UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A multiple-objectives approach to address motorized two-wheeled vehicle emissions in Delhi, India Badami, Madhav Govind 2001

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2001-610578.pdf [ 19.03MB ]
Metadata
JSON: 831-1.0099650.json
JSON-LD: 831-1.0099650-ld.json
RDF/XML (Pretty): 831-1.0099650-rdf.xml
RDF/JSON: 831-1.0099650-rdf.json
Turtle: 831-1.0099650-turtle.txt
N-Triples: 831-1.0099650-rdf-ntriples.txt
Original Record: 831-1.0099650-source.json
Full Text
831-1.0099650-fulltext.txt
Citation
831-1.0099650.ris

Full Text

A MULTIPLE-OBJECTIVES A P P R O A C H TO ADDRESS M O T O R I Z E D TWO-WHEELED V E H I C L E EMISSIONS I N D E L H I , INDIA  by  M A D H A V GOVIND B A D A M I  B. Tech. (Mech. Engg.), Indian Institute of Technology, Madras, 1977 M . S. (Mech. Engg.), Indian Institute of Technology, Madras, 1981 M . E . Des. (Environmental Science), The University of Calgary, 1994  A THESIS SUBMITTED IN P A R T I A L F U L F I L M E N T OF THE REQUIREMENTS FOR THE D E G R E E OF DOCTOR OF PHILOSOPHY in T H E F A C U L T Y OF G R A D U A T E STUDIES (School of Community and Regional Planning) We accept this thesis as conforming to the required standard  T H E UNIVERSITY OF BRITISH C O L U M B I A April 2001 © Madhav Govind Badami, 2001  UBC Special Collections - Thesis Authorisation Form  http://www.library.ubc.ca/spcoll/thesauth.html  In p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f t h e requirements f o r an advanced degree a t the U n i v e r s i t y o f B r i t i s h Columbia, I agree t h a t t h e L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and study. I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y purposes may be g r a n t e d by the head o f my department o r by h i s o r h e r r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t copying o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l gain s h a l l not be a l l o w e d without my w r i t t e n p e r m i s s i o n .  The U n i v e r s i t y o f B r i t i s h Columbia Vancouver, Canada  1 of 1  4/26/01 2:10 PM  11  ABSTRACT  Motor vehicle activity is growing and air quality is deteriorating rapidly in Indian and other less-industrialized country cities. The contribution of transport to air pollution is increasing. Motorized two-wheeled (M2W) vehicles, mainly powered by two-stroke engines, account for a significant proportion of motor vehicle activity and emissions. These issues are important because, in addition to local health and welfare impacts, they have important implications for energy security, acidification and climate change. The challenge in terms of M 2 W vehicles is to address their emissions while minimizing adverse policy impacts for vehicle users, since these vehicles provide mobility to millions. The dissertation illuminates key aspects of the M 2 W vehicle air pollution problem, and reports on policy-relevant research related to M2W vehicle emissions in Delhi. It investigates contributory factors and the institutional setting, and explores the policy implications of critical vehicle user choices and perspectives. The dissertation proposes an analytic framework for effective policy-making and implementation, and multiple objectives and measures to characterize the impacts of policy alternatives. Information sources include published and unpublished literature on various aspects of the problem, discussions with decision makers, industry representatives and researchers, and a questionnaire survey of, and in-depth interviews with, M2W vehicle users. The dissertation demonstrates the importance of considering system-wide emissions due to vehicle activity, technology-human behaviour-political institution interactions, in-use realities and institutional constraints, and implementation issues including how vehicle users are affected by and respond to policies. In addition to these issues, the policy-analytic framework incorporates a wide range of policy impacts, and the concerns of various actors and affected groups, to address transport air pollution effectively and equitably over the long term. It is argued that policy-making and implementation should be adaptive and flexible, and promote continual learning, for policy effectiveness. While considering implementation issues will lead to robust policies, policies that minimize reliance on expensive technologies and institutional mechanisms, and that are impervious to in-use realities and constraints, should be implemented. Since technological measures can be neutralized over time, and given multiple  Ill  transport impacts and constrained resources, the aim should be to achieve transport synergies, in addition to improving air quality.  