AW~~~N ..X 4.l'*'; N- jo t _t y -g. . 14 * o, 6 - J,; '>;" j -;is | F;1S ' I ~,...s_...... ..,,_j-.>|_ | j | l , ; ; ; e _; i; B ~~~~~~~~~~. ; ., 6 | 3| i l ; X N r e .4 .4 .|r i ! _ Is ' s . . . ' e; ; |~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ; i li -- ! r; S X : £:>e X; $ _ ' ! : i:~~F 2- .|- J,.,....... :'~~ 1r f''" ' '^ ' . 'e ' ' c Air Pollution from Motor Vehicles Standards and Technologies for Controlling Emissions Air Pollution from Motor Vehicles Standards and Technologies for Controlling Emissions Asif Faiz Christopher S. Weaver Michael P.Walsh With contributions by Surhid P Gautam Lit-Mian Chan The World Bank Washington, D.C. © 1996 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 November 1996 The findings, interpretations, and conclusions expressed in this publication are those of the authors and do not necessarily represent the views and policies of the World Bank or its Board of Executive Directors or the countries they represent. Some sources cited in this paper may be informal documents that are not readily available. 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 dissemination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy por- tions for classroom use is granted through the Copyright Clearance Center, Inc., Suite 910, 222 Rosewood Drive, Danvers, Massachusetts 01923, U.S.A. The complete backlist of publications from the World Bank is shown in the annual Index of Publications, which contains an alphabetical title list (with full ordering information) and indexes of subjects, authors, and countries and regions. The latest edition is available free of charge from Distribution Unit, Office of the Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, The World Bank, 66, avenue d'I6na, 75116 Paris, France. Cover photos: Asif Faiz Asif Faiz is currently chief of the Infrastructure and Urban Development Operations Division of the World Bank's Latin America and the Caribbean Country Department I. Christopher S. Weaver and Michael P. Walsh coauthored this book as consultants to the World Bank. Library of Congress Cataloging-in-Publication Data Faiz, Asif. Air pollution from motor vehicles: standards and technologies for controlling emissions/ Asif Faiz, Christopher S. Weaver, Michael P. Walsh, with contributions by Surhid Gautam and Lit-Mian Chan. P. cm. Includes bibliographical references (p. ). ISBN 0-8213-3444-1 1. Motor vehicles-Pollution control devices. 2. Automobiles- Motors-Exhaust gas-Law and legislation-United States. I. Weaver, Christopher S. II. Walsh, Michael P. III. Title. TL214.P6F35 1996 363.73'1-dc2O 95-37837 CIP Contents Preface xlii Acknowledgments xvii Participants at the lTNEP Workshop xix Chapter 1 Emission Standards and Regulations 1 International Standards 2 US Standards 2 UN Economic Commission for Europe (ECE) and European Union (EU) Standards 6 Country and Other Standards 9 Argentina I 1 Australia II Brazil 12 Canada 13 Chile 14 Cbina 15 Colombia 15 Eastern European Countries and the Russian Federation 15 Hong Kong 16 India 1 7 Japan 18 Republic of Korea 18 Malaysia 19 Mexico 19 SaudiArabia 19 Singapore 19 Taiwan (China) 20 Thailand 20 Compliance with Standards 21 Certiffcation or Type Approval 21 Assembly Line Testing 22 In-Use Surveillance and Recall 22 Warranty 23 On-Board Diagnostic Systems 23 Alternatives to Emission Standards 23 References 24 Chapter 2 Quantifying Vehicle Emissions 25 Emissions Measurement and Testing Procedures 25 Exhaust Emissions Testing for Light-Duty Vehicles 25 Exhaust Emissions Testing for Motorcycles and Mopeds 29 Exhaust Emissions Testingfor Heavy-Duty Vehicle Engines 29 V vi Air Pollution from Motor Vehicles Crankcase Emissions 32 Evaporative Emissions 32 Refueling Emissions 33 On-Road Exhaust Emissions 33 Vehicle Emission Factors 33 Gasoline-Fueled Vehicles 37 Diesel-Fueled Vehicles 39 Motorcycles 43 References 46 Appendix 2.1 Selected Exhaust Emission and Fuel Consumption Factors for Gasoline-Fueled Vehicles 49 Appendix 2.2 Selected Exhaust Emission and Fuel Consumption Factors for Diesel-Fueled Vehicles 57 Chapter 3 Vehicle Technology for Controlling Emissions 63 Automotive Engine Types 64 Spark-Ignition (Otto) Engines 64 Diesel Engines 64 Rotary (Wankel) Engines 65 Gas-Turbine (Brayton) Engines 65 Steam (Rankine Engines) 65 Stirling Engines 65 Electric and Hybrid Vehicles 65 Control Technology for Gasoline-Fueled Vehicles (Spark-Ignition Engines) 65 Air-Fuel Ratio 66 Electronic Control Systems 66 Catalytic Converters 67 Crankcase Emissions and Control 67 Evaporative Emissions and Control 67 Fuel Dispensing/Distribution Emissions and Control 69 Control Technology for Diesel-Fueled Vehicles (Compression-Ignition Engines) 69 Engine Design 70 Exhaust Aftertreatment 71 Emission Control Options and Costs 73 Gasoline-Fueled Passenger Cars and Light-Duty Trucks 73 Heavy-Duty Gasoline-Fueled Vehicles 76 Motorcycles 76 Diesel-Fueled Vehicles 76 References 79 Appendix 3.1 Emission Control Technology for Spark-Ignition (Otto) Engines 81 Appendix 3.2 Emission Control Technology for Compression-Ignition (Diesel) Engines 101 Appendix 3.3 The Potential for Improved Fuel Economy 119 Chapter 4 Controlling Emissions from In-Use Vehicles 127 Inspection and Maintenance Programs 127 Vehicle Types Covered 129 Inspection Procedures for Vehicles with Spark-lgnition Engines 130 Exhaust Emissions 131 Evaporative Emissions 133 Motorcycle White Smoke Emissions 133 Inspection Procedures for Vehicles with Diesel Engines 133 Institutional Setting for Inspection and Maintenance 135 Centralized I/M 136 Decentralized lIM 137 Comparison of Centralized and Decentralized IIM Programs 138 Inspection Frequency 140 Vehicle Registration 140 Roadside Inspection Programs 140 Contents vii Emission Standards for Inspection and Maintenance Programs 141 Costs and Benefits of Inspection and Maintenance Programs 144 Emission Improvements and Fuel Economy 149 Impact on Tampering and Misfueling 151 Cost-Effectiveness 153 International Experience with Inspection and Maintenance Programs 154 Remote Sensing of Vehicle Emissions 159 Evaluation of Remote-Sensing Data 162 On-Board Diagnostic Systems 164 Vehicle Replacement and Retrofit Programs 164 Scrappage and Relocation Programs 165 Vehicle Replacement 165 Retrofit Programs 166 Intelligent Vehicle-Highway Systems 167 References 168 Appendix 4.1 Remote Sensing of Vehicle Emissions: Operating Principles, Capabilities, and Limitations 171 Chapter 5 Fuel Options for Controlling Emissions 175 Gasoline 176 Lead and Octane Number 176 Fuel Volatility 179 Olefins 180 Aromatic Hydrocarbons 180 DistiUation Properties 181 Oxygenates 182 Sulfur 183 Fuel Additives to Control Deposits 184 Reformulated GasolUne 184 Diesel 186 Sulfur Content 187 Cetane Number 188 Aromatic Hydrocarbons 188 Other Fuel Properties 189 Fuel Additives 190 Effect of Diesel Fuel Properties on Emissions: Summary of EPEFE Results 191 Alternative Fuels 193 Natural Gas 195 Liquefied Petroleum Gas (LPG) 200 Methanol 202 Etbanol 204 Blodiesel 206 Hydrogen 210 Electric and Hybrid-Electric Vehicles 211 Factors Influencing the Large-Scale Use of Alternative Fuels 213 Cost 213 End-Use Considerations 215 Lffe-Cycle Emissions 216 Conclusions 218 References 219 Appendix 5.1 International Use of Lead in Gasoline 223 Appendix 5.2 Electric and Hybrid-Electric Vehicles 227 Appendix 5.3 Alternative Fuel Options for Urban Buses in Santiago, Chile: A Case Study 237 Abbreviations and Conversion Factors 241 Country Index 245 viii Ar Polutionfrom Motor Vehicles Boxes Box 2.1 Factors Influencing MotorVehicle Emissions 34 Box 2.2 Development of Vehicle EmissionsTesting Capability inThailand 36 Box 3.1 Trap-Oxidizer Development in Greece 72 Box A3.1 .1 Compression Ratio, Octane, and Fuel Efficiency 90 Box 4.1 Effectiveness of California's Decentralized Smog Check" Program 128 Box 4.2 Experience with British Columbia's AirCare I/M Program 129 Box 4.3 On-Road Smoke Enforcement in Singapore 142 Box 4.4 ReplacingTrabants andWartburgs with CleanerAutomobiles in Hungary 167 Box 5.1 Gasoline Blending Components 176 Box 5.2 Low-Lead Gasoline as aTransitional Measure 178 Box 5.3 Use of Oxygenates in Motor Gasolines 182 Box 5.4 CNG in Argentina: An Alternative Fuel for Buenos Aires Metropolitan Region 196 Box 5.5 Brazil's 199OAlcohol Crisis: the Search for Solutions 207 Box 5.6 Electric Vehicle Program for Kathmandu, Nepal 214 Box 5.7 Ethanol in Brazil 216 Box 5.8 Compressed Natural Gas in New Zealand 217 Figures Figure 2.1 Exhaust Emissions Test Procedure for Light-Duty Vehicles 26 Figure 2.2 Typical Physical Layout of an EmissionsTesting Laboratory 27 Figure 2.3 U.S. EmissionsTest Driving Cycle for Light-DutyVehicles (FTP-75) 27 Figure 2.4 Proposed U.S. Environmental ProtectionAgency US06 EmissionsTest Cycle 28 Figure 2.5 European Emissions Test Driving Cycle (ECE-1 5) 30 Figure 2.6 European Extra-Urban Driving Cycle (EUDC) 30 Figure 2.7 European Emissions Test Driving Cycle for Mopeds 31 Figure 2.8 Relationship between Vehicle Speed and Emissions for Uncontrolled Vehicles 35 Figure 2.9 Effect of Average Speed on Emissions and Fuel Consumption for European Passenger Cars without Catalyst (INRETS Driving Cycles; Fully Warmed-Up In-use Test Vehicles) 39 Figure 2.10 Cumulative Distribution of Emissions from Passenger Cars in Santiago, Chile 40 Figure 2.11 Effect of Average Speed on Emissions and Fuel Consumption for Heavy-Duty Swiss Vehicles 42 Figure 2.12 Effect of Constant Average Speed and Road Gradient on Exhaust Emissions and Fuel Consumption for a 40-ton Semi-TrailerTruck 43 Figure 2.13 Cumulative Distribution of Emissions from Diesel Buses in Santiago, Chile 44 Figure 2.14 Smoke Opacity Emissions from Motorcycles in Bangkok,Thailand 46 Figure 3.1 Effect of Air-Fuel Ratio on Spark-Ignition Engine Emissions 66 Figure 3.2 Types of Catalytic Converters 68 Figure 3.3 Effect of Air-Fuel Ratio on Three-Way Catalyst Efficiency 69 Figure 3.4 Hydrocarbon Vapor Emissions from Gasoline Distribution 70 Figure 3.5 Nitrogen Oxide and Particulate Emissions from Diesel-Fueled Engines 71 Figure A3.1.1 Combustion in a Spark-Ignition Engine 81 Figure A3.1.2 Piston and Cylinder Arrangement of a Typical Four-Stroke Engine 84 Figure A3.1.3 Exhaust Scavenging in a Two-Stroke Gasoline Engine 85 Figure A3.1.4 Mechanical Layout of a Typical Four-Stroke Engine 86 FigureA3.1.5 Mechanical Layout of aTypical Two-Stroke Motorcycle Engine 86 Figure A3.1.6 Combustion Rate and Crank Angle for Conventional and Fast-Burn Combustion Chambers 89 Contents im. Figure A3.2.1 Diesel Combustion Stages 102 FigureA3.2.2 Hydrocarbon and Nitrogen Oxide Emissions for Different Types of Diesel Engines 103 FigureA3.2.3 Relationship betweenAir-Fuel Ratio and Emissions for a Diesel Engine 106 Figure A3.2.4 Estimated PM-NO,Trade-Off overTransientTest Cycle for Heavy-Duty Diesel Engines 109 Figure A3.2.5 Diesel Engine Combustion ChamberTypes 110 Figure A3.2.6 Bus Plume Volume for Concentration Comparison between Vertical and Horizontal Exhausts 116 Figure A3.2.7 Truck Plume Volume for Concentration Comparison between Vertical and Horizontal Exhausts 116 Figure A3.3.1 Aerodynamic Shape Improvements for an Articulated Heavy-Duty Truck 120 Figure A3.3.2 TechnicalApproaches to Reducing Fuel Economy of Light-DutyVehicles 121 Figure 4.1 Effect of Maintenance on Emissions and Fuel Economy of Buses in Santiago, Chile 130 Figure 4.2 Schematic Illustration of the IM240Test Equipment 132 Figure 4.3 Bosch Number Compared with Measured Particulate Emissions for Buses in Santiago, Chile 134 Figure 4.4 Schematic Illustration of a Typical Combined Safety and Emissions Inspection Station: Layout and Equipment 137 Figure 4.5 Schematic Illustration of an Automated Inspection Process 138 Figure 4.6 Cumulative Distribution of CO Emissions from Passenger Cars in Bangkok 143 Figure 4.7 Cumulative Distribution of Smoke Opacity for Buses in Bangkok 143 Figure 4.8 Illustration of a Remote Sensing System for CO and HC Emissions 160 Figure 4.9 Distribution of CO Concentrations Determined by Remote Sensing of Vehicle Exhaust in Chicago in 1990 (15,586 Records) 161 Figure 4.10 Distribution of CO Concentrations Determined by Remote Sensing of Vehicle Exhaust in Mexico City 161 Figure 4.11 Distribution of HC Concentrations Determined by Remote Sensing of Vehicle Exhaust in Mexico City 161 Figure 5.1 Range of Petroleum Products Obtained from Distillation of Crude Oil 186 Figure 5.2 A Comparison of the Weight of On-Board Fuel and Storage Systems for CNG and Gasoline 199 FigureA5.2.1 Vehicle Cruise Propulsive Power Required as a Function of Speed and Road Gradient 228 Tables Table 1.1 Progression of U.S. Exhaust Emission Standards for Light-Duty Gasoline-Fueled Vehicles 3 Table 1.2 U.S. Exhaust Emission Standards for Passenger Cars and Light-Duty Vehicles Weighing Less than 3,750 PoundsTest Weight 4 Table 1.3 U.S. Federal and California Motorcycle Exhaust Emission Standards 5 Table 1.4 U.S. Federal and California Exhaust Emission Standards for Medium-Duty Vehicles 6 Table 1.5 U.S. Federal and California Exhaust Emission Standards for Heavy-Duty and Medium-Duty Engines 7 Table 1.6 European Emission Standards for Passenger Cars with up to 6 Seats 9 Table 1.7 European Union 1994 Exhaust Emission Standards for Light-Duty Commercial Vehicles (Ministerial Directive 93/59/EEC) 10 Table 1.8 ECE and Other European Exhaust Emission Standards for Motorcycles and Mopeds 10 Table 1.9 Smoke Limits Specified in ECE Regulation 24.03 and EU Directive 72/306/EEC 11 Table 1.10 European Exhaust Emission Standards for Heavy-Duty Vehicles forType Approval 11 Table 1.11 Exhaust Emission Standards (Decree 875/94), Argentina 12 Table 1.12 Exhaust Emission Standards for MotorVehicles, Australia 13 Table 1.13 Exhaust Emission Standards for Light-DutyVehicles (FTP-75Test Cycle),Brazil 13 Table 1.14 Exhaust Emission Standards for Heavy-DutyVehicles (ECE R49Test Cycle), Brazil 14 Table 1.15 Exhaust Emission Standards for Light- and Heavy-Duty Vehicles, Canada 14 Table 1.16 Exhaust Emission Limits for Gasoline-Powered Heavy-Duty Vehicles (1983), China 15 Table 1.17 Proposed Exhaust Emission Limits for Gasoline-Powered Heavy-Duty Vehicles, China 16 Table 1.18 List of Revised or New Emission Standards and Testing Procedures, China (Effective 1994) 16 x Air PoTlutlonfmm Motor Vebicles Table 1.19 Emission Limits for Gasoline-Fueled Vehicles for Idle and Low Speed Conditions, Colombia 16 Table 1.20 Exhaust Emission Standards for Gasoline- and Diesel-Fueled Vehicles, Colombia 17 Table 1.21 Summary of Vehicle Emission Regulations in Eastern Europe 17 Table 1.22 Exhaust Emission Standards for Gasoline-Fueled Vehicles, India 18 Table 1.23 Motorcycle Emission Standards, Republic of Korea 18 Table 1.24 Emission Standards for Light-DutyVehicles, Mexico 19 Table 1.25 Exhaust Emission Standards for Light-DutyTrucks and Medium-DutyVehicles by Gross Vehicle Weight, Mexico 20 Table 1.26 Exhaust Emission Standards for Motorcycles,Taiwan (China) 21 Table 1.27 Exhaust Emission Standards, Thailand 21 Table 2.1 Estimated Emission Factors for U.S. Gasoline-Fueled Passenger Cars with Different Emission Control Technologies 37 Table 2.2 Estimated Emission Factors for U.S. Gasoline-Fueled Medium-Duty Trucks with Different Emission Control Technologies 38 Table 2.3 Estimated Emission and Fuel Consumption Factors for U.S. Diesel-Fueled Passenger Cars and Light- Duty Trucks 41 Table 2.4 Estimated Emission and Fuel Consumption Factors for U.S. Heavy-Duty Diesel-Fueled Trucks and Buses 41 Table 2.5 Emission and Fuel Consumption Factors for Uncontrolled U.S.Two- and Four-Stroke Motorcycles 45 Table 2.6 Emission Factors for Uncontrolled European Motorcycles and Mopeds 45 Table 2.7 Emission and Fuel Consumption Factors for UncontrolledThai Motorcycles 45 TableA2.1.1 Exhaust Emissions, European Vehicles, 1970-90 Average 49 TableA2.1.2 Exhaust Emissions, European Vehicles, 1995 Representative Fleet 49 TableA2.1.3 Estimated Emissions and Fuel Consumption, European Vehicles, Urban Driving 50 TableA2.1.4 Estimated Emissions and Fuel Consumption, European Vehicles, Rural Driving 51 TableA2.1.5 Estimated Emissions and Fuel Consumption, European Vehicles, Highway Driving 52 Table A2.1.6 Automobile Exhaust Emissions, Chile 53 Table A2.1.7 Automobile Exhaust Emissions as a Function ofTest Procedure and Ambient Temperature, Finland 53 TableA2.1.8 Automobile Exhaust Emissions as a Function of Driving Conditions, France 53 Table A2.1.9 Automobile Exhaust Emissions and Fuel Consumption as a Function of Driving Conditions and Emission Controls, Germany 53 TableA2.1.10 Exhaust Emissions, Light-Duty Vehicles and Mopeds, Greece 54 Table A2.1.11 Hot-Start Exhaust Emissions, Light-Duty Vehicles, Greece 54 TableA2.1.12 Exhaust Emissions, Light-Duty Vehicles and 2-3 Wheelers, India 54 TableA2.2.1 Exhaust Emissions, European Cars 57 TableA2.2.2 Estimated Emissions and Fuel Consumption, European Cars and Light-Duty Vehicles 57 Table A2.2.3 Estimated Emissions, European Medium- to Heavy-Duty Vehicles 58 Table A2.2.4 Exhaust Emissions, European Heavy-Duty Vehicles 58 Table A2.2.5 Exhaust Emissions and Fuel Consumption, Utility and Heavy-Duty Trucks, France 58 TableA2.2.6 Exhaust Emissions, Santiago Buses, Chile 59 Table A2.2.7 Exhaust Emissions, London Buses, United Kingdom 59 Table A2.2.8 Exhaust Emissions, Utility and Heavy-Duty Vehicles, Netherlands 59 Table A2.2.9 Automobile Exhaust Emissions as a Function of Driving Conditions, France 59 Table A2.2.10 Automobile Exhaust Emissions and Fuel Consumption as a Function ofTesting Procedures, Germany 60 TableA2.2.11 Exhaust Emissions, Cars, Buses, and Trucks, Greece 60 Table A2.2.12 Exhaust Emissions, Light-Duty Vehicles and Trucks, India 60 Contents xi Table 3.1 Automaker Estimates of Emission Control Technology Costs for Gasoline-Fueled Vehicles 74 Table 3.2 Exhaust Emission Control Levels for Light-Duty Gasoline-Fueled Vehicles 75 Table 3.3 Recommended Emission Control Levels for Motorcycles in Thailand 76 Table 3.4 Industry Estimates of Emission Control Technology Costs for Diesel-Fueled Vehicles 77 Table 3.5 Emission Control Levels for Heavy-Duty Diesel Vehicles 78 Table 3.6 Emission Control Levels for Light-Duty Diesel Vehicles 78 Table A3.1.1 Effect of Altitude on Air Density and Power Output from Naturally Aspirated Gasoline Engines in Temperate Regions 87 TableA3.1.2 Cold-Start and Hot-Start Emissions with Different Emission ControlTechnologies 91 TableA3.1.3 Engine Performance and Exhaust Emissions for a Modified Marine Two-Stroke Engine 93 TableA3.1.4 Exhaust Emissions and Fuel Economy for a Fuel-Injected Scooter 94 TableA3.1.5 Moped Exhaust Emissions 97 TableA3.3.1 Energy Efficiency of Trucks in Selected Countries 122 Table A3.3.2 International Gasoline and Diesel Prices 124 Table A3.3.3 Gasoline Consumption byTwo- andThree-Wheelers 125 Table 4.1 Characteristics of Existing I/M Programs for Heavy-Duty Diesel Vehicles in the United States, 1994 136 Table 4.2 Estimated Costs of Centralized and Decentralized I/M Programs in Arizona, 1990 139 Table 4.3 Schedule of Compulsory Motor Vehicle Inspection in Singapore by Vehicle Age 141 Table 4.4 Inspection and Maintenance Standards Recommended forThailand 145 Table 4.5 Distribution of Carbon Monoxide and Hydrocarbon Emissions from 17,000 Short Tests on Gasoline Cars in Finland 145 Table 4.6 In-Service Vehicle Emission Standards in the European Union, 1994 146 Table 4.7 In-Service Vehicle Emission Standards in Argentina, New Zealand, and East Asia,1994 147 Table 4.8 In-Service Vehicle Emission Standards in Poland, 1995 148 Table 4.9 In-Service Vehicle Emission Standards for Inspection and Maintenance Programs in Selected U.S. Jurisdictions, 1994 148 Table 4.10 U.S. IM240 Emission Standards 149 Table 4.11 Alternative Options for a Heavy-Duty Vehicle I/M Program for Lower Fraser Valley, British Columbia, Canada 150 Table 4.12 Estimated Emission Factors for U.S. Gasoline-Fueled Automobiles with Different Emission Control Technologies and Inspection and Maintenance Programs 151 Table 4.13 Estimated Emission Factors for U.S. Heavy-Duty Vehicles with Different Emission Control Technologies and Inspection and Maintenance Programs 152 Table 4.14 U.S. EPA's I/M Performance Standards and Estimated Emissions Reductions from Enhanced I/M Programs 153 Table 4.15 Effect of Engine Tune-Up on Emissions for European Vehicles 153 Table 4.16 Tampering and Misfueling Rates in the United States 154 Table 4.17 In-Use Emission Limits for Light-Duty Vehicles in Mexico 158 Table 4.18 Remote Sensing CO and HC Emissions Measurements for Selected Cities 163 Table 5.1 Incremental Costs of Controlling Gasoline Parameters 185 Table 5.2 Influence of Crude OilType on Diesel Fuel Characteristics 187 Table 5.3 Influence of Diesel Fuel Properties on Exhaust Emissions 190 Table 5.4 Properties of Diesel Test Fuels Used in EPEFE Study 192 Table 5.5 Change in Light-Duty Diesel Vehicle Emissions with Variations in Diesel Fuel Properties 192 Table 5.6 Change in Heavy-Duty Diesel Vehicle Emissions with Variations in Diesel Fuel Properties 193 Table 5.7 Toxic Emissions from Gasoline and Alternative Fuels in Light-Duty Vehicles with Spark-Ignition Engines 194 xdi Air Podltion from Motor Vebhcles Table 5.8 Wholesale and Retail Prices of Conventional and Alternative Fuels in the United States, 1992 194 Table 5.9 Properties of Conventional and Alternative Fuels 195 Tables 5.10 Inspection and Maintenance (Air Care) Failure Rates for In-Use Gasoline, Propane, and Natural Gas Light-Duty Vehicles in British Columbia, Canada, April 1993 195 Table 5.11 Emissions Performance of Chrysler Natural Gas Vehicles 198 Table 5.12 Emissions from Diesel and Natural Gas Bus Engines in British Columbia, Canada 198 Table 5.13 Emissions from Diesel and Natural Gas Bus Engines in the Netherlands 198 Table 5.14 Comparison of Emissions and Fuel Consumption for Five Modern Dual-Fueled European Passenger Cars Operating on Gasoline and LPG 201 Table 5.15 Pollutant Emissions from Light- and Heavy-Duty LPG Vehicles in California 201 Table 5.16 Standards and Certification Emissions for Production of M85 Vehicles Compared withTheir Gasoline Counterparts 203 Table 5.17 Average Emissions from Gasohol and Ethanol Light-Duty Vehicles in Brazil 205 Table 5.18 Physical Properties of Biodiesel and Conventional Diesel Fuel 208 Table 5.19 Costs of Substitute Fuels 214 Table 5.20 Comparison of Truck Operating Costs Using Alternative Fuels 215 Table 5.21 Alternative Fuel Vehicles: Refueling Infrastructure Costs and Operational Characteristics 217 Table 5.22 Aggregate Life-Cycle Emissions for Gasoline-Fueled Cars with Respect to Fuel Production, Vehicle Producion, and In-Service Use 218 Table 5.23 Aggregate Life-Cycle Emissions from Cars for Conventional and Alternative Fuels 218 TableA5.1.1 Estimated World Use of Leaded Gasoline, 1993 224 TableA5.2.1 Characteristics of Electric Motors for EVApplications 229 TableA5.2.2 Goals of the U.S. Advanced Battery Coalition 231 TableA5.2.3 Specific EnergiesAchieved and Development Goals for Different BatteryTechnologies 232 TableA5.2.4 Relative Emissions from Battery-Electric and Hybrid-Electric Vehicles 234 TableA5.2.5 Examples of Electric Vehicles Available in 1993 234 TableA5.3.1 Emissions of Buses with Alternative Fuels, Santiago, Chile 238 TableA5.3.2 Economics ofAlternative Fuel Options for Urban Buses in Santiago, Chile 238 Preface Because of their versatility, flexibility, and low initial Air Pollution in the Developing World cost, motorized road vehicles overwhelmingly domi- nate the markets for passenger and freight transport Air pollution is an important public health problem in throughout the developing world. In all but the poorest most cities of the developing world. Pollution levels in developing countries, economic growth has triggered a megacities such as Bangkok, Cairo, Delhi and Mexico boom in the number and use of motor vehicles. Al- City exceed those in any city in the industrialized coun- though much more can and should be done to encour- tries. Epidemiological studies show that air pollution in age a balanced mix of transport modes-including developing countries accounts for tens of thousands of nonmotorized transport in small-scale applications and excess deaths and billions of dollars in medical costs rail in high-volume corridors-motorized road vehicles and lost productivity every year. These losses, and the will retain their overwhelming dominance of the trans- associated degradation in quality of life, impose a signif- port sector for the foreseeable future. icant burden on people in all sectors of society, but es- Owing to their rapidly increasing numbers and very pecially the poor. limited use of emission control technologies, motor ve- Common air pollutants in urban cities in developing hicles are emerging as the largest source of urban air countries include: pollution in the developing world. Other adverse im- pacts of motor vehicle use include accidents, noise, * Respirable particulate matter from smoky diesel ve- congestion, increased energy consumption and green- hicles, two-stroke motorcycles and 3-wheelers, house gas emissions. Without timely and effective mea- burning of waste and firewood, entrained road sures to mitigate the adverse impacts of motor vehicle dust, and stationary industrial sources. use, the living environment in the cities of the develop- * Lead aerosol from combustion of leaded gasoline. ing world will continue to deteriorate and become in- * Carbon monoxide from gasoline vehicles and burn- creasingly unbearable. ing of waste and firewood. This handbook presents a state-of-the-art review of * Photochemical smog (ozone) produced by the re- vehicle emission standards and testing procedures and action of volatile organic compounds and nitrogen attempts to synthesize worldwide experience with ve- oxides in the presence of sunlight; motor vehicle hicle emission control technologies and their applica- emissions are a major source of nitrogen oxides and tions in both industrialized and developing countries. It volatile organic compounds. is one in a series of publications on vehicle-related pol- * Sulfur oxides from combustion of sulfur-containing lution and control measures prepared by the World fuels and industrial processes. Bank in collaboration with the United Nations Environ- * Secondary particulate matter formed in the atmo- ment Programme to underpin the Bank's overall objec- sphere by reactions involving ozone, sulfur and ni- tive of promoting transport development that is trogen oxides and volatile organic compounds. environmentally sustainable and least damaging to hu- * Known or suspected carcinogens such as benzene, man health and welfare. 1,3 butadiene, aldehydes, and polynuclear aromatic xiii xiv Air Poution from Motor Vehicles hydrocarbons from motor vehicle exhaust and oth- * Transport demand management and market in- er sources. centives. Technical and economic measures to dis- courage the use of private cars and motorcycles In most cities gasoline vehicles are the main source and to encourage the use of public transport and of lead aerosol and carbon monoxide, while diesel vehi- non-motorized transport modes are essential for re- cles are a major source of respirable particulate matter. ducing traffic congestion and controlling urban In Asia and parts of Latin America and Africa two-stroke sprawl. Included in these measures are market in- motorcycles and 3-wheelers are also major contributors centives to promote the use of cleaner vehicle and to emissions of respirable particulate matter. Gasoline fuel technologies. As an essential complement to vehicles and their fuel supply system are the main transport demand management, public transport sources of volatile organic compound emissions in near- must be made faster, safer, more comfortable, and ly every city. Both gasoline and diesel vehicles contrib- more convenient. ute significantly to emissions of oxides of nitrogen. Gasoline and diesel vehicles are also among the main * Infrastructure andpublic transport improvements. sources of toxic air contaminants in most cities and are Appropriate design of roads, intersections, and traf- probably the most important source of public exposure fic control systems can eliminate bottlenecks, ac- to such contaminants. commodate public transport, and smooth traffic Studies in a number of cities (Bangkok, Cairo, Jakar- flow at moderate cost. New roads, carefully targeted ta, Santiago and Tehran, to name five) have assigned pri- to relieve bottlenecks and accommodate public ority to controlling lead and particulate matter transport, are essential, but should be supported concentrations, which present the greatest hazard to only as part of an integrated plan to reduce traffic human health. Where photochemical ozone is a prob- congestion, alleviate urban air pollution, and im- lem (as it is, for instance, in Mexico City, Santiago, and prove traffic safety. In parallel, land use planning, Sao Paulo), control of ozone precursors (nitrogen ox- well-functioning urban land markets, and appropri- ides and volatile organic compounds) is also important ate zoning policies are needed to encourage urban both because of the damaging effects of ozone itself and development that minimizes the need to travel, re- because of the secondary particulate matter formation duces urban sprawl, and allows for the provision of resulting from atmospheric reactions with ozone. Car- efficient public transport infrastructure and services. bon monoxide and toxic air contaminants have been as- signed lower priority for control at the present time, An integrated program, incorporating all of these el- but measures to reduce volatile organic compounds ex- ements, will generally be required to achieve an accept- haust emissions will generally reduce carbon monoxide able outcome with respect to urban air quality. Focus and toxic substances as well. on only one or a few of these elements could conceiv- ably make the situation worse. For example, building new roads, in the absence of measures to limit transport Mitigating the Impacts of Vehicular Air demand and improve traffic flow, will simply result in Pollution more roads full of traffic jams. Similarly, strengthening public transport will be ineffective without transport Stopping the growth in motor vehicle use is neither fea- demand management to discourage car and motorcycle sible nor desirable, given the economic and other ben- use and traffic engineering to give priority to public efits of increased mobility. The challenge, then, is to transport vehicles and non-motorized transport (bicy- manage the growth of motorized transport so as to max- cles and walking). imize its benefits while minimizing its adverse impacts on the environment and on society. Such a management strategy will generally require economic and technical Technical Measures to Limit Vehicular Air measures to limit environmental impacts, together with Pollution public and private investments in vehicles and transport infrastructure. The main components of an integrated This handbook focuses on technical measures for con- environmental strategy for the urban transport sector trolling and reducing emissions from motor vehicles. will generally include most or all of the following: Changes in engine technology can achieve very large re- ductions in pollutant emissions-often at modest cost. * Technical measures involving vehicles and fuels. Such changes are most effective and cost-effective when These measures, the subject of this handbook, can incorporated in new vehicles. The most common ap- dramatically reduce air pollution, noise, and other proach to incorporating such changes has been through adverse environmental impacts of road transport. the establishment of vehicle emission standards. Preface xv Chapter 1 surveys the vehicle emission standards that the low cost of repairs result in a very slow turnover of have been adopted in various countries, with emphasis the vehicle fleet, with large numbers of older polluting on the two principal international systems of standards, vehicles remaining in service for long periods of time. those of NorthAmerica and Europe. Chapter 2 discusses The role of fuels in reducing vehicle emissions is re- the test procedures used to quantify vehicle emissions, viewed in chapter 5, which discusses both the benefits both to verify compliance with standards and to esti- achievable through reformulation of conventional gaso- mate emissions in actual use.This chapter also includes line and diesel fuels and the potential benefits of alterna- a review of vehicle emission factors (grams of pollutant tive cleaner fuels such as natural gas, petroleum gas, per kilometer traveled) based on investigations carried alcohols, and methyVethyl esters derived from vegeta- out in developing and industrial countries. ble oils. Changes in fuel composition (for example, re- Chapter 3 describes the engine and aftertreatment moval of lead from gasoline and of sulfur from diesel) technologies that have been developed to enable new are necessary for some emission control technologies to vehicles to comply with emission standards, as well as be effective and can also help to reduce emissions from the costs and other impacts of these technologies. An existing vehicles. The potential reduction in pollutant important conclusion of this chapter is that major re- emissions from reformulated fuels ranges from 10 to 30 ductions in vehicle pollutant emissions are possible at percent. Fuel modifications take effect quickly and be- relatively low cost and, in many cases, with a net sav- gin to reduce pollutant emissions immediately; in addi- ings in life-cycle cost as a result of better fuel efficiency tion, they can be targeted geographically (to highly and reduced maintenance requirements. Although the polluted areas) or seasonally (during periods of elevated focus of debate in the industrial world is on advanced pollution levels). Fuel regulations are simple and easy to (and expensive) technologies to take emission control enforce because fuel refining and distribution systems levels from the present 90 to 95 percent control to 99 are highly centralized.The use of cleaner alternative fu- or 100 percent, technologies to achieve the first 50 to els such as natural gas, where they are economical, can 90 percent of emission reductions are more likely to be dramaticatly reduce pollutant emissions when com- of relevance to developing countries. bined with appropriate emission control technology. Hydrocarbon, carbon dioxide, and nitrogen oxide Hydrogen and electric power (in the form of batteries emissions from gasoline fueled cars can be reduced by and fuel cells ) could provide the cleanest power sourc- 50 percent or more from uncontrolled levels through es for running motor vehicles with ultra-low or zero engine modifications, at a cost of about U.S.$130 per emissions. Alternative fuel vehicles (including electric car. Further reductions to the 80 to 90 percent level are vehicles) comprise less than 2 percent of the global ve- possible with three-way catalysts and electronic engine hicle fleet, but they provide a practical solution to urban control systems at a cost of about U.S.$600 - $800 per pollution problems without imposing restrictrions on car. Excessive hydrocarbon and particulate emissions personal mobility. from two-stroke motorcycles and three-wheelers can be Technical emission control measures such as those de- lowered by 50 to 90 percent through engine modifica- scribed in this handbook do not, by themselves, consti- tions at a cost of U.S.$60 - $80 per vehicle. For diesel en- tute an emission control strategy, nor are they sufficient gines, nitrogen oxide and hydrocarbon emissions can to guarantee environmentally acceptable outcomes over be reduced by 30 to 60 percent and particulate matter the long run. Such measures can, however, reduce pollut- emissions by 70 to 80 percent at a cost less than ant emissions per vehicle-kilometer traveled by 90 per- U.S.$1,500 per heavy-duty engine.After-treatment sys- cent or more, compared with in-use uncontrolled tems can provide further reductions in diesel vehicle vehicles.Thus a substantial improvement in environmen- emissions although at somewhat higher cost. tal conditions is feasible, despite continuing increases in Measures to control emissions from in-use vehicles national vehicle fleets and their utilization. Although are an essential complement to emission standards for technical measures alone are insufficient to ensure the new vehicles and are the subject of chapter 4. desired reduction of urban air pollution, they are an in- Appropriately-designed and well-run in-use vehicle in- dispensable component of any cost-effective strategy for spection and maintenance programs, combined with limiting vehicle emissions. Employed as part of an inte- remote-sensing technology for roadside screening of grated transport and environmental program, these mea- tailpipe emissions, provide a highly cost-effective sures can buy the time necessary to bring about the means of reducing fleet-wide emissions. Retrofitting en- needed behavioral changes in transport demand and the gines and emission control devices may reduce emis- development of environmentally sustainable transport sions from some vehicles. Policies that accelerate the systems. retirement or relocation of uncontrolled or excessively polluting vehicles can also be of value in developing countries where the high cost of vehicle renewal and Acknowledgments This handbook is a product of an informal collabora- Emaad Burki (Louis Berger International, Washington, tion between the World Bank and the United Nations D.C., USA); David Cooper (University of Central Florida, Environment Programme, Industry and Environment Orlando); John Lemlin (International Petroleum Indus- (UNEP IE), initiated in 1990.The scope and contents of try Environmental Conservation Association, London); the handbook were discussed at a workshop on Auto- Setty Pendakur (University of British Columbia, Canada); motive Air Pollution-Issues and Options for Develop- Kumares Sinha (Purdue University, Indiana, USA); ing Countries, organized by UNEP IE in Paris in January Donald Stedman (University of Colorado, Denver), and 199I.The advice and guidance provided by the work- by Antonio Estache, Karl Heinz Mumme, Adhemar Byl, shop participants, who are listed on the next page, is and Gunnar Eskeland (World Bank) proved invaluable in gratefully acknowledged. the preparation of this work In addition, we made gen- It took nearly five years to bring this work to comple- erous use of the literature on this subject published by tion, and in the process the handbook was revised four the Oil Companies' European Organization for Environ- times to keep up with the fast-breaking developments mental and Health Protection (CONCAWE) and the Or- in this field. The final revision was completed in June ganization for Economic Cooperation and Development 1996. This process of updating was greatly helped by (OECD). the contributions of C. Cucchi (Association des Con- We owe very special thanks to Jose Carbajo,John Flo- structeurs Europeans d'Automobiles, Brussels); Juan Es- ra, and Anttie Talvitie at the World Bank, who kept faith cudero (University of Chile, Santiago); Barry Gore with us and believed that we had a useful contribution (London Buses Ltd., United Kingdom); P Gargava (Cen- to make.We gratefully acknowledge the support and en- tral Pollution Control Board, New Delhi, India); A.K. couragement received from Gobind Nankani to bring Gupta (Central Road Research Institute, New Delhi, In- this work to a satisfactory conclusion. Our two collabo- dia); Robert Joumard (Institute National de Recherche rators, Surhid P Gautam and Lit-Mian Chan spent end- sur les Transports et leur S&urite, Bron, France); Ricar- less hours keeping track of a vast array of background do Katz (University of Chile, Santiago); Clarisse Lula information, compiling the data presented in the book, (Resource Decision Consultants, San Francisco); A.PG. and preparing several appendices. Our debt to them is Menon (Public Works Department, Singapore); Laurie great. Michaelis (Organization for Economic Co-operation and We would like to acknowledge the support of Jeffrey Development/International Energy Agency, Paris); Peter Gutman, Anthony Pellegrini, Louis Pouliquen, Richard Moulton (Global Resources Institute, Kathmandu); Scurfield and Zmarak Shalizi at the World Bank, who Akram Piracha (Pakistan Refinery Limited, Karachi); Zis- kept afloat the funding for this work despite the delays sis Samaras (Aristotle University, Thessaloniki, Greece); and our repeated claims that the book required yet an A. Szwarc (Companhia de Tecnologia de Saneamento other revision. Jacqueline Aloisi de Larderel, Helenc Ambiental, Sao Paulo, Brazil); and Valerie Thomas (Prin- Genot, and Claude Lamure at UNEP IE organized and fi ceton University, New Jersey, USA). We are specially nanced the 1991 Paris workshop and encouragecl us to grateful to our many reviewers, particularly the three complete the work despite the delays.We would like t(' anonymous reviewers whose erudite and compelling record the personal interest that Ibrahim Al Assaf, until comments induced us to undertake a major updating recently the Executive Director for Saudi Arabia at thc and revision of the handbook. We hope that we have World Bank, took in the conduct of the work and the en- not disappointed them. Written reviews prepared by couragement he offered us. xvii xviii AirPollution from MotorVebicies Paul Holtz provided editorial assistance and advice. lines. Without their understanding, this would still Jonathan Miller, Bennet Akpa,Jennifer Sterling, Beatrice be an unfinished manuscript. Very special thanks to Sito, and Catherine Ann Kocak, were responsible for art- our wives, Surraya Faiz, Carolyn Weaver, and Evelyn work and production of the handbook. Walsh. In closing we are grateful for the patience and Asif Faiz support our families have shown us while we toiled Christopher S.Weaver to finish this book. Many weekends were consumed Michael P.Walsh by this work and numerous family outings were can- celed so that we could keep our self-imposed dead- November 1996 Participants at the UNEP Workshop The workshop onAutomotive Air Pollution - Issues and Options for Developing Countries, sponsored by the United Nations Environment Programme, Industry and Environment (UNEP IE), was held in Paris,January 30-31, 1991.The titles of the particpants reflect the positions held at the time of the workshop. Marcel Bidault Tamas Meretei Chief, Directorate of Studies and Research Professor, Institute of Transportation Renault Industrial Vehicles, France Sciences, Hungary David Britton Juan Escudero Ortuzar International Petroleum Industry Executive Secretary Environmental Conservation Association Special Commission for the IPIECA, United Kingdom Decontamination of the Santiago Metropolitan Region, Chile Asif Faiz HighwaysAdviser Peter Peterson Infrastructure and Urban Development Division Director, Monitoring Assessment and The World Bank, US.A. Research Centre (MARC) UNEP/GEMS, United Kingdom He1lne Genot Senior Consultant John Phelps UNEP IE, France Technical Manager, European Automobile Manufacturers Association, France Barry Gore Vehicle Engineer Claire van Ruymbeker London Buses Ltd., United Kingdom Staff Scientist, Administration forAir Quality, Mexico M. Hublin President, Expert