WORLD BANK TECHNICAL PAPER NO. 373 V/TP3qS Work in progress for public discussion 'Sept. 1qq Vehicular Air Pollution t~~r I/(/i'/mw .S .t-cf/ l8'y''/ A/i,,/i .m~ 1// P ,t/ 1 /1{/1/ (.m, ( wt Ila~~~~~~ ~~~~- '1 jj d'%57 .-T_ ~~r p i c L jw. r i; * ,77777 77777 -At S!~~~~~ /X / (hi, i/i- _/r In /w P r / / Su1/ ///i,/ P G (111(11tlz RECENT WORLD BANK TECHNICAL PAPERS No. 289 Dinar, Seidl, Olem, Jorden, Duda, and Johnson, Restoring and Protecting the World's Lakes and Reservoirs No. 290 Weijenberg, Dagg, Kampen Kalunda, Mailu, Ketema, Navarro, and Abdi Noor, Strengthening National Agricultutal Research Systems in Eastern and Central Africa: A Frameworkfor Action No. 291 Valdes and Schaeffer in collaboration with Errazuriz and Francisco, Sutrveillance of Agricutltural Price and Trade Policies: A Handbookfor Chtile No. 292 Gorriz, Subramanian, and Simas, Irrigation Management Transfer in Mexico: Process and Progress No. 293 Preker and Feachem, Market Mechanisms and the Health Sector in Central and Eastern Eutrope No. 294 Valdes and Schaeffer in collaboration with Sturzenegger and Bebczuk, Suirveillance of Agricultutral Price and Trade Policies: A Handbookfor Argentina No. 295 Pohl, Jedrzejczak, and Anderson, Creating Capital Markets in Central and Eastern Eutrope No. 296 Stassen, Small-Scale Biomass Gasifiersfor Heat and Power: A Global Review No. 297 Bulatao, Key Indicatorsfor Family Planning Projects No. 298 Odaga and Heneveld, Girls and Schools in Suib-Saharan Africa: From Analysis to Action No. 299 Tamale, Jones, and Pswarayi-Riddihough, Technologies Related to Participatory Forestry in Tropical and Suibtropical Coluntries No. 300 Oram and de Haan, Technologiesfor Rainfed Agricuilture in Mediterranean Climates: A Review of World Bank Experiences No. 301 Mohan, editor, Bibliography of Puiblications: Technical Department, Africa Region, Jutly 2987 to April 1995 No. 302 Baldry, Calamari, and Yameogo, Environmental Impact Assessment of Settlement and Development in the Upper Leraba Basin No. 303 Heneveld and Craig, Schools Couint: World Bank Project Designs and the Quality of Primanj Education in Suib-Saharan Africa No. 304 Foley, Photovoltaic Applications in Ruiral Areas of the Developing World No. 305 Johnson, Edutcation and Training of Accoutntants in Suib-Saharan Anglophone Africa No. 306 Muir and Saba, Improving State Enterprise Performance: The Role of Internal and External Incentives No. 307 Narayan, Toward Participatory Research No. 308 Adamson and others, Energy Use, Air Polluttion, and Environmental Policy in Krakow: Can Econonmic Incentives Really Help? No. 309 The World Bank/FOA/UNIDO/Industry Fertilizer Working Group, World and Regional Suipply and Demand Balancesfor Nitrogen, Phosphate, and Potash, 1993/94-1999/2000 No. 310 Elder and Cooley, editors, Stustainable Settlement and Development of the Onchocerciasis Control Programme Area: Proceedings of a Ministerial Meeting No. 311 Webster, Riopelle and Chidzero, World Bank Lendingfor Small Enterprises 1989-1993 No. 312 Benoit, Project Finance at the World Bank: An Overview of Policies and Instrlments No. 313 Kapur, Airport Infrastructure: The Emerging Role of the Private Sector No. 314 Valdes and Schaeffer in collaboration with Ramos, Surveillance of Agriciultural Price and Trade Policies: A Handbookfor Ecutador No. 316 Schware and Kimberley, Information Technology and National Trade Facilitation: Making the Most of Global Trade No. 317 Schware and Kimberley, Information Technology and National Trade Facilitation: Gutide to Best Practice No. 318 Taylor, Boukambou, Dahniya, Ouayogode, Ayling, Abdi Noor, and Toure, Strengthening National Agricultural Research Systems in the Hutmid and Sub-hutmid Zones of West and Central Africa: A Frameworkfor Action No. 320 Srivastava, Lambert, and Vietmeyer, Medicinal Plants: An Expanding Role in Development No. 321 Srivastava, Smith, and Forno, Biodiversity and Agricullture: Implicationsfor Conservation and Developnment No. 322 Peters, The Ecology and Management of Non-Timber Forest Resouirces No. 323 Pannier, editor, Corporate Governance of Putblic Enterprises in Transitional Economies (List continues on the inside back cover) WORLD BANK TECHNICAL PAPER NO. 373 Vehicular Air Pollution Experiencesfrom Seven Latin American Urban Centers Bekir Onursal Surhid P Gautam The World Bawk Washington, D.C. Copyright © 1997 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing September 1997 Technical Papers are published to communicate the results of the Bank's work to the development community with the least possible delay. The typescript of this paper therefore has not been prepared in accordance with the proce- dures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. Some sources cited in this paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data in- cluded in this publication and accepts no responsibility whatsoever for any consequence of their use. The boundaries, colors, denominations, and other information shown on any map in this volume do not imply on the part of the World Bank Group any judgment on the legal status of any territory or the endorsement or acceptance of such boundaries. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissem- ination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center, Inc., Suite 910, 222 Rosewood Drive, Danvers, Massachusetts 01923, U.S.A. ISBN 0-8213-4016-6 ISSN: 0253-7494 Bekir Onursal is senior environmental specialist in the Environment Unit of the World Bank's Latin America and the Caribbean Region Technical Department. Surhid P. Gautam is a consultant in the Country Operations Department I of the Latin America and the Caribbean Region. Library of Congress Cataloging-in-Publication Data Onursal, Bekir. Vehicular air pollution : experiences from seven Latin American urban centers / Bekir Onursal, Surhid P. Gautam. p. cm. -(World Bank technical paper : no. 373) Includes bibliographical references. ISBN 0-8213-4016-6 1. Motor vehicles-Motors-Exhaust gas-Environmental aspects- Latin America-Case studies. 2. Air-Pollution-Latin America- Case studies. 3. Air quality management-Latin America-Case studies. 4. Urban pollution-Latin America-Case studies. I. Gautam, Surhid. II. Title. III. Series. TD886.50555 1997 97-18346 363.738'7'098-dc21 CIP CONTENTS CHAPTER 1: VEHICULAR AIR POLLUTION: AN OVERVIEW .1................................ Urbanization in Latin America and the Caribbean .............................................2 Motorization in Latin America and the Caribbean .............................................5 CHAPTER 2: AIR POLLUTANTS AND THEIR EFFECTS ......................................15 Pollutant Emissions from Motor Vehicles ....................................................15 Hydrocarbon compounds.............................................................18 O zone.19...................................................................................................................................................... 19 Sulfur~~~~~~~~~~~~~~~- dioxide.19 c Dunisper Plutn Eision o i oltn s............................................................23 Healthla.Effects...of..Air..Pollutants.25.... 23 Nitogenodioxide.2 Polycyclic aromatic hydrocarbons.......................................................27 EeuieSulfury dixide2 Partoviculae mattersh2p8 The Ortloocobos.12 Re f erences.............................................................2 Choofurcabns.................b........................................................3 iv Contents Quantifying the Health Effects of Air Pollutants .................... .............................................. 31 Environmental Effects of Air Pollutants .................................................................. 36 Nitrogen oxides .................................................................. 36 Ozone .................................................................. 36 Sulftur dioxide .................................................................. 36 Particulate matter .................................................................. 37 Chlorofluorocarbons .................................................................. 37 Carbon dioxide .................................................................. 37 Ambient Air Quality Standards and Monitoring .................... .............................................. 37 Ambient air quality standards .................................................................. 37 Ambient air quality monitoring .................................................................. 38 Annex A: Estimated Emission Factors for U.S. Vehicles ........................ .......................................... 41 Annex B: Ambient Air Quality Standards and WA'HO Guidelines .................................................................. 45 References .................................................................. 51 CHAPTER 3: ABATEMENT MEASURES FOR VEHICULAR AIR POLLUTION .............. ..................... 55 Vehicle-Targeted Measures .................................................................. 59 Emission Standards and Related Measures for New Vehicles .................................................................. 59 Emission standards .................................................................. 59 Certification, assembly line testing, recall, and warranty .................................................................. 63 Emission Standards and Inspection Programs for In-Use Vehicles .............................................................. 65 Emission standards .................................................................. 65 Inspection and maintenance programs .................................................................. 66 Roadside inspection programs ................................................................... 69 Emission Restrictions on Imported Vehicles .................................................................. 70 Vehicle Emission Fees .................................................................. 70 Differential Vehicle Registration Fees and Taxes .................... .............................................. 70 Vehicle Retrofit Programs .................................................................. 71 Vehicle Replacement or Scrappage Incerntives .................................................................. 72 Fuel-Targeted Measures .................................................................. 73 Gaasoline Standards .................................................................. 73 Lead .................................................................. 73 Volatility .................................................................. 80 Benzene and other aromatic hydrocarbons .................................................................. 80 Reformulated gasoline .................................................................. 82 Oxygenated gasoline .................................................................. 84 Diesel Fuel Standards .................................................................. 87 Sulfur .................................................................. 87 Cetane number .................................................................. 88 Aromatic hydrocarbons and density .................................................................. 88 Alternative Fuels .................................................................. 89 Compressed natural gas .................................................................. 90 Liquefied petroleum gas .................................................................. 92 Methanol .................................................................. 94 Ethanol .................................................................. 96 Fuel Taxes .................................................................. 98 Tradable Permits and Credits .................................................................. 100 Transport Management Measures .................................................................. 101 Driving Bans .................................................................. 101 On-Street Parking and Trading Restrictions .................................................................. 104 Traffic Priority Measures for Buses .................................................................. 104 Ride Sharing .................................................................. 105 Staggered Work Hours .................................................................. 105 Speed Limits and Other Traffic Management Measures ................................................ .................. 106 Contents v Land Use Planning and Cotitrols ................................................ 106 Road Pricing ................................................ 107 Area Licensing ................................................ 109 Provision of Public Transport Services ................................................ 109 Promotion of Nonmotorized Transport ................................................ 112 Provision of Off-Street Parking ................................................ 112 Annex A: Vehicular Emission Standards in the Unitedl States ................................................ 113 New Vehicles .................................................: 113 In-Use Vehicles ................................................ 114 References .............................................. ...119 CHAPTER 4: CASE STUDIES ................................................ 123 Mexico City ................................................ 125 Ambient Air Quality ................................................ 127 Sources of Pollutants ................................................ 130 Institutional Responsibilities ................................................ 134 Federal institutions ................................................ 134 State of Mexico institutions ................................................ 136 Federal District institutions ................................................ 137 Metropolitan institutions ................................................ 137 Implemented Measures ................................................ 140 Vehicle emission standards .................................................;. 141 Vehicle inspection programs ................................................ 142 Fuel-targeted measures ................................................ 145 Transport management ................................................ 148 Air quality monitoring ................................................ 149 Evaluation of Implemented Measures ................................................ 150 Vehicle emission standards ................................................ 150 Vehicle inspection programs ................................................ 151 Fuel-targeted measures ................................................ 152 Transport management ................................................ 155 Air quality monitoring ................................................ 156 Santiago .................................................157 Ambient Air Quality ................................................ 159 Sources of Pollutants ................................................ 163 I ns titutional R esponsibilities ................................................. 165 National institutions ................................................ 165 Regional institutions ................................................ 167 Local institutions ................................................ 167 Measures Implemented ................................ 168 Vehicle emission standards ................................. 168 Vehicle inspection programs ................................ 170 Fuel-targeted measures ................................ 171 Transport management ................................ 172 Air quality monitoring ................................ 174 Evaluation of Implemented Measures ................................. 176 Vehicle emission standards ................................1 76 Vehicle inspection programs ................................ 177 Fuel-targeted measures ................................ 178 Transport management ................................ 180 Air quality monitoring ................................ 181 Siao Paulo .................................. 183 Ambient Air Quality ................................. 185 Sources of Pollutants ................................ 186 vi Contents Institutional Responsibilities ................................ 189 Federal institutions ................................ 190 State institutions ................................1 90 Local institutions ................................1 91 Implemented Measures ................................1 91 Vehicle emission standards ................................1 91 Vehicle inspection programs ................................ 194 Fuel-targeted measures ................................ 195 Transport management ................................ 196 Air quality monitoring ................................ 196 Evaluation of Implemented Measures ................................ 197 Vehicle emission standards ................................ 197 Vehicle inspection programs ................................ 198 Fuel-targeted measures ................................1 99 Transport management ................................ 202 Air quality monitoring ................................ 203 Belo Horizonte .................................205 Ambient Air Quality ................................ 207 Sources of Pollutants ................................ 208 Institutional Responsibilities ................................ 211 Federal institutions ................................ 211 State institutions ................................ 211 Metropolitati institutions ................................ 211 Local institutions ................................ 212 Implemented Measures ................................ 212 Vehicle emission standards ................................ 212 Vehicle inspection programs ................................ 212 Fuel-targeted measures ................................ 212 Transport management ................................ 212 Air quality monitoring ................................ 213 Evaluation of Implemented Measures ................................ 213 Vehicle emission standards ................................ 213 Vehicle inspection programs ................................ 213 Fuel-targeted measures ................................ 213 Transport management ................................ 214 Air quality monitoring ................................ 214 Buenos Aires .................................217 Ambient Air Quality ................................ 219 Sources of Pollutants ................................ 220 Institutional Responsibilities ................................ 221 National institutions ................................ 221 Provincial institutions ................................ 222 Metropolitan institutions ................................ 222 Local institutions ................................ 222 Implernented Measures ................................ 222 Vehicle emission standards ................................ 222 Vehicle inspection programs ................................ 224 Fuel-targeted measures ................................ 226 Tra nsport manag ement ................................ 227 Air quality monitoring ................................ 229 Evaluation of Implemented Measures ................................ 229 Vehicle emission standards ................................ 229 Vehicle inspection programs ................................ 230 Contents vii Fuel-targeted measures ................................................................. 230 Transport management ................................................................. 232 Air quality monitoring ................................................................. 233 Rio dejaneiro ................................................................. 235 Ambient Air Quality ................................................................. 237 Sources of Pollutants ................................................................. 240 Institutional Responsibilities ................................................................. 242 Federal institutions ................................................................. 242 State institutions ................................................................. 242 Local institutions ................................................................. 243 Implemented Measures ................................................................. 244 Vehicle emission standards ................................................................. 244 Vehicle inspection programs ................................................................. 244 Fuel-targeted measures ................................................................. 244 Transport management ................................................................. 244 Air quality monitoring ................................................................. 245 Evaluation of Implemented Measures ................................................................. 245 Vehicle emission standards ................................................................. 245 Vehicle inspection programs ................................................................. 245 Fuel-targeted measures ................................................................. 246 Transport management ................................................................. 246 Air quality monitoring ................................................................. 246 Santafe de Bogota ................................................................. 249 Ambient Air Quality ................................................................. 251 Sources of Pollutants ................................................................. 254 Institutional Responsibilities ................................................................. 256 National institutions ................................................................. 256 Regional and local institutions ................................................................. 256 Implemented Measures ................................................................. 257 Vehicle emission standards ................................................................. 257 Vehicle inspection programs ................................................................. 258 Fuel-targeted measures ................................................................. 259 Transport management ................................................................. 260 Air quality monitoring ................................................................. 261 Evaluation of Implemented Measures ................................................................. 262 Vehicle emission standards ................................................................. 262 Vehicle inspection programs ................................................................. 262 Fuel-targeted measures ................................................................. 262 Transport management ................................................................. 263 Air quality monitoring ................................................................. 264 References ................................................................. 265 CHAPTER 5: CONCLUSIONS ................................................................. 269 Planning Tools for Air Quality Management .................................................................. 272 Air quality monitoring ................................................................. 272 Emission inventories and dispersion moclels ................................................................. 273 Epidemiological studies ................................................................. 273 Air Pollution Control Measures ................................................................. 274 Vehicle-targeted measures ................................................................. 274 Fuel-targeted measures ................................................................. 276 Transport management measures ................................................................. 279 Formulation and Implementation of Urban Air Quality Managemenit Strategy ....................................... 281 References ................................................................. 281 viii Contents Flgures CHAPTER 1: VEHICULAR AIR POLLUTION: AN OVERVIEW 1.1 Population residiig in urban areas, 1950-2010 ....................................................................................2 1.2 Global distribution of the motor vehicles fleet, 1994 ............................................................................ 7 1.3 Per Capita Income and Motorization in Latin America and the Caribbean, 1970, 1980, and 1990 ...... 11 CHAPTER 3: ABATEMENT MEASURES FOR VEHICULAR AIR POLLUTION Fuel-Targeted Measures 3.1 Relatio nship b etween lead addition a nd octane number .................................................................... 74 3.2 Timelines for eliminatinig leaded gasoline in Latin America and the Caribbean .......... ................... 76 Transport Management Measures 3.3 Pollutant emission rates front vehicles as a functiorn of vehicle speed ............................................. 102 CHAPTER 4: CASE STUDIES Mexico City 4.1 Ambient lead concentrations in the MCMA, 1990-94 ...................................................................... 131 4.2 Share of air pollutant emissions from motor vehicles and other sources in the MCMA, 1994 ...... 131 4.3 Relation between ambient HC. NO, and maximum ozone concentrations in the MCMA ........... 133 4.4 Distribution of vehicles by model-year in the MCMA, 1994 ...................... ....................................... 133 4.5 Prices of Magna Sin and Nova in the MCMA, 1990-96 ................................. ................................... 148 4.6 Ambient concentrations of formaldehyde in the MCMA before and after adding MTBE ............ 154 4.7 Ambient concentrations of ozone in Mexico City before and after adding MTBE ......... ............... 154 Santiago 4.8 Annual ambient concentrations of PM-10 in the SMA, 1989-94 ................. .................................... 160 4.9 Annual average concentrations of SO, in the SMA, 1989-92 ........................................................... 162 4.10 Share of air pollutant emissions from motor vehicles and other sources in the SMA, 1992 .......... 164 4.11 Effects of maintenance on pollutant emissions from buses in the SMA ............. ............................. 178 Sio Paulo 4.12 Average ambient pollutant concentrations in the SPMR, 1983-95 .................................................. 187 4.13 Share of air pollutant emissions from road-based motor vehicles and other sources in the SPMR, 1995 ................................................................................... 188 Belo Horizonte 4.14 Road-based vehicles in Belo Horizonte ........................................... ........................................ 210 Rio dejaneiro 4.15 Air pollutants by emission source iTn the RJMR, 1978 ....................................... ................................ 241 4.16 Types of vehicles in Rio deJaneiro, 1994 ................................................................................... 242 Santafe de Bogota 4.17 Ambient TSP concentrations in Santafe de Bogota, 1986-89 .................... ...................................... 252 4.18 Ambient SO2 concentrations in Santafe de Bogota, 1986-88 ........................................................... 253 Tables CHAPTER 1: VEHICULAR AIR POLLUTION: AN OVERVIEW 1.1 Urbanization in Latin America and the Caribbean, 1970-94 ............................................................... 3 1.2 Urban agglomerations with populations of more than 1 million people in Latirn America and the Caribbean, 1994 .............. ...................... -............................................. 4 1.3 Motor vehicle fleets and motorization in Latin America and the Caribbean, 1970-90 ...................... 6 1.4 Motor vehicle fleets and motorizationi in selected Latin American urban areas, 1994 ......... ............. 7 1.5 Per capita income in Latin America and the Caribbean, 1970-94 ................. ..................................... 9 CHAPTER 2: AIR POLLUTANTS AND THEIR EFFECTS 2.1 Exhaust emissions from unconitrolled light-duty gasoline-fueled vehicles at different driving modes ................................................................................... 16 Contents ix 2.2 Exhaust emissions lfroin unicontrolle(d heavy-duty vehicles at different driving modes ......... ........... 17 2.3 Exhatist emissions of toxic air poollutalits fromt gasoline-fueled cars .................................................. 19 2.4 Typical factors affecting motor vehiicle emiiissiotn rates ...................................... .................................. 22 2.5 List of modifications to adopt MOBILE5a to Monterrey, Mexico ...................................................... 23 2.6 Equivalent model-years for Monterrey. Mexico and U.S. vehicles .................. ................................... 24 2.7 Blood lead levels olfdifferent urban popullation groups in selected Latin American and Caribbean countries ...................................................................................... 32 2.8 Ranges and central estimates for dose-response coefficients ....................... ...................................... 34 2.9 Alternative mortality subgroups from idifferent studies ..................................... ................................ 35 Annex A: Estimated Emission Factors for U.S. Vehicles A.1 Estimated emission factors for U.S. gasoline-fueled passenger cars with different emission control technologies ..................................................................................... 41 A.2 Estimated emission factors for U.S. gasoline-fueled medium-duty trucks with different emission control technologies ......................... ............................................................ 42 A.3 Emission and fuel consumption factors for U.S. diesel-fueled passenger cars and light-duty trucks with different emission control techniologies ..................................................... ...................... 43 A.4 Emission and fuel consumption factors for U.S. heavy-duty diesel-fueled trucks and buses with different emission control technologies ..................................................................................... 43 Annex B: Ambient Air Quality Standards and MWHO Guidelines B.1 Air quality standards and MWHO guidelinie for carbon monoxide ...................................................... 45 B.2 Air quality standards and WAHO guidelinle for nitrogen dioxide ................... .................................... 46 B.3 Air quality standardIs and WHHO gui(elinle for ozone .......................................................................... 47 B.4 Air quality standards and WHO guidelinie for sulfur dioxide ...................... ...................................... 48 B.5 Air quality standards and WHO guiidelinle for TSP ................................................................. ............ 49 B.6 Air quality standards for PM-10 ........................... .......................................................... 50 B.7 Air quality standards and WHO guidelinle for lead ............................................................................. 50 CHAPTER 3: ABATEMENT MEASURES FOR VEHICULAR AIR POLLUTION 3.1 Typical measures for the abatement of air pollution from motor vehicles ............ ........................... 56 Vehicle-Targeted Measures 3.2 Estimates for performance and costs of exhaust emission control technologies for light-duty gasoline-fueled vehicles ..................................................................................... 63 3.3 Estimates for performance and costs of exhaust emission control technologies for heavy-duty diesel-fueled vehicles ..................................................................................... 64 3.4 Estimated costs of centralized and decentralized inspection and maintenance programs in Arizona ...................................................................................... 68 3.5 Motor vehicle inspection schedule in Singapore, by vehicle age ................... .................................... 68 Fuel-Targeted Measures 3.6 Standards, quality, and consumptioni of leadled gasoline in Latin America and the Caribbean, 1996 ...................................................................................... 77 3.7 Standards and quallity for Reid vapor pressure of gasoline in Latin America and the Caribbean, 1996 .......... ........................................................................... 81 3.8 Standards and quallity for benzene and arormatic hydrocarbons content of gasoline in Latin America and tihe Caribbean ..................................................................................... 81 3.9 Baseline parameters for gasoline in the Ulnite(d States (1990 average) .............. ................................ 83 3.10 Properties of severely reformulated and( city gasoline in Finland .................. .................................... 83 3.11 Compositional constraints for regular and Type 2 gasoline grades in Thailand ............................... 84 3.12 Mexico's refinery investment program. 1C996-98 ................................................................................ 85 3.13 Oxygenate requirements for motor vehicle fLels ................................................................................ 86 3.14 Standards for aromatic HC, PAH, and( dlenisity of diesel fuel in Finland, Sweden, and the United States ..................................................................................... 89 3.15 Properties of some conventional and alterniative fuels ...................................... ................................. 90 3.16 Prices of some conventional and alternative fuels in the United States .............. .............................. 90 3.17 Exhaust ernissions from CNG- and diesel-fuieled buses ................................. ..................................... 91 x Contents Transport Management Measures 3.18 Pollutant emissions by different transport modes in Mexico City .................. ................................. 103 3.19 Effects of bus priorities oni bus exhaust emissions ............................................................................ 104 3.20 Examples of congestion pricing, 1994 ................................................................................ 108 Annex A: Vehicular Emission Standards in the United States A.1 U.S. emission standards for new passenger cars ................................................................................ 114 A.2 U.S. emission standards for new light-duty trucks ............................................................................. 115 A.3 U.S. emission standards for new heavy-duty vehicles ........................................................................ 116 A.4 U.S. Federal evaporative emission standards for new vehicles ...................... .................................... 117 A.5 U.S. emission standards for in-use vehicles ................................................................................ 117 A.6 U.S. emission standards for in-use vehicles based on the IM240 test ................ .............................. 118 CHAPTER 4: CASE STUDIES Mexico City 4.1 Number of days with high ozone concenitrations in the MCMA, 1988-95 ............. ......................... 128 4.2 Pollutant emissions by source in the MCMA, 1994 ........................................................................... 132 4.3 Pollutant emission rates from various transport modes in the MCMA ............................................ 134 4.4 Institutional responsibilities for vehicular air pollution control in the MCMA ........... ................... 135 4.5 Exhaust emission standards for new cars in Mexico ......................................................................... 141 4.6 Exhaust emission standards for new light-duty commercial vehicles in Mexico ........... .................. 141 4.7 Proposed exhaust emission standards for new urban microbuses in Mexico (3,000-5,000 kilograms) ................................................................................ 142 4.8 Exhaust emission standards for new heavy-duty commercial vehicles in Mexico using gasoline or alternative fuels ................................................................................ 142 4.9 Exhaust etnission standards for new heavy-duty diesel-fueled vehicles in Mexico .......... ................ 143 4.10 Exhaust emission standards for in-use light-duty gasoline-fueled vehicles in Mexico ......... ........... 143 4.11 Exhaust CO and HC emission standards for in-use commercial, multiple-use, and service vehicles and for trucks fueled with LPG, CNG, methanol, anid mixture of these fuels with gasoline or diesel fuel in Mexico ......................... ....................................................... 144 4.12 Exhaust smoke emission standards for in-use diesel-fueled vehicles in Mexico ............................. 144 4.13 Gasoline quality standards in Mexico ................................................................................ 146 4.14 IMECA values for the MCMA ................................................................................ 150 4.15 Significance of the IMECA values ................................................................................ 151 4.16 Typical fuel quality parameters of Magna Sin and Nova, 1995 ....... . ................................................ 153 Sanitiago 4.17 Ambient ozone levels exceedinig the 1-hour Chilean standard in the SMA, 1992-93 .................... 161 4.18 An nual average concentrations of SO2 in the SMA, 1987-91 ............................................................ 162 4.19 Institutional responsibilities for air pollution control in the SMA ................................................... 166 4.20 Exhaust emission standards for new vehicles in Chile ................................................. ..................... 169 4.21 Exhaust emission standards for in-use vehicles registered before the effective dates in Chile ...... 170 4.22 Exhaust emission standards for in-tse vehicles registered after the effective dates in Chile ......... 171 4.23 Gasoline quality standards in Chile . ................................................................................ 172 4.24 Projected refinery investmenits required to eliminate lead from gasoline ............. ......................... 173 4.25 Air quality indices and pollutant concentrations .............................................................................. 175 4.26 Air quality indices and human health effects ................................................................................ 176 4.27 New vehicle registrations in Chile. 1992-95 ................................................................................ 177 4.28 Comparison of exhaust emissions between controlled and uncontrolled heavy-duty vehicles ................................................................................ 177 4.29 Average pollutant emission factors for in-use vehicles in the SMA, 1988-89 ............ ...................... 178 4.30 Typical gasoline quality parameters in Chile, 1995-96 .......................................... ........................... 179 4.31 Estimated emissions from diesel- and CNG-fueled buses ..............................I.................................. 180 Sao Paulo 4.32 Sources of TSP and PM-10 in ambient air at four stations in the SPMR, November 1986-Novetnber 1987 ................................................................................ 189 4.33 Daily trip distribution in the SPMR, 1990 .189 Contents xi 4.34 Exhaust emission standards for new light-duty vehicles in Brazil ................... ................................. 192 4.35 Exhaust emission standards for new dliesel-fueled vehicles in Brazil ................. .............................. 193 4.36 Exhaust emission standards for in-use light-duty vehicles with spark-ignition engines in the SPMR .............................................................................. 194 4.37 Ambient air quality indices for the SPMR ............................................................................. 197 4.38 Air quality categories for the SPMR ......................... .................................................... 197 4.39 Average emission factors for new light-duty vehicles in the SPMR ................... ............................... 199 4.40 Estimnated reductions in pollutant emissionis due to the inspection and maintenance program in the SPMR, 1994 ....................................................................... .200 4.41 Effect of the ethanol content of gasohol on vehicle exhaust emissions .............. ............................ 200 4.42 Ambient concentrationis of aldehydes in Sao Paulo and four U.S. cities ......................................... 201 Belo Horizonte 4.43 Annual average TSIP concentrations in the BHMR, 1985-92 ........................................................... 208 4.44 Ambient PM-10 conceritratiorns in the BHMR, 1995-96 ................................................................... 209 4.45 Estimates of transport-related pollutant emissions in the BHMR, 1991 .......................................... 209 4.46 Bus inspection results for Belo Horizonte, 1988-93 ......................................................................... 214 Buenos Aires 4.47 Ambient CO concentrations in Buenos Aires ................................................................................. 220 4.48 Amnbient lead concentrations in Buenos Aires ................................................................................. 220 4.49 Estimates of emissions frorn gasoline- and diesel-fueled vehicles in the Federal Capital and the BAMA, 1993 ................................................................................. 221 4.50 Exhaust emission standards for new liglht-duty vehicles in Argentina ............... .............................. 223 4.51 CO, HC, and NO. exhaust ernission standards for new diesel-fueled heavy-duty vehicles in Argentina ................................................................................. 224 4.52 PM exhaust emission standards for new diesel-fueled vehicles in Argentina ............ ...................... 224 4.53 Exhaust emission standards for in-tuse vehicles with spark-ignition engines in Argentina ............. 224 4.54 Exhaust emission standards for in-use vehiicles with diesel engines in Argentina .......... ................ 224 4.55 Fuel consumption in the Federal Capital, Province of Buenos Aires, and Argentina, 1996 .......... 226 4.56 Gasoline quality standards in Argentina . ................................................................................. 227 4.57 Conversion of vehicles to CNG in the BAMA and Argentina ........................ ................................... 227 4.58 Typical fuel prices in Argentina, December 1996 and October 1994 .............................................. 228 4.59 Passenger trips in the BAMA, 1970, 1992, and 1994 ......................................................................... 232 Rio (ie Janeiro 4.60 Annlual average concentrationis of TSP in the Municipality of Rio dejaneiro and the RJMR. 1987-93 ................................................................................. 238 4.61 Concentrations of TSP in the Municipality of Rio deJaneiro, 1994 ................................................ 239 4.62 Anniual average concentrations of SO, in the RJMR, 1980-84 ......................................................... 240 4.63 Estimated industrial emissions of PM and SO2 in the Municipality of Rio dejaneiro and the RJMR, 1978, 1981, and 1983 ................................................................................. 241 4.64 Transport modes in Rio deJaneiro, 1994 .................................................................................. 242 4.65 Black smoke test results of heavy-duty vehicles in the RJMR during two periods .......... ................. 246 Santafe de Bogota 4.66 Estimates of emissions from fixed and mobile sources in Santafe de Bogota, 1991 ....................... 254 4.67 Age of vehicle fleet in Santafe de BogotA, 1991 ................................................................................ 255 4.68 Exhaust emission standards for new andl in-utse gasoline-fueled vehicles in Colombia .................. 257 4.69 Exhaust smoke opacity standards for newv and in-use diesel-fueled vehicles in Colombia ............. 258 4.70 Exhatust emission standards for gasoline-and(i diesel-fueled vehicles imported, assembled, or manufactured in Colombia ................................................................................. 258 4.71 Typical quality of gasoline consumed in Colombia, 1996 ................................................................. 259 4.72 Gasoline quality standards in Colombia ................................................................................. 260 4.73 Diesel fuel quality standards in Colombia ................................................................................. 260 CHAPTER 5: CONCLUSIONS 5.1 Maximum ambient pollutant concentrationis in selected Latin American urban centers ................................................................................. 270 xii Contents Boxes CHAPTER 2: AIR POLLUTANTS AND THEIR EFFECTS 2.1 Commonly used air dispersion models for vehicular emissions ......................................................... 25 CHAPTER 3: ABATEMENT MEASURES FOR VEHICULAR AIR POLLUTION 3.1 Public awareness and education ............................................................................ 57 Vehicle-Targeted Measures 3.2 Pollution control technologies for vehicles with four-stroke spark-ignition engines .......... .............. 60 3.3 Pollution control technologies for diesel-fueled vehicles ................................................................... 62 Fuel-Targeted Measures 3.4 How was lead eliminated froml gasoline in Costa Rica? ...................................................................... 79 3.5 Liquefied petroleum gas in Canada ........................................................................... 94 3.6 Proalcohol program in Brazil ........................................................................... 97 CHAPTER 4: CASE STUDIES Santiago 4.1 Air pollution control efforts in Chile, 1916-present ......................................................................... 165 Maps CHAPTER 4: CASE STUDIES Mexico City metropolitan area .......................................................................... 126 Santiago metropolitan area .......................................................................... 158 Sao Paulo metropolitan region .......................................................................... 184 Belo Horizonte metropolitan region .......................................................................... 206 Buenos Aires metropolitan area .......................................................................... 218 Rio dejaneiro metropolitan region .......................................................................... 236 Santafe de Bogota .......................................................................... 250 FoREwoRD D t eteriorating air quality is a major en- cludes contributions from government represen- vironmental problem in many Latin tatives, consultants, and nongovernmental orga- American urban centers, including nizations (NGOs) from various Latin American megacities like Mexico City, Saio countries. The report was also aided by Bank Paulo, Santiago, and Buenos Aires. Most air pol- experts who are working on the vehicular air lution in Latin America, whLich occurs in highly pollution problem in Latin America. The report urbanized areas with unfavorable topographical is intended mainly for policymakers in Latin and meteorological condiitions, is caused by America and the Caribbean who are responsible motor vehicles. During 1970-90 the growth of for controlling and managing air pollution, as the region's vehicle fleet was phenomenal: it well as for technical audiences and citizens grew by about 250 percent, reaching 37 million groups. The detailed case studies should also vehicles. With new economic and trade liberal- interest policymakers in other parts of the world. ization policies and the formation of various The report has five chapters. Chapter 1 pro- trade blocks in the region, the vehicle fleet is vides an overview of vehicular air pollution, fo- expected to grow even more quickly over the cusing on population, urbanization, and next few decades. If appropriate measures are motorization trends and economic and trade not taken soon, vehicular air pollution in Latin developments in Latin America and the Carib- America and the Caribbean is likely to worsen, bean. Chapter 2 discusses sources, properties, posing a great threat to human health and emission characteristics, dispersion, and environ- welfare. mental and health effects of vehicular air pol- The idea for this study emerged during a 1993 lutants. This chapter also quantifies the health World Bank-sponsored Urban Environment Con- effects of air pollution and provides ambient air ference in Quito, Ecuador. During the confer- quality standards adopted in Latin American and ence the severity of air pollution in Mexico and Caribbean countries. Chapter 3 presents the Sao Paulo and the pollution control measures vehicle- and fuel-targeted and transport manage- taken in both urban centers were presented. ment measures used to control vehicular air Participants at the conference, most of whom pollution, drawing on examples from the United were high-level government officials from States, European Union countries, and others. throughout Latin America, felt that these and Chapter 4 provides case studies for seven Latin other city experiences on vehicular air pollution American urban centers: Mexico City, Santiago, should be more widely disseminated. This idea Sao Paulo, Belo Horizonte, Buenos Aires, Rio was subsequently accepted by the World Bank, de Janeiro, and Santaf6 de Bogoti. Each case and the Environment Unit: of the Latin Ameri- study analyzes ambient air quality, sources of can and the Caribbean Region (LATEN) was pollution, institutional responsibilities, and mea- entrusted to undertake this work. sures implemented to abate vehicular air pollu- This study, which took almost three years to tion, and evaluates these measures. Chapter 5 complete, is a result of extensive research. It in- draws conclusions based on the findings of the xiii xiv Foreword report and provides lessons learned from imple- in Latin America and the Caribbean and can mentation of air pollution measures in various help policymakers design air pollution control urban centers. strategies. I would like to thank Bekir Onur-sal (Senior Environmental Specialist, LATEN) and Surhid Sri-Ram Aiyer Gautam (Consultant, LAl CO) for preparing this Director report. Extensive help was provided by various Latin America and the Caribbean Region colleagues in the Bank who have been duly iden- The World Bank tified and acknowledged. My hope is that this 1818 H Street, N.W. report provides insight to vehicular air pollution Washington, D.C. 20433 USA AcKNOWLEDGMENTS his report was prepared under the aus- ment; Peter Brandriss for his technical com- T pices of the Environment Unit of the ments and editing; Mercedes Aleman for incor- Latin America and the Caribbean Re- porating the editorial changes; and Cynthia gional Office of the World Bank. Many Stock for designing and typesetting the report. individuals, both from within the Bank and from We are also grateful to Odil Tunali for research public and private institutions in the region, conducted during initiation of this project and made important contributions for which the au- Laura Alvarez and Patricia Lee for their admin- thors are deeply grateful. We owe special thanks istrative support. to Sri-Ram Aiyer and William Partridge for their Several other people, listed below, provided support; Asif Faiz for his encouragement in start- information, reviewed chapters, and made sug- ing this work and for his review and suggestions gestions without which this report would have on the draft report; Douglas Graham, Magda been incomplete. We are grateful for their valu- Lovei, and Masami Kojima for their thorough able contributions. reviews; Paul Holtz for editing the entire docu- Chapter 1 Chapter 4 Valeriano Garcia (World Bank) Sonia Plaza (World Bank) Mexico City Dan Biller (World Bank) Michael Walsh (Independent Consultant, Joachim von Amsberg (World Bank) Virginia) Sandra Rosenhouse (World Bank) Fernando Menendez Garza (Eco-infraestructura y Financiamiento Consultores, Mexico City) Chapter 2 Jesdis Berumen Garcia (Pemex, Mexico City) Maureen Cropper (World Bank) Jose RosasJaramillo (Pemex, Mexico City) Nathalie Simon (World Bank) Sergio Sanchez (Secreratarfa del Medio James Listorti (World Bank) Ambiente DF, Mexico City) Olga Corrales (World Bank) Eduardo Olivares Lechuga (Secrerataria del Livia Benavides (World Bank) Medio Ambiente DF, Mexico City) Bruce Carlson (World Bank) Gloria Yafiez Rodriguez (DDF, Mexico City) - Carl-Heinz Mumme (World Bank) Chapter 3 Christopher S. Weaver (Engine, Fuel Santiago and Emissions, Engineering Inc., California) Ricardo Katz (Universidad de Chile, Santiago) Lit-Mian Chan, (Engine, Fuel and Emissions, Christian Santana (CONAMA-RM, Santiago) Engineering Inc., California) Gianni L6pez (CONAMA-RM, Santiago) Dan Biller (World Bank) Patricia L6pez (CONAMA, Santiago) John Flora (World Bank) Ignacio Olaeta (SESMA, Santiago) xv xvi Acknowledgments Estanislao Gacitua (World Bank) Nestor Nazer (Subsecretarfa de Mineria, Fernando Manibog (World Bank) Buenos Aires) Almudena Manterola (World Bank) Alberto Cuesta (Businessman, Buenos Aires) Jorge Rebelo (World Bank) Sdo Paulo Gerhard Menckhoff (World Bank) Gabriel Murgel Branco (CETESB, Sao Paulo) Alfred Szwarc (CETESB, Sao Paulo) Rio dejaneiro Luiz Carlos da Costa (CETESB, Sao Paulo) Sergio Margulis (FEEMA, Rio dejaneiro) Olimpio de Melo AlvaresJr. (CETESB, Saio Jose Antonio Sales (FEEMA, Rio dejaneiro) Paulo) Antonio Carlos Dias (FEEMA, Rio dejaneiro) Jorge Rebelo (World Bank) Jorge Rebelo (World Bank) Fabiano de Franco (Consultant, Washington, D.C.) SantaJf de Bogotd Eduardo Uribe Botero (DAMA, Santafe de Belo Horizonte Bogota) Beverly Wen Yuh Liu (FEAM, Belo Horizonte) Bernardo Garcia Martinez (Ministerio del Jorge Rebelo (World Bank) Medio Ambiente, Santafe de Bogoti) Stella Bastidas (Ministerio del Medio Buenos Aires Ambiente, Santafe de Bogoti) Maria Dolores Mazzola (SRNAH, Buenos German G6mez (Consultant, Santafe de Aires) Bogoti) Roberto Domecq (Secretaria de Transporte Juan David Quintero (World Bank) Automotor, Buenos Aires) Osvaldo Galuppo (Iteveba, Buenos Aires) Norma Cadoppi (Fundaci6n Argentina We also would like to thank the many authors SIGLO21, Buenos Aires) whose work we have freely quoted in this report. Juan Morrone (ADEFA, Buenos Aires) Finally, we would like to thank our families for Alberto Moran (Instituto Nacional de la their patience and support during a long project Administraci6n Puiblica, Buenos Aires) that consumed many evenings and weekends. Anibal Sar (YPF, Buenos Aires) Carlos Calvo (YPF, Buenos Aires) Bekir Onursal Jorge Elias (La Naci6n, Buenos Aires) Surhid Gautam June 1997 ABSTRACT L atin America and the Caribbean is the particulate matter, and carbon monoxide are most urbanized region in the develop- pollutants of greatest concern. This report ana- ing world with a rapidly growing mo- lyzes the pollutants emitted by motor vehicles, tor vehicle fleet. The number of urban their effects, and measures targeted to vehicles, areas with populations exceeding 1 million was fuels, and transport management to control 43 in 1994 and is expected to increase to 52 in them. Case studies for seven urban areas Mexico 2010. Airborne pollutant levels in many urban City, Santiago, Sao Paulo, Belo Horizonte, areas far exceed the national, regional, or local Buenos Aires, Rio de Janeiro, and Santafe de standards and World HealthLOrganization gLide- Bogota are presented to illustrate how these lines. The main source of air pollution in these measures have been used in the region and how urban areas is motor vehicles, especially those they can be strengthened. which are old and poorly maintained. Ozone, xvii EXECUTIVE SUMMARY I n recent years policymiakers in metropoli- people in the region. These policy measures typi- tan areas throughout the developing world cally focus on vehicles, fuels, and transport man- have been working to monitor and miti- agement. Seven case studies are presented to gate worsening levels of air pollution. For illustrate how these policies have been used in several reasons efforts to stem air pollution have the region and how they can be strengthened. been especially intense in Latin America and the Formulating and implementing a comprehen- Caribbean. The region's urbanization rate is the sive and effective urban air quality management highest in the developing world. Population strategy require a coordinated effort among na- growth has given the region four of the woIrld's tional, regional, and local institutions represent- ten most populated urban centers-Sao Paulo. ing the variousjurisdictions in urban areas. At a Mexico City, Buenos Aires, and Rio dejaneiro- minimum, institutions responsible for environ- as well as thirty-nine others with populations ex- ment, transport and traffic, fuel quality and sup- ceeding 1 million people. And motorization, with ply, industry, and health issues should large numbers of old and poorly maintained ve- participate. The participation of the scientific hicles as well as growing domestic and imported and business communities, labor unions, and vehicles, is on the rise. As a result airborne pol- nongovernmental organizations is also impor- lutant levels in many of the region's urban cen- tant. Moreover, a participation plan should be ters far exceed national, regional, or local developed to ensure that the public's views are standards and World Health Organization guide- identified and incorporated into the lines. The pollutants of greatest concern are decisionmaking process. ozone, particulate matter, and carbon monox- Formulation and implementation of specific ide. In Mexico City, Sao Paulo, and Santiago pollution control measures generally have been these and other air pollutants threaten human hampered by unclear or overlapping institu- health and quality of life. Other urban centers tional responsibilities; inadequate equipment, of the region-such as Belo Horizonte, Buenos technical expertise, and human and financial Aires, Rio dejaneiro, and Santafe de Bogota- resources; weak financial management; lack of are expected to face similar problems. Wlhat ac- political will; and limited public support or par- tions should the region's policymakers take to ticipation. The institutions responsible for these prevent and control these problems? efforts must strengthen their human and finan- Because most of the region's urban air pollu- cial resources and their management systems if tion is contributed by motor vehicles, this repor-t they are to implement and enforce an effective attempts to answer that qtuestion by analyzing air quality management strategy. The exact com- the pollutants emitted by motor vehicles, their position of the strategy cannot be the same for effects, and the measures that can be used to all urban centers, but should be based on the control them. By helping the region's policy- specific air pollution problems and other char- makers to formulate and implement strategies acteristics of each urban center. Any air pollu- to control pollutants, this report hopes to lower tion abatement and control strategy must be pollution levels and raise the quality of life for guided by a careful evaluation of emissions and xix xx Executi7le Summasy ambient air quality data, air dispersion models, benzene) should be lowered to reduce human and estimated costs and benefits of different exposure. Diesel fuel standards should include measures. Still, some general guidelines can be limits for parameters that affect emissions of recommended for most Latin American and sulfur, particulate matter, and carcinogens. In Caribbean urban areas. locations where compressed natural gas or liq- Vehicle-targeted measures should adldress both uefied petroleum gas are available, conversion new and in-use vehicles. Establishing emission of high-use vehicles (such as buses) to these al- standards for new vehicles is an effective mea- ternative fuels should be considered. Fuel taxes sure in vehicle-manufacturing countries where should be used to promote use of cleaner alter- state-of-the-art pollution control technologies native fuels or cleaner grades of the same fuels have not yet been adopted. These standards must and to reduce fuel consumption and associated be achievable in terms of technology availability pollutant emissions. and must weigh the potential benefits against By themselves transport management measures the additional cost to consumers. Imported ve- are insufficient to eliminate air pollution prob- hicles should be subject to international stan- lems, but are important as a complement to ve- dards for new model-year vehicles. Compliance hicle- and fuel-targeted measures in designing should be monitored both for imported vehicles air quality management strategies. Traffic flow and newly manufactured domestic vehicles. should be improved through carefully planned Given the composition and age of the vehicle infrastructure investment, traffic management, fleet in Latin America and the Caribbean, poli- road pricing, high-occupancy vehicle restric- cies should be developed that keep old and tions, and other measures that reduce travel poorly maintained vehicles off the road. Emis- time and pollutant emissions. Driving bans for sion standards for in-use vehicles should be es- responding to emergency episodes of extreme tablished and verified through periodic air pollution should be designed to prevent inspection and maintenance programs and road- strategies that circumvent the bans, should side inspections. Mexico City's experiences with avoid exemptions that make the bans porous inspection and maintenance programs are valu- (and so defeat their purpose), and should re- able. These measures can be complemented with ward the use of clean vehicles. Growth in the programs to purchase and scrap the oldest and number of vehicles in circulation should be most polluting vehicles. slowed by improving the quality, efficiency, ac- Fuel quality standards compatible with interna- cessibility, and value of public transport, and tional standards should be established, and mea- by improving conditions for nonmotorized sures to meet these standards should be designed transport. In the long term land use planning and implemented. Unleaded gasoline with mini- and control measures should be used to relieve mal sulfur content should be used in new model- pressures on urban centers and to create multi- year vehicles equipped with catalytic converters. nucleated urban areas. The lead content of gasoline used by older ve- Although some measures, such as eliminating hicles should be reduced and ultimately elimi- lead from gasoline, should be used as extensively nated to minimize human exposure to airborne as possible, no single set of measures can be rec- lead. In doing so, however, it is also important ommended for the entire region. Each urban to evaluate the health and environmental im- area has specific pollution problems and unique pacts of the reformulated gasoline-especially environmental, physical, social, and economic its effect on emissions from vehicles not factors,andsomustdevelopitsownmixofpol- equipped with catalytic converters. Oxygenates lution control strategies. Measures that improve can be used both as a replacement for lead to transport management may share common enhance the octane number, and as an additive principles and dynamics, but should be tailored to reduce carbon monoxide emissions in urban to each area. Even fuel standards, which every areas (especially at high altitudes) with high country should have, may vary according to lo- ambient carbon monoxide concentrations. cal conditions. Whatever the approach, the Where feasible, ether-based oxygenates should tools and knowledge are available to make sub- be used rather than alcohol-based oxygenates. stantial gains at reasonable cost in most cities, In addition, the vapor pressure of gasoline and concerns about the environment, human should be lowered to reduce hydrocarbon emis- health, and quality of life provide a compelling sions and ambient ozone concentrations, and case for making vehicular pollution control a the aromatic hydrocarbon content (inicluding top priority. VEHICULAR AIR POLLUTION: AN OVERVIEW A ir pollution is the presence of pol- (VOCs), ozone, sulfur dioxide (SO2), particu- lutants in the atmosphere from late matter (PM), and lead. In large urban cen- anthropogenic: or natural sub- ters such as Mexico City, Sao Paulo, and Santiago, stances in quantities likely to air pollutants threaten human health and qual- harm human, plant, or animal life; to damage ity of life. Because of deteriorating air quality, human-made materials and structures; to bring other urban centers in the region such as Belo about changes in weather or climate; or to in- Horizonte, Buenos Aires, Rio de Janeiro, and terfere with the enjoyment of life or property Santafe de Bogota are expected to face similar (Cooper and Alley 1986; Elsom 1987). The problems. Preventing and controlling air pollu- amount of pollutants released to the atmosphere tion in rapidly growing urban centers in the re- by fixed or mobile anthropogenic sources is gen- gion will likely be major challenge for the future. erally associated with the level of economic ac- Air pollution is worse in locations with unfa- tivity. Meteorological and topographical vorable topographical or meteorological char- conditions affect dispersion and transport of acteristics. In Sao Paulo, Santiago, Mexico City, these pollutants, which can result in ambient and Rio de Janeiro, for example, meteorologi- concentrations that may harm people, struc- cal factors such as thermal inversions restrict tures, and the environment. In general, the ef- dispersion of pollutants and result in high am- fects on people are most intense in large urban bient pollutant concentrations. Unfavorable to- centers with significant emission sources, unfa- pography (a problem in Mexico City, Santiago, vorable dispersion characteristics, and high and some other urban centers) and wind direc- population densities. tion have similar effects. Although pollutant Deterioration of air quality is a major environ- emissions in Santiago are about 16 percent of mental problem in many large urban centers in those in 5ao Paulo, the magnitude and severity both industrial and developing countries. Al- of air pollution episodes are similar. The can- though urban air quality in industrial countiies yon effect of tall buildings in downtown areas has been controlled to some extent during the also has a significant influence on ambient air past two decades, in many developing countries concentration of pollutants emitted (such as CO it is worsening and becoming a major threat to and NO.) from vehicles. The health effects of the health and welfare of people and the envi- pollutants depend on many factors, including ronment (WHO/UNEP 1992). The LatinAmeri- the number and age group of exposed people can and the Caribbean region is no exception. and their health status, ambient concentrations Air quality in many urban centers of the region and types of pollutants, and dose-response func- exceeds national, regional, or local standards tions. and World Health Organization (WHO) guide- Increasing urbanization and industrialization lines. The most common urban air pollutants in result in more energy demand, which generally the region include nitrogen oxides (NO and leads to higher emissions of air pollutants. Emis- NO2, collectively represented as NO,), carbon sions from fixed sources such as refineries, power monoxide (CO), volatile organic compounds and industrial plants, commercial and residen- 1 2 Chapter 1 tial buildings, chemical and fuel storage facili- percent (with an urban population of 315 mil- ties, and gasoline stations are the main sources lion; Table 1.1), much greater than in Eastern of air pollution in some cities of the world. But Europe (63 percent, with an urban population for most Latin American urban centers, motor of 61 million), Africa (32 percent, with an ur- vehicles are the main contributor to deteriorat- ban population of 206 million), and Asia (31 ing ambient air quality. Motor vehicles account percent, with an urban population of 974 mil- for 99 percent of total CO emissions, 54 percent lion; UN 1993). of hydrocarbons, and 70 percent of NO, in Much of the air pollution in Latin America and Mexico City; 96 percent of CO, 90 percent of the Caribbean occurs in major urban centers. In hydrocarbons, 97 percent of NO,, and 86 per- 1990 four urban centers-Sao Paulo, Mexico City, cent of SO2 in Sao Paulo; and 94 percent of CO, Buenos Aires, and Rio dejaneiro-ranked second, 83 percent of hydrocarbons, and 85 percent of fourth, eighth, and tenth among the world's ten NO,, in Santiago. Vehicular air pollution is also largest urban agglomerations (UN 1993). In 1994 becoming severe in many other Latin American forty-three of the region's cities had populations urban centers such as Santafe de Bogota. Fur- of more than one million people (Table 1.2). thermore, emissions from natural sources (such The main cause of the region's high urban- as fugitive dust) contribute to pollutant emis- ization has been an unequal land tenure system, sions in some urban centers of Latin America, which does not provide enough land for most including Santiago and Mexico City. families in rural areas to feed themselves. In addition, during the 1960s and the first half of the 1970s improvements in production led to Urbanization in Latin America economic growth that strengthened and ex- and the Caribbean panded the wage-earning middle class. The in- creasing per capita gross domestic product Latin America and the Caribbean has long been (GDP) was sufficient to generate high expecta- the most urbanized region in the developing tions among marginalized segments of the popu- world and this trend shows no sign of changing lation. Although jobs in urban centers did not (Figure 1.1). In 1950 only 42 percent of the pay well, they provided higher living standards region's population was living in urban centers; than could be obtained in rural areas. As a re- by 1990 this percentage had risen to about 72 sult there were massive population movements Figure 1.1 Population residing in urban areas, 1950-2010 Percent 100 Latini America and the Caribbeani 80- 60 _ r, _urop 40 20 - f 0 I I I 1950 1960 1970 1980 1990 2000 2010 Sotrce.: UN 1993. Ve)iicularAirPollution: An Ovenriewv 3 Table 1.1 Urbanization in Latin America and the Caribbean, 1970-94 Annual urban p5opulation growth Urbaon ppuldation Urban o)(pulalinn (percent) (thousands) (firwelt of total poptlation) 1970- 1980- 1990- Region/rountry 1970 1s 980 199( 1994 1970 1980 1990 1994 s8 90 94 South America 114,651 163,670 221,121 - 60.0 68.2 75.2 - 3.6 3.1 - Argentina 18,784 23,401 27,829 30,096 78.4 82.9 86.1 88.0 2.2 1.7 1.6 Bolivia 1,762 2,467 3,656 4,176 40.8 44.2 51.2 58.0 3.4 3.9 3.2 Brazil 53,500 80,334 112,116 122,507 55.8 66.2 75.2 77.0 4.1 3.4 2.7 Chile 7,150 9,054 11,145 12,040 75.2 81.2 84.6 86.0 2.4 2.1 1.8 Colombia 12,218 16.957 22,604 26,136 57.2 63.9 70.0 72.0 3.3 2.9 2.7 Ecuador 2,392 3,815 5,937 6,496 39.5 47.0 56.3 58.0 4.8 4.5 3.6 French Guiana 33 49 73 - 67.4 70.8 74.6 - 4.0 4.1 - Guyana 209 228 261 - 29.4 30.1 32.8 - 0.9 1.4 - Paraguay 872 1,312 2,030 2,496 37.1 41.7 47.5 52.0 4.2 4.5 4.4 Peru 7,574 11,168 15,041 16,704 57.4 64.6 69.8 72.0 4.0 3.0 2.6 Suriname 171 158 200 - 45.9 44.8 47.5 - -0.8 2.4 - Uruguay 2,306 2,484 2,751 2,880 82.1 85.2 88.9 90.0 0.7 1.0 0.9 Venezuela 7,680 12,510 17,478 19,504 72.4 83.3 90.5 92.0 5.1 3.4 2.9 Central America and Mexico 36,189 53,796 74,486 - 53.8 60.2 65.7 - 4.1 3.3 - Belize 61 72 96 - 50.9 49.4 50.5 - 1.7 2.9 - Costa Rica 687 985 1,429 1,617 39.7 43.1 47.1 49.0 3.7 3.8 3.3 El Salvador 1,414 1,880 2,296 2,520 39.4 41.5 44.4 45.0 2.9 2.3 2.7 Guatemala 1,864 2,587 3,628 4,223 35.5 37.4 39.4 41.0 3.3 3.4 4.0 Honduras 759 1,317 2,244 2,726 28.9 36.0 43.7 47.0 5.7 5.5 4.9 Mexico 29,705 44,485 61,317 66,375 59.1 66.3 72.6 75.0 4.1 3.3 2.8 Nicaragua 970 1,497 2,197 2,604 47.2 53.4 59.8 62.0 4.4 3.9 4.2 Patiama 729 973 1,279 1,404 47.6 49.7 52.9 54.0 2.9 2.8 2.7 The Caribbean 11,604 15,549 19,870 - 46.7 53.4 59.1 - 3.0 2.5 - Barbados 89 100 115 - 37.1 40.1 44.7 - 1.2 1.4 - Cuba 5,129 6,592 7,809 - 60.2 68.1 73.6 - 2.5 1.7 - Dominican Reptiblic 1,781 2,877 4,329 4,864 40.3 50.5 60.4 64.0 4.9 4.2 3.1 Haiti 893 1,269 1,855 2,170 19.8 23.7 28.6 31.0 3.6 3.9 4.0 Jamaica 776 998 1,266 1,375 41.5 46.8 52.3 55.0 2.5 2.4 2.1 Trinidad & Tobago 612 682 801 858 63.0 63.0 64.8 66.0 1.1 1.6 1.7 Other coun- tries 2,324 3,031 3,695 - 54.6 62.4 68.5 - 2.7 2.0 - Latin America andl the Caribbean 162,444 233,015 315,477 - 57.4 65.0 71.5 - 3.7 3.1 - World 1,352,143 1,752,063 2,282,367 2.520,585 36.6 39.4 43.1 45.0 2.6 2.7 2.3 - Not available. So(Urce: UN 1993; World Bank 1996. from rural areas to cities. During the 1980s, how- * Employment opportunities in urban cen- ever, urban population increased ata slower pace ters decreased partly as a result of the than during the previous decade. This slowdown recessionary effects of the debt crisis. Struc- in urbanization can be attributed to the follow- tural adjustment policies also adversely af- ing factors: fected urban employment. 4 Chapter 1 Table 1.2 Urban agglomerations with populations of more than 1 million people in Latin America and the Caribbean, 1994 C(ontry tUrlban aglomeration Polndation (thousands) Argentina Buenos Aires 12,034 C6rdoba 1,258 Rosario 1,165 Bolivia La Paz 1,143 Brazil Belem 1,109 Belo Horizonte 4,199 Brasilia 3,201 Campinas 2,119 Curitiba 2,422 Fortaleza 2,378 Goiania 2,141 Manaus 1,637 Porto Alegre 3,569 Recife 2,614 Rio de Janeiro 11,467 Salvador 2,758 Santos 1,333 Sao Paulo 20,113 Chile Santiago 5,249 Colombia Barranquilla 1,113 Cali 1,724 Medellifn 1,710 Santafe de Bogota 5,452 Costa Rica SanJose 1,186 Cuba Havana 2,219 Dominican Republic Santo Domingo 2,510 Ecuador Guayaquil 1,947 Quito 1,691 Guatemala Guatemala City 1,676 Haiti Port-au-Prince 1,217 Mexico Guadalajara 3,088 Guadalupe 1,498 Mexico City 15,453 Monterrey 2,738 Naucalpan 1,693 Puebla de Zaragoza 1,182 Nicaragua Managua 1,092 Paraguay Asunci6n 1,022 Peru Lima 7,266 Uruguay Montevideo 1,318 Venezuela Caracas 2,924 Maracaibo 1,551 Valericia 1,211 Source: UN 1993 and 1996. * With fewer children born in rural areas, Based on urbanization data for 1994 (see Table migration to urban centers decreased. In 1.1), the region can be divided into four groups. addition, birth rates in urban centers fell The first group includes countries with an ur- at an even faster rate. banization level greater than 85 percent-Argen- * Some firms moved out of crowded, expen- tina, Chile, Uruguay, and Venezuela. These sive urban centers (such as Sao Paulo) to South American countries have had the highest nearby small towns, forming polycentric urbanization levels in the region since 1970. Al- urban areas (Oxford Analytica, 6June 1996). though the pace of urbanization in Argentina, VehicularAirPollution: An Ov7erviewv 5 Chile, and Uruguay was lower than the regional Quito), and one in Bolivia (La Paz), Dominican average during 1970-80 and 1980-90, in Ven- Republic (Santo Domingo), Nicaragua (Man- ezuela it was higher, mainly because of heavier agua), and Paraguay (Asunci6n) have popula- migration from rural areas. Within this country tions greater than 1 million. Santa Cruz in Bolivia group three urban centers in Argentina (Buenos and Santiago de Caballero in Dominican Repub- Aires, C6rdoba, and Rosario), three in Venezu- lic are within the 500,000 to 1 million popula- ela (Caracas, Maracaibo, and Valencia), and one tion range. each in Chile (Santiago) and Uruguay The fourth group includes countries with ur- (Montevideo) have populations exceeding I banization levels below 50 percent-Barbados, million. La Plata, Mar del Plata, Mendoza, and Costa Rica, El Salvador, Guatemala, Guyana, San Miguel de Tucuman in Argentina and Haiti, Honduras, Suriname. Among them, Haiti Barquisimeto, Ciudad Guyana, and Maracay in and Honduras have much higher urban growth Venezuela have populations between 500,000 rates mainly because of migration from rural and 1 million. areas. Guatemala City in Guatemala, SanJose in The second group includes countries with an Costa Rica, and Port-au-Prince in Haiti have urbanization level between 70 and 85 percent- populations larger than I million. San Pedro Sula Brazil, Colombia, Cuba, French Guiana, Mexico, and Tegucigalpa in Honduras have populations and Perm. In 1990 these six countries accounted between 500,000 and 1 million. for about 70 percent of the region's population. Among them, French Guiana historically had high levels of urbanization. During 1970-80 and Motorization in Latin America 1980-90 this group's urban population growth and the Caribbean rate was similar to the regional average, except for French Guiana (where it was higher) and An overview. The motor vehicle fleet in Latin Cuba (where it was lower because urbanization America and the Caribbean increased from 10.6 occurred earlier; Villa 1992). Within the second million vehicles in 1970 to 25.2 million in 1980 group, fourteen urban centers in Brazil (Belem, and 37.1 million in 1990 (Table 1.3). In 1994 Belo Horizonte, Brasilia, Campinas, Curitiba, the fleet reached 42.8 million vehicles and con- Fortaleza, Goiania, Manaus, Porto Alegre, Recife, sisted of 31.8 million cars and 11.0 million trucks Rio dejaneiro, Salvador, Santos, Sao Paulo), six and buses (AAMA 1996). The region's share of in Mexico (Guadalajara, Guadalupe, Mexico the global fleet in 1994 was only 6.8 percent over- City, Monterrey, Naucalpan, and Puebla de all; 6.6 percent for cars, and 7.3 percent for Zaragoza), four in Colombia (Barranquilla, Cali, trucks and buses (Figure 1.2). Cars constituted Medellin, and Santafe de Bogota), and one each between 70 and 90 percent of the national mo- in Cuba (Havana) and Peru (Lima-Callao) have tor vehicle fleets, a distribution similar to that populations exceeding 1 million.Jaboata,o,Joao in the United States (75 percent) and Canada Pessoa, Macei6, Natal, Sao Luis, Teresina, and (79 percent). Argentina, Brazil, Mexico, and Mossor6 in Brazil; Cartagena in Colombia; Chi- Venezuela together accounted for 84 percent of huahua, CiudadJuirez, Le6n, Merida, Mexicali, the cars and 78 percent of the truLcks and buses San Luis Potasi, Tampico, Tijuana, and Torreon in the region (AAMA 1996). in Mexico; and Arequipa and Trujillo in Peru Latin America and the Caribbean has higher have populations between 500,000 and 1 million. motorization levels than other developing re- The third group includes countries with mod- gions except Eastern Europe.' For example, the erate urbanization levels (between 50 and 69 region's motorization level is about three times percent). Within this group three countries- that of Asia (excluding Japan) and four times Dominican Republic, Nicaragua, and Trinidad that of Africa. Motorization in the region in- and Tobago-had urbanization levels greater creased from 38 vehicles per 1,000 people in than 60 percent in 1994. Among them, Nicara- 1970 to 70 vehicles per 1,000 people in 1980, gua had the highest and Trinidad and Tobago and 84 vehicles per 1,000 people in 1990 (see the lowest annual rates of urban population Table 1.3). In 1994 motorization in the region growth. Belize, Bolivia, Ecuador, Jamaica, Panama, and Paraguay were below the 60 per- cent urbanization level. Within the third group 1. Motorizationi is defined as the number of motor ve- two urban centers in Ecuador (Guayaquil and hicles per thtousand people. 6 Chapter 1 Table 1.3 Motor vehicle fleets and motorization in Latin America and the Caribbean, 1970-90 Motorization Annurial growth raite of Utehicles in cihrculation (vehicles in firtlaltion vehicles in cirrulation (thousandsi per 1,000 e)eonld) (pierrent) Region/rountry 1970 198( 199(9 1970 1980 1990 1970-80 1980-90) South America 8,098 18,738 25,028 42 78 85 8.8 2.9 Argentina 2,357 4.176 5,786 98 148 179 5.9 3.3 Bolivia 5 64 317 1 12 44 2.9 17.4 Brazil 3,540 10,160 13,063 37 84 87 11.1 2.5 Chile 326 588 1,017 34 53 77 6.1 5.6 Colombia 295 753 1,381 14 28 42 9.8 6.3 Ecuador 91 215 240 15 26 23 8.9 1.1 French Guiana 10 - - 204 - - - - Paraguay 28 67 110 12 21 26 9.1 5.1 Peru 359 489 625 27 28 27 3.1 2.5 Suriname 24 32 48 65 91 114 2.9 4.1 Uruguay 209 261 258 74 89 83 2.2 -0.1 Venezuela 854 1.933 2,183 81 128 113 8.5 1.2 Central America and Mexico 2,133 5,773 10,958 32 65 97 10.6 6.6 Belize - 12 5 - 82 26 - -8.4 Costa Rica 58 175 250 34 77 82 11.7 3.6 El Salvador 64 140 160 18 31 30 8.1 1.3 Guatemala 70 334 230 13 48 25 16.9 -3.7 Honduras 36 71 133 14 19 26 7.0 6.5 Mexico 1,792 4,847 9,882 36 72 117 10.5 7.4 Nicaragua 53 68 74 26 24 20 2.4 0.9 Panama 60 126 224 39 64 93 7.6 5.9 The Caribbean 410 665 1098 20 27 39 5.0 5.1 Barbados 23 29 45 96 116 175 2.3 4.5 Cuba - - - - - - - - Dominica 3 - 4 43 - 56 - - Dominican Rep. 60 145 270 14 25 38 9.2 6.4 Haiti 20 37 55 4 7 8 6.4 4 Jamaica 63 133 112 34 63 46 7.8 -1.7 Trinidad & Tobago 90 1l68 342 93 155 276 6.4 7.4 Other countries 151 153 270 35 32 50 0.1 5.8 Latin America and the Caribbean 10,641 25,176 37,084 38 70 84 9.0 4.0 W'orld 246,368 411,0(76 582,982 67 92 110 5.3 3.6 - Not available. a. Excltudes two- and three-wheelers. Sourre: MVMA 1971, 1982, and 1992 for number of vehicles; UN 1993 for population data. reached 89 vehicles per 1,000 people (AAMA tained 58 percent of the motor vehicle fleet in 1996). Chile, the Buenos Aires Metropolitan Area con- The overwhelming concentration ofrmotorve- tained 51 percent of the fleet in Argentina, and hicles in urban centers is an important contrib- the Santafe de Bogota Metropolitan Area con- uting factor to the urban air pollution problem tained 28 percent of the fleet in Colombia. The in Latin America (Table 1.4). In 1994, for ex- three largest metropolitan areas in Brazil (Sao ample, the Santiago Metropolitan Area con- Paulo, Rio dejaneiro, and Belo Horizonte) col- VehicularAirPollution: An Oveviewt 7 Figure 1.2 Global distribution of the motor vehicle fleet, 1994 Percent 100 90 - Cars 80 78.2 Buses and truecks '70 66.1 60 50 40 30 20 -13.6 10 5.7 '7.7 6.6C '73 376.2 2.1 20 F7 _7 _ 3 OECD Eastern Europe Latin America Africa Asia and the Caribbean Note: The percentages do not add to 100 because not all countries are included (e.g. Turkey, Greece).Japan is included in the OECD country group rather than in Asia. .Sorce: AAIA 1996. lectively had 45 percent of the national motor ceeded the national level by 98 percent in vehicle fleet (31, 10, and 4 percent, respectivelv). Santafe de Bogota, 68 percent in Santiago, 56 Motorization levels in these urban areas are percent in Belo Horizonte, 54 percent in Rio de much higher than the respective national levels Janeiro, 43 percent in Buenos Aires, and 27 per- (see Table 1.4). The 1994 motorization levels in cent in Mexico City. This concentration is even the Sao Paulo Metropolitan Area was 2.9 times greater in the urban core of metropolitan areas. higher than the national average in Brazil. The Along with rising incomes in highly urbanized motorization level in metropolitan areas ex- areas of the region, the vehicle fleet has grown Table 1.4 Motor vehicle fleets and motorization in selected Latin American urban areas, 1994 Motor vehirnefleet Motorization Uirban share At urban areas Total of national (molayr vehile fieet Urbanl Urban area (million) fleet (%) peor thousand 1e5ple) Nattional Mexico City 3.00 24 177 1.27 Santiago 0.64 58 133 1.68 Sao Paulo 4.70 31 277 2.89 Rio deJaneiro 1.55 10 148 1.54 Belo Horizonte 0.62 4 150 1.56 Buenos Aires 2.92 51 237 1.43 Santafe de Bogota 0.46 28 89 1.98 Note: Excludes two- and three-wheelers. Sourcrf Auithors' estimates based on data from Chapters I an(i 4. 8 Chapter I at a rapid pace. To respond to increasing de- by 137 percent (at an annual average of 9.0 per- mand for motor vehicles, some counitries in the cent), while the global motor vehicle fleet grew region have developed and expanded domestic by only 67 percent (at an annual average of 5.3 automotive industries. Between 1990 and 1993 percent). The increase in motorization was 84 the number of vehicles assembled and manu- percent in the region and 37 percent in the factured in the region increased by 55 percent world (see Table 1.3). During the 1970s growth (from 1.949 million vehicles in 1990 to 3.014 in per capita incomes was fairly high for most million vehicles in 1993; Piquini 1995). Trade countries in the region (Table 1.5). This growth liberalization policies, international trade agree- largely resulted from a postwar economic boom ments, and national trade pacts also have pro- (1950-73) in the United States, Europe, and moted lower prices and increased sales of new Japan. But even when economic growth in the motor vehicles. But while the number of new OECD countries faltered after the first oil shock vehicles (with lower emissions) is increasing, (1973-74),mostLatin American and Caribbean most older, poorly maintained, and more pol- countries continued to grow at a fast pace. The luting vehicles have remained in use and the av- oil shock had a limited impact because the re- erage age of national motor vehicle fleets is still gion was experiencing a substantial trade expan- high. The average age of cars is thirteen years in sion and commodities boom. Brazil, which had Costa Rica, twenty years in Venezuela, and liberalizedits economyin the 1960s, avoided the twenty-three years in Paraguay. Motor vehicles full impact of the oil shock through massive bor- older than ten years old make up 50 percent of rowing in the 1970s. Oil-exporting countries like the fleet in Argentina, 60 percent in Ecuador, Ecuador, Mexico, and Venezuela benefited from and 64 percent in El Salvador. In Lima about 75 the windfall. The region was also able to main- percent of vehicles are more than ten years old, tain a fairly high level of investment and imports and in Mexico City the average vehicle age is 8.5 from industrial countries (Teitel 1992). years. These older cars are responsible for an Between 1980 and 1990 the motor vehicle fleet inordinately high share of motor vehicle emis- in Latin America and the Caribbean increased sions and resulting deterioration of ambient air by 47 percent (at an annual average of 4.0 per- quality in urban areas. cent), while the global motor vehicle fleet grew Despite the growth of the vehicle fleet, trans- by 42 percent (at an annual average of 3.6 per- port infrastructure and services in Latin America cent). The region's growth in motorization was and the Caribbean often have not been im- the same as the global increase (20 percent), proved or adequately maintained. As a result although because of unfavorable economic con- many urban centers in the region are plagued ditions it was considerably lower than in the by congestion and by a subsequent worsening 1970s. In the 1980s, for a number of reasons, of air quality from vehicular emissions. This prob- the rapid economic growth of the previous de- lem has not been adequately addressed for vari- cades was not sufficient to ensure the long-term ous reasons, including weaknesses in legal and sustainability of the region's economy. First, ex- institutional frameworks, administrative bottle- cessive protectionism and government control necks, insufficient financial resources, lack of air created a rigid economic structure unable to quality management planning, and inadequate respond to global changes. Second, the region's implementation of policies for air pollution inward-lookingdevelopmentstrategygreatlydis- control. couraged exports. Third, many countries be- came incapable of providing efficient social Motor vehicle oumership. Motor vehicle ownership services because of the increasing burden of in- is affected by various factors including econom- efficient public sectors and tax systems. Finally, ics, culture, and geography. Economic factors, unlike the first oil shock, the shock of 1980 had especially per capita income, appear to be the strong adverse effects on the region. An increase driving force behind the growth of motor vehicle in interest rates, caused by the introduction of ownership in Latin America and the Caribbean floating interest rates and deregulation of the as well as elsewhere (Faiz and others 1990; Faiz, banking system in the United States, exposed Gautam, and Burki 1995; O'Brian and manycountriesintheregiontohigh-costfinanc- Karmokolias 1994; Stares and Zhi 1997). ing. The oil shock also reduced the volume of Between 1970 and 1980 the motor vehicle fleet international trade and led to a fall in commod- in Latin America and the Caribbean increased ity prices. Instead of adjusting to this changing VehicularAirPollution: An Overview 9 Table 1.5 Per capita income in Latin America and the Caribbean, 1970-94 (GNI'per rariilfa Annual growvth rite of GNP (U1.S. eollat rsr tapil (per enl) R?gion/rounthy 1970 1980 1990 1994 1970-80 1980-90 1990-94 South America Argentina 1,160 2,590 2,370 8,110 0.7 -1.8 6.9 Bolivia 180 570 630 770 1.9 -2.6 1.9 Brazil 420 2,160 2,680 2,970 5.9 0.6 0.3 Chile 720 2,290 1,940 3,520 0.5 1.1 5.9 Colombia 340 1,260 1,260 1,670 4.0 1.1 3.1 Ecuador 290 1,100 980 1,280 5.3 -0.8 1.6 French Guiana 940 200 - - -0.4 - - Guyana 370 690 330 530 6.4 -5.8 10.5 Paraguay 260 1,410 1,110 1,580 5.9 -1.3 -0.5 Peru 450 1,080 1,160 2,110 0.2 -2.0 3.2 Suriname 530 2,770 3,050 860 6.8 -5.0 0.6 Uruguay 820 2,620 2,560 4,660 3.2 -0.9 3.9 Venezuela 980 3,910 2,560 2,760 2.2 -2.0 0.6 Central America and Mexico Belize - 980 1,990 2,530 - 2.5 1.1 Costa Rica 560 1,390 1,900 2,400 2.6 0.6 - El Salvador 300 670 1,110 1,360 1.3 -0.6 4.3 Guatemala 360 1,080 900 1,200 2.8 -2.1 -0.2 Honduras 280 560 590 600 0.5 -1.2 0.7 Mexico 670 1,980 2,490 4,180 3.1 -0.9 0.7 Nicaragua 430 760 420 340 -2.9 - - Panama 730 1,730 1,830 2,580 1.2 -2.0 5.9 The Caribbean Barbados 570 3,270 6,540 6,560 3.2 1.4 -1.1 Cuba 530 - - - - - - Dominica 280 640 2,220 2,800 -3.1 3.0 2.0 Dominican Republic. 