10 days, that is the days between the 20th and 27th of Ramadan. The movements of the visitors and their vehicles in the central area of Makkah have been reflected on the concentrations of particulates in the city air. This is very clear from the differences between the TSP concentrations recorded in the atmospheres of the three sites of measurements and the significant increase of TSP during the last 10 days as compared to those found at the beginning of the month (fig. 1). Here, it should be noted that TSP reached a concentration of 665µg/m3 during the 20th of Ramadan including the night of the 21st. This concentration is much higher than the Saudi Air Quality Standard of 340 µg/m3. Fig. 1 also indicated that TSP in the atmosphere of the residential area, 6 km out of the city centre, was only 25% higher than the daily limit of 120µg/m3, previously recommended by WHO [3]. 500 450 400 350 300 250 200 150 100 50 0

1st,10 days last 10 days

central area

Figure 1:

city center

residential

TSP in Makkah atmosphere during Ramadan 1424H (Oct/Nov 2004.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

322 Air Pollution XIX Daily average concentrations of TSP in the air of the central area of Makkah reached levels as much as twice of that previously recommended limit of WHO [3]. Moreover, TSP concentrations in the central area during the peak time of the days 20th–27th of Ramadan reached three times more than those found in the air of the residential area and reaching about 5 times of that previously recommended by WHO [3]. Here, it should be noted that WHO removed their guide line for particulate matters from recent publication WHO [4] and stated that there is no safe limit can be recommended for particulates above 20 µg/m3. These high levels of TSP extended to cover the surrounding air of Makkah city centre (fig. 1). Moreover, the daily average concentrations of PM10, during the last 10 days of the month, reached levels of more than 250µg /m3 with a maximum value of 800µg /m3 at midnight of 21st of the month. In the absence of Saudi standard for PM10, these concentrations may be compared to the European standard of 50µg /m3. 3.2 Particulate concentrations during Hajj period The results of the particulates monitoring program during the pilgrimage period (Hajj) show clearly the impact of transportation and the movements of pilgrims on the variations in concentrations of particulate in the atmospheres of the Makkah and Mina valley. Results indicated three different trends of daily variations in TSP concentrations in the air of the three sites of measurements as shown in fig. 2. It may be seen that the lowest concentrations at the three locations were recorded on the 9th of Zulhijah in 1424H (31st Jan., 2004). On that day all pilgrims moved to Arafat valley, 19 km to the SE of Makkah and 12 km of Mina valley. Here, it should be noted that the high levels of TSP reflected the movements of the pilgrims with their vehicles in Makkah, Mina and back to stay in Makkah. In other words, the highest levels of particulates were recorded in the places they live at or move to. Fig. 2 also shows that TSP concentrations

TSP,µg/m3

900 800

Residential

700

City center

600

Mina

500 400 300 200 100 0 5zulh

Figure 2:

6zulh

7zulh

8zulh

9zulh 10zulh Date, Zulhijah 1424

11zulh

12zulh

13zulh

14zulh

TSP in Makkah and Mina atmospheres during pilgrimage session, Zulhijah 1424H (Jan / Feb 2004).

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

323

increased to levels of > 200% during their residence in Mina and >300% during their stay in Makkah as compared to concentrations recorded on the 9th of Zulhijah when they moved to Arafat valley. Here, it should be noted that the early morning movements of vehicles in Makkah and Mina on the 9th of Zulhijah resulted in TSP concentrations twice that recorded in the residential area of Alazizia on the same day. This furthermore confirms the contribution of transportation in polluting the air of Makkah and Mina with particulate matters. Results of the present work show that PM10 concentrations in Mina valley followed the same pattern with regard to TSP daily variation during the period of the pilgrims’ occupancy of the valley. The recorded PM10 concentration in the atmosphere of the valley on the 9th of Zulhijah in 1424H (31st Jan. 2004) was less than 50µg /m3 and peaked to 137µg/m3 on the return of the pilgrims back to Mina on the 10th of Zulhijah (1st Feb. 2004). Furthermore, fig. 3 shows the concentrations of TSP, PM10 and PM2.5 in the Mina valley during the pilgrimage period of 1425H (January 2005). These figures clearly confirm the influence of the vehicles and movements of pilgrims on the concentrations of particulates in the air of the valley. PM10 daily concentrations in the atmosphere of Mina valley ranged between 191–262µg/m3 during the pilgrims’ occupancy of Mina. These concentrations represent 34%– 40% of TSP. This percentage of PM10 is similar to that found in several urban areas such as Cairo (Nasralla [5] and Nasralla et al. [6]), Athens (Chaloulakou et al. [7]), China (Ying et al. [8]) and several other cities of the world (WHO [4, 9]). Furthermore, high levels of PM2 5 have been recorded in the atmosphere of the valley representing 34%–42% of the recorded PM10 concentrations. One of the interesting findings of this work is that the levels of PM2 5 and PM10 at 4.5 m above the ground surface did not vary much than those recorded at 1.6 m at the entrance of one of the pilgrim’s tents. Knowing that pilgrims usually sleep in tents as well as outside and around streets, it can be concluded that pilgrims are exposed to high levels of these air pollutants with deleterious effects on their health. Here, it should be noted that WHO [10] stated that there is no safe limit for the exposure to PM10 levels above 20µg/m3. 800 TSP,4m 700

PM10, 4m PM10,1.6m

concentraion,µg/m3

600

PM2 5,1.6m

500 400 300 200 100 0 6Zulh

7Zulh

8Zulh

9Zulh

10Zulh

11Zulh

12Zulh

13Zulh

14Zulh

Date , Hijry

Figure 3:

TSP, PM10 and PM25 in Mina Valley during pilgrimage 1425H, Jan 2005.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

324 Air Pollution XIX Similarly, pilgrims were exposed to high levels of PM10 during their residence in Makkah close to the central area (fig. 4). These daily concentrations of PM10 recorded in the atmosphere of Makkah may be compared to the European United Community (EUC) air quality standard of 50µg/m 3. Moreover, it may be seen that the pattern of PM10 variation in Makkah air during the different occasions of pilgrimage was very similar during both years (2004 and 2005) of the study (see fig. 4). This confirms the reflection of man’s activities with reference to transportation on the concentrations of PM10 in the air of Makkah central area.

PM10, micrograms/m3

180 160

1424H

140

1425H

120 100 80 60 40 20 0 a

b

c d e periods of pigrimage occasions

f

Legend: a, before pilgrims arrival to Makkah, 24th–29th of Zulkedda b, residence in Makkah before pilgrimage, 1st–6th of Zulhijah c, leaving to Mina and Arafat, 7th and 8th of Zulhijah d, Arafat day, 9th of Zulhijah e, residence in Mina and moving between Mina and Makkah, 10th–12th of Zulhijah f, back to stay in Makkah, 14th–20th of Zulhijah Figure 4:

PM10 in Makkah central area during various pilgrimage occasions. 250

18

CO

200 PM10 , µg/m3

20

PM10

16 14 12

150

10 8

100

6 4

50

2 0

0 hr

03:00

06:00

09:00

12:00

15:00

18:00

21 00

00 00

Time,hr

Figure 5:

Diumal variation of CO and PM10 in Makka central area, Zulhijah 1425H, 14 Jan 2005.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

325

This conclusion is furthermore confirmed by the diurnal variation of PM10 and CO in Makkah air (figs. 5 and 6). These figures indicated, to a great extent, similar diurnal cycles for both air pollutants. Furthermore, the increase of PM10 concentrations in Makkah air during night time is very similar to that recorded in Mina valley. This cannot only be explained by the movements of vehicles 24 hours daily during pilgrimage, but is possibly due to the nocturnal formation of a stable and stagnant atmosphere and the long periods of calm or light winds during night times. 300

20

250

PM10

18

CO

16 14

PM10,ug/m3

200

12 150

10 8

100

6 4

50

2 0

0

hr

03:00

06 00

09:00

12 00

15:00

18 00

21:00

00 00

time,hr

Figure 6:

Diurnal variation of CO and PM10 in the atmosphere of Makka central area, 14 Zulh, 24 Jan 2005.

3.3 Estimated health risks in Mina during Hajj season Analysis of particulate samples collected from the Mina atmosphere during “Tashreek days” indicated that PM10 contains high levels of nitrates (NO3-), sulphates (SO4- -), ammonium (NH4+) and chloride (Cl-) compared to those encountered in TSP (table 1). The high levels of nitrates, sulphates, ammonium and chloride recorded in PM10 samples are in accordance with previous findings in several European urban areas (Clarke [11] and WHO [9]). Table 1:

Average concentrations of PM10 and TSP anions in Mina. Ion

TSP, %

PM10, %

TSP, µg /m3

PM10, µg/m3

NO3SO4- NH4+ Cl-

2.1 2.7 1.8 1.4

4.9 6.1 3.8 1.9

7.8 10 6.7 5.2

6.3 7.9 4.9 2.5

A trial was made to assess the health risks for pilgrims during their residence in Makkah and Mina due to the exposure to PM10.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

326 Air Pollution XIX Calculations were conducted according to eqns. (1), (2), (3), (4) given by WHO [10], (see table 2). % increase in daily mortality = (0.07 ± 0.012) x PM10…… .……(1) % change in hospitals admission = (0.084 ± 0.033) x PM10… .....(2) % change in cough = (0.455 ± 0.228) x PM10 ….…………. ……(3) ..(4) % change in symptom exacerbation = (0.345 ± 0.162) x PM10 . Table 2:

Estimated risks for pilgrims health during residence in Mina, pilgrimage 1425H (2005). % increase in mortality

% change in hospitals admission

Mina

13 - >16

16 - >23

Makkah

8 - 14

10 - 14

% change in cough 86 >123 51 - 77

% change in symptom exacerbation 65 - >101 39 -59

The high temperature and relative humidity, existence of high concentrations of other air pollutants and the required heavy physical activities and high particulate concentrations are likely to cause adverse health effects during “Tashreek days” in Mina and during their residence in Makkah central district. It is important to keep in mind that more than 350,000 pilgrims were aged over 50 years and more than 75,000 over 65 years of age CDSI [12] and are thus more susceptible to the adverse effects of air pollutants on health. It is recommended to conduct thorough investigations and source apportionment studies to evaluate the exact contribution of the different types of vehicles to the problem of PM10 in Makkah and Mina. In fact, well planned air quality management and an alternative transportation system are urgently recommended for Makkah and holy places.

4 Conclusion TSP, PM10 and PM2 5 in the atmosphere of the Makkah and Mina valley were measured during Ramadan and Hajj periods. PM10 daily concentrations in Mina valley ranged between 191–262µg/m3. Analysis of PM10 indicated high levels of sulphates, ammonium, nitrates and chlorides as compared to TSP. The concentrations of TSP and PM2 5 in different periods of time were also estimated.

Acknowledgement The authors acknowledge the support of the Custodian of the Tow Holy Mosques Institute of Hajj Research, Umm Al-Qura University, Makkah for financial support and providing facilities.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

327

References [1] Nasralla, M.M. Carbon monoxide and photochemical oxidants in Mina valley during pilgrimage, Arab Gulf Journal of Scientific Research, 4 (1), pp. 193-201, 1986. [2] Nasralla, M.M., Air pollution in subtropical Saudi urban area, Environment International, 9, pp. 255-264, 1983. [3] WHO, Air Quality Guidelines, 1st Edition, World Health Organization, Regional Office for Europe, Copenhagen, Denmark, 1987. [4] WHO, Air Quality Guidelines, Global Update 2005, Particulate matters, ozone, nitrogen dioxide and sulfur dioxide, World Health Organization (WHO), Regional Office for Europe, Copenhagen, Denmark, 2006. [5] Nasralla, M.M. Carcinogenic, toxic and microbial contaminants in Cairo air, Final Report, Academy of Science and Technology, Cairo, Egypt, 1997. [6] Nasralla, M.M , Ali, E. A, Hassan, W. H. and Shahat F., Size distribution and chemistry of particulate matters in Cairo atmosphere, 3rd International Conference on the Environment, NRC, Cairo, 2006. [7] Chaloulakou A., Kassomenos P., Grivas G., Spyrellis N., Particulate matter and black smoke concentration levels in central Athens, Greece. Environment International, 31, pp. 651-659, 2005. [8] Ying Wang, Guoshun Zhuang, Xingying Zhang, Kan Huang, Chang Xu, Aohan Tang, Jiamin Chen, Zhisheng An, The ion chemistry seasonal cycle and sources of aerosol in Shanghai. Atmospheric Environment, 33, pp. 843853, 2006. [9] WHO, Air Quality Guidelines, 2nd Edition, World Health Organization (WHO), Regional Office for Europe, Copenhagen, Denmark, 2000. [10] WHO, Guidelines for Air Quality, World Health Organization (WHO), Geneva, 2000a. [11] Clarke, A.G., Particle size and chemical composition of urban aerosols, The Science of the Total Environment, 23(5), pp. 15-24, 1999. [12] CDSI, (Central Department of Statistics & Information), Results of statistical data of pilgrims for Hajj season in 1425, Ministry of Economic & Planning, 2005.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

This page intentionally left blank

Section 5 Emissions studies

This page intentionally left blank

Air Pollution XIX

331

Effect of biodiesel and alkyl ether on diesel engine emissions and performances D. L. Cursaru1, C. Tănăsescu1 & V. Mărdărescu2 1 2

Petroleum-Gas University of Ploieşti, Romania Transilvania University of Braşov, Romania

Abstract Over the last years numerous attempts have been made to minimize the amount of toxic and harmful exhaust gases from diesel powered vehicles. The rapidly exhausted fossil sources coupled with increasing price of petroleum together with the public awareness concerning the environmental protection, are the main reasons that have made many scientists to search for alternative and renewable energy sources. According to the recent EU regulations starting with 1st of January 2010, 5.75wt% of classical diesel fuel must be replaced with more environmental friendly fuels. The most used biofuel is biodiesel (fatty acid methyl esters), mainly synthesized by catalytic transesterification of fatty glycerides. Roughly 10 wt% of glycerol is obtained as by-product in catalytic transesterification of fatty glycerides and there are many researching directions in order to find new applications for the increasing availability of glycerol as a low-cost feedstock. In our study we have tested the energetic and ecologic performances (the exhaust emissions as CO, CO2 and NOx) of biodiesel or alkyl ether-diesel blends by testing the fuels on a Diesel engine 392-L4-DT/104 at different engine speeds. The emission tests were measured by using a FTIR SESAM 1.4 equipment. Engine tests were run on the same engine in the same day in order to have the same atmospheric conditions (96 kPa pressure and 29oC) within the three repetitions of each test and average of measured values were taken. The measured CO emissions of biodiesel and alkyl ether-diesel blends were found to be 15% and 37% lower than that of diesel fuel, respectively. Keywords: biodiesel, alkyl ether, diesel engine, emissions.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110311

332 Air Pollution XIX

1 Introduction In the past century, fuels derived from fossil resources, were the global source of energy. The rapidly depletion of the petroleum, the oscillating increase in the oil price per barrel together with the worldwide effort to protect the environment, are the main reasons that have made many scientists to concentrate their efforts to search for alternative and renewable energy sources in order to reduce harmful emissions. Concerning diesel fuel, the newest regulations related to the environmental protection are more restrictive. In order to fulfill these requirements it is obviously essential to control the injection and combustion processes or to find alternative fuels that may offer an approach to reduce harmful emissions without drastically modifications of the engine power and fuel consumption. In our study we try to find alternative fuels as biodiesel and alkyl ether in order to reduce the dangerous emissions. It is well known that biodiesel is a very interesting alternative fuel because does not contain carcinogens such poly aromatic hydrocarbons and nitrous poly aromatic hydrocarbons. More than this, when burned, produces pollutants which are less aggressive to human health. The emissions of carbon monoxide and particulate matter are much lower than corresponding to classic diesel fuel, although a slightly increase in emissions of nitrogen oxides (NOx) was observed by using biodiesel. The increasing of nitrogen oxides emissions could be attributed to the lowered in-cylinder soot levels thus lower radiation heat transfer resulting in higher in-cylinder temperatures. By using biodiesel, it is important to improve the calorific value, the horsepower output and to reduce the emission of nitrogen oxides (NOx). The objective of the present study is to investigate the effects of biodiesel and alkyl ether addition in diesel fuel over diesel engine emissions and performances. The fatty acid methyl esters are produced by catalytic transesterification of the fatty glycerides existing into sunflower oil in the presence of methanol. The glycerol is the main by-product obtained in the transesterification reaction and its production is equivalent to approximately 10 wt% of the total fatty acid methyl esters (biodiesel), therefore the glycerol was etherified with isobutene and the main product – the alkyl ether was also used for diesel fuel blending. The fatty acid methyl ester or alkyl ether-diesel fuel blends were tested in a diesel injection engine in order to measure the engine power, specific fuel consumption, smoke density and the amount of CO, CO2 and NOx from the gases from the exhaust pipe of the car.

2 Experimental The experimental study was focused on two researching routes; the first one was carried on synthesis and characterization of fatty acid methyl esters and alkyl ethers, while the second route consisted of engine performances of fatty acid methyl ester or alkyl ether-diesel blends and comparison to classic diesel fuel.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

333

2.1 Fatty acid methyl ester synthesis Transesterification of the fatty glycerides existing into sunflower oil was realized in a batch reactor using potassium hydroxide as catalyst via a method given elsewhere [1–3]. The reaction was carried out at 60°C, atmospheric pressure, for 2 hours, under vigorous agitation in order to achieve the maximum conversion. In order to obtain the fatty methyl ester by transesterification we used 100 wt% excess methanol, keeping the molar ratio sunflower oil to methanol at 1:6 and the catalyst concentration of 1%. The crude methyl ester was separated by glycerol by gravity and the catalyst was eliminated by hot water washing. The residual water in biodiesel was removed by distillation at 100°C and reduced pressure for 30 min. The properties of the sunflower oil and fatty acid methyl ester such acid value, density, viscosity, lubricity, and pour point were determinate according to ASTM standards and are given in table 1. Table 1:

Properties of vegetable oil, methyl ester, glycerol and alkyl ether. Results

Density [25 C, kg/m3] Viscosity [40 C, cSt] Lubricity WS1.4, [  m ] Acid value [mgKOH/g] Pour point, [ C]

Methods

Sun flower oil

Methyl ester

Glycerol

Alkyl ether

918 31.86 123 0.12 -13

857 4.38 210 0.30 -5

975 1 6.3 300 0 16 -22.0

831.0 180 -20.0

ASTM D-1298 ASTM D-455 ASTM D-6079 ASTM D-1980 ASTM D-2500

2.2 Alkyl ether synthesis Roughly 10 wt% of glycerol is obtained as by-product in catalytic transesterification of fatty glycerides and there are many researching directions in order to find new applications for the increasing availability of glycerol as a low-cost feedstock. In our present contribution the alkyl ether was synthesized by catalytic etherification of glycerol with isobutene in a stainless steel Berghoff autoclave. The reaction was carried out at 80°C, for 5 hours under vigorous agitation, with 1.5:1 isobutene:glycerol molar ratio [4]. The properties of the glycerol and alkyl ether such acid value, density, viscosity, lubricity, and pour point were determinate according to ASTM standards and are given in table 1. 2.3 Materials Diesel fuel used for our investigations was obtained by hydrodesulfurization of diesel fuel from atmospheric distillation plant and it was delivered by PetrotelLukoil Refinery. Its properties, especially the sulfur content and lubricity, are depicted in table 2. The sunflower oil, methanol and potassium hydroxide used in our investigation are of analytical grade and were provided by Sigma-Aldrich. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

334 Air Pollution XIX Table 2: Characteristics Density [25 C, kg/m3] Sulfur [wt%] Viscosity at 40 C, cSt Cetane number Pour point, C Copper corrosion Distillation T 90% T 95% Hydrocarbon type Aromatics Polyaromatics Lubricity WS1.4,  m

Properties of diesel fuel. Diesel Fuel 838.0 0.0076 2.29 51.9 -17 1a C 334.7 354.9 % vol 22.1 6.2 535

Methods ASTM D-1298 ASTM D-2622 ASTM D-455 ASTM D-613 ASTM D-2500 ASTM D-130 ASTM D-86 ASTM D-1319 ASTM-D6079

2.4 Engine tests The energetic and ecologic performances of diesel fuel, fatty acid methyl ester or alkyl ether-diesel blends were investigated by testing the fuels on a Diesel engine 392-L4-DT/104 at different engine speeds. The parameters of diesel engine are given in table 3. The emission tests are measured by using FTIR SESAM 1.4 equipment. Engine tests were done according to internal combustion engines examine and test standards (RO 6635/87). Engine performances and exhaust emission tests were carried out at 1600, 1800 and 2000 RPM. Engine tests were run on the same engine in the same day in order to have the same atmospheric conditions (96 kPa pressure and 29oC) within the three repetitions of each test and average of measured values were taken [5–12]. Table 3:

Nameplate parameters of the diesel engine.

Engine Model Engine type Engine total displacement Compression ratio Bore/Stroke Maximum power Injection model Fuel injection pump Maximum power Maximum toque Needle opening pressure

392-L4-DT seria 104 four stroke, 4 cylinder in line, water cooled 3922 cm3 17.5:1 102 mm/120 mm 68 kW at 2600 rpm KBEL BOSCH-DLLA 150 P44 RO-PES 4A 90D 410 RS 2240 76 kW@2800 rpm 255 Nm@1600 rpm 240 bar

3 Results and discussion 3.1 Engine performances The engine performances such variation of power, specific fuel consumption, variation of smoke density and the exhaust emissions as CO, CO2 and NOx of WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

335

prepared alkyl ether or methyl ester mixed with diesel fuel were studied in comparison with diesel fuel. 3.1.1 Variations of power The variation of engine power with engine speed for diesel fuel and for mixtures diesel fuel-methyl ester, diesel fuel-alkyl ether is presented in figure 1. From figure 1 it is obviously that the power output of classic diesel fuel is higher than the power developed in the case of blended fuels but, in general, the maximum power values of all three fuels are very close. By addition of methyl ester or alkyl ether in diesel fuel the engine power decreases with about 4% than the engine output when classic diesel fuel is used. The engine power decreases proportional with the rising of the oxygen content from the mixture. Another possible explanation for this decrease could be due to the fuel flow problems, as higher density and higher viscosity [13]. Diesel fuel 90 wt% Diesel fuel+10wt% Biodiesel 90 wt% Diesel fuel+10wt% Ether

45 44 43

Engine power, kW

42 41 40 39 38 37 36 35 1600

1700

1800

1900

2000

Engine speed, RPM

Figure 1:

Engine power versus engine speed.

3.1.2 Specific consumption The evolution of the specific consumption with engine speed for all three fuels is depicted in figure 2. For smallest engine speed 1600 RPM, the highest specific consumption was recorded for classical diesel fuel but for higher engine speeds the specific consumption for diesel fuel decreases while for mixtures diesel fuelmethyl ester and diesel-alkyl ether we observed a high specific consumption. A higher density and viscosity compared to diesel fuel, as well as a lower heating value coupled to a bad fuel injection atomize, could be a possible explanation for this evolution. 3.1.3 Variations of smoke density The variation of smoke density produced with engine speed during the test for all three fuels is depicted in figure 3. The maximum smoke density was measured at WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

336 Air Pollution XIX Diesel fuel 90 wt% Diesel fuel+10 wt% Biodiesel 90 wt% Diesel fuel+10 wt% Ether

224

Specific consumption, g/kWh

222 220 218 216 214 212 210 208 1600

1700

1800

1900

2000

Engine speed, RPM

Figure 2:

Specific consumption versus engine speed. Diesel fuel 90 wt% Diesel fuel+ 10 wt% Biodiesel 90 wt% Diesel fuel+ 10 wt% Ether

220

Smoke density, mg/m

3

200

180

160

140

120

100 1600

1700

1800

1900

2000

Engine speed, RPM

Figure 3:

Smoke density versus engine speed.

2000 RPM and from figure 3 it is obvious that the smoke density is much lower for mixture diesel fuel-alkyl ether. The smoke density measured for diesel fuel was found to be 161.55 mg/m3 while for mixture diesel fuel-alkyl ether was 106.09 mg/m3 at 1600 RPM. The maximum smoke density was measured at 2000 RPM as 196.3 mg/m3 for diesel fuel, 134.19 mg/m3 for diesel fuel blended with alkyl ether and 168.31 mg/m3 for diesel fuel blended with biodiesel. The average amount of smoke density decrease was determined as 33% for diesel fuel blended with alkyl ether and as 13.4% for diesel fuel blended with biodiesel compared to diesel fuel.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

337

3.1.4 CO emission The variation of CO produced by running the diesel engine using all three fuels is presented in figure 4. By increasing the engine speed we observed a diminish of CO emissions with about 21wt.% for diesel fuel, with 15 wt.% for mixture diesel fuel-biodiesel, but the most spectacular decreasing of 37 wt.% was noticed for the mixture alkyl ether-diesel fuel. The average reduction of CO emissions was 17 wt.% less than diesel fuel. 1600

Diesel Fuel 90wt% Diesel Fuel+10wt% Biodiesel 90wt% Diesel Fuel+10wt% Ether

1500 1400

CO content, ppm

1300 1200 1100 1000 900 800 700 600 500 1600

1700

1800

1900

2000

Engine speed, RPM

Figure 4:

CO emissions versus engine speed.

3.1.5 CO2 emission In figure 5 is presented the variation of CO2 emissions with engine speed using diesel fuel and diesel fuel blended with biodiesel or alkyl ether. The maximum CO2 was produced for all three fuels at 1800 RPM, respectively 7.6 wt.% when 7,8

Diesel Fuel 90 wt% Diesel Fuel+10 wt% Biodiesel 90 wt% Diesel Fuel+10 wt% Ether

7,6 7,4 7,2 7,0

CO2, %

6,8 6,6 6,4 6,2 6,0 5,8 5,6 5,4 1600

1700

1800

1900

2000

Engine speed, RPM

Figure 5:

CO2 emissions versus engine speed.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

338 Air Pollution XIX diesel fuel was running the diesel engine, 6.5 wt.% when diesel fuel blended with 10 wt.% biodiesel was used and 5.9 wt.% when diesel fuel blended with 10 wt.% alkyl ether. The highest CO2 emission was obtained from the use of diesel fuel while the lowest CO2 emission was recorded by using diesel fuel blended with ether. 3.1.6 Exhaust temperatures In figure 6 is depicted the variation of exhaust temperature with engine speed. Maximum exhaust temperatures were measured for 1800 RPM. These temperatures were determinate as 510 0C for diesel fuel, 505 0C for diesel fuel blended with biodiesel and 495 0C for diesel fuel blended with alkyl ether. The exhaust temperature for diesel fuel blended with biodiesel is lower as 1% at all test period than diesel fuel and for diesel fuel blended with biodiesel is lower as 3% than diesel fuel. This could be because the methyl ester and alkyl ether have lower heating values than diesel fuel. 525

Diesel Fuel 90wt% Diesel Fuel+10 wt% Biodiesel 90wt% Diesel Fuel+10 wt% Ether

520

0

Exhaust Temperature, C

515 510 505 500 495 490 485 480 1600

1700

1800

1900

2000

Engine speed, RPM

Figure 6:

Exhaust temperature versus engine speed.

3.1.7 NOx changes The variation of NOx with engine speed for all three fuels is shown in figure 7. The maximum content of NOx corresponding to all three fuels was recorded at 1800 RPM. NOx emitted by biodiesel blend are higher than the one corresponding to diesel fuel, while the NOx emissions related to ether blend are much lower than the one associated to diesel fuel and biodiesel blend. This tendency can be explained as a consequence of the advanced injection derived from the physical properties of biodiesel (viscosity, density, compressibility).

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

339

Diesel fuel 90wt% Diesel fuel+10wt% Biodiesel 90wt% Diesel fuel+10wt% Ether

1100

NOx content, ppm

1000

900

800

700

600 1600

1700

1800

1900

2000

Engine speed, RPM

Figure 7:

Exchange of NOx.

4 Conclusions In recent years investigations have been done in order to fulfill the EU regulations concerning the environmental protection, mainly to reduce harmful diesel fuel emissions. In our experimental study we investigate the effect of addition of two different compounds; methyl ester and alkyl ether in diesel fuel, on diesel engine performances and exhaust emissions. The following conclusions may be drawn from the result of the present study:  Fatty acid methyl ester obtained by catalytic transesterification of the fatty glycerides existing into the sunflower oil is a renewable energy resource;  Alkyl ether is also a renewable energy resource and a low-cost feedstock;  The maximum power output was observed, for all engine speeds, for diesel fuel while a slightly decreasing of the power output was recorded for diesel fuel blended with alkyl ether or biodiesel;  The emissions of CO and CO2 were significantly reduced by blending classic diesel fuel with biodiesel or alkyl ether;  The NOx emissions were slightly increased with the use of biodiesel blend with respect to those of the classic diesel fuel, while the emissions recorded for alkyl-diesel fuel blend were lower than for the other two fuels;  The first results of our investigations on engine performances and exhaust emissions are positive, however, additional tests are necessary to find the optimum biodiesel or alkyl ether percents which must be added to diesel fuel in order to obtain maximum performances of diesel engine.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

340 Air Pollution XIX

Acknowledgement The authors are grateful for financial support to Postdoctoral project CNCSISUEFISCSU, project number PN II-RU 6/2010.

References [1] Anastopoulos, E.L., Zannikos, F., et al., Influence of aceto acetic esters and di-carboxylic acid esters on diesel fuel lubricity, Tribology International, 34, pp. 749-755, 2001. [2] Benjumea, P., Agudelo, J., Agudelo, A., Basic properties of palm oil-diesel blends, Fuel, 87, pp. 2069-2075, 2008. [3] Caynak, S., Gürü, M., Bicer, A., Keskin, A., Icingür, Y., Biodiesel production from pomace oil and improvement of its properties with synthetic manganese additive, Fuel, 88, pp. 534-538, 2009. [4] Mangourilos, V., Conversion of glycerol to additives for green fuels, PhD Thesis, Petroleum-Gas University of Ploiesti, Romania, 2009. [5] Kegl, B., Effect of biodiesel on emissions of a bus diesel engine, Bioresource Technology, 99, pp. 863-873, 2008. [6] Buyukkaya, E., Effects of biodiesel on a DI diesel engine performance, emission and combustion characteristics, Fuel, 89, pp. 3099-3105, 2010. [7] Rakapoulos, D.C., Rakapoulos, C.D., Giakoumis, E.G., Dimaratos, A.M., Kyritsis, D.C., Effects of butanol-diesel fuel blends on the performance and emissions of a high-speed DI diesel engine, Energy Conversions and Management, 51, pp. 1989-1997, 2010. [8] Rakapoulos, C.D., Rakapoulos, D.C., Hountalas, D.T., Giakoumis, E.G., Andritsakis, E.C., Performance and emissions of bus engine using blends of diesel fuel with bio-diesel of sunflower or cottonseed oils derived from Greek feedstock, Fuel, 87, pp. 147-157, 2008. [9] Rakapoulos, C.D., Antonopoulos, K.A., Rakapoulos, D.C., Experimental heat release analysis and emissions of a HSDI diesel engine fueled with ethanol-diesel fuel blends, Energy, 32, pp. 1791-1808, 2007. [10] Rakapoulos, C.D., Antonopoulos, K.A., Rakapoulos, D.C., Hountalas, D.T., Giakoumis, E.G., Comparative performances and emissions study of a direct injection Diesel engine using blends of Diesel fuel with vegetable oils of bio-diesels of various origins, Energy Conversions and Management, 47, pp. 3272-3287, 2006. [11] Ghojel, J., Honnery, D., Al-Khaleefi, K., Performance, emissions and heat release characteristics of direct injection diesel engine operating on diesel oil emulsion, Applied Thermal Engineering, 26, pp. 2132-2141, 2006. [12] Kumar, M.S., Kerihuel, A., Bellettre, J., Tazerout, M., Ethanol animal fat emulsions as a diesel engine fuel-Part 2: Engine test analysis, Fuel, 85, pp. 2646-2652, 2006.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

341

[13] Utlu, Z., Kocak, M.S., The effect of biodiesel fuel obtained from waste frying oil on direct injection diesel engine performance and exhaust emissions, Renewable Energy, 33, pp. 1936-1941, 2008. [14] Lapuerta, M., Armas, O., Rodriguez-Fernandez, J., Effect of biodiesel fuels on diesel engine emissions, Progress in Energy and Combustion Science, 34, pp. 198-223, 2008.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

This page intentionally left blank

Air Pollution XIX

343

Emissions of selected gas pollutants in the application of the additive EnviroxTM F. Bozek1, J. Mares1, H. Gavendova1 & J. Huzlik2 1 2

Civil Protection Department, University of Defence, Czech Republic Transport Research Centre, Czech Republic

Abstract The value of selected emission factors was monitored in the operation of an older type of engine testing bench using diesel and compared with the same parameters monitored under similar conditions with the addition of the additive, EnviroxTM. It was found that the additive based on CeO2 nanoparticles reduces emissions of hydrocarbons CxHy, and NOx, while emissions of CO2 remain comparable or slightly lower and CO emissions even significantly increase. Dependence of tested emissions on reduced torque TMR, engine power P and revolutions f was observed as well. Keywords: additive, emission factor, nano-particle, environment, cerium dioxide, carbon mono-oxide, nitrogen oxide.