Group on Emissions and Zissis C. Samaras Energy Associate Professor European Automobile Manufacturers Aristotle University, Thessaloniki, Greece Association, France Kumares C. Sinha Claude Lamure Professor of Transport Engineering, Director Purdue University, Indiana, US.A. National Institute forTransport and Safety Research (INRETS), Michael P. Walsh France International Consultant Arlington,Virginia, US.A. Jaqueline Aloisi de Larderel Director UNEP IE, France xix 1 Emission Standards and Regulations Motor vehicle emissions can be controlled most effec- adopted by Canada and Mexico. Other countries and ju- tively by designing vehicles to have low emissions from risdictions that have adopted U.S. standards, test proce- the beginning. Advanced emission controls can reduce dures or both include Brazil, Chile, Hong Kong, Taiwan hydrocarbon and carbon monoxide emissions by more (China), several Western European countries, the Re- than 95 percent and emissions of nitrogen oxides by 80 public of Korea (South Korea), and Singapore (for mo- percent or more compared with uncontrolled emission torcycles only). The generally less-stringent standards levels. Because these controls increase the cost and and test procedures established by the United Nations complexity of design, vehicle manufacturers require in- Economic Commission for Europe (ECE) are used in the ducements to introduce them.These inducements may European Union, in a number of former Eastern bloc involve mandatory standards, economic incentives, or a countries, and in some Asian countries. Japan has also combination of the two. Although mandatory standards established a set of emission standards and testing pro- have certain theoretical disadvantages compared with cedures that have been adopted by some other East economic incentives, most jurisdictions have chosen Asian countries as supplementary standards. them as the basis for their vehicle emissions control In setting limits on vehicle emissions, it is important to programs. Vehicle emission standards, now in effect in distinguish between technology-forcing and technology- all industrialized countries, have also been adopted in following emission standards. Technology-forcing stan- many developing countries, especially those where rap- dards are at a level that, though technologically feasible, id economic growth has led to increased vehicular traf- has not yet been demonstrated in practice. Manufactur- fic and air pollution, as in Brazil, Chile, Mexico, the ers must research, develop, and commercialize new tech- Republic of Korea, and Thailand. nologies to meet these standards. Technology-following Because compliance with stricter emission standards standards involve emission levels that can be met with usually involves higher initial costs, and sometimes demonstrated technology. The technical and financial higher operating costs, the optimal level of emission risks involved in meeting technology-following standards standards can vary among countries. Unfortunately, the are therefore much lower than those of technology-forc- data required to determine optimal levels are often un- ing standards. In the absence of effective market incen- available. Furthermore, economies of scale, the lead- tives to reduce pollution, vehicle manufacturers have time required and the cost to automakers of developing little incentive to pursue reductions in pollutant emis- unique emission control systems, and the cost to gov- sions on their own. For this reason, technology-forcing ernments of establishing and enforcing unique stan- emission standards have provided the impetus for nearly dards all argue for adopting one of the set of all the technological advances in the field. international emission standards and test procedures al- The United States has often set technology-forcing ready in wide use. standards, advancing emissions control technology The main international systems of vehicle emission worldwide. Europe, in contrast, has generally adopted standards and test procedures are those of North Amer- technology-following standards that require new emis- ica and Europe. North American emission standards and sion control technologies only after they have been test procedures were originally adopted by the United proven in the U.S. market. States, which was the first country to set emission stan- Incorporating emission control technologies and dards for vehicles. Under the NorthAmerican FreeTrade new-vehicle emission standards into vehicle production Agreement (NAFFA), these standards have also been is a necessary but not a sufficient condition for achieving 2 Ar Pollutionfrom Motor Veblc1es low emissions. Measures are also required to ensure the mandated cleaner fuels, and added cold temperature durability and reliability of emission controls throughout standards. The California Air Resources Board (CARB) the vehicle's lifetime. Low vehicle emissions at the time established even more stringent regulations under its of production do little good if low emissions are not Low-Emission Vehicle (LEV) program. maintained in service. To ensure that vehicle emission Efforts are now being made to attain global harmoni- control systems are durable and reliable, countries such zation of emission standards. Emissions legislation is be- as the United States have programs to test vehicles in ser- ing tightened in many member countries of the vice and recall those that do not meet emission stan- Organization for Economic Co-operation and Develop- dards.Vehicle emission warranty requirements have also ment (OECD). Harmonization of emission standards been adopted to protect consumers. among countries can reduce the costs of compliance by avoiding duplication of effort. Development of a new emission control configuration typically costs vehicle International Standards manufacturers tens of millions of dollars per vehicle Vehicle emission control efforts have a thirty-year histo- model, and takes from two to five years. By eliminating ry. Legislation on motor vehicle emissions first ad- the need to develop separate emission control configu- dressed visible smoke, then carbon monoxide, and rations for different countries, harmonization of emis- later on hydrocarbons and oxides of nitrogen. Reduc- sion standards can save billions of dollars in tion of lead in gasoline and sulfur in diesel fuel received development costs. Such harmonization would greatly increasing attention. In addition, limits on emissions of facilitate international exchange of experience with re- respirable particulate matter from diesel-fueled vehi- spect to standards development and enforcement activ- cles were gradually tightened. Carcinogens like ben- ities, particularly between industrialized and zene and formaldehyde are now coming under control. developing countries. For light-duty vehicles, crankcase hydrocarbon controls The independent standards development and en- were developed in the early 1960s, and exhaust carbon forcement activities of the California Air Resources monoxide and hydrocarbon standards were introduced Board require a staff of more than 100 engineers, scien- later in that decade. By the mid-1970s most industrial- tists, and skilled technicians, along with laboratory op- ized countries had implemented some form of vehicle erating costs in the millions of dollars per year. The total emission control program. state budget for Califormia's Mobile Source Program is Advanced technologies were introduced in new U.S. U.S.$65 million a year. This figure substantially exceeds and Japanese cars in the mid- to late 1970s.These tech- the entire environmental monitoring and regulatory nologies include catalytic converters and evaporative budget of most developing nations. emission controls. As these developments spread and Harmonization of emission standards in North Amer- the adverse effects of motor vehicle pollution were rec- ica was an important aspect of the NAFTA involving ognized, worldwide demand for emission control sys- Canada, Mexico, and the United States.The ECE and the tems increased. In the mid-1980s, Austria, the Federal EU have established common emission regulations for Republic of Germany and the Netherlands introduced much of Europe. The United Nations Industrial Devel- economic incentives to encourage use of low-pollution opment Organization (UNIDO) is supporting work to vehicles. Australia, Denmark, Finland, Norway, Sweden, harmonize emission regulations in southeast Asia. A and Switzerland adopted mandatory vehicle standards proposal submitted by the United States would expand and regulations. A number of rapidly industrializing the ECE's functions by creating an umbrella agreement countries such as Brazil, Chile, Hong Kong, Mexico, the under which any country could register its emission Republic of Korea, Singapore, and Taiwan (China) also standards, testing procedures, and other aspects of its adopted emission regulations. vehicle emission regulations as international standards. In 1990, the European Council of Environmental A mechanism would also work toward regulatory com- Ministers ruled that all new, light-duty vehicles sold in patibility and the eventual development of consensus the EU in 1993 meet emission standards equivalent to regulations. Agreement has already been reached on 1987 U.S. levels. They also proposed future reductions harmonized emission requirements for some engines to reflect technological progress. While Europe moved used in off-highway mobile equipment. toward U.S. standards, the United States, particularly California, moved to implement even more stringent US. Standards legislation. Also, in 1990, the U.S. Congress adopted California was the first U.S. state to develop motor vehi- amendments to the Clean AirAct that doubled the du- cle emission standards and, because of the severe air rability requirement for light-duty vehicle emission con- quality problems in Los Angeles, remains the only state trol systems, tightened emission standards further, with the authority to establish its own emission stan- Emission Standards and Regulations 3 dards. In the past several decades California has often Table 1.1 Progression of U.S. Exhaust Emission established vehicle emission requirements that were lat- Standards for Light-Duty Gasoline-Fueled Vehicles er adopted at the U.S. federal level. The national effort (grams per mile) to control motor vehicle pollution can be traced to the Carbon Nitrogen 1970 Clean Air Act, which required a 90 percent reduc- Modelyear monoxide Hydrocarbons oxides tion in emissions of carbon monoxide, hydrocarbons, Pre-1968 and nitrogen oxides from automobiles.The Act was ad- (uncontrolled) 90.0 15.0 6.2 justed in 1977 to delay and relax some standards, im- 1970 34.0 4.1 - pose similar requirements on trucks, and mandate 1972 28.0 3.0 - vehicle inspection and maintenance programs in areas 1973-74 28.0 3.0 3.1 with severe air pollution. Further amendments to the 1975-76 15.0 1.5 3.1 Act, passed in 1990, further tightened vehicle emission 1977 15.0 1.5 2.0 requirements. Because of the size of the U.S. auto market, vehicles 1980 7.0 0.41 2.0 meeting U.S. emission standards are available from most 1981 3.4 0.41 1.0 international manufacturers. For this reason, and be- 1994-96 (Tier 1) 3.4 (4.2) 0.25 (0.31) 0.4 (0.6) cause U.S. standards are generally considered the most 2004 (Tier 2)b 1.7 (1.7) 0.125a (0.125) 0.2 (0.2) innovative, many other countries have adopted U.S. - Not applicable standards. Note: Standards are applicable over the 'useful life" of the vehicle, which is defined as 50,000 miles or five years for automobiles. The dura- bility of the emissions control device must be demonstrated over this dis- Light-duty vehicles. The U.S. emission standards for pas- tance within allowed deterioration factors. Figures in parenthesis apply senger cars and light trucks that took effect in 1981 were to a useful life of 100,000 mile, or ten years beyond the first 50,000 miles. later adopted by several countries including Austria, a. Non-methane hydrocarbons. Brazil, Canada, Chile, Finland, Mexico, Sweden, and b. The U.S. Environmental ProtectionAgency (EPA) could delay im- Switzerland. Compliance with these standards usually plementation of tier 2 standards until 2006. required a three-way catalytic converter with closed- Source. CONCAWE 1994 loop control of the air-fuel ratio, and it provided the im- emissions of nitrogen oxides to be 60 percent less than petus for major advances in automotive technology the U.S. federal standards applied in 1993. Useful-life re- worldwide. The 1990 Clean Air Act amendments man- quirements are extended from 80,000 to 160,000 kilo- dated even stricter standards for light-duty and heavy- meters to further reduce in-service emissions. duty vehicles, and also brought emissions from nonroad Requirements for low-temperature testing of carbon vehicles and mobile equipment under regulatory con- monoxide emissions and for on-board diagnosis of emis- trol for the first time. sion control malfunctions should also help reduce in- The evolution of U.S. exhaust emission standards for service emissions. light-duty, gasoline-fueled vehicles is traced in table In response to the severe air pollution problems in 1.1. In addition to exhaust emission standards, U.S. reg- Los Angeles and other California cities, CARB in 1989 es- ulations address many other emission-related issues, in- tablished stringent, technology-forcing vehicle emis- cluding control of evaporative emissions, fuel vapor sion standards to be phased in between 1994 and 2003. emissions from vehicle refueling, emissions durability These rules defined a set of categories for low-emission requirements, emissions warranty, in-use surveillance vehicles, including transitional low-emission vehicles of emissions performance, and recall of vehicles found (TLEV), low-emission vehicles (LEV), ultra low- not to be in compliance. Regulations that require on- emission vehicles (ULEV), and zero-emission vehicles board diagnostic systems that detect and identify mal- (ZEV). These last two categories are considered as fa- functioning emission systems or equipment are also voring natural gas and electric vehicles, respectively. being implemented. Table 1.2 summarizes the emission limits for passenger The 1990 Clean Air Act amendments mandated im- cars and light-duty vehicles corresponding to these low- plementation of federal emission standards identical to emission categories. 1993 California standards for light-duty vehicles.These In addition to being far more stringent than any pre- Tier I emission standards (to be phased in between vious emission standards, the new California standards 1994 and 1996) require light-duty vehicle emissions of are distinguished by having been designed specifically volatile organic compounds to be 30 percent less and to accommodate alternative fuels. Instead of hydrocar- 1. As U.S. standards are used by many other countries and are con- bons, the new standards specify limits for organic emis- sidered a benchmark for national standards around the world, they are sions in the form of non-methane organic gas (NMOG) treated as de-facto international standards. which is defined as the sum of non-methane hydrocar- 4 Air Pollutionfrom Motor Vebicles Table 1.2 U.S. Exhaust Emission Standards for Passenger Cars and Light-Duty Vehicles Weighing Less than 3,750 Pounds Test Weight (grams per mile) 50,000 miles or five years 100,000 miles or ten years Carbon Carbon Year monoxide Nitrogen monoxide Nitrogen Standard implemented 75"/20'F Hydrocarbons oxides 75°F Hydrocarbons oxides Passenger car' (rier 0) 1981 3.4/- 0.41 1.0 - Light-duty trucka (Tier 0) 1981 10/- 0.80 1.7 - Tier lb 1994-6 3.4/10.0 0.25 NMHC 0.4 4.2 0.31 NMHC 0.6 Tier 2 2004 1.7/3.4 0.125 NMHC 0.2 - - - California Low-EmissionVchicle/Federal Clean-fuel Fleet programs Transitional low-emission vehicle (TLEV) 1994c 3.4/10 0.125 NMOG 0.4 4.2 0.156 NMOG 0.6 Low-emission vehicle (LEV) 1997c 3.4/10 0.075 NMOG 0.2 4.2 0.090 NMOG 0.3 Ultra low-emission vehicle (ULEV) l997c 1.7/10 0.040 N.MOG 0.2 2.1 0.055 NMOG 0.3 Zer-emission vehicle (ZEV) I998c 0 0 0 0 0 0 - Not applicable NIMHC = non-methane hydrocarbons NMOG = non-methanc organic gases Note: The federal Tier I standards also specify a particulate matter limit of 0.08 gram per mile at 50,000 miles and 0.10 gram per mile at 100,000 miles. The California standards also specify a maximum of 0.01 5 gram per mile for formaldehyde emissions for 1993 standard, transitional low- emission, and low-emission vehicles, and 0.008 grams per mile for ultra low-emission vehicles. Likewise, for benzene, a limit of 0.002 gram per mile is specified for low-emission and ultra low-emission vehicles. For diesel vehicles, a particulate matter limit of 0.08 gram per mile is specified for 1993 standard, transitional low-emission, and low-emission vehicles, and 0.04 gram per mile for ultra low-emission vehicles at 100,000 miles. a. Except for California. b. Equivalent to California 1993 model year standard. c. To be phased in over a ten-year period; expected year of phase-in. Source: CONCAWE 1994, Chan and Weaver 1994 bons, aldehydes, and alcohol emissions, and thus ac- try offer would not include California's more-restrictive counts for the ozone-forming properties of aldehydes ULEV and ZEV standard, which are required under Mas- and alcohols tests that are not measured by standard hy- sachusetts and New York law. drocarbon tests.The new standards also provide for the non-methane organic gas limit to be adjusted with reac- Motorcycles. Current U.S. and California emission stan- tivity adjustment factors.These factors account for the dards for motorcycles are summarized in table 1.3. Un- differences in ozone-forming reactivity of the NMOG like other vehicles, motorcycles used in the U.S. can emissions produced by alternative fuels, compared meet these emission standards without a catalytic con- with those produced by conventional gasoline. This verter. The most important effect of the U.S. federal provision gives an advantage to clean fuels such as nat- emission standards has been the elimination of two- ural gas, methanol, and liquified petroleum gas, which stroke motorcycles, which emit large volumes of hydro- produce less reactive organic emnissions. carbons and particulate matter. California standards, The 1990 CleanAirAct amendments also clarified the though more stringent than the federal ones, can still be rights of other states to adopt and enforce the more met without a catalytic converter. Motorcycle standards stringent California vehicle emission standards in place in the United States are lenient compared with stan- of federal standards. New York and Massachusetts have dards for other vehicles because the number of motor- done so. In addition, the other states comprising the cycles in use is small, and their emissions are 'Ozone Transport Region" along the northeastern sea- considered insignificant compared with other mobile board of the United States (from Maine to Virginia) have emission sources. agreed to pursue the adoption of the California stan- dards in unison.This has prompted the auto industry to Medium-duty vebicles. In 1989, CARB adopted regula- develop a counter-offer, which is to implement Califor- tions that redefined vehicles with gross vehicle weight nia's LEV standard throughout the U.S.The auto indus- ratings between 6,000 and 14,000 pounds as medium- Emission Standards and Regulations 5 Table 1.3 U.S. Federal and California Motorcycle Exhaust Emission Standards (grams per kilometer) Engine type/size Standard (cubic centimeters) Carbon monoxide Hydrocarbons U.S. Federal 1978 50-170 17.0 5.0 170-750 17.0 5+0.0155 (D.170)a More than 750 17.0 14.0 1980 to present All models 12.0 5.0 California 1978-79 50-169 17.0 5.0 170-750 17.0 5+0.0155 (Dl170)a More than 750 17.0 14.0 1980-81 All models 12.0 5.0 1982-February 1985 50-279 12.0 1.0 More than 280 12.0 2.5 March 1985-1987 50-279 12.0 1.0 More than 280 12.0 1.4 1988 to present 50-279 12.0 1.0 280-699 12.0 1.0 More than 700 12.0 1.4 a. D is the engine displacement in Lubic centimecers. Source: Chan and Weaver 1994 duty vehicles. Previously, vehicles under 8,500 pounds ulations for heavy-duty vehicle engines are summarized gross vehicle weight were defined by both the CARB in table 1.5. The 1991 and 1994 emission standards and by the U.S. Environmental Protection Agency (EPA) were established by regulations adopted in 1985. En- as light duty, while those weighing more than 8,500 gines meeting the 1994 standards are now being sold. pounds gross vehicle weight were defined as heavy The 1990 Clean Air Act amendments established still duty and subject to emission standards based on an en- more stringent particulate levels for urban buses, and a gine dynamometer test.The U.S. EPA still classifies vehi- new standard of 4.0 g/bhp-hr for nitrogen oxides will des according to the old system, though vehicles take effect in 1998.The U.S. EPA has also adopted low- weighing between 6,000 and 8,500 pounds are subject emission vehicle and ultra low emission vehicle stan- to somewhat less stringent standards (table 1.4). dards for heavy-duty vehicles covered under the Clean- CARB recognized that large pickup trucks, vans, and Fuel Fleet program. In July 1995 the U.S. EPA and the En- chassis have more in common with light-duty trucks gine ManufacturersAssociation agreed that the limits on than with true heavy-duty vehicles. Light-duty trucks nitrogen oxides and hydrocarbons equivalent to the ul- are subject to more rigorous emission control require- tra low emission vehicle standard would become man- ments than larger vehicles. Medium-duty gasoline- and datory for all engines in 2004. At present, the only alternative-fueled vehicles are tested using the same heavy-duty engines capable of meeting these standards procedure as light-duty vehicles, but with heavier simu- use methanol or natural gas as fuel. Engine manufactur- lated weight settings. Medium-duty vehicles that have ers expect to be able to meet the standards using diesel diesel engines or that are sold as incomplete chassis engines with exhaust gas recirculation by 2004. have the option of certifying under the heavy-duty en- gine testing procedures instead. CARB has also estab- Evaporative emissions. Evaporative emission limits ap- lished LEV and ULEV emission standards for these ply to vehicles fueled by gasoline or alcohol fuels. Both engines. Presently, the only engines capable of meeting CARB and the EPA limit evaporative hydrocarbon emis- the ultra low emission vehicle standards use natural gas sions from light-duty vehicles to 2.0 grams per test, or methanol as fuel. which is considered effectively equivalent to zero (a small allowance is needed for other, non-fuel related or- Heavy-duty vehicles Limits on pollutants from heavy- ganic emissions from new cars, such as residual paint duty engines were adopted by the United States in solvent). California also applies this limit to motorcy- 1970. The current transient test procedure was intro- cles, but the U.S. EPA does not regulate motorcycle duced in 1983. Current U.S. and California emission reg- evaporative emissions. New, more stringent evaporative 6 Air Pollutionfrom Motor Vebicles Table 1.4 U.S. Federal and California Exhaust Emission Standards for Medium-Duty Vehicles (grams per mile) 50,000 miles orfive years 120,(000 miles or eleven years Year Carbon Nitrogen Carbon Nitrogen Standard (FTP-75) implemented monoxide Hydrocarbons oxides monoxide Hydrocarbons oxides U.S. federal 1983 10.0 0.80 1.7 - -- California/U.S.Tier I 1995/1996a 0-3,750 pounds 3.4 0.25 NMHC 0.4 5.0 0.36 NMHC 0.55 3,751-5,750 pounds 4.4 0.32 NMHC 0.7 6.4 0.46 NMHC 0.98 5,751-8,500pounds 5.0 0.39NMHC 1.1 7.3 0.56 NMHC 1.53 8,500-10,000poundsb 5.5 0.46 NMHC 1.3 8.1 0.66NMHC 1.81 10-14,000 poundsb 7.0 0.60 NMHC 2.0 10.3 0.86 NMHC 2.77 California Low-Emission Vehicle/Federal Clean-Fuel Fleet programs Low-emission vehicle (LEV) l998a 0-3,750 pounds 3.4 0.125 NMOG 0.4 5.0 0.180 NMOG 0.6 3,751-5,750 pounds 4.4 0.160 NMOG 0.7 6.4 0.230 NMOG 1.0 5,751-8,500 pounds 5.0 0.195 NIMOG 1.1 7.3 0.280 NMOG 1.5 8,501-I0,000 poundsb 5.5 0.230 NMOG 1.3 8.1 0.330 NMOG 1.8 10-14,000 poundSb 7.0 0.300 NMOG 2.0 10.3 0.430 NMOG 2.8 Ultra low-emission vehicle (ULEV) 1998, 0-3,750 pounds 1.7 0.075 NIMOG 0.2 2.5 0.107 NMOG 0.3 3,751-5,750 pounds 2.2 0.100 NMOG 0.4 3.2 0.143 NMOG 0.5 5,751-8,500 pounds 2.5 0.117NMOG 0.6 3.7 0.167NMOG 0.8 8,501-10,000poundsb 2.8 0.138NMOG 0.7 4.1 0.197NMOG 0.9 10-14,000poundsb 3.5 0.180NMOG 1.0 5.2 0.257NMOG 1.4 - Not applicable Note: NMHC-Non-methane hydrocarbons, NMOG-Non-methane organic gas. Emission standards for medium-duty vehicles also include limits for particulate matter and aldehyde emissions. a. Expected year of phase-in. b. California non-diesel vehicles only. All U.S. and California diesel-fueled vehicles weighing more than 8,500 pounds are subject to heavy-duty testing procedures and standards. Source: CONCAWE 1994 test procedures scheduled to take effect during the mid- transportation control measures. As of mid-1995, just 1990s will have the same limit of 2.0 grams per test, but two vehicle models were certified as inherently low- running-loss emissions will be limited to 0.05 grams per emission vehicles, and both were fueled by compressed mile.The new standard is nominally the same as the old natural gas (CNG). one, but more severe test conditions under the new test procedures will impose much greater compliance re- UN Economic Commission for Europe (ECE) quirements on manufacturers. and European Union (EU) Standards Evaporative and refueling emissions have become a The vehicle emission standards established by the ECE more significant fraction of total emissions as a conse- and incorporated into the legislation of the EU (former- quence of the steady decline in exhaust hydrocarbon ly the European Community) are not directly compara- emissions. To address this problem, and to encourage ble to those in the United States because of differences the introduction of vehicles using cleaner fuels, the U.S. in the testing procedure.2 The relative emissions mea- EPA has defined a special category of vehicles called in- sured using the two procedures vary with the vehicle's herently low-emission vehicles (ILEVs). These vehicles must meet the ultra low-emission standard for emissions 2. Besides the member states of the EU, China, the Czech Republic, of nitrogen oxides and the low-emission vehicle stan- Hong Kong, Hungary, India, Israel, Poland, Romania, Saudi Arabia, Sin- dards for carbon monoxide and non-methane organic gapore, Thailand, the Slovak Republic, and countries in the former gas. They must also exhibit inherently low evaporative U.S.S.R. and the former Yugoslavia require compliance with ECE reg- emissions by passing an evaporative test with the evap- ulations. Austria, Denmark, Finland, Norway, Sweden, and Switzer- orative control system disabled. Gasoline-fueled vehicles land have adopted U.S. standards. Following their admission into the European Union in 1995,Austria, Finland, and Sweden must comply cannot meet this standard. Inherently low-emission vehi- with EU regulations; a four-year transitional period has been agreed af- cles are eligible for certain regulatory benefits, including ter which national emission standards must either be harmonized exemption from 'no-drive" days and other time-based with EU regulations or renegotiated. Emission Standards and Regulations 7 Table 1.5 U.S. Federal and California Emission Standards for Heavy-Duty and Medium-Duty Engines Exbaust emissions (g/bbp-br) Total Hydrocarbons Nitrogen Carbon Particulate Smoke bydrocarbons (non-metbane) oxides monoxide matter Formaldebyde (pacftya U.S. Federal heavy-duty regulation 1991 HDV diesel 1.3 - 5.0 15.5 0.25 - 20/15/50 1991 LHDV gasoline 1.1 - 5.( 14.4 - - b 1991 MHDV gasoline 1.9 1.2 5.0 37.1 - b 1994 HDV diesel 1.3 0.9 5.0 15.5 (.1( - 20/15/50 c b 1994 LHDV otto 1.1 1.7 5.0 14.4 - - - c b 1994 MHDV otto 1.9 1.2 5.0 37 1 - - - b 1994 transit bus 1.3 1.2 5.( 15.5 0.07 - 20/15/50 b 1996 transit bus 1.3 1.2 5.0 15.5 0.05 - 20/15/50 b 1998 HDV diesel 1.3 1.2 4.0 15.5 0.10 - 20/15/50 b 1998 transit bus 1.3 1.2 4.0 15.5 0.05 - 20/15/50 d 2004(proposed)HDVOpt.A 1.3 2.4 15.5 0.10 - 20/15/50 2004 (proposed) HDV Opt. B 1.3 2.5 de15.5 0.10 - 20/15/50 Federal clean fuel fleet standards regulation d LEV - Federal fuel NR 3.8 14.4 0.10 - 20/15/50 d LEV-California fuel NR 3.5 14.4 0.10 - 20/15/50 d ILEV NR 2.5 14.4 0.10 0.050 20/15/50 d ULEV NR 2.5 7.2 0.05 0.025 20/15/50 Califbmia heavy-duty regulation 1991 HDV diesel 1.3 1.2 5.0 15.5 0.25 0.10 20/15/50 1991 LHDVotto 1.1 0.9 5.0 14.4 - 0.10 - h f g 1991 MHDVotto 1.9 1.7 5.0 37.1 - (.10 - 1994HDVdiesel 1.3 1.2 5.0 15.5 0.10 0.10 20/15/50 1994urban bus 1.3 1.2 5.0 15.5 0.07 0.10 20/15/50 Optional bus std. 1994 1.3 1.2 0.5-3.5 15.5 0.07 0.1() 20/15/50 1996 urban bus 1.3 1.2 4.0 15.5 0.(5 0.05 2(0/15/50 Optional bus std. 1996 1.3 1.2 0.5-2.5 15.5 0.05 0.05 20/15/50 California medium-duty regulation d Tier I NR 3.9 14.4 0.10 - 20/15/50 d LEV 1992-2001 NR 3.5 14.4 0.10 0.05 20/15/50 d 2002-2003 NR 3.0 14.4 0.10 0.05 20/15/50 d ULEV 1992-2(03 NR 2.5 14.4 0.10 0.05 20/15/50 2004+Opt.A. NR 2.5 d,e 14.4 0.10 0.05 20/15150 2004+Opt.AB NR 2.4 14.4 0.10 0.(5 20/15/50 d SULEV NR 2.0 7.2 0.05 0.05 2(0/15/5(0 - Not applicable NR = Not regulated; HDV= Heavy-duty vehicle; LHDV= Light heavy-duty vehicle (<14,000 lb. GVW); MHDV= Medium heavy-duty vehicle (>14,00() lb.fiVW); ILEV= Inherently low-emission vehicle; LEV= Low-emission vehicle; ULEV= Ultra low-emission vehicle; SULEV= Super ultra low-emission vehicle. a Acceleration/lug/peak smoke opacity. b Non-methane hydrocarbon (NMHC) standard applies instead of total hydrocarbion (THC) for natural gas engines only. c Replaced by'medium dutyW vehicle classification beginning 1995. d These standards (NMHC+NO,) limit the sum of NMHC and NO, emissions. e NMHC limited to (1.5 g/bhp-hr. f Use of NIMHC instead of THC standard is optional for diesel, LPC., and natural gas engines. g Methanol-fueled engines only. From 1993-95, limited to 0.10 g/bhp-hr, subsequently to 0.05 g/bhp-hr. h Includes spark-ignition gasoline and alternative fuel engines, except those derived from heavy-duty diesels. iOptional standards. Engines certified to these standards may earn emission credits. iOptional standards for diesel and diesel-derived engines and engines sold in incomplete medium-duty vehicle chassis. Source. CONCAWE 1995 8 Air Pollution from Motor Vebicies emission control technology, but test results in grams EEC, this regulation was not implemented in national per kilometer are generally of the same order. legislation by any European country, in anticipation of Until the mid-1980s, motor vehicle emission regula- the adoption of the Consolidated Emissions Directive, tions in Europe were developed by the ECE for adop- 91/441/EEC.This latter directive was adopted by the tion and enforcement by individual member countries. Council of Ministers of the European Community in It had been a common practice for the EU to adopt stan- June 1991. Under the Consolidated Emission Directives, dards and regulations almost identical to those issued exhaust emission standards for passenger cars (includ- by the ECE. In terms of stringency (i.e. level of emission ing diesel cars) are certified on the basis of the new control technology required for compliance) the Euro- combined ECE-1 5 (urban) cycle and extra-urban driving pean standards have lagged considerably behind the cycle (EUDC). In contrast to previous directives, a com- U.S. standards. Much of this lag has been caused by the mon set of exhaust emission standards (including dura- complex, consensus-based approach to standard setting bility testing) were applied to all private passenger cars used by the ECE and by the difficulty of obtaining agree- (both gasoline and diesel-engined) irrespective of en- ment between so many individual countries, each with gine capacity. The standard also covers vehicle evapo- its own interests and concerns.With the recent shift to rative emissions. Limit values for passenger car decision procedures requiring less-than-unanimous emissions are shown in table 1.6. These limits became agreement within the European Union, it has been pos- effective July 1, 1992 for new models, and on December sible to adopt more stringent emission standards. The 31,1992 for all production. stringency of the most recent EU emission standards is In March 1994, the Council of Ministers of the Euro- now closer to that of the U.S. standards. For all practical pean Community adopted Directive 94/12/EC which purposes the ECE no longer promulgates standards that provides for more stringent emission limits for passen- have not been agreed first by the EU. ger cars from 1996 onwards (table 1.6). These standards Unlike the U.S. standards, the ECE emission standards again differentiate between gasoline and diesel vehicles, apply to vehicles only during type approval and when but require significant emission reductions from both the vehicle is produced (conformity of production). fuel types.These standards make separate provisions for Once the vehicle leaves the factory and enters service, direct-injection (DI) diesel engines to meet less-stringent the manufacturer has no liability for its continued com- standards for hydrocarbons plus oxides of nitrogen and pliance with emission limits. Surveillance testing, recall for particulate matter, until September 30,1999. campaigns, and other features of U.S. emissions regula- In contrast to previous directives, production vehi- tion are not incorporated in the European regulatory cles must comply with the type approval limits.There is structure. As a result, manufacturers of such vehicles also a durability requirement for vehicles fitted with pol- have little incentive to ensure that the emission control lution control devices. Implementation of these emis- systems are durable enough to provide good control sion standards by EU member States is mandatory and throughout the vehicle's lifetime. unlike previous directives, not left to the discretion of individual national governments. Directive 94/12/EC Light-duty vehicles. These vehicles were the first to be also required that new proposals must be prepared be- regulated, beginning in 1970, to conform to the original fore June 30,1996 to implement further reductions in ECE Regulation 15. The regulation was amended four exhaust emissions by June 1, 20003 (CONCAWE 1995). times for type approval (ECE 15-01, implemented in Limit values for emissions of gaseous pollutants from 1974, ECE 15-02 in 1977, ECE 15-03 in 1979, and ECE light-duty trucks and commercial vehicles were also 15-04 in 1984) and twice for conformity of production established in the Consolidated Emissions Directive, (1981 and 1986). Regulation ECE 15-04 was applied to but the actual limits were identical to the limits estab- both gasoline and diesel-fueled light-duty vehicles, lished in ECE 15.04, and did not include a limitation on whereas earlier regulations applied only to gasoline-fu- eled vehicles.The emission limits included in these reg- 3. In June 1996, the European Commission proposcd to adopt the ulations were based on the ECE 15 driving cycle (van following exhaust emission limits for cars to become effective in years 2000 and 2005 (Walsh 1996a; Plaskett 1996). Ruymbeke and others 1992). * Gasoline-fueled (g/km) 2000 2005 The ECE did not adopt emission standards requiring co 2.30 1.00 three-way catalytic converters until 1988 (ECE regula- HC 0.20 0.10 tion 83), and then only for vehicles with engine dis- NOx 0.15 0.08 placement of 2.0 liters or more. Less stringent standards * Diesel-fueled (g/km) co 0.64 0.50 were specified for smaller vehicles, in order to encour- HC+NOX 0.56 0.30 age the use of lean-burn engines. Although ECE 83 was NOx 0.50 0.25 also adopted as European Community Directive 88/76/ P.M 0.05 0.025 Emission Standards and Regulations 9 Table 1.6 European Union Emission Standards for Passenger Cars with up to 6 Seats (ECE15+EUDC test procedure, grams per kilometer) ab 91/441/EEC 94/12/EC Conformity of l)pe approval production Gasoline Diesel CO 2.72 3.16 2.2 1.0 HC + NO, 0.97 1.13 0.5 0.7 PM 0.14 0.18 - 0.08d Evap. Emissions (g/test) 2.0 2.0 2.0 - - Not applicable Note: Directive 94/1 /EC applies to both type approval and conformity of production. a. Effective dates: (DAlI light-duty vehicics except direct-ignition (D) diesels;new modelsJuly 1,1992,all models Dec.31,1992. (ii) DI diesels,July 1, 1994. b. Effective dates: (i) Gasoline and IDI diesels; new models Jan 1,1996, all models Jan 1 1997. (ii) Dl diesels Oct 1,1999. c. DI diesel limits until June 30,1994 were 1.36 g/km (HC+NO,) and 0.19 g/km (PM). d. Less stringent standards apply to DI diesel until Sept 30,1999:0.9 g6km (HC+NOx) and 0.10 g/km (PM). Source: CONCAWE 1995 particulate matter emissions. Ministerial Directive 93/ to require catalytic converters, at least on two-stroke 59/EEC did finally modify the emission limits for light engines. trucks and commercial vehicles. Table 1.7 shows the emission standards established for light trucks and com- Heavy-duty engines. European regulation of heavy-duty mercial vehicles by this directive. These standards be- vehicle engines has lagged behind U.S. standards for the came effective with the 1994 model year. For light same reasons as that for light-duty engines. ECE regula- trucks with reference mass less than 1,250 kg, the stan- tion 49.01, for gaseous emissions and ECE regulation dards are equivalent to those established for passenger 24.03 for black smoke emissions (table 1.9), in effect un- cars by the Consolidated Emissions Directive; heavier til July 1992, was comparable in stringency to U.S. regu- vehicles are allowed somewhat higher emissions. Light lations from the 1970s, and could be met with little or no truck standards comparable in strictness to the passen- effort by diesel-engine manufacturers. The Clean Lorry ger car standards of 93/59/EEC have not yet been pro- Directive (91/542/EEC), compulsory throughout the EU, posed. reduces particulate and gaseous emissions for heavy-duty The present European emissions standards for passen- vehicles in two stages.The first-stage standards (Euro 1), ger cars and light commercial vehicles are comparable to which took effect in July 1992, are comparable in strin- the U.S. standards adopted in the early 1980s. The emis- gency to 1988 U.S. standards, while the second-stage sion control technologies required to meet these standards (Euro 2) are comparable to 1991 U.S. levels (ta- standards are similar. The main emission control ble 1.10). An even more stringent third-stage standard is requirements for gasoline vehicles include three-way under discussion, as is a change from the current steady- catalytic converters with feedback control of the air-fuel state emissions testing procedure to a transient cycle sim- ratio through an exhaust gas oxygen sensor. The 1996 ilar to the one used in the United States (Baines 1994). European emissions standards for passenger cars are more stringent, following the example set by the U.S. Tier 1 standards. Country and Other Standards Motorcycles. Although the ECE has issued emission This section summarizes the vehicle emission standards standards for motorcycles (ECE regulation 40.01) and adopted by a number of individual countries, as of early mopeds (ECE Regulation 47), these regulations are only 1995. Because emission standards often change, readers now being adopted in the EU (table 1.8). In addition, who require precise information are advised to obtain Austria and Switzerland have established their own up-to-date information from the legal authorities of the technology-forcing emission standards for motorcycles country involved. The Oil Companies' European Orga- and mopeds.The moped standards are sufficiently strict nization for Environmental Protection and Health 10 Air Pollution from Motor Vehicles Table 1.7 European Union 1994 Exhaust Emission Standards for Light-Duty Commercial Vehicles (Ministerial Directive 93/59/EEC) (grams per kilometer) Exhaust emissions Carbon Hydrocarbons + Particulate Vebicle category Reference mass (kg)' monoxide nitrogen oxides mattert Light trueksc RM < 1,250 Type-approval 2.72 0.97 0.14 Conformity of production 3.16 1.13 0.18 1,250 < RM < 1700 Type-approval 5.17 1.4 0.19 Conformity of production 6.0 1.6 0.22 RM > 1,700 Type-approval 6.9 1.7 0.25 Conformity of production 8.0 2.0 0.29 a. Reference mass (RM) means the mass of the vehicle in running order less the uniform mass of a driver of 75 kg and increased by a uniform mass of 100 kg. b. Diesel vehicles only. c. Includes passenger vehicles with seating capacity greater than six persons or reference mass greater than 2,500 kg. Source: CONCAWE 1995 Table 1.8 ECE and Other European Exhaust Emission Standards for Motorcycles and Mopeds (grams per kilometer) Testing Regulation, engine type Carbon monoxide Hydrocarbons Nitr_gen oxides pr)cedure ECE 40 Two-stroke, less than 100 kilograms 16.0 10.0 - ECE cycle T-wo-stroke, more than 300 kilograms 40.0 15.0 - ECE cycle Four-stroke, less than O00 kilograms 25.0 7.0 - ECE cycle Four-stroke, more than 300 kilograms 50.0 10.0 - ECE cycle ECE 40.01 Two-stroke, less than 100 kilograms 12.8 8.0 - ECE cycle Two-stroke, more than 300 kilograms 32.0 12.0 - ECE cycle Four-stroke, less than 100 kilograms 17.5 4.2 - ECE cycle Four-stroke, more than 300 kilograms 35.0 6.0 - ECE cycle ECE 47 for mopeds Two-wheel 8.0 5.0 - ECE cycle Three-wheel 15.0 10.0 - ECE cycle Switzerland Two-stroke 8.0 3.0 0.10 ECE 40 Four-stroke 13.0 3.0 0.30 ECE 40 Moped 0.5 0.5 0.10 ECE 40 Austria Motorcycles (<50 cc and >40 km/h) Two stroke (before Oct. 1,1991) 13.0 6.5 2.0 ECE 40 Two stroke (from Oct. 1, 1991) 8.0 7.5 0.1 ECE 40 Four stroke (before Oct. 1,1991) 18.0 6.5 1.0 ECE 40 Four stroke (from Oct. 1,1991) 13.0 3.0 0.3 ECE 40 Motorcycles (<50 cc) Two stroke (before Oct. 1,1990) 12.0-32.0 8.0-12.0 1.0 ECE 40 Two stroke (from Oct. 1, 1990) 8.0 7.5 0.1 ECE 40 Four stroke (before Oct. 1,1990) 17.5-35.0 4.2-6.0 0.8 ECE 40 Four stroke (from Oct. 1, 1990) 13.0 3.0 0.3 ECE 40 Mopeds (<50 cc and <40 km/h) 1.2 1.0 0.2 ECE 40 From Oct. 1,1988 - Not applicable Note: This table does not show ECE4() and ECE40.1 limits for Reference Weight, R (motocycle weight+75 kg) of more than 1)0 kg and less than 30(1 kg. Furthermore only limits for type approval are shown in this table. See CONCAWE (1995) for additional information and applicable limits for conformity of production. Source: CONCAWE 1992; 1994; 1995 Emissi)n Standards and Regulations 11 Table 1.9 Smoke Limits Specified in ECE Regulation 24.03 and EU Directive 72/306/EEC (smoke emission limits under steady state conditions) Nominalflow (liters/second) Absorption coefficient (m') 42 2.26 100 1.495 200 1.065 Note: Intermediate values are also specified. Opacity under free acceleration should not exceed the approved level by more than 0.5 m-1 Although the free acceleration test was intended as a means of checking vehicles in service it has not proved entirely successful.A number of different methods have been proposed by various countries, but there is no generally accepted alternative method of in- service checking. Source: CONCAWE 1994 Table 1.10 European Exhaust Emission Standards for Heavy-Duty Vehicles for Type Approval (grams per kilowatt bour) Effective date Carbon Nitrogen Particulate Regulation New models Allproduction monoxide oxides Hydrocarbons matter ECE 49 (13-mode) 14.0 18.0 3.5 a ECE 49.01 (88/77/EEC) April 1988 October 1990 11.2 14.4 2.4 (13.2) (15.8) (2.6)b a Clean lorry directive (91/542/EEC) Stage 1 (Euro 1) July 1992 October 1993 4.5 8.0 1.1 0.36-0.61c (4.9) (9.0) (1.23) (0.40-0.68) Stage 2 (Euro 2) October 1995 October 1996 4.0 7.0 1.1 0.15-0.25c (4.0) (7.0) (1.10) (0.15-0.25) Stage 3 (Euro 3) 1999 (tentative) n.a. 2.5 5.0 0.7 less than 0.12 n.a. = Not available a. Smoke according to ECE Regulation 24.03, EU Directive 72/306/EEC. b. Figures in parentheses are emission limits for conformity of production. c. Depending on engine rating. Source: Havenith and others 1993; CONCAWE 1994 (CONCAWE) in Brussels, has also published a series of and for all used vehicles appear to be based on ECE regu- reports summarizing vehicle emission and fuel stan- lations. Another regulation required the retirement of dards worldwide. The most recent such report (CON- buses older than 10 years (about 3,500 buses) in 1995. CAWE 1994) is a comprehensive source of information Municipal emission standards in the Capital Federal are on motor vehicle emission regulations and fuel specifi- embodied in Ordinance No. 39,025 and appear to be cations worldwide. tighter than national emission limits. It is understood that Argentine standards conform closely to Brazilian stan- Argentina dards although implementation is delayed because of the Decree 875/94 of National Law 2254/92 issued in 1994 current limited availability of unleaded gasoline. establishes national emission standards for new and used motor vehicles (table 1.1 1).The Decree also assigns the Australia Secretaria de Recursos Naturales y Ambiente Humano as In Australia's federal system of government, the power the responsible agency for enforcing and updating these to introduce motor vehicle legislation, including emnis- standards.These emission limits were reinforced by the sion regulations, lies with state governments.This is the Joint Resolutions 96/94 and 58/94 issued by the Secretar- opposite of the situation in the United States. The Aus- ies of Transport and Industry in March 1994. Emission tralian Transport Advisory Council (ATAC) is composed limits are established with a different compliance sched- of federal and state transport ministers who meet twice ule for trucks and urban passenger transport vehicles. In a year. The Council can agree to adopt emission and addition, emission limits for particulate matter are being safety standards which, while not binding on the states, established for the years 1996 and 2000. The exhaust are usually adopted in state legislation. States have acted emission standards for new light-duty gasoline-fueled ve- unilaterally when agreement is not reached within the hides, new heavy-duty vehicles, diesel-fueled vehicles, Council, however. 