350 1,190 830 1,330 3.3 -0.4 2.9 Haiti 110 280 370 230 1.8 -2.3 -9.6 Jamaica 670 1,090 1,500 1,540 -2.8 -0.4 3.5 Trinidad & Tobago 860 5,010 3,610 3,740 3.9 -6.0 -0.7 - Not available. a. At current prices. b. At constant prices. Sourre: World Bank 1972, 1983, 1992, and 1996. situation, the region continued to borrow tn- Argentina, Bolivia, Ecuador, Mexico, Pert, and der unfavorable conditions to finance high gov- Venezuela (Tables 1.3 and 1.5). ernment deficits that led to spiraling inflation, Since the mid-1980s, however, economic re- which affected income distribution. The high forms in the region have been impressive. These inflation also resulted in reduced savings and reforms have included implementation of ma- investments, which hampered economic growth jor stabilization programs, opening up of econo- (Edwards, 1995). Despite the decline in per mies to international competition, and capita income during the 1980s, inflationary con- privatization of a large number of state-owned ditions led to the emergence of upper-income enterprises. The decline in per capita income groups whose vehicle acquisitions helped in- during 1982-86 was subsequently recovered in crease motorization levels in countries such as many countries. Most countries, especially the 10 Chapter 1 advanced reformers, experienced respectable vehicles in 1993), and Argentina (0.34 million growth during the early 1990s. After accelerat- vehicles in 1993). Smaller numbers of motor ve- ing in the late 1980s, inflation has declined sub- hicles are produced in Chile, Colombia, Peru, stantially throughout the region. The economic Umuguay, and Venezuela. Although most motor recovery was also assisted by the massive inflow vehicle manufacturers in the region serve their of foreign capital in the beginning of 1991, domestic markets, Mexico exports 44 percent of coupled with expansion of exports and rising its production and Brazil exports 24 percent. productivity in the manufacturing sector However, Brazil sells 88 percent of its exports (Edwards 1995). Despite this progress, several within the region (70 percent to Argentina problems persist a decade after the debt crisis. alone), while Mexico sell only 9 percent of its These problems include severe deterioration of exports within Latin America and the Caribbean the physical infrastructure, increases in poverty, (90 percent of its exports go to the United States and double-digit inflation in manv countries, and Canada; AAMA 1995; Piquini 1995). In addition, because economic reforms were From its establishment in the late 1950s not accompanied by the modernization of po- through its development in the following two litical institutions, political tensions (and in decades, the automotive industry in the region some cases civil unrest) have developed in such was marked by strong government involvement. countries as Brazil, Guatemala, Haiti, Penm, and Local industries emerged and flourished thanks Venezuela. to government import substitution policies, Cross-sectional regression analyses were car- which were politically attractive because they ried out on per capita income and motorization generated local employment opportunities and levels for the same set of twenty-one countries independence from international markets. Lo- of the region for 1970, 1980, and 1990. The re- cal industries were protected by both tariff and sults show a strong correlation between these two nontariff barriers, which in some cases banned variables, confirming that per capita income is a importation of foreign-made vehicles, leaving significant factor for motorization (Figure 1.3).2 participation in local production as the only The income elasticity coefficients for the three option available to foreign vehicle manufactur- years were 1.59, 1.09, and 1.18, respectively (im- ers. Governments also provided large tax incen- plying that a 1 percent increase in incomes would tives to local automotive industries. As a result lead to 1.59, 1.09, and 1.18 percent increases in of these protections and incentives, local auto- vehicle ownership). These elasticity coefficients motive industries thrived in Argentina, Brazil, are similar to those found in other studies (rang- Chile, Colombia, Ecuador, Mexico, Uruguay, and ing from 1.02 to 1.59) conducted for various Venezuela. In 1980 total production for these years and groups of countries during 1968-90 eight countries reached a record 2.2 million. (Stares and Zhi 1997). Other Latin American and Caribbean countries with smaller markets and lower per capita in- log (ntimber of vehicles in circulation/l,000 people) ,,, comes did not develop their own industries but -1.97 + (1.18 x log GNP/capita) allowed vehicle imports (Sanchez 1992). The log (number of vehicles in circulationi/l,000 people)1,, emergence of the debt crisis in 1981 caused the -1.78 + (1.09 x log GNP/capita) region's economies to spiral downward, however. log (miimber of vehicles in circiliatioi/ 1,000 people) 97 = Vehicle production dropped by 22 percent in -2.85 + (1.59 x log GNP/capita) just one year (to 1.8 million in 1981) and re- mained below the 2 million mark until the end Motor vehicle supply. The great majority of new of the decade (Piquini 1991). vehicles sold in Latin America and the Carib- With economic liberalization, most countries bean are produced within the region, though a in the region began to move away from the im- small percentage are imported from outside, port substitution, state-controlled development mostly from OECD countries. The main motor model. In Chile this process was initiated in 1975 vehicle producers in the region are Brazil (1.39 and followed by privatization and deregulation million vehicles in 1993), Mexico (1.08 million after the debt crisis of 1982. Mexico in 1988, Argentina in 1989, Venezuela in 1990, and Co- lombia in 1991 took decisive steps to liberalize 2. The r2 and t statistics were 0.82 and 9.2 for 1970, 0.81 their economies. In 1993 vehicle sales in the re- and 9.1 for 1980, and 0.82 and 9.2 for 1990. gion increased to 2.4 million (59 percent more Veh'wular Air Pollution: An Overview 11 Figure 1.3 Per capita income and motorization in Latin America and the Caribbean, 1970, 1980, and 1990 Vehicles per 1,000 people 10 . ~~~~~~ 70 ..... 100 .... 1,000.......10,000...... ............... ..........per .....pi ... (....5). .. ....... . ... Vehicles . .. per 1,00 people.. .......... ........ .. ...... . . . . ......... ....... 1,000 -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.... .Z~.. ... ...... .. . 10~~~~~~~~~~~~~~~~~~~MC *.q *5OI 100...... ... .. .00 .....0...00.....0.. ...... AN e apt 5 Vehicles per 1,000 people.... 1.000.~~~~~~~~~~~~~~~~~~~~~~~~" 100 1.000 10,000 GNP per capita() year) icolpe cpita inc0popleda. 12 Chapter I than the 1990 level and 13 percent more than 1990, foreign companies were permitted to have the 1980 level) and vehicle production increased up to 100 percent ownership under certain con- to 3.0 million (55 percent more than the 1990 ditions. Trade liberalization reduced the licens- level and 38 percent more than the 1980 level). ing requirements for importation and tariffs Although Latin American markets are now open (Sanchez 1992). NAFTA has further liberalized to imports, they are not totally free of restric- Mexico's trade policies. tions. For example, in Mexico import of vehicles The Argentine government's economic policy is restricted to vehicle manufacturers. InArgen- of 1989 included price stabilization through tina there are import quotas and taxes. Brazil's dollarization of the economy, tax reform, and market is free of quotas, but a 35 percent tax is privatization. In 1991 a sectoral pact cut the price charged for all imported vehicles. Import taxes of vehicles by about 30 percent. As a result ve- also exist in Colombia and Venezuela (Piquini hicle production increased fourfold in three 1995). years. Brazil's domestic automotive markets, Latin American countries have also entered which had been closed to external competition, into trading blocs. In 1993 Mexico signed the opened in 1990, and imported vehicles ac- North American Free Trade Agreement counted for 8 percent of sales in 1993. Because (NAFTA) with the United States and Canada, of competition from foreign imports, vehicle allowing expansion and modernization of its prices dropped by 25 percent (in dollar terms) automotive industry as well as integration with between 1992 and 1994 (Piquini 1995). In tradi- the U.S. industry. That same year Colombia, tionally closed economies like Colombia and Ecuador, and Venezuela-members of the Venezuela, imported vehicles account for about Andean Pact that also includes Bolivia. Chile, 40 percent of the market (O'Brian and and Peru-formed a trading bloc by agreeing Karmokolias 1994). Since the early 1990s growth on a common automotive policy. Argentina, Bra- in the region's automotive industry has been so zil, Paraguay, and Umiguay forned the South- spectacular that companies stnrggling with over- ern Cone Common Market (Mercosur), which capacityjust a few years before are now striving became effective onJanuary 1, 1995. That agree- to find extra capacity. In 1994 car manufactur- ment established free circulation of goods, ser- ers were planning to invest $3.1 billion in Ar- vices, financial resources, and workers; gentina, Brazil, and Mexico (Piquini 1995). eliminated nontrade barriers; set a single tariff for goods; and implemented similar The Outlook macroeconomic policies in several areas, includ- ing transport. Through agreements signed in The urban growth rate in Latin America and the 1986 and 1990 and followed by Mercosur, Ar- Caribbean is expected to slow as a result of de- gentina and Brazil have almost integrated their clining overall population growth, less migration automotive industries (Piquini 1995). Thus the from niral areas, and the lower birth rates of region's previously closed automotive industry the earlier migrants. Still, by 2010 about 80 per- has become subject to intense competition and cent of the region's population is expected to a demand for higher-quality vehicles. live in urban areas (see Figure 1.1; UN 1993). Mexico is a good example of successful eco- Moreover, nine urban areas in addition to the nomic restructuring. The Mexican program fo- forty-three listed in Table 1.2 will have popula- cused on four interrelated issues: price tions greater than 1 million: MendozainArgen- stabilization, privatization, domestic deregula- tina; Santa Cruz in Bolivia; Sao Luis in Brazil; tion, and trade liberalization (Sanchez 1992). Santiago de Caballero in Dominican Republic; Prices were stabilized through social pacts and Tegucigalpa in Honduras; Ciudad Juarez and the public deficit was controlled through tax re- Le6n in Mexico; and Barquisimeto and Maracay form, major cuts in public sector investment, in Venezuela. elimination of subsidies (including on gasoline), As a result of economic growth and trade lib- and an aggressive privatization progr-am. Deregu- eralization, vehicle sales in Latin America and lation lifted requirements on the number of ve- the Caribbean are expected to grow consider- hicle models produced by each company in ably to reach 4.5 million a year in 2000 (about Mexico, specific vehicle components produced three times more than in 1990). This will con- in Mexico, and operation of the Mexican truck- sist of about 3.3 million passenger cars, 900,000 ing industry. With the foreign investment law of light-duty commercial vehicles, and 300,000 VehicularAirPollution: An Ovierviewv 13 heavy-duty commercial vehicles (Piquini 1995). O'Brian, Peter and Yannis Karmokolias. 1994. "Radi- Most of these vehicles, especially cars, will en- cal Reform in the Automotive Industry." IFC ter the urban fleet. Far fewer vehicles will be Discussion Paper 21. International Finance Corpo- retired from the urban fleet than added, resuilt- ration, Washington, D.C. ing in a much greater number of vehicles on Oxford Analytica. 1996. "Latin America: Urban Ex- the road, more congestion, and more fuel con- pansion." June 6, 1996, The United Kingdom. sumption. Unless adequate air pollution con- Piquini, Marco. 1991. The Motor Industries of South trol measures are adopted soon, urban air America and Mexico: Opportunity and Challengesfor the quality will worsen considerably causing adverse Future. E.I.U. Special Report 2096. Economic Intel- health effects not only for highly vulnerable ligence Unit, London. population groups (children, the sick, and the . 1995. The Motor Industries of South America and old) but for the general population as well. Mexico: Poised for Grotth. E.I.U. Research Report. Economic Intelligence Unit, London. Sanchez, Enrique P. 1992. "The Auto Industry in Latin References America: The Challenge of Adjusting to Economic Reform." Business Economics 27:25-30. AAMA (American Automobile Manufacturers Asso- Stares, S. and L. Zhi (eds.). 1997. Cfiina's Urban Trans- ciation). 1995. World Motor Vehicle Data. Washing- port Development Strategy: Proceedings of a Symposium ton, D.C. in Beijing, November 8-10, 1995. World Bank Discus- . 1996. World Motor Vehicle Data. Washington, sion Paper No. 352, Washington, D.C. D.C. Teitel, S. 1992. Towards a New Development Strategy for Cooper, C.D. and EC. Alley. 1986. Air Pollution Con- Latin America: Pathways from Hirschman's Thought. trol: A Design Approval. Prospect Heights, Waveland Inter-American Development Bank, Washington, Press, Illinois. D.C. Edwards, Sebastian. 1995. Crisis and Reform in Latin UN (United Nations). 1993. World Urbanization Pros- America: From Despair to Hope. Oxford University pects. New York. Press, New York. . 1996. Demographic Yearbook, 46th Issue. Depart- Elsom, D. 1987. Atmospheric Pollution: Causes, Effects, ment for Economic and Social Information and and Control Policies. Basil Blackwell, Oxford, U.K. Policy Analysis, NewYork. Faiz, Asif, K. Sinha, M. Walsh, and A. Varma. 1990. Villa, Miguel. 1992. "Urbanizaci6n y Transici6n "Automotive Air Pollution: Issues and Options for Demografica en America Latina: Una Resefia Del Developing Countries." Policy Research Working Perfodo 1930-1990." Conference proceedings of Paper 492, U'orld Bank, Washington D.C. The Peopling of the Americas, International Union for Faiz, Asif, Surhid Gautam, and Emaad Burki. 1995. the Scientific Study of Population, Veracruz, Mexico. "Air Pollution from Motor Vehicles: Issues antic WHO/UNEP (World Health Organization/United Options from Latin American Countries." The Sci- Nations Environment Programme). 1992. Urban Air ence of the Total Environment, 169:303-10. Pollution in Megacities of the World Basil Blackwell, MVMA (Motor Vehicle Manufacturer'sAssociation). Oxford, U.K 1971. World Motor Vehicle Data. Washington, D.C. World Bank. 1972. World BankAtlas. Washington, D.C. - 1982. World Motor Vehicle Data. Washingtotn, . 1983. World Bank Atlas. Washington, D.C. D.C. . 1992. World Bank Atlas. Washington, D.C. 1992. World Motor Vehicle Data, Washi ngton. . 1996. From Plan to Market, World Development D.C. Report, The World Bank, Washington, D.C. AIR POLLUTANTS AND THEIR EFFECTS A ir pollutants can be grouped in two cat- Pollutant Emissions from Motor Vehicles egories: p'imaiy, if emitted directly into the atmosphere, and secondary, if Motor vehicle emissions result from fuel com- formed in the atmosphere as a result bustion or evaporation. The most common types of chemical reactions (such as hydrolysis, oxida- of transport fuels are gasoline (in leaded or un- tion, or photochemical reactions) that involve leaded form) for light-duty vehicles (such as primary pollutants. Primary pollutants emitted cars) and diesel fuel for heavy-duty vehicles (such by motor vehicles include carbon dioxide (CO,), as buses and trucks). Other commercial fuels carbon monoxide (CO), hydrocarbon com- used in light-duty vehicles include alcohols (such pounds (HC), sulfur dioxide (SO2), nitrogen as ethanol and methanol), gasoline-alcohol mix- oxides (NO.), particulate matter (PM), and lead. tures, compressed natural gas (CNG), and liq- Secondary pollutants associated with motor ve- uefied petroleum gas (LPG). For heavy-duty hicle emissions include nitrogen dioxide (NO,), vehicles other commercially available fuels in- photochemical oxidants (for example, ozone), clude gasoline, CNG, and LPG. and sulfuric or nitric acids and their salts (that Emissions from motor vehicles with spark- is, sulfate and nitrate aerosols). NO2 is formed ignition engines (for example, gasoline-fueled through oxidation in the air of nitric oxide vehicles) are from the exhaust, engine crank- (NO), a gaseous pollutant formed at high com- case, and fuel system (carburetor, fuel line, and bustion temperatures and emitted by motor ve- fuel tank). CO2 and watervapor (H20), the main hicles. Ozone (03) is formed from NO, and products of combustion, are emitted in vehicle reactive HC in the presence of sunlight. SO2 and exhaust. The major pollutants emitted from NO. can react with atmospheric moisture, oxy- gasoline-fueled vehicles are CO, HC, NO., and gen, and PM to form sulfuric or nitric acid or lead (only for leaded gasoline fuel). In addition, their salts. SO2 may also be present in exhaust gases. The This chapter provides an overview of the air conditioning system, tires, brakes, and other sources, properties, and emission characteristics vehicle components also produce emissions. of vehicular air pollutants. Factors that affect For a given fuel quality, concentrations of dispersion of these pollutants, such as meteo- many of these pollutants are influenced by such rological and topographical factors, are then factors as the air-fuel ratio in the cylinder at the reviewed. A discussion of health and environ- time of combustion, ignition timing, combustion mental effects and some quantification of effects chamber geometry, engine parameters (for ex- on human health is then presented. Finally, the ample, speed, load, and engine temperature), ambient air quality standards adopted by differ- and use of emission control devices. Vehicles with ent Latin American countries are provided and electronic fuel injection engines electronically the design options for ambient air quality moni- maintain an air-fuel ratio of about 14.7:1 (that toring systems are discussed. is, 14.7 grams of air per gram of gasoline, which 15 16 Chapter 2 is the stoichiometric ratio for the air-gasoline changes. Losses from the carburetor, called hot mixture) to achieve complete combustion. soak emissions, occur when a hot engine is Higher ratios ("lean" mixtures) produce less HG stopped. Evaporative emissions from the carbu- and CO emissions, while lower ratios ("rich" retor have been greatly alleviated with the ad- mixtures) produce more CO and HC emissions vent of electronic fuel injection engines that from unburned or partially burned ftiel. Igni- maintain fuel under pressure and prevent its tion timing also affects the combustion process. escape from the system. Hot soak and diurnal The air-fuel ratio and ignition timing are readily emissions are controlled by on-board activated adjustable, both in design specifications and field carbon canisters. Evaporative emissions also oc- tune-up adjustments (Bellomo and Liff 1984). cur during refueling through displacement and Light-duty gasoline-fueled vehicles not equipped spillage and account for a large proportion of with pollution control devices have the highest HC emissions. exhaust emissions during acceleration, followed Crankcase blow-by (also called running loss by deceleration, cruising, and idling cycles (Table emissions) are unburned or partially burned fuel 2.1). Thus frequent cycle changes, as required components that, under pressure, escape from by stop-and-go traffic patterns in congested ur- the combustion chamber, pass the pistons, and ban areas, increase pollutant emissions. At enter the crankcase. In older model-yearvehicles higher cruising speeds HC and CO emissions these emissions were vented to the atmosphere. decrease, while NO, and CO, emissions increase. In newer model-years they are controlled by re- Three-way catalytic converters installed on gaso- cycling to the engine through the intake system line-fueled vehicles can reduce CO and HC (Faiz, Weaver, and Walsh 1996). emissions by about 90 percent and NO, emis- The pollutants from diesel-fueled vehicles are sions by 70 percent from uncontrolled levels (see PM (including smoke), NO., SO2, CO, and HC. Chapter 3). Most of these pollutants are emitted from the Evaporative emissions are HC vapors lost di- exhaust. Because diesel engines operate at high rectly to the atmosphere, mainly from the fuel air-fuel ratios (about 30:1), they tend to have tank and carburetor. Fuel tank losses consist pri- low HC and CO emissions. They have consider- marily of the more volatile fractions of fuel dis- ably higher PM emissions than gasoline-fueled placed from the vapor space above the liquid vehicles, however. For heavy-duty vehicles, CO, fuel in the fuel tank and mainly occur as a result HC, and NO., emissions in the exhaust also vary of a temperature change in the fuel tank; they with driving modes, engine speed, and load. may also be caused by diurnal temperature Table 2.2 shows the effects of different driving Table 2.1 Exhaust emissions from uncontrolled light-duty gasoline-fueled vehicles at different driving modes (parts per million) Afode CO HC NO, (cO2 Idling 16 1.3 0.1 68 Accelerating 0-15 mph 2,997 536 62 10,928 0-30 mph 3.773 757 212 19,118 Cruising 15 mph 67 5.1 0.8 374 30 mph 30 3.0 2.0 323 45 mph 28 2.9 4.2 355 60 mph 29 2.9 6.4 402 Decelerating 15-0 mph 1,902 344 21 5,241 30-0mph 1,390 353 41 6,111 mpb = Mile per hotur. Sowurre: Bellomo and Liff 1984. Air Pollutants and Their Effects 17 Table 2.2 Exhaust eniissions from uncontrolled heavy-duty vehicles at different driving modes (parts per million) (;asolineilf1iied vehicle Diesel-fueled vehird Aode C() H(; NOy (,0 HOC NOx Idling 69,000 5.300 30 Trace 400 60 Accelerating 29,000 1,600 1,020 1,000 200 350 Cruising 27,000 1,0(0 650 Trace 100 240 Decelerating 39,000 10,0(0 20 Trace 300 30 Source: Faiz and others 1994. modes on pollutant emissions from uncontrolled oxygen affect CO formation. Conversion of CO heavy-duty vehicles. Two-way catalytic converters to CO2 in the atmosphere is slow and takes two used in diesel-fueled vehicles can reduce CO to five months (Masterson, Slowinski, and emissions by 80 percent and a large portion of Stanitski 1985). HC present in PM emissions (see Chapter 3). Motor vehicles (especially cars) are the main Ambient air temperature also affects the emis- contributors to anthropogenic CO emissions. sion rates of pollutants from vehicles. Higher Worldwide anthropogenic CO emissions for 1995 temperatures promote evaporative emissions are estimated at 350 million tons, 59 percent of that contain the more volatile fraction of HC in which were contributed by the transport sector, gasoline but result in lower emissions of HC and 39 percent by the residential and commercial NO, from the vehicle exhaust. CO emissions sectors, and 2 percent by the industrial and tend to be higher at lower temperatures. PM and power sectors. Developing countries contribute SO2 emissions are not affected by ambient air 50 percent of global CO emissions. In develop- temperature. ing countries the transport sector accounts for In addition, at higher altitudes, where air den- 53 percent of CO emissions and the residential sity is lower, the fuel-air ratio of the mixture ad- and commercial sectors, 46 percent. In Latin mitted to the engine of vehicles equipped with American urban centers motor vehicles are re- carburetors or continuous fuel injection systems sponsible for 60 to more than 90 percent of CO becomes larger. A higher ratio lowers engine emissions (OECD/IEA 1991). power output and increases CO and HC emis- sions from gasoline-fueled vehicles. Engine ad- Nitrogen oxides. Nitrogen oxides include nitric justments for higher altitudes are required to oxide (NO), nitrogen dioxide (NO2), nitrous reduce these emissions. For vehicles equipped oxide (N,O), dinitrogen trioxide (N203), and with electronic fuel injection systems, which au- nitrogen pentoxide (N2O5). Nitrogen oxides are tomatically compensate for altitude changes, produced by natural phenomena such as light- such adjustments are not necessary (Faiz, Weaver ning, volcanic eruptions, and bacterial action in & Walsh 1996). the soil and by anthropogenic sources such as combustion of fuels in internal combustion en- gines, thermal power plants, industrial and heat- Air Pollutants ing facilities, and incinerators. NO and NO2, collectively represented as NO., are the main Carbon monoxide. Carbon monoxide (CO), a col- nitrogen oxides emitted by vehicles. About 90 orless and odorless gas that is slightly denser than percent of these emissions are in the form of air, is emitted by natural and anthropogenic NO. NO is produced in the vehicle engine by sources. Anthropogenic sources form CO from combustion of nitrogen at high temperatures. incomplete combustion of carbonaceous fuels NO,, formed by oxidation of NO, has a reddish- in motor vehicles, heating and industrial facili- brown color and pungent odor. In the atmo- ties, thermal power plants, and incinerators. sphere it may be involved in a series of reactions Residence time and turbulence in the combus- (in the presence of ultraviolet radiation) that tion chamber, flame temperature, and excess produce photochemical smog, reducing visibil- 18 Chapter 2 ity. It may also react with moisture in air to form emissions from gasoline-fueled vehicles with no nitric acid (HNO3) aerosols. In the lower atmo- emission controls originate in the exhaust sys- sphere (troposphere) NO, forms ozone by re- tem, 13 to 25 percent come from the crankcase acting with HC. In the upper atmosphere it blow-by, and 20 to 32 percent evaporate in the reacts with chlorine monoxide to form chlorine fuel lines, fuel tank, and carburetor. Methane nitrate, which releases ozone-destroying chlorine constitutes 5 to 15 percent of HC emissions from atoms upon reaction with hydrogen chloride. vehicles not equipped with catalytic converters Motor vehicles are the main contributors to and up to 40 percent of exhaust HC from cata- anthropogenic NO, emissions. Worldwide an- lyst-equipped vehicles. This is because the cata- thropogenic NO, emissions for 1995 are esti- lysts are less effective in oxidizing methane than mated at 93 million tons, 43 percent of which other hydrocarbons. In the presence of ultra- were contributed by the transport sector, 32 per- violet radiation, NMHC and NO, react with oxy- cent by the power sector, 12 percent by the in- gen to form ozone in the troposphere. The dustrial sector, 8 percent by the residential and reaction time varies from less than an hour to commercial sectors, and 5 percent by other several days depending on the reactivity of the sources. Developing countries contribute 26 NMHC. percent of global NO, emissions. In developing Motor vehicles emit toxic HC, including ben- countries the transport sector accounts for 49 zene, 1,3-butadiene, aldehydes, and polycyclic percent of NO, emissions, and the power sector, aromatic hydrocarbons (PAH). Benzene is an 25 percent; the industrial sector, 11 percent; the aromatic HC present in gasoline. About 85 to residential and commercial sectors, 10 percent; 90 percent of benzene emissions come from and other sources, 5 percent (OECD/IEA 1991). exhaust and the remainder comes directly from gasoline evaporation and through distribution Hydrocarbon compounds. Hydrocarbon com- losses (Faiz, Weaver, and Walsh 1996). The ben- pounds (HC) are defined chemically as com- zene in exhaust originates both from partial com- pounds consisting of carbon and hydrogen. In bustion of other aromatic HC compounds in air quality studies, however, the term hydrocar- gasoline such as toluene and xylene; and from bons is often extended to include a variety of the benzene already in gasoline. Benzene con- other volatile organic compounds (VOCs) such stitutes 63 to 85 percent of the toxic emissions as alcohols and aldehydes. Most HC are not di- in exhaust from gasoline-fueled cars equipped rectly harmful to health at concentrations found with fuel-injected engines and new technology, in ambient air. Through chemical reactions in and 36 to 65 percent from older model-year cars the troposphere, however, they play an impor- equipped with carbureted engines and catalytic tant role in forming NO2 and ozone, which are converters (Table 2.3; AQIRP 1991). Aldehydes health and environmental hazards. Among the and 1,3-butadiene are not present in gasoline, various HC, methane (CH4) does not participate diesel fuel, ethanol, or methanol but are present in these reactions. Because the remaining HC, in their exhaust emissions as partial combustion nonmethane hydrocarbons (NMHC), are rieactive in products. Aldehydes are also formed in the at- forming secondary air pollutants, they are the mosphere from other mobile source pollutants focus of air quality studies (Horowitz 1982). and have a high photochemical reactivity in HC are emitted from natural and anthropo- ozone formation. The major types of aldehydes genic sources. Natural sources include anaero- formed include formaldehyde and acetaldehyde. bic decomposition of plants in swamps and Combustion of ethanol favors acetaldehyde marshes, seepage from natural gas and oil fields, emissions and combustion of methanol favors and emissions from trees. The first two sources formaldehyde emissions. Controlled gasoline- mainly produce methane, and the third source fueled cars have higher emissions of formalde- produces photochemically reactive HC hyde than acetaldehyde (see Table 2.3). (Horowitz 1982). Anthropogenic emission Uncontrolled gasoline-fueled vehicles emit 0.6 sources include motor vehicles, gasoline and to 2.3 grams of aldehydes per liter and uncon- solvent storage tanks and transfer stations, pe- trolled diesel-fueled vehicles emit 1 to 2 grams troleum refineries, and chemical and petro- of aldehydes per liter (Wijetilleke and chemical plants. HC emissions from motor Karunaratne 1995).PAHare emitted ata higher vehicles occur from unburned fuel or from par- rate in the exhaust of diesel-fueled vehicles than tial combustion of fuels. About 55 percent of HC gasoline-fueled vehicles. Air Pollutants and TheirEffects 19 Table 2.3 Exhaust eniissions of toxic air pollutants from gasoline-fueled cars Bei,zene 1, 3 Btuldiette Formaldehyde Acetal(dehyde 1989 model-year cars Toxic emissionsh (mg/km) 4.4-10.8 0.3-0.9 0.7-2.1 0.4-0.8 Percent of toxic emissions 60-85 6-10 7-24 2-8 1983-85 model-year U.S. carsa Toxic emissionsb (mg/km) 7.7-14.4 0.6-2.6 4.4-11.9 1.5-2.7 Percent of toxic emissions 36-66 2-11 22-49 7-13 a. The 1989 model-year U.S. cars are equipped with fuiel-inlected engines and new technology. The 1983-85 model-year U.S. cars are old catalyst-equipped cars with carbureteed enginies. b. Total toxic emissions range from 6.3 to 13.1 mg/km for 1989 model-year cars and from 16.9 to 25.0 mg/klm for 1983-85 model-year cars. ,Source: AQIRP 1991. Ozone. Ozone is a colorless gas that occurs in two converted to sulfur trioxide (SO) by reacting separate layers of the atmosphere. Ozone in the with oxygen. SO2 and SO3 react with the mois- outer (stratospheric) layer of the atmosphere is ture in air to form sulfurous (H,S03) and sulfu- generated by photolysis of oxygen or naturally ric (H2SO4) acids, which may be transported by occurring HC, and protects the earth from ul- winds many hundreds of kilometers before fall- traviolet rays. In the lower (tropospheric) layer, ing to earth as acid rain. Sulfates may also be ground-level ozone is formed 'by the reaction of produced through reaction of these sulfur com- VOCs and NO, with ambient oxygen in the pres- pounds with metals present in PM. ence of sunlight and high temperatures. Annual global emissions of SO2 are estimated Ground-level ozone is a major constituent of at 294 million tons. Of this, 160 million tons are smog in urban areas and motor vehicles are the from anthropogenic sources. About 90 percent main anthropogenic emission source of its pre- of these emissions originate from the northern cursors. Areas downwind of urban centers may hemisphere; the United States and the repub- also be affected by ground-level ozone as winds lics of the former Soviet Union are the two big- carry VOCs and NO,, from their original sources. gest sources (UNEP 199 1). The transport sector's The reactions that form ground-level ozone also contribution to global S02 emissions is estimated produce small quantities of other organic and at 2 to 6 percent. inorganic compounds such as peroxyacetyl- nitrate (PAN) and nitric acid (Horowitz 1982). Particulate matter. Particulate matter (PM) con- Ground-level ozone concentrations depend on sists of fine solids and liquid droplets, other than the absolute and relative concentrations of its pure water, that are dispersed in air. PM origi- precursors and the intensity of solar radiation, nates from natural as well as anthropogenic which exhibits diurnal and seasonal variations. sources. Natural sources include wind-blown soil Thermal inversions increase ground-level ozone dust, volcanic ash, forest fires, sea salt, and concentrations (World Bank 1996). pollens. Anthropogenic sources include thermal power plants, industries, commercial and resi- Sulifr dioxide. Sulfur dioxide (SO2) is a stable, dential facilities, and motor vehicles using fossil nonflammable, nonexplosive, colorless gas that fuels. can be detected by taste at concentrations as low Total suspended particulates (TSP) are particles as 1,000 4.g/m3 or by smell at concentrations with an aerodynamic diameter of less than above 10,000 Rig/mt. It is extremely soluble in 70pm.' PM larger than 10pim in diameter results water. S02 is produced by the combustion of sul- from physical actions such as wind erosion or fur-bearing fossil fuels for thermal power gen- grinding operations and tend to settle near their eration, heating, cooking, and transportation. Petroleum refining and ore smelting are addi- tional sources. In the atmosphere SO2 may be 1. lpm = 1 micrometer = 1 micron = 10' mieter. 20 Chapyer 2 emission source. PM with an aerodynamic diam- by oxygen deficiency during the fuel combus- eter of 10m or less, known as suspended inhalable tion or expansion phase. Blue, gray, and white particulate matteror PM-10, remains in the atmo- smoke are caused by the condensed HC in the sphere for longer periods because of its low exhaust of diesel-fueled vehicles. Blue or gray settling velocity. PM-10 can penetrate deeply smoke results from vaporized lubricating oil, and into the respiratory tract and cause respiratory white smoke occurs during engine start-up in illnesses in humans. PM with an aerodynamic cold weather. Diesel fuel additives such as diameter of 2.5-10 pim or less is defined as fine barium, calcium, or magnesium reduce smoke particles (PM-2.5), while larger PM is called coarse emissions, but increase PM sulfate emissions. particles. These additives may also increase PAH emissions. Coarse particles are generally emitted from Copper-based additives can reduce PM emis- wind-blown dust, vehicles traveling on unpaved sions, but may catalyze the reaction between HC roads, materials handling, and crushing and and trace amounts of chlorides in diesel fuel to grinding operations. Nearly all PM emitted by form dioxins, which are emitted in the exhaust motor vehicles consists of fine particles and a (Faiz, Weaver, and Walsh 1996). The use of large fraction of these particles has an aerody- barium and copper in diesel fuel additives is of namic diameter less than 1 lim. PM-2.5 results concern because of their toxic properties. from combustion of fossil fuels in power gen- eration and manufacturing facilities, residential Lead. Motor vehicles fueled with leaded gaso- fireplaces and wood stoves, and agricultural line are the main source of lead in ambient air. burning. PM-2.5 can also be formed in the at- Tetraethyl lead is added to gasoline to increase mosphere as aerosols from chemical reactions the fuel's octane number, which improves the that involve such gases as SO2, NO,, and VOC. antiknock characteristics of the fuel in spark- Sulfates, which are commonly generated by con- ignition engines. About 70 to 75 percent of this version from primary sulfur emissions, make up lead is transformed into inorganic lead in ve- the largest fraction of PM-2.5 by mass. (Pope and hicles' engines upon combustion and emitted others 1995). PM-2.5 can also form as a result of to the atmosphere through the exhaust pipe solidification of volatile metal salts as crystals along with 1 percent of the organic lead that following cooling of hot exhaust gases from ve- passes through the engine unchanged. The rest hicles in ambient air (Winchester 1989). PM-2.5 of the lead remains trapped within the exhaust can remain suspended in the air and travel long system. Organic lead emissions usually occur as distances. vapor, while inorganic lead is emitted as PM, of- Gasoline-fueled vehicles have lower PM emis- ten less than 1 pm in size. sion rates than diesel-fueled vehicles. PM emis- Inorganic lead in ambient air also originates sions from gasoline-fueled vehicles result from from emissions from coal combustion and vari- unburned lubricating oil, and ash-forming fuel ous lead-based industries such as lead smelters and oil additives (Faiz, Weaver, and Walsh 1996). and lead battery plants. Although lead in gaso- For vehicles fueled by leaded gasoline, lead com- line accounts for less than 10 percent of all re- pounds accountfor a major portion of PM emis- fined lead production, about 80 to 90 percent sions. PM emitted by diesel-fueled vehicles of lead in global ambient air originates from consists of soot formed during combustion, combustion of leaded gasoline (GEMS 1988). heavy HC condensed or adsorbed on the soot, and sulfates. These emissions contain polycyclic Carbon dioxide. Carbon dioxide (CO2) is a green- aromatic hydrocarbons (PAH). In older diesel- house gas.2 Between 1970 and 1992 CO, con- fueled vehicles the contribution of soot to PM centrations in the atmosphere increased from emissions is between 40 and 80 percent. With the 325 ppm to 356 ppm (WRI 1994). The increase advance of emission control measures in en- in CO2 concentrations, which is associated with gines, however, the contribution of soot has been reduced considerably. Heavy HC, referred to as the soluble organic fraction of PM, originate from 2. Greenihoiise gases are gases that absorb some of the lubricating oil, unburned fuel, and compounds heat radiated from the earth's surface, which would othier- formed during combustion (Walsh 1995). wise escape into the space. This process raises the tempera- ttire of the atmosphere. Besides C02, greenhouse gases Black smoke, associated with the soot portion incluide water vapor, methane, nitrous oxide, ozone, and of PM emitted by diesel-fueled vehicles, is caused other trace gases. Air Pollutants and Their Effects 21 global warming, results mainly from increased after the ban on CFC use in the United States in combustion of fossil fuels (including motor 1984 and the signing of the Montreal Protocol vehicle fuels) and land use (including defores- in 1987. The Montreal Protocol calls for restric- tation). tions on the production and consumption of Between 1970 and 1992 energy-related global CFCs and halons according to the following time- CO2 emissions increased from about 15 billion table: a return to 1986 levels by 1989, 80 per- tons to 22 billion tons. During this period (C,, cent of 1986 levels by 1994, and 50 percent of emissions grew by 28 percent in OECD coun- 1986 levels by 1999. Developing countries fol- tries and 82 percent in non-OECD countries. low a less restrictive schedule. The protocol was The largest CO, emitters were the United States, strengthened by the London Agreement of 1990, the former Soviet Union, and China (EIA which included a faster timetable and complete 1994a). phaseout of certain ozone-depleting chemicals. In 1992 Latin America and the Caribbean The 1992 Copenhagen amendments, which are emitted about 1.1 billion tons of CO,, less than not yet in effect, further shortened the phase- any other non-OECD region except Africa. C,O., out schedule. In 1991 global CFC emissions were emissions resulted mainly from combustion of estimated to be 400,000 metric tons, of which oil and also from natural gas and coal. Brazil and only 3.5 percent (or 14,000 metric tons) came Mexico together contributed 52 percent of the from Latin American countries (WRI 1994). region's CO2 emissions. Other major CO2 emit- ters were Venezuela, Argentina, and Colombia, with a total contribution of about 26 percent. Quantifying PoHutant Emissions The transport sector's share in CO2 emissions from these five countries was 36 percent (EIA Air pollution in urban areas originates from 1994b). mobile, stationary, and natural sources. The con- tribution of each source type is site-specific. The Chlorofluorocarbons. Chlorofluorocarbons (CFCs) systematic collection and compilation of detailed are used mainly in aerosols, solvents, foam blow- information about pollutant emissions in a given ing, refrigerators, and air conditioners. The area is referred to as an "emissions inventory." source of CFC emissions from motor vehicles is An inventory should contain as much informa- the freon gases used in air conditioners. The tion as necessary on the type of sources, compo- contribution of motor vehicles to global CFC sition of emissions, and rates of emissions for emissions is estimated at about 28 percent (Faiz, various pollutants. The inventory must be kept Weaver, and Walsh 1996). current to study the impacts of changes in pol- Unlike other pollutants, CFCs are unaffected lution sources (such as an increase in the ve- by natural cleansing mechanisms such as rain. hicle fleet) and the effectiveness of air pollution CFCs emitted into the atmosphere rise to the control strategies on ambient air quality stratospheric layer within ten years and are esti- (Rossano and Rolander 1976). Only a few ur- mated to remain there for about 400 years. CFC, ban centers (Mexico City, Sao Paulo) in Latin molecules struck by ultraviolet radiation release America and the Caribbean maintain up-to-date chlorine atoms, which destroy ozone by form- emission inventories. In some urban centers ing chlorine monoxide. Furthermore, when a (Belo Horizonte, Rio de Janeiro, Santiago, free oxygen atom reacts with a chloride mol- Santafe de Bogota) emission inventories were ecule, an oxygen molecule is formed and a chlo- developed as part of specific environmental rine atom is released to destroy more ozone. projects but have not been updated to be used Ozone destruction in the stratosphere implies as a continuing planning tool for air quality more exposure to ultraviolet radiation with a management. In other urban centers (Buenos wvavelength range of 295 nnm to 300 nm (UV-B), Aires) no systematic effort has been made to which is biologically the most damaging (USEPA develop emission inventories. 1995) .3 Developing an accurate emissions inventory Global CFC emissions increased exponentially for motor vehicles is a complex task. Emissions during the 1960s and early 1970s but declined modeling of a motor vehicle fleet consists of quantifying emission-producing activities through a travel demand model or other means 3. t nm = I nanometer = 10` meter. of estimation; providing data on vehicle, fuel, 22 Chapter 2 operating, and environmental characteristics to gasoline vehicles, light-duty diesel vehicles, light- the computer model; running the emission rate duty diesel trucks, heavy-duty diesel vehicles, and model to predict activity-specific emission fac- motorcycles. Emission factors derived from tors for a given vehicle, fuel, operating, and en- MOBILE5a under specific conditions for U.S. vironmental characteristics (that is, emission gasoline-fueled passenger cars and trucks, and rates from mobile sources); multiplying each U.S. diesel-fueled passenger cars, trucks, and activity estimate by its activity-specific emission buses are shown in Annex A to this chapter. factor; and stumming the estimated emissions for These results illustrate the effects control tech- all activities. Typical vehicle, fuel, operating, and nologies can have on pollutant emissions under environmental characteristics that affect motor these conditions. Because MOBILE5a cannot es- vehicle emission rates are shown in Table 2.4. timate PM and S02 emissions, in 1995 the USEPA Ambient concentrations of pollutants can be introduced another model, PART5, for these pol- predicted through estimation of emissions from lutants. PART5 uses input parameters similar to polltition sources and dispersion modeliing. A those for MOBILE5a. commonly used source for estimating emission MOBILE5a can be used to estimate emission rates is the USEPA's emission factor database factors and total emissions from vehicles in a (Compilation of Air Pollutant Emission Factors, developing country only after certain modifica- commonly referred asAP42). Because manyfac- tions to the model have been made. The tors affect pollutant emission rates from vehicles, MOBILE5a modifications made in a recent air the USEPA has developed a comptiter model pollution study in Monterrey, Mexico are shown that estimates pollutant emissions under selected in Table 2.5. One of these modifications involved conditions. A recent version of this model. called matching exhaust and evaporative emission con- MOBILE5a, estimates emission factors for ex- trol technologies between U.S. and Mexican ve- hatist, evaporative, hot-soak, and diturnal emis- hicles (Table 2.6). The results indicate that a sions for each vehicle type and model-year and typical model-year vehicle in Monterrey (for the total emissions for the vehicle fleet. The air example, a typical 1985 model-year light-duty pollutants modeled in MOBILE5a are CO, HC, gasoline-fueled vehicle) has the same rate of and NO,. Input data include such parameters pollutant emissions as an older typical model- as the type and model-year of motor vehicles, year vehicle in the United States (for example, vehicle-kilometers traveled per yeal, speed char- a typical 1972 model-year light-duty gasoline-fu- acteristics, fuel type and vapor pressuire, inspec- eled vehicle). Another major difference that was tion and maintenance program featuires, taken into accountwas the types of vehicle fleets tampering effects, and summer and winter diur- in Mexico and the United States. The Mexican nal temperatures. The vehicle types considered registration data revealed a larger fraction of in the model are light-duity gasoline vehicles, older cars in Monterrey than in the United States light-duty gasoline trucks (two types). heavy-duty (Radian Corporation 1995). Table 2.4 Typical factors affecting motor vehicle emission rates Vehicle operating Environmental Vehkle parameters Fi,>0rp()I'IMel: conditions Parameters Vehicle class Ftiel t%pe Average vehicle Altitude Model-year Oxygen content speed Humidity Fuel delivery system Fuel volatility Load (such as air Ambient temperature Emissions control system Sulfuir contelnt conditioner, Diurnal temperature Onboard computer control Benzenie conitent heavy loads) changes system Olefin and aromatic Cold or hot start mode Control system tampering HC conltenlt Inspectiorn and mainteTnance Lead and metals history content Soturce: Adapted from Guelnsler 1994. Air Pollutants and Their Effects 23 Table 2.5 List of modifications to adopt MOBILE5a to Monterrey, Mexico Para meter Afodfirtalion Basic exhaust emission rates of HC, CO, and NOX: Matched appropriate exhaust and evaporative HC evaporative emissions emissions control technology between the United States and Mexico (exhaust from inspection data, evaporation from assump- tions made for the Mexico City vehicles) Registration distribution Based on registration and inspection data Vehicle kilometer traveled by vehicle type Based on Mexican records and inspection data Tampering offsets and tampering rates MOBILE5a correction based on Mexican vehicle techlnology Inspection and maintenance program credits MOBILE5a correction based on Mexican vehicle technology Temperature and Reid Vapor Pressure (RVP) MOBILE5a correction factors based on equiva- corrections for exhaust emissions lent Mexican vehicle technology Fraction of carburetor/fuel injection vehicles MOBILE5a correction factors based on equiva- lent Mexican vehicle technology Speed, air conditioning, extra load, towing MOBILE5a correction factors based on equiva- lent Mexican vehicle technology Operating mode corrections MOBILE5a correction factors based on equiva- lent Mexican vehicle technology Fleet kilometer accumulation distribution MOBILE5a values as a function of vehicle age (data from Mexican authorities) Crankcase emissions MOBILE5a relationships for equivalent Mexican technology Evaporative emissions (vehicles with RVP, tempera- MOBILE5a relationships for equivalent Mexican ture, fuel tank level, vehicle driving conditions) vehicle technology Running loss relationships with RVP and temperature MOBILE5a relationships for equivalent Mexican vehicle technology Miles per day and trips per day as a function of MOBILE5a relationships for equivalent Mexican vehicle kilometers vehicle technology Anti-tampering program benefits MOBILE5a values for equivalent Mexican vehicle technology Idle emissions MOBILE5a relationships for equivalent Mexican vehicle technology Sourre: Radiani Corporation 1995. Dispersion of Air Pollutants in forming secondary pollutants. In addi- tion, higher temperatures promote evapo- Ambient concentrations of pollutants result nlot rative emissions from vehicles. only from the magnitude of pollutant emissions * Solar radiation enhances the formation of but also from the way in which primary polltit- secondary pollutants such as ozone. The in- ants are transported and dispersed and react with tensity of solar radiation is affected by the each other in the atmosphere to form second- amount of cloud cover. Urban centers in ary pollutants. These mechanisms are affected wann, sunny locations with high traffic den- by a number of meteorological factors: sities (such as Mexico City and Sao Paulo) tend to be especially prone to the forma- * Wind moves air pollutants from one loca- tion of ozone from emissions of HC and tion to another. The trajectory followed by NO,. airborne pollutants and the extent of their * Ceiling height, or mixing height or depth, dilution depend on wind direction and is the height above the earth's surface at speed. which relatively vigorous vertical mixing * Higher temperatures increase the reactiv- occurs. It is used to represent the disper- ity of primary pollutants in the atmosphere sion capacity of the atmosphere. The mix- 24 Chapter 2 Table 2.6 Equivalent model-years for Monterrey, Mexico and U.S. vehicles Mexican vehicle U.S. vehicle model-vear mod(el-year 7Iype la Typ'e 2' 7Tpe f 1971 1968 1968 1971 1972 1968 1968 1971 1973 1971 1971 1971 1974 1971 1971 1971 1975 1971 1971 1971 1976 1971 1971 1971 1977 1971 1971 1971 1978 1971 1971 1971 1979 1971 1971 1971 1980 1971 1971 1971 1981 1972 1972 1972 1982 1972 1972 1972 1983 1972 1972 1972 1984 1972 1972 1972 1985 1972 1972 1972 1986 1972 1972 1972 1987 1972 1972 1972 1988 1975 1974 1974 1989 1975 1974 1974 1990 1980 1974 1974 1991 1980 1974 1974 1992 1981 1974 1974 1993 1988 1977 1977 1994 1988 1981 1981 1995 1989 1981 1981 a. Light-dutty gasoline vehicles. b. Light-dcity gasolinie trucks, heavy-duty gasoline vehicles, light-duty diesel vehicles, light-duty diesel trucks, and heavy-duty diesel vehicles. c. Motorcycles. Source: Radian Corporation 1995. ing height varies by the time of the day and * Precipitation affects ambient pollutant con- month of the year. For example, in centrations because it washes out pollut- Santiago, Chile, the mixing height is as low ants, particularly PM, from the air. In the as 300 meters durirg the winter and as high presence of acidic pollutants, humidity con- as 1,000 meters during the summer. The tributes to corrosion of limestone buildings, mixing height is affected by theimal inver- sculptures, and metallic structures. sion. Under normal conditions, when the * Local topography also influences the trans- air temperature decreases with altitude, hot portation and dispersion of air pollutants. pollutant gases rise to high altitudes. Un- Urban centers with relatively level topog- der isothermal conditions, however, when raphy, such as Buenos Aires, have better there is no change of temperature with al- wind dispersion. The presence of an ocean titude, an inversion layer forms above the coastline can lead to onshore and offshore ground, trapping primary pollutants and diurnal wind patterns that help disperse enhancing formation of secondary pollut- pollutants. Hills and mountains that sur- ants in the atmosphere. These conditions round urban centers often act as downwind are of greatest concern when wind speeds barriers. For example, in the Mexico City are low. Thermal inversions are observed Metropolitan Area the surrounding moun- in many Latin American urban centers, tains tend to limit air circulation, trapping including Mexico City, Santiago, Sao Paulo, pollutants within a valley. Daytime winds in and Rio deJaneiro. the Valley of Mexico carry pollutants from Air Pollutants and Their Effects 25 Box 2.1 Commonly used air dispersion models for vehicular emissions Among the air dispersion models discussed below. onilv the Urban Airshed Model (UAM) is able to simulate an entire uirban area (the others are microscale modtels) and incorporate reactive pollutants. The UAM is recom- mended for urban areas with an ozone problem. CALINE 3 (California Line Sour-e Model-Version 3). This USEPA-approved Gaussian model is used to predict dispersion of nonreactive pollutanits near highways and main streets. It is not suitable for urban areas with complex terrain, urban street canyoni conditions (btuiltling heights over three stories), wind speeds less than 1 m/sec, or distances over 10 km. It also containis an algorithm for deposition and settling velocity so that PM concentrations can be predictedl. CAL1NE4. This model updates CALINE3 but has not been approved by the USEPA. It accounts for queuing, delays, excess emission rates due to modes, an(l cruising. HIWAY2. This USEPA-developed Gaussian model lacks the sophistication of GALINE3 or CALINE4 but is easier to use. It provides ambient polltitant concenitrations based on uniform emission rates (that is, it does not take into account interrupted flows at intersectionis). It is not suitable for complex terrain, street canyon condi- tions, or calm atmospheric conditions. TEXIN2. This Gaussian model incorporates features of MOBILE and CALINE3. It accounts for queuing, delays, excess emission rates due to modes, and cruisinig. It is not suitable for simulating conditions in which the winid speed is less than 1 m/sec and receptors are at heights above 10 meters. UAM (Urban Airslhed Model). This USEPA-approved model is an urban-scale, three-dimensional, grid-type nu- merical simulation model that can be used for mobile and fixed emission sources. The model incorporates a photochemical kinetics mechanism for urban atmospheres. It is designed for computing short-term (one or two days) ozone concentrations resulting from emissionis of NO. and volatile organic compounds. Simulation of CO is optional. Somnre: USEPA 1993. the heavily industrial areas of the north and reactive pollutants (such as NO.) are particu- northwest to the populated areas, causing larly difficult to simulate because they form sec- severe air pollution. P'olluted air trapped ondary pollutants. Typical inputs to such by the high mountains to the south recir- dispersion models include information on emis- culates and exacerbates the air pollutioni. sion sources (for example, frequency distribu- This area tends to experience the highest tion of emissions from major sources in the area frequency of ozone standard violations, under study), meteorological parameters (wind with levels in excess of the Mexican stan- speed and direction, vertical atmospheric tem- dard more than 88 percent of the days in a perature profile, radiation intensity), and a ki- year. netic mechanism to describe the rates of * Within an urban center, buildings and atmospheric chemical reactions as a function of other strmctures can have a great effect on pollutants present. Validation and fine tuning the dispersion of air pollutants. The "street of models with actual monitoring data are nec- canyon" effect occurs when tall buildings essary. prevent wind dispersion of low-level emis- sions (Bellomo and Liff 1984). Health Effects of Air Pollutants Using actual or estimated pollutant emission data, air dispersion models predict ambient air Air pollutants emitted by motor vehicles have a concentrations based on atmospheric and topo- number of adverse effects on human health. graphic conditions. Typical air dispersion mod- Inhalation is the main route of exposure to air els used in the United States for vehicular pollutants originating from motor vehicle emis- emissions are described in Box 2.1. Among the sions. Other exposure routes-drinking water various pollutants emitted by vehicles, chemically contamination, food contamination, and absorp- 26 Chapter 2 tion through the skin-are also possible. Expo- linked to NO2 occurs at thejunction of the con- sure by inhalation directly affects respiratory, ducting airway and the gas exchange region of nervous, and cardiovascular systems of humans, the lungs. The upper airways are less affected resulting in impaired pulmonary functions, sick- because NO, is not very soluble in aqueous sur- ness, and even death. faces. Exposure to NO2 is linked with increased susceptibility to respiratory infection, increased Carbon monoxide. CO absorbed throughi the lungs airway resistance in asthmatics, and decreased reduces the blood's capacity to transport avail- pulmonary function. Short-term exposure to able oxygen to the tissues. CO bonds with he- NO2 has been associated with a wide range of moglobin (Hb) to form carboxvhemoglobin lower respiratory illnesses in children (cough, (COHb), which lowers the oxygen level in blood. ninny nose, and sore throat are among the most Because more blood is needed to supply the same common), as well as increased sensitivity to amount of oxygen, the heart must work harder. urban dust and pollen. Health effects of occu- The relationship between ambient CO con- pational exposure to NO2 range from inflamma- centrations in air and COHb levels in blood de- tion of the mucous membrane of the pends mainly on the duration of exposutr-e and tracheobronchial tree to bronchitis, broncho- the pulse rate of the exposed person (that is, pneumonia, and acute pulmonary edema. Ni- intensity of physical effort). Body size. IaLng con- tric and nitrous acids or their salts are present dition, and barometric pressure also affect CO in the blood and urine after exposure to NO2 uptake. The COHb level is normally about 1.2 (Romieu 1992). to 1.5 percent, but it can reach 4 to 7 percent Lung function is affected by 30-minute expo- among one-pack-a-day cigarette smokers. At sure to a NO2 concentration of 560 pg/M3 with about 5 percent the COHb level begins to in- exercise, 940 pig/m2 in asthmatic people, and duce adverse health effects. Some studies have above 1,300 pig/ml for a 10- to 15-minute expo- shown that impaired judgment starts at even sure in healthy people. Eleven epidemiological lower COHb levels of 3.2 to 4.2 percent (Romieu studies of long-term exposure found that a 30 1992). pg/m3 increase in indoor NO2 concentrations CO uptake impairs perception and thinking, from gas stoves causes respiratory illnesses to slows reflexes, and may cause drowsiness, angina, increase by 20 percent among children under unconsciousness, or death (Romieu 1992). Ex- 12 years of age (Romieu 1992). The relation- posure of pregnant women to CO has been ship between outdoor NO2 exposure and acute linked to low birth weights and retarded postna- health effects, however, has not been demon- tal development. The synergistic effect of CO strated consistently from epidemiological stud- with other pollutants promotes illness in people ies because of other intervening factors such as with respiratory problems. Increased concentra- exposure to other pollutants, smoking habits, tions of CO are also associated with reduced and indoor exposure to NO2. In one study ex- worker productivity and general discomfort. posure to a daily mean NO2 concentration of An exposure to concentrations of 45 mg/m3 244 pg/mi was associated with sore throats of CO for more than two hours adversely, affects among adults (Schwartz and Zeger 1990). a person's ability to makejudgments. Two to four hours of exposure at 200 mg/mi raises the Benzene. About 50 percent of inhaled benzene is COHb level in the blood to 10 to 30 percent absorbed. Part of the absorbed benzene is ex- and increases the possibility of headaches. Ex- haled by respiration and eliminated through the posure to 1,000 mg/mr of CO raises the COHb urinary tract. Benzene maintained in the human level in blood to more than 30 percent and body is concentrated in the fat tissue and bone causes a rapid increase in pulse rate leading to marrow. coma and convulsions. One to two hotirs of ex- Benzene has toxic and carcinogenic effects. posure at 1,830 mg/m3 results in 40) percent The toxic effects are associated with the central COHb in blood, which may cause death (MARC nervous system as well as the hematological and 1991). immunological systems. Toxic effects on the ner- vous system have been observed following expo- Nitrogen dioxide. NO2 is an irritating gas that is sure to concentrations higher than 3,200 mg/ absorbed into the mucous membrane of the res- mi3 (1,000 ppm). Occupational studies of high- piratory tract. The most adverse health effect level exposure to benzene have found that it can - Air Pollutants and Their Effects 27 damage the respiratory tract, lung tissue, and reported at hourly average ozone concentrations bone marrow and can cause death. Carcinogenic in the range of 160 pg/mr to 300 pg/m3. In a effects include leukemia. The risk of lifetime study conducted among schoolchildren in exposure to 1 pig/M3 of benzene is estimated to Mexico City, acute and subacute effects of ozone range from 0.08 to 10 excess leukemia deaths on lung functions were reported. Because the per million (Romieu 1992). decrease in lung functions was smaller than that observed in another study, it was suggested Polycydic aromatic hydrocarbons. PAH, absorbed that children chronically exposed to ozone in the lungs and intestines and metabolized could tend to develop a tolerance to it (Romieu in the human body, are mutagenic and car- 1992). cinogenic. Epidemiological studies have identi- A study on the long-term health effects of fied 50 percent greater risk of bladder cancer ozone exposure in southern California found among truck drivers and delivery men exposed that it may reduce pulmonary function (Detels to diesel engine exhaust. Itis also estimated that and others 1987). The synergistic effects of 9 of 100,000 people exposed to 1 pg/mi of ozone and other pollutants (sulfates and NO2), benzo [a] pyrene, a PAH, over a lifetime, would and absence of a threshold value for ozone have develop cancer. There no known threshold level also been reported (Romieu 1992). for carcinogenic effects of benzo[a]pyrene Based on results of different studies in the Los (Romieu 1992). Angeles area (Krupnick, Harrington, and Ostro 1990; Schwartz and Zeger 1990; Whittemore Aldehydes. Aldehydes are absorbed in the respi- and Korn 1980), Ostro (1994) estimated the ratory and gastrointestinal tracts and metabo- effects of ozone on respiratory symptoms (for lized. Once metabolized they are excreted from example, chest discomfort, coughing, wheezing, the human body. Adverse health effects of form- sore throat, cold, and flu), eye irritation inci- aldehyde include eye and nose irritation (at a dents, and asthma attacks. For a 1 pig/m3 in- concentration of 0.06 mg/m3), irritation of mu- crease in the annual average of 1-hour daily cous membranes and alteration in respiration maximum ozone, Ostro predicted 28 to 97 (at a concentration of 0.12 mg/m3), coughing, respiratory symptom days per person per year, nausea, and shortness of breath. The threshold 23 to 30 eye irritations per adult per year, and for tissue damage is about I mg/m3. Occupa- 39 to 190 asthma attacks per asthmatic person tional exposure to formaldehyde is associated per year. with risk of cancer (Romieu 1992). Su3ur dioxide. SO2, an irritating gas that is ab- Ozone. One of the most widespread traffic-in- sorbed in the nose and aqueous surfaces of the duced air pollutants is ozone formed in the tro- upper respiratory tract, is associated with re- posphere, a principal ingredient of urban smog. duced lung function and increased risk of mor- Adverse health effects of ozone have been ob- tality and morbidity. Adverse health effects of served for exposure periods as short as five min- SO2 include coughing, phlegm, chest discomfort, utes. These effects become much more and bronchitis. Some of the SO2 emissions from pronounced during longer exposure periodls mobile or fixed sources are transformed in the (for example, over six hours) at moderate exer- atmosphere into sulfate aerosols (discussed be- cise levels. Changes in pulmonary function have low), which are also associated with mortality and been reported for one- to three-hour exposures morbidity. during exercise (Romieu 1992). Ozone can SO2 exacerbates the effects of PM, and vice cause severe damage to lung tissues and impair versa. The World Health Organization (WHO) defenses against bacteria and viruses. has determined that the effects of 24-hour hu- Short-term adverse health effects have been man exposure to SO2 include mortality at ambi- observed from hourly exposures to ozone con- ent concentrations above 500 pg/m3 and centrations as low as 200 pg/m3. These effects increased acute respiratory morbidity at ambi- include eye, nose, and throat irritation, cough- ent concentrations above 250 pg/mr3. Annual ex- ing, throat dryness, thoracic pain, increased posure to SO2 causes increased respiratory mucous production, chest tightness, lassitu(de, symptoms or illness at ambient concentrations malaise, and nausea. A decrease in pulmonary above 100 pg/m3 (WHO 1987). In recent stud- functions in children and young adults has been ies, however, the adverse effects of SO2 have been 28 Chapter 2 observed at lower concentrations (Romieu near the fine airways. PM-2.5 is a larger health 1992). concern because it can evade the human body's Correlations between exposure to SO, and respiratory defense system and reach the lung mortality have been established through stud- tissue, where it can remain imbedded for years, ies performed in different parts of the world, or in the case of soluble particles, be absorbed including England, France, Greece, and Poland. into the bloodsteam (ALA 1997). Particle depo- In the London study a correlation between am- sition increases with mouth breathing. PM in bient SO2 concentrations and mortality from ambient air has been associated with increased chronic bronchitis was found at an SO,, level of mortality, morbidity, and reduced lung function. 172 pg/min (and at a smoke level of 80 jig/mi) Adverse health effects have been observed in among men above 65 years of age and women both children and adults. These effects are as- between 45 and 65 years of age (Chinn and oth- sociated with coughing and respiratory diseases ers 1981). In another study associations between such as pneumonia, asthma, and bronchitis. ambient SO2 concentrations and respiratory PM and SO2 often occur together in ambient deaths in Marseilles and Lyon and between am- air and may have synergistic effects with other bient SO2 concentrations and cardiovascular pollutants emitted by motor vehicles. Studies that deaths in Marseilles were observed among indi- compare cities have found that SO2 and PM to- viduals over 65 years of age. In these studlies the gether account for 4 percent of the variation in outcomes were observed at daily averages for death rates from cardiovascular diseases. Al- ambient concentrations of SO2 and TSP of 78 though many other factors (such as differences pig/mi3 and 106 pig/m3, respectively, and monthly in smoking habits and type of occupation) may SO averages above 182 pig/m3 (Derriennic and also be important, studies conducted in differ- others 1989). In the Athens study ambient SO2 ent parts of the world indicate a relatively con- concentrations and mortality were correlated at sistent association between long-term exposures mean daily SO and black smoke levels of 85 pg/ in residential communities polluted by PM and m3 and 63 jig/m3, respectively (Hatzakis and SO2andincreasedmortalityrates (Romieu 1992). others 1986). In the Cracow study a relationship Increased mortality and respiratory diseases was found between ambient SO2 and PM con- are associated with PM-2.5 and sulfate air pollu- centrations and mortality for men (Krzyzanowski tion at levels commonly found in the U.S. cities and Wojtyniak 1982). Based on the literature, (4 pig/mi to 20 pig/m3 for sulfate and 10 pig/mr3 Ostro (1994) estimated that a 10 pg/ m:' drop in to 25 jig/ml for PM-2.5). Two recent studies con- ambient SO2 concentrations would be associated ducted in the United States indicate that long- with 0.20 to 1.21 percent drop-equivalent to term exposure to an extra 10 pg/M3 of PM-2.5 is fifteen to eighty-seven deaths per I million associated with 5 to 10 percent increase in over- people. all mortality and a higher increase in cardiores- A statistically significant correlation between piratorymortality (Dockery and Pope 1993; Pope ambient SO2 concentrations and acute health and others 1995). Based on data calculated over effects (coughing) was demonstrated for chil- a six-year period in the United States, Ostro dren in Watertown, Massachusetts (Schwartz and (1989) found an association between a 1 pg/ml others 1991). In another study performed on increase in the annual mean of PM-2.5 concen- nursing school students in Los Angeles, a sig- trations and a 3.2 percent increase in acute res- nificant association was observed between am- piratory diseases in adults aged 18 to 65 years. bient SO2 concentrations and chest discomfort Sulfate aerosols, especially those that are acidic, (Schwartz, Hasselblad, and Pitcher 1988). Based are considered one of the likely causative agents on these data, Ostro (1994) estimated that a 10 in the association between PM-2.5 and health pg/m3 change in ambient concentrations of 82 effects in the eastern United States. The main would cause ten to twenty-six cough incidents health effects of sulfate aerosols include chronic among 100,000 children, and five to fifteen chest bronchitis and asthma. discomfort incidents among 100 adults. An air pollution incident in London during the winter of 1958-59, in which smoke and SO2 Particulate matter. Through nasal breathing, PM levels exceeded 500 pg/m3, affected the health greater than 10 pm in diameter is deposited in of exposed people with preexisting heart and the extrathoracic part of the respiratory tract, lung diseases, the elderly, and children under while the 2.5 pm to 10 pm fraction is deposited five years of age. Further evaluation of data for Air Pollutants and TheirEffects 29 this incident and for subseq[uent years found a Korn 1980; Pope and others 1991; Ostro and strong correlation between ambient PM concen- others 1991). Respiratory symptoms (including trations and daily mortality iin London, with no chest discomfort, coughing, wheezing, sore threshold level of ambient pollutant concentra- throat, cold, and flu) were associated with ambi- tions for adverse health effects (Ostro 1994). ent concentrations of PM-10, using haze as a Based on Ostro's analysis, the increase in mor- surrogate measure for fine particles (Krupnick, tality for a 10 pg/m3 increase in PM-10 concen- Harrington, and Ostro 1990). Ostro (1994) esti- tration was 0.29 to 0.33 percent from the London mated that a 10 pig/m3 change in the ambient study (Schwartz and Marcus 1990), 0.44 to 0.94 concentrations of PM-10 would cause two to percent from a Steubenville study (Schwartz and twenty-seven asthma attacks among ten asth- Dockery 1992b), 0.49 to 1.47 percent from an matic people, and one to three respiratory symp- Ontario study (Plagiannakos and Parker 1988), tom days per person per year. 0.73 to 1.51 percent from a Santa Clara County Several epidemiological studies have been study (Fairley 1990), 0.96 to 1.44 percent from carried out to determine the health effects of a Philadelphia study (Schwartz and Dockery PM in Latin America. For example, in 1980 a 1992a), and 0.96 to 2.06 percent from a study of study was conducted in two public elementary "100 U.S. metropolitan areas" (Ozkaynak and schools in Mexico City, one in the industrial area Thurston 1987). Fora 10 jig/m3 increase in PM- of Xalostoc (with high ambientPM and SO2 con- 10 concentrations, the corresponding range for centrations) and the other in the less industrial- the number of deaths would be between forty- ized suburban area of San Lorenzo. Children five and ninety-one people per million. from Xalostoc were found to have a lower pul- Based on 24-hour exposure, smoke at 250 jig/ monary function than those from San Lorenzo, m3 is associated with increased acute respiratory although no difference was observed in terms morbidity among adults, and TSP at 180 jig/ml of acute or chronic respiratory symptoms and level and PM-10 at 110 jig/m3 with decrements illness. During 1983-84 the prevalence of respi- in lung functions among children. Increased ratory symptoms was investigated in three com- respiratory symptoms or illness would be ex- munities southwest of Mexico City. In two of pected at an annual mean exposure to 100 jig/ these communities, which were affected by PM m3 of smoke, and decrements in lung function emissions from a cement plant, the incidence would be expected at an annual mean exposure of chronic cough was found to be significantly to 180 pig/m3 of TSP (WHO 1987). However, related to the length of residency. Among sub- more recent studies suggest that health may be jects living away from a major emission source, affected even at lower concentrations (Romieu acute respiratory symptoms, possibly related to 1992). ozone exposure, were more frequent. During Dockery, Speizer, and Stram (1989) investi- November 1985 toJune 1986, another studywas gated the effects of PM on lower respiratory ill- conducted involving 6- to 13-year-old school chil- ness in children of six cities in the U.S. cities. dren from three different areas of Mexico City: The study found that chronic cough, bronchi- Xalostoc from the northern area, Morazan from tis, and chest illness were positively associated the central area, and Pedregal from the south- with TSP, PM-15, PM-2.5, and sulfate aerosol. ern area. The study found that respiratory ill- Frequency of earache was associated with ambi- nesses were higher in the northern and southern ent PM concentrations; and children with histo- areas (with high PM concentrations) than in the ries of wheeze and asthma had a much higlher central area (Romieu, Weitzenfeld, and Einkel- prevalence of respiratory symptoms, and the man 1992). symptom rates were stronger among children For different areas within the Rio deJaneiro with hyperactive airways. Using data from this metropolitan region, a statistically significant study, Ostro (1994) estimated that a 10 pig/ni3 association was observed between the average change in the ambient concentration of PM-10 annual PM concentrations in ambient air and would cause eight to twenty-four chronic bron- infant mortality from pneumonia (Penna and chitis incidents among 1,000 children age 17 and Duchiade 1991). Another Brazilian study on the under. health effects of ambient PM-10 concentrations Ambient concentrations of PM and exacerba- was conducted between 1984 and 1987 in Vila tion of asthma attacks were found to be related Parisi (near Cubatao), where industry is the main among children and adults (Whittemore and source of air pollution. The study found that a 30 Chapter 2 decrease in the annual arithmetic average of change in the ambient concentration of PM-10 ambient PM-10 concentrations from 186 jig/m3 would cause three to nine chronic bronchitis in- to 151 jig/ml (a result of industrial pollution cidents among 10,000 people older than 25. control measures) lowered the share of respira- Exposure to diesel engine exhaust fumes was tory problem-related emergency room visits found to decrease pulmonary function, but the from 31 to 23 percent, and of bronchitis- and effects were reversible after a few days without asthma-related emergency room visits from 15 exposure. In other studies associations between to 11 percent. Another study conducted in PM and pulmonary cancer in animals were ob- Ctubatio explored the effects of air polluttion on served (Romieu 1992). Based on a review of pre- the pulmonary functions of 6-year-old schoolchil- vious research, the International Agency for dren. Tests performed on 480 children in 1983 Research on Cancer (IARC) concluded that die- and 630 children in 1985 found an improvement sel PM emissions had possible carcinogenic ef- in pulmonary functions that was attributed to fects on humans. Confirmatory studies suggest the reduction in ambient PM concentrations that soot in diesel PM emissions was primarily (Romieu, Weitzenfeld, and Finkelman 1992). responsible for lung cancer in rats and, at high The effects of air pollution in Santiago and PM concentrations, the mutagenic compounds Los Andes (a city in Chile considered non- adsorbed onto sootwould playalesserifany role polluted) on 300 schoolchildren, aged 9 to 13 in tumor development. This finding indicates were studied during November 1987 to March the importance of controlling not only the or- 1988. Respiratory symptoms (coughing, hoarse- ganic compounds present on soot particle sur- ness, wheezing, nocturnal respiratory malaise) faces but also the soot particles themselves were found to be significantly higher in Santiago. (Walsh 1995). The results of this study suggested an associa- tion between these symptoms and PM-10 levels Lead. Most lead in ambient air is in the form of (Romieu 1992). fine particles with an aerodynamic diameter of In another study, daily ambient PM-10 con- less than 10 microns (PM-10). Ambient air also centrations in Santiago were correlated with to- contains organic lead compounds as gases. Mo- tal mortality as well as with mortality by tor vehicles are the major source of lead in am- population subgroups (all men, all women, and bient air in many Latin American urban centers, all people over sixty-four) and due to respira- where leaded gasoline is still used. Adults retain tory and cardiovascular disease (Ostro and oth- 20 to 60 percent of airborne particles, and chil- ers 1995). This study found that mortality among dren have a lung deposition rate that can be 2.7 the elderly and men due to respiratory and car- times higher than that of adults on a unit body diovascular diseases is more responsive to mass basis. The proportion of lead absorbed changes in ambient PM-10 concentrations than from the gastrointestinal tract is about 10 to 15 is total mortality. The results indicate that the percent for adults and up to 50 percent for chil- estimated impact of PM-10 on mortality in dren. Lead absorption increases in diets with low Santiago is consistent with that found in other levels of calcium, vitamin D, iron, and zinc. Lead studies in the United States. absorbed in the human body is distributed Long-term exposure to PM also was recently among bones, teeth, blood, and soft tissues. Most found to be related to decrements in lung func- of it is concentrated in bones (70 percent for tion or chronic respiratory disease. Lower lung children and 95 percent for adults). Unabsorbed function was associated with exposure to higher lead is excreted in the feces, and 50 to 60 per- ambient PM (as well as higher SO,, NO,, and cent of the absorbed lead is discharged through HC) concentrations based on tests performed the urinary tract. Organic lead is mainly ab- on nonsmokers from two different communities sorbed by the lungs through the respiratory tract in southern California (Detels, Tashkin, and and also through the skin (Romieu 1992). Based Sayre 1991). In another study conducted on non- on a review of epidemiological studies, an in- smokers in California, statistically significant re- crease of I p1g/M3 in lead concentrations in lationships between ambient concentrations of ambient air was associated with an increase in TSP and ozone were found with several respira- blood lead levels of 0.3 jpg/dl to 0.5 Vig/dl tory disease outcomes, including chronic bron- (Brunekreef 1986). chitis (Abbey and others 1993). Based on these Adverse effects of lead exposure have been results, Ostro (1994) estimated that a 10 plg/m3 observed in small children, women of reproduc- AirPollutants and TheirEffects 31 tive age, and male adults. Newborns and young ments of Mexican women of reproductive age, children are most vulnerable. Exposures to lev- the amount of time living in Mexico City was els of lead commonly encountered in urban found to be a strong determinant of bone lead environments constitute a significant hazard for levels. In another study conducted between 1990 children, especially those less than 6 years old. and 1992, the average blood lead level in school- Children with high levels of lead accumulated children decreased from 15.4 pig/dl to 10.2 pg/ in their baby teeth experience lower intelligence dl most likely in response to lower atmospheric quotients (IQs), short-term memory loss, read- lead concentrations resulting from the intro- ing and spelling underachievement, impairment duction of unleaded gasoline to the Mexican of visual motor function, poor perception inte- market. In 1995 the Pan American Health Or- gration, disruptive classroom behavior, and im- ganization (PAHO) and the Mexican authori- paired reaction time (USEPA 1990). Based on a ties estimated that about 800,000 women in the recent review of epidemiological studies, an in- reproductive age group in Mexico City had blood crease in children's blood lead level of 10 pg/dl lead levels above 15 pg/dl, and that each year was associated with a fall of 2.5 IQ points (CDC about 25,000 women would be delivering babies 1991). Adult women of reproductive age are also with blood lead above this level. Blood levels of a high risk group because lead levels of preg- different urban population groups in selected nant women are closely correlated with those of Latin American and Caribbean countries are newborns. People who are exposed to lead on presented in Table 2.7. the job, such as traffic police inhaling airborne Based on data in the literature, Ostro (1994) lead particles, also suffer adverse health effects. established relationships between lead levels in Among adults lead levels in blood are linked to air and effects on human health. Based on this an increased incidence of high blood pressure. analysis Ostro estimated that a 1 jig/ml increase No threshold level for the adverse health effects in ambient lead levels would cause a 0.975 IQ of lead has been identified. point decrement per child, twenty to sixty-five Several studies in Mexico City have demon- premature deaths and eighteen to fifty nonfatal strated correlations between exposure to lead heart attacks among 100,000 40- to 59-year-old in the environment and lead levels in the hu- males, and forty-five to ninety-eight hypertension man body. One study conducted on women of cases among 1,000 20- to 70-year-old males. reproductive age and on children less than 5 years old in two districts of Mexico City, one in- Chlorofluorocarbons. Exposure to increased UV- dustrial and the other residential, showed that B radiation is suspected to increase the risk of the blood lead level increased with age, the per- skin cancer and eye illnesses (especially cata- centage of children with blood lead levels ex- racts) and to adversely affect the immune sys- ceeding 10 pg/dl was higher in the industrial tem. During September and October 1991 a zone, and a significant correlation existed be- large human population in the southern tip of tween lead concentration in ambient air and in South America was exposed to WV-B as a result dust from the streets and from the home envi- of the Antarctic ozone hole. The affected loca- ronment (for example, windows, furniture, and tion with the greatest population density was dirt on children's hands; Romieu and others Punta Arenas in southern Chile (population of 1995). In a previous study, exposure to traffic- 110,000). A review of medical records by an in- related pollution was found to be the main de- ternational team of experts concluded a greater terminant of high levels of lead in blood in a frequency of dermatologic visits for warts. Eye sample of ninety children in Mexico City. Chil- and skin examinations on fishermen, shepherds, dren who lived on or close to high-traffic streets and hospital workers did not, however, reveal any had higher lead levels in blood than children association with increased UV-B exposure who resided in low-traffic areas. The major (USEPA 1995). source of lead was identified as the tetraethyl lead added to gasoline (Romieu, Weitzenfelcl, and Finkelman 1992). In addition, children from Quantifying the Health Effects of Air public schools were found to have higher blood Pollutants lead levels than children from private schools, indicating socioeconomic differences in lead Quantifying the health effects of air pollutants exposure. From bone lead content measure- enables researchers and policymakers to esti- 32 Chapter 2 Table 2.7 Blood lead levels of different urban population groups in selected Latin American and Caribbean countries srnnpe Range Average > 10Wg/ldl (Countr? PoIndation Age size (r'g/dl) (ug/dl) (perent) Brazil Adults 15-49 149 2.8-27.2 11.8 ± 5.2 75 Children 4-5 199 0.6-35.7 9.6 + 4.6 30 Chile Babies 1 200 0.5-18.0 4.3 + 1.8 5 Ecuador Children 7 64 17.0-54.0 28.8 100 Babies 0.1 27 6.0-20.0 14.4 60 Women Pregnant 83 - 18.4 60 Mexico Children <5 200 1.0-31.0 9.0 ± 5.8 28 Adults 15-55 200 1.0-39.0 9.7 ± 6.2 37 Adults 15-45 3,309 5.0-62.2 10.6 42 Trinidad and Tobago Women - 94 1.2-14.4 4.8 ± 2.0 2 Babies 0.1 94 0.0-8.7 3.4 ± 1.6 0 Uruguay Children 2-14 48 1.0-31.0 9.5 30 - Not available. Source: Lacasania and others 1996. mate the potential health benefits of pollution dT= (b)(POPi) (dA)(V,) control measures. These benefits are compared with the costs of pollution control measures to where b is the slope of the dose-response func- evaluate their feasibility or to determine the rela- tion. The dose-response function can be ex- tive merits of alternative measures. Health ben- pressed in absolute terms or in terms of efits from reduced levels of air pollution can be percentages: quantified using a four-step approach: (:Change in the number of affected persons per ' The change in health effect resulting from exposed population per year = (b) (change in the change in air pollution is estimated. ambient pollutant concentration) The relationship between air pollution and Percentage change in the number of affected health effects, called the dose-response persons in a year = (b') (change in ambient pol- ftunction, can be developed or obtained lutant concentration). from published epidemiological studies. * The population exposed and susceptible to Epidemiological studies have shown that a 1 air pollution (POP,) is determined. pg/m3 change in ambient PM-10 concentrations * The change in the air pollution level (dA) results in a change of 0.062 to 0.130 percent in from the current to a target level is speci- total mortality per year, with a central value of fied. The target level may be based on the 0.096 percent (that is, b' = 0.096; Ostro 1994). national orlocal airqualitystandards, WHO Thus a 10 pg/mi reduction in ambient PM-10 guidelines, USEPA standards, a percentage concentrations would be expected to lower to- change (for example, a 10 percent reduc- tal mortality by 0.96 percent a year. tion) in the ambient pollutant level, or a The coefficient b can be derived from coeffi- percentage reduction in the total emission cient b'. For mortality, the following relationship of a pollutant. relates ff to b: * The unit economic value of the effect (VY) , is developed for the reduced risk of health( effects. This could be based on willingness Using the above example and assuming a to pay, medical treatment costs, or lost of crude mortality rate of 0.007, the central value productive days and years (Ostro 1994). of coefficient b for mortality would be (0.096) (1/100) (0.007) = 6.72 x 10-' for a 1 JIg/mi The benefit of reduced air pollution (dT) can change in the ambient concentration of PM-10. be expressed mathematically as: This means that if the ambient PM-10 concen- Air Pollutants and Their Effects 33 tration is reduced by 1 p1g/rn3, about seven fewer fects of meteorological conditions from the ef- people per 1 million would be expected to die fects of air pollutant emissions. And because of prematurely. the high correlation among pollutants, it is of- The coefficient b or b' can be derived from ten not possible to distinguish the effects of in- epidemiological studies, which can be per- dividual pollutants (such as SO,, sulfates, and formed either through time-series or cross-sec- sulfuiric acid aerosols) from those of PM. Fur- tional analyses. Time-series analyses examine the thermore, time-series analyses need to take into correlation between fluctuations in pollution consideration normal weekly fluctuations in and health within a single population, and cross- medical services (LSHTM and St. GHMS 1995). sectional analyses make comparisons between The literature finds a strong association be- different populations with different population tween many air pollutants and negative health exposures. Time-series analyses address short- effects, mainly respiratory system problems. term (acute) health effects; cross-sectional anaIv- Based on a review of this information, Ostro ses pertain mostly to long-term health (1994) summarized the dose-response coeffi- consequences. Most epidemiological research cientbassociatedwiththehealthimpactsofPM- has been dominated by time-series analysis be- 10, SO, ozone, lead, and NO2 (Table 2.8). cause it avoids the problem of interpopulation Ostro's investigation did not include CO because confounding (for example, due to variations in little quantitative dose-response informationwas smoking prevalence) that afflicts cross-sectional available for this pollutant due to its rapid dissi- correlations, and it can be performed using his- pation in the environment. Further research is torical data sets for a single population, which needed to establish the magnitude of the asso- are readily available in industrial countries. ciations, possible interactions between pollutants In conducting time-series analyses, it is impor- and other variables, possible threshold (no ef- tant to remove term fluctuations in the outcome fect) levels, and the biological mechanisms of variables related to meteorological factors, tem- action (LSHTM and St. GHMS 1995). poral trends, and seasonal variations. For ex- Information on dose-response functions for ample, ozone concentrations are positively cities in developing countries-with the excep- correlated with high temperatures, number of tion of Santiago, Beijing, and Sao Paulo-is not sunshine hours, and aeroallergens; and high readily available in the literature. The dose-re- concentrations of NO9, other gaseous pollutants, sponse relationship between PM-10 concentra- and PM occur together in still weather. For this tions and mortality in Santiago was found to be reason time-series analyses need to identify sea- fairly consistent with other analyses (Ostro and sonal weather characteristics and infectious dis- others 1995). These studies also found a stron- ease epidemics, and account for increases in ger correlation betwveen PM-10 and mortality pollutant concentrations under certain meteo- among the elderly, people with heart and lung rological conditions (LSHTM and St. GHMS diseases, and men (Table 2.9). Despite similar- 1995). It is also worth noting that respiratory ity in these results, extreme caution is advised infections are more common in winter, when in extrapolating the reported dose-response re- levels of pollutants such as SO, and PM are gen- lationships to other situations. erally at their peak. Research findings suggest One reason these studies might not be appli- that at least part of the health effects attributed cable to other circumstances is that the dose- to fluctuations in ambient PM concentrations response function may not be valid beyond the may actually be due to spikes in respiratory viral range developed in the original study. Most epi- infections (Lamm, Hall, and Engel 1994). In demiological studies have assumed a linear dose- time-series analyses clusters of days with high response function and suggested that there is rates of death or hospital admission also need no conclusive evidence of threshold values for to be accounted for by the serial correlation of pollutant concentrations below which no adverse variables such as temperature. For example, low health effects occur. This implies that, even be- temperatures are associated with higher cardio- low the ambient air quality standards, air pollut- vascular mortality and, since one cold day tends ants might still have adverse health effects. to follow another, days with high cardiovascular Another concern is associated with the impact mortality tend to cluster as well. Although of populations' baseline conditions on the dose- autoregressive models are useful, in most in- response functions. Differences between the stances it is difficult to separate the health ef- baseline conditions of two populations (that is, 34 Chtapter 2 Table 2.8 Ranges and central estinates for dose-response coefficients (change in risk of health effect per exposed person for one unit change in ambient polluttant concentration in a year) Polluotants and aSsociated heeIlth effeas Ranges of estimates Gntral estimaWt PM-10 Premature mortality 4.47 x 10-6 - 9.10 x 10-6 6.72 x 10' Hospital admissions for respiratory ililnesses 6.57 x 104 - 15.60 x 10' 12.00 x 10e Emergency room visits 12.83 x 105 - 34.25 x 10- 23.54 x 1e Restricted activity days 4.04 x 10-2 - 9.03 x 10-2 5.75 x 10-2 Lower respiratory illnesses in childreni (per child) 8.00 x 104 - 23.80 x 104 16.90 x 104 Asthma attacks (per asthmatic) 1.63 x 10-2 - 27.30 x 10-2 3.26 x 10-2 Respiratory symptoms 9.10 x 10-2 - 27.40 x 10- 18.30 x 10Y Chrornic bronchitis (above age 25) 3.06 x 10- - 9.18 x 10- 6.12 x 10- S02 Respiratory symptoms (per child) 1.00 x 10-5 - 2.62 x 10-' 1.81 x 10-5 Chest discomforts (per adults) 5.00 x 10-3 - 15.00 x 10-3 10.00 x 10-3 Ozone (1-hour max.) Hospital admissions for respiratory illnesses 3.80 x 10-3 _ 12.10 x 10-3 7.70 x 10- Minor restrictions in activity 17.00 - 51.00 34.00 Days with respiratory malaise symptoms 28.11 - 96.60 54.75 Eye irritations 23.40 - 29.90 26.60 Asthma exacerbations (per asthmatic) 38.69 - 189.80 68.44 Lead IQ decrement (per child) 0.975 Hypertension (per male aged 20-70) 4.48 x 10 - 9.78 x 10-2 7.26 x IT2 Premature mortality (per male aged 40-59) 2.00 x 10-4 - 6.50 x 10-4 3.50 x 104 Nonfatal heart attacks (per male age(d 40-59) 1.80 X 10-4 - 5.00 x 10-4 3.40 x I0-4 NO, Respiratory symptoms 6.02 - 14.42 10.22 ANote: The unit is pg/sn3 for PM-10, SO2, and lead and ppm for ozone and NO2. Source: Ostro 1994. between those considered in the original study given ambient concentration of an air pol- and those in the population to which results are lutant, populations in Latin America and extrapolated) may result in different dose-re- the Caribbean are at a greater health risk sponse functions. Most of the dose-response than populations in the United States be- functions have been established through epide- cause they spend more time outdoors. Thus miological studies conducted in industrial coun- extrapolation of a dose-response function tries-particttlarly in the United States and to a developed for a North American popula- lesser extent in Britain, Germany, and Canada- tion to a Latin American or Caribbean where the baseline conditions are somewhat dif- population would likely underestimate ferent from those of Latin American and health effects. Caribbean countries. The following baseline con- * Differences in exposure to pollutants from ditions are especially important: other sources. For example, exposure to cooking fires and space heating in devel- * Differences in lifestyles that affect exposure oping countries is a risk factor for a wide to outdoor air pollutants. Populations that range of health outcomes, especially for spend more time outdoors and are more chronic lung diseases in adults (particularly exposed to air pollutants are at a greater women) and acute respiratory diseases in health risk than others. For example, for a children (Chen and others 1990; Smith Air Pollutants and TheirEffects 35 Table 2.9 Alternative mortality subgroups from different studies (percentage change in the iiumber of affected persons in a vear for 10 jig/in' change of PM-10) Utah Birmingham Plhiladelphia Health effert San tiagr'l Sfi l'ala' Beijing Valey Alabama' Pennsylvania Total mortality 0.7 0.7g 1.6 1.1 1.3 Total mortality (over 65 years) 1.0 1.3 1.8 Respiratory disease 1.3 4.3 Cardiovascular disease 0.8 2.0 1.7 1.8 Chronic obstructive pulmonary disease 1.8 3.4 Chrornic lung disease 1.6 Pneumonia 2.0 All other 0.5 0.6 Male 1.0 Female 0.5 NVot,: Blank spaces inidicate that no data are available. a. Ostro and others 1995. b. Saldiva and others 1994. c. Xii and others 1994. d. Pope, Schwartz, and Ransom 1992. e. Schwartz 1993. f. Schwartz and Dockery 1992a. g. The coefficient is associated with the stmnmer months. 