1 Introduction Escalating activities of human society in order to ensure a higher standard of living brings along a number of negative externalities. Among the important, and by the lay and professional community strongly discussed externalities, belongs the increase of burden to the atmosphere through the implementation of industrial and agricultural activities, energy production, waste management and household management. To the increase in ambient air pollution also contributes a significant level of mobile assets, consisting primarily of passenger and goods transport [1]. It is estimated that transport contributes to overall air pollution by carbon monoxide (CO) and carbon dioxide (CO2) emissions by 37%, a mixture of nitrogen oxide and nitrogen dioxide (NOx) by 30%, and volatile organic compounds (VOCs), about 24%. Besides the above, the quantitative aspect of the WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110321

344 Air Pollution XIX dominant pollutant is the traffic which is also producer of other, mainly health and eco highly harmful substances [2]. Emitted pollutants have often nonquantifiable impact on morbidity and mortality of the population, ecosystem function and value of social assets [3]. Because transportation is currently one of the world’s most dynamically developing sectors, it is necessary, in accordance with the principles of sustainable development, to pay enormous attention to minimize emissions. This requirement is compounded by the use of personal automobiles at the expense of public transport and permanently increasing the ratio of road haulage transport relative to rail transport [1]. The submitted paper is devoted to the evaluation of the quantities of CO2, CO, NOx and VOCs in the form of unburned hydrocarbons CxHy emissions from fuels used in conventional diesel engines in comparison to the addition of fuel additive based on cerium dioxide (CeO2).

2 The analysis of current state The amount of pollutants emitted while driving a motor vehicle is dependent on many factors. If the typical way of driving for each driver is left aside, as well as the nature of terrain and weather conditions, the current emissions are dependent upon, in particular: a) The type of engine and its technical parameters [4]; b) Principles of oxidation catalyst effect [5]; c) The type and amount of biodiesel added to the base fuel [6]; d) The composition and quality of used motor oil [7]; e) The type and composition of additives added to the basic fuel [8]. The calculation of the i-th pollutant emission is based on the knowledge of an emission factor Ef mi [g kg-1] which is consistently with eqn (1) given by weight of the i-th pollutant per a unit mass of consumed fuel [9]:

Ef mi  mi  mF

1

 yid 

M i nd M F  NF

(1)

where mi [g] is the mass of the i-th pollutant, mF [kg] weight of fuel, Mi molar molecular weight of the i-th pollutant [g mol-1], MF [kg mol-1] molar molecular weight of fuel, nd [mol] substance amount of dry exhaust gas, NF [mol] substance amount of consumed fuel, and finally y id molar fraction of the i-th pollutant in dry exhaust gasses. The industry producing and supplying fuels operates with a number of additives that can be added to the diesel fuel. These can be divided into three main areas depending on their nature [10]: a) Refinery additives; b) Safety increasing or legally required additives; c) Additives for improvement of technical parameters and increase of fuel performance. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

345

(8x+2y) CeO2 + 2 CxHy = (4x+y) Ce2O3 + 2x CO2 + y H2O

(2)

4 CeO2 + C = 2 Ce2O3 + CO2

(3)

2CeO2 + CO = Ce2O3 + CO2

(4)

2 Ce2O3 + 2 NO = 4 CeO2 + N2

(5)

4 Ce2O3 + 2 NO2 = 8 CeO2 + N2

(6)

CeO2 regeneration catalyst is carried out in accordance with chemical formula (7). 2 Ce2O3 + O2 = 4 CeO2

(7)

Statistically validated operational tests carried out by Oxonica Company provide evidence that the recommended dosage of 5 to 10 ppm w/w CeO2 can achieve relevant reductions in fuel consumption (about 5-12%) present reduction of emissions of CO2, CO, NOx, CxHy and particulate matters. The additive is also compatible with all diesel common additives [11, 12].

3 Problem solution 3.1 Applied methods and devices Tests to determine emission levels were carried out on the engine testing bench of VOP-026 Šternberk electric eddy current brake Schenk 0900 kW, operating in the range of 0–6000 revolutions min-1. Diesel NM-54 was used as the primary fuel which served as an alternative version for the comparative tests mixed with 2.510-4 volumes of additives. The concentrations of CeO2 found in diesel fuel by inductively coupled plasma atomic emission spectroscopy was 7.6 ppm w/w which corresponds to EnviroxTM suppliers’ requirement. For the actual test a diesel engine was used, four stroke, naturally-aspirated engine Tatra T3 930-31 with direct injection, air-cooled, engine cylinder capacity of 1.9104 cm3, cylinder diameter/stroke 120/140 mm, OHV distribution and a compression ratio of 1:16. The engine had 12 cylinders in two separate lines at 90°. Rated engine output was 235 kW  10% at 2.2103 min-1 with a maximum torque of 1.13103 N m at revolutions 1.4103  200 min-1. Emission testing was performed by a combined device for analysis of combustion gas composition ECOM – JN, equipped by electrochemical sensor of an English company City Technology which enabled the determination of CO, NO, NO2 and O2 concentrations. Sensor types, ranges and uncertainty in determination of individual quantities are listed in table 1. A sample of combustion gas was taken by a vacuum pump tube probe analyzer. The current air mass was led from the probe tube by unheated tube to filters and water separators analyzer and then to each pollutant sensors. It was possible to determine the concentration of CO and NOx by applying unheated tube between the probe and analyzer as possible combustion gas condensation in the traffic route did not affect their value. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

346 Air Pollution XIX Table 1:

Ranges and uncertainties in determination of measured quantities.

Pollutant NO NO2 CO O2

Uncertainty of measurements 20 % range 100 % range 2% 5% 2% 5% 2% 5% 2% 5%

Range (ppm) 0 – 2.0103 0 – 2.0102 0 – 1.0104 0 – 2.1105

CxHy content was tested by analyzer operating on the principle of flame ionization (FID). The principle uses the effect that the burning of hydrocarbons in the hydrogen flame of the combustion chamber of the analyzer burner gets ionized bond C-H. If the electrodes placed in the burner are energized, the value of flowing current is proportional to the number of free ions including organic matter content in the sample. Into the FID unit was the sample of gases transported through a tube heated vacuum pump analyzer. Location of measuring point where the combustion gas velocity was measured with the Prandtl probe simultaneously with gas temperature measured with thermocouple is evident from fig. 1. left hand right hand side of engine gas fumes exhausting

heat removal shield measured engine

measuring point (pipeline)

gas fumes exhausting to the chimney

Figure 1:

Location of sampling probes.

3.2 Outcomes and discussion 3.2.1 Mass balance Calculation of emission factors combustion of fuel, also based hydrogen (10), oxygen (11) and entering into the combustion

was based on eqn (8) characterizing the on the carbon material balance, eqn (9), nitrogen (12), on the amount of substance process NVd [mol] and on the process

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

347

of exiting nd [mol] of the dry gas. eqn (13) eventually (14) and finally based on measurements of the concentration of contaminants and oxygen. Fuel combustion in the engine was considered under simplified conditions in the absence of trace amounts of polycyclic aromatic hydrocarbons, N2O, NH3, SO2 etc. e/2 N2+ CaHbOc+ [a(1-z/2)+b/4-c/2+d(1-w/2)] O2 = = a(1-z) CO2 + az CO+ (b/2) H2O + d(1-w) NO2+ ew NO

(8)

a   N F  n F   n CO 2  n CO

(9)

b  N F  n F   2 n H 2O

(10)

2 ( N O2  nO2 )  c   N F  n F   n H 2 O  2nCO 2  n CO  2n NO 2  n NO

2

N

N2

 n N2



n NO2  n NO

(11) (12)

N Vd  N N 2  N O2  N CO2

(13)

n d  n N 2  nO2  N CO2  nCO2  nCO  n F  n NO2  n NO

(14)

In eqns (9)–(14) N [mol] with the corresponding subscript represents amount of substance in the process of entering gases N2, O2 and CO2 and n [mol] subscript represents amount of substance in the process of exiting exhaust gases, i.e. N2, O2, CO2, CO, NO2, NO and fuels F. To simplify the record of other formulas a substitution (15) was introduced where symbols a, b, c are stoichiometric coefficients in eqn (8).



b  2c 4a



  4



2

(15)

Molar fractions Yi d a y id of i-th component in dry inlet air or combustion gas are defined by relations (16) or (17):

Yi d  N i  ( N Vd ) 1 ´

(16)

y id  n i  ( n d ) 1

(17)

where Ni [mol] is the amount of substance of i-th component in the inlet ni [mol] is the amount of substance of i-th component in output, symbol NVd [mol] has the same meaning as in eqn (13) and the symbol nd [mol] as in eqn (14). Based on the mass balance relations with the acceptance of relations (15)-(17) eqn (18) and (19) can be derived. These are needed to calculate emission factors for i-th pollutant. The meaning of symbols used in these eqns is in respect of previous signs in eqns (9)-(17). WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

348 Air Pollution XIX nd  NF





a  1    1  YOd2



YOd2  y Od 2  1  YOd2



 YOd y   1  2 2  2 d CO



 yd d   y NO  NO  a  1     1  a  YOd2  y Fd 2  2 





(18)

3.2.2 Molar fraction of unmonitored components Because the concentration of water vapor and CO2 in combustion gases were not monitored, it was necessary to express the molar fractions from the mass balance. After adjustment formula (20) was obtained for the molar fraction of water and formula (21) for the molar fraction of carbon dioxide.



1   1 YOd2

n d  NVd 



1   1 y

y Hd 2O 

 2



y yd yd  CO  1       NO  1   y Fd 2 2 2





YOd2  y Od 2  1  YOd2  

(19)

d d  YOd  y CO y NO d  1  2   y NO   YOd2  y Fd 2 2  2  2 (20) 1    1  YOd2





 YOd2  1  YOd2  d YOd2  y Od 2    1  YOd2    y CO  1  2  2    d 1    1  YO2



d y CO 2

d O2



d NO2







 yd d   y NO  NO  YOd2  y Fd 2  2 (21) 

3.2.3 Calculation of fuel composition Stoichiometric coefficients in eqn (8) can be determined if the relative proportion i of each i-th fuel components is known. It is clear that relations (22) and (23) stand good:



 H  AC  C  AH

(22)



 O  AC  C  AO

(23)

It is obvious that the molecular weight of fuel F related to one carbon atom is given by eqn (24):

 F  AC    AH    AO  100  AC  C 1

(24)

where C, H and O means a relative content of carbon, hydrogen and oxygen in the fuel, AC, AH and AO [g mol-1] corresponding molar atomic weights and constants β and  have the same meaning as in eqn (15). WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

349

Using eqn (22) and (23) was derived the formula (25) for constant  and after substitution into relation (15) the eqn (26) for constant :

 O   AC    AH  100   O  AO

(25)

  O   AC    AH   4 2  100   O   AO

(26)

 

After substituting from relations (22) and (23) into formula (24) applies for

F eqn (27) from where it is possible easy to directly express stoichiometric coefficients a, b, c of eqn (8).

F 

M F 100   AC    AH   a 100   O

(27)

3.2.4 Calculation of emission factors Emission factors for the i-th contaminant were calculated according to eqn (28) which was obtained by substitution of nd × (NF)-1 from relation (18) in eqn (1) and by application of eqn (27) for F = MF × a-1. y id  Ef

i m







YOd2  y Od 2  1  YOd2 

Mi

F

d  YOd y CO  1  2 2  2





 1    1  YOd2



 yd d   y NO  NO  a  1     1  a  YOd2  y Fd 2  2 





(28)

Calculated values of emission factors under varying conditions of engine operation in the use of diesel NM 54 with or without additives are presented in table 2. From there it is obvious that the application of additives compared to clear diesel fuel amounted total decrease in emissions of CxHy about 12% and NOx emissions of about 8.5%, while emissions of CO notably increased by approximately 22.5%. Differences of emission factors of CO2 were for both alternatives statistically insignificant. Summary reduction of CO2 emissions after applying additives by about 0.3% roughly correlates with the slightly (1.5%) reduced fuel consumption during the entire measurement. The actual value of fuel consumption depends on engine operating conditions. Fuel savings of a maximum about 3% were observed in the application of additives when reduced torque TMR890; 1030 N m, engine power P186; 214 kW, and revolutions f1800; 2200 min-1. Due to the significant increase in concentrations CO in the combustion products the carbon balance of the burnt process varies within the measurement errors. Oxonica company, engaged in long-term tests in urban and suburban transport with an additive EnviroxTM, advertises, depending on operating conditions and type of engine in the application of an additive, reduction of CxHy emissions within 6-14%, 1-7% of CO and NOx by 11%. Reduction of CO2 WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

350 Air Pollution XIX Table 2:

Parameters of the engine and emission factors for selected pollutants with and without additive.

Test ML

f

No

TMR

TCG TEP

P

FC

-1 [%] [min-1] [N m] [K] [K] [kW] [kg kWh ]

C Hy

Ef m x

[g kg-1]

Ef mCO Ef mNO x Ef mCO 2 [g kg-1]

[g kg-1]

[g kg-1]

NM-54 1

100.0 2200

891.0 793 853 205.3

0.244

2.38

23.01

49.80

3 084

2

100.0 1399 1100.0 769 828 161.2

0.213

2.17

24.45

31.86

3 082

3

60.1

1803

993.0 773 833 187.5

0.226

2.42

27.66

48.41

3 076

4

54.4

1799

821.6 663 720 154.8

0.220

2.89

26.40

56.63

3 077

5

50.0

1799

616.1 593 629 116.1

0.231

3.97

26.97

56.83

3 073

6

46.8

1800

408.4 527 550 77.0

0.264

5.64

28.78

51.43

3 065

7

100.0 2199

932.3 764 821 214.7

0.235

3.01

27.77

36.18

3 074

48.98

41.81

3 036

NM-54 + EnviroxTM 1

100.0 2201

927.8 780 840 213.8

0.236

4.63

2

100.0 1399 1109.2 771 836 162.5

0.214

3.29

35.01

34.60

3 062

3

63.4

1799 1022.3 767 826 192.6

0.219

2.01

26.62

43.21

3 079

4

57.2

1799

817.7 673 721 154.0

0.219

1.54

23.16

49.09

3 086

5

53.3

1801

616.3 595 633 116.2

0.227

1.90

22.98

47.81

3 085

6

50.3

1801

408.0 520 549 76.9

0.260

3.17

25.48

44.35

3 077

7

100.0 2201

934.6 772 832 215.4

0.233

3.25

45.17

45.81

3 046

emissions should be in accordance with a reduction in fuel consumption, which is declared by the company to be in an interval of 5-12% [12]. ML [%] motor load, f [min-1] engine revolutions, TMR [N m] reduced torque, TCG [K] temperature of combustion products, TEP temperature of exhaust pipe, P [kW] engine power, FC [kg kWh-1] fuel consumption, Efmi [g kg-1] emission factor for the i-th contaminant. Results obtained by us are in compliance with data reported by the Oxonica company solely for CxHy and NOx emissions however for contaminants CO2 and CO are markedly different. The fact that there is no declared reduction of fuel consumption and hence reduction of CO2 and that CO emission factor even strongly increased may be partially explained by short-term sampling after the addition of EnviroxTM into diesel fuel. Moreover, increased production of CO could have been caused by oxidation of carbon residue in the engine. Also it could have been caused by an older type of engine with production year 1986, in which the mentioned effect does not sufficiently appear. For these reasons, there are planned additional verification tests for the newer engine type with a sampling period of at least 200 engine hours of the engine operation with fuel containing the investigated additive. At the same time the dependence of emission factor values were monitored as a function of TMR reduced torque, engine power P and engine revolutions f. According to theoretical assumptions, the majority of emissions after an initial WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

351

slight increase tend to decrease with increasing TMR and reach a minimum of around 180 kW engine power. The example in fig. 2 illustrates the graphical dependence of the emission factor values for NOx as a function of TMR. Format of the trend line was evaluated by linear regression second degree polynomial with the corresponding regression equation with reliability value R and they are included within the graph. For yet unknown reasons the maximum of CO2 emissions was observed in engine power around 180 kW, when the fuel consumption is on the minimal.

EfmNOx

[g kg‐1] 60 40 20 0

NM‐54 y = ‐0.0001x2 + 0.138x + 13.667 R = 0.842 NM‐54 + adidivum y = ‐0.00007x2 + 0.098x + 15.931 R = 0.920

300.0

500.0 NM‐54

Figure 2:

700.0

900.0

1100.0 TMR[N m]

NM‐54 + aditivum

The dependence of emission factor Ef mNO2 on reduced torque TMR for both test alternatives.

Also decline in value of emission factors at the engine revolutions of f  1900 min-1 is consistent with the theory because these conditions lead to efficient use of fuel. In contrast to this theory is the growth in CO2 emissions when applying fuel additives, the increase in CxHy emissions, if the clear diesel fuel was used and NOx emissions in both alternatives of test in discussed area of engine revolutions. Therefore it will be meaningful to verify these gaps with additional planned experiments under conditions listed above.

4 Conclusions A methodology was developed for measuring and calculating emissions determination of CxHy, NOx, CO and CO2 in the engine exhaust. With its use it was found that the additive EnviroxTM based on dispersed nanoparticles of CeO2 reduces the value of emission factors for CxHy by approximately 12% and NOx by around 8.5%. Nevertheless the reduction of CO2 corresponding with lower fuel consumption in the range 5–12% declared by Oxonica company could not be established. Maximum fuel savings of about 3% were found out only WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

352 Air Pollution XIX under optimum conditions of engine operation with a corresponding decrease in CO2 emissions by only circa 1%. In contrast with announced lower amount of CO emissions an increase of almost 23% was observed. At the same time was monitored dependence of emission factors on reduced torque, engine power and engine revolutions. With some exceptions the referred functions were in accordance with theoretical expectations. The divergence between CO2 a CO emissions production and the data stated by Oxonica Company as well as data collected in some dependency of emission factors on selected engine characteristics will be necessary to verify with further tests on newer type of engine and after sufficiently long period of engine operation with the addition of EnviroxTM fuel efficiency additive.

References [1] Bozek, F. et al., Genotoxic Risk for Population in Vicinity of Traffic Communication Caused by PAHs Emissions, Proc. of the Conf. Transactions on Environment and Development, 6(3), pp. 186-195, 2010. [2] Adamec, V. et al., Traffic, Health and Environment. Grada Publishing, a.s.: Prague, p. 58, 2008. ISBN 978-80-247-2156-9. [3] Peters A. et al. Associations between Mortality and Air Pollution in Central Europe. Environ. Health Perspect. 108, pp. 283-287, 2000. [4] Knecht, W. Diesel Engine Development in View of Reduced Emission Standards. Energy, 33(2), pp. 264-271, 2008. [5] Clerc, J. C. Catalytic Diesel Exhaust After-treatment. Applied Catalysis B: Environmental, 10(1-3), pp. 99-115, 1996. [6] Pandey, R. K. et al. Automobile Emission Reduction and Environmental Protection through Use of Green Renewable Fuel. Journal of Water, Energy and Environment, 7, pp. 65-70, 2010. [7] Lappas, A. A. et. al. Production of Low Aromatics and Low Sulphur Diesel in a Hydrodesulphurization Pilot Plant Unit. Global Nest: the International Journal, 1(1), pp. 15-22, 1999. [8] Gürü, M. et al. Improvement of Diesel Fuel Properties by Using Additives. Energy Conversion and Management, 43(8), pp. 1021-1025, 2002. [9] Dzung, H. M., Thang, D. X. Estimation of Emission Factors of Air Pollutants from the Road Traffic in Ho Chi Minh City. Journal of Science, Earth Sciences, 24, pp. 184-192, 2008. [10] Oxonica. Technical note 1. Diesel Fuel Additives and EnviroxTM Fuel Efficiency Additive. Oxford: Oxonica, p. 3, 2008. [11] Wakefield, G. et al. Envirox™ Fuel-Borne Catalyst: Developing and Launching a Nano-Fuel Additive. Technology Analysis  Strategic Management, 20(1), pp. 127-136, 2008. [12] Attfield, M. Envirox™ Fuel-Borne Catalyst. Oxford: Oxonica, p. 49, 2009.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

353

Non-thermal plasma abatement of trichloroethylene with DC corona discharges A. M. Vandenbroucke1, A. Vanderstricht1, M. T. Nguyen Dinh2, J.-M. Giraudon2, R. Morent1, N. De Geyter1, J.-F. Lamonier2 & C. Leys1 1

Department of Applied Physics, Ghent University, Belgium Unité de Catalyse et Chimie du Solide, Université des Sciences et Technologies de Lille, France

2

Abstract The decomposition of trichloroethylene (TCE) in air by non-thermal plasma was investigated with a multi-pin-to-plate direct current (DC) discharge at atmospheric pressure and room temperature. The effects of various operating parameters on the removal efficiency (RE) were examined. The experiments indicated that for low energy densities higher removal could be obtained with positive corona. For negative corona and 10% relative humidity (RH) a maximum RE of 99.5% could be achieved at 1100 J L-1. Formation of byproducts was qualitatively analyzed in detail with FT-IR spectroscopy and mass spectrometry. Detected by-products for negative corona operated at 300 J L-1 and 10% RH include dichloroacetylchloride, trichloroacetaldehyde, phosgene, ozone, HCl, Cl2, CO and CO2. The highest RE for TCE was achieved with a relative humidity of 19%. Keywords: non-thermal plasma, volatile organic compounds, trichloroethylene, by-products.

1 Introduction Many organic solvents used in metal, semiconductor and chemical industry have good physical and chemical properties and are therefore useful during the manufacturing process. Despite their good commercial value, most of them have a high volatility causing them to easily evaporate at ambient conditions. As a result the process waste gases are frequently contaminated with volatile organic WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110331

354 Air Pollution XIX compounds (VOCs). Trichloroethylene (TCE) is an example of such a VOC that is widely used for degreasing metal parts and semiconductors. It is also an ingredient in paint removers, adhesives and spot removers. Inhalation of TCE can however cause headaches, lung irritation and dizziness. In more severe cases nerve, kidney and liver damage have been reported. The International Agency for Research on Cancer (IARC) has determined that TCE is “probably carcinogenic”, proving its adverse effects to humans. Due to limitations of traditional methods (incineration, catalytic oxidation and adsorption) for the abatement of low concentrations of VOCs, non-thermal plasma (NTP) generated in atmospheric pressure discharges have been investigated as an energy and cost effective alternative [1–3]. In a NTP, the energy delivered to the system is almost entirely used to accelerate plasma electrons instead of heating up the gas flow like is the case for incineration (750–1150 °C) and catalytic oxidation (250– 500°C). While the gas stream remains at room temperature, the highly energetic electrons collide with background molecules (e.g. N2, O2, H2O) with formation of active plasma species (ions, radicals,...). These latter species are responsible for the oxidative abatement of VOCs. Kim has reviewed NTP techniques for the destruction of air pollutants [4]. In this work, a multi-pin-to-plate electrode configuration is used to generate a DC corona or glow discharge for the abatement of TCE. This electrode geometry was developed by the group of Akishev et al. [5–8] and was successfully tested for the removal of SO2 and NOx [9]. Vertriest et al. applied a large scale laboratory multi-pin-to-plate negative glow discharge for the removal of VOCs and discovered that molecules containing a double carbon bond have the lowest energy requirement for decomposition [2]. A recent review [10] summarizes DCexcited NTPs for VOC abatement. In the present study, a DC corona/glow discharge operated at atmospheric pressure was used for the oxidation of small amounts of TCE in air. The effect of the discharge polarity, initial TCE concentration and water content on the removal efficiency have been investigated. By-products from TCE abatement were qualitatively analyzed with Fourier transform infrared (FT-IR) spectroscopy and mass spectrometry (MS).

2 Experimental The experimental set-up is illustrated in Figure 1. A compressor delivers ambient air to an air dryer (MSC-Air, Model Compact 10) which controls the relative humidity at approximately 12%. The humidity of the air can be changed by turning off the air dryer or by using dry air from a cylinder (Air Liquide, Alphagaz 1). A bubbler system is used to set the TCE concentration in the gas stream. The initial TCE concentration is controlled by changing the flow rate (Bronkhorst, EL-FLOW) of air through a 0.5 L bottle containing liquid TCE. The bottle is kept at a temperature of 30.0 ± 0.5 °C in a thermostatic water bath. Experiments are carried out with a total air flow of 2 L min-1 and with varying inlet concentrations of TCE (420 – 4000 ppm). The multi-pin-to-plate plasma source is based on the concept of a negative DC glow discharge [7, 9]. It consists of 5 aligned pins which were positioned 28 mm of each other. The interWIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Figure 1:

355

Experimental set-up.

electrode gap in this configuration and the total reactor length are 9 mm and 200 mm respectively. The discharge is powered with a 30 kV/20 mA DC power supply and generated at atmospheric pressure and room temperature. For a total gas flow of 2 L min-1, the gas residence time in the plasma reactor is 2.16 seconds. A high voltage probe (Fluke 80K-40, division ratio 1/1000) measures the voltage applied to the inner electrode. The discharge current is determined by recording the voltage signal across a 100 Ω resistor placed in series between the counter electrode and the ground. Each pin is ballasted with a 1.5 MΩ resistor. The fraction of the total electrical power dissipated in these resistors amounts to 10% at most. The mass spectrometer used in these experiments is a Quadrupole MS (Omnistar GSD 301 O2 Pfeiffer Vacuum) equipped with a Faraday cup and a SEM ChanneltronTM) detector. Balzers Quadstar 200 (QMS 200) software (Pfeiffer Vacuum) is applied for collecting and displaying data. Only the maximum peak intensities and the corresponding m/z numbers are collected. The peaks are represented as bar lines over the corresponding m/z. The qualitative identification of the by-products is achieved using Scan Bargraph mode and a SEM voltage of 1600 V in the mass range 0 – 200 m/z and an acquisition rate of 1.5 scan min-1. A resolution of 50 with electron ionization of 60 V was adopted. The decomposition efficiency of TCE and the formation of by-products is also determined with a FT-IR spectrometer (Bruker, Vertex 70). The optical path length of the adjustable gas cell and the resolution of the spectrometer are set at 0.80 m and 4 cm-1 respectively. For each spectrum, 40 samples are averaged from 600 to 4800 cm-1. The mercury-cadmium-telluride (MCT) detector is nitrogen cooled and OPUS (Bruker) software is used to collect and analyze the obtained spectra. Formation of ozone is analyzed by an ozone monitor (Envitec, Model 450). The temperature and the water content of the inlet gas stream are measured with a climate meter (Testo 445). The RE (%) of TCE is calculated from: RE 

1‐

TCE in x 100                                            1 TCE out

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

356 Air Pollution XIX where TCE in is the concentration of TCE introduced in the reactor and TCE out is the TCE concentration in the effluent gas. The energy density (J L-1) is calculated as: ε 

P                                                                  2 Q

where P is the discharge power (W) and Q the gas flow rate (L s-1) through the plasma reactor. The selectivity to CO2 is defined as: SCO2

CO2  x 100                                          3 2 x  TCE in  x RE

where CO2 is the concentration of carbon dioxide detected in the effluent gas as a result of total TCE oxidation.

3 Results and discussion 3.1 Effect of discharge polarity on TCE removal efficiency Figure 2 presents the experimental results of the effect of the discharge polarity on the RE. The polarity of the DC corona discharge depends on whether the electrode pins are connected to the DC high voltage source (positive corona) or to the ground (negative corona). The mechanism for corona generation differs for both discharges [11]. The discharge of a positive corona consists of thin current filaments, also referred to as streamers. At a certain threshold voltage the streamer-like discharge transitions to an unstable spark discharge regime. For the negative corona, increasing the applied voltage results in the formation of a

Figure 2:

TCE removal efficiency versus energy density (temperature: 20°C, relative humidity: 10%, TCE inlet concentration: 1600 ppm).

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

357

steady state diffusive glow discharge and in a subsequent sparking [12]. By increasing the energy density, the RE for both negative and positive corona are increased (Figure 2). However, with positive corona and for energy densities up to 330 J L-1 higher efficiencies are obtained compared to negative corona. Overall, positive corona removes about 17% more TCE. At higher energy densities (> 330 J L-1) sparking occurs for the positive corona and a maximum RE of 80% is reached. The higher removal obtained with positive corona can be attributed to the fact that streamer discharges have a higher efficiency in the generation of chemical active species that are responsible for the oxidation of TCE [13]. At high energy densities (> 500 J L-1) the negative corona shifts to the glow regime which gradually fills the entire gap with active plasma volume. In this mode, almost complete removal of TCE is established at 1100 J L-1. For both discharges (negative and positive corona) the selectivity to CO2 does not exceed 10%. However, higher CO2-selectivities are obtained with negative corona. 3.2 Identification of by-products For negative corona, by-products are qualitatively analyzed and identified with FT-IR and MS. Table 1 shows the by-products and their IR absorption bands that are detected with FT-IR spectroscopy at an energy density of 300 J L-1. Due to incomplete oxidation, TCE is partially converted to dichloroacetylchloride and phosgene. Hydrogen chloride, carbon monoxide and carbon dioxide are detected as complete oxidation products. Ozone, a typical by-product of NTP treatment, is mainly formed by a three body collision by the following reactions:

Table 1:

O2

+

e-

2O

+

e-

(4)

O

+

O2

O3

+

M

(5)

Assignments of the observed infrared frequencies of TCE oxidation (negative corona, energy density: 300 J L-1, relative humidity: 10%). Compound TCE (Cl2C=CHCl) Dichloroacetylchloride (Cl2CH-COCl) Phosgene (COCl2)

Ozone (O3) Hydrogen chloride (HCl) Carbon monoxide (CO) Carbon dioxide (CO2)

Infrared frequency (cm-1) 3169 (C-H); 3099(C-H); 1592, 1560 (C=C); 1253 (CH-Cl); 945, 784, 634 (C-Cl); 850 (C-H) 1822, 1788 (C=16O); 1232 (C-H); 1079, 991 (C-C); 800, 742 (C-Cl2); 1783 (C=18O) 1832 (C=16O); 1791(C=18O); 1682 [2(C-Cl2)]; 852, 662 (C-Cl2) 1047, 1029 (O=O=O) 3049 – 2732 (H-Cl) 2178 (C≡16O); 2114 (C≡18O) 2359 (16O=C=16O); 2343 (18O=C=18O)

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

358 Air Pollution XIX MS results additionally confirm the formation of trichloroacetaldehyde and chlorine. After continuous operation of the plasma system, the spherical segments of the anode surface were covered with a brown carbonaceous deposit. The deposit is however not analyzed in the present work. With positive corona and under identical experimental conditions similar by-products were identified. The by-product analysis indicates that TCE is only partially mineralized. This demonstrates one of the shortcomings of NTP which can be improved by catalyst hybridization [14-16]. 3.3 Effect of initial TCE concentration Figure 3 shows the effect of the initial TCE concentration on the RE for both negative and positive corona. Experimental results indicate that for negative corona the RE of TCE is little influenced by the initial concentration. For positive corona, the RE decreases as the initial concentration of TCE increases. The latter observation is consistent with other studies where the effect of the initial concentration is studied [17-19]. It is believed that higher initial TCE concentration results in less energy available for the decomposition of each TCE molecule which causes a decrease in RE [19].