12 Air Pollution from Motor Vehicles Table 1.11 Exhaust Emission Standards (Decree 875/94), Argentina Emission limits for new light-duty gasoline or diesel vehicles Carbon monoxide Hydrocarbons Nitrogen oxides Carbon mon(xide" Hydnocarbonsa Modelyear (g/km) (glkm) (glkm) in low gear (% v) in low gear (ppm) 1994 24.0 2.1 2.0 3.0 660 1995 12.0 1.2 1.4 2.5 400 1997 2.0 0.3 0.6 0.5 250 1999 2.0 0.3 0.6 0.5 250 Emission limits for new heaty-duty gasoline or diesel vehicles Carbon monoxide Hydrocarbons Nitrogen oxides Carbon monrxidea Hydrocarbonsa Model Year (g/kWb) (g/kWh) (glk W) In low gear (C v) in low gear (ppm) 1995 11.2 2.4 14.4 3.0 660 1997 11.2 2.4 14.4 2.5 400 a. For gasoline vehicles. Source: Boletin Oficial 1994 The Council is advised on vehicle emission matters standards equivalent to those adopted in the United by a hierarchy of committees: the Motor Transport States in 1981 are scheduled to take effect in 1997; Groups (comprising senior federal and state public ser- compliance with these standards usually requires a three- vants), the Advisory Committee on Vehicle Emissions way catalytic converter and electronic fuel injection with and Noise (ACVEN) comprising lower-level federal and feedback control of the air-fuel ratio. More-stringent lim- state public servants, and the ACVEN Emissions Sub- it values will be introduced by 2000 and will match the Committee, which includes public servants, representa- U.S. standards. Crankcase emissions have been prohibit- tives from the automotive and petroleum industries as ed since 1977; evaporative emissions are limited to 6 well as consumers.The Committee also provides advice grams per test. Brazilian regulations, like U. S. regula- to the Australian Environment Council, which has some tions, require an emissions warranty of 80,000 kilome- emissions responsibilities. ters for light-duty vehicles and 160,000 kilometers for Before 1986, passenger car emission standards were heavy-duty vehicles. Alternatively, emissions must be 10 based on the 1973-74 U.S. requirements (ADR27). percent below the set limits.The Brazilian fuel program Since January 1986, manufacturers are required to meet also promotes the use of ethanol, both in pure form and a standard equivalent to 1975 U.S. requirements as an additive for gasoline. Ethanol, although considered (ADR37). Current requirements for commercial gaso- a cleaner-burning fuel than gasoline, can result in exces- line-fueled vehicles are based on regulations from New sive emissions of aldehydes, especially acetaldehyde. For South Wales and Victoria (table 1.12). Proposals are be- this reason, the 1992 and 19'37 standards limit aldehyde ing considered to introduce stricter emission standards emissions as well as emissions of hydrocarbons, carbon for passenger cars (equivalent to 1980 U.S. standards) monoxide, and nitrogen oxides (table 1.13). beginning in January 1996. Smoke opacity limits (ADR Control of heavy-duty diesel emissions has lagged be- 30 and ADR 55) apply to diesel-fueled vehicles. hind that of light-duty vehicles. Limits on smoke emis- sions took effect in 1987 for buses and in 1989 for trucks. Brazil These limits follow European standards and test proce- The Brazilian emissions control program, PROCONVE, dures. The first limits on gaseous emissions from diesel was introduced by the national environmental board engines, also based on European practice, took effect in CONAMA in May 1986. The first emission standards for 1993. More-stringent standards, based on the Clean Lorry light-duty vehicles took effect in 1987, but these stan- legislation of the European Union, were recently adopt- dards were lenient enough to be met by engine modifica- ed. These provided for the first-stage standards in 80 per- tions alone. More stringent emission standards, com- cent of new buses in 1994 and 80 percent of all new parable to those adopted by the United States in 1975, heavy-duty vehicles by 1996. The second-stage limits took effect in 1992. Compliance with these standards (comparable in stringency to current U.S. standards for usually requires an open-loop catalytic converter, elec- heavy-duty engines) are to be phased in between 1998 tronic fuel injection, or both.The Brazilian Congress has and 2002 (table 1.14). A system of prototype and produc- also enacted new legislation (No. 8723) effective Octo- tion certification, based on the U.S. procedure has been ber 1, 1993, setting strict emission standards for passen- established. Certification takes about 180 days.AII manu- ger vehicles for the rest of the decade. Exhaust emission facturers must submit statements specifying emissions of Emission Standards and Regulations 13 Table 1.12 Exhaust Emission Standards for Motor Vehicles, Australia (grams per kilometer) Evaporative Carbon Nitrogen Particulate Test emissions Regulation Effective date monoxide Hydrocarbons oxides matter procedure (grams per test) Passenger cars ADR 27A/B/C July 1976 24.2 2.1 1.9 - FTP 75 2.0 (Canister) January 1982 22.0 1.91 1.73 - FTP 75 6.0 (SHED)b ADR 37a January 1986 9.3 0.93 1.93 - FFP 75 2.0 (SHED) (8.45) (0.85) (1.75) - (FTP 75) (1.9 SHED) Proposed standards January 1996 4.34 0.26 1.24 2.0 FTP 75 January 2000 2.11 0.26 0.63 2.0 FTP 75 Commercial vehicles (gasoline) NSW (Clean Air Act) and Victoria (statutory rules) 1992 9.3 0.93 1.93 - FTP 75 2.0 (SHED) - Not applicable Note: Figures in parentheses apply to certification vehicles. a. The higher figures apply to production vehicles,which must meet the limits from 150 kilometers to 80,000 kilometers or for five years,which- ever occurs first. b. SHED = Sealed Housing for Evaporative Determinations. Source: CONCAWE 1994 Table 1.13 Exhaust Emission Standards for Light-Duty Vehicles (FrP-75 Test Cycle), Brazil (grams per kilometer) Carbon Particulate Carbon monoxide Evaporative Year effective monoxide Hydrocarbons Nitrogen oxides Aldebydes MatteSt at idle (%v)b (gramspertest) 1988a 24 2.1 2.0 - - 3.0 1989b 24 2.1 2.0 - - 3.0 - 1990C 24 2.1 2.0 - - 3.0 6.0 1992d 24 2.1 2.0 - - 3.0 6.0 1992e 12 1.2 1.4 0.15 0.05 2.5 6.0 1994 12 1.2 1.4 0.15 0.05 2.5 6.0 1997 2 0.3 0.6 0.03 0.05 0.5 6.0 2000 Limits in line with U.S. standards - Not applicable Notes: Effective January 1,1988, no crankcase emissions permitted. Effective January 1,1990, evaporative emissions limited to 6 grams per test (SHED). Exemptions possible for manufacturers with production less than 2,000 vehicles per year. a. New cars only. b. For certain specified models. c. For all models, except those not derived from light-duty vehicles. d. For models not derived from light-duty vehicles. e. For models not covered in footnote d. f. For alcohol-fueled vehicles only. g. For diesel-fueled vehicles only. h. Idle CO for alcohol in gasohol-fueled vehicles. i. Evaporative emissions expressed as propane for gasohol-fueled and ethanol for alcohol-fueled vehicles. Source: CETESB 1994; CONCAWE 1994 all models. Manufacturers of light-duty trucks (over 2,000 1.15). Limits for heavy-duty trucks are expected to be kg GVW) have the option to choose either the LDV or tightened in line with U.S. standards between 1994 and HDV test procedures for certification. 1996. The Canadian federal government has announced plans to bring passenger car emission standards in line Canada with the limits established in the U.S. Clean AirAct, pur- New standards for cars, light-duty trucks, and heavy suant to the provisions of NAFTA. Standards of 0.25 duty trucks were introduced in 1987, bringing Canadi- grams per mile for hydrocarbons, 3.4 grams per mile for an standards in line with then current U.S. limits (table carbon monoxide, and 0.4 grams per mile for nitrogen 14 Air Polutionfrom Motor Vebices Table 1.14 Exhaust Emission Standards for Heavy-Duty Vehicles (ECE R49 Test Cycle), Brazil (grams per kilowatt bour) Percent Carbon Particulate Vebicle class Effective date vehicles monoxide Hydrocarbons Nitrogen oxides matter Smoke K All vehicles January 1, 1989 - - - _ - 2.5b January 1, 1996 20 11.2 2.45 14.4 80 4.9 1.2 9.0 0.7/0.4a January 1,2000 20 4.9 1.2 9.0 0.7/0.4 80 4.0 1.1 7.0 0.15 January 1,2002 100 4.0 1.1 7.0 0.15 All imports January 1,1994 100 4.9 1.2 9.0 0.7/0.4a 2.5b January 1,1996 100 4.9 1.2 9.0 0 7/0.42 January 1,1998 100 4.0 1.1 7.0 0.15 Urban buses January 10,1987 - - - - - 2 .5b January 3,1994 20 11.2 2.45 14.4 - 2. 5b 80 4.9 1.2 9.0 - January 1 1996 20 11.2 2.4 14.4 - January 1, 1998 80 4.9 1.2 9.0 0.7/0.42 20 4.9 1.2 9.0 0.7/0.4a January 1,2002 80 4.0 1.1 7.0 0.15 100 4.0 1.1 7.0 0.15 - Not applicablc Note: * k=soot (g/m3) 'x gas flow (I/sec), applies to all vehicles. a. Particulatc emissions (PAM) 0.7 g/kWh for engines up to 85 kWh; 0.4 gIkWh for engines above 85 kWh. Crankcase emissions must be nil, except for some turbocharged diescl cngines if there is a technical justification. b. Applies from this datc onwards. Source: CETESB 1994; CONCAWE 1994 Table 1.15 Exhaust Emission Standards for Light- and Heavy-Duty Vehicles, Canada Year Carbon Nitrogen Diesel Testing Vehicle type effective monoxide Hydrocarbons oxides particulates procedure Light-duty vehicles (grams per kilometer) Cars and light-dut trucks 1975-87 25.00 2.00 3.10 - FrP 75 Cars 1988 2.11 0.25 0.62 0.12 FTP 75 Trucks less than 1,700 kilograms 1988 6.20 0.20 0.75 0.16 FTP 75 Trucks more than 1,700 kilograms 1988 6.20 0.50 1.10 0.16 FIP 75 Heavy-duty vehicles (grarns per brake horsepower-hour) Less than 6,350 kilograms 1988 14.4 1.1 6.0 - U.S. transient Morc than 6,350 kilograms 1988 37.1 1.9 6.0 - U.S. transient Diesels 1988 15.5 1.3 6.0 0.6 U.S. transient 1994 15.5 1.3 5.0 0.1 U.S. transient - Not applicable Source: CONCAWE 1994 oxides will probably be required as of 1996. The Feder- for vehicle emissions and fuels. Based on the California alTransport Minister and representatives of the automo- model, the provincial standards applicable to the dense- tive industry have agreed that cars sold in Canada from ly populated southern regions of the province are the 1994 to 1996 will meet the same emissions standards as most stringent in Canada. those sold in the United States.The U.S. manufacturers have committed themselves to market 1991 and subse- Chile quent model heavy-duty engines meeting U.S. standards Chilean authorities have adopted regulations requiring in Canada in the absence of specific regulations. all new light-duty vehicles to meet 1983 U.S. emission Air quality legislation enacted by British Columbia in limits.These regulations have been applied to the Santi- 1994 gives the province the authority to set standards ago metropolitan area and surrounding regions since Emission Standards and Regulations 15 September 1992, and were extended nationally in Sep- Table 1.16 Exhaust Emission Limits for Gasoline- tember 1994. Regulations have also been adopted re- Powered Heavy-Duty Vehicles (1983), China quiring new heavy-duty trucks and buses to be Idk equipped with engines meeting U.S. or European emis- Carbon monoxide Hydrocarbons sion standards. U.S. 1991 or Euro 1 standards were re- Vehicles (% V) (ppm) quired for buses in Santiago beginning September 1993, New (% 2(pm and for all heavy-duty vehicles in September 1994. U.S. New 5.0 2500 1994 or Euro 2 standards will be required for Santiago In-use 6.0 3000 buses in 1996, and nationwide in 1998. An emissions Imports 4.5 1000 test facility for certification and enforcement purposes Source. CONCAWE 1994 is under development. been adopted for heavy-duty gasoline engines, and China consideration is being given to establishing mass emis- Regulation of motor vehicle emission in China has been sion limits for these engines in grams per kilowatt guided by legislation enacted by the Standing Commit- hour. Revised or new emission standards and testing tee of the National People's Congress in 1979 and 1987 procedures that came into force in 1994 are listed in and the State Council in 1991: the "Environmental table 1.18. Protection Law of the People's Republic of China" Chinese motor vehicle regulations require that all do- (1979), the "Law of the People's Republic of China on mestically produced vehicle models must be listed in the Prevention and Control of Air Pollution" (1987), and the "Index of Enterprises Producing Motor Vehicles and the "Detailed Rules and Regulations for the Law of the their Products" issued annually by China National Auto- People's Republic of China on the Prevention and Con- motive Industry Corporation (CNAIC) and the Ministry trol of Air Pollution" (1991). Based on these laws, the of Public Security. Before a vehicle model is listed in the National Environmental Protection Agency, which is Index, the vehicle should pass an approval test carried responsible for formulating emission standards and test- out by "The Type Approval Organization for New Motor ing procedures, has issued 11 motor vehicle emission Vehicle Products." Recently, the responsibility for issu- control standards and formulated a Management Proce- ing the index and type approvals has been transferred to dure and Technical Policy to control emissions. the Auto Industry Bureau of the Ministry of Machinery. Standards for light-duty vehicles, adopted in 1979, Included in the approval tests are idling emissions tests are equivalent to ECE regulation 15-03 and include for gasoline-fueled vehicles and free acceleration smoke testing procedures for type approval and conformity tests for diesel-fueled vehicles. In addition, a full load of production. The standard test procedure lasts 13 smoke test is required for diesel engines. In case of minutes and has four cycles with no intermission. imported motor vehicles the type approval tests are Each cycle covers 15 working phases (idling, accelera- i m tion, deceleration, steady speed, and so on). These conducted by authorized laboratories of the State standards were developed by the Changchun Automo- Administration for Import and Export Commodity tive Research Institute and submitted to the State Envi- Inspection. ronmental Protection Administration by the Chinese Automotive Industry Federation (CSEPA 1989).The En- Colombia vironmental ProtectionAdministration adopted perfor- New emission standards for gasoline- and diesel-fueled mance targets for motorcycles in 1985 (GB 5366-85), vehicles were established in 1996 by Resolution 5 of the, and exhaust emission standards for light-duty vehicles Ministry of the Environment and Transport.These stan- in 1989 (GB 11641-89). The light-duty vehicle stan- dards establish carbon monoxide and hydrocarbons dards apply to cars, passenger vans, and light-duty emission limits for two mean-sea-level ranges of 0 to freight vehicles (reference mass 3,500 kilograms or 1,500 meters, and 1,501 to 3,000 meters (table 1.19). less) operating at a minimum speed of 50 kilometers Additional emission standards have been adopted by an hour. the Ministry of the Environment for light, medium, and Exhaust emission standards for heavy-duty vehicles, heavy-duty gasoline- and diesel-fueled vehicles to come adopted in 1983 (Regulations No. GB 3842-83, 3843- into effect with model year 1997 (table 1.20). 83, and 3844-83) consist only of carbon monoxide and hydrocarbon limits determined at idle and apply both Eastern European Countries and to new and in-use vehicles (table 1.16). China is con- the Russian Federation sidering legislation for heavy-duty gasoline-engine ve- hicles based on two runs of the U.S. 9-mode cycle used Most eastern European and central Asian countries in- by U.S. EPA during 1970-1983. Proposed limits are giv- cluding Russia use some combinations of ECE and EU en in table 1.17. Idle emission standards have also regulations, as shown in table 1.21. 16 Alr Pollutionfrom Motor Vebicles Hong Kong also tightened effective April 1, 1995. All new passenger New cars sold in Hong Kong are required to meet either cars and taxis must comply with 1990 U.S. standards or U.S. or Japanese emission standards or the new consol- equivalent EU and Japanese standards. Similar require- idated European limits (91/441/EEC). Each of these reg- ments will apply to medium goods vehicles and light ulations requires the use of three-way catalytic buses. For goods vehicles and buses with a design converters with electronic control systems. All new weight of 3.5 tonnes or more either the 1990 U.S. stan- dards or the Euro 1 standards will apply. Emissions stan- cars were required to meet these standards as ofJanuary dards have no b stablis for moorycles. 1992. Light-duty diesel vehicle emission standards were Table 1.17 Proposed Exhaust Emission Limits for Gasoline-Powered Heavy-Duty Vehicles, China (grams per kil(natt bour) Carbon Hydrocarbons + Nitrogen oxides Year Vebicle monoxide (g/kWb) (g/kWlb) Up to 1997 Certified before 1992 80 32 Produced after 1992 50 20 Type approved after 1992 40 16 1997 Certified before 1992 50 20 Produced after 1992 34 13.6 Type approved after 1992 28 11 Source: CONCAWE 1994 Table 1.18 List of Revised or New Emission Standards and Testing Procedures, China (Effective 1994) Number Title GB 14761.1-93 Emission standard for exhaust pollutants from light-duty vehicles GB 14761.2-93 Emission standard for exhaust pollutants from gasoline engine of road vehicles GB 14761.3-93 Emission standard for fuel evaporative emissions from road vehicle with gasoline engine GB 14761.4-93 Emission standard for pollutants from crankcase of vehicle engines GB 14761.5-93 Emission standard for pollutants at idle speed from road vehicle with gasoline engine GB 14761.6-93 Emission standard for smoke at free acceleration from road vehicles with diesel engine GB 14761.7-93 Emission standard for smoke at full load from automotive diesel engines GB/T 14762-93 Measurement method for exhaust pollutants from gasoline engine of road vehicles GB/T 14763-93 Measurement method of fuel evaporative emissions from road vehicles with gasoline engine GB/T 3845-93 Measurement method for pollutants at idle speed from road vehicles with diesel engine GB/T 384693 Measurement method for smoke at free acceleration from road vehicles with diesel engine GB/T 3847-93 Measurement method for smoke at full load from automotive diesel engines GB 14621-93 Emission standard for exhaust emissions from motorcycles GB/T 14622-93 Measurement method for exhaust emissions from motorcycles under running mode GB/T 5466-93 Measurement method for exhaust emissions from motorcycles under idle speed Source: Walsh 1995 Table 1.19 Emission Limits for Gasoline-Fueled Vehicles for Idle and Low Speed Conditions, Colombia Carbon monoxide (°/) Hydrocarbons (ppm) Modelyear above msl: 0-1500 m 1501-3000 m above msl:0-1500 m 1501-3000 m 2001 or newer 1.0 1.0 200 200 1998-2000 2.5 2.5 300 300 1997-1996 3.0 3.5 400 450 1995-1991 4.5 5.5 750 900 1990-1975 5.5 6.5 900 1000 1974 or older 6.5 7.5 1000 1200 msl = mean sea lcvel Source: Onursal and Gautam 1996 Emission Standards and Regulations 17 Table 1.20 Exhaust Emission Standards for Gasoline and Diesel-Fueled Vehicles, Colombla Vebicle category Unit Carbon monoxide Hydrocarbons Nitrogen oxides Light-duty vehicles g/km 2.3 0.25 0.62 Medium-duty vehicles g/km 11.2 1.05 1.43 Heavy-duty vehicles g/bhp-hr 25.0 10a a. Sum of HC and NO, emissions. Source: Onursal and Gautam 1996 Table 1.21 Summary of Vehicle Emission Regulations, Eastern Europe Implementation Country Vehicle type date Regulation Comments Czech and Slovak Passenger cars Type approval 89/458/EEC Republics 01.10.92 All vehicles 01.10.93 Light-duty vehicles As above 83/351/EEC Heavy-duty vehicles As above 91/542/EEC Hungary Passenger cars July 1992 ECE R83 For imported vehicles Heavy-duty vehicles 1990 ECE R49 Steady-state CO 14, HC 3.5, and NOx 18g/kWh ECE R24 Full load smoke Ordinance 6/1990 Free acceleration smoke Poland Passenger cars July 1995 ECE R83.02; 93/59/EC Heavy-duty vehicles Oct 1993 ECE R49.02; 91/542/EC 1988 ECE R24.03 Motorcycles Nov. 1992 ECE R40.01 Mopeds Nov. 1992 ECE R47 The Russian Gasoline passenger cars 1986 OST 37.001 054-86 Similar to ECE R15.04 Federation (without catalytic converters) Gasoline passenger cars 1986 OST 37.001 054-86 Conforms to ECE R83 (with catalytic converters) Diesel engines-exhaust emis- 1981 OST37.001 234-81 CO 9.5,HC3.4,NO. 14.35 per sions bhp-hr (ECE R49 test mode) Diesel engines-black smoke 1984 GOST 17.2 01-84 Full load smoke; emission limits as emissions follows: Nominal flow Smoke limit (1/sec) (opacity %) <42 60 100 45 >200 34 Source: CONCAWE 1995 India procedures tailored to Indian driving conditions (table The union government enacted a revised Motor Vehicle 1.22). Evaporative emissions are not regulated. Confor- Act in 1990, making emission regulations a federal gov- mity of production tests have also been developed. In erinent responsibility. India has established limits on addition, deterioration factors and endurance tests have carbon monoxide emissions (at idle) for gasoline-fueled been prescribed. cars, motorcycles, and three-wheelers; diesel smoke From April 1, 1996, all two-stroke engines in two- emissions are limited to 75 Hartridge units at full load. and three-wheelers would be required to comply with New emiission standards for gasoline-fueled cars took ef- the tighter emission standards shown below: fect in 199 Emissions from dieselne-fueles came toodef- * Three-wheelers fect in 1991. Emissions from diesel vehicles came under CO: 6.75 g/km; HC+NO,: 5.41 g/km control in 1992 based on ECE R49 regulations. These * Two-wheelers limits are similar to the ECE 15-04 limits but with test CO: 4.50 g/km; HC+NO,: 3.50 g/km 18 AIr Pollutionfrom Motor Vehicles Table 1.22 Exhaust Emission Standards for Gasoline-Fueled Vehicles, India (grams per kilometer) Reference mass (kilograms) Carbon monoxide Hydrocarbon s Two- and three-wheel vehicles Lessthan 150 12 8 150-350 12+(18-(R-150)/200) 8+(4'(R- 150)/200) More than 350 30 12 Light-duty vehicles Less than 1,020 14.3 2.0 1,020-1,250 16.5 2.1 1,250-1,470 18.8 2.1 1,470-1,700 20.7 2.3 1,700-1,930 22.9 2.5 1,930-2,150 24.9 2.7 More than 2,150 27.1 2.9 R = Reference mass. Source: India 1989 Japan Because of the differences in test procedures, a direct Japan revised its emissions test procedures for light-duty comparison of Japanese emission standards with those vehicles in 1991. The new test procedure, resembling applied to the U.S. and Europe cannot be made. the new ECE emissions test cycle, consists of a series of Republic of Korea low- and moderate-speed accelerations and decelera- tions at 20 kilometers per hour to 40 kilometers per Passenger car and light-truck emission standards equal hour, as well as a higher-speed component reaching up to current U.S. standards have been in effect since to 70 kilometers per hour. It will apply to passenger cars 1987.These standards apply to passenger cars using ei- and light- and medium-duty trucks (up to 2.5 tons gross ther gasoline or hquified petroleum gas, with engine displacement greater than 0.8 liter and gross vehicle weight).The test procedure for heavy-duty engines has dipaentgaerhn .ltradgosvhce weight) Thee toies proedure forheavy-duty six den est hs weight less than 2.7 tons. Standards for heavy-duty gas- alsother b teenamodited frome previouse tesix-m testtolv oline and liquified petroleum gas engines, based on the another steady-state, engine dynamometer test involving U.S. heavy-duty transient test procedure, are similar to 13 operating modes (these modes are different from the those in effect for heavy-duty gasoline engines in the ECE 13-mode test).The units of measurement have also Uie ttsbfr 90 ev-uydee etpo beenchaned,fromgram pe tes andppmto gams United States before 1990. Heavy-duty diesel test pro- been changed, from grams per test and ppm to grams cedures and emission standards are similar to those of per kilometer and grams per kilowatt hour, making it Japan. The Korean government is also moving to dis- easier to compare Japanese standards with U.S. and ECE courage the use of diesel engines in medium-duty emission standards. In addition to these changes, emis- trucks in favor of gasoline or liquified petroleum gas en- sion limits on nitrogen oxides (already among the most gines with more effective emissions control. Legisla- stringent worldwide) are to be further tightened, and tion has also been introduced for two-stroke and four- limits on diesel particulate emissions have been intro- stroke motorcycles that would require the use of cata- duced. Smoke limits were reduced by 20 percent in lytic converters. Emission limits for two- and four- 1993 for light- and medium-duty diesel vehicles and stroke motorcycles are summarized in table 1.23; the more stringent smoke limits were expected for heavy- test procedure however, is not known (CONCAWE duty passenger vehicles. Detailed information on Japa- 1994; UNIDO 1990). nese emission standards is available in CONCAWE 1994. Table 1.23 Motorcycle Emission Standards, Republic of Korea Two-stroke Four-stroke Period Carbon monoxide (% v) Hydrocarbons (ppm) Carbon monoxide (% v) Hydrocarbons (ppm) 1/91 to 12/92 5.5 1100 5.5 450 1/93/12/95 4.6 1100 4.5 450 1/96 onwarda 3.6 450 3.6 400 a. Proposed. Source: CONCAWE 1994 Emission Standards and Regulations 19 Malaysia turers agreed on catalyst-forcing standards for gasoline- In Malaysia, vehicle emission regulations based on ECE fueled microbuses. New microbuses are required to In Mlaysa, ehice eissin reulalonsbasd onECE meet these standards. Similar standards will apply to all 15.04 were introduced in September 1992. A further re- medium-ds. Similestandards of June 1992 quirement that all new gasoline-fueled vehicles be new medium-duty vehicles in 1994. As of June 1992, new heavy-duty diesel vehicles were required to be equipped with catalytic converters has been temporar.- equipped with engines meeting 1991 U.S. emission stan- ly postponed. dards at sea level, with an additional test of smoke opac- Mexico ity conducted at Mexico City's altitude (2,000 meters). Work on evaluating the feasibility of meeting 1994 U.S. New passenger cars and light-duty trucks sold in Mexi- emission standards at Mexico City's altitude is planned. co have been subject to exhaust emission standards (generally based on U.S. standards) since 1975 (table SaudiArabia 1.24). Until the 1991 model year, however, these stan- Saudi Arabia has adopted emission standards equivalent dards were loose enough to be met without the use of to ECE R15.03.Annual inspections of vehicle emission a catalytic converter or other advanced emission con- control systems is required in Jeddah, Riyadh, and Dam- trol technology. The 1991 and 1992 model years were man. Evaporative emissions are limited to 6.0 grams per a transition during which standards were made suffi- test (SHED). ciently stringent to require catalytic converters but not the full range of emission control technology required Singapore in the United States. Despite the transition period, most Singapore introduced European (ECE R 15-04) emission Mexican automakers equipped their vehicles with U.S.- standards for passenger cars in 1986. Since July 1,1992, model emission controls instead of a less-sophisticated all new gasoline-fueled vehicles registered in Singapore system that would be used for only two years. New cars have been required to comply either with ECE 83 or and light trucks since model year 1993 have been re- current Japanese emission regulations. In July 1993, the quired to meet exhaust emission standards that are limits of the Consolidated Emissions Directive replaced equivalent to 1987 U.S. standards. Starting in 1995, all ECE 83. New diesel-fueled vehicles have been required cars, light commercial vehicles and light trucks were re- to meet the smoke limits of ECE R 24.03 since January quired to meet an evaporative emissions standard of 2.0 1991, and used diesel vehicles imported to Singapore grams per test as well. Mexico has not yet adopted other have been required to meet the same standard since Jan- elements of the U.S. regulations: emissions durability, uary 1992. New motorcycles have been required since emissions warranty requirements, and in-service testing October 1991 to comply with U.S. emission standards with recall of vehicle models found to be violating emis- stipulated in the U.S. Code of Federal Regulations (40 sion standards in service.As part of the NAFTA, the Unit- CFR 86.410-80) before they can be registered for use in ed States, Canada, and Mexico are in the process of Singapore. harmonizing vehicle emission standards. Since 1982, all in-use vehicles have been required to In the past few years, Mexican authorities have estab- undergo a periodic, compulsory mechanical inspec- lished emission standards for medium- and heavy-duty tion.This is to ensure that vehicles on public roads are vehicles (table 1.25). In 1991 authorities and manufac- maintained properly, are roadworthy, and do not pol- Table 1.24 Emission Standards for Light-Duty Vehicles, Mexico (grams per kilometer) Year Carbon monoxide Hydrocarbons Nitrogen oxides Pre-1975 (uncontrolled) 54.0 5.5 2.3 1975 29.2 2.5 2.3 1976 24.2 2.1 2.3 1977 24.2 2.6 2.3 1988 22.2 2.0 2.3 1990 18.0 1.8 2.0 1991 7.0 0.7 1.4 1993 2.11 0.25 0.62 Note: Evaporative hydrocarbon emissions are not regulated. Source: World Bank 1992 20 Air Pollutionfrom Motor Vehicles Table 1.25 Exhaust Emission Standards for Light-Duty Trucks and Medium-Duty Vehicles by Gross Vehicle Weight, Mexico (grams-per kilometer) Vebicle type Carbon monoxide Hydrocarbons Nitrogen oxides Light-duty trucks, less than 2,727 kilograms 1991-93 22.0 2.0 2.3 1994 8.75 0.63 1.44 Light-duty trucks, 2,728-3,000 kilograms 1991 35.0 3.0 3.5 1992-93 22.0 2.0 2.3 1994 8.75 0.63 1.44 Medium-duty vehicles, 3,000-3,857 kilograms 1992 28.0 2.8 2.8 1993 22.0 2.0 2.3 1994 8.75 0.63 1.44 Urban transport minibuses, 3,001-5,500 kilograms 1991 10.0 0.6 1.5 1992 3.0 0.3 1.0 Note: Evaporative emissions are not regulated. a. Standards applied only to highly polluted areas through 1992; minibuses outside critical areas were not regulated until 1993. Source. World Bank 1992 lute the environment. Exhaust emissions are checked Diesel engines have been required to comply with during these inspections. Enforcement inspection is smoke emission limits since 1984. Since July 1993, die- also conducted daily by spot checking vehicles on the sel engines have been required to meet emission limits road. Detailed information on Singapore emission stan- of 6.0 grams per bhp-hour for nitrogen oxides and 0.7 dards is given in CONCAWE 1994. grams per bhp-hour for particulate matter, based on the U.S. heavy-duty transient cycle. The nitrogen oxides Taiwan (China) standard is the same as the 1988 U.S. standard, while All new cars sold in Taiwan (China) have been required the standard for particulate matter is slightly more to meet ECE regulation 15.04 emission standards since lenient. July 1987. In July 1990, the regulation was tightened to require compliance with 1983 U.S. emission standards. Thailand New models and all imports were required to meet The rapid growth in the vehicle fleet has compelled the these standards immediately, and existing domestically Royal Thai Government to quickly establish emission produced models were allowed waivers of up to three standards. New gasoline-fueled vehicles have been re- years. Beginning July 1994, all car models sold in Tai- quired to be fitted with catalytic converters since 1993. wan (China) were required to meet U.S. standards. Thailand has adopted test cycles and emission standards Emission standards for new motorcycle engines have conforming to ECE/EEC regulations for light-duty gaso- also been established-1991 standards are some of the line and light- and heavy-duty diesel vehicles (table most stringent in the world (table 1.26). Compliance 1.27). Emission standards for motorcycles equivalent to with motorcycle standards has required significant en- ECE R40 were introduced inAugust 1993 and soon after gine modifications, including the use of air injection in revised to comply with ECE R40.01 regulations. Third- four-stroke engines and the installation of catalytic con- phase controls similar to the Taiwanese regulations are verters on two-stroke engines. Electric motorcycles being introduced over the period 1994-1997. have been available since May 1992 but with modest In addition to the standards themselves, procedures sales. for verifying compliance and for corrective action to Motorcycle durability requirements have been in force deal with non complying vehicles also need to be de- since November 1991. All new motorcycles are required veloped.The Thai Institute of Standards has established to demonstrate that they can meet the relevant standards laboratory facilities to measure emissions from light- for a minimum of 6,000 kms. It is expected that the dura- duty gasoline-and diesel-fueled vehicles and motocy- bility requirement will be increased to 20,000 kms from cles. Facilities for diesel engines are scheduled to be January 1, 1998 and that the share of electric powered completed soon after. They will be used only to verify motorcycles will be mandated at 5 percent. compliance by new vehicles-as in Europe, but unlike Emission Standards and Regulations 21 Table 1.26 Exhaust Emission Standards for Motorcycles, Taiwan (China) ECE 40 (driving cycle) ECE 40 (idle) Carbon Hydrocarbons + Carbon Acceleration monoxide nitrogen oxides monaxide smoke (grams per (grams per (percent by Hydrocarbons (opacity Year implemented Classification kilometer) kilometer) volume) (ppm) percentage) 1984 New 8.8 6.5 4.5 7,000 In-use - - 4.5 9,000 - 1988 Prototype 7.3 4.4 4.5 7,000 15 Production 8.8 5.5 4.5 7,000 15 In-use (inspection) - - 4.5 9,000 30 1991 Prototype 3.75 2.4 4.5 7,000 15 Production 4.50 3.0 4.5 7,000 15 In-use (inspection) - - 4.5 9,000 30 1997 (proposed) Prototype 075a 0.8a 4.5 7,000 15 - Not applicable a. From January 1994 onwards, the limits are CO (1.5 g/km), HC (1.2 g/km), and NO, (0.4 g/km). It is proposed to reduce the 1994 limits by half as of January 1997. Source: Chan and Weaver 1994; CONCAWE 1992 Table 1.27 Exhaust Emission Standards, Thailand Vehicle Effective date Cycle Equivalent limits Gasoline Jan. 1,1995 ECE 83 ECE 91/411/EEC Light diesel Jan. 1,1995 ECE 83 ECE 91/411/EEC Heavy diesel Jan. 1,1993 ECE 49 ECE 91/542/ (A)/EEC (Euro 1) Jan. 1,2000 ECE 49 ECE 91/542 (H)/EEC (Euro 2) Motorcycles Jan. 1,1994 ECE 40 see table 1.26 footnote (a) Jan. 1,1997 ECE 40 Source: CONCAWE 1994 in the United States, manufacturers will bear no re- should be regulated to perform this maintenance prop- sponsibility for continuing emissions compliance in erly. A comprehensive compliance program should in- consumer use. clude the items outlined in the following sections. Compliance with Standards Certification or Type Approval Most countries require some form of certification or type Stringent emission standards are of little value without approval by vehicle manufacturers to demonstrate that a program to ensure that vehicle sellers and buyers com- each new vehicle sold is capable of meeting applicable ply with them. A comprehensive compliance program emission standards. Usually, type approval requires emis- should cover both new and in-use vehicles; it should as- sion testing of prototype vehicles representative of sure that attention is given to emission standards at the planned production vehicles. Under ECE and Japanese vehicle design stage prior to mass production; it should regulations, such compliance is required only for new ve- ensure quality on the assembly line; and it should deter hicles (though new regulations require manufacturers to the manufacture of non-conforming vehicles through demonstrate in each case the durability of catalytic con- an enforceable warranty and recall system. Further- verters up to 80,000 kilometers). U.S. regulations require more, vehicle owners should be encouraged to carry that vehicles comply with emission standards through- out maintenance on emission control devices as re- out their useful lives when maintained according to the quired by the manufacturer, and the service industry manufacturers specifications. 