1993). Dockery and others (1993) fomnd a young and old are more affected by air higher mortality risk from air pollution pollution. For example, in Chile the popu- among people with an occupational expo- lation above 65 years of age is 6 percent of sure to respiratory hazards. Human expo- the total population; in the United States sure to lead occurs not only through this share is 12.9 percent. ambient air, but also through food, water. * Differences in the cause of death. Extrapo- and, in children, soil and dust. lating the effect of air pollution on total • Differences in smoking habits. Dockery and death rates from one study in the United others (1993) found a higher mortality risk States to a population in a developing coun- from air pollution in smokers than in non- try would yield incorrect results if the dis- smokers. tribution of causes of death differs greatly * Differences in geographic and climatic con- between the two populations. For example, ditions-such as altitude, extreme temper-a- half the deaths in the United States are ture, and htumidity-that may exacerbate caused by cardiovascular disease or respi- the adverse health effects of air pollution. ratory illness, whereas less than 20 percent * Differences in nutritional status and defi- of deaths in Delhi, India are attributable ciencies in vitamins C and E, which weaken to these causes. Even if the people in Delhi the body's defenses against air pollutants. reacted to a given change in ambient pol- * Differences in access to medical care that lutant concentration the same way as in the influence the relationship between pollu- United States, the effect of the change on tion levels and hospital admissions or emer- total mortality would be lower (Cropper gency room visits. Lack of appropriate and Simon 1996). medical surveillance and treatment can * Differences in public awareness of air pol- increase the susceptibility of people at r isk lution. Exposure to outdoor air pollutants (such as the elderly and people with previ- can be reduced through preventive mea- ous respiratory diseases). sures. For example, billboards in some * Differences in age distributions. Popula- Latin American cities (such as Mexico City tions with a higher proportion of very and Sao Paulo) and television, radio, and 36 Ciapter 2 newspapers are used to inform the public probably easier to apply in developing countries. about ambient air quality. This approach does not take into consideration * Differences in PM composition or in the personal discomfort and suffering, however, and mixture of air pollutants with potential provides a lower estimate of health values than synergetic effects. the willingness to pay approach (Cropper and Oates 1992). Once the health response to a given change in ambient pollution is estimated, valuation tech- niques are used to quantify the benefits of re- Environmental Effects of Air Pollutants duced air pollution on human life and health for a selected population. The human capital The environmental effects of pollutant emissions and willingness to pay approaches are the two from motor vehicles include global climate most commonly used techniques for valuing pre- changes from greenhouse gases, acidification of mature death. The human capital approach is soil and surface waters, adverse effects on some based on the discounted flow of income from plant (including crop) and animal species, and the person during the years he fails to live. This damage to buildings and structures. approach is, however, less useful for valuing per- sons with no income (children, retired people, Nitrogen oxides. Strong evidence suggests that housewives). The willingness to pay approach NO, can harm the environment through nitrate measures people's willingness to pay for reduc- formation and acidification of surface waters. tions in their risk of dying. This may be estimated The nitrate concentration of ice in Greenland through the wvage differentials between riskier has doubled since the pre-1900 era (Herron and saferjobs (the underlying assumption is that 1982). The nitrate concentration of precipita- risk-averting people move from higher-paying, tion in Europe increased steadily between 1955 risky jobs to lower-paying, safer jobs). Another and 1979 (Faiz and others 1990). In 305 lakes in way of estimating willingness to pay involves ask- Norway the nitrate concentration doubled be- ing people how much they would pay to reduce tween 1974 and 1986 and was apparently associ- their risk of dying. Because it is hypothetical, ated with a decrease in fish populations however, this approach might lead to higher es- (Derwent 1988). A threefold increase in the ni- timates. Of the two techniques, willingness to trate aerosol concentration was detected in am- pay may be more difficult to apply in develop- bient air at Chilton, U.K. (Watkins 1991). ing countries because of its extensive data re- Although not quantified, NO. is also believed to quirements. affect corrosion of materials such as buildings Because valuation of premature mortality in- and other humanmade structures. N20, as a volves putting a monetary value on human life, greenhouse gas, contributes to global climate it may raise moral and ethical issues. This should warming by absorbing infrared radiation. not be a concern, however, because the focus of valuation is not a specific person, but a statisti- Ozone. Ozone damages plants and vegetation. cal entity in a selected population. In addition, The degree of damage is affected by ozone con- these estimates are intended to evaluate an air centration and extent of exposure. Greater dam- pollution management project or rank air pol- age is also observed with higher light intensity lution management alternatives within a given and humidity. Tobacco, a very sensitive plant, can urban center or country, not across countries. It be injured by exposure to an hourly ozone con- should also be noted that the human capital and centration of 80 mg/m3 or higher. Other sensi- willingness to pay approaches may not be appro- tive plants, including beans, spinach, and clover priate for cultures where perceptions of life, may be injured by exposure to 200 mg/mi of health, and death more strongly reflect nonma- ozone for several hours. Ozone is believed to be terial values. the major contributor to forest damage in the Morbidity effects due to air pollution can be United States (for example, the pines of San valued by gauging people's willingness to pay to Bernardino Forest in southern California), Ger- avert the disease or the cost of illness approach, many, and Central Europe (Watkins 1991). which involves estimating of medical expenses for disease prevention and treatment and lost Sulfur dioxide. Environmental effects of SO2 re- wages. Of the two approaches, cost of illness is sult either directly or indirectly from acid depo- Air Pollutants and TheirEffects 37 sition. SO2 at concentrations as low as 800 Jg/ century this warming is expected to raise the m' harms crops such as wheat, oat, barley, and Earth's temperature rise by about 3°C, and could cotton. SO2, along with ozone and NO., has dam- induce changes in rainfall patterns, shifts in cli- aged forests in Europe (for example, in the matic zones, and a rise in sea levels. Czech Republic). It also has caused the acidifi- cation of soils and lakes in Europe (in Finland) and North America and has damaged marble Ambient Air Quality Standards structures and monuments (in Athens) by form- and Monitoring ing calcium sulfate (gypsum). Ambient air quality standards. Air quality standards Particulate mafter. Environmental effects of PM are set to protect society and the environment include soiling and degradation of visibility. from the harmful effects of air pollutants. They Large-size PM falls out of the atmosphere based are designed to achieve a given desirable level on its settling characteristics and atmospheric of air quality, and frequently serve as a reference conditions (especially wind). PM that settles out base for other standards such as emission stan- of the air adversely affects people's welfare by dards or fuel quality standards (UN 1987). As accumulatingvirtually anywhere-on buildings. such, they do not take into account the costs or windows, cars, laundry, even inside houses-leav- benefits associated with these pollutants. ing dirty deposits, requiring frequent cleaning, Ambient air quality standards are of two types: and damaging some materials. primary and secondary. Primary ambient air PM which remains suspended in the air is of quality standards are established to protect the smaller size and includes PM emitted by motor most vulnerable groups of population, namely vehicles as well as PM formed in the atmosphere the young, the old, and people in poor health. such as sulfate or nitrate aerosols. PM in the at- Short-term standards and guidelines are estab- mosphere absorbs and disperses light, and hence lished to control acute effects that result when reduces visibility. The presence of nitrates in PM- high levels of pollution persist for short periods. 2.5 also darkens the color of the sky. The visibil- Typical short-term standards are for 1-hour, 8- ity reduction resulting from urban pollution is hour, and 24-hour averages of pollutant concen- readily noticeable in Latin American urban cen- trations. Long-term standards and guidelines are ters such as Mexico City and Santiago. Because designed to protect human health from regular PM-2.5 can travel long distances, reduction of exposure to high levels of pollution over a long visibility is also a regional problem. For example, period of time (typically one year; WHO/UNEP 20 percent of visibility reduction in Rocky Moun- 1992). Secondary air quality standards are estab- tain National Park (United States) is attributed lished for nonhealth impacts such as those in- to pollution generated in Los Angeles, about 600 volving soil, crops, vegetation, human-made kilometers away. materials, animals, wildlife, atmospheric visibil- ity, property damage, transportation hazards, Chlorofluorocarbons. Though injurious UV-B ef- and effects on the economy and personal com- fects have been documented on individual spe- fort (Cohn and McVoy 1982). cies within marine ecosystems, the nature and The United States was the first country to es- extent of ecosystem responses to UV stress are tablish ambient air quality standards. The 1970 not well understood. The effects of UV-B radia- U.S. Clean AirActAmendments of the 1967 U.S. tion on marine ecosystems and crops (such as Air Quality Act established national primary and rice production) are now being investigated secondary ambient air quality standards. The act (USEPA 1995). In addition, CFCs and halons also gave states the option of establishing their contribute to global climate warrning. own ambient quality standards, provided they were equal to or more stringent than the na- Carbon dioxide. C02 forms an iinsulating blanket tional standards. In 1978 lead was added to the around the Earth that prevents the escape of list of restricted pollutants, and in 1979 the ozone heat. This greenhouse effect causes an increase standard was made more stringent. In 1983 the in the earth's temperature. About 50 percent of USEPA revoked the national ambient air quality global warming is caused by CO,; the rest is standards for HC because this pollutant was caused by methane, ground-level ozone, CFCs, deemed safe at or near ambient levels and no and nitrous oxide. By the end of the twenty-first consistent quantitative relationship was found 38 Chapter 2 nationwide between ambient ozone concentra- for comparison. Ambient air quality standards tions and hydrocarbon air quality levels. The (STDRF) are expressed in milligrams per cubic USEPA indicated, however, that HC shotild con- meter (mg/m3) for CO and micrograms per tinue to be controlled because of their contri- cubic meter (jg/m3) for all other pollutants at bution to ozone formation and the resultant reference conditions of 25°C temperature and health and welfare effects of ozone. In 1987 the 760 mm Hg pressure. For an urban center with ambient standard for TSP was abandoned and a t°C temperature and p mm Hg pressure, the one for PM-10 was established. With the revision air quality standard (STD) can be obtained of the National Ambient Air Qualitv Standards through the following relationship: in 1989 secondary standards were abandoned for all pollutants except SO3. Ambient air qual- STD= (STDRF)(p) ( 298)/[(760)(273 + t)]. ity standards for PM-10 are under review. The Ambient air quality standards under the ref- USEPA has proposed new standards for PM-2.5 erence conditions can be also expressed in terms (12.5 pg/m3 to 20 pg/m3 on an annual average of parts of pollutant per million parts of air basis). These standards are estimated to save (ppm) using the following relationship: about 20,000 lives (especially among the elderly STDRF (expressed in ppm) = (STDRFexpressed and those with existing heart and lung diseases) in jig/m3) (24,500) (101)/ (MWM and result in fewer hospital admissions (by over 9,000 a year) and reduced risk of respiratory dis- where MW is the molecular weight of the pol- eases (over 60,000 fewer incidences a year of lutant. chronic bronchitis, over 250,000 fewer inci- dences of asthma attacks, and over 250,000 cases Ambient air quality monitoring. Ambient air qual- of respiratory symptoms). The new standards are ity monitoring involves measuring pollutants to also expected to improve haze and reduce soil- determine ambient air concentrations. During ing and material damage effects. A final deci- the 1960s, following identification of the adverse sion on the new standards will be taken byJuly human health effects of certain air pollutants 1997. (CO, NO,, ozone, SO2, TSP, lead), many indus- Many other countries also have established trial countries adopted national ambient air their own ambient air quality standards, includ- quality standards and initiated ambient air qual- ing most Latin American countries. Primary air ity monitoring. Most of these monitoring efforts quality standards have been established in every focused on SO2 and TSP measurements. During South American country except Guyana, Para- the 1970s, with the emergence of motor vehicles guiay, Pemi, Suriname, and Umiguay. No Central as the main source of air pollution in urban ar- American or Caribbean country has such stan- eas, air quality monitoring incorporated other dards, however. Ambient standards are being traffic-related pollutants such as CO, NO2, and prepared in Guatemala, Paraguay, Peru, and lead. During this period and mostly in the 1980s Uruguay. Primary air quality standards have been urban air quality monitoring was initiated in established mostly through national legislation some developing countries, especially in Latin and in some cases through regional (provincial) America and Asia. In recent years greater em- or local legislation. These standards are some- phasis has also been placed on the monitoring times complemented by additional legislation on of photochemical oxidants (especially ozone) pollutant standards that trigger certain actions and volatile organic chemicals (VOCs) and their by society. For example, attention, alert, emer- precursors as well as PM-10 (instead of TSP, gency, and critical levels of air pollutants (such as which has a limited value for health effect as- CO, NO2, ozone, SO2, TSP, and PM-10) have been sessments; WHO/UNEP 1992). established in Sao Paulo. Brazil and Chile are the Ambient air quality monitoring in urban ar- only Latin American and Caribbean cotntries that eas may have a number of objectives. One is to have established secondary ambient air quality generate information on the spatial and tempo- standards. ral distribution of air pollution in urban areas. Ambient air quality standards for various Latin Monitoring data are then compared against air American countries are presented in Annex B quality standards to identify potential risks to to this chapter for CO, NO2, ozone, SO,. TSP, human health or the environment. Data indi- PM-10, and lead. Related information on U.S. cating high pollutant concentrations in certain standards and WHO guidelines are also included parts of an urban area (such as a business dis- Air Pollutants and TheirEffects 39 trict) or during certain periods of the day (peak by two mobile stations in areas not covered by morning hours) or year (certain months) en- stationary stations. The monitoring data in these able policymakers to take the necessary measures stations are recorded on magnetic tapes and aimed at reducing pollution at these locations subsequently sent to a central computer system. or during these periods. Monitoring data also Fully automated monitoring with data transmission help policyrnakers evaluate the effectiveness of includes all the above automatic features of the control measures they implement. In addi- monitoring as well as transmission of the elec- tion, ambient air quality monitoring can be per- tronic output of the instrument, possibly by a formed to inform the public on short notice telephone line, to a computer at a central con- about air quality, especially when pollutant con- trol room where data are stored and processed centrations reach or are about to reach high lev- (Rossano and Thielke 1976). In Latin America els. Such warnings allow the public to take the and the Caribbean such systems are used in Sao necessary preventive measures against health- Paulo, Mexico City, and Santiago. The monitor- related risks. Ambient air quality monitoring, if ing network in Sao Paulo includes twenty-five conducted over many years, can also be used to stationary stations that generate monitoring data generate pollutant trends as an input to medium- for PM-10, S02, NO2, CO, and HC. The data are or long-term policy decisions for air quality man- immediately sent by telephone to a central sta- agement. tion for computer processing. The air quality Ambient air quality monitoring involves sam- information is disseminated to the public using pling of pollutants, analysis of the samples, data billboards on major roads and through press collection, and data transmission and control. and communication agencies. The monitoring It can be performed using any of a number of network in Mexico City is similar to that in Sao systems. Manually operated samplingrefers to sam- Paulo. This network includes thirty-two stations plingwith aninstrumentforwhich the start and that monitor ozone, SO2, NO,, CO, and PM-10. end of the sampling period are actuated manu- In Santiago the monitoring network consists of ally. Analyses of samples may be conducted five fully automated stations measuring ambient manually or automatically. Semiautomaticsampling concentrations of CO, SO2, NO2, ozone, HC, and involves a sampling instrument equipped with PM-10. automatic sampling capabilities, but which re- The design of an urban air quality monitor- quires manual removal of samples. Analyses of ing program should reflect its objectives and samples may be conducted manually or auto- include decisions on sampling, measurement, matically. Such systems have been used in Latin data collection, and data transmission issues. The America for many years. For example, a six-sta- design should also consider the combined ef- tion semiautomatic monitoring system has been fects of pollutant emission source configurations operated in Santiago since 1981. However, these and the meteorological and topographical fea- stations have lost their importance since a fully tures of each urban area. For this reason pollu- automated monitoring network was installed. In tion monitoring should be accompanied by Sao Paulo and Mexico City semiautomatic sam- atmospheric monitoring that includes such pa- pling is conducted to complement the more rameters as wind speed, wind direction, tempera- advanced systems described below. In Sio Pauilo ture, and humidity. Air monitoring of an entire seven of eighteen stations are used to monitor urban area requires selection of the station lo- SO2; the remaining eleven are used for PM-1 0. In cations to cover the varying levels of air pollu- Mexico City the semiautomatic system consists tion across the urban area. However, fixed of nineteen stations monitoring TSP and PM-10. monitoring stations only provide air quality in- Automatic monitoring involves sampling and formation for their location and are expensive analysis performed by an instrument that pro- to install in every part of a city. Because resources duces printed data sheets, magnetic tape recorcl- are limited, installations should be prioritized ings, and charts without human intervention based on the significance of air pollution in except for regular maintenance. Some urban terms of human health risks and ease of mea- areas in Latin America use automatic monitor- surement. In some cases fixed monitoring sta- ing systems. For example, Santiago's six auto- tions may be complemented by mobile stations matic monitoring stations measure ambient to spatially extend the monitoring effort. levels of CO, SO2, NO2, ozone, HC, and PM-10. Every sampling station should be located to In Sao Paulo automatic monitoring is conducted ensure that the information it collects is repre- 40 Chapter 2 sentative of its surrounding area. The issues re- in Sao Paulo the designation for the air quality lated to pollutant measurement should include categories consists of "bad" (index over 200) in types and concentrations of pollutants, required the "attention" level, "very bad" (index over 300) averaging times for sampling, presence of other in the "alert" level, and "critical" (index over 400) pollutants that may cause interference with the in the "emergency" level. In addition, air qual- monitor, and availability and qualifications of ity data are generally disseminated to the press. available personnel and laboratories. In addition, The most comprehensive efforts to monitor in selecting the monitoring system, factors such ambient air quality have occurred in Sao Paulo, as functional characteristics (measuring ranges, Mexico City, and Santiago. Ambient air quality accuracy, repeatability), operational character- monitoring in many other cities has been ham- istics (complexity in operation and response to pered by inadequate financial resource commit- external conditions such as temperature and ments and lack of technical resources and spare humidity), requirements for technical expertise parts. For example, until 1982 the monitoring and maintenance, availability of spare parts, and effort conducted by the Municipality of Buenos physical dimensions should be considered. Aires included twelve stations for gaseous pol- In terms of data collection, continuous moni- lutants and twenty-one stations for TSP measure- toring instruments that present data in bar charts ments. This effort was significantly curtailed, have the advantage of providing a permanent however, following budget cuts and a reduction and visual record of information, but they re- in staff. The municipality's current effort in- quire considerable staff to maintain. Fully auto- cludes monitoring of three pollutants (NO, NO2, mated systems can communicate the measured and SO2) at one stationary station and two pol- data from each monitoring station to the cen- lutants (TSP and lead) by three mobile stations. tral control room. An electronic map in the con- In Belo Horizonte the air quality monitoring sta- trol room allows the display of monitoring tions established in 1984 experienced opera- stations where ambient pollutant concentrations tional difficulties and were shut down in 1988 exceed a specified level. Such systems help pro- because of a lack of funding. A second monitor- vide short-term warnings to the public about high ing effort initiated in 1991 was suspended in 1992 concentrations of pollutants at specific locations. for the same reason. In Rio de Janeiro an auto- Air quality data can be disseminated to the pub- matic monitoring station provided by the WHO lic through electronic billboards located in in 1975 could not be fully used because of tech- publicly visible areas or to radio and television nical difficulties and a lack of spare parts. The stations. In Latin America such systems are avail- operation of another automatic monitoring able in Sao Paulo and Mexico City. In these cit- station installed in 1986 under the Brazilian- ies real-time air quality data are presented using Japanese Technical Cooperation Project also was indices and associated qualifiers. For example, hampered by a lack of spare parts. ANNEX A ESTIMATED EMISSION FACTORS FOR U.S. VEHICLES Table A.1 Estimated emission factors for U.S. gasoline-fueled passenger cars with different emission control technologies (grams per kilometer) Fuel consum)tlion (liters per 1 00 Type of control (.0 Mlethanea NMHC NOXa N20 C02 kilomneters) Advanced three-way catalyst control 8.4 Exhaust 6.20 0.04 0.38 0.52 0.019 200 Evaporative 0.09 Running loss 0.16 Resting 0.04 Total emissions 6.20 0.04 0.67 0.52 0.019 200 Early three-way catalyst control 10.6 Exhaust 6.86 0.05 0.43 0.66 0.046 254 Evaporative 0.14 Running loss 0.16 Resting 0.06 Total emissions 6.86 0.05 0.79 0.66 0.046 254 Oxidation catalyst control 16.7 Exhaust 22.37 0.10 1.87 1.84 0.027 399 Evaporative 0.39 Running loss 0.17 Resting 0.06 Total emissions 22.37 0.10 2.49 1.84 0.027 399 Non-catalyst control 16.7 Exhaust 27.7 0.15 2.16 2.04 0.005 399 Evaporative 0.70 Running loss 0.17 Resting 0.06 Total emissions 27.7 0.15 3.09 2.04 0.005 399 Uncontrolled 16.7 Exhaust 42.67 0.19 3.38 2.7 0.005 399 Evaporative 1.24 Running loss 0.94 Resting 0.06 Total emissions 42.67 0.19 5.62 2.7 0.005 399 a. Estimated using the USEPA's MOBILE5a model for the following conditions: temperature of 24'C, vehicle speed of 31 kilometers an hour, gasoline RVP of 9 psi, and no inspection and mainteniance program in place. Sourre: Faiz, Weaver, and Walsh 1996. 41 42 Chapter 2 Table A.2 Estimated emission factors for U.S. gasoline-fueled medium-duty trucks with different emission control technologies (grams per kilometer) Fuel con sumption (lit-r per 1(0 Type of control (X'P Alethane3 NMHfO NOIa N/) CO2 kilometers) Three-way catalyst control 34.5 Exhaust 10.2 0.12 0.83 2.49 0.006 832 Evaporative 0.38 Running loss 0.17 Resting 0.04 Total emissions 10.2 0.12 1.42 2.49 0.006 832 Refueling 0.24 Non-catalyst control 35.7 Exhaust 47.61 0.21 2.55 3.46 0.006 843 Evaporative 2.16 Running loss 0.94 Resting 0.08 Total emissions 47.61 0.21 5.73 3.46 0.006 843 Refueling 0.25 Uncontrolled 50.0 Exhaust 169.13 0.44 13.56 5.71 0.009 1,165 Evaporative 3.93 Running loss 0.94 Resting 0.08 Total emissions 169.13 0.44 18.51 5.71 0.009 1,165 Refueling 0.32 a. Estimated using the USEPA's MOBILE5a model for the following conditions: temperature of 24'C, vehicle speed of 31 kilometers an hour, gasoline RVP of 9 psi, and no inispection and maintenance program in place. Sourre: Faiz, Weaver, and Walsh 1996. Air Pollutants and Their Effects: Annex A 43 Table A.3 Emission and fuel consumption factors for U.S. diesel-fueled passenger cars and light-duty trucks with different emission control technologies (grams per kilometer) Fuel consumption (liters per 100 Vehil& type (CO HG NO, (XJ2 kilometers) Passenger cars Advanced control 0.83 0.27 0.63 258 9.4 Moderate control 0.83 0.27 0.90 403 14.7 Uncontrolled 0.99 0.47 0.99 537 19.6 Light-duty trucks Advanced control 0.94 0.39 0.73 358 13.0 Moderate control 0.94 0.39 1.01 537 19.6 Uncontrolled 1.52 0.77 1.37 559 23.3 Source: MOBILE5a estimates adapted from Faiz, Weaver, and Walsh 1996. Table AA4 Emission and fuel consumption factors for U.S. heavy-duty diesel-fueled trucks and buses with different emission control technologies (grams per kilometer) I'tel consumption (liters per I 00 lVehicle type CO H(. NQ, PM (202 kilometers) U.S. heavy-duty diesel trucks Advanced control 6.33 1.32 5.09 982 35.7 Moderate control 7.24 1.72 11.56 991 35.7 Uncontrolled 7.31 2.52 15.55 1,249 45.5 U.S. 1984 measurements Single-axle tractors 3.75 1.94 9.37 1.07 1,056 Double-axle tractors 7.19 1.74 17.0 1.47 1,464 Buses 27.40 1.71 12.40 2.46 1,233 New York City vehicles Medium-heavy trucks 2.84 23.28 2.46 53.8 Transit buses 5.22 34.89 2.66 80.7 Sourre: MOBILE5a estimates adapted from Faiz, Weaver, and Walsh 1996. L ANNEX B AMBIENT AIR QALITY STANDARDS AND WORLD HEALTH ORGANIZATION GUIDELINES Table B.1 Air quality standards and WHO guideline for carbon monoxide (milligrams per cubic meter) 7imne-wreiglted average Other Country I hour 8 hours 24 hotIrs aver7Agng times Argentina 57.3 11.5 Alert level 114.5 17.2 Alarn level 137.4 34.4 Emergency level 171.8 57.3 City of Buenos Aires 3.0 15.0 (20 minutes) Province of Buenos Aires 45.8 17.2 9.2 Bolivia 30.0 10.0 Brazil Primary standards 40.0 10.3 Secondary standards 40.0) 10.3 Sao Paulo Attention level 17.2 Alert level 34.4 Emergency level 45.8 Critical level 57.3 Chile 40.0 10.3 Santiagoa Good level 10.3 Regular level 21.8 Bad level 34.4 Critical level 45.8 Dangerous level 57.3 Colombia 5.t) 15.0 Santafe de Bogota 38.4k' 11.5h Ecuador 40.0 10.0 Mexico 12.6 Mexico Citya Satisfactory level 12.6 Unsatisfactory level 25.2 Bad level 35.5 Very bad level 57.3 Peru 40.0' 20.0' Venezuela 10.0(1 40.0e United States 40.0 10.0 W%HO 30.0 10.0 60 (30 minutes)f 100 (15 minutes)f Note: 1 ppm = 1.145 milligrams per cubic meter (mg/nt'). A blank space indicates that no standard was established. a. Time-wveighted averages indicate maximum values for each category. b. This standard reflects 13"C temperatire and 560 mm Hg pressure for Santafe de Bogota. c. Proposed standards. d. 50 percentile (of the 8-hour mean, 50 percent must he below the indicated value). e. 99.5 percentile (of the 8-hour mean, 99.5 percent imist be below the iidicated valie). f. The 15-minute and 30-minute values are established to maintain the carboxyhemoglobit level in blood less than 2.5 to 3.0 percent inl non-smoking populations. Sourre: WHO/UNEP 1992; personal comnitications with technical experts from different countries. 45 46 Chapter 2 Table B.2 Air quality standards and WHO guideline for nitrogen dioxide (micrograms per cubic meter) Time-weighted avernge Other (Lountry I hoir f hours 24 hours averaging times Argentiiaa 846 Alert level 1.128 282 Alarm level 2.256 564 Emergency level 752 City of Buenos Aires 100 400 (20 minutes) Province of Buenos Aires 376 188 94 Bolivia 400 150 Brazil Primary standards 320 190 Secondary standards 320 100 Sao Paulo Attention level 1,130 Alert level 2.260 Emergency level 3.000 Critical level 3,750 Chile 100 Santiagob Good level 470 Regular level 1,290 Bad level 2,110 Critical level 2,930 Dangerous level 3,750 Colombia 100 Santafe de Bogota 77C Ecuador 100 Mexico 395 Mexico Cityb Satisfactory level 395 Unsatisfactory level 1.241 Bad level 2,068 Very bad level 3.760 Peru 200d Venezuela ioo" 300" United States 100 WHOg 400 1501" Note: 1 ppm = 1,880 micrograms per cubic meter (pg/lmn). A blank space indicates that no standard was established. a. All standards in Argentina are for NO, (expressed as NO2). b. Time-weighted averages indicate maximutimi valies for each category. c. This standard reflects 13NC temnperature and( 560 mm Hg pressure for Santafe de Bogota. d. Proposed standards. e. 50 percentile (of the 24-hour meat,, 50 percent must be below the indicated valie). f. 95 percentile (of the 24-hour mean, 95 percenit mnist be below the indicated valie). g. WHO guidelines to protect vegetation (in the presence of SO2 and ozone not exceeding 30 jig/mr and 60 hg/rni, respectively) are 30 1gg/m' for the anntial mean and( 95 jkg/mi for the 4-hour mean. h. The 24-hour WHO guideline is establishe(d to protect against chroniic exposure, and the 1-houtr WHO guideline is designed to provide a margin of protection for- liung fiuctioni in asthmatic people from short-term exposures. Somrre: WHO/UNEP 1992; personal communications with technical experts from different countries. Air Pollutants and Their Effects: Annex B 47 Table B.3 Air quality standards and WHO guideline for ozone (micrograms per cubic meter) Tine-roeighted Other (Counlhy avemge (1 lnhoir) averaging tlies Argentina 20()0 Alert level 30() Alarm level 400 Emergency level 800 City of Buenos Aires 30 (24 hours) 100 (20 minutes) Bolivia 236i Brazil Primary standard 160 Secondary standard 160 Sao Paulo Attention level 400 Alert level 800 Emergency level 1,000 Critical level 1,200 Chile 160 Santiagoa Good level 16i Regular level 470 Bad level 780 Critical level 1,090 Dangerous level 1,400 Colombia 170 Santafe de Bogota 131t) Ecuador 200 Mexico 220 Mexico Citya Satisfactory level 220 Unsatisfactory level 460 Bad level 700 Very bad level 1,200 Peru 400c (30 minutes) 200c (8 hours) Venezuela 240'I ------------------------------------------------------- United States 235 WHO 150-200 100-120 (8 hours)' Note: 1 ppm = 2,000 micrograms per cubic meter (jig/iil). A blank space indicates that no standard was established. a. Time-weighted averages indicate maximum values for each category. b. This standard reflects 13°C temperature and 560 mm Hg pressure for Santafe de Bogoti. c. Proposed standards. d. 99.98 percentile (of the 1-hour mean, 99.98 percenit Ilutlst he below the indicated value). e. The 8-hour WHO guideline is intended to lesseni the potenitial for adverse acute and chronic effects and to provide an additional margin of protection. Soure: WH(O/UNEP 1992; personal communications witil teclihical experts from different countries. 48 Chapter 2 Table BA4 Air quality standards and WHIO guideline for sulfur dioxide (micrograms per cubic meter) Tim*-zveighted averageote Country I hou(r 8 houwn 24 hours I year averaiging times Argentina 70 (30 days) Alert level 2,6001 780 Alarm level 13,000 Emergency level 26,000) City of Buenos Aires 500 (20 minutes) 70 (30 days) Province of Buenos Aires 780 78 Bolivia 365 80 Brazil Primary standards 365 80 Secondary standards 100 40 Sio Paulo Attention level 800 Alert level 1,600 Emergency level 2,100 Critical level 2,620 Chile 365 80 Santiago( secondary standards) North zone 700 260 60 South zone 1,000 365 80 Santiagoa Good level 365 Regular level 929 Bad level 1,493 Critical level 2,056 Dangerous level 2,620 Colombia 400 100 1,500 (3 hours) Santaf6 de Bogota 307' 77) 1,152b (3 hours) Ecuador 400 80 1,500 (3 hours) Mexico 338 78 Mexico City'a Satisfactory level 338 Unsatisfactory level 910 Bad level 1,456 Very bad level 2,600 Peru 858c 172' Venezuela 80"i 200e 250F 3659 United States 365 80 1,300 (3 hiouirs) WHII 35Oi 100-150 40-60 500 (10 minutes)1 Note: I ppm = 2,600 micrograms per ctubic isieter (pg/mn5). A blank space inidicates that no standard was established. a. Time-weighted averages indicate maximumin valuies for each category. b. This standard reflects 13"C temperatuire and( 560 mm Hg pressuire for Sanitaff de Bogotd. c. Proposed standards. d. 50 percentile (of thie 24-houir meals, 50 perceuit muiist be below the inidicated value). e. 95 percentile (of the 24-houir mean, 95 per-cent miiust be below thle inidicated v'alue). f. 98 percentile (of tlse 24-houir mean, 98 percent isslist be below the indicated valuie). g. 99.5 percentile (of the 24-houir mean, 99.5 per-cenit muist be below the indicated valute). h. WHO guiidelines for combined exposuire toS502 ali(l TSP are 50 gIg/m5 for annutal mean and 125 hg/in5 for 24-houir mean.To protect vegetation, WHO guiidelines for SO., ar-e 330 pg/rn2 for tise annual mean and 100 pg/in3 for the 24 hour mean. i. The 10-minuite WHO guiidelinie incorporates a pr-otection factor of 2, and the 1-hiouir WHO gutideline is a derived equtivalent figutre. Source: WHO/UNEP 1992; persosial communtiicationis withi techniical experts from differenit cotuntries. Air Pollutants and Their Effects: Annex B 49 Table B.5 Air quality standards and WHO guideline for TSP (micrograms per cubic meter) Time-weighted aver(Oge Oher (Cniontry 24 hours I year averaging times Argentina 150 (1 month) City of Buenos Aires 500 (20 minutes) 150 (30 days) Province of Buenos Aires 150 Bolivia 260 75a Brazil Primary standards 240 80a Secondary standards 150 60a Sao Paulo Attention level 375 Alert level 625 Emergency level 875 Critical level 1,000 Chile 260 75a Colombia 400 1OOa Santafe de Bogota 307b 77a,b Ecuador 250 80a Mexico 260 75 Mexico City" Satisfactory level 260 Unsatisfactory level 445 Bad level 630 Very bad level 1,000 Peru 350d 150d Venezuela 75e 150f 2601" '"HO' 150-230 60-90 Note: A blank space indicates that no standard was established. a. Geometric mean. b. This standard reflects 13°C temperatuire and 560 mm Hg pressure for Santafe de Bogota. c. Time-weighted averages indicate maximum values for each category. d. Proposed standards. e. 50 percentile (of the 24-hour mean, 50 percent must be below the indicated valie). f. 95 percentile (of the 24-hour mean, 95 percent must be below the indicated valie). g. 98 percentile (of the 24-hour mean, 98 percent must be below the indicated value). h. 99.5 percentile (of the 24-hour mean, 99.5 percent riuilst be below the indicated value). i. WHO guideline for combined exposure to SO2 and TSP is 125 pg/mi for 24-hour mean. Sourre: WHO/UNEP 1992; personal communications with tecliinical experts from different countries. 50 Chlapter 2 Table B.6 Air quality standards for PM-10 (micrograms per cubic meter) 7ime-wveighted average (Country 24 hours I year Bolivia 150 50a Brazil Primary standar(ds 150 50 Secondary standards 150 50 Sao Paulo Attention level 250 Alert level 420 Emergency level 500 Critical level 600 Chile 150 Santiagoh Good level 150 Regular level 195 Bad level 240 Critical level 285 Dangerous level 330 Mexico 150 50 Mexico Cityb Satisfactory level 150 Unsatisfactory level 350 Bad level 420 Very bad level 600 United States 150 50 Note: A blank space indicates that no standard was established. a. Geometric mean. b. Time-weighted averages indicate maximumil values for each category. S.ource: WHO/UNEP 1992; personal commutiicationls with technical experts from different cotintries Table B.7 Air quality standards and WHO guideline for lead (micrograms per cubic meter) ;ountry I houtr 24 hours 3 months Other averaging times Argentina City of Buenos Aires 1 10 (20 minutes) Province of Buenos Aires 16 15 10 (30 days) Bolivia 1.5 Ecuador 1.5 Mexico 1.5 Peru 15a (30 days) 5a (1 year) Venezuela 1.5b 2' United States 1.5 WsHO 0.5-1.0 (1 year)'" Note: I ppm = 2,600 micrograms per cubic meter (gg/m3). A blank space indicates that no standard was established. a. Proposed standards. b. 50 percentile (of the 24-hour mean, 50 percent must be below the indicated value). c. 95 percentile (of the 24-houir mean, 95 percent must be below the indicated valie). d. WHO giuideline incorporates a protection factor of two and is based on the assumption that 98 percent of the populationi will have a lead concentration in )blood( less than 29 mg/dl. Soturre: WHO/UNEP 1992; personal commuiicationis with technical experts from different countries. Air Pollutants and Their Effects 51 References Detels, R., D.P. Tashkifi, andJ. W. Sayre. 1991. "The UCLA Population Studies of COPD: X. A Cohort Abbey, David E., Floyd Petersen, Paul K. Mills, and Stuidy Changes in Respiratory Function Associated W. Lawrence Beeson. 1993., "Long Term Ambient with Chronic Exposure to SO,, NO,, and HC." Concentrationis of Total Suspended Particiilates. AmeticanJournal of Public Health 81:3,350-9. Ozone and Sulfur Dioxide and Respiratory Symp- Dockery, Douglas W.C., and Arden Pope III, 1993. toms in a Non-Smoking Population." AirhiVes of "An Association Between Air Pollution and Mor- Envziionmental Health 48:33-46. tality in Six U.S. Cities." Ner EnglandJournal of Medi- ALA (American Lung Associationi). 1997. "Gambling *ine 329: 1753-59. With Health II. WAho Loses UJnder New Health Stan- Dockery, Douglas W., F.E. Speizer, and( D.O. Stram. dards for Particulate Matter." Washington, D.C. 1989. "Effects of Inhalable Particles on Respiratory AQIRP (Air Quality Improvement Research Pro- Health of Children." American Reviewo of Respiratory gram). 1991. "Exhaust Emissions of Toxics Air Pol- Disease 139:587-594. lutants Using Reformulated Gasolines." Technical EIA (Energy Information Administration). 1994a. Bulletin 5. Coordinating Research Council Inc., Energ Use and Carbon Emissions: Non-OECD Goun- Atlanta, Georgia. tries. DOE/EIA-0587. Washington, D.C. Bellomo, Salvatore L. and Sally D. Liff. 1984. Funda- . 1994b. Energy Use and Carbon Emissions: Some mentals ofAir QualityforHighwayPlanningand Pm ject International Comparisons. DOE/EIA-0579 (94). Development. U.S. Department of Transportation Washingtont, D.C. Federal HighwayAdministration, McLean,Virginia. Fairley, David 1990. "The Relationship of Daily Mor- Brunekreef, B. 1986. ChilIdhoodExposure toEnvirionmen- tality to Suspended Particles in Santa Clara County, tal Lead. MARC Report 34. Monitoring and Assess- 1980-1986." Environmental Health Perspectives ment Research Centre, King's College, University 89:159-68. of London, London. Faiz, Asif, Kumares Sinha, Michael Walsh, and A. CDC (Center for Disease Control). 1991. Strategic Varma. 1990. "Automotive Air Pollution: Issues and Plan for the Elimination of Childhood Lead Poi- Options for Developing Countries." Policy Research soning. U.S. Department of Health and Humani Working Paper 492, World Bank, Washington, D.C. Services, Washington, D.C. Faiz, Asif, Kumares Sinha, and Surhid Gautam. 1994. Cheni B., C. Hong, M. Pandey, and K Smith. 1990. "Air Pollution from Motor Vehicles: Issues, Char- "Iridoor Air Pollution in Developing Countries." acteristics, and Trends" (Draft). World Bank, Wash- World Health Statistics Qua terly Report 43: 127-38. ington, D.C. Chinn, S., V Florey, W.E Beeson, and E Peterson. Faiz, Asif, Christopher S.Weaver, and Michael Walsh. 1981. "Ambient Air Pollution and Cancer in Cali- 1996. AirPoUlutionfrom Motor Vehicles: Standards and fornia Seventh Day Adventists." Archives of Environ- Technologiesfor ControllingEmissions. World Bank and mental Health 46:271-80. United Nations Environment Programme, Wash- Cohn, L.F. and G.R. McVoy. 1982.EnvironmentalAnaiv- i ngton, D.C. sis of Transportation System. VA'iley-Interscience, New GEMS (Global Environment Monitoring System). York. 1988. Assessment of Urban Air Quality. United Na- Cropper Maureen L. and W.E. Oates. 1992. "Envi- tions Environment Programme and World Health ronmental Economics: A Survey." Journal of Eco- Organization, Geneva. nomic Literatire 30: 675-740. Guensler, Randall. 1994. "Data Needs for Evolving Cropper, Maureen L. and Nathalie B. Simon. 1996. MotorVehicleEmission ModelingApproaches." In "Valuing the Health Effects of Air Pollution." DEC Thomas E Wholley, ed., Transportation Planningand Notes 7, World Bank, Washington, D.C. Air Quality 1, Proceedings of the National Conference. Derriennic, F., S. Richardson, A. Mollie, and J. American Society of Civil Engineers, New York. Lellouch. 1989. "Short-Term Effects of Sulphur Hatzakis, A., K Katsouyannli, A. Kalandidi, N. Day, Dioxide Pollution orn Mortality in Two French .Cit- and D. Trichopoulos. 1986. "Short-Term Effects of ies." InternationalJournal ofEpidemiologt, 18:186-97. Air Pollution onl Mortality in Athens." International Derwent, D. 1988. "Better Way to Control Pollutioni." Joutrnal of Epidemiology 15:73-81. Nature 331: 575-78. Herron, M.M. 1982. "Impurity Sources of F, Cl, NO3, Detels, Roger, Donald P. Tashkin, James W. Sayre, andSO4 in Greefiland atidAitarctic Precipitatioti." Stanley N. Rokaw, Frank J. Massey, Annie H. Jou-rnal of Geophysical Research 87:3052-60. Coulson, and David H. Wegman. 1987. "The UCLA Horowitz, Joel L. 1982. "Air Quality Analysis for Ur- Population Studies of Chronic Obstructive Respi- ban Transportation Planning." MIT Press, Cam- ratory Disease" Clhest 92:594-603. bridge, Massachusetts. 52 Chiapter 2 Krupnick, Alain J., Winstoti HarriTigton. and Bart Penna, M.L.E and M.P. Duchiade. 1991. "Air Pollu- Ostro. 1990. "Ambient Ozone and Acute Health tion and Infant Mortality fTom Pneumonia in the Effects: Evidence from Daily Data." Journal of Envi- Rio de Janeiro Metropolitan Area." Butlletin of the ronmentalEconomies and Management 18:1-18. Pan-American Health Organization 25:47-54. Krzyzanowski, M., and B. Wojtyniak. 1982. "Ten-Year Plagiannakos, Takis, andJohti Parker. 1988. "An As- Mortality in Sample of an Adult Population in Re- sessment of Air Pollution Effects on Human Health lation to Air Pollution." Journal of Ep5idemiology and in Ontario." Ontario Hydro, Ontario, Canada. Communitv Health 36:262-8. Pope III, C.Arden, Joel Schwartz, and Michael R. Lacasafia, M., I. Romieu, R. McConnell. y Grnpo de Ransom. 1992. "Daily Mortality and PM-10 Pollu- trabajo sobre plomo de la OPS. 1996. "El Probletna tion in Utah Valley." Archives ofEnvi?nnmentalHealth de Exposicion al PIOmo en America Latina y El 42(3):211-17. Caribe." Centro Panamericano dce Ecologia Pope III, C.Arden, Douglas W. Dockery, John D. Humana y Salud, Organizaci6n Mundial de la Spengler, and Mark E. Raizenne. 1991. "Respira- Salud, Metepec, Estado de Mexico. tory Health and PM10 Pollution: A Daily Time-Se- Lamm, S.H., TA. Hall, and A. Engel 1994. "PM-10 ries Analysis." American Reviewv of Respiratory Disease Particulate: Are they Major Determinanit of Pedi- 144:688-74. atric Respiratory Admissions in Utah Couinity, Utah Pope III, C.Arden , M.J. Thun, N.M. Namboodiri, (1985-1989)." Annals of Occupational Hygiene 38 D.W. Dockery, J.S. Evans, EE. Speizer, and C.WU (Suppl. 1): 969-72. Heath Jr. 1995. "Particulate Air Pollution as a Pre- LSHTM (London School of Hygiene and Tropical dictor of Mortality in a Prospective Study of U.S. Medicine) and St. GHMS (St. George's Hospital Adults." American Journal of Critical Care Medicine Medical School). 1995. "Review of Met,hodIs Pro- 151(3):669-74. posed, and Used, for Estimating the Population Radian Corporation. 1995. "Mobile Sources Estimate He'alth Risks of Exposure to Urban Air Population." for Monterrey, N.L., Mexico." Austin, Texas. Report prepared for the World Bank. Lond(on. Romieu, Isabelle. 1992. "Epidemiologic Studies of the MARC (Monitoring and Assessment Research Cen- Health Effects of Air Pollution due to Motor Ve- ter). 1991. "Health Implications ofAveragingTimes hicles." David E. Mage and Oliver Zali, eds. Motor Shown." University of London, King's College. VelhicleAirPolluttion: Public Health Impact and Control Masterson, '.L., E. Slowinski, and C.L. Stanitski. Measures. World Health Organization and Depart- 1985. Chemical Principles. CBS College Publishing, ment of Public Health, Geneva. New York. Romieu, Isabelle, H. Weitzenfeld, andJ. Firikelman. OECD/IEA (Organization for Econotmic Coopera- 1992. "Urban Pollution in Latin America and the tion and Development/International Energy Caribbean: Health Perspective." World Healthl Sta- Agency). 1991. Greenhouse Gas Emissions: Tlte Energy tistics Qutarterly 43:153-67. Dimension. OECD, Paris. Romieu, Isabelle, Eduardo Palazuelos, Mauricio Ostro Bart. 1989. "Estimating the Risk of Smoking, Hernandez, Gustavo Olais, and Roberto Tapia. Air Pollution and Passive Smoke in Acute Respira- 1995. "Issues of Lead Contamination: The Experi- tory Conditions. Risk Analysis 9:189-91. ence of Mexico." Paper for U.S.- and Mexico-spon- 1994. "Estimating the Health Effects of Air sored International Workshop on PhasingLead Ott Pollutants. A Method with an Application to of Gasoline, March 14-15, Smithsonian Institution, Jakarta." Policy Researc'h Working Paper 1301, WA'ashington, D.C. World Bank, Washington D.C. Rossano, A.T and TA. Rolander. 1976. "The Prepa- Ostro, Bart, M.L. Lipsett., M.B. Wiener, andJ.C. Selner. ration of an Air Pollution Source Inventory." In M.J. 1991. "Asthmatic Response to Airborne Acid Aero- Suess and S.R. Craxford, eds., Manual on Urban sols." AmericanJouwrnal of Public Health 81: 694-702. Air Quality Management. WHO Regional Publica- Ostro, Bart,Jose Miguel Sanchez, Carlos Aranida, and tions, European Series 1, Copenhagen. Gunnar S. Eskeland. 1995. "Air Pollution and Mor- Rossano, A.T. andJ.E Thielke. 1976. "The Design and tality: Results from Santiago, Chile." Policy Re- Operation of Air Quality Surveillanice System." In search Working Paper 1453, W'orld Bank, MJ. Suess and S.R. Craxford, eds., Manual on Ur- Washington D.C. ban Air Quality Management. WHO Regional Pub- Ozkaynak, Halik, and G.D. Thurston. 1987. "Asso- lications, European Series 1, Copenhagen. ciations Between 1980 U.S. Mortalitv Rates and Roth, PM. 1990. "Effects of Transportation of Ozone Alternative Measures of Airborne Particle Concen- in Cities." TR News 148:11-16. Transportation Re- tration." Risk Analysis 7:449-61. search Board, Washington, D.C. AirPollutants and TheirEffects 53 Saldiva, Paulo, C. Arden Pope III, Joel Schwartz, .1993. Supplement B to tile Gideline on Air Quality Douglas W. Dockery AJ. Liclhi erfels AJ.,J.M. Salge, Models (Revised). Office of Air Quality Planining and I. Barone, and G.M. Bohmn. 1994. "Air Pollition Standards, Research Triangle Park, North Carolina. and Mortality in Elderly People: A Time Series . 1995. Stratospheric Ozone Depletion. A Focus on Study of Sao Paulo, Brazil." Amterican Review Res1pi- EPA's Research Office of Research and Development. ratoty Disease 144:688-74. WVashington, D.C. Schwartz,Joel. 1993. "Air Pollution and Daily Mortal- Walsh, Michael P. 1995."Global Trends in Motor Ve- itv in Birmni ngiamn, Alabama.' AmericanJournal ELli- hicle Pollution Control: Challenges and Opportu- demiology 137:1136-47. nities in 1995." In Proceedings of Third International Schwartz, Joel and Douglas W. Dockerv. 1992a. ' n- Conferenceon theAutomotiveIndustry and theEnviron- creased Mortality in Phiiladelpiiia Associated witlh ment. March 9-10, Geneva. Daily Air Pollution Conceintratioris." American Re- Watkins. J.H. 1991. Air Pollution from Road Vehicles. view of Respiratory Disease 145:600-4. Transport and Road Research Laboratory, Depart- 1992b. "Particulate Air Pollution and Dajilv ment of Transport, HMSO Publication Centre, Mortality in Steubenville, Ohio." AmericanJournal Londoni. of Epidemiology 135:12-20. Whittemore, Alice S. and Edward L. Korn. 1980. Schiwartz,Joel and Allan Marcus 1990. "Mortality anid "Asthma and Air Pollution in the Los Angeles Air Pollution in Londoni: a Time Series Analvsis." Area." AmericanJournal of Public Health 70:687-96. American Joutnal of Epidemiology 131:185-194. WHO (World Health Organization). 1987. "Sulfur Scnvwartz, Joel and Scott Zeger. 1990. "Passive Smnok- Dioxide and Particulate Matter." In Air Quality ing, Air Pollution, ancl Acute Respiratory Symptoms Guidelines for Europe. WHO Regional Publications, in a Daily Study of Student Nurses." American Re- European Series 23, Copenhagen. vien'y of Respiratory Disease 141:62-7. WHO/UNEP (World Health Organization/United Schwartz, Joel, V Hasselblad, and H. Pitcher. 1988. Nations Environment Programme). 1992. Urban Air "Air Pollution and Morbidity: A Further Analysis Pollution in Megacities of the World. Blackwell Refer- of the Los Angeles Student Nurses Data." Journal ence, Oxford, U.K. of the Air Pollttioni Con trol Association 38:158-62. Wijetilleke, Lakdasa and Suhashini A.R. Karunaratne. Schwartz,Joel, David Wypij, Douglas Dockery,James 1995. "Air Quality Management. Considerationsfor Ware, Scott Zeger, John Spengler, and Benjamin Developing Countries." Technical Paper 278, World Ferris,Jr. 1991. "Daily Diaries of Respiratory Synmp- Bank, Washington, D.C. toms and Air Pollution: Methodological Issues and( Winchlester, J.W. 1989. "Coarse and Fine Particles in Results." EnvironmentalHealthPerspectives90:181-7. the Atmosphere." In E.E. Pickett, ed., The Atmo- Smith, KR. 1993. "Combustion, Air Pollution and spieric Pollution. Hemisphere Publishing Corpora- Health in Developing Countries." Annual Review of tion, New York, Ener,f and the Environment 18:529-66. W"orld Bank. 1996. "Ground-Level Ozone." Pollution UN (Uniited Nations). 1987. Effects and Control of Prevention and Abatement Handbook (Draft). Transbo7tndar-' Air Pollution: Report Prepared, within Washington, D.C. the Framewvork of the Convention on Long-Range WARI (World Resource Institute). 1994. World Resources Tr-ansbouindary Air Pollttion. New York. 1994-1995: A Guide to the GlobalEnvironment. United UNEP (United Nations Environimenit Programimle). Nations Environiment Programme and United Na- 1991. Enviionmental Data Book (1991/1992). Basil tions Development Programme, New York. Blackwell, Oxford, U.K. Xu, Xiping, Jun Gao, Douglas Dockery, and Yuda USEPA (United States Environimenital Protection Chen. 1994. "Air Pollution and Daily Mortality in Agency). 1990. Air Quality Cr-iteriafor Lead: Suppl5le- Residential Areas of Beijing, China." Archives of ment to the 1986 Addendum. Office of Research and Environmental Health. 49: 216-22. Development, Washington, D.C. ABATEMENT MEASURES FOR VEHICULAR AIR POLLUTION De esigning a strategy to abate vehicu- quality. Since pollution can be viewed as a nega- lar air pollution in an urban area re- tive externality whose costs are not fully borne quires a good understanding of the by polluters, market-based incentives based on nature and magnitude of the air pol- the "polluter pays" principle impose a price on lution problem and the applicability of various polluting activities, thereby internalizing the cost abatement measures. Policy measures can be of the externality. Some polluters may prefer to classified as command-and-control or market- pay the "price" imposed by the regulator rather based incentives. The choice of a measure de- than lower their level of pollution, while others pends on its costs and benefits, as well as on the may find it cheaper to modify their current ac- monitoring and enforcement: capabilities of the tivities in ways that reduce or eliminate pollu- responsible institutions. More often than not tion. The higher the government sets the price these measures, if well used, complement rather of pollution, the greater is the reduction in pol- than substitute for one another. lution, assuming adequate institutional support, Command-and-control measures for curtailing monitoring, and enforcement (Hamrin 1990). vehicular air pollution mainly rely on regulatory Market-based incentives that have been used to options. Such measures include emission stan- control vehicular air pollution include vehicle dards that set a legal ceiling on the quantity or taxes, fuel taxes, and congestion charges. concentration of pollutants discharged from Common measures for reducing air pollution vehicles, standards that specify fuel quality for from motor vehicles are shown in Table 3.1. motor vehicles, requirements to use a certain These measures, which targetvehicles, fuels, and technology (such as smaller nozzles for fueling transport management, are classified as com- unleaded gasoline), testing and certification mand-and-control measures, market-based in- nrles for new vehicles, inspection requirements centives, and additional measures. These for in-use vehicles, and traffic restrictions. measures are sometimes accompanied by actions Market-based incentives rely on market forces that promote public awareness and education to bring about improvements in ambient air (Box 3.1). 55 56 Chlapter 3 Table 3.1 Typical measures for the abatement of air pollution from motor vehicles 7Ypee ofmeasure Command-and-conmol mneasmesv Afzrket-based incentives Addilional measutres Vehicle-targeted Emission standards andc related Vehicle emission fees measures for new vehicles Differential vehicle registra- * Emission standar(ds tiotn fees and taxes * Certification Vehicle retrofit programs * Assembly line testing Vehicle replacement or * Recall scrappage incentives * WAarranty Emission standards and inspec- tion programs for in-use vehicles * Emission standards * Inspection and miaintenance programs * Roadside inspection programs Emission restrictions on imported vehicles Fuel-targeted Gasoline standards Fuel taxes * lead Tradable permits and credits * volatility * benzene and other aromatic hydrocarbons * reformulated gasoline * oxygenated gasoline Diesel fuel standards * sulfur * cetane number * aromatic hydrocarbons and density Alternative fuels * compressed natural gas * liquefied petrolenm gas * methanol * ethanol Transport Driving bans Road pricing Provision of public managernent On-street parking and tradling Area licensing transport services restrictions Promotion of non- Traffic priority measures for buses motorized transport Ride sharing Provision of off-street Staggered work hours parking Speed limits and other traffic management tneasures Land use planning an d controls Soutrce: Based on Carbajo 1994. Abatement Measures for Vehicular Air Pollution 57 Box 3.1 Public awareness and education Preparing and implementing an effective urban air pollution control strategy requires education at all levels of society, including policyrnakers, implementinig agencies, manufacturing and service industries, and the public. Education of policymakers is fundamnerital in developing an integrated urban air pollution control strategy that includes cost-effective measures to control vehictilar emissions. Education programs expose policymakers to such topics as the type and effectiveness of different air pollution control measures, analytical tools for evaluat- ing these measures, and lessons learned from different countries. For example, an October 1996 seminar in Buenos Aires organized by USEPA wvas designed for policymakers from the environmental, health, and trans- port agencies responsible for formulating strategies to reduce vehicular emissions. The seminar focused on identifying main air pollutants and their health effects, interpreting emissions inventory and ambient air quality (lata. evaluating pollution control measures, developing options for reducing and eliminating lead from gaso- line, and preparinig an urban transport policy. Introducing a motor vehicle emissions control techniology in a developing country generates a growing need for trained and experienced vehicle mechaniics and technicians. A cadre of inspectors and technicians is also required to effectively manage and operate inspection and mairitenance programs. At every level of vehicle emission control, efforts should be made to provide appropriate and thorough training of officials, inspectors, techniciatis, atid operators. Training may iniclude such measures as specialized courses, technical meetings, individual instruction, and in-service training. For example, in Sao Paulo mechanical training courses for servic- ing vehicles failing the inspections have been provided to technical school instructors. International organiza- tions could also play a catalytic role in organizing trainirig programs. An executive training program on motor vehicles and the environment, recently establishedl by the International Road Federation, provides a useful training opportunity for senior officials in developing countries. If measures to reduce urban air pollution are to be effective, they must be supported by the public. Public participation through roundtables, seminars, and meetings enhances formulation of a sound package of mea- sures and eases resistance to environmental protection by special interest groups. Public education and aware- ness campaigns explaininig the general nature of air pollution, its adverse health effects, and the ways in which the public can help reduce pollution from motor vehicles increase the degree of success in reaching the desired objectives during implementation of these measures. Several Latin American countries have taken various steps to educate the public on the adverse effects of vehicuilar air pollution. In Chile, for example, an inforniation campaign was organized to help people understandl the drawbacks of using leaded gasoline in cars fitted with catalytic converters. In Honduras, with the help of the Swiss government, numerous public meetings were held and pamphlets were distributed to inform people about the harmful health effects of lead in gasoline. In Trinidad an(d Tobago local NGOs have written extensively in local newspapers on the health effects of lead. In Uruguay several seminars and workshops have been organized to foster public awareness of vehicular air pollution. These included a roundtable discussion on lead in gasoline, a conference oin vehicular air pollution, a roundtable on atmospheric contamination, and a presentation by a local NGO on vehicular air pollution. In Venezuela several NGOs have prepared papers for public dissernination on the health risk of lead and written articles in local newspapers. Contitiuous public awareness campaigns about the negative effects of leaded gasoline have been organized in Barbados and Costa Rica. VEHICLE-TARGETED MEASURES Emission Standards and Related Measures chapter. New vehicle emission standards in cer- for New Vehicles tain Latin American countries have been set us- ing the U.S. standards and, to a lesser extent, Emission standards are established to limit dis- the European Union's standards. The emission charge of air pollutants from new motor vehicles. standards established in Argentina, Brazil, Chile, These standards are usually accompanied with Colombia, and Mexico are presented and dis- certification, assembly line testing, recall, and cussed in Chapter 4. warranty requirements. The establishment of emission standards for new motor vehicles has been a major impetus Emission standards. Emission standards for new for research and development of engine designs vehicles are generally specified by vehicle class and pollution control technologies. Major ad- and engine type (based on fuel). Vehicle classes vances in this area have come from engine modi- include light-duty vehicles (passenger cars, light- fications, catalytic converters, on-board canisters, duty trucks, light-duty commercial vehicles), me- and on-board diagnostics for gasoline-fueled dium-duty vehicles, and heavy-duty vehicles. vehicles (Box 3.2); and engine modifications, Engine types include spark-ignition and diesel. trap oxidizers, and catalytic converters for die- Spark-ignition engines use such fuels as gaso- sel-fueled vehicles (Box 3.3). line and low-emission fuels (such as CNG and Emission standards for new vehicles must be LPG), and diesel engines use diesel fuel. In ad- backed by stable government policies because dition, some emission standards restrict emis- of their implications for the automotive indus- sions from vehicles operating in a specific try, petroleum refiners, and consumers. To meet geographical area (such as urban buses). Stan- the standards, vehicle manufacturers must in- dards for new gasoline-fueled vehicles (or ve- vest in research and development that involves hicles equipped with spark-ignition engines) modifying vehicle technology, investing in the generally limit CO, HC, and NO, emissions in parts and labor associated with the production the exhaust and evaporative emissions. In Brazil, of the modified design, and testing for and cer- where alcohol-based fuels are used in spark-igni- tifying compliance. Because this effort is exten- tion engines, aldehyde emissions are also Jim- sive, vehicle manufacturers must have sufficient ited. Standards for new diesel-fueled vehicles lead time for product research and develop- generally limit CO, HC, NO., PM, and smoke ment-at least three to four years in industrial emissions. countries (Nill 1995). Forshorterperiods of lead All industrial countries have established emis- time, a phased-in compliance program, such as sion standards for new motor vehicles; the that used in Brazil, can be used to allow vehicle United States was the first country to do so. An manufacturers to decide on the combination of overview of U.S. emission standards for new pas- vehicle types and models to introduce into the senger cars, light-duty trucks, and heavy-duty market. In addition, vehicle manufacturers pre- duty vehicles is presented in Annex A to this fer to respond to fewerchanges ofmoderate emis- 59 60 Chapter 3 Box 3.2 Pollution control technologies for vehicles with four-stroke spark-ignition engines Formation of pollutants in four-stroke gasoline engines depends on two major factors: the air-fuel ratio and ignition timing. The air-fuel ratio of 14.7:1 is called stoichiomnetic because the mixture has exactly enough air to completely burn the fuel. At air-fuel ratios below stoichiometric high levels of CO and HC are emitted as a result of incomplete combustion of the fuel. CO anid HC emissions decrease, but NO, emissions rise as the air-fuel ratio increases below the stoichiometric value. At ratios slightly higher than stoichiometric (between 14.7 and about 17.6) the CO and HC emissions remlaini low and NO,, emissions remain high. At higher ratios, the CO and HC emissions rise moderately and NO, emnissions decrease sharply. Ignition timing affects the formation of HC and NO,. In addition, the air-fuel ratio andl ignitioni timinig affect fuel consumption. Pollution control technolo- gies for vehicles with four-stroke spark-igniition engines include: Engine modifications. The most commonlv' used engine modifications techniques include: * Increasing air-fuel ratios above 17:1 (lean burn engines). Satisfactory performance of these engines can be maintained by enriching the mixture during idling, acceleration, and high speed (ECMT 1990). * Replacing the conventional carbuiretor and distributive ignition technology with electronic fuel injection and ignition systems to ensure precise control of engine combustion and emission of air pollutants. * Using exhaust gas recirculation technology to recycle a portion of the exhaust gas from the engine to the incoming air-fuel mixture to reduce peak temperatures in the combustion chamber and lower NO, emis- sions from vehicle exhaust. This techniology can also be used in conjunction with catalytic converters. Electronic fuel injection and ignitioni systems reduce HC and CO emissions but are less effective for NO, emissions. When fuel ignition timing andl exhatust gas recirculation are also electronically controlled, however, NO,, exhaust emissions can be reduced with nio change and, in some cases, with an improvement in fuel economy (Faiz 1990). Vehicles equipped with electronic fuel injection systems also do not require engine adjustments at different altittudes, whereas older vehicles equipped with carburetors or automatic injection systems require such adjustments to compensate for the amount of oxygen in the air (Faiz, Weaver, and Walsh 1996). Catalytic converters. Catalytic converters remove pollutants from the engine exhaust. Two types of catalytic con- verters are common in motor vehicles: those with oxidation (two-way) catalysts and others with oxidation-reduc- tion (three-way) catalysts. Two-way catalytic converters oxidize CO arid HC to CO2 and water in the presence of platinum or a palladium catalyst. The effectiveness of the catalytic reaction depends on the exhaust gas tem- perature, air-fuel ratio, and types of HC present in the exhaust gas. Two-way catalytic converters are not fully efficient at temperatures below 250"C for oxi(lizing CO and below 250'C to 340"C for oxidizing HC. The air-fuel ratio in the engine is mainitained above stoichionietric to ensure presence of oxygen in the exhaust for the sion reductions than multiple changes of small ply. Nevertheless, such standards are still promul- emission reductions to achieve a long-run target. gated. For example, a 1996 ordinance by the This approach allows manufacturers to best uti- Municipality of the Metropolitan District of Quito lize their resources and maintain their competi- bans circulation of 1997 and newer model-year tive position in the market. Furthermore, vehicle vehicles not equipped with an evaporative con- manufacturers dislike technology-based stan- trol system, and of 1998 and newer model- year dards, which may affect their competitive position vehicles not equipped with a catalytic converter. in the market.' Instead vehicle manufacturers In developing countries the availability and prefer to have the government set emissions stan- practicality of a pollution control technology dards, but to choose for themselves hiow to com- must be considered prior to establishing emis- sion standards for new vehicles. Because most 1. Technology-based standards require vehicle mauitifac- of the automotive indstry in Latin American turers to tuse a specified polltition control teclhinology (such countries is linked to international automotive as catalytic converters). companies from industrial countries, complying Abatement Measuresfor 1ehbicularAirPollution: Vehicle-Targeted Measures 61 oxidation of CO and HC. Highly reactive HC like formialdehyde and olefins are oxidized more effectively than less reactive ones, whiereas short-chaini paraffins like methiane, ethane, and propane are difficult to oxidize. Two- way catalytic converters typically remove about 80 percent of both unburned HC and CO. Three-way catalytic converters use a combination of platinum, palladium, and rhodium catalyst. These con- verters oxidize CO and HC to CO, and water. andl re(luce nitric oxide to nitrogen. Oxidation and reduction reactions are optimum at a narrow, stoichionietric range of the air-fuel ratio. In gasoline-fueled cars this range is mainitainied through the use of exhaust sensors (also knowni as oxygen sensors) that provide feedback to the electronic injectioni system. Three-way catalytic converters remove about 90 percent of unburnied HC and CO, and about 70 percent of NO, in the exhaust stream from the engine. Reduction of NO2 to nitrogen is less efficient at temperatures above 400GC. Vehicles equipped with catalytic converters miust be ftieled with unleaded gasoline to avoid deposition and poisoning of the catalysts by lead emissions. In adltition, emissions of sulfur and phosphorous from the fuel reduce catalytic activity. Catalysts can also be damaged by excessive temperature caused by combustion of com- bustible materials in the engine exhaust. To prevent such damage, an electronic fuel injection system with an oxygen sensor in the exhaust (as a feedback) is uLsed. Such a system is called a "closed-loop catalytic converter." Catalytic converters have been used in the United States since 1975. They have also been used in Japan since the 1970s and in Europe since the 1980s. Compliance with current emission standards for light-duty gasoline- fueled vehicles in these countries requires the use of three-way catalytic converters. On-board canisters. On-board activated carbon canisters are used to adsorb hot soak and diurnal emissions (which are HC evaporated from gasoline) at removal efficiencies greater than 90 percent. The canisters are purged with the intake air, which directs the removed HC to the engine for combustion. Use of on-board canisters has been cotnpulsory in the United States andJapan since 1970s, and more recently in member countries of the Europeani Union, as well as Switzerland (ECMT 1990). On-board diagnostics. These systems idenitify antI tliagniose emission-related malfunctions in vehicles equipped with electronic enginie control systenms while thie vehicle is actually being driven, and warn the driver with an indicator liglht (Faiz, Weaver, and Walsh 1996). On-board diagnostics not only alert the vehicle operator to take necessary corrective actions, but help service techlnicianis analyze and repair malfunctions, and facilitate detec- tionI of problems durinig periodic inspection and( uainitenanice (Walsh 1995). California has required on-board (liagnostics for some time, and in 1990 the U.S. Clean Air Act was amended to phase in such systems for all new light-dutv vehicles in the Unite(d States by 1998 (Faiz, WNeaver, and Walsh 1996). with a new standard usually requires adapting an reformulation of fuels, which may involve refin- existing control technology rather than researchi- ery modifications. In such cases refiners need to ing a new technology. However, this effort still be given enough lead time to make the necessary involves redesigning vehicles to incorporate the process modifications and produce higher-qual- necessary pollution control technologies. It may ity fuels. also require auxiliary manufacturers to develop In setting up emission standards, the ability and supply some of the needed components. In of potential consumers or society in general addition, emission standards for imported ve- (such as through tax incentives) to afford the hicles must be compatible with those for domes- higher costs of environmentally friendly vehicles tically manufactured vehicles to avoid adverse must be considered. In addition, consumers impacts on the local manufacturing industry. must bear the increased maintenance cost of Making emission standards more stringent also vehicles and the cost of additional fuel for any may affect petroleum refiners. For example, tight- reduction in fuel economy associated with emis- ening emission limits for lead and sulfur requir-es sions control technology. Typical performance 62 C,lapter, 3 Box 3.3 Pollution control technologies for diesel-fueled vehicles Uncontrolled diesel-fueled vehicles emit tihe same types of pollutants as uncontrolled gasoliine-fueled vehicles. But since diesel-fueled enginies operate at higher air-fuel ratios than spark-ignition engines, uncontrolled diesel- fueled vehicles emit less CO and HC. However, utncotitrolled diesel-fueled vehicles emit more PM and smoke, whichi containi unburned HC and other organics. The most commonily used pollution control technologies for diesel-fueled vehicles include: Enigine modifications. Engine modifications include: * Optinizinig combustion using high-pressure fuel injection with electronic control and an altered combus- tion chamber design. These techiniques can reduce HC and smoke emissions. * Turbochiarging and charge cooling. Tuirbocharging involves compressing the intake air to the combustion chamber, allowing more fuel to be burnt and increasing the power output. Turbocharging lowers PM emissions but iticreases NO, emissions. Charge cooling lowers the temperature of the hot air from the turbocharger exhaust, reducing NO, emissions. * Injection timing at the top or somewhiat before the top of the compression stroke (also called top-dead- center) achieves optimum fuel econotniv. Injection timing before top-dead-center increases the maximum temperature and pressure in the cvlittder and results in greater NO, emissions and lower PM and HC emissions. Retarding fuel injection to lower peak temperatures and pressures during combustion reduces NO, emissions but increases fuel consumption as well as smoke and HC emissions, especially under high- load conditions. High fuel injection pressures, with electronic controls, improve fuel atomization and fuel- air mixing and offset the effects of retarded injection timing by increasing the injection rate (Faiz, Weaver, and Walsh 1996). * Using exhaust gas recirculatiort to replace sonie of the intake air, reducing the available oxygen and combustion temperature. This tecltni(ltie can lower NO, emissions, but causes engine wear due to higher PM emissions. X Minimizing the amount of lubricating oil that leaks from various sources (such as piston rings, valve guides, or the turbocharger) into the enginie to reduce HC emissions (ECMT 1990). Trap oxidizers. Trap oxidizers remove PM from diesel engine exhaust gases. They include a filtration system to collect PM frotn exhaust gases and an oxidation system to prevent clogging of the filter through regeneration by oxidation. Trap oxidizers can remnove about 9() percent of PM from the exhaust stream of diesel engines. Trap and catalyst techriologies can also be used jointly (Walshi 1995). Cataltic convertes. Two-way catalytic converters are used in diesel-fueled vehicles, mostly in cars and, more recently, in heavy-duty vehicles. Besides removing up to 80 percent of CO and volatile organic compounds, they can oxidize a large portion of HC present in the soluble organic portion of PM emissions (soluble organic compounids make up 30 to 70 percent of PM) and decrease odor in exhaust emissions by oxidizing such com- pounds as aldehydes. The sulfur contenit in diesel fuel must be low (less than 0.05 percent) to reduce formation of sulfuric acid and sulfates which contribute to PM emissions (Faiz, Weaver, and Walsh 1996). Two-way catalytic converters are cominionlv used in industrial countries. Almost all ligit-duty diesel-fueled vehicles manufactured in Austria, France, and Germany (about 500,000 cars a year) are equipped with two-way catalytic converters. By 1997 all new light-duity diesel-fueled vehicles in Europe will come with two-way catalytic converters to meet the new European standards. In the United States diesel-fueled trucks equipped with two-wvay catalytic converters were marketed in 1994 to tmeet tne 0.1 g/bhp-h PM standard. The samne technology is used for diesel-fueled buses to meet the 0.07 g/bhlp-h PM standard (Walsh 1995). Abatement Measure.s for Vehiiciular Air Pollution: Vehicle-Targeted Measures 63 Table 3.2 Estimates for pierformance and costs of exhaust emission control technologies for light-duty gasoline-fueled vehicles (Charge Estimaled Emnission sl(ndar'r in fitel rost per (grans per kiloteter) economy vehicle (Control level Control tecrrh oloq (.(0 H(; NOx (%) (J) Non-catalyst Ignition timing, air-futel ratio controls control, air injection, EGR 1.5 1.5 1.9 -5 130 Two-way catalyst Oxidation catalyst, ignitioni tiniTng, EGR 7.0 0.5 1.3 -5 380 Three-way catalyst Three-way catalyst, closed-loop carburetor 2.1 0.25 0.63 -5k' 630 Lean-burn engine Oxidation catalyst, EFI, fast-burn combustion chamber 1.0 0.25 0.63 15 630 U.S. Tier I Three-way catalyst, EFI, EGR 1.3 0.16 0.25 5 800 Notf: EGR is exhatust gas Tecirciulation; EFI is electronic fiuel inijectiori; U.S. Tier 1 is discussed in Annex A to this cliapter. a. At 80,000 kilometers. b. Change it ftsel economy is +5 percent (instead of -5 percenit) if EFI is used instead of a carburetor. So.rce: Faiz, Veaver, and Walsli 1996 (based on USEPA 1990). and costs of pollution control technologies are conditions should reflect as closely as pos- shown in Table 3.2 for light-duty gasoline-fueled sible actual fuel types and vehicle operat- vehicles and in Table 3.3 for heavy-duty diesel- ing conditions. Consideration should be fiueled vehicles. given to matching the test fuels with the ac- tual fuels used by vehicles and to measur- Certification, assembly line testing, recall, and war- ing CO and HC emissions at colder ranty. Emission standards for new vehicles are temperatures and higher altitudes, evapo- ineffective without a comprehensive program that rative emissions at higher temperatures, and ensures compliance. For this reason, new vehicle cold start and high load operations (Walsh emission standards are generally accompanied 1995). Among Latin American and Carib- by requirements for certification, assembly line bean countries, Brazil conducts certification testing, recall, and warranty. Manufacturers testing by regulatory authorities. Newvehicle should be required to: models manufactured in Argentina are taken to Brazil for certification. Certify that each of their new vehicles corn- * Perform assembly line testing during pro- plies with emission standards during a speci- duction to ensure compliance with appli- fied amount of use or time. For example, cable emission standards. Assembly line in Argentina and Brazil manufacturers of testing allows government authorities to light-duty vehicles are required to certify identify noncompliant vehicles and take in writing that their product will comply necessary measures (such as revoking of with emission standards for five years or certification). 