Figure 3:

TCE removal efficiency versus energy density for different initial TCE concentrations ((a) negative corona, (b) positive corona, temperature: 20°C, relative humidity: 10%).

3.4 Effect of humidity The effect of humidity is of great interest for practical applications in industry since process gas consists of ambient air that usually contains water vapour. The influence of relative humidity (RH) on TCE removal is shown in Figure 4. For dry air and 12% RH the removal of TCE is comparable. However, at 19% RH the RE is improved with 15 to 20% over the range of energy densities tested. The water content of the waste gas has opposite effects on the chemical and physical nature of the plasma. Collisions between H2O and plasma species such as electrons, oxygen radicals and O(1D) are responsible for the production of OH radicals which act as highly reactive oxidizing agents for VOCs. However, when the relative humidity increases, the electron mean energy is lowered by the

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Figure 4:

359

TCE removal efficiency versus energy density for different relative humidities (negative corona, temperature: 20°C, TCE inlet concentration: 500 ppm).

electronegative character of water. The subsequent quenching effect of activated chemical species has an adverse effect on the RE [20]. The latter effect was not observed during the present study due to the limited water content of the gas stream. Guo et al. determined the concentration of OH radicals during the decomposition of toluene in air with a dielectric barrier discharge operated at atmospheric pressure and with varying RH. The highest toluene RE was achieved at a RH of 20% which corresponded with the maximum OH yield [21]. Although higher RH were not examined in this study, the RE reached a maximum value at RH of almost 20%.

4 Conclusions In the present study a DC multi-pin-to-plate plasma source was utilized to decompose dilute concentrations of TCE in air. This VOC is often used in industry to degrease metal and semiconductors. The polarity of the plasma discharge affected the removal efficiency. At low energy densities, higher RE were achieved with positive corona. Degradation products with negative corona include dichloroacetylchloride, trichloroacetaldehyde, phosgene, HCl, Cl2, CO and CO2. A maximum RE of 99% was achieved with negative corona at 1100 J.L-1 and a relative humidity of 19%.

References [1] Magureanu M., Mandache N.B. & Ruset C., Pulsed multipoint-to-plate corona discharge for VOC abatement. Journal of Advanced Oxidation Technologies, 7(2), pp. 128-132, 2004. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

360 Air Pollution XIX [2] Vertriest R., Morent R., Dewulf J., Leys C. & Van Langenhove H., Multipin-to-plate atmospheric glow discharge for the removal of volatile organic compounds in waste air. Plasma Sources Science & Technology, 12(3), pp. 412-416, 2003. [3] Agnihotri S., Cal M.P. & Prien J., Destruction of 1,1,1-trichloroethane using dielectric barrier discharge nonthermal plasma. Journal of Enironmental Engineering, 130(3), pp. 349-355, 2004. [4] Kim H.H., Nonthermal plasma processing for air-pollution control: A historical review, current issues, and future prospects. Plasma Processes and Polymers, 1(2), pp. 91-110, 2004. [5] Akishev Y., Deryugin A., Karal’nik V., Kochetov I., Napartovich A. & Trushkin N., Numerical simulation and experimental study of an atmospheric pressure direct-current glow discharge. Plasma Physics Reports, 20(6), pp. 511-524, 1994. [6] Akishev Y., Grushin M.E., Kochetov I., Napartovich A., Pan’kin M.V. & Trushkin N., Transition of a multipin negative corona in atmospheric air to a glow discharge. Plasma Physics Reports, 26(2), pp. 157-163, 2000. [7] Akishev Y., Goossens O., Callebaut T., Leys C., Napartovich A. & Trushkin N., The influence of electrode geometry and gas flow on coronato-glow and glow-to-spark threshold currents in air. Journal of Physics DApplied Physics, 34(18), pp. 2875-2882, 2001. [8] Callebaut T., Kochetov I., Akishev Y., Napartovich A. & Leys C, Numerical simulation and experimental study of the corona and glow regime of a negative pin-to-plate discharge in flowing ambient air. Plasma Sources Science & Technology, 13(2), pp. 245-250, 2004. [9] Akishev Y., Deryugin A., Kochetov I., Napartovich A. & Trushkin N., DC glow discharge in air flow at atmospheric pressure in connection with waste gases treatment. Journal of Physics D-Applied Physics, 26(10), pp. 16301637, 1993. [10] Morent R., Leys C., Dewulf J., Neirynck D., Van Durme J. & Van Langenhove H., DC-excited non-thermal plasmas for VOC abatement. Journal of Advanced Oxidation Technologies, 10(1), pp. 127-136, 2007. [11] Chang J.S., Lawless P.A. & Yamamoto T., Corona Discharge Processes. IEEE Transactions on Plasma Science, 19(6), pp. 1152-1166, 1991. [12] Akishev Y., Grushin M., Kochetov I., Karal’nik V., Napartovich A. & Trushkin N., Negative corona, glow and spark discharges in ambient air and transitions between them. Plasma Sources Science & Technology, 14(2), pp. S18-S25, 2005. [13] Van Durme J., Dewulf J., Sysmans W., Leys C. & Van Langenhove H., Abatement and degradation pathways of toluene in indoor air by positive corona discharge. Chemosphere, 68(10), pp. 1821-1829, 2007. [14] Magureanu M., Mandache N.B., Parvulescu V.I., Subrahmanyam C., Renken A. & Kiwi-Minsker L., Improved performance of non-thermal plasma reactor during decomposition of trichloroethylene: Optimization of the reactor geometry and introduction of catalytic electrode. Applied Catalysis B-Environmental, 74(3-4), pp. 270-277, 2007. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

361

[15] Harling A.M., Glover D.J., Whitehead J.C. & Zhang K., The role of ozone in the plasma-catalytic destruction of environmental pollutants. Applied Catalysis B-Environmental, 90(1-2), pp. 157-161, 2009. [16] Van Durme J., Dewulf J., Sysmans W., Leys C. & Van Langenhove H., Efficient toluene abatement in indoor air by a plasma catalytic hybrid system. Applied Catalysis B-Environmental, 74(1-2), pp. 161-169, 2007. [17] Magureanu M., Mandache N.B. & Parvulescu V.I., Chlorinated organic compounds decomposition in a dielectric barrier discharge. Plasma Chemistry and Plasma Processing, 27(6), pp. 679-690, 2007. [18] Kang H.C., Decomposition of chlorofluorocarbon by non-thermal plasma. Journal of Industrial and Engineering Chemistry, 8(5), pp. 488-492, 2002. [19] Chang M.B. & Yu S.J., An atmospheric-pressure plasma process for C2F6 removal. Environmental Science & Technology, 35(8), pp. 1587-1592, 2001. [20] Futamura S., Zhang A.H. & Yamamoto T., The dependence of nonthermal plasma behavior of VOCs on their chemical structures. Journal of Electrostatics, 42(1-2), pp. 51-62, 1997. [21] Liao X.B., Guo Y.F., He J.H., Ou W.J., Ye D.Q., Hydroxyl radicals formation in dielectric barrier discharge during decomposition of toluene. Plasma Chemistry and Plasma Processing, 30(6), pp. 841-853, 2010.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

This page intentionally left blank

Air Pollution XIX

363

Monitoring of atmospheric dust deposition by using a magnetic method A. Kapička, E. Petrovský & H. Grison Institute of Geophysics ASCR, Czech Republic

Abstract Several studies showed that atmospheric dust contains significant portions of minerals characterized by ferrimagnetic properties. These minerals, mostly iron oxides, can serve as tracers of industrial pollutants in soil layers. In our paper we have investigated magnetic properties of depth soil profiles from the Ore Mountains (Czech Republic), which belong to a highly contaminated, so-called Black Triangle in central Europe. Emissions are determined by considerable concentration of big sources of pollution (power plants burning fossil fuel, metallurgical and chemical industry). Increased values of magnetic susceptibility (25100  10 -5 SI) were clearly identified in the top-soil layers. Thermomagnetic analyses and SEM observation indicate that the accumulated anthropogenic ferrimagnetics dominate these layers. Magnetic enhancement is limited to depths of 47 cm below the soil surface, usually in F-H or on the top of Ah soil horizons; deeper soil horizons contain mainly magnetically weak materials and are characterized by much lower values of susceptibility (up to 20  10-5 SI). Significant magnetic parameters (e.g. Curie temperature TC) and SEM results of contaminated topsoils are comparable with magnetic parameters of atmospheric dust, collected (using high-volume samplers) at the same localities. Keywords: atmospheric dust, topsoil pollution, magnetic susceptibility.

1 Introduction Results of regular monitoring provide information on the temporal development of air pollution and spatial distribution of the pollutants concentration over the territory of Czech Republic. Distribution of the pollutants is very complex and depends upon several factors. Emissions from local heating, traffic and partly WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110341

364 Air Pollution XIX medium sized sources are of local extent, while those from major sources can be transported over long distances, often even outside the country territory. Dust particles are a significant threat to human health and, thus, represent major problem in air pollution. Industrial dust particles are emitted, among others, by combustion (of fossil fuel in stationary sources and fuel in traffic). About two thirds of PM10 and half of finer, more harmful PM2.5 fraction are produced by electric plants. Traffic is another significant source of these particles and, moreover, causes redistribution of already deposited particles back to the atmosphere. Generally, two different approaches are used to measure the amount of dust deposited on a surface; Determination of the quantity of dust deposited in terms of weight, or determination of the soiling of a surface, by a change in its properties. The former approach uses high-volume samplers. This method determines average dust concentrations and comprises the collection of dust by drawing a constant flow rate of ambient air through a collector. Data are usually collected over a 24-hour period and results are expressed in mass of dust per unit volume of air per 24 hrs. A selective inlet may be fitted to a high-volume sampler to restrict the particle size being sampled (for example, to ensure only PM10 particles are sampled). Measurement of magnetic properties of contaminated soil surface exemplifies second approach for monitoring of air pollution. Currently, ambient air pollution caused by suspended particles represents a major problem not only in the Czech Republic but almost throughout the whole of Europe. With regard to health risks, especially fine particles represent the most difficult challenge. Analyses of measured data show that the PM10 limit values are markedly exceeded in a number of sites throughout the country [3]. Contamination due to deposition of particulate matter shows significant variability within the Czech Republic. National parks, such as the Giant Mountains (NE Bohemia) and Šumava (SW Bohemia) belong to the cleanest areas. There are no major sources of atmospheric pollution in the surrounding area and relatively low concentrations of atmospheric dust are due to long-range transport. The average annual concentration of PM10 is in both the regions < 15 μg/m3, which is well below the allowance limit. On the other hand, the Ore Mountains region, which is part of the “black triangle” area [4], belongs to the most polluted parts of the Czech Republic. This area is on the junction between the Czech Republic, Germany and Poland and contains numerous sources of atmospheric pollution, such as major power plants burning high-sulphur brown coal, chemical industry, incinerators and big heat plants, intense automobile and rail traffic, deposits of power-plant ashes, waste dumps, deposits of overburdens, etc. The average annual concentration of PM10 (35  46 μg/m3) in the area of the Ore Mts. foredeep belongs to the highest within the Czech Republic. The situation on the top of the Ore Mts. is somewhat better, with an average annual concentration of PM10 at about 20 μg/m3 [3]. In this study, we demonstrate the application of a magnetic method to assess contamination of soils due to atmospheric deposition in the Ore Mountains region. Soil magnetometry has several advantages. It is fast and, contrary to WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

365

stationary monitoring stations, allows data acquisition at a large number of sites, thus enabling better delineation of areas with different amounts of deposited dust. Areas with higher imissions can be thus targeted for sampling for more detailed and standardized chemical analyses.

2 Magnetic study of soil contamination due to atmospheric dust deposition Deposited dust accumulates mostly in topsoils and sediments. However, in order to assess the contamination due to industrial activities, discrimination between anthropogenic and natural (weathering of geological basement, pedogenic processes, etc.) contributions is necessary. This can be achieved by analyzing the magnetic properties of contaminated samples collected in the areas of concern. Magnetic susceptibility (k) measured in a low magnetic field is one of the most important parameters used in environmental magnetism. In the case of soil samples, this integral parameter represents combined contributions of diamagnetic (iron-free silicates and carbonates), paramagnetic (silicates containing Fe and Mn), antiferromagnetic (e.g. hematite) and ferrimagnetic (e.g. maghemite and magnetite) minerals. Magnetic susceptibility is composition and concentration dependent parameter and primarily can be considered to reflect concentration of (strongly magnetic) ferrimagnetic substances. Interpretation of magnetic data depends on the samples examined. It is relatively easy and straightforward in the case of passive biomonitors, such as tree leaves, needles, or peat bogs, where atmospheric dust particles are deposited on (magnetically weak) diamagnetic substances (e.g. [5]). However, availability of these collectors is rather limited to certain localities. The other disadvantage is low concentration of the deposited material and, thus, the need for highlysensitive instruments, which are not available for the in-situ monitoring. Contrary to that, the soil surface represents the most obvious trap to atmospheric fallout. Consequently, topsoil measurements are often used in assessing the soil contamination due to atmospheric deposition of dust particles (e.g. [6]). Fig. 1 shows typical vertical distribution of magnetic susceptibility in forest soil from localities distributed evenly over the territory of the Ore Mountains, Czech Republic. Low-field magnetic susceptibility was measured on depth soil cores, using a Bartington MS2C sensor. A significant increase of magnetic susceptibility in the uppermost (4  7cm), organic L-F and Ah horizons is commonly observed. In deeper soil layers, magnetic susceptibility is considerably less. Relative contribution of ferrimagnets of natural origin in topsoils is mainly due to two factors: weathering of geological basement and neoformation of ferrimagnetic minerals during pedogenic processes. Mineral composition of underlying rocks is the primary factor influencing mineral populations in the developed soils. However, vertical distribution of magnetic susceptibility can easily identify sites with significant geological contribution to magnetic enhancement of topsoils [7]. The other contribution results from oxidationreduction processes in soils, transforming magnetically weak low-crystalline FeWIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

366 Air Pollution XIX mag susc [10-5 SI] 0

20

40

60

80

100

0 5 10 KH-5

depth [mm]

15

KH-6 KH-7

20

KH-8 25

KH-10 KH-12

30

KH-13 35

Litv 2

40 45

Figure 1:

Typical depth soil profiles of magnetic susceptibility from Ore Mountains region in Czech Republic.

oxides and Fe-hydroxides to magnetically strong maghemite/magnetite. However, these particles of pedogenic origin are prevailingly in a form of very fine, superparamagnetic (SP) or single-domain (SD) grains and their relative significance can be estimated using frequency-dependent magnetic susceptibility kFD [8]. Atmospheric dust of industrial origin contains Fe-oxides, magnetite (Fe3O4), maghemite (γ-Fe2O3) and hematite (-Fe2O3). These are most often produced during combustion of fossil fuel (in, e.g., power plants) by decomposition of pyrite or Fe-rich clay minerals present in coal, followed by high-temperature oxidation of iron [2]. Also, emissions from industrial units such as steel and cement works, as well as traffic represent significant sources of anthropogenic ferrimagnets (e.g. [9–11]). Magnetic properties of industrially derived ferrimagnets are different from those of natural origin. In terms of morphology, they are typically of spherical shape, with Fe-oxides sintered on Al-Si phase. Their typical size varies between about 2 and 50 μm and from magnetic point of view multi-domain (MD) structures prevail [12].

3 Results and discussion Increased concentration of atmospherically deposited ferrimagnetics of anthropogenic origin was found in the topsoil layers. However, we still have to prove the presence of ferrimagnetics in the collected PM10 samples and to show that this substance has magnetic properties similar to those of the topsoil layer. Therefore, thermomagnetic analyses (temperature dependence of magnetic susceptibility) were carried out using KLY-4S Kappabridge equipped with CS-3 WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

367

furnace (AGICO, Brno, Czech Republic). Both samples of quartz microfiber filters with PM10 and magnetic extracts of soil samples were measured at temperatures from 20 to 700ºC in order to detect temperatures of magnetic phase transitions of major ferrimagnetic minerals. Samples of the L-F soil horizons were dominated by magnetite with Curie temperature of about 580ºC (Fig. 2). Magnetic composition of bottom soil layers is always more complex, showing a whole sequence of transformations of magnetically weak minerals at elevated temperatures. In the case of PM10 samples, despite very minute concentration, magnetite could be identified as well. This finding is of crucial importance for justification of magnetic monitoring of atmospherically deposited dust using soil magnetometry in the region of the Ore Mountains. 4.0E-06 3.5E-06

mag susc [m3/kg]

3.0E-06 2.5E-06

KH(7), KH(11) - topsoils 2.0E-06 1.5E-06

KH(7), KH(11) - subsoils

1.0E-06 5.0E-07 0.0E+00 0

100

200

300

400

500

600

700

o

T [ C]

Figure 2:

Temperature dependence of magnetic susceptibility for topsoil and subsoil samples from different localities in the Ore Mts.

Anthropogenic ferrimagnetic particles in atmospheric dust have typical morphology. Those resulting from combustion processes are typically spherules with the size from 5 to about 50 μm (e.g. [12]). We examined using SEM magnetic extract of the topsoil samples the as well as PM10 samples. Typical result is depicted in Fig. 3. Spherical Fe-rich particles, with variable grain size, were found in magnetic extracts of topsoil samples. Despite the fact that the PM10 samples contain only smaller particles, below 10 μm, similar particles were also identified. Deeper mineral soil horizons were free of these Fe-rich spherules, which are presumably of industrial origin. Monitoring of PM10 is performed using stationary high-volume samplers located at the Ore Mountains piedmont at the altitude of 300 m a.s.l. (Lom) and close to the summit (Rudolice, 780 m a.s.l.). On both sites, sampling is during 24 hrs and the air flow through the sampler is about 720 m3/day. Long-term monitoring clearly showed that the highest concentrations of atmospheric dust in WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

368 Air Pollution XIX Table 1:

PM10 concentrations in sampler filters from two localities in the Ore Mts.

Locality

Date

Rudolice (780 m) Lom (300 m) Rudolice (780 m) Lom (300 m) Rudolice (780 m) Lom (300 m)

Figure 3:

Oct. 24, 2008 Oct. 25, 2008 Oct. 26, 2008

Conc. [μg m-3] 17 68 23 67 17 47

Date Jan.13, 2009 Jan.15, 2009 Jan.17, 2009

Conc. [μg.m-3] 29 138 8 225 17 92

SEM of ferrimagnetic spherules in topsoil magnetic extract.

this region are in the winter period [3]. Therefore, for demonstration of PM10 concentration variations with different altitude, typical values obtained in October 2008 and January 2009 were used. PM10 concentration sampled on 24, 25, 26 October 2008 and 13, 15, 17 January 2009 in both the localities are listed in Table 1. High concentrations in the piedmont are attributed to major sources of air pollution in the close neighbourhood (power plants, smelters). Comparison with the summit concentrations suggests limited transport of PM10 to higher altitudes.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

369

-3

k [10 SI] -100

0,2

0,4

0,6

0,8

1

1,2

0

depth [mm]

815m a.s.l.

443m a.s.l

100

851m a.s.l. 200

300

400

Figure 4:

Depth soil profiles of magnetic susceptibility at sites with different altitude (443 m, 815 m and 851 m a.s.l) in the Ore Mts.

Vertical distribution of magnetic susceptibility was measured to depths of 30  40 cm on forest soils along the whole altitude profile from piedmont to the summit using MAGPROX SM400 kappameter (ZHInstruments, Brno, Czech Republic [13]). Profiles were measured in the altitudes of 443 m, 815 m and 851 m a.s.l. In all the cases the profiles show clearly magnetically enhanced superficial (L-F) pedozones (Fig. 4). At the same time, magnetic susceptibility in deeper mineral horizons is much lower and practically constant. This pattern suggests dominance of atmospherically deposited anthropogenic ferrimagnetics and the negligible effect of natural minerals of lithogenic origin (e.g. [14]). The maximum value of magnetic susceptibility along the depth profiles seems to depend upon the altitude, the higher the altitude, the lower magnetic enhancement in the topsoil layer. This finding reflects qualitatively the differences in PM10 concentrations in the piedmont and summit sites.

4 Conclusions Our results prove that measurements of topsoil magnetic susceptibility can help in assessing the spatial distribution of soil contamination due to atmospheric deposition of pollutants over the area of the Ore Mts. Topsoil horizons are magnetically enhanced due to the presence of a magnetic mineral phase which shows similar characteristics as that found in the collected PM10 samples, and which is presumably of industrial origin. Magnetic susceptibility of the enhanced topsoil layer shows qualitatively similar altitude dependence as concentration of atmospheric PM10. If certain rules are obeyed (estimating the significance of ferrimagnetic particles of lithogenic and/or pedogenic origin), a magnetic method can be used for relatively fast and cheap assessment of soil contamination due to atmospheric deposition of pollutants. Contrary to routinely used stationary PM10 WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

370 Air Pollution XIX samplers, soil magnetometry allows acquisition of large datasets, covering the area in concern with much larger density. Hence, it allows delineation of significantly contaminated areas for more targeted sampling for detailed chemical analyses, which are more expensive and time consuming. However, soil magnetometry can only be considered as a proxy and approximate method, which is site specific and has to be always calibrated using available environmental data.

Acknowledgements This study was supported by the Grant Agency of the Czech Republic through grant No. 205/07/0941 and the Grant Agency ASCR through project No. IAA300120701

References [1] Kapička, A., Jordanova, N., Petrovský, E. & Ustjak, S., Magnetic stability of power-plant fly-ashes in different soil solutions. Phys. Chem. Earth (A), 25(5), pp. 431-436, 2000. [2] Flanders, P.J., Collection, measurements and analysis of airborne magnetic particulates from pollution in the environment. J. Appl. Phys., 75, pp. 59315936, 1994. [3] Statistical environmental yearbook of the Czech Republic 2006. Ministry of the Environment of the Czech Republic (MZP CR), Praha, 2007. [4] Hykyšová, S. & Brejcha, J., Monitoring of PM10 air pollution in small settlements close to opencast mines in the North-Bohemian Brown Coal Basin. WIT Transactions on Ecology and the Environment, 123, pp. 387398, 2009. [5] Zhang, C.X., Huang, B.C., Piper, J.D.A. & Luo, R.S., Biomonitoring of atmospheric particulate matter using magnetic properties of Salix matsudana tree ring cores. Sci. Tot. Environ., 393, pp. 177-190, 2008. [6] Petrovský, E., Kapička, A., Jordanova, N., Knab, M. & Hoffmann, V., Magnetic susceptibility - a proxy method of estimating increased pollution of different environmental systems. Environ. Geol., 39(3-4), pp. 312-318, 2000. [7] Kapička, A., Jordanova, N., Petrovský, E. & Podrázský, V., Magnetic study of weakly contaminated forest soils. Water Air Soil Pollut., 148(1-4), pp. 31-44, 2003. [8] Maher, B.A. & Taylor, R.M., Formation of ultrafine-grained magnetite in soils. Nature, 336, pp. 368-370, 1988. [9] Zheng, Y. & Zhang, S.H., Magnetic properties of street dust and topsoil in Beijing and its environmental implications. Chinese Sci. Bull., 53, pp. 408417, 2008. [10] Hoffmann, V., Knab, M. & Appel, E., Magnetic susceptibility mapping of roadside pollution. J. Geochem. Explor., 66, pp. 313-326, 1999.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

371

[11] Kim, W., Doh, S.J., Park, Y.H. & Yun, S.T., Two-year magnetic monitoring in conjunction with geochemical and electron microscopic data of roadside dust in Seoul, Korea. Atmos. Environ., 41, 7627-7641, 2007. [12] Strzyszcz, Z., Magiera, T. & Heller, F., The influence of industrial immisions on the magnetic susceptibility of soils in Upper Silesia. Stud. Geophys. Geod., 40, pp. 276-286, 1996. [13] Petrovský, E., Hůlka, Z. & Kapička, A., A new tool for in situ measurements of the vertical distribution of magnetic susceptibility in soils as basis for mapping deposited dust. Environ. Technol., 25, pp. 1021-1029, 2004. [14] Magiera, T., Strzyszcz, Z., Kapička, A., Petrovský, E. & MAGPROX Team, Discrimination of lithogenic and anthropogenic influences on topsoil magnetic susceptibility in Central Europe. Geoderma, 130, pp. 299-311, 2006.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

This page intentionally left blank

Air Pollution XIX

373

Improving car environmental and operational characteristics using a multifunctional fuel additive E. Magaril Department of Economics and Organization of Chemical Industries, Ural Federal University, Russia

Abstract Vehicle modernization has been developed towards the growing necessities of speed, power, efficiency, ergonomics, and design etc. The requirements, nowadays of environmental safety and operational efficiency of vehicles are being brought to the forefront. The aim of this work is to increase efficiency and reduce the harmful environmental impact of automobile transport by improving the quality of fuels during its operation. The improvement of the quality of fuels by means of highly-effective additives is the most rapidly implemented and lowcost method. According to the settled requirements of the properties of additives and the analysis of the catalytic and physicochemical properties of the substances, the universal content of the additive to gasoline and diesel fuel was found and the technology of its production was proposed. In addition, the additive was thoroughly tested in the laboratory, test bench, traffic operation and experimental-industrial checkout. It was found that the application of the additive in minor quantities significantly improves operational and environmental properties of fuels and engine characteristics. Keywords: multifunctional additive, carbonization suppression, fuel consumption, environmental and operational characteristics of vehicles.

1 Introduction Transport with its consumption of engine fuel of about 2 billion tons per year is one of the most important factors which determine the state of the world economy and, as a consequence, geopolitics. The most important problems facing humanity in this connection are the problems of environmental WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110351

374 Air Pollution XIX deterioration and exhaustion of the petroleum and natural gas resources. Environmental safety and efficiency of vehicle operation are influenced by many factors such as the size of the vehicle fleet, the quality of car maintenance, the fuel’s characteristics, the road network, traffic management, environmental conditions and regulations. The fastest and least expensive way to improve the environmental and operational characteristics of vehicles of the current state of the vehicle fleet and the quality of engine fuels is the introduction of highly effective fuel additives affecting the individual fuel characteristics or its quality in common (multifunctional additives). It should be noted that the current state of the petroleum-refining methods cannot provide the required level of such fuel characteristics as its cleaning and lubricating properties, which can be easily improved by using the appropriate additives. At the present time there are about 2.500 brands of fuel additives on the world market and only in the United States, their consumption exceeds considerably 100 tons a year. In Russia, the improvement of the quality of engine fuels which influences the operational and environmental performance of vehicles has a special significance under the conditions of the rapid growth of the number of vehicles simultaneously with the steady quality of fuels or even its worsening. The decision taken in this regard should be based on scientifically sound conclusions about the influence of physicochemical properties of fuels and fuel additives on the operating and environmental characteristics of vehicles.

2 The requirements for the multifunctional fuel additive Based on the theoretical investigations of the influence of fuel quality on the operating and environmental characteristics of vehicles, [1], the data of the catalytic and physicochemical properties of substances we have formulated the requirements for the properties of additives which can improve vehicles’ environmental and operational characteristics. 2.1

The requirements for carbonization suppression in engines

The maximum temperature reached during the combustion of fuel-air mixture in an engine, is determined by other equal conditions by heat transfer to the cylinder walls, which is mainly determined by their temperature, [1, 2]. Due to the high thermal conductivity of metal, the wall temperature does not differ greatly from the cooling water temperature. However, the carbon deposit formation during the operation reduces significantly the heat transfer through the walls, because it’s coefficient of thermal conductivity in 1-2103 times and heat capacity about 3 times lower than that of metal. Carbon deposit layer even of low thickness causes an increase of the temperature of engine’s walls because of its insulating properties. Our calculations showed that after the carbon deposition an increase of the wall’s temperature by 200 degrees and more takes place depending on the thickness of the layer, the fraction of carbon in the amorphous state in it, as well as a coolant temperature. It reduces of the heat radiation from the volume of the chamber through the walls by 5–10% or more and leads to the WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

375

increase of peak temperature of the combustion chamber. When the process of carbonization in the combustion chamber reaches certain equilibrium in its thickness, to provide operation of a gasoline engine without detonation in practice the octane number of gasoline usually should be increased by 10 points. In addition, carbon deposit, as a reaction-active substance, inclines to selfignition at 330-360º. Scorching during the burning of fuel mixture and emission of the products of combustion, carbon deposit, due to its low thermal conductivity, keeps the heat and propensity to interact with oxygen in the inlet fuel with the formation of “hot spots” causing the initiation of combustion before sparking. The “surface ignition” appears. Premature combustion in the compression stroke, in turn, leads to the significant increase in energy consumption for the compression, the increase of fuel specific consumption and the decrease of engine power. The rate of heating of the working mixture in the compression stroke grows as a result of the surface ignition that gives additional opportunity for detonation in gasoline engines. In addition, the carbonization on the working surfaces of the engine increases significantly the friction that leads to the growth of the fuel specific consumption and the reduction of engine power. Elimination of carbonization, therefore, because of mitigation of the temperature regime in the engine, will reduce the requirements for the octane number of gasoline used up to 10 points, and provide the increase of the power of gasoline and diesel engines and the reduction of fuel specific consumption. Furthermore, the removal of carbon deposit in gasoline engines is accompanied by the almost complete elimination of emission of polycyclic aromatic hydrocarbons, precursors of carbon deposit, including benzo(a)pyrene, one of the strongest carcinogens. It should be also noted that the reduction of the fuel specific consumption will lead to a corresponding reduction of emission and the amount of “greenhouse gases” (CO2, CH4, etc.). The calculated dependences of equilibrium constants of the reactions determining the formation of oxides of nitrogen and carbon (NO and CO) in the engine on temperature, [1, 2] show that the decrease of peak temperature in the combustion chamber due to enhancement of heat transfer through the walls with carbon deposit elimination in its turn will lead to a considerable decrease of nitrogen and carbon’s oxides emission. The amount of carbon deposit on the wall’s surface is determined by ratio of the condensation rate of aromatic and unsaturated hydrocarbon in the wall boundary layer and the rate of oxidation and gasification of products of condensation (and their precursors) by oxygen, water and carbon dioxide. Therefore, the additive should catalyze effectively the reactions of oxidation and gasification. 2.2

The requirements for the decrease of unburned combustibles’ emission

The additive should decrease the amount of unburned combustibles in outlet gases – hydrocarbons and their oxy derivatives. As such substances are generated mainly in the wall boundary layer in the low temperature area, the WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

376 Air Pollution XIX additive should catalyze full oxidation and gasification (under the lack of oxygen) of hydrocarbons and their oxy derivatives. 2.3

The requirements for the decrease of nitrogen and carbon’s oxides emission

The additive should provide decomposition and (or) reduction of nitrogen oxide and oxidation of carbon monoxide, in other words, it should catalyze the reactions: 2NO → N2 + O2 2NO + 2CO → N2 + 2CO2 2CO + O2 → 2CO2 2.4

The requirements for the decrease of soot emission

The additive should suppress soot production. In principle, it is possible by means of facilitation of nucleation of soot production that leads to increasing of its dispersion. The rate of soot burning in a flame front is proportional to the specific surface of soot particles, therefore increasing of its dispersion leads to increasing of the degree of its burning. The catalysis of carbon burning gives another possibility to the growth of the degree of soot burning in the flame front. 2.5

The requirements for the enhancement of the detergent properties of fuels

To ensure the cleanness of the fuel system of vehicles, as well as undisturbed operation of engines with fuel injection the high detergent properties of fuels are required, which cannot be reached by methods of oil refining but can be achieved easily by application of fuel additives. The additive should effectively improve the detergent properties of fuel. It is possible if the additive has a high surface activity, providing the tar solubilization or formation of the durable adsorption layers on the metal surface, preventing the deposition of tar. 2.6

The requirements for the increase of the cetane number of diesel fuel

The value of the cetane number of diesel fuel affects a number of operational parameters of diesel engines: the time of cold start, the rate of use of fuel, the efficiency of the engine. The additive should enhance the cetane number of diesel fuel, i.e. it should be a catalyst for ignition. 2.7

The requirement for the improvement of the inlet mixture formation in the engine

Our studies showed that the spraying of fuel of given quality can be improved by the introduction of surfactants into the fuel, which reduce surface tension of the fuel-air interface. Therefore the additive should possess the high surface activity.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

2.8

377

The requirement for the improvement of the low temperature properties of diesel fuel

The additive should decrease the limiting temperature of filterability of diesel fuel, which defines the possibility of its utilization under the low temperatures which is a great importance in Russia. 2.9

The requirement for the increase of the wear-resistant properties of fuel

The high pressure fuel pump of diesel engines undergoes wear of the pump internal elements. The anti wear properties of diesel fuel with reduced sulfur content in it decrease sharply, and can be improved only by introduction of additives which are the surface-active substances. Therefore, the additive should possess the surface activity and improve lubricating properties of diesel fuel. 2.10 The requirement for the solubility in the fuel The additive should form a true solution in the fuel under all conditions during its application. Otherwise, it will negatively affect the engine performance. 2.11 The requirement for the unavailability of additional pollutants The additive should not be the source of making new toxic substances which are not produced in fuel without additives. 2.12 The requirement for the accessibility and low cost of raw materials The additive should be synthesized from available materials and it should not affect significantly the cost of fuel. Based on the stated requirements and screening of the data on the catalytic properties of substances, the composition of oil-soluble multifunctional additive and technology of its production was developed, and its optimal concentration in fuel is 9.25 ppm for gasoline and 27.75 ppm for diesel fuel was found.