22 AMrPollutionfrom MotorVebicles As part of the U.S. certification process, vehicle In-Use Surveillance and Recall manufacturers are required to operate a prototype ve- U.S. emission regulations require vehicles to continue hicle in an accelerated durability test driving program, to meet emission standards throughout their useful for distances ranging from 6,000 kilometers (for small lives, if maintained according to manufacturers' specifi- motorcycles) to 160,000 kilometers (for passenger cations. U.S. authorities have instituted extensive test- cars) or 192,000 kilometers (for light trucks). Heavy- ing programs to guard against increases in emissions duty engines are similarly tested, using an accelerated resulting from defective emission controls in customer durability procedure on an engine dynamometer. Emis- use. Several hundred vehicles per year are temporarily sion tests on the prototype vehicle or engine undergo- procured from consumers for testing. A questionnaire ing this accelerated testing are used to establish and physical examination of the vehicle are used to de- deterioration factors for each pollutant for a given en- termine whether the vehicle has been properly main- gine family.To determine compliance with the regula- tained.The vehicles are then adjusted to specifications tions, the emissions from low-mileage prototype and tested for emissions. Average emissions from all vehicles are multiplied by the established deterioration properly maintained vehicles of a given make and mod- factor; the result is required to be less than the applica- el tested must comply with the applicable emissions ble standard. Durability demonstrations, maintenance, standard within a range that allows for statistical uncer- and emissions testing are normally conducted by vehi- tainty. Otherwise, the manufacturer may be ordered to cle manufacturers, but national testing authorities will recall the vehicles for repairs or modifications to bring occasionally test certification vehicles on a spot check them within the standard. Hundreds of thousands of ve- basis. hides are recalled in this manner each year. The advantage of a certification program is that it can The expense and consumer dissatisfaction generated influence vehicle design prior to mass production. Ob- by emission recalls have induced manufacturers to de- viously, it is more cost-effective if manufacturers identi- velop far more durable and effective emission control fy and correct problems before production actually systems and to establish internal emnission targets for begins. As a practical matter, the certification process new vehicles that are much stricter than legal standards. deals with prototype cars (sometimes almost hand- Most manufacturers design for a margin of at least 30 made) in an artificial environment of careful mainte- percent (and preferably 50 percent) between the certi- nance, perfect driving conditions, and well-trained driv- fied emissions and the standard in order to provide a ers using ideal roads or dynamometers. As a result, reasonable allowance for in-use deterioration. Thus, to vehicles that fail to meet emission standards during cer- ensure in-use compliance with a standard of 0.2 grams tification will almost certainly fail to meet standards in per mile for nitrogen oxides, most manufacturers would use. The converse, however, is not true-it cannot be set a development target of 0.1 grams per mile. As a re- said with confidence that vehicles that pass certification sult, the emissions performance of vehicles in use has will inevitably perform well in use. improved significantly. Emissions surveillance programs provide extensive Assembly Line Testing data on vehicle emissions in the real world, as opposed to the artificial world of prototype development and The objectives of assembly line testing are to enable reg- certification. They thus provide important information ulatory authorities to identifv certified production vehi- to air quality planning authorities, vehicle regulators, ses that do not comply with applicable emission and vehicle manufacturers on the actual effectiveness standards, to take remedial actions (such as revoking of emission control programs and technologies.The im- certification and recalling vehicles) to correct the prob- portance of such real-world data cannot be overestimat- lem, and to discourage the manufacture of noncomply- ed. Surveillance programs in the United States have ing vehicles. been instrumental in identifying the causes of high ve- Assembly line testing provides an additional check hicle emissions in use, which in turn has led regulators on mass-produced vehicles to assure that the designs and manufacturers to take actions to correct these prob- found adequate in certification are satisfactorily translat- lems. In the absence of systematic measurements of ac- ed into production, and that quality control on the as- tual vehicle emissions in use, government and industry sembly line is sufficient to provide reasonable assurance can all too easily assume that no problems exist. that cars in use will meet standards. The main advan- Surveillance and recall programs have had their diffi- tage of assembly line testing over certification is that it culties. Surveillance programs are expensive and poten- measures emissions from real production vehicles. tially are subject to sampling bias, because citizens must However, assembly line testing provides no measure of be induced to lend their vehicles to the government for vehicle performance over time or mileage-a substan- testing. Surveillance testing of heavy-duty engines re- tial and inevitable shortcoming. quires that they be removed from the vehicle and tested Emission Standards and Regulations 23 on an engine dynamometer-an extremely expensive have been introduced in which manufacturers who are procedure. As a result, very little of this type of testing able to do better than the emission standards on one ve- has been done to date, and data on in-use emissions hicle or engine model can generate credits that can be from heavy-duty vehicles are correspondingly sparse. used to offset higher-than-standard emissions from an- Recalls are not fully effective either-in the United other model of vehicle or engine.The manufacturer can States, on average, only 55 percent of the vehicles re- choose whether to use the credits in the same year (av- called are actually brought in by their owners for repair. eraging), sell them to another manufacturer (trading), The lag time between identification of a nonconform- or save them against possible need in a subsequent year ing class and the manufacturer's recall notice can be (banking). well over a year. Mandatory recalls are possible only Another promising strategy for achieving maximum when a substantial number of vehicles in a class or cat- emission reductions at minimum cost is to establish dif- egory exceeds the standards limit; this may preclude re- ferentiated emission standards for heavily used vehicles calls of serious but less frequent failures. in highly-polluted areas. In the United States, for exam- ple, fleet vehicles in major urban areas will be covered Warranty by the "Clean Fuel Vehicle' program, a special program that requires vehicles certified to lower emission stan- Warranty programs are intended to provide effective re- course to consumers against manufacturers when indi- dards. Special strict emission standards have also been vidual vehicles fail to meet in-use standards, and to established for gasoline-fueled minibuses in highly pol- discourage the manufacture of such vehicles. Warran- luted areas in Mexico. In Chile, buses in Santiago were ties attempt to assure the remedy of defects in design or required to meet emission standards earlier than other workmanship that cause high emissions. vehicles. An extension of the differentiated emission standard On-Board Diagnostic Systems approach would be to offer economic incentives for ur- ban vehicles to adopt more stringent emission controls. Increasingly complicated vehicle engine and emissions By calibrating the size of the incentive to the expected control systems have made the diagnosis and repair of use of the vehicle, the most stringent and expensive malfunctioning systems more difficult.With present in- emission controls could be applied where they would spection and maintenance program designs, many emis- be most effective. sions-related malfunctions can go undetected in Some countries in Western Europe, notably Germany, modern vehicles. This is especially true for malfunc- have made effective use of tax incentives to encourage tions related to nitrogen oxides, because present in- buyers to choose vehicles certified to more stringent spection and maintenance programs do not test for emission standards than the minimum requirements. In these emissions.To improve the effectiveness of emis- the United States, consumers are being encouraged to sions control diagnosis, the United States has recently purchase low-emitting vehicles by making these vehi- adopted second-generation requirements for on-board cles exempt from transportation control measures, such diagnosis of emissions-related malfunctions. as mandatory no-drive days.This approach is also being used in Mexico City to encourage owners of commercial vehicles to switch to cleaner fuel systems, such as liqui- Alternatives to Emission Standardsfldptoemgsadnurla. fied petroleum gas and natural gas. Because they are mandatory and universal, traditional From a theoretical standpoint, a vehicle emissions tax vehicle emission standards lack flexibility and may thus would be an ideal economic incentive for controlling impose higher costs in return for lower benefits than emissions. Although such a tax would pose formidable more flexible approaches. To avoid disrupting the mar- implementation problems, a properly implemented ket, universal standards must be set at a level that nearly emission tax could encourage vehicle owners to pur- all vehicles can meet, which implies that some vehicles chase clean cars (leading manufacturers to compete in are capable of meeting stricter standards but have no in- cleanliness as they now compete in fuel economy), and centive to do so. In most cases, standards do not differ- it would encourage them to maintain their cars properly entiate between vehicles according to use-the taxicab so that they continue to be clean in use. An alternative in a highly polluted urban center meets the same stan- to the emissions tax that might be easier to implement dards as a car used in a remote rural area. Emissions con- would be to impose a high indirect tax (e.g. on fuel), and trol for the latter vehicle has little or no social utility and then to offer a rebate on this tax based on a vehicle's social resources are wasted. emission performance in a representative test such as A number of approaches have been taken to reduce the IM240 described in Chapter 4. This would create an these drawbacks. In the United States, for example, pro- incentive for drivers to undergo the test (in order to re- grams for emissions averaging, trading, and banking ceive the rebate) rather than to avoid it. 24 Air Poflutfonfrom Motor Vebicles Very similar in theoretical effect to a vehicle emis- .1995. Motor Vehicle Emission Regulations and sions tax would be the provision of mobile-source emis- Fuel Specifications in Europe and the United sion reduction credits, which could be traded to States-1995 Update. Brussels. stationary sources or other vehicle owners in lieu of CSEPA (China State Environmental Protection Adminis- meeting emission regulations. If combined with a suffi- tration). 1989,'Emission Standards for Exhaust Pollu- ciently tight limit on overall emissions, such a program tion from Light-Duty Vehicles." National Standard would provide an incentive for those who could reduce GB11641-89, Beijing. emissions cost-effectively to do so, in order to sell the Havenith, C., J.R. Needham, A.J. Nicol, and C.H. Such. resulting reductions (or rights to emit) to others for 1993."Low Emission Heavy-Duty Diesel Engine for Eu- whom reducing emissions would be more costly. Credit rope."SAE Paper 932959.Warrendale, Pennsylvania. programs of this kind are now being implemented in a India, Ministry of Surface Transport. 1989. 'Report of number of jurisdictions in the United States. The Working Group on RoadTransport for the Eighth Plan (1990-95)." Government of India, New Delhi. Onursal, B. and S. Gautam. 1996. 'Vehicular Air Pollu- References tion: Experience from Seven LAC Urban Centers."A World Bank Study (forthcoming),Washington, D.C. Baines, T.M. 1994. Personal Communication. U.S. Envi- Plaskett, L. 1996.Airing the Differences."Financial ronmental Protection Agency,Washington, D.C. Times,June 26,1996, London. Boletin Oficial. 1994. Section No. 27.919. Buenos Aires, UNIDO (United Nations Industrial Development Orga- Argentina. nization). 1990. "Control and Regulatory Measures CETESB. 1994. Relatorio de Qualidade do Ar no Esta- Concerning Motor Vehicle Emissions in the Asia-Pa- do de Sso Paulo - 1993. Companhia de Tecnologia cific Region. Report from a Meeting at the Korea In- de Saneamento Ambiental, Sao Paulo, Brazil. stitute of Science and Technology, Seoul. Chan, L.M. and C.S.Weaver. 1994.'Motorcycle Emission van Ruymbeke, C., R.Joumard, R.Vidon, and C. Provost. Standards and Emission Control Technology. Depart- 1992.'Representativity of Rapid Methods for Measur- mental Paper Series No. 7, Asia Technical Depart- ing Pollutant Emissions from Passenger Cars." IN- ment,TheWorld-Bank,Washington, D.C. RETS Report LEN9219. Bron, France. CONCAWE (Conservation of Clean Air and Water in Eu- Walsh, M.P. 1995. 'Technical Notes." 3105 N. Dinwiddie rope). 1992. Motor Vehicle Emission Regulations Street, Arlington, Virginia. and Fuel Specfflcations-1992 Update. Report 2/ _. 1996a.'Car Lines." Issue 96-3, 3105 N. Dinwiddie 92, Brussels. Street, Arlington,Virginia. . 1994. Motor Vehicle Emission Regulations and World Bank. 1992. "Transport Air Quality Management Fuel Specifications-1994 Update. Report 4/94, in the Mexico City MetropolitanArea." Sector Report Brussels. No. 10045-ME.The World Bank,Washington, D.C. 2 Quantifying Vehicle Emissions There are several procedures for measuring vehide Vehicle emission factors for a given jurisdiction emissions for regulatory purposes. The most commonly should ideally be based on emission measurements per- used are the U.S. federal, the United Nations Economic formed on a representative sample of in-use vehicles Commission for Europe (ECE), and the Japanese test pro- from that area. Such data collection is expensive, how- cedures. The U.S. and European tests are also used ex- ever, and requires facilities that few countries possess. tensively in other countries.These test procedures have Without specific data, preliminary planning can be many common elements. For light-duty vehicles, includ- based on MOBILE5 or COPERT estimates for similar ing motorcycles, emissions are measured by operating technologies, or on data from countries with similar the vehicle on a chassis dynamometer while collecting vehicle fleet characteristics. its exhaust in a constant-volume sampling system.Test- ing for heavy-duty vehide engines is done on an engine dynamometer. EImissions Measurement and The main difference among the procedures is the Testing Procedures driving cycle (for vehicles) or operating cycle (for Motor vehice emissions are highly variable. In addition heavy-duty engines). The European and Japanese pro- to differences among vehicles, differences in operating cedures test in a series of steady-state operating condi- conditions ang ehions fromce in operto tions, while the U.S. procedures involve transient conditions can cause emissions from a given vehide to variations in speed and load more typical of actual driv- change by more than 100 percent. A consistent and rep- ing. None of the tests fully reflect real-world driving licable test procedure is required if emission regulations patterns, however. More representative driving cycles or incentive systems are to be enforceable.To ensure im- have been developed, and are being considered for provement in emissions, testing should be repre- adoption in the United States (AQIRP 1996). sentative of in-use conditions or severe enough to Vehicle emissions are affected by driving patterns, ensure effective emission control system performance traffic speed and congestion, altitude, temperature, and under all conditions. Current procedures do not always other ambient conditions; by the type, size, age, and achieve this objective.This chapter includes the key fea- condition of the vehicle's engine; and, most important- tures of the various emission test procedures. Detailed ty, by the emnissions control equipment and its mainte- descriptions of these procedures can be found in official nance. Emission factors are estimates of the pollutant documents - Part 86 of the Code of Federal Regulations emiissions produced per kilometer traveled by vehicles for the U.S. procedures and the various ECE standards of a given class. Computer models estimate vehicle available through the United Nations ECE secretariat in emission factors as functions of speed, ambient Geneva. temperature, vehicle technology, and other variables. The U.S. EPA's MOBILE5 is probably the most widely ExhaustEmissionsTestingforLight-DutyVehicles used. It is based on emission tests carried out on in-use Three test procedures are presently used to measure vehicles as part of the U.S. EPA's emissions surveillance the emissions of light-duty vehicles: the U.S. federal test program. The European Union's COPERT model in- procedure (FTP), the European test procedure estab- cludes a wide cross-section of European vehicles and is lished by the United Nations Economic Commission for based on an extensive program of emission tests carried Europe (ECE) regulation 83, and the Japanese test pro- out in several European countries. cedure. The U.S. procedure has now been adopted 25 26 Air Pollutionfrom Motor Vehicles throughout North America, and is also used in Brazil, Although it is intended to represent typical urban Chile, Republic of Korea, Taiwan (China), and some driving (based on a transient cycle representative of Western European countries. The European test proce- driving patterns in Los Angeles, California), the driving dure and emissions standards are used in the European cycle used for the U.S. test procedure does not cover Union, most Eastern European countries, China, and the full range of speed and acceleration conditions that India, where it includes an Indian driving cycle.Though vehicles experience.When the cycle was adopted in the primary used in Japan, the Japanese procedure and stan- early 1970s, chassis dynamometers had limited capabil- dards are accepted by several East Asian countries. ities that made it necessary to use low speeds and All three test procedures measure exhaust emis- acceleration rates.The top speed in the U.S. cycle is 91 sions produced while the vehicle is driven through a kilometers per hour and maximum acceleration is 5.3 prescribed driving cycle on a chassis dynamometer. kilometers per hour per second (1.47 m/sec2), both are Emissions are sampled by means of a constant volume lower than what most vehicles can achieve on the road. sampling (CVS) system (figure 2. 1). The specific driv- As the FTP cycle does not cover the full range of pos- ing cycle differs, however. Because emissions in urban sible speed and acceleration conditions, emissions areas are the principal concern of control programs, under off-cycle conditions are effectively uncontrolled all testing is based on vehicles operating in stop-and-go (AQIRP 1996). Manufacturers can and do take advan- driving conditions typical of urban areas.The layout of tage of this to increase the power output and perfor- a typical emissions testing laboratory is shown in fig- mance of their vehicles under off-cycle conditions. As a ure 2.2. result, vehicle emissions may increase dramatically un- der these conditions. For example, most gasoline pas- US. procedure. In the U.S. federal test procedure (FTP- senger car engines use a rich mixture and shut off 75), the vehicle is driven on a chassis dynamometer ac- exhaust gas recirculation at or near full throttle, causing cording to a predetermined speed-time trace (driving huge increases in CO emissions. Such increases associ- cycle) while exhaust emission samples are diluted, ated with high power and load conditions (such as hard cooled, and collected in the constant volume sampling acceleration, high speed operations, or use of accesso- system. The driving cycle, lasting 2,475 seconds, re- ries), can soar as high as 2,500 times the emission rate flects the varying nature of urban vehicle operation (fig- noted for stoichiometric operations. Although most ve- ure 2.3). The average driving speed is 31.4 kilometers hicles spend less than 2 percent of their total driving per hour (excluding the ten-minute "hot soak" between time in severe enrichment conditions, this can account 1,370 and 1,970 seconds when the engine is shut off). for up to 40 percent of total CO emissions (Guensler The test begins with a cold start (at 20° to 300 Celsius) 1994). NOX and HC are also increased. after a minimum 12-hour soak.The emission results re- The U.S. EPA and CARB have studied in-use driving ported are calculated as the weighted average of emis- patterns and found that high speed and high accelera- sions measured during three phases: cold start, hot tion driving are not uncommon, and their inclusion in a stabilized, and hot start. test procedure can greatly affect overall emission mea- Figure 2.1 Exhaust Emissions Test Procedure for Light-Duty Vehicles CVS Sampling System Air Filter Dilution Tunnel Air Pump Atmospheric Air Dynamometer Rolls Constant Volume - E qtnt Sampling Pump Source: Weaver and Chan 1995 Quantfying Vebicle Emissions 27 Figure 2.2 Typical Physical Layout of an Emissions Testing Laboratory 30 m (98.5 ft.) ° ° ° OOOOO - - -- - -................... , ..........F 0allbratn 0 ases Sample Heavy-duty Chassis Dynamometer CalibratIo Gases Bags 9 m . ' ~~~~~~~Heaydy CVS an I Particutate tune l\Y - ..................... bay doors ________ J_ vyu _,~ __ __ __ __ _ and_________F 0 0 O O g Wlndow e ZUght-dutyCVSand I / I :l :Partculate tunnel Motorcycle iii/ Analy/zer Dynamotr Benhes - Ught-duty Chassis (2()ft. ................ .......-- :Drmoer: aOOoaO!. I ~~~~~~~~~~Dynamonmeter Om 1: ! \ "" . ............. Sampie o O °Saple B a g s~ . . p . a sI s0 0 0 0 Fustorage B_10 . ~~(26 ft.=) ~ ~~(33 ft.) I (39 ft.) Source: Weaver and Chan 1995 Figure 2.3 U.S. Emissions Test Driving Cycle for Light-Duty Vehicles (FTP-75) 100 80 for motorcycles < 170cc 80 J40 20 0 I - - ------_ _ - -,- T - 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 rime (seconds) Source: CONCAWE 1994 28 AirPolUutionfrom MotorVebicles surements. CARB further concluded that these effects cedure. The proposed revisions also include testing may be responsible for emissions in the South Coast Air with the air conditioner on, to better reflect actual emis- Basin of California being as much as 7 percent higher sions during warm weather. for nitrogen oxides, 20 percent higher for hydrocar- Until recently the U.S. emissions test was only bons, and 80 percent higher for carbon monoxide than carried out at ambient temperatures of about 20 to projected by the current emission models. 30°C. Although it was known that hydrocarbons and es- Another failing of the FTP-75 is the poor simulation pecially carbon monoxide emissions increase greatly at of air conditioner operation. The air conditioner does low temperatures because of increased cold start en- not run during the FTP; instead, dynamometer load is richment and slower catalyst light-off, these low tem- increased to simulate the additional load on the engine perature emissions were unregulated. As a result low due to the air conditioner compressor. EPA tests of sev- temperature carbon monoxide emissions are five to ten eral vehicles with the air conditioner running showed times the emissions of carbon monoxide at normal am- NO, emissions nearly double those without the air con- bient conditions. Recent legislation in the United States, ditioner, although fuel consumption increased only 20 however, requires carbon monoxide emissions testing percent. Apparently, manufacturers have not optimized at both -7°C and normal ambient temperature, bringing for low emissions with the air conditioner on, since this source of excess emissions under control for all they know that the vehicle will not be tested in that new vehicle models by 1996. condition. A revised version of the U.S. test procedure has been European procedure.The emissions test procedure for developed and proposed for adoption by the U.S. EPA. European passenger cars was defined by ECE regulation This procedure includes an additional driving cycle, 15 and consists of three tests. Like the U.S. procedure, called the US06, constructed so that when combined the first test measures the exhaust emissions produced with the existing FTP (figure 2.4) it includes the full in a driving cycle on a chassis dynamometer.The differ- range of speed and load conditions found in actual driv- ence is the driving cycle used.The European test proce- ing. The US06 cycle includes higher speeds and more dure is a simplified representation of the driving cycle aggressive driving patterns than in the current U.S. pro- in a typical European urban center (such as Rome),with Figure 2.4 Proposed U.S. Enviromnental Protection Agency US06 Emissions Test Cycle 140 130 120 110 100 90 80 CV 70 0 60 50 .2 40 130 20 0 0 50 100 150 200 250 300 350 400 450 500 550 600 Time (seconds) Source: Weaver and Chan 1995 QuantotyIng Vehicle Emissions 29 15 linked driving modes (figure 2.5). The maximum Exhaust Emissions Testingfor speed for the European test cycle is 50 kilometers per Motorcycles and Mopeds hour. The vehicle is allowed to soak for at least six The emissions test procedure for motorcycles in the hours at a temperature between 20 and 300C. It is then United States is the same as for light-duty passenger started and, after idling for 40 seconds, is driven cars, except that the maximum speed is reduced for through the test cycle four times without interruption. motorcycles with an engine displacement of less than The cycle lasts 780 seconds and totals 4.052 kilometers, 170 cc engine displacement (figure 2.7). For passenger making the average speed 18.7 kilometers per hour. cars and motorcycles with displacement over 170 cc, The second test samples tailpipe carbon monoxide con- the maximum speed is 91 kilometers per hour, while for centrations immediately after the last cycle of the first motorcycles under 170 cc it is 58.7 kilometers per hour. test.The third test measures crankcase emissions. Testing is done on a single-roll dynamometer equipped Compared with the U.S. procedure, the European with a flywheel to simulate the inertia of motorcycle driving cycle is simple, consisting of stable speeds and rider, and with a clamp to hold the motorcycle up- linked by uniform accelerations and decelerations.The right. The European test procedure for motorcycles is absence of real-world driving effects reduces the se- the same as for light-duty vehicles, but does not include verity of the test. The Institut National de Recherche the extra-urban driving cycle. sur les Transports et Leur Securite (INRETS) in France As with passenger cars, present emission test cycles has determined that the European test cycle, charac- for motorcycles are inadequate. The operating charac- terized by uniform transitions between steady states, teristics of many motorcycles make acceleration rates underestimates actual emissions by 15-25 percent particularly significant. Observations in Bangkok when compared with more realistic driving at the indicate that motorcycle acceleration rates in traffic same average speed (Joumard and others 1990). Also, often exceed 12 kilometers per hour per second (3.3 the maximum acceleration rate in the European m/sec2), more than twice the maximum acceleration procedure is 3.75 kilometers per hour per second rate in the U.S. procedure, and three times that in the (1.04 m/sec2)-significantly less than the U.S. proce- European procedure (Chan and Weaver 1994). dure, and thus even less representative of real driving. ECE regulation 47 defines the test procedure for emis- This rate is sustained for just four seconds during the sions from vehicles with less than 50 cc engine displace- first brief peak in the driving cycle; rates for the other ment and an unladen weight of less than 400 kilograms two accelerations in the cycle are 2.61 and 1.92 kilo- (figure 2.7). Such vehicles are almost entirely mopeds. 2- meters per hour per second (0.73 and 0.53 m/sec , re- The maximum speed in this test cycle is 50 kilometers spectively), less than one fifth the rate commonly per hour, or the maximum speed the vehicle can reach observed in actual driving. at wide-open throttle if less than 50 kilometers per hour. To account for higher vehicle speeds outside of ur- The United States does not regulate emissions from mo- ban centers, EU emission test procedures now include torcycles with less than 50 cc engine displacement, but an extra-urban driving cycle (figure 2.6). The extra- the U.S. EPA has recently proposed doing so.The test cy- urban driving cycle (EUDC) lasts 400 seconds at an av- cle would be the same as for motorcycles with an engine erage speed of 62.6 kilometers per hour, with a maxi- displacement between 50 and 170 cc. mum speed of 120 kilometers per hour.The maximum acceleration rate, however, is only 3 kilometers per Exhaust Emissions Testingfor hour per second (0.83 m/sec2). Since the maximum Heavy-Duty Vehicle Engines speed and the maximum acceleration rate are still con- Heavy-duty, on-highway vehicles are also tested by U.S., siderably less than in actual driving, emissions at off-cy- European, and Japanese test procedures. The U.S. pro- cle conditions remain uncontrolled. As testing is only cedure (1985) tests under transient conditions while the conducted between 20° and 30°C, low-temperature car- European (ECE 49) and Japanese (13-mode test) proce- bon monoxide emissions are not controlled. dures use steady state tests. (These procedures are de- scribed in detail in CONCAWE 1994, and in official Japanese procedure.The Japanese test procedure is sim- documents.) All three procedures measure exhaust emis- ilar to the European one. Until 1991 the main test for sions from the engine alone (removed from the vehicle), light-duty vehicles was a 10-mode driving cycle simulat- operating over a specified cycle on an engine dynamom- ing congested urban driving.This was replaced with the eter. The U.S. procedure involves transient changes in 10.15 mode cycle, which adds a segment reaching 70 ki- speed and load to mimic actual road operation.The Euro- lometers per hour. Both tests are hot-start procedures pean and Japanese procedures measure emissions at a (i.e. the vehicle is already warmed up). A separate, 11- number of specified steady-state conditions, and com- mode test cycle measures cold-start emissions. bine these according to a weighting scheme. Results of 30 Air Pollution from Motor Vehicles Figure 2.5 European Emissions Test Driving Cycle (ECE-15) (repeatedfour times) 8.0 _ Z- 50 0 40 ~340 0. CD, a) 0 0 20 40 60 80 100 120 140 160 180 200 Time (seconds) Source: CONCAWE 1994 Figure 2.6 European Extra-Urban Driving Cycle (EUDC) 120 - o 100 80 80 Declutch 60 - speed of Q a- 50 kM/h v. 40 - or lower Gearchanges as shown (D 20- 0- 0 50 100 150 200 250 300 350 400 Time (seconds) Source: CONCAWE 1994 Quant4fyng Vebicle Emissions 31 Figure 2.7 European Emissions Test Driving Cycle for Mopeds 50 7 one cycle 40- 0 _E 0) 20- C,, WOT 10 0 0 112 224 336 448 Time (Seconds) WOT = Wide-open throttle 4 cydes warm up, 4 cycles measurement Source: CONCAWE 1994 the U.S. and European tests are reported in mass of pol- Testing in-use vehicles with engine dynamometers is lutant emissions per unit of work output rather than difficult since the engine must be removed from the ve- emissions per vehicle-kilometer. This is because of the hicle. Accordingly, emissions from in-service heavy- wide range of sizes and applications in heavy-duty vehi- duty vehicles are usually measured with the whole vehi- cles, each of which would otherwise have to be tested in- cle operating on a chassis dynamometer. A number of dividually on a chassis dynamometer. Emissions per test cycles have been developed for this purpose, in- kilometer are strongly affected by vehicle size and fuel cluding a chassis version of the U.S. heavy-duty transient consumption and would require different standards for test cycle and various cycles simulating bus operations. each vehicle category. Regulation based on work output These include the New York City cycle and the Santiago allows one set of standards to be applied to engines used bus cycle developed in Chile (Steiner 1989). in a broad range of vehicles.The Japanese test is reported The comparative advantages of transient and steady- in terms of pollutant concentration corrected to standard state test procedures have often been debated during the conditions, which has a similar effect. past few decades (Cornetti and others 1988). U.S. author- The U.S. heavy-duty transient test consists of engine ities adopted the transient procedure (replacing an earli- speed and load transients that were selected to simulate er 13-mode steady-state test) in 1985, arguing that steady- intracity truck operations.This is because trucks greatly state tests do not adequately measure the air-fuel ratio outnumber buses in U.S. urban areas. Engine testing during transient operations such as acceleration. Diesel yields pollutant emissions per unit of work output of the particulate matter and hydrocarbon emissions are sensi- engine (in grams per horsepower-hour).The U.S. EPA for- tive to the test cycle used, particularly in transient condi- mula for converting grams per horsepower-hour to tions. Particulate matter and hydrocarbon emissions in grams per mile utilizes a conversion factor of about three transient tests are generally found to be higher than in horsepower-hours to a mile, based on fuel consumption. steady state tests. Emissions of nitrogen oxides show bet- However, tests by the U.S. EPA and other organizations ter correlation between the two types of tests. suggest that using the U.S. EPA transient test for bus en- Regulations based on a specific emissions test usu- gines may underestimate actual bus chassis emissions by ally only control emissions in the operating modes ex- a factor of three to six (Alson,Adler, and Baines 1989). perienced during that procedure. Since vehicles 32 Air Pollutionfrom Motor Vehicles operate in a variety of speed-load conditions, it is im- Evaporative Emissions portant that testing procedures reflect these condi- United States and California. The U.S. procedure for tions. Emission control strategies based on a light-duty vehicles measures evaporative emissions procedure that measures emissions at a limited num- from simulated diurnal heating and cooling and evapo- ber of operating conditions may be insufficient. Elec- ration from the carburetor under hot-shutdown condi- tronic engine control systems can be programmed to tions. Evaporative emission requirements apply to undermine a steady state test cycle, by minimizing passenger cars, light-duty trucks, and heavy-duty vehi- emissions only at the test points.This is much more dif- cles using gasoline or other volatile fuels (but not die- ficult in a transient test cycle. In response to these ar- sel). California also tests motorcycles. guments, European authorities are developing an Evaporative emissions are measured by placing the appropriate transient test cycle (Baines 1994). vehicle in an enclosure called Sealed Housing for Evap- Adopting the U.S. test cycle in Europe would pro- orative Determination (SHED), which captures all va- mote international standardization, but it may not be pors emitted from the vehicle. The diurnal portion of the optimal solution. Most of the work during the U.S. the evaporative test measures emissions from the vehi- test cycle is produced near rated speed and load (Cor- cle as the temperature in the gasoline tank is increased netti and others 1988). While the U.S. EPA considers from 15.6 to 28.90C. This simulates the warming that this to be typical of truck operations in U.S. cities, it is occurs as the temperature rises during the course of a not typical of European driving patterns or long dis- day. tance driving patterns in the United States. The The hot soak portion of the evaporative test mea- development of high torque-rise engines, overdrive sures emissions from the vehicle for one hour following transmissions, road speed governors, and the increasing the exhaust emissions test. The vehicle is moved from concern for fuel economy are resulting in U.S. trucks the dynamometer into the SHED as soon as the driving spending more time operating near peak torque speed, test is complete.The sum of diurnal and hot-soak emis- even in urban areas. Since the transient test contains lit- sions total grams per test, which is the regulated quan- tle operation at this speed, manufacturers are able to tity. Individual diurnal and hot soak test results can also calibrate their engines for best performance and fuel be used to translate grams per test into grams per kilo- economy at peak torque, rather than for least emissions. meter. The formula is: Trucks that run mostly near peak torque conditions diurnal grams/test + N * (hot soak grams/test) rather than near rated speed will produce more pollut- grams/kilometer = average kilometers driven per day (2.1) ant emissions than estimated from the emission test re- sults. Thus, the argument is strong for a revised test where N is the average number of trips per day. cycle that would give increased emphasis to peak torque operating conditions. Given the significantly Evaporative test procedures have recently been higher proportion of trucks and buses in the traffic reevaluated in the United States.This is because testing streams in developing countries, there may be a case to programs indicate that fuel evaporation is a larger develop test cycles that are more representative of op- source of emissions of volatile organic compounds than erating conditions (low power to weight ratio and slow- had previously been recognized and that under some er speeds) in developing countries. circumstances significant evaporative emissions were coming even from vehicles equipped with evaporative Crankcase Emissions control systems meeting U.S. EPA and CARB standards. In failing to test the evaporative control system under There is no common procedure for measuring crank- conditions as severe as in actual use, existing diurnal case emissions across countries. European regulations and hot-soak test procedures may have limited the de- specify a functional test to confirm the absence of vent- gree of evaporative control achieved. ing from the crankcase, while U.S. regulations simply These issues have been addressed in the latest Cali- prohibit crankcase emissions. Uncontrolled crankcase fornia and U.S. EPA regulations, which have established emissions have been estimated from measurements of more elaborate evaporative test procedures.These pro- volatile organic compound concentrations in blowby cedures are intended to be more representative of real- gases. On vehicles with closed crankcase ventilation world vehicle evaporative emissions. The California systems, crankcase emissions are assumed to be zero. procedure includes a 72-hour triple diurnal test cycle in Adoption of European standards for controlling a SHED that ranges between 18 and 41°C. Running loss- crankcase emissions may be more appropriate for de- es are measured by operating the vehicle on a chassis veloping countries, given the similarities of engine dynamometer through three consecutive U.S. driving technology. cycles in the SHED at 410C.The U.S. EPA test cycle is Quant/ifyng VebJcle Emissions 33 similar but involves less extreme temperatures. These sampler (CVS), gas analyzers, a sensor and measuring tests have required manufacturers to design higher-ca- system for fuel consumption, an optical sensor to mea- pacity evaporative emissions control systems that sure vehicle speed and distance travelled, and a data pro- achieve better in-use control even under extreme cessor on a laptop computer for online collection, with conditions. real-time processing and evaluation of data.This system Both the CARB and the U.S. EPA limit evaporative has been used to measure vehicle exhaust emissions in hydrocarbon emissions to 2.0 grams per test, which is motorway and rural highway traffic. Measured emissions considered effectively equivalent to zero (a small have been within 10 percent of laboratory data.The sys- allowance is needed for other, non-fuel related organic tem can be used with all carbon fuels to directly measure emissions from new cars, such as residual paint sol- exhaust emissions of carbon monoxide, hydrocarbons, vent).The new test procedures have the same 2.0 grams nitrogen oxides, and carbon dioxide in either grams per per test limit, with a separate limit on running-loss emis- second or grams per kilometer (Lenaers 1994). sions of 0.05 gram per mile. Although the standards are nominally the same as before, the more severe testing conditions impose more stringent requirements on Vehicle Emission Factors manufactus rer. Pollutant emission levels from in-service vehicles vary depending on vehicle characteristics, operating condi- Europan Uion.Unti recntly he Eropen teting tions, level of maintenance, fuel characteristics, and am- procedures did not provide for evaporative emission bient conditions such as temperature, humidity, and measurements as evaporative hydrocarbon emissions bi. codtin suha eprtre uiiyn measremets a evaoratve hdrocrbonenusions altitude.The emission factor is defined as the estimated were not regulated. This was remedied by the Consoli- atiueTem sonfcrisdiedsthetmtd average emission rate for a given pollutant for a given dated Emidssions Directive issued by the Council of class of vehicles. Estimates of vehicle emissions are ob- Ministers of the European Community in June 1991 .The taied by multiplying an estimate of the distance tray- directive established evaporative emission limits based eled by a given class of vehicles by an appropriate on tests similar to the former U.S. SHED test procedure. emission factor. Because of the many variables that influence vehicle JKapan. The Japanese evaporative test procedure meca-y emissions (see box 2.1), computer models have been de- sures hot-soak emissions only; diurnal emissions, run- emsin(eebx2 )coptrodlhaeend- sures hot-esoakd remission lonly; diurena memissions r veloped that estimate emission factors under any combi- ning losses, and resting losses are not measured. The ' ~~~~~~~~~~~nation of conditions. Two of the most advanced models test uses carbon traps connected to the fuel system at are the U.S. EPA's MOBILE series (the current version is points where fuel vapors may escape into the atmo- arhMOBILE5a), and the EMFAC model developed by GARB. sphere. The vehicle IS driven at 40 kilometers per hour MOIEa,ndteMFCmelevopdbCAB onspchassiTh vehiclmetis d orive 40 kileters pher thou e Both models use statistical relationships based on thou- on a chassis dynamometer for 40 minutes, then the en- gine is stopped, the exhaust is sealed, and preweighed sands of emission tests performed on both new and used carbon traps are connected to the fuel tank vent, air vehicles. In addition to standard testing conditions, cleaner, and any other possible vapor sources. After one many of these vehicles have been tested at other tem- peratures, with different grades of fuel, and under differ- hour, the traps are reweighed. ent driving cycles. Relationships have been developed Refueling Emissions for vehicles at varying emission control levels, ranging from no control to projections of in-use performance of Testing for refueling emissions involves measuring future low-emission vehicle fleets. concentrations of volatile organic compounds in the Although accurate emission factors and an under- vapors vented from gasoline tanks during the refueling standing of the conditions that affect them are obviously process and observing spillage frequency and volume. important for air quality planning and management, data Emissions from the underground storage tank vent are for in-service vehicles are surprisingly poor. Even in the monitored by measuring the flow rate and concentra- United States, where systematic emission measurements tions of volatile organic compounds in gases emitted have been carried out on in-service vehicles for more from the vent. than a decade, there is considerable uncertainty about Oln-Road Exhaust Emissions the applicability of the results. The most important sources of uncertainty are the sensitivity of vehicle emis- To obtain emissions data that is directly representative of sions to the driving cycle, the wide variety of driving pat- actual traffic conditions and driving patterns, a number terns, and the effects of sampling error, given the highly of on-board systems have been developed and tested. A skewed distribution of emission levels among vehicles system developed by the Flemish Institute for Techno- equipped with emission controls.The U.S. sampling sur- logical Research includes a miniature constant volume veys indicate that a small fraction of 'gross" emitters in 34 Ar Pollutionfrom Motor Vebicles Box 2.1 Factors Influencing Motor Vehicle Emissions 1.Vehicle/Fuel Characteristics * Engine type and technology-two-stroke, four-stroke; Diesel, Otto, Wankel, other engines; fuel injection, turbocharging, and other engine design features; type of transmission system * Exhaust, crankcase, and evaporative emission control systems in place-catalytic converters, exhaust gas recirculation, air injection, Stage II and other vapor recovery systems * Engine mechanical condition and adequacy of maintenance . Air conditioning, trailer towing, and other vehicle appurtenances * Fuel properties and quality-contamination, deposits, sulfur, distillation characteristics, composition (e.g., aromatics, olefin content) additives (e.g., lead), oxygen content, gasoline octane, diesel cetane * Alternative fuels * Deterioration characteristics of emission control equipment * Deployment and effectiveness of inspection/maintenance (WM) and anti-tampering (ATP) program 2. Fleet Characteristics . Vehicle mix (number and type of vehicles in use) * Vehicle utilization (kilometers per vehicle per year) by vehicle type. * Age profile of the vehicle fleet * Traffic mix and choice of mode for passenger/goods movements * Emission standards in effect and incentives/disincentives for purchase of cleaner vehicles * Adequacy and coverage of fleet maintenance programs * Clean fuels program 3. Operating Characteristics * Altitude, temperature, humidity (for NO, emissions) * Vehicle use patterns-number and length of trips, number of cold starts, speed, loading, aggressiveness of driving behavior * Degree of traffic congestion, capacity and quality of road infrastructure, and traffic control systems * Transport demand management programs Source: Faiz and others 1995; Faiz and Aloisi de Larderel 1993 the vehicle fleet are responsible for a large fraction of to- MOBILE 4.1 predictions of CO/NOX and NMHC/NOX ra- tal emissions. These are generally vehicles in which tios were in closer agreement with the observed ratios ermission controls are malfunctioning, tampered with, or than were MOBILE5 predictions. damaged. It is difficult to represent this minority