80,000 kilometers (whichever comes first). * Recall vehicles in case of noncompliance The requirement for heavy-duty vehicles is with emission standards. Recall programs for five years or 160,000 kilometers. An al- allow manufacturers to design and produce ternative requirement is that the vehicle's vehicles with a margin of safety to avoid the emissions are 10 percent below the levels costs and negative publicity associated with specified in vehicle emission standards. recalls. Certification, which involves emission test- * Provide a warranty to consumers for defec- ing of prototype vehicles prior to procluc- tive design or workmanship of vehicles' tion, allows manufacturers to identify and emissions control equipment. Warranty correct problems prior to initiation of mnss programs allow manufacturers to design production of vehicles. Testing is usually and produce vehicles with controlled qual- done by manufacturers and also by govern- ity, and consumers to have a recourse for ment authorities through spot checks. Test noncompliant vehicles resulting from 64 Chapter3 Table 3.3 Estimates for performance and costs of exhaust emission control technologies for heavy-duty diesel-fueled vehicles Estimated Emission standarda Fuel cost '_ (g/bh p-h) economy engine Control level Controls requiired NO. PM (%) ($) Uncontrolled None (PM level depends on smoke controls anid main- 9.0-16.0 0.75-3.70 0 0 tenance level) Minimal control Injection timing, smoke limiter 8.0 0.5-0.75 -3 to 0 0-200 Moderate control Injection timing, combustion optimization 6.0 0.5 -5 to 0 0-1,500 1991 U.S. standard Variable injection timinlg, high pressure fuiel injection, combustion optimization, charge-air cooling 5.0 0.25 -5 to 5 1,000-3,000 Lowest diesel Electronic fuel injection, charge- standards under air cooling, combustion consideration optimization. exhiaust gas recirculation. catalytic con- verter or PM trap 2.0-4.0c 0.05-0.10 -10 to 0 2,000-6,000 Alternative-fuel Gasoline/three-way catalyst, forcing natural gas/lean-burn, natural gas/three-way catalyst, tnetliaiol-diesel 2.0 0.04 -30 to 0 0-5,000 a. At fuill uisefil life. b. Potential fuel economy improvements result fiom addition of turbocharging and intercooling to naturally aspirated engines. c. Not yet demonistrated in prodtictioni vehicles. Sowrce: Adapted from Faiz, Weaver, and W7alslh 1996. manufacturers' fault. Harmonization of test Exhaust emission test procedures for light- procedures among the trading Latin Ameri- duty vehicles include the United States test pro- can cotuntries is important. The implemen- cedure (FTP-75), the European test procedure, tation of international trade agreements and the Japanese test procedure. All three pro- would contribute to the regional harmoni- cedures measure exhaust emissions while the zation of emission standards and certifica- vehicle is driven through a specified driving cycle tion programs, avoiding unnecessary on a chassis dynamometer. Each procedure is investment in time and money for product based on a different driving cycle. In Latin Ameri- development and emission testing. Such can countries FTP-75 is the most commonly used agreements have already affected the au- procedure. This procedure simulates different tomotive manufacturing industty in Latin urban driving conditions up to a speed of91 kilo- America.2 meters an hour with an average speed of 31.4 kilometers an hour. However, it fails to cover all speeds and acceleration conditions, or air condi- tioner operation. These factors may cause sub- 2. For example. Brazil's exports are mainily targeted to tanel hger essions mai Wave and other Mercosuir coutntries (Argenitina, Paraguay, arid ud - U- v guay). In 1993 Brazil exported 331,522 vehicles, of whichi Walsh 1996). 75 percent went to the other Mercosur coutntries and only Test procedures for exhaust emissions from 13 percent went to other countries in the regioni. Similarly, heavy-duty vehicles have been established in the because of the North American Free Tradle Agreement United States, Europe, andJapan. All three pro- (NAFTA), in 1993, 91 percent of Mexico's ve'hicle exports cedures measure exhaust emissions of the en- (471,483) wenit to the United States and Canada. Onily 9 percent went to counitries in Latin America andi the Carib- gine removed from the vehicle using a chassis bean (AAMA 1995). dynamometer. Latin American countries use ei- Abatement Meas?res /or Velticular Air Pollution: Vehicle-Targeted Measures 65 ther the U.S. or European procedure. The U.S. emission standards in Argentina, Brazil, Chile, procedure, which replaced an earlier 13-mode Colombia, and Mexico. In addition, the environ- steady-state test, uses changes in transient speed mental standards promulgated in Bolivia in 1995 and load conditions to simulate in-city driving, limit HC and CO emissions from motor vehicles whereas the European procedure uses steady- with spark-ignition engines and smoke emissions state conditions aggregated according to a weight- from gasoline- and diesel-fueled motor vehicles. ing scheme. To standardize emissions from These standards are set by vehicle class and heavy-duty vehicles, which have a wide range in model-year. In Costa Rica CO and HC emissions sizes and applications, the test results are re- are limited from in-use gasoline-fueled vehicles ported in mass of pollutant emissions per unit of and smoke emissions are limited from in-use die- work output (grams per kilowatt-hour or grams sel-fueled vehicles. In Venezuela the current regu- per break horsepower-hour) rather than mass of lations restrict only smoke emissions from pollutant emissions per distance traveled diesel-fueled vehicles. Nicaragua is preparing in- (grams/kilometer). The USEPA uses a conver- use vehicle emission standards. sion factor of about 0.54 grams per break horse- Regulations that include emission standards power-hour to each gram per kilometer (Faiz, for in-use vehicles also specify procedures for Weaver, and Walsh 1996). compliance testing. The most common inspec- Evaporative emissions from gasoline-fueled tion test involves measuring CO and HC emis- vehicles are measured either by collection in ac- sions when the vehicle is idling. In the United tivated carbon traps or by putting the vehicle in States and some European countries (for ex- an airtight housing (SHED) and measuring the ample, Finland, Germany, Sweden) this test has HC concentration. The SHED method is favored been supplemented with a second measurement in the United States, Europe, and Latin America when the engine is running at 2,500 rpm with (CONCAWE 1994). no load. This test, although appropriate for older model vehicles equipped with mechanical car- buretors or fuel injection systems, does not give Emission Standards and Inspection satisfactory results for vehicles using electronic Programs for In-Use Vehicles air-fuel ratio control systems because it passes many vehicles that would otherwise fail a com- Emission standards are established to control prehensive test. To reduce false failures, some discharge of air pollutants from in-use vehicles inspection and maintenance programs using the through proper vehicle maintenance. Compli- idle/2,500 rpm test procedure require precon- ance with these standards are verified through ditioning of the vehicle at 2,500 rpm with no inspection and maintenance programs and road- load before the idle test. This preconditioning side inspections. helps ensure that the control system is in nor- mal closed-loop operation and the catalytic con- Emission standards. Emission standards for in-use verter is adequately warmed. Tests conducted in vehicles are generally set based on the vehicle Finland have shown that 95 percent of the ve- model-year for selected vehicle classes and en- hicles equipped with catalytic converters reach gine types. Older model-years are usually allowed stabilized readings in three minutes. In Austria more lenient emission limits. CO and HC are preconditioning of the vehicle prior to emission usually the pollutants limited for in-use vehicles testing is mandatory. Because of the problems with spark-ignition engines. New emission stan- with the idle/2,500 rpm testing procedure, other dards for in-use vehicles in polluted areas of the tests have been developed (such as ASM5015, United States also limit NO, emissions. Emission ASM2525, ASM2, IM240). IM240, the most ad- standards for in-use diesel-fueled vehicles usu- vanced of these tests, was recently introduced in ally focus on smoke emissions. the United States. The IM240 test measures pol- In general, industrial countries have estab- lutants in vehicle exhaust over a four-minute pe- lished emission standards for in-use motor ve- iiod based on the U.S. federal certification test hicles. The U.S. federal and the Arizona and procedures and is conducted by operating the Florida state emission standards are presented vehicle in a transient cycle simulating actual stop- in Annex A to this chapter. Some Latin Ameri- and-go driving conditions. This new test is can and Caribbean countries have also established claimed to be about three times more accurate in such standards. Chapter 4 presents in-use vehicle identifying noncompliant vehicles than the idle/ 66 Chlapter 3 2,500 rpm test. The main drawback of this proce- ing in-use vehicle emission standards. For ex- dure is its high capital cost requirement. Accord- ample, in Costa Rica all new light-duty vehicles ing to the USEPA, the test equipment costs entering circulation must be equipped with cata- $150,000 for the IM240 test and $5,000 to $12,000 lytic converters, but only 8 percent of in-use light- for the idle/2,500 rpm test. In addition, use of duty vehicles have them. In Colombia, catalytic the test equipment is more complex for the converter-equipped light-duty vehicles make up IM240 test than for the idle/2,500 rpm test (Faiz, less than 5 percent of the fleet and less than 10 Weaver, and Walsh 1996). percent of new vehicles entering circulation. In A successful measurement of diesel smoke Brazil 46 percent of light-duty vehicles are requires that measurements be taken when en- equipped with catalytic converters, and all new gines are under high-load conditions because the vehicles must be equipped with this technology. low smoke levels emitted under idle or light-load In Uruguay 43 percent of new vehicles are conditions do not easily allow distinction of clean equipped with catalytic converters (Alconsult In- vehicles from smoky vehicles. These standards ternational Ltd. 1996). generally require measurements be made follow- Technology-based requirements can also be ing the Bosch or Hartridge methods, or using an established for in-use vehicles. For example, ver- opacimeter. The Bosch method measures soot tical exhausts installed on in-use diesel-fueled and other dark materials in the smoke by pass- buses can reduce the concentration of exhaust ing a sample volume of smoke througlh filter pollutants at breathing level by 65 to 87 percent paper and determining the level of darkness on through dispersion (Faiz, Weaver, and Walsh the paper isingaphotoelectric device. However, 1996). This measure has been required in the this method responds poorly to smoke that is urban areas of many countries to lessen expo- not black. A Hartridge meter draws a continu- sure from diesel bus emissions. For example, ver- oUs sample of the vehicle exhaust emissions into tical exhausts have been required for buses in a chamber and measures the attenuation of a Santiago since 1987. A similar requirement is beam of light shining through the chamber. A being considered for buses in Santafe de Bogoti. Hartridge meter requires frequent cleaning from exposure to heavy oil smoke. An opacimeter Inspection and maintenance programs. Inspection measures the attenuation of a beam ot light shin- and maintenance programs involve periodic ing through the smoke plume. This method is measurement of emissions from in-use vehicles more accurate than the Bosch method because by regulatory authorities (or their designates) it measures both the effects of light absorption against established emission standards. Such (by soot) and light scattering (by oil or fuel drop- programs: lets) of the entire exhaust emissions. Emission standards for in-use vehicles should * Provide regulatory authorities with an ef- be based on the potential emission reductions fective tool for enforcing established stan- that can be achieved through proper mainte- dards for in-use vehicles. nance of these vehicles (given the age distribu- * Put pressure on vehicle owners to bring tion of the fleet and extent of use of pollution their vehicles into compliance (through control technologies). Many of the vehicle fleets maintenance) prior to inspection. in Latin America and the Caribbean are old and * Force gross polluters-vehicles unable to poorly maintained. For example, the average age meet the standards through maintenance of cars is 13 years in Costa Rica (1993), 23 years in or repair-out of circulation. For example, Paraguay (1994), and 12 years in Venezuela an exhaust test of 60,000 vehicles in the (Alconsult International Ltd. 1996). Vehicles United Kingdom found that 12 percent of older than 10 years constitute 50 percent of the cars were responsible for 55 percent of pol- motor vehicle fleet in Argentina, 60 percent in lutant emissions (Reynier 1995). A recent Ecuador, and 64 percent in El Salvador (both in survey of five European cities found that 1996). In Lima about 75 percent of vehicles are about 10 percent of the vehicle fleet is re- more than 10 years old, and in Santafe de Bogota sponsible for about 60 percent of CO and about half of the public transport fleet is more 45 percent of HC emissions, and that about than 20 years old (both in 1996). 90 percent of HC and NO, emissions origi- The extent of pollution control technologyuse nate from vehicles that are not equipped is also an important consideration in establish- with catalytic converters (van der Straaten Abatement Measu res for VehicularAirPollution: Vehicle-Targeted Measures 67 1995). The situation is similar for urban grams allows vehicle owners to drive shorter dis- centers in Latin America and the Carib- tances for inspection. However, inspection results bean. For example, in 1991 in Mexico City may lack the accuracy of automated centralized the most polluting 25 percent of the fleet stations. In addition, inspections usually are not accounted for 47 percent of CO and 62 performed very promptly because the main func- percent of HC emissions from mobile tion of these garages is to repair vehicles. Fur- sources (Beaton, Bishop, and Stedman thermore, the regulatory authority must employ 1992). a larger surveillance team in order to control the * Identify in-use vehicles that require recalls quality of service provided by the large number by vehicle manufacturers. of private garages. * Identify in-use vehicle component defects The design of the inspection and maintenance and failures covered by vehicle manufac- program affects the share of the in-use vehicle turers' warranty programs. fleet inspected at centralized stations relative to * Provide feedback to regulatory authorities private garages. For example, in NewJersey, where on the need for investigations and further vehicle owners have a choice between free cen- tests by vehicle manufacturers on vehicles tralized inspections and paid inspections in pri- with consistently high emissions. vate garages, 80 percent of drivers have chosen * Discourage vehicle owners from tampering centralized inspection. But in Mexico City, just with emission control systems or misfueling 20 percent of the vehicle fleet was inspected at their vehicles because failing an inspection centralized public stations in the early 1990s be- acts as a strong deterrent. Experience in cause of the associated inconvenience: the ve- the United States has shown that faulty, or hicle owners had to first go to an agency of the tampered vehicles can emit up to twventy Treasury to pay for the inspection and then take times more pollutants than properly main- the receipt to a centralized inspection station. In tained vehicles. addition, there were often lines at the central- ized stations. People who used the centralized Inspection and maintenance programs are stations perceived this service to be more honest usually of two types: centralized programs, which because of the potential for fraud by private ga- require all inspections be done in stations spe- rages. Integrating fee payment and inspection at cialized in emissions testing and with no vehicle the same facilitywould reduce any inconvenience repair capability, and decentralized programs, associated with central inspection centers which require both emissions testing and vehicle (McGregor and Weaver 1994). repairs be done in private garages. In central- Centralized inspections are much cheaper for ized programs vehicles must be inspected at one vehicle owners because of their favorable of a small number of high-volume inspection economy of scale. The nine centralized, contrac- stations. If a vehicle fails an inspection at a cen- tor-nm programs in the United States cost, on tralized station, then the driver has to take it else- average, less than $8 an inspection. The central- where to be repaired and return it to the station ized programs in Arizona, Connecticut, and to have it reinspected. The centralized inspec- Wisconsin use dynamometers and loaded test tion stations are govemment-controlled and rtin procedures and charge fees ranging from by either government employees orindependent $7.50-10. In California's decentralized idle/ contractors. Governments typically franchise a 2,500 rpm program inspection fees average about single contractor to build and operate the inspec- $40. In most cases IM240 testing in centralized tion station in a given area and allow the contrac- facilities costs less than $20 a vehicle in the United tor to charge a set fee that allows for profit and States (McGregor and Weaver 1994). For com- recovery of operating costs and capital during the parative purposes Table 3.4 presents some cost franchise period. Governments also may issue estimates for centralized and decentralized pro- multiple franchises in an area to reduce its de- grams in the state of Arizona. Based on these esti- pendence on any one contractor (for example, mates, as well as on its greater effectiveness, in Mexico City franchises were awarded to five Arizona decided to retain a centralized system. contractors). Franchises are usually awarded us- A centralized program is likely to have higher ing a competitive bidding process. start-up costs than a decentralized program be- The large number of private garages in dle- cause of the costs associated with constructing centralized inspection and maintenance pro- centralized inspection stations. In Maryland, for 68 Chapter3 Table 3.4 Estimated costs of centralized and decentralized inspection and maintenance programs in Arizona Derentrzlized Gentralized Centralized nautomazted Vamriable matnatal lane atruomated lane inspection station Tests per hour 6-12 20-30 2-5 Tests per man-lhour 1.5-4.0 6-15 1-3 Labor cost per test $2.00$5.33A $0.53:$1.33a $4.00-$12.00b Equipment cost per test $0.50-$1.00 $0.50-$1.00 $3.00-$5.00 Facilities cost per test $2.00-$3.00 $1.50-$2.00 $5.00-$10.00 Data collection cost per test $0.50-$ 1.00 $0.10-$0.25 $0.504$1.00 Administration cost per test $0.35-$0.50 $0.20-$0.30 $1.00$2.00 Total cost' $6.00-$10.18 $3.08-$4.63 $15.00-$28.50 a. At $8.00 an lhour. b. At $12.00 an hour. c. Does niot incltude overhead or profit. ,Source: Rothe 1990. example, a private contractor was selected clude licensing for numerous stations, certifica- through competitive biddingto constnrctten new tion of repair facilities, and supervision of inspection stations, with an average of five lanes stations. each. These stations are expected to inspect 1.6 The frequency ofvehicle emissions testingvar- million vehicles a year. A large portion of the fee ies by location. In the United States vehicles are charged by each station is retained to recover its inspected annually in forty-five of the fifty states. investment and operating costs; the rest is re- In Japan new vehicles require inspection after turned to the state to cover the administrative the first three years and every two years thereaf- costs of the program (Faiz, Weaver; and Walsh ter. In the European Union the inspection sched- 1996). ule depends on the vehicle type: for heavy-duty A centralized program does not alvays have diesel vehicles an annual inspection is required high investment costs. In New Jersey, for ex- regardless of the vehicle's age; for spark-ignition ample, the state incorporated its emissions test- vehicles an annual inspection is required for ve- ing program into an existing centralized safety hicles older than three years; and for light-duty inspection program, thus avoiding the cost of diesel vehicles an annual inspection is per- new construction (Faiz, Weaver, and Walsh formed after the vehicle is four years old and ev- 1996). Decentralized programs are less expen- ery two years thereafter. In Singapore a different sive to set up than centralized programs but are inspection schedule is used for different type of more expensive to operate. Operating costs in- vehicles (Table 3.5). Table 3.5 Motor vehicle inspection schedule in Singapore, by vehicle age Ipe (of vehicle l.sxs tlhoun three years T'hree to ten years More than ten yeaers Motorcycles and scooters Exempt Anniually Annually Cars and station wagons Exempt Every two years Annually Private hire cars Antuially Annually Not required Public service vehicles Taxis Every six months Every six monthis Not required Public tranisport buses Everv six moniths Every six months Every six months Trucks and goods vehicles Anntually Arnniually Annually All other heavy vehicles Annually Annually Annually .Sourre: Faiz, Weaver, and Walsh 1996. Abatement Measures f or VehlicularAirPollution: Vehicle-Targeted Measures 69 In the Mexico City Metropolitan Area (MCMA) private cars and trucks, and twice a year for taxis a voluntary inspection and maintenance program and buses. was introduced in 1988. In 1989 inspection of in- In Sao Paulo the Program to Control Air Pol- use vehicles was made compulsory with one test lution from Motor Vehicles (PROCONVE) au- per year; in 1990 the number of inspections was thorizes state and local governments to inspect increased to two per year. This program encom- in-use vehicles for pollutant emissions. The Sao passed all gasoline-fueled and diesel-fueled ve- Paulo inspection and maintenance program is hicles. Inspections were carried out at about 1,650 expected to start in 1997 (see Chapter 4). authorized private and 26 high-volume govern- In Costa Rica the Transportation Law speci- ment-operated centralized inspection stations. fies the inspection schedule of motor vehicles. However, it became increasingly clear that the The schedule is annual for all light-duty vehicles private garages were not performing inspections except cars less than five years old, every two properly, and sinceJanuary 1996 all inspections years for cars less than five years old, and every in the MCMA have taken place at one of the cen- six months for public transport vehicles (taxis, tralized facilities. Although this change has buses, and microbuses) and heavy-duty trucks. caused serious congestion problems at the sta- Inspections are conducted at government-ap- tions, those problems are expected to be over- proved private garages (Weaver 1996). come through an expanded system that will be The on-board emission diagnostics installed put in place in 1997. in newer model vehicles in the United States are In the Buenos Aires Metropolitan Area three expected to become common in other countries different jurisdictions are involved in vehicle as well, including those in Latin America and the inspections. Periodic inspections of exhaust Caribbean. These systems will likely simplify pe- emissions from commercial vehicles operating riodic inspection tests, perhaps requiring no more within the Federal Capital are conducted by a than a check of the system's indicator light, and private firm under a ten-year concession agree- will help service technicians to identify and re- ment with the Municipality of Buenos Aires. Pe- pair malfunctions (Faiz, Weaver, Walsh 1996). riodic inspections of private cars registered in the Federal Capital have not started yet. Howv- Roadside inspection programs. In urban centers ever, the Municipality of Buenos Aires has be- where periodic inspection and maintenance pro- gun the bidding process for concessioning of grams are not implemented (such as Belo periodic inspection services to private firms. The Horizonte, Rio de Janeiro, Santafe de Bogota, Province of Buenos Aires recently started peni- and Sao Paulo) roadside inspections help iden- odic inspections of private and commercial ve- tify grossly polluting vehicles. These inspec- hicles under concession agreements between tions-performed on the roadside and weight provincial authorities and private firms. The in- stations or safety inspection stations-have fo- spection schedule is set at once a year for pri- cused mostly on visible smoke emissions from vate cars and trucks, and twice a year for taxis, diesel-fueled vehicles, especially buses and private chauffeured cars, and buses. Commer- trucks. Roadside inspections involve stopping cial vehicles operating between the Federal Capi- vehicles that appear to emit excessive amounts tal and a province or between two provinces are of smoke and testing the exhaust emissions us- inspected at private stations certified by national ing a smoke opacimeter. Vehicles found to be in authorities. Inspection frequency is set at six violation of smoke emission standards are fined months for buses and twelve months for trucks. or prohibited from circulation until the prob- Inspections include safety and mechanical checks lem is resolved. as well as measurement of smoke emissions for Roadside inspections have also been used to diesel-fueled vehicles and CO and HC emissions supplement inspection and maintenance pro- for spark-ignition vehicles. grams in Latin American urban centers such as Several years ago periodic inspection and Mexico City and Santiago. Because the time and maintenance was made mandatory for all ve- place of these inspections cannot be predicted hicles in Santiago. There are two inspection sta- they are difficult to defeat through tampering of tions for buses, three for taxis and trucks, and vehicle components that affect exhaust emissions. twenty for private cars. All inspection stations Roadside inspections for invisible CO and HC are owned and operated by private concession- emissions involve stopping vehicles at random, aires. The inspection schedule is once a year for which inconveniences drivers of vehicles that are 70 Chapter 3 in compliance with the standards. However, a new information about the actual performance of in- technology that measures concentrations of CO use vehicles, especially those which are exces- and HC by infrared remote sensing coupled with sively worn, poorly maintained, or used with a video camera shows promise in identifying gross tampered emission control equipment. To ensure polluters without stopping vehicles. This tech- that the imported in-use vehicle is in acceptable nology, however, is limited to measuring only in- condition, a recent certification of emission test- stantaneous emissions (which may not be ing in the country of origin maybe required. As a representative of the vehicle's overall emissions) supplement or alternative to this requirement, in one lane of a road. The potential application all imported in-use vehicles may be subject to of remote sensing for emissions compliance ap- emission testing in the importing country pears to be bringing gross polluters to inspec- tion stations. Vehicle Emission Fees Emission Restrictions on Imported A vehicle emission fee is a charge levied by the Vehicles government on the amount of pollutants emit- ted by the vehicle. A uniform fee can be set by Many countries impose restrictions on imported vehicle type (such as heavy-duty vehicles) based vehicles as a measure to limit emissions of air on average emissions determined from driving pollutants. In general, these restrictions are as- cycle emission tests. Another option may be to set sociated with the model-year or age of the ve- a fee for each vehicle based on its emission fac- hicle, use of a certain emission control technology, tors and the distance traveled since the last in- or compliance with in-country new vehicle emis- spection. This scheme encourages vehicle sion standards. For example, in Costa Rica all owners to adjust the engine before inspection to motor vehicles imported to the country afterJani- reduce emissions, but also creates an incentive ary 1995 must be equipped with emissions con- to cheat by tampering with the odometer. Avaria- trol devices and meet the emission standards for tion of this scheme, which involves levying fees new vehicles. In Bolivia all motor vehicles that on new vehicle models based on their expected enter into circulation should meet the emission lifetime emissions, may induce demand for standards established in 1995. In Ecuador, al- cleanervehicles, but does not provide any incen- though no emission standards have been set for tive for vehicle owners to properly maintain their new vehicles, import of used vehicles is prohib- vehicles or drive less. Emission fees are not cur- ited effective August 1996. In El Salvador all ve- rently imposed on vehicles anywhere in the world, hicles imported after January 1998 must be although the European Union is considering the equipped with an emission control system. In introduction of emission fees for heavy-duty trucks Nicaragua emission standards are being pre- (Carbajo 1994). pared for imported vehicles according to the le- gal framework established in 1996. Imported vehicles should be checked to en- Differential Vehicle Registration Fees sure compliance with local requirements. Some and Taxes countries in Latin America and the Caribbean (such as Argentina) require certification testing The vehicle registration fee or taxing scheme of vehicles in their country of origin. A visual can be designed so that it influences consumers inspection of the emission control label (located to buy fuel-efficient and low-polluting vehicles under the hood in U.S.-manufactured vehicles) (such as vehicles equipped with catalytic convert- confirms that the vehicle complies with emission ers) rather than energy-inefficient and highly standards in the country of origin (McGregor polluting vehicles. This goal can be achieved by and Weaver 1994). This information and a vi- using differential vehicle registration or taxation sual check of all emission control equipment (Carbajo 1994). listed in the label would be sufficient for inspect- The vehicle registration fee scheme can be ing a new vehicle at the point of entry to the part of a vehicle replacement program aimed at importing country. Although review of the emis- taking old polluting vehicles out of circulation. sion label is useful, it may not provide accurate For example, in Singapore car owners who re- Abatement Measures for VehicularAirPollution: Vehicle-Targeted Measures 71 place their cars within ten years with new cars are buses exceeding the 40-Hartridge-unit standard entitled to exemption from registration fees. This were taken from service for revision. Fuel injec- is a significant benefit because car registration tion pumps were also recalibrated and injectors fees are very high in Singapore. They include a were replaced to reduce other air emissions from 45 percentimportdutyon the open marketvalhie these buses. In 1996 about 1,000 heavy-duty die- of the car, a registration fee of $640, and an sel-fueled vehicles installed catalytic converters additional registration fee of 150 percent on the to be exempt from the two-day driving ban in open market value of the car. In addition, dutring Mexico City. In addition, under a government- its use each car owner has to pay an annual road sponsored program about 27,000 buses and tax based on the vehicles' engine capacity trucks and 1,300 commercial microbuses have (Carbajo 1994). their fuel systems converted to LPG. Annual vehicle taxes can favor less-polluting The urban bus retrofit program in Romania vehicles through differential taxation based on involves installing a deep-bed particle filter in the use of pollution control technologies or ve- the exhaust system of diesel-fueled buses. This hicle age. For example, in Germany vehicles that technology requires the addition of an iron- are not equipped with emission control technolo- based catalyst to the fuel to lower the combus- gies to comply with European Union regulations tion temperature of soot. Deep-bed filtration can are subject to a higher tax (based on 100 cubic reduce black smoke emissions from heavy-duty centimeters of engine displacement) than ve- diesel-fueled vehicles by about 85 percent, but hicles that comply. Similarly, vehicles equipped with a fuel penalty of 2 to 6 percent. The in- with catalytic converters are given a tax credit stalled system costs about $1,500 in Romania that is positively correlated with engine displace- (Bloom 1997). ment volume. Another technology, which has been tested in The design of the vehicle taxing scheme can England, shows some promise for retrofitting also incorporate the vehicle age to favor less- diesel-futeled urban buses. A secondary filter, polluting and newer vehicles. But higher taxes which supplements the conventional cartridge- for older vehicles would likely be regressive and filter, can capture the light fraction of lubricat- politically difficult to implement. In some coun- ing oil. The captured oil is vaporized by a heating tries, however, the vehicle tax declines with the element and recycled to the fuel inlet. This tech- age of the vehicle. This is the case for Mexico, nology is claimed to reduce the PM emissions where vehicles ten years or older are exempt by about 50 percent and cost about $500 (The from the annual vehicle tax. This scheme has Economist, 23 November 1996). adverse environmental consequences because it A ceramic coating technology has been ap- encourages the presence of old, highly polhlt- plied to in-use diesel-fueled buses in the United ing cars in the fleet. States and is being tested on diesel engines in Mexico City. When ceramic coating is applied to combustion area components of a diesel en- Vehicle Retrofit Programs gine, the coated engine emits 50 to 90 percent less black smoke and generates a fuel savings of Retrofit programs are implemented to bring in- up to 9 percent. When combined with engine use vehicles manufactured without emission con- retarding, this technology can remove up to 40 trols in compliance with emission standards. percent of NO.. Use of this technology with an Retrofitting may take different forms inclucding oxidation catalyst can remove up to 70 percent engine work, installing pollution control compo- PM, up to 50 percent of CO, and up to 40 per- nents, or changing fuel systems to burn cleaner cent NO, (Viola 1996). fuels. These programs, which are generally associ- In Argentina about 400,000 vehicles have their ated with incentives, target grossly polluting high- fuel systems converted to CNG, about 270,000 of use vehicles, especially diesel-fueled urban buses. which circulate in Buenos Aires. CNG-fueled ve- Between 1991 and 1995 about 4,100 urban hicles in Buenos Aires include a large percent- buses in Mexico City were retrofitted with new age of taxis and light-duty trucks, and some engines to reduce pollutant emissions. With es- private cars. The Argentine government's differ- tablishment of the Permanent Program for Cor- ential taxation policy for these fuels was the in- recting Excessive Black Smoke in early 1992, centive for these conversions. 72 Chapter3 Vehicle Replacement or Scrappage Congress/OTA 1992). Between 1990 and 1995, Incentives 9,200 heavily polluting vehicles were removed from circulation in southern California. The Vehicle replacement or scrappage programs are fourth phase of this program (SCRAP IV) is un- designed to eliminate the most polluting velhicles der way. This phase targets pre-1975 vehicles, (gross polluters) that are responsible for a dis- which can emit 50 to 100 times more pollutants proportionate share of vehicular air pollutant per kilometer traveled than new vehicles and emissions in a given area. These programs pro- which account for disproportionately high vide financial incentives to scrap or replace such amounts of pollutant emissions among mobile vehicles. A classic example of such a program is sources in the Los Angeles Basin. According to from Hungary, where the City of Budapest initi- the estimates made by UNOCAL, taking these ated the "green-two-stroke" program, which pro- old cars out of circulation will eliminate 45,000 vided public transport passes to car owners in kilograms of HC, 14,000 kilograms of NOV, and exchange for voluntary elimination of high-pol- 273,000 kilograms of CO emissions each year over luting cars. This program also allowed car owners the next three years (Oil and GasJournal, 1995). to sell their Trabants and Wartburgs (former East In February 1994 a car replacement program German two-stroke engine models) to the city for was introduced in France. This program entitled a higher price than the prevailing market rate owners of cars older than ten years to replace and to use the money as part of the downpayment them with a new car for $936. Within a year on new cars equipped with emission controls. In 450,000 old cars had been replaced by new cars February 1994 more than 14,000 car owners ap- that complied with the latest emission standards. plied to exchange their cars for public transport This program generated a noticeable reduction passes at the rate of a four-year pass for each in motor vehicle pollutant emissions-within a Trabant and a six-year pass for each Wartburg. year CO emissions had fallen by 6 percent In 1990, as part of the South Coast Recycled (Reynier 1995). In 1996 the Municipality of Atuto Project (SCRAP), the Union Oil Company Quito (Ecuador) eliminated about 1,000 diesel- of California (UNOCAL) initiated a program to fueled buses older than twenty years from circu- scrap 7,000 pre-1971 cars for $700 each (U.S. lation for about $2,560 per bus. FUEL-TARGETED MEASURES Fuel standards to control air pollutant emissions add lead compounds-such as tetra-ethyl lead from vehicles typically target lead, volatility, and (TEL) or tetra-metyl lead (TML)-to gasoline. benzene and other aromatic hydrocarbons in The octane improvement of gasoline decreases gasoline and sulfur, cetane number, aromatic with each increment of lead added, however hydrocarbons, and density in diesel fuel. Produc- (Figure 3.1). With the exception of two small tion and use of reformulated and oxygenated plants in Russia and in Germany, Octel Ltd. is gasoline also have been required in some coun- the only company in the world that produces tries. In addition, alternative fuels have been gasoline lead additives (Thomas 1995a). used in place of conventional fuels (gasoline and Besides boosting octane, TEL lubricates the diesel fuel) in some countries as an urban air exhaust valve seats of vehicles. This lubrication pollution control measure or as part of a national prevents wear (called "valve seat recession") energy policy. The most common alternative under severe driving conditions (prolonged high fuels include compressed natural gas (CNG), liq- speed, towing, and hilly terrain) of older model uefied petroleum gas (LPG), methanol, ethanol, vehicles manufactured with soft valve seats. This and mixtures of gasoline and ethanol (gasohol). concern was the reason lead was not completely eliminated from gasoline in the United States Gasoline Standards (where the limit for lead was 0.0267 g/liter for several years) and Europe (where the limit for Standards for gasoline have been set in many lead is 0. 15 g/liter). industrial and some developing countries to Lead antiknock compounds in gasoline, the reduce or eliminate the adverse health and main source of ambient lead concentrations in environmental effects associated with its han- most urban environments, have adverse health dling and use. These standards affect fuel refin- impacts (see Chapter 2). Reducing or phasing ers and distributors, vehicle manufacturers, and out lead in gasoline is the most effective way to consumers. The most common gasoline stan- lower ambient lead concentrations. Lead in gaso- dards affecting motor vehicle emissions include line was first restricted in 1967 in some urban parameters for lead, volatility, benzene, and centers (Moscow, Leningrad, Kiev, Baku, other aromatic hydrocarbons. In addition, stan- dards are established for reformulated and oxyv- genated gasoline. 3. The octane number ranges between 0 for n-heptane and 100 for iso-octane. The higher a gasoline's octane is, Lead. Poorly timed combustion of the gasoline- the better is its antiknock performace. Gasolines have two air mixture causes a knocking effect in spark- octane ratings: research octane number (RON), which mea- ignition vehicles. This effect, which lowers an sures anitiknock perfornanice at low engine speeds; and engine's efficiency and useful life, is reduce M61 . . . . . . . . . . . . . . . . . . . 661 ........ ... - .. .. ... ...... ------------ Z,661 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1661 ............. - - .............. - ............ 0661 . . . . . . . - - - - - - - - - - - . . . . . . . . . . . . . . . . . . . . . . - - - - - - - - - - - - - - - - .. . . . . . . . . . . . . . . . . . . . 6861 0 ------- ........... ....... .......... ...... - - ................ .. ..... ..... 8861 ------------------- ....... - ........... - ........... ...................... 4861 : ,:., : , : : : , , , - - - - - - - - - - - - - - - - - - - - - - - - - - . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . bb 9861 ------------------- . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I . - . . . . . . . . . . I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C861 7 . . . . . . . . . . . . . .. . . . . . . . . . . . . . - - - - - - - - - - . . . . . . . . . . . . . . . . . . . . . . . . . . I . . . . . . . . . . . . . . . . . . OP f,861 U . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . .. .... 0 F861 Z861 b ............... ........ ...... - - ................. ..................... - ................... ......... -.-- ........... r m CZ 40 u 'M CZ (U m cl: u " 6. cli 0 CZ CZ m C>, ct 1) 0 > N u 0 0 U m cl, u CZ = :R -a E bl, ct .0 ho :Z > E PQ CZ CZ 0 E bc 0 P -0 CZ Cd rz u m E Qd E < 0 u U0 z Abatement Measures for Vehicular Air Pollution: Fuel-Targeted Measures 77 Table 3.6 Standards, quality, and consumption of leaded gasoline in Latin America and the Caribbean, 1996 Leaded gaso6ine lReq,11al'f/ Mlatkeled ronsumnltion h01ad onmtent lead content (lercentage (Cousntry (gram/llilr) (granm/liter) of total gasoline) Barbados - 0.79 60 Chile 0.f0 0.31 72 Dominican Republic 0.40 0.40 69 Ecuador (0.84 0.50 76 Jamaica - 0.77 70 Mexico 0.06-0.08 0.11 44 Netherlands Antillesa - 0.23 50 Panama 0.76 0.45-0.63 7 Paraguay 0.15 0.20 99 Peru 0.74 0.41-0.62 75 Trinidad and Tobago 0.40 0.15-0.45 99 Uruguay - 0.58 96 Venezuela 0.85 0.30-0.40 100 -Not available. a. Data is only for Aruba. Source: Alconsult International Ltd. 1996; Pesce 1996. may involve modifying, expanding, or adding a The Chilean government's policy for short-tern catalytic reforming unit.6 Increasing the sever- refinery upgrading includes remodeling exist- ity' of the catalytic reforming unit, however, in- ing catalytic cracking units (replacing the cata- creases the yield of aromatic hydrocarbons, lysts and operating at conditions thatwouldyield which results in high benzene emissions with a product with the maximum achievable octane adverse health effects. By investing in alkylation' number), constructing new low-pressure catalytic or isomerization'' units the octane number can reforming units, and converting existing high- also be increased. Another option to refiners pressure catalytic reforming units to isomeriza- involves investing in an oxygenate plant that can tion units. The government's long-term policy produce or blend an oxygenate with gasoline. includes adding a continucus regeneration unit to the catalytic reforming units and producing methyl tertiary-butyl ether (MTBE) and di-iso- 6. Catalytic reforming iinvolves catalytic reaction of lowv- propyl ether (DIPE). Refinery upgrading in Ar- octane naphtha (35-55 RON) with hydrogen tinder pres- gentina has involved investments for catalytic sure (4.4-34 atmospheres) and at high temperatiures reformer modifications and for new reforming, (450-510oC) to produce a high-octane naphtha (90-102 isomerization, alkylation, and MTBE units. Simi- RON). 7. Severity can be increased by reducing pressure or in- lar modifications have taken place in Mexican creasing temperature. (see Chapter 4) and Venezuelan refineries. 10 Re- 8. Alkylation involves reaction of olefins having three to finery upgrade investments are under way in Ec- five carbon atoms in their stnicttires (that is, C, to C.) wvitl uador and Nicaragua and are planned in isobutanie in the presence of a liquiid catalyst (suchi as stilfu- Argentina, Costa Ricajamaica, Mexico, Trinidad ric or hydrofluoric acid). The CG to C, olefins (propylene A n-btutene, isobtitene) are produced tnainly froni the fluid and Tobago, Uruguay, and Venezuela. catalytic cracking (FCC) Ullit. The alkylation reaction pro(1- Refinery upgrading, especially when a new ucts (alkylates) are high octane blendstocks (92-97 RON). unit is required, is capital intensive. Estimates Alkylation is suitable only for refineries equipped with FCC tiits. 9. Isomerization involves producing an isonmer from light naphtha (that is, the C5 and Q fraction from the atmospheric 10. The refinery upgrade investmenits in Venezuiela have distillation tiiit). This process increases the RON frolim been made to produce unleaded gasoline for the export 70-78 to 85-90. market. 78 Chapter 3 for refinery investments to increase gasoline's higher emissions of aldehydes (such as acetal- octane number are provided in Chapter 4 in the dehyde) than gasoline but result in lower CO Mexico City and Santiago sections. It costs $0.02 and HC emissions. Ethanol (an alcohol-based to $0.03 a liter to remove lead from gasoline with oxygenate) is more volatile than ether-based initial levels of 0.6 g/liter or more, and $0.01 to oxygenates. $0.02 a liter for initial levels of about ().15 g/ The minimum engine octane requirement of liter (World Bank 1996a). The refining cost of older model gasoline-fueled vehicles (equipped removing lead from gasoline is lower for com- with carburetors or continuous fuel injection plex (deep conversion) refineries thani for less systems) is influenced by altitude. The octane sophisticated (for example, hydroskimming) re- number requirement decreases by about 10 units fineries, which require more expensive processes for an altitude change from sea level to 1,000 (USEPA 1995)." meters (Faiz, Weaver, and Walsh 1996). By sup- Some Latin American and Caribbean coun- plying lower octane gasoline in urban centers at tries meet the local market demand for unleaded higher altitudes savings can be achieved in re- gasoline by importing. The imported gasoline finery investments or in foreign exchange or naphtha is either blended with locally pro- needed for importing oxygenates, high-octane duced low-octane unleaded gasoline or naphtha naphtha, or high-octane gasoline. If leaded gaso- (as in Costa Rica, Dominican Republic, El Salva- line is still used in the country, the minimum dor, and Nicaragua) or sold in the local market octane requirement may be met by reduced lead (as in Barbados, Honduras, Paraguay, St. Kitts, addition to gasoline. In Latin America such a and St. Lucia). In some countries (Chile, Gua- strategy may be beneficial for countries (such as temala, Uruguay) locally produced low-octane Bolivia, Chile, Colombia, Ecuador, and Mexico) gasoline or naphtha is blended with an imported with urban centers at high altitudes (such as La oxygenate (MTBE or ethanol) to produLce a high- Paz, Santiago, Santaf6 de Bogota, Quito, Mexico octane gasoline. In Costa Rica high-octane gaso- City, Monterrey, Puebla de Zaragoza, and Le6n). line is produced also by blending imported The implications for motor vehicle compo- high-octane gasoline with imported MTBE. The nents of shifting from leaded to unleaded gaso- series of events that led to the elimination of line also should be considered. Because lead lead from gasoline in Costa Rica are presented serves as a lubricant between exhaust valves and in Box 3.4. their seats, unleaded gasoline may cause valve Although reducing or eliminating lead in gaso- seat recession under severe driving conditions line definitely has positive health impacts, con- in vehicles without hardened seat valves. But this sideration should also be given to the health problem is not likely to occur tinder normal driv- impacts of the gasoline produced (after refin- ing conditions and in newer vehicles with hard- ery modifications) to meet lower lead require- ened valve seats. Moreover, a lead concentration ments. The reforming process increases the of 0.05 g/liter of gasoline is sufficient to pre- content of benzene and other aromatic com- vent seat valve recession in vehicles without hard- pounds in the gasoline blend. Benzene is carci- ened seat valves. And the presence of lead in nogenic. In addition, this and other- aromatic gasoline increases the risk of other vehicle main- compounds are toxic. Olefins and alkylbenzenes tenance problems that may be more severe than (such as m-xylene and trimethylbenzene), when valve seat recession (Thomas 1995a) 1 evaporated, have high reactivity for ozone for- A country's lead elimination strategy should mation. Lighter hydrocarbons have higher vola- include a realistic time schedule that takes into tility than the other hydrocarbons in the gasoline account, among other factors, the availability and blend. Oxygenates, upon combustiotn, produce capabilities of domestic petroleum refineries. In most Latin American and Caribbean countries production of unleaded gasoline to meet octane 11. Hydroskimming refineries inchiide atmosplheric and number requirements is not likely to be imme- vactumn distillationi, hydrodesulftirizationi, catalytic refonr- ing, and bleniding and possibly isonierizationi operations. Deep conversion refineries iicitl(le the same openations as hydroskimming refineries and other operations sucth as flhiid 12. Lead scavengers stich as ethylene chloride or ethyl- catalytic cracking, coking, hydr-ocrakinig, isomerfization, alky- ene dibroniide foriim corrosive halogetiated compotinds that lationi, polymerization, and oxygenate (sucih as MTBE or degrade exhatist valves, spark pltigs, mufflers, and exhatist TAME) prodtiction. pipes. Abatement Measures for Vehicular Air Pollution: Fuel-Targeted Measures 79 Box 3.4 How was lead eliminated from gasoline in Costa Rica? A series of events led to the elimination of lead from gasoline in Costa Rica: * Promulgation of Decree 19,088-S-MEIC-MIRENEM, which established a seven-year period to eliminate lead from gasoline (1989). * Introduction of 95-octane super gasoline, which is imported (1990). * Undertaking of technical studies to evaluate alternative options for producing unleaded gasoline with 88 octane (1991-92). * Promulgation of the Transit Law, which establishes requirements for motor vehicle registration and emis- sion standards (1993). * Opinion survey of regular gasoline consumers to gather their views about the introduction of unleaded gasoline to the market and to devise a publlic awareness campaign strategy (1994). * Introduction of an environmentally friendly super gasoline (1994). This gasoline is produced by blending imported high-octane unleaded gasoline with 10 percent by volume MTBE content (also imported). * A massive campaign on the benefits of enivironmentally friendly super gasoline through television an- nouncements and distribution of pamphlets (1994) * Establishment of an inter-institutional commission for all entities involved in the elimination of lead from gasoline, including vehicle importers, the agency in charge of regulating prices, the Costa Rican refinery (RECOPE), fuel consumers, and ministries of the environment, energy, public works, and transportation (1994). * Evaluation of the impacts of lead elimination from regular gasoline (1995). * Use of unleaded gasoline in RECOPE's entire fleet to prove to consumers that unleaded gasoline has no adverse impacts on engines (1995). * Reduction of the price difference between regular and unleaded gasoline to 5 percent (1995). * Promulgation of Decree 24,637-MIRENEM-S. which prohibited the use of leaded gasoline in public insti- tution vehicles (1995). * Introduction of a new unleaded gasoline called "bio-plus" (April 1996). This was done without any public announcement to allow time to clean the storage and distribution network of any residual lead. Bio-plus is prepared by blending naphtha produced by RECOPE and imported high-octane naphtha. * Announcement of the introduction of bio-plus to the market and associated media campaigns through television and press (May 1996). Source: Ministry of the Environment and Energy of the Republic of Costa Rica 1996. diately possible because of the lack of adeqttate and final sale of the unleaded gasoline to the refinery capacity and infeasibility of suitable pro- consumer (such as labeling, testing, reporting cess reconfiguration. Furthermore, investments requirements, and prohibition of misfueling) as for certain refinery reconfigurations may take well as penalties for violation. While compliance two to five years to materialize. In the interim it with fuel specifications has to be ensured and might be possible to modify certain operating enforced, prohibition of misfueling is not likely parameters of the refinery process units. In ad- to be effective if gasoline pricing creates incen- dition, gasoline can be blended with oxygenated tives for misfueling. To avoid misfueling, the two additives or imported unleaded gasoline. Any gasoline grade should be taxed differentially so deficit in the octane number could then be rem- that, at a minimum, unleaded gasoline is not sold edied by blending with lead additives. at a higher retail price than leaded gasoline. A lead elimination strategy with a realistic Finally, any lead elimination strategy should schedule also should include policies other than be considered separately from the vehicle-tar- establishing fuel standards for unleaded gaso- geted policy option that requires catalytic con- line. Such options may include other command- verters on gasoline-fueled vehicles. Although the and-control measures that are necessary to catalytic converter-equipped vehicles cannot nin monitor and enforce production, distribution, with leaded gasoline (lead poisons the catalyst), 80 Chlapter 3 the opposite is not tnre. That is, vehicles de- to attain national requirements. The range for signed to rmn on leaded gasoline can also use the maximum values of RVP for these classes is unleaded gasoline. between 10.1 psi and 14.5 psi (CONCAWE 1994). Some Latin American and Caribbean coun- Volatility. Because of its toxicity and ozone and tries have limited the vapor pressure of leaded smog formation effects the vapor pressutre of and unleaded gasoline. In most instances the gasoline is limited. Gasoline with a low vapor quality of marketed gasoline meets these require- pressure-as measured by Reid vapor pressure, ments (Table 3.7). or RVP-has lower evaporative and figitive emis- The USEPA estimates that a U.S. refiner's sions than gasoline with a higher vapor pressure. long-term cost of meeting the 9.0 psi RVP limit In tests performed on European vehicles not for gasoline is $0.0038 a liter (based on a price equipped with pollution control devices, evapo- of $20 a barrel of crtde oil). Deducting the ben- rative emissions were found to nearly double efits resulting from the savings in fuel losses when the RVP of gasoline increased from 9 psi through evaporation and improved fuel to 11.9 psi (McArragher and others 1988). In economy, the net cost to the consumer is esti- tests performed on vehicles with evaporative mated to be $0.0012 a liter (Faiz, Weaver, and controls in the United States, diurnal emissions Walsh 1996). increased by more than five times (as a result of Ambient temperatures must be taken into the saturation of the charcoal canister) and hot- consideration when establishing RVP limits for soak emissions by 25 to 100 percentfor the same fuels. Because a fuel's volatility increases with increase in RVP (USEPA 1987). In tests per- temperature, lower RVP limits may be required formed as part of the U.S. Air Quality Improve- in tropical countries with little variation in sea- ment Research Program, a 1 psi reduction in sonal and diurnal temperatures. In countries RVP from 9 psi to 8 psi decreased total evapora- where differences between summer and winter tive emissions by 34 percent, CO emissions by 9 temperatures are more pronounced, separate percent, and exhaust HC emissions by 4 percent, volatility limits may be established for each sea- without affecting NO, emissions (Faiz, Weaver, son. In colder countries less stringent RVP lim- and Walsh 1996). Refueling emissions also in- its should be used to avoid problems associated crease when fuels have higher RVP. Refueling with fuel volatilization, which affect engine emissions from gasoline were about 30 percent startup and driveability. higher when the fuel's RVP increased from 9.3 psi to 11.5 psi (Braddock 1988). Benmene and other aromatic hydrocarbons. Aromatic In the United States vapor pressure limits for hydrocarbons are hydrocarbons that contain one gasoline were first implemented in 1989. These or more benzene rings in their structures. Al- limits, which established maximum RVP values though aromatic hydrocarbons help raise the for gasoline during the summer months (May octane number of gasoline, their content in gaso- through September), varied between 9.5 psi and line is often restricted because of their toxic 10.5 psi for different states. In 1992 a more strin- properties. They also contribute to higher CO gent set of limits was imposed. Accordingly, the and HC in exhaust emissions from motor ve- RVP of gasoline was restricted to 9.0 psi for the hicles (AQIRP 1990). cooler northern states during May through Sep- Some countries regulate benzene and aro- tember and to 7.8 psi for the warmer southern matic hydrocarbon content of gasoline. For ex- states duringJune through September. In 1995 ample, benzene is limited to 1.0 percent by RVP limits in the ozone nonattainment areas volume in the United States; 3.0 percent in Italy (which are served by reformulated gasoline) of and New Zealand; 3.5 percent in Thailand; and the northern and southern states were reduced 5.0 percent in Australia, all EU countries except to 8.1 psi and 7.2 psi, respectively. In 1996 the Italy, and the Republic of Korea. The aromatic maximum RVP value for reformulated gasoline hydrocarbon content is limited to 25 percent by used in the Los Angeles metropolitan area was volume in the United States, 33 percent in Italy, set at 7.0 psi. 50 percent in Thailand, and 55 percent in the The European Union has established eight Republic of Korea. In Latin America and the volatility classes for gasoline from which each Caribbean benzene in gasoline is limited in a member country must specify one for each de- few countries, and aromatic hydrocarbons only fined period of the year and for a defined region in Mexico (Table 3.8). Abatement Measuitts fotr Vehicular Air Pollution: Fuel-Targeted Measures 81 Table 3.7 Standards and quality for Reid vapor pressure of gasoline in Latin America and the Caribbean, 1996 (potnds per square inch) (asoliie slandard Alarketed gasaline quality Country Leaded Unleaded Leaded Unlea,ded Argentina - - - 8.5-12.0 Bolivia 9.0 - - 8.0-8.5 Brazil - 1.0 - 9.8-10.9a Chile 10.0-12.5" 10.0-12.5" 7.6-13.5 7.0-13.5 Colombia - 8.5 - 7.5-8.5 Costa Rica 10.0 10.0' - less than 10 Dominican Republic 9.5 - 9.5 9.5 Ecuador - - 9.2 6.5-8.4 El Salvador 10.0 10.0 8.0-9.0 Honduras 10.0 - - - Jamaica - - 10.0 10.0 Mexico 6.5-8.5' 6.5-8.5e 7.5 7.9-10.1 Nicaragua 10.0 - - Panama 10.0 - - - Paraguay 7.5-10.0 - 8.0 8.0 Peru 10.0 - 6.3-9.0 6.9-8.7 Trinidad and Tobago 9.5 - 8.1 8.1 Venezuela 9.5 - 7.7 -INot available. a. The fuel is a mixture of 22 percent by volume etha1nol with 78 percent gasoline. b. The lower limit is for summer months; the higher limit is for winter months. c. Proposed limit. d. Applicable to metropolitan areas. e. For the Mexico City metropolitan area. In 1998 the cutlrrent limit (6.5-9.5 psi) for the Monterrey and Guadalajara metropolitan areas will be reduced to 6.5-8.5 psi. Sotrce: Alconsult International Ltd. 1996; Berumen 1996; Ruiz 1996. Table 3.8 Standards and quality for benzene and aromatic hydrocarbons content of gasoline in Latin America and the Caribbean (percent by volume) (;asolinelI1 .,tdard (mne,x.) Marketed gasoline quality Armatlic Aromatir (Coun try Benzene hlydrocadons lenzene hydrocarlon.s Argentina 4.0 - 0.8-2.5 16.8-44.4 Chile 5.0 - 1.4-1.8 19.2-49.0 Colombia Unleaded regular 0.8 - 0.64-0.80 22.1-24.4 Unileaded extra 0.9 - 0.91-0.93 23.2-28.3 Mexico 2.o0 30"1 0.45-2.1 25.0-33.3 Trinidad and Tobago 5.0 - 5.0 -Not available. a. For the Mexico City, Monterrey, and Guadalajara meti-opolitan areas. For the rest of the country, the standard for benzene is 4.9 percent by volime. b. For the Mexico City, Monterrey, and Gitadalajara MetTopolitan areas. Smuare: Alconsult International Ltd. 1996; Berimeni 1996: Ecopetrol 1996; Ruiz 1996. 82 Chapter 3 A U.S. oil-automotive industry research group tent) and industry-average gasoline on catalytic evaluated the effects of the aromatic hydro- converter-equipped older and newer cars, indi- carbons content of gasoline on toxic emissions cate that reformulated gasoline emits 12 to 27 (benzene, 1,3-butadiene, formaldehyde, and ac- percent less NMHC, 21 to 28 percent less CO, etaldehyde) from catalytic converter-equipped and 7 to 16 percent less NO, than the industry- cars. Reducing the aromatic hydrocarbons con- average gasoline. However, adding 11 percent tent of gasoline from 45 to 20 percent lowered MTBE as an oxygenate increased aldehyde emis- toxic emissions by 34 to 36 percent in 1989 sions by 13 percent (Faiz, Weaver, and Walsh model-year cars and 7 to 17 percent in 1983-85 1996). Aldehydes, through photochemical reac- model-year cars. Benzene accounted for 60 to tions, may contribute to the formation of peroxi- 85 percent of toxic emissions from the newer acetyl nitrates, which are health hazards cars and 36 to 66 percent of toxic emissions from (Wijetilleke and Karunaratne 1992). the older cars (AQIRP l991b). The same reduc- In the United States reformulated gasoline is tion in the aromatic hydrocarbons content low- prepared to reduce emissions of ozone (which ered CO by 14 and HC emissions by 6 percent form volatile organic compounds, or VOCs) in the exhaust (AQIRP 1990). during the high ozone season and emissions of toxic air pollutants during the entire year. The Reformulated gasoline. Reformulated gasoline is 1990 Clean Air Act amendments required that gasoline with lower emission characteristics than effectiveJanuary 1, 1995, reformulated gasoline conventional gasoline. A reformulated gasoline be used in those parts of the United States with may include any of the following properties listed the worst ozone problems (where the popula- below (Wijetilleke and Karunaratne 1992). Most tion exceeds 250,000 people).13 Gasoline con- research in reformulating gasoline, however, has sumption in these areas-which cover six entire focused on reducing volatility, reducing sulfur, and states and portions of twelve other states that adding oxygenates (Faiz, Weaver, and Walsh 1996). include twelve major urban areas-accounts for i30 percent of all gasoline sold in the United e An octane nuimber consistent witli the com- ro States. The USEPA's final rule on reformulated pression ratios Of vehilCes . No or a minimum amount of heavy metals gasoline, announced in December 1993, man- dates that all reformulated gasoline contain a exposure mminimum 2.0 percent of oxygen by weight and exAosmallufractioneofvolatiecompo a maximum of 1.0 percent of benzene by vol- A small vracton of volaule cotiponents sior ume. The presence of heavy metals (such as lead ofdunbned gvasotine or ftigitive emissions or manganese) is not allowed. The amount of of Smallnamourntdgasoflbne and other aro- detergents is not specified, although their pres- aSmatic hroconts ofor bzeduead ohervaror- ence is required to prevent accumulation of de- matic hydrocarbons for redutced evapora- posits in engines or vehicle fuel suipply systems. tive and exhaust emissions and lower The average for Tl, sulfur, and olefins content exhaust emissions of CO and unburned HG must not exceed the 1990 average (Table 3.9). * A small fraction of olefins for lower photo- Implementation of this program consists of chemical reactivity in evaporative emissions plem e f phas which consithe and for lower gum formation and plugging period betweenJanuary 1, 1995 andJanuary 1, of fuel injectors 2000, targets a 15 to 17 percent reduction in VOC * An increased fraction of alkylates for clean and toxic emissions from motor vehicles based on the average 1990 U.S. baseline (see Table * A small amount of sulfur for higher cata- lyAstmeficenc amoundf sulo r sulfur emissios i 3.9). The second phase starts in January 2000 lyst efficiency and lower sulfi-ir emissions in and targets a 25 to 29 percent reduction in VOC the exhaust. . emissions, 20 to 22 percent reduction in toxic * Presence of oxygen for lower exhaust emis- emissions, and 5 to 7 percent in NO, emissions, sions of CO and HC Ggain and 5 to averce m U.S eine. * Additives for cleaning deposits. reducing again based on the average 1990 U.S. baseline. misfire, and improving fuel delivery to help reduce emissions. 13. These areas are those where the U.S. ambient air qual- Exhaust emission tests, performed with refor- ity standard is exceeded for ozone (also called non-attain- mulated gasoline (with 11 percent MTBE con- ment areas for ozone). Abatenzent Measures for- Vehicular Air Pollution: Fuel-Targeted Measures 83 (Table 3.10). Both initiatives were supported by Table 3.9 Baseline parameters tax incentives. for gasoline in the United States In Thailand regular leaded gasoline was the (1990 average) only gasoline grade available for light-duty mo- Additive Standatlrdi tor vehicles until 1993, when two additional BenTzene 1.6 percent by volumie grades of reformulated gasoline (called "pre- Aromatic hydrocarbons 28.6 percent by voltinie mium leaded" and "premium unleaded") were Olefins 10.8 percent by volume introduced. Each grade includes two types, Type Sulfur 338 ppm 1 for rural areas and Type 2 for urban areas 1 67"C (Table 3.11). The requirements for Type I gaso- Reid vapor pressure 8.7 psi line are the same as for Type 2 gasoline, except that Type 1 does not have any requirement for a. T., is the temperature at which 90 percent of ftel is evaporated. T,,, has been related to engine deposits anl the minimum MTBE content. engine oil diluition. In Latin America and the Caribbean reformu- S.lowne: CONCAWE 1994. lated gasoline has been produced in Mexico since 1992, when standards were established to control the content of benzene, aromatic hydro- carbons, and olefins in gasoline and the 1986 Refiners must certify their reformulated gaso- RVP specification was tightened. The specifica- line through the use of a "simple model" (appli- tions for benzene, aromatic hydrocarbons, and cable only for 1995-97) or a "complex model", olefins were tightened again in 1994. Argentina's both of which were specified by the USEPA. 1996 fuel quality standards also require refor- The U.S. reformulated gasoline program in- mulated gasoline (see the Mexico City and cludes an "antidumping" rule that also restricts Buenos Aires sections of Chapter 4). the properties of non reformulated gasoline pro The production of reformulated gasoline re- duced in the United States. Accordingly, during quires refinery modifications that are dictated 1995-97 olefins, sulfur, and T9,, are not allowed by the configuration of the refinery and proper- to exceed their values for 1990 by more than 25 ties of the crude oil processed at the refinery. percent. Subsequently, emissions of NOX, ben- These modifications may involve such unit op- zene, and toxics are limited by their respective erations as reforming, alkylation, isomerization, values for 1990, and VOC emissions are to be and oxygenates production or blending. Each controlled by regional RVP limits (CONCAWE modification has a different effect on the com- 1994). position and properties of gasoline and, there- Examples of other countries which produce fore, on the concentration of pollutants in the reformulated gasoline include Finland and Thai- exhaust emissions from motor vehicles. The in- land. Finland's reformulated gasoline program cremental cost of producing reformulated gaso- was implemented in two stages. An interim re- line depends on the extent of refinery formulated gasoline (called "city gasoline") and modifications; the cost to U.S. refiners ranges a more severely reformulated gasoline, with re- from $0.008 to $0.013 a liter to meet the first duced levels of benzene and sulfur, introduced phase of the federal program, about $0.03 a li- in January 1993 and March 1994, respectively ter to meet the second phase of the federal pro- Table 3.10 Properties of severely reformulated and city gasoline in Finland Severety Additives Gity gasoline reTnduletled gsolihne Oxygen (percent by weight) 2.0-2.7 2.0-2.7 Betizene (maximuml) percent by volhunie) 3.0 1.0 Sulfur (mnaximum ppm by weight) 400 100 RVP (summer/winter) (maximum psi) 10.1/13.0 10.1/13.0 Sourre: CONCAWE 1994. 84 Chapten 3 Table 3.11 Compositional constraints for regular and Type 2 gasoline grades in Thailand 7Ype 2 gasoline (0-eformuthtded) legidar P'emium Premium Adlditive gasoline leaded unleaded Lead (maximum gram/liter) 0 O. 15a 0.013 Sulfur (maximum percent by weight) 0.15 0.15 0.10 Benzene (maximum percent by volume) 3.5 3.5 3.5 Aromatic hydrocarbons (maximum percenit by voltime) - 50 50b MTBE (percent by volume) Minimum - 5.5 5.5 Maximum 10.0 10.0 10.0 - No limit is established. a. This limit was rediuced to 0.013 gram/liter effectiveJaniuary 1995. b. This limit wvill be reduced to 35 percent by volume effective January 2000. Source: CONCAWE 1994. gram, and about $0.05 a liter to meet the sec- verter-equipped vehicles than for older uncon- ond phase of the California program4 (USEPA trolled vehicles because catalytic converters al- 1994; Faiz, Weaver, and Walsh 1996). To comply ready remove a large fraction of pollutants from with Mexican gasoline quality requirements combustion gases (a large fraction of CO and HC Pemex invested $344 million in fourteen projects by two-way catalytic converters and CO, HC, and at different refineries (see the Mexico City sec- NOC by three-way catalytic converters). Thus a tion in Chapter 4). In addition, Pemex has un- reformulated gasoline program is more suitable dertaken ten new projects requiring a total for urban centers or countries where most in- investment of about $1 billion between 1996 and use vehicles lack catalytic converter technology. 1998 (Table 3.12). To meet the Argentine fuel Design and implementation of a gasoline re- quality requirements YPF has invested $170 mil- formulation program must consider a set of com- lion in refinery modifications that included alky- mand-and-control measures and, possibly, some lation, isomerization, and oxygenation (MTBE market-based incentives. In the United States the and TAME) units (Tanco 1996). Esso is planning measures and incentives included in this pro- to increase its reformer capacity, and Eg3 is plan- gram include prohibition of the sale of conven- ning to add a high severity reformer and MTBE tional gasoline by refiners, distributors, or and isomerization units (Alconsult International retailers for use in ozone nonattainment areas; Ltd. 1996). establishment of certification procedures for Before launching a reformulated gasoline pro- reformulated gasoline; specification of sampling, gram, an evaluation should be made of refor- testing, and record-keeping requirements to mulated gasoline production capabilities that prevent violation of regulations; establishment considers crude properties, existing refinery of a credit system (see section on market-based configurations, and oxygenate supplv options. incentives, below); and penalties for violations. In addition, a careful analysis should be made In establishing the requirements for reformu- of the impacts on vehicle emissions and ambi- lated gasoline in the United States the energy ent air quality. This analysis should ensulre that requirements, health and environmental im- the reformulated gasoline does not create other pacts, and costs of achieving emission reductions emissions or ambient concentrations of pollut- were taken into consideration. ants which pose serious health threats. Given these considerations, the benefits of reformu- Oxygenated gasoline. Oxygenates are added to lated gasoline are more modestfor catalytic con- gasoline to increase its oxygen content and, therefore, enhance cleaner combustion in mo- tor vehicles. Combustion of oxygenated gasoline 14. The seconid phase of the California gasorline program produces lower CO and HG exhaust emissions is expected to restllt in about 30 percent redutction in HC than straight gasoline. NO, emissions, however, and toxic air emissionscompared with the fiels sold in 19o. may increase. Oxygenates also have higher Abatement Measutres for Vehicutlar Air Pollution: Fuel-Targeted Measures 85 Table 3.12 Mexico's refinery investment program, 1996-98 (;7aparily Investment Eshtmated Refinet/pm1pert (barreh n day) (million $) roDmIAetion date Saiina Cruz Alkylatiorn 14,100 57 1996 Isomerizationi 13,400 30 1996 Tula Hydrodesulfurizatioin of intermediate distillates 25,000 72 1996 Alkylation 7,500 32 1996 Isomerization 15,000 26 1996 H-Oil 50,000 623 1997 Salamanca Hydrodesulfurization of intermediate distillates 25,000 73 1996 Alkylatiori 3,400 26 1996 Isomerization 12,000 25 1996 Madero Isomerization 12,000 47 1998 Source: Berumen 1996. blending octane numbers than most other gaso- Oxygenated fuel tests performed on 1989 line components except for aromatic hydrocar- model-year gasoline-fueled U.S. vehicles with 10 bons. Oxygenates, however, can potentially percent (byvolume) ethanol blends, 15 percent reduce fuel economybecause of their lowervoli- (by volume) MTBE blends, and 17 percent (by metric energy content than conventional gaso- volume) ETBE blends"8 had similar effects on line. pollutant exhaust emissions: a reduction of 5 Oxygenates are either alcohol-based or ether- percent in HC, 4 percent in NMHC, and 9 per- based. Alcohol-based oxygenates include metha- cent in CO emissions. A 5 percent increase in nol5 and ethanol;"' ether-based oxygenates NO, emissions was also noted with oxygenate include methyl tertiary butyl ether (MTBE), blending (AQIRP 1991a). In Mexico City the ethyl tertiary butyl ether (ETBE), and tertiary addition of 5 percent MTBE to leaded gasoline amyl methyl ether (TAME). The most commonly reduced CO emissions by 15 percent and HC used oxygenates are MTBE and ethanol. Of the emissions by 12 percent from non-catalyst gaso- two groups, ether-based oxygenates are pre- line-fueled vehicles. This addition did not in- ferred for processing and environmental reasons crease NO, emissions (CMPCCA 1995). (for example, alcohol-based oxygenates have Gasoline quality standards in some countries higher evaporative emissions than ether-based establish limits for oxygen or oxygenate content oxygenates because of their higher blending (Table 3.13). In Latin America and the Carib- RVP).17 bean these limits have been set in Argentina, Brazil, and Mexico, and are proposed in Costa Rica. However, despite the absence of such regu- lations, increasing amounts of oxygenates have been used in the 1990s because countries in the 15. Methanol is obtained in a refiniery from the synthesis region are rapidly eliminating or reducing lead gas 1)rodulced by steam reforming.s gas prduced y stcan refoiiing.in gasoline. The predominant types of oxygen- 16. Ethanol is produced from fennenitation of renewable resources such as sugar cane or corn. ates used are ethanol in Brazil, MTBE and TAME 17. Ether-based oxygenates are produced from reactions between alcohols and olefins: MTBE from a reaction be- tween methanol and isobutylene; ETBE from a reaction between ethanol and isobutvlene; and TAME from a reac- tion between methanol and isoamylenie. Because isobuty- 18. The oxygen content in the gasoline-oxygenate blends lene is more readily available than isoamylene in a refinery, was 3.7 percent (by weight) for ethanol, 2.7 percent for productioni of MTBE or ETBE is preferred over TAME. MTBE, and 2.7 percent for ETBE. 86 Chapter 3 Table 3.13 Oxygenate requirements for motor vehicle fuels (COm71hr/o3ggenate tqpe' Sitanderd Latin American cotunitries Argentina MTBE (maximumil percent by volume) 15.0 Ethanol (maximrimn percent by volume) 5.0 Isopropyl alcohol (maximum percent by volume) 5.0 Tertiary butyl alcolol (maximum percent by volume) 7.0 Isol)utyl alcohol (maximlum percerit by volume) 7.0 Oxygen (maxinimiim percent by weight) 2.7 Brazil Ethaniol (niniiiuimn percent by volume) 22.0 Costa Rica Oxygen (minimiumin percent by weight) 2.02 Mexico Oxygen (minininiiii percent by weight) 1*0h (maximium percent by weight) 2.0b Others Republic of Korea Oxygen (minimiumi percent by weight) 0.5 South Africa Oxygen for 95 RON (maximum percent by weight) 2.8' Oxygen for 91 RON (maximum percent by weight) 3.7' Thailand MTBE (milimulil percent by volume) 5.5 MTBE (maximumii percent by volume) 10.0 Utnited Statesf Oxygeni (rnaxinmiunT percent by weight) 2.7 a. Proposed standard. b. Applies to metropolitan areas (Mexico Citv. Guadalajara, Monterrey). No standards have been established for the rest of the coun1try. c. Corresponds to 7.5 percent alcohol by voluine (85 percent ethanol and 15 percent propanol). d. Corresponds to 9.5 percent alcohol by voliutie (85 percent ethanol and 15 percent propanol). e. For refornulated gasoline requiired in urban areas. f. "Gasohol" consistinig of gasoline with 10 percent alcohol by volume is permitted. This fuiel contains 3.5 percent oxygen by weight. Soturre: CONCAWE 1994; DDF 1996; Ministiv of the Environmenit and Energy of the Repuiblic of Costa Rica 1996; Resolition 54/96 of the Argentine Ministry of Economy. in Mexico, and MTBE in some other countries of 2.7 percent for a minimum of four months in of the region. MTBE production units have been areas where ambient CO concentrations are 9.5 built in Argentina, Mexico, and Veneztiela. Many ppm or higher based on two-year CO data. The cotntries of the region-including C hile, Costa act also requires labeling of oxygenated gasoline, Rica, Guatemala,Jamaica, Peru, and Uruguay- established a credit system (discussed under the increase the octane number of gasoline by im- section on market-based incentives, below), and porting MTBE. introduced enforcement measures and penal- In the United States the oxygenated gasoline ties if the requirements are not met. If the am- program was established by the 1990 Clean Air bient air quality for CO cannot be reached in a Act amendments to reduce vehicular emissions nonattainment area, then the 2.7 percent oxy- in areas with the worst CO problems (that is, gen requirement is raised to 3.1 percent. areas where the national ambient air quality stan- U.S. refiners and marketers prefer ether- dard for CO is exceeded). The act requires the blended gasoline over alcohol-blended gasoline use of gasoline with a minimum oxygen content because of its lower vapor pressure and octane- Abatement Aleastres for Vehicular Air Pollution: Fuel-Targeted Measures 87 boosting ability. In addition, ether-based gaso- asphaltic and carbonaceous content (Wijetilleke line is not prone to phase separation or to water and Karunaratne 1992). extraction. Ether-based oxygenated gasoline In Canada and the United States the sulfur costs between $0.01 and $0.03 a liter more than content of diesel fuel is limited to 0.05 percent straight gasoline (Faiz, Weaveir, and Walsh 1996). by weight. In 1993 the European Union estab- Use of oxygenated gasoline may be beneficial lished a sulfur limit for diesel fuel of 0.2 percent for traffic-congested Latin American and Carib- by weight effective October 1, 1994 and 0.05 bean urban centers with high ambient CO percent by weight effective October 1, 1996. In concentrations (such as Buenos Aires) and for Japan the sulfur content of diesel fuel was lim- urban centers that are located at high altitu(les ited 0.2 percent in 1992 and 0.05 percent in May and have significant fraction of the motor ve- 1997. Thailand also has adopted a phased re- hicle fleet equipped with carburetors or continu- duction of sulfur in diesel fuel used in urban ous fuel injection systems (such as Quito and La areas, from 0.25 percent inJanuary 1996 to 0.05 Paz). Development of an oxygenated gasoline percent by January 2000. In the Republic of program requires a careful evaluation of the Korea the sulfur content of diesel fuel will be impacts on vehicle emissions and ambient air reduced to 0.05 percent by 1998 (CONCAWE quality. This analysis needs to include NO. and 1994). reactive HC emissions, especially in urban cen- Some Latin American countries regulate the ters where ozone is a health problem. sulfur content of diesel fuel. In Argentina the sulfur content of diesel fuel is limited to 0.25 percent by weight. In Brazil it was limited to 0.5 Diesel Fuel Standards percent by weight for urban areas and 1.0 per- cent for rural areas, but in October 1996 the 0.5 Standards for diesel fuel have been established percent limit was reduced to 0.3 percent and in many industrial and some developing coun- the 1.0 percent limitwas reduced to 0.5 percent. tries. The main diesel fuel parameters affecting In Chile the limit is 0.3 percent for the Metro- pollutant emissions are sulfur content, cetane politan Region (which includes Santiago) and number, aromatic hydrocarbons, and densitv. 