Keywords: Urban Transport, Air Quality, Less-Industrialized Countries, Motorized twowheeled vehicles, Policy Making and Implementation, Objectives.  Human Dimensions, Multiple  iv TABLE OF CONTENTS Abstract  ii  Table of Contents  iv  List of Tables  ix  List of Figures  x  Glossary  xiii  Dedication  xvii  Acknowledgments CHAPTER I ,  xviii Introduction  1  1.1  Background, and the Public Policy Challenge  1  1.2  Research Objectives and Approach 1.2.1 Contributory Factors 1.2.2 The Institutional Setting 1.2.3 Policy-Analytic Framework 1.2.4 Vehicle User Choices and Perspectives 1.2.5 Multiple Policy Objectives Level of Analysis Methodology Structure of Dissertation  3 5 5 6 7 7 9 10 11  1.3 1.4 1.5 C H A P T E R II  The Transport Air Pollution Problem in Delhi, India  13  2.1 2.2 2.3  13 13 18 18 26  2.4 2.5  Introduction Urbanization and Motor Vehicle Activity in Delhi Air Pollution and Health Effects in Delhi 2.3.1 Air Pollution in Delhi 2.3.2 Air Pollutants and their Health Effects 2.3.3 Studies Linking Air Pollution and Health and Economic Costs in Delhi The Role of Transport in Delhi's Air Pollution Delhi' s Motor Vehicle Activity and Air Pollution in a Wider Context 2.5.1 Motor Vehicle Activity and Urban Air Pollution in India and Other LICs 2.5.2 Transport, Energy Security and Climate Change  31 33 39 40 42  2.5.3  2.6 C H A P T E R III  Regional Acidification and Tropospheric Ozone 2.5.4 Other Impacts of Motor Vehicle Activity and Urbanization Summary and Conclusions  Transport Air Pollution In India: A Discussion Of Contributory Factors 3.1 3.2 3.3  Introduction Methodology Factors Contributing to Motorized Two-wheeled Vehicle Air Pollutant Emissions in the Indian Context 3.3.1 Vehicle Technology 3.3.2 Vehicle Operation, Maintenance, and Disposal 3.3.3 Ineffective Monitoring and Enforcement 3.3.4 Congestion, and Road Availability and Condition 3.3.5 Fuel and Lubricating Oil Quality 3.3.6 Fuel and Oil Pricing and Adulteration 3.3.7 Motor Vehicle Activity 3.3.8 The Role of Government Policy Conclusions ;  3.4 CHAPTER IV  45 46 49  52 52 53  54 54 60 62 62 64 68 69 73 80  Transport Air Pollution in India: A Discussion of the Institutional Setting  88  4.1  88  4.2 4.3 4.4 4.5  Introduction 4.1.1 The Urban Challenge in Asian LICs, and the Role of Institutional Factors 4.1.2 Chapter Objectives and Outline Methodology Actors, Responsibilities and Roles Institutional Constraints Policy-Making and Implementation: Actors' Interactions 4.5.1 Vehicle Emission Standards and Fuel Quality Improvements 4.5.2 In-use Vehicle Emissions Monitoring and Control 4.5.3 Vehicle Scrappage 4.5.4 Public Interest Litigation and the Supreme Court  88 90 91 92 97 106 107 Ill 113 113  4.6 CHAPTER V  A Policy Analytic Framework For Prevention And Control Of Air Pollutant Emissions From Motorized Two-wheeled Vehicles In The Indian Context 5.1 5.2 5.3 5.4  5.5  5.6 5.7 CHAPTER VI  4.5.5 NGOs and the Media Conclusions  Introduction Methodology Criteria for Good Policy Analysis 5.3.1 Important Contextual Characteristics Applying Systematic and Value-Focused Thinking to Motorized Two-wheeled Vehicle Air Pollution in the Indian Context 5.4.1 Prioritizing Air Pollution Impacts and Air Pollutants 5.4.2 Prioritizing Motorized Two-wheeled Vehicle and Transport System Sources 5.4.3 Policy Options Analytical Framework for Evaluating Policies Targeted at Motorized Two-wheeled Vehicle Air Pollutant Emissions in the Indian Context 5.5.1 Determination of Motor Vehicle Activity and Emissions 5.5.2 Determination and Valuation of Emissions Impacts The Multiple-Objectives Approach Conclusions  Motorized Two-wheeled Vehicle User Choices And Perspectives: Implications For Air Pollution Prevention And Control 6.1  6.2  6.3  Introduction 6.1.1 Rationale and Objectives 6.1.2 Research Questions Methodology 6.2.1 Research Instruments 6.2.2 Study Participants, Household Characteristics and Vehicle Ownership 6.2.3 Implementation of the Study 6.2.4 Analysis and Discussion