accu- Emission factors calculated by the MOBILE models are rately in a sample of reasonable size. Another concern is based on average speeds, ambient temperature, diurnal the potential for sampling bias: owners must agree to temperature range, altitude, and fuel volatility; changes in have their vehicles tested, and owners of the worst vehi- these input assumptions alter the resulting emission fac- cles may be less likely to do so. A consensus is develop- tors. Exhaust pollutant emission factors increase mark- ing that the combined effect of these problems has edly at low temperatures, while evaporative emissions of caused existing models to underestimate motor vehicle volatile organic compounds increase with increasing emission factors by a substantial margin. temperature. Evaporative emissions of volatile organic A study under the U.S. Air Quality Improvement Re- compounds also increase as gasoline volatility and diurnal search Program (AQIRP 1995) compared real world vehi- temperature range increase. Hydrocarbon and carbon de emissions to values calculated using the U.S. EPA monoxide emissions per vehicle-kilometer tend to in- MOBILE computer models (MOBILE4.1 and MOBILE5). crease at low average speeds, such as in congested city In general, the MOBILE models predicted emissions rates driving, while emissions of nitrogen oxides tend to in- within + 50 percent although at one site MOBILE5 over- crease at high speeds, which corresponds to higher load predicted rates to a much greater extent. MOBILE5's pre- conditions.The relationship between average speed and dictions were consistently higher than those of emissions estimated by MOBILE5 for uncontrolled motor MOBILE4.1 at both test sites (Fort McHenry tunnel under vehicles is shown in figure 2.8. Low average speeds are Baltimore Harbor and Tuscarora tunnel in the mountains due to traffic congestion, and the increase in emissions of Pennsylvania). Both MOBILE modes underpredicted under these conditions is due to the stop-and-go pattern light-duty non-exhaust emission rates, which constituted of traffic flow in congested condition. approximately 15-20 percent of the total light-duty non- Other emission models have been developed, though methane hydrocarbon emissions. For light-duty vehicles, none incorporates as much data on in-use emissions as Quant4fying Vebicle Emissions 35 Figure 2.8 Relationship between Vehicle Speed and Emilssions for Uncontrolled Vehicles 900 - _ 800 - - Carbon Monoxide 700 - - 600 ,500 0 400 - - Heavy-duty Gasoline 300- - 200 - ' %,Heavy-duty Diesel 1 00 - - - - 0 - 4-Passenger Car 0 10 20 30 40 50 60 70 80 90 Speed (km/hr) 50 -_ 45 Hydrocarbons 40 35 30 5' 25 - 20 Heavy-duty Gasoline 15 - Passenger 10 Car S - Heavy-duty Diesel 0 0 10 20 30 40 50 60 70 80 90 Speed (km/hr) 35 -- Nitrogen Oxides 30- - 25- - b, 20 - Heavy-duty Diesel e z 15 - 10 - _ 5 - _ Heavy-duty Gasoline 20 4 Passenger Car O - I 1i I I I I I I I 0 1 0 20 30 40 50 60 70 80 90 Speed (km/hr) Source: Chan and Weaver 1995 36 Air Pollutionfrom Motor Vebicles MOBILE5 and EMFAC.The COPERT model (Andrias and quently been used to estimate vehicle emission factors others 1992) applies a methodology developed by the for uncontrolled vehicles in such countries as Chile, In- CORINAIR working group on emission factors to calcu- donesia, and Mexico. This provides only a rough esti- late emissions from road traffic in the EU (Eggleston and mate, however, since the technologies and others 1991). COPERT can be used to estimate vehicle characteristics of today's vehicles, even without emis- emission factors for carbon monoxide, non-methane hy- sion controls, are significantly different from the uncon- drocarbons, methane, oxides of nitrogen, total particulate trolled (pre-1970) U.S. vehicles that were used to matter, ammonia, and nitrous oxide. Fuel consumption develop the MOBILE models. In addition, emission fac- estimates are also provided. Emission factors are estimat- tors for motorcycles and heavy-duty vehicles have re- ed for urban, rural, and highway driving with an average ceived little attention in the past and are supported by automobile speed of 25 kilometers per hour, 75 kilome- limited data.They should be considered rough estimates ters per hour, and 100 kilometers per hour, respectively. for vehicles in the United States, and even less represen- COPERT accounts for cold-start emissions and evapora- tative of vehicles in developing countries. One weak- tion losses, and uses an average trip length of 12 kilome- ness of MOBILE5 is that it does not estimate particulate ters. Less extensive emission factor models have also matter (PM) emissions. Although a PM model based on been developed for vehicles in Chile (Turner and others MOBILE5a became available in 1994, this model does 1993), Indonesia (IGRP 1991), andThailand (Chongpeer- not account for particulate emissions deterioration in apien and others 1990). The MOBILE4 model has been use, and thus underestimates real-world emissions. adapted to estimate emissions from the Mexican vehicle Since motorcycles and heavy-duty vehicles are among fleet (Radian Corporation 1993). the most significant vehicular sources of air pollution in The Swiss Federal Environment Department has com- many developing countries, and PM emissions are among missioned the development of a data bank of vehicle ex- the most pressing concerns, it is clear that continued re- haust and evaporative emissions for both regulated and liance on MOBILE5 will be insufficient.Vehicle emission unregulated pollutants. Information on nearly 300 differ- factor models should be developed on the basis of emis- ent compounds, including specific hydrocarbons, aIde- sion tests carried out under local conditions and should hydes, phenols, polycyclic aromatic hydrocarbons, and reflect actual in-use performance of vehicles. several inorganic compounds, is available in the data A key requirement for effective long-run emissions bank (Brunner and others 1994). The data bank current- control is an ongoing program that monitors in-service ly contains about 16,000 emission factors, classified by vehicle emissions in an appropriate emissions laborato- six main vehicle categories (gasoline-fueled, diesel-fu- ry. It is essential to measure actual vehicle emissions be- eled, and other light-duty vehicles; gasoline-fueled, die- fore and after control measures are implemented to sel-fueled, and other heavy-duty vehicles; and two- know whether they are effective and how they could be wheelers). improved. As mentioned elsewhere, surveillance testing Extensive emissions testing of in-use vehicles is re- is an important element of U.S. in-service emissions test- quired to develop and validate an emission factor model. ing.This testing is performed to verify compliance with For lack of better data, the MOBILE5 emissions have fre- emissions durability requirements. Testing programs Box 2.2 Development of Vehicle Emissions Testing Capability in Thailand The Royal Thai government has adopted an action plan that addresses the air pollution and noise problems caused by road ve- hicles. Among the measures included in this plan is the provision of a vehicle emissions laboratory. The primary purpose of this laboratory is the measurement and development of in-use vehicle emission factors. Another important function is to devel- op improved vehicle emission short tests for use in the planned inspection and maintenance program.The laboratory will be capable of measuring (using constant volume sampling under simulated transient driving conditions) exhaust emissions of car- bon monoxide, oxides of nitrogen, hydrocarbons, and particulate matter from two- and four-stroke motorcycles, three-wheel taxis (tuk-tuks), light-duty gasoline and diesel vehicles, and heavy-duty diesel vehicles weighing up to 21 metric tons. The laboratory will also analyze driving patterns so that Bangkok-specific driving cycles can be established. The laboratory is ex- pected to cost about $2 million. Equipment costs will be financed by the World Bank. Plans call for three test cells, one each for motorcycle, light-duty vehicle/light truck, and heavy-duty truck/bus testing.Testing equipment will include chassis dynamometers, constant volume sampling and dilution tunnel units, gas analyzer instruments, and data acquisition and control hardware. Each test cell will have its own dynamometer, so the entire weight and power range of vehicles can be tested.Three-wheelers will be tested on the motorcycle dynamometer. A set of driving cycles representative of Bangkok and otherThai traffic conditions will be developed, to provide representa- tive emission factors for the local vehicle population.The laboratory will provide the flexibility to run standard tests (such as the U.S., European, and Japanese certification cycles) as well as custom-designed cycles. Source: Chan and Weaver 1994 Quantifying Vebicle Emissions 37 have also been carried out on in-use vehicles in Chile (Es- hydrocarbons, nitrogen oxides, and carbon monoxide. cudero 1991; Sandoval, Prendez, and Ulriksen 1993), Emissions of the greenhouse gases carbon dioxide and Finland (Laurikko 1995), and Greece (Pattas, Kyriakis, nitrous oxide were also estimated: the former based on and Samaras 1985), among others. Development or ex- typical fuel economy and fuel carbon content, the latter pansion of emission laboratories for such testing is pro- based on the data available for different types of emis- ceeding in several developing countries, including sions control systems. Brazil, Iran, Mexico, and Thailand (box 2.2).The results An inventory of emission factors for gasoline-fueled obtained by these laboratories are expected to add sig- vehicles derived from the COPERT model and individual nificantly to the knowledge of vehicle emission charac- studies in Europe and several developing countries is teristics and compilation of appropriate emission factors presented in appendix 2.1. These emission factors are for use in developing countries. likely to be more representative of conditions in devel- oping countries, although there is considerable variation Gasoline-Fueled Vehicles among emission factor measurements, even for similar Emission factor estimates for U.S. gasoline-fueled pas- vehicles and test conditions.This variation indicates the senger cars and medium-duty trucks equipped with dif- importance of basing emission factor estimates on actual ferent levels of emission control technology are measured emissions for a vehicle fleet rather than rely- presented in tables 2.1 and 2.2. These estimates incor- ing on data or estimates from other sources. This not porate MOBILE5a results for methane, non-methane only ensures accurate emission factors, it provides an Table 2.1 Estimated Emission Factors for U.S. Gasoline-Fueled Passenger Cars with Different Emission Control Technologies (grams per kilometer) Non-methane Fuel efficiency Carbon volatile organic Nitrogen Nitrous Carbon (liters per Type of control monoxide compounds Methane oxides oxide dioxide 100 kilometers) Advanced three-way catalyst control Exhaust 6.20 0.38 0.04 0.52 0.019 200 8.4 Evaporative 0.09 Running loss 0.16 Resting 0.04 Total emissions 6.20 0.67 0.04 0.52 0.019 200 Early three-way catalyst Exhaust 6.86 0.43 0.05 0.66 0.046 254 10.6 Evaporative 0.14 Running loss 0.16 Resting 0.06 Total emissions 6.86 0.78 0.05 0.66 0.046 254 Oxidation catalyst Exhaust 22.37 1.87 0.10 1.84 0.027 399 16.7 Evaporative 0.39 Running loss 0.17 Resting 0.06 Total emissions 22.37 2.48 0.10 1.84 0.027 399 Non-catalyst control Exhaust 27.7 2.16 0.15 2.04 0.005 399 16.7 Evaporative 0.70 Running loss 0.17 Resting 0.06 Total emissions 27.7 3.08 0.15 2.04 0.005 399 Uncontrolled Exhaust 42.67 3.38 0.19 2.7 0.005 399 16.7 Evaporative 1.24 Running loss 0.94 Resting 0.06 Total emissions 42.67 5.62 0.19 2.7 0.005 399 Note: Estimated with the U.S. EPA MOBILESa model for the following conditions: temperature, 24 °C; speed, 31 kilometers per hour; gasoline Reid Vapor Pressure, 62 kPa (9 PSI); and no inspection and maintenance program in place. Source: Chan and Reale 1994;Weaver andTumer 1991 38 Air Polutionfrom Motor Vebicles Table 2.2 Estimated Emission Factors for U.S. Gasoline-Fueled Medium-Duty Trucks with Different Emission Control Technologies (grams per kilometer) Non-metbane Fuel efficiency Carbon volatile organic Nitrogen Nitrous Carbon (liters per 7)pe of control monoxide compounds Methane (xides oxide diaxide 100 kilometers) Three-way catalyst control Exhaust 10.2 0.83 0.12 2.49 0.006 832 34.5 Evaporative 0.38 Running loss 0.17 Resting 0.04 Refueling 0.24 Total emissions 10.2 1.41 0.04 0.52 0.006 832 Non-catalyst control Exhaust 47.61 2.55 0.21 3.46 0.006 843 35.7 Evaporative 2.16 Running loss 0.94 Resting 0.08 Refueling 0.25 Total emissions 47.61 5.73 0.21 3.46 0.006 843 Uncontrolled Exhaust 169.13 13.56 0.44 5.71 0.009 1,165 50.0 Evaporative 3.93 Running loss 0.94 Resting 0.08 Refueling 0.32 Total emissions 169.13 18.50 0.44 5.71 0.009 1,165 Note: Estimated with the U.S. EPAs MOBILE5a model for the following conditions: temperature, 24 "C; speed, 31 kilometers per hour; gasoline Reid Vapor Pressure, 62 kPa (9 PSI); and no inspection and maintenance program in place. Source: Chan and Reale 1994;Weaver andTurner 1991 important baseline against which the effectiveness of without emission controls.This effect can be observed emission control programs can be measured. in figure 2.10, which shows cumulative distributions for Emission factors are strongly influenced by the way a hydrocarbons, carbon monoxide, nitrogen oxides, and vehicle is driven-in particular, by the average speed particulate matter emissions from a sample of Chilean and the degree of acceleration and deceleration in the cars tested in 1989.The lower curve in each plot is the driving cycle (Joumard and others 1995). The results of cumulative percentage of cars with emissions greater emission tests on a number of European vehicles, using than a given level, while the upper curve is the percent- a variety of driving cycles, are shown in figure 2.9 (jou- age of total emissions accounted for by these cars. Only mard and others 1990). Average emissions per 20 percent of the vehicles, for example, had hydrocar- kilometer increase sharply at the low average speeds bon emissions above 1.2 grams per kilometer, but these typical of highly congested stop-and-go urban driving. vehicles were responsible for 40 percent of total hydro- Emissions are minimized in free-flowing traffic at mod- carbon emissions. Similar patterns were found for other erate, speeds, then increase again under the high-speed pollutants: 10 percent of the vehicles accounted for 37 driving conditions typical of European motorways. percent of total PM emissions, 20 percent of the vehicles Emissions were higher in the transient test cycles; accounted for 43 percent of total carbon monoxide steady-state cycles gave much lower emissions per emissions, and 20 percent of the vehicles accounted for kilometer. The European test cycle, characterized by 35 percent of total emissions of oxides of nitrogen. Since uniform transitions between steady states, underesti- owners of the worst polluting vehicles may be less likely mates actual emissions by about 15 percent. to volunteer them for testing, the real distribution could Pollutant emissions are affected by the vehicle's level be even more skewed. This distribution has important of maintenance, and the highest-polluting vehicles are consequences for emissions control strategy. An inspec- responsible for a disproportionate share of total emis- tion and maintenance program that identifies the worst sions (chapter 4). Emission-controlled vehicle fleets 10 to 20 percent of the vehicles and requires that they show the most skewed distribution of emissions, but the be repaired or retired could reduce overall emissions uneven distribution is significant even for populations significantly. Quant4fyng Vebicle Emissions 39 Figure 2.9 Effect of Average Speed on Emissions and Fuel Consumption for European Passenger Cars without Catalyst (INRETS Driving Cycles; Fully Warmed-up In-use Test Vehicles) 100 600 * Gasoline * Gasoline 80 * Diesel 500 0 Diesel ~400 60 g~~~~~~~~~~~~300 140 ° 20- E 200 0~~~~~~~~~~~~~ o) 20 100 0w ,*_ *i.@ @1 .I*1 I 0 I I I I I .I I I 0 30 60 90 120 0 30 60 90 120 Average speed (km/h) Average speed (km/h) 12 - 4- * Gasoline * Gasoline 510 - Diesel | l sesel x 8 0 6 ~~~~~~~~~~~~~~2 E ~ ~ ~ ~~11 ' z 0 I I I I I I I I I 0 30 60 90 120 0 30 60 90 120 Average speed (knVh) Average speed (km/h) 25 --250 - *Gasoline * Gasoline D 20 V Diesel 200 0 Diesel U)15 150 E * 0 110 0 ~~~~~~~00 0 30 60 90 120 0 30 60 90 120 Average speed (kmr/h) Average speed (km/h) Source: Adapted from joumard and others 1990 Diesel-Fueled Vehicles Similar estimates for U.S. heavy-duty diesel-fueled trucks Emission factors for diesel-fueled vehicles are strongly and buses are given in table 2.4. Estimates for diesel-fu- affected by differences in engine technology, vehicle eled vehicles in Europe and other regions for a variety of size and weight, driving cycle characteristics, and the emissions control levels are summarized in appendix state of maintenance of the vehicles. Emission factor 2.2.These emission factors may be more representative estimates for diesel-fueled passenger cars and light-duty of diesel-fueled vehicles operating in developing coun- trucks in the United States are summarized in table 2.3. tries. There is, however, considerable variation among 40 Air PoUutionfrom Motor Vebicles Figure 2.10 Cumulative Distribution of Emissions from Passenger Cars in Santiago, Chile Distribution of CO emissions Distribution of PM emissions 100 100- _ _ _ - 90--- 90 -- 1-= - -- 90it% of emissions Di o of emissions 80 - Hor % of vehicles N 80 (g/km) 70- 70-- -- -- 60--___ - 60 ……-- &50 5 40 4 30--- 30 20- 20 10 - o___ 0 20 40 60 0 0. 0.6 10.9 1 0 30 50 0.1 0.3 0.5 0.7 0.9 1.1 Carbon monoxide (gtab) Particulate matter (gAmm) Distribution of HG emissions Distribution of NO x emissions 100 ~~~~~~~~100 - - 90 ~~~~~~90 - Order t % of emissions - %oof emhiss io % of vehicles 80 p r%l Ofvehidles 70 - ~~~~~70- -_ 40- senger 4 car 40 ih rcsi al . r oprbe gos eil egtrtnso 0t 0tn.h ms 30 20 ~~~~~20 - - - 0.2 0.6 1 1.4 1.8 2.2 2.6 0 123 0.4 0.8 1.2 1.6 2 2.4 2.8 0.8 1.5 2.5 3.5 Hydrocarbons (gllkm) Nitrogen oxides (g/1km) Source: Turner and othcrs 1993 the estimates and measurements in the tables and ap- and appendix 2. 1. This is not true for the estimates for pendix because of differences in cycle conditions, differ- heavy-duty vehicles. Although heavy-duty diesel-fueled ences in the sample population, and different estimation and gasoline-fueled vehicles are covered by similar emis- techniques.This indicates again the importance of actual sion standar-ds and test procedures, the average charac- emiission measurements on the population of interest in teristics of the vehicles themselves differ considerably. order to develop realistic emiission factors. Unlike heavy-duty gasoline-fueled vehicles, heavy-duty The emiission factor estimates for diesel-fueled pas- diesel-fueled vehicles are primnarily large trucks with senger cars and light trucks in table 2.3 are comparable gross vehicle weight ratings of 10 to 40 tons. The emis- to the estimates for gasoline-fueled vehicles in table 2.1 sion factors in table 2.4 are therefore more representa- Quanttfylng Vebicle Emissions 41 Table 2.3 Estimated Emission and Fuel Consumption Factors for U.S. Diesel-Fueled Passenger Cars and Light-Duty Trucks (grams per kilometer) Carbon Nitrogen Particulate Carbon Fuel consumption Vebicle type monoxide Hydrocarbons axides matter dioxide (liters/I00 km) 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 - Not applicable Note: MOBILE5 estimates. Source: Chan and Reale 1994; Weaver and Turner 1991 ;Weaver and Klausmeier 1988 Table 2.4 Estimated Emission and Fuel Consumption Factors for U.S. Heavy-Duty Diesel-Fueled Trucks and Buses (grams per kilometer) Carbon Nitrogen Particulate Carbon Fuel consumption Vehicle type monoxcide Hydrocarbons oxides matter dioxide (7iters/100 km) 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 - Doubleaxle 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 - Not applicable Note: MOBILE5 estimates. Source: Chan and Reale 1994; Weaver andTurner 1991;Weaver and Klausmeier 1988 tive of large trucks (and buses) than smaller medium- increase sharply with increasing grades, particularly for duty trucks and vans. The opposite is true for the gaso- nitrogen oxides and hydrocarbons. There is also some line-fueled vehicle emission factors in table 2.2. correlation of carbon monoxide and particulate emtis- Average pollutant emissions from heavy-duty diesel- sions with speed and road gradient (Roumegoux 1995). fueled vehicles are especially sensitive to the speed There are also large variations in particulate and hy- and acceleration characteristics of the driving cycle. drocarbon emission factors for uncontrolled vehicles. Emissions per kilometer vary according to the average This is partly the result of differences in maintenance, cycle speed. For heavy-duty diesel-fueled vehicles, which can have a tremendous effect on particulate emis- emissions of carbon monoxide, hydrocarbons, and sions. In testing buses in Chile, for example, particulate particulate matter increase at low average speeds, due matter emissions from well-maintained buses were more to the stop-and-go driving associated with congested than 80 percent less than the average particulate emis- traffic. Emissions of nitrogen oxides and fuel consump- sions for the entire bus fleet. Cumulative probability dis- tion increase between 40 and 50 kilometers per hour tributions for particulate matter, hydrocarbon, and (figure 2.11). nitrogen oxides emissions from the Chilean bus fleet The effects of steady-state vehicle speed and road gra- illustrate the nature of the problem (figure 2.13). The 10 dient on emissions and fuel consumption of a 40-ton semi- percent of the buses with highest emissions are respon- trailer truck are shown in figure 2.12. On negative grades, sible for 25 percent of total particulate emissions, while the emissions of all pollutants are insignificant, but they 20 percent of the buses produced 40 percent of the 42 Air Pollution from Motor Vebicles Figure 2.11 Effect of Average Speed on Emissions and Fuel Consumption for Heavy-Duty Swiss Vehicles 25 25 220 E 20 .2 ~~~~~~~~~~~~0 LI) ~veag spe LI)/h Avrg spe0knh 10 10x 0 0~~~~~~~~~~0 1 0 20 30 40 50 60 70 80 90 100 0 20 3 40 5 60 7 9 10 Average speed (krrvh) Average speed (km/h) 30 0.7 25 0.6 ~120 0.5 0 ~ ~ S4 ( / 15 g~~~~~~~~~~~~~ 0.4 0~ 30_ ¶10 10 Eo 0.2 ol lo~~~~~ z 0 5 0.1 10 20 30 40 50 60 70 80 90 100 0 1- 16 10 20 30 40 50 60 70 80 90 100 Average speed (kff/h) Average speed (km/Vh) 50 -z 40 30 0 20 ii10 U. 0 10 20 30 40 50 60 70 80 90 100 Average speed (km/Vh) Source.: OFPE 1988 QuantlJtylng Vehlcle Emlsslons 43 Figure 2.12 Effect of Constant Average Speed and Road Gradient on Exhaust Emissions and Fuel Consumption for a 40-ton Semi-Trailer Truck 250 5000 8200 .4000 O 150 c -a4% 3000 oL 0 E 100 - * 2000 * 2% . t* *t22% .* 0% 0 20 40 6080100120 140 020 40 6080 100120 140 Steady-state speed (km/h) Steady-state speed (km/h) 6000 5000 50 ^ T54 04000 Z 3000 0 MO_ E looot o% M . ,,, .,,, .,2 . -2%l 0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 .140 Steady-state speed (knVh) Steady-state speed (km/ah) 50D .102000 6E 000 2 4000 0 1000 2O-%%40% 3 23000 ^ * ; J 100 0-Y 0 -2% 0 20 4086080 100120140 02Z040 60 80100120 140 Steady-state speed (km/h) Steady-state speed (kmlh) c c 2~~~~~~~~~~000 4 .9 2000 ~ ~ ~ ~ ~ ~ ~ ~ . 40 SrE: Romeoux E19950 Q6000 2%~~~~~~~~2% e0 p n o heusit e I i - -i g c es i 0 204060 80100 120 140 0 20 40 6080 100120140 Steady-state speed (knVh) Steady-state speed (km/h) emissios.The 0 pecn- ftebss ihlws s Ize vehcle indvlpigcutie4priclryi sions (generally those with the best mnaintenance) pro- East and South Asia.While they are responsible for a rel- duced only 7 percent of particulate emissions. atively small fraction of the total vehicle kilometers trav- eled in most countries, they are major sources of air Motorcycles pollution-particularly two-stroke engines running on a Two- and three-wheeled vehicles, such as motorcycles mixture of gasoline and lubricating oil. It has been esti- and auto-rickshaws, constitute a large portion of motor- mated that uncontrolled motorcycles in industrialized 44 Air Polution from Motor Vebicles countries emit 22 times the mass of hydrocarbons and cycle emission regulations developed by U.S. EPA and 10 times as much carbon monoxide as automobiles con- CARB. Data on uncontrolled emissions from U.S. motor- trolled to U.S. 1978 levels (OECD 1988). In Taiwan (Chi- cycles are from this period (table 2.5). These data na), hydrocarbon emissions from two-stroke engine include the only particulate emission measurements on motorcycles were 13 times higher than the hydrocar- motorcycles available in the technical literature, and bon emissions from new four-stroke motorcycles and they allow an assessment of the relationship between more than 10 times higher than the hydrocarbon emis- smoke opacity and particulate emissions.The data are sions from in-use passenger cars. Carbon monoxide mostly for high- powered, large-displacement touring emissions from two-stroke motorcycle engines were motorcycles, which are common in the United States similar to those from four-stroke engines (Shen and Hua- but not in most developing countries. Motorcycles used ng 1991). in developing countries seldom exceed 250 cc in en- Data on motorcycle emissions are scarce. Available gine displacement. Due to their smaller size and weight, data pertain mostly to uncontrolled U.S. motorcycles, these motorcycles would be expected to have lower though limited information is also available from emissions than large touring motorcycles. Europe and Thailand. The average emissions for uncontrolled two-stroke motorcycles were 9.9 grams per kilometer of hydrocar- United States. Between 1970 and 1980, considerable bons, 16.1 grams per kilometer of carbon monoxide, study of motorcycle emissions was prompted by motor- 0.022 grams per kilometer of nitrogen oxides, and Figure 2.13 Cumulative Distribution of Emissions from Diesel Buses in Santiago, Chile Distribution of HC emissions Distributon of NOxemissions Combined autobus and taxibus samples Combined autobus and taxibus samples 100 100 - -- I t 80 -% of emissions 80 - _ - _ -% of emissions .____ % of vehicles S _ ---%of vehides 640 _ _ _ _ _ 0 - - - - - - - - - - - 20 -20- t X 0t CT B t. .. ... .. 0 6 01 I 1117 0 1 2 4 6 0 2 4 6 8 10 12 1 3 5 1 3 5 7 9 1 Hydrocarbons (ghkm) Nitrogen oxides (glkm) Distribution of particulate emissions Combined autobus and taxibus samples 100 -C I so- - % of emissions % of vehides so &40 20 - - - - - - - 0- 0 2 4 6 8 10 1 3 5 7 9 11 Particulate matter (g/km) Source: Tumer and others 1993 Quant4fying Veb/cle Emissions 45 0.281 grams per kilometer of particulate matter. Com- particulate emissions for two-stroke motorcycles. A cor- pared with emissions from four-stroke motorcycles, relation between measured particulate matter emis- emissions from uncontrolled two-strokes are about ten sions and smoke opacity is given by: times higher for hydrocarbons, carbon monoxide emis- PM=0.066 + (0.015 * OP) (2.2) sions are similar, and emissions of nitrogen oxides are less (though motorcycle emissions of nitrogen oxides where PM = particulate matter (g/km), are always small compared to those from other vehi- and OP = smoke opacity for a 3 path length (percent). cles). Particulate matter emissions from the two-stroke motorcycles tested were also about five to ten times Europe. Limited data exist on uncontrolled motorcycle those of four-stroke motorcycles. PM emissions data are emissions in Europe. A report prepared by the Swiss available for only five motorcycles, however, and cover Office of Environmental Protection in 1986 reviewed a wide range, from 0.082 grams per kilometer to 0.564 several studies on motorcycle emissions and concluded grams per kilometer. In addition, three of the particu- that average exhaust emissions for uncontrolled motor- late measurements were not taken under the U.S. pro- cycles and mopeds were within the range of European cedure but as a weighted combination of measurements standards. The average emissions for 35 uncontrolled under different steady-state operating conditions.Thus, four-stroke motorcycles, 40 uncontrolled two-stroke there is some uncertainty about how well these data motorcycles, and 141 two-stroke mopeds are shown in represent real-world driving conditions. In addition, all table 2.6. The motorcycles tested were in consunmer measurements were made on new or nearly-new motor- use, so these data might be representative of real motor- cycles that were properly adjusted and maintained. cycle emissions.The emissions for European two-stroke Actual emissions in consumer service would be expect- and four-stroke motorcycles are about twice the levels ed to be significantly higher. of new, uncontrolled motorcycles in the United States. The U.S. emissions data can be used to develop an es- Although the data are not strictly comparable because timate of the relationship between smoke opacity and of the differences in the test cycles, they do suggest that Table 2.5 Emission and Fuel Consumption Factors for Uncontrolled U.S. Two- and Four-Stroke Motorcycles (grams per kilometer) Carbon Nitrogen Particulate Fuel economy Engine type monoxide Hydrocarbons oxides matter (liters per 100 kilometers) Two-stroke 16.1 9.9 0.022 0.206 4.7 Four-stroke 23.5 2.0 0.135 0.048 5.2 Four-stroke with displacement less than 250 cc 15.0 1.0 0.206 0.020 2.9 Source: Chan andWeaver 1994 Table 2.6 Emission Factors for Uncontrolled European Motorcycles and Mopeds (grams per kilometer) Vebicle type Engine type Number tested Carbon monoxide Hydrocarbons Nitrogen oxides Motorcycle Four-stroke 35 40.0 5.9 0.2 Two-stroke 40 24.6 19.0 0.035 Moped Two-stroke 141 10.0 6.0 0.06 Source: Chan and Weaver 1994 Table 2.7 Emission and Fuel Consumption Factors for Uncontrolled Thai Motorcycles (grams per kilometer) Fuel economy Engine type Carbon monoxide Hydrocarbons (liters per 100 kilometers) Four-stroke 19.0 2.9 1.6 Two-stroke 28.1 14.6 2.5 Source: Chan and Weaver 1994 46 Air Polutionfrom Motor Vebicles Figure 2.14 Smoke Opacity Emissions from Motorcycles in Bangkok, Thailand .2 '- __ _ _- _____ ____ 70 --- - - __ E_ __ .0 o o 50=_== = E E O 40 12 20 (0 E o 10 _ _ 0 10 20 3 40 s0 60 70 s0 80 100 Smoke opacity (percent) Source: Chan and Weaver 1994 average in-use emissions from uncontrolled motorcy- and 28.1 grams per kilometer, respectively; the averages cles are higher than emissions from new, properly ad- for the four-stroke motorcycles were 2.9 and 19.0 grams justed motorcycles (Chan and Weaver 1994). per kilometer, respectively. These values are similar to but lower than the average for European motorcycles Thailand. A cumulative distribution plot of accelera- reported in table 2.6. Since the Thai data are for new tion smoke opacity for 167 randomly selected motorcy- motorcycles and the European data are for motorcycles des in Bangkok is shown in figure 2.14. More than 95 in use, this is not surprising. In addition to having lower percent of the motorcycles tested were equipped with emissions, four-stroke motorcycles also had much bet- two-stroke engines.The mean smoke opacity, corrected ter fuel economy than the two-strokes, averaging 1.6 to a three-inch path length, was 61 percent-four times liters per 100 kilometers compared with 2.5 liters per the opacity for the smokiest of the uncontrolled U.S. motorcycles with opacity measurements ranging from 3 to 18 percent. By extrapolating the correlation in equa- tion 2.2 it can be estimated that the 61 percent average References smoke opacity in Bangkok is equivalent to average par- ticulate matter emissions of about 1.0 grams per kilome- Alson, J., J. Adler, and T. Baines. 1989. "Motor Vehicle ter (figure 2.14). Emission Characteristics and Air Quality Impacts of The Thai Department of Pollution Control has gath- Methanol and Compressed Natural Gas'" in D. Sper- ered data on hydrocarbon and carbon monoxide emis- ling, ed. Alternate Transportation Fuels: An Envi- sions for 17 Thai-produced motorcycles based on the ronmental and Energy Solution. Greenwood Press, European test cycle. These data were obtained from Wesport, Connecticut. manufacturers and thus probably represent new, prop- Andrias, A., D. Zafiris, Z. Samras, and K-H. Zierock. erly-adjusted motorcycles (table 2.7). Of the 17 motor- 1992. 'Computer Program to Calculate Emissions cycles, two had four-stroke engines and the rest had from Road Traffic-User's Manual,' COPERT, EC Con- two-strokes. Average hydrocarbon and carbon monox- tract No. B4-3045 (91) 10 PH (DG XI/B/3), European ide emissions for the two-stroke motorcycles were 14.6 Commission, Brussels. Quantifying Vebicle Emissions 47 AQIRP 1995. 'Real World Automotive Emissions - Re- Faiz,A., and J.Aloisi de Larderel. 1993. 'Automotive Air sults of Studies in the Fort McHenry and Tuscarora Pollution in Developing Countries: Outlook and Con- Mountain Tunnels." Air Quality Improvement Re- trol Strategies". The Science of the Total Environ- search Program, Technical Bulletin No. 14. Auto/Oil ment, 134:325-344. Industry Research Council,Atlanta, Georgia. Guensler, R. 1994. "Loop Holes forAir Pollution," ITS Re- -___ 1996. "Dynamometer Study of Off-Cycle Exhaust view, Volume 18, Number 1. University of California, Emissions." Air Quality Improvement Research Pro- Berkeley. gram, Technical Bulletin No. 19. Auto/Oil Industry IGRP (Indonesian German Research Project). 1991."En- Council, Atlanta, Georgia. vironmental Impacts of Energy Strategies for Indone- Baines, T. 1994. "Personal Communication." U.S. Envi- sia." VWS Report on Assessment of the Emission ronmental ProtectionAgency, Office of Mobile Sourc- Coefficients of the Traffic Sector in Jawa.Jakarta. es,Washington, D.C. Joumard, R., L. Paturel, R. Vidon, J. P. Guitton, A. Saber, Brunner, D., K. Sclapfer, M. Ros, and E Dinkel. 1994. and E. Combet. 1990. Emissions Unitaires de Pollu- "Data Bank on Unregulated Vehicle Exhaust and ants des Vhicules L6gers. Report 116(2nd ed.) Insti- Evaporative Emissions." 3rd International Sympo- tut National du Recherche sur les Transports et Leur sium on Transport and Air Pollution, Poster Proceed- Securite (INRETS), Bron, France. ings, INRETS,Arcueil, France. Joumard, R., P. Jost, J. Hickman, and D. Hassel. 1995. Chan, L.M. and M. Reale. 1994."Emissions Factors Gen- "Hot Passenger Car Emissions Modelling as a erated from In-House MOBILE5a." Engine, Fuel, and Function of Instantaneous Speed and Acceleration'" Emissions Engineering. Sacramento, California. The Science of the Total Environment, 169:167-174. Chan, L.M. and C.S.Weaver. 1994."Motorcycle Emission Laurikko, J. 1995. "Ambient Temperature Effect on Auto- Standards and Emission Control Technology." motive Exhaust Emissions: FTP and ECE Test Cycle Departmental Paper Series, No. 7,Asia Technical De- Responses." The Science of the Total Environment, partment,The World Bank,Washington, D.C. 169:195-204. Chan, L.M. and C.S. Weaver. 1995. "Figures Generated Lenaers, G. 1994. "A Dedicated System for On-the-road with the Use of MOBILE5a," Engine, Fuel, and Emis- Exhaust Emissions Measurements on Vehicles." 3rd sions Engineering, Sacramento, California. International Symposium on Transport and Air Pollu- Chongpeerapien, T., S. Sungsuwan, P. Kritiporn, S. Bura- tion, Poster Proceedings, INRETS,Arcueil, France. nasajja. 1990. "Energy and Environment, Choosing OECD (Organization for Economic Cooperation & De- the Right Mix." Resource Management Associates Re- velopment). 1988. Transport and the Environment. search Report No. 7, Thailand Development Re- Paris. search Institute, Bangkok. OFPE (I'Office federal de la protection de l'environne- CONCAWE (Conservation of Clean Air and Water in Eu- ment). 1988. "Emissions Polluantes du Trafic Routier rope). 1994. Motor Vehicle Emission Regulations Prive de 1950 a 2000." Les cahiers de l1environne- and Fuel Specifications-1994 Update. Report 4/94. ment 55, Berne, Switzerland. Cornetti, G., K. Klein, G. Frankle, and H. Stein. 1988. Pattas, K., N. Kyriakis, and Z. Samaras. 1985. "Exhaust "U.S. Transient Cycle Versus ECE R.49 13 - Mode Emission Study of the CurrentVehicle Fleet inAthens Cycle. SAE Paper 880715, Society of Automotive (Phase II)." Final Report. 3 volumes. University of Engineers,Warrendale, Pennsylvania. Thessaloniki, Greece. Eggleston, H., N. Gorissen, R.Joumard, R. Rijkeboeer, Z. Radian Corporation. 1993. "Revision of the MOBILE- Samaras, and K. Zierock. 1991. "CORINAIR Working MEXICO (Mobile Mexico) Model Report." Departa- Group on Emission Factors for Calculating 1990 mento del Distrito Federal, Mexico City, Mexico. Emissions From Road Traffic." Volume 1. Methodolo- Roumegoux, J.P 1995. "Calcul des Emissions Unitaires gy and Emission Factors. Final Report, EC Study Con- de Polluants des Vehicules Utilitaires", The Science of tract B4-3045(91)10PH. Commission of the the Total Environment, 169: 273-82. European Communities. Brussels. Sandoval, H., M. Prendez, and P Ulriksen eds. 1993. Con- Escudero, J. 1991. "Notes on Air Pollution Issues in San- taminacion Atmosferica de Santiago: Estado Actual tiago Metropolitan Region," Letter #910467, dated y Soluciones, University of Chile, Santiago. May 14, 1991. Comisi6n Especial de Descontami- Shen, S., and Huang, K. 1991 ."Response to Transport-In- naci6n de la Region Metropolitana. Santiago. duced Air Pollution: The Case of Taiwan." in M.L. Birk Faiz, A., S. Gautam, and E. Burki. 1995. "Air Pollution and D.L. Bleviss, eds., Driving New Directions: from Motor Vehicles: Issues and Options for Latin Transportation Experiences and Options in Devel- American Countires." The Science of the Total Envi- oping Countries. International Institute for Energy ronment, 169:303-310. Conservation,Washington, D.C. 48 Air Poiluttonfrom Motor Vebicles Steiner,A. 1989. "Operacion y Resultados del Programa Report to the California Air Resources Board, Engine, de Mediciones de la UMEVE." Paper Presented at the Fuel, and Emissions Engineering Inc., Sacramento, Seminar on Contaminacion Atmosferica Debido a California. Motores de Combustion Interna, Santa Maria Univer- Weaver, C.S., and S. Turner. 1991. "Mobile Source sity, Valparaiso, Chile. Emission Factors for Global Warming Gases." Report Turner, S.H., C.S. Weaver, and M.J. Reale. 1993. "Cost to U.S. EPA Contract 68-W8-0113. Sacramento, and Emissions Benefits of Selected Air Pollution Con- California. trol Measures in Santiago, Chile." Engine, Fuel, and Weaver, C.S., and L.M. Chan. 1995."CompanyArchives" Emissions Engineering, Inc. Sacramento, California Multiple Sources. Engine, Fuel, and Emissions Engi- Weaver, C.S., and R. Klausmeier. 1988. "Heavy-Duty neering Inc., Sacramento, California. Diesel Vehicle Inspection and Maintenance Study." Appendix 2. 