3 The testing results of the additive application The additive was thoroughly tested in the laboratory, test bench, traffic operation and experimental-industrial checkout. The studies of influence of multifunctional additive on the process of an engine’s carbonization showed that the additive concentration of 9.25·ppm in gasoline and 27.75 ppm in diesel fuel was enough to eliminate carbon deposits almost completely. In addition, a positive effect of the additive on the mixture formation in engines, the decrease of the probability of dripping on the walls with subsequent formation of carbon deposit, was found. The testing of the influence of the additive on the fuel consumption on a diesel engine of locomotive 2TE116 № 42 showed that the reduction of the fuel

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

378 Air Pollution XIX consumption after 9 hours of operation on fuel with the additive was 6.9%. The combustion chamber of one of the cylinders was unsealed before and after the testing run on fuel with the additive. It was found that amount of carbon deposit in the combustion chamber on the valves and piston crown decreased sharply (fig. 1).

a)

b) Figure 1:

The outward of piston’s crown of diesel engine of locomotive 2ТЭ116 №42: a) Before the additive application; b) After 9 hours of testing run on fuel with the additive.

The additive with the concentration of 27.75 ppm under on-peak load during the test of the diesel bench engine reduced the soot content by 50%. CO, CH and NO2 were decreased, respectively, by 14.6, 37.2 and 20.6%. On the fig. 2 the effect of the additive on the opacity of exhaust gases is presented. Introduction of the additive to the diesel fuel increases considerably the cetane number (fig. 3), improves the lubricating and low-temperature properties of fuel.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

379

Operation testing of the influence of the multifunctional additive on the gasoline specific consumption was carried on vehicles which had travelled 20000 km before the test. The test results are shown in fig. 4. Vehicles with gasoline engines had the octane number of fuel used decreased by 6-10 points after the additive usage.

Exhaust opacity by Hartridge, %

Exhaust opacity by Hartridge, %

60

Figure 2:

1 1680 min-1

40

2

20

0 20

40

60 80 Effective power, Ne, kW

40

100

1

30

2

2800 min-1

20 10 0 0

50

100

150

Effective power, Ne, kW

Influence of the additive on the soot emissions. (1- without the additive; 2 - with the additive.)

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

380 Air Pollution XIX

51 50 49 Cetane number

48 47 46 45 44 43 42 41 0

20

Figure 3:

40 60 80 Additive concentration C, mg/kg

120

Dependence of cetane number on additive concentration.

6.64 5.91 9.25 8.80 13.25 11.76 13.25 11.87 15.91 15.09 8.91 7.98 7.00 6.29 7.14 6.35 7.25 6.87

Daewoo Nexia Ford Focus Nissan Maxima Nissan Maxima VAZ 2121 GAZ 3102 GAZ 3102 VAZ 21099 VAZ 21099 0

5

10 15 20

specific gasoline consumption, l/100 km

Figure 4:

100

Daewoo Nexia

12.3

Ford Focus

4.9

Nissan Maxima

11.2

Nissan Maxima

10.4

VAZ 2121

5.2

GAZ 3102

10.4

GAZ 3102

10.2

VAZ 21099

10.9

VAZ 21099

5.2 0

5

10

15

specific consumption reduction, %

Influence of the additive on the specific consumption of gasoline. (□ - without the additive; ■ - with the additive.)

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

381

Technical checkout of the engines after the operation testing showed significant purification of spark plugs, valves and cylinder-piston group of the testing vehicles from carbon deposits. In fig. 5 pictures of the Nexia SOHC engines of UZ-Daewoo’s vehicles after applying of the additive are presented.

a)

b) Figure 5:

The outward of cylinder-piston group of Nexia SOHC engines of UZ-Daewoo’s vehicles after the testing run of 1935 km, a) controlling vehicle and b) testing vehicle after the run with the application of the additive.

The influence of the additive on the emissions of benzo(a)pyrene by vehicles with gasoline engines idling was investigated. It was found that the reduction of the emission of polycyclic aromatic hydrocarbons was  95% (table 1), which increases durability of the exhaust catalytic converters as well as has an Table 1:

The influence of the additive on the emissions of benzo(a)pyrene and other multiring aromatic hydrocarbon. Vehicle GAZ – 3102, idling.

Gasoline Without additive With additive Change, %

Content in burnt gases, mg/m3 Scope of multiring aromatic benzo(a)pyrene hydrocarbon lighter then benzo(a)pyrene 0.00023 0.00583 0.00001 -95.7

0.00030 -94.9

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

382 Air Pollution XIX enormous environmental concern. Bench tests of vehicles with gasoline engines has shown that the application of the additive reduces carbon oxide emissions by 15-30%, oxides of nitrogen – up to 26%, hydrocarbons by 8-35%. Cleaning the fuel system from tar deposits takes place as a result of the detergent effect of the additive. The additive has extremely high detergent properties, more efficient than ones of the best detergents (Paradine-50, MPA-85), available on the world market. With the introduction of the additive detergent properties of fuel increase in 2.33 times at a rate almost 20 times smaller than for commonly used additives. As the surfactant, the additive concentrates on the surface of gasoline decreasing the vapor pressure of gasoline and, as a result the loss from evaporation (fig. 6). The combination of the influence of additives on the vapor pressure and surface tension provides a valuable technical effect, the reduction of the vaporization loss of gasoline, and at the same time, the improvement of the fuel-air mixture formation. 4.5

65 SVP of Gasoline kPa 55

4

Vaporization loss, % mass

45

3.5 35 3

25

Surface tension, mN/m 2.5

15 5

2 0

5

10

15

20

Concentration of additive, ppm

25

30

surface tension ,mN/m (research octane number 80) surface tension,mN/m (research octane number 92) SVP of gasoline, kPa Vaporization loss of gasoline,% mass

Figure 6:

The influence of the additive on the value of surface tension at the 15°, on the SVP of gasoline and the vaporization loss of gasoline.

The atomic-absorption analysis of content showed no additional toxic components in exhaust gases, not observed before the application of the additive. The high efficiency of the developed additive was confirmed by the experimental-industrial application. The additive was used in the production of WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

383

about 360 thousands tons of gasoline and 40 thousands tons of diesel fuel. The additive is synthesized from available, inexpensive components and its production does not require highly qualified specialists and sophisticated equipment. The additive can be introduced into the fuel at all stages: during the production, in a finished fuel, at the petrol stations and directly into the fuel tanks of vehicles. Known and applied monofunctional additives are much less efficient than that proposed, at a consumption rate of 20 - 100 times higher and considerably more expensive. Table 2 compares the efficiency of the developed multifunctional additive and available on the world market additives. Table 2:

Influence of the developed multifunctional additive on the operational and environmental characteristics of vehicles and on the properties of motor fuels (concentration of the additive 9.25 ppm in gasoline and 27.75 ppm in diesel fuel).

Characteristics

Gasoline engines, gasoline 512%

Diesel engines, diesel fuel 47%

Analogs

In agreement with patent data less efficient by the rate 0.1–0.5% Reduction of carbon ~ –95100% ~ – 95100% less efficient by the rate 0.05–0.10% deposits Detergent properties, 2.33 2.33 The same efficiency by decrease of cleaning the rate 0.025–0.05% time, times Reduction of emission The same effect by the СО concentration in 50–100 NOx times more. Analogs 1517% 1530% CH with influence on 2022% 2026% benzo(a)pyrene and benzo(a)pyrene are 3537% 835% its analogs unknown soot 9596%  4050%  Decrease of 6–10 items The same efficiency by  requirements to octane the rate 50-100 times number of gasoline more Cetane number +3 items The same efficiency by  the rate 18-36 times more The same efficiency by Lubricating properties 25%  the rate 0.05 – 0.10% (coefficient of static friction) There are no analogs. Vaporization loss of 1520%  gasoline Fuel consumption

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

384 Air Pollution XIX

4 Conclusion Application of the developed multifunctional additive will quickly improve the environment in big cities and solve in some countries the problem of a shortage of high octane gasoline, with significant savings in fuel consumption without any capital investment in the automobile industry and oil refining.

References [1] Magaril, E.R. Influence of the quality of engine fuels on the operation and environmental characteristics of vehicles: monograph [in Russian]. KDU: Moscow, 2008. [2] Magaril, E.R., Magaril R.Z. Motor fuels [in Russian]. Second edition. KDU: Moscow, 2010.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Section 6 Global and regional

This page intentionally left blank

Air Pollution XIX

387

Application of methanotrophic biofilters to reduce GHG generated by landfill in Quebec City (Canada) N. Turgeon1, Y. Le Bihan1, G. Buelna1, C. Bourgault2, S. Verreault3, P. Lessard2, J. Nikiema4 & M. Heitz4 1

Centre de recherche industrielle du Québec (CRIQ), Canada Université Laval, Département de génie civil et de génie des eaux, Quebec, Canada 3 Ville de Québec, Canada 4 Université de Sherbrooke, Department of Chemical Engineering and Biotechnological Engineering, Faculty of Engineering, Canada 2

Abstract Compared to carbon dioxide (CO2), methane (CH4) is a strong greenhouse gas (GHG) and landfills are one of the major anthropogenic sources of atmospheric CH4 produced by anaerobic degradation of organic waste. In Canada as in many countries around the world, programs and regulations are implemented to force capture and burning of landfill gas (LFG). However, when thermal oxidation (flaring or energetic valorisation) is not possible (i.e. low CH4 concentration or flowrate), microbial methane oxidation by methanotrophic biofilters represents a new technology that holds great promises for GHG reduction and air pollution control of LFG. Exploratory work done in CRIQ laboratories (Quebec Canada) allowed testing different types of mediums (organic and inorganic) for the design of methanotrophic biofilters. Following this initiative, pilot scale project was undertaken in 2009. The objective was to evaluate, using a prototype installed in a closed landfill (Beauport, Quebec City), the technical and economic feasibility of implantation of methanotrophic biofilter for the treatment of LFG. Testing protocol has been implemented over a period of 83 d (from September to November 2009). The collected data were used to evaluate conversion rates (up to 80%) and the maximum elimination capacity (ECmax = 66 g CH4/m3/h). Large-scale technology demonstration work is planned for 2011-2012 to WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110361

388 Air Pollution XIX validate, over 12 months, the GHG reduction cost established for methanotrophic biofilter (CAD$16-20/t CO2eq). Keywords: methane, landfill gas, greenhouse gas, methanotrophic bacteria, biofilters, nitrous oxide.

1 Introduction Methane (CH4) is a colorless, odourless, flammable gas that has a 100-year global warming potential (GWP) of 21 [1]. Since 1750, the overall average atmospheric concentration of CH4 has increased by 157% [2]. After carbon dioxide (CO2), CH4 is the next most plentiful greenhouse gas (GHG) that can be attributed to human activities (raising livestock, intensive agriculture, industrial processes, extraction of combustible fossil fuels, coal mines, the incomplete combustion of combustible fossil fuels and waste management) [3]. Landfill gases (LFGs) are produced when the organic portion of landfill waste biodegrades in the absence of oxygen. LFGs may contain substantial concentrations of CH4 (up to 55-60% v/v) and CO2 (40-45% v/v) [4]. Figure 1 drawn from Jensen and Pipatti [5] shows a representation of LFG production as a function of time. Methane production continues until the organic waste is completely degraded by methanogenic bacteria, which may take up to 100 years for certain sites. Over the course of the stable production phase of CH4 (i.e., ~20 years), the produced LFGs may be utilized or incinerated using a flare stack. In Quebec (Canada), capture and incineration are mandatory for sites having a landfill capacity more than 50,000 metric tons/year according to the Regulation Respecting the Landfilling and Incineration of Residual Materials (REIMR) that

Figure 1:

Scholl Canyon model representation of landfill degradation.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

389

went into effect as part of the Quebec Action Plan 2006–2012 [6]. For other landfill locations (smaller ones or those set up prior to 2006), or when the quality of the LFGs (i.e. CH4 < 25% by volume) does not allow further thermal oxidation without adding a supplementary gas (propane or natural gas), incineration is not mandatory and the LFGs produced at these locations are generally emitted directly into the atmosphere, thereby generating considerable GHG emissions. Methanotrophic bacteria belong to a sub-group that falls under the physiological group of methylotrophic bacteria and are unique in their capability to use CH4 as a source of carbon and energy. This characteristic has also been the subject of various studies especially in order to identify the primary environmental factors that govern the biological oxidation process of CH4 (temperature, moisture, oxygen, nitrogen, pH, etc.) [8–13]. Among the biological treatment processes, biofiltration has been for several years a recognized approach for controlling industrial atmospheric emissions such as odours and volatile organic compounds (VOCs) [14–16]. Given the context of fighting climate change, biofiltration of CH4 now appears to be a very promising method to reduce GHGs. This process consists of passing the gas stream to be treated through a filtering bed packed with porous materials on which microorganisms are fixed. Under aerobic conditions, compounds such as CH4 are transformed into molecules that are less harmful to the environment (CO2, H2O, salts and biomass) according to the biological oxidation reaction described in equation 1. CH4 + O2  CO2 + H2O + Cells, with   2,   1,   2 G0 = -780 kJ mol-1 CH4

(1)

Contrarily to passive oxidation systems (e.g., landfill biocovers), methanotrophic biofilters are engineering reactors where conditions such as the concentration of CH4, air flow rate, humidity, temperature, pH, inoculation, and dosage with nutrients are controlled especially to ensure higher and more stable elimination capabilities and conversion rates. A review of the literature by Nikiema [17] made it possible to identify about 20 published studies bearing specifically on the biofiltration of methane. Most of these projects were performed in laboratories (prototype) with synthetic gases (mixture of natural gas and compressed air). Exploratory work conducted in laboratories (30-liter biofilters) has enabled various types of filter bed (organic and inorganic) pertaining to biofilter design [18] to be tested. Based on this experimentation, a technological demonstration project was carried out in collaboration with Quebec City in 2009. The primary objective of this project was to evaluate the technical-economic feasibility of establishing a biofiltration process to treat LFGs and reduce GHG emissions by means of a prototype installed at a landfill site.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

390 Air Pollution XIX

2 Material and methods 2.1 Description of the Beauport Landfill Site (Quebec, Canada) According to the greenhouse gas emissions balance performed in 2007 for the Quebec City metropolitan area, the Beauport landfill site contributes ~8% of the total GHG emissions of the City’s corporate inventory (i.e. ~18,000 t CO2eq/year). Beauport is a closed site where 918,000 metric tons of waste were buried between 1979 and 1986. In 2005, a capture system was installed to control LFG migration toward neighbouring homes and businesses. Since then, the LFGs captured in this way (~ 240 m3/h, CH4 = 3% to 18% v/v) have been discharged into the atmosphere without any treatment. 2.2 Description of the Biofiltration Prototype The technological demonstration projects were carried out over a period from July to November 2009. The experimental biofiltration prototype used for these tests was installed directly at the Beauport site, near the existing LFG pumping station (Figure 2:). This prototype consisted primarily of a pre-filter (500 mm x 600 mm x 1,400 mm) equipped with atomization nozzles to ensure humidification of the gas stream, a fan followed by a BiosorTM biofilter (950 mm x 1,000 mm x 1,000 mm) with organic filter (mixture of peat moss and wood chips) (Figure 3:). Two reservoirs, each with a volume of 350 L, made it possible to recover by means of gravity the scrubbing water (pre-filter) and the nutrient solution (biofilter), and then to recycle these effluents using centrifugal pumps.

Figure 2:

Experimental setup (closed landfill Beauport, Quebec City).

Laboratory-cultivated methanotrophic bacteria were applied by liquid inoculation directly on the biofilter’s medium. The dosed addition of nutrients (N, P, oligoelements) necessary for methanotrophic bacteria to grow was performed by incorporating slow-dissolving granulated fertilizer directly into the biofilter or added as needed to the reservoir containing the nutrient solution. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Figure 3:

391

Flow chart of pilot scale Methanotrophic Biofilter. 1-LFG from biogas station; 2-ambient air; 3-exhaust; 4-conditioning chamber; 5-methanotrophic biofilter; 6-water; 7-nutrient solution.

The biofiltration system was started on July 24, 2009. However, problems pertaining to air infiltration at the ventilator were noted in early September. Testing was re-started on September 3, 2009 and continued up to November 25, 2009 (total duration: 83 days). 2.3 Analytical monitoring Characterizing the primary GHGs (CH4, N2O, and CO2) was performed using a multi-gas analyzer and gas chromatography using a flame ionization detector equipped with an automatic sampling system (GC-FID Varian 3800). This device was installed on site inside a mobile laboratory for the duration of the tests. The parameters used to describe the results are defined in table 1, namely the inlet load (IL) in gCH4/m3/h, the CH4 conversion rate (X) in % and the elimination capacity (EC) in gCH4/m3/h. The temperature of the LFGs and the biofilters was monitored using RTD sensors connected to an automatic data acquisition system (DAQ PRO-5300, EQ-3015).

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

392 Air Pollution XIX Table 1:

Determination of the quantitative parameters.

Parameters

Methods of determination xQ C CH IL V

IL : Volumetric inlet load (gCH4/m3/h) X

X: Conversion (%)

C CH

C CH C CH

EC : Elimination capacity (gCH4/m3/h)

 x 100

EC = IL x X 3

C CH : Inlet methane concentration in gCH4/m ; C CH : Outlet methane concentration in gCH4/m3; Q = Volumetric flow of LFG in m3/h; V: Biofilter volume in m3

3 Results and discussion 3.1.1 Characterization of the LFGs Table 2 depicts the characteristics of landfill gas captured and discharged at the LFG station during the testing period. The relatively high concentration of oxygen (7 to 9% v/v) reveals the presence of air infiltration in the Beauport site’s capture network. Calculating the direct GHG emissions shows that the LFG station discharges about 1,270 t CO2eq/year into the atmosphere, or about 7% of the total annual emissions modeled for the Beauport site. Table 2: Period Flow Temperature CH4 CO2 N2O O2

LFG characterization. July 24, 2009 to Nov. 24, 2009 Average:165 Nm3/h (141 to 192 Nm3/h) 30 to 40oC 6.4% (1.7 to 13.1% v/v) 4.1% (2.6 to 13.1% v/v) 3 ppmv (0 to 8 ppmv) 7 to 9% v/v

3.1.2 Monitoring the Biofiltration Prototype Figure 4: depicts the elimination capacity (EC) as a function of the inlet load (IL). The data points represented by ◊ correspond to the start-up phase of the biofiltration process (Sept. 4, 2009 to Sept. 14, 2009), or about 10 days. The data points represented by correspond to the loading phase of the biofiltration prototype. The results obtained show an average conversion rate (Xaverage) of 53% and a maximum (Xmax) going up to 80%. The maximum elimination capacity (ECmax) recorded over the course of testing was 66 g/m3/h for an inlet load (IL) of about 95 g/m3/h. These results are comparable to the best purification performance identified by Nikiema et al. [19].

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Figure 4:

393

Elimination capacity (EC) as a function of CH4 inlet load (IL).

Figure 5: depicts the trend of the temperature in the middle of the filter bed (uninsulated biofilter) and the inlet concentration of CH4 as a function of time. Over the first 30 days of operation (month of September), the concentration of CH4 at the biofilter’s inlet was maintained at < 2% v/v (i.e., 20,000 ppmv) by adjusting the flow rate of the diluting ambient air. Then, the dilution valve was closed until the end of testing to treat the non-diluted LFGs. An increase of the CH4 concentration at the biofilter’s inlet was accompanied by a significant and spontaneous increase of the biofilter’s temperature (from 35 to 60 °C) due to the exothermic CH4 degradation process. Temperatures exceeding 50oC within the biofilters were also noted during the studies under similar operation conditions [20]. In early November, a drop in temperature was recorded from 55 to 25 °C. This cooling of the biofilter was associated with a decrease in the exterior temperature as well as a lack of nutrient solution causing a decrease in the CH4 conversion rates. These results show that the bio-oxidation of CH4 is highly dependent on temperature and nutrients. Accordingly, these parameters represent two important indicators pertaining to the operation of the methantrophic biofiltration process. The biological treatment processes may generate nitrous oxide (N2O) emissions due to an incomplete nitrogen transformation process [21, 22]. Figure 6 depicts the presence of N2O at the biofilter’s outlet with concentrations varying between 3 and 307 ppmv (average: 97 ppmv). Since N2O is a powerful GHG (GWP= 310), it is important to consider these direct emissions when quantifying the process’s GHG reductions (section 3.1.3). WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

394 Air Pollution XIX

Figure 5:

Figure 6:

Inlet CH4 concentration and temperature of biofilter.

Monitoring of nitrous oxide (N2O) concentration.

3.1.3 Evaluating the costs associated with reducing GHG emissions Table 3 shows an established scenario for calculating the costs associated with reducing GHG emissions using BiosorTM methanotrophic biofiltration technology in Quebec (Canada). When considering a site already equipped with LFG capture systems, the unit costs for reducing GHGs is estimated to be about CAD$16-20/t CO2eq. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Table 3:

395

GHG’s emission reducing cost: Methanotrophic Biofilter.

C(CH4)in C(CH4)out IL : EC : X: C(N2O)in: C(N2O)out: GHG reduction (CH4 bio-oxidation)1 GHG emission (N2O)2 GHG emission (CO2)3 GHG reduction Total annual cost (Methanotrophic Biofilter)4

7.0% 2.1% 91.4 g CH4/m3/h 64.0 g CH4/m3/h 70% 0 ppmv 100 ppmv 11.7 t CO2eq /m3/year 0.7 t CO2eq/m3/year 11.0 t CO2eq/m3/year ~ 180-220 $/m3/year

GHG reduction cost

CAD$16-20/t CO2eq

1

GWP: 21 2 GWP: 310 3 CO2 emission from biomass: not computed (UNFCCC methodology) 4 Capital costs (financing) and operation costs (change of biofilter’s medium after 4-5 years, maintenance).

4 Conclusion This project’s primary objective was to validate the technical and economic feasibility of BiosorTM biofiltration technology to treat CH4 and reduce GHGs generated by landfill gas (LFG) emissions originating from a site that has been closed for more than 15 years. A biofiltration prototype equipped with a monitoring system was thus installed at the Beauport landfill site (Quebec, Canada) and monitored for a period of 83 days. The results obtained enabled one to establish the technology’s conversion rates (Xmax:80%) as well as a maximum elimination capacity (ECmax: 66 g/m3/h for an inlet load of 95 g/m3/h). Monitoring the biofilter’s temperature constitutes a significant and simple indicator for the process’s proper functioning. The reduction costs associated with implementing a methanotrophic biofilter in a landfill site already equipped with a capture system were estimated to be about CAD$16-20/t CO2eq. Fullscale demonstration projects are planned in order to validate that, in a Nordic climate situation, biofiltration constitutes a robust and cost-effective alternative for treating LFGs and reducing GHG emissions.

Acknowledgements This project was completed subsequent to agreements between Centre de recherche industrielle du Québec (CRIQ), Université Laval and Université de Sherbrooke, and made possible thanks especially to the financial participation of CRIQ’s internal research program and the City of Quebec. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

396 Air Pollution XIX

References [1] Intergovernmental Panel on Climate Change (IPCC). Climate Change 1995: The Science of Climate Change. Summary for Policymakers and Technical Summary of the Working Group I Report. p. 22. Cambridge (UK): Cambridge, University Press, 1995. [2] World Meteorological Organization. WMO Greenhouse Gas Bulletin: Main Greenhouse Gases Reach Highest Level Ever Since Pre-Industrial Time. November 2009. No. 868. Available online at: http://www.wmo.int /pages/mediacentre/press_releases/pr_868_en html [3] Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B, Tignor, M. and Miller, H.L., Climate Change 2007: The Physical Science Basis, Contribution of Working Group 1 to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC), Cambridge (UK): Cambridge University Press, 2007. [4] Environment Canada, National Inventory Report 1990-2008, Part 1: Greenhouse gas sources and sinks in Canada, The Canadian Government’s Submission to the UN Framework Convention on Climate Change, 2010. [5] Jensen, EF, Pipatti, R., CH4 Emissions from Solid Waste Disposal, February 2003. Available online at: http://www.ipccnggip.iges.or.jp/public/gp/bgp/5_1_CH4_Solid_Waste.pdf [6] Le Québec et les changements climatiques, un défi pour l’avenir. Plan d’action du Québec 2006-2012 sur les changements climatiques. Juin 2008. Disponible en ligne à: [7] http://www.mddep.gouv.qc.ca/changements/plan_action/2006-2012_fr.pdf [8] Humer, M., Lechner, P., Microbial methane oxidation for the reduction of landfill gas emission, Journal of Solid Waste Technology and Management, Volume 27, NOS 3 & 4, November 2001. [9] Dammann, B., Steese, J., Stregmann, R., Microbial oxidation of methane from landfills in biofiltres, Proceedings Sardinia 99, Seventh International Waste Management and Landfill Symposium, S Margherita di Pula, Cagliari, Italy, 4-8 October 1999. [10] Park, S.Y., Brown, K.W., Thomas, J.C., The use of biofiltres to reduce atmospheric methane emissions from landfills: Part 1. Biofilter Design, Water, Air, and Soil Pollution, 155: 63-85, 2004. [11] Haubrichs, R., Widmann, R., Evaluation of aerated biofilter systems for microbial methane oxidation of poor landfill gas, Waste Management, 26: 408-416, 2006. [12] Nikiema, J., Heitz, M., The use of inorganic packing materials during methane biofiltration, Hindawi Publishing Corporation, International Journal of Chemical Engineering, Article ID 573149, 8 pages, Volume 2010 [13] Scheulz, C., Bogner, J. E., De Visscher, A., Gelbert, J., Hilger, H. A., Huber-Humer, M., Kjeldsen, P., Spokas, K., Microbial methane oxidation processes and technologies for mitigation of landfill gas emissions, Waste management & research, Volume 27, Issues 5, August 2009. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

397

[14] Devinny, J.S., Deshusses, M.A., Webster, T.S., Biofiltration for Air Pollution Control. Lewis Publishers, Boca Raton, Florida, 1999. [15] Kennes, C., Veiga, M.C., Bioreactors for Waste Gas Treatment. Kluwer Academic Publishers, 312 p., 2001. [16] Delhoménie, M.-C., Heitz, M., Biofiltration of Air: A Review, Critical Reviews in Biotechnology, 25:1-20, 2005. [17] Nikiema, S. J., Atténuation des émissions de gaz à effet de serre par biofiltration du méthane : optimisation des paramètres opératoires, thèse de doctorat ès sciences appliquées, Université de Sherbrooke, 2008. [18] Turgeon, N., Bourgault, C., Buelna, G., Le Bihan, Y., Verreault, S., Lessard, P., Nikiema J., Heitz, M., Développement d’un procédé de biofiltration pour le traitement du CH4 et la réduction des GES, Salon des technologies environnementales du Québec, Centre des Congrès de Québec (Québec, Canada), 16 mars 2010. [19] Nikiema, J., Brzezinski, R., Heitz, M., Elimination of methane generated from landfills by biofiltration: a review, Rev. Environ. Sci. Biotechnol, 6:261-284, 2007. [20] Steese, J., Stegmann, R., Design of biofilters for methane oxidation, Proc. Sardinia 2003, Ninth International Waste Management and Landfill Symposium, S. Margherita di Pula, Cagliari, Italy; 6 -10 October 2003. [21] Czepiel, P., Crill, P., Harriss, R., Nitrous Oxide Emissions from Municipal Wastewater Treatment, Environ. Sci. Technol., 29 (9), pp 2352-2356, 1995 [22] Kimochi, Y., Inamori, Y., Mizuochi M., Xu K.-Q., Matsumura M., Nitrogen removal and N2O emission in a full-scale domestic wastewater treatment plant with intermittent aeration, Journal of Fermentation and Bioengineering, Volume 86, Issue 2, pp 202-206, 1998.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

This page intentionally left blank

Air Pollution XIX

399

A study of the atmospheric dispersion of an elevated release with plume rise in a rural environment: comparison between field SF6 measurements and computations of Gaussian models (Briggs, Doury and ADMS 4.1) C. Leroy1, F. Derkx2, O. Connan1, P. Roupsard1, D. Maro1, D. Hébert1 & M. Rozet1 1

Laboratoire de Radioécologie de Cherbourg-Octeville, IRSN/DEI/SECRE, France 2 Veolia Environnement Recherche et Innovation, Département Environnement et Santé, France

Abstract In order to reduce uncertainties and enhance the knowledge of elevated releases atmospheric dispersion in a rural plain, the French Institute for Radioprotection and Nuclear Safety (IRSN), in collaboration with VEOLIA, carried out six weeks of experimental campaigns between November 2008 and July 2009 in the vicinity of an energy recycling unit. The atmospheric dispersion of the plume was studied by SF6 tracer injection in the 40 m high stack. Maximal values of experimental Atmospheric Transfer Coefficient (ATCmax) and horizontal dispersion standard deviations (h) were compared to the results of the first generation Gaussian models, Doury and Briggs, and to the results of the last generation Gaussian models, ADMS 4.1 (Atmospheric Dispersion Modelling System). Several modelling parameterization for ADMS 4.1 computations were tested and revealed an overestimation of the h with the building option or with the integration of a surface roughness file. Consequently, ADMS 4.1 was used without model options. The Doury and Briggs models were combined with Holland formulation for the effective height calculation. In neutral atmospheric conditions and in summer unstable conditions (class A, B and C according to Pasquill classification), the ADMS 4.1 model is appropriate to estimate ATCmax WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110371

400 Air Pollution XIX value but overestimates in some cases. It was noticed that, during wintry periods and in class C atmospheric conditions, all of these models overestimated ATCmax for distances from the release point comprised between 0 and 2000 m. To estimate the atmospheric dispersion of an industrial release with a commercial model, as ADMS 4.1, without a prior comparison with an experimental data base dedicated to the studied site, can induce a poorly suitable modelling parameterization and leads to uncertainties difficult to quantify on the dispersion conditions. Keywords: atmospheric dispersion, rural environment, SF6 tracer release, Gaussian models.