of Study Results Results and Discussion 6.3.1 Factors Contributing to Choice of M2W Vehicle as Travel Mode  118 122  125 125 126 127 128  131 131 135 138  139 141 148 153 159  161 161 161 162 163 163 165 170 172 173 173  vii 6.3.2 6.3.3 6.3.4 6.3.5 6.3.6  6.4 C H A P T E R VII  Structuring Multiple Objectives And Measures For Policies To Prevent And Control Motorized Two-wheeled Vehicle Air Pollutant Emissions In The Indian Context 7.1 7.2  7.3 7.4 7.5 7.6 C H A P T E R VIII  Daily Travel Activity M2W Vehicle Purchase M2W Vehicle Operation and Maintenance M2W Vehicle Disposal and Replacement Participant Perspectives on Transport Air Pollution Prevention and Control Policies 6.3.6.1 Current Policies 6.3.6.2 Vehicle Technology Improvements 6.3.6.3 Fuel/Oil Quality Improvements 6.3.6.4 Mandated Vehicle Scrappage ... 6.3.6.5 Ranking of Policy Alternatives 6.3.6.6 Policy Impacts, Measures to Enhance Policy Attractiveness, and Desirable Policy Characteristics 6.3.6.7 Public Transit and Cycling Conclusions  Introduction 7.1.1 Chapter Objectives and Outline Methodology 7.2.1 The Multiple-Obj ectives Approach for Eliciting, Clarifying and Structuring Public Values 7.2.2 Methodology Used in the Present Study Structuring Multiple Objectives Mean-Ends Objectives Network Measures Conclusions  181 185 192 195 197 201 202 205 208 209  211 215 219  220 220 221 222  222 225 227 233 236 238  Conclusions  242  8.1 8.2  242  8.3 8.4 8.5 8.6 8.7  Introduction Technology-Political InstitutionHuman Behaviour Interactions The Institutional Setting Vehicle User Perspectives Policy-Analytic Framework Institutional Reform and Capacity Building Prevention is Better than Cure  243 244 245 248 250 254  viii  8.8  Suggestions for Further Research  References Appendix I  256 261  Comparison of Indian National Ambient Quality Standards (NAAQS) with WHO/US/California Ambient Air Quality Standards  281  Appendix II  Air Pollutants and their Health Effects  285  Appendix III  Indian Motor Vehicular Emission Standards since 1991  290  Appendix IV  Indian Motor Gasoline Specifications (Current and Proposed for 2000) - Comparison of Key Characteristics  292  List of Interviewees in Government Agencies, Industry, and Research Institutions  293  Informed Consent Form for Discussions with Government Decision Makers, Experts and Vehicle and Fuel Industry Representatives  294  Appendix VII  Proposed Rules for Vehicle Scrappage/Phase-out in Delhi ....  297  Appendix VIII  M 2 W Vehicle User Survey Informed Consent Form  300  Appendix I X  M 2 W Vehicle User Survey Questionnaire  304  Appendix X  M 2 W Vehicle User Interview Protocol  313  Appendix X I  M 2 W Vehicle User Survey Supplementary Questionnaire  318  Appendix V  Appendix V I  ix  LIST OF TABLES Table 2.1 Table 2.2  Table 3.1 Table 3.2 Table 3.3  Table 5.1 Table 5.2  Table 6.1 Table 6.2  In-use Indian Motor Vehicular Emission Factors - Modal Comparison Emission Factors for In-use Indian Cars and Buses, Compared to M2W Vehicle In-use Indian M 2 W Vehicle Fuel Economy and Exhaust Emission Factors Compared to Europe, Early 1990s Evolution of Indian M2W Vehicle Exhaust Emission Standards Compared to Taiwan Quality of Indian Gasoline Compared to California and Europe Policy Instruments for Prevention and Control of Transport Air Pollution Approximate Toxicity Weighting Factors for Selected Pollutants  37 38  57 58 67  140 154 170  Table 6.10  Vehicles Owned by Participants'Households Breakdown of Participant Responses to Study Instruments Questions Related to Mode Choice Questions Related to Daily Travel Activity Questions Related to M2W Vehicle Purchase Choices and Motivations Questions Related to M2W Vehicle Operation and Maintenance Questions Related to M2W Vehicle Disposal and Replacement Questions Related to Perspectives on Policies Participants' WTP for M2W Vehicle Technology Improvements Participants' WTP for Fuel-Oil Quality Improvements  Table 7.1 Table 7.2  Overall Fundamental Objectives Hierarchy Measures for Characterization of Policy Impacts  230 239  Table 6.3 Table 6.4 Table 6.5 Table 6.6 Table 6.7 Table 6.8 Table 6.9  171 175 182 185 190 195 198 203 206  X  LIST OF FIGURES Figure 2.1 Figure 2.2 Figure 2.3 Figure 2.4 Figure 2.5 Figure 2.6 Figure 2.7 Figure 2.8 Figure 2.9 . Figure 2.10 Figure 2.11 Figure 2.12 Figure 2.13 Figure 2.14 Figure 2.15 Figure 2.16 Figure 