0.5 percent for the rest of the country. In May 1998 these limits will be reduced to 0.2 and 0.3 Sulfir. The presence of sulfur compounds in die- percent, respectively. In Colombia the current sel fuel results in SO, and PM emissions from limit is 0.4 percent. This will be reduced to 0.1 the exhaust of diesel-ftieled vehicles. Metal sul- percent in 1998 and 0.05 percent in 2002. In fates and sulfuric acid in the form of PM consti- Mexico the limit for the low-sulfur diesel fuel tute 1 to 3 percent sulfur emissions from (called Diesel Sin) used in urban areas is the heavy-duty diesel-fueled vehicles and 3 to 5 per- same as in the United States (0.05 percent). In cent of sulfur emissions from light-duty diesel- 1995 thesulfurcontentofDiesel Sinwasbetween fueled vehicles. They also account for about 10 0.03 and 0.05 percent (DDF 1996). The limit percent of PM emissions from these vehicles for high-sulfur diesel fuel used for intercity trans- (Faiz, Weaver, and Walsh 1996). These emissions port was 0.5 percent. However, the high-sulfur have adverse health and environmental impacts diesel fuel was eliminated from the market in (see Chapter 2). February 1997. In catalyst-equipped diesel-fueled vehicles, the Along with sulfur content standards, tax in- oxidation catalyst converts SO, from the engine centives are used to lower the sulfur content of into SO. But the presence of SO2 in the engine diesel fuel sold in some European countries. For exhaust reduces the catalyst's efficiency at oxi- example, in 1992 Denmark introduced a tax in- dizing CO and HC. centive for lower sulfur grade diesel fuel (with a The quality of diesel fuel produced in devel- maximum of 0.05 percent) and in 1993 for "no oping countries is generally lower than in indus- sulfur" grade diesel fuel (with a maximum of trial countries. Because of the higher demand 0.005 percent). In Sweden low sulfur and low for diesel fuel in developing countries refiners aromatic grades of diesel fuel receive tax breaks. have expanded the distillate cut from the atmo- Since 1993 the use of sulfur-free diesel fuel in spheric distillation unit to include the heavier Finland has been encouraged by a tax incentive fraction. As a result, diesel fuel in developing (CONCAWE 1994). countries generally has a higher sulfur and more The sulfur content of diesel futel can be re- 88 Chapter 3 duced by hydrodesulfurization."' Low-pressure weather conditions (Faiz, Weaver, and Walsh hydrodesulfurization units are capable of remov- 1996). ing 65 to 75 percent of the sulfur, lowering aro- The cetane number of diesel fuel is affected matic levels by 5 to 10 percent, and increasing by crude fuel properties and refinery operations. the cetane number by 1 to 2 points. However, Diesel fuels with straight-chain hydrocarbons newer high-pressure hydrodesulfurization units have higher cetane numbers than those with can remove more than 95 percent of sulfur and branched hydrocarbons and those with a higher lower aromatic levels by 20 to 30 percent (Faiz, aromatic content. The cetane number is con- Weaver, and Walsh 1996). In Europe the cost of trolled by maintaining a proper boiling range reducing sulfur in diesel fuel from 1.0 percent of the cut from the atmospheric distillation unit. to 0.05 percent ranges from $0.009 to $0.014 a Inclusion of the lighter naphtha fraction into liter (CONCAWE 1989). this cut reduces the cetane number. Some refin- eries, such as those in the United States, cata- Cetane number. The cetane number is a measure lytically crack heavy oil to produce gasoline and of the spontaneous ignition quality of diesel fuel blend the cracked stream containing high per- at the same temperature and pressure as in the centages of aromatic hydrocarbons into the die- combustion chamber of the engine.2" The cet- sel cut. However, this practice increases the ane number of diesel varies from 43 to 57 (with aromatic HC content and reduces the cetane an average of 50) in Europe, but is lower in the number. To boost the cetane number, these re- United States (40 to 44). The cetane index, oc- fineries may need to switch to hydrocracking or casionally used instead of cetane numbers, is an segregate the highly aromatic distillate streams approximation of the cetane number based on for use in other products such as fuel oil. The an empirical relationship with gravity and vola- addition of certain barium or calcium derivatives tility parameters. in diesel fuel suppresses smoke emissions but The higher the cetane number, the shorter increases PM-sulfate emissions. Copper-based the delay between injection and ignition and the diesel fuel additives may increase dioxin better the ignition quality. A higher cetane num- emissions. ber is desirable because it improves combustion, The cetane number of diesel fuel is regulated reduces HC, CO, and PM emissions especially in some Latin American countries. For example, during vehicle warm-up, improves cold starting, the required minimum cetane number is 40 in reduces white smoke after starting, and lowers Brazil, 45 in Chile and Colombia, 48 (for low- noise levels (OECD/IEA 1993). At lower cetane sulfur diesel fuel) and 45 (for high-sulfur diesel numbers, especially at cetane numbers below 45, fuel) in Mexico, and 50 in Argentina. These lim- black smoke and HC and CO emissions increase. its compare with minimum cetane number re- It is estimated that a two-point reductioni in the quirements of 40 in the United States, 47 in cetane number from a European median of 50 Finland (for reformulated diesel), and 47 and could increase NO, emissions by 2 percent and 50 for the two grades of urban diesel fuel in Swe- PM emissions by up to 6 percent. The emission den (CONCAWE 1994). In 1995 the typical ob- benefits, however, may not be significant beyond served value for the cetane number of diesel fuel the current European levels. Diesel fuiels with was 49 in Chile and 55.4 (for low-sulfur diesel high cetane numbers, which tend to be paraf- fuel) in Mexico (Ruiz 1995; DDF 1996). finic, could also cause driving problems in cold Aromatic hydrocarbons and density. The aromatic HC content of diesel fuels depends on crude properties and refinery operations. Blending of 19. Hydrodesullfisrization involves catalytic reaction of sul- the diesel cut from the catalytic conversion and fur-containinig oil molecules (the diesel fraction from the tmospheric distillation tanitstincreasessthe aro atmospheric distillation unit and light oils fiom catalytic atmospheric disullation units increases the aro- cracking and coking units) with hydrogen. This process re- matic HC content of diesel ftuels. moves sulfur as hydrogen sulfide. The aromatic HC content, cetane number, 20. The cetane ritimber of a firel is determiiiedi by corn- and density of diesel fuels are closely related. paring its ignition property with that of a fuiel imiixtire con- Diesel fuels with a higher aromatic HC content sisting of n-cetane (a straight-chain paraffitiic lsvli-ocarbon whose assigned cetane number is 100) anid heptamethyl tend to be denser and have lower cetane ntm- nonane (a branched paraffiniic hyvdrocarhont whlose cetane bers. A high aromatic HC content is associated nuimber is 15). with difficulty in cold engine starting, and in- Abatemrent Menses for Vehlticutlar Air Pollution: Fuel-Targeted Measures 89 Table 3.14 Standards for aromatic HC, PAH, and density of diesel fuel in Finland, Sweden, and the United States Aromatic H(; PAH (maximum (maximmn l)ensify Country % 15y voltume) % by volume) (Ng/in') Finlanid-ReformtilatedI diesel fuel 20 Sweden-Urban diesel 1 5 0.02 800-820 Urban diesel 2 20 0.1 800-820 United States (California) 10 1.4 824-860 Nole: A blank space indicates that nco standard wvas established. Aromatic HC is aromatic hydrocarbolns; PAH is polycyclic aromatic hydrocarbons. !Sonwte: CONCAWE 1994. creased HC, NO., and noise emissions. Exhaust ral gas (CNG, composed mainly of methane), emissions from combustion of highly aromatic liquefied petroleum gas (LPG, composed of pro- diesel fuels contain carbonaceous PM that in- pane and butane), methanol (made from natu- clude soot, soluble organic fraction, and PAH. ral gas, coal, or biomass), ethanol (made from These emissions can be reduced with diesel fuel grain or sugarcane), vegetable oils, electricity, additives. Higher density diesel fuels result in hydrogen, synthetic liquid fuels (derived from black smoke and PM emissions (Faiz, Weaver, hydrogenation of coal), and various fuel blends Walsh 1996). such as gasohol. A number of alternative fuels In Europe the aromatic HC content of diesel have been used commercially in an effort to cur- fuel has been regulated in Finland and Sweden tail air pollution in urban centers (Tables 3.15 (Table 3.14). Sweden has established aromatic and 3.16). Among these, CNG and LPG reduce HC and PAH limits for two diesel fuel grades, pollutant emissions and provide some economic which are supported by tax incentives. The regu- benefits for certain applications and in specific lation also specifies the density requirement for locations. Alcohol fuels (ethanol and methanol), these fuels. Diesel fuel specifications in Califor- because of their lower ratio of carbon to hydro- nia limit aromatic HC and PAH content, fuel gen (4:1 for ethanol, 3:1 for methanol, and 6:1 density, and other parameters such as sulfur (see for diesel fuel and gasoline), result in lower CO2 Table 3.14). A diesel fuel with different param- emissions than gasoline and diesel fuel but are eters is allowable as long as its emissions are not more expensive. Hydrogen and methanol burn greater than those of this reference fuel efficiently and are intrinsically low-polluting fu- (CONCAWE 1994). els but are uneconomic at the present time. The aromatic HC content of diesel fuel is lim- There are no greenhouse effects associated with ited to 20 percent by volume in Colombia. In the use of hydrogen because its combustion does Mexico the limit is 30 percent by volume for low- not produce CO2. However, hydrogen is difficult sulftir diesel fuel. In 1995 the low-sulfur Mexi- to store in motor vehicles and its production can diesel fuel contained 26.15 percent aromatic from fossil fuels would probably generate high HC (DDF 1996). Density of diesel fuel is limited pollutant emissions (ECMT 1990). In addition, between 820 kg/mi and 880 kg/m3 in Brazil and on a unit energy basis hydrogen is about twice between 830 kg/m3 and 870 kg/m3 in Chile. as expensive as gasoline at the wholesale level and more than thirty times as expensive at the retail level. Use of electricity as a fuel substitute Alternative Fuels for road-based vehicles can also reduce pollut- ant emissions considerably, although this reduc- Since the 1970s rising oil prices and greater con- tion is often very costly. In addition, electric cars cerns about the health and environmental ef- have considerably shorter driving ranges than fects of air pollution have prompted the cars fueled by conventional fuels. development of alternative fuels for the trans- The potential environmental benefits ofalter- port sector. These fuels include compressed natu- native fuels must be considered with caution. In 90 Chapter 3 Table 3.15 Properties of some conventional and alternative fuels Pf,eriy (,Gasuline Diesel Methane Prop)ane' Methanol E,:thanol Energy contenit (LHV) (MJ/kg) 44.0 42.5 50.0 46.4 20.0 26.9 Liquid density (kg/liter) 0.72-0.78 0.84-0.88 0.4225 0.51 0.792 0.785 Liquid energy density (MJ/liter) 33.0 36.55 21.13 23.66 15.84 21.12 Gas energy density (MJ/liter) At atmospheric pressure n.a. n.a. 0.036 0.093 n.a. n.a. At 2,900 psi pressure ni.a. n.a. 7.47 n.a. n.a. n.a. Boiling point ('C) 37-205 140-360 -161.6 -42.15 65 79 Research octane number 92-98 25 120 112 106 107 Motor octane numnber 80-90 n.a. 120 97 92 89 Cetane number 0-5 45-55 0 2 5 5 n.a. Not applicable. Note: LHV is loNver heating valie. a. Most of the LPG sold in the U.S. is propane. Sourwe: Weaver 1995. many instances the same or even better emis- bulk shipment by sea or train, it is liquefied and sion reductions can be achieved with conven- maintained at cryogenic temperatures (that is, tional fuels in vehicles equipped with advanced as LNG). It is also transported in a compressed emissions control systems. form in cylinders for short distances. Natural gas is stored as a liquid or compressed gas. However, Compressed natural gas. After coal, natural gas is for use as a transport fuel it is stored in vehicles the world's most abundant fossil fuel with proven in heavy cylinders as a gas pressurized to about reserves estimated to be twice those of petro- 3,000 psi and at ambient temperatures (that is, leum. In Latin America and the Caribbean, the as CNG). largest proven natural gas reserves and produc- Almost all the CNG-fueled light-duty vehicles tion are located in Mexico and Venezuela. Other in the world have been retrofits from gasoline- countries with high natural gas production are fueled vehicles. Such retrofits allow the vehicle Argentina, Trinidad and Tobago, Brazil, Colom- to be fueled with either CNG or gasoline. Ve- bia, and Bolivia. hicle conversion for CNG fueling includes in- Natural gas contains predominantly methane stallation of at least one high-pressure storage (95 to 99 percent) with the balance made up by cylinder, a gas fueling outlet, piping from the other gases (such as ethane and propane). It is cylinder to the engine, a CNG supply gauge and used mainly in the industrial and domestic sec- a switch on the dashboard indicating the fuel tors, as well as in the transport sector. Large quan- entering the engine, a pressure regulator to re- tities of natural gas are transmitted and duce the tank's gas pressure to atmospheric pres- distributed by land through gas pipelines. For sure, and an emergency fuel shut-off system. In Table 3.16 Prices of some conventional and alternative fuels in the United States (foirth quarter 1992 dollars) (;asot,in Afethanol Ethainol LP, CN(; W'holesale ($/liter) 0.13-0.18 0.08-0.11 0.34-0.38 0.07-0.12 0.07-0.13 W\tholesale ($/therm) 0.41-0.54 0.56-0.74 1.70-1.91 0.29-0.53 0.26-0.52 Retail ($/liter) 0.26-0.35 0.21-0.24 - 0.25-0.29 0.11-0.24 Retail ($/therm) 0.78-1.06 1.41-1.62 - 1.12-1.29 0.41-0.93 - Not available. Note.: 1 therm = 0.009478 MJ. Source: Faiz, Weaver, and Walsh 1996. Abatement Measure's for Vehicutlar Air Pollution: Fuel-Targeted Measures 91 Table 3.17 Exhaust emissions from CNG- and diesel-fueled buses (grams per kilowatt houir) Bus p)C -H;C NO, PM CNG-fueled city bus 0.4 2.1 4.3 < 0.05 CNG-fueled regional btis 2.5 3.1 2.9 < 0.05 Diesel-ftueled bus 4(1 1.2 14.0 0.55 1996 European Union Standard 4.0 1.1 7.0 0.15 Note: The tests were performed in the Netherlands on tmo biuses equipped wvith DAF GKL1 160 enigines. These engines were converted to lean-buirn naturally-aspirated enigines writh anl oxidation catalyst for CNG testing. .Swurce. Prensa Vehiculor 70. October 15, 1994. addition, modifications to the carburetor or the designed to burn CNG. Compared with diesel- fuel injection system are needed. fueled buses, CNG-fueled buses have lower NO., In the United States tests performed on dual- CO, and PM emissions but higher HC emissions, fuel light-duty vehicles with model-years rang- which mostly consist of methane (Table 3.17). ing from 1983 to 1987 (with sophisticated Combustion of CNG does not emit SO2. controls) show that as much as a 97 percent re- A major concern with light-duty dual-fuel ve- duction in CO emissions can be achieved with hicles is the size and weight of the CNG storage CNG over gasoline if the vehicle is properly cali- cylinder. Depending on its material construction, brated with leaner air-fuel ratios. These tests also a full CNG cylinder weighs more than twice as confirmed the results of previous tests conducted much as and up to five times as much as a full on pre-1981 model-year dual-fuel light-duty ve- gasoline tank to provide the same amount of hicles. In addition, tests on the newer vehicles energy. Space and weight considerations in light- indicate that CO emissions can be as low as 0.17 duty dual-fuel vehicles restrict the maximum gram/km, NO. emissions can be significantly number of CNG cylinders in the trunk to two or higher (as much as 195 percent) if design three. This limits the travel range of vehicles tradeoffs are not implemented, and NMHC before refueling. A typical fueling by "fast fill" emissions can be higher or lower (Alson, Adler, systems takes five to ten minutes. In addition, and Baines 1989). Exhaust emissions from com- CNG is less readily available than gasoline at re- bustion of CNG do not contain benzene, 1,3 tail fuel stations in urban centers and may not butadiene, formaldehyde, or lead. There are no be available at all outside urban areas. For these evaporative or running loss emissions with CNG reasons gasoline- or diesel-fueled cars generally because the fuel system is sealed. are not converted to CNG by private vehicle own- Very few light-duty vehicles are of original ers. However, in countries where CNG is cheaper equipment manufacture (OEM; that is, designed than gasoline and the trunk space is not very to use CNG). In the United States OEM CNG- important, owners of light-duty commercial ve- fueled vehicles are mostly operated by gas com- hicles prefer the CNG conversion. This is the panies. Optimization of the engine design for case for taxi drivers in Buenos Aires, where ve- these vehicles (to reduce pollutant emissions) hicle conversion costs are recovered in about two includes raising compression ratios to take ad- years through fuel savings. A further concern vantage of their higher octane rating, advanced about the additional weight of CNG is associated timing, and leaner operation. Tests performed with the reduction in acceleration rates, need in the United States on pick-up trucks designied for longer braking distances, and decreases in to burn CNG indicated emission reductions of the fuel efficiency of vehicles. up to 99 percent for CO and about 30 percent Manyof the concerns that existwith light-duty for NMHC relative to comparable gasoline-fieled CNG-fueled vehicles are not relevant for CNG- vehicles. However, the CNG-fueled vehicles fueled buses. Since CNG-fueled urban buses have yielded higher NO, emissions (Alson, Adles; and space for CNG cylinders in the chassis area, they Baines 1989). are not affected by the travel range restriction. CNG has also been used as a transport fuel in However, CNG cylinders and their associated retrofitted diesel-fueled buses or buses originally tubing add additional weight on buses. In addi- 92 Chapter 3 tion, CNG-fueled buses can be refueled through pilot program. It included the supply of an ad- .slow fill" systems overnight at a central station. ditional 2.5 million cubic meters a day of natu- CNG's potential as a transport fuel is strongly ral gas (enough to fuel 45,000 vehicles) by the affected by its price in the local market. Although national petroleum company Pemex, constnrc- natural gas is much cheaper to produce than tion of a new service station for CNG fueling gasoline, it is much more expensive to distrib- and rehabilitation of an existing station, and ute. In addition, it requires costly compression conversion of some private and commercial ve- to about 3,000 psi. A CNG refueling station that hicles, police cars, and microbuses. In addition, serves 100 cars a day can cost as much as a LNG plant was constructed to supply fuel for $300,000, $150,000 of which is for compression buses and trucks (CMPCCA 1995). About 6,000 equipment and fuel pumps. Gas compression light-duty vehicles in Venezuela and 3,000 in and refueling costs add between $0.016 and Trinidad and Tobago use CNG. $0.024 to the price of natural gas energy equiva- CNG's market potential as a motor vehicle fuel lent of a liter of gasoline (CCEEB 1990). Costs could be increased if vehicle manufacturers com- incurred by vehicle owners for CNG retrofitting mit to CNG-fueled vehicle development pro- range from $1,000 to $1,500 for light-duty ve- grams that lead to technological advances in hicles, $2,000 to $3,000 for light-duty trucks and engine, storage cylinder, and vehicle designs. vans, and $5,000 to $7,000 for buses. A new bus designed to use CNG costs 15 to 20 percent more Liquefiedpetroleumgas. Liquefied petroleum gas than a similar diesel-fueled bus. (LPG) is a by-product of natural gas treatment CNG has been used in many countries. About at cryogenic temperatures, whereby liquefied 300,000 vehicles in Italy, 200,000 vehicles in Rus- components are purified and then maintained sia, 55,000 vehicles in the United States, 45,000 as a pressurized liquid for ease of handling. LPG vehicles in New Zealand, and 40,000 vehicles in is also a by-product of crude oil refining. It con- Canada run on CNG. In the United States many sists mainly of propane and butane, with some utility companies operate vehicles powered ex- minor quantities of propylene and other hydro- clusively by CNG. carbons. Propane and butane can be used sepa- Among Latin American and Caribbean coun- rately or in mixtures. Since propane has a tries, Argentina uses CNG most extensively as a superior knock resistance (with an octane num- motor vehicle fuel. In addition, CNG-fueled ve- ber of 112, which is superior to gasoline but lower hicles circulate in Brazil, Colombia, Mexico, than natural gas), it is preferred over butane for Peru, Trinidad and Tobago, and Venezuela. use as a transportation fuel. Since the initiation of Argentina's 1985 tax ex- LPG is produced in countries with natural gas emption program for natural gas, about 400,000 and petroleum exploitation or refining capabili- vehicles have been retrofitted to burn CNG. ties. LPG is distributed overland by pipelines, rail- About 270,000 of these vehicles are circulating road cars, and trucks. Where LPG is transported in the Buenos Aires Metropolitan Area. Over 80 by sea in tankers and barges, it is received in percent of about 46,000 taxis, many light-duty bulk storage and distribution terminals. Uses of trucks and chauffeured short-term cars, and LPG include residential and commercial heat- some private cars and buses circulating in Buenos ing, chemical manufacturing, agricultural prod- Aires are fueled with CNG. Elsewher e in the re- uct drying, industrial processing, and transport. gion, General Motors recently introduced CNG- About 3.9 million vehicles operate on LPG ethanol dual-fuel capability in two of its new worldwide. LPG is very popular in Asian coun- vehicle models in Brazil. In Sao Paulo about tries, includingJapan, Thailand, and the Repub- 3,000 taxis run on the CNG-ethanol dual-fuel lic of Korea. In Japan alone some 1.5 million system, and 80 urban buses are fueled witlh CNG. cars are fueled with LPG. In addition, many In Rio de Janeiro about 2,000 taxis have been three-wheelers in Asian countries have been con- retrofitted to use CNG since 1990, and 200 buses verted to LPG. The use of LPG is also growing are CNG-fueled. CNG has not been used as a quickly in Europe (for example, in the Nether- transportation fuel in Santiago because it is not lands and Austria). The United States and locally available. However, use of CNG in buses Canada are estimated to have 330,000 and will be considered when natural gas is made avail- 140,000 LPG-fueled vehicles, respectively. About able from Argentina through a pipeline. Mexico 90 percent of these fleets are commercially City's CNG program was initiated in 1992 as a operated. Abatement Measiirrs for VehicularAirPollution: Fuel-Targeted Measures 93 In Latin America and the Caribbean LPG is ing LPG vehicles requires additional fittings to used extensively as a motor vehicle fuel in Ven- attach the dispensing nozzle to the vehicle. To ezuela (68,200 barrels per day or bbd) and avoid overpressuring, an overfill prevention de- Suriname (36,011 bbd). Other countries of the vice is also required to ensure that the LPG cyl- region using LPG include ]Bolivia (7,448 bhdb), inder is not filled more than 80 percent (CCEEB Costa Rica (1,250 bbd), Dominican Republic 1990). (6,850 bbd), Mexico (3,000 bbd), and Paraguay LPG-fueled vehicles have lower CO, CO., and (957 bbd). In addition, LPG is uised to some ex- PM emissions than gasoline-fueled vehicles and tent in Colombia. El Salvador (2 bbd), Peru. and do not emit lead or SO2. CO emissions are 25 to St. Kitts-Nevis (12 bbd; Alconsult International 80 percent lower. Evaporative emissions from Ltd. 1996; Berumen 1996).2 1 To date about LPG-fueled vehicles are negligible because of the 28,300 passenger and freight vehicles and vans tight seals for the LPG fuel system. NO, emis- in Mexico City have been converted to LPG. Of sions from LPG-fueled vehicles not equipped these, more than 95 percent belong to commer- with catalytic converters are higher than those cial companies. Pemex supports this program from similar gasoline-fueled vehicles because of through production of LPG. Ten LPG fueling the higher flame temperature. However, LPG- stations have been constructed to fuel about fueledvehicles with catalytic converters produce 8,000 vehicles on a daily basis. In addition, most NO, emissions similar to gasoline-fueled vehicles commercial companies that have converted their and less than diesel-fueled vehicles. HC emis- fleets to LPG have their own storage and fuiel- sions from the exhaust of LPG-fueled vehicles, ing facilities (Sanchez 1996). In Santaf6 de although exceeding those from gasoline-fueled Bogota a fuel switching program for buses and vehicles in certain cases, are less reactive in form- taxis has been initiated by the Santafe de Bogota ing ozone (CCEEB 1990). District Environmental Agency in coordination LPG can deliver up to 85 percent of gasoline's with the national oil company Ecopetrol. kilometers per liter. The engine power and ac- Throughout the world, most LPG-fueled ve- celeration performance of LPG-ftieled vehicles hidces are conversions from gasoline-fueled cars, are similar to gasoline-fueled vehicles. LPG, be- although there are some OEM LPG-fueled ve- cause of its ability to vaporize at low tempera- hicles. For example, Nissan and Toyota inJapan tures, has an advantage over other liquid fuels and Hyundai in the Republic of Korea manu- for cold weather starting. This characteristic pre- facture LPG-fueled taxis, and Ford in the United vents much of the engine wear and crankcase States manufactures LPG-fueled medium-duty oil dilution by fuel leaks during starting and trucks. Most gasoline-to-LP'G conversions decli- warm-up periods (CCEEB 1990). cate the engines to LPG use because of differ- Especially in LPG-producing countries, LPG ences in optimal settings for either fuel. is a cheaper fuel than gasoline or diesel fuel. However, dual-fuel LPG vehicles are also avail- The wholesale price of propane closely follows able. An LPG-fueled car or truck essentially has those of gasoline and diesel fuel. In the United the same engine as a gasolin,e-fueled vehicle with States the wholesale price of a liter of propane the exception of the carburetor or fuel injec- is about half that of a liter of gasoline.22 How- tion system. This is because LPG, which enters ever, this relative price advantage may not be the engine as a gas, requires metering through maintained at the retail level because of the throttling instead of the complex liquid-to-gas higher transportation and storage costs and atomization system used for gasoline. Storage of greater retail margin (due to lack of economies LPG at moderately high pressure (up to 200 psi) of scale) for LPG. requires the use of cylinders, which take about LPG is an attractive automotive fuel because 45 percent more space than a gasoline tank hold- it is cheaper than gasoline or diesel fuel in many ing an equivalent amount of fuel. Although LPG locations. In addition, the cost of converting a tanks are heavier, the fuel is lighter than gaso- gasoline- or diesel-fueled vehicles to LPG is low. line, so the total weight of the tank and fuel is The cost of converting light-duty gasoline-fueled about 20 kilograms more for a mid-size LPG-fui- vehicles to LPG ranges from $2,000 to $3,000, eled car than for a gasoline-fueled car Reftuel- 22. In mid-1996 the Gulf Coast wholesale price of LPG 21. All consumption figures are for 1995. varied between $0.073 and $0.092 a liter. 94 Chlapter 3 depending on the size of the fuel cylinders. For As a transport fuel, methanol contains 48 per- high-use vehicles (taxis, light-duty commercial cent of gasoline's and 43 percent of diesel fuel's trucks, urban buses) this cost usuallv can be re- energy content per liter (see Table 3.15). How- covered within a few years. A new LP-futeled ever, this lower energy density is somewhat com- medium-dtity tmick costs about $1,000 mor-e than pensated by methanol's higher octane number, a similar gasoline-fueled tmick. which allows higher engine compression ratios Canada is an example of a countrv where suc- and improved fuel efficiency. As a result about cessful conversion to LPG (propane) has oc- 1.8 liters of methanol are required to travel the curred (Box 3.5). In the United States LPG is same distance as 1.0 liter of gasoline. Thus used mainly in commercial fleets, including light- methanol-fueled vehicles require more frequent and medium-sized commercial trucks (for ex- refueling for a given fuel storage tank capacity ample, for carrier services and newspaper distri- or greater fuel storage capacity (and cost) for bution), taxis, police cars, and buises (school the same refueling frequency. buses, urban buses, and special service buses). Because methanol has a higher auto-ignition In Houston, Texas, conversion of the urban bus temperature than diesel fuel, it does not ignite fleet was initiated in 1990. Emission tests per- in a diesel engine under normal operating con- formed on these buses indicated full compliance ditions. For this reason the use of methanol in a with Califormia's stringent standards for NMHC, diesel engine requires installation of either a CO, NOR, and PM (J.E. Senior Constiltants Inc. spark plug to convert the engine into a spark- 1993). ignition engine or a heating element (called a "glow plug") to raise the temperature in the com- Methanol. Methanol, or methyl alcohol, is a bustion chamber above the auto-ignition tem- chemical that can be produced from natural gas, perature of methanol. liquid hydrocarbons, coal, or biomass. The most Combustion of methanol does not produce economical way to produce methanol is from SO2 or lead emissions. As compared to gasoline, natural gas. Methanol is a feedstock for the pro- combustion of methanol in motor vehicles pro- duction of chemicals such as formaldelhyde, sol- duces lower emissions of benzene and polycyclic vents, and acetic acid. In the past fifteen years, aromatic hydrocarbons. Methanol-fueled ve- it has also been increasingly used to produce hicles, however, emit unburned methanol and MTBE, an octane-boosting gasoline additive. considerably higher levels of formaldehyde than The use of methanol as a dedicated transport gasoline-fueled vehicles. Despite the high reac- fuel has been limited. tivity of fonnaldehyde, the ozone forming po- Box 3.5 Liquefied petroleum gas in Canada Thie Canadian government began promoting the use of propane in the transport sector in 1980 by providing a $280 grant to vehicle owners to convert to propane or purchase a new propane-fueled vehicle. This program was bolstered by provincial governments through such incentives as the removal of sales taxes on propane, propane-fueled vehicles, and propane conversion kits. By 1985 about 130,000 vehicles, most of which were commercial (urban truck fleets and taxis), had been converted to propane and were served by about 4,000 propane fueling stations. The payback perio(ls were about three years with the incenitives and four years with- out. Because of the plentiful supply of propane, the federal government's financial incentives wvere removed in March 1985 and efforts were directed toward working with provincial governments and the propane industry to foster continued growth in the demand for propane as a transport fuel. The government's new measures in- cluded a research and development partniership between the federal government of Ontario arid Chrysler Canada to provide propane technology on vehicles manufactured in Canada; demonstration project to pro- nmote the use of propane in urban buses in Ottawa; a grant to the industry association for mnarket development; public information campaigns; and use of propane within thle Canadian federal fleet. The use of propane in the transport sector is expected to continue to grow as retail fuel pricing differentials do not diminish. Sourre: Sathaye, Atkinsoni, and Meyers 1989; Saiive 1989. Abatement Measut res for Vehicular Air Pollution: Fuel-Targeted Measures 95 tential of total HC emissions is less for metha- cent less ozone-forming HC emissions (CCEEB nol than gasoline. Tests performed on catalytic 1990). converter-equipped methanol-fueled light-duttv Methanol has been used in professional rac- vehicles indicate that NO, emissions are at least ing cars because it allows fast acceleration and twice as high as their gasoline counterparts. As speed in high-performance engines. There are compared to diesel fuel, methanol-fueled heavy- only about 800 dedicated methanol-fueled ve- duty engines emit substantially less PM and, for hicles worldwide. About 600 passenger cars- some engine designs, less NO,. Methanol-fueled manufactured by Ford and Volkswagen-have buses, however, require evaporative emissions been operated under the California Energy controls that are not needed with diesel-fueled Commission's (CEC) methanol vehicle program. buses as well as oxidation catalysts for control- This program has shown the potential applica- ling CO, methanol, and formaldehyde emissions. tion for methanol-fueled vehicle technology (CCEEB 1990). under warm weather conditions, while indicat- Because of its low volatility and high heat of ing areas for improvement. In 1986 the commis- vaporization, methanol is associated with diffi- sion initiated a new phase aimed at developing cult engine starting at ambient temperatures flexible-fuel vehicles that can operate on gaso- below 10GC. This also causes high emissions of line, methanol, ethanol, or mixtures of these CO, methanol, and formaldehyde. In addition, fuels. methanol-fueled vehicles have higher mainte- In Brazil methanol was used as a component nance costs than gasoline-fueled vehicles be- of the transport fuel blend MEG during 1990-93 cause methanol and the formic acid it forms are to overcome the ethanol shortage caused by a corrosive chemicals and can dissolve materials drop in sugar cane production. During this time such as solder, aluminum, and rLbber. Experi- about 7 billion liters of MEG were consumed in ence with certain engine designs suggests that the State of Sao Paulo, mostly in the Sao Paulo the engine life for these designs is likely to be metropolitan area. This fuel-a mixture of 33 considerably shorter than for gasoline-fueled percent methanol, 60 percent ethanol, and 7 engines. Although the excessive engine wear percent gasoline-has properties and fuel con- problem has been substantially solved by fuel and sumption characteristics similar to ethanol. The lubricant additives, methanol-fueled cars still emissions from both fuels are the same, except require frequent oil and filter changes. that acetaldehyde emissions from MEG are about Another disadvantage of methanol is associ- 50 percent lower and formaldehyde emissions ated with its safety characteristics. Methanol are about 5 percent higher. Use of MEG in etha- burns with a nearly invisible flame during day- nol-fueled light-duty vehicles does not require light, making detection and control of fires dlif- any special engine modification or calibration. ficult. In addition, at ambient temperatures The use of methanol in diesel-fueled heavy- methanol vapors in the ftiel tank are in the ex- duty vehicles also has been studied. Since 1983 plosive range. Furthermore, methanol is a toxic the Golden Gate Transit Company has been chemical. Its adverse health effects include skin operating two methanol-fueled buses in San irritation, visual disturbances (including blincd- Francisco. One of these buses was modified with ness), loss of muscular coordination, dizziness, a spark plug and the other with a glow plug. The nausea, abdominal cramps, delirium, coma, con- technical problems encountered with these vulsions, respiratory failure, cardiac arrest, and buses have included fuel pump breakdowns and death. frequent spark plug or glow plug failures. Par- To improve the cold start, flame luminosity, ticulate emissions from these buses are lower and vapor explosivity problems of pure metha- than from comparable diesel-fueled buses, but nol, a blend consisting of 85 percent methanol CO and formaldehyde emissions are higher. Vari- and 15 percent gasoline (M85) was developed ous transit authorities in the Los Angeles area and tested on vehicles. Use of the M85 fuel re- also have demonstrated the application of quires simple modifications on gasoline-fueled methanol on heavy-duty diesel vehicles (urban cars, including changes to the carburetor (or fuiel buses, school buses). In addition, several diesel injection system) and the materials in the fuel engine manufacturers (Caterpillar, Cummins, delivery system, and an increase in the compres- Ford, Detroit Diesel, and Navistar) are conduct- sion ratio. M85-fueled cars have CO emissions ing research on the use of methanol in diesel similar to gasoline-fueled cars but 30 to 44 per- engines. 96 Chapter 3 A major drawback of methanol as a transport achieved through use of a palladium catalyst fuel is its relatively high and volatile price rela- relocated near the engine. For evaporative emis- tive to conventional transport fiiels. The price sions, a charcoal canister placed near the front of methanol in the world market increased from wheel was first used. This control was subse- $0.06 a liter in the early 1980s and about $0.17 a quently improved through the installation of liter in the late 1980s to about $0.47 a liter in thermal insulation between the carburetor and 1994. Methanol would have to cost $0(.10 a liter the engine. to compete with the spot price of gasoline of Ethanol does not emit S02 and has negligible $0.18 to $0.20 a liter on an equivalent energy PM emissions. SO2, PM, CO, and NO, emissions efficiency basis (Faiz, Weaver, and Walsh 1996). from uncontrolled ethanol-fueled vehicles are There is little prospect for methanol to become less than those from gasoline or gasohol-ftieled price-competitive with conventional transport vehicles of the same model-year. Because of fuels unless world oil prices increase substantially. ethanol's lower CO emissions, 10 percent etha- nol blends (or MTBE gasoline blends) in gaso- Ethanol. Ethanol, or ethyl alcohol, is a chemical line are required in certain U.S. cities. that can be produced by fermentation of sugar Uncontrolled ethanol-fueled vehicles emit more extracted from biomass such as sugarcane or aldehydes (especially acetaldehyde) from the ex- corn, or by catalytic hydration of ethylene. It is haust than similar gasoline-fueled vehicles. How- marketed as a final product, industrial feedstock, ever, these emissions can be controlled with the or transport fuel. As a transport fuel ethanol has use of a catalytic converter. Cold start emissions been used as a dedicated fuel in hydrous form from ethanol- and gasohol-fueled vehicles are or in combination with gasoline (the resulting not a problem in countries with a hot climate fuel is called "gasohol") or gasoline and metha- (such as Brazil) but may be a concern for other nol (the resulting fuel is called MEG: see above) locations. Quantitative data on typical emissions in anhydrous form. Ethanol marketed as a trans- from ethanol-, gasohol-, and gasoline-fueled ve- port fuel is derived from biomass in most coun- hicles are shown in Chapter 4 in the section on tries and its price as a transportation fuel is tied Sao Paulo. to food prices. In Brazil the ethanol industly (the In the United States subsidies ensure that etha- world's largest) is based almost entirely on sugar nol is produced from biomass (mostly by corn cane, whereas more than 80 percent of the U.S. fermentation) instead of ethylene hydration. ethanol plant capacity (the world's second larg- Ethanol production in the United States costs est) uses corn as the input material. about $0.33 a liter ($1.24 a gallon), including As a transport fuel, ethanol's properties lie the subsidy (Rendleman and Hohmann 1993). between methanol and gasoline. The energy Since a liter of ethanol contains two-thirds the density of ethanol is about two-thirds that of gaso- energy of a liter of gasoline, the wholesale price line but greater than that of methanol. Ethanol of ethanol at the plant gate, on an energy equiva- has about the same octane number as methanol lent basis, ranges from $0.39 to $0.61 a liter. The (see Table 3.15). Because it is water soluble, it State of California's Energy Commission has con- must be kept free of water if it is blended with cluded that ethanol production from conven- gasoline for use as a transport fuel. Its corrosion tional feedstocks (such as corn) would not be characteristics are milder than those of economical in California. In addition, the U.S. methanol's but more severe than gasoline's. Department ofAgriculture has argued that, since Ethanol or gasohol can be used in spark-igni- much of the benefit of ethanol subsidies goes to tion engines with the same type of engine modi- large producers and retailers instead of farm- fications required for methanol-fueled vehicles. ers, subsidized ethanol production is an ineffi- The exhaust emissions control technology used cient way of raising farm income. Furthermore, in Brazil for ethanol- or gasohol-fueled light-duty the subsidy program raises beef and pork prices vehicles has improved over the years. The cur- in the United States (CCEEB 1990). rent technology includes multipoint fuel injec- In Latin America and the Caribbean ethanol tion, mapped electronic ignition, and is used in Brazil (Box 3.6) andParaguay as alight- closed-loop three-way catalytic converter-s with duty motor vehicle fuel. In Brazil ethanol is used feedback control of the air-fuel ratio (Branco and instead of gasoline either as an anhydrous form Szwarc 1993). Control of aldehydes to nmeet strin- in a blend with gasoline (78 percent gasoline gent emission requirements is expected to be with 22 percent ethanol) or in hydrous form Abatement Measures. for Vehicular Air Pollution: Fuel-Targeted Measures 97 Box 3.6 Proalcohol program in Brazil The Proalcohol Program in Brazil has been the largest alternative fuels program in the world to date. This program was initiated by the Brazilian governinenit in 1975 in response to a sharp increase in world oil prices that adversely affected Brazil's foreign debt and ecolnOmic growth and a rapid decrease in world sugar prices that put the recently expanded and modernize(d Brazilian sugar industry in a difficult financial situation. The Proalcohol Program was implemented in two phases. The first pliase, implemented during 1975-79, aimed at increasing the percentage of ethaniol in gasohol to 20 percent nationwide. The second phase, carried out dur- ing 1979-85, focused on providing 100 percent hydrous ethianol (containinig 96 percent ethanol and 4 percent water by volume) to ethanol-dedicated vehicles. Ethanol produced by private distillers was p)ur(chasedl by government entities (first the National Petroleum Council and then by Petrobras, the national petroleumln company of Brazil) on a sugar-equivalent basis accord- ing to fixed prices and quotas. Incenitives for producers included credit subsidies for up to 75 percent of invest- ment costs and government assurance of a 6 percent return on investments for supplying ethanol to the government. In 1979 the goals of the program's first phase were reached because the 20 percent ethanol blend did not require major modifications in vehicle or fuieling station technologies, and the ethanol supply was provided through use of excess capacity and rapid construction of new distilleries. Auto manufacturers demon- strated the feasibility of using ethanol as a dedicated fuel in spark-ignition engines, but did not commit them- selves to mass production. The second phase of the program promoted the utse of hydrous ethanol in light-duty vehicles. It was initiated as a result of increasing world oil prices and Brazilian foreign debt. Specific goals of this phase included increas- ing prodtiction of ethanol-dedicated vehicles to 50 percent of new vehicle sales and supplying and distributing 10.6 billion liters of ethanol by 1985. The second( goal required an investment of $5 billion for a 150 percent increase in alcohol production capacity. The necessary funds were provided by both domestic sources (fuel taxes, vehicle licensing fees, and so on) and international financial institutions. Consumer incentives to buy new cars included lower purchase taxes and registration fees, smaller downpayments, longer repayment periods, and lower fuel costs. For example, taxi drivers buying new vehicles were exempt from taxes-the equivalent of about a 50 percent discount. The pump price of ethaniol was guaranteed to be no more than 65 percent of gasohol, providing ethanol a 20 percent advantage in cost per kilometer. However, the actual ratio of the etha- nol-gasoline price varied over time. The auto manufacturing industry invested hleavily in tooling and research and development on ethanol- fueled vehicles. However, the first generation ethaniol-dedicated new vehicles and retrofits caused customer dissatisfaction because of the lack of technical expertise by auto manufacturers in producing alcohol-fueled cars and by unauthorized mechaniics in conducting poor vehicle conversions. The auto manufacturinig industry regained public confidence by improving enginie qualitvy. As world sugar prices rose, the government raised the price of ethanol in the late 1980s from 40 percent of the price of gasoline toward the 65 percent limit. This measure resulted in a drop in ethanol-fueled vehicle sales. The incentives for distillers and consumers were removed and then restored, in some cases more strongly, raisinig the demand for ethanol-dedicated cars in 1984 and 1985. Between 1976 and 1985 investment in the Proalcohol Program totaled $3.7 billion for thie Brazilian govern- ment and $2.7 billion for thie private sector, antd the foreign exchange savings were $8.9 billion. In 1983 the real cost of ethaniol to replace a barrel of gasoline was estimated at $40 to $65 in the southeast regions and $100 in the northeast regions. Other implications of Proalcohol included the use of additional land for sugar cane plantations, creation of newjobs (about 900,000 direct jobs), increased agricultural productivity for sugar cane (from 46 tons a hectare before the program to 54 tons a hectare in 1987, and up to 94 tons a hectare in the State of Sao Paulo), increased ethanol conversion productivity from sugar cane (from 57 liters a ton before Proalcohol to 71 liters a ton in 1987), increased enviroinmenital emissions from distilleries located in rural areas, and decreased motor vehicle pollution in urban areas for ethanol-to-gasolitne (or ethanol-to-gasohol) substitution. In addition, in- creasing national detnand for diesel fuel and penetration of ethanol in the transport fuels market put consider- (Box continus on the follnoing page.) 98 Chapter 3 Box 3.6 (continued) able pressure on Petrobras' refinery operations. Greater quantities of surplus gasoline had to be exported to foreign markets, particularly to the United States. Although the Proalcohol program in Brazil has been a remarkable technical success, it has also become an economic burden. In 1995 subsidies to the program totaled $1.9 billion, or $0.15 a liter. Based on current gasoline price projections, it is expecte(d that the use of ethanol as a gasoline additive will completely dominate the ethanol fuel by the turn of the centurv. Howvever, with about 4.5 million ethanol-dedicated cars still driving the streets of Brazil, a steady and reliable supply of this fuel will continue to be needed if the engines of these vehicles are not modified. Sonrce: Faiz, Weaver, and Walsh 1996; Sathave. Atkinison, and Meyers 1989; Trindade and Garvalho 1989. (with 4 percent water content). In 1995, 11.3 ferentialwhichis expected to provide incentives million motor vehicles in Brazil used ethanol or for converting diesel-fueled vehicles to CNG. ethanol blend. In Paraguay about 10 percent of Differential fuel taxes can be used between cars are fueled with ethanol. Although air qual- leaded and unleaded gasoline or between low- ity improvements can be achieved through the and high-sulfur diesel fuel to promote consump- use of ethanol as an alternative fuel, the Brazil- tion of the cleaner fuel. A tax differential be- ian experience indicates that there is a delicate tween leaded and unleaded gasoline is balance among sugar cane farmers, alcohol dis- commonly used as an economic incentive to tillers, petroleum refiners, and auto manufac- curtail air pollution. As a percentage of pretax ttirers. Given a sharp rise in oil prices, countries price, unleaded gasoline is taxed lower than that would potentially have an interest in em- leaded gasoline in all European Union countries barking on a similar program would be those except Greece. This has contributed to a sub- with food surpluses and energy deficits. stantial increase in the use of unleaded gasoline in these countries. For example, the consump- Fuel Taxes tion of unleaded gasoline in the European Union rose from 1 percent of the gasoline mar- Fuel taxes are intended to serve as a proxy for ket in 1986 to 41 percent in 1991 and to 54 per- vehicle emission taxes in order to reduce vehicle cent in 1992. In Germany, for example, although travel and air pollutant emissions. Ftiel taxes can premium leaded gasoline is still available (sale be used differentially among various fuels to of regular leaded gasoline is banned), it is con- promote the consumption of cleaner ftiels. For siderably more expensive than unleaded gaso- example, in Mexico the retail price differ-ential line because higher taxes are levied on it (T6pfer between diesel fuel and CNG (3.4 percent in 1995). In 1992 unleaded gasoline accounted for 1994 and 1.1 percent in 1995 in favor of CNG) 85 percent of the gasoline market in Germany, was widened (to 11 percent in 1996) by differ- 76 percent in the Netherlands, and 47 percent ential taxation to promote greater use of CNG- in the United Kingdom (CONCAWE 1994). fueled vehicles (O'Ryan 1996). In the Buenos In Thailand, before elimination of lead in Aires Metropolitan Area more than 80 percent gasoline, unleaded gasoline was taxed at a lower of gasoline-fueled taxis have had their fuel sys- rate than leaded gasoline to nearly equalize the tems converted to CNG in response to the Ar- prices of these grades at the pump (the price of gentine government's heavy taxation of gasoline unleaded gasoline is only $0.004 a liter less than and tax exemption for CNG. In contrast, less that of leaded gasoline). As a result the switch than 2 percent of the diesel-fueled buses in the from leaded to unleaded gasoline occurred in same area have had their fuel system converted less than five months (Bartone and others 1994). to CNG because until recently both diesel fuel In Singapore a differential tax caused unleaded and CNG were exempt from fuel taxes and had gasoline to cost $0.10 a liter less than leaded about the same retail price. However, a recent gasoline. tax placed on diesel fuel has created a price dif- In Latin American and the Caribbean coun- Abatement Measuirs for Vehicular Air Pollution: Fuel-Targeted Measures 99 tries where both leaded and unleaded gasoline Differential taxation policy has been used in are available, the leaded grade typically has a several European and Latin American countries lower tax and retail price than the unleaded to promote use of diesel fuels with lower sulfur grade. For example, inJanuary 1996, Peru's tax content. For example, in Denmark the lower- on low-octane gasoline was $0.225 a liter for the sulfur diesel fuel (containing 0.05 percent sul- leaded grade and $0.293 for the unleaded grade, fur by weight ) was given a $0.015 a liter tax with respective retail prices of $0.383 and $0.468 incentive compared to regular diesel fuel (con- a liter (Alconsult International Ltd. 1996). In taining 0.20 percent sulfur by weight). The tax Costa Rica unleaded gasoline had the same re- incentive for the low-sulfur diesel fuel used by tail price as leaded gasoline when it was first public buses was even greater ($0.045 a liter). marketed in 1989, but the price ratio of unleaded In Sweden, the low-sulfur diesel fuel marketed to leaded gasoline (which is determined by a in urban areas (containing 0.02 percent sulfur regulatory agency) was increased as high as 1.15 by weight) was taxed $0.025 a liter less than the in 1992 ($0.400 a liter versus $0.348 a liter) .2 high-sulfur diesel fuel (containing 0.1 to 0.2 This ratio was then reduced to 1.09 in early 1996 percent sulfur by weight) used for intercity trans- ($0.825 a liter compared with $0.755 a liter) just port. The tax incentive between the ultralight before lead was eliminated from gasoline. In diesel fuel (containing 0.001 percent sulfur by Urugutay the retail price of unleaded gasoline is weight) used by urban buses and the high-sul- about 5 percent higher than leaded gasoline of fur fuel was $0.058 a liter (CONCAWE 1994). In the same octane number. Brazil the two grades of diesel fuel (0.5 percent For some Latin American and Caribbean sulfur by weight for intercity transport and 0.3 countries, however, differences in retail leaded percent sulfur by weight for urban transport) and unleaded gasoline prices are small or non- are taxed differentially to yield the same retail existent. In Ecuador the respective taxes are price. The same policy was followed in Mexico $0.063 and $0.069 a liter and the retail prices for the two diesel fuel grades (0.5 sulfur by weight are $0.330 and $0.349 a liter (Alconsult Interna- for intercity transport and 0.05 percent sulfur tional Ltd. 1996). In Mexico the 1991 price gap by weight for urban transport) until the high- of $0.09 a liter between unleaded and leaded sulfur grade was totally eliminated from the gasoline grades has been reduced to $0.02 a li- market in February 1997. ter as a result of higher taxes on leaded gasoli ne Lower taxes and retail prices for motor vehicle (see Chapter 4). In Chile, retail prices of the 93- fuels also encourage more fuel consumption, octane leaded and unleaded gasoline grades which results in increased air emissions. Among have been the same since September. In Barba- Latin American and Caribbean countries, Ven- dos leaded and unleaded gasoline grades also ezuela has the lowest tax and retail price for low- sell at the same price. In El Salvador unleaded octane leaded gasoline (a tax of $0.016 a liter gasoline sold for $0.003 to $0.021 (0.6 to 5.2 and a retail price of $0.106 a liter in January percent) more a liter than leaded gasoline in 1996). In comparison, the fuel tax and retail 1994-95.24 However, in 1996, before lead was price for low-octane leaded gasoline in Paraguay eliminated from gasoline injuly, unleaded gaso- were $0.283 and $0.497 a liter, and in Uniguay line sold for up to $0.011 a liter less than leadled $0.431 and $0.790 a liter (Alconsult Interna- gasoline. tional Ltd. 1996). High taxes on fuels alter the market demand for energy-efficient vehicles, which in turn af- fect the design and production mix of newer 23. At the request of the national refinery (RECOPE). models and, possibly, the market share ofvehicle the regulatory agency determines fiuel prices in Costa Rica manufacturers. Evidence from industrial coun- by analyzing the cost strcttire of all petroleumit derivatives tries suggests that higher gasoline taxes and in the country. In addition to this analysis done at least once a year, the regulatory agency adjusts fuiel prices based on prices are associated with higher fuel efficiency. variations in internatioial fuel prices and the excliange rate. This is the case for Italy, which has high gaso- 24. SinceJanuary 1994 the Energy, Mines, and Hydrocar- line taxes (74.3 percent in 1993) and a fuel-effi- bons Department of the Ministry of Economy has deter- cient vehicle fleet. mined, on a weekly basis, the maximum prices that tiel importers and local refiners can bill to distributors. Distribu- Some lessons can be learned from the lead tor and retailer margins are not controlled by the goverin- elimination program in the United States. Dur- ment. ing implementation of this program no fiscal 100 Cllapter3 incentives were established to encourage the use ing and auditing mechanisms. A system was es- of unleaded gasoline. This lead to the contami- tablished for refiners to make a quarterly report nation and misfueling of gasoline even though of the amount of lead actually used, the amount the USEPA established some command-and-con- of leaded gasoline produced, and the amount of trol measures to prevent such events (Hillson lead the refiner was allowed to use. The USEPA 1995). In Mexico the large price difference be- received as many as 900 quarterly reports from tween unleaded and leaded gasoline grades in refiners and reviewed them for errors and self- 1991 also encouraged the use of leaded gaso- reported violations. In addition, the USEPA re- line, even in cars equipped with catalytic con- viewed reports sent by lead additive manuf-acturers. verters (such misfueling renders the catalyst USEPA officials expanded their monitoring of ineffective). lead trading through on-site audits of refiners. Tax incentives have also been adopted to en- These audits involved close scrutiny of refinery courage the use of cleaner fuels through vehicle records of gasoline production and lead use, or facility conversions. For example, since July banking, and trading. The audits revealed that 1994 the state of Connecticut (United States) some refiners had greatly overstated their pro- has been providing a corporate tax break to any duction volume to falsely indicate compliance commercial fleet owner that converts its vehicle with the lead standard. The USEPA issued civil fleet from gasoline to cleaner fuels, such as CNG penalties for violations, including a number of or electricity. For every dollar spent on convert- multimillion-dollar enforcement cases. As a re- ing vehicles to cleaner fuels or building ftuel sta- sult of the audits, refiners began taking greater tions with CNG, fleet owners have been allowed care in preparing their reports (Hillson 1995). to deduct 50 cents from their corporate profit The U.S. experience with the tradable permit tax. Moreover, after the conversion the new fuel system indicates that establishment of strong is exempt from the state fuel tax ($0.08 a liter). oversight coupled with financial penalties for violations are necessary for the success of such a system. Tradable Permnits and Credits The marketable (tradable) credit system has also been used for reformulated and oxygenated In the early 1980s a tradable permit system for fuel programs in the United States. Under such lead was initiated in the United States. This sys- a system the credit is given to refiners, blenders, tem established rights for refiners to use certain or importers who certify that the quality of their amounts of lead additives and allowed them to gasoline is superior than that established in the trade unused lead rights to other refiners. The 1990 amendments to the Clean Air Act for re- 1985 rule, which reduced the lead standard in formulated or oxygenated gasoline. The param- gasoline from 0.291 gram a liter (1.1 grams a eters used to determine fuel quality are oxygen, gallon) to 0.027 gram a liter (0.1 gram a gallon), aromatic hydrocarbons, and benzene under the also allowed refiners to bank lead rights. These reformulated fuel program and oxygen under rights could be used or sold to others through the oxygenated fuels program. The credit may the end of 1987. During this period trading of then be used by the same person or transferred lead rights became very common in the United to another person for use within the same States. Lead rights, which were worth about $0.05 nonattainment area during the allotted period a gram of lead, provided significant incentives specified in the regulations. No transfer of credit for unscrupulous companies to sell false lead is allowed between nonattainment areas. The rights at discount prices (Hillson 1995). marketable credit system establishes require- However, the tradable permit system for lead ments for the issuance, application, and trans- was monitored by the USEPA through report- fer of credits. TRANSPORT MANAGEMENT MEASURES Many large cities, including those in Latin uncontrolled, then the emission reductions America, suffer from traffic congestion caused would be even greater: about 70 percent for NO, by rapidly rising ownership of private cars. In and over 99 percent for HC and CO. In addi- addition, the increasing number of public trans- tion, a shift to the public transport mode sup- port vehicles, which use urban roads to cope with presses extensive demand for road use and the transportation needs of quickly growing low- improves the traffic flow, which further reduces and middle-income urban populations, contril- pollutant emissions. utes to traffic congestion. The package of transport management Traffic congestion burdens productivity be- options chosen must be tailored to each urban cause of the cost of time lost during travel. In center because many factors-such as physical addition, the slow movement of motor vehicles infrastructure, characteristics of the urban on congested roads results in higher fuel con- transport system, layout of the urban area, and sumption and increased rates of pollutant emis- transport demand-need to be taken into con- sions. Typical relationships between vehicle sideration. In all cases the selected options must speeds and emission rates of major air pollut- be able to meet specific needs, beneficial and ants from gasoline-fueled cars, gasoline-fueled fair to the community as a whole, flexible enough heavy-duty vehicles, and diesel-fueled heavy-dutty to respond to changing situations, simple and vehicles are shown in Figure 3.3. As the figure inexpensive to enforce, and easy for users to shows, HC and CO emissions from all vehicle understand and comply with (ECMT 1990). types and NO, emissions from heavy-duty ve- hicles fall when traffic speeds reach 50 to 60 ki- lometers an hour-a speed that is feasible Driving Bans wherever road safety and the physical layout of the urban area permit. Safety and layout con- Measures have been taken to curtail air pollu- cerns, in turn, can be achieved through trans- tion in urban centers by banning vehicle circu- port management measures. lation based on vehicle type, day of the week, Transport management options are mostly di- time of the day, or location. Vehicle bans have rected toward promoting the use of public trans- been used in different forms in a number of port in lieu of private cars. The use of public urban centers. One of the best-known types of transport vehicles can reduce emissions per pas- vehicle ban is the "odds and evens" scheme. senger traveled. For example, in Mexico City the Under this scheme vehicles with license plates use of a bus in lieu of a catalytic converter- ending in an odd number are allowed to drive equipped private car is estimated to result in a on odd dates and vehicles with license plate end- reduction of 40 percent in NO. emissions, 95 ing in an even number are allowed to drive on percent in HC emissions, and 98 percent in CO even dates. emissions per passenger-kilometer traveled Since 1989 Mexico City has used another form (Table 3.18). If the private car is assumed to be of driving ban in which cars are restricted from 101 102 Chapter 3 Figure 3.3 Pollutant emission rates from vehicles as a function of vehicle speed CO (g/km) Carbon Monoxide 900 800 - Heavy-duty 700 - ehicled 600 - 500 - 400 - Hea 300 - diesel-fueled 200 - '.vehicle 100 - Passenger car * ... 0- I I Ir 0 10 20 30 40 50 60 70 80 90 Speed (km/hr) HC (g/km) Hydrocarbons 50 - 45 - 40 - 35 - 30 - 25 Heav-dty Heavy-duty gasoline-fueled vehicle 20 - diesel-fueled 15 - v,ehicle 10 --...asne.a 5 ...... 0- 0 10 20 30 40 50 60 70 80 90 Speed (km/hr) NO. (g/kmn) Nitrogen Oxides 35 _ 30 25 20 - ' * . ., Heavy-duty diesel-fueled vehicle 15 10 Heavy-duty gasoline-fueled vehicle . - F ; -; -- -- - T-- Passenger car 0 --- --- -r TI - … …__- I ~ - _ -- --- -I _ - I 0 10 20 3() 40 50 60 70 80 90 Spee(d (km/hr) Note: Emission rates are for vehicles not equipped with pollutiot) control devices. Sourre: Faiz, Weaver, and WValsh 1996. Abatement Measuresfor VehicndtarAirPollution: Transport Management Measures 103 Table 3.18 Pollutant emissions by different transport modes in Mexico City (grams per passenger-kilometer traveled) Transport mode CO H( NOQ Private car (catalytic converter-equipped) 45.2 4.4 1.0 Taxi 96.9 9.5 2.2 Microbus 9.7 1.2 0.2 Urban bus 0.7 0.2 0.6 Note: For comparison, tnicks are estimated to emit 20.77 grains of NO, per passenger-kilometer traveled, 7.71 grams of HC per passenger-kilometer traveled, and 18.76 grams of CO per passenger-kilometer traveled. So2ir1e: CMPCGA 1995. driving one day of a weekday depending on their Truck passage on narrow roads and on weak or license plate number. This ban has not proved low bridges is also restricted based on weight, effective because many families in Mexico City width, or length limitations. Some urban areas have purchased a second car with a different li- have implemented bans on truck circulation in cense plate number for use on the banned week- an effort to maintain smooth traffic flows dur- day. In addition, most of these second cars, which ing the daytime or during rush hours. In certain were brought in from other parts of Mexico, cities (for example, Windsor in the United King- adversely affected ambient air quality because dom) truck circulation is totally banned. they were older models with no or limited pol- Closure to traffic of traffic-congested commer- lution control equipment. As a result of the ap- cial areas or environmentally sensitive residen- parent failure of this driving ban, a two-day ban tial areas of urban centers reduces vehicular air has been in place since December 1995. Under pollution in those sections. For commercial ar- this program 40 percent of the vehicles (except eas this measure is even more effective if pedes- buses) in the MCMA are prohibited from circu- trians in the banned area benefit from improved lation when the air pollution level exceeds a cer- public transport services. Traffic bans have been tain level (IMECA 250). This emergency measure implemented in some Latin American cities, has only been applied once, and it received quite including Buenos Aires, C6rdoba, Lima, a negative response from the public. Mendoza, and San Jose. A similar program was implemented in Buenos In SanJos6 (Costa Rica) the municipality closed Aires for a short period, but was suspended be- the Central Avenue to traffic. To achieve this the cause it was found to be ineffective. In addition, municipality, which was in poor financial condi- in Buenos Aires and Santiago taxis with no pas- tion, convinced local businesses to finance 50 sengers are prohibited from entering certain percent of the required works. The municipality downtown areas. As a result of this program and also prohibited street vendors on the Central Av- of parking ban measures, the traffic flow in a enue and enforced this rile strictly. This pedes- forty-block section of Santiago was reduced by trian area provides a nice environment for the 30 percent. Vehicle bans are also used in some people of San Jose. In Cartagena, however, fol- European cities such as Bologna, Cologne, and lowing the traffic ban in certain sections of the Rouen (ECMT 1990). The government of Ath- colonial part of the city, streets became greatly ens, after implementing a program that allowed crowded by informal traders. Following com- cars to enter the downtown area only on alter- plaints by local businesses and residents, the ban nate days, initiated a three-month experimen- was lifted within three years of implementation. tal total ban for cars, taxis, and motorcycles in In some German cities streets in certain down- April 1995 (The Times, 11 April 1995). town areas have been closed to traffic since the The movement of heavy trucks, although nec- late 1970s. This measure has brought remarkable essary for some businesses in city centers, gen- success in increasing the number of pedestrians erates air and noise pollution. Trucks can also and decreasingmotorvehicle traffic. For example, be dangerous to other road users, including pe- within one year pedestrian flow had increased by destrians and bicyclists. In addition, they dam- 40 percent in Banberg and 25 percent in Aachen age roads because of their heavy axle loadings. (Hass-Klau 1993). 104 Chaplter 3 On-Street Parking and Trading divert street trading to more suitable locations. Restrictions In addition, the creation of pedestrian streets attracted both pedestrians and street traders. As In many urban areas road capacities are reduced a result the traffic flow in the central area im- mostly by on-street parking and stopping of ve- proved (World Bank 1986). hicles to load and unload passengers and goods. Especially during peak hours, these movements slow down the traffic flow and raise pollutant Traffic Priority Measures for Buses emissions from motor vehicles. Traffic flows in heavily congested streets can be improved by Traffic priority measures for buses in urban cen- imposing on-street parking restrictions, which ters increase the attractiveness of bus services can be implemented at all times, during the day- by reducing transit time and increasing the reli- time, or only at peak hours. These restrictions ability of service. These measures also have require installation of signs and a firm enforce- environmental benefits: they curtail fuel con- ment effort through fining, wheel clamping, and sumption and pollutant emissions from road towing. Although on-street parking restrictions transport because they allow faster speeds for are simple and inexpensive to implement, they buses and eliminate cars from traffic by shifting can generate public opposition. For this reason, users to bus transport. Types and effectiveness in devising a strategy for parking restrictions, ofvarious bus traffic priority measures are shown priority should be given to business and shop- in Table 3.19. Among these, the most effective ping traffic instead of commuters who can use measures include exclusive bus routes and lanes, public transport. contra-flow lanes for buses, and bus flow lanes To improve traffic flow in the central area during rush hours. of San Jose, a World Bank-financed project Bus routes and exclusive bus lanes that are implemented parking restrictions and banned physically separated from general traffic by parking along priority bus routes. The project means of barriers-also called "busways"-have also installed parking meters to encourage been effectively used in some Latin American quick turnover, designed formal loading and urban centers. For example, in Santafe de unloading spaces, and strengthened enforce- Bogota two lanes in each direction of a major ment. These measures proved effective-the avenue have been physically separated from gen- travel speeds of buses increased and the flow of eral traffic for use as a busway. Busways are also the remaining traffic improved (World Bank used in such Latin American cities as Curitiba, 1986). Porto Alegre, and Lima. Street trading in crowded streets can adversely Another priority measure involves dedicating affect traffic flows because it attracts pedestri- certain lanes of an existing route to buses. Bus ans. This problem can be avoided by introduc- lanes are very common in European and North ing controls on street trading. In Lima, Peru American cities, and are also used in some Latin street trading was tackled by setting up a special American cities (such as Buenos Aires and force to enforce street trading regullations and Santiago). In some cities bus lanes are used solely Table 3.19 Effects of bus priorities on bus exhaust emissions (percent) Pofrtion f daily bus Redurtion in bzus Redution in bits 7jpe of measutre travel affecled journey times exhitztast emtissions Peak period with flow bus lanes 5 15 20 Contra-flow lanes (all-day) 2 30 35 Motorway privileges 1 50 60 Signial preemption 20 10 12 Bus routes 2 50 60 Priority turns and other measures 5 5 7 tSource: ECMT 1990. Abatement Measures for Vehicuilar A ir Pollution.: Transport Management Measures 105 during rush hours (Washington, D.C.). Effective fective in locations where traffic congestion operation of bus lanes requires enforcement to spreads to the suburbs of a metropolitan area. ensure that the lanes are not occupied by cars. Successful car/van pooling programs establish Bus lanes cut bus and car travel times by up to traffic priorities for high-occupancy vehicles, 30 percent in Bangkok. In Manila the travel parking privileges for ride sharers, assistance in time of buses was halved because bus lanes in- matching commuters, and an employer incen- creased their average speed from 9 to 18 kilo- tive program (ECMT 1990). For example, in meters an hour. major traffic corridors leading into Washington, The contra-flow lane measure involves estab- D.C. high-occupancy vehicle lanes and routes are lishing an exclusive bus lane counter to the di- maintained in the direction toward the city dur- rection of general traffic. This measure is ing the morning rush hours and in the opposite effective because it ensures that bus drivers and direction during the evening rush hours. The general traffic stay in their lane to avoid head- minimum occupancy requirement in vehicles on collisions. using these routes is set at two or three people The benefits of bus priority measures outweigh depending on the corridor. Special parking lots their costs only in highly congested corridors. for ride sharers are also provided at suburban In addition, busways and bus lanes are not very areas near the high-occupancy vehicle route. effective in corridors where traffic is interrupted Assistance in matching commuters is provided by the movement of crossing vehicles. To reducec through telephone hotlines. Employer incen- such delays, buses in certain cities are fitted with tives-such as company-provided or arranged mechanisms that modify the traffic signals for transportation, or free parking space for high- priority passage (that is, the traffic signal stays occupancy vehicles-encourage employees to or turns green when buses approach). share rides. Bus priority measures cost less than subways as a means of public transport. For example, an exclusive bus lane requires investment of less Staggered Work Hours than $1 million per kilometer, a busway in an existing corridor costs about $2 million per kilo- Traffic congestion in urban areas generally oc- meter, and a new busway costs $7 million to $12 curs before and after work, school, and shop- million per kilometer. These costs compare with ping hours. One measure to relieve congestion $22 million to $60 million per kilometer for an and reduce vehicular air pollution involves stag- elevated metro and $50 million to $165 million gering of these hours. In addition, staggered per kilometer for an underground metro. Metros work hours reduce the load on public transport have several key advantages over bus priority systems. Staggered work hours have been used, measures, however, they can operate at higher on a compulsory basis by some governments and speeds (28 to 37 kilometers an hour for metro private businesses. For example, in the Repub- compared with 15 to 22 kilometers an hour for lic of Korea staggered work hours are used by a busways or bus lanes) and carry more passen- major television company. In Singapore there gers (50,000 to 75,000 passengers an hour per are different peak hours for two-shift schools and lane for metro compared with 10,000 to 25,000 businesses. In many European cities shop clos- passengers an hour per lane for bus priority ings are delayed on the busiest days of the year measures; UN 1994). to smooth out shopping and other peak hours (ECMT 1990). In cases where the nature of employment does Ride Sharing not require all employees to start and finish work together, staggered work hours have been used Another option for reducing congestion and on a voluntary basis under "flextime" arrange- curtailing air pollution involves setting a require- ments. However, widely staggered work hours are ment for the minimum number of people trav- not acceptable for many businesses because they eling in light-duty vehicles (cars and vans). Such reduce the interaction among workers or with a measure necessitates building special infi-a- clients. Another impediment to staggered hours structure that includes, at a minimum, special arises in situations that require multipurpose routes or lanes in an urban area. Ride sharing trips such as taking a child to school on the way (also called car or van pooling) is especially ef- to work (Gwilliam 1995). 106 Chlapter 3 Speed Limits and Other Traffic fies areas for residential, educational, commer- Management Measures cial, industrial, and recreational facilities, plan- ning controls future population densities in Speed limits in urban areas are set priimarily for these areas that affect the demand for transport road safety, but vehicle speeds also affect fuel services and the associated air pollution. Effec- consumption and pollutant emissions from ve- tive land use planning reduces the amount of hicle exhausts. Fuel consumption starts to in- travel for work and other activities by bringing crease at speeds above 60 kilometers an hour. schools, offices, shops, and recreational facili- Although HC and CO emissions are not sensi- ties closer to residential areas. Environmentally tive to speeds above 60 kilometers an hour; NO, sound land use policies create mixed-use multi- emissions increase slightly for gasoline-fueled nucleated urban areas that promote walking for vehicles and sharply for diesel-fueled vehicles. short trips and mass transit for long trips. Still, In residential areas traffic speeds can be effec- even though land use planning can create a de- tively controlled by a combination of speed limits sirable urban environment for the future, it can- and careful design of road layouts (ECMT 1990). not solve immediate traffic-related air pollution Other traffic management measures are in- problems. tended to establish a smooth traffic flow and In Curitiba success in maintaining acceptable increase road safety by minimizing conflicting air quality levels can be attributed to a master movements between vehicles and between ve- plan that effectively integrated land use, road, hicles and pedestrians. Some traffic manage- and transport policies. This plan, prepared in ment measures include installation of signals at 1965, created a structure for urban growth based intersections, re-routing of traffic, prohibition on linear-guided land development with limited of conflicting turns, designation of one-way physical expansion of the central area. Land use streets, and segregation of motorized and non legislation encouraged urban growth along five motorized traffic. These measures are much less main axes served by express buses. Prior to the costly than building additional infrastructure. construction of these roads, the municipal gov- Traffic flow can be improved through such modi- emient acquired land along or near these roads fications to existing infrastructure as widening and built high-density housing for about 17,000 of roads to provide turning lanes, extra lanes low-income families. Development permits lim- for express bus or high-occupancy vehicles, con- ited the ratio of total floor area to plot size based striction of short road links at critical locations; on the distance from roads served by public buses. and building footbridges, flyovers, or tunnels for For example, the maximum ratio was set at six pedestrians. Traffic flow also can be improved for developments along the five major axes and substantially by changing the direction of traffic four for developments close to roads served by on certain roads during rush hours. This ap- other urban buses. The city center was relieved proach has been widely adopted in the United from commercial pressures and became friend- States and is common in several Latin American lier to pedestrians. In addition, the city's open and Caribbean countries. A more sophisticated leisure space increased by two orders of magni- traffic management measure includes traffic tude in twenty-three years (from 0.5 square control systems that use vehicle detectors and meterperinhabitantin 1970 to 50square meters link traffic signals to computers to optimize traf- per inhabitant in 1993; Rabinovitch 1993). fic flow. Although such systems are expensive, Many other urban centers in Latin America they are effective in enhancing traffic flows and the Caribbean, however, suffer from trans- (World Bank 1986). The traffic management port and environmental problems caused by measures chosen for a given urban area should growth without proper consideration for urban reflect the physical layout of the urban area, the planning. For example, in Asunci6n most streets density of street space, and the characteristics constructed in the growing parts of the metro- of the urban transport system. politan area dutring the past twenty years are not connected to the city center. As a result the few avenues that extend the suburbs to the city cen- Land Use Planning and Controls ter are congested with heavy traffic caused by trucks, buses, and cars. This transport problem Land use planning establishes patterns for land is expected to be addressed as part of the new development in urban centers. Because it speci- urban development plan for Asunci6n. The plan Abatement Measuresfor VehicularAirAPollution: Transport Management Measures 107 also calls for the closure of the city's historic area businesses, and communities exposed to air and to motor vehicle traffic. noise pollution from the traffic. Road pricing Preparation and implementation of an urban involves charging a fee to motorists for the use plan for a metropolitan area requires coordina- of a road in an effort to encourage the use of tion among various administrative entities rep- public transport, high-occupancy vehicles, or resenting differentjurisdictions within the area. congestion-free alternate routes, or to induce In Latin American and Caribbean urban cen- driving at off-peak hours. If motorists are ters such coordination tends to be rather weak. charged a fee for the delay they cause by using a In addition, the regional trend toward decen- particular road, they will use the road only if the tralization of administrative and institutional benefits exceed the toll. The main advantage of responsibilities will compound this problem road pricing is that it encourages vehicle own- unless appropriate actions are taken. For ex- ers to find ways to reduce congestion. This has a ample, one factor contributing to the transport direct impact on ambient air quality because less problem in Asunci6n has been the lack of coor- congestion means reduced pollutant emissions dination among the Ministry of Public Works and from vehicles. the twenty-four municipalities in the Asunci6n The setting of road prices is a technically, ad- metropolitan area. In Venezuela, following pro- ministratively, and politically challenging task. mulgation of the 1989 law that transferred cen- Technically, it is difficult to estimate accurately tral responsibilities to local authorities, each of the social and other costs caused by congestion the five municipalities forming the city of Caracas and associated air pollution. In addition, pre- prepared a different urban policy, urban devel- dicting the effects of road pricing on congestion, opment plan, and urban transport plan accord- traffic patterns, and public transport or high- ing to its respective priorities. However, the occupancy vehicle use may be complex. central government is still responsible for cer- In considering this policy option, the availabil- tain transport-related activities. For example, the ity of non congested alternative routes and the Ministry of Transportation and Communications efficiency of the existing public transport system is responsible for motor vehicle administration must be carefully reviewed. In addition, existing and traffic enforcement (such as issuance of infrastructure must permit isolation of a toll road fines), and the Ministry of Finance is respon- without impeding traffic flow on other roads and sible for establishing parking tariffs.'25 In Novem- installation of entry points to the toll road. Toll ber 1996 the five municipalities have finally booth systems suffer from being land-intensive agreed to form a metropolitan transport authlor- because they require large plots of open space ity for the Caracas metropolitan area to bring for toll plazas and are labor-intensive because about an integrated solution to the city's trans- they require hiring toll operators. The design port management problems. of toll roads also requires an enforcement sys- For many urban centers in developing coun- tem-such as a police force or electronic de- tries, land use controls have not been able to vices-that identifies drivers who evade road curtail traffic densities and vehicular air polul- charges. More important, collection of fees at tion. Implementation of urban plans has suf- toll booths may cause additional congestion in fered from disregard for regulations, inadequate urban areas because it is time-intensive. This enforcement, and public opposition. congestion can be reduced through prepayment methods (such as monthly fees) and appropri- ate identification of vehicles. Under area licens- Road Pricing ing schemes vehicles entering an urban center during peak hours display a daily or monthly li- Congestion in urban area roads imposes costs cense that traffic authorities can check without not only on road users but also on pedestrians, stopping traffic. Electronic road pricing with automatic vehicle identification requires a tran- sponder (known as a "tag," which may be based 25. The Ministry ofTransportation and Communications on optical and infrared systems, induced loop issues licenses to vehicles entering tthe traffic but does not systems, radio frequency and microwave systems, check their emissions. Because the Ministry of Finanice es- tablished the same parking rate for the entire city of Caratcas, or smart card systems) that stores a unique iden- vehicles are not discouraged from enterinig congested sec- tification code for each vehicle, an interrogator tions of the city. that reads the transponder and decodes its iden- 108 Chapter 3 tification, and a computer system that transmits, electric enforcement system) and five-year main- analyzes, and stores the data. Electronic road use tenance of the system cost $44.2 million and charging is superior to manual approaches from $11.6 million, respectively (Hau 1992). In Hong the perspective of road users, road auithorities, Kong the electronic road pricing system, tested and society as a whole (Hau 1992). on 2,500 vehicles, allowed for more accurate Consideration must also be given to the tim- road use charges and eliminated toll booth-in- ing and duration of road charges. Short charg- duced congestion by mounting electronic plates ing periods are likely to cause congestions just underneath vehicles, recordingvehicles atcharg- before or after the restricted period, while long ing points equipped with electronic loops, and charging periods may lead to under-uitilization billing monthly road charges. of road capacity. In Singapore, for example, en- Preliminary experience with road pricing on try to the central business district was subject to a Los Angeles highway has proven to be effec- a fee that originally was charged between 7:30 tive. Before road pricing was introduced, an A.M. and 9:30 A.M. for low-occupancy vehicles eight-lane highway (four lanes in each direction) (fewer than four people per vehicle). Because was experiencing severe congestion, especially this measure resulted in major congestion right during peak hours. Through a franchise agree- after this restricted period, the restricted period ment between the state of California and a pri- was extended to 10:15A.M. The extension proved vate contractor, four additional lanes (two in effective because it eliminated the congestion each direction) were constructed in the middle (World Bank 1986). of the original highway. An electronic road pric- In Latin America, road pricing is used in ing system using radio technology was selected Buenos Aires (see Chapter 4). Examples of road to identifyvehicles on the newly constructed 16- pricing strategies used in other parts of the world kilometer highway. This technology uses over- are shown in Table 3.20. Singapore has been the head antennas and a small, windshield-mounted pioneer in road pricing. In 1995 its labor-inten- transponder to collect tolls electronically with- sive area licensing scheme (with toll booths at out stopping at toll booths. twenty-six entry points) was converted to an elec- The system's sophisticated electronics can tronic road pricing system. The bids received for handle up to 2,500 vehicles per hour per lane, installation ranged from $22.7 million for a and can recognize vehicles at speeds well in ex- simple automatic vehicle identification system cess of legal speed limits. The units are more to $90.6 million for a sophisticated "smart card" than 99.9 percent accurate. The private contrac- system with numerous options. The authorities tor can set road prices based on traffic conges- adopted the smart card system as the basis for tion with the provision that it must return to the their electronic road pricing system. This system state half the proceedings above an agreed rate is based on a two-way communications link be- of return. The current pricing system, estab- tween an on-board unit (consisting of a smart lished for vehicles occupied by only one or two card, smart card reader, and transponder) and passengers, has prices ranging from $0.25 for a roadside antenna. Installation (including the low-traffic periods to $2.50 for peak traffic hours. Table 3.20 Examples of congestion pricing, 1994 731ye of congeslion /prcing Location Year introduced Flat charge to enter the central or (lownitown Singapore 1975 area of a metropolitan area during rtsh hours Hong Kong Tested in 1983-85 anid abandoned Flat charge to enter the central or (lownitown Bergen 1986 area of a metropolitan area Oslo 1990 Trondheim (Norway) 1991 Stockholm 1997 Rush hour toll surcharge on an intercity Paris 1992 expressway Source: Gomez-Ibafiez and Small 1994. Abatement Measuresfor VehicnlarAirPollution: Transport Managemenf Measures 109 No charges are levied from high-occupancy ve- Provision of Public Transport Services hicles.26 Another feature of this system is that road maintenance and operation of the clhar-ge Public transport systems aim to efficiently move system, includingenforcement, are the responsi- large numbers of people using buses, hility of the franchised company. Fines for driv- microbuses, vans, trolleybuses, tramways, metros, ers abusing the payment system start at $100 and or trains. The provision of public transport can go up to $500. Since this system was imple- affect air quality in urban centers. Since conges- mented in December 1995, traffic has been flow- tion-induced air pollution in urban centers is ing very smoothly at high speeds in the toll mostly attributed to the growing number of pri- highway. The older portion of the highway is still vate cars, promotion of public transport can be congested (Warriner 1996). considered an air pollution control measure. Although public transport vehicles also emit air pollutants, their contribution to pollution per Area Licensing person transported is much less than that of pri- vate cars. Area licensing is a charge that can be applied Among public transport modes, rail transport for the use of roads in designated sections of an is more attractive than road-based transport be- area during specified times of the day or days of cause it operates over protected rights of way, the week. The area licensing scheme introduced carries more passengers per trip, and provides a in Singapore in 1975 is a good example of this faster and more reliable service in highly charge. This scheme originally targeted passen- congested areas and over longer distances. In ger cars but was subsequently modified to in- addition, electrified rail transport is more envi- clude all vehicles except ambulances, fire ronmentally attractive, especially in cases where engines, police vehicles, and public buses. It was the primary fuel source for electricity genera- used as part of a larger package of measures for tion is natural gas. However, rail-based transport improving public transport and the environ- requires a higher investment and cannot serve ment. Under this scheme vehicle owners were as many locations as road-based transport ser- required to buy special licenses to enter certain vices. Among the rail-based public transport sys- restricted zones during morning rush hours. tems, metros offer the highest capacity at the Other measures included in the policy package highest speed, but require very high construc- included free passage into the restricted zone tion costs and sophisticated technology. Metros for car pools, increased parking charges within are mostly used for short trips within the central the restricted zone, strict enforcement at the area of congested urban centers (for example, twenty-eight entry points into the restricted zone, Buenos Aires) and by those who live or work near and progressively more onerous taxes on the the metro stations. import, purchase, and registration of cars Buses are the least expensive and most flex- (Bernstein 1991). ible way of meeting a range of demands in ur- Since 1991 Singapore has used a "weekend ban areas. The attractiveness of buses in car" scheme. Under this scheme weekend car congested areas can be increased through bus drivers enjoy a 70 percent reduction in road taxes priority measures and through provision of ex- and a tax rebate of up to $9,600. Weekend cars, press buses operating over long distances in ar- identified by their red license plates, are allowed eas not served by rail. Express buses can be to circulate only between 7 P.M. and 7 A.M. on enhanced by "park and ride" facilities at the sub- weekdays, after 3 P.M. on Saturdays, and all day urban end of the route and a pleasant walking on Sundays and public holidays. These vehicles environment at the city center. In developing can also be driven outside these periods for up countries smaller buses (such as minibuses) are to five days a year by displaying a special day li- both cost-effective and popular with users. These cense. Each additional day license costs $13 vehicles, which cost less per seat than large buses, (Carbajo 1994). can operate at higher speeds than large buses on narrow and congested streets and are finan- cially viable in low-density areas, providing fre- 26. Vehicles with three or more occupants are given a quent service despite low demand. transponder with a special code and asked to utse certainl In Latin America and the Caribbean most ur- lanes. ban public transport services are provided by 110 Chapter 3 buses, minibuses, and vans. In some South port ridership. Fare elasticities have not been American cities (Buenos Aires, Montevideo, Rio determined for Latin American urban centers, de Janeiro, Santafe de BogotA, and Santiago) but in European cities fare elasticities of -0.3 trolleybuses used to be operated but were sub- are common (ECMT 1990).2" During non-rush sequently taken out of service. Only Mexico City, hours fare elasticities are higher because of the Recife, Sao Paulo, and Valparaiso are now served availability of other transport modes and less con- by trolleybuses. Metro has been used in Buenos gestion. Aires since 1913, in Mexico City since 1969, in Simplification of the fare structure through Sao Paulo since 1974, in Santiago since 1975, in flat or zonal fares along with the use of prepaid Rio dejaneiro since 1979, and in Caracas since passes or integrated fares between different 1982. In other cities (Belo Horizonte, Porto transport modes can increase the perceived Alegre) electrified urban railways-known as value of public transport and hence its ridership. "surface metros"-are dedicated to passenger Such a scheme, introduced first in Stockholm transport. Light rail transit is used in a few cities in the early 1970s, has been used in many large including Campinas, Btienos Aires, and Mexico European cities. In London, mostly as a result City. In Buenos Aires and Mexico City these sys- of restructuring and integration of fares between tems are connected to the metro. In addition, the metro and bus system in 1983, ridership for some South American cities (Buenos Aires, Rio these modes increased by 30 percent within five dejaneiro, Sao Paulo) are served by a suburban years while car commuting dropped by 17 per- rail system (UN 1994). cent (ECMT 1990). Some fare integration has To ease congestion and reduce the contribu- been established for the bus-metro and bus-rail tion of buses to air pollution, bus capacities can systems in Sao Paulo and for the rail-metro sys- be increased. For example, in Porto Alegre buses tems in Rio de Janeiro. However, lack of fare are fit with an extension unit during rush hours integration between different modes of trans- (UN 1994). In Curitiba large-capaciqt buses (for port-a result of weak transport regulatory in- 270 persons) have been used to provide express stitutions and the desire of private bus companies service. for a large share of the fares-is common in To curtail air pollution in urban areas, use of many other urban centers in the region, includ- public transport in lieu of private cars can be ing Buenos Aires and Santiago. promoted through such incentives as improve- The private sector can play an important role ments in the quality of public transport services, in promoting public transport. For many decades reduction of fares, and simplification of fare public transport services in the urban centers of structures. These incentives are intended to re- developing countries, including those in Latin duce riders' overall cost of using the public trans- America, were provided by large government port system.2" entities under the pretext of "protection from For users one of the most important attributes unfair competition." These public entities suf- of a public transport system is the extent and fered from cost-ineffectiveness, competition for quality of its service. Because public transport is revenues, and inflexibility in coping with chang- less flexible in timing and routing than private ing conditions (for example, the ability to hire transport, its appeal is limited. Public transport's and lay off staff). Subsidies, financed by taxes appeal can be increased by providing a dense levied from all citizens-public transport riders network of frequent and reliable service with and others-were justified on the grounds of convenient interchanges. Service elasticities have providing the public with satisfactory service at not been determined for Latin American urban affordable fares. However, these subsidies did not centers. In European cities, however, service elas- produce the expected results because they did ticities of 0.4 are common (ECMT l990) .2 not affect the transport patterns of car drivers Lower fares also would encourage public trans- who placed a high value on the comfort and convenience of driving their private cars. Fur- thermore, only about half of these subsidies were 27. The overall cost is the sum of the riders 'cost of time for accessing, waiting for, anid riding ptiblic transport, in addition to the cost of the fare. 28. In these cities a 10 percent improvemenit in public 29. Assuiming a fare elasticity of -0.3 and modal split of transport services woild increase piblic transport ri(ership 50:50 between private and public transport, a 10 percent by 2 percent assumting a 50:50 split betweeni the private and( fare redtictiots wouldl increase ptiblic transport ridership public transport. by 1.5 percenit. Abatement Measures for VehirndlarAirPollution: Transport Management Measures 111 actually reflected in lower FEares, and in many routes were auctioned. The selection criteria instances they resulted in larger staffs, wages, and favored bigger and newer buses as well as larger unit costs. Since large government-owned pub- transport companies. Through this process, bus lic transport companies provided low-quality ser- service to users was considerably improved. vices, people willing to pay higher prices for In Montevideo the public transport system better services were forced to use private cars or included tramways until the 1950s. These were taxis. This pattern increased congestion and ad- replaced by trolleybuses until the late 1970s, and versely affected bus services, used mostly by low- then by buses. Until 1975 the municipality pro- and middle-income people. vided about 40 percent of public transport ser- Ample evidence exists of the efficiency of vices, which were passed over to the cooperatives transport services provided by private companies, formed by public employees. The municipality which operate with a profit motive and staff ac- still regulated bus routes, tariffs, and schedules, countability. For example, World Bank studies however. In 1990, because most buses were old indicate that in Bangkok, Calcutta, and Istanbul and the transport system was not responding to the costs of private bus companies are 50 to 60 the public's needs, new measures were imple- percent of those of publicly owned and oper- mented. These included restructuring previous ated bus companies. These studies also found bus routes, creating new bus routes, and direct- no evidence that private bus companies are less ing the express bus system (diferenciales) to sub- safe or poorer service providers than public com- urbs away from the city center, adjusting tariffs panies, or are only willing to operate on the most based on cost studies, renovating the bus fleet, profitable routes (World Bank 1986). constructing bus terminals, and installing signals In Latin America and the Caribbean private at bus stops. Currently, mass transport services bus companies operate in many urban centers, are provided by 1,460 buses owned by five pri- including Buenos Aires, Guatemala City, vate companies (three cooperatives and two in- Kingston, Porto Alegre, San Jose, Santafe de corporated companies). These buses account for Bogota, Santiago, and Sao Paulo. After Santiago's 63 percent of the daily trips in Montevideo. Only public bus companywas dissolved, private trans- 20 percent of trips are made by private cars, 12 port companies started providing a variety of percent by taxis, and 4 percent by other modes. public transport services with thirty-five seat Bus subsidies are equivalent to 8 percent of bus buses, fifteen seat minibuses, and shared-ride company billings to cover rides by students and taxis. Between 1978 and 1984 the number of retired people. public transport vehicles increased by 50 percent Curitiba is Latin America's best example of for buses, nearly 100 percent for minibuses, and an urban center with an efficient public trans- 300 percent for shared-ride and regular taxis port system. During the 1960s Curitibawas slowly (Meyer and Gomez-Ibafiez 1991). moving toward a car-dominated city with increas- In Asunci6n about 70 to 80 percent of daily ing congestion, while its bus system was inad- trips are made by public transport. The public equate to curb the use of private cars. Following transport system, which is totally private, consists the 1965 master plan, express bus services were of 2,400 buses operating on 165 routes under set established in the 1970s on restricted lanes along tariffs. The main transport management prob- each of the city's five axes. Express buses were lem appears to be the lack of planning of routes, complemented by interdistrict and feeder buses which is caused by the lack of coordination operating on other routes. Each type of bus was among various transport authorities (the Minis- color-coded. The public transport system was try of Public Works and the twenty-four munici- fully integrated through large bus terminals at palities in the Asunci6n metropolitan area). the end of each of the five express busways for Public transport in Lima used to be provided people to transfer to interdistrict, feeder, or by 4,000 old and poorly maintained buses un- intermunicipal buses. Smaller bus terminals were der a controlled tariff system. Following an built every two kilometers along each express abrupt policy change, implemented through leg- route. These terminals were equipped with news- islative decree, import of buses was liberalized paper stands, public telephones, post offices, and and bus routes and tariffs were set free. As a re- small commercial facilities. A single fare was es- sult the number of registered buses, most of tablished for all buses running within Curitiba. which were small (12 to 24 passenger capacity), A prepayment scheme eliminated the need for increased drastically, to 43,000. Because these the bus crew to collect fares and freed up space microbuses increased traffic congestion, eleven for more passengers. Special tubular station plat- 112 Clhapter 3 forms, raised to the same height as the express quality of walking environment, making city cen- bus floor, reduced boarding and deboarding ters more friendly to pedestrians, and closing time by 300 percent. A bus-activated electronic certain streets to motor vehicle traffic. Use of traffic light system was established as a bus pri- bicycles can be enhanced by constructing ority measure. In the downtown area priority was bikeways, promoting domestic bicycle manufac- given to pedestrians over private cars. In 1992, turing, and reducing import taxes on bicycles. 