1 Selected Exhaust Emission and Fuel Consumption Factors for Gasoline-Fueled Vehicles Table A2.1.1 Exhaust Emissions, European Vehicles, 1970-90 Average (grams per kilometer) Traffic type and engine/emission control cbaracteristics Carbon monoxide Hydrocarbons Nitrogen oxides Particulate matter Urban Two-stroke 32.9 20.2 0.26 0.00 More than 2,000 cc 32.0 3.0 2.00 0.01 1,400-2,000 cc 31.0 2.9 1.80 0.10 Less than 1,400 cc 30.0 2.8 1.70 0.14 Catalyst without 02 sensor 6.6 0.9 1.00 0.00 Catalyst with 02 sensor 1.5 0.2 0.27 0.00 Rural highway Two-stroke 21.6 13.0 0.33 0.00 More than 2,000 cc 21.0 2.2 2.49 0.01 1,400-2,000 cc 20.0 2.1 2.29 0.01 Less than 1,400 cc 19.0 2.0 2.09 0.01 Catalyst without 02 sensor 4.5 0.5 1.15 0.00 Catalyst with 02 sensor 1.2 0.1 0.24 0.00 Motorway Two-stroke 22.7 13.0 0.42 0.00 More than 2,000 cc 21.5 2.3 3.03 0.01 1,400-2,000 cc 20.5 2.2 2.83 0.01 Less than 1,400 cc 19.5 2.1 2.63 0.01 Catalyst without 02 sensor 5.0 0.5 1.35 0.00 Catalyst with 02 sensor 1.3 0.1 0.32 0.00 Source: Metz 1993 Table A2.1.2 Exhaust Emissions, European Vehicles, 1995 Representative Fleet (grams per kilometer) Emission regulations/controls Carbon monoxide Hydrocarbons Nitrogen oxides Carbon dioxide Particulate matter ECE 15-03 31.5 3.57 2.29 188 ECE 15-04 24.1 2.97 2.40 192 Three-way catalytic converter 5.2 0.32 0.40 247 Diesel 0.7 0.13 1.22 188 0.14 - Not applicable a. Induded for comparison. Source: Joumard and others 1995 49 50 Air PoUution from Motor Vebicles Table A2.1.3 Estimated Emissions and Fuel Consumption, European Vehicles, Urban Driving (grams per kilometer) Carbon Non-metbane Nitigen Nitrous Fuel monoxide volatile org compounds oxides Methane oxide Ammonia Enap. consumption Vehicle type (CO) (NM-VOC) (NOx) (CH4) (N20) (NH3) emissions (liter/I00 km) Passenger Cars Two-stroke 20.7 15.2 0.3 0.150 0.005 0.002 1.13 14.93 LPG 6.9 11.8 1.9 0122 0.000 0.000 - 8.78 Pre-1971 (Pre PCE) Less t..han 1,400 cc 56.8 4.2 1.7 0.224 0.005 0.002 1.13 12.99 1,400-2,000 cc 56.8 4.2 1.9 0.224 0.005 0.002 1.13 15.38 More than 2,000 cc 56.8 4.2 2.4 0.224 0.005 0.002 1.13 19.23 1972-76 (ECE15/00-01) Less than 1,400 cc 41.6 3.3 1.7 0.224 0.005 0.002 1.13 11.63 1,400-2,000 cc 41.6 3.3 1.9 0.224 0.005 0.002 1.13 13.89 More than 2,000 cc 41.6 3.3 2.4 0.224 0.005 0.002 1.13 15.63 1977-79 (ECE15/02) Less than 1,400 cc 35.4 3.3 1.5 0.224 0.005 0.002 1.13 10.64 1,400-2,000 cc 35.4 3.3 1.7 0.224 0.005 0.002 1.13 12.82 More than 2,000 cc 35.4 3.3 2.0 0.224 0.005 0.002 1.13 15.87 1980-84 (ECE15/03) Less than 1,400 cc 22.3 3.3 1.6 0.224 0.005 0.002 1.13 10.64 1,400-2,000 cc 22.3 3.3 1.9 0.224 0.005 0.002 1.13 12.82 More than 2,000 cc 22.3 3.3 2.5 0.224 0 005 0.002 1.13 15.87 1985-89 (ECE15/04) Less than 1,400 cc 21.4 2.6 1.6 0.224 0.005 0.002 1.13 5.71 1,400-2,000 cc 21.4 2.6 1.9 0.224 0.005 0.002 1.13 10.53 More than 2,000 cc 21.4 2.6 2.3 0.224 0.005 0.002 1.13 18.29 Improved conventional Less than 1,400 cc 13.5 1.8 1.4 0.224 0.005 0.002 1.13 7.94 1,400-2,000 cc 7.8 1.5 1.5 0.224 0.005 0.002 1.13 10.53 Oxidizing catalyst without I sensor Less than 1,400 cc 15.7 1.6 1.1 0.224 0.005 0.002 1.13 8.70 1,400-2,000 cc 7.6 0.4 0.9 0.224 0.005 0.002 1.13 10.53 3 way catalyst with I sensor Less than 1,400 cc 4.1 0.4 0.4 0.062 0.050 0.070 0.23 8.47 1,400-2,000 cc 4.1 0.4 0.4 0.062 0.050 0.070 0.23 10.31 More than 2,000 cc 4.1 0A4 0.4 0.062 0.050 0.070 0.23 12.99 Light-duty vehicles 46.1 4.7 3.1 0.230 0.oo6 0.002 1.03 17.86 Motorcycles Less than 50 crn3 10.0 5.9 0.1 0.100 0.001 0.001 0.28 2.40 More than 50 cm3, two-stroke 22.0 14.9 0.1 0.150 0.002 0.002 0.45 4.00 More than 50 cm3, four-stroke 20.0 2.8 0.3 0.200 0.002 0.002 0.45 5.08 n.a. = Not available - Not applicable Notes * Average driving speed 25 kilometers per hour. . Emission factors in g/km are derived from the COPERT model for 1990, utilizing the CORINAIR methodology for road traffic emissions. The pollutants included are: CO, NMVOC, NO., CH4, N20, NH3. Fuel consumption is also estimated. Total VOC (or HC) emission factors may be obtained by adding NM-VOC and Ci4 factors. * Cold-start emissions are calculated for urban driving conditions only, taking into account the monthly variation of the average minimum and maximum temperatures. For this specific application, average temperatures in Brussels (Belgium) were used,with the following monthly distribution: Month 01 02 03 04 05 06 07 08 09 10 11 12 Tmin (°C) -1.2 0.3 2.2 5.1 7.9 10.9 12.1 12.2 10.6 7.3 3.1 0.2 Tmax (C) 4.3 6.7 10.3 14.2 18.4 22.0 22.7 22.3 20.5 15.4 8.9 5.6 * Evaporative losses are estimated in g/km; REID vapor pressure of gasoline 80 kPa (from October to March) and 65 kPa (from April to September) with monthly tem- peratures given in note 3. * For cold-starts and evaporation losses, average trip length equals 12 km. Source: Samaras 1992 Selected Exbaust Emission and Fuel Consumption Factorsfor Gasoline-Fueled Vebicles 51 Table A2.1.4 Esttmated Emissions and Fuel Consumption, European Vehicles, Rural Driving (grams per kilometer) Carbon Non-methane Nitr gen Nitrnoxv Fuel monoxide volatile org. compounds oxides Metbane oxide Ammnioa lvap. consumption Vebicle type (CO) (NM-VOC) (NOx) (CH4) (N20) (NH.) emissions (liter/lOG km) Passenger Cars Two-stroke 7.5 7.24 1.0 0.040 0.005 0.002 0.14 8.77 LPG 3.1 0.6 2.6 0.035 0.000 0.000 - 6.02 Pre-1971 (Pre ECE) Less than 1,400cc 18.5 1.53 2.1 0.025 0.005 0.002 0.14 7.35 1,400-2,000 cc 18.5 1.53 2.8 0.025 0.005 0.002 0.14 8.93 More than 2,000cc 18.5 1.53 4.3 0.025 0.005 0.002 0.14 10.64 1972-76 (ECE 15100-01) Less than 1,400cc 14.8 1.23 2.1 0.025 0.005 0.002 0.14 5.88 1,400-2,000cc 14.8 1.23 2.8 0.025 0.005 0.002 0.14 6.94 More than 2,000 cc 14.8 1.23 4.3 0.025 0.005 0.002 0.14 7.75 1977-79 (ECE 15/02) Less than 1,400cc 7.9 1.03 2.2 0.025 0.005 0.002 0.14 6.02 1,400-2,000 cc 7.9 1.03 2.5 0.025 0.005 0.002 0.14 6.85 More than 2,000cc 7.9 1.03 2.8 0.025 0.005 0.002 0.14 8.47 1980-84 (ECE 15/03) Less than 1,400cc 8.3 1.03 2.3 0.025 0.005 0.002 0.14 6.02 1,400-2,000 cc 8.3 1.03 2.8 0.025 0.005 0.002 0.14 6.45 More than 2,000cc 8.3 1.03 3.4 0.025 0.005 0.002 0.14 8.00 1985-89 (ECE 15/04) Less than 1,400cc 4.7 0.83 2.2 0.025 0.005 0.002 0.14 5.78 1,400-2,000 cc 4.7 0.83 2.7 0.025 0.005 0.002 0.14 6.45 More than 2,000cc 4.7 0.83 2.9 0.025 0.005 0.002 0.14 8.00 Improved conventional Less than 1,400cc 6.5 0.73 2.2 0.025 0.005 0.002 0.14 6.37 1,400-2,000cc 2.3 0.63 2.7 0.025 0.005 0.002 0.14 7.25 Oxidizing catalyst without I sensor Less than 1,400cc 5.5 0.53 1.7 0.025 0.005 0.002 0.14 6.58 1,400-2,000 cc 3.6 0.13 1.5 0.025 0.005 0.002 0.14 7.52 3-way catalyst with I sensor Less than 1,400cc 1.4 0.12 0.3 0.020 0.050 0.100 0.03 6.25 1,400-2,000cc 1.4 n.a. 0.3 n.a. 0.050 0.100 0.03 7.30 More than 2,000cc 1.4 0.12 0.3 0.020 0.050 0.100 0.03 8.85 Light-duty vehicles 15.0 1.7 2.7 0.040 0.006 0.002 0.14 9.03 Motorcydes Less than 50cm3 10.0 6.00 0.1 0.100 0.001 0.001 0.00 2.40 More than 50cm3, two-stroke 22.0 15.05 0.1 0.150 0.002 0.002 0.06 4.00 More than 50 cm3, four-stroke 20.0 3.00 0.3 0.200 0.002 0.002 0.06 5.08 - Not applicable n.a. = Not available Notes * Average driving speed, 75 kilometers per hour. * Emission factors In g/km are derived from the COPERT model for 1990, utilizing the CORINAIR methodology for road trAffic emissions. The pollutants Included are: CO, NMVOC, NO., CH4, N20, NH3. Puel consumption Is also estimated. Total VOC (or HC) emission factors may be obtained by adding NMVOC and CH4 factors. . Cold-start emissions are calculated for urban drIving conditions onlytaking into account the monthly variation of the average minimum and maximum temperatures. For this specific application,aveage temperatures In Brussels (Belgium) were used.with the following monthly distribution: Month 01 02 03 04 05 06 07 08 09 10 11 12 Tmin (OC) -1.2 0.3 2.2 5.1 7.9 10.9 12.1 12.2 10.6 7.3 3.1 0.2 Tmax (IC) 4.3 6.7 10.3 14.2 18.4 22.0 22.7 22.3 20.5 15.4 8.9 5.6 * Evaporative losses are estimated In g/km; REID vapor pressure of gasoline 80 kPa (from October to March) and 65 kPa (from April to September) with monthly tem- pertures given In note 3. * For cold-starts and evaporation osses, average trip length equals 12 km. Source: Samaras 1992 52 AirPoilutionfromMotorVebicles Table A2.1.5 Estimated Emissions and Fuel Consumption, European Vehicles, Highiway Driving (grams per kilometer) Carbon Non-methane Nitr(gen Nitrous Fuel monoxide volatile org. compounds oxides Methane oxide Antmoia Evap. consunption Vehicle type (CO) (NM-VOC) (NO,) (CH4) (N20) (NH3) emissions (liier/JOO km) Passenger Cars Two-stroke 8.7 0.93 0.7 0.025 0.005 0.002 0.14 7.58 LPG 9.8 0.5 2.9 0.025 0.000 0.000 - 7.23 Pre-1971 (Pre-ECE) Less than 1,400cc 15.5 1.23 2.0 0.026 0.005 0.002 0.14 8.33 1,400-2,000cc 15.5 1.23 3.1 0.026 0.005 0.002 0.14 10.20 More than 2,000 cc 15.5 1.23 5.5 0.026 0.005 0.002 0.14 11.76 1972-76 (ECE 15/00-0]) Less than 1,400 cc 18.6 1.13 2.0 0.026 0.005 0.002 0.14 6.49 1,400-2,000 cc 18.6 1.13 3.1 0.026 0.005 0.002 0.14 8.06 More than 2,000 cc 18.6 1.13 5.5 0.026 0.005 0.002 0.14 8.85 1977-79 (ECE I5/02) Less than 1,400 cc 8.3 0.93 2.9 0.026 0.005 0.002 0.14 6.85 1,400-2,000 cc 8.3 0.93 3.3 0.026 0.005 0.002 0.14 7.94 More than 2,000 cc 8.3 0.93 3.7 0.026 0.005 0.002 0.14 9.43 1980-84 (ECE 15/03) Less than 1,400cc 7.9 0.93 3.3 0.026 0.005 0.002 0.14 6.85 1,400-2,000 cc 7.9 0.93 3.8 0.026 0.005 0.002 0.14 7.94 More than 2,000 cc 7.9 0.93 4.5 0.026 0.005 0.002 0.14 9.43 1985-89 (ECE 15104) Less than 1,400cc 4.3 0.73 2.7 0.026 0.005 0.002 0.14 6.37 1,400-2,000cc 4.3 0.73 3.5 0.026 0.005 0.002 0.14 6.98 More than 2,000 cc 4.3 0.73 3.7 0.026 0.005 0.002 0.14 9.35 Improved conventional Less than 1,400cc 10.5 0.83 2.4 0.026 0.005 0.002 0.14 9.26 1,400-2,000cc 6.7 0.73 3.7 0.026 0.005 0.002 0.14 10.42 Oxidizing catalyst without A sensor Less than 1,400cc 8.4 0.53 1.9 0.026 0.005 0.002 0.14 9.01 1,400-2,000 cc 6.7 0.23 1.6 0.026 0.005 0.002 0.14 10.87 3-way catalyst with A sensor Less than 1,400cc 3.1 0.12 0.5 0.020 0.050 0.001 0.03 8.70 1,400-2,000 cc 3.1 0.12 0.5 0.020 0.050 0.001 0.03 10.20 More than 2,000cc 3.1 0.12 0.5 0.020 0.050 0.001 0.03 12.99 Light-duty vehicles 12.0 1.0 3.2 0.025 0.002 0.006 0.14 8.54 Motorcycles Lessthan50cm3 10.0 6.00 0.1 0.100 0.001 0.001 0.00 2.40 More than 50 cm3, two-stroke 22.0 15.05 0.1 0.150 0.002 0.002 0.06 4.00 More than 50 cm3, four-stroke 20.0 3.00 0.3 0.200 0.002 0.002 0.06 5.08 - Not applicable Notes * Average driving speed, 100 kilometers per hour. * Emission factors in g/km are derived from the COPERT model for 1990, utilizing the CORINAIR methodology for road traffic emissions. The pollutants included are: CO, NMVOC, NO., CH4, N20, NH3. Fuel consumption is also estimated. Total VOC (or HC) emission factors may be obtained by summing up NMVCC and CH4. * Cold-start emissions are calculated for urban driving conditions only taking into account the monthly variation of the average minimum and maximum temperatures. For this specific application, average temperatures in Brussels (Belgium) were used, wvith the following monthly distribution: Month 01 02 03 04 05 06 07 08 09 10 11 12 Tmin (IC) -1.2 0.3 2.2 5.1 7.9 10.9 12.1 12.2 10.6 7.3 3.1 0.2 Tmax (OC) 4.3 6.7 10.3 14.2 18.4 22.0 22.7 22.3 20.5 15.4 8.9 5.6 * Evaporative lossesare estimated in g/km;RElD vaporpressure of gasoline from 80 kPa (from OctobertolMarch) and 65 kPa (fromApril to September) with monthly temperatures given in note 3. * For cold-starts and evaporation losses, average trip length equals 12 km. Source: Samaras 1992 Selected Exbaust Emission and Fuel Consumption Factorsfor Gasoline-Fueled Vebicles 53 Table A2.1.6 Automobile Exhaust Emissions, Chile (grams per kilometer) Vebide type Carbon monoxide Hydrocarbons Nitrogen aoides Particulate matter Private car 26.0 1.00 1.2 0.07 Taxi 28.0 1.50 1.4 0.06 Note: Measured using the U.S. federal testing procedure. Source: Escudero 1991 Table A2.1.7 Automobile Exhaust Emissions as a Function of Test Procedure and Ambient Temperature, Finland (grams per kilometer) 53 5 Hydrocarbons Nitrogen oxides ECE-15 22°C 2.60 0.27 0.27 .70 11.22 1.09 0.65 -20°C 17.81 2.79 0.62 FTP-75 22°C 1.40 0.13 0.16 -7°C 5.34 0.50 0.29 -20°C 8.58 1.25 0.31 Note: Based on emission tests on cars with a three-way catalytic converter. Source: Laurilkko 1995 Table A2.1.8 Automobile Exhaust Emissions as a Function of Driving Conditions, France (grams per kilometer) Traffic type Carbon monoxide Hydrocarbons Nitrogen oxides Carbon dioxide Congested urban 94.1 10.70 1.6 520.73 Free-flowing urban 29.3 3.52 1.8 189.54 Highway 19.4 2.45 2.2 149.05 Motorway 16.0 1.09 3.0 153.48 Source: Joumard and others 1990 Table A2.1.9 Automobile Exhaust Emissions and Fuel Consumption as a Function of Driving Conditions and Emission Controls, Germany (grams per kilometer) Carbon Nitrogen Carbon Fuel consumption (Ii- Vehicle type monoxide Hydrocarbons oxides diaoide ters/lOOkm) European test procedure (low-speed urban driving with cold start) Catalyst with 02 sensor 6.27 0.81 0.59 274 11.90 Catalyst without 02 sensor 15.34 1.93 0.94 234 9.80 No catalyst 17.67 2.62 1.29 243 10.50 U.S. federal test procedure (mixed urban driving) Catalyst with 02 sensor 3.02 0.27 0.39 204 8.90 CatalystwithoutO2 sensor 11.76 1.35 0.88 180 7.80 No catalyst 12.14 1.84 1.63 182 7.90 Rural highway Catalyst with 02 sensor 0.98 0.06 0.28 136 5.90 CatalystwithoutO2sensor 3.98 0.42 1.15 130 5.60 No catalyst 4.89 0.76 1.94 126 5.50 Motorway Catalyst with O2 sensor 5.13 0.14 0.75 193 8.30 Catalyst without 02 sensor 12.72 0.60 1.74 189 8.20 No catalyst 12.78 0.94 3.18 178 7.70 Source: Hassel and Weber 1993 54 Air PoUutionfrom Motor Vebices Table A2.1.10 Exhaust Emissions, Light-Duty Vehicles and Mopeds, Greece (grams per kilometer) Vebide type Carbon monoxide Hydrocarbons Nitrogen axides Light-duty 45.8 1.6 1.60 Motorcycle 21.4 3.4 0.11 Moped 14.0 10.4 0.05 Note: Mcasured using the ECE-15 testing procedure. Source: Pattas, Kyriakis, and Nakos 1993 Table A2.1.11 Hot-Start Exhaust Emissions, Light-Duty Vehicles, Greece (grams per kilometer) Engine capacity/Emission standanl Carbon monoxdde Hydrocarbons Nitrogen oxides Less than 1,400 cc Pre-control (1971) 60.02 5.15 1.23 ECE 15-00 (1971-75) 58.17 4.80 1.42 ECE 15-02 (1975-79) 44.37 3.77 1.58 ECE 15-03 (1980-84) 31.33 3.60 1.81 ECE 15-04 (1985-present) 25.50 2.08 2.08 1,400-2,000 cc Pre-control 73.61 4.87 1.06 ECE 15-00 54.51 4.67 1.82 ECE 15-02 51.20 3.65 1.75 ECE 15-03 35.94 3.27 1.95 ECE 15-04 27.99 2.11 2.06 More than 2,000 cc Pre-control 77.12 5.60 1.40 ECE 15-00 91.97 5.85 0.77 ECE 15-02 21.28 1.62 2.28 ECE 15-03 81.50 3.32 1.02 ECE 15-04 30.18 2.14 2.04 Note: Measured using the ECE-15 testing procedure, average speed 18.7 kilometers per hour. Source: Pattas, Kyriakis, and Nakos 1993 Table A2.1.12 Exhaust Emissions, Light-Duty Vehicles and 2-3 Wheelers, India (grams per kilometer) Carbon Hydro- Nitrogen Sulfur Vebicle type monoxide carbons oxides dioxide Car/jeep 23.8 3.5 1.6 0.1 Taxi 29.1 4.3 1.9 0.1 Two-wheeler 8.2 5.1 - - Auto-rickshaw (3-wheeler) 12.5 7.8 - 0.0 Light-duty vehicles 40.0 6.0 3.2 0.08 Motorcycles 17.0 10.0 0.07 0.02 - Not applicable n.a. = Not available Sources: Biswas and Dutta 1994; Bose 1994; Gargava and Aggarwal 1994 Selected Exhaust Emission and Fuel Consumption Factorsjbr Gasoline-Fueled Vebicles 55 References ants des Vehicules LUgers. Report 116 (2nd ed.) Insti- tut National du Recherche sur les Transports et leur Biswas, D. and S.A. Dutta. 1994. "Strategies for Control Securite (INRETS), Bron, France. of Vehicular Pollution in Urban Areas of India." Pro- Joumard, R., P. Jost, J. Hickman, and D. Hassel. 1995. ceedings of a Workshop on The Energy Nexus-In- "Hot Passenger Car Emissions Modelling as a Func- dian Issues & Global Impacts (April 22-23,1994), tion of Instantaneous Speed and Acceleration." The University of Pennsylvania, Philadelphia. Science of the Total Environment, 169:167-74. Escudero,J. 1991. "Notes on Air Pollution Issues in San- Laurikko,J. 1995."Ambient Temperature Effect on Auto- tiago Metropolitan Region", Letter # 910467, dated motive Exhaust Emissions: FTP and ECE Test Cycle May 14, 1991. Comisi6n Especial de Descontami- Responses." The Science of the Total Environment, naci6n de la Region Metropolitana. Santiago, Chile. 169:195-204. Gargava, P., and A. Aggarwal. 1994. "Prediction of Im- Metz, N. 1993. "Emission Characteristics of Different pact of Air Environment on Planning the Control Combustion Engines in City, on Rural Roads, and on Strategies for Automobile Pollution in an Indian Highways." The Science of the Total Environment, Coastal City." Poster Proceedings, 3rd International 134:225-35. International Symposium on Transport and Air Pollu- Pattas, K., N. Kyriakis, and C. Nakos. 1993. "Time De- tion, INRETS,Arcueil, France. pendence of Traffic Emissions in the Urban Area of Hassel, D. and F-J. Weber. 1993. "Mean Emissions and Thessaloniki."The Science of the Total Environment Fuel Consumption of Vehicles in Use with Different 134:273-84. Emission Reduction Concepts." The Science of the Samaras, Z. 1992."COPERT Emission Factors'" Informnal Total Environment 134: 189-95. Communication. Directorate-General for Environ- Joumard, R., L. Paturel, R. Vidon, J. P. Guitton, A. Saber, ment, Nuclear Safety, and Civil Protection, Commis- and E. Combet. 1990. Emissions Unitaires de Pollu- sion of the European Communities, Brussels. Appendix 2.2 Selected Exhaust Emission and Fuel Consumption Factors for Diesel-Fueled Vehicles Table A2.2.1 Exhaust Emissions, European Cars (grams per kilometer) Traffic type/emission control Carbon monoxide Hydrocarbons Nitrogen oxides Particulate matter Urban with catalyst 0.05 0.08 0.70 0.20 without catalyst 1.30 0.10 0.90 0.30 Rural with catalyst 0.02 0.10 0.61 0.18 without catalyst 0.60 0.10 0.79 0.29 Highway with catalyst 0.20 0.10 0.77 0.27 without catalyst 0.70 0.10 0.97 0.37 Source: Metz 1993 Table A2.2.2 Estimated Emissions and Fuel Consumption, European Cars and Ilght-Duty Vehicles (grams per kilometer) Particulate Fuel consumption Traffic and vebicle type Carbon monoxide Hydrocarbons Nitrogen oxides matter (liters/Il Okm) Urban Passenger cars Less than 2,000 cc 1.0 0.306 0.7 0.362 10.0 More than 2,000 cc 1.0 0.306 1.0 0.362 10.0 Light-duty vehicles 2.4 0.506 1.7 0.333 14.08 Rural Passenger cars Less than 2,000 cc 0.5 0.105 0.4 0.131 5.05 More than 2,000 cc 0.5 0.105 0.7 0.131 5.05 Light-dutyvehicles 0.8 0.205 1.2 0.131 8.40 Motorway Passenger cars Less than 2,000 cc 0.4 0.105 0.5 0.170 6.17 More than 2,000 cc 0.4 0.105 0.9 0.170 6.17 Light-dutyvehicles 0.6 0.105 1.3 0.160 7.87 Notes: * Average driving speeds for urban, rural and motorway are 25 km/hour, 75 kn/hour and 100 kn/hour, respectively. * Emission factors in glkm are derived from the COPERT model for 1990, utilizing the CORINAIR methodology for road traffic emissions. The pollutants included are: CO, HC, NOJ,TPM. Fuel consumption is also estimated. Source: Samaras 1992 57 58 Air Pollutionfrom Motor Vebicles Table A2.2.3 Estimated Emissions, European Medium- to Heavy-Duty Vehicles (grams per kilometer) Fuel Carbon Nitrogen Particulate consumption Vehicle type monoxide Hydrocarbons oxides matter CH4 N20 NH3 (liters/lOOkn) Urban 3.5-16.0 tons 18.8 2.79 8.7 0.95 0.085 0.030 0.003 27.03 More than 16.0 tons 18.8 5.78 16.2 1.60 0.175 0.030 0.003 43.48 Rural 3.5-16.0 tons 7.3 0.76 7.4 0.82 0.010 0.030 0.003 22.22 More than 16.0 tons 7.3 2.58 14.8 1.40 0.080 0.030 0.003 38.46 Motorway 3.5-16.0tons 4.2 0.62 6.0 1.67 0.020 0.030 0.003 18.18 More than 16.0 tons 4.2 2.27 13.5 1.25 0.070 0.030 0.003 34.48 Notes: * Average driving speed for urban: 25 km/h; rural: 75 km/h; and highway: 100 km/h. * Emission factors in g/km are derived from the COPERT model for 1990, utilizing the CORINAIR methodology for road traffic emissions. The pollutants included are: CO, NO.,TPM. Fuel consumption is also estimated. Source: Samaras 1992 Table A2.2.4 Exhaust Emissions, European Heavy-Duty Vehicles (grams per kilometer) Engine type and vehicle loading Carbon monoxide Hydrocarbons Nitrogen oxides Particulate matter Natural aspiration,3.5-16.0 tons 3.41 0.61 6.58 0.55 Turbo-charged 3.5-16.0 tons 2.00 0.57 13.07 0.37 16.0-38.0 tons 4.21 1.06 26.90 0.71 Turbo-charged with inter-cooling, 16.0-38.0 tons 5.37 1.00 16.90 0.61 Source: Sawer 1986 Table A2.2.5 Exhaust Emissions and Fuel Consumption, Utility and Heavy-Duty Trucks, France (grams per kilometer) Avg. Speed Fuel Consmp. Carbon Nitrogen Particulate Vehicle type km/h km/l monoxide Hydrocarbons oxides matter Empty Utiliq truck (3.5t), IDI 76.5 11.0 1.0 0.6 1.6 0.5 123.7 6.1 1.7 1.6 1.9 2.0 Heavy-duty truck (19t), DI 68.9 4.3 2.6 0.8 15.5 0.5 88.4 3.9 2.8 0.7 12.0 0.4 Serni-trailer (40t), DI 69.2 4.0 1.9 1.1 6.7 0.9 88.0 3.7 1.7 1.0 7.4 0.9 Loaded Utility truck (3.5t), IDI 74.2 9.3 1.2 0.9 1.6 0.9 117.7 5.8 1.7 1.8 1.9 2.3 Heavy-dutytruck(19t),DI 66.8 3.5 3.1 0.8 16.4 0.5 84.7 3.4 3.8 0.7 13.4 0.5 Semi-trailer(40t),DI 62.2 2.3 3.2 1.1 10.7 1.4 75.6 2.4 3.0 1.0 10.1 1.3 Source: Roumegoux 1995 Selected Exbaust Emission and Fuel Consumption Factorsfor Diesel-Fueled Vebicles 59 Table A2.2.6 Exhaust Emissions, Santiago Buses, Chile (grams per kilometer) Testing procedure Carbon monoxide Hydrocarbons Nitrogen ocides Particulate matter Santiago cycle (CADEBUS) 5.70 1.40 5.40 2.50 Source: Escudero 1991 Table A2.2.7 Exhaust Emissions, London Buses, United Kingdom (grams per kilometer) Testingprocedure Carbon monoxide Hydrocarbons Nitrogen oxides Particulate matter London Bus Limited In-service simulation Laden 7.09 1.19 28.89 1.69 Unladen 6.61 0.92 32.37 1.47 Test cycle Laden 5.64 0.67 22.50 1.36 Unladen 5.64 0.62 14.07 0.58 Source: Gore 1991 Table A2.2.8 Exhaust Emissions, Utility and Heavy-Duty Vehicles, Netherlands (grams per kilometer) Testing procedure and loading Carbon monoxide Hydrocarbons Nitrogen oxides Particulate matter Urban Less than 3.5 tons 3.0 1.3 1.3 1.2 3.5-5.5 tons 4.0 2.0 6.0 1.5 5.5-12.0 tons 10.0 7.0 10.0 3.5 12.0-15.0 tons 13.0 9.0 13.0 5.0 More than 15.0 tons 16.0 12.0 20.0 7.0 Rural Less than 3.5 tons 1.5 0.7 1.3 0.6 3.5-5.5 tons 2.0 1.0 6.0 1.0 5.5-12.0 tons 4.0 2.5 10.0 2.0 12.0-15.0 tons 4.5 3.0 13.0 2.5 More than 15.0 tons 5.0 3.5 20.0 3.0 Highway Less than 3.5 tons 0.9 0.5 1.4 0.5 3.5-5.5 tons 1.0 0.8 7.0 0.9 5.5-12.0 tons 1.5 2.0 13.0 1.8 12.0-15.0 tons 2.0 2.3 15.0 2.0 More than 15.0 tons 2.0 2.5 25.0 2.5 Source: Veldt 1986 Table A2.2.9 Automobile Exhaust Emissions as a Function of Driving Conditions, France (grams per kilometer) Traffic type Carbon monoxide Hydrocarbons Nitrogen aorides Particulate matter Carbon dioxide Congested urban 3.29 1.04 2.70 0.68 588 Free-flowing urban 1.05 0.29 0.76 0.29 225 Highway 0.61 0.16 0.57 0.19 179 Motorway 0.61 0.09 0.56 0.25 166 Source: Joumard and others 1990 60 Ar Pollutionfrom Motor Vebicles Table A2.2.10 Automobile Exhaust Emissions and Fuel Consumption as a Function of Testing Procedures, Germany (grams per kilometer) Carbon Nitrogen Particulate Carbon Fuel consumption Testing procedure monoxide Hydrocarbons oxides matter dioxide Cliters/lOOkm) European (ECE-15) 1.00 0.17 0.91 0.115 215 38.30 With cold start (ETK) 1.00 0.17 0.19 0.115 215 38.30 Extra urban driving cycle (EUDC) 0.27 0.05 0.55 0.081 128 4.90 ETK + EUDC 0.54 0.10 0.68 0.093 160 6.70 U.S. federal (FTP-75) 0.61 0.10 0.70 0.092 166 6.40 Highway 0.25 0.04 0.48 0.052 115 4.40 Motorway 0.33 0.05 0.83 0.119 179 6.90 Source: Hassel and Weber 1993 Table A2.2.11 Exhaust Emissions, Cars, Buses, and Trucks, Greece (grams per kilometer) Vehicle type Carbon monoxide Hydrocarbons Nitrogen oxides Passenger car 1.34 1.81 0.69 Urban bus 21.16 5.57 10.40 Other buses 4.95 2.15 5.94 Trucks 6.19 2.68 7.43 Source: Pattas, Kyriakis, and Nakos 1993 Table A2.2.12 Exhaust Emissions, Light-Duty Vehicles and Trucks, India (grams per kilometer) Vehicle type Carbon monoxide Hydrocarbons Nitrogen oxides Sulfur dioxide Particulate matter Eight-duty vehicles 1.1 0.28 0.99 0.39 2.0 Heavy-dutytruck 12.70 2.10 21.0 1.50 3.0 Source: Biswas and Dutta 1994; Gargava and Aggarwal 1994 Selected Exhaust Emission and Fuel Consumption Factorsfor Diesel-Fueled VehIcles 61 References ants des V6hicules L0gers. Report 116 (2nd ed.) Insti- tut National du Recherche sur les Transports et leur Biswas, D. and S.A. Dutta. 1994. "Strategies for Control Securite (INRETS), Bron, France. of Vehicular Pollution in Urban Areas of India," Pro- Metz, N. 1993. "Emission Characteristics of Different ceedings of a the Workshop onThe Energy Nexus- Combustion Engines in City, on Rural Roads, and on Indian Issues & Global Impacts (April 22-23, Highways." The Science of the Total Environment 1994), University of Pennsylvania, Philadelphia. 134:225-35. Escudero,J. 1991."Notes on Air Pollution Issues in San- Pattas, K., N. Kyriakis, and C. Nakos. 1993. 'Time De- tiago Metropolitan Region", Letter# 910467, dated pendence of Traffic Emissions in the Urban Area of May 14, 1991, Comisi6n Especial de Decontami- Thessaloniki." The Science of the Total Environment naci6n de la Region Metropolitana, Santiago. 134: 273-84. Gargava, P., and A. Aggarwal. 1994. "Prediction of Im- Roumegoux, J.P 1995. "Calcul des Emissions Unitaires pact of Air Environment on Planning the Control de Polluants des V6hicules Utilitaires,"The Science of Strategies for Automobile Pollution in an Indian the Total Environment, 169:273-82. Coastal City." 3rd International Symposium on Trans- Samaras, Z. 1992."COPERT Emission Factors." Informal port and Air Pollution, Poster Proceedings, INRETS, Communication. Directorate-General for Environ- Arcueil, France. ment, Nuclear Safety, and Civil Protection, Commis- Gore, B.M. 1991 (draft). "Vehicle Exhaust Emission sion of the European Communities, Brussels. Evaluations." Office of the Group of Engineers, Lon- Sawer,J.M. 1986.A Review of Diesel Engine Emissions don Buses Limited, London. in Europe," Report DP 86/1946, Ricardo Consulting Hassel, D. and F-J. Weber. 1993. "Mean Emissions and Engineers, Shoreham-by-Sea, England. Fuel Consumption of Vehicles in Use with Different Veldt, C. 1986. "Emissions from Road Transport", Dis- Emission Reduction Concepts." The Science of the cussion Paper for the OECD Workshop on Compari- Total Environment, 134:189-95. son of Emission Inventory Data, MT-TNO, Joumard, R., L. Paturel, R. Vidon, J. P Guitton, A. Saber, Apeldoorn, Netherlands (October 22-24). and E. Combet. 1990. Emissions Unitaires de Pollu- 3 Vehicle Technology for Controlling Emissions The principal pollutant emissions from vehicles truck.These emissions can be controlled by substituting equipped with spark-ignition gasoline engines include a four-stroke engine or an advanced two-stroke design unburned hydrocarbons, carbon monoxide, and nitro- that uses fuel injection, at a cost of about U.S.$60 to gen oxides in the exhaust. Emissions of respirable par- U.S.$80 per vehicle.This change also reduces fuel con- ticulate matter (PM) can also be considerable, sumption by 30 to 40 percent. Further control of motor- particularly from two-stroke engines. Lead aerosol emis- cycle and three-wheeler emissions can be achieved with sions from combustion of leaded gasoline are also signif- catalytic converters. icant and have important impacts on public health. The most significant emissions from diesel-fueled Evaporation of gasoline in the fuel system, the escape of vehicles are particulate matter, nitrogen oxides, and hy- gasoline vapors during refueling, and the escape of drocarbons. Particulate matter emissions from uncon- blowby losses from the crankcase contribute additional trolled diesel engines are six to ten times those from hydrocarbon emissions. gasoline engines. Diesel smoke is also a visible public In new automobiles, carbon monoxide, hydrocar- nuisance. Emissions of other pollutants from diesel en- bon, and nitrogen oxide emissions can be reduced by gines are generally lower than those for comparable gas- 50 percent or more from uncontrolled levels through oline engines. Compared with similar vehicles with engine modifications, at a cost of about U.S.$130 per uncontrolled gasoline engines, light-duty diesel vehicles car. Fuel consumption may increase slightly. Hydrocar- without emission controls emit about 90 percent less bon and carbon monoxide reductions of 90 to 95 per- hydrocarbons and carbon monoxide and about 50 to 70 cent and nitrogen oxide reductions of 80 to 90 percent percent less nitrogen oxides. Heavy-duty diesel vehicles are possible with three-way catalysts and electronic en- emit 50 to 100 percent more nitrogen oxides than their gine control systems that cost about U.S.$600 to gasoline counterparts, but 90 to 95 percent less hydro- U.S.$800 per car. Such devices have little impact on fuel carbons and 98 percent less carbon monoxide. Both economy. Lean-burn techniques combined with an oxi- light- and heavy-duty diesel- fueled vehicles are consid- dation catalytic converter can achieve comparable erably more fuel efficient than their gasoline counter- hydrocarbon and carbon monoxide reductions, a 60 to parts (15 to 40 percent for light-duty diesels, as much as 75 percent reduction in nitrogen oxides, and a 10 to 15 100 percent for heavy-duty ones) and therefore emit percent improvement in fuel economy. less carbon dioxide. Two-stroke gasoline engines-used in motorcycles Diesel engine emissions of nitrogen oxides and and three-wheelers, predominantly in Asia and Europe, hydrocarbons can be reduced by more than 50 percent and formerly in some automobiles in Eastern Europe- and emissions of particulate matter by more than 75 are a special case. Hydrocarbon emissions from two- percent from uncontrolled levels through engine stroke engines are high because a significant part of the design changes, improved fuel injection systems, turbo- air-fuel mixture escapes unburned into the exhaust. Par- charging, and charge air cooling. These changes ticulate emissions from two-strokes are also excessive improve fuel economy (diesel fuel-efficiency has im- because oil is mixed with the fuel, and recondenses into proved by 30 percent since the 1 970s) but increase oil particles in the exhaust. Hydrocarbon emissions from engine costs. Particulate matter and hydrocarbon a single two-stroke motorcycle can exceed those from emissions can be further reduced through the use of three uncontrolled passenger cars and particulate mat- low-sulfur fuel and an oxidation catalytic converter.The ter emissions can exceed those from a heavy-duty diesel lowest particulate matter emissions (a 95 percent reduc- 63 64 Air Pollution from Motor Vehicles tion from uncontrolled levels) are possible with the use bon emissions occur in the vehide exhaust, the engine of trap-oxidizers, but the reliability of these systems has crankcase, the fuel system, and from atmospheric vent- not been proven conclusively. ing of vapors during fuel distribution and dispensing. Particulate matter emissions from gasoline engines are caused by the condensation of oil vapor in the ex- Automotive Engine lTypes haust. These particulate matter emissions are usually Pollutant emiissions from motor vehices are determined small for four-stroke engines. Two-stroke engines and by the vehicle's engine type and the fuel it uses. Spark-ig- four- strokes with excess oil consumption can exhibit nition and diesel engines are the two most common en- high particulate matter emissions. gines. Engine technology, emission characteristics, and The use of lead as an antiknock additive in gasoline emission control technologies for these two basic engine is being discontinued in many countries for environ- types are discussed in detail in appendices 3.1 and 3.2. mental reasons.Where these compounds are still in use, Measures to improve fuel economy and directly reduce lead aerosol emissions from gasoline engines are the emissions of carbon dioxide are presented in appendix major source of airborne lead in the environment.A re- 3.3. Other engine technologies and advanced vehide pro- view of lead additives in gasoline is presented in an ap- pulsion systems have been reviewed by Watkins (1991), pendix to chapter 5. Brogan and Venkateswaran (1993), Kimbom (1993), Ma- son (1993), MacKenzie (1994), and OTA (1995). Diesel Engines Most heavy-duty trucks and buses have diesel engines, Spark-Ignition (Otto) Engines as do some light-duty vehicles and passenger cars. Die- Most passenger cars and light-duty trucks use spark-igni- sel engines in light-duty vehicles are common in Eu- tion (Otto cycle) gasoline engines. These engines are also rope, (about 20 percent of the light-duty fleet) and in used in heavy-duty trucks and buses in some countries, parts of southeast Asia. Diesel engines, unlike spark-ig- including China, Mexico, the Russian Federation and oth- nition engines, do not premix fuel with air before it en- er republics of the former Soviet Union, and the United ters the cylinder. Instead, the fuel is injected at high States. Other fuels used in spark-ignition engines indude pressure near the top of the compression stroke. Once natural gas, liquified petroleum gas, alcohols, and hydro- injected, the fuel is heated to ignition by the com- gen (see chapter 5). pressed air in the cylinder, eliminating the need for a Spark-ignition gasoline engines have either a two- separate spark-ignition system. stroke or four-stroke design. Two-stroke engines are There are two types of diesel engine: indirect and di- cheaper, lighter, and can produce greater power output rect injection. In an indirect injection (IDI) diesel en- per unit of displacement, so they are widely used in gine the fuel is injected into a pre-chamber where small motorcycles, outboard motors, and small power ignition occurs and combustion then spreads to the equipment.Two-stroke engines emit 20 to 50 percent of main combustion chamber. Indirect injection technolo- their fuel unburned in the exhaust, resulting in high gy is mainly used for small, high-speed applications emissions and poor fuel economy. Because the crank- such as passenger cars, where low noise and high per- case pumps the air-fuel mixture through the engine, formance are important. two-stroke engines require that oil be mixed with the In a direct injection (DI) engine the fuel is sprayed air-fuel mixture to lubricate bearings and pistons. Some directly into and ignited in the combustion chamber. of this oil appears as white smoke in the exhaust, result- These engines are generally used in medium and large ing in high emissions of particulate matter. trucks and give higher power output and better fuel All gasoline engines currently used in automobiles economy but they are considerably noisier. Develop- and larger vehicles use the four-stroke design, although ments in reducing noise and improving performance advanced two-stroke engines are being developed. have led to the use of these engines in passenger cars, These advances pertain to fuel injection, combustion, although there is considerably less experience with and the lubrication system. Advanced two-stroke en- these engines in small applications (Holman 1990). gines under development would achieve lower emis- Compared with gasoline spark-ignition engines, sions and fuel consumption than four-stroke engines heavy-duty diesel engines have lower carbon monoxide and retain the two-stroke's advantages of lower weight and hydrocarbon emissions but higher nitrogen oxide and cost per unit of power output. emissions. They are up to 100 percent more fuel-effi- The main pollutant emissions from spark-ignition gas- cient, resulting in lower emissions of carbon dioxide. oline engines are hydrocarbons, carbon monoxide, and Light-duty diesels exhibit better fuel efficiency and low- nitrogen oxides. Carbon monoxide and nitrogen oxides er carbon monoxide, hydrocarbon, and nitrogen oxide are only emitted in the vehicle exhaust, while hydrocar- emissions than their gasoline counterparts. Vehicle Technologyfor Controlling Emissions 65 Particulate matter emissions from diesel engines are engines. Some research has been done on closed-cir- considerably higher than from gasoline engines. Diesel cuit, low-temperature Rankine engines as a bottoming emissions-in the form of black smoke-are a major cycle for internal combustion engines in heavy-duty source of high ambient concentrations of particulate trucks.These engines would use the wasted heat in the matter in most large cities of the developing world. 0th- truck's exhaust to produce additional power, thereby er pollutant emissions from diesel vehicles include sulfur increasing overall efficiency. dioxide and noise. Particulate matter and noise emissions result from the combustion process.These emissions can Stirling Engines be reduced by modifying the engine and combustion sys- Stirling engines have been of interest for many years.They tem (appendix 3.1). Diesel engines meeting current U.S. are theoretically capable of achieving high fuel efficiency, and future European emission standards are smokeless and have demonstrated low emission levels. Currently when properly maintained, have better fuel efficiency, available Stirling engines are not practical for automotive are less noisy, and emit less nitrogen oxides and hydrocar- use, however, because of their high cost, poor transient bons than the uncontrolled diesel engines sold in devel- response, and poor power-to-weight and power-to-vol- oping countries. ume ratios. Rotary (Wankel) Engines Electric and Hybrid Vebicles A rotary engine utilizes a triangular rotor which turns Electric vehicles have been pursued because of their within an elliptical combustion chamber.The motion of mechanical simplicity and the absence of direct pollut- the triangular rotor varies the volume of the space be- ant emissions, although emissions from the power tween the rotor and the chamber wall and performs the source should be taken into account. The potential to compression and expansion functions of a piston in a recover kinetic energy during braking can contribute to conventional engine. Rotary engines are smaller, lighter increased fuel efficiency. Electric vehicles are used in and simpler than reciprocating piston engines. Carbon specialized applications, but current battery technolo- monoxide and hydrocarbon emissions are significantly gy is inadequate for electric vehicles to compete with higher from rotary engines as compared to conventional internal combustion vehicles in most applications. engines; emissions of nitrogen oxides are about the Although improved batteries are being researched, a same. Production models of passenger cars and motorcy- breakthrough that would make electric vehicles com- cles have been built with rotary engines. petitive appears unlikely in the near future (OTA 1995). Hybrid vehicle designs, in which an internal combus- Gas-Turbine (Brayton) Engines tion Brayton or Stirling engine would supplement the Gas-turbine engines are used in aircraft, stationary batteries, are being developed. In this design, the engine applications, high-speed trains, and marine vessels. supplies average power while batteries supply surge These engines have high output in relation to engine power for acceleration and absorb power during size and low emissions because of a low-pressure com- braking. Running under steady state conditions at its bustion process. Gas-turbine engines have been tested most efficient point,the engine in a hybrid vehicle could in road vehicles since the 1960s, but no commercially have very low emissions, and such vehicles could more viable vehicle system has been developed. Drawbacks than double the fuel efficiency of present vehide of gas turbines for road vehicles include high costs, designs. Electric and hybrid vehicles are discussed in poor transient response, and inefficiency, particularly at appendix 5.2. light loads.The problem of poor efficiency at light loads is especially severe in passenger vehicles, which com- Contol Technology for Gasoline-Fueled monly use less than 10 percent of the maximum power output in highway cruise conditions. Vehicles (Spark-Ignition Engines) Emissions from spark-ignition engines can be reduced Steam (Rankeine) Engines through changes in engine design, combustion condi- Steam engines were used in early automobiles. These tions, and catalytic aftertreatment. Some of the engine engines lost favor to spark-ignition engines because of and combustion variables that affect emissions are the the efficiency of gasoline engines, the time required to air-fuel ratio, ignition timing, turbulence in the combus- raise steam pressure, the need to refill with both high- tion chamber, and exhaust gas recirculation. Of these, purity water and fuel, and safety concerns over the the most important is the air-fuel ratio.These topics are high-pressure boiler. Like other engines using external discussed briefly below, and in detail in appendix 3.1. combustion, steam engines exhibit low pollutant emis- Engine-out pollutant emissions can be reduced sub- sions compared with uncontrolled internal combustion stantially from uncontrolled levels through appropriate 66 Air Pollutionfrom Motor Vebicles Figure 3.1 Effect of Air-Fuel Ratio on Spark-Ignition Engine Emissions l CO ) XNOX Lean Misfire coNO Limit a) CD a) CO H C 0 -4 ~ Richer 1.0 Leaner _ >mo- Normalized air/fuel ratio ) Source: Weaver and Chan 1995 engine design and control strategies. This involves ratio of 17.60:1I)-are lean burn.The generalized vari- tradeoffs arnong engine complexity, fuel economy, pow- ation of emissions with air-fuel ratio for a spark-igni- *zr, and emissions.The use of catalytic aftertreatment al- tion engine is shown in figure 3.1. lows a furtlier order-of-magnitude reduction in poollutant ernissions and, by reducing the need for en- Electronic Control Systems gine-out control, an improvemnent in power and fuel Electronic control technology for stoichiometric en- economy at a given emissions level. gines using three-way catalysts has been extensively de- veloped. Nearly all engine em-ission control systems A ir-Fuel Ratio used in the United States since 1981 incorporate com- The air-fuiel ratio has an important effect on engine puter control of the air-fuiel ratio. Similar systems have power, efficiency, and emissions. The ratio of air to 'oeen used in Japan since 1978 and in Europe since the fuel in the combustible mixture is a key design param- late 1980s.These systems measure the air-fuel ratio in eter for spark-ignition engines.An air-fuel mixture that the exhaust and adjust the air-fuel mixture going into has exactly enough air to burn the fuel, with neither the engine to maintain stoichiometry. In addition to the air nor fuel left over, is stoichiometric, and has a air-fuel ratio, computer systems control features that normalized air-fuel ratio (k) of 1 .0 . Mixtures with were controlled by vacuum switches or other devices in more air than fuel are lean, with Xs higher than 1.0; earlier emission control systems. These include spark those with more fuel are rich, with ks less than 1.0. A timing, exhaust gas recirculation, idle speed, air injec- mixture with a X of 1.5 has 50 percent more air than tion systems, and evaporative canister purging. needed to burn all the fuel. Engines using lean mix- The stringent air-fuel ratio requirements of three-way tures are more efficient than those using stoichiomet- catalysts made advanced control systems necessary. But ric mixtures. There are a number of reasons for this, the precision and flexibility of the electronic control sys- including less heat loss, higher compression ratios Eem can reduce etnissions even in the absence of a cata- (lean mixtures knock less readily), lower throttling lytic converter. Many control systems can self-diagnose losses at part load, and favorable thermodynamic engine and control system problems. Such diagnostics properties in burned gases. Engines designed to burn are mandatory in the United States.The abflity to warn veryleanmixtres- mor thn 1. (nuericair-uel the driver of a malfunction and assist the mechanic in its diagnosis can improve maintenarice quality. Self-diagnos- I . The numerical value of the stoichiometric air-fuel ratio for gaso- tic capabilities are becomiing increasingly sophisticated line is 14.7: 1, corresponding to a X of 1.00. and important as engine control systems become more Vehicle Technology for Controlling Emisstons 67 complex. Computer-controlled engine systems are also carbon monoxide, palladium and rhodium promote the more resistant to tampering and maladjustment than me- reduction of nitric oxide (NO) to nitrogen and oxygen. chanical controls.The tendency for emissions to increase For efficient NO reduction, a rich or stoichiometric air- over time is thus reduced in computer-controlled fuel ratio is required.At optimal conditions a three-way vehicles. catalyst can oxidize hydrocarbons and carbon monoxide and reduce nitrogen oxides.The window of air-fuel ratios Catalytic Converters in which this occurs is narrow, and there is a tradeoff be- The catalytic converter is one of the most effective tween nitrogen oxide and hydrocarbon/carbon monox- emission control devices available. The catalytic con- ide control even within this window. The variation of verter processes exhaust to remove pollutants, achiev- three-way catalyst efficiency with the normalized air-fuel ing considerably lower emissions than is possible with ratio Q.) is shown in figure 3.3.To maintain the precise in-cylinder techniques. Vehicles with catalytic convert- air-fuel ratio required, gasoline cars use exhaust (L) sen- ers require unleaded fuel, since lead forms deposits that sors (also known as oxygen sensors) with electronic con- poison"'the catalytic converter by blocking the access trol systems for feedback control of the air-fuel ratio. of exhaust gases to the catalyst. A single tank of leaded gasoline can significantly degrade catalyst efficiency. Lean nitrogen-oxide catalysts. Conventional three-way Sulfur and phosphorous in fuel can also poison the cat- catalysts are ineffective at reducing nitrogen oxides un- alytic converter. Converters can also be damaged by ex- der lean conditions. This has restricted the use of ad- cessive temperature, which can arise from excess vanced lean-burn engines in passenger vehicles. oxygen and unburned fuel in the exhaust. Because of their superior fuel efficiency and low carbon The catalytic converter comprises a ceramic sup- monoxide emissions, lean-burn engines are otherwise port, a washcoat (usually aluminum oxide) to provide a an attractive technology. Researchers have developed very large surface area and a surface layer of precious zeolite catalytic materials that reduce nitrogen oxide metals (platinum, rhodium, and palladium are most emissions,usingunbumedhydrocarbonsintheexhaust commonly used) to perform the catalyst function. Cata- as the reductant. Although the lean nitrogen-oxide cat- lysts containing palladium are more sensitive to the sul- alyst is typically about 50 percent effective-consider- fur content of gasoline than platinum/rhodium catalysts ably less than a three-way catalyst under stoichiometric (ACEA/EUROPIA 1995). conditions-the benefit is still significant. A few auto- Two types of catalytic converters are commonly used mobile models using lean nitrogen-oxide catalysts have in automotive engines: oxidation (two-way) catalysts been introduced in Japan. control hydrocarbon and carbon monoxide emissions Crankcase Emissions and Control and oxidation-reduction (three-way) catalysts control hydrocarbons, carbon monoxide, and nitrogen oxides The blowby of compressed gases past the piston rings (figure 3.2).A new type of catalytic converter is the lean consists mostly of unburned or partly-burned hydrocar- nitrogen-oxide catalyst, which reduces nitrogen oxide bons. In uncontrolled vehicles, the blowby gases were emissions in lean conditions where a three-way catalyst vented to the atmosphere. Crankcase emission controls is ineffective. involve closing the crankcase vent port and venting the crankcase to the air intake system via a check valve. Con- Two-way catalysts. Oxidation catalysts use platinum, trol of these emissions is no longer considered a signifi- palladium, or both to increase the rate of reaction be- cant technical issue. tween oxygen, unburned hydrocarbons, and carbon monoxide in the exhaust.This reaction would normally Evaporative Emissions and Control proceed slowly. Catalyst effectiveness depends on its Gasoline is a relatively volatile fuel. Even at normal temperature, the air-fuel ratio of the mixture, and the temperatures, significant gasoline evaporation occurs if mix of hydrocarbons present. Highly reactive hydrocar- gasoline is stored in a vented tank.The most common bons such as formaldehyde and olefins are oxidized measure of gasoline volatility is the Reid vapor pressure more effectively than less-reactive ones. Short-chain (RVP), which is the vapor pressure measured under paraffins like methane, ethane, and propane are among standard conditions at an air-to-liquid ratio of 4:1 and a the least reactive hydrocarbons and are difficult to oxi- temperature of 37.80C. Gasoline volatility is normally dize. adjusted to compensate for variations in ambient tem- perature. When temperatures are below freezing 0°C, Three-way catalysts.Three-way catalysts generally use a gasoline is usually adjusted to an RVP of about 90 kPa combination of platinum, palladium, and rhodium. In (13 psi) to increase fuel vaporization. This level of vol- addition to promoting the oxidation of hydrocarbons and atility would cause vapor lock in vehicles at tempera- 68 Air PoUutionfromMotorVehicles Figure 3.2 Types of Catalytic Converters Secondary air Mixture-formation "t = =1 systemOxdto SINGLE BED OXIDATION CATALYTIC CONVERTER Reduction Oxidation i _ ~~~~~~~ ~ ~ ~~~catalyst cat,alyst Mixture-formationr Secondary air DUAL BED OXIDATION CATALYTIC CONVERTER t r ~~~~~~~Lambda Mixture-formationl E 'l. system . . 