2.17 Figure 3.1 Figure 3.2 Figure 3.3  Population Growth, Delhi Urban Area Motor Vehicle Growth in Delhi, 1971-1996 Population and Motor Vehicle Growth Rates in Delhi, 1941-2001 Delhi's Motor Vehicles Compared to Other Major Indian Cities, 1994 Annual Average Sulphur Dioxide Levels in Delhi, 1978-1996 Annual Average Nitrogen Dioxide Levels in Delhi, 1978-1996 Annual Average Particulate Levels in Delhi, 1978-1996 Peak 24-Hour Sulphur and Nitrogen Dioxide Levels at Various Sites in Delhi, 1994 Peak 24-Hour SPM Levels at Various Sites in Delhi, 1994 8-Hour Mean and 1-Hour Maximum Ozone Levels in Delhi, 1997 Transport Share of Air Pollutant Emissions in Delhi - Data from NEERI and CPCB, 1981-1996 Air Pollutant Emissions Inventory for D e l h i - C P C B , 1996 Motor Vehicular Exhaust Emissions in Delhi - M o d a l Shares Passenger Motor Vehicular V K M and P K M in Delhi - Modal Comparison Motor Vehicle Growth in India, 1961 -1994 Petroleum Products Consumption in India, 1970-1995 Mechanical Mode Share Changes in Delhi, 1957-1994  15 15 16 16 22 22 23 23 24 2,4 36 36 37 38 44 44 48  Door-to-door Journey Times by Different Modes in Delhi .... Factors Influencing Transport Emissions Factors Influencing the Extent and Nature of Motor Vehicle Activity  72 86  Figure 4.1  Actors, Roles and Responsibilities  99  Figure 5.1  Percentage of Residential Households in Delhi Owning At Least One M 2 W Vehicle, by Income Group Distribution of M2W Vehicle Owning Households and M 2 W Vehicle Trips by Household Income in Residential Households in Delhi Distribution of All Trips and M 2 W Vehicle Trips by Residential Households in Delhi, by Purpose  Figure 5.2  Figure 5.3  87  132  132 133  xi  Figure 5.4 Figure 5.5  Figure 5.6 Figure 5.7  Figure 6.1 Figure 6.2 Figure 6.3 Figure 6.4 Figure 6.5 Figure Figure Figure Figure Figure  6.6 6.7 6.8 6.9 6.10  Figure 6.11 Figure 6.12 Figure 6.13 Figure 6.14 Figure 6.15 Figure 6.16 Figure 6.17 Figure 6.18 Figure Figure Figure Figure Figure  6.19 6.20 6.21 6.22 6.23  Distribution of Trips for All Purposes and Work Trips by Mode in Residential Households in Delhi Percentage of All Trips and Work Trips Performed on M 2 W Vehicles by Residential Households in Delhi, by Income Group Policy-Analytic Framework for Evaluating Policies Targeted at Motor Vehicle Emissions Analytical Framework for Evaluation of Policies Targeted at M 2 W Vehicle Air Pollutant Emissions Gender and Age Distribution of Participants Distribution of Participants by Household Size and Earners Per Household Distribution of Participants by Household Income Distribution of Survey Households by Household Income in RITES/ORG (1994) Distribution of Participants by Per Capita Household Income Participants'Journey-to-Work Trip Lengths Participants'Opinions and Use of Bus Service Participants' Opinions and Use of Bicycle Commuting Participants'Involvement in Road Accidents Rating of Factors Influencing Choice of M 2 W Vehicle as Travel Mode Participants' Journey Times for Work Trips - M 2 W Vehicle versus Bus Modes Used for Work and Education Trips by Participants and their Households Rating of Factors Influencing Choice of M 2 W Vehicle Model ..' Distribution of M 2 W Vehicle Types in Participants'Households Annual Production of Different M 2 W Vehicle Types in India, 1960-1995 Sales of Different M 2 W Vehicle Types in Delhi and India, 1980-1995 Frequency and Quantity of Fuel/Oil Fills on Participants'Vehicles Oil/Fuel Ratios and Oil Purchase Mode Used by Participants M 2 W Vehicle Disposal Ages Planned by Participants Minimum Acceptable Scrappage Period Participants'Ranking of Policy Alternatives Rating of Bus Service Improvements Distribution of Trips by Distance in Delhi  133  134 143 144 166 166 167 167 168 177 177 178 178 179 179 187 187 188 188 189 191 191 196 210 210 216 216  Figure 7.1  Means-Ends Objectives Network  Xlll  GLOSSARY AIAM AIIMS APM AQ ASRTU BaP BAU BC BIS CBA CE CFC CIRT CNG CoHB CoP COPD CPCB CRRI CSE  cv  DCB DDA DPCC DTC DUA ECE EPEFE ESCAP ESI FE FTP GEMS GNP Gol GVRD GVW HSU I&M IARC IDRC IIP  Association of Indian Automobile Manufacturers, Mumbai and New Delhi, India All India Institute of Medical Sciences, New Delhi, India Administered Pricing Mechanism Air Quality Improvement Association of State Road Transport Undertakings, New Delhi, India Benzo(a)pyrene Business-as-usual (scenario) British Columbia Bureau of Indian Standards, New Delhi, India Cost-benefit Analysis Cost-effectiveness Chlorofluorocarbons Central Institute of Road Transport, Pune, India Compressed Natural Gas Carboxyhaemoglobin Conformity of Production Chronic Obstructive Pulmonary Disease Central Pollution Control Board, New Delhi, India Central Road Research Institute, New Delhi, India Centre for Science and Environment, New Delhi, India Contingent Valuation Delhi Cantonment Board, New Delhi, India Delhi Development Authority, New Delhi, India Delhi Pollution Control Committee, New Delhi, India Delhi Transport Corporation, New Delhi, India Delhi Urban Area Economic Commission for Europe European Programme on Emissions, Fuels, and Engine Technologies United Nations Economic and Social Commission for Asia and the Pacific Employees' State Insurance Corporation (India) Fuel Efficiency Improvement Federal Test Procedure (US) Global Environment Monitoring System Gross National Product Government of India Greater Vancouver Regional District Gross Vehicle Weight Hartridge Smoke Unit Inspection and Maintenance International Agency for Research on Cancer, Lyon, France International Development Research Centre, Ottawa Indian Institute of Petroleum, Dehra Dun, India  xiv  IIT Delhi INR IP ITO JNU JTW LCV LIC LPG M-0 M2W M3W MCD MoEF MoF MoH Mol MoPNG MoST MQ MRTS MTBE NAAQM NAAQS NCR NCTD NDMC NEERI NGO NMT  occ  OECD ORG PAH persons/ha PKM PM ppm PUC Q R&D RFG RITES RPM  Indian Institute of Technology, Delhi, India Indian Rupee Interview Protocol Income Tax Office, New Delhi, India Jawaharlal Nehru University, New Delhi, India Journey to Work Light Commercial Vehicle Less-industrialized country Liquefied Petroleum Gas Multiple-objectives Motorized two-wheeled (vehicle) Motorized three-wheeled (vehicle) Municipal Corporation of Delhi, Delhi, India Ministry of Environment and Forests, Government of India, New Delhi, India Ministry of Finance, Government of India, New Delhi, India Ministry of Health, Government of India, New Delhi, India Ministry of Industry, Government of India, New Delhi, India Ministry of Petroleum and Natural Gas, Government of India, New Delhi, India Ministry of Environment and Forests, Government of India, New Delhi, India Survey Questionnaire Mass Rapid Transit System (New Delhi, India) Methyl tertiary-butyl ether National Ambient Air Quality Monitoring Programme (India) National Ambient Air Quality Standards (India) National Capital Region (India) Government of the National Capital Territory of Delhi, New Delhi, India New Delhi Municipal Corporation, New Delhi, India National Environmental Engineering Research Institute, Nagpur, India Non-governmental organization Non-motorized transport Oil Co-ordination Committee, Government of India, New Delhi, India Organization for Economic Cooperation and Development Operations Research Group, Baroda, India Polycyclic Aromatic Hydrocarbons persons per hectare Passenger-kilometres Particulate Matter parts per million Pollution Under Control (Certificate) (New Delhi, India) Question Research and Development Reformulated Gasoline Rail India Technical and Economic Services Ltd., New Delhi Respirable Particulate Matter  XV  RVP SAE SOF SPM SQ TERI TSM TSP TVS UBC UNEP USEPA VKM VOC WHO WTAC WTP  Reid Vapour Pressure Society of Automotive Engineers (USA) Soluble Organic Fraction Suspended Particulate Matter Supplementary Questionnaire Tata Energy Research Institute, New Delhi, India Transport System Management Total Suspended Particulates T. V . Sundaram Iyengar Group of Companies, India The University of British Columbia, Vancouver, B C United Nations Environment Programme United States Environmental Protection Agency Vehicle-kilometres Volatile Organic Compounds World Health Organization Willingness to Accept Compensation Willingness to Pay  % vol. ug/dl ug/m urn °C O3 2-T cc. Cdn$ CH4 CO CO2 dB(A)  Percentage on a volume basis Micrograms per decilitre Micrograms per cubic metre micrometre Degrees Celsius Ozone (Lubricant for) two-stroke spark-ignition air-cooled gasoline engines Cubic centimetres Candian Dollar Methane Carbon Monoxide Carbon Dioxide Unit of relative intensity of sound, using the A weighting network, which discriminates against low frequency sounds Grams per kilometre Grams per litre Grams per cubic metre Grams per passenger-kilomtre