270-passenger express buses (with five lateral In addition, nonmotorized traffic can be im- doors that facilitated passenger entry from and proved through provision of additional security. exit to newly accommodated tubular platforms) Lack of security threatens nonmotorized traffic were introduced to replace the lower-capacity in certain sections of Latin American cities. For (110 passengers) buses. The average life of buses example, although bike paths are available in has been maintained at three years, nearly one- Lima, they are not used extensively for fear of third the average for other Brazilian cities. The personal attack (World Bank 1996b). Safe walk- older buses were recycled to serve as classrooms ing or cycling is also a concern in certain sec- for vocational training (for example, carpentry, tions of Rio deJaneiro. hairdressing, word processing). As a result of Curitiba has done well at promoting non- these measures public transport ridership in motorized transport. Its road network is designed Curitiba increased dramatically: since 28 percent in a way that allows pedestrians and bikers to of express bus riders are previous private car share the roads efficiently with cars. By the end drivers, the city's fuel consumption has been cut of 1992 the city had 150 kilometers of bike paths by 25 percent. Curitiba's public transport system integrated with the public transport network and now serves 1.3 million passengers a day and at- with green areas. In the city center pedestrians tracts two-thirds of the city's population. Al- are given higher priority than private cars though Curitiba has the highest per capita car (Rabinovitch 1993). In Santaf6 de Bogota and ownership in Brazil, its air pollution is one of Rio dejaneiro certain avenues are closed to traf- the lowest among Brazilian urban centers fic for bikers and pedestrians on Sundays and (Rabinovitch 1993). holidays. In Denmark there are 4.2 million bi- Curitiba's integrated transport network is cycles among 5.2 million people, and in managed by a mixed-capital company created Copenhagen about one-third of commuters ride by the municipal government. This company is to work on bicycles. responsible for developing bus timetables and frequencies, implementing new bus routes, cal- culating the necessary number of buses, moni- Provision of Off-Street Parking toring the network's performance, training drivers and conductors, responding to hus rid- In areas where on-street parking is restricted or ers' suggestions and complaints, and managing banned, off-street parking can be used to provide interstate and municipal bus terminals, public parking in congested urban areas, especially parking systems, and paving programs. Fares, where business activities are predominant. Con- which are some of the cheapest in Brazil. are set struction of off-street parking facilities should only by this company and the municipal government be permitted at suitable locations through land based on monitoring of the number of bus pas- use regulations, and operation of these facilities- sengers. The buses are operated by private com- including setting of parking fees-should be left panies through a permit system on specific to the private sector. Taxes on parking may pro- routes. Since 1987 the city has paid tlhe bus com- vide the public sector with additional control over panies based on the number of kilometers of traffic flows. Raising the cost or restricting the transport provided (Rabinovitch 1993). supply of off-road parking while enforcing on-road parking constraints may be an effective way to reduce traffic flows, but the combined use of these Promotion of Nonmotorized Transport measures would likely meet with opposition from local businesses. Restriction of off-street parking Promotion of nonmotorized transport can re- has been used effectively in Seoul (World Bank duce the use of motor vehicles and associated 1996b). In downtown Singapore parking fees have air pollutant emissions. Walking can be pro- been significantly increased to discourage cars moted by providing sidewalks, improving the from entering the city. ANNEX A VEHICULAR EMISSION STANDARDS IN THE UNITED STATES New Vehicles the United States. Separate emission standards for new light-duty Emission standards for new cars in the United trucks were first established by the U.S. Environ- States were first established by the 1968 Clean mental Protection Agency (USEPA) for 1975 Air Act, which was amended in 1970. These stan- model-year (Table A.2).' These standards, which dards limited CO and HC emissions from vehicle limited CO, HC, and NO, emissions, were tight- exhaust gases (Table A.1). In 1971 evaporative ened starting with the 1979 model-year. PM was emissions were controlled and in 1973 NO. was added to the list of regulated pollutants for 1982 added to the list of regulated pollutants. The model-year diesel-fueled vehicles. More stringent CO, HC, and NO, standards were tightened start- CO and HC standards for 1984 model-year ve- ing with 1975 model-year cars. Compliance with hicles required use of two-way catalytic convert- the 1975 standards required car manufacturers ers. The more stringent NO5 standard for 1988 to install oxidation catalysts and refiners to pro- model-year vehicles required three-way catalytic dtuce unleaded gasoline to service these cars. In converters. Separate emission standards were es- 1977 the Clean Air Act was amended and re- tablished for two different weight classes of light- quired a 90 percent reduction in HC in 1980 duty trucks starting with 1991 model-year. The and a 90 percent CO and 75 percent NO, re- 1990 amendments to the CleanAirActspecified duction in 1981. This led to the automotive in- more stringent emission standards and certifica- dustry to introduce three-way catalytic converter tion requirements starting with 1994 model-year technology. Along with these standards, the au- light-duty trucks. thorities established a certification requirement U.S. emission standards for CO and HC ex- for manufacturers to build vehicles in compli- haust gases from new heavy-duty vehicles2 were ance with these standards for 80,000 kilometers established in 1970 (Table A.3). In addition, opac- or five years (whichever comes first), a manu- ity' standards for diesel-fueled heavy-duty ve- facturer recall program for failing vehicle corin- hicles were set at 40 percent during acceleration ponents, and a warranty program providing and 20 percent during lugging. In 1974 exhaust consumers with an effective recourse for non- emission standards were revised to substitute the complying vehicles. Emission standards were sum of HC and NO, emissions for HC emissions. tightened considerably in later years. In addi- tion, the 1990 amendments to the 1968 Clean Air Act extended the certification requirement 1. Lighit-duty tncks are trucks with gross velhicle weight for 1996 and newer model-year vehicles to below 2,722 kilograms. This weight limit was extented to 160,000 kilometers or ten years (whichever 3,856 kilograms starting with 1979 model-year vehicles. comes first). The state of California, becauise of 2.Haydyvhilsretoewhgosvhce weight of 2,722 kilograms or more for model-years before the poor air quality in the LosAngeles area, gen- 1979 and 3,856 kilograms for model-years 1979 and after. erally has adopted more stringent standards than 3. Opacity measures the percentage of light blocked by the federal standards, which apply to the rest of smoke enmissions. 113 114 Chapto,3 Table A.1 U.S. emission standards for new passenger cars (grams per kilometer) AJodel-year (:0 H(. NO, Pf Evaf). (grams/lesi) Pre-control 56 9.3 3.9 6.0 1970a 21 2.6 1971a 21 2.6 60h 1972a 17 1.9 2.0 1973-743 17 1.9 1.9 2.0 1975-76 9.3 0.9 1.9 2.0 1977 9.3 0.9 1.2 2.0 1978-79 9.3 0.9 1.2 6.0' 1980 4.4 0.25 1.2 6.0 1981 2.1 0.25 0.62 2.0 1982-86(i 2.1 0.25 0.62 0.37 2.0 1987-93 2.1 0.25 0.62 0.12 2.0 1994 -95e 2.1 0.16g 0.25 0.05 2.0 1996-2003 2.6k O.199 0.37" 0.06 2.0 2004 and newerd 1.1 0.0789 0.12 0.05 2.0 Note: A blank space indicates that no stan(lard was established. National emission standards for cars were set in the 1968 Clean Air Act, which was amended in 1970. a. Pre 1975 standards are expressed as equivalent 1975 test vaitles. b. Using the carbon canister trap method. c. Using the sealed housing evaporative determiiiation (SHED) method. 6.0 grams/test by SHED method represents approximately 70 percent less emissions thani 2.0 grams/test by the carbon trap method. d. High altitude standards for 1982-83 modlel-years are 4.8 g/km for CO, 0.35 g/km for HC, 0.62 g/km for NO,, and 2.6 grams/test for evaporative emissions. Starting with 1984 model-year, all cars must meet standards at high altittude. e. Standards apply to 40 percent of produiction for 1994 model-year and 80 percent of prodtuction for 1995 model-year. f. Additionial CO limit of 6.2 g/km at -7"C. g. Nionmethane hydrocarbons. h. Diesel-ftieled passenger cars lhave separate NO, limits. The NO, limit is 0.62 g/km for 5 years or 80,000 kilometers, and 0.78 g/km for 10 years or 160,000 kilometers. i. Revised test procedures with additional test proceduires are established. j. Implementation of these standards are at tJSEPA's discretion. .So1erce: CONGAWE 1994. In addition, more stringent opacity standards for were established. The 1987 revisions of CO and heavy-duty diesel-fueled vehicles were adopted HC limits required use of two-way catalytic con- (20 percent duringacceleration, 15 percent dur- verters. Introduction of a PM limit and a more ing lugging,4 and 50 percent at maximum stringent NO, limit for gasoline-fueled heavy- power). These opacity standards are still in ef- duty vehicles in 1990 required use of three-way fect. In 1979 exhaust emission standards were catalytic converters in these vehicles. The certi- again revised and the testing procedure for HC fication requirement for manufacturers to com- emissions was changed for gasoline-fueled ve- ply with emission standards of 1991 and thereafter hicles, resulting in higher readings for equiva- was set at 193,000 kilometers. lent emissions. In 1984 a standard for NO, Effective April 1993, the USEPA has specified emissions and a transient emissions testing pro- evaporative emission standards with new test cedure were introduced. In 1985 separate limits procedures. These standards, phased in over the for diesel-fueled vehicles were implemented and 1996-99 model-years, apply to both light-duty and the standard for the sum of HC and NO, was elimi- heavy-duty vehicles (Table A.4). nated. In addition, standards for evaporative emissions for gasoline-fueled heavy-duty vehicles In-Use Vehicles 4. Lugging simLulates the uip hill mziovemiienit of a fiully Emission standards for in-use vehicles in the loaded heavy-duty vehicle. United States are established at the federal and Abatemolent Measuresfor Vehicular AirPolluttion: Annex A 115 Table A.2 U.S. emission standards for new light-duty trucks (grams per kilometer) A11ode-year (() HC NOr PA a 1970-74 Same as passenger cars 1975-78 12.4 1.24 1.93 1979-81 11.2 1.06 1.43 1982 11.2 1.06 1.43 0.37 1983 11.2 1.06 1.43 1984-86 6.2 0.50 1.43 0.37 1987 6.2 0.50 1.43 0.16 1988-90 Less than 2,722 kg 6.2 0.50 0.75 0.16 2,722-3,856 kg 6.2 0.50 1.06 0.16 1991-93 Less thani 1,701 kg (6.2 0.50 0.75 0.16 1,702-3,856 kg 6.2 0.50 1.06 0.08 1994-2002 Less than 1,701 kg 5 years or 80,000 km 2.1 0.16c 0.25 0.05 10 years or 160,000 km 2.6 0.19' 0.37 0.06 1,701-2,608 kg 5 years or 80,000 km 2.7 0.20' 0.49 0.05 10 years or 160,000 km 3.4 0.25' 0.60 0.06 2,609-3,856 kg 5 years or 80,000 km 3.1 0.24' 0.68 10 years or 160,000 km 4.5 0.35' 0.95 0.07 Noie: A blank space indicates that no standard was established. a. PM standards apply to diesel-fuieled trmcks only andl are relaxed for vehicles with gross vehicle weight over 1.701 kilograms. Limits are 0.31 g/km for 1987 model year and 0.28 g/km for 1988-90 model-years. b. Standards are phased in over a three-year pertiod starting with 1994 model-year. c. Nonmethane hydrocarbons. d. USEPA must decide by 1997 whether to apply these limits or set different standards. Sorwre: CONCAWE 1994. state levels and vary by state (Table A.5). The gram. The standard for basic inspection and main- 1990 amendments to the Clean Air Act called for tenance program was modeled on the use of a the introduction of enhanced inspection and simple idle test. The USEPA's standard for the maintenance programs in polluted nonattain- enhanced inspection and maintenance program ment areas for ozone with populations over (commonly known as IM240) was based on an- 200,000 people as well as in "ozone transport nual testing of all 1968 and newer model-year regions" with populations over 100,000 people."' passenger cars and light-duty trucks at central- States were required to submit their inspection ized inspection stations. In these areas the scheme and maintenance programs by November 1993 allowed a steady-state test for 1968-85 model-year and promulgate the necessary legislation. Of 181 vehicles but required the IM240 test for 1986 and areas of the United States identified to establish newer model-year vehicles. Under the same act emissions testing programs, 95 moderately pol- more stringent emission standards, known as luted areas were required to implement a basic "Tier 1," were to be phased during 1994-96 and inspection and maintenance program and 82 to be completely implemented for all 1997 and more polluted areas were required to implement newer model-year vehicles (Table A.6). The in- an enhanced inspection and maintenance pro- spection and maintenance procedures also re- quire pressure and purge checks on the carbon canister and a visual inspection of the catalytic con- 5. Nonattainmenit areas are those where riationial ambi- verter, fuel inlet, and evaporative emission con- ent air quality standards are not exceeded. trol systems (CONCAWE 1994). 116 Chapter3 Table A.3 U.S. emission standards for new heavy-duty vehicles (grams per break horsepower-hour) Model-wear ((o HC NO HMC + NO, PM Fvnp. (grnms/test) 1970-73a 63.6 6.55 1974-78 40 16 1979-83 25 1.5" 10 1984 Gasoline-fueled vehicles Transient d 25 1.5 10.7 10 Idle 0.5% Diesel-fueled vehicles 15.5 5 9.0 1985-86 Gasoline-fueled vehicles Option Ae 37.1 1.9 10.6 3.0' Option Be 40.0 2.5 10.7 3.0 Diesel-fueled vehicles 15.5 1.3 10.7 1987-89 Gasoline-fueled vehicles Less than 6,350 kg 14.4 1.1 10.6 3.0 Idleg 0.5% More than 6,350 kg 37.1 1.9 10.6 4.0 Diesel-fueled vehicles 15.5 1.3 10.7 Idleg 0.5% 1990 Gasoline-fueled vehicles Less than 6,350 kg 14.4 1.1 6.0 4.0 Idleg 0.5% More than 6,350 kg 37.1 1.9 6.0 4.0 Diesel-fueled vehicles 15.5 1.3 6.0 0.6 4.0 Idleg 0.5% 1991-92 1 Gasoline-fueled vehicles Less than 6,350 kg 14.4 0.9 5.0 More than 6,350 kg 37.1 1.7 5.0 Diesel-fueled vehicles 15.5 1.3 5.0 0.25 1993"I Buses 15.5 1.3 5.0 0.10 1994-95 h Diesel-fueled vehicles 15.5 1.3 5.0 0.10 Buses 15.5 1.3 5.0 0.07 1996-97" Buses 15.5 1.3 5.0 0.05 1998 and newer I 3,856-12,700kg 15.5 1.3 3.15 0.10 Buses 15.5 1.3 4.0 0.05 Note: A blank space indicates that no standard was established. a. Standards apply to gasoline-fuieled vehiicles only. b. Alternative standards are 25 g/bhp-h for CO and 5 g/bhp-h for the sum of CO + NO,. c. HC measulrement method changed startinig with 1979 model-year gasoline-fuieled vehicles. The new method (FID) results in higher readings for equivalenit emissions measured with the former method (NDIR). d. A new transient test procedure is introduiced. The standard for NO, is interim. e. Different dynamometer schedules are use(d for options A and B. Option A and B limits apply at the beginning and end of the durability test schedules. f. For heavy-duty vehicles with gross vehicle weights of 3,856-6,350 kilograms. For heavier vehicles the limit is 4.0 grains/test. g. For heavy-dtity gasoline-fuieled vehicles usinig the catalyst technology. h. From 1991 on, the NMHC limit of 1.2 g/bhlp-h applies for natuiral gas engines instead of the HC limit. Sourre: CONCAWE 1994; CONCAWE 1995. Abatte ielt Measures for Vehicular Air Pollution: A nnex A 117 Table AA4 U.S. Federal evaporative emission standards for new vehicles Imfdemnen taOon schedl7dl' 7hree-day Sutl)plementart Spitbark Share o!f diurnal two-day Running test prnductiwn (;Vlv l)urobilit'l hot soatk test diurnal test loss (grams Yealr (percent) (kilograms) (kiltmoeters) (grams/test) (grams/test) (grams/km) liquid/test) 1996 20 < 2,727 ' 2.0 2.5 0.031 1.0 1997 20 2,727-3,864 192.000 2.5 3.0 0.031 1.0 1998 90 3,865-6,364 192.000 3.0 3.5 0.031 1.0 1999 100 > 6,364 192,000 4.0 4.5 0.031 NVote: A blank space indicates that no standard was established. Diuriial losses occur when the vehicle is stationary with the engine off. These losses are dne to vapor emissions from the fuiel tank at ambient temperature changes during a 24-hotir period. Evaporative losses are measured after tfisee-day ambient temperattire cycles of 22.2-35.6°C usitig the Sealed Houising for Evaporative Determination (SHED) method which is supplemented by two-day diurnal test with similar temperature cvcles. Runininig losses occur while the vehicle is normally driven. The spitback test simulates fuel losses durinig vehicle refuieling. a. Methanol-fuieled vehicles are required to complv the implemenitation schedule starting with the 1998 model-year. However, manifactuirers selling less than 10,000 vehicles a vear do not have to comply tntil the 1999 model-year. b. GVW is gross vehicle wveiglit. c. Durability for 1996 model-vear vehicles is definedl as follows: light-duty vehicle two years or 38,400 kilometers if tlse evaporative emission control device costs less than $200: eight years or 128,000 kilometers if deemed 'specified major emission components." For light-duty trucks the durability reqtuirements are specified as ten years or 160,000 kilometers for vehicles less than 1,705 kilograms, and 192.000 kilometers for heavier vehicles. Soure: CONCAWE 1994. Table A£5 U.S. emission standards for in-use vehicles CO HC Smole Jtrisdiction/vehirle type (percentage) (ppm) (percent opacity) Federal (USEPA inspection and mairitetiatice program) Passenger cars (198 land iewer models) 1.2 220 State of Arizona Gasoline-fueled passenger cars and light-dutty trucks 1967-71 5.5 500 1972-74 5.0 400 1975-78 2.2 250 1979 2.2 220 1980 and later model-years 1.2 220 Diesel-fueled vehicles 50 State of Florida Gasoline-fueled vehicles Less than 2,722 kilograms 1975-77 5.0 500 1978-79 4.0 400 1980 3.0 300 1981 atid later mo(lel-years 1.2 220 2,722-4,536 kilograms 1975-77 6.5 750 1978-79 5.5 600 1980 4.5 400 1981-84 3.0 300 1985 and later model-years 1.2 220 Diesel-fueled vehicles Cruise mode 20 Idle mode 5 Note: A blank space indicates that no standard was established. CO and HC emission standards for gasoline-fueled vehicles apply both at idle and 2,500 rpm. Smoke teastntments of diesel-fieled vehicles are made at free acceleration (in this test the engine is rapidly accelerated from idle to full speed). .houn-r: Adapted from Faiz, Weavet; andl Walsh 1996. 118 Chapter3 Table A£6 U.S. eniission standards for in-use vehicles based on the IM240 test (grams per kilometer) I'ehirle Iype/mnodel-year (( HC NOQ Light-duty vehicle 198I-1993 12.4 0.5 1.2 1994-1996' 12.4/9.3 0.5/0.4' 1.2/0.9 1997 and later model-vears 9.3 0.4' 0.9 Light-duty truck Less than 2,722 kilograms 1986-1993 12.4 0.7 2.2 1994-1996' 12.4/9.3 0.7/0.4h 2.2/1.2 1997 anic later model-vears 9.3 0.4" 1.2 Light-duty truck More thiani 2,722 kilograms 1986-1993 12.4 0.5 2.2 1994-1996 12.4/9.3 0.5/0.4 2.2/1.6 1997 and later model-years 9.3 O.4P 1.6 a. During the 1994-96 phase-in period, a percentage of vehicles designated by each state is required to comply with the standards for the 1986-93 model-year vehicles and tihe remaining vehicles are required to comply with the standards for the 1997 and riewer model-year vehicles. b. Non-methane hydrocarbonis. Sourre: Adapted from Faiz, WVeaver, and Walsh 1996. Abatement Measures for Vehicular Air Pollution 119 References Branco. Gabriel Murgel, and Alfred Szwarc. 1993. "Flexible Fuel or Flexible Fuel Vehicle?" Paper pre- AAMA (American Automobile Manifactuirers Asso- sented at the Tenth International Symposium on Alco- ciation). 1995. World Motor Vi'hirle Data.. Washing- hol Fzuels, Colorado Springs, Colorado. Published tori, D.C. by CETESB, Sao Paulo, Brazil. AQIRP (Air Quality Improvemenit Research Pro- Carbajo,Jose 1994. "The Role of Market-based Incen- gram). 1990. "Initial Mass Exhaust Emissions fromii tives." Transportation, Water, and Urban Develop- Reformulated Gasolines." Technical Bulletin 1. ment Department. World Bank, Washiington, D.C. Coordinating Research Council Inc.,Atlanita, Geor- CCEEB (California Council for Environmental and gia. Economic Balance). 1990. Alternative Motor Vehicle . 1991a. "Emissions Results of Oxygenated Fuels to Improve Air Quality: Options and Implications Gasolines and Changes in RVP." Technical Bulle- for California. San Francisco, California. tin 6. Coordinated Research Council Inc., Atlanta, CMPCCA (Comisi6n Metropolitana Para la Georgia. Prevenci6n y de la Contaminaci6n Ambiental en - . 1991b. "Exhaust Emissions ofToxicsAir Pol- el Valle de Mexico). 1995. 'La Contaminaci6n lutants Using Reformulated Gasolines." Technical Atmosferica en el Valle de Mexico. Acciones Para Bulletin 5. Coordinatinig Research Council Inc., su Control, 1988-1994." Mexico City. Atlanta, Georgia. CONCAWE. 1989. "Economic Consequences of Lim- Alconsult Interriationial Ltd. 1996. "Database for the iting Benzene/Aromatics in Gasoline." The Hague. Study Concerning the Elinlination of Lead in Gaso- . 1994. Motor Vehicle Emission Regulations and line in Latin America and the Caribbean." Calgary, Fuel Specifications. 1994 Updates. Report No. 4/94. Alberta, Canada. Brussels. Alsoti,Jeffrey A.,J.M. Adler, arid TM. Bainies. 1989. DDF (Departamento del Distrito Federal). 1996. "Motor Vehicle Emission Characteristics andl Air "Programa par Mejorar la Calidad del Aire en el Quality Impacts of Methanol and Compressed Valle de Mexico 1995-2000." Mexico City. Natural Gas." In Daniel Sperling, ed., QtioroLIm Ecopetrol (Empresa Colombiana de Petr6leos). 1996. Books. Alternative Transportation Fuels: An Environ- "Programa de Mejoramiento de Caladid de las mental and Energy Solution, New York. Gasolinas Colombianas." Paper presented at the Bartone, Carl J., Jaiis Berstein, Josef Leitmann, andl World Bank seminar Elimination of Lead in Gasoline Jochen Eigen. 1994. "Toward Environmenital Strat- in Latin America and the Caribbean, September 4-6, egies for Cities: Policy Considerations for Urban Santiago. Environimental Management in Developing Couni- ECMT 1990. (European Conference of Ministers of tries." Urban Management Program Discttssioin Transport) Transport Policy and the Environment. Paper 13. World Bank, Washington, D.C. Published jointly with the OECD. Paris. Beaton, S.P., G.A. Bishop, and D.H. Stedman. 1992. The Economist. 1996. "Dirty Diesels. Oil Change on "Emission Characteristics of Mexico CityVehicles." the Buses." London. Joutrnal of Air and Waste Management Association 42: Faiz, Asif, Christopher S. Weaver, and Michael Walsh. 1424-29. 1996. Air Pollution from Motor Vehicles: Standards and Berstein,Janis 1991. "AlternativeApproachestoPol- TechnologiesforConttollingEmissions.WorldBankand lution Control and Waste Managemenit: Regulatory United Nations Environment Programme, Wash- and Economic Instrumenits." Infrastructtire and ingtoni, D.C. Urban Development Department Discussion Paper Faiz, Asif, K. Sinha, M. Walsh, and A. Varma. 1990. INU79. WAorld Bank, Washingtoni, D.C. "Autotmotive Air Pollution: Issues and Options for Bloom, L. Richard. 1997. Personal Conimunicationis. Developing Countries." Policy Research WAorking Automotive Engineered Systems Division, 3M, St. Paper 492. World Bank, Washington D.C. Paul, Minnesota. GEMS/WHO. 1988. Assessment of Uthan Air Quality. Braddock,J.N. 1988. "Factors Influencing the Coin- United Nations Envirotiment Programme and position and Quantity of Passenger Car Refueling World Health Organization, London. Emissions - Part II." SAE Paper 880712. SAE Inter- Gomez-lbaiiez,J.A., and K.A. Sniall. 1994. Road Pric- natiorial, Warrendale, Pennsylvaniia. ing for Congestion Management: A Sutrvey of Interna- Berumen, Jesuis. 1996. "Evoluci6n Historica de las tional Practice. National Cooperative Highway Gasolitias en Mexico." Paper presented at the Research Program. National Academy Press, Wash- World Bank seminarElimination of Lead in Gasoline ington, D.C. in Latin America and the Caribbean, September 4-6, Gwilliam, Ken. 1995. "The Energy and Environmental Santiago. Impacts of Urban Traffic Congestion: Transport 120 Chapter 3 Policies and Regulations." Paper presented at the OECD/IEA (Organization for Economic Coopera- W'orld Bank Confewenceon Transport, Ereigr, andEn- tion and Development/International Energy vironment, April 24-25, Washingtoni, D.C. Agency). 1993. Cars and Climate CGange.Energy and Hamnrin, R.D. 1990. "Policy Options for Comparative Environment Series. Paris. Air Pollution Study in Urban Areas. "EnviroTnment Oil and GasJournal. 1995. "Unocal Extends Califor- Working Paper 28. World Bank, Enviro>inmeTnt De- nia Clean Air Campaign." 93(2):22-23. partmnent, W'ashirigtoni, D.C. O'Ryan, R. 1996. "TransportAir QualityManiagemenit Hass-Klau, C. 1993. "Impact of Pedestrianizatioti and Project for the Mexico City Metropolitan Area." Traffic Calring on Retailing: A Review of thie Evi- W'orld Bank Internal Mission Report, Washington, dence from Germaniy and U.K." Transp)orf Policy 1 D.C. (1). Pesce, Ernesto. 1996. "Reducci6n del Plomo en las Hau, Timnothy D. 1992. "Congestion Charging Mecha- Gasolinas." Paper presented at the World Batik nisms for Roads. An Evaluation of Current Prac- seminar Elimination of Lead in Gasoline in Latin tice." Policy Researchi Working Paper 1071. World America and the Caribbean, September 4-6, Bank, W'ashington, D.C. Santiago. Hillson, Marc R. 1995. "On Phasing Lead Out of Gaso- Rabinovitch, Jonas. 1993. "Urban Public Transport line: Imnplemenitation and Enforcemenit Issues." Management in Curitiba, Brazil." UNEPIndustry and Paper presented at the US. and Mexico .Sp5onsored Environment 16(1-2):17-20. International Workshop on Phasing Lead Out of Gaso- Rendleman, C.M. and N. Hohman. 1993. "The Im- line, March 14-15, The U.S. Environmtletital Protec- pact of Production Iinovations in the Fuel Etha- tioti Agency and the Mexican Secretariat of nolIndustry."Agribusiness,9(3):217-31. Environment, Natural Resources, and Fisheries, Reyneir, J.P. 1995. "Automobile Policy for the Year W'ashingtori, D.C. 2000." In Proceedings of ThrirdInternational Conference J.E. Senior Consultants Inc. 1993. "Houstoni Metro's on the Automotive Industry and the Environment, LNG Program Updated." The Clearn Fuels Report March 9-10, Environmental Matters and Knibb, 5(5):106-08. Niwvot, Colorado. Gormezano and Partners (U.K), Geneva. McArragher,J.S., WE. Betts,J. Brandt, D. Kiessling, Rothe,V.E. 1990. "AReport on MotorVehicle Inspec- G.F. Marchesi, K. Owen, J.K. Pearson. KP. Schug, tion Effectiveness." Arizona Department of Envi- and D.G. STnelgrove. 1988. "Evaporative Emissions ronmental Quality, Phoenix, Arizona. from Modern European Vehicles and( their Con- Ruiz,Jose Antonio. 1996. "The Chilean Case: Improve- trol." SAE Paper 880315, SAE Intternational, ment itAirrQualityand Elimination ofLead in Gaso- W'arrendale, Pennsylvania. line." Paper presented at the World Bank seminar McGregor, Dotiglas B. and Christopher S. Weaver. Elimination ofLead in GasolineinLatinAmerica and tte 1994. "Implementing Vehicle ITnspectio(n and On- Caribbean, September 4-6, Santiago. Road Enforcement Programs in Costa Rica. Phase Sanchez, Sergio. 1996. Personal Communications. II Report." Enginie, Fuel, and Emnissionis Enginieer- CMPCCA (Comision Metropolitana para la Preven- ing, Inc., Sacramento, California. ci6n y Control de la Contaminiaci6n Ambiental en Meyer,J.R. andJ.A. Goinez-lbafiez. 1991. "Transit Bus el Valle De Mexico), Mexico City. Privatization and the Deregulation Arounid the Sathaye, Jayant, B. Atkinson, and S. Meyers. 1989. World: Some Perspectives and Lessons." Interna- "Promoting Alternative Transportation Fuels: The tionalJournal of Transport Economics 18(3). Role of Government in New Zealand, Brazil, and Ministry of the Environnient and EnerPg of the Re- Canada." In Daniel Sperling, ed., Alternativie Trans- public of Costa Rica. 1996. "Docurnen to (le Trabajo portation Fuels: An Environmental andEnergy Soliution. de la Primera Reuni6ni de Puntos Focales para la Quorum Books, New York. Eliminaci6n del Plomo en la Gasolina.' Paper pre- Sauve, Robert. 1989. "Compressed Natural Gas and sented at the World Bank seminar Elimination of Propane in the Canadian Transportation Energy Lead in Gasoline in Latin Amenra and the Ca7ribbean, Market." In Daniel Sperling, ed., Alternative Trans- September 4-6, Santiago. portation Fuels: An Environmental and Energy Solu- Nill, E. 1995. "The Impact of Legislation on Autonio- tion. Quorum Books, New York. tive Industry Efliciency." In Proceedings of Third In- Thomas, Valerie. 1995a. "The Elimination of Lead in ter-national Conference on fte Automotive lindustry and Gasoline." Annual Review Energy Environment the Environment, March 9-10, Enviroitnetital Mat- 20:301-24. ters and Knibb, Gormezano and Partners (U.K.), . 1995b. "National and International Efforts Geneva. to Phase Out Lead in Gasoline: A Status Report." Abatement Measures for Vehicular Air Pollution 121 Paper presented at the U.S. and Mexico Sponsored Proceedings. Marchl 14-15, 1995, EPA-160R-95-001, International Workshop on Phasing Lead Ouzt of Gaoo- June 1995, Washington, D.C. line, March 14-15, The U.S. Environmerital Protec- van derStraaten,Y 1995. "The Impact of Legislationi tion Agency and thie Mexican Secretariat of on Vehicle Design and Production." Proceedings of Enviroiinienit. Natural Resources, and Fisheries. Tlhird International Conference on the Automotive In- Washington, D.C. dust7y and the Environment, March 9-10, Environ- The Times. 1995. "Vehicles Banined from iAtthenis Cen- menetal Matters and Kniibb, Gormezano and ter.' London. Partners (U.K.), Geneva. T6pfer, Klaus. 1995. "Letter of the Chairman of the Viola, Denise M. 1996. Personal Communications. UN Commission on Sustainiable Developmen t.' Engelliard Corporation, Iselin, NewJersey. Presented at the US. and Mexico Sponsored Inteoria- Walsh, M.P. 1995. "Global Trends in Motor Vehicle tional Workshop on Plhasing Lead Outt of Gasoline. Pollution Control: Challenges and Opportunities March 14-15, The U.S. Environmetital Protection in 1995." Proceedings of Third International Conference Agency and the Mexican Secretariat of Environ- on tieAutomotiveIndustrv and theEnvironment, March ment, Natural Resources, and Fisheries, Washing- 9-10, Environmental Matters and Knibb, ton, D.C. Gormezano and Partners (U.K), Geneva. Trindade, Sergio C., and Arnoldo Vieira de Carvalho. Warriner, P. 1996. Personal Communications. Califor- Jr. 1989. "Transportation Fuels Policy Issues and nia Department of Transportation, Sacramento. Options: The Case of Ethanol Fuels in Brazil." In Weaver, Christopher. 1995. "Reducing Transport-Re- Daniel Sperling, ed., Alternativie T7ansportationFtelh: lated Pollutant Emissions Fuel and Vehicle Tech- An Environmental and Energy Soluttion. Quoriin nology Options." Paper presented at the World Books, New York. Bank Conference on Transport, Energy, and Environ- UN (United Nations). 1994. TheDevelopment of UYb-1an ment, April 24-25, Washington, D.C. Puiblic Transportin Latin America andthe World. United . 1996. Personal Communications. Engine, Nations Economic Commissioni for Latin Atieicira Fuel, and Emissions Engineering, Inc., Sacramento, and the Caribbean, Santiago. California. U.S. Congress/OTA. 1992. "Retiring Old Cars: Pro- WHO/UNEP (World Health Organization /United grams to Save Gasoline and Reduce Emissions." Nations Environment Programme). 1992. Urban Air OTA-E536 UT.S. Government Printing Office, Wash- Pollution in Megacities of the World. United Nations ington. D.C. Enviroriment Program. Basil Blackwell, Oxford, U.K USEPA (United States Environmnental Protection Wijetilleke, Lakdasa, and Suhashini A. Karunaratne. Agency). 1987. "Draft Regulatory Impact Analysis: 1992. "Control and Management of Petroleum Control of Gasoline Volatility and Evaporative Fly- Fuels Related Air Pollution." World Bank, Wash- drocarbon Emissions from New Motor Vehicles." ington, D.C. Office of the Mobile Sources, Washington, D.C. World Bank. 1986. "Urban Transport: A World Bank 1990. 'Volatile Organic Compounds from O(i- Study." Washington, D.C. road Vehicles: Sources and Control Options." Re- . 1992. World Development Repart 1992: Develop- port prepared for the Long-Ranige Transboundarv ment and the Environment. Oxford University Press, Air Pollutioin Convention, Sponsored by the United NewYork. Nations Economic Comm ission for Europe - . 1994. "Sao Paulo Integrated Urban Transport (AATC), Washiingtoni, D.C. Project." World Bank, Washington, D.C. . 1994. Reformulated Gasoline: A Major Step To- - . 1996a. "Removal of Lead from Gasoline: Tech- w0ards Cleaner Air Betterfor Youtr Health and the Envi- nical Consideratioris." Pollution Prevention and ionment. Washingtorn, D.C. Abatement Handbook (Draft). WAashington, D.C. - 1995. "U.S. and Mexico Sponsored Interima- . 19961). "Sustainable Transport: Sector Review tional Workshop on Phasing Lead OutofGasolinie." and Lessons of Experience." Washiiigton, D.C. 4 CASE STUDIES A n ~increasing number of urban centers hicular air pollution problem in seven Latin in Latin America are beset with vehii- American urban centers: clar air pollution. Despite efforts to MeioCt alleviate them, ambient levels of air MexticgoCt polluitants are rising as vehicle fleets grown pos- * Santiaglo ing a health risk to urban populations and dam- * Belo Horizonte aging natural resouLrces. Thiree urban centers oBeo ie in Latin America-Mexico City, Santiago, and i Buejnos eires Sa.o Paulo-hiave particularly bad air pollution * Rioaf dej Boneiro problems. Several air pollutants in these urban SatfdeBgt centers exceed the national ambient air quiality Each study analyzes amnbient air quality, sources standards and World Health Organization of air pollution, institutional responsibilities (at (WHO) guidelines. In other urban center-s- the national, regional, and local levels), measures Belo Horizonte, Buenos Aires, Rio de Janeiro, implemented to curtail vehicular air pollution, and Santaf6 de Bogot~i-national ambient air and evaluation of these measures. These measures quality standards and WHO guiidelines are ex- are described in terms of vehicle emission stan- ceeded for at least one pollutant. dards and inspection programs, fuel-targeted This chapter presents case stuidies of the ve- measures, and transport management. 123 i MEXICO CITY IBRD 28160 . To 0eetoro T. Zcualtopon \OTepeji ,> *+ -) / _HIDALGO HIDALGO / Tizayuca / E ~~~~~ E Xf I C6 0 ~~~~~~~~ 6e~~~~~~~~~~~~~~~~~~ 32 / KlJZT6 S Teo 6huacan ThkmrJpwrJsprorSuc~~~~~ ~~~~ dbth X. 30 C 35064 7~~~~~~~ K,TLAX CA LA 0 ~~~28. \ M E Map Design Unit of The World Bank >/eF r The bosnrdaries,calors.dexnominaJtions 5 , y, aJnd any other infoJrmation shown on 'i tis map do not imply,on the port of / >: . bc.'§ The Wodld Bonk Groasp,any judoment ./ .o o. on the legal status of any berritorr.or t ~~UNITED STATES OF AMERICA * AIR QUAL17Y MONITORING STATIONS \ t \ \\ 1.~~~~~~~~~LAGUNILLA 12.XALOSTOC 22.BENITOJUARE_ 32.COACALCO \\\ _r \ ~~~~~~~~~~2.VALLEJO 13. MERCED 23.TAXQUEN~A 33. SHACIENDA < 99 \ fR ~~~~~~~~~~3. SANTA URSULA 14. PEDREGAL 24.1INSURGENTES 34.PORTALES >\\ \ N i ~~~~~~~~ ~~4.TACUBA 15. CERRO DE LAESTRELLA 25.CUITLAH-UAC 35.XOCHIMILCO ) t < 9 S~~~~~~~~~.ENEP-ACAtLAN 16.PLATEROS 26.TULTITLAN 36.CERRO DELTEPEYAC 9< \ t ~~~Gulf of Mexico 6.L.OSLZAURELES 17. HANGRES 27. ATIZAPAN 37. FELIPE ANGELES \\ \ I ~~~~~~~~~7. LA PRESA 18. UAM17ZTAPALAPA 28. CHAPINGO 38. MUSEO DE LA MEXICO 8.LAVIL.LA 19.ARAGON 29.CUAJIMALPA 39.LOMAS l ~~~~~~~~~~9. SAN AGUSTIN 20. NETZAHUALCOYOTL 30. TLALPAN 40. MUSEO TECNOLOGICO / MoxitoCity <7 1~~10.AZCAPOTZALCO 21.1IMP 31 .TLAHUAC 41.NETZAHUALCOYOTL SUR Mec CI i 11 .TLANEPANltA ; Pacific sJ; BUILT-UP AREAS ::: 7: 5 : j~~~~~*Ei eE MEXICO CITY METROPOLITAN AREA BOUNDARY Ocea ______t t t0; AIN ROADS :~~~~~~~~~~~~~~~~~~ j_- STATE BOUNDARIES i00;:::;0;04000:0 j ; 4 0 0 0 : 0 S;i9\- - INTERNATIONAL BOUNDARIES JULY 1997 Case Studies: Mexico City 127 The Mexico City Metropolitan Area (MCMA) is of cold, dense, thermally stable air prevents pol- the most populated urban area in Latin America lutants emitted at ground level from dispersing and the Caribbean. Located at the southern end upward and produces high pollutant concentra- of the Valley of Mexico at an average altitude of tions, frequently reaching critical levels. Inver- about 2,240 meters, the MCMA has expanded sions occur in the MCMA throughout the year to fill the full width of the valley, and now coin- but are especially severe during the dry winter prises about 1,200 square kilometers adminis- season (from November to March). On most tered by the Federal District and seventeen days the trapped pollutants are released when municipalities. High mountains constrain ftr- the sun heats the cold air and destroys the ther- ther growth, except to the north and southleast. mal inversion at around 9 A.M. to 11 A.M. The average height of the surrounding moun- Some of the short-term health effects of air taini ranges is about 3,200 rneters, with some pollution in the MCMA include eye irritation, peaks exceeding 5,000 meters. The city was origi- migraine, and irritation and inflammation of the nally situated on a series of artificial islands in a upper respiratory tract resulting in shortness of shallow lake surrounded by pine forests. But to- breath, sore throat, coughing, and hoarseness. day, as a result of uncontrolled urban growth, Documentation by Mexican authorities shows a 99 percent of the original lake and 75 percent clear positive correlation between the level, du- of the forests have disappeared, exposing more ration, and frequency of air pollution episodes than 40,000 hectares of land to wind erosion. a and the incidence of the symptoms (DDF 1996). major source of PM in ambient air. Urban expansion in the MCMA has been caused by high rates of population growth and Ambient Air Quality migration from rural areas. From less than I million people at the beginning of the century, The most critical air pollutants in the MCMA the MCMA's population had grown to more than are ozone and its precursors NO, and NMHC, 17 million in 1994, representing 18 percent of and PM. In addition, ambient CO concentrations Mexico's population. Between 1930 and 1980 the exceed the Mexican standard in traffic-con- MCMA became the center for Mexico's rapid gested streets. Although SO2 and lead were of industrial growth, with the number of industrial concern in the late 1980s and early 1990s, their establishments increasing by about twelve-fold. ambient concentrations have been below the The MCMA's good infrastructure and educated Mexican air quality standards since 1992. work force attracted industry, and migrants came Ambient ozone concentrations in the MCMA because of its employment opportunities. The have consistently exceeded the Mexican 1-hour MCMA now accounts for 36 percent of Mexico's standard of 0.11 ppm (220 I.g/m3) despite sig- domestic production and consumes 17 percent nificant control efforts of NO, and NMHC emis- of domestic energy production. sions. Violations of the standard have been Emissions from about 3 million vehicles, observedyearroundbecause the numberofday- 31,000 industries, and 12,000 service-related fa- light hours and the direct angle of the sun in the cilities in the MCMA are the principal anthlo- MCMA are not heavily affected by seasonal varia- pogenic source of air pollution. The area's tions.' Since 1988 ambientozone concentrations topography and climate also play a key role in have exceeded the 1-hour standard on 89 to 97 making the pollution levels unusually severe. percent of the days of the year (Table 4.1). Al- Because of its high altitude, air in the MCMA is though the number of days with very high ambi- 23 percent less dense than at sea level. This re- ent ozone concentrations increased during duces the combustion efficiency of fuels in inter- 1988-92, the situation substantially stabilized nal combustion engines that are not adjusted to during 1992-95. For example, the number of days account for the altitude and results in higher lev- with ozone concentrations above 710 Wing/m els of CO and HC emissions. 'The mountains srll- peaked in 1992 with eleven days, but none oc- rounding the MCMA tend to limit air circulation curred in 1994 or 1995. The highest ozone con- as well, trapping pollutants within the Valley of centration ever recorded in the MCMA was on Mexico. During the daytime winds carry indlus- trial emissions from the north and northwest to poptilated areas in the city. 1. By contrast, inimany other Latin American urban Thermal inversions limit air circulation even centers violatiois of 1-hour ozone standards occur only further. A thermal inversion occurs when a laver seasonally. 128 Chapter 4 Table 4.1 Number of days with high ozone concentrations in the MCMA, 1988-95 (Gwler thean (Gwrater than Gratru than (rnater than Year 220) g/grln 465 1g/rm' 588 peg/m3 710 pg/rmn 1988 326 67 1 1 1 1989 '329 15 3 0 1990 '328 84 27 3 1991 353 173 56 8 1992 333 123 37 1 1 1993 '324 80 14 1 1994 344 93 4 0 1995 '324 88 6 0 ANote: The Mexican standard is 220 pig/m:' (0 11 ppm) for 1-houir averaging time. The 220 jig/m3, 465 pg/mr, 590 pg/mr, and 710 jig/rn9measiures used here correspondl to IMECA values (described later) of 100, 200, 250, and 300, respectively. Source: DDF 1996. March 6,1992, with a value of 955 ptg/m" (DDF western section to thirteen days in the central 1996). In 1995 the two highest ozone concentra- section (DDF 1997a). tions in the MCMA were 698 pg/mn (Pedregal) In 1992 the annual average concentration of in the southwestern section and 690 pig/ma NO2 ata downtown station (244 pg/m3) was more (BenitoJuarez) in the central section. The high- than twice the USEPA's annual average standard est concentrations in each of the other sections of 100 4g/m3 (there is no corresponding Mexi- were 588 pig/m3 (Azcapotzalco) in the northwest, can standard; LANL and IMP1994). In 1995 the 562 pg/in3 (Uam lztapalapa) in the southeast, annual average NO2 concentrations in the MCMA and 552 pig/m3 (Xalostoc) in the northeast (DDF ranged from 56 pig/mi3 (Xalostoc) in the north- 1997a). In 1995 ozone concentrations were above eastern section to 87 pig/mr (BenitoJuarez) in the Mexican 1-hour standard on 89 percent of the central section (DDF 1997a). the days, above 465 pg/m3 on 24 percent of the Ambient monitoring of HC conducted in 1992 days, and above 588 jig/m3 on 2 percent of the and 1993 indicates that the highest concentra- days (see Table 4.1). tions were observed in the northeastern indus- Besides its adverse health effects and ozone trial section (Xalostoc), the central section and nitrate fonning characteristics, NO, pollut- (Merced), and the northwestern industrial, com- tion in Mexico City produces a brownish color mercial, and commuter section (Tlalnepantla). smog that reduces visibility. Ambient concentra- In March 1992, HC concentrations ranged be- tions of NO, in the MCMA are lower during tween 2.0 and 7.2 parts per million carbon April through September because of more fre- (ppmC) in Xalostoc, between 2.4 and 6.2 ppmC quent rain which transforms NO,, into nitrates in Merced, and between 1.6 and 4.7 ppmC in and nitric acid, and because of increased solar Tlalnepantla (LANL and IMP 1994). HC con- radiation and higher temperatures which pro- centrations were higher during morning hours mote conversion of NO2 to ozone. In 1992 the and averaged about 3.5 ppmC. This average con- peak 1-hour NO, level at a downtown station was centration exceeds those observed in Los Ange- 602 jg/mr, a concentration well over the Mexi- les in the late 1970s and early 1980s (2 ppmC can standard of 395 jig/ni3 (0.21 ppm; LANL and and 1 ppmC, respectively; DDF 1996). IMP 1994). In 1995 the highest 1-hour concen- Although the photochemical ozone formation tration in the MCMA was 835 pg/rnm (Enep- potential of benzene is much smaller, it has carci- AcatlAn) in the northwestern section. The highest nogenic effects on humans. At Merced and concentrations in each of the other sections were Xalostoc, 3-hour average concentrations of ben- 658 pg/nil (Benitojuarez) in the centei; 624 jig/ zene averaged between 50 and 60 parts per billion mi3 (Plateros) in the southwest, 525 pig/r:n in the (ppb), with a maximum of 86 ppb. Concentra- northeast (Xalostoc), and 466 pig/ml in the tions higher than those monitored at these sta- southeast (Taxquena). The 1-hour Mexican stan- tions would be expected at street level (LANL dard was exceeded from one day in the north- and IMP 1994). Case Studies: Mexico City 129 TSP is considered another air pollutant of recorded since 1988. Ambient PM-1 concentra- concern in the MCMA. Since 1986 ambient TSP tions are highest in the northeastern and south- concentrations have exceeded the Mexican 24- eastern sections of the MCMA, where industry hour standard of 260 pg/ml, especially in the is concentrated, and lowest in the southwestern northeastern and southeastern sections, with clay section. Ambient PM-10 concentrations over the time concentrations nearly twice as high as those standard decreased from 40 percent of the total at night. During 1990 and 1991 the highest 24- measurements in 1988 to 16 percent in 1994 and hour average TSP concentrations recorded va.- 13 percent in 1995 (DDF 1996). The 24-hour ied between 1,100 pg/mr3 and 1,300 lg/m3 in an PM-10 concentrations for the entire MCMA dur- industrial area located in the northeastern sec- ing 1990-92 were considerably lower than those tion (LANL and IMP 1994). These concentra- during 1988-89, increased slightly in 1993, but tions are among the highest recorded in any city decreased again in 1994. In 1995 the Mexican of the world. In 1993 ambient TSP concentra- 24-hour standard was exceeded in all sections tions exceeded the Mexican 24-hour standard of the MCMA except in the southwest. The high- on about 64 percent of the sampling days in the est PM-10 concentrations in the MCMA were 252 southeastern section, 30 percent of the sampling pig/m3 (Xalostoc) in the northeastern section, days in the northeastern section, and 10 percent 238 pig/m3 (Tlahuac) in the southeastern sec- of the sampling days in the central section. In tion, 206 pig/m3 (Tlalnepantla) in the northwest- 1995 the highest 24-hour average TSP concen- ern section, 187 pig/m3 (Merced) in the central tration in the MCMA was 727 pig/m3 (Xalostoc) section, and 143 pg/mi (Pedregal) in the south- in the northeastern section. The highest cOn- western section. The Mexican 24-hour standard centrations in each of the other sections of the was exceeded on 16 percent of the sampling days MCMA were 597 pg/m3 (Cerro de la Estrella) in the northeastern section, 6 percent of the in the southeast, 555 jig/m3 (Tlalnepantla) in sampling days in the southeastern section, and the northwest, 390 pig/m3 (Lomas) in the southl- 3 percent of the sampling days in the central west, and 357 pg/m3 (Merced) in the center. The and northwestern sections. Ambient PM-10 con- Mexican 24-hour standard was exceeded on 82 centrations did not exceed the PM-10 standard percent of the sampling days in the northeast- in the southwestern section (DDF 1997a). PM- ern section, 46 percent of the sampling days in 10 concentrations were lower during the nights the southeastern section, 18 percent of the sam- and rainy months. pling days in the central section, 12 percent of In 1995 annual average PM-10 concentrations the sampling days in the northwestern section, exceeded the Mexican standard of 50 pg/mr3 in and 2 percent of the sampling days in the souith- all sections of the MCMA, except the southwest- western section (DDF 1997a). ern section. The highest annual average PM-10 In 1993 annual average TSP concentrations concentrations were 87 pg/m3 (Netzahualc6yotl) in all sections of the MCMA exceeded the Mexi- in the northeastern section, 67 pig/m3 (Cerro can standard of 75 pg/m3. These concentrations de la Estrella) in the southeastern section, 60 were about 340 ,g/m3 at a southeastern station, pig/m3 (Ttultitlan) in the northwestern section, about 200 pg/m3 at central, northern, and nortlh- 51 pg/m3 (Merced) in the central section, and eastern stations, and about 120 Rtg/m3 at a south- 44 pg/m3 (Pedregal) in the southwestern sec- ern station. Annual average TSP concentrations tion (DDF 1997a). in 1993 were 28 percent less than those regis- Ambient concentrations of CO vary accord- tered in 1988 but were slightly higher than those ing to time of day and traffic flow. In 1992 8- in 1992 (CMPCCA 1995a). In 1995 annual aver- hour ambient CO concentrations were measured age TSP concentrations were still higher than as high as 27.5 mg/m3 at a monitoring station the Mexican standards: 375 jig/m3 (Xalostoc) in heavily influenced by local traffic (Cuitlahuac) the northeastern section, 244 pig/m3 (Cerro de in the northwest of the MCMA. This concentra- la Estrella) in the southeastern section, 173 jig/ tion is more than twice the Mexican 8-hour av- m3 (Merced) in the central section, 172 pig/rni erage standard of 12.6 mg/m3 (11 ppm). Since (Tlalnepantla) in the northwestern section, and 1992 ambient CO concentrations in the MCMA 119 pig/m3 (Lomas) in the southwestern section have been showing a decreasing trend (CMPCCA (DDF 1997a). 1995a). The monitoring data for 1995 show that Ambient PM-l concentrations over the Mexi- the Mexican 8-hour average standard was ex- can 24-hour standard of 150 ,ug/m3 have been ceeded in the northeastern section on three 130 Chlapter 4 days, the central section on one dav, and the jig/m3 to 0.6 pg/m3at Pedregal; LANL and IMP northwestern section by at least one day. The 1994). Since 1992 ambient lead concentrations highest 8-hour CO concentrations in the MCMA have been below the Mexican lead standard of were 18.9 mg/mrn (Netzahualc6yotl) in the 1.5 p1g/M3 for quarterly average (CMPCCA northeastern section, 17.1 mg/mr' (Merced) in 1995a). The average lead concentration for the the central section, 13.6 mg/m3 (Vallejo) in the fourth quarter of 1994 was about 0.15 pg/m3 northwestern section, 12.5 mg/rn' (Taxquena) (Pedregal) in the southwestern section and 0.6 in the southeastern section, and 9.2 mg/mr3 jg/m3 (Xalostoc) in the northeastern section (Plateros) in the southwestern section (DDF (DDF 1996).Ambientlead concentrations in air 1997a). Because most ambient concentrations have decreased significantlyas a result of reduc- are measured by monitoring stations located at tion of lead in regular gasoline and introduction roof tops, however, the restulting data may not of unleaded gasoline (Figure 4.1). be representative of the higher concentrations that people are exposed to on canyon-type streets (LANL and IMP 1994). Sources of Pollutants Ambient levels of SO2 in the MCMA have been declining as a result of the lower sulftur content Main sources of pollutant emissions in the of conventional fuels and use of alternative fu- MCMA are road transport, industry, service-re- els. During the second half of the 1980s the an- lated facilities, and wind-blown dust. Based on nual average ambient concentration of SO, the emissions inventory for 1994, motor vehicles decreased from 165,ug/m3 to 130 gg/mr' but was are by far the most important source of pollut- still well over the Mexican standard of 78 gg/m3 ant emissions. In 1994 road-based motor vehicles (0.03 ppm; LANL and IMP 1994). In 1995 the contributed to 99 percent of CO, 54 percent of maximum annual average SO2 concentrations HC, 70 percent of NO5, 27 percent of SO2, and were within the Mexican standard and ranged 4 percent of PM emissions in the MCMA (Fig- from 44 jg/m3 in the southwestern section (Santa ure 4.2). As shown in Table 4.2, which details Ursula) to 62 pig/mrn (Arag6n) in the central sec- the sources of pollutant emissions, private cars tion of the MCMA (DDF 1997a). alone emit about 44 percent of CO, 25 percent Since March 1991 ambient SO., levels have of HC and NO, 39 percent of non-dust PM, and been lower than the Mexican 24-hour- standard 13 percent of SO2 in the MCMA. of 338 pg/rml (0.13 ppm; CMPCCA 1995a; DDF Most CO emissions from vehicles are caused 1996). In 1995 the maximum 24-hotr average by incomplete fuel combustion, especially in S02 concentrations were 224 pig/mr (Arag6n) vehicles that are poorly maintained, uncon- in the northeastern section, 200 pig/m3 trolled (for example, vehicles not equipped with (Lagunilla) in the central section, 182 pig/m3 catalytic converters) or not adapted for opera- (Vallejo) in the northeastern section, 179 pg/m3 tion at high altitudes. After motor vehicles, in- (Uam Iztapalapa) in the southeast section, and dustry is the second largest emission source for 130 jig/m3 (Santa Ursula) in the southwestern NO,. In addition to motor vehicle exhaust emis- section (DDF 1997a). sions, major anthropogenic sources of HC in- In 1988 some ambient lead concentrations in clude evaporative emissions from gasoline, the MCMA exceeded 10 jg/m3 in the northwest- industrial processes, distribution and use of LPG, em, northeastern, and southeastern sections and and use of solvents. 2.5 pig/mi' in the southeastern and centr-al sec- The HC to NO, ratio in ambient air is impor- tions of the MCMA (DDF 1996). Between 1988 tant for identifying the limiting pollutant for and 1992, however, quarterly average concentra- ozone formation. For the MCMA this ratio was tions of lead in ambientair decreased by83 per- determined as varying between 13 and 50 cent in the northwestern section (from 3.0 ptg/ ppmC/ppm NO, (DDF 1996).2 Based on Mexi- mi3 to 0.5 p1g/Mr3 at Tlalnepantla), 79 percent in can Petroleum Institute plots that relate maxi- thesoutheasternsection (frorn2.4 pg/rn4toO.5 mum concentrations of ozone to the mg/ml at Cerro de la Estrella), 71 percent in the concentrations of HC and NO, in ambient air northeastern section (from 3.9 pg/rn' to 1.1 mg/ m3 at the Xalostoc), 60 percent in thie center 2. By Contrast, in the Soutt Coast Basin of California (from 1.5 pg/m3 to 0.6 pg/m3 at Merced),and 43 this ratio wvas aboult 12 in the late 1970s and mid-1980s, percent in the southwestern section (from 1.05 and abouit 9 in the 1990s. Case Studies: Mexico City 131 Figure 4.1 Ambient lead concentrations in the MCMA, 1990-94 Micrograms per cubic meter 3.5 3.0 Xalostoc 2.5 2.0- 1.5 USEPA stndard 1.5 Rg/m3 3-month average 1.0 Pedregal 0.5 0.0 1990 1991 1992 1993 1994 S.nirre: DDF 1996. Figure 4.2 Share of poHlutant emissions from motor vehicles and other sources in the MCMA, 1994 Percent of total emissions 100 - - ZE ~~~~~~~~~~~~~~~~Other 90 sources 80 Heavy-duty 70 _ | . .. | I / t, . R vehicles 60 Light-duty 50 _'' . * ii2. e t ^ jr svehicles 40 _ 30 _S 20 _ 10 _ _* _ 5 10 CO HC NO, SO2 PM Lead (2,358) (1,026) (129) (45) (452) (0.8) Air pollutants (tons/y ear) SouIrce: DDF 1996. 132 Chapter 4 Table 4.2 Pollutant eniissions by source in the MCMA, 1994 (CO HC NO, PM SO2 Fmission somrce 7;Tns/year Percent 'Vms/year Perrent Tons/year Perent Tons/year Percent Tons/year Percent Road transport 2,346,811 99.5 554,749 54.1 89,784 69.8 18,779 4.2 12,138 26.7 Private cars 1,044,008 44.3 253,866 24.8 31,913 24.8 10,321 2.4 6,062 13.3 Taxis 529,530 22.5 126,575 12.3 15,982 12.4 613 0.1 3,073 6.8 Vans, pick-ups, and microbuses 432,451 18.3 120,955 11.8 16,989 13.2 1,488 0.3 1,832 4.0 Ruta-100 buses 5,655 0.2 2,337 0.2 6,751 5.3 1,900 0.4 366 0.8 State of Mexico buses 59,110 2.5 2,837 0.3 5,077 3.9 2,195 0.5 502 1.1 Tmicks 276,057 11.7 48,179 4.7 13,072 10.2 2,262 0.5 303 0.7 Other transport' 1,686 0.1 570 0.1 2,003 1.6 63 0.0 62 0.1 Industry 8,696 0.4 33,099 3.2 31,520 24.5 6,358 1.4 26,051 57.3 Service facilities 948 0.0 398,433 38.8 5,339 4.1 1,077 0.2 7,217 15.9 Vegetation 0 0.0 38,909 3.8 0 0.0 0 0.0 0 0.0 Natural dust 0 0.( 0 0 0 0.0 425,337 94.2 0 0.0 Total 2,358,141 100.0 1,025,760 100.0 128,646 100.0 451,614 100.0 45,468 100.0 a. Includes rail and air transport modes. Sourre: DDF 1996. (Figure 4.3), formation of ozone in the MCMA PM-10 also includes secondary particles that can be more effectively controlled by reducing form in the atmosphere in the presence of solar concentrations of NO, than HC. For this reason radiation. the MCMA's recent air quality control strategy for The main source of lead exposure in the ozone emphasizes control of NO, emissions from MCMA is through inhalation of lead in ambient motor vehicles, which are responsible for about air, which results largely from combustion of 70 percent of total NO, emissions. However, high leaded gasoline in motor vehicles. People also concentrations of HC in the MCMA are likely to ingest lead from metal salts leached from metal be correlated with extensive nonautomotive use cans containing acidic food products (such as of LPG, which consists mainly of propane and chili sauces and fruit juices) and lead enamel butane. Because propane and butane are less pottery used for eating and drinking. reactive for ozone formation than most of t'he vola- About 3 million motor vehicles are estimated tile HC present in gasoline, control of HC emis- to circulate and contribute to air pollution in sions (especially from gasoline production, the MCMA. Of the vehicles registered in Mexico storage, distribution, and use) is probably more City, 71.2 percent are private cars, 17.8 percent important than the HC to NO, ratio suggests. are freight vehicles, 5.4 percent are taxis, 1.9 Industry contributes 57 percent and service- percent are buses, 1.1 percent are government related facilities contribute 16 percent of SO2 vehicles, and 2.6 are other vehicles (DDF 1996). emissions in the MCMA. SOQ emissions from About 42 percent of the vehicle fleet in the these sources originate from sulfur in industrial MCMA is more than ten years old, and 68 per- gas oil, which is limited to 2 percent by weight. cent of the vehicle fleet was manufactured be- Industrial gas oil constitutes only 2 percent of fore 1991 and thus is not equipped with catalytic total energy consumption in the MCMA. S02 converters (Figure 4.4). The deep recession emissions from the transport sector are from Mexico experienced during the 1980s partly ex- sulftur in gasoline and diesel ftuel, which together plains the large proportion of older vehicles in make up 53 percent of total energy consump- the MCMA. Tax exemptions for vehicles older tion in the MCMA (DDF 1996). than ten years and low repair costs also have en- About 94 percent of PM in air is believed to couraged prolonged vehicle ownership. These originate from natural sources, mainly soil ero- oldervehicles emit a considerable proportion of sion and dust from unpaved and paved surfaces. pollutants, and poor maintenance greatly con- However, the main contributor to PM-10 is an- tributes to air pollution. For example, based on thropogenic sources, especially motor vehicles. the results of remote sensing tests conducted on Case Studies: Mexico City 133 Figure 4.3 Relation between ambient HC, NO., and maximum ozone concentrations in the MCMA NO,( (ppm) 0.35 0.375 0.35 0.350 0.30 - 0.30 0.325 0.25 -0.300 0.25 0.20 0.20~~~~~~~~~~0.7 0.15 C | / |/ / //l ;urves representing equal 0 200 0.15 concentrations of - 0.175 0.10 maximum ozone levels 0.150 0.10 0.125 0.05 0.100w 0.05 0.00 0.00 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 HC (ppmC) Sotuie: DDF 1996. Figure 4.4 Distribution of vehicles by model-year in the MCMA, 1994 1991 and later 1986-90 1981-85 1976-80 1971-75 1970 0 5 10 15 20 25 30 35 Percent Sotirre: DDF 1996. 134 Chlapter 4 Table 4.3 Pollutant emission rates from various transport modes in the MCMA (grams of pollutatnt per passenger-kilometer) (.f) ___________ _ Nf 11*'ith \lithoit IVithz Without WVith Wilthout nl'll)yitit cntalYuic catalytic catealytic catal,ytwf catalYtir l';hirle type flznverl- converter converter converter converter converter Private cars 4.70 45.20 0.47 4.47 0.40a 1.00 Taxis 10.00 96.85 1.00 9.57 0.86a 2.14 Vans 1.11 13.40 0.14 1.30 0.08a 0.20 Gasoline-fueled microbtises 0.04 0.79 0.02 0.09 0.06a 0.16 LPG-fueled microbuses 0.11 0.91 0.01 0.06 0.03 0.07 Urban buses - 0.70 - 0.20 - 0.60 - Not available. a. Three-vay catalytic converter. Sourre: DDF 1996. motor vehicles in 1991, the highest emitting 25 percent of the fleet, they are used by only 15 per- percent of the fleet in the MCMA was found to cent of the population (CMPCCA 1995a).3 Taxis be responsible for 47 percent of CO emissions are estimated to carry more than 1 million passen- from mobile sources. Twelve percent of the fleet gers a day (World Bank 1992). Vans and microbuses contributed 50 percent of HC emissions, and 25 consume about 18 percent of gasoline. percent of the fleet accounted for 62 percent of Buses carried 991 million passengers in 1993. HC emissions (CMPCCA 1995a; Beaton, Bishop, The MCMA's electricity-driven public transport and Stedman 1992). system-which consists of the metro, trolley- While the older vehicles have remained in cir- buses, and light trains-does not emit pollutants. culation, sales of new vehicles in the MCMA have The metro provides public transport services on increased considerably since 1987. For example, a 175-kilometer network. Trolleybuses and light the annual increase for vehicle sales was 19.8 trains together transport only 1 percent of the percent between 1987 and 1992, a result of lower city's passengers. About 350 trolleybuses operate interest rates and more attractive financing con- on 13 lines with a coverage of 429 kilometers, 69 ditions, as well as of simplification of licensing kilometers of which are elevated. Trolleybuses formalities and slight increases in vehicle prices carry an average of 3 10,000 passengers a day and (CMPCCA 1995a). During the pastfewyears the light trains, about 42,000 passengers a day. The size of the MCMA's vehicular fleet has been grow- light train system consists of 15 trains that oper- ing by nearly 10 percent a year (DDF 1996). ate on the 12.5-kilometer line between Taxquefia Pollutant emission rates for typical private cars, and Xochimilco (CMPCCA 1995a). taxis, vans, microbuses, and urban buses in the MCMA are presented in Table 4.3. Among these, Institutional Responsibilities taxis and private cars have much higher- emis- sion rates than road-based mass transport ve- The institutions responsible for controlling ve- hicles. Table 4.3 also shows pollutant emission hicular air pollution in the MCMA are shown in rates for typical vehicles equipped with catalytic Table 4.4 and discussed below. converters. About 36 million person-trips are made every Federal institutions. Over the past twenty years day in the MCMA, of which 21.4 percent are by Mexico has developed a comprehensive legal private cars. Unimodal trips represent about 80 framework for environmental protection. The percent of the total and consist of about 60 per- original approach, reflected in the 1971 federal cent by buses, 29 percent by private cars, 5 per- cent by taxis, 4 percent by metro, and 2 percent by the combination of trolleybuses, bicycles, and 3. About 75 percent of the private cars in the MCMA are motorcycles. Althouglh private cars make tup 71.2 estimated to be operationial oni aniy giveni day. Case Studies: Mexico City 135 Table 4.4 Institutional responsibilities for vehicular air pollution control in the MCMA Federal District Iesponsibility SEiAMRNAP State ofAMexio I)Del/rnrment WI)) Fuel standards SEMARNAP in colla- boration with the Secretariat of Com- merce and Industrial Development (SECOFI), the Mexican Petroleum Company (Pemex), andl the Mexican Petroleum Institute Emissions standards SEMARNAP in collabora- for new vehicles and tion with SECOFI, Sec- certifying their retariats of Energy, observance by Mines, and State-Owned automotive industry Industries (SEMIP), and Pemex Emissions limit for SEMARNAP in collabora- circulating vehicles in tion with DDF, State of critical pollution areas Mexico, and Mexican and the rest of the Petroleum Institute country and determin- ing procedures and equipment to verify observance of such limits Observance of emission State of Mexico in 17 DDF in the Federal limits for private cars and MCMA municipalities, District, except for public transport; limiting except for federal public federal public vehicles, circulation of vehicles; vehicles, considering the considering the opinions setting up or authorizing opinions of SEMARNAP. of SEMARNAP. The the setting up of obligatory The Federal Ministry Federal Ministry of Tranis- inspection centers, super- of Transport and Corn- port and Communications vising the operation of these munications is respon- is responsible for federal inspection systems and limit- sible for federal public public vehicles ing the circulation of veh- vehicles icles which do not comply with emission standards Air monitoring system SEMARNAP is respon- State of Mexico is respon- In 1986 the former SEDUE sible for certifying air sible for air monitoring established an automatic quality monitoring in the State, excluding air quality monitoring systems set up by local the 17 MCMA munici- netwvork in the MCMA. governments outsidle palities. System is subject In 1990 the responsibility the MCMA to certification by for its expansion, opera- SEMARNAP tion, maintenance, and quality certification be- camne DDF's responsibility Implementinig State of Mexico in 17 DDF in Federal District emergency measures MCMA municipalities Source: General Law of Ecological Balaince and Proiectioii of the Environment; Menc'ndez 1996. 136 Chtapter 4 law dealing with environmental protection, had ment in Mexico at the federal level. As a cabinet- a relatively narrow focus on pollution prevention level department of the Mexican federal govern- and control as well as on the negative health im- men t, it oversaw programs related to pacts of pollution. In line with this focus, various environmental protection, conservation of natu- secretariats of the federal government were made ral resources, regional and urban development, responsible for setting and enforcing environ- housing, and indigenous people. SEDESOL de- mental standards, with the Secretariat of Health veloped a network of state and local vehicle in- and Public Assistance playing a coordinating role. spection and maintenance programs and, with An undersecretariat within the Secretariat of the assistance of other agencies, managed the Health and Public Assistance was made respon- inspection and maintenance program for fed- sible for formulating, implementing, and moni- eral public transport vehicles. Itwas also respon- toring air pollution control measures. sible for setting fuel specifications for the A significant shift in the approach to environ- automotive sector (Rowley 1994). mental protection occurred in 1982 with the cre- Two semiautonomous agencies within ation of the Secretariat of Urban Development SEDESOL were responsible for environmental and Ecology (SEDUE) as the federal environ- protection in Mexico: the National Institute of mental institution. In December 1988 the Gen- Ecology (INE) and the Office of the Federal At- eral Law of Ecological Balance and Protection torney for Environmental Protection (PROFEPA). of the Environment, which constitutes the cur- INE was responsible for overall environmental rent legal framework for environmental manage- policy formulation and implementation, re- ment in Mexico, was adopted. This law made search, natural resource conservation and for- SEDUE responsible for formulating Mexico's estry, hazardous waste cleanup, ecosystem environmental policy and setting national stan- management, environmental regulation, and dards, in close collaboration with other federal standards development (including setting fuel secretariats. The law also allowed a gradual shift quality specifications and developing vehicle of air pollution control responsibilities from the emission standards). It also assisted the State of federal level to state and municipal agencies. It Mexico and Federal District Department in en- gave authority to the State of Mexico and Fed- forcing environmental standards and operating eral District Department (DDF) to enforce en- the air quality monitoring system in the MCMA. vironmental standards in the MCMA, although PROFEPA, the environmental enforcement arm there have been some exceptions to this basic of SEDESOL, was responsible for monitoring scheme (for example, control of industrial air environmental compliance and enforcement of pollution in the MCMA originally was assigned environmental regulations in cases in which such to SEDUE). authority had not been vested in other Mexican Within a year of the creation of SEDUE, federal, state, or local government agencies Mexico entered a period of economic recession (Rowley 1994). and debt restructuring that required severe fis- In December 1994 SEDESOL was divided into cal austerity. As a result SEDUE's professional the Secretariat of the Environment, Natural staff was reduced and its nonsalaiy operating Resources, and Fisheries (SEMARNAP) and the budget was cut, severely weakening its capacity Secretariat of Social Development. SEMARNAP to carry out the functions assigned to it by law. became responsible for industrial pollution con- As of early 1992, SEDUE had not been able to trol, natural resources (mines, soil erosion, for- establish emission standards for heavy-duty die- ests, bulk water distribution, sanitation) sel-fueled vehicles or for in-use vehicles retrofit- management, fisheries, federal environmental ted with catalytic converters, had not conducted standards setting, and enforcement of industrial an independent analysis of the quality of fuels solid and hazardous waste regulations. Urban produced by Petroleos Mexicanos (Pemex, the planning and infrastructure were assigned to the state-owned petroleum company), and had not Secretariat of Social Development. Along with been able to enforce industrial pollttion stan- these changes, INE was totally integrated with dards (World Bank 1992). In December 1992 SEMARNAP. PROFEPA has remained a semi- SEDUE's functions were legally transferred to the autonomous agency reporting to SEMARNAP. Secretariat of Social Development (SEDESOL). Thus SEDESOL became responsible for environ- State of Mexico institutions. The Ecological Law mental protection and natural resource manage- approved by the State of Mexico legislature in Case Studies: Mexico City 137 September 1991 has expanded the responsibili- * The General Directorate of Pollution Pre- ties and powers of the State's Secretariat of Ecol- vention and Control is responsible for pub- ogy. The Secretariat of Ecology contributes to lic and private vehicle emissions testing, hay the MCMA's air pollution control program and no circula (the one-day driving ban), and supervises private vehicle inspection stations gasoline station vapor recovery programs. through a private contractor: It has delegated re- * The General Directorate of Urban Auto- sponsibility for the inspection of public service mobile Transport oversees vehicle registra- vehicles to the State of Mexico Transport Coin- tion grants and controls drivers licenses and mission (COTREM). The Secretariat of Ecoloy bus and taxi licenses. is also responsible for strengthening the MCMA's * The General Directorate of Operations is inspection and maintenance system and extend- responsible for traffic management and ing it to the rest of the State of Mexico. In addi- traffic signal systems. tion, it is responsible for monitoring air, except * The General Transport Coordinating Of- in the seventeen municipalities of the MCMA. fice regulates public transport and parking State transport-related institutions include: lots (including tariffs) and coordinates the *. Metropolitan Commission, which oversees operation of the metro, bus, and trolleybus the operation of urban and suburban buses. systems. COTREM, which is involved in the promo- The General Planning Secretariat carries ' . . ~~~~~~~~out overall transport planning for the Fed- tion, operation, and maintenance of the erat DistritDpartment. State toll road system. * The General Directorate of Public Works *Ministry for Urban Development and Pub- involved in the formula- prepares, constructs, and supervises pub- lic Works, which IS mvice works.rmla tion of transport policy and planning; lic works. tionofratransprefort pol and planning; * The General Directorate of Urban Services infrastructure for urban and suburban buises and parking; planning and constmic- sets criteria and standards and maintains tion of transport infrastructure; and issu- public works. ing bus, microbus, and taxi licenses. * The Commission for Highways and Urban Local municipalities, which issue drivers li- Transport constructs metro, bus infrastnmc- censes. ture, and parking lots. In the MCMA the Collective Transport Ser- FederalDisti-ct institutions. The Secretariat of the vice operates the metro, the Electric Transport Environment is the key environmental agency Service runs the trolleybus and light rail services, in the Federal District. It serves as the city's en- and Metropolitan Services manages the parking vironmental authority and acts as the coordinat- infrastructure (World Bank 1992). Ruta-100 ing body for the Technical Secretariat of buses, which had been operated by the Electric CMPCCA. The Secretariat of the Environment Transport Service within Federal District, ceased includes the General Directorate of Environmen- operation in 1995. In the future, the private sec- tal Projects and the General Directorate of Envi- tor will provide this service. ronmental Protection and Control. The General Directorate of Environmental Projects identifies Metropolitan institutions. Designing and imple- and prepares environmental projects. It is the menting of a program to reduce transport air basic support groulp for the Technical Secretariat pollution in the MCMA is complicated because of CMPCCA. It is made up of about forty experts national, regional, and local institutions are in- in such areas as environmental and civil engineer- volved and coordination among them is difficult. ing, chemistry, biology, sociology, communica- The problem is local in the sense that issues like tions, economics, and law. The General Direc- vehicle inspection, traffic management, and pub- torate of Environmental Protection and Control lic transport are in the hands of Federal District provides the enforcement function in the Fed- Department, state, and municipal authorities. eral District and includes thirty inspectors anil The problem is regional because the metropoli- administrative personnel (Men6ndez 1996). tan area, as currently defined, includes the Fed- The specific responsibilities of the main Fed- eral District and seventeen municipalities in the eral District transport management and air qual- neighboring State of Mexico. With the rapid ur- ity control agencies are as follows: banization and growth of industry throughout 138 Clzapter 4 the entire Valley of Mexico, the problem is re- eral Comptroller, and Energy; the Mexican Petro- gional in a broader sense and requires common leum Institute; the State of Mexico; and the Fed- action acrossjurisdictional boundaries. From an eral District Department. Although the institutional and legal standpoint, the subject is commission is responsible for preventing and national because the Federal District is not only controlling all aspects of environmental degra- the national capital, but its city government is dation in the MCMA, its primary focus has been also a direct arm of the federal government. As on air pollution. CMPCCA's main functions are: such its mayor is not elected, but is nominated by the president of Mexico and functions as a * To define antipollution policies, programs, delegate of the president.4 projects, and actions to be carried out byfed- The Federal District mayor took the first met- eral entities in the MCMA ropolitan air qualitymanagementinitiative when * To establish criteria and guidelines to pre- he established an informal institutional arrange- vent and control pollution in the MCMA ment in response to the president of Mexico's * To give its views on programs, projects, and December 1988 inaugural address. Under this budgetary allocations of federal govern- arrangement the Federal District Department ment entities, State of Mexico, the Federal took the lead in planning an air quality program District and its municipalities, and affected for the MCMA without creating new bureau- interest groups and individuals cratic structures and by relying on private sector * To propose to the authorities actions to expertise. The mayor appointed the chief of his prevent and control environmental emer- General Coordination for Environmental Pro- gencies in the MCMA grams to prepare an air pollution control pro- * To support environmental research, tech- gram. To help develop the program, two nological development, and training pro- intergovernmental groups were formed: a steer- grams ing committee consisting of senior officials of * To define ways to secure resources to form the Federal District Department, State of a fund to finance the programs, projects, Mexico, SEDESOL, some federal secretariats, and actions approved by CMPCCA Pemex, Mexican Petroleum Institute, and the * To develop CMPCCA's internal rules and Federal Power Commission; and a technical sec- regulations retariat consisting of staff members of the agen- * To carry out other actions related to its ob- cies making up the steering committee, working jectives, as directed by the president of on a part-time or temporary basis. Members of Mexico. the technical secretariat and teams of local and foreign experts, directed by Mexican consultants CMPCCA comprises a president, a technical prepared the program. Although these arrange- secretariat, and an advisory council. CMPCCA's ments worked relatively well for program pre- presidency was established for a two-year term paration, an effective long-run air quality rotating among the mayor of the Federal Dis- management program required that institutional trict, the governor of the State of Mexico, and arrangements be strengthened, complemented, the secretary of SEMARNAP. Although the and made more permanent. mayor and governor have both served as presi- On January 8, 1992, the president of Mexico dent of the commission, the secretary of issued a decree creating CMPCCA as a penna- SEMARNAP recently declined the presidency. nent entity to define, coordinate, and monitor This decision, which will make localjurisdictions implementation of public policies, programs, and responsible for solving their own environmen- projects to control pollution in the MCMA. The tal problems, is in line with a national trend to- agencies comprising the CMPCCA include the ward decentralization of responsibilities and Secretariats of Health, Treasury and Public Credit, resources from federal to local institutions. Social Development (currently SEMARNAP), CMPCCA's technical secretariat comprises a Public Education, Commerce and Industrial De- technical secretary and one technical represen- velopment, Communications and Transport, Fed- tative from each member agency. It is respon- sible for identifying measures to curtail air pollution in the MCMA and specifying sched- ules, budgets, and institutional responsibilities 4. In 1997 the mayor of Mexico City is schediuled to be for implementing these measures. The Depart- elected. ment of Environmental Projects within the Fed- Case Studies: Mexico City 139 eral District Department's Environment Secre- nated policy development and implementation. tariat is the coordinating body for the CMPCCA's The council consisted of a president, a technical technical secretariat. By presidential decree, the secretary, a full-time technical team, and working costs of this coordinating group are assumed by groups. One of the working groups was charged the Federal District Department. with developing a transport master plan for the The Federal District's environmental secretary MCMA (defined as the Federal District and has been appointed by the president of Mexico twenty-one instead of seventeen municipalities). as CMPCCA's technical secretary for a two-year Critical to this plan was formulation of a program term. The key functions of the technical secre- of travel demand management to help achieve tary are to prepare programs and projects to both transport and air management objectives. obtain and administer grant, loan, and budget Through the council's efforts, transport services funds; ensure that these funds are used in an started across the Federal District's boundary in effective and timely manner; coordinate the July 1991. activities of the working groups; and train tech- To better address transport issues between the nical personnel to ensure the quality and conti- State of Mexico and the Federal District, inJuly nuiity of actions to deal with pollution problems. 