2 ^ ~~~~~~~~catalyst _ . _ ^ i ~~~NOX, HC, CO SINGLE BED THREE-WAY CATALYTIC CONVERTER Source: Wijetilleke and Karunaratne 1992 Vehicle Technology for Controlling EmIssions 69 Figure 3.3 Effect of Air-Fuel Ratio on Three-Way Catalyst Efficiency 100 . .....l 80- C~~H 40 window of 16 20 0 1 0 . ...___._....._ 0.97 0.98 0.99 1.00 1.01 1.02 1.03 Normalized air/fuel ratio (k) 14.3 14.4 14.5 14.7 14.8 15.0 15.1 Numerical air/fuel ratio (A) Source: Weaver and Chan 1995 tures exceeding 30°C. At these temperatures, gasoline tank truck is refilled at the bulk terminal. Sources and RVP is ideally kept below 70 kPa (10 psi). Gasoline with magnitude of hydrocarbon vapor emissions from gaso- an RVP of 75 kPa (11 psi) will produce about twice the line distribution and dispensing are shown in figure 3.4. evaporative emissions of gasoline with an RVP of 60 kPa Technology to reduce gasoline distribution emis- (8.7 psi). sions involves two types of controls. One method con- The four primary sources of evaporative emissions from trols vapors displaced from the receiving tank by vehicles are diurnal (daily) emissions, hot-soak emissions, venting them to the delivery truck tank.This is known resting losses, and running losses. as Stage I and is about 95 percent effective, reducing va- Diurnal and hot-soak emissions have been controlled for por emissions from 1.14 grams per liter dispensed to some time in the United States, and such controls were in- 0.06 grams per liter. International experience shows duded in the Consolidated Emissions Directive adopted by that the cost of retrofitting fuel storage tanks, delivery the European Community in 1991. Evaporative emissions trucks, and service stations with vapor recovery devices are controlled by venting the fuel tank (and, in carbureted is small and the payback period based on the benefit of vehicles, the carburetor bowl) to the atmosphere through a fuel savings alone, is two to three years. Two alterna- canister of activated charcoal. Hydrocarbon vapors are ad- tives are available to control fuel vapors displaced from sorbed by the charcoal, so little vapor escapes to the air.The the vehicle tank during refueling (Stage 11 control). One charcoal canister is regenerated or"purged" by drawing air alternative modifies the gasoline dispensing system to through it into the intake manifold when the engine is run- capture vapors.The other alternative captures vapor on ning. Adsorbed hydrocarbons are stripped from the char- board the vehicle in a charcoal canister similar to that coal and burned in the engine. used for controlling evaporative emissions. Fuel Dispensing/Distribution Emissions Control Technology for Diesel-Fueled and Control CnrlTcnlg o islFee Vehicles (Compression-ignition Engines) As with evaporative emissions, emissions from fuel dis- tribution are significant only for vehicles using volatile The principal pollutants emitted by diesel engines are fuels, such as gasoline.These emissions result from fuel nitrogen oxides, sulfur dioxide, particulate matter, and vapor contained in the headspace of the vehicle fuel hydrocarbons. Diesels also produce carbon monoxide, tank.This vapor is displaced as fuel is added during re- smoke, odors, and noise. Fuel quality affects diesel fueling. Vapor emissions also occur when the service emissions, the main factors being fuel density, sulfur station tank is refilled from a tank truck, and when the content, aromatic content, and certain distillation char- 70 Air Pollutionfrom Motor Vebicles Figure 3.4 Hydrocarbon Vapor Emissions from Gasoline Distribution Vent Emissions: 1.14 gliter from gasoline deliveries I O.2 Miter from breathina losse Emissions from refilling 'Refuelng Emissions tank truck: 1.14 g/itr Filipipe Issions: 1.44 g/iter Spillage Emissions: 0.08 gAiter Gas Pump Total emissions: 3.92 gAiter Refueling emissions: 1.64 gAiter Note: 62 kPa (9 PSI) Reid vapor pressure Source: Weaver and Chan 1995 acteristics. Engine variables with the greatest effect on involves complex tradeoffs among nitrogen oxide, hy- diesel emission rates are the combustion chamber de- drocarbon, and particulate matter (PM) emissions. sign, air-fuel ratio, rate of air-fuel mixing, fuel injection Most engine manufacturers have followed a broadly timing, compression ratio, and the temperature and similar approach to reducing diesel emissions, although composition of the charge in the cylinder.These factors the specific techniques used differ considerably from are discussed in detail in appendix 3.2. one manufacturer to the next.This typical approach in- cludes the following major elements: Engine Design * Reducing parasitic hydrocarbon and PM emissions There is a tradeoff between nitrogen oxide and par- (those not directly related to the combustion pro- ticulate control measures in diesel vehicles. This cess) bni injectio nze msa v rol tradeoff is shown in figure 3.5 for three different levels and oil consumption of diesel technology.The tradeoff is not absolute-both * Reducing PM emissions and improving fuel effi- nitrogen oxides and particulate matter emissions can be ciency and power output through turbocharging reduced simultaneously. There are limits on the extent and by refining the match between the turbocharg- to which either can be reduced, however, without in- er and the engine creasing the other. To minimize all pollutants simulta- * Reducing emissions of PM and nitrogen oxides by neously requires optimization of fuel injection, fuel-air cooling the compressed-charge air with aftercoolers mixing, and combustion processes over the range of * Further reducing nitrogen oxides to meet regulato- operating conditions. ry targets by retarding fuel injection timing over Reduced nitrogen oxide and particulate emissions most of the speed-load range.A flexible timing sys- have resulted from an improved understanding of the tem minimizes the adverse effects of retarded tim- diesel combustion process and the factors affecting ing on smoke, starting, and light-load hydrocarbon pollutant formation and destruction in the cylinder. emissions Modifying the diesel combustion process is complex; it * Further reducing nitrogen oxides in light-duty vehi- has a direct impact on cost, fuel economy, power and cles by recirculating exhaust gas under light-load torque output, cold starting, and visible smoke, and it conditions Vehicle Tecbnology for Controlling Emissions 71 • Reducing the PM increase resulting from retarded and combustion system, and costs to manufacturers and timing by increasing the fuel injection pressure and engine purchasers have been sizable. But the benefits in injection rate the form of cleaner and more efficient engines have * Improving air utilization (and reducing hydrocar- been significant. bon and PM emissions) by minimizing parasitic vol- umes in the combustion chamber-such as the ExhaustAftertreatment clearance between piston and cylinder head and Another approach to reducing pollutant emissions is to the clearance between the piston and the walls of use a separate process to eliminate pollutants from the the cylinder exhaust after it leaves the engine, but before it is emit- * Optimizing in-cylinder air motion through changes ted into the air. Aftertreatment systems include particu- in combustion chamber geometry and intake air late trap-oxidizers and diesel catalytic converters, both swirl to provide adequate mixing at low speeds (to of which have been used in vehicles.Work is under way minimize smoke and PM) without over-rapid mix- on lean nitrogen oxide catalysts for diesel engines, but ing at high speeds (which would increase hydrocar- success has been limited. bons, nitrogen oxides, and fuel consumption); and * Controlling smoke and PM emissions in full-power Trap-oxidizers. A trap-oxidizer system has a particulate operation and transient accelerations by improving filter (the trap) in the engine exhaust stream and some the governor curve shape and limiting transient means of burning (oxidizing) collected particulate mat- smoke (frequently through electronic governor ter from the filter. Manufacturing a filter capable of col- controls). lecting soot and other particulate matter from the exhaust stream is straightforward, and effective trapping These changes have reduced PM emissions from die- media have been developed and demonstrated. The sel engines by more than 80 percent and emissions of ni- main problem of trap-oxidizer system development is trogen oxides by 50 to 70 percent compared with how to remove the soot effectively and regenerate the uncontrolled levels. Fuel efficiency has increased mark- filter. Diesel particulate matter consists of solid carbon edly compared with older engines. These emission re- coated with heavy hydrocarbons. This mixture ignites at ductions and fuel economy improvements have 500 to 600°C, well above the normal range of diesel en- required complete redesign of large parts of the engine gine exhaust temperatures (1 50-400°C). Special means Figure 3.5 Nitrogen Oxide and Particulate Matter Emissions from Diesel-Fueled Engines I1- 0.9 0.8 0.7 - 0.6 - 1985 ENGINE 1988 Standards 0.5 E w) 0.4- 0.3- 0.2 1994 Standards . 0 2 4 6 8 Nitrogen oxides (g/bhp-hr) Source: Wcaver and Chan 1995 72 Air Pollutionfrom Motor Vebicles are therefore needed to ensure ignition. Once ignited, manufacturers are working on quasi-passive systems, in however, this material burns at temperatures that can which the system usually regenerates passively without melt or crack the particulate filter. Initiating and control- intervention, but the active system remains as a backup. ling regeneration without damaging the trap is the cen- No catalytic coating has sufficiently reduced trap re- tral problem of trap-oxidizer development. generation temperature to permit reliable passive regen- A number of trapping media have been tested or pro- eration in heavy-duty diesel service. But catalyst coatings posed, including cellular ceramic monoliths, woven ce- have a number of advantages in active systems. The re- ramic-fiber coils, ceramic foams, corrugated multi-fiber duced ignition temperature and increased combustion felts, and catalyst-coated, stainless-steel wire mesh. The rate resulting from the catalyst imply that less energy is most successful trap-oxidizer systems use either the ce- needed from the regeneration system. Regeneration will ramic monolith or the ceramic-fiber coil traps (Feutlin- also occur spontaneously under most duty cycles, greatly ske 1989; Holman 1990; Knecht 1991). reducing the number of times the regeneration system Many techniques for regenerating particulate trap- must operate. Spontaneous regeneration also provides oxidizers have been proposed, and much development insurance against regeneration system failure. Finally, a effort has been invested. Regeneration techniques can trap catalyst may simplify a regeneration system. be divided into passive and active approaches. Passive To date, trap-oxidizer systems have been used in only systems attain the conditions required for regeneration a few engine and vehicle models. Traps were installed as a result of normal vehicle operation.This requires a on one diesel passenger car model sold in California in catalyst (as either a coating on the trap or a fuel addi- the early 1990s, but the systems were not durable and tive) to reduce the ignition temperature of the collected were withdrawn after two years.Trap-oxidizer systems particulate matter. Regeneration temperatures of 420 °C were standard devices on new, heavy-duty, U.S. bus have been reported with catalytic coatings, and lower engines certified to meet 1993 PM standards, and they temperatures can be achieved with fuel additives. Ac- have been retrofitted to buses in a number of U.S. urban tive systems monitor particulate matter in the trap and areas, including New York City and Philadelphia. They trigger specific actions to regenerate it when needed.A have also been deployed in a number of demonstration variety of approaches to trigger regeneration have been projects in Europe, including transit buses in Germany proposed, including diesel-fuel burners, electric heat- andAthens, Greece (Pattas and others 1990).The Athens ers, and catalyst injection systems. program was so successful that proposals were made to Passive regeneration is difficult on heavy-duty vehi- fit trap-oxidizers to all buses in the city (box 3.1). cles. Regeneration temperatures must be attained in Addition of a trap-oxidizer would add substantially to normal operation, even under light-load conditions. the initial cost of a diesel engine, and would increase Currently, no purely passive regeneration system is un- fuel consumption and maintenance costs. Engine man- der consideration for heavy-duty applications. Some ufacturers anticipate strong market resistance to this Box 3.1 Trap-Oxidizer Development in Greece Urban buses are responsible for more than half the traffic-produced smoke in downtown areas of major Greek cities. Since the service life of these vehicles often approaches 15 years, the possibility of retrofitting urban buses with trap-oxidizer sys- tems was considered by the Athens Bus Corporation, utilizing the following features: * Wall-flow ceramic monoliths for filtration, with more than 90 percent filtration efficiency, * A regeneration system using cerium-based fuel additives and exhaust gas throttling, and * Bypass control of the regeneration system for protection against filter melting or cracking. A pilot phase was initiated in 1989 to determine the service life of the filters and the feasibility of the trap-oxidizer system. A Greek manufacturer produced the systems, and 1 10 buses were retrofitted and put in normal service.Two years of pilot operation indicated that the normal service life of the filter exceeded 100,000 kilometers-more than a year of bus opera- tion.The trap-oxidizer system represented 3 to 5 percent of the market price of a new bus. And the operational cost (fuel penalty plus fuel additive) was 2 percent of the cost of fuel. On the basis of these findings, the Greek Ministry for the Environment recommended retrofitting the entire Athens urban bus fleet with traps, sponsoring similar actions in other major Greek cities, and adopting U.S. regulations for particulate emis- sions for all new urban buses sold in Athens. Unfortunately, the Greek Ministry ofTransport was unable to raise the 1 billion drachmas needed to fit the remaining 1,700 city buses. Source: Pattas and others 1990; Hope 1992 Vebicle Tecbnologyfor Contrvlling EmissIons 73 technology. Their success in reducing engine-out PM achieved, but results have been discouraging, partly be- emissions from new diesel engines has greatly dimin- cause water in the exhaust inhibits the catalyst and part- ished the interest in trap-oxidizers. ly because the sulfur in diesel poisons the catalyst. In-cylinder diesel PM control has greatly reduced PM Current catalyst formulations also require that the ex- emission levels. Progress has been most effective in re- haust be hydrocarbon-enriched to achieve reasonable ducing the soot portion of PM emissions, so the soluble efficiency, thus increasing fuel consumption and possi- organic portion of particulate matter now accounts for bly emissions. Despite these problems, some analysts a much larger share. Depending on engine and operat- expect viable lean nitrogen-oxide catalysts for diesel ing conditions, the soluble organic portion may ac- engines to be developed by the late 1990s. count for 30 to 70 percent of PM emissions. Vertical exhausts. The exhaust pipes on heavy-duty ve- Oxidation catalysts. Like a catalytic trap, a diesel catalyt- hicles are either vertical (so that the exhaust is emitted ic converter oxidizes a large portion of the hydrocar- above the vehicles) or horizontal.Although the choice bons present in the soluble organic portion of PM of exhaust location does not affect overall pollutant emissions, as well as gaseous hydrocarbons, carbon emissions, it can have a significant effect on local con- monoxide, odor (from organic compounds such as aIde- centrations of pollutants. A vertical exhaust pipe reduc- hydes), and mutagenic emissions. Unlike a catalytic trap, es the concentration of exhaust pollutants at breathing the oxidizing catalytic converter does not collect solid level, reducing human exposure to high local concen- particulate matter, which passes through in the exhaust. trations. Vertical exhausts can reduce exposure to high This eliminates the need for a regeneration system. local concentrations of pollutants by 65 to 87 percent Oxidation catalytic converters have been used in (Weaver and others 1986). Vertical exhausts also make light-duty vehicles and demonstrated to be effective for it easier to enforce on-road smoke limitations. heavy-duty applications.They have little effect on nitro- Many heavy-duty trucks and some buses are designed gen oxide emissions, but can reduce volatile organic with vertical exhausts from the beginning.There is no compound and carbon monoxide emissions by up to 80 technical reason why all trucks and buses could not be percent.The durability of oxidation catalytic converters so equipped. Retrofitting vertical exhausts to trucks on heavy-duty engines has yet to be determined, but it originally equipped with horizontal exhausts is feasible is likely to be acceptable.These catalysts have a negligi- in many cases, but can be impractical because of limits ble effect on fuel consumption. imposed by truck design or use. Retrofitting buses is The main difficulty with using oxidation catalytic more complex, but also feasible. In Santiago, Chile, reg- converters on heavy-duty diesel engines is that they can ulations adopted in 1987 required vertical exhausts on cause the formation of sulfuric acid and sulfates from all buses, and led to a large number of retrofits. sulfur dioxide in the exhaust. If fuel sulfur levels are sig- nificant, these compounds can add considerably to par- ticulate mass. Fuel with less than 0.05 percent sulfur by Emission Control Options and Costs weight is required for diesel catalysts to perform well. However, the linkage between fuel sulfur content and This section discusses motor vehicle emission controls the higher conversion rates of SO2 (gaseous) to S03 achievable with current and foreseeable technology, (particulate matter) in vehicles equipped with oxida- and estimates the costs of achieving these controls .The tion catalysts may be due to the high operating temper- focus is on technology rather than regulations (see ature of the catalyst (above 3500C) obtained during Chapter 1 for a discussion of vehicle emission standards static dynometer tests. Actual on-the-road operating and regulations). conditions tend to result in lower catalyst temperature and hence a lower rate of sulphate particle emissions Gasoline-Fueled Passenger Cars (CONCAWE Review 1994). and Ligbt-Duty Trucks Many technologies that improve automotive fuel Lean nitrogen-oxide catalysts. Since diesel engines op- efficiency such as fuel injection, electronic control of erate with lean air-fuel ratios, three-way catalytic con- spark timing, advanced choke systems, and improved verters do not reduce the emissions of nitrogen oxides. transmissions, also reduce exhaust emissions.And some Research is underway on zeolite-based, lean nitrogen- emission control requirements have improved fuel oxide catalysts that reduce nitrogen oxide emissions efficiency. In the absence of tight emission standards using unburned hydrocarbons in the exhaust.A 20 per- and controls, it is unlikely that these advanced engine cent reduction in nitrogen oxide emissions has been technologies would have been applied to automobiles. 74 Air Pollution from Motor Vehicles In industrialized countries, passenger cars and light- tion of unleaded fuel and catalytic converters in 1975 duty trucks are responsible for a larger share of total mo- coincided with substantial fuel economy gains. bile source pollutant emissions than any other vehicle Table 3.2 summarizes six possible levels of emission category. Many jurisdictions have adopted strict limits on control for light-duty vehicles which range from the emissions from new light-duty vehicles (chapter 1).As a simple controls used in the UJnited States and Japan in result of these emission regulations, several levels of con- the 1970s (and, until the early 1990s, in Europe) to the trol technology have been developed that can be classi- most sophisticated systems now available. fied according to effectiveness, complexity, and cost. Emission controls range from those achievable through Non-catalyst controls. Emissions standards at this control simple air- fuel ratio and timing adjustments to standards level can be met by four-stroke gasoline engines without requiring feedback-controlled fuel injection systems with emissions aftertreatment and correspond to U.S. stan- exhaust gas recirculation and three-way catalysts. dards of the early 1970s. Exhaust emission controls for The costs of emission control systems are controver- gasoline vehicles involve modification in carburetor de- sial. Industry estimates of cost and fuel consumption sign and setting the air-fuel ratio to minimize carbon changes for gasoline vehicles in various engine and monoxide and hydrocarbon emissions, while nitrogen treatment configurations are given in table 3.1. Esti- oxide control is achieved by retarding ignition timing, ex- mates by other parties give somewhat different results haust gas recirculation, or both. Diesel vehicles require (compare table 3.2, based on estimates developed by no modifications to meet these standards.Two-stroke gas- Christopher Weaver for the U.S. EPA). oline engines would either be eliminated or forced to As vehicle technology is pushed to achieve low adopt fuel injection. Crankcase and evaporative emission pollution levels, common international elements are controls for gasoline vehicles are also needed. emerging. In every case, the least polluting vehicles use The non-catalyst approach avoids the complexities catalytic converters. Since these systems are poisoned and costs of catalysts and unleaded fuel. Total system by lead and the phosphorous in most engine oils, they cost would be about U.S.$130 for a passenger car- foster the introduction of unleaded gasoline and cleaner U.S.$60 for air injection, U.S.$20 for evaporative con- oils, reducing overall lead pollution. To optimize these trols, and U.S.$5 for crankcase controls. The remaining systems, better air-fuel and spark management systems U.S.$45 covers engine modifications and cold-start have evolved, leading to increased use of both electron- emission controls. ics and fuel injection.These advances also increase fuel efficiency and lower carbon dioxide emissions. Oxidation catalyst. Unleaded gasoline increases the tech- Improved emissions in the United States have been nological feasibility of more stringent exhaust emissions accompanied by improved fuel economy.The weighted control. With two-way catalysts, air injection, and me- fleet average in 1967 was 14.9 miles per gallon, com- chanical air-fuel ratio controls (such as a standard carbure- pared with 27.3 miles per gallon in 1987, an increase of tor), total system cost is about U.S.$380-$205 for the 83 percent. Correcting for vehicle weight reductions, catalytic converter, U.S.$60 for the air injection system, the improvement was about 47 percent.The introduc- and U.S.$25 for crankcase and evaporative controls.The Table 3.1 Automaker Estimates of Emission Control Technology Costs for Gasoline-Fueled Vehicles (percent) Technology Engine cost increase Fuel consumption change Lean-burn engine with carburetor and conventional ignition 1.0 -2 Pulse air and exhaust gas recirculation 4.5 3 Lean-burn engine with carburetor and programmed ignition 2.0 1 Recalibrated conventional engine with electronic fuel injection 8.0 2 Lean-burn engine with electronic fuel injection 9.0 -7 Lean-burn engine with oxidation catalyst 4.5 -3 Open loop, three-way catalyst carburetor 4.1 2 Lean-burn engine, closed loop, electronic fuel injection, variable intake oxidation catalyst 15.0 -7 Closed loop, electronic fuel injection, three-way catalyst 13.0 3 Note: Baseline is a small vehicle with 1.4-liter conventional carburetor engine meeting ECE 15/04 standard. Source: ECMT 1990 Vehicle Tecbnologyfor Controlling Emissions 75 Table 3.2 Exhaust Emission Control Levels for Light-Duty Gasoline-Fueled Vehicles Emission standard Estimated cost Percent Fuel economy per vehicle Control level Grams per kilometer4 controlledb Controls required (percent) (US. dollars) Non-catalyst Hydrocarbons-1.5 66 Ignition timing -5 130 controls Carbon monoxide- 15 63 Air-fuel ratio Nitrogen oxides- 1.9 11 Air injection Exhaust gas recirculation Oxidation Hydrocarbons-0.5 89 Oxidation catalyst -5 380 catalyst Carbon monoxide-7.0 83 Ignition timing Nitrogen oxides- 1.3 39 Exhaust gas recirculation Three-way Hydrocarbons-0.25 94 Three-way catalyst -5 (carburetor) 630 catalyst Carbon monoxide-2.1 95 Closed-loop carburetor 5 (electronic fuel Nitrogen oxides-0.63 71 or electronic fuel injection injection) Lean-burn engine Hydrocarbons-0.25 94 Oxidation catalyst 15 630 Carbon monoxide- 1.0 98 Electronic fuel injection Nitrogen oxides-0.63 71 Fast-burn combustion chamber U.S. tier I Hydrocarbons-0. 16 96 Three-way catalyst 5 800 Carbon monoxide- 1.3 97 Electronic fuel injection Nitrogen oxides-0.25 88 Exhaust gas recirculation California Hydrocarbons-0.047 99 Electric three-way catalyst tJnknown More than 1,000 low-emission Carbon monoxide-0.6 99 Electronic fuel injection vehicle standard Nitrogen oxides-0.13 94 Exhaust gas recirculation a. At 80,000 kilometers. b. Compared with uncontrolled levels. Source: U.S. EPA 1990 remaining U.S.$90 covers electronic ignition timing, high- for the catalytic converter, U.S.$60 for the air injection energy ignition, and cold-start emission controls. system, and U.S.$25 for crankcase and evaporative con- trols.The remaining U.S.$280 is allocated for the fuel in- Three-way catalyst/lean-burn engine. This level is jection system and electronic controls, which also help equivalent to 1981 U.S. emissions standards. It is essen- to improve performance and fuel economy. tially the world standard of the early 1990s-many oth- er countries have adopted them, and current Japanese US. tier I.These standards, adopted in amendments to and ECE regulations require a similar level of control. In the U.S. Clean AirAct, reflect the state-of-the-art in emis- gasoline-fueled vehicles, a three-way catalyst (in con- sion control for light-duty vehicles. These standards junction with exhaust gas recirculation, control of were implemented in California in the early 1990s and spark timing, and other measures) reduces emissions of in the rest of the United States in the mid-1990s.These carbon monoxide, hydrocarbons, and nitrogen oxides. standards require a three-way catalyst, with the air-fuel This is achievable using a stoichiometric carburetor sys- ratio controlled through electronic fuel injection and tem with closed-loop electronic trim, though the trend air-fuel ratio feedback. This system costs about is to use fully electronic systems with fuel injection2. U.S.$800. Compared with vehicles meeting the 1981 These levels have also been met with lean-bum technol- U.S. standards, vehicles meeting this emission level ogy (with an oxidation catalyst) at a similar cost, but have more precise air-fuel ratio control, more precious with better fuel economy and lower carbon monoxide. metal in the catalytic converter, better evaporative emis- Total system cost for a gasoline-fueled passenger car sion controls, and better durability and reliability to or light-duty truck is estimated at U.S.$630-U.S.$265 meet in-use requirements. Contemporary light-duty die- sel vehicles in the United States are capable of meeting the particulate matter, hydrocarbon, and carbon mon- 2. Diesel vehicles only requirc engine modifications, control oxide standards, but require a higher nitrogen oxide optimization, and possibly exhaust gas recirculation to achieve these standard-about 0.5 grams per kilometer. emnission levels. 76 Ar Pollutionfrom Motor Vehicles California low-emission vehicle (LEV). These stan- Motorcycles dards, to be implemented in California beginning in In the past, motorcycles were subject to lenient emis- 1997, and in other parts of the United States in 2001, ex- sion controls or none at all.This reflects their minor con- ceed the current state-of-the-art in emission control for tribution to emissions in most industrial countries and gasoline vehicles. Compliance with these standards will the difficulty and expense of installing emission controls require even more precise control of the air-fuel ratio, on small, heavily-loaded engines. In many Asian cities, heavier catalyst loadings, and better engine-out emis- however, motorcycles and three- wheelers are responsi- sion controls than the current state-of-the-art, and pos- ble for a large fraction of hydrocarbon and PM emissions. sibly new technologies or alternative fuels. For large Recommended emission control levels based on an as- vehicles, it appears that preheated catalytic converters sessment of motorcyde emissions in Thailand are sum- may be required to reduce emissions when the vehicle marized in table 3.3 (Weaver and Chan 1994). is started cold.This will pose a significant challenge to The first step in controlling emissions from motorcy- automakers. The total cost is unknown, but will most cles vehicles is eliminating the excessive emissions likely exceed U.S.$1,000. from two-stroke engines.This can be done by switching to a four-stroke design or to a two- stroke design incor- Heavy-Duty Gasoline-Fueled Vehicles porating timed fuel injection and crankcase lubrication. Heavy-duty vehicles with gasoline engines can use the This would reduce hydrocarbon and particulate matter emission control technologies outlined above. Because emissions by about 90 percent, at a cost of about of differences in vehicle size and testing procedures, U.S.$60 per vehicle.Additional emission reductions are however, emissions standards cannot be directly com- possible with improved four-stroke engine design and pared. Because of their smaller numbers, heavy-duty gas- calibration and through the use of catalytic converters. oline vehicles receive less regulatory attention, and Catalytic converters are used on two-stroke motorcy- emissions standards are less strict. Emission standards for cles in Taiwan (China) and on mopeds in Austria and these vehicles are also lax because they typically operate Switzerland. Catalyst-forcing standards for four-stroke at high loads, making it difficult to ensure catalyst dura- engines have not been adopted in any jurisdiction. bility. Design solutions for the catalyst durability problem exist, and recent California regulations will require vehi- Diesel-Fueled Vehicles des under 14,000 pounds gross vehicle weight to meet As with gasoline-fueled vehicles, diesel engine emis- low-emission vehicle standards. The cost of meeting sion reductions have accompanied improvements in emissions standards in larger vehicles is expected to be fuel efficiency.Although measures such as retarding in- 50 to 100 percent more than for passenger cars because jection timing increase fuel consumption, these have of the larger size of the equipment required. been offset by gains from turbocharging, charge-air Table 3.3 Recommended Emission Control Levels for Motorcycles in Thailand Emission standard Estimated cost Percent Fuel economy per vehicle Control level Grams per kllometera controlledb Controls required (percent) (US. dollars) Eliminate Hydrocarbons-5 .0 66 Four-stroke engine or advanced 30-40 60-80 two-stroke Carbon monoxide-12.0 50 two-stroke Nitrogen oxides-NR Particulate matter-0. 15 50-90 Non-catalyst Hydrocarbons-1.0 90 Four-stroke or two stroke with 0 80-100 controls Carbon monoxide-12.0 50 catalyst, ignition timning, air-fuel Nitrogen oxides-0.5 200 ratio control Particulate matter-0. 15 50-90 Oxidation Hydrocarbons-0.5 98 Four-stroke or advanced two- stroke, -5 80-100 catalyst or Carbon monoxide-2.0 80 ignition timing, air-fuel ratio advanced Nitrogen oxides-0.5 200 control, catalytic converter technology Particulate matter-0.05 85 or electronic fuel injection -Not applicable NR = Not regulated. a. At 80,000 kilometers. b. Compared with uncontrolled two-stroke. Source: Weaver and Chan 1994 Vebicle Tecbnologyfor Contr)lllng Emissions 77 Table 3.4 Industry Estimates of Emission Control Technology Costs for Diesel-Fueled Vehicles (percent) Technology Engine cost increase Baseline engine, no emission control equipment 0 Injection timing retard 0 Low sac volume and valve covering nozzle Minimal Turbocharging 3-5 Charge cooling 5-7 Improved fuel injection 13-15 High-pressure fuel injection with electronic control 14-16 Variable geometry turbocharging 1-3 Particulate trap 4-25 Source: ECMT 1990 (based onTonkin and Etheridge 1987) cooling, and improved fuel injection equipment. The tion, and charge- air cooling. A tighter U.S. particulate costs of these features are offset by lower fuel con- matter standard (0.13 g/kWh; 0.10 g/bhp-hr) took ef- sumption. Industry estimates of cost increases for en- fect in 1994.This standard applied to diesel engines in gine modifications to meet emission standards are urban buses in 1993, and a limit of 0.07 glbhp-hr was shown in table 3.4. adopted for 1994 and 1995.After 1995, U.S. urban bus- Uncontrolled emissions of nitrogen oxides from es will be required to meet an emission standard of 0.07 heavy-duty diesel engines range from 12 to 21 grams per g/kWh (0.05 g/bhp-hr). kilowatt-hour (9 to 16 grams per brake horsepower- Further reductions are being contemplated. The hour) when measured using the U.S. transient or Euro- 1990 U.S. Clean Air Act amendments require reduced pean 13-mode cycle. Particulate matter emissions on emissions of nitrogen oxides for all heavy-duty truck the transient cycle are typically 1-5 g/kWh (0.75-3.7 g/ and bus engines (5.4 glkWh; 4.0 g/bhp-hr) in 1998. bhp-hr), but are significantly lower on the European The stringent nitrogen oxides and PM standards steady-state cycle. Engines that have been tampered adopted by the United States would not have been possi- with or poorly maintained may emit higher PM emis- ble for diesel engines until very recently. Developments sions in the form of smoke. in fuel-injection rate-shaping, and the potential use of ex- By moderately retarding fuel injection timing from haust gas recirculation mean that nitrogen oxide emis- the optimal point, nitrogen oxides can be reduced to sion levels of 2.6 g/kWh (2.0 g/bhp-hr) may now be less than 11.0 g/kWh (8 g/bhp-hr).This may require up- achievable, in combination with low PM emissions. Us- graded fuel-injection equipment. Particulate matter ing selective exhaust gas recirculation and extensive en- emissions can be limited through smoke opacity stan- gine optimization, nitrogen oxide levels as low as 2.6 g/ dards under acceleration and full-load conditions.Achi- kWh (2.0 g/bhp-hr) have been achieved in the laboratory evable peak acceleration smoke opacity levels range (Needham, Doyle, and Nicol 1991).Translating research from 25 to 35 percent, while steady- state smoke opaci- results into marketable engines takes time, but this may ty of less than 5 percent (the limit of visibility) is readily be feasible by the end of the 1990s. achievable.This minimal control level is comparable to Further reductions in nitrogen oxide and particulate that required of California engines in the 1970s and Eu- matter emissions are possible with alternative fuels. ropean engines until the early 1990s. Pre-production methanol direct-injection engines us- Moderate control of 8 g/kWh (6 g/bhp-hr) of nitro- ing glow-plug assisted compression ignition have pro- gen oxides and 0.7 g/kWh (0.5 g/bhp-hr) of particulate duced nitrogen oxide emissions below 2.9 g/kWh (2.1 matter requires further optimization of the injection g/bhp-hr), with efficiency comparable to that of a reg- timing and the overall combustion system. This corre- ulated diesel engine. Heavy-duty, lean-burn, natural-gas sponds to 1990 U.S. federal standards.The next level of engines have achieved nitrogen oxide levels below 2.5 control corresponds to the 1991 U.S. standards, which g/kWh (1.9 g/bhp-hr), with energy efficiency about 10 have also been adopted in Canada and Mexico.Achiev- percent worse than the diesel. Spark-ignition engines ing this level of nitrogen oxides control while meeting using natural gas, liquefied petroleum gas, and gasoline PM emission standards requires major engine design with three-way catalysts, stoichiometric air-fuel ratios, modifications.These include variable fuel-injection tim- and closed-loop control have achieved nitrogen oxide ing, high-pressure fuel injection, combustion optimiza- emissions below 1.5 g/kWh (. 1 g/bhp-hr) at low mile- 78 Air PoUutton from Motor Vebicles Table 3.5 Emissions Control Levels for Heavy-Duty Diesel Vehicles Emissions limit atfull useful iffe Estimated Grams Fuel cost per per brake economy4 engine Control level Grams per kilowatt-hour borsepower-bour Controls required (percent) (US. dollars) Uncontrolled Nitrogen oxides- 12.0 to 21.0 9.0 to 16.0 None (PM level depends on 0 0 Particulate matter- 1.0 to 5.0 0.75 to 3.70 smoke controls & maintenance level) Minimal control Nitrogen oxides- 11.0 8.0 Injection timing -3 to 0 0-200 Particulate matter-0.7 to 1.0 0.5 to 0.75 Smoke limiter Peak smoke-20 to 30 percent opacity Moderate control Nitrogen oxides-8.0 6.0 Injection timing -5 too 0-1,500 Particulate matter-0.7 0.5 Combustion optimization 1991 U.S. standard Nitrogen oxides-6.7 (7.0) 5.0 Variable injection timing -5 to 5 1,000-3,000 (Euro 2) Particulate matter-0.34 (0.15)b 0.25 High-pressure fuel injection Combustion optimization Charge-air cooling Lowest diesel Nitrogen oxides-2.7 to 5.5C 2.0 to 4.0 Electronic fuel injection -10 to 0 2,000-6,000 standards under Particulate matter-0.07 to 0.13 0.05 to 0.10 Charge-air cooling consideration Combustion optimization Exhaust gas recirculation Catalytic converter or particulate trap Alternative-fuel Nitrogen oxides-less than 2.7 2.0 Gasoline/three-way catalyst -30 to 0 0-5,000 forcing Particulate matter-less than 0.07 0.04 Natural gas lean-burn Natural gas/three-way catalyst Methanol-diesel Note: Kilowatt-hours are converted to brake horsepower-hours by multiplying by 0.7452. a. Potential fuel economy improvements result from addition of turbocharging and intercooling to naturally aspirated engines. b. Euro-2 emissions are measured on a steady-state cycle that underestimates PM emissions in actual driving. Actual stringency of control require- ments is similar to that of U.S. 1991. c. Not yet demonstrated in production vehicles. Source: Weaver 1990 Table 3.6 Emission Control Levels for Light-Duty Diesel Vehicles Fuel Estimated cost Emissions limit atfull useful Iffe Reductiona economy per engine Control level (grams per kilometer) (percent) Controls required (percent) (US. doUars) Uncontrolled Nitrogen oxides-1.0 to 1.5 0 None (PM level depends 0 0 Particulate matter-0.6 to 1.0 0 on smoke controls & maintenance level) Moderate control Nitrogen oxides-0.6 40 Injection timing -5 to 0 0-500 Particulate matter-0.4 33 Combustion optimization 1988 U.S. standard Nitrogen oxides-0.6 40 Variable injection timing -5 to 0 100-200 (EU Directive (HC+NO,:0.97) 78 Combustion optimization 91/441/EEC) Particulate matter-0.13 (0.14) Exhaust gas recirculation Advanced diesel Nitrogen oxides-0.5 40 Electronic fuel injection -10 to 0 200-500 technology Particulate matter-0.05-0.08 92 Combustion optimization Exhaust gas recirculation Catalytic converter or particulate trap a. Compared with uncontrolled levels. Source: Weaver 1990 Vehicle Tecbnology for Controlling Emissions 79 age, and acceptable catalyst durability has been demon- Hope, K. 1992. "Urban Air Pollution: Greeks Battle to strated in some cases. Particulate matter emissions Defeat the Nefos." Financial Times, March 11, 1992. with these fuels are also low. Since they do not form London. soot or condensable organic compounds, particulate Kimbom, G. 1993. "Some Examples of Alternative En- emissions derive only from the lubricating oil. With a gines" in Towards Clean & Fuel Efficient Automo- catalytic converter, these are less than 0.05 g/kWh biles. Proceedings of an International Conference (0.04 g/bhp-hr). held in Berlin (March 25-27,1991), OECD, Paris. The emission control levels achievable for heavy- Knecht,W 1991. "Reduction of Heavy-duty Diesel Emis- duty diesel engines, the costs of achieving these levels, sions andApplication of ParticulateTraps and Oxida- and the corresponding effects on fuel economy are tion Catalysts." in The Diesel Engine-Energy Stake shown in table 3.5. These estimates prepared for an and Environment Constraints,TUV Brussels, Rhein- OECD study differ from those based on industry land/OPET. sources. MacKenzie,JJ. 1994. The Keys to the Car-Electric and Potential emission controls for light-duty diesel vehi- Hydrogen Vehicles for the 21st Century. World cles range from none to moderate control (applied in Resources Institute,Washington, D.C. Europe until recently) to the stringent controls typical Mason, J.L. 1993. IC Engines & Fuels for Cars & Light of California and the rest of the United States. Given the Trucks: 2015." in Transportation & Global Climate collapse in demand for diesel vehicles in the U.S., no Change, eds. L. Greene & DJ. Santini, American manufacturer considered it worthwhile to develop an Council for an Energy Efficient Economy, Washing- emissions control system to meet the California limits. ton, D.C. Similar emission standards for diesel-fueled passenger Needham, J.R., D.M. Doyle, and Aj. Nicol. 