Hydrocarbons Kilometre Kilometres per hour Kilometres per litre Kilopascal Milligrams per cubic metre Millilitre Nitrous Oxide Nanograms per cubic metre  3  g/km g/1 g/m g/pass-km HC km km/h km/1 kPa mg/m ml N 0 ng/m 3  3  2  3  xvi  N0 N0 OH Pb PMi PM PM S0 S0 2  X  1 0 2 5  2  X  T90  us$  Nitorgen Dioxide Nitrogen Oxides Hydroxyl radical Lead Suspended particulate matter of diameter 1 micrometre or less Suspended particulate matter of diameter 10 micrometres or less Suspended particulate matter of diameter 2.5 micrometres or less Sulphur Dioxide Sulphur Oxides Temperature in degrees Celsius at which 90% of the fuel evaporates and is recovered in the distillation process United States Dollar  \  To my parents Radha and Govind Badami  ACKNOWLEDGMENTS Many persons helped make this dissertation possible. To all of them, I express my deepest gratitude. I thank the following individuals in particular for their guidance, help, encouragement and support: Dr. Tim McDaniels, School of Community and Regional Planning, U B C Dr. V . Setty Pendakur, School of Community and Regional Planning, U B C Dr. Clyde Hertzman, Department of Health Care and Epidemiology, U B C Aditya Badami and Anita Rau Badami Suman and Harish Badami, New Delhi Mr. T. M . Balaraman, Bajaj Auto Limited, Pune Dr. Kamal Bhattacharyya, Air Quality Department, G V R D Dr. Ranjan Bose, Tata Energy Research Institute, New Delhi Dr. T. Chandini, Ministry of Environment and Forests, Government of India, New Delhi Dr. Hadi Dowlatabadi, Carnegie Mellon University Keisuke Enokido, U B C Centre for Human Settlements Mr. Mahesh Gaur, Transportation Research and Injury Prevention Programme, IIT Delhi Dr. Prodipto Ghosh, Asian Development Bank, Manila Dr. Henry Hightower, School of Community and Regional Planning, U B C Dr. Thomas A. Hutton, School of Community and Regional Planning, U B C Mr. N . V . Iyer, Bajaj Auto Limited, Pune Dr. Rajive Kumar, All India Institute of Medical Sciences, New Delhi Dr. Milind Kandlikar, Carnegie Mellon University Dr. Ajay Mathur and colleagues, Tata Energy Research Institute, New Delhi Dr. Dinesh Mohan, Transportation Research and Injury Prevention Programme, IIT Delhi Mr. M . N . Muralikrishna and colleagues, TVS-Suzuki Limited, Hosur M y colleagues in the U B C Centre for Human Settlements Dr. R. K . Pachauri, Tata Energy Research Institute, New Delhi Dr. Sudarsanam Padam, Central Institute of Road Transport, Pune Mr. N . R. Raje, Indian Oil Corporation R & D Centre, Faridabad Dr. A . T. K . Rau, Armed Forces Research and Referral Hospital, New Delhi Mr. T. S. Reddy, Central Road Research Institute, New Delhi Mr. A . S. Subramanian, Tata Engineering and Locomotive Company Limited, Pune The participants in my vehicle user survey and interviews Dr. Geetam Tiwari, Transportation Research and Injury Prevention Programme, IIT Delhi Mr. Saurabh Yadav, IIT Delhi I thank the following institutions for their financial and/or institutional support, and/or hospitality: Centre for India and South Asia Research, Institute of Asian Research, U B C Centre for Integrated Study of the Human Dimensions of Global Change, Department of Engineering and Public Policy, Carnegie Mellon University  xix  International Development Research Centre Social Sciences and Humanities Research Council Tata Energy Research Institute, New Delhi The University of British Columbia Transportation Research and Injury Prevention Programme, Indian Institute of Technology, Delhi U B C Centre for Human Settlements  1  CHAPTER I INTRODUCTION  1.1  BACKGROUND, AND THE PUBLIC POLICY C H A L L E N G E  Global motor vehicle numbers and activity are growing rapidly. If present trends continue, the global motor vehicle fleet is likely to double from its 1990 level of around 700 million by 2020. While the O E C D countries account for the bulk of global motor vehicle activity, much of the growth will likely be concentrated in the less-industrialized countries (LICs), including in Asia. Motor vehicle numbers are doubling every 7-10 years in many Asian LICs. B y 2040, rapidly industrializing LICs could have as many vehicles as North America and Western Europe. A n important characteristic of motor vehicle activity in Asian LICs is the predominance of motorized two-wheeled (M2W) vehicles (Faiz et al 1992; Sathaye, Tyler and Goldman 1994; Walsh 1991a; Walsh 