1994 the Metropolitan Transport Council was The representatives from member agencies are replaced by the Metropolitan Commission for mid-level government officials backed by their Transport and Highway Administration institutional resources. This arrangement se- (COMETRAVI) in accordance with an agree- cures the necessary participation and contribu- ment signed by the Ministry of Communications tion of all involved agencies. These technical and Transport, the State government of personnel have accrued considerable experience Mexico, and the Federal District Department. from working together to solve MCMA's coin- COMETRAVI's organizational structure is simi- plex environmental issues, and are exposed to lar to CMPCCA's except that it has fewer profes- the experiences of the world's most advanced sional staff (about ten members), it does not environmental agencies and to international have a advisory council (so all its study propos- environmental information (Menendez 1996). als are only technical), its technical secretary CMPCCA has established an advisory counicil does not have executive authority, and its staff to ensure citizen participation in the design and operates independently of federal agencies and implementation of air pollution control mea- local governments. Authority lies with the secre- sures for the MCMA. The advisory council con- taries of transport from the local governments sists of representatives of the scientific, in the State of Mexico and the Federal District environmental, industrial, business, and NGO (Menendez 1996). COMETRAVI developed a communities. Representatives from the federal. 1995-96 work program that includes homogeniz- Federal District, and State of Mexico govern- ing transport laws and regulations; developing ments are also invited as needed to participate technical standards for vehicle manufacturing; in advisory council discussions. The advisory regularizing vehicle permitting, improving bus council can suggest air pollution control mea- routes, and modernizing license plates; tariffing sures to the CMPCCA's technical secretariat, and and financing public and freight transport for evaluates and approves the secretariat's pro- fleet renewal; supervising traffic and enforcing; posed measures before implementation. developing road infrastructure; saving energy The MCMA's jurisdictional complexity had and protecting the environment; regulating resulted in overlapping transport management freight transport; developing a master plan for responsibilities until 1989. The sharp division public transport and roads; and preventing ac- between the Federal District Department and cidents. A working group has been formed for the State of Mexico government in planning, each element of this program. The working coordinating, and supplying transport services group on the environment includes the techni- resulted in a disparate, inefficient, and costly (to cal secretariat of the CMPCCA. Because users) transport network, demonstrated by the COMETRAVI does not have any executive au- massive interchanges at the Federal District's thority or the ability to commit budgets to imple- boundaries. In 1989 a Metropolitan Transport ment proposed projects, COMETRAVI's Council was formed by the federal secretary for technical secretary relays important recommen- transport and communications, the governor of dations made by the working group to the local the State of Mexico, and the mayor of the Fed- governments' transportation secretaries for ne- eral District to provide a framework for cooirli- gotiation. 140 Chapter 4 Implemented Measures and communication. Of these, the measures dealing with oil industry and fuels and the trans- Although efforts to control air pollution in the port sector were the most relevant for curtailing MCMA go back many years, it was only in 1989 vehicular air pollution. Many of the components that an emergency program was initiated after of the integrated program have been financed widespread public concern over increasing lev- through a $220 million World Bank loan for els of air pollution. The program was coordi- the MCMA Transport Air Quality Management nated by a steering committee and a technical Project. secretariat, established by the federal govern- The oil industry and fuels component of the ment and consisting of the representatives from integrated program included two types of mea- Federal District Department; State of Mexico; sures: changes in the refining and distribution Secretariats of Ecology and Development, Fi- infrastructure to reduce emissions from these nance, Programming and Planning. and Indus- sources, and changes in fuel composition to re- try and Agriculture; Pemex; Mexican Petroleum duce pollutant emissions from other sources. Institute; and National Bank of Public Works and The first set of measures was aimed at reducing Services (BANOBRAS). The following measures pollutant emissions directly by shutting down a were directed toward curtailing vehiicular air refinery located in the MCMA, adding floating pollution: roofs to fuel storage tanks, and installing vapor recovery systems for filling stations to reduce HC * Tighter emission standards for vehicles, vapor emissions. The second set of measures effective from model-year 1991. included supplying unleaded gasoline for cata- * A rotating driving ban called "don't drive lytic converter-equipped vehicles, reformulating today" (hoy no circula) aimed at prohibit- leaded gasoline to reduce pollutant emissions, ing circulation of one-fifth of the vehicles and desulfurizing diesel fuel and fuel oil. each weekday. The program's transport-related emission con- * Addition of 5 percent MTBE to gasoline to trol measures involved steps to reduce pollutant make up for the reduction in the lead con- emissions per kilometer traveled and encourage tent of gasoline and to reduce HC and CO use of public transport while limiting private car emissions. use. Specific measures included developing new * A vehicle inspection and maintenance pro- emission standards for gasoline-fueled vehicles, gram. requiring emission controls by catalytic convert- * Replacement of some bus engines with ers; replacing high-use vehicles (taxis and engines meeting the California 1990 emis- microbuses) with new vehicles meeting emission sion standards and rehabilitation of these standards; converting gasoline-fueled cargo buses to improve their comfort, appear- trucks to LPG or CNG; expanding and improv- ance, and mechanical condition. ing the inspection and maintenance program; * Expansion and improvement of the air continuing the "don't drive today" program for quality monitoring network. private vehicles and expanding it to taxis and public transport vehicles; improving parking and Some of these measures were implemented traffic management systems; expanding the through a $22 million World Bank loan, Mexico metro and surface electric transport systems; and First Urban Transport Project (for example, authorizing of new private bus routes. purchase of some Ruta-100 bus engines and of At the end of 1995 CMPCCA prepared a Pro- air quality monitoring equipment). gram to Improve the Quality of Air in the Valley In September 1990 the Mexican government of Mexico for 1995-2000 (hereafter referred to announced an expanded program called the as the new program). The new program includes Integrated Program Against Air Pollution in the a diagnosis of the air pollution problem in the MCMA. This program consisted of forty-two spe- MCMA and considers strategic actions in the cific measures (including continuation of mea- following areas: improving and using new tech- sures under the 1989 Emergency Program), nologies and cleaner fuels in motor vehicles, in- grouped into the following categories: oil indus- dustry, and the service sector; providing safe and try and fuels; transport sector; private industry efficient public transport services; integrating and services; thermoelectric plants; reforestation metropolitan policies for urban development, and sanitary measures; and research, education, transport, and environment; providing eco- Case Studies: Mexico City 141 Table 4.5 Exhaust eniission standards Table 4.6 Exhaust emission standards for new cars in Mexico for new light-duty commercial vehicles (grams per kilometer) in Mexico Aodlel-year Mt) HC 1j() (grams per kilometer) 1975 29.0 2.5 lihicle weighif/model-year CO HNQ, 1976 24.0 2.1 Gross vehicle weight 1977-88 24.0 2.6 2.3 below 2,727 kg 198S9 22.0 2.0 2.3 1990-93 22.0 2.0 2.3 1990 18.0 1.8 2.0 1994 and newer 8.75 0.63 1.44 1991-93 7.0 0.7 1.4 Gross vehicle weight 1994andnewer 2.11 0.25 0.62 from2,728-3,000kg Note: A blank space indicates that nio standard was 1990-91 35.0 3.0 3.5 established. Evaporative hvdrocarbon emissions were 1992-93 22.0 2.0 2.3 not regtilated before 1995. Subseclitently, all cars were 1994 and newver 8.75 0.63 1.44 required to meet the 2.0 grams/test emission standard. Somrce: World Bank 1992; Dijio 0/idial of October 19. m Dietio Ofiil of October 19, 1988. 1988. they agreed to install evaporative control systems nomic incentives; inspecting and enforcing mo- compatible with those in the United States. Start- tor vehicles and industry; and encouraging infor- ing in 1995 all cars were required to meet the mation, environmental education, and social evaporative emission standard of 2.0 grams/test. participation. The new program identifies spe- As part of NAFTA, the United States, Canada, cific measures that include estimated benefits and Mexico are harmonizing vehicle emission and costs, budgets, and institutions responsible standards. for implementation. In 1988 emission standards were established for new commercial vehicles with a gross weight Vehicle emission standards. Since 1975 new cars below 2,727 kilograms and for vehicles between sold in Mexico have been subject to exhaust 2,728 and 3,000 kilograms (Table 4.6). Manu- emission standards that include CO, HC, and facturers were allowed to comply with these stan- NO, (Table 4.5). In general, these emission stan- dards through an "average fleet emission" until dards have been based on U.S. standards. Be- the 1992 model-year. This average allowed the fore model-year 1991, Mexican standards were marketing of some new vehicles with emissions lenient enough to be met without use of a cata- higher than the standard as long as other ve- lytic converter or other advanced emission con- hicles marketed by the same manufacturer had trol technologies. The standards for 1991-93 lower emissions, therefore achieving compliance model-years represented a transition period be- with the average fleet emission. In 1993 emis- cause they were sufficiently stringent to require sion standards for these two weight categories were use of catalytic converters but did not inclttde merged for 1994 and subsequent model-year ve- the full range of emission controls required in hicles. In addition, the weight limit of light-duty the United States. Beginning with the 1994 commercial vehicleswas extended to 3,857 kilo- model-year all new cars were subject to exhaust grams.5 In 1995 evaporative emissions for light- emission standards equivalent to those of the duty commercial vehicles were limited to a United States. These standards necessitated the maximum of 2.0 grams/test.6 installation of three-way catalytic converters and Exhaust emission standards for new urban computerized engine control systems but lacked transport microbuses were proposed for polluted requirements for evaporative emission controls, emission warranties, or recall of vehicle parts that did not comply with emission standards. During 5. The regtilationi NOM-CCAT-004-ECOL/ 1993 was pub- the 1991-93 transition period most Mexican car lished in the Diario 0/idioil of October 22, 1993. Commer- cial vehicles are definied as light-duity tnicks designed for manufacturers still equipped their vehicles with commercial, tnultiple, or service tuse and fueled with gaso- U.S.-equivalent emission controls. Even the ah- linie or alterniative ftuels (sticih as LPG and CNG). sence of formal evaporative emission standards. 6. Evaporative emissionis were not regulated until 1995. 142 Chlapter 4 Table 4.7 Proposed exhaust emission Table 4.8 Exhaust eniission standards standards for new urban niicrobuses for new heavy-duty comnmercial vehicles in Mexico (3,000-5,000 kilograms) using gasoline or alternative fuels (grams per kilometer) in Mexico AModel-year C( HC NQ0 (grams per break horsepower-hour) 1991 10.0 0.6 1.5 eMhicle rveightl/model-year Co HC NQ,) 1992 3.0 0.3 1.0 Gross vehicle weight .Source: World Bank 1992. from 3,858-6,350 kg 1995-97 14.4 1.1 5.0 1998 and newer 14.4 1.1 4.0 Gross vehicle weight above 6.350 kg areas but have not been enacted (Table 4.7). Be- 1995-97 37.1 1.9 5.0 cause of the critical air quality levels in the MCMA, 1998 and newer 37.1 1.9 4.0 however, vehicle manufacturers agreed to com- Note: All model-year vehicles indicated in this table ply with these standards one year earlier than is shotild meet the 3.0 grams/test for evaporative emnissionis. indicated in Table 4.7. In practical terms, that Source: Men6ndez 1996. meant that microbuses sold in the MCMA had to incorporate a catalytic converter starting with the 1992 model-year (Menendez 1996). In October 1995 emnission standards were established for new CNG, methanol, and mixtures of these fuels with commercial vehicles using gasoline or alterna- gasoline or diesel fuel (Table 4.11). tive fuels with gross weights ranging from 3,858 In 1993 standards limiting smoke emissions kilograms to 6,350 kilograms and greater than were established for in-use diesel-fueled vehicles. 6,350 kilograms (Table 4.8). The smoke opacity (expressed by absorption Until 1993 new diesel-fueled vehicles were coefficient) was limited according to the nomi- subject only to smoke opacity testing.7 Exhaust nal flow of a vehicle's exhaust gas. These stan- emission standards for CO, HC, NO,, and PM dards were revised in 1996 and limited smoke were established starting with the 1993 model- emissions based on the vehicle weight category, year diesel-fueled vehicles. These standards were model-year, and gas flow from the exhaust (Table further tightened first for the 1994-97 and then 4.12). for the subsequent model-years (Table 4.9). Emission standards were established for in-use Vehicle inspection programs. The Mexican govern- gasoline-fueled cars, trucks, and commercial, ment views inspection of in-use vehicles as a cru- multiple-use, and service vehicles with a gross cial element in controlling air pollution in the vehicle weight below 2,727 kilograms. These stan- MCMA. Inspection of in-use vehicles was initi- dards, which limit CO and HC emissions in ex- ated in the 1970s. At that time the main objec- haust, were tightened in 1996 (Table 4.10). In tive was to detain vehicles that were emitting addition, in 1996 emission standards were estab- excessive amounts of smoke based on visual in- lished for in-use commercial, multiple-use, and spections. In August 1988 a voluntary periodic service vehicles and for trucks fueled wvith LPG, inspection and maintenance program was intro- duced as a measure to curtail air emissions from gasoline- and diesel-fueled in-use vehicles in the MCMA. InJanuary 1989 periodic inspection of in-use vehicles was made compulsory, with one 7. The opacity standard for heavy-duty vehicles was set at 70 units using the Hartridge method. However, with the test per year charged at a cost equivalent to one establishment of the Permanent Program foir Correcting day of the minimum wage. In 1990 the number Excessive Black Smoke in early 1992, this standard was low- of periodic inspections was increased to two per ered to 40 units for Ruta-100 buses. The Haritridge method year, with the second test free of charge. In 1991 draws a contintiotis sample of the vehicle exhatist into a one annual inspection was established, but sub- chamber to measure the attenuatioll of a light-shining beam. This methiod can measuire the smoke level contributed by quently icreased to two ispections a year. soot (throtigh light absorption) antI oil or fuiel dtroplets Emission testing of in-use gasoline-fueled ve- (thirough light scatteritig). hicles involves determining CO and HC concen- Case Studies: Mexico City 143 Table 4.9 Exhaust eniission standards for new diesel-fueled vehicles in Mexico (grams per break horsepower-hotur) Vehict' type/model-year (:) HC NOV PMI 1993 15.5 1.3 5.0 0.25 1994-97 Heavy-duty urban buses 15.5 1.3 5.0 0.07 Medium-duty urban buses and others 15.5 1.3 5.0 0.10 1998 and newer Heavy-duty urban buses 15.5 1.3 4.0 0.05 Medium-duty urban buses and others 15.5 1.3 4.0 0.10 Neote: These standards apply to diesel-fueled vehicles withi gross weight greater than 3,857 kilograms. Heavy-duty urban buses have a gross weight greater than 15,000 kilograms and an engine power greater than 250 hp, and carry 15 or more passengers. All diesel-fiueled vehicles should also meet the opacity test of 15 percent at idle, 20 percent at acceleration, and 50 percent at peak power. Sourre: Walsh 1996; NOM-CCAT-007-ECOL/1993 published in Diario Oficial of October 22, 1993. trations in the tailpipe exhaust at 2,500 rpm with- tion consists of measuring smoke opacity as the out load. Although this procedure is commonly engine is accelerated rapidly without load. This used and very effective in identifying high emis- test procedure is capable of detecting high PM sions from older vehicles, it cannot reliably iden- and HC emissions from diesel engines. tify emission-related defects in late-technology To test high-use passenger and freight vehicles, vehicles equipped with electronic air-fuel con- government-operated inspection stations are trol systems. For diesel-fueled vehicles the inspec- equipped with dynamometers that simulate driv- Table 4.10 Exhaust emission standards for in-use light-duty gasoline-fueled vehicles in Mexico O(0 (% by volume) HG (ppm) Before After Before After January January January January Vehicle type/model-year 1996 1996 1996 1996 Cars and commercial vehicles 1979 and older 6.0 4.0 700 450 1980-86 4.0 3.5 500 350 1987-93 3.0 2.5 400 300 1994 and newer 2.0 1.0 200 100 Multiple-use and service vehicles and light-, medium-, and heavy-duty trucks 1979 and older 6.0 5.0 700 600 1980-85 5.0 4.0 600 500 1986-91 4.0 3.5 500 400 1992-93 3.0 3.0 400 350 1994 and newer 2.0 2.0 200 200 Note: These standards apply to gasoline-fuieled in-use vehicles with gross weight below 2,727 kilograms. The maximum oxygen content is limited to 6.0 percent by volume tot model-years before 1994 and 15.0 percent by volume for model-years 1994 and newer. The minimuim and maximumni CO+ 02 dilutioni limits for all model-years are 7.0 anti 18.0 percent by volume. Smnirce: Walsh 1996; NTE-CCAT-003/88 published in l)ioio i(Oirial of June 6, 1988; NOM-EM-102-ECOL-1966 published in D)iario Oficial of July 24, 1996. 144 Chapter 4 Table 4.11 Exhaust CO and HC emission Table 4.12 Exhaust smoke emission standards for in-use commercial, standards for in-use diesel-fueled multiple-use, and service vehicles and for vehicles in Mexico trucks fueled with LPG, CNG, methanol, Absorption and mixtures of these fuels with gasoline Vehicle woeighi/model-year coeffirient (m') or diesel fuel in Mexico Gross vehicle weight below 2,727 kg (;O H(; 1995 and older 1.99 Model-year (% ;y volume) (ft/mn) 1996 and newer 1.07 1986 and older 2.0 200 Gross vehicle weight above 2,727 kg 1987-93 1.0 150 1990 and older 1.99 1994 and newer 0.75 100 1991 and newer 1.27 Note: The maximum oxygen content is limited to 6.0 Source: NOM-045-ECOL-1995 published in )iario Oficial percent by volume, and the minimtim andc maximum ofjJaiuary 8, 1996. CO + CO2dilution limits for all model-years are 7.0 and 18.0 percent by volume. Source: NOM-EM-102-ECOL-1966 published in D)iaro Qficial of July 24, 1996. role of inspection and maintenance, at the end of 1995 the Federal District Department decided to close all private garage inspection stations in favor of a completely centralized inspection ing conditions in the MCMA under load. In 1993 system. these stations were modernized with computer- SinceJanuary 2, 1996, all periodic inspections ized equipment that included high-precision gas in the Federal District have taken place at one analyzers and a data processing system that can of the existing centralized facilities. Twenty-six immediately print out emissions test results and such facilities maintain about 142 lanes, result- compare them with applicable emissions stan- ing in serious congestion. This problem is ex- dards. Along with this procedure, stickers are pected to be overcome, however, when the issued to passing vehicles to allow circulation for expanded system becomes operational in 1997. a period of six months. A central verification The inspection frequency has been reduced to center equipped with dynamometers allows once per year to relieve congestion at inspec- measurements of pollutant emissions from tion centers, but taxis and other high-use vehicles heavy-duty passenger or freight vehicles under continue to be tested twice a year. simulated driving conditions in the MCMA. Fur- In close coordination with vehicle manufac- thermore, since 1994 nine emission evaluation turers, the Federal District Department has been centers in the Federal District have been used moving to expand the existing centralized in- to provide drivers with a technical assessment of spection network by adding thirty-six centralized their vehicles free of charge. stations. A bidding document has been prepared Until 1996 emissions testing of private vehicles for the construction and operation of these fa- in the MCMA was carried out at 1,650 autho- cilities in 1997 using accelerated simulation rized private and twenty-six high-volume govern- mode (ASM) testing.8 The bidding documents ment-operated inspection stations. The periodic do not put any restrictions on the maximum num- inspection and maintenance program with the ber of lanes (with the minimum being three) or private garage system, however, proved ineffec- facilities any one company will be able to win. In tive. Although the failure rate was 16 percent at addition, any company winning a bid must have the centralized government-operated inspection as part of its team a former owner of one of the centers, it was only about 9 percent at private pnivategaragesthatcarriedouttinspections. Fund- garages. Performance monitoring of a large ing for the centralized facilities will be entirely number of private inspection stations was also private, with the payback coming out of the fee virtually impossible with available resources. Sta- tions conducting improper or fraudulent inspec- tions were taken to court on several occasions 8. The ASM tests, which require dynamometer loading with the courts usually agreeing to shutt them of vehicles, yield good correlations with Federal Test Proce- down. As part of MCMA's new air polltution con- dures (FTP) for pollutants emitted in the exhaust, especially trol program and in recognition of the critical NO,. Case Studies: Mexico City 145 for each test. Fees will be 7'0 pesos ($9.40) for of Nova consumed in the MCMA was gradually carsand 100 pesos ($13.50)forhigh-usevehicles, reduced from 0.14-0.28 g/liter in 1986 to and 10 pesos ($1.35) of the lfee will be submitted 0.08-0.15 g/liter in 1991, 0.06-0.08 g/liter in to the government to pay for the stickers. A spe- 1992, and 0.03-0.06 g/liter in the winter of 1994 cial committee has been set, Up with prominent (Berumen 1996). Production of Magna Sin be- industrialists and academics, who will help se- gan in 1988 to meet the fuel requirements of lect the winning bidders. new light-duty vehicles equipped with catalytic The existing stations are being upgraded so converters. In 1995 the lead content of Magna that all will have a dynamometer and be capable Sin was 0.00017 g/liter. The share of Magna Sin of running the ASM test in 1997. Existing sta- in total gasoline sales in the MCMA was only 2 tions equipped with the necessary equipment percent in 1989, but it increased to 37 percent in ASM began testing in mid-1996. Test results are 1994,44 percent in 1995, and 48 percent in May not used to pass or fail vehicles but rather to gather 1996 (CMPCCA 1995a, DDF 1996, Berumen necessary data for setting newv emission standards 1996) ." Magna Sin's market share is increasing for in-use vehicles, as well as standards that will as more catalytic-converter equipped vehicles are exempt in-use vehicles from the one-day and twvo- marketed and some of the old vehicles in circu- day driving ban programs (see below). The test lation are retiring. procedure includes the two-stage idle test for One component of Pemex's production of private cars and steady-state loaded test for higlh- Magna Sin involved increasing the production use vehicles. Independent auditors are present capacity of the Madero refinery from 15,000 to at each inspection station during tests. These 20,000 barrels a day. In addition, the octane de- auditors change stations monthly, and their con- ficiency caused by reduced lead content in gaso- tracts are funded by companies operating the sta- line was made up for by converting the catalytic tions. Each test is recorded on a video in real reforming units for naphtha from a semi- time, and tapes and computer printouts are spot regenerative to a continuous catalytic regenera- checked to assure that tests are valid. Evaporative tive system at various Pemex refineries. test procedures will be developed later. The State Investment requirements were $12 million at of Mexico plans to use the same inspection ap- each of the Salamanca, Cadereyta, and Mina- proach as the Federal District, with about a one- titlan refineries (in August 1993), $36 million at year delay. the Salina Cruz refinery (inJuly 1994), $32 mil- Roadside inspections complement periodic lion at the Tula refinery (in August 1994), and vehicle inspections. Since December 1, 1992, the $54 million at the Madero refinery (in Septem- Federal District Department, the State of ber 1995). Three isomerization units were also Mexico, and the Federal Secretariat of Commu- constructed to convert low-octane n-pentanes nications and Transport have been jointly con- and n-hexanes to higher-octane hydrocarbons. ducting an on-the-road inspection program The conversions were performed at the La under PROFEPA's supervision. This program, Cangrejera Petrochemical Complex at a cost of which targets excessively polluting in-use vehicles $52 million (in August 1994), at the Minatitlan in the MCMA, is staffed by fifty inspectors who refinery at a cost of $28 million (in December use eleven patrol vehicles and twenty mobile 1994), and at the Cadereyta refinery at a cost of emission testing instruments. Vehicles barred $36 million (in December 1994; Rosas 1996). from circulation are taken to one of thirtv-six The closure of Pemex's 18 de Marzo refinery in detention centers (CMPCCA 1995a). the MCMA in March 1991 somewhat lowered production of transport fuels, but it also consid- Fuel-targeted measures. In 1994,56 percent of the erably reduced air emissions. fuel consumed in the MCMA (44.4 million li- The 23 percent lower air density at MCMA's ters a day) was for the transport sector, mostly in altitude adversely affects combustion of fuels in the form of leaded gasoline (called Nova), un- internal combustion engines, resulting in higher leaded gasoline (called Magna Sin), and diesel fuel. All gasoline and diesel fuel sold in Mexico is produced by Pemex. Alternative fuels used in the MCMA include LPG and CNG. 9. The share of Magna Sin in gasoline sales in May 1996 was 59 percenit nationwide, 51 percenit in Gtuadalajara, 88 Fuel quality has improved as a result of percent in Monterrey, and 98 percent in the northern bor- Pemex's refiner-y investments. The lead content der area (Berumeni 1996). 146 Chapter 4 CO emissions. To reduce CO emissions from ve- percent by volume, olefins content to a maximum hicles and boost the octane number of gasoline, 10 percent by volume, benzene content to a maxi- since 1989 Pemex has produced leadled and un- mum content 1.0 percent by volume, and Reid leaded gasoline containing 5 percent MTBE by vapor pressure to a maximum 7.8 psi (Rosas volume. The MTBE raises the oxygen content in 1996). To ensure compliance with gasoline speci- gasoline to the legally required range of 1 to 2 fications, CMPCCA has been inspecting refiner- percent. National production of oxygenates ies and gas stations. started with the construction of MTBE plants at CMPCCA is considering further improve- three refineries: Salina Cruz (atacost of$15 mil- ments in the qualityofmotorvehicle fuels utsing lion in December 1994 for a 30,000 metric ton a a cost-effective approach. Currently, an experi- year capacity), Cadereyta ($12 million in Decem- ment is being conducted in Monterrey with a ber 1994 for a 30,000 metric ton capacity), and new, high quality "premium" gasoline. This fuel, Tula ($12 million in September 1995 for a 90,0000 which was introduced in several stations as a re- metric ton capacity). In addition, TAME plants placement for Nova, has the following charac- were constructed at Salina Cruz ($13 million for teristics: 92 research octane number, maximum a 60,000 metric ton capacity) and Cadereyta ($19 limits of 1 percent benzene, 32 percent aromat- million fora 90,000 metric ton capacitv) to supply ics, 12 percent olefins and an oxygenate con- additional oxygenates for gasoline (Rosas 1996). tent of between 1 and 2 percent, and no lead. To limit air emissions of ozone-forming HC and This fuel costs about 20 percent higher than carcinogenic benzene, in December 1992 Pemex Nova. established reformulated fuel specifications for In 1986 the sulfur content of the Mexican die- Magna Sin and Nova (CMPCCA 1995a). Based sel fuel marketed in the MCMA was reduced on these specifications, the Mexican gasoline from 1 percent to 0.5 percent by weight with the quality standards were established (Table 4.13). introduction of Special Diesel fuel. Pemex later Controlled parameters include vapor pressure, initiated a program to further reduce SO2 emis- aromatics, olefins, benzene, sulfur, lead, and sions by supplying a lower sulfur diesel fuel to oxygen. meet the fuel quality requirements of diesel-fu- For the high-pollution winter seasons, start- eled vehicles equipped with advanced emission ing with the 1993-94 winter season, Pemex tight- controls. As a result the sutlfur content of the die- ened gasoline specifications. For both Magna Sin sel fuel was reduced to 0.05 percent by weight and Nova the olefins content was limited to a with the introduction of Diesel Sin fuel in Octo- maximum 12 percent by volume and the ben- ber 1993. On December 2, 1994, fuel standards zene content to a maximum 1.5 percent by vol- for the Special Diesel were published in Diario ume. In addition, the aromatics content of Nova OficiaL These standards limit the sulfur content was limited to a maximum 25 percent by volume. to 0.05 weight percent and specify the require- In 1997 Pemex will begin producing Magna Sin ments for other parameters as well (for example with aromatics content limited to a maximum 25 maximum aromatics content of 30 percent by vol- Table 4.13 Gasoline quality standards in Mexico Fuel parameter Magna Sin Novu Reid vapor pressure (psi) 6.5-8.5a 6.5-8.5a Aromatics (maximumn percent by volume) 30 30 Olefins (maximnum percent by volume) 151) 15 Benzene (maximum rpercent by volume) 2 2 Oxygen (percent by weiglht) 1-2 1-2 Sulfur (maximum percent by weight) 0.10 0.15 Lead (g/liter) 0.0026 0.06-0.08 Note: Futel standards for Magna Sin and Nova incltude other parameters not included in this table. a. The Reid vapor pressure of Magna Sini and Nova prior to these standards was specified at 7.0-9.5 psi. b. After January 1998, the maximutm permissible value will be 12.5 percent by volttme. Source: Diano Oficial of Decenmber 2, 1994. Case Studies: Mexico City 147 uime and minimum cetane number of 48). To problems. Two vehicles converted to use LNG provide this low sulfur fuel, three hydro- were tested. These vehicles received the fuel in desulfurization plants were constructed: two at liquid form and evaporated it before combustion Tula refinery with 20,000 barrel a day capacity (Sanchez 1996). and one in the Salamanca refinery with 25,000 Measures are being adopted to reduce HC barrel a day capacity. Total investment for this emissions from gasoline storage, distribution, project was $115 million. In addition, desulfur- and sales activities in the MCMA. Four Pemex ization units at other Pemex refineries (such as gasoline storage distribution terminals have in- Salina Cruz) have been producing low-sulfur die- stalled vapor recovery systems. In addition, all sel fuel (CMPCCA 1995a). In February 1997 the gasoline stations and Pemex's gasoline distribu- sulfur content of diesel fuel used throughout tion trucks have been required to install vapor Mexico was reduced to 0.05 percent. recovery systems. Gas stations collect the stored Pemex's future investments to meet the str-in- vapors and return them to Pemex's distribution gent motor vehicle fuel quality requirements in centers. Mexico will total about $1 billion. These invest- The federal government controls fuel pricing ments involve construction of alkylation, isomer- in Mexico. The retail price of motor vehicles is ization, and hydrodesulfLirization plants atvarious the same at any location in Mexico. It includes refineries (Berumen 1996). the producer price (which is determined accord- In the MCMA the use of LPG as a transport ing to a reference price based on international fuel began in February 1992 with the conversion price to reflect the opportunity cost of the fuel), of some buses operated by the public sector or a specific tax, and a value added tax (which is by the private sector under a concessioning charged on the producer price). Since the re- agreement. Since then Pemex has been supply- tail price is fixed to control inflation and the ing LPG for high-use vehicles. Six main and fotur producer price changes according to market provisional LPG fueling stations were con- conditions, the specific tax is adjusted each stmicted to service about 8,000 vehicles a day. In month so that the desired retail price is obtained. addition, most commercial companies, which Since January 1995 each state can collect a converted their fleet from petroleum-based fu- charge on fuel prices based on environmental els to LPG, have their own storage and fueling considerations to finance specific pollution con- facilities. About 27,000 buses and trucks have trol measures (O'Ryan 1996). been now converted to LPG. In addition, about Before November 1991 the retail prices of 1,300 LPG-fueled microbuses equipped with Nova and Magna Sin were set at $0.24 a liter three-way catalytic converters are circulating and $0.33 a liter, respectively. The high price for (Sanchez 1996). More than 200 conversion kit Magna Sin was intended to restrict its use be- models have been certified. Each fuel conver- cause it was in limited supply. However, the large sion and catalytic converter installation costs to price difference between the two gasoline grades vehicle owners between $2,000 and $3,000 (Viola encouraged misfueling of catalyst-equipped ve- 1996). Vehicles that have converted to LPG and hicles with leaded gasoline, which impairs pol- installed a catalytic converter are exempt from lution control efficiency. In November 1991 the the circulation ban in the MCMA. government raised the prices of Nova by 55 per- Use of CNG in the MCMA was initiated in 1992 cent and Magna Sin by 25 percent, bringing them through a pilot program. This program included to $0.37 a liter and $0.42 a liter, respectively. In an additional supply of 2.5 million cubic meters April 1992, the price of Magna Sin was lowered to of natural gas by Pemex, enough to fuel about $0.40 a liter. InJanuary 1995 a uniform gasoline 45,000 high-use vehicles; rehabilitation of an ex- surcharge of $0.01 a liter was introduced in the isting service station owned by the Mexican Pe- MCMA. The funds generated by this surcharge troleum Institute and construction of a new are channeled to an Environmental Tmist Fund service station for CNG fueling in the northern to finance CMPCCA's programs."' Since May 1996 section of the city; and retrofitting of forty-seven the surcharge has been increased and differenti- private and commercial vehicles, microbuses, and police cars for CNG. An LNG plant was also constructed to fuel 10. Since its initiationi iniJanuary 1995 the gasoline sir- buses and tmicks. However, the new plant could chiarge program has raised more than $11 million for the not be operated because of legal and financial Erivironimental Trust Ftunid. 148 Chlapter 4 ated by gasoline type. The surcharge was raised ciations, car manufacturers, and the Na- from $0.01 to $0.04 for Nova and to $0.02 for tional Bank of Public Works and Services. Magna Sin, reducing the price difference be- In the Federal District the program sought tween unleaded and leaded gasoline grades. to replace all pre-1985 taxis over a two-year Magna Sin is now only 2.4 percent more expen- period with newer vehicles meeting 1991 sive than Nova (Menendez 1996). Retail prices emission standards (or 1993 standards if ve- of Magna Sin and Nova in the MCMA since 1990 hicles are replaced after October 1991). In are shown in Figure 4.5. the State of Mexico the program sought to The fuel taxing policy used to result in higher replace all pre-1982 model-year taxis with prices for CNG than gasoline, which discouraged 1982 or later model-year vehicles. By De- use ofCNGbymotorvehicles. However, recently cember 1993 about 47,000 taxis had been the policy was revised to maintain the CNG price replaced by new model vehicles equipped at 35 percent of the gasoline price. with catalytic converters. * A Federal District regulation required that Transport management. The modernization of all pre-1977 model-year trucks be replaced. high-use vehicles (taxis, trucks, buses, micro- About 15,000 trucks were replaced by emis- buses, and vans) has been an important compo- sion-controlled vehicles meeting the new nent of traffic management policy in the MCMA. emission standards under financing by the World Bank. Priority was given to replace- * In 1990 taxis were estimated to emit high ment by LPG- and CNG-fueled trucks. levels of pollutants per passenger-kilome- * Between 1991 and 1995 about 4,100 Ruta- ter traveled. The taxi fleet was old, with 60 100 buses were retrofitted with new engines percent of vehicles more than ten years old to reduce emissions. About 75 percent of and 93 percent more than six years old. The these buses were operating during week- taxi modernization program was initiated in days and 50 percent on Sundays and holi- 1991 based on an agreement among the days. The remaining ones were being kept Federal District, State of Mexico, taxi asso- for maintenance. With establishment of the Figure 4.5 Prices of Magna Sin and Nova in the MCMA, 1990-96 Pesos per liter 3.0 2.5- 2.0- 1.5 - Magna Sint 1.0 - 0.5 ~gwV Nova 0.0 1990 1991 1992 1993 1994 1995 1996 Source: Benmmen 1996. Case Studies: Mexico Cit 149 Permanent Program for Correcting Exces- duced a new program called doble no circula sive Black Smoke in early 1992, Ruta-100 (don't drive for two days), which is used when buses exceeding the 40-Hartridge-unit stan- emergency air pollution levels (above IMECA dard were taken from service for revision. 250) are reached. When in place, the program Fuel injection pumps were also recalibrated prohibits 40 percent of the vehicles from circu- and injectors were replaced to reduce other lation. Buses, low-emission vehicles, and trmcks air emissions from Ruta-100 buses. Since equipped with an oxidation catalyst are ex- Routa-100 buses ceased operation in 1995, empted from the ban. As a result of this mea- private sector will proxide this service in the sure about 1,000 diesel-fueled heavy-duty vehicles future. Among the former 2,500 Routa-1)0 have been retrofitted (Viola 1996). The emer- buses, only those conmplying with the 1994 gency measure has generated a quite negative emission standards will be sold to the pri- public opinion. vate sector for circulation in the Federal The following measures have been imple- District and the rest of the MCMA. mented in the 1990s for rail-based mass trans- * By December 1993 there were about 10,000 port vehicles in the MCMA: microbuses replaced with vehicles equipped with catalytic converters in the MCMA. * Between 1989 and 1994 the metro system * In 1993 vans equipped with catalytic con- was extended by 37 kilometers to provide verters were substituted for more than services on a line that runs from Iztapalapa 27,000 old polluting vans. In 1994 there to Garibaldi. Metro construction on Line were an average of 400 substitutions a IOis also underway. This new line will pro- month. vide transport services from Guerrero in downtown Mexico City to the municipality The 1995 economic crisis in Mexico adversely of Catepec in the State of Mexico. affected the vehicle modernization program * Modernization of the Federal District because high interest rates reduced demand for trolleybus fleetwas carried out in 1991-94. financing and development banks discontinued * Twelve new trains equipped with modern credit programs. The vehicle modernization technology were purchased for the light program is being revitalized by emphasizing at- train system (CMPCCA 1995a). tractive financing conditions for replacement of microbuses by large buses. To reduce congestion in the MCMA, 200 kilo- Another component of the transport manage- meters of roads (including the beltway and main ment policy in the MCMA includes circulation interior roads) were modernized. Lanes of ac- bans. One type of circulation ban is based on cess roads from the city center to five different the model-year of certain vehicle classes. For highways were widened, and nine bridges on example, since 1992 circulation of taxis oldler major roads and eight underpasses on Line 8 of than model-year 1986 and microbuses older than the metro system were constructed. Further- model-year 1984 have been prohibited in the more, between 1990 and 1994,1.7 million square Federal District. In addition, in November 1989 meters of road surface were paved in an effort an emergency measure for traffic management to reduce dust levels. was implemented in the MCMA. Called how no circula (don't drive today), it prohibited circula- Air quality monitoring. Air quality monitoring in tion of vehicles in the MCMA one workday of the MCMA began in the mid-1970s with mea- the week based on the last digit of the vehicle's surement of TSP and SO2. At that time the main license plate. The program was designed to re- air quality problems were perceived to be dust duce gasoline consumption, traffic congestion, storms and combustion of high sulfur fuels at and emission of pollutants from mobile sour-ces. industrial sources. Since 1986 air quality moni- Initially, all vehicles except high-use vehicles, fire toring has been conducted by manual and auto- trucks, ambulances, and police cars were prohib- matic networks. Both networks were designed ited from circulation one workday per week. In and operated to conform with criteria estab- late 1991 the program was extended to include lished by the WHO, and the U.S. and German vans, taxis, and microbuses. Circulation of taxis environmental protection agencies. was also prohibited on alternate Saturdays. Until recently the automatic network con- In December 1995 the government intro- sisted of twenty-five stations in five different sec- 150 Clhapter 4 tions of the MCMA (southwest, northwest, south- Evaluation of Implemented Measures east, northeast, and center). These stations con- tinuously monitor ozone (in sixteen stations), Despite a fast growing vehicle fleet, the urban air SO2 (in seventeen stations), NO, (in ten stations), quality management strategy in the MCMA has CO (in seventeen stations), PM-1 0 (in eleven sta- effectively reduced ambient concentrations of tions), and atmospheric parameters such as tem- some pollutants (such as lead, SO2, and CO) and perature, wind direction and speed, and relative somewhat stabilized the upward trend of ambi- humidity (in ten stations). These stations are con- ent concentrations of other pollutants (such as nected to a central data processing facility. Seven ozone and PM-10). Successful formulation and new automatic stations were recently added to implementation of this strategy can be attributed this network, bringing the total number to thirty- to a coordinated effort that has involved govern- two. The automatic air monitoring network is veri- ment institutions representing the variousjuris- fied every six months by the USEPA. dictions in the MCMA, research agencies, and The manual system consists of nineteen sta- the nongovernmental community. tions that monitor ambient TSP and PM-10 con- centrations. The TSP samples are also analyzed Vehicle emission standards. As a result of exhaust for such parameters as nitrates, sulfates, lead, emission standards, all new vehicles that have and other heavy metals. According to criteria entered into traffic in the MCMA since 1991 are established by USEPA standards, samples are equipped with catalytic converters. These newer collected once every six days for a 24-hour pe- vehicles constitute about 32 percent of road- riod. During winter, however, sampling frequency based motor vehicles in the MCMA. Installation increases to once every three days. of catalytic converters has enabled a 90 to 95 per- Ambient concentrations of HC are measured cent reduction in CO and HC emissions and a 60 through special programs jointly conducted by to 65 percent in NO, emissions from gasoline- Pemex and the USEPA. In addition, there are fueled cars, taxis, vans, and microbuses (see Table two mobile air monitoring units and a remote 4.3). detection system used for research and siting of The Mexican emission standards for new ve- new air monitoring stations. hicles, however, are less stringent than those in Air quality data in the MCMA are expressed as effect in the United States, and this gap may widen IMECA values (Tables 4.14 and 4.15). IMECA unless standards are tightened to keep pace with values are calculated from hourly averages of pol- technological advances (Walsh 1996). Emission lutant concentrations using a scale from zero to standards for in-use gasoline-fueled vehicles with 500, with a reference value of 100 for the air qual- model-year 1994 and older are relatively lenient ity standard of the pollutant. These valutes are compared with those in the United States. displayed at electronic billboards located next to Emission tests conducted on new vehicles at principal roads of the city and transmitted by ra- sea level would not yield the same results at high- dio stations. They are also published every day by altitude locations like the MCMA. Because the local newspapers and in monthly and yearly re- MCMA does not have a vehicle testing labora- ports of the CMPCCA. tory, new vehicles or engines are not certified Table 4.14 IMECA values for the MCMA PoJlluttant (lime-rveighled average) (~~~) VNO 03 SO TVi I'Ml-J0 1iA11iE(A (8-hour) ( 1-hour) (1-hotur) (24-houzr) (24-hour) (24-hour) 3 3 (Pl3) vailuie (mng/rn ) (pg'/rn?') (pg/m ) (pg/r ) (pg/rn ) (pg/rn') 100 12.6 395 220 338 260 150 200 23.8 1,236 465 695 445 350 300 34.9 2,077 710 1,469 630 420 500 57.2 :3,760 1,200 2,600 1,000 600 Sourre: DDF 1997b. Case Studies: Mexico City 151 Table 4.15 Significance of the IMECA values IMECA value Air queality Humantn /heailIh e/f rts 0-100 Satisfactory A favorable situation for physical activities 101-200 Unsatisfactory Some anniovanice to sensible people 201-300 Bad Increased aninoiyance and ititolerance for people who exercise and have somle respiratory problems 301-500 Very bad General increase in intolerance among the public Sioure: LANL and IMP 1994. for emissions. Testing for certification of new ve- Based on test results at government-owned sta- hicles and engines is carried out by manufactuir- tions, the modernization of the inspection sys- ers with some oversight by government officials. tem resulted in a 61 percent reduction of CO Occasional certification tests are carried out at emissions and a 71 percent in HC emissions the Mexican Petroleum Institute's laboratory at (CMPCCA 1995a). the request of either government or industry A noncomputerized inspection and mainte- (Walsh 1996). nance system, such as the one used by private Exempting low-emission vehicles from the stations and the previously government-owned one-day or two-day driving ban provides a good stations in the MCMA, leaves the pass-fail deci- incentive to have only clean vehicles for use on sion for each vehicle up to the individual me- high pollution days. This measure may also chanic, who must read the results and compare stimulate the modernization of the fleet by en- them with emission standards. Based on the U.S. couraging people to replace their old, high-pol- experience, such a design was associated with a luting vehicles with new ones that could qualify high percentage of improper inspections, ren- for the exemption. Emission standards for ex- dering the periodic inspection and maintenance emption needs to be announced sufficiently in program ineffective. Anecdotal evidence and advance to allow vehicle owners to adjust their limited data available on failure rates in the behavior accordingly. Mexican program suggest that the same is true for the MCMA. Vehicle inspection programs. During the second Procedures for vehicle inspection payments half of 1993 about 2.8 million vehicles were in- differed for government-operated stations and spected at private and government-operated in- private stations. For testing at government- spection stations. Of these vehicles about 2.3 operated stations, vehicle owners paid the fee percent were barred indefinitely from circula- at one of the forty-one treasury offices located in tion. Among vehicles passing the test, almost 25 the MCMA and presented the receipt at the test- percent needed maintenance and a second in- ing station. Cash payments were made directly at spection within six months in order to comply private inspection stations. As such, testing at the with emission standards. For those vehicles per- government-operated stations was less time-con- mitted to circulate, 71 percent were private cars, suming and more convenient. 8 percent were public transport vehicles, 18 per- Government-operated and private inspection cent were freight vehicles, and 3 percent were stations also provided different type of services. emergency vehicles, official vehicles, or motor- Government-operated stations did not provide cycles. The average age of these vehicles was 8.5 repair services, but private stations did. So, ve- years (CMPCCA 1995a). hicles that failed the emissions test at government- The computerized system motivated many operated stations had to be taken to private service drivers to take their vehicles to maintenance stations for repair. Based on a survey of service before inspection. The fail rate in the first in- receivers and inspection technicians in early spection of vehicles was 42 percent during the 1990s, government-operated stations were per- first six months after the introduction of the ceived as being stricter and more straightforward computerized inspection system in 1993. This than the private service stations. With the mod- rate dropped to 26 percent at the end of 1994. ernization of the government-operated inspec- 152 Chapter 4 tion stations, human intervention in determin- rels a day in mid-1996 and accounts for 48 per- ing test results has been eliminated. cent of the gasoline consumed in the MCMA. Federal District Department's recent decision Magna Sin accounts for a higher percentage of to conduct periodic vehicle inspections at high- total gasoline sales in wealthier Mexican cities volume automatic inspection stations is based on (88 percent in Monterrey) and in the northern the above-mentioned factors. This scheme would border area (98 percent; Berumen 1996). eliminate many of the problems encountered at Because of the reduction of lead levels in private inspection stations. Considering previous leaded gasoline and higher market share of un- private inspection station owners in bids would leaded gasoline, the lead concentration in am- somewhat compensate those who lost their work bient air in the MCMA has decreased almost 90 and would also increase acceptability of the percent during the past five years. Because there government's decision. is less lead in ambient air, the average lead level Emission testing of the actual motor vehicle in blood of children examined in the MCMA fleet in the MCMA will allow the government to fell by 66 percent between 1991 and 1994 (from establish realistic standards for in-use vehicles 17.5 pg/dl to 6.0 pig/dl of blood (CMPCCA and for the one-day and two-day driving ban pro- 1995a). grams. This approach will be more effective Large price differences between Magna Sin than adopting the in-use vehicle emission stan- and Nova in the early 1990s caused persistent dards used in industrial countries, where motor misfueling of vehicles in the MCMA (that is, vehicle fleets are likely to have substantially dif- Nova was used in vehicles equipped with cata- ferent characteristics, emission rates, and driv- lytic converters). Frequent supply problems with ing patterns. Magna Sin also resulted in misfueling. Misfueling The periodic inspection and maintenance damages the catalyst and oxygen sensor in cata- program in the MCMA has increased public lytic converter-equipped vehicles, permanently awareness about the contribution of vehicle increasing emissions. A survey carried out be- emissions to air pollution and the importance tweenJanuary and March 1992 on 500 taxis and of vehicle maintenance for curtailing these emis- buses showed that 95 percent had misfueled. sions. As a result ambient CO concentrations Another survey carried out that year by indepen- decreased by 68 percent between 1991 and 1994. dent consultants detected lead in the tailpipes Similarly, there were twenty-three air quality vio- of 12 percent of the catalyst-equipped private lations for CO registered in 1992, but none in vehicles and 13 percent of the public transport 1993 or 1994 (CMPCCA 1995a). vehicles sampled (O'Ryan 1996). To reduce Through the roadside inspection program, an misfueling, a sufficient supply of Magna Sin was average of 2,700 vehicles are fined and 700 ve- maintained to meet demand in the MCMA and hicles are barred from circulation every month. the price difference between Magna Sin and About 73 percent of these vehicles are trmcks Nova was reduced. The use of differential sur- and buses (CMPCCA 1995a). charges also helped reduce the retail price gap to 2.4 percent in mid-1996. Fuel-targeted measures. Between 1986 and 1991 These surcharges provide revenue for the the lead content of Nova was reduced by about Environmental Trust Fund to implement pollu- 45 percent. This was achieved mainly by adding tion control projects with little or no financial MTBE to boost the octane number of the gaso- recuperation. This arrangement illustrates a nice line. Since 1992 the lead content of Nova has example of the "polluter pays" concept: those been lower than the maximum allowable lead who pollute the environment by using fuels, es- content in Mexico (0.08 g/liter) and the Euro- pecially the dirtier one, should pay for environ- pean Union countries that still use leaded gaso- mental improvement. At the start-up of the line (0.40 g/literin Portugal and 0.15 g/liter in Environmental Trmst Fund in 1995, one project other European Union countries). being considered by CMPCCA involved financ- After Magna Sin was introduced in Mexico to ing the installation of vapor recovery systems in meet the fuel requirements of catalytic converter- gasoline stations. Another project recently an- equipped vehicles, consumption of Nova in the nounced involves substituting microbuses for MCMA decreased from 90,000 barrels a day in buses. The TrLst Fund will provide financial guar- 1989 to 56,000 barrels a day in mid-1996. Con- antees to individual owners so that commercial sumption of Magna Sin increased to 52,000 bar- banks will consider them creditworthy. The goal Case Studies: Mexico City 153 Table 4.16 Typical fuel quality parameters of Magna Sin and Nova, 1995 fagna Sin Nova Futel pfirameters Quality Stndalrd a Quality Standard a Reid vapor pressure (psi) 7.9-8.2 6.5-8.5 8.1 6.5-8.5 Aromatics (percent by volume) 24.7-28.6 30 22.4 30 Olefins (percent by volume) 8.3-12.3 15"1 8.5 15 Benzene (percent by volume) 0.9-1.2 2 1.2 2 Oxygen (percent by weight) 0.92-1.139 1-2 0.78 1-2 Sulfur (percent by weight) 0.035-0.070 0.10 0.07 0.15 Lead (g/liter) 0.00017 0.0026 0.03 0.06-0.08 a. The lead standard for Magna Sin and the aromatics. olefitis, benzene, and sulfuir standards for Magna Sin and Nova specify the maximum values. b. Afterjanutary 1988. the maximum permissible valtue will be 12.5 percent by volume. S,,rUe: Berumen 1996; DDF 1996. of this project is to improve the public transport (0.070 percent by weight) is higher than in the senrice by offering higher capacity vehicles. Fi- United States (1990 industry average of 0.0338 nancing of educational campaigns is also being percent by weight). Unleaded gasoline with a considered (Menendez 1996). high sulfur content reduces the efficiency of At a retail price for Nova of $0.373 a liter, the catalytic converters and promotes higher emis- new fuel taxing policy would result in a retail sions of CO and ozone precursors (HC and price for CNG of $0.21 a liter (including a comn- NO,,). pression cost of $0.08 a liter) on a gasoline- Studies conducted by the Mexican Petroleum energy equivalent basis. This price differential Institute indicate that leaded gasoline contain- should promote CNG-conversion of gasoline- ing 5 percent MTBE by volume reduces exhaust ffueled vehicles, especially those that travel ex- emissions from noncatalytic gasoline-fueled ve- tensively (such as gasoline-fueled microbuses hicles by about 15 percent for CO and 11 per- and taxis). However, assuming an annual con- cent for HC without increasing NO, emissions sumption of 300,000 liter of diesel fuel per btus, (CMPCCA 1995a). However, Bravo and others the taxing policy would not provide sufficient in- (1991) found that adding 5 percent MTBE to centive for a private investor to spend an addi- gasoline increased ambient peak concentrations tional $10,000 to $15,000 for a new CNG-fueled of formaldehyde in the southwestern section of bus instead of a new diesel-fueled-bus. This as- the MCMA and shifted the peak from I 1-12A.M. sessment is based on a retail price of diesel fiuel to 10-11 A.M. (Figure 4.6). A slight increase in of $0.292 aliterfor diesel fuel and taking account ambient ozone concentrations in the morning 11 percent greater fuel efficiency of diesel fuel and evening hours was also observed after add- compared to gasoline and 10 percent greater ing MTBE (Figure 4.7)." Further ambient air engine efficiency for diesel engines than CNG measurements in March 1992 indicate that form- engines. Use of CNG instead diesel fuel can be aldehyde concentrations were as high as 33 ppb promoted by imposing a surcharge on diesel fuel. in the northwestern section (Tlalnepantla) and In 1995 typical Magna Sin and Nova fuel pa- 18 ppb at the center (Merced) of the MCMA rameters all complied with the established stan- (LANL and IMP 1994). Because Magna Sin has dards, except for oxygen content (Table 4.1(6). been reformulated since 1992, the effect on Nevertheless, the Reid vapor pressure of gaso- fuel's volatile and reactive HC emissions (which line marketed in Mexico is higher than in the southern United States (a maximum of 7.2 psi where use of reformulated gasoline is required and a maximum of 7.0 psi in the Los Angeles 11. Sampling was condtucted on 29 weekdays duiring the Metropolitan Area). Further reduction of tin- December 1998-Febmiary 1989 period wheni no MTBE was added to gasolinie, and on 56 weekdays duiring the Decem- leaded gasoline volatility would help redulce ber 1989-Febniary 1990 period when gasoline containied 5 ambient ozone levels in the MCMA. In addition, percent MTBE. Meteorological conditions during both sam- the sulfur content of the Magna Sin in Mexico plin.g periods wvere similar (Bravo and others 1991). 154 Chapter 4 Figure 4.6 Ambient concentrations of formaldehyde in the MCMA before and after adding MTBE Parts per billion 25 20- Before MTBE 15 - 10 5 0 10-11 11-12 12-13 13-14 14-15 Period of sampling (hour of the day) Smorce: Bravo and others 1991. Figure 4.7 Ambient concentrations of ozone in the MCMA before and after adding MTBE Parts per billion 160 140 - 120 - e E 100 80 - Before MTBE 0 1 3 5 7 9 11 13 15 17 19 21 23 Period of sampling (hour of the day) Source: Bravo and others 1991. Case Studies: Mexico City 155 contribute to formaldehyde and ozone concen- reasons. First, the National Institute of Ecology trations in ambient air) caused by adding 5 per- did not announce the mandatory requirement of cent oxygenates cannot be predicted without installing a vapor recovery system until Septem- evaluation of the entire fuel matrix. ber 1995, although it was decided in October Diesel fuel quality parameters have complied 1994. Second, the initial decision to provide non- with the Mexican standard. 'The typical sulfuir refundable funds for the installation of vapor re- content of Special Diesel sold in the MCMA in covery systems was replaced by the decision to 1995 was 0.041 percent by weight. Elimination provide interest-free loans. This decision was ve- of the regular diesel fuel throughout Mexico (in hemently opposed by gasoline station owners, February 1997) must have eliminated higher SO, who argLed that the low profit margin was not emissions caused by misfueled vehicles circulat- sufficient to pay back the credit. It was only when ingin the MCMA orintercityvehicles fueled with CMPCCA intervened and threatened to shut regular diesel fuel entering the MCMA. In 1995 down all the gas stations not equipped with vapor Diesel Sin had a cetane index of between 55.0 recovery systems under the smog alert mandatory and 59.0, and an aromatics content of 25.9 to 28.1 shut-down program that some gas stations applied percent by volume (DDF 1996). The quality of for the credit. The Finance and Public Credit Diesel Sin is similar to diesel fuel in the United Secretariat has recently mandated interest pay- States and better than the average diesel fuel in ments on the credits provided by the Environ- Europe. Further reformulation of diesel fuel mental Trmst Fund. This is expected to further should focus on reducing its aromatics content. delay adoption of the vapor recovery project. This would reduce PM emissions in the form of soot, whiclh has carcinogenic effects, reduces Nis- Transport management. Modernization of the mo- ibility, and causes soiling. tor vehicle fleet has reduced pollutant emissions Limited data from testing of five in-use light- by replacing some of the old, heavily polluting duty trucks converted to LPG indicate that none in-use vehicles in the MCMA. For example, the of the five vehicles met the emission standards to taxi replacement program to meet the 1991 stan- which they were originally certified. Conversion dards is estimated to have reduced average emis- efficiencies were less than 20 percent for NO, sions in the replaced vehicles by 76 percent for (for four of these vehicles) and 50 to 80 percent CO, 60 percent for HC, and 9 percent for NON. for CO and HC. The low NO, efficiency suggests The average reduction in emissions from taxi improper converter installation (Walsh 1996). In replacement to meet the 1993 standards is esti- addition, although LPG is a more desirable mo- mated to be 96 percent for CO, 84 percent for tor vehicle fuel than gasoline or diesel fuel, it is HC, and 60 percent for NO.. The replacement not totally environmentally benign because its of old gasoline-fueled trucks with newer trucks photochemical components contribute to ozone meeting the 1993 standards is estimated to have formation. Based on preliminary estimates, LPG lowered average emissions for replaced vehicles losses during handling, distribution, and use are by 85 percent for NO., 81 percent for CO, and responsible for 20 to 30 percent of the ozone in 49 percent for PM (World Bank 1992). Replac- the MCMA. Asolution to this problem maybe to ing engines in about 4,100 Ruta-100 buses re- reformulate LPG by reducing its butane and lbu- duced emissions of black smoke and eliminated tylene content (DDF 1996). 41,000 tons of air pollutants. Inspection and The gasoline vapor recovery program is ex- maintenance of all Ruta-100 buses further re- pected to reduce HC emissions by 29,000 tons duced black smoke emissions. Adjusting and to 36,000 tons a year. This would also reduce the replacing the injection system in these buses re- formation of ozone in ambient air and recover duced their emissions by 67 percent for NO., 50 gasoline, which would otherwise be lost to the percent for HC, 25 percent for CO, and 14 per- atmosphere. The Environmental Trust Fund was cent for CO,. to be used to finance the installation of a vapor Despite the modernization efforts for these recovery system in the gasoline stations operat- vehicles, no action has been taken for the old, ing in the MCMA. Despite collection of more than highly polluting cars circulating in the MCMA. A $11 million, the vapor recovery system has been vehicle taxing policy disfavoring older cars in the installed in only nine of the MCMA's 360 gaso- MCMA would create an incentive to export these line stations. Significant delays in using the Envi- vehicles outside the MCMA. Because this policy ronmental Trust Fund have occurred for two may be politically difficult to implement, a 156 Chlapter 4 scrappage program may be considered instead. emissions by up to 50 percent (through oxida- Such a program, however, would bring a cost bur- tion of the soluble organic fraction of PM; Viola den to the government and also create a demand 1996). Implementation of the driving ban pro- for older vehicles in the local market. This de- grams is expected to reduce annual emissions mand may be controlled by prohibiting importa- from the most polluting vehicles by 71,100 tons tion of older vehicles to the MCMA. for CO, 5,250 tons for HC, and 1,350 tons for Circulation bans in the MCMA have had mixed NO, (DDF 1996). results. Banning older microbuses and taxis from Other transport management measures in the circulation has eliminated emission of pollutants MCMA have had the following effects: from these highly polluting sources of the fleet in the Federal District. Despite its initial success * Opening the newly constructed 37 kilome- based on reduced traffic congestion and curtailed ters of Line 8 of the metro system has pro- air pollution, however, the one-day driving ban vided transport services to an additional 300 program did not achieve the desired results be- million passengers a year, eliminating about cause households circumvented the ban by buy- 100,000 vehicles a day from traffic and re- ing a second car. According a recent study, 22 ducing pollutant emissions from mobile percent of vehicle owners in the MCMA pur- sources by about 22,000 tons. chased a second vehiclejust to bypass the require- * Constructing Line 10 of the metro system is ment (CMPCCA 1995b). These included 170,000 expected to replace 230 buses, 1,160 inexpensive old vehicles brought in from sur- microbuses and 18,000 private cars a day. rounding regions were purchased, mostly by This substitution is expected to reduce pol- households having more driver's licenses than lution in the MCMA by 4 percent. cars. The acquired second car was not only used * Renovating trolleybuses and purchasing fif- as a one-day replacement for the principal ve- teen new trains has improved public trans- hicle but was also driven during the other (per- port services in the MCMA. Each train can mitted) days of the week by other members of the carry about 350 people and travel at an aver- family. As a result the total kilometers traveled by age speed of 50 kilometers an hour. Two many households increased. Accelerated gaso- trains can be coupled at peak hours line consumption and increased air pollution to provide additional services. With the after this driving ban suggest that the intended introduction of these trains, the wait for results were not achieved. trains at rush hours was reduced to three The Mexican government has not entirely minutes. given-up the one-day driving ban, however, be- * Modernizing the road network should have cause it is estimated that eliminating this pro- reduced congestion, although no quanti- gram will increase daily gasoline consumption tative data are available (CMPCCA 1995a). by 132,000 liters and increase daily vehicle cir- culation by 385,000, resulting in a 756-ton in- Privatization of Ruta-100 buses is expected to crease in daily pollutant emissions. Thus the provide a more efficient mass transport service government is trying to use both the one-lay and to the public. This would attract some additional two-day ban programs as an incentive to mod- riders who currently drive private cars, resulting ernize the fleet, switch to cleaner fuels, and in- in a net reduction in pollutant emissions. stall pollution control equipment. For example, since 1992 all freight, passenger, and service ve- Air quality monitoring. The automatic air quality hicles that have switched to CNG or LPG are monitoring network in the MCMA is one of the exempt from the one-day driving ban. With the most sophisticated in the world in terms of cover- introdtuction of the two-day driving ban, there is age and quality. This network now includes thirty- an incentive for vehicles manufactuired between two automatic stations supported by nineteen 1980 and 1990 to retrofit with catalytic convert- manual air monitoring stations and can provide ers. For example, installation of an oxidation cata- hourly air quality information to the public lyst on in-use trucks, which exempts these vehicles through billboards placed next to major roads from the ban, would reduce their average CO and radio stations. USEPA verifies the network and HC emissions by up to 70 percent and PM twice a year. SANTIAGO IBRD 28164 .i,0> 1 o*o v i SANTA ISABEL Re h i i @o \; FRANCO S MAIRIPORA , . -vd v 'I ~~DA ROCHA... this snp pmoordoced &yl the MapDe, CAJAMAR ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ntof Th. W.1orl.ok. I The boundaries, colors. clsoninations and any other inbonatio shams GUARULHOSARUAI on this OS do not The World Bank Grasp, sny judgmet on the Iv~a/ f f~~~~~~~~~~~~~~~~~~~~~~~~~~tatus of any territoryp ':Z-Sont.. d~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~o an .y endo,se,nent Re0- --pt. f smr _ houndr.s. PAULO~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~A AL SA ~ ~ ~ ERA/. UACECEUn EMUGA4n omim Case Studies: Sao Paulo 185 Sao Paulo, the largest city in Brazil, contains 9.5 was exceeded at these six stations. The frequency million people. The Sao Paulo Metropolitan of violation ranged from 2 percent of the sam- Region (SPMR), the second largest urban cen- pling days at Penha and 20 percent at Osasco. ter in Latin America and the Caribbean (after The 24-hour "alert" level of 625 jig/m3 was ex- the Mexico City Metropolitan Area), contains ceeded once at Parque Ibirapuera with a con- about 17 million people. The SPMR is an im- centration of 685 jg/mr3 and the 24-hour portant economic center that generates about "attention" level of 375 jg/mr3 was exceeded half of Brazil's gross national product (CETESB once at S.B. do Campo with a concentration of 1996). It spans 8,051 square kilometers and in- 387 gg/m3(CETESB 1996). cludes thirty-eight municipalities besides the In 1995 annual average ambient PM-10 con- Municipality of Sao Paulo (which occupies 1,577 centrations in the SPMR, monitored by twenty- square kilometers). The SPMR's population den- one monitoring stations, ranged from 60 jg/M3 sity averages 2,110 inhabitants per square kilo- (at Maud) to 105 jig/m3 (at Guarulhos). The meter, rising to an average of 6,380 in the city annual standard of 50 jig/m3 was exceeded at and reaching as high as 10,000 in certain down- all twenty-one stations. The highest 24-hour av- town areas. erage concentrations, which ranged between 184 Ecologically, the surface of the SPMR consists jig/m3 (at Cerqueira Cesar) and 298 jg/m3 (at of urbanized areas, forests, natural vegetation, Cambuci), also exceeded the 24-hour standard and a hydrological system consisting of three riv- (150 jg/m3) at all twenty-one stations. At ers. The SPMR lies at altitudes ranging from 650 Cambuci and Osasco violations of the 24-hour meters to 1,200 meters. The SPMR's climate is standard occurred on 16 and 17 percent of the moderate, with dry and cold winters (averaging sampling days. The 24-hour "attention" level of 8°C) and hot and humid summers (averaging 250 jig/m3 was exceeded at S.B. do Campo five 300C). Annual precipitation averages 1,500 mil- times, and at Cambuci and Parque Ibirapuera limeters, with most rainfall occurring during once each (CETESB 1996). October and March. Because of thermal inver- In 1995 two of the seven stations monitoring sions, ambient air quality in Sao Paulo worsens smoke in the SPMR recorded average ambient during the winter (May through September). levels above the annual standard of 60 jg/m3 Sao Paulo's rapid industrialization and urban- (103 jig/m3 at Campos Eliseos and 68 jg/mr at ization have caused serious environmental prob- Tatuape). The 24-hour standard of 150 jg/mr lems.2' Federal and state authorities began was exceeded at five of these stations, with the addressing industrial pollution during the 1970s, highest concentration being 245 jg/m3 at Cam- with some success. Air pollution from motor ve- pos Eliseos. At this station the 24-hour standard hicles is now the environmental priority for the was exceeded on 12 percent of the sampling days SPMR. (CETESB 1996). The latest ambient NO2 data for the SPMR are from 1993 when average concentrations were Ambient Air Quality below the annual standard of 100 jg/mr. An- nual average concentrations were 99 jg/m3 at Pollutants that exceed the ambient air quality Congonhas, 85 jg/m3atParque D. Pedro II, and standards in the SPMR are TSP, PM-10, smoke, 62 jig/rn3 at Cerqueira Cesar. However, the 1- NO., ozone, and CO. In 1995 annual average hour standard of 320 jig/m3 was exceeded at ambient TSP concentrations ranged from 65 Congonhas (on 125 of the 214 monitored days, jg/m3 to 131 tg/m3. The annual standard of with a maximum 1-hour concentration of 784 80 gg/M3 was exceeded at six of nine monitor- jg/m3), Parque D. Pedro II (on 57 of the 158 ing stations: Osasco (131 gg/m3), Parque D. monitored days, with a maximum 1-hour con- Pedro II (1 16 jig/m3), S.B. do Campo (94 pg/ centration of 1,097 gg/m3), and Cerqueira Csar m3), Santa Amaro (85 jg/rn3), Sao C. do Sul (on 9 of the 180 monitored days, with a maxi- (85 jtg/M3i),andParque Ibirapuera (84 jg/m3). mum 1-hour concentration of 568 jg/m3). The 24-hour TSP standard (240 jtg/m3) also These concentrations were below the 1-hour "attention" level of 1,130 jtg/m3 (CETESB 1994b). 21. There are about 30,000 industries in the SPMR In 1995 the 1-hourozone standard of 160 jg/ (Romano and others 1992). m3 was exceeded three times at Mo6ca, thirty 186 Chlapter 4 times at Lapa, and twenty times at Parque D. 1.51 gg/m'sin 1987 (CETESB 1996).22 Ambient Pedro II. Maximum 1-hour ambient ozone con- concentrations of lead, which used to be a seri- centrations were 763 .tg/m3 at Lapa. 269 I.g/m3 ous environmental problem in the SPMR, have at Parque D. Pedro II, 218 Vg/m3at Mo6ca, and fallen to about 0.10 [.g/m2 as a result of the 131 1zg/m3 at Congonhas. The 1-hour "attention" Proalcohol program as well as the elimination level of 200 ng/m3 for ozone was exceeded on of lead from gasoline (Branco 1995). six days at Lapa and on one day each at Parque D. Pedro II and Mo6ca. The highest ozone lev- Sources of Pollutants els are observed during September and Octo- ber. Ozone levels are relatively low during the The main source of air pollution in the SPMR is winter (CETESB 1996). motor vehicles. In 1995 the fleet consisted of In 1995 ambient CO concentrations exceeded about 5.16 million vehicles (3.3 million gasohol- the 8-hour standard of 9 ppm (10 mg/1n3) at fueled light-duty vehicles, 1.5 million ethanol- four stations in the SPMR: Parque D. Pedro II fueled light-duty vehicles, and 360,000 on forty-two days (with a maximum of 18 ppm), diesel-fueled heavy-dutyvehicles). These vehicles Congonhas on thirty-four days (maximum 19 were responsible for 96 percent of CO, 90 per- ppm), Cerqueira Cesar on twenty-four days cent of HC, 97 percent of NOx, 86 percent of (maximum 17 ppm), and Centro on fourteen SO,, and 42 percent of PM emissions (Figure days (maximum 18 ppm). The 8-hour "atten- 4.13). The light-duty vehicles, which are mostly tion" level of 15 ppm was exceeded at Centro cars, contributed 68 percent of CO and 69 per- on five days, Congonhas on four days, Parque cent of HC emissions. Gasohol-fueled vehicles D. Pedro II on three days, and Cerqueira Cesar were responsible for higher emissions than etha- on one day (CETESB 1996). nol-fueled vehicles. In addition, diesel-fueled In 1995 ambient S02 concentrations in the heavy-duty vehicles contributed 82 percent of SPMR, monitored by twenty-nine stations, were NO, and 77 percent of SO2 emissions. below the annual (80 p.g/m3) and 24-hotur (365 In 1995 industry contributed 46 percent of all pg/m3) standards. The highest annual average PM emitted in the SPMR, but its estimated con- concentration (46 .tg/m2) was at Campos tribution to ambient PM-10 concentrations at Eliseos and the highest 24-hour average con- monitoring stations was only 10 percent mainly centration (179 pg/m3) was at Guarulhos because emissions from industrial stacks are dis- (CETESB 1996). persed before reaching the stations' receptors Average ambient concentrations of TSP, PM- (CETESB 1996). By contrast, street-level emis- 10, smoke, and SO2 in the SPMR for the 1983-95 sions may expose the public to very high PM-10 period are presented in Figure 4.12. During this concentrations. And because the PM in vehicle period annual average ambient concentrations emissions are toxic and within the inhalable dropped by 29 percent for TSP (from 125 .g/ range, they pose greater risks to the public than m3 to 89 ,ug/m3), 18 percent for smoke (from those generated by most industrial sources 83 tig/m3 to 68 jig/mi), and 69 percent for S02 (Szwarc 1993). In 1995 PM-10 in ambient airwas (from 61 .tg/m3 to 19 .tg/m3). These reductions contributed mostly by diesel-fueled vehicles (30 were the result of industrial pollution control percent), resuspended dust (25 percent), and measures and the Proalcohol Program, which aerosols (25 percent). The contribution of gaso- promoted the use of alcohol as a motor vehicle hol-fueled vehicles was much less (10 percent; fuel (see Chapter 3). Annual average PM-10 con- CETESB 1996). centrations, which fluctuated between 50 Itg/ The 1986-87 data from four monitoring sites m3 and 80 g.g/m' during 1983-95, have been in the SPMR show that PM-10 constitutes 49 to increasing since 1992. 61 percent of TSP and is largely contributed by No major changes were observed in ozone, motor vehicles and road dust (Table 4.32). The NO2, and CO levels for the 1983-95 period. data also show that fine particles (PM-2.5), which Although no recent data are available, monitor- are not regulated but constitute a major health ing in 1983-87 found high ambient concentra- tions of NMHC. For example, annual average NMHC ambient concentrations at Parque D. 22. Both concentratiorns are expressed as propane Pedro II increased from 0.74 gg/mi in 1983 to equivalenlt. Case Studies: Sao Paulo 187 Figure 4.12 Average ambient pollutant concentrations in the SPMR, 1983-95 TSP Micrograms per cubic metelT 140 120 100 80_ 60- 40- 20- 0 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 Micrograms per cubic meter PM-10 80 60 - 40 - 20 - 20 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 Micrograms per cubic meter Smiioke 100 80 60 40- 20 - 0 IIII I I 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 Micrograms per cubic meter SO2 40 _ \ 20 - 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 .So(rce: CETESB 1996. 188 Chapter 4 Figure 4.13 Share of air pollutant emissions from road-based motor vehicles and other sources in the SPMR, 1995 Percent of total emissions 100 ^ As . / oo ~.... ... . . CCtO~~~~~~~~~~~~~>k~~~~~~~~.. ~~~~~~~~~~~RENCN GUIANA~~~~~~~~~~~~~~~~~~~~~~...... Lt R % aIB>>l::tA~ I >>>> WSANTA LUZIA IBIRITEIBE6gpg$S:§ --- z BRAZIL PERtUf N E BO RA SIUA~ E~' f- CH E1"\) H~~~~For,zne/ iis~~~ 5.-$$$g$ BIT-PRA O - 3 4 5 / -*-NI INENTONALBONARE L)0 RUGUAY 6 95 44°05' _ - 7 4 0 This nsop .ros p-dJoduc by thRs MpDesign Unit ofThe Womldms Bonk, de mirsofions o nd, my d# ihur neofia sh,o,. on this map do not wpm,iMON poet f The Worl Boe Growp, ~ tats~ ony htrrilo.y or acceptanceCoNTsGc SABARA "I -5% BETIM . ) ~~~~BRAZIL 20'00' ~ ~ ~ ~ ~ /BELO HORIZONTE IBIRITE /- / ~~METROPOLITAN AREA . . ........-, AIR QUALITY MONITORING STATIONS / ~~~~~~~~~'~-N.- RIVERS ,5 ~~~~~~~~~~~~BUILT-UP AREAS \ I - - - ~~~~~~~~~MUNICIPAL BOUNDARIES 0 1 2 3 I INTERNATIONAL BOUNDARIES KILOMETERS AA4'5' 44 N.0 \ 43' 555 JULY 1997 Case Studies: Belo Horizonte 207 Belo Horizonte, the capital of the State of Minas 82) and Lagoa Santa. The highest monitored Gerais and third largest city in Brazil, had a popu- 24-hour TSP concentrations were 474 jig/m3 in lation of 2.06 million people in 1993. The Belo Belo Horizonte, 253 pig/m3 in Betim, 361 jig/ Horizonte Metropolitan Region (BHMR) com- ml in Contagem, 337 jig/m3 in Pedro Leopoldo, prises the city of Belo Horizonte and twenty-two and 521 jig/m3 in Vespasiano. Between 1984 and other municipalities. In 1993 the BHMR's popu- 1988 annual average TSP concentrations ex- lation was 3.57 million people, spread over an ceeded the 80 pg/M3 national and state standard area of 5,824 square kilometers (Almanaque in Betim (with a maximum annual average of Abril 1995). By 2000 the population of Belo 123 pig/M3 in 1987), Contagem (181 pg/m3 in Horizonteisexpectedtoreach2.95rmillionand 1985-86), Pedro Leopoldo (96 pig/mr in the BHMR 5.16 million. The BHMR generates 1991-92), and Vespasiano (118 pg/m3 in 1987). about half the gross domestic product of the Annual average TSP concentrations in Belo State of Minas Gerais and is considered Brazil's Horizonte and Lagoa Santa were in compliance third most important economic region (World with the standard (Table 4.43). Bank 1995). Ambient monitoring of TSP resumed in 1991 The BHMR is located at an altitude of about at five monitoring stations in Belo Horizonte, 850 meters in the foothills of the Serra do Curral Contagem, and Pedro Leopoldo. During a mountains which extend from southwest to the twelve-month period in 1991-92 ambient 24- northeast. Belo Horizonte is located in a lower hour TSP concentrations at all the stations ex- area, with well-demarcated openings through the ceeded the standard at least once and, in some hills to the municipalities of Contagem, Betim, cases, up to three times. The highest monitored and Ibirite to the west and the municipality of 24-hour TSP concentrations were 293 p1g/M3 Santa Luzia to the northeast. The municipali- at station 75 in Belo Horizonte, 604 pg/m3 at ties of Pedro Leopoldo and Vespasiano are lo- station 35, 561 pg/m3 at station 37, and 345 cated on an open plain in the northern part of pg/m3 at station 44 in Contagem, and 299 pg/ the BHMR (Leme Engenharia 1991). m3 at station 124 in Pedro Leopoldo (FEAM The BHMR has a subtropical climate, with 1992). Ambient TSP concentrations were high- monthly average temperatures of 19"C dur-ing est during the dry months Uune through Sep- the winter and 230C during the summer. The tember) and lowest during the rainy months annual precipitation, about 1,450 millimeters, (December through February; Liu, de Morais, falls almost entirely between December and and da Silveira 1992). The annual TSP standard March. Winds are predominantly from the east was exceeded in Contagem (107 pg/m3 at a and northeast, with an annual average speed of commercial site and 101 p1g/M3 at a residential 2.1 meters per second. Thermal inversions oc- site) and Pedro Leopoldo (96 pig/m3 at a resi- cur during the dry season (June through Sep- dential site). Annual average TSP concentra- tember; Leme Engenharia 1991). tions were high at another residential site in Contagem and at a curbside in Belo Horizonte, although they did not exceed the standard (see Ambient Air Quality Table 4.43). In a study conducted by the municipality of Ambient air quality data for the BHMR is lim- Belo Horizonte, ambient TSP and SO2 concen- ited. The available data show that ambient TSP trations were monitored at five stations on eight and PM-10 concentrations are in violation and different days in 1992. The 24-hour TSP stan- S02 concentrations are in compliance with dard was exceeded at two stations, with a con- Brazil's air quality standards. No ambient moni- centration of 253 p1g/M3 at Praca da Estaiao (the toring data is available for other air pollutants Station Square) and 243 pg/mr3 at a school site such as CO, NO9, and ozone. (Colegio Marconi). The geometric average of Between 1984 and 1988 ambient TSP concen- the 8-day TSP concentrations for these five sta- trations were monitored in Belo Horizonte, tions ranged from 80 jig/m3 to 156 plg/M3. Dur- Betim, Contagem, Lagoa Santa, Pedro Leopoldo ing the same period maximum 24-hour SO, and Vespasiano. On at least one day during this concentrations for the five stations remained period the 24-hour national and state TSP stan- below the national and state standard of 365 pg/ dard of 240 p1g/M3 was exceeded at all munici- m3 (ranging between 55 pg/m3 at the Colegio palities except Belo Horizonte (at stations 71 and Marconi and 292 tpg/m3 at the Station Square). 