1 Introduction Predicting the dispersion of accidental releases into the atmosphere and estimating their consequences for the population is a major challenge. Because of their simplicity and rapidity of calculation, Gaussian models are ideally suitable tools for this problem. The Gaussian model is a simplified solution of the diffusion transport equation, which describes the spatial evolution of the concentration of a pollutant in the event of a constant release under uniform meteorological conditions. The use of a Gaussian model requires the standard deviation of the dispersion to be determined. For first generation Gaussian models, such as the models of Pasquill [1], Briggs [2] and Doury [3], the standard deviations for dispersion have been determined from experimental campaigns and are valid for the experimental conditions under which they were established, mainly from releases at ground level and over flat or slightly hilly terrain. A new generation of Gaussian models, such as the ADMS 4.1 model developed by CERC [4], has made it possible to determine the dispersion of industrial releases into the atmosphere as a function of the characteristics of the atmospheric boundary layer and the characteristics of the site: buildings, roughness, etc. The aim of this study is to evaluate the ability of the ADMS 4.1 model to reproduce dispersion phenomena for an elevated release with plume rise, by comparing the dispersion calculations with those deduced from an experimental data base and from calculations with first generation Gaussian models. In order to acquire an understanding experimental data base of near-field atmospheric dispersion, IRSN in collaboration with VEOLIA, carried out six weeks of experimental campaigns in the vicinity of an energy recycling unit between November 2008 and July 2009. The dispersion of the atmospheric releases was studied with injections of SF6, a passive tracer, via the 41 m-high stack. Ground samplings allow evaluating the maximal values of the Atmospheric Transfer Coefficient (ATCmax) and the horizontal dispersion standard deviation (h) up to a distance of 4.2 km from the release. Measurements were carried out during a wide range of meteorological conditions, which allow us to evaluate the plume dispersion during neutral and unstable atmospheric conditions.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

401

2 Experimental campaigns: equipment and methods 2.1 Experiment site The atmospheric dispersion campaigns were held from 17 to 28 November 2008, from 19 to 29 January 2009 and from 29 June and 10 July 2009 in the vicinity of an Energy Recycling Unity (EUR) on a rural plain, characterized by wooded surfaces and vegetable and cereal growing. The main building of the EUR is a 36 m high building, and is equipped with a 41 m high stack, that is 5 m above the build roof. The discharge conditions are given in table 1. Temperature and flow rate features induce a plume rise. Table 1:

Discharge conditions of the EUR.

Release height Stack diameter Flow rate Temperature Release velocity

41 m 1.3 m 60000 Nm3.h-1 145°C 20 m.s-1

2.2 SF6 release methods, sampling and measurements To study the atmospheric dispersion of the EUR release, a passive tracer, the SF6, was injected through the stack. Releases were realized with duration of 30 minutes and a constant generation rate of 5.4 g.s-1. The system used a SF6 bottle (Messer, France), connected to a mass flowmeter (Sierra 820) and installed on a balance to control the released mass. For ground measurements, atmospheric sampling were carried out into Tedlar bags using thirty autonomous gas sampling devices, spaced every 2 to 3° along axis perpendicular to the mean wind direction. Gas sampling devices allow doing consecutively 2 or 5 sample collections of nine minutes each. Among these samplings, it was chosen for data processing, the sampling for which, there was the best agreement between the wind direction and the transit time of the SF6 plume. SF6 analyses were conducted by Gas Chromatography with electron capture detector (CPG-ECD, AUTOTRAC 101 Tracer Gas Monitor, Lagus Applied Technology Inc). The detection limit is 25 ppt, with an accuracy of ± 3%. 2.3 Acquisition of meteorological data Instrumental meteorological devices were set on the industrial site. Turbulent parameters and wind direction and velocity were evaluated at a 10 m height with an ultrasonic anemometer (Young 81000), operating at 20 Hz. The turbulent parameters (friction velocity, kinematic heat flux and Monin Obukhov length, roughness length) were derived via eddy-correlation. Roughness length and inverse length of Monin-Obukhov were correlated according to the works of Golder [5] to determine the atmospheric stability. Temperature, atmospheric WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

402 Air Pollution XIX pressure, global radiation and pluviometry were measured at a 1.5 m height with a meteorological station (AHLBORN). Meteorological and turbulent data used for dispersion modelling are averaged on the duration of discharge and sampling.

3 Experimental results 3.1 Data processing The measurements of SF6 concentration are used to determine the maximal atmospheric transfer coefficients value (ATCmax) evaluated for each measurement axis, expressed according to eqn (1): t1

ATCmax 



X ( M , t ).dt

t0

(1)

t '1



q(t ).dt

t '0

- X(M,t): maximal SF6 concentration (ppb), measured along a radial, - q(t): SF6 release rate, in m3.s-1, - t’0, t’1: instant of the beginning and end of source emission in s, - t0, t1: instant of the beginning and end of measurement in s. The SF6 tracer dispersion was also studied in terms of plume form, in order to check if the Gaussian dispersion is correctly represented by the experimental results. To achieve this, a Gaussian was fitted to our field data (for each radial performed). 3.2 The data base description Among the forty four releases studied during the three campaigns, only measurement radials for which there was a good coherence between the sampling location and the wind direction were used for the comparison between the ATCmax which was observed (ATCo) and the ATCmax which was predicted (ATCp) (exploitation rate: 75%). According to the Pasquill classification, twenty radials were realized in neutral atmospheric condition (class D), and eighteen in unstable conditions: ten in class C, four in class B and four in class A. Measurements were carried out at distances from the release point ranging from 100 m to 4160 m. Some SF6 releases were performed, during the UVE stops, with temperature less than 32°C. Values of experimental h and ATCo for various distances are indicated in table 2 for neutral atmospheric conditions and in table 3 for unstable atmospheric conditions.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

403

In neutral conditions and for release performed at 145°C, the maximal impact of the plume, is located at a distance ranging from 1000 to 1500 m from the release point with an ATCo in the order of 5.0 10-6 to 1.0 10-5 s.m-3; for greater distance, the value of ATCo gradually decreases to reach 10-7 s m-3 at 3250 m. It can be noticed an atypical value of 1.1 10-6 at 3640 m. In the case of cold releases and whatever the distance, the ATCo value is one order of magnitude greater than in the case of warmer releases. During winter days, in unstable conditions (Class C and B), the ATCo is about 10-7 s m-3 at 400 m from the release and increases with the distance up to a factor of five. During summer days in stability class C, an opposite behaviour is observed; the plume impact is maximal at 245 m from the release with an ATCo value of 7.6 10-6 s m-3 and decrease to reach a value of 1.1 10-6 s.m-3 at 2090 m. In atmospheric stability class A, ATCo is weaker of one order of magnitude than in class C. The standard deviation of the horizontal dispersion varied between 36 and 280 m, it can be noticed that the evolution of h in function of the release distance is similar in neutral and unstable condition. Table 2:

Date 08/07/09 07/07/09 23/01/09 22/01/09 22/11/08 22/01/09 22/11/08 23/01/09 08/07/09 22/01/09 08/07/09 24/01/09 19/11/08 24/11/08 20/11/08 07/07/09 07/07/09 27/01/09 21/11/08 23/01/09

ATCo and h deduced from experimental campaigns in neutral atmospheric conditions and related experimental conditions: distance from release, and wind speed at 10 m height. Grey lines correspond to SF6 release performed with cold temperature. Distance from release (m) 100 104 570 730 850 1480 1590 1650 1650 1670 1706 1760 1800 2020 2090 2180 2660 3250 3515 3640

Discharge temperature (°C) 145 145 145 145 106 146 17 145 145 145 145 143 17 145 15 145 145 145 13 150

Stability Class

U (m.s-1)

ATCo (s.m-3)

h (m)

D D D D D D D D D D D D D D D D D D D D

3.8 7.3 11.4 9.9 5.6 9.8 5.8 11.4 4.8 9.9 3.8 3.4 3.7 3.6 4.4 7.3 7.3 3.0 7.2 11.6

0 4.1 10-6 1.2 10-5 3.7 10-6 1.0 10-5 4.0 10-6 1.1 10-5 2.8 10-6 2.9 10-6 2.0 10-6 1.2 10-6 2.5 10-7 5.0 10-6 8.1 10-7 8.6 10-6 9.6 10-7 6.7 10-7 3.2 10-7 2.0 10-6 1.1 10-6

– – 36 70 60 86 100 – 82 132 83 80 165 110 80 152 280 165 250 250

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

404 Air Pollution XIX

Table 3:

ATCo and h deduced from experimental campaigns in unstable atmospheric conditions and related experimental conditions: distance from release, wind speed at 10 m height. Grey lines correspond to SF6 release performed with cold temperature.

27/11/08 27/01/09 27/11/08 21/01/09 27/11/08 27/01/09

Distance from release (m) 400 730 1010 1215 1940 2995

Discharge temperature (°C) 145 143 145 146 145 143

07/07/09 06/07/09 07/07/09 06/07/09

245 629 2073 2090

28/01/09 28/01/09

Stability Class

U (m.s-1)

ATCo (s.m-3)

h (m)

C C C C C C

4.3 3.9 4.3 3.7 4.3 3.9

1.1 10-7 1.7 10-7 1.5 10-7 3.4 10-7 5 10-7 6.0 10-7

– 57 – 90 100 178

145 145 145 145

C C C C

5.5 5.6 5.5 5.1

7.6 10-6 5.0 10-6 2.3 10-6 1.1 10-6

– 44 101 146

1260 4160

143 143

B B

3.2 3.2

3.1 10-7 4.7 10-7

70 230

01/07/09 01/07/09

912 1000

30 32

B B

3.2 1.9

1.7 10-6 7.5 10-6

96 105

03/07/09 02/07/09 03/07/09 02/07/09

700 1395 1714 3080

145 145 145 145

A A A A

3.23 2.3 3.23 2.3

4.0 10-7 5.6 10-7 4.8 10-7 8.0 10-8

– – 150 –

Date

4 Comparison with Gaussian models 4.1 Description of ADMS 4.1, Briggs and Doury models Three Gaussian plume atmospheric dispersion models were used to predict the ATC around the site, the ADMS 4.1 model [4], the Briggs model [2] and the Doury model [3]. The main difference between these models is that ADMS 4.1 uses a more modern method of boundary layer scaling, the Monin-Obukhov length, which allows for vertically inhomogeneous turbulence in the atmosphere to be modelled. ADMS 4.1 uses a Gaussian concentration distribution to calculate the dispersion of releases under stable and neutral conditions [6, 7] and a skewed distribution in the case of unstable conditions [8, 9]. Moreover, it contains several modules which enable to take into account the effects of buildings, topography and roughness on the dispersion and on the trajectory of the plume [10, 11]. Several modelling parameterization for ADMS 4.1 computations were tested and revealed a strong overestimation of the plume width with the building option or with the integration of a surface roughness file. Consequently, ADMS 4.1 was WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

405

used without these model options. The meteorological and micro-meteorological data set allow testing the sensitivity of the model to the meteorological parameter inputs. For each atmospheric stability condition, the modelling configuration, that allows the best correlation between ATCo and ATCp, was retained for comparison with ATCp calculations performed with Doury and Briggs models. In neutral conditions, the following meteorological parameters were considered: wind velocity and direction, temperature, humidity, global radiation and roughness length measured on site. In unstable conditions, the following meteorological parameters were considered: Monin Obhukhov length, kinematic heat flux, wind velocity and direction, temperature, humidity and a constant value of roughness length (z0 = 0.02 m). The Briggs’ formulations of standard deviations for dispersion are function of the Pasquill stability classes and of the distance from the release point [2]. Pasquill stability classes consider 6 atmospheric stability ranges from very unstable (A), to very stable (F). On the other hand, the Doury standard deviations are function of the transfer time and of only two classes of atmospheric stability: normal diffusion and weak diffusion [3]. Normal diffusion is defined by a vertical temperature gradient less than or equal to -0.5°C/100 m and corresponds to unstable or neutral atmospheric conditions. Weak diffusion is defined by a vertical temperature gradient greater than -0.5°C/100 m and is equivalent to stable or very stable atmospheric conditions. The Doury and Briggs models were combined with Holland formulation [12, 13] for the calculation of the effective height ( H eff ) :

H eff  z s  h

(2)

- z s : source height (m), - h : plume rise (m) :

h  2.2

Vs d u

 418

W u

(3)

- VS : discharge velocity (m/s), - d : stack diameter (m), - u : wind velocity at the discharge height (m/s), - W : thermal flow (W) The wind velocity at the discharge height was calculated according to the Businger-Dyer relationship [14]. 4.2 Comparison with the experimental results In order to compare our experimental results to the dispersion calculations with Briggs, Doury, ADMS 4.1 models, the ratios h_p/ h_o and ATCp/ATCo are shown for neutral conditions in figures 1 and 2 and, for unstable conditions, in figures 3 and 4. The statistical parameters (mean, median, standard deviation) of those ratios are given in tables 4 and 5. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

406 Air Pollution XIX Briggs_17 C

Doury_17 C

ADMS_17 C

Briggs_145 C

Doury_145 C

ADMS_145 C

h_p/h_o

1.E+01

1.E+00

1.E-01 0

1000

2000

3000

4000

Release distance (m)

Figure 1:

Comparison of the ratios h_p/ h_o for Briggs, Doury and ADMS 4.1 models as a function of the release distance and of the temperature discharge in neutral atmospheric conditions (class D according to Pasquill classification). Briggs_17 C Briggs_135 C

Doury_17 C Doury_135 C

ADMS_17 C ADMS_135 C

1.E+02

ATCp/ATCo

1.E+01

1.E+00

1.E-01

1.E-02 0

1000

2000

3000

4000

Release distance (m)

Figure 2:

Comparison of the ratios ATCp/ATCo for Briggs, Doury and ADMS 4.1 models as a function of the release distance and of the temperature discharge in neutral atmospheric conditions (class D according to Pasquill classification).

In neutral conditions, the most appropriate model to provide the h is the Briggs model (h_p/ h_o mean = 1.19). The Doury model underestimates the WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

407

calculation of h of 35% in mean, while the ADMS 4.1 model tends to overestimate h of 37% in mean. In terms of ATC, ADMS 4.1 achieves a good agreement with the measurements (ATCp/ATCo mean = 1.19). For warm releases, the Doury and the Briggs models overestimate the ATC value up to a factor 5. For cold releases, the ATC results for the Doury and the Briggs models are much better; the mean ratio ATCp/ATCo is respectively of 1.51 and of 0.84. Thus, the Holland formulation for plume rise does not seem to be appropriate for this discharge configuration. Table 4:

Comparison of the mean, the median and the standard deviation calculated for the ratios h_p/ h_o and ATCp/ATCo for Briggs, Doury and ADMS 4.1 models in neutral atmospheric condition (class D according to Pasquill classification).

h_p/ h_o ATCp/ATCo release at T = 145°C ATCp/ATCo release at T < 32°C

Table 5:

h_p/ h_o ATCp/ATCo Winter days ATCp/ATCo Summer days

mean median standard deviation mean median standard deviation mean median standard deviation

ADMS 4.1 1.37 1.37 0.59 1.02 0.54 1.61 0.40 0.40 0.17

Briggs 1.19 1.18 0.33 3.11 1.00 5.05 0.84 0.73 0.38

Doury 0.65 0.57 0.42 5.08 2.35 7.99 1.51 1.40 0.70

Comparison of the mean, the median and the standard deviation calculated for the ratios h_p/ h_o and ATCp/ATCo for Briggs, Doury and ADMS 4.1 models in unstable atmospheric condition (class D according to Pasquill classification). mean median standard deviation mean median standard deviation mean median standard deviation

ADMS 4.1 1.80 1.66 0.78 6.17 3.40 6.51 1.53 0.84 2.22

Briggs 1.81 1.57 0.47 18.31 11.71 19.27 2.44 1.49 2.85

Doury 0.83 0.80 0.36 21.81 16.80 18.68 11.94 8.25 12.58

In unstable conditions, the most appropriate model to provide the h is the Doury model (h_p/ h_o mean = 0.83), whereas the Briggs and the ADMS 4.1 models, which deliver similar results (h_p/ h_o mean = 1.8), overestimate this parameter (fig.3).

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

408 Air Pollution XIX Briggs_winter

Doury_winter

ADMS_winter

Briggs_summer

Doury_summer

ADMS_summer

h_p/h_o

1.E+01

1.E+00

1.E-01 0

1000

2000

3000

4000

Release distance (m)

Figure 3:

Comparison of the ratios h_p/ h_o for Briggs, Doury and ADMS 4.1 models as a function of the release distance and of the temperature discharge in unstable atmospheric conditions (classes A, B, C according to Pasquill classification).

Briggs_winter

Doury_winter

ADMS_winter

Briggs_summer

Doury_summer

ADMS_summer

1.E+02

ATCp/ATCo

1.E+01 1.E+00 1.E-01 1.E-02 0

1000

2000

3000

4000

Release distance (m)

Figure 4:

Comparison of the ratios ATCp/ATCo for Briggs, Doury and ADMS 4.1 models as a function of the release distance and of the temperature discharge in unstable atmospheric conditions (classes A, B, C according to Pasquill classification).

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

409

In unstable conditions on winter days, none of the models accurately evaluate the ATCo at distances less than 2000 m from the release point; ADMS 4.1 overestimates the ATCo value of a factor 2 to 20 and the models of Briggs and Doury of a factor 6 to 50. For radials performed at 2995 and 4160 m, the ATCp/ATCo ratio calculated with Briggs and ADMS 4.1 models is lower than two, whereas the one calculated with Doury model varied between three and five. In unstable conditions on summer days, the ADMS 4.1 models is the most adapted to evaluate the ATCo; the mean of the ATCp/ATCo ratio is about 1.53 and the median about 0.84. The Briggs model over-estimates, in mean, the ATCo with a factor 2.5 and the Doury model with a factor 12. The Holland formulation for the effective height calculation of the plume, combined with Briggs model, in this discharge configuration, allows a better agreement between the ATCo and the ATCp in unstable condition during summer days than in winter days. This formulation does not suit to the dispersion calculations with the Doury model.

5 Conclusion This study allowed us to obtain ATC values in rural environments in the case of an elevated release for different atmospheric stability conditions. In neutral condition and in unstable conditions during summer days, whatever the distance from the release point, the most appropriate model to simulate the SF6 plume dispersion is ADMS 4.1. It can be noticed however that ADMS 4.1 tends to overestimate the plume width. To reach this result a parametrical study was necessary; used with the building option or with the integration of a surface roughness file, the ADMS 4.1 model highly overestimates the h. Consequently, ADMS 4.1 was used without model options. During winter days, in unstable conditions (Class C and B), for distance less than 2000 m, the ATCo value is inferior of one order of magnitude than during summer days. The three models overestimate the ATCo in this condition up to a factor 50. Although, the Holland formulation for the plume rise calculation, combined with Briggs model, allows a better agreement between the ATCo and the ATCp than combined with the Doury model, this formulation does not seem to be appropriate for this discharge configuration. Other plume rise formulations should be tested. To estimate the atmospheric dispersion of an industrial release with a commercial model, as ADMS 4.1, without a prior comparison with an experimental data base dedicated to the studied site, can induce a poorly suitable modelling parameterization and leads to uncertainties difficult to quantify on the dispersion conditions.

References [1] Pasquill, F., Estimation of the dispersion of windborne material. Meteorol. Mag., 90, pp. 33-49, 1961. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

410 Air Pollution XIX [2] Briggs, G.A., Diffusion estimation of small emissions. Contribution No.79. Atmospheric Turbulence and Diffusion Laboratory, Oak Ridge, TN, 1973. [3] Doury, A., Le vadémécum des transferts atmosphériques, rapport n°CEADSN-440, CEA, 1981. [4] CERC, ADMS 4.0 User Guide. Cambridge Environmental Research Consultants (CERC), 2009. [5] Golder, D., Relations among stability parameters in the surface layer. Boundary Layer Met., 3, pp. 47-58, 1972. [6] Weil, J.C., Updating applied diffusion models. J. Clim. Appl. Met. 24, pp. 1111-1130, 1985. [7] Hunt, J.C.R., Turbulent diffusion from sources in complex flows. Ann. Rev. Fluid. Mech. 17, pp. 447-458, 1985. [8] Lamb, R.G., Diffusion in the convective boundary layer; atmospheric turbulence and air pollution modelling, ed. F.T.M. Nieuwstadt and H. Van Dop, D. Reidel, Dordrecht, pp. 159-230, 1982. [9] Carruthers, D.J., Weng, W.S., Dyster, S.J., Singles, R. & Higson, H., Complex terrain module. Tech. rept. ADMS 4.0, P14/010/09. Cambridge Environmental Research Consultants (CERC), 2009. [10] Carruthers, D.J., Weng, W.S., Hunt, J.C.R., Holroyd, R.J., McHugh, C.A. & Dyster S.J., Plume/Puff spread and mean concentration module specifications. Tech. rept. ADMS 4.0, P10/01V/09, P12/01V/09. Cambridge Environmental Research Consultants (CERC), 2009. [11] Jackson, P.S., Hunt, J.C.R., Turbulent wind flow over a low hill. Q. J. R. Meteorol. Soc. 101, pp. 929-955, 1975. [12] Carpenter, S., Report on a full scale study of plume rise at large electric generating stations, 60th Annual Meeting of the Air Pollution Control Administration, Cleveland, pp. 67-82, 1967. [13] Brummage, K., The calculation of atmospheric dispersion from stack, Atmos. Environ., 2, pp. 197-224, 1968. [14] Stull, R.B., An Introduction to Boundary Layer Meteorology, Kluwer Academic Publishers, pp. 385, 1988.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

411

Ozone pollution during stratosphere-troposphere exchange events over equatorial Africa K. Ture & G. Mengistu Tsidu Department of Physics, Addis Ababa University, Ethiopia

Abstract Both natural factors such as thunderstorm events and anthropogenic activities contribute to very high ozone production. On the flight route from Johannesburg to Vienna enhanced ozone and relative humidity spikes were observed by MOZAIC (Measurements of Ozone by Airbus In service airCraft). MOZAIC recorded high resolution in-situ ozone and relative humidity at a flying altitude of 250-200 hPa at equatorial Africa. This area is one of the lightning hot spot regions of the world. We report introduction of enhanced ozone of stratospheric origin into the troposphere during two events and the resulting pollution. Vertical cross-section of potential vorticity over the region of interest showed high PV intrusion below the tropopause level. Both OLR (Outgoing longwave radiation) and vertical wind have indicated presence of strong convection. Cloud water content transport and high latent heat have confirmed the existence of thunderstorm activity coupled to PV intrusion. The two distinct events are characterized by very low ozone within the thunderstorm cloud and very high ozone of stratospheric origin outside the thunderstorm cloud. The events have produced ambient air pollution Keywords: thunderstorm, scavenging, pollution, ozone spikes, PV intrusion.

1 Introduction Ozone has different impacts depending on where it resides. Stratospheric ozone, where approximately 90% of the atmospheric ozone is found, prevents the sun’s ultraviolet radiation reaching the surface of the earth. In the troposphere, ozone is direct green house gas [1] while in the boundary layer, it is a pollutant, which has harmful effect on human, animal and crops. The health impact of ozone depends WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110381

412 Air Pollution XIX on the age group and health status of individuals. Since there is no discernible threshold below which no adverse health effects occur, no level would eliminate all risk. Thus, a zero risk standard is not possible. However, the current WHO (World Health Organization) AQG (Air quality guidelines) in the year 2000 for O3 provide a guideline value of 120µg/m3 (60 ppb), based on controlled human exposure studies, for a maximum 8-hour concentration. The AQG also provides two concentration-response tables, one for health effects estimated from controlled human exposure studies and one from epidemiological studies. No guideline for long-term effects was provided. Since the time these guidelines were agreed, there is sufficient evidence for their reconsideration. Issues to be considered are: the averaging time(s) for the shortterm guidelines and their associated levels, the concentration-response functions used in the tables, the outcomes included in the concentration-response tables, whether a long-term guideline and/or complementary guidelines (e.g. restricting personal activity) should be adopted [2]. With regards to the sources of ozone at a given area; it is determined by transportation and/or in-situ production by different mechanisms. Photochemical production is the dominant source of ozone in the stratosphere. In the troposphere, ozone is produced from biomass burning, natural emission from vegetation and soil, lightning, NOx emissions, and other anthropogenic sources such as emissions related to the combustion of fossil-fuel for energy, industrial, transport and domestic uses. Over the tropics, Africa is an important reservoir of ozone precursor sources allowing ozone to build up through active photochemistry exacerbated by high solar radiation. Africa contributes a significant amount to the global emissions from the first three categories, while emissions from fossil fuel combustion are important only on the regional scale. Regener [3] and Junge [4] consider the stratosphere to be the main source from which ozone enters the troposphere via tropopause exchange processes. Ozone is transported from the lower stratosphere into the upper troposphere through tropopause folding [5, 6] and is exchanged with the troposphere via diabatic processes and turbulent diffusion [7], mixing processes and convective erosion during the breakup of stratospheric filaments [8, 9]. The most important sources of ozone precursors over equatorial Africa are biomass burning, biogenic and lightning. African biomass burning activities, generally categorized as savanna, forest and agricultural waste burning, are driven by the “slash and burn” agricultural practices that take place during the dry seasons (late November to early March) in the northern hemisphere (NH), and July to October in the southern hemispheric (SH). The dynamic processes allow redistribution of such emissions on a more global scale. During the TRACE-A campaign, plumes loaded with high O3 over the Atlantic were attributed to biomass burning emissions from Africa. More recently high CO mixing ratios over the Indian Ocean have been attributed to African biomass burning [10]. Thunderstorms inject NOx mainly into the relatively clean upper troposphere. Measurements of O3 in clouds indicate that both production and loss mechanisms exist. Locally within the cloud the concentrated NO reacts with WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

413

ambient O3 to produce NO2 which reduces the O3 abundance accordingly. Lightning NOx is also responsible for a large fraction of the O3 produced in the troposphere [11]. Dickerson et al. [12] measured enhanced mixing ratios inside a cloud compared to the boundary layer value; Poulida et al. [13] measured high O3 near but not within a thunderstorm anvil. Both attributed this effect to intrusion of stratospheric air. Hauf et al. [14] detected a drop in O3 on entering the anvil core of a thunder storm cloud. Similar measurements by Ridley et al. [15] and Huntrieser et al. [16] also showed no systematic increase of in-cloud O3 concentrations. In contrast, Drapcho et al. [17] found a correlation between a decrease in O3 and an increase in NO2 caused by the production of NO, followed by its reaction with O3. Recent laboratory studies of arc discharges also indicate a large ratio of NO/NO2 and minimal O3 production [18]. According to Price [19] and Price and Asfur [20], equatorial Africa is one of lightning hot spot regions of the world. Marufu et al. [21] showed that 27% of the troposphere O3 abundance observed over Africa is caused by lightning induced nitrogen oxides (LNOx). On few occasions of several MOZAIC measurements during the period 199697, ozone and relative humidity spikes were observed at equatorial Africa. The main focus of this work is to investigate the causes and sources for the enhancements. These enhancements are spikes, covering very small region and dispersed. We used ozone records of GOME (Global Ozone Mapping Experiment) in the troposphere during this time from the European Centre for Medium-Range Weather Forecasts (ECMWF)-ERA-interim as complimentary information. Two processes were clearly distinguished: stratospheric intrusion and thunderstorm activities happening within the same time frame, between 12 and 18 UTC. The paper is structured as follows. In Section 2, the data and methodology used in this work will be presented. Section 3 gives a brief description of both MOZAIC ozone and relative humidity spike observations. In Section 4, the existence of stratospheric and troposphere exchange during the events is presented. Also mechanism of transport of ozone of stratospheric origin down to the ground level is discussed. Finally, the conclusion will be presented in Section 5.

2 Data and methodology Measurements of ozone in the MOZAIC program are taken every four seconds from take-off to landing. Based on the dual-beam UV absorption principle (Model 49-103 from Thermo Environment Instruments, USA), the ozone measurement accuracy is estimated to be ± [2 ppbv + 2%] [22]. From the beginning of the program in 1994, the measurement quality control procedures have remained unchanged to ensure that long-term series are free of instrumental artefacts. Instruments are laboratory calibrated before and after the flight periods, the duration of which is generally 12 to 18 months. The laboratory calibration is performed with a reference analyzer which is periodically cross checked at the National Institute of Standards and Technology in France. Additionally during WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

414 Air Pollution XIX the flight operation period, the zero and the calibration factor of each instrument are regularly checked using a built-in ozone generator. Furthermore, comparisons are made between aircraft when they fly close in location and time, which happen several times a month. Ozone measurements from the MOZAIC program were validated by comparisons with the ozone sounding network [23]. Data are recorded from aircraft take-off to landing, providing vertical profiles and cruise data between 8 and 12.5 km altitude. In this study only cruise data between 20 S–20 N were considered. In addition to MOZAIC data records, additional data such as potential vorticity, O3 mass mixing ratio, vertical and meridional wind fields, cloud cover, cloud liquid water content and relative vorticity data derived from ECMWF ERA-intrim data set were used to investigate the observed spikes during the events. The ERA-intrim data sets are given on a horizontal resolution of 1.5×1.5 degree and variable vertical resolution in pressure level with a 6-hourly analysis frequency. The vertical levels used for this analysis were partitioned in such a way that there are 6 levels within 900-775 hPa layer at an interval of 25 hPa, 11 levels within 750250 hPa layer at an interval of 50 hPa, 7 levels within 250-100 hPa layer at interval of 25 hPa, 2 levels at 70 and 50 hPa each. The ERA-Interim data assimilation has T255 horizontal resolution, better formulation of background error constraint, reprocessed ozone profiles from GOME data from the Rutherford Appleton Laboratory from 1995 onwards and CHAMP GPS radio occultation measurements, reprocessed by UCAR. ERA-Interim uses mostly the sets of observations acquired for ERA-40, supplemented by data for later years from ECMWF’s operational archive. EUMETSAT provided reprocessed winds and clear-sky radiances from Meteosat-2 (1982-1988) for ERA-40 and has reprocessed later Meteosat data for ERA-Interim [24]. The latent heat data set was retrieved from Mirador, Goddard earth sciences data and information center. Mirador contains a series of land surface parameters simulated from the Common Land Model (CLM) V2.0 model in the Global Land Data Assimilation System (GLDAS). The data are in 1.0 degree resolution and range from 1979 to the present. The temporal resolution is 3-hour [25]. The pressure and temperature profiles used to identify tropopause level, based on the methodology proposed in reference [26], were obtained from NASA Goddard Space Flight Center’s Laboratory for Atmospheres.

3 MOZAIC ozone and relative humidity observations O3 volume mixing ratio of exceeding 100 ppbv were observed in very few MOZAIC flights during 1996-97 over the equatorial Africa region (20 S–20N) within 250-200 hPa altitude range. The enhancements are shown in both panels of Fig. 1. This altitude level is below tropopause level marked by black cross lines shown in Fig. 2. High relative humidity at this cruise level over the region of elevated O3 VMR is also observed. The question needs to be answered is then from where the high O3 VMR of up to 190 ppbv at 21 UTC originates. To answer this question, analysis of 5 days back trajectory on isentropic surfaces from online NOAA HYSPLIT was made. Then the observed ozone enhancement WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

415

Figure 1:

MOZAIC cruise ozone (solid line) and relative humidity (dash line) spikes observed on 27 February 1996 and 27 March 1997.