1994). Because of the concentration of motor vehicular and other energy-intensive activities in LIC megacities, air quality in these cities is deteriorating rapidly, and could soon rival Mexico City's, where levels of several air pollutants already exceed WHO limits by more than a factor of two. Further, because of the large populations of urban poor, who suffer from inadequate nutrition and limited medical care, significant health impacts ensue. The contribution of transport to air pollution in LIC cities is significant and growing (Brandon and Ramankutty 1993; CPCB 1995; CPCB 1996; CSE 1996; Faiz et al 1992; Romieu, Weitzenfeld and Finkelman 1991; WHO/UNEP 1992). The trends in motor vehicle activity and urban air quality in the LICs are abundantly evident in India. India's motor vehicle fleet increased from only 665,000 in 1961, and 5.4 million as late as 1981, to over 27 million in 1994. The fleet is likely in the range of 40 million as of 2000. M 2 W vehicles are the most rapidly growing vehicle type in India, and represent around 67% of motor vehicles nationally. India arguably has the largest population of this vehicle type of any country. In the Indian capital, Delhi, motor vehicles grew 20% annually in the 1970s and 1980s, as against a population increase of 5-6% per annum. While motor vehicle numbers no longer appear to be increasing at the same pace, they are still growing at around 8% per annum. In 1996, around 2.6 million motor vehicles were registered in the city.  2  Of these, about 1.7 million were M2W vehicles. If current trends persist, Delhi will likely have around 5.2 million motor vehicles by 2005. Around 3.4 millions of these will likely be M 2 W vehicles ( A I A M 1994a; A I A M 1995; ASRTU/CIRT 1997; Faiz et al 1992; Mohan et al 1997; TERI 1997; WHO/UNEP 1992). Air quality in Delhi is poor, and deteriorating. In particular, annual average suspended particulate (SPM) levels, which are strongly correlated with respiratory and cardiovascular diseases, have been routinely around five times the World Health Organization (WHO) guideline limit since the 1980s. Daily average S P M levels exceed W H O limits almost every day, with peak levels as high as 6-10 times the WHO limit at many sites. Daily average sulphur dioxide (SO2) and nitrogen dioxide (NO2) levels exceed W H O guideline limits on several days of the year, at several sites. Ozone appears to be a major problem, especially in winter (CPCB 1995; CPCB 1996; CSE 1996; WHO/UNEP 1992). The contribution of transport to air pollution is growing in Delhi, as in many other Indian and L I C cities (CPCB 1996; CSE 1996; Faiz et al 1992). Because M 2 W vehicles are used intensively, and are for the most part powered by highly polluting two-stroke engines, these vehicles play an important role in transport air pollution, particularly on a passenger-kilometre basis (ASRTU/CIRT 1997; Bose 1996; CPCB 1997; GoI/ESCAP 1991; IIP 1994; Shah and Nagpal 1997). The rapid growth in motor vehicle activity in Delhi and other L I C cities has important implications for local well-being, in terms of air pollution, road safety, land use, access and mobility, and other transport impacts. This growth also has implications for regional and global issues such as energy security, acidification and climate change. While the OECD countries account for about two-thirds of global commercial energy consumption due to transport, their demand growth is expected to be flat or growing slowly. On the other hand, LIC transport energy demand, currently only around one-third that in the OECD, could increase as much as three times in as many decades. Transport already consumes around 45% of the world's oil, and is the fastest growing end-use category. India's and China's combined oil consumption accounts for 6.6% of the world's, but is increasing at 6.4% per annum, while world oil consumption is increasing at 1.5% per annum (Brandon and Ramankutty 1993; Flavin and Lenssen 1991; Griibler 1994; Holdren 1990; TERI 1997).  