208 Chtapter 4 Table 4.43 Annual average TSP concentrations in the BHMR, 1985-92 (micrograms per cubic meter) Year Mnnihiparlit station 1985-86 1987 1988 1991-92 Belo Horizonte 43 68 - - - 71 49 52 41 - 75 - - - 80 82 - 63 58 - Betim 33 92 123 112 - Contagem 35 96 90 90 76 37 - - - Iola 44 - - - 107 45 181 - - - 63 - - 85 - Lagoa Santa 113 29b- - Pedro Leopoldo 123 97 98 82 - 124 - - - 96' Vespasiano 112 98 118 - - Not available. iVote: Except where indicated, the sampling petiod varies from ten to twelve months and the number of 24-hour sampling days varies from thirty-seven to fifty-two. a. Sampling period is nine months, and the nu1mber of 24-hour sampling days is twenty-eight. b. Data are for 1985. Sampling period is teni months, and the number of 24-hour sampling days is thirty-two. c. Sampling period is eight months, and the intmber of 24-hour sampling days is thirty. Soirre: Data derived from FEAM 1992. The arithmetic average of 24-hour SO2 concen- station was 276 jig/ml (Table 4.44). The 24-hour trations ranged from 19 pg/M3 to 38 pg/m3 for SO2 standard (365 pg/m3) was not exceeded at four of the stations but was 120 pig/m3 at the any of the stations. The maximum SO2 concen- Station Square (the national and state SO, stan- trations measured were 130 pg/M3 at Belo dard for the annual average is 80 pig/m3; Horizonte, 74 pg/mi at Contagem, and 87 jg/ Fundacao Christiano Ottoni 1992). At the Sta- m3 at Betim (FEAM 1996). tion Square contributors to the SO., levels in- clude emissions from buses, trucks, and trains. In 1995-96 ambient PM-10 and SO, concen- Sources of Pollutants trations were continuously monitored during various months at Belo Horizonte, Contagem, Air emission inventories for the BHMR, last com- and Betim. During the five months monitored piled in 1986 for fixed sources and 1991 for at the Belo Horizonte station, where air quality mobile sources (Table 4.45), indicate that indus- is affected by local traffic, the 24-hour PM-10 try and motor vehicles are the main emission standard (150 pg/m3) was exceeded once with sources (Leme Engenharia 1991). Since 1986 a concentration of 174 pg/mi. At the Betim sta- some industries have installed air pollution con- tion, located in an area remote from the local trol equipment and since 1991 the vehicle fleet traffic but close to Petrobras' Gabriel Passos re- has expanded considerably. Although the finery, the maximum 24-hour PM-10 concentra- amounts and sources of emissions have changed, tion during an eight-month monitoring period most air pollution is still caused mainly by in- was 119 pg/M3. At the Contagem station, where dustry and vehicles (Liu 1995). air quality is affected mostly by local traffic (es- Contagem and Belo Horizonte are the mu- pecially diesel-fueled vehicles) and to a certain nicipalities most affected by vehicular air pollu- extent by local industries, the 24-hour PM-10 tion. Air quality in Contagem also suffers from standard was exceeded ten times during a ten- the emissions of local industries as well as from month monitoring period. The maximum 24- air pollution drifting from Belo Horizonte. The hour PM-10 concentration monitored at this main industrial activities affecting air quality in Case Studies: Belo Horizonte 209 Table 4.44 Ambient PM-10 concentrations in the BHMR, 1995-96 (micrograms per ctubic meter) Belo HorizontIe Betim Contagem 24-lhtzr Alaon IhI 24-hour Monthly 24-hour Monthly Yeair AMon th maiximum eu11lge mrnximuum average maximum average 1995 July 70 28 102 38 129 51 August 99 46 119 66 212 67 September 174 57 102 43 231 76 October - - 53 19 133 64 November - - 83 32 96 49 December - - 45 26 116 52 1996 January 44 23 71 48 87 44 February 47 29 58 36 47 36 March - - - - April - - - - - May - - - - 135 59 June - - - - 276 105 - Not available. Source: FEAM 1996. the BHMR are located in the municipalities of located in Vespasiano recently reduced its PM Contagem and Betim, west of Belo Horizonte. emissions by installing control equipment. These industries include automotive, ceramics, Air emissions from the transport sector have glass, and cement manufacturing. The main air increased with the rapidly growing number of pollutants from these industries are TSP and vehicles in the BHMR. For example, between S02. In addition, NO, emissions from Petrobrais' 1980 and 1991 the number of buses, cars, and refinery in Betim are a concern. Other major trucks in Belo Horizonte increased by 71, 72, air emission sources include an iron and steel and 90 percent, respectively (Figure 4.14). In plant located in the municipality of Belo 1991 motor vehicles circulating in the city ac- Horizonte near Contagem (with SO2 and PM counted for 88 percent of the vehicles in the emissions) and a cement plant in the munici- BHMRAbout60percentofthevehiclefleetand pality of Pedro Leopoldo north of Belo 50 percent of the urban buses in circulation are Horizonte (with PM emissions). A cement plant more than eight years old (Liu 1995). Table 4.45 Estimates of transport-related pollutant emissions in the BHMR, 1991 (thousand tons per year) Emission soutrce (O H(C NO. 02 I'M Aldehydes Exhaust emissions Ethanol-fueled vehicles 22.5 2.3 1.9 2.0 Gasohol-fueled vehicles 122.6 7.2 4.7 0.2 0.8 1.0 Diesel-fueled vehicles 25.0 4.7 25.4 3.2 1.5 1.9 Evaporative emissions 12.6 Crankcase emissions 2.8 Road emissions (from tires) 1.7 Transfer operationsa 2.9 Total 170.1 32.5 32.0 3.4 4.0 4.9 Note: A blank space indicates insignificant emissions. a. From fuel stations. Souerre: Leme Engenharia 1991. 210 Chapter 4 Figure 4.14 Road-based vehicles in Belo Horizonte Number of vehicles (thousands) 400 - 393.8 _ _ 1980 350 _ 1991 300 - 250 - 228.6 200 1706 150- 100 63.1 50 -33. 3.7 6.3 57 0 Cars Trlicksa Buses Other) a. Incltudes light- and heavy-dtty trucks. b. Incltudes vans, agricultural and constmuction vehicles, motorcycles, and bicycles. Source: IBGE 1980 and 1991. Motorvehicle fuels in the BHMRinclude etha- dominated by road-based motorized modes, ad- nol (which contains 4 percent water) and gaso- versely affects the BHMR's environment (World hol (a blend of 22 percent anhydrous ethanol Bank 1995). and 78 percent gasoline) for light-duty vehicles The BHMR's road network is radial, and all and diesel fuel for buses, tnicks, and some other ten of its main corridors intersect the beltway vehicle types. In 1991, 55 percent of the motor that surrounds it. The three busiest corridors are vehicles in the BHMR were fueled with gasohol, Amazonas Avenue, Ant6nio Carlos Avenue, and 28 percent with ethanol, and 17 percent with Cristiano Machado Avenue. Municipal and in- diesel fuel (Leme Engenharia 1991). PM and ter municipal bus services are provided by 3,949 SO2were emitted mostly by diesel- and gasohol- vehicles that are privately operated by 74 com- fueled vehicles (see Table 4.45). Diesel-fueled panies on 425 routes (CBTU-Demetr6 1995). vehicles accounted for 38 percent of PM emis- Bus service is generally of low speed, unreliable, sions and 94 percent of S02 emissions from trans- and infrequent at peak hours because of con- port-related sources. Gasohol-fueled vehicles gestion. In 1991, 5,655 taxis were registered in were responsible for 20 percent of PM and 6 Belo Horizonte, carrying 75,000 passengers a day percent of SO9 emissions. PM emissions caused (IBGE 1991). by tire movement on the roads were also signifi- The suburban train, known as Demetr6, be- cant (43 percent). gan operating in 1986. The system consists of a Some 3.2 million person-trips take place each single 16.1-kilometer line that runs parallel to day in the BHMR. Of these, 68 percent are by the main east-west highway. This line carries bus, 25 percent are by private car, 1.7 percent about 50,000 passengers a day, considerably less are by urban train, and 5.3 percent are by other than the 250,000 originally planned. Ridership modes (mainly bicycling and walking). At peak is low because of the network's length, which is periods the main roads in the BHMR are con- far shorter than was originally planned because gested. This level of urban transport activity, of budget problems. Demetro is integrated with Case Studies: Belo Horizonte 211 seventy-four bus lines at just one terminal formulating environmental legislation; helping (Eldorado), that generates about 60 percent of municipalities with environmental protection, its demand. For passengers who are able to walk conservation, and improvement efforts; estab- to and from the stations, Demetr6 offers a com- lishing environmental priority areas; enforcing fortable and reliable mode of transport. But if (directly or through delegation of authority) their origins and destinations are beyond a rea- environmental policies; and fostering public sonable walking distance, the lack of stations and awareness of environmental issues. The council's limited integration with bus services make the twenty-five members are drawn from various system impractical and expensive. The lack of parts of society, including government offices, integration with buses discourages rail trips and civic associations, and universities. Before FEAM encourages bus and car use adding to conges- was established the council directed the Foun- tion and transport air pollution during peak dation for Technology Center of Minas Gerais hours (World Bank 1995). to conduct air pollution research for the BHMR FEAM was established in 1987 to conduct stud- ies and research on the environment and to help Institutional Responsibilities ensure its protection, conservation, and improve- ment in accordance with the council's directives. Federal institutions. The federal environmental FEAM is responsible for conducting research on institutions are described in the section on Sao pollution control and standard setting; propos- Paulo. The Brazilian Urban Train Company ing measures for environmental protection, con- (CBTU), through its Belo Horizonte subsidiaiy servation, and improvement; offering specialized (STU-BH), operates Demetr6. Although the environmental services, including environmen- subsidiary is currently under the jurisdiction of tal training; developing public education activi- the federal government, it is expected to be ties on environmental issues; supporting transferred to the State of Minas Gerais in 1997. municipalities in developing and implementing local environmental programs; and enforcing State institutions. The State Secretariat of Science, environmental legislation on behalf of the State Technology and the Environment was created Council for Environmental Policy. It also moni- in 1987 to plan, organize, direct, coordinate, tors air quality in the BHMR. execute, and control all sectoral activities related The Roads Directorate of Minas Gerais regu- to scientific and technological development and lates intermunicipal bus services and the road environmental protection. In the environmen- network in the BHMR. tal area, the secretariat's responsibilities include formulating state policies and directives, elabo- Metropolitan institutions. In 1993 the State of rating plans and programs to improve the envi- Minas Gerais established the BHMR Metropoli- ronment, stimulating, sponsoring basic and tan Council (AMBEL) to foster integration of applied research, coordinating activities by en- metropolitan transport services, approve and tities under the secretariat, and initiating and supervise implementation of the BHMR master coordinating the execution of state environmen- plan and its budget, establish guidelines for tar- tal policies and programs. The entities under iffs on metropolitan public services (including the secretariat are the State Council for Environ- transport), and administer the Metropolitan mental Policy (COPAM), State Council of Development Fund. The council has a number Science and Technology (CONECIT), and Car- of sectoral technical committees, including the tographic Coordination Council (CONCAR). Water Resources and Environmental Commit- The entities affiliated with the secretariat are the tee, Intermunicipal Transport and Road Com- State Foundation for the Environment (FEAM), mittee, and Land Use Committee. Foundation for Technology Center of Minas For the implementation of the World-Bank Gerais (CETEC), Institute of Weights and Mea- financed Belo Horizonte Metropolitan Transport surements of Minas Gerais (IPEM), and Re- Decentralization Project, the State of Minas search Support Foundation of Minas Gerais Gerais and the municipalities of Belo Horizonte, (FAPEMIG). Betim, and Contagem signed an agreement to The State Council for Environmental Policy create a permanent integrated planning group was created in 1977. Its responsibilities include that will function as a precursor to a regional trans- developing the state's environmental policies: portcoordination commission (World Bankl995). 212 Chapter 4 Local institutions. Municipalities are the most using the Ringelmann scale for vehicles important environmental institutions at the lo- equipped with naturally-aspirated diesel engines cal level. The Municipality of Belo Horizonte has and the filtration method (through an equip- a Motor Vehicle Pollution Branch within the ment inserted directly into the tailpipe) for tur- Municipal Secretariat of Environmental Control. bocharged vehicles. A notice is given during the This secretariat has been involved in signing first inspection if a bus does not comply with stan- agreements with the state government, conduct- dards. If the problem is not corrected within a ing air quality monitoring studies, passing local certain period, fines are imposed. Three tech- environmental legislation, and enforcing laws nicians and two policemen conduct the inspec- and regulations within itsjurisdictioni. tions. The Oxygen Operation program was The Belo Horizonte Transport and Transit recently extended to trucks and intercity buses. Company, a municipal agency, regulates bus ser- The Municipality of Contagem's Environmen- vices within the Municipality of Belo Horizonte. tal Secretariat plans to initiate a similar testing It is also responsible for traffic management and program for buses and trucks. control in the municipality. The BHMR does not have a periodic inspec- tion and maintenance program that measures emissions at inspection stations. However, such Implemented Measures a program is being prepared under the World Bank-financed transport management project Vehicle emission standards. The national regtula- for Belo Horizonte, mentioned earlier. tory measures for light-and heavy-duty vehicles are described in the Sao Paulo section. These Fuel-targeted measures. Fuel policies-including measures, which include emission standards for type, quality, supply, and price-are determined light-and heavy-duty vehicles, also apply to the at the national level. For the most part the dis- BHMR cussion in the Sao Paulo section is valid for the In 1988 the Environment Secretary of the BHMR. Municipality of Belo Horizonte signed the "Oxy- Most cars in the BHMR use either ethanol or gen Operation" agreement with the State of gasohol. The average lead content of gasoline Minas Gerais to reduce air emissions from die- was reduced from 0.25 g/liter in 1977 to 0.15 g/ sel-fueled vehicles. This agreement required liter in 1979, to 0.09 g/liter in 1983, and to 0.06 compliance with Municipal Law 4253/1985 and g/liter in 1987. Since 1991 gasohol has been its Decree 5893/1988. Article 25 of this decree lead-free. limits smoke emissions from diesel-fueled ve- Most heavy-duty vehicles (buses and trucks) hicles to No. 2 on the Ringelmann scale or its use diesel fuel. Two grades of diesel fuel, with equivalent."' different sulfur contents, are available. The low- sulfur grade is used by diesel-fueled vehicles cir- Vehicle inspection programs. Since 1988 the Mu- culating in the BHMR. In October 1996 the nicipality Belo Horizonte's Environment Secre- sulfur content of this diesel fuel was reduced tariat has been testing buses circulating in Belo from 0.5 to 0.3 percent by weight. The high-sul- Horizonte as part of the Oxygen Operation pro- fur grade is allowed for vehicles operating out- gram. The tests are performed at bits terminals side the BHMR. The sulfur content of this fuel and involve measurement of smoke emissions was reduced from 0.9 to 0.5 percent by weight. Both grades of diesel fuel are sold at the same price. 30. Smioke opacity measurement usiig thie Riigeiani Transport management. Of the ten main roads in scale is an old technique that relies oni visuiallv miatchinoig the color of a smoke with a series of sha(les, ranging from t white (Rinigelmann No. 0) to black (Ringelmuann No. 5), a reserved bus lane. Buses in the segregated printed on a whiite card. By lookirig tlirottgih a hole in the busway move at about 25 kilometers/hour, al- card at a smoke plitne, the user cats judge whiich shade of though this speed drops to 9 kilometers/hour gray the smoke is closest to, arsd titis assigti a Ringelrnanni in the central business district. nuimber to the smoke. Ringelmnatin No. I corresponds to roxughly 20 percent opacity, No. 2 to 40 perceint opacity, No Bis routes are allocated to private operators 3 to 60 percent opacity, No. 4 to 80 percent opacity, and( No. on the basis of competitive bidding and regu- 5 to 100 per-cent opacity. lated (including the setting of tariffs) by the Belo Case Studies: Belo Horizonte 213 Horizonte Transport and Transit Company in the Between February 1992 and February 1993 the Municipality of Belo Horizonte and by the Roads municipality of Belo Horizonte contracted the Directorate of Minas Gerais across the munici- Federal University of Minas Gerais to monitor palities within the BHMR. Both entities, which ambient TSP and SO, concentrations. Five moni- have a clearinghouise (Camara de Compensacio) toring stations in Belo Horizonte were used for for bus tariffs, pay for bus services according to this purpose. a standard cost formula that they approve. This Air monitoring in the BHMR, which for many approach is used to compensate bus companies years had been hampered by lack of funding, that charge tariffs below the cost estimated with started again in May 1995 as a result of a 1994 the standard formula. No subsidies are paid to legal agreement between Petrobras, the State bus operators, however (World Bank 1995). Council for Environmental Policy, and FEAM. The World Bank-financed transport project The investment cost of the monitoring network will extend the urban rail system by 12 kilome- ($400,000)-which includes the automatic sta- ters. constrmct nine new rail stations and improve tions in Belo Horizonte, Betim, and Contagem- an existing station, integrate the rail system with was financed by Petrobras. These three stations urban buses serving the BHMR by constructing are capable of continuously monitoring ambi- three terminals, and introduce traffic engineer- ent SO, and PM-10 levels and meteorological ing measures (signals, signs, busways, bus lanes, parameters (wind direction and speed, tempera- and shelters for bus passengers) and a road traf- ture, and relative humidity). The monitoring fic control system (such as a centralized signal data are sent by telephone lines to FEAM and control system). In addition, the project will PetrobrAs. The system generates real-time data upgrade the BHMR's masteir plan and prepare as well as daily, weekly, monthly, and annual av- an integrated land use, urban transport, and air erages. FEAM is responsible for operating the qualitv strategy (World Bank 1995). monitoring network (FEAM 1996). Air quality monitoring. Air quality monitoring in Evaluation of Implemented Measures the BHMR was initiated in 1984 with a monitor- ing network established by the Foundation for Vehicle emission standards. The evaluation of emis- Technology Center of Minas Gerais at commer- sion standards in the Sao Paulo section is also cial, industrial, curbside, and residential loca- valid for the BHMR. In summary, the emission tions. The pollutants of concern were TSP and limits imposed on manufacturers of new light- SO,. Because SO2 monitoring equipment was not and heavy-duty vehicles have helped reduce ve- available, total sulfation raites were measuried hicular emissions. instead. TSP concentrations were measured us- ing high-volume samplers and analyzed gravi- Vehicle inspection programs. Between June 1988 metrically. The number of stations in operation and July 1993, 65 percent of the 25,515 buses varied from twenty-three to thirty-four a year for inspected were found to be in compliance, 12 total sulfation and from seven to twelve for TSP. percent were brought to compliance after the Many of the stations suffered operational diffi- first failure, and 23 percent were fined. With the culties. The operation of the network was stos- implementation of the Oxygen Operation pro- pended in 1988 because of lack of funding. gram the noncompliant buses were reduced In 1991 FEAM introduced a new air quality from 53 percent in 1988 to 19 percent in 1993 monitoring network that used high-volume (Table 4.46). The increase in compliance is as- equipment to sample TSP in the BHMR. This sociated with a reduction in pollutant (especially network included five monitoring stations, thli-ee PM) emissions. were in Contagem (stations 35, 37, and 44) and The inspection and maintenance program one each in the municipalities of Belo Horizonte being developed under the World Bank-financed (station 75) and Pedro Leopoldo (station 124). project is expected to establish an inspection All field and laboratory work as well as evahla- scheme for in-use vehicles in the BHMR. Such a tion of the results were conducted by the Foun- program would reduce emissions by encourag- dation for Technology Center of Minas Gerais, ing proper maintenance of these vehicles. with the results sent to FEAM each month. The monitoring network was suspended in early 1993 Fuel-targeted measures. The evaluation of fuel because of insufficient funding. types presented in the Sao Paulo section also 214 Chlapter 4 Table 4.46 Bus inspection results for Belo Horizonte, 1988-93 hiutl number In coml/ianc Noofied-r*rrecled Fined 'eslr i)insp)ecled Number P'crcent Numnber 1ercent Nutmber Perrent 1988 6.899 3,231 47 1,227 18 2,441 35 1989 2.618 1.827 70 257 10 534 20 1990 4,788 3,080 65 482 10 1,216 25 1991 4,622 3,358 73 450 10 814 18 1992 4,155 3,124 75 389 10 642 15 1993 2.443 1,974 81 179 7 290 12 Total 25,515 16,594 65 2,984 12 5,937 23 Note: Data for 1993 onily coverJanullary-Juinie. Soutre: Prefeittura Mutnicipal de Belo Hofizonte 1993. holds for the BHMR. In summary, the use of Transport management. Because the BHMR Met- ethanol and gasohol in place of gasoline has re- ropolitan Council was forned only recently, poli- duced CO, HC, PM, and S02 emissions but in- cies and regulations affecting urban transport creased aldehyde emissions in the exhaust of are developed without a formal coordinating light-duty vehicles that are not equipped with arrangement among different government agen- catalytic converters. Evaporative (HC) emissions cies. As a result there has been a lack of consis- have decreased for ethanol-fueled vehicles, but tency in the fares charged by similar or have increased for gasohol-fueled vehicles. The competing transport modes, an absence of cri- shift from leaded to unleaded gasoline tndoubt- teria to prioritize investments, duplication of edly has reduced lead emissions. Because of the investments, and limited modal integration. At lack of reliable ambient air quality dlata in the times these shortcomings have resulted in bus BHMR, the effects of these measures cannot be services operating in corridors served by quantified. Demetr6, expensive and time-consuming dis- Based on fuel quality data for 1996 the Reid agreements on the construction of a light rail vapor pressure and sulfur content of gasohol transportsystem,andwidelyvaryingsubsidypoli- appear to be high. Lowering the vapor pressure cies that are often based on non economic con- would reduce evaporative emissions from gaso- siderations. Creation of a regional transportation hol-fueled vehicles not equipped with evapora- coordination commission would ensure that the tive controls, and lowering the sulfulr content recently established cooperation among the Belo would reduce CO, HC, and NO, emissions from Horizonte Transport and Transit Company, the gasohol-fueled vehicles equipped with catalytic Roads Directorate of Minas Gerais and the Belo converters. Horizonte Subsidiary of the Brazilian Urban The addition of lighter and heavier fractions Train Company continue irrespective of the po- to meet market demand for diesel ftuels increases litical parties in power. NO. and PM emissions from diesel-fueled ve- The World Bank-financed transport project is hicles. Lowering the sulfur content of the urban expected to alleviate traffic congestion, which is diesel fuel must have reduced SO., and sulfate growing by 20 percent a year. The greatest an- emissions as well as the formation of second- ticipated benefit of this project will result from ary sulfates in the BHMR. Although the the reduction in car and bus traffic due to the government's policy of setting the same price new train links. In addition, the traffic manage- for both diesel grades is intended to reduce ment, control, and safety components of this misfueling, the requirement to use low-sulfur project will reduce accidents and vehicle emis- diesel is not observed by all trucks and buses cir- sions. culating in the BHMR because it is not enforced. Therefore, the health and environmental ben- Air quality monitoring. During the 1980s ambient efits of these regulatory measures have not been air monitoring in the BHMR focused on TSP fully attained (Liu 1995). instead of PM-10 because the Brazilian PM-10 Case Studies: Belo Horizonte 215 standards were not established untilJune 1990. sentative of the equipment's foreign manufac- In addition, it was expensive to import PM-10 turer was deemed not competent to provide the monitoring equipment, and Brazilian manufac- technical assistance necessary to quickly resolve turers did not fabricate it because there was little the maintenance-related problems (FEAM domestic demand (FEAM 1996). 1996). During the 1980s and early 1990s technical Because of its remote location,the air quality problems and a lack of financial resources ham- data from the Betim station does not accurately pered the operation of the air monitoring net- reflect the level of human exposure to air pol- work in the BHMR. The quality of the available lutants in that municipality. Relocation of the data is also suspect because no quality control station to a populated part of the municipality and quality assurance procedures were used. The would provide a better information about pol- available data indicate that Belo Horizonte. lutant exposure levels. The BHMR also needs a Betim, Contagem, and Pedro Leopoldo are most more widespread, permanent air quality moni- affected by industrial and vehicular air pollution, toring network that would at least measure am- mostly during the dry season. TSP levels in Belo bient concentrations of PM-10 and SO2, and Horizonte originate mainly from vehicles, espe- which could be expanded to monitor ambient cially diesel-fueled vehicles. Emissions from die- concentrations of NO2, CO, ozone, and alde- sel-fueled vehicles are also an important source hydes. In addition, the BHMR's outdated emis- of the high ambient SO2 levels observed at the sions inventory needs to be updated to provide Station Square. In Betim, Contagem, and Pedi-o data for planning the expansion of the monitor- Leopoldo industries are also important emission ing network and devising pollution control mea- sources. sures. Furthermore, because industrial sources Continuous operation of the three air moni- in some parts of the BHMR are believed to be toring stations installed in 1995 has not been important contributors to air pollution, ajoint possible because maintenance-related problems strategy for industrial and vehicular pollution have caused frequent shutdowns. For example, control needs to be considered. Funds to develop during March-April 1996 the stations were shut and implement pollution control strategies are down because of an electrical discharge which currently lacking, but are essential for improv- damaged microcomputers. Also the local repre- ing ambient air quality in the BHMR. BUENOS AiREs IBRD 28882 19-W 58'30' BOLMAL ( - opone l: <; mpana ARGENTINA r / : StA:AT; B.-S:ff:4;AdSS;:S:fA;400St:04 AmAT,AN T I R E V. Rosa ARGENTINA D. ~~~~~~~ ~BUENOS AIRES METROPOLITAN AREA. A MR QUWTYMONPOCRING STATONS 34'30'.....0 ON ---METROPOLITANAREA BOUNDARY .t~ E O PARIMDOS BOUNDAPIE S --INTERNATIONAL BOUNDARIES: - ..~~~~~~~~~~K 0Cuiiuelas;: ;0f0Sft;00:00000000000 S Vicnl 35*00XlXXt0 0 o'00040;; Ti Mop w r A m ano by the Map Design Unit 004 j 0;0:2000000 ;0 0;00:00) 0 f 00f;00 :00 ;; / :;i t;::::A141tnC00 lt .4 ~ ~ ~ ~ ~ ~~E V.Roso EnsenadaaT r 0000000t 3500' ~ j.|.< \ , :. Caiula S. Viet AIQll0OlatGxTO 3 : 00' -00 ofThe WrldBak The boundaries, colors, denominations and any aother information shown .on this maop do not imply, on the part of Te Word Bank Group, 0 5 1 5 2 any judgment on the legal ? l ~~ status of any territory, KILOMETESR or any endorsement or acceptance of such boundaries. 6r> \t0 Wj00000j00320000000;00;00000004j00000004S000' P58'30' 5 JULY 1997 Case Studies: Buenos Aires 219 The Buenos Aires Metropolitan Area (BAMIA) between 8A.M. and 4P.M. and lowest between 12 comprises twenty-five municipalities in the Prov- A.M. and 8 A.M. on weekdays. In 1994 the highest ince of Buenos Aires and the city of Buenos Aires 8-hour average concentration was 17.3 ppm and (the Federal Capital ofArgentina).31 In 1991 thie in 1995, 12.4 ppm. These concentrations ex- BAMA contained 10.9 million people, 3.7 times ceeded the national standard of 10 ppm and the the population of the Federal Capital and about WHO guideline of 9 ppm. The highest 1-hour one-third of Argentina's population (Pre-ATAM CO concentration observed at this location was 1994b). Argentina's economic center, it is lo- 30 ppm in 1994 and 33 ppm in the first five cated on the southern bank of the Plata River, months of 1995 (FAS21 1995). In 1996, 1-hour 160 kilometers from the Atlantic Ocean at an concentrations as high as 68 ppm were recorded altitude of about 25 meters on a flat topogra- in the La Recoleta district, 43 ppm in the down- phy. The BAMA covers 3,880 square kilometers, town business area, 40 ppm in the Once district, of which 200 square kilometers are in the Fed- and 38 ppm in the Palermo district were re- eral Capital, and has a humid subtropical climate corded (Table 4.47). Ambient CO concentra- with average temperatures ranging from 1 "C tions exceeded the 1-hour WHO guideline (26 during the winter (June-August) to 24"C dlir- ppm) at all these locations, but exceeded the 1- ing the summer (December-February). Precipi- hour national standard (50 ppm) only at La tation averages 1,000 millimeters annually, and Recoleta. varies from 60 millimeters in July to 120 milli- In 1994 short-term NO, concentrations var- meters in March. Air emissions are generally well ied from 26 pg/m3 to 447 pg/M3. During No- dispersed by southeasterly and southwesterly vember and December the short-term winds. However, the winds are not sufficient to concentrations exceeded the city's 20-minute disperse vehicular air pollution in narrow, heavily NO, standard of 400 pg/m3. The annual aver- traveled streets during weekday commuting and age NO, concentration was 157 pig/mi (DVA business hours. 1995). This level exceeds the annual NO,stan- dard of 94 pg/mi for the Province of Buenos Aires.33 Ambient Air Quality In 1994 short-term ambient TSP concentra- tions, which varied from 62 j1g/M3 (in Septem- Limited ambient air quality data are available ber) to 335 p1g/m3 (in November), were below for Buenos Aires. These data are based on daily the city's 20-minute standard of 500 pig/m3. manual samples of NO., lead, TSP, and SO, col- Monthly average TSP concentrations were be- lectedbythe Municipality ofBuenos Aires at one tween 111 p.g/m3 (in September) and 218 g.g/ monitoring station, continuous samples of CO m3 (in November). The city's 30 day standard of collected by Fundaci6n Argentina Siglo 21 at 150 gg/m3 was exceeded only in November. The another monitoring station, and short-term annual average (146 pg/m3) was higher than the samples of CO collected by Xilix at roadside ar- WHO guideline of 60 pig/ml to 90 pg/M3 (DVA eas in the downtown business district and three 1995).34 other areas in the Federal Capital.32 These data Limited monitoring data indicate that ambi- indicate that CO, NO, and TSP are the pollut- ent lead used to be an environmental concern, ants of concern for Buenos Aires. No infor-ma- especially in downtown Buenos Aires, where traf- tion is available on ambient ozone or PM-10 fic is dense. Evidence of lead pollution in Buenos levels. Aires was reported in the late 1980s (Table 4.48). Ambient CO concentrations at a traffic-dense Short-term lead concentrations monitored dur- location in downtown Buenos Aires are highest ing five months in 1994 varied from 1.1 pg/M3 to 9.2 pg/ni3. These concentrations complied with the city's 20-minute standard of 10 p1g/M3 (DVA 1995). However, ambient lead concentra- 31. Some other studies define BAMA to cover forty-thlee municipalities in the Province of Buenos Aires and the citv of Btienos Aires in an area spanninig 16,770 square kilome- ters. In 1991, the population of this area was 12.4 millioni 33.ThereisnocorrespondinglocalornationalNO,stan- (Pre-ATAM 1994b). dard. 32. NO, is used instead of NO2 as the pollutant pararn- 34. There is 11o correspondiiig local, provincial, or na- eter in air quiality standards and monitorinig data in Argen- tional TSP standard. tina. 220 Chapter 4 Table 4.47 Ambient CO concentrations in Buenos Aires (parts per million) 7'ine Location August 1993 Jutne 1994 October 1995 M1ay 1996 9 A.M. Downtown (Corrientes & Madero) 18 21.5 19.3 24.2 6 P.m. Downtowtn (Corrientes & Madero) 26 22.4 26.1 28.2 9 %.m. Downtown (Corrientes & Pellegrini) 23 27.2 21.6 26.5 6 P.M. Downitown (Corrientes & Pellegrinii) 17 25.6 27.1 27.8 9 A.M. Dowtntown (Esineralda & C6rdoba) 25 22.4 31.2 27.4 6 P.M. Downltown (Esnieralda & C6rdoba) 20 24.8 29.4 30.5 9 .M.m. Downtowni (Corrientes & Leand(ro N. Alem) 26 24.3 29.3 28.3 6P.mi. Downtown (Corrientes & Leand(ro N. Alem) 26 33.1 31.8 42.5 9 A.M. Downtown (San Martin & Tucuiw3n) 28 29.2 26.5 24.3 6 P.NI. Downtown (San Martin & Tuciuinin) 33 30.5 27.6 36.3 9 A.M. Palermo (Sta. Fe &Juan B.Justo) 28 26.7 31.5 32.4 6 P.m. Palermo (Sta. Fe &Juan B.Justo) 28 37.2 33.4 37.8 9 A.M. Once (Bme. Mitre & Pueyrred6n) 22 31.8 35.6 33.2 6 P.M. Once (Bme. Mitre & Pueyrred6n) 30 39.1 36.3 40.3 9 A.M. La Recoleta (F. Alcorta & Pueyrred6n) 31 34.2 30.6 35.4 6 P.M. La Recoleta (E Alcorta & Puevrred6n) 62 58.0 67.1 68.2 Sot,rce: Cantini 1996 (based on monitoring dlata from Xilix). tions must have fallen considerably since lead major traffic corridors. Between 1994 and 1995 was eliminated from gasoline in late 1995. the motor vehicle fleet increased by 5.4 percent In 1994 the highest recorded short-term SO2 in the Federal Capital and by 3.0 percent in the concentration was 31 pig/mi, much lower than Province of Buenos Aires, largely because of in- the city's 20-minute standard and WHO's 10- creased numbers of private cars. In 1995 the minute guideline of 500 pg/M3. The annual av- BAMA contained about 2.6 million private cars, erage SO, concentration of 10 jig/mt was also 46,000 taxis, and 16,000 buses."5 About 19 per- lower than the WHO's gLideline of 40 plg/M3 to cent of the 5.9 million motor vehicles in circula- 60 jig/ml (DVA 1995). tion in Argentina were in the Federal Capital. These included 953,000 cars, 134,000 trucks, and 11,000 buses (ADEFA 1996). In the BAMA gaso- Sources of Pollutants line-fueled vehicles emitted about 85 percent of CO, 80 percent of HC, and 40 percent of NO, Motor vehicles are the main source of air pollu- from road-based sources, and diesel-fueled ve- tion in downtown Buenos Aires and the BAMA's hicles emitted 98 percent of SO2 and 63 percent of PM (Table 4.49). Gasoline-fueled vehicles contributed to a higher share of CO, HC, and Table 4.48 Ambient lead concentrations NO, emissions in the Federal Capital than the in Buenos Aires BAMA because they constitute a higher propor- tion of the vehicle fleet in the Federal Capital. (micrgramsperCcbiccmter)Public transport services in the BAMA are pro- L.oration amplieril tron- vided by buses, metro, and suburban trains. In Downown- heav trafic) 8 A.. -I)ermt3f 1994 about 15,000 buses operated on 299 lines Downtown (heavy traffic) 7 A.M. - 85.M. 1.7 and were run by 240 privately owned companies. Downtown (medium traffic) 24 hotii.s 1.5 About 2,000 of these buses (on 93 lines) oper- Suburban (medium traffic) 24 houirs 1.0 ated exclusively within and were regulated by a Suburban (low traffic) 24 hotuTrs 0.3 specific municipality in the Province of Buenos Standard for the City of Buenos Aires 24 hours 1.0 Soure: Caridi and others 1989. 35. Estimated based on data from pre-ATAM 1994 and ADEFA 1996. Case Studies: Buenos Aires 221 Table 4.49 Estimates of emissions from gasoline- and diesel-fueled vehicles in the Federal Capital and the BAMA, 1993 (thousanids of tons) (:0) I-IL _______________Nos '2 PM Velhile lpe ITr BAAMA F(. llAi\A iT: BAAIA 1( BAMA FC BAAMA Gasoline-fueled 247.7 861.7 9.5 36.9 6.8 26.4 04 1.6 1.3 5.0 Diesel-fueled 27.2 156.1 1.6 9.2 6.8 39.0 11.7 67.2 1.5 8.6 Total 274.9 1,017.8 11.1 46.1 13.6 65.4 12.1 68.8 2.8 13.6 Note: FC is the Federal Capital andc BAMA is the Beneios Aires Mletiopolitan Area. iSmrre: Derived from World Banik 1995. Aires. Another 3,000 (on 59 lines) operated center. In the microcenter area taxis represent across municipality boundaries within the Prov- more than 50 percent of all circulating vehicles, ince of Buenos Aires and were regulated by the followed by private cars (25 percent) and buses province. The remaining I 0,000 buses were rini (less than 15 percent). But in the macrocenter by 123 companies (on 147 lines) within the Fed- area private cars account for 46 percent of traf- eral Capital and on routes extending from it. fic, followed by taxis (40 percent) and buses (6 Besides providing regular buis services, these buis percent; World Bank 1995). companies also operate express buses On a typical workday in 1994 about 17.6 mil- (diferenciales) on many lines in the BAMA. Buses lion passenger trips took place in the BAMA. In operating within and through the Federal Capi- addition, about 200,000 daily trips were made tal are regulated bv the National Commission of by trucks transporting more than 600,000 tons Transport Control under the jurisdiction of the of goods. Of the passenger trips, 56.5 percent Municipality of Buenos Aires and the Secretariat were made by public transport modes, consist- of Public Works and Transportation, a national ing of buses (47.1 percent), subturban train (4.9 entity formed inJune 1996.*" In addition, char- percent), and metro (4.5 percent). The share ter buses have recently become popular for corII- of private cars was 32.4 percent. About 71 per- mtuters living in stuburbs 20 kilometers to 30 cent of the bus commuters used the federal bus kilometers outside of the downtown area. system; the rest used the provincial and munici- The fully electrified 44-kilometer metro net- pal bus systems. work is owned by the Municipality of Buenos Aires and operated by a private firm under a Institutional Responsibilities twenty-year concession. Only 18 percent of thie suburban railway, which has a total length of 900 National institutions. The Secretariat of Natural kilometers on six different systems, is electrifiecl. Resources and Sustainable Development, a cabi- It is owned by the state and operated by fomr net-level agency reporting directly to the presi- private firms tinder ten-year concessions. dent, formulates national environmental policy The taxi fleet includes about 38,600 registered in Argentina.37 The secretariat, which includes and 7,000 illegally operated taxis. In addition, subsecretariats for human environment and chauffeured short-term cars (reinises) have cap- nattiral resources, is the national institution tured an increasing share of the taxi market. responsible for setting minimum environmen- About 155 agencies, withaboutfive remiseseaclh, tal standards-including emission standards operate legally, and about 750 agencies operate for vehicles-in Argentina. Provincial authorities illegally. The share of taxis in the vehicle fleet establish their own standards based on this decreases with further distance from the city minimum. A number of other institutions also deal with environmental issues at the national level in co- 36. In 1996 the National Coinmissioti of Tranisport C(on- trol was formed to replace the National Commatission( ot Automotive Control (CONTA), ar(i the Secretariat of Pub- 37. In 1996 the Secretariat of Nattiral Resources and Stis- lic Works and Transportation was formned to replace tile taimableDevelopmentwasformed to replacetiheSecretariat Secretariat of Eniergy and Transpor-tation. of Natiral Resources and Htiman Enivironimenit. 222 Chapter 4 ordination with the Secretariat of Natural Re- Metropolitan institutions. A number of unsuccess- sources and Sustainable Development. These ftil attempts have been made to improve coordi- institutions include the Secretariat of Industry, nation among the federal government, the Commerce, and Mining (industry and mining), Province of Buenos Aires, and the Municipality Ministry of Education (environmental educa- of Buenos Aires on transport-related issues in tion), Ministry of Health and Social Action (en- the BAMA. One such attempt occurred in 1991 vironmental standard setting, resettlement, with the signing of an agreement between the environmental impact assessment of inirastruc- economy minister and the mayor of Buenos Aires ture projects), Secretariat of Public WVorks and to create a temporary planning unit, called "pre- Transportation (inspection of commercial ve- ATAM", to define and develop an independent hicles), Ministry of External Relations and In- and autonomous Transit Authority for the Met- ternational Trade (international aspects of the ropolitan Area. The Transit Authority was con- environment), and Secretariat of Agrictulture, ceived as consisting of representatives from all Cattle, and Fisheries (environmental research). three government jurisdictions and was to be This fragmentation of responsibilities constrains responsible for the planning, regulation, and the efficiencies of Argentina's environmental control of all activities involving the urban trans- management framework and sometimes causes port system in the BAMA. Although the legisla- coordination conflicts and disputes over the al- tion to create the authority passed in the Senate, location of resources (World Bank 1995). it failed to gain broad political support. In late The Secretariat of Natural Resources and Sus- 1996 it was converted into a new planning unit tainable Development hosts the Federal Coun- for Urban Transport of the BAMA (called sel of the Environment, the forum in which all "TUAMBA"). national and provincial governments discuss common themes. In 1993, the Secretariat nego- Local institutions. The Municipality of Buenos tiated the Federal Environmental Pact with all Aires has established ambient air quality stan- the provinces. However, few provincial con- dards and emission standards for vehicles, and gresses have ratified this pact, a political instru- has been monitoring ambient air quality in the ment that signals a willingness to coordinate Federal Capital through the Atmospheric Moni- provincial and national efforts. toring Laboratory of its Environmental Hygiene The Secretariat of Public Works and Transpor- Department. The mtnicipality does not have any tation finances, constructs, and maintains na- authority to regulate the national bus system. tional highways, subsidizes the operation of However, it maintains the road infrastructure suburban railvay lines, regulates the national bus and regulates the operation of taxi services lines operated by private bus companies, and within Federal Capital. It also owns the infrastruc- inspects commercial vehicles operating under ture of the Buenos Aires metro system. The the nationaljurisdiction. The traffic police divi- operation of this system, which is under a twenty- sion of the federal police department is respon- year concession to a private firm, is subsidized sible for traffic control and enforcement in the by the national government. The other munici- Federal Capital. palities of the BAMA are responsible for air and noise pollution, traffic management, and regu- Provincial institutions. In principle, most environ- lation of bus and taxi services within their re- mental matters that are not expressly delegated spective municipalities. to the national government are the responsibil- ity of the provinces. The Province of Btienos Aires has one of the best environmental management Implemented Measures systems in Argentina. The provincial Secretariat of Environmental Policy is responsible for envi- Vehicle emission standards. National Decree 875/ ronmental matters within the province, includ- 94 of Regulation 2254/92 established emission ing air pollution from fixed and mobile sources. and noise standards for new and used motor The provincial Ministry of Public Works and Ser- vehicles and made the Secretariat of Natural vices regulates the operation of bus lines that Resources and Human Environment responsible cross municipalities within the province. It is also for implementing and updating these standards. responsible for constructing and maintaining the For new light-duty vehicles, Decree 875/94 lim- principal roads in the province. ited exhaust emissions for CO, HC, and NO, Case Studies: Buenos Aires 223 Table 4.50 Exhaust eniission standards for new light-duty vehicles in Argentina lehicl tvpe and (:0 H( NQ, PM e.fpli/ee dte qf standard (g/krn) (perrent)" (g/km) (14)mr (g/km) (g/km) Passenger vehicles BeforeJuly 1994' 3.0 600 July 1994' 24.0 3.0 2.1 600 2.0 November 1995'1 12.0 2.5 1.2 400 1.4 January 19961 12.0 2.5 1.2 400 1.4 0.373 January 1997"' 2.0 0.5 0.3 250 0.6 0.124 Commercial vehicles Before July 1994' 3.0 600 Jtuly 1994' 24.0 3.0 2.1 600 2.0 January 1998 1 6.2 0.5 0.5 250 1.43 0.16/0.31e Januarv 1999'1 2.0 0.5 0.3 250 0.6 0.124 ANte: A blank space indicates that no standard was established. CO, HC, and NO, emissions from the exhaust are to be tested using USEPA's Methiod FIP-75. a. Measured at idle only for vehiicles with spark-ignition enginies. b. Applicable only to diesel-fuieled vehicles. c. Applicable only to domestically produced vehicles. d. Applicable both to domestically produced and( imported vehiicles. e. The 0.16 g/km Iiimit is applicable to light-duty diesel-fiueledl vehicles with reference weight (weight of the vehicle plus 136 kg) not more than 1,700 kg. The 0.31 g/km limit is applicable to light-duty diesel-fueled vehicles with reference weight more than 1,700 kg. Swornr Decree 875/94, publishied in Bollin Oficial 27,919 of July 27, 1994, and Decree 779/95, published in Bok.tin Oficial 28,281 of November 29, 1995. from all domestically produced and imported light-duty vehicles that requires manufacturers effective June 1994. 3 The decree also set more of light-duty vehicles to comply with emission stringent limits for these pollutants according standards for 80,000 km or five-years (whichever to a timetable, eliminated crankcase emissions comes first). Alternatively, this requirement can effectiveJuly 1994, and limited evaporative emis- be replaced by a 10 percent reduction of emis- sions from light-duty vehicles with spark-ignition sion levels from those specified in Table 4.50. engines to 6.0 grams/test effective January 1, Small light-duty vehicle manufacturers (all 1995. This decree was superseded by Decree manufacturers producing less than 1,000 779/95 of Law 24,449/94. The new decree es- vehicles a year and some specialized manufac- tablished CO, HC, and NO, standards for light- turers producing less than 2,500 vehicles a duty passenger and commercial vehicles based year) are exempted from the certification re- on a more stringent compliance schedule. In quirement. addition, the decree set PM standards for die- For new diesel-futeled urban buses and other sel-fueled vehicles (Table 4.50)." The require- vehicles circulating on roads under the national ments for crankcase and evaporative emissions jurisdiction, Joint Resolution 58/94 and 96/94 remained the same. Decrees 875/94 and 779/ by the Secretariats of Transportation and Indus- 95 also established a certification system for new try established CO, HC, and NO, emission stan- dards effective July 1994. The resolution tightened these standards according to a time- 38. Light-duty vehiicles are those with a net weight (witl- table, set additional standards for smoke emis- out load) up to 2,722 kilograms and gross weighlt (with pas- sions as a function of exhaust gas flow rate and sengers and maximum freighit) up to 3,856 kilograms. PM standards based on engine power, and elimi- 39. Light-duty passenger vehicles are those designe(d to nated evaporative emissions. Decree 779/95 re- carry up to twelve passengers or derived from those vehicles confirmed these standards for all new to carry freight. Light-duty commercial vehiicles are those designed to carry freight . or derived from those vehicles. diesel-fueled heavy-duty vehicles In Argentna They also incIlide all lighlt-duty vehicles designed to carry (Tables 4.51 and 4.52). The decree also requires more than twelve passengers. that compliance with the standards be guLaran- 224 Ctapter 4 Table 4.51 CO, HC, and NO. exhaust Table 4.54 Exhaust emission standards emission standards for new diesel-fueled for in-use vehicles with diesel engines heavy-duty vehides in Argentina in Argentina (grams per kilowsatt-hour) Barharach Absor)tion Ve,ohite type and effertive dat7e (° 1i( NO, E fective date index coeffiiet (m') Urban buses November 1995 6 2.62 July 1994 11.2 2.4 14.4 July 1997 5 2.62 Janiuary 1995 4.9 1.23 9.0 Note: Measured at free acceleration. January 1998 4.0 1.1 7.0 Sourre: Decree 779/95, puiblished in Bol-eti Oficial 28,281 Other heavy-duty vehicles of November 29, 1995. July 1994 11.2 2.4 14.4 January 1996 4.9 1.23 9.0 January 2000 4.0 1.1 7.0 Note: Exhaust emissions are to be tested usinig EEC.s 13- teed by manufacturers for five years or 160,000 iode cycle. On July 1, 1994, crankcase emissions were kilometers, whichever comes first. Alternatively, eliminiated wiith the exception of turbocharged vehicles. this requirement can be replaced by a 10 per- Sourre:Joint Resoltutioni 96/94 and 58/94, published in Boletin Ofirial27,854 of Marclh 21, 1994, andl Decree 779/ cent reduction of emission levels. 95, published in Boletlb OfiriaId 28,281 of November 29, For in-use vehicles with spark-ignition engines 1995. (such as gasoline- or CNG-fueled vehicles), De- cree 875/94 established distinct CO and HC emission standards for vehicles manufactured in 1983-91, 1992-94, and after 1994. Decree 779/ Table 4.52 PM exhaust emission 95 reconfirmed these standards (Table 4.53). standards for new diesel-fueled vehicles For in-use diesel-fueled vehicles, smoke emis- in Argentina sions are regulated by Decree 779/95. This de- (grams per kilowatt-houir) cree, which supersedes Decree 875/94, requires diesel-fueled vehicles that enter circulation af- E ngne poenvr 1'71771t power EJffectivJe d ate IJetowEi 85pkW 1e0'nge I) owl terJuly 1, 1994 to meet the meet the standards Ff.felive ate Ie1ow 5 kW bove 5 kIVspecified in Table 4.54. January 1996 0.680 0.400s January 2000 0.225 ).150 Vehide inspectionprogramts. In 1995 national Law Note: Exhauist emissions are to be tested uising EEC's 13- 24,449 and its Decree 779/95 established the mode cycle. requirement for periodic inspection of in-use Source:Joint Resolition 96/94 and 58/94. piblished in c p Bolelin Oficial 27,854 of March 21, 1994, and Decree 779/ vehles In Argentia. Accordingly, localjurisdic- 95, published in Boletin Ofrcial 28,281 of November 29, tions were given the responsibility of inspecting 1995. private and commercial vehicles providing intrajurisdictional services and the national ju- risdiction of inspecting commercial vehicles pro- viding interjurisdictional services.'" About half the provinces in Argentina have adopted this Table 4.53 Exhaust emission standards decree, and the rest are doing so. Past and for in-use vehicles with spark-ignition present measures taken in the BAMA-which engines in Argentina includesjurisdictions that fall under the respon- C( HC sibility of the Secretariat of Transportation, the Date manutictcured (pterrent) (pplin) City of Buenos Aires, and the Province of Buenos January 1, 1983- Aires-are discussed below. December 31, 1991 4.5 900 January 1, 1992- December 31, 1994 3.0 60()0 AfterJanuary 1, 1995 2.5 400 Note: Measured at idle. Source: Decree 779/95, published in Boleti77 Oicial 28,281 40. Vehicles that provide inteijurisdictional services in of November 29, 1995. Argentinia are those operating between the Federal Capital an(d a province, or between two or more provinces. Case Studies: Buenos Aires 225 The first legislation for periodic inspection of commercial vehicles operating under the na- public transport vehicles operating under the tionaljurisdiction (Galuppo 1997). national jurisdiction was adopted in 1944. This Since October 1994 these periodic inspections was followed by three national regulations (in have been supplemented by unannounced in- 1974, 1980, and 1989) that required inspection spections-conducted at major bus terminals at a central station located in the Federal Capi- and public roads-of urban buses operating tal. This inefficient system required all public within the national jurisdiction. These unan- transport vehicles in Argentina to travel to nounced inspections, which focus on smoke and Buenos Aires. Because the physical capacity of noise emissions, are administered by the Na- this central station was limited, the waiting pe- tional Commission of Transport Control (the riod for inspections was long. Furthermore, since technical arm of the Secretariat of Transporta- inspections were based on visual, manual, and tion) and take into consideration public com- auditive measures, the results were affected by plaints made to the commission's customer thejudgment of the inspectors. During the last service division. Buses that fail the inspection phase of the centralized inspection program are prohibited from circulating tntil the prob- about 1,200 public transport vehicles were in- lem is corrected and the vehicle has been rein- spected each month (GCTV1[ 1994). spected at one of the thirty inspection stations. Inspection of in-use public transport vehicles All commercial vehicles registered in the City operating under the national jurisdiction was of Buenos Aires are required to have a periodic decentralized in 1992 by the Secretariat of inspection atastation owned and operated bya Transportation's Resolution 417, which also ex- private firm, Sociedad Argentina de Control tended the inspections to cover trucks. These Tecnico de Automotores (SACTA), under a ten- inspections included safety and mechanical year concession from the Municipality of Buenos checks as well as measurement of smoke emis- Aires. These inspections aim to ensure compli- sions for vehicles with diesel engines and C,O ance with the safety standards for the vehicles and HC emissions for vehicles with spark-igni- as well as environmental standards for emissions. tion engines. Inspection frequency was set at six Exhaust emission inspections are limited to mea- months for buses and twelve months for trucks, suring CO and HC for CNG- and gasoline-fu- irrespective of the age of the vehicles. Manage- eled vehicles and opacity for diesel-fueled ment of the inspection system, including ap- vehicles. Compliance with all inspection require- proval of private inspection centers, was assigned ments is required to renew the license for taxis to the Executive Consultant for National Trans- and to renew registration for other commercial port (GCTVI 1994).4' vehicles (both of which are issued by SACTA). The first private station to inspect commer- Inspection fees are $37 for taxis, $42 for trucks, cial vehicles operating under the national jur-is- and $52 for buses (Galuppo 1997). diction was opened in February 1993. By SACTA's inspection station, designed for 1,200 November 1994 there were sixty-eight approved vehicles a day, cost $5 million. Operating since private inspection stations in Argentina, each October 1991, the facility covers 12,600 square with an average investment cost for inspection meters and includes 2,200 square meters of cov- equipment of about $80,000 (GCTVI 1994). By ered inspection area with nine inspection lanes mid-1996 there were thirty approved private sta- (three for light-duty vehicles and six for heavy- tions in the BAMA inspecting about 150,000 duty vehicles) and 800 square meters for an administrative building. The facility will be re- turned to the Municipality of Buenos Aires (at no charge) when the concession ends (SACTA 1994). 41. Te Excut,'e Cnsulant fr Naiona Traspor Roadside inspections of commercial vehicles (CENT) is a consulting firm created by the National Tecih- R nological University following an agreemeTit between the registered in the City of Buenos Aires are con- university and the Secretariat of Energy and Transportationl. ducted by SACTA, supported by federal police. CENT's management of inspection stations in Argentina is These inspections aim to verify that vehicles have conducted throughl ani executive director and twelve ie- undergone the required annual inspection and gional auditing entities located at the uniiversity's varions to ensure that they are in good operating condi- academic units throughouit the counitry. Currenitly, CENT s steering committee includes members from the Secretai iat tion (SACTA 1994). of Public Works and Transportation, the uniiversity, and( the Periodic inspection of private cars registered National Commission of Transport Control. in the City of Buenos Aires has not started yet. 226 Chlapter 4 This service will be provided by a private firm no statistics are available, the share of the BAMA concessioned from the Municipality of Buenos in total fuel consumption in Argentina is esti- Aires. The bidding process for the periodic in- mated to be 37 percent for gasoline, 30 percent spection of in-use vehicles is under way accord- for diesel fuel, and 51 percent for CNG. Of the ing to the requirements of national Law 24,449 313 fuel stations serving CNG in the BAMA 47 although no specific legislation has been passed percent are located in the Federal Capital (Petnsa in the Federal Capital (Galuppo 1997). Vehicular, February 15,1997). The Province of Buenos Aires enacted pro- Unleaded gasoline was introduced in 1993. vincial Law 11,430 and its complementary Law Addition of lead to gasoline was eliminated in 11,787, and promulgated Decree 4.103 for the late 1995 as a result of the Argentine govern- inspection of in-use vehicles. The basic require- ment's tax incentive program for petroleum re- ments of this provincial legislation are the same finers. Unleaded gasoline production in as the national legislation. Private cars and trucks Argentina has been achieved by refinery modi- are scheduled for inspection once a year, while fications, including installation of MTBE units taxis, remises, and buses are scheduled for in- atYPF's La Plata and Lujan de Cuyo refineries.42 spection twice a year. The inspection strategy Various Argentine refiners (such as Eg3, Esso, entails dividing the province into eleven zones Refisan, and YPF) have also used the imported (five of which are in the BAMA) and MTBE as a source of oxygenate for increasing concessioning construction and operation of the octane number of gasoline. In 1996 the pre- twenty-five high-volume, high-technology sta- vious leaded gasoline storage, distribution, and tions to eleven private firms (one firm in each transfer systems were used to sell unleaded gaso- zone). Each station consists of four lanes. Only line under the label of "leaded gasoline" to avoid one of the concessioned stations is currently any potential damage to vehicle catalysts by re- operational and the rest are under conistrulction. sidual lead that might have been left in these Annual inspections for 1997 are scheduled be- systems. Although tax incentives to refiners will tween March andJuly according to the last digit be removed, unleaded gasoline production is ex- of the vehicle's license plate number. Inspection pected to increase to satisfy the market demand fees are set at $36 for vehicles less than 2,500 (Alconsult International Ltd. 1996). kilograms in gross weight and $75 for heavier Regulation 54/96 of the Ministry of Economy vehicles (Galuppo 1997). and Public Works and Services specifies gasoline and diesel fuel quality standards in Argentina Fuel-targeted measures. Gasoline, diesel fuel, and effective September 31, 1996. The parameters CNG are the transport fuels available in Argen- specified for gasoline include the research oc- tina. In 1996 motor vehicles in the Federal Capi- tane number, and oxygenates, oxygen, benzene, tal consumed about 14 percent of gasoline, 8 and lead content (Table 4.56). The regulation percent of diesel fuel, and 26 percent of CNG also limits the sulfur content of diesel fuel to a consumed in Argentina (Table 4.55). Although maximum of 0.25 percent by weight and the ce- tane number to a minimum of 48 according to ASTM Method D976 and 50 according to ASTM Method D613. Table 4.55 Fuel consumption in the The Argentine transport sector began using Federal Capital, Province of Buenos CNG in response to a 1985 tax exemption pro- Aires, and Argentina, 1996 gram designed to encourage substitution of CNG (thouisanids of cubic meters) for petroleum fuels. By the end of 1996, 395,800 vehicles in Argentina had their fuel systems con- FelyeCpial .mAie A in verted to use CNG; 265,200 of those vehicles Futel I3se C.a/hail Bluenos Aires Agegnflina Regular gasolinea 190 921 2,402 Super gasolinea 702 1,635 3,943 Diesel fuel 818 4,085 10.706 42. MTBE production capacities are 1,433 barrels a day CNG 278,000 462,000 1.080,000 atthe La Platarefineryand 1,110 barrelsadayatthe Lujani a. Unleaded grades. Production of leaded gasoline wvas de Ctivo refinery. In 1995 the La Plata refinery produced discontinued in late 1995. 885 barrels a day of MTBE and the Lujan de CtGyo refinery Sourre: World Bank data. produced 1,110 barrels a (lay (Alconsilt International Ltd. 1996). Case Studies: Buenos Aires 227 the BAMA are CNG-fueled. The use of CNG in Table 4.56 Gasoline quality standards buses is very limited (less than 200 buses). The in Argentina remaining taxis and urban buses, as well as in- Fitel parameter Standard tercity buses and most trucks, run on diesel fuel. Research octane number (RON; minimium) Some trucks are also fueled with gasoline. For common, normal, or regular gasoline 83 CNG and diesel fuel are less expensive motor For super, extra, or special gasolinle 93 vehicle fuels than gasoline. Both CNG and die- Oxygenates (maximum percent by volume) sel fuel were exempt from fuel taxes until late MTBE 15 1996, when a 59 percent fuel tax was imposed Ethanol 5 on diesel fuel. The fuel tax on gasoline is the Isopropyl alcohol 5 highest among the available motor vehicle fuels Tertiary-butyl alcohiol 7 in Argentina. The fuel tax on various gasoline Isobutyl alcohol 7 grades which ranged from 127 to 136 percent Oxvgen (maximum percent by weight) 2.7 g Benzene (maximum percent by volume) 4 in October 1994, has been increased to between Lead (maximum gram/liter) 165 and 189 percent (Table 4.58). In leaded gasoline 0.2 In unleaded gasoline 0.013 Transport management. Buenos Aires has traffic Swince: Ministry of Economy and Public Works and problems common to large urban centers world- Sernices 1996. wide. Illegal parking of vehicles and the loading and unloading of delivery trucks on busy streets during unauthorized hours forces taxis, buses, tourist buses, and official vehicles to stop in the were in the BAMA (Table 4.57). During 1995-96, middle of the road to drop off and pick up pas- 62 percent of the conversions were for private sengers. The problem is aggravated by taxis cruis- cars, 19 percent for light-duty trucks, 13 percent ing at slow speeds in search of passengers and for taxis, 0.7 percent for official vehicles, and 5 by cars trying to enter parking lots during rush percent for other vehicles. Although most con- hours. versions have been for private cars, it is estimated National Decree 2254 of 1992 required urban that less than 7 percent of private cars in the buses older than 10 years and operating in the BAMA are fueled with CNG. However, CNG is Federal Capital to retire byJanuary 1, 1996. In used by a larger fraction of taxis, remises, and 1994 the national Congress declared a state of other light-duty commercial vehicles. For ex- emergency for transit in the Federal Capital and ample, an estimated 80 to 90 percent of taxis in ordered that solutions to the congestion proh Table 4.57 Conversion of vehicles to CNG in the BAMA and Argentina (thousanids) IA IA Argentina Year Annual Cwunulative Annual Cum2lanihve t984 0.1 0.1 0.1 0.1 1985 1.6 1.7 2.0 2.1 1986 1.8 3.5 3.4 5.5 1987 5.0 8.5 7.9 13.4 1988 5.3 13.8 7.4 20.8 1989 10.6 24.4 14.9 35.7 1990 23.2 47.6 31.9 67.6 1991 27.2 74.8 4t.0 108.6 1992 25.9 100.7 40.5 149.1 1993 35.3 136.0 52.7 201.8 1994 44.3 180.3 67.3 269.1 1995 36.7 217.0 54.7 323.8 1996 48.2 265.2 72.0 395.8 S0ureP: Prensa Vehicular, February 15, 1997. 228 Chapter 4 Table 4.58 Typical fuel prices in Argentina, December 1996 and October 1994 Valtue TTipical Net Fuel added Stalion station Fi(el tYpe tU7nit price tarx tax pnft price December 1996 Extra unleaded gasoline $/liter 0.2940 0.4865 0.0617 0.1538 0.9960 Super unleaded gasoline $/liter 0.2580 0.4865 0.0542 0.1473 0.9460 Regular unleaded gasoline $/liler 0.2152 0.3878 0.0452 0.1078 0.5860 Diesel fuel $/liter 0.2035 0.1200 0.0427 0.0628 0.4290 CNG - 0.0000 - - 0.3060 October 1994 Unieaded gasolinie $/liter 0.2573 0.3496 0.0460 0.1401 0.7930 Super leaded gasoline5 $/liter 0.2660 0.3496 0.0480 0.1234 0.7870 Regular leaded gasoline5 $/liter 0.1975 0.2509 0.0360 0.2006 0.5850 Diesel fuel $/liter 0.1800 0.0000 0.0320 0.0520 0.2640 CNG $/n3 - 0.0000 - - 0.2620 -Not available. a. Productioni of leaded gasoline was discontinuiiedl in late 1995. Source: Prensa V7ehicular October, 15 1994 andi Febmary 15, 1997. lem be investigated. Subsequently, an emergency leges for official vehicles and limiting the wait- committee comprising representatives of the ing period of vehicles that transport money to national Secretariat of Transportation and the or from financial institutions, and restricting the Municipality of Buenos Aires was created. To al- use of some streets to pedestrian traffic. leviate congestion, a weekday traffic restriction, A number of traffic management measures based on the license plate numbers of private were suggested to improve circulation of private cars, was implemented in downtown Buenos cars, taxis, and buses in the macrocenter area of Aires. This restriction was lifted in a few months, Buenos Aires; optimizing the traffic light system however, because it was ineffective. to improve traffic flow in the main arteries; cre- The commission also developed a modal hi- ating a central traffic command center that can erarchy for the microcenter area, that would give modify the programming of traffic lights in case on-foot and metro modes the highest priority, of unanticipated events (such as accidents, dem- followed by buses, taxis, and private cars. A key onstrations, or emergencies), inform motorists element of the proposed transport policy was the of traffic problems, and propose alternative segregation of different transport modes to im- routes; extending the recently implemented re- prove traffic flows. One measure introduced a versible lane system (use of the lanes is based bus lane program that was to be extended to on the main flow of traffic) to other arteries; the main arteries. The program was initiated by separating the lanes used by buses and taxis; in- establishing two short bus lanes on a pilot basis troducing exclusive bus routes and extending (World Bank 1995). In addition, the circulation bus lanes to the main arteries; initiating a bus of private cars in the microcenter area was pro- ticket purchasing system before boarding the hibited between 10 A.M. and 8 P.M. to encourage buses, establishing differential parking tariffs to pedestrian travel on streets with narrow side- discourage long-term parking of private cars at walks. Since Febriary 1995 a traffic restriction the city center and, at the same time, creating has prohibited taxis without passengers from parking facilities in the peripheral zones with entering the microcenter area of Buenos Aires. good access to the city center; designing and Other traffic management measures consid- implementing an effective traffic enforcement ered for the microcenter included allowing bus system; limiting parking of official, diplomatic, traffic only on specified streets and avenues, pro- and medical vehicles to locations that do not moting walking by expanding sidewalks and al- significantly affect the flow of traffic; updating lowing only one lane for buses (except for stops), the traffic network for the truck fleet with prohibiting special permits and parking privi- proper signs, and eliminating freight loading Case Studies: Buenos Aires 229 and loading activities during certain periods; and the number of functioning monitoring stations implementing a public education program for declined every year. In 1994 the effort included motorists. daily monitoring of NO, NO2, and SO2 at a moni- Major infrastructure investments have been toring station in the City of Buenos Aires (in undertaken to link the Federal Capital to its sub- Palermo) and periodic monitoring of TSP and urbs in the Province of Buenos Aires. In 1996 lead by three mobile monitoring stations (DVA construction of highways to the western and 1995). southeastern suburbs was completed. The high- Ambient CO concentrations in the Federal way to Pilar in the west has been operating as a Capital have been monitored by Fundaci6n Ar- toll road and the highwayto LaPlatain the south- gentina Siglo 21 since 1992. Currently, this ef- east has opened to traffic, but the toll booths fort is being conducted at three automatic are yet to be installed. In addition, the highways monitoring stations with continuous sampling to Cafiuelas in the southwest and to Ltjan in at one-minute intervals. Two of these stations are west, both in the Province of Buenos Aires, are located in the port area and the third one in a under construction. These highways also will be narrow street in downtown Buenos Aires. The operated as toll roads. All four of these highways sampling network is maintained by the Univer- are interconnected. sity of Buenos Aires. Since December 1993 moni- Argentine Railways used to be the sole pro- toring results from the downtown station have vider of rail-based passenger and cargo services been disseminated to the public through local in Argentina (except for the Buenos Aires metro, newspapers, television and radio stations, and which was served by Subterraneos de Buenos news agencies. Aires). In 1991 the Argentine government called In addition, since 1993 short-term ambient CO for bids to award operating concessions for vari- concentrations have been monitored by Xilix, a ous rail packages in Argentina to encourage pri- private environmental consulting firm, at the vate participation in the railway sector. Passenger intersections of heavy traffic areas. Five of ther.? services within the BAMA, including the Buenos monitoring locations are in the downtown busi- Aires metro, were reorganized under the Rail- ness area, and one each in the LaRecoleta, Palermo, way Restructuring Coordination Unit. In 1994 a and Once districts of the Federal Capital. twenty-year concession went into effect for the In 1994 Instituto Pro Buenos Aires, in an ef- metro, combined with one suburban railway sys- fort to develop an environmental map of the city, tem, and ten-year concessions for some other measured ambient air concentrations of TSP, suburban railway systems (World Bank 1995). CO, NO2, and HC using a mobile laboratory. In With the concessioning of the two remaining addition, in 1996 Greenpeace conducted an railway systems in January and May 1995, the ambient air quality monitoring study in Buenos entire rail passenger system in the BAMA is now Aires. The results of this study are not available. operated by the private sector. Air quality monitoring. The first ambient air qtual- Evaluation of Implemented Measures ity monitoring in the Federal Capital was con- ducted in 1964 by the Municipality of Buenos Vehicle emission standards. The 1997 exhaust emis- Aires. Until 1973 an ambient air monitoring pro- sion standards for light-duty vehicles, which are gram was jointly carried out by the Municipality based on the USEPA's 1987 emission standards, of Buenos Aires and the national Ministry of will require that vehicles be equipped with three- Public Health. Initially, two indicator pollutants way catalytic converters as well as exhaust gas were monitored each day for about a year. Fol- recirculation and injection timing control tech- lowing evaluation of the monitoring results, a nology. These standards are comparable to the network of twelve stations for monitoring pol- 1992 Chilean and 1994 Mexican standards. The Iltant gases and twenty-one stations for moni- evaporative emission standard of 6 grams/test toring TSP was designed and operated. SO2, NO,, is the same as the Brazilian standard but not as CO, TSP, smoke, and HC were monitored oII a stringent as the Chilean or Mexican standard daily basis. Lead, ozone, and aldehydes were (2 grams/test). monitored less frequently. In 1982 the monitor- The 2000 emission standards for diesel-fueled ing program was considerably weakened by bud- urban buses and 1998 standards for other die- get cuts and personnel reductions. As a result sel-fueled vehicles, which are based on the 230 Cltapter 4 USEPA's 1994 emission standards, require exten- public roads. The buses inspected included 74 sive use of variable fuel injection timing, high percent of the buses that had generated com- injection pressure, low-temperature charge-air plaints from the public. Of the inspected buses, cooling, and combustion optimization. These 1,148 (or 16 percent) were found to be non- standards are comparable to the 1996 Chilean compliant. The noncompliant buses represented standards. 25 percent of the buses denounced by the pub- The CO standards for in-use vehicles with lic. DuringJanuary-May 1995 the noncompli- spark-ignition engines are comparable to the ance rate of inspected buses dropped from 24 Chilean or Mexican standards. The HC stan- to 10 percent (GIIE 1995). Fines for noncompli- dards are more lenient, however. ant buses range from 10,000 to 30,000 times the Because there is no emissions monitoring labo- passenger ticket.43 The collected fines are incor- ratory to perform all the required emission tests porated into the National Commission of Trans- in Argentina, nationally manufactured new ve- port Control's budget. The commission is hicles are being tested in Brazilian state labora- considering lowering the fines, however. tories in the presence of personnel from the The City of Buenos Aires' periodic inspection National Institute of Industrial Technology. For program has had mixed results. In 1994, al- imported vehicles, Argentina requires certifica- though nearly all taxis were inspected, only 10 tion of emissions from an official vehicle testing percent of the registered trucks and 8 percent laboratory in the country of origin. Establish- of the registered passenger transport vehicles ment of a certification laboratory is under con- were inspected at SACTA's inspection station sideration. (SACTA 1994). To reduce noncompliance, the municipal Secretariat of Urban Planning and Vehicle inspectionprograms. A 1991 survey of gaso- Environment, along with the municipal police, line-fueled vehicles on public roads in the Fed- carry out sporadic road checks. In April 1994, eral Capital found that 66 percent of the vehicles 235 of 898 randomly selected vehicles (26 per- were in violation of the CO emission standard. cent) failed to meet the emission standards. The A similar survey in 1992 found that 90 percent rate of noncompliance was 17 percent for buses of gasoline-fueled vehicles had CO emissions (80 of 483), 38 percent for trucks (96 of 252), over the standard (Moran 1992). Although gaso- and 36 percent for taxis (59 of 163; World Bank line-fueled private cars have been the main 1995). Strengthening inspections for heavy-duty source of air emissions in the Federal Capital vehicles would encourage compliance with the and certain parts of the BAMA, until recently emission standards and improve air quality. they were excluded from the periodic inspec- tion and maintenance program. The recent pe- Fuel-targeted measures. Fuel quality standards in riodic inspection requirement for private cars Argentina establish the quality requirements of should reduce polhltant emissions. The scheme gasoline and diesel fuel produced by Argentine adopted in the Federal Capital and the Province refiners or imported to Argentina. These stan- of Buenos Aires will allow for efficieiit provision dards are intended to ensure supply of fiuels ca- of these services at highly automated, high-volume pable of meeting the 1997 motor vehicle inspections by private concessionaires. emission standards and to reduce pollutant con- The regular and unannounced inspections centrations in ambient air. In 1996 gasoline pro- and heavy fines have played a major role in re- duced by different Argentine refiners contained ducing the number of noncompliant buses and 3 to 15 percent MTBE by volume, 0.47 percent trucks in circulation. Between February 1993 and to less than 3 percent benzene by volume, and a November 1994,33,500 buses and 82,000 trucks maximum of 0.013 g/liter of lead (Alconsult operating within the national jurisdiction were International Ltd. 1996). These parameters met inspected at sixty-eight inspection stations ap- the Argentine quality standards for gasoline (see proved by the Executive Consultant for National Table 4.58). Transport. In the first nine months of 1994 an average of 3,015 buses and 8,570 trucks were in- spected each month (GCTVI 1994). In addition, between October 1994 and May 1995. 7,268 ur- 43.The average ticketprice btses operatingbetween the ban buses operating within the national juris- Federal Capital and the suburbs in the Province of Buenos diction were inspected at bus terminals and on Aires is $0.65. Case Studies: Buenos Aires 231 Elimination of lead from gasoline must have by Refsan) was higher than allowed in the reduced ambient lead concentrations in the United States (the 1990 baseline value for BAMA, especially in the Federal Capital where reformulated gasoline is 28.6 percent by public exposure is highest. No recent ambient volume). Reduction of aromatics content air quality data are yet available, however, to con- would reduce emissions of NO., HC, and firm this assumption. benzene. To make up for the octane that previously was * Gasoline produced by Esso had a much provided by tetraethyl lead, Argentine refiners higher sulfur content (1,000 ppm) than have resorted to the addition of MTBE and, pos- gasoline produced by Eg3 (181 ppm to 288 sibly, to some refinery modifications in refomiu- ppm) or YPF (less than 500 ppm), or the lating gasoline." Because no leaded gasoline is 1990 baseline value of gasoline in the presently available in Argentina, vehicles not United States (338 ppm). Higher stilfur equipped with catalytic converters are using un- content reduces the efficiency of catalytic leaded gasoline. YPF is investigating the emis- converters. sion effects of using unleaded gasoline in uncontrolled vehicles. Although this investiga- As a result of the government's fuel taxing tion is of major interest to YPF, a parallel study policy CNG is the least expensive motor vehicle by an independent group would ensure the pub- fuel in Argentina, and is especially popular for lic credibility of the findings. Based on a 1996 use in taxis circulating in Buenos Aires (see Table study of gasoline quality in Argentina (Alconsult 4.58). A typical fuel cost for a taxi is about $0.04/ International Ltd. 1996), the following param- kilometer if operated with CNG and $0.1 1/kilo- eters are noted to have higher values than gaso- meter if operated with gasoline. In addition, line in the United States: newer CNG-fueled taxis cost about 30 percent less than diesel-fueled taxis. This requires install- @ The Reid vapor pressure of medium- and ing a CNG tank and associated connections and high-octane gasoline produced by YPF a new fueling system. More than half of the 40 (11.5 psi) and Eg3 (10 psi) was higher than or so conversion kits in the market are locally allowed in the United States (the 1992 lim- produced. The conversion costs $1,000 to $1,500, its for unleaded gasoline during summer depending on the size and weight of the CNG months are 9.0 psi for the northern states tank. and 7.8 psi for most of the southern states; Although gasoline is more expensive than the limits for reformulated gasoline are CNG, mainly because of the differential tax even stricter). Lowering the vapor pressure policy applied to these two fuels, it still generally in gasoline would reduce volatile HC emis- is not preferred by private car owners because sions that contribute to ozone formation. the CNG tank is heavy (a minimum of 80 kilo- * The olefinic content of low-octane gasoline grams) and takes up space. Choosing to convert produced by Eg3 (22.3 percent by volume) private cars to CNG is often associated with ex- was higher than allowed in the United tensive travel needs, where the economy of CNG States (the 1990 baseline value for refor- outweighs its inconvenience and conversion cost. mulated gasoline is 10.8 percent by vol- Although CNG-fueled buses are less polluting uime). Reduction of olefinic compounds than diesel-fueled buses in terms of NO., CO, would reduce emissions of reactive HC that and PM emissions, not even 2 percent of the contribute to ozone formation. urban buses operating in the BAMA are fueled * The aromatics content of medium-octane with CNG. CNG-fueled buses are not generally gasoline produced by Eg3 (44.4 percent by favored mainly because fuel price incentives to volume) and high-octane gasoline pro- convert diesel-fueled buses to CNG were not cre- duced by most refiners (35.6 percent by ated until recently (see Table 4.58). In addition, YPF, 37.5 percent by Eg3, and 40.0 percent CNG-fueled buses have lower resale value than diesel-fueled buses (because CNG service stations are not very common outside of urban centers), 44. YPF is producing MTBE at its La Plata and Liijani (le companies operating CNG-fueled buses must Cuvo refineries, and in the near future will produce TAME invest about $1.5 million to have their own CNG at its La Plata refinery. EG3, Esso, and Refisan are ptircias- filling stations, CNG tanks need to be filled daily, ing MTBE (Alconsult International Ltd.). it is slower to fill a CNG tank than a diesel fuel 232 Chlapter 4 tank, CNG cylinders add extra weight to the bus (as much as 1,200 kilograms), and CNG-fueled Table 4.59 Passenger trips in the BAMA, buses cost more than diesel-fueled buses (by 1970, 1992, and 1994 about 15 to 20 percent). Until 1995 the CNG (percent) program caused a 12 percent reduction in con- 1ode,e 1970 1992 1994 sumption of diesel fuel in favor of CNG in the Private car 15.4 25.0 32.4 BAMA, which should correspond to a 6 percent Taxi 6.7 3.2 4.0 reduction in PM emissions (World Bank 1995). Bus 54.3 56.7 47.1 The fuel tax adjustment increasing the price of Train 7.0 3.9 4.9 diesel fuel relative to CNG is expected to en- Metro 5.4 2.7 4.5 courage more diesel-fueled vehicles to convert Othera 11.2 8.5 7.1 to CNG, and therefore result in lower PM, CO, Total 100.0 100.0 100.0 and NO, emissions. Number of daily passengers (millions) 17.4 18.1 17.6 T*-ansport management. The overlappingjurisdic- a. Includes trips by foot, charter and school bus, tion of the federal government, the Province of motorcycle, and bicycle. Buenos Aires, and the Municipality of Buenos Aires complicates the traffic management in the BAMA. This situation has created inefficiencies in cooperation, planning, operation, administra- attributed to the deteriorating service provided tion, regulation, and control of the transport by metro and suburban trains under public sec- system, yielding high social costs in the form of tor management. Between 1992 and 1994, how- congestion, pollution, suboptimal investment, ever, the metro and suburban train ridership and decreasing levels of services. Coordination increased by 62 and 22 percent, respectively. Rid- and planning of transport services among the ership increased further in 1995 and 1996. For agencies responsible for road management and example, ridership for the first nine months of public transport provision from all three juris- 1996 was up by 38 percent for metro and 93 per- dictions in the BAMA is necessary to improve cent for the suburban railway system as com- transport management and reduce vehicular air pared to the same period in 1993. Although part emissions. of these increases is due to a reduction in fare Although the number of passenger trips made evasion, privatizing these systems improved their in the BAMA in 1970, 1992, and 1994 did not service reliabilityand security (particularly atsta- change significantly, the number of trips made tions) at no increase in fares. This is confirmed by private cars is increasing (Table 4.59). Pas- by a 1995 government survey, which found that senger trips by private cars increased by 69 per- 96 percent of the riders on four suburban lines cent between 1970 and 1992 (2.4 percent a year) felt the service was as good as or better than be- and by 26 percent between 1992 and 1994 (12.3 fore, with a majority stating that it was better. percent a year). The sharp increase in recent The 1994 traffic restriction based on the li- years can be attributed to greater ownership of cense plates of private cars was lifted because it private cars. Since 1991 the number of private did not yield the desired results. Although traf- cars in circulation in the BAMA has increased fic restrictions in the microcenter area of Buenos by about 100,000 a year because of higher in- Aires have reduced congestion, they have not comes, a new financing system for cars (four-to been totally successful because cars with special five-year loans with an annual interest rate of 15 permits and official vehicles continue to circu- to 20 percent), and stable car prices resulting late, and traffic enforcement has not been strict. from an agreement among the government, car In addition, circulation of certain bus lines in manufacturers, dealers, and automotive indus- the microcenter area has contributed to conges- try unions. tion. Pilot bus lanes have proved worthwhile and Between 1970 and 1992 the share of public should be extended to other arteries. Expand- transport in total passenger trips fell slightly, by ing the restricted zone for private cars and about 1 percent (see Table 4.59). Although btis strengthening enforcement also would help re- ridership increased by about 9 percent. metro duce congestion and air pollution in Buenos and suburban train ridership decreased by 48 Aires. Furthermore, better traffic engineering and 45 percent, respectively. This drop can be measures (such as elimination of bottlenecks and Case Studies: Buenos Aires 233 signalization) as well as measures to increase the sity to be controlled. When the toll road to the attractiveness of public transport (such as reha- southwest and the second toll road to the west bilitation of the metro and suburban trains) are completed, congestion and vehicular air would achieve the same objectives. emissions are expected to be further reduced in With the bus retirement program adopted in the BAMA. 1992, about 3,500 urban buses 10 years and older were taken out of circulation byJanuary 1996. Air quality monitoring. Air quality data for Buenos This measure has reduced air pollutant enis- Aires is limited to manual sampling of NO,, TSP, sions from the bus fleet. lead, and SO2 at one monitoring station; con- Compared with regular buses, diferencial tinuous sampling of CO at three monitoring sta- buses offer faster and more comfortable service tions; and short-term monitoring of CO at heavy to commuters because they stop less frequently traffic intersections. In addition, air quality data and guarantee a seat for every passenger. The for lead are available for certain months of the fares for diferencial buses are higher than those year. However, data is not collected for certain for regular buses. For suburban commuters, key air pollutants such as ozone. An air quality charter buses provide faster and more conve- monitoring network needs to be established that nient service because of reliable schedules and includes all the above-mentioned pollutants and guaranteed seats. covers the entire BAMA. An emissions inventory The new toll road to Pilar and the highway to for the Federal Capital and BAMA also needs to La Plata have increased travel speeds and re- be developed. The data generated from these lieved the traffic load on alternate roads. As a efforts would allow determination of potential result, emission of pollutants from vehicular exposure levels of air pollutants and design of a sources must have decreased. Establishing tolls rational air pollution control strategy for the on the highway to La Plata will allow traffic den- BAMA. RiO DE JANEIRO 22?0C' 430' 300 134P30' 2 Ioo MINAS GERAIS BRAZIL RIO DE JANEIRO METROPOLITAN REGION ...* AIR QUALITY MONITORING STATIONS Pog < (aroibo, JJ r > sN a.AIR SHEDS f P ,7 / do Sul ,. RIVERS AND STREAMS O SELECTED CITIES AND TOWNS ®E STATE CAPITAL _/ ,F. ....f.T. MUNICIPALITY BOUNDARIES' / < Ovole.;. ~ \ | 2 X z _ RIO DE JANEIRO METROPOLITAN AREA BOUNDARY - v - - - STATE ROUNDARIES INTERNATIONAL BOUNDARIES (INSET) X s , f ) ,/ IPETROPOIIS 'WOnly .n.icLp.litei to-nd jithin t1h. M -ropofl-on A or- dopidId. : 9 o,6 2 R 180 ~~~~~~~~~~~~~~~~~~~~~~~~~D E JAN I R O ( Teres6polis 22-30 ~ \ ac oeir .