Figure 2:

Potential vorticity vertical cross section at 25.5 E between 20 S20 N on 27 February 1996 (left panel) and 27 March 1997 (right panel) at 18 UTC. (The black cross-marks indicate tropopause level).

was traced along the trajectory and matched with SAGE ozone record (not shown) to determine whether horizontal transport contributed to the enhancement or not. The result of this analysis indicates that this air mass in the five days period had not encountered O3 rich air mass of similar concentration before 21 UTC. The other possible source of O3 rich air mass is stratospheric intrusion. Since the relative humidity content of the stratosphere is very low in addition to stratospheric intrusion there has to be other physical process happening in conjunction with stratospheric intrusion in the troposphere. Thunderstorm activity would be one of the most likely events that could produce high relative humidity record in the order of 100% in the vicinity of high ozone observation. To determine whether this is the case or not, we have investigated potential vorticity, vertical velocity, O3 VMR, OLR, cloud cover, cloud liquid water content and latent heat flux data set of the same day. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

416 Air Pollution XIX

4 Results and discussion To address the premises outlined above, first we investigated the existence of stratospheric intrusion at the area of the flight prior to the plane traversing the route. The two events occurred during 222030 -222454 UTC on 27 March 1997 and 213832-214716 UTC on 27 February 1996. Therefore we used ECMWF data at 12 and 18 UTCs ahead of the events. The potential vorticities for 27 February 1996 and 27 March 1997 at 18 UTC are depicted in Figure 2. High potential vorticity value greater than 2 PVU is shown well below the tropopause marker in both panels of figures. This shows the existence of stratospheric intrusions which are capable of inducing high concentration ozone of stratospheric origin to the troposphere. Secondly we investigated the existence convective air movement during this time. For this vertical wind (not shown here) and OLR values with in this time frame (12-18 UTC) were checked. OLR which is a proxy of deep convection for 27 February 1996 is shown in Figure 3 (right panel). Both vertical wind fields and OLR values for both days show the existence convection. Since no lightning data available during this time, we have used additional proxy parameters, such as cloud cover, cloud liquid water content and latent heat, which are indicators of the existence of thunderstorm activity. Nearly overcast cloud cover was observed at both 12 and 18 UTC (not shown) with higher values being at 12 UTC.

Figure 3:

Latent heat (left panel) and OLR (right panel) on 27 February 1997.

Thunderstorm is associated with release of very high latent heat. On 27 February 1996, very strong latent heat in the region of cloud cover was observed. For instance, at 15.5 S, the latent heat records at 09, 15, 12 and 18 UTC were 342.52, 502.50, 267.5, 57.38 W/m2 respectively (see Figure 3, left panel). Again for the 27 March 1997, the latent heat flux shows similar features (not shown here). From these records it is apparent that thunderstorm dissipation likely occurred between 12 and 15 UTC because the heat flux grew from 09 to 12 UTC WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Figure 4:

417

Ozone vertical cross section at 25.5 E on 27 February 996 (left panel) and 27 March 1997 (right panel) at 18 UTC.

and then decreased in between 12 and 15 UTC. In addition cloud liquid water transport, another indicator of thunderstorm activity. Observation of cloud liquid water content during both days at 12 and 18 UTC shows a clear ascent of cloud liquid water content during 18 UTC. According to Price and Asfur [20], equatorial Africa is one of the three thunderstorm hot spot regions (South America, Africa, and Southeast Asia) where most thunderstorm events take place around 14.00 UTC. Therefore our observations are in agreement with previous finding. The vertical ozone mass mixing ratio profile is shown in Figure 4 (for clarity the altitude is restricted to 900-100 hPa layer). During thunderstorm, electrical current causes the split in the oxygen molecules to atomic oxygen then further reaction of oxygen atom with oxygen molecule produces ozone. Ozone produced by this mechanism is very low. The lightning event also favors high nitrogen oxide (NO) production. NO reacts with ozone which leads to the production of nitrogen dioxide (NO2) which is manifested in a decrease of O3 concentration. The net result is reduction of ozone within the thunderstorm cloud. Two clearly identified ozone events are shown in both panels of Figure 4. Very low ozone concentration in the region of cloud cover shows scavenging during both events over the whole equatorial latitudes. However, significant O3 VMR loss were confined to latitude bands of 5-16 S, 7-8 N on 27 February 1996 and 8-16 S on 27 March 1997. Theses regions were also sites of intense thunderstorm activity as revealed by proxy parameters. On the other hand, during both MOZAIC observations and GOME data, enhanced O3 VMR was detected at upper troposphere. These were associated with PV intrusions. As discussed in the preceding paragraphs, part of the intruded stratospheric O3 VMR was able to descend down to boundary layer over the cloud free regions.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

418 Air Pollution XIX

5 Conclusions Enhanced MOZAIC ozone and relative humidity spikes observed over equatorial Africa in the year 1996-97 have been analyzed. Stratospheric intrusion coupled to thunderstorm activity found to be the key physical process that attributed to spiky O3 VMR and relative humidity. Potential vorticity values exceeding 2 PVU observed below the tropopause level confirms stratospheric intrusion. The vertical air movement, OLR value, very high latent heat release, cloud liquid water transport indicate the existence of thunderstorm activity during the PV intrusion. ECMWF ERA-interim ozone data were used to investigate the possible ozone distribution within the troposphere during these two dynamical events. Very low ozone of the order about 20 ppbv within the thunderstorm cloud was observed during these events. The low ozone level observed within the thunderstorm cloud is in agreement with previous reports. The other observed feature was very high ozone of stratospheric origin (>200 ppbv) over the upper equatorial troposphere during both observations. Our analyses have revealed that these were associated with PV intrusions. The intruding stratospheric ozone was able to make its way down to boundary layer outside the thunderstorm cloud. These processes have resulted in ambient air pollution with O3 VMR exceeding the proposed WHO guideline limit of 60 ppbv.

References [1] Prather, M. & Ehhalt, D., 2001. Atmospheric Chemistry and Greenhouse Gases in Climate Change 2001: The Scientific Bases, (Eds.) by J. T. Houghton et al., Cambridge University Press, Cambridge. [2] Report on a WHO Working Group, Health Aspects of Air Pollution with Particulate Matter, Ozone and Nitrogen Dioxide. Bonn, Germany 13–15, 2003. [3] Regener, V. H., Vertical flux of atmospheric ozone, J. Geophys. Res., 62, pp. 221–228, 1957. [4] Junge, C. E., Global ozone budget and exchange between stratosphere and troposphere. Tellus, 14, pp. 363–377, 1962. [5] Danielsen, E. F., Stratospheric-tropospheric exchange based on radioactivity, ozone and potential Vorticity. J. Atmos. Sci., 25, pp. 502– 528, 1968. [6] Danielsen, E. F., Hiskind, R. S., Gaines, S. E., Sachse, G. W., Gregory, G. L., and Hill, G. F., Three-dimensional analysis of potential vorticity associated with tropopause folds and observed variations of ozone and carbon monoxide. J. Geophys. Res., 92, pp. 2103–2111, 1987. [7] Lamarque, J. F. and Hess, P. G., Cross-tropopause mass exchange and potential vorticity budget in a simulated tropopause folding. J. Atmos. Sci., 51, pp. 2246–2269, 1994.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

419

[8] Appenzeller, C., Holton, J. R., and Rosenlof, K.H., Seasonal variation of mass transport across the tropopause, J. Geophys. Res., 101(D10), pp.15071–15 078, 1996. [9] Gouget, H., Vaughan, G., Marenco, A., and Smit, H. G. J., Decay of a cutoff low and contribution stratosphere-troposphere exchange. Q. J. R. Meteorol. Soc., 126, pp. 1117–1141, 2000. [10] Sauvage, B., Thouret, V., Cammas, J.-P., Gheusi, F., Athier, G., and Nédélec, P., Tropospheric ozone over Equatorial Africa: regional aspects from the MOZAIC data. Atmos. Chem. Phys. Discuss., 4, pp. 3285–3332, 2004. [11] Grewe, V., Impact of climate variability on tropospheric ozone. Sci. Total Environ., 374, pp.167–181, 2007. [12] Dickerson, R. R. et al., Thunderstorms: An important mechanism in the transport of air pollutants. Science, 235, pp. 460–464, 1987. [13] Poulida, O. et al., Stratosphere-troposphere exchange in a midlatitude mesoscale convective complex 1. Observations. J. Geophys. Res., 101(D3), pp. 6823–6836, 1996. [14] Hauf, T. et al., Rapid vertical traces gas transport by an isolated multitude thunderstorm, J. Geophys. Res., 100(D11), pp. 22957–22970, 1995. [15] Ridley, B. A. et al., Distribution of NO, NOx, NOy, and O3 at 12 km altitude during the summer monsoon season over New Mexico, J. Geophys. Res., 99(D12), pp. 25519–25534, 1994. [16] Huntrieser, H. et al., Transport and production of NOx in electrified thunderstorms: Survey of previous studies and new observations at midlatitudes, J. Geophys. Res., 103(21), pp. 28247–28264, 1998. [17] Drapcho, D. L. et al., Nitrogen fixation by lightning activity in a thunderstorm. Atmos. Environm., 17(4), pp. 729–734, 1983. [18] Wang, Y. et al., Nitric oxide production by simulated lightning: Dependence on current, energy, and pressure. J. Geophys. Res., 103(D15), pp. 19149–19159, 1998. [19] Price, C., Lightning Sensors for Observing, Tracking and Now Casting Severe Weather. Sensors, 8, pp.157–170, 2008. [20] Price, C., and Asfur, M., Can Lightning Observations be used as an Indicator of Upper-Tropospheric Water Vapor Variability? American Metrological Society, pp. 291–298, 2006. [21] Marufu, L., Dentener, F., Lelieveld, J., Andreae, M. O., and Helas, G., Photochemistry of the African troposphere: Influence of biomass-burning emissions, J. Geophys. Res., 105(2248) , pp. 14513–14530, 2000. [22] Thouret, V., Marenco, A., Nédélec, P., and Grouhel, C., Ozone climatologies at 9-12 km altitude as seen by the MOZAIC airborne program between September 1994 and August 1996, J. Geophys. Res., 103(25), pp. 653–679, 1998a. [23] Thouret, V., Marenco, A., Logan, J. A., Nédélec, P., and Grouhel, C., Comparisons of ozone measurements from the MOZAIC airborne program and the ozone sounding network at eight locations, J. Geophys. Res., 103(25), pp. 695–720, 1998b. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

420 Air Pollution XIX [24] ECMWF, Website, uk http://wcrp.ipsl.jussieu fr/Workshops/Reanalysis 2008/Documents/V1- 102_ea.pdf. [25] Mirador, Website, NASA Goddard, http://mirador.gsfc nasa.gov. [26] McCalla, C., Objective Determination of the Tropopause Using WMO Operational Definitions, U.S. Department of Commerce National Oceanic and Atmospheric Administration National Weather Service Metrological Center, October 1981.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Section 7 Economics of air pollution control

This page intentionally left blank

Air Pollution XIX

423

Environmental tools of atmospheric protection in the Czech Republic O. Malíková & M. Černíková Faculty of Economics, Technical University of Liberec, Czech Republic

Abstract The Czech’s relationship to the environment and its protection has undergone a relatively long development. In the late 20th century the Czech Republic changed the political system as well as its environmental behaviour. At the beginning of this period the Czech Republic was one of the most polluted regions of Europe. With the change of a political situation in 1989, problematic status of the air pollution in the Czech Republic has been significantly improved. Current status and positive progress of the environment in the context of a sustainable development is guaranteed by the EU legislation also in the Czech Republic. The aim of this paper is to discuss the status of atmospheric protection and its efficient tools used in the Czech Republic. In accordance with the sustainable development requirements, advanced countries establish systems of environmental tools affecting interaction between manufacturing corporations and the major components of the environment, especially air. The paper describes the system of atmospheric protection in the Czech Republic and administrative tools which are quite efficient and well applicable in a corporate environment. The tools however bring plenty of problems as well. The paper analyzes increased attention to economic tools, especially the potential of a tax system. Another issue that is recently discussed quite frequently is the so-called ecological tax reform which could also be used as another important environmental tool. A new dimension of the environmental regulation of atmospheric protection is represented by voluntary activities of the companies. For a successful combination of atmospheric environmental protection instruments in the Czech Republic it is necessary to assess their economic efficiency and environmental performance. The Czech Republic has proved the system for effective atmospheric protection which brings significant results. Despite the improvements, there are still developed and implemented new tools for atmospheric protection. Keywords: atmospheric protection, environmental tools, administrative tools, WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110391

424 Air Pollution XIX economic tools, tax reform, voluntary activities by companies, economic efficiency, environmental performance.

1 Introduction The contemporary activity of economic entities is always more or less adherent to the deterioration of all components of the environment, mainly air quality. The extent of not only local and regional, but also national and international negative impact on the environment demonstrates that we simply cannot rely on market mechanisms, hoping they will regulate such an impact on the environment automatically. The deficit in the functioning of market mechanisms is caused by a collective ownership of environmental goods that are globally used by the whole community [1]. In context of requirements on a sustainable development of a civilized community it is necessary to systematically control and manage interactions between the economic development and the state of the environment. In the activities of economic entities, especially in the developed industrial areas, the market balance has been impaired due to various reasons which lead to the occurrence of negative externalities relating to air pollution. The conflict between the interests of the polluter and concerns of the company results in a necessity for the management of the conflict using various environmental tools of atmospheric protection. This article presents relevant groups of environmental tools implemented in the Czech Republic environment and examines approaches to the evaluation of the success relating to the implementation of this atmospheric protection regulation.

2 Atmospheric quality protection in the Czech Republic The relationship with the environment has been developing over a relatively long period in the Czech Republic. At the end of the 80´s the Czech Republic was facing not only problems related to the change of a political system, but also a big devastation of the environment as a result of a complete absence of an ecological policy and other aspects in a socialist government decision-making. For many decades the main priority of ‘building the system of socialism’ remained a quantitative growth of mainly primary sector and heavy industry without any respect to the environmental protection requirements; due to this attitude the Czech Republic was one of the most polluted regions in Europe with the highest emission of a wide range of pollutants in the air. The change of a political system in 1990 brought about a significant improvement of air quality in the Czech Republic. Figure 1 shows the development of air pollution by major pollutants in the Czech Republic within the past twenty years [2]. The main reasons for a more satisfactory air condition in the Czech Republic were mainly the decrease of industrial activities in the first few years after November 1989, enforcing key laws related to the environmental protection, founding relevant environmental institutions and creating an effective system of

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

425

1 600 000 1 400 000 1 200 000 1 000 000

solid particles (ton/year) SO2 (ton/year)

800 000

NOx (ton/year) CO(ton/year)

600 000 400 000 200 000

08

07

20

06

20

05

20

20

03 04 20

02

20

01

20

00

20

20

98 99 19

97

19

96

19

95

19

94

19

93

19

92

19

91

19

90

19

19

19

89

0

Source: author´s submission, the Czech statistical Office [3]. Figure 1:

Development of air pollution by major pollutants in the Czech Republic.

environmental protection. The sulphur oxide air pollution was significantly reduced. Companies emitting these pollutants to the air were forced to take measures in a relatively short time to meet strict limits, which in some cases lead to cuts in production or even its closing. Other pollutants that did not reach such high levels are still present in the air and their reduction is problematic for many reasons not only in the Czech Republic, but also in Europe – the situation of air pollution has been steady for the last few years. A number of measures have been taken leading to a higher atmospheric protection since the Czech Republic joined the EU. A present state and a satisfactory development of atmospheric protection in the Czech Republic is guaranteed by the need to conform to the European Union legislation.

3 Environmental tools of atmospheric protection In the past few decades multiple environmental tools were implemented in the Czech Republic with various levels of effectiveness in relation to the air quality. These tools may be divided into several groups. 3.1 Normative (administrative) tools Normative tools of the environmental policy are based on the compulsory authority of state administration bodies. Such tools especially cover:   

directives (orders and prohibitions), limits (material, time), rules and technical standards.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

426 Air Pollution XIX Regulation system of atmospheric protection is based on the legislatively supported administrative measures intended to affect the behavior of polluters, their control and imposition of possible sanctions (fines, penalties) in case of non-performance of the legislative requirements. These tools are environmentally effective, however they are accompanied with quite a lot deficiencies [4]. Direct regulation requires extensive administration and subsequent inspection. This means a risk of inefficient bureaucratic management, corruption and needless expenses. Despite many disadvantages and difficulties, normative tools have been so far the most expanded and the most widely used tools in the field of atmospheric protection. Direct regulation is especially effective in the areas where the goals must be achieved immediately. The current requirements for environmental maturity of economic subjects however require new attitudes that offer multiple options and promote voluntary approaches. 3.2 Economic tools Economic tools of atmospheric protection are based on the indirect affecting the behavior of economic subjects causing harm to the environment [5]. Each economic subject (companies, households) may decide whether it is more advantageous to expend some money on mitigation or elimination of impacts on the environment or to damage the environment and pay for it in the form of fees and taxes. New environmental protection system that has been developing in the Czech Republic since 1990 makes use of various economic tools for the environmental protection. Amongst the most significant economic tools we count for instance the fees for atmospheric protection; in the recent years this issue is also more increasingly considered in the governmental tax system. 3.2.1 Fees designated for the atmospheric protection The most widespread form of economic tools of atmospheric protection in the Czech Republic are the tools of so called negative stimulation, especially the fees designated for the atmospheric protection. The amount of such fees usually depends on the quantity and concentration of the emissions released to the air. The fees are considered as environmental costs of companies – polluters. The fee should motivate the companies - polluters to adopt such measures that will contribute to atmospheric protection. The amount of the fee should be based on the costs for the elimination of damage, so that no economic advantage is given to those organizations that pay lower amount than the potential environmental damage. The fee should be also a tool for giving a priority in the market to those companies that act responsibly and protect the atmosphere. Their market position should be then stronger [6]. 3.2.2 The Czech Republic tax system Another economic tool that is contributing to atmospheric protection is the Czech Republic tax system, with many more standard functions (allocation, distribution, fiscal). In this way the functions of the tax system are extended thus the tax system not only acts as a fiscal tool that brings certain tax revenues

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

427

into public budgets, but also a tool of the state environmental policy aimed at the limitation of air quality deterioration [7]. As for the current tax legislation of the Czech Republic, environmental aspects are incorporated in the direct taxes section in form of multiple tax allowances, disallowances or exemptions facilitating the environmental protection process [8]. The crux of the environmental regulation is currently moving mainly to the area of indirect taxes. Indirect taxes have imminent effect on the price of the goods they are imposed on – thus they can greener production with lower emissions. The move to indirect taxes should be – in accordance with the principle of tax neutrality – compensated by reduction of other taxes. In the end this should contribute to the growth of national economy. Reduction of labor costs should help to maintain competitiveness of our economy. At the same time it should represent a kind of stimulus attracting foreign investors. In the EU countries this conception (so called ecological tax reform) has been discussed for many years as one of the potential approaches to improvement of the quality of environment and also to lowering air emissions. Ecological tax reform should be implemented in the Czech Republic in three successive phases in the period 2007–2017. Just as in other EU countries also in the Czech Republic the reform should be based on the principle of tax neutrality, i.e. revenues should be used for the reduction of labor costs. The first phase of the ecological tax reform was completed by introduction of the Council Directive 2003/96/EC into the Czech Republic legislation. Since 2008 the existing tax system of the Czech Republic was amended by taxation of energy products and electricity through the Act No. 261/2007 Coll. The subject of taxation covers natural gas, solid fuels and electricity [9]. Reduction of labor costs in 2008 was realized by abatement of income tax of both physical and legal entities. Quite extensive amendment to the income tax legislation changed the system for determination of tax base and tax allowances and also lowered the social insurance rates. The second phase of the ecological tax reform should be realized in the period 2010 – 2013. During this phase tax rates and allowances for solid fuels and natural gas should be changed. The main goal is to revise the existing fees and other tools with an option to transform the selected environmental fees to environmental taxes. Special attention should be paid to traffic, as one of the main air polluters, but also to administration of environmental taxes or transformed fees, especially to the principle of tax neutrality. The third phase is scheduled to the period of 2014–2017. In general this phase should correspond with the amended EU directives. Its conception should be based on the analysis of the two previous phases, assessment of all their effects and impacts and probably it will also be based on the political situation in the Czech Republic in the forthcoming decade [10]. So far the economic tools do not represent the core area of atmospheric protection in the Czech Republic. In practice there is still a significant majority of normative tools, while economic tools are rather perceived as a certain supplement to the normative ones. At creation of specific atmospheric protection WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

428 Air Pollution XIX policies, both groups of tools are not considered as equal. The main reason for the existence of economic tools is mainly their fiscal effect [11]. 3.3 Voluntary environmental activities by organizations With the requirement for a sustainable development and the constantly increasing environmental awareness new voluntary approaches to atmospheric protection have been emerging since the eighties of the last century. Voluntary activities support the strategy of preventive approaches realized by companies beyond the scope of the statutory legislation framework with the aim to mitigate negative impacts of their activities on the environment, to consolidate their market position, to improve their competitiveness as well as the company image [12]. This is a significant turnover in the perception of atmospheric protection. Small effect of sanctions or various subsequent recourses leads to a deflection from these traditional tools and inclination to voluntary preventive actions that are introduced into decision-making processes at the company level, affecting each and every activity of the company that is related to the environment. Voluntary activities are consolidated at the international level and the wide range of tools implemented also by business entities in the Czech Republic [13]. Voluntary environmental activities of companies may be considered as voluntary tools and recommended approaches. Implementation of voluntary tools into company practice is a good opportunity, however there may be some risks and doubts that will probably appear and will be solved in near future. Should the voluntary activities be a fullfledged instrument for the meeting of atmospheric protection goals of companies, the companies must be mature and responsible enough to be able to deal with the environmental protection individually, but at the same time in accordance with the goals and directions of the governmental policy. This is one of the reasons why we must state that in Czech Republic the process of implementation of voluntary activities stands at the beginning of a relatively long way.

4 Effectiveness of atmospheric protection tools To achieve the set environmental goals, we can use a wide range of tools combining especially the normative and economic tools, but also some voluntary approaches of companies. Each of these tools has its positives, but the implementation always brings some negatives as well. While preparing its own atmospheric protection conception the government decides about the effective and suitable combination of all the tools in order to achieve the selected environmental goals [14]. Should this selection be rational, certain criteria must be considered in order to choose suitable tools in context with the goals set (see Fig. 2.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

429

Criteria for the environmental tools  assessment

Other effects: Economic effect Environmental effects

-legislative framework

static

-administrative functionality

dynamic

-political acceptance -competitive perfection

Source: author’s submission Figure 2:

Potential criteria for the atmospheric protection tools assessment.

Environmental effects – the tools must demonstrably fulfill the requirements that were set in the field of atmospheric protection. Economic effect – implementation of the tools should be connected with minimum community costs (administrative, legislative, inspection, etc), but incentives for the elimination of environmental deterioration should be clear and obvious – so called static effect. At the same time there should be some space left for innovations of companies in this field (implementation of more environmentally friendly technologies, green products) - this may give a strong boost to the economic effect of the environmental tools (dynamic effect) [15]. Legislative framework – all activities connected with the implementation of environmental tools cannot be realized within a single insulated legislative framework - national legislation must be interconnected with EU legislation and environmental protection issues must be dealt with in a complex way. Administrative functionality – there is a requirement for simple implementation and administration of environmental tools. Good knowledge of the tools is a necessary presumption for proper administration. Political acceptance - in a broad sense this refers to a general feasibility of the environmental tools. In the strict sense of the word particular environmental tools must be accepted by various groups concerned (political groups, state administration, companies, private parties concerned, etc) for their successful implementation. A necessary precondition is their political transparency and comprehensibility. Competitive perfection – the tools must not create any competitive barriers or limits that would be in conflict with the applicable international legislation (especially the European regulations and directives on restrictive trade practices). The tools should not prevent new entities from entering the national (or international) market or tend to creation of monopolies [16]. In order to achieve the set goals in atmospheric protection area, it is necessary to make use of a suitable combination of the available tools. Should the selection be rational, certain criteria must be considered in terms of suitability of particular

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

430 Air Pollution XIX tools. Besides environmental effects we must also analyze economic effect and other aspects of each environmental tool to be implemented.

5 Conclusion Companies should assume responsibility for contamination and deterioration of the air caused by their activities and should include these negative externalities into their costs. Internalization of externalities is especially possible through tools that are implemented by government and relevant authorities through the national environmental policy. Normative tools are relatively used with more frequency, while economic environmental tools bear quite a big potential. There is however also a real risk that the primary environmental effect of these tools will be reduced by other secondary effects that are also expected from the economic tools. This is also supported by the fact that the main motivation of government, while implementing economic tools, is their fiscal effect. Low efficiency of economic tools reflects inadequate evaluation of economical aspects of environmental protection. Also the interconnection between environmental goals and tax legislation is not optimal as the current tax incentives are rather indistinctive and not much motivating for the companies. The conception of the ecological tax reform – with its main idea “from taxation of labor to taxation of environmentally unfriendly products and services” – is quite interesting and the European countries adopt it as a potential tool for the satisfaction of the sustainable development goals in a long-term horizon. Implementation of reform conceptions however always bring some risks, for instance it is not possible to get some accurate estimation of revenues from this reform, the direction to internalization of negative externalities is not obvious and also the basic attribute of the reform – tax neutrality is, as far as I think, rather controversial. Companies should set a new dimension in their interaction with the environment and through a wide selection of voluntary activities control, evaluate and guide their environmental behavior.

Acknowledgements This article was worked up as one of the outputs of the research project “Environmental Tax Reform in the Context of Environmental Policy of the Czech Republic”, which was implemented at the Faculty of Economics of Technical University in Liberec in 2011 with the financial support from the Technical University in the competition supporting specific projects of academic research (student grant competition).

References [1] Villlacampa, Y. & Brebbia, C.A., (eds). Ecosystems Sustainable Development VIII, WIT Press: Ashurst, 2011.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

and

Air Pollution XIX

431

[2] Zprávy o životním prostředí v jednotlivých letech. Ministerstvo životního prostředí, Online. www.mzp.cz [3] Stav životního prostředí v jednotlivých letech. Český statistický úřad, Online. www.czso.cz/csu/edicniplan nsf [4] Wiener, J., Global Environmental Regulation: Instrument Choice in Legal Kontext. Yale Law journal, (108), p. 983, 1999. [5] Endres, A., Umweltökonomie: Arbeits - und Übungsbuch, Kohlhammer: Stuttgart, pp. 120–150, 2007. [6] Nástroje environmentální politiky. Centrum pro otázky životního prostředí, Online. www.enviwiki.cz [7] Ščasný, M., Administrativní a vyvolané náklady poplatků za znečisťování ovzduší a energetické daně. Proc. Of the 1st Int. Conf. On Modelování dopadů environmentální regulace, ed. M. Ščasný, COŽP Univerzity Karlovy: Praha, pp. 9–11, 2010. [8] Bosquet, B. Environmental tax reform: does it work? Ecological Economics, 1(34), pp. 19–32, 2000. [9] Marková, H. Daňové zákony 2011. Grada Publishing: Praha, pp 215–228, 2011. [10] Principy a harmonogram ekologické daňové reformy. MŽP ČR, Online. wwwmzp.cz [11] Poterba, J., Tax policy and the Ekonomy, MITT Press: Massachusetts, pp. 150–182, 2006. [12] Burrit, R. & Schaltegger, S., An Introduction to Corporate Environmental Management Striving for Sustainability, Greenleaf Publishing: Sheffield, pp. 58–72, 2003. [13] Remtová, K., Dobrovolné environmentální aktivity. Orientační příručka pro podnikatele Planeta, 6(14), pp 3–27, 2006. [14] Stehling, F., Ökonomische Instrumente der Umweltpolitik zur Reduzierung stofflicher Emissionen, Universität Ulm: Ulm, pp. 150–170, 1999. [15] Burrit, R., & Schaltegger, S., An Introduction to Corporate Environmental Management Striving for Sustainability, Greenleaf Publishing: Sheffield, pp. 85–112, 2003. [16] Fitzgerald, J., Assessing Vulnerability of Selected Sectors under Environmental Tax Reform: The Issue of Pricing Power. Journal of Environmental Planning and Management, 3(52), pp. 413–533, 2009.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

This page intentionally left blank

Air Pollution XIX

433

Environmentally related impacts on financial reporting: the case of pollution permits in Czech legislative conditions J. Horák & O. Malíková Faculty of Economics, Technical University of Liberec, Czech Republic

Abstract The Czech Republic was one of the biggest air polluters in Europe before the year 1989. Czech industry was restructured from heavy to light industry as a consequence of the transition from a centrally planned economy to a market economy after the political, social, economic and other changes that took place that year. The Czech Republic became a member of the European Union in 2004. Due to the membership it was crucial to implement environmental legislation that ensures the reduction of greenhouse gas emissions. The submitted paper deals with problems of pollution permits from the view of Czech accounting legislation. The purpose of financial accounting is to generate financial information about a company in order to provide a basis for transparency and accountability relationships with stakeholders. Financial reporting is used by managers to communicate the dated financial information to external parties. The main environmental topics in a company’s financial accounting are the recognition, measurement and disclosure of environmentally related economic impacts on business and a company value. Local financial accounting standards differ between jurisdictions and can substantially influence the economic results of a company. When examining sustainable development and accountability it is crucial to investigate how environmental issues are dealt with by financial reporting rules, whether and when environmentally induced financial outlays should be classified as assets or as expenses. The aim of this paper is to discuss the influence of pollution permits on Czech financial reporting with the view of accounting entity that pollutes air by greenhouse gases. The paper describes problems of a present situation of recording accounting transactions connected with pollution permits such as a purchase or donation of permits, their evidence in Czech financial accounting. The paper analyzes implementation of the EU WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110401

434 Air Pollution XIX environmental legislation into Czech national law and it is focused on Czech companies that have an obligation to obtain pollution permits. This situation influences financial accounting and reporting of financial statements. Keywords: air pollution, pollution permits, emission allowances, environmental accounting, financial reporting, intangible fixed assets, expenses, emissions trading, balance sheet, greenhouse gases, environmental legislation.

1 Introduction The submitted paper deals with problems of air pollution by greenhouse gases and the tools that should decrease their amount in the atmosphere. The first part of the paper is focused on emissions of CO2 in the Czech Republic, the second section discusses auctioning of pollution permits (allowances) in the European Union and the third part analyses described problems in Czech conditions with a focus on financial reporting. The Kyoto Protocol is an international environmental treaty that concentrates on the decreasing of greenhouse gases such as carbon dioxide (CO2), nitrous oxide (N2O), methane (CH4), perfluorocarbons (PFCs), hydroflourocarbons (HFCs) and sulphur hexafluoride (SF6) in the atmosphere [1]. Since the European Union joined the Kyoto Protocol, it was important to create a specific tool (the Emissions Trading Scheme) that could ensure reduction of greenhouse gases produced by companies situated in the European Union. The system focuses on companies with installations with a net heat supply more than 20MW (e.g. power plants), steel production (capacity exceeds 2.5 tons per hour), cement production (capacity exceeds 50 tons per day), glass production (capacity exceeds 20 tons per day), production of ceramic products (capacity exceeds 75 tons per day) and production of paper if capacity exceeds 20 tons per day [2]. Airlines flying to or from Europe will join the Emissions Trading Scheme in 2012. With ratification of the 2002 Kyoto Protocol all European member states have to reduce their emissions of CO2 to 8% below the emissions level reached in 1990 by 2012 [3]. Due to meeting of criteria of the Kyoto Protocol by joined countries there are other systems that should ensure reduction of emission of greenhouse gases all over the world such as systems valid in the USA, Australia, Japan, New Zealand etc. The submitted paper focuses on the European Union especially on the Czech Republic and analyses the impact of emission allowances on financial reporting.