1991. 'The cars have recently been adopted in Europe, where die- Low Nitrogen Oxide Truck Engine." SAE Paper sel cars are a large part of the market.Vehicles meeting 910731, Society of Automotive Engineers, Warren- these control levels are thus likely to be developed in dale, Pennsylvania. the near future. Estimates of cost and emission control OTA (Office of Technology Assessment, U.S. Congress). effectiveness for light-duty diesel-fueled automobiles 1995. Advanced Automotive Technology Visions of are provided in table 3.6. a Super-Efficient Family Car OTA-ETI-638. U.S. Gov- ernment Printing Office,Washington, D.C. References Pattas, K., Z. Samaras, N. Patsatzis, C. Michalopoulous, 0. Zogou, A. Stamatellos, and M. Barkis. 1990. "On-Road ACEA/EUROPIA (European Automobile Manufacturer's Experience with Trap Oxidiser Systems Installed on Association/European Petroleum Industry Associa- Urban Buses. SAE Paper 900109, Presented at the So- tion). 1995. European Program on Emissions, Fu- ciety of Automotive Engineers (SAE) International els, and Engine Technologies. EPEFE Report, Congress and Exposition, Detroit, Michigan. Brussels. Tonkin, PR. and P. Etheridge. 1987. "A New Study of the Brogan, Jj. and S.R. Venkateswaran. 1993. "Fuel-Effi- Feasibility and Possible Impact of Reduced Emission cient Low-Emission Propulsion Technology" in To- Levels from Diesel Engined Vehicles," Report No. wards Clean & Fuel Efficient Automobiles. DP87/0927, Shoreham by Sea, Ricardo Consulting Proceedings of an International Conference held in Engineers, U.K. Berlin (March 2 5-27,1991), OECD, Paris. U.S. EPA (United States Environmental Protection Agen- CONCAWE Review. 1994."Oxidation Catalysts for Die- cy). 1990. "Volatile Organic Compounds from On- sel Cars,"Volume 3, Number 2, October, Brussels. Road Vehicles: Sources and Control Options." Report ECMT (European Conference of Ministers ofTransport). prepared for the Long-range TransboundaryAir Pollu- 1990. Transport Policy and the Environment. Or- tion Convention. United Nations Economic Commis- ganisation for Economic Cooperation and Develop- sion for Europe.Washington, D.C. (Draft) ment (OECD). Paris. Watkins, L.H. 1991. Air Pollution from Motor Vehicles. Feutlinske, H. 1989. "Possibilities for Reducing Pollutant State-of-the-Art Review/I. Transport and Road Re- Emissions from Diesel Engines-Comments on the search Laboratory. HMSO. London. Development Status of Particle Filters." Report 4, Weaver, C.S., RJ. Klausmeier, L.M. Erickson, J. Gallagh- 48th International Congress, International Union of er, and T. Hollman. 1986. "Feasibility of Retrofit Tech- Public Transport. Brussels. nology for Diesel Emissions Control." SAE Paper Holman, C. 1990. "Pollution from Diesel Vehicles." Brief- 860296, Society of Automotive Engineers, Warren- ing Document, Friends of the Earth, London. dale, Pennsylvania. 80 Air Poiution from Motor Vehicles Weaver, C.S. 1990. 'Emissions Control Strategies for Sacramento, California, Engine, Fuel, and Emissions Heavy-Duty Diesel Engines." EPA report 460/3- 90- Engineering, Inc. 001. U.S. Environmental Protection Agency. (Draft) Weaver, C.S. and L.M. Chan. 1995. "Company Archives" Ann Arbor, Michigan. Multiple Sources. Engine, Fuel and Emissions Engi- Weaver, C.S. and L.M. Chan. 1994. 'Motorcycle Emis- neering, Inc. Sacramento, California. sion Standards and Emission Control Technology." Wijetilleke, L and S. Karunaratne. 1992. "Control and Report to the Royal Thai Ministry of Science, Tech- Management of Petroleum Related Air Pollution." nology, and the Environment, and The World Bank. The World Bank,Washington, D.C. (Draft) Appendix 3. 1 Emission Control Technology for Spark-Ignition (Otto) Engines This appendix provides a more detailed technical dis- (PM), and lead (where leaded gasoline is used), as well cussion to supplement the general information on as other toxics such as benzene, 1,3 butadiene, and spark-ignition engine emissions and control technolo- formaldehyde. gies in chapter 3. Exbaust Emissions Combustion and Pollutant Formation in Exhaust emissions are caused by the combustion pro- Spark-Ignition Engines cess. Figure A3. 1.1 shows the combustion process in an Otto-cycle engine.After the initial spark, there is an igni- Gasoline engines emit carbon monoxide (CO), nitrogen tion delay while the flame kernel created by the spark oxides (NO,), hydrocarbons (HC), particulate matter grows.The flame then spreads through the combustion Figure A3.1.1 Combustion in a Spark-Ignition Engine 9 Flame Kernel Ignition growth Flame spread Late combustion Source: Weavcr and Chan 199 5 81 82 Air Pollutionfrom Motor Vebicies chamber. The rate of spread is determined by the flame speeds may be too low for combustion to be completed speed, which is a function of air-fuel ratio, temperature, during the power stroke, or combustion may not occur. and turbulence level.The increase in volume of the hot These conditions also cause high hydrocarbon emissions. burned gases behind the flame front presses the un- Unburned hydrocarbons emitted from the cylinder burned charge outward. Overall cylinder pressure in- continue to react in the exhaust only if the temperature is above 6000C and oxygen is present. So hydrocarbon creases, increasing the temperature of the unburned emissions from the tailpipe may be significantly lower charge. Finally, the remaining elements of the unburned than the hydrocarbons leaving the cylinder.This effect mixture burn out as the piston descends. The main ex- is important at stoichiometric or near-stoichiometric haust pollutants are discussed below. conditions because of the higher exhaust temperatures experienced. Nitrogen oxides.The two main nitrogen oxides emitted from combustion engines are nitric oxide (NO) and ni- Particulate matter. Unlike diesel engines, PM emissions trogen dioxide (NO2). Most nitrogen oxides from from spark-ignition engines are generally not regulated. combustion engines-90 percent-are nitric oxide. They frequently are not measured or reported, leading This gas is formed from nitrogen and free oxygen at many to assume they are negligible.This is not the case. high temperatures.The rate of formation is a function of Particulate matter emissions from four-stroke spark-igni- oxygen availability, and is exponentially dependent on tion engines result from unburned lubricating oil in the the temperature. Most nitrogen oxide emissions form exhaust and from ash-forming fuel and oil additives early in the combustion process, when the piston is such as tetra-ethyl lead.Although PM emission rates for near the top of its stroke (top-dead- center) and temper- spark-ignition engines are low compared with diesels, atures are highest. Nitrogen oxide emissions are con- emissions can be significant when poor maintenance or trolled by reducing the flame temperature (by retarding engine wear lead to high oil consumption, or when oil combustion, diluting the reacting mixture, or both) and is mixed with the fuel, as in two-stroke engines. by minimizing the time that burned gases stay at high temperatures. Toxicpollutants.Toxic chemicals emitted by spark-igni- tion engines include lead compounds, benzene, 1,3 Carbon monoxide. Carbon monoxide emissions are butadiene, and aldehydes. Lead emissions are caused by caused by the combustion of rich mixtures, where the tetra-ethyl lead, used as an octane enhancer in leaded air-fuel ratio, k is less than 1.0. In such mixtures, there gasoline. Benzene is one of the many hydrocarbons in is insufficient oxygen to convert all the carbon to car- gasoline-engine exhaust, accounting for about 4 per- bon dioxide.A small amount of carbon monoxide is also cent of total hydrocarbons. Benzene in the exhaust is emitted under lean conditions because of chemical ki- caused by fuel coming through unburned and by de- netic effects. Carbon monoxide emissions are con- alkylation of other aromatic compounds. 1,3 butadiene trolled in the engine by adjusting the air-fuel ratio of the is a product of partial hydrocarbon combustion. Both charge entering the cylinder. benzene and 1,3 butadiene are carcinogens, so expo- sure to them is cause for concern. Studies in the United Hydrocarbons. Hydrocarbon emissions result from ele- States indicate that motor vehicles are responsible for ments of the air-fuel mixture that have not finished burn- most human exposure to benzene and 1,3 butadiene ing at the time the exhaust valve opens. Hydrocarbon (U.S. EPA 1990). emissions are composed of unburned fuel and products Aldehydes are intermediate products of hydrocarbon of partial combustion, such as ethylene and formalde- combustion. Highly reactive, they form other products hyde. Hydrocarbon sources include crevice volumes, during combustion. The small amounts of aldehyde such as the space between the piston and cylinder wall emissions found in gasoline, diesel, and other engines above the piston ring, and the quenched layer immediate- using hydrocarbon fuels are caused by the quenching of ly next to the combustion chamber walls. Unburned mix- partially reacted mixture (because of contact with a ture is forced into these crevices during compression and cold surface, for instance). For fuels containing ethanol combustion, and emerges late in the expansion and dur- and methanol, the primary oxidation reactions proceed ing the exhaust stroke.This is the major source of hydro- through formaldehyde and acetaldehyde, respectively, carbon emissions from four-stroke engines. In two-stroke so these compounds are often found in significant con- engines, fuel mixing with the exhaust during scavenging centrations in the exhaust. Both formaldehyde and ace- and misfire at light loads are sources of hydrocarbons.Ab- taldehyde are irritants and suspected carcinogens. normal operation in a four-stroke engine, such as a misfir- ing cylinder, can cause significant quantities of unburned Nitrous oxide. Uncontrolled auto engines emit a few fuel to pass into the exhaust. In lean mixtures, flame milligrams per kilometer of nitrous oxide (N20). If the Emission Control Technologyfor Spark-Ignition (Otto) Engines 83 engine has a catalytic converter, increased nitrous oxide greatly reduced in sealed fuel systems, such as those used formation occurs because of the reaction of nitric oxide with fuel injection systems. Such systems may have higher and ammonia with the platinum in the catalyst. No running losses, however, due to the recirculation of hot more than 5 to 10 percent of the nitric oxide in exhaust fuel from the engine back to the fuel tank. is converted into nitrous oxide in this way. The conver- Evaporative emissions can be reduced by venting the sion in the catalyst is highly temperature dependent.As fuel tank and the carburetor to a canister containing ac- the catalyst warms up after a cold start, nitrous oxide tivated carbon.This material adsorbs volatile emissions levels increase to about 5.5 times the inlet level at from the fuel system when the engine is not running. 360°C. Emissions then decrease to the inlet level at When the engine is running, intake air is drawn through 460°C. At higher temperatures, the catalyst destroys ni- the canister, purging it of hydrocarbons, which then trous oxide instead of forming it (Prigent and Soete form part of the fuel mixture fed to the engine.A larger 1989). Nitrous oxide is thus formed primarily during canister can also be used to control refueling vapor cold starts of catalyst-equipped vehicles. emissions, or these emissions can be controlled by cap- turing them through the refueling nozzle and conduct- Crankcase Emissions ing them back to the service station tank. All piston engines experience leakage or blowby of compressed gas past the piston rings. In spark-ignition Engine Design engines, this leakage consists of unburned or partly burned air-fuel mixture and contains unburned Spark-ignition vehicle engines are either two stroke or hydrocarbons. In older vehicles, blowby gases were four stroke. The distinction is important for emissions vented to the atmosphere. Hydrocarbon emission levels since two-stroke engines emit more hydrocarbons and from the crankcase were about half the level of uncon- particulate matter than four-stroke engines of similar trolled exhaust emissions.These emissions are now con- size and power. Two-stroke engines are less fuel effi- trolled by venting the crankcase to the air intake system cient than four strokes, but have higher power output, by means of a positive crankcase ventilation (PCV) quicker acceleration, and lower manufacturing costs. valve. The unburned hydrocarbons in the blowby gas Because of their performance and cost advantages, two- are recycled to the engine and burned. The flow of air stroke engines are used extensively in motorcycles and through the crankcase also helps to prevent condensa- in small power equipment such as chainsaws and out- tion of fuel and water in the blowby gases, thus reduc- board motors. They are particularly common in small ing oil contamination and increasing engine life. One motorcycles (50 to 150 cc engine displacement),where significant drawback of the PCV system is that the recy- their poor fuel economy is of less importance to a con- cling of blowby gases tends to foul the intake manifold. sumer.They were also used in some small automobiles, An effective gasoline detergent can eliminate this prob- particularly in Eastern Europe. lem. The piston and cylinder of a typical four-stroke en- gine are shown in figure A3.1.2. Engine operation takes place in four distinct steps: intake, compression, power, Gasoline-fueled vehicles emit a significant amount of hy- and exhaust, with each step corresponding to one drocarbons as evaporative emissions from their fuel sys- stroke of the piston (180 degrees of crankshaft rota- tem. Four main sources of evaporative emissions have tion). During intake, the intake valve admits a mixture been identified: breathing (diurnal) losses from fuel of air and fuel, which is drawn into the cylinder by the tanks caused by the expansion and contraction of gas in vacuum created by the downward motion of the piston. the tank with changes in air temperature; hot-soak Figure A3.1.2 shows the piston near the end of the in- emissions from the fuel system when a warm engine is take stroke, approaching bottom-dead-center. During turned off; running losses from the fuel system during compression, the intake valve closes, and the upward vehicle operation; and resting losses from permeation motion of the piston compresses the air-fuel mixture of plastic and rubber materials in the fuel system. Refu- into the combustion chamber between the top of the eling emissions consist of gasoline vapor displaced piston and the cylinder head. from the fuel tank when it is filled. The compression stroke ends when the piston reach- Evaporative and refueling emissions are strongly affect- es top-dead-center. Just before this point, the air-fuel ed by fuel volatility. Diurnal emissions also vary depending mixture is ignited by a spark from the spark plug and be- on the daily temperature range and the amount of vapor gins to burn. Combustion of the air-fuel mixture takes space in the fuel tank. Hot-soak emissions result from the place near top-dead-center, increasing the temperature conduction of heat from the warm engine to the carbure- and pressure of the trapped gases. During the power tor, which is open to the atmosphere. Hot-soak losses are stroke, the pressure of the burned gases pushes the pis- 84 Air Pollution from Motor Vebicles Figure A3.1.2 Piston and Cylinder Arrangement of a Typical Four-Stroke Engine intake valve _ .apuexhaust valve intake port '7 exhaust port TDC BDC Intake Stroke Source: Weaver and Chan 1995 ton down, turning the crankshaft and producing power. ing charge. In the process, mixing between exhaust gas As the piston approaches bottom-dead-center again, the and the charge takes place, so some of the fresh charge exhaust valve opens, releasing the burned gases. During is also emitted in the exhaust. If fuel is already mixed the exhaust stroke the piston ascends toward top-dead- with the air in the charge (as in all current two-stroke center, pushing the remaining burned gases out the motorcycle engines), this fuel will be lost in the ex- open exhaust port.The exhaust valve then closes and haust.The loss of fuel during scavenging is one of the the intake valve opens for the next intake stroke. main causes of high hydrocarbon emissions from two- In a four-stroke engine, combustion and the resulting stroke motorcycle engines. power stroke occur once every two revolutions of the The other reason for high hydrocarbon emissions crankshaft. In a two-stroke engine, combustion occurs from two strokes is their tendency to misfire under low- in every revolution of the crankshaft. Two-stroke load conditions. At low loads, the amount of fresh engines eliminate the intake and exhaust strokes, leav- charge available to scavenge the burned gases from the ing only the compression and power strokes. Rather cylinder is small, so a significant amount of the burned than occupying distinct phases of the cycle, exhaust gas remains in the cylinder to dilute the incoming and intake occur simultaneously (figure A3.1.3). As the charge. The resulting mixture of air, fuel, and exhaust piston approaches bottom-dead-center in the power burns less readily and is more difficult to ignite than an stroke, it uncovers exhaust ports in the wall of the cylin- air-fuel mixture alone.At light loads, the mixture some- der. The high-pressure combustion gases blow into the times fails to ignite, allowing the fuel vapor in the cylin- exhaust manifold. As the piston gets closer to the bot- der to pass unburned into the exhaust.These occasional tom of its stroke, intake ports are uncovered, and fresh misfires are the cause of the popping sound made by air-fuel mixture is forced into the cylinder while the ex- two-stroke engines under light-load conditions. haust ports are still open. Exhaust gas is scavenged The mechanical systems required by four-stroke en- (forced) from the cylinder by the pressure of the incom- gines to open and close their intake and exhaust valves Emission Control Technology for Spark-Ignition (Otto) Engines 85 Figure A3.1.3 Exhaust Scavenging in a Two-Stroke Gasoline Engine , spark plug some fresh _ mixture leaks fresh P r > ~ into exhaust air/fuel mixture : I exhaust ' g T ' <_ As above + low oil consumption Asabove+trap * 0- 1 I I 2 4 6 8 10 10.7 NO(gfbhp.hr) Note: Current engines improve significantly on these trends meeting U.S. 1998 standards without a trap. Source: Adapted from Latham and Tomkin 1988 tion chamber, and thus play an important role in air-fuel intake port. These engines typically use moderate-to- mixing and emissions. A number of different combus- high injection pressures, and three to five spray holes tion chamber designs are used in diesel engines.Virtual- per nozzle. Low-swirl engines rely primarily on the fuel ly all commercial diesel engines make use of one of the injection process to supply the mixing. They typically common combustion chamber designs shown in figure have very high fuel injection pressures and six to nine A3.2.5.The most fundamental difference between the spray holes per nozzle. Wall-wetting DI engines also combustion chambers is between the indirect-injection have fairly high swirl, but the injection system is (IDI) and direct-injection (Dl) designs. In an indirect-in- designed to deposit the fuel on the combustion cham- jection engine, fuel is injected into a separate 'precham- ber wall, where it vaporizes and burns relatively slowly. ber,' where it mixes and partly burns before jetting into In the IDI engine, the mixing between air and fuel is the main combustion chamber above the piston. In the driven primarily by air swirl induced in the prechamber more common direct-injection engine, fuel is injected as air is forced into it during compression, and by the tur- directly into a combustion chamber hollowed out of the bulence induced by the expansion out of the precham- top of the piston. DI engines can be further divided into ber during combustion. These engines typically have high-swirl, low-swirl (quiescent chamber), and wall- better high-speed performance than DI engines, and can wetting designs. The latter has many characteristics in use cheaper fuel injection systems. For this reason, they common with indirect-injection systems. have been used extensively in passenger car and light Fuel-air mixing in the direct-injection engine is lim- truck applications. Historically, IDI diesel engines have ited by the fuel injection pressure and any motion im- also exhibited lower emission levels than DI engines. parted to the air entering the chamber. In high-swirl DI With recent developments in DI engine emission con- engines, a strong swirling motion is imparted to the air trols, however, this is no longer the case. Disadvantages entering the combustion chamber by the design of the of the IDI engine are the extra heat and frictional losses 110 Air Pollutionfrom Motor Vebicles Figure A3.2.5 Diesel Engine Combustion Chamber Types Direct injection Indirect injection low swirl Direct injection Direct injection high swirl w s w s wall-wetting Source: Weaver and Chan 1995 due to the prechamber.These result in a 5 to 15 percent combustion. To reduce these volumes, engine makers reduction in fuel efficiency compared to a Dl engine. Be- have reduced the clearance between the piston and cyl- cause of this, nearly all heavy-duty truck engines are of inder head through tighter production tolerances, and the DI type, and there is an increasing trend toward DI have moved the top compression ring toward the top of engines in passenger cars and light trucks. the piston. This increases the temperature of the top ring and poses design problems for the piston top and DI Combustion Chamber Design cooling system. These problems have been addressed Changes in the engine combustion chamber and related through redesign and the use of more expensive mate- areas have demonstrated a major potential for emission rials. The higher piston ring temperature also requires control. Design changes to reduce the crevice volume higher quality engine oil to avoid formation of damag- in DI diesel cylinders increase the amount of air avail- ing deposits. able in the combustion chamber. Changes in combus- tion chamber geometry -such as the use of a reentrant Combustion chamber shape. Numerous test results in- lip on the piston bowl-can markedly reduce emissions dicate that, for high swirl DI engines, a reentrant com- by improving air-fuel mixing and minimizing wall im- bustion chamber shape (in which the lip of the pingement by the fuel jet. Optimizing the intake port combustion chamber protrudes beyond the walls of the shape for best swirl characteristics has also yielded bowl) provides a substantial improvement in perfor- significant benefits.A number of light- and medium-duty mance and emissions over the previous straight-sided engines now incorporate variable air intake systems to bowl designs. Nearly all of the manufacturers of high- optimize swirl characteristics across a broader range of swirl engines are developing or using this approach. engine speeds. Similar improvements in the performance of low-swirl DI engines have come primarily through refinements to Crevice volume.The crevice volume includes the clear- the classic 'mexican hat' combustion chamber shape. ance between the top of the piston and the cylinder head, and the 'top land'-the space between the side of Intake air swirl. Optimal matching of intake air swirl the piston and the cylinder wall above the top com- with combustion chamber shape and other variables is pression ring.The air in these spaces contributes little critical for emissions control in high-swirl engines. to the combustion process.The smaller the crevice vol- Swirl is determined mostly by the design of the air in- ume, the larger the combustion chamber volume can be take port. Unfortunately, the selection of a fixed swirl for a given compression ratio. Thus, reducing the crev- level involves some tradeoffs between low-speed and ice volume increases the amount of air available for high-speed performance. At low speeds, higher swirl Emission Control Technologyfor Compression Ignition (Diesel) Engines 111 provides better mixing, permitting more fuel to be in- the crankshaft and connected to individual injection jected and thus greater torque output at the same nozzles at the top of each cylinder by special high-pres- smoke level. However, this can result in too much swirl sure fuel lines. These pump-line-nozzle injection sys- at higher speeds, impairing the airflow to the cylinder. tems can be further divided into two subclasses: Attaining an optimal swirl level is most difficult in "distributor" fuel pumps, in which a single pumping el- smaller light-duty and medium-duty DI engines, as these ement is mechanically switched to connect to the high- experience a wider range of engine speeds than do pressure fuel lines for each cylinder in turn, and"in-line" heavy-duty engines. One solution to this problem is to pumps having one pumping element per cylinder, each vary the swirl ratio as a function of engine speed. A connected to its own high-pressure fuel line. Distribu- number of production light- and medium-duty engines tor-type systems are less costly and are commonly used now use this approach with a noticeable reduction in in light- duty engines. In-line fuel injection pumps have PM and NO, emissions (Shimada, Sakai, and Kurihara better durability and can reach higher injection pres- 1986). sures.They are much more common in engines used for heavy-duty vehicles. Fuel Injection The most common alternative to the pump-line noz- The fuel injection system in a diesel engine includes the zle injection systems are systems using unit injectors, in machinery by which the fuel is transferred from the fuel which the individual fuel metering and pumping ele- tank to the engine, then injected into the cylinders at ment for each cylinder is combined in the same unit the right time for optimal combustion, and in the cor- with the injection nozzle at the top of the cylinder.The rect amount to provide the desired power output.The pumping elements in a unit injector system are general- quality, quantity, and timing of fuel injection determine ly driven by the engine camshaft. the engine's power, fuel economy, and emission charac- Worldwide, many more engines are made with teristics so that the fuel injection system is one of the pump-line-nozzle injection systems than with unit injec- most important components of the engine. tors.This is primarily due to the higher cost of unit in- The fuel injection system normally consists of a low- jector systems. Due to the absence of high- pressure pressure pump to transfer fuel from the tank to the sys- fuel lines, however, unit injectors are capable of higher tem, one or more high-pressure fuel pumps to create injection pressures than pump-line-nozzle systems. the pressure pulses that actually send the fuel into the With improvements in electronic control, these systems cylinder, the injection nozzles through which fuel is in- offer better fuel economy at low emission levels than jected into the cylinder, and a governor and fuel meter- the pump-line-nozzle systems. For this reason, most of ing system. These determine how much fuel is to be the new heavy-duty engines produced in the United injected on each stroke, and thus the power output of States are now equipped with unit injectors (Balek and the engine. Heitzman 1993). The major areas of concentration in fuel injection A new type of fuel injection system has recently been system development have been on increased injection introduced.This system uses electro-hydraulic actuators pressure, increasingly flexible control of injection tim- supplied with fuel from a "common rail" instead of me- ing, and more precise governing of the fuel quantity in- chanically driven pumping elements. These systems al- jected. Systems offering electronic control of these low great flexibility in control of fuel injection rate and quantities, as well as fuel injection rate, have been intro- timing. duced. Some manufacturers are also pursuing technolo- gy to vary the rate of fuel injection over the injection Fuel injection pressure and injection rate. High fuel period, in order to reduce the amount of fuel burning in injection pressures are desirable to improve fuel the premixed combustion phase. Reductions in NO, atomization and fuel-air mixing, and to offset the effects and noise emissions and maximum cylinder pressures of retarded injection timing by increasing the injection have been demonstrated using this approach (Gill rate.A number of studies have been published on the ef- 1988). Other changes have been made to the injection fects of higher injection pressures on PM or smoke emis- nozzles themselves, to reduce or eliminate sac volume sions.AII show marked reductions as injection pressure and to optimize the nozzle hole size and shape, number is increased. High injection pressures are most impor- of holes, and spray angle for minimum emissions. tant in low-swirl, direct-injection engines, since the fuel injection system is responsible for most of the fuel-air Injection system types.The most common diesel fuel in- mixing in these systems. For this reason, low-swirl en- jection system consists of a single fuel pump (typically gines tend to use unit injector systems, which can mounted at the side of the engine) driven by gears from achieve peak injection pressures in excess of 1,500 bar. 112 Air Polution from Motor Vehicles The injection pressures achievable in pump-line- quires varying the rate of injection during the injection nozzle fuel injection systems are limited by the mechan- stroke. This represents a difficult design problem for ical strength of the pumps and fuel lines, as well as by mechanical injection systems, but should be possible pressure wave effects, to about 800 bar. Improvements using electro-hydraulic injectors. Another approach to in system design to minimize pressure wave effects, and the same end is split injection, in which a small amount increases in the size and mechanical strength of the lines of fuel is injected in a separate event ahead of the main and pumping elements have increased the injection fuel injection period. Experimental data show a marked pressures achievable in pump-line-nozzle systems sub- beneficial effect from reducing the initial rate of injec- stantially from those achievable a few years ago. tion (Gill 1988). A substantial reduction in NO, emis- The pumping elements in nearly all present fuel in- sions could be achieved through this technique with no jection systems are driven by a fixed mechanical linkage significant adverse impacts on HC or PM emissions.As from the engine crankshaft.This means that the pump- an additional benefit, engine noise and maximum cylin- ing rate, and thus the injection pressure, are strong der pressures for a given power output are reduced. functions of engine speed.At high speeds, the pumping element moves rapidly, and injection pressures and in- Low sac/sacless nozzles.The nozzle sac is a small inter- jection rates are high.At lower speeds, however, the in- nal space in the tip of the injection nozzle.The nozzle or- jection rate is proportionately lower, and injection ifices open into the sac so that fuel flowing past the pressure drops off rapidly.This can result in poor atom- needle valve first enters the sac and then sprays out of ization and mixing at low speeds, and is a major cause the orifices. The small amount of fuel remaining in the of high smoke emissions during lugdown. Increasing sac tends to burn or evaporate late in the combustion cy- the pump rate to provide adequate pressure at low cle, resulting in significant PM and HC emissions.The sac speeds is impractical, as this would exceed the system volume can be minimized or eliminated by redesigning pressure limits at high speed. the injector nozzle. One manufacturer reported nearly a A new type of in-line injection pump provides a par- 30 percent reduction in PM emissions through elimina- tial solution to this problem (Ishida, Kanamoto, and tion of the nozzle sac. It is also possible to retain some Kurihara 1986).The cam driving the pumping elements of the sac while designing the injector nozzle so that the in this pump has a non-uniform rise rate, so that pump- tip of the needle valve covers the injection orifices when ing rate at any given time is a function of the cam angle. it is closed. This valve-covers-orifice (VCO) injector de- By electronically adjusting a spill sleeve, it is possible to sign is now common on production engines designed select the portion of the cam's rotation during which for North American emissions standards. fuel is injected, and thus to vary the injection rate. Injec- tion timing varies at the same time, but the system is de- Engine Control Systems signed so that desired injection rate and injection Traditionally, diesel engine control systems have been timing correspond fairly well. A 25 percent reduction closely integrated with the fuel injection system, and in PM emissions and a 10 percent reduction in HC emis- the two systems are often discussed together.These ear- sions has been obtained using this system, with virtually lier control systems (still in use on many engines) are no increase in NO, emissions. Fuel injection pumps in- entirely mechanical. In recent years an increasing num- corporating this technology are now used in a number ber of computerized electronic control systems have of production engines. The same approach has also been introduced in diesel engines. With the introduc- been applied to unit injector systems, using an electron- tion of these systems, the scope of the engine control ically controlled spill valve. system has been greatly expanded. Another approach to increasing injection pressure at low engine speeds is the use of electro-hydraulic"com- Mechanical controls. Most current diesel engines still mon rail" injection systems. Through appropriate de- rely on mechanical engine control systems. The basic sign and control schemes, such systems can control and functions of these systems include basic fuel metering, maintain fuel injection pressures nearly independently engine speed governing, maximum power limitation, of engine speed (Hower and others 1993). torque curve 'shaping", limiting smoke emissions dur- ing transient acceleration, and (sometimes) limited con- Initial injection rate and premixed burning. Reducing trol of fuel injection timing. Engine speed governing is the amount of fuel burned in the premixed combustion accomplished through a spring and flyweight system phase can significantly reduce total NO, emissions.This which progressively (and quickly) reduces the maxi- can be achieved by reducing the initial rate of injection, mum fuel quantity as engine speed exceeds the rated while keeping the subsequent rate of injection high to value.The maximum fuel quantity itself is generally set avoid high PM emissions due to late burning. This re- through a simple mechanical stop on the rack control- Emission Contn)l Technologyfor Compression Ignition (Diesel) Engines 113 ling injection quantity. More sophisticated systems al- electronically controlled transmission also help to re- low some "shaping" of the torque curve to change the duce fuel consumption, and will thus likely reduce in- maximum fuel quantity as a function of engine speed. use emissions. Since the effect of these technologies is Acceleration smoke limiters are needed to prevent to reduce the amount of engine work necessary per ki- excessive black smoke emissions during transient accel- lometer, rather than the amount of pollution per unit of eration of turbocharged engines. Most are designed to work, these effects are not necessarily reflected in dyna- limit the maximum fuel quantity injected as a function mometer emissions test results. of turbocharger boost, so that full engine power is de- veloped only after the turbocharger comes up to speed. Turbocharging and Intercooling Many pump-line-nozzle fuel injection systems incor- A turbocharger consists of a centrifugal air compressor porate mechanical injection timing controls. Since the mounted on the same shaft as an exhaust gas turbine. injection pump is driven by a special shaft geared to the By increasing the mass of air in the cylinder prior to crankshaft, injection timing can be adjusted within a compression, turbocharging correspondingly increases limited range by varying the phase angle between the the amount of fuel that can be burned without exces- two shafts, using a sliding spline coupling.A mechanical sive smoke, and thus increases the potential maximum or hydraulic linkage slides the coupling back and forth power output.The fuel efficiency of the engine is im- in response to engine speed and load signals. proved as well. The process of compressing the air, In mechanical unit injector systems, the injectors are however, increases its temperature, increasing the ther- driven by a direct mechanical linkage from the cam- mal load on critical engine components. By cooling the shaft, making it very difficult to vary the injection tim- compressed air in an intercooler before it enters the cyl- ing. Formerly, some California engines with unit inder, the adverse thermal effects can be reduced.This injectors used a mechanical timing control that operat- also increases the density of the air, allowing an even ed by moving the injector cam followers back and forth greater mass of air to be confined within the cylinder, with respect to the cam. Although effective in limiting and thus further increasing the maximum power poten- light load hydrocarbon and PM emissions, these sys- tial. Turbocharging and intercooling offer an inexpen- tems have proved to be troublesome and unpopular sive means to simultaneously improve power-weight among users. ratios, fuel economy, and control of NO, and PM emis- sions (OECD/IEA 1993). Electronic controls. The advent of computerized elec- Increasing the air mass in the cylinder and reducing tronic engine control systems has greatly increased the its temperature can reduce both NO. and PM emissions potential flexibility and precision of fuel metering and as well as increase fuel economy and power from a given injection timing controls. In addition, it has made possi- engine displacement. Most heavy-duty diesel engines are ble whole new classes of control function, such as road equipped with turbochargers, and most of these have in- speed governing, alterations in control strategy during tercoolers. In the United States virtually all engines were transients, synchronous idle speed control, and adap- equipped with these systems by 1991, and in Europe tive learning, including strategies to identify and com- turbocharging and aftercooling are expected to become pensate for the effects of wear and component-to- standard feature on all heavy-duty engines. Recent devel- component variation in the fuel injection system. opments in air charging systems for diesel engines have By continuously adjusting the fuel injection timing to been primarily concerned with increasing the turbo- match a stored 'map" of optimal timing vs. speed and charger efficiency, operating range, and transient re- load, an electronic timing control system can signifi- sponse characteristics; and with improved intercoolers cantly improve on the NOX/PM emissions and NO, to further reduce the temperature of the intake charge. emission/fuel consumption tradeoffs possible with stat- Tuned intake air manifolds (including some with vari- ic or mechanically-variable injection timing. Most elec- able tuning) have also been developed to maximize air tronic control systems also incorporate the functions of intake efficiency in a given speed range. the engine governor and the transient smoke limiter. This helps to reduce excess PM emissions due to me- Turbocharger refinements. Turbochargers for heavy- chanical friction and lag time during engine transients, duty diesel engines are already highly developed, but ef- while simultaneously improving engine performance. forts to improve their performance continue.The major Potential reductions in PM emissions of up to 40 per- areas of emphasis are improved matching of turbo- cent have been demonstrated with this approach (Wade charger response characteristics to engine require- and others 1987). ments, improved transient response, and higher Other electronic control features, such as cruise con- efficiencies. Engine/turbocharger matching is especial- trol, upshift indication, and communication with an ly critical because of the inherent conflict between the 114 Ar PAllutionfrym Motor Vehicles response characteristics of the two types of machines. The most common type of low-temperature charge- Engine boost pressure requirements are greatest near air cooler rejects heat directly to the atmosphere the maximum torque speed, and most turbochargers through an air-to-air heat exchanger mounted on the are matched to give near optimal performance at that truck chassis in front of the radiator. Although bulky and point.At higher speeds, however, the exhaust flow rate expensive these charge air coolers are able to achieve is greater, and the turbine power output is correspond- the lowest charge air temperatures-in many cases, only ingly higher. Boost pressure under these circumstances 10 or 15°C above ambient. An alternative approach is can exceed the engine's design limits, and the excessive low- temperature air to water intercooling, in which the turbine backpressure increases fuel consumption.Thus basic water-air intercooler is retained, but with drastical- some compromise between adequate low speed boost ly reduced radiator flow rates to lower the water temper- and excessive high speed boost must be made. ature.This water is then passed through the intercooler before it is used for cooling the rest of the engine. Variable geometry turbocharger. Because of the inher- ent mismatch between engine response characteristics Intake manifold tuning. Tuned intake manifolds have and those of a fixed geometry turbocharger, a number of been used for many years to enhance airflow rates on engine manufacturers are considering the use of variable high performance gasoline engines, and are now used geometry turbines instead (Wallace and others 1986). In on some diesel engines as well A tuned manifold pro- these systems the turbine nozzles can be adjusted to vides improved airflow and volumetric efficiency at vary the turbine pressure drop and power level in order speeds near its resonant frequency at the cost of re- to match the engine's boost pressure requirements. duced volumetric efficiency at other speeds.A variable- Thus, high boost pressures can be achieved at low en- resonance manifold has been shown to improve airflow gine speeds without wa