2 Emissions of CO2 in the Czech Republic The Czech Republic was one of the largest polluters in Europe before 1990. This situation was caused by the structure of economy that was primarily focused on a heavy industry. After the political changes in 1989 there came changes in the structure of economy with specialization in light industry and services. Figure 1 shows the development of total emissions of CO2 in the EU 27, EU 15 and in the Czech Republic. The trend of emissions of CO2 in European countries are decreasing, the only exception was the period between the years WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

435

2003 and 2006. This situation could have been caused by economic conditions before the financial crisis that occurred approximately from 2007 to 2010 and a possible efficiency of emission allowances. During the financial crisis the amount of production of goods and services dramatically decreased in all member states of the European Union.

Figure 1:

Total emissions of CO2 in EU 27, EU 15 and in the Czech Republic from 1990 to 2009. (Source: http://dataservice.eea.europa.eu /PivotApp/pivot.aspx?pivotid=475 [4].)

Figure 2 shows emission shares of CO2 in member states of the European Union. As can be seen, the Czech Republic (approximately 10 000 000 inhabitants) is the eighth largest polluter in EU 27 with emissions of CO2 in the amount of 113,388 Tg (million tons). The emission share of the Czech Republic is 3.0%. The largest polluter of the EU 27 is Germany with a share of 21.0% and the least polluting country is Malta with a share of 0.1%. The table 1 shows the amount of emission allowances of five companies with the highest amount of CO2 pollution in the Czech Republic. These companies have obtained presented allowances in the second phase of implementation of the European Union Emission Trading Scheme (2008 – 2012). The largest Czech polluter is a company called ArcelorMittal Ostrava a. s. This company is a producer of hot metal, steel and rolled products. Power plants of the company CEZ, a. s., which produces electrical energy, takes the second to the fourth position of the largest polluters. The latter is the largest polluter of CO2 in the Czech Republic in general. The fifth largest polluter is a coal power plant of Sokolovska uhelna, a. s. – Vresova.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

436 Air Pollution XIX

Figure 2:

Emission shares of CO2 by country in EU 27 in 2009. (Source: http://dataservice.eea.europa.eu /PivotApp/pivot.aspx?pivotid=475 [4].)

Table 1: Order 1. 2. 3. 4.

Five largest polluters of CO2 in the Czech Republic. Company

Amount of Allowances (pcs) 6,958,508 6,696,795 6,335,369 5,309,817

ArcelorMittal Ostrava a.s. CEZ, a. s. – Power plant Pocerady CEZ, a. s. – Power plant Prunerov 2 CEZ, a. s. – Power plant Tusimice 2 Sokolovska uhelna, a. s. – Power plant 5. 4,478,948 Vresova Source: http://www.energostat.cz/emisni-povolenky-jako-vazny-problemenergetiky-v-cr.html [5].

3 Emission trading in the European Union The main goal of the European Union Emission Trading Scheme (EU ETS) is to reduce and control the air pollution by carbon dioxide (CO2) caused by all member states in the EU. A trade with allowances is a market based tool [6]. This trading scheme is one of the largest in the world and it is connected with meeting the criteria set in the Kyoto Protocol that the EU joined. This system is based on a cost-effective and an economically efficient method that tries to

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

437

ensure competitiveness of the European companies obliged to obtain allowances for their production [7]. Due to the costs, time period for preparation, information of companies that are obliged to obtain pollution permits, preparing infrastructure for monitoring of actual emissions, their reporting and verification and of course due to other important reasons the process of implementation was divided into three basic phases that will take place 15 years in total. The first phase started on 1st January 2005 and finished on 31st December 2007. This three-year phase was important for a preparation of the following two phases. This time period led to the establishment of price of carbon (the price of allowances is generally set by supply and demand at the time of transaction) and set of caps on national allocations of allowances. These caps have made traded allowances scarce and have forced all companies to reduce their emissions of green house gasses or buy other allowances. However, these transactions increase the costs of the companies and influence the profit or loss of that subject. All the EU member states had to prepare their own national allocation plans that set the total quantity of allowances for each period according to the ETS Directive. At least 95% of allowances were given to the companies free of charge during this period. The second phase of implementation is dated from 1st January 2008 to 31st December 2012. This five-year period is crucial for all the EU member states that must implement rules into the national law and must comply with emissions targets under the Kyoto Protocol. At least 90% of all allowances were given to the companies free of charge during this period. It is possible to state that present allocation scheme is based mainly on grandfathering of allowances because only 10% of allowances are being auctioned. The final eight-year phase will start on 1st January 2013 and will be crucial for trading with the allowances in comparison with last two phases that were based mainly on their grandfathering. The system of auctioned allowances will increase and change the structure of the market. All involved companies will have to control the quantity of allowances and improve them according to their actual needs. If there is a lack of them, they must buy appropriate amount of allowances or decrease emissions of greenhouse gases or decrease their market share during this time period [8]. In Cihakova et al. [9] the authors present that the empirical studies encourage the implementation of free market tools in order to achieve the solution of determinate environmental problems and free market environmentalism is not based on taxation or other kinds of environmental regulations. As the system of implementation of emission allowances is set by the EU, a question arises of whether this situation could not harm the industry and competitiveness of European companies in comparison with other producers mainly from Asia that are not obliged to reduce their emissions of greenhouse gases.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

438 Air Pollution XIX

4 Pollution permits – environmental issue in financial accounting and reporting under the Czech legislation Financial accounting is a tool to collect the information required to be disclosed to external stakeholders of a company, whereas a financial report is a platform for sharing this information. Financial report consists of financial statements which give a picture of a financial position of the company (balance sheet), its performance (income statement), cash flow etc. In the past, environmentally related issues were not included in financial reporting. This has changed with the increasing extent of environmental costs, environmental management started to become an important issue in financial markets during the 1990s [10]. The main environmental topics in a company’s financial accounting are the recognition, measurement and disclosure of environmentally related economic impacts on business and a company value. Environmental issues in financial accounting and reporting are concerned with revenues and expenses influencing the performance which is shown in the income statement, as well as with assets and liabilities which are shown in the statement of the financial position (balance sheet). In order to meet the objectives of financial statements and to be comparable, reports are assumed to be prepared on the basis of general accounting concepts and principles but the economic results of a company can differ substantially because they are influenced by local financial accounting rules which differ among jurisdictions. 4.1 Pollution permits and tradable emission allowances as an accounting item In all countries, companies are allowed to pollute the environment as long as they do not exceed legal emission standards. The right to pollute is specifically certified by emission allowances. The total amount of pollution is strictly limited through the total number of pollution permits issued. An emission allowance is a certified right that allows a company to discharge a certain amount of pollution into the natural environment within a specified time limit. Emission allowance can be understood either as a licence to pollute or as a right to emit a specific level of pollutants in a specific period of time [10]. In 2003, the definitions were set by the European Parliament and the Council: “...‘allowance’ means an allowance to emit one tonne of carbon dioxide equivalent during a specified period, which shall be valid only for the purposes of meeting the requirements of … Directive and shall be transferable in accordance with the provisions of … Directive;…” [11]. Greenhouse gas emissions permit means the permit issued and it “... shall include a description of: (a) the installation and its activities including the technology used; (b) the raw and auxiliary materials, the use of which is likely to lead to emissions of gases …; (c) the sources of emissions of gases … from the installation; and (d) the measures planned to monitor and report emissions …” [11].

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

439

Proposed by the European Financial Reporting Advisory Group (EFRAG), emission allowance is recognised as intangible assets. According to Mladek, intangible assets are recognized by applying the following criteria:  It arises from law, contract or some other claims or obligations.  It does not arise from law or contract but is separable (can be sold, transferred, licensed, rented or exchanged) either individually or together with another assets or liabilities.  It does not arise from law or contract, is not separable but is the difference between the amount paid for a business and the fair value of that business’s net assets [12]. 4.2 Emission allowances under Czech accounting legislation Financial accounting in the Czech Republic is regulated by national legislation in three levels: Accounting act, Decree of the Czech Republic Ministry of Finance and Czech accounting standards. In terms of emission allowances, changes in the Czech accounting legislation took place in October 2005 responding to the Act No. 695/2004 Coll., on Conditions for Trading in Greenhouse Gas Emissions, in which Directive 2003/87/EC on Emissions Trading was transposed [2]. In the Czech accounting legislation there is no definition of an asset; intangibles do not exist in current assets and an item is recognized as a fixed intangible asset if its useful lifetime exceeds one year and acquisition costs exceed the value limit specified by the accounting entity. Rules for recognition of emission allowances differ – they are recognized as an intangible fixed asset always, regardless of their acquisition costs and useful lifetime [13, 14]. Currently in the Czech Republic the emission market as well as the auction is not active. Since 2006 related first holders have acquired emission allowances in the form of government grants. Such emission allowance is valued by replacement costs. In the Czech Republic the costs of the emission allowances are set by the Energy Regulatory Office according to the chosen spot market and its weighted arithmetical average of the settlement prices and volume of the emission trade. Since 2010 BleuNext Market in France has been functioning. There the trade volume of the emission allowances is the largest in the European Union. The average costs of the emission allowance were calculated for 358.25 CZK – 14.83 EUR in the year 2010 and 349.84 CZK – 14.48 EUR in the year 2011 per allowance [15, 16]. When the emission allowance is consumed the operating expenses increase and at the same time the value of the government grant decreases together with the increase of the operating revenue. A financial report (balance sheet) has to show the decrease of the asset value when an emission allowance is consumed. We can say that in a situation of consumption of the granted emission allowances in a full amount, the profit (loss) of accounting period is not influenced by these transactions. Acquisition of the emission allowances by the purchase is valued in acquisition costs, which includes all additional cost of the acquisition (i.e. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

440 Air Pollution XIX commissions etc.). A consumption of the allowances is valued by the FIFO method (“first in, first out”) only. If the company pollutes more than how many emission allowances it holds, it is recommended to disclose this difference as an estimated accrued item; on the other hand it is not recommended to create commissions for emission allowances. 4.3 Illustrative example of the accounting and reporting the emission allowances in Czech legislative conditions In the year 20XX operator receives government grant of 100 emission allowances valued by 40 EUR per allowance. He sells 30 allowances for 24 EUR per allowance. Then he purchases 10 allowances for 20 EUR per allowance and 10 for 28 EUR per allowance. Surrender of allowances is estimated at 85 pieces. On the balance sheet date, the market value of purchased allowances decreases to 16 EUR/piece. A received government grant increases the value of intangible fixed assets and grants related to assets to 4,000 EUR. A sale of 30 allowances increases bank account and revenues by a sale price of 720 EUR and at the same time an expense is increased and intangible assets decreased by acquisition costs of these allowances which is 1,200 EUR. By the same value of 1,200 EUR the grant is reduced and it influences the revenue as well. Purchases of the allowances at the emission market increase value of intangible fixed assets by 200 EUR (1st purchase) and 280 EUR (2nd purchase) and reduce our bank account. At the end of the year 20XX we have to disclose estimated accrued item related to surrender of purchased allowances which will be made in April 20XX+1. This transaction is valued according to FIFO valuation method (i.e. 10 pieces for 20 EUR/piece and 5 pieces for 28 EUR/piece). At the balance sheet date, the market value of purchased allowances decreases to 16 EUR/piece, we have to create adjustments to remaining emission allowances (5 pieces) which influence intangible fixed assets as well as operating expenses by 60 EUR (calculated according to the FIFO methods as well). The following table 2 illustrates the impact of emission allowances accounting on a financial reporting namely on a balance sheet. Table 2:

Balance Sheet reporting of emission allowances and related items. Balance Sheet as of 31st December, 20XX (in EUR) ASSETS LIABILITIES Fixed assets Equities

Emission allowances Adjustments to allowances

3280 60

Grants related assets

Current assets Bank account

Profit

240

∑ ASSETS 3460 Source: Authors’ own calculation.

320

Other/current liabilities 2800

Estimated accruals

340

∑ LIABILITIES

3460

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

441

The balance sheet shows a result of the chosen transaction connected with the emission allowances only. This simple example confirms that these transactions significantly influence the structure and values of reported asset and liability items. The performance of the company is influenced as well, namely in the part of operating income.

5 Conclusion It can be said that the Czech legislation has adopted European Union rules concerning pollution and greenhouse gas emission. The directive 2003/87/EC on Emissions Trading was transposed to the Czech Act No. 695/2004 Coll., on Emissions Trading. It has influenced the Czech financial accounting rules as well. Emissions of CO2 are decreasing and the situation could be caused either by accepted emission legislation or by a recession of national as well as international markets. Concerning the emission trades, the Czech market is not still active and the cost of the emission allowances is set by Energy Regulatory Office according to the chosen spot market and its weighted arithmetical average of the settlement prices and the volume of the emission trade. We could argue that from the point of view of the financial accounting, the recognition of emission allowances as intangible assets which is proposed by EFRAG and generally accepted is not entirely accurate. This item could be recognized also as marketable securities or inventory. In the Czech accounting system emission allowances are disclosed as fixed asset because intangibles are not placed in current assets. Finally, we can state that emission allowances and related transactions significantly influence a financial position and a performance of a company which is reported in its financial statements.

Acknowledgement This paper was written in accordance with a research project “Analysis of Processing Accounting Information with Focus on Satisfaction with Present Accounting Software Products” solved by Technical University of Liberec, Faculty of Economics.

References [1] Townley, R. S., So Much Carbon, So Little Time: State Options for Effective Regulation of Mobile Source Emissions of Greenhouse Gases. The University of Memphis Law Review, 39(1), pp. 193-228, 2008. [2] Act No. 695/2004 Coll., on Conditions for Trading in Greenhouse Gas Emissions. [3] Cló, S., The effectiveness of the EU Emissions Trading Scheme. Climate Policy, 9(3), pp. 227-241, 2009. [4] European Environment Agency, http://dataservice.eea.europa.eu WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

442 Air Pollution XIX [5] ENERGOSTAT, www.energostat.cz [6] Energy policy and the transition to a low-carbon economy. OECD Economic Surveys, 2009(13), 2009. [7] Nimmo, B. & Armellini, A., EU temporarily shuts emissions trading system; member states blamed. MCClatchy – Tribune Information Services, 2011. [8] Benz, E., Löschel, A. & Sturm, B., Auctioning of CO2 emission allowances in Phase 3 of the EU Emissions Trading Scheme. Climate Policy, 10(6), pp. 705-718, 2010. [9] Čihaková, S.A., Černíková, M. & Dubová, M., Historical contamination and brownfield management: what does econimic theory say about it? Villacampa, Y. & Brebbia, C.A. (eds). Ecosystems and Sustainable Development VIII, WIT Press: Ashurst, pp. 339-352, 2011. [10] Schaltegger, S. & Burritt, R., Contemporary Environmental Accounting: Issues, Concepts and Practice, Greenleaf Publishing Limited: UK, pp. 162203, 2000. [11] Directive 2003/87/EC of the European Parliament and of the Council, of 13 October 2003 establishing a scheme for greenhouse gas emission allowance trading within the Community and amending Council Directive 96/61/EC. Official Journal of the European Union, L 275/32-46 Online. eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2003:275:0032: 0032: EN:PDF [12] Mládek, R., Postupy účtování podle IFRS / IFRS Policies and Procedures, Leges: Praha, pp. 225-252, 2009. [13] Act No. 563/1991 Coll., on Accounting. www mvcr.cz [14] Decree of Czech Republic Ministry of Finance No. 500/2002 Coll., on Accounting of Entrepreneurs. www mfcr.cz [15] Average cost of emission allowances 2010. Energy Regulatory Office, http://www.eru.cz/user_data/files/cenova%20rozhodnuti/CR%20teplo/cena %20povolenky_2010.pdf [16] Average cost of emission allowances 2011. Energy Regulatory Office, http://www.eru.cz/user_data/files/sdeleni_elektro2/Oceneni%20povolenky_ darovaci%20dan_2011.pdf

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Section 8 Health effects

This page intentionally left blank

Air Pollution XIX

445

Analysis of lung cancer incidence relating to air pollution levels adjusting for cigarette smoking: a case-control study P. R. Band, H. Jiang & J. M. Zielinski Population Studies Division, Environmental Health Science and Research Bureau, Environmental and Radiation Health Sciences Directorate, Healthy Environments and Consumer Safety Branch, Health Canada, Ottawa, Canada

Abstract A case-control study (lung cancer: 2711; age and sex matched non-lung cancer controls: 2711) encompassing the years 1986-2004 was undertaken further to the results of increased mortality and cancer incidence in Windsor, Canada. The objective was to investigate associations between lung cancer incidence and exposure to air pollutants controlling for cigarette smoking and duration of residence. A nominal file of cases and controls ascertained by the Ontario Cancer Registry was obtained from Cancer Care Ontario; smoking information and addresses of residence within Windsor were obtained respectively from medical charts at the Windsor Regional Cancer Centre and from the Windsor City Directories. For each case and control and for each air pollutant, a cumulative exposure was calculated based on duration of residence at each address location; the centroid of postal codes of addresses was used as proxy for residential location. For each location, annual NO2 and SO2 levels were estimated, using Land Use Regression based on results from 54 monitors within Windsor. Results obtained from conditional likelihood regression for matched or stratified data, (including sex, age at diagnosis, numbers of cigarette smoked per day and duration of smoking as covariates) are presented. Keywords: air pollution, cigarette smoking, lung cancer, case-control study.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110411

446 Air Pollution XIX

1 Introduction Long-term exposure to air pollution particles (Dockery et al. [1], Pope et al. [2, 3], Beeson et al. [4]), nitrogen dioxide (Nyberg et al. [5]) and nitrogen oxides (Nafstad et al. [6, 7]) has been associated with an increased lung cancer risk. In January 2003, the Governments of Canada and of the United States launched the Canada-US Border Air Quality Strategy (BAQS) to assess health effects of populations exposed to air pollution in these border regions. The following key objectives were determined: collecting human health data; analyzing evidence of human health impacts and assessing the risk to human health exposed to air pollution. The municipality of Windsor, situated in the most southern part of the Province of Ontario, Canada, was selected as one of BAQS target because of its proximity to the city of Detroit, U.S.A, and because of traffic pollution with about 10,000 trucks crossing the bridge linking Windsor to Detroit each day. In a preliminary study, we reported an increased risk of lung cancer incidence and mortality in males and females residents of Windsor (Band et al. [8]). Results assessing health effects of air pollution exposure have serious limitations inherent to mortality and cancer incidence studies. In particular, information on smoking and duration of residence at specific residential locations is lacking. In this paper, we present preliminary results of a case-control study assessing the relationship between lung cancer and chronic exposure to SO2 and NO2, controlling for cigarette smoking and residential duration.

2 Feasibility study Prior to undertaking the project, a feasibility study was conducted to: a) review a random sample of 30 medical charts to verify if smoking information was reported by staff physicians and/or nurses; b) obtain precise information on medical file location and accessibility; c) obtain precise information on address and residential duration at given addresses in Windsor from public community records. Thirty randomly selected medical charts from lung cancer cases (20) and from non-lung cancer controls (10) diagnosed prior to 1990 were reviewed. Of the lung cancer cases, the number of cigarettes smoked per day and the years of smoking were available for 18 with one information missing; one case was a non-smoker. Of the controls, five were never smokers, one had no smoking information and in four, intensity and duration of smoking were available. Thus, in total, smoking information was recorded for 27 (90%). Medical charts were in electronic format at the Windsor Regional Cancer Centre since April 1999. Prior to that date, medical charts of patients no longer seen at the centre were stored out of site at a privately owned Records Management Company in Windsor. A review of the accessibility of data included in the Property Assessment files, Voter’s list, Windsor City Land Registry and Windsor City Directories was undertaken. The Windsor City Directories, available since the late 1800 were the only useful and readily available database from which address information over

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

447

time could be accurately obtained. It was thus concluded from the feasibility study that adequate information was available to proceed to the planned casecontrol study.

3 Method 3.1 General Population data for the City of Windsor and for enumeration areas within the City of Windsor was obtained from Statistics Canada annually from 1986 to 2004. A nominal file of cases and controls were obtained from the Ontario Cancer Registry, Cancer Care Ontario. All 2,711 lung cancer cases occurring during 1986 to 2004 inclusive among residents of the City of Windsor were included in the study. 2,711 randomly selected cancer controls and residents of the City of Windsor at time of diagnosis during the same time period were matched on age, sex and year of diagnosis. Lung cancer, non melanoma skin cancer and multiple primary cancers were excluded from the control group. Smoking information (smoking status, average number of cigarettes smoked per day, total number of years smoked, and similar information for pipe and cigars) was obtained from the medical charts of patients at the Windsor Regional Cancer Centre. Address at residence and residential duration was obtained from the Windsor City Directories as follows: using the patient’s name and address at diagnosis from the medical charts, addresses were then verified backward in the Windsor City Directories every five years until the patient’s name could no longer be found; addresses were then verified annually to obtain the last known address of residence in the City of Windsor. For each subject, the total number of years lived at a given address of residence was derived from this information. 3.2 Exposure levels Annual levels of NO2 and SO2 for the year 2004 were estimated by the Land Use Regression model for each postal code in the City of Windsor based on results from 54 monitors set-up under the BAQS (Wheeler [9]). Land Use Regression (LUR) uses measured air pollution concentrations from networks as the response variable, and land use types within circular areas, called buffers, as predictors of changes of the measured concentrations ((Jerrett [10]). The variables used to create the LUR included: land-use groups (such as open space, industrial, commercial and residential land use); road network groups (such as length of local and major roads, length of primary highways and expressways), population and dwelling groups, and Detroit and Windsor industrial point sources variables (Wheeler [9]). The 1986-2003 pollution data was limited to observations from the National Air Pollution Surveillance (NAPS) Network [11], established as a joint Federal and Provincial program in 1969 to provide accurate and long-term air quality data of a uniform standard throughout Canada. NAPS sites only had up to 5 monitors during the years of our study compared to 54 monitors used in the BAQS study. To obtain annual estimates for each location for the entire WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

448 Air Pollution XIX study period (1986-2004) information from both the BAQS LUR model and NAPS sites were used. Firstly, the LUR estimates were adjusted using the average air pollution levels from the 2004 BAQS monitors and the average of the 2004 NAPS monitors (adjusted estimates = LUR estimates / BAQS average × 2004 NAPS average). Secondly, the average yearly values for the remaining years (1986-2003) were adjusted to the 2004 adjusted values (adjusted yearly estimates = adjusted estimates / 2004 NAPS average × NAPS yearly average). 3.3 Cases and controls exposure profile From 1986 to the year of diagnosis, for each incident case and control and for each air pollutant (NO2 and SO2) an exposure profile was created consisting of the annual estimates, as described above, at each of the subject’s residence (the centroid of the address postal was used as proxy for the residence location). For each year for which address information was unavailable for a subject we imputed missing values using the observed control mean imputation method (Weinberg et al. [12]). To investigate possible associations between long-term exposure to ambient pollution and lung cancer risk the average annual pollutant levels during the study period (i.e. from year 1986 to the year of diagnosis) was used for each pollutant as surrogate measure of exposure (individual pollution score). 3.4 Statistical analyses Data analysis was conducted using conditional likelihood regression for matched or stratified data (Breslow and Day [13]). All regression models for the evaluation of air pollution effects included four covariates: sex, age at diagnosis, number of cigarettes smoked per day, and duration of cigarette smoking. These factors were included as stratification variables in the regression model. The pollution scores (as described above) were used as proxies for pollution levels. Analyses were done by categorizing exposures to various pollutants into discrete categories. Odds ratios and associated confidence intervals were calculated using matched conditional logistic regression (Breslow and Day [13]). The models were fitted using the PECAN module in the Epicure software package, which calculates parameter estimates using conditional analytic methodology (Preston et al. [14]).

4 Results This report includes all subjects diagnosed from January 1st 2000 to December 31st 2004 inclusive, whose medical charts were accessible in electronic format. The 1422 subjects (711 lung cancer cases and 711 controls) represent 28 percent of the entire study population. The characteristics of cases and controls are shown in Table 1. Smoking information (ever/never) was obtained in over 80% of cases and controls, with data on number of cigarettes smoked per day and

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Table 1:

Characteristics of cases and controls. Males Cases Controls # % # %

Age (years) < 60 60 - 69 ≥ 70 Average Smoking Status Never smoker Ever smoker Unknown Cigarettes per day Unknown 0 1-9 10 - 19 20 - 29 ≥ 30 Average Duration of smoking (years) Unknown 0 1 - 24 25 - 34 35 - 44 ≥ 45 Average Residence duration (years) < 10 10 - 19 ≥ 20 Average Total number

Females Cases Controls # % # %

74 19.0 115 29.6 200 51.4 68.4

74 19.0 115 29.6 200 51.4 68.4

88 27.3 94 29.2 140 43.5 66.1

88 27.3 94 29.2 140 43.5 66.1

12 3.1 310 79.7 67 17.2

93 23.9 257 66.1 39 10.0

23 7.1 253 78.6 46 14.3

169 52.5 128 39.8 25 7.8

112 28.8 14 3.6 8 2.1 24 6.2 132 33.9 99 25.4 25.9

126 32.4 103 26.5 13 3.3 25 6.4 81 20.8 41 10.5 14.0

81 25.2 23 7.1 11 3.4 44 13.7 121 37.6 42 13.0 18.7

65 20.2 169 52.5 14 4.3 19 5.9 45 14.0 10 3.1 6.3

145 37.3 16 4.1 21 5.4 36 9.3 70 18.0 101 26.0 39.1

134 34.4 103 26.5 55 14.1 32 8.2 25 6.4 40 10.3 18.8

100 31.1 23 7.1 29 9.0 38 11.8 55 17.1 77 23.9 35.4

65 20.2 169 52.5 29 9.0 23 7.1 17 5.3 19 5.9 10.7

50 12.9 33 8.5 82 21.1 66 17.0 257 66.1 290 74.6 27.4 30.6 389

45 14.0 26 8.1 77 23.9 75 23.3 200 62.1 221 68.6 26.0 28.7 322

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

449

450 Air Pollution XIX years of smoking being known for about two thirds of the subjects. The average number of cigarette smoked per day and the average years of smoking of lung cancer cases was about twice that of male controls, and three times that of female controls. The majority of cases and controls lived over 20 years in the City of Windsor. No significant relationships were observed between exposure to SO2 and lung cancer (data not shown). Results for NO2 are presented in Table 2. Significant associations between NO2 levels and lung cancer were noted for males only (Table 2a). Significance persisted after controlling for residential duration, smoking duration and number of cigarette smoked per day, using the categories described in Table 1. Table 2a:

Lung cancer odds ratios (ORs) with 95% confidence intervals (CI) by mean NO2 levels for male subjects with full smoking information. NO2 level 3

(ug/m )

Cases

Controls

Odds ratio

p-

(n=231)

(n=243)

(95% CI)

value

Stratification by Age < 21.0

30 (13%) 58 (24%) 1.00

21.0-23.5

84 (36%) 81 (33%) 1.92 (1.13-3.28)

0.02

23.5-26.0

63 (27%) 64 (26%) 1.83 (1.05-3.20)

0.03

26.0-28.5

45 (19%) 35 (14%) 2.37 (1.26-4.43)

0.01

9

0.04

>28.5

(4%)

5

(2%)

3.36 (1.03-10.9)

Stratification by Age and Residence Duration < 21.0

30 (13%) 58 (24%) 1.00

21.0-23.5

84 (36%) 81 (33%) 2.05 (1.19-3.54)

0.01

23.5-26.0

63 (27%) 64 (26%) 1.90 (1.08-3.33)

0.03

26.0-28.5

45 (19%) 35 (14%) 2.43 (1.29-4.55)

0.01

9

0.04

>28.5

(4%)

5

(2%)

3.41

1.05-11.1)

Stratification by Age, Residence Duration, Duration of smoking and Cigarettes per day < 21.0

30 (13%) 58 (24%) 1.00

21.0-23.5

84 (36%) 81 (33%) 3.04 (1.38-6.69)

0.01

23.5-26.0

63 (27%) 64 (26%) 1.47 (0.67-3.21)

0.34

26.0-28.5

45 (19%) 35 (14%) 2.80 (1.18-6.61)

0.02

9

0.05

>28.5

(4%)

5

(2%)

5.49 (1.04-29.0)

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

Table 2b:

451

Lung cancer odds ratios (ORs) with 95% confidence intervals (CI) by mean NO2 levels for female subjects with full smoking information. NO2 level 3

(ug/m )

Cases

Controls

Odds ratio

p-

(n=209)

(n=244)

(95% CI)

value

Stratification by Age < 21.0

30 (14%) 41 (17%) 1.00

21.0-23.5

89 (43%) 98 (40%) 1.23 (0.71-2.15)

0.46

23.5-26.0

49 (23%) 64 (26%) 1.01 (0.55-1.85)

>0.5

26.0-28.5

32 (15%) 29 (12%) 1.53 (0.77-3.06)

0.23

9

>0.5

>28.5

(4%) 12 (5%)

1.07 (0.40-2.88)

Stratification by Age and Residence Duration < 21.0

30 (14%) 41 (17%) 1.00

21.0-23.5

89 (43%) 98 (40%) 1.26 (0.72-2.21)

0.42

23.5-26.0

49 (23%) 64 (26%) 1.00 (0.54-1.83)

>0.5

26.0-28.5

32 (15%) 29 (12%) 1.59 (0.79-3.17)

0.19

>28.5 9 (4%) 12 (5%) 1.06 (0.39-2.86) Stratification by Age, Residence Duration, Duration of smoking and Cigarettes per day < 21.0 30 (14%) 41 (17%) 1.00

>0.5

21.0-23.5

89 (43%) 98 (40%) 2.14 (0.91-5.03)

0.08

23.5-26.0

49 (23%) 64 (26%) 1.42 (0.56-3.58)

0.46

26.0-28.5

32 (15%) 29 (12%) 1.54 (0.59-4.02)

0.38

9

>0.5

>28.5

(4%) 12 (5%)

1.22 (0.26-5.64)

5 Discussion Reviews of lung cancer risk from air pollution exposures have reported associations with exposure to particulate and gases including SO2 and NO2 (Cohen and Pope [15], Katsouyanni and Pershagen [16]). Significant relationships with SO2 have been found by some authors (Dockery et al. [1], Beeson et al. [4] , Jedrychowski et al. [17]) but not all (Nyberg [5], Nafstad [6], Beelan [18] ), and this report. Similarly positive relationships with NO2 have been noted (Filleul et al. [19], Naess et al. [20]), but not in a study examining air pollution related to exposure to traffic (Beelen et al. [18]). In this study, a significant association between lung cancer and long-term exposure to NO2 was clearly shown in males, but not in females. Although WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

452 Air Pollution XIX limited by the fact that a subset of the study population has been analysed at this time and thereby that caution must be exercised in interpreting the data, these results are nonetheless strengthened by the fact that a) the number of cases and controls are relatively large; b) cumulative exposure was assessed for each subject based on residential exposure estimates; c) odds ratios were adjusted for smoking intensity and duration, and for residential duration in the City of Windsor. Final conclusion will have to await analyses of the full study population.

6 Disclaimer Parts of this research are based on data provided by Cancer Care Ontario. However, the analysis, conclusions, opinions and statements expressed are those of the authors and not necessarily those of Cancer Care Ontario.

Acknowledgment The financial support from the Clean Air Regulation Act, Health Canada, is gratefully acknowledged.

References [1] Dockery DW, Pope A, Xiping X, Spengler JD, Ware JH, Fay ME, Ferris BG, Speizer FE. An association between air pollution and mortality in six U.S. cities. N Engl J Med, 329, pp.1753-1759, 1993. [2] Pope CA, Thun MJ, Namboodiri MM, Dockery DW, Evans JS, Speizer FE, Heath Jr CW. Particulate air pollution as a predictor of mortality in a prospective study of U.S. adults. Am J Respir Crit Care Med, 151, pp.669-674, 1995. [3] Pope CA, Burnett RT, Thun MJ, Calle EE, Krewski D, Ito K, Thruston GD. Lung cancer, cardiopulmonary morality and long-term exposure to fine particulate air pollution. JAMA, 287, pp. 1132-1141, 2002. [4] Beeson WL, Abbey DE, Knutsen SF. Long-term concentrations of ambient air pollutants and incident lung cancer in California adults: results from the AHSMOG study. Environ Health Perspect, 106, pp. 813-823, 1998. [5] Nyberg F, Gustavsson P, Järup L, Bellander T, Berglind N, Jakobson R, Perhsagen G. Urband air pollution and lung cancer in Stockholm. Epidemiology, 11, pp. 487-495, 2000. [6] Nafstad P, Håheim LL, Wisloff T, Gram F, Oftedal B, Holme I, Hjermann I, Leren P. Urban air pollution and mortality in a cohort of Norwegian men. Environ Health Perspect, 112, pp. 610-615, 2004. [7] Nafstad P, Håheim LL, Oftedal B, Gram F, Holme I, Hjermann I, Leren P. Lung cancer and air pollution: 27 year follow up of 16209 Norwegian men. Thorax, 58, pp. 1071-1076, 2003.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

453

[8] Band PR, Zielinski JM, Jiang H, Liu L. Canada-US border air quality strategy (BAQS): preliminary results of mortality and cancer incidence in Windsor, Canada. In: “Air Pollution X1V” Longhurst JWS and Brebbia CA Eds. Wit Press, Southampton, Boston, pp 777-83, 2006. [9] Wheeler AJ, Smith-Doiron M, Xu X, Gilbert NL, Brook JR. 2008. Intra-urban variability of air pollution in Windsor, Ontario--measurement and modeling for human exposure assessment. Environ Res 106, pp. 7-16, 2008. [10] Jerrett M, Arain A, Kanaroglou P, Beckerman B, Potoglou D, Sahsuvaroglu T, Morrison J, Giovis C. A review and evaluation of intra-urban air pollution exposure models. J Expo Anal Environ Epidemiol 15, pp. 185-204, 2005. [11] National Air Pollution Surveillance (NAPS) Network Annual Dada Summary. Her Majesty the Queen in Right of Canada (Environment Canada). Online. http://www.etc-cte.ec.gc.ca/publications /napsreports_e.html. [12] Weinberg, C. R., E. S. Moledor, D. M. Umbach, and D. P. Sandler. Imputation for exposure histories with gaps, under an excess relative risk model. Epidemiology 5, pp. 490-7, 1996. [13] Breslow NE, Day NE. Statistical methods in cancer research. Volume 1. The design and analysis of case-control studies. International Agency for Research on Cancer. IARC Scientific Publication no. 32, Lyon, France, 1986 [14] Preston, D. L., Lubin, J. H., Pierce, D. A., and McConney, M. Epicure User's Guide. Seattle, Washington: Hirosoft International Corporation, 2000. [15] Cohen AJ, Pope CA III. Lung cancer and air pollution. Environ Health Perspect 103 (Suppl 8), pp. 219-224, 1995. [16] Katsouyanni K, Pershagen G. Ambient air pollution exposure and cancer. Cancer Causes Control 8, pp. 284-191, 1997. [17] Jedrychowski W, Becher H, Wahrendorf J, Basa-Cierpialek Z. A case-control study of lung cancer with special reference to the effect of air pollution in Poland. J Epidemiol Community Health 44, pp.114-120, 1990. [18] Beelen R, Hoek G, van den Brandt PA, Goldbolm A, Fischer P, Schouten LJ, Armstrong b, Brunekreef B. Long-term exposure to traffic-related air pollution and lung cancer risk. Epidemiology, 19, pp. 702-710, 2008. [19] Filleul L, Rondeau V, Vandentorren S, Le Moual N, Cantagrel A, Annesi-Maesano I, Charpin D, Declercq C, Neukirch F, Paris C, Vervloet D, Brochard P, Tessier J-F, Kauffmann F, Baldi I. Twenty five year mortality and air pollution: results from the French PAARC survey. Occup Environ Med, 62, pp. 453-460, 2005. [20] Naess Ø, Nafstad P, Aamodt G, Claussen B, Rosland P. Relation between concentration of air pollution and cause-specific mortality: four-year exposures to nitrogen dioxide and particulate matter pollutants in 470 neighborhoods in Oslo, Norway. Am J Epidemiol, 165, 435-443, 2007.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

This page intentionally left blank

Air Pollution XIX

455

Comparison of fungal contamination between hospitals and companies food units C. Viegas1, M. Almeida1, C. Ramos1, R. Sabino2, C. Veríssimo2 & L. Rosado2 1

Higher School of Health Technologies of Lisbon, Polytechnic Institute of Lisbon, Portugal 2 National Institute of Health Dr Ricardo Jorge, Mycology Laboratory, Portugal

Abstract A descriptive study was developed to compare air and surfaces fungal contamination in ten hospitals’ food units and two food units from companies. Fifty air samples of 250 litres through impaction method were collected from hospitals’ food units and 41 swab samples from surfaces were also collected, using a 10 by 10 cm square stencil. Regarding the two companies, ten air samples and eight surface samples were collected. Air and surface samples were collected in food storage facilities, kitchen, food plating and canteen. Outdoor air was also collected since this is the place regarded as a reference. Simultaneously, temperature, relative humidity and meal numbers were registered. Concerning air from hospitals’ food units, 32 fungal species were identified, being the two most commonly isolated genera Penicillium sp. (43.6%) and Cladosporium sp. (23.2%). Regarding yeasts, only Rhodotorula sp. (84.2%) and Trichosporon sp. (15.8%) were isolated. Regarding the analyzed surfaces from the same places, 21 fungal species were identified, being also Penicillium sp. (69.1%) and Cladosporium sp. (8.25%) the genera most frequently found. Candida parapsilosis (36.3%) and Rhodotorula sp. (25.7%) were the most prevalent yeast species. In the two companies, nine fungal species were identified in air, being Cladosporium sp. the most frequent genus (71.2%) followed by Penicillium sp. (13.0%). Only one yeast species, Candida famata, was identified. Eight filamentous fungi and three yeasts were identified in the analyzed surfaces, being

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line) doi:10 2495/AIR110421

456 Air Pollution XIX Penicillium sp the most frequently isolated mould (97.2%) and Candida famata the most frequent yeast (42.9 %). Aspergillus species, such as A. ochraceus, A. versicolor, A. candidus, A. fumigatus and A. niger were also isolated in hospitals’ food units, whereas companies’ food units only A. glaucus was isolated. No positive association was observed (p>0.05) among fungal contamination and the following parameters: temperature, relative humidity and number of meals served in hospitals’ and companies’ food units. Keywords: air, surfaces, fungal contamination, hospitals, companies.

1 Introduction Fungal spores are complex agents that may contain multiple hazardous components. Health hazards may differ across species because fungi may produce different allergens and mycotoxins, and some species can infect humans [1]. Exposure effects to fungi are dependent on the species present, the metabolic products produced, concentration and duration of the exposure, and also individual susceptibility [2]. The main source of fungi in office environments is outdoor air. As outdoor air is often filtered before entering in the ventilation system and fungi settle due to lower air velocities in buildings than outdoors, common indoor fungal levels are expected to be lower than levels in outdoor air [1]. However, in hospital settings there are diverse possible sources of fungal contamination including ventilation or air-conditioning systems, decaying organic material, dust ornamental plants, food, water and, particularly, building works in and around hospitals [3, 4]. Despite the possibility of adverse health effects due to exposure to fungal products, no health-based exposure limits have yet been proposed. In part this is due to the difficulty of accurately characterizing cumulative fungal spore concentrations [5] and also because, epidemiological studies have failed to establish a causal relation of the extent of fungal presence, exposure time and specific effects on health or frequency and severity of symptoms reported. Studies tend to show only existence of a link between exposure to fungi and development of symptoms, especially respiratory ones [2]. Attempts have been made only to identify fungi responsible for specific symptoms attributed to mould exposure, such as allergenic [6], inflammatory [7] or mycotoxic [8] effects. In Portugal, the prevalence of diseases such as asthma and rhinoconjunctivitis in the general population varies from 15% to 25% and from 10% to 15%, respectively, and in recent years has been increasing [9]. Various causes have been considered, including indoor air pollution caused by fungal contamination. This investigation was designed to compare air and surfaces fungal contamination in ten hospitals’ food units and two food units from companies and explore possible associations with independent variables.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

457

2 Materials and methods A descriptive study was developed to compare fungal contamination of air and surfaces analyzed from ten hospitals’ food units and two food units from companies. Fifty air samples of 250 litres were collected through impaction method from hospitals’ food units. Forty one swab samples from surfaces were also collected, using a 10 by 10 cm square stencil. Regarding the two companies, ten air samples and eight swab samples were collected using the methodology previously described. Air and swab samples were collected in food storage facilities, kitchen, food plating and canteen. Simultaneously, temperature and relative humidity were also monitored through the equipment Babouc, LSI Sistems and according to the International Standard ISO 7726 – 1998. The number of meals served in each case was also registered. Air samples were collected at one meter tall with a flow rate of 140 L/minute, onto malt extract agar supplemented with the antibiotic chloramphenicol (MEA), in the facilities, and also outdoor, since this is the place regarded as reference. Concerning surfaces samples they were collected by swabbing the surfaces of the same indoor places, using a 10 by 10 cm square stencil disinfected with 70% alcohol solution between samples according to the International Standard ISO 18593 – 2004. Subsequently, all the collected samples were incubated at 27 ºC for 5 to 7 days. After laboratory processing and incubation of the collected samples, quantitative (CFU/m3 and CFU/m2) and qualitative results were obtained, with identification of the isolated fungal species. Whenever possible, filamentous fungi were identified to the species level, since adverse health effects vary according to fungal species [10, 11]. Identification of filamentous fungi was carried out on material mounted in lactophenol blue and achieved through morphological characteristics listed in illustrated literature [11] and yeasts were identified through biochemical API test [12]. Tables with frequency distribution of isolated fungal species were made with the obtained data. Fungal concentration dependence in the two monitored environmental parameters – temperature and relative humidity– and also number of meals served was analyzed.

3 Results Concerning air from hospitals’ food units, 32 species of fungi were identified, being the two most commonly isolated genera Penicillium sp. (43.6%) and Cladosporium sp. (23.2%). Regarding yeasts, only Rhodotorula sp. (84.2%) and Trichosporon sp. (15.8%) were isolated. Considering surfaces from the same places, twenty one fungal species were identified, being also Penicillium sp. (69.1%) and Cladosporium sp. (8.25%) the genera most frequently found. Candida parapsilosis (36.3%) and Rhodotorula sp. (25.7%) were the most prevalent yeasts species. In the two companies, nine fungal species were identified in air, being Cladosporium sp. the most frequent genus (71.2%) followed by Penicillium sp. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

458 Air Pollution XIX (13.0%). Concerning yeasts only Candida famata was identified. Eight fungal species and three yeasts were identified in the analyzed surfaces, being moulds and yeasts most frequently found Penicillium sp. (97.2%) and Candida famata (42.9%), respectively. Regarding hospitals’ food units, there was coincidence between prevailing genera in indoor and outdoor air/environment. Nevertheless, all ten food units presented fungal species different from the ones isolated from outdoor. Moreover, nine from the ten food units presented Aspergillus species. Regarding the comparison of concentrations found in air, for indoor and outdoor environments, nine of the ten food units showed higher levels of contamination in indoor air. Concerning companies’ food units, there was no coincidence between prevailing genera in indoor and outdoor and the two food units presented fungal species different from the ones isolated from outdoor. However, only one of them presented Aspergillus species and none showed higher levels of contamination in indoor air when compare with outdoor air levels. Aspergillus species, such as A. ochraceus, A. versicolor, A. candidus, A. fumigatus and A. niger were isolated in hospitals’ food units, whereas companies’ food units only A. glaucus was isolated. There was no significant relation (p>0,05) between fungal contamination and temperature, relative humidity and number of meals served in hospitals’ and companies’ food units.

4 Discussion The most predominant genus found in hospital air was Penicillium. Regarding this genus, there are different potential risks associated with their inhalation, due to the toxins release [13]. Regarding the most frequent genus in companies’ air – Cladosporium – is probably the fungus that occurs more frequently around world, especially in temperate climates [14] such as in Portugal and is deeply connected to indoor condensation problems [15]. Both of the referred genus were also the more frequent in a study realized in a Portuguese poultry [16]. It is suggested that fungal levels found indoors should be compared, quantitatively and qualitatively, with those found outdoors, because the first are dependent on the last [2]. Nevertheless, when it comes to fungal levels, it should be taken into account that indoor and outdoor environments are quite different which, by itself, justifies diversity of species between different spaces. However, with regard to fungal contamination, there are no stipulated thresholds which makes essential to compare fungal levels indoors and outdoors. Thus, indoor air quality that significantly differs from the outdoor could mean that there are infiltration problems and that exist a potential risk for health. It is worth mentioning that as outdoor air is a major source of the fungi found indoors, it is no coincidence that, in the case of hospitals’ food units, the prevailing genera, Penicillium sp. and Cladosporium sp., are the same in both these environments [13]. Nonetheless, in the companies’ case there was no coincidence between prevailing genera in indoor and outdoor air. Besides that, WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

459

all the monitored food units (hospitals and companies) had one or more spaces with fungal species that differed from the ones isolated outdoor, suggesting fungal contamination from within [13]. Moreover, according to American Industrial Hygiene Association (AIHA), in 1996, for determination of biological contamination in environmental samples, The confirmed presence of the species Aspergillus flavus and Aspergillus fumigatus (both identified in hospitals’ food units), requires implementation of corrective measures [17]. Regarding comparison of spore concentrations found in air, for indoor and outdoor environments, nine of the ten hospitals’ food units showed higher levels of contamination in indoor air, whereas all the companies’ food units presented higher levels of fungal concentrations in outdoor air. This fact could be explained by the discrepancy of the number of institutions analyzed or maybe because possible sources of fungal dissemination include hospitals’ ventilation or air-conditioning systems [4, 18]. Taking into account what is mentioned in Portuguese law, this value was 500 CFU/m3 is the maximum reference concentration in indoor air, was exceeded only in four indoor spaces from the hospitals’ food units analyzed in this study. Regarding what is mentioned in Portuguese Technician Norm NT-SCE-02, the presence of opportunistic fungi from Aspergillus genus, shows a lack of air quality in indoor space. Aspergillus species are frequently present on food and thus can be an indirect source of airway or digestive tract colonization of the patients and workers [19]. Results related to environmental variables are not consistent with what is expected [20]. It was found that the relation between the fungal air contamination and the temperature, relative humidity, and also number of meals served was not statistically significant (p>0,05). This may be justified by the effect of other environmental variables also influencing fungal spreading, namely patients and workers, who may carry a great diversity of fungal species [21], as well the developed activities that may also affect fungal concentration [22].

5 Conclusions With this study, it was possible to characterize fungal distribution in ten hospitals’ food units and two food units from companies and evaluate the association of environmental variables and also number of meals served with this distribution. It was also possible to observe that hospitals’ food units presented more evidence that fungal contamination comes from within than in the companies’ case. Unlike other studies, environmental variables monitored (temperature and relative humidity) and also number of meals served, did not show the expected association with fungal concentration, which may possibly have resulted from other variables not investigated in this study.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

460 Air Pollution XIX

References [1] Eduard, W., Fungal spores: A critical review of the toxicological and epidemiological evidence as a basis for occupational exposure limit setting. Informahealthcare, 2008. [2] Goyer N, Lavoie J, Lazure L & Marchand G., Bioaerosols in the Workplace: Evaluation, Control and Prevention Guide. Institut de Recherche en Santé et en Sécurité du Travail du Québec, 2001. [3] Van Den Bergh M, Verweij P & Voss A., Epidemiology of nosocomial fungal infections: invasive aspergillosis and the environment. Diagnostic Microbiology and Infectious Disease, 34, 3, pp. 221–227, 1999. [4] Boukline A, Lacroix C, Roux N, Gangneux J & Derouin F., Fungal contamination of food in hematology units. J Clin Microbiol, 38, pp. 4272– 4273, 2000. [5] Bartlett K, Kennedy S, Brauer M, Van Netten C &Dill B., Evaluation and predictive model of airborne fungal concentrations in school classrooms. Ann Occup Hyg, 48, pp., 547–554, 2004. [6] Cruz A, Saenz de Santamaria M, Martinez J, Martinez A, Guisantes J & Palácios R., Fungal allergens from important allergenic fungi imperfecti. Allergol Immunopathol, 25, pp., 153–158, 1997. [7] Rylander R, Persson H, Goto H, Yuasa K & Tanaka S., Airborne beta (1,3)glucans may be related to symptoms of sick buildings. Indoor Environment, 1, pp., 263–267, 1992. [8] Hodgson M, Morey P, Leung W, Morrow L, Miller D, Jarvis B, Robbins H, Halsey J & Storey E., Building associated pulmonary disease from exposure to Stachybotrys chartarum and Aspergillus versicolor. J Occup Environm Med, 40, pp., 241 – 249, 1998. [9] Nunes C & Ladeira S., Estudo da Qualidade de Ambiente Fúngico em Escolas e Edifícios Públicos no Algarve. Revista Portuguesa Imunoalergologia,15, pp. 411–422, 2007. [10] Rao C, Burge H & Chang J., Review of quantitative standards and guidelines for fungi in indoor air. J Air Waste Manage Assoc., 46, pp. 899– 908, 1996. [11] Hoog C, Guarro J, Gené G & Figueiras M., (2th ed). Atlas of Clinical Fungi. Centraalbureau voor Schimmelcultures, 2000. [12] Ghannoum M, Hajeh R, Scher R, Konnikov N, et al., A large-scale North American study of fungal isolates from nails: The frequency of onychomycosis, fungal distribution and antifungal susceptibility patterns. J. Am. Acad. Dermatol, 43, pp. 641–648, 2000. [13] Kemp P, Neumeister-Kemp H, Esposito B, Lysek G & Murray F., Changes in airborne fungi from the outdoors to indoor air; Large HVAC systems in nonproblem buildings in two different climates. American Industrial Hygiene Association, 64, pp. 269–275, 2003. [14] Cooley J, Wong W, Jumper C & Straus D., Correlation between the prevalence of certain fungi and sick building syndrome. Occup. Environ Med, 55, pp., 579–584, 1998. WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

Air Pollution XIX

461

[15] Garret M, Rayment P, Hooper M, Abranson M & Hooper B., Indoor Airborne Fungal Spores, House Dampness and Associations with Environmental Factors and Respiratory Health in Children. Clinical and Experimental Allergy, 28, pp., 459–467, 1998. [16] Viegas C, Veríssimo C, Rosado L & Silva Santos C., Poultry fungal contamination as a public health problem. Environmental Toxicology III. WIT Transactions on Biomedicine and Health, 2010. [17] American Industrial Hygiene Association: Field Guide for the Determination of Biological Contaminants in Environmental Samples. AIHA, 1996. [18] Beggs C & Kerr K., The threat posed by airborne micro-organisms. Indoor and Built Environment, 9, 5, pp. 241–245, 2000. [19] Sarfati J, Jensen H & Latgé J., Route of infections in bovine aspergillosis. J Med Vet Mycol, 34, pp., 379–383, 1996. [20] Kakde U, Kakde H & Saoji A., Seasonal Variation of Fungal Propagules in a Fruit Market Environment, Nagpur (India). Aerobiologia, 17, pp. 177– 182, 2001. [21] Scheff P, Pulius V, Curtis L & Conroy L., Indoor air quality in a middle school, Part II: Development of emission factors for particulate matter and bioaerosols. Applied Occupational and Environmental Hygiene, 15, pp. 835–842, 2000. [22] Buttner M & Stetzenbach L., Monitoring Airborne fungal spores in an experimental indoor environment to evaluate sampling methods and the effects of human activity on air sampling. Applied and Environmental Microbiology, 59, pp. 219–226, 1993.

WIT Transactions on Ecology and the Environment, Vol 147, © 2011 WIT Press www witpress com, ISSN 1743-3541 (on-line)

This page intentionally left blank

Air Pollution XIX

463

Author Index Almeida M. .............................. 455 Amorim J. H. ............................. 13 Apon F. .................................... 137 Band P. R. ................................ 445 Barbiere M. .............................. 185 Barnes J. H. .......................... 1, 277 Belis C. .................................... 185 Bellagamba S. .......................... 199 Booth C. A. .............................. 117 Borowiak A.............................. 185 Borrego C. ................................. 13 Borůvková J. ............................ 173 Bourgault C.............................. 387 Bozek F. ................................... 343 Božnar M. Z. .............................. 47 Briant R...................................... 37 Buelna G. ................................. 387 Bugajny C. ................................. 37 Burns L. ................................... 137 Cairnsa E.................................. 239 Callén M. S. ............................. 149 Cascão P. ................................... 13 Černíková M. ........................... 423 Chatterton T. J. ........................ 277 Chen H. ...................................... 93 Chen J. B............................ 59, 307 Chrysoulakis N. ......................... 13 Connan O. ................................ 399 Cortes L. .................................. 301 Crosby C. J. ............................. 117 Čupr P. ..................................... 173 Cursaru D. L. ........................... 331 Dacre H. ..................................... 81 Damiani F. ............................... 199 De Geyter N. ............................ 353 De Simone P. ........................... 199 Derkx F. ................................... 399 Di Molfetta V........................... 199

Edokpayi C. A. ............................ 1 Eijk A. R. A. ............................ 129 Faizal M. .................................. 231 Falcón Y. I. .............................. 301 Fraaij A. L. A........................... 219 Fullen M. A. ............................ 117 Gadrat M. ................................... 37 Gavendova H. .......................... 343 Gerboles M. ............................. 185 Giraudon J.-M.......................... 353 Grašič B. .................................... 47 Grison H. ................................. 363 Gulia S. ...................................... 71 Hayes E. T. .............................. 277 He J. J. ............................... 59, 307 Hébert D. ................................. 399 Heitz M. ................................... 387 Hoek G..................................... 257 Holoubek I. .............................. 173 Horák J..................................... 433 Huzlik J.................................... 343 Iqbal M. ................................... 231 Janssen N. ................................ 257 Jiang H. .................................... 445 Julia S. ..................................... 231 Kapička A. ............................... 363 Kapus M. ................................. 185 Kareš R. ................................... 173 Kastek M. ................................ 161 Keuken M. P. ................... 129, 257 Khare M. .................................... 71 Khatib J. M. ............................. 117 Klánová J. ................................ 173 Kohoutek J. .............................. 173 Kukučka P. .............................. 173

464 Air Pollution XIX Lagler F.................................... 185 Lamonier J.-F........................... 353 Lao J. ......................................... 25 Le Bihan Y............................... 387 Leroy C. ................................... 399 Lessard P.................................. 387 Leys C. ..................................... 353 Lima M. M. C. ......................... 105 Liu N. ................................. 59, 307 Longhurst J. W. S. ............... 1, 277 Lopes M. .................................... 13 López J. M. .............................. 149 Lu J. ......................................... 239 Luckhurst D. A. ....................... 117 Magaril E. ................................ 373 Malíková O. ..................... 423, 433 Malkina-Pykh I. G. .................. 267 Mărdărescu V........................... 331 Mares J..................................... 343 Maro D..................................... 399 Martinez E. .............................. 301 Martins J. ................................... 13 Mastral A. M............................ 149 Mengistu Tsidu G. ................... 411 Methven J. ................................. 81 Miranda A. I. ............................. 13 Mlakar P. ................................... 47 Morent R. ................................. 353 Munir S. ..................................... 93 Murena F.................................. 287 Nagendra S. ............................... 71 Nguyen Dinh M. T. .................. 353 Nikiema J. ................................ 387 Nugraha T. ............................... 231 Olowoporoku A. O. ............. 1, 277 Ottelé M. .................................. 219 Paglietti F. ................................ 199 Peake D. ..................................... 81 Petrovský E. ............................. 363 Piątkowski T. ........................... 161

Polakowski H........................... 161 Prokeš R................................... 173 Pykh Y. A. ............................... 267 Ramos C. ................................. 455 Rodrigues V. .............................. 13 Ropkins K. ................................. 93 Rosado L. ................................. 455 Roupsard P............................... 399 Rozet M. .................................. 399 Sabino R. ................................. 455 Searle D. E. .............................. 117 Seigneur C. ................................ 37 Seroji A. R. .............................. 319 Tan G. F. M. ............................ 239 Tănăsescu C. ............................ 331 Tang H. .................................... 137 Tavares R. .................................. 13 Teixidó O. .................................. 25 Tharumakulasingam K. ........... 239 Tomaz E................................... 211 Ture K...................................... 411 Turgeon N. ............................... 387 Ueda A. C. ............................... 211 Urciuolo M. ............................. 287 Ursem W. J. N. ........................ 219 van Bohemen H. D. ................. 219 van den Elshout S. ................... 257 Vandenbroucke A. M. ............. 353 Vanderstricht A........................ 353 Veríssimo C. ............................ 455 Verreault S. .............................. 387 Viegas C. ......................... 247, 455 Vijay P. ...................................... 71 Voogt M. H. ............................. 129 Winspear C. P. ......................... 117 Worsley A. T. .......................... 117 Xia D. S. .................................. 307

Air Pollution XIX

Yang L. .................................... 137 Yap D....................................... 239 Yu Y. ................................. 59, 307 Yuliarto B. ............................... 231

465

Zandveld P. ...................... 129, 257 Zhao S. P. .......................... 59, 307 Zielinski J. M. .......................... 445

This page intentionally left blank

...for scientists by scientists

Environmental Health Risk VI Edited by: C.A. BREBBIA, Wessex Institute of Technology, UK and M. EGLITE, Riga Stradins University, Latvia

Health problems related to the environment are causing increasing concern all over the world. The health of the population depends upon society’s ability to ensure good quality air, water, soil, and food and to eliminate or considerably reduce hazards from the human environment. Society’s ability to achieve these objectives is greatly dependent on the development of modelling and interpretive techniques that allow decision-makers to assess the risk posed by various factors, as well as to suggest improvements. Environmental Health Risk VI contains contributions presented at the Sixth International Conference on the Impact of Environmental Factors on Health. The successful biennial series began in 1997 and covers such topics as: Air Pollution; Water and Soil Quality Issues; Risk Prevention and Monitoring; Ecology and Health; Food Safety; Toxicology Analysis; Occupational Health; Control of Pollution Risk; Mitigation Problems; Disaster Management and Preparedness; Epidemiological Studies and Pandemics; Radiation Fields; Waste Disposal; Industrial Safety and Hygiene; Social and Economic Issues; Accidents and Man-made Risks; The Built Environment and Health; Designing for Health; Contamination in Rural Areas; Climate Change and Adaptation; Educational Projects; Environmental Education and Risk Abatement. WIT Transactions on Biomedicine and Health, Vol 15 ISBN: 978-1-84564-524-3 eISBN: 978-1-84564-525-0 Published 2011 / 512pp / £220.00

We are now able to supply you with details of new WIT Press titles via E-Mail. To subscribe to this free service, or for information on any of our titles, please contact the Marketing Department, WIT Press, Ashurst Lodge, Ashurst, Southampton, SO40 7AA, UK Tel: +44 (0) 238 029 3223 Fax: +44 (0) 238 029 2853 E-mail: marketing@witpress com

...for scientists by scientists

Biological Monitoring Theory and Applications Edited by: M.E. CONTI, University of Rome ‘La Sapienza’, Italy

Provides the reader with a basic understanding of the use of bioindicators both in assessing environmental quality and as a means of support in environmental impact assessment (EIA) procedures. The book primarily deals with the applicability of these studies with regard to research results concerning the basal quality of ecosystems and from an industrial perspective, where evaluations prior to the construction of major projects (often industrial plants) are extremely important. Environmental pollution and related human health concerns have now reached critical levels in many areas of the world. International programs for researching, monitoring and preventing the causes of these phenomena are ongoing in many countries. There is an imperative call for reliable and cost-effective information on the basal pollution levels both for areas already involved in intense industrial activities, and for sites with industrial development potential. Biomonitoring methods can be used as unfailing tools for the control of contaminated areas, as well as in environmental prevention studies. Human biomonitoring is now widely recognized as a tool for human exposure assessment, providing suitable and useful indications of the ‘internal dose’ of chemical agents. Bioindicators, biomonitors, and biomarkers are all well-known terms among environmental scientists, although their meanings are sometimes misrepresented. Therefore, a better and full comprehension of the role of biological monitoring, and its procedures for evaluating polluting impacts on environment and health, is needed. This book gives an overview of the state of the art of relevant aspects of biological monitoring for the evaluation of ecosystem quality and human health. Series: The Sustainable World, Vol 17 ISBN: 978-1-84564-002-6 eISBN: 978-1-84564-302-7 Published 2008 / 256pp / £84.00

...for scientists by scientists

Air Pollution XVIII Edited by: C.A. BREBBIA, Wessex Institute of Technology, UK and J.W.S. LONGHURST, University of the West of England, UK

Air Pollution is one of the most challenging problems facing the international community; it is widespread and growing in importance and has clear and known impacts on health and the environment. The human need for transport, manufactured goods and services brings with it impacts on the atmospheric environment at scales from the local to the global. The rate of development of the global economy brings new pressures and the willingness of governments to regulate air pollution is often balanced by concerns over the economic impact of such regulation. Science is the key to identifying the nature and scale of air pollution impacts and is essential in the formulation of policies for regulatory decision-making. Continuous improvements in our knowledge of the fundamental science of air pollution and its application are necessary if we are to predict, assess and mitigate the air pollution implications to local, regional, national and international economic systems. The Eighteenth Annual Meeting in the successful series of International Conferences dealing with Modelling, Monitoring and Management of Air Pollution discussed papers dealing with a wide variety of topics, including: Air Pollution Modelling; Air Quality Management; Emission Studies; Monitoring and Measuring; Aerosols and Particles; Innovative Indoor Air Quality Techniques; Indoor Air Pollution; Exposure and Health Effects; Air Pollution Mitigation. WIT Transactions on Ecology and the Environment, Vol 136 ISBN: 978-1-84564-450-5 eISBN: 978-1-84564-451-2 Published 2010 / 464pp / £176.00

WIT eLibrary Home of the Transactions of the Wessex Institute, the WIT electronic-library provides the international scientific community with immediate and permanent access to individual papers presented at WIT conferences. Visitors to the WIT eLibrary can freely browse and search abstracts of all papers in the collection before progressing to download their full text. Visit the WIT eLibrary at http://library.witpress.com

...for scientists by scientists

Urban Transport XVII Urban Transport and the Environment in the 21st Century Edited by: A. PRATELLI, University of Pisa, Italy and C.A. BREBBIA, Wessex Institute of Technology, UK

Over the last sixteen years, the Wessex Institute of Technology has convened annually the International Conference on Urban Transport and the Environment in order to facilitate the sharing of research on a subject of growing importance to cities around the world. Urban transportation systems can enhance or degrade the quality of urban life by their impact on the environment as well as their operation and accessibility. Municipal authorities put a priority on transportation systems that minimize ecological and environmental impacts, are sustainable, and help to improve the socio-economic fabric of the city. At the same time, systems must ensure the safety and security of the public while retaining system efficiency. All of these concerns are addressed by research presented at the conference and contained in this book. Topics covered include: Urban Transport Planning and Management; Transportation Demand Analysis; Traffic Integration and Control; Intelligent Transport Systems; Transport Modelling and Simulation; Land Use and Transport Integration; Public Transport Systems; Environmental and Ecological Aspects;Air and Noise Pollution; Safety and Security; Energy and Transport Fuels; Economic and Social Impact; and Advanced Transport Systems. WIT Transactions on the Built Environment, Vol 116 ISBN: 978-1-84564-520-5 eISBN: 978-1-84564-521-2 Published 2011 / 736pp / £316.00

WITPress Ashurst Lodge, Ashurst, Southampton, SO40 7AA, UK. Tel: 44 (0) 238 029 3223 Fax: 44 (0) 238 029 2853 E-Mail: [email protected]

This page intentionally left blank

This page intentionally left blank

This page intentionally left blank

This page intentionally left blank

This page intentionally left blank

This page intentionally left blank

This page intentionally left blank

This page intentionally left blank