3  Air pollution due to M 2 W and other motor vehicles, and more generally, transportenergy-environment linkages in India and other LICs, are therefore issues worthy of public policy attention. M 2 W vehicles play an important role in transport air pollution in Delhi and other Asian L I C cities, but they also provide mobility to millions who have few other attractive options (Sathaye, Tyler, and Goldman 1994). Thus, the public policy challenge in terms of M 2 W vehicle emissions is to address the problem while minimizing adverse policy impacts for vehicle users. This challenge is made more daunting by the fact that transport air pollution is far more difficult to control than stationary source emissions, because it is more complex in its causes and effects, and involves a large number and variety of motor vehicles (Faiz et al 1992), and the choices of millions of vehicle users. Finally, the institutional setting for policy-making and implementation in the Indian and LIC contexts is characterized by serious technological, financial and administrative constraints (Author's survey and interviews 1997; Brandon and Ramankutty 1993; Douglass and Lee 1996; Faiz et al 1992; Hardoy, Mitlin and Satterthwaite 1992). It is this challenge that provides the rationale for this dissertation, which focuses on policy analysis of air pollution due to M2W vehicles in Delhi. This focus is all the more relevant because the city's rapid growth in motor vehicle activity, particularly in terms of M 2 W vehicles, and its deteriorating air quality, are features shared by many other Indian and L I C cities.  1.2  RESEARCH OBJECTIVES AND APPROACH  This dissertation has two objectives. The first is to contribute to a more thorough understanding of, and to the academic policy analysis literature on, transport air pollution in the Indian and LIC contexts. The second objective is to inform policy-making and implementation for prevention and control of M2W vehicle air pollution in those contexts. The dissertation attempts to achieve these objectives by illuminating key aspects of the M 2 W vehicle air pollution problem, presenting an analytic framework for addressing the problem effectively, and reporting on policy-relevant research that the author conducted on M 2 W vehicle emissions in Delhi.  4  The dissertation addresses the following specific research questions: •  What are the technological, institutional, and vehicle user behavioural factors that contribute to air pollution due to M 2 W vehicle activity in Delhi?  •  What is the institutional setting for prevention and control of transport air pollution in the Indian context? Who are the actors and what are their roles, responsibilities and interactions in terms of policy-making and implementation? What are the institutional barriers and constraints? What are the implications of all of the above for transport air pollution prevention and control in the Indian context?  •  What is an appropriate analytic framework for thinking systematically about, and for enabling effective policy-making and implementation with regard to air pollution from M 2 W vehicles, given the characteristics of the Indian context?  •  What are the important vehicle user preferences, choices, and motivations that influence M 2 W vehicle activity and air pollution in Delhi? What are M 2 W vehicle user perspectives on various current, proposed and possible policies, in terms of how they would be affected by and would respond to these policies? What measures would likely make these policies more attractive to users? What are the implications for policy-making and implementation, and how can policies be better designed in light of user preferences, choices and perspectives?  •  What are appropriate objectives and measures on the basis of which to evaluate policies targeted at M 2 W vehicle emissions in the Indian context? In this regard, can tools employed with success in Western settings to clarify and structure public policy problems be used to achieve similar ends when applied to a highly complex issue such as transport air pollution in the LIC context?  •  What are the broader implications of this study for transport air pollution prevention and control, urban transport, and urban environmental policy and planning in India?  The dissertation research focuses on M2W vehicles in Delhi, but aims to have relevance for transport air pollution generally in the Indian and L I C contexts. Following is a brief introduction to the approaches used to answer the research questions, and their contribution to fulfilling the dissertation objectives.  5  1.2.1  Contributory Factors  Transport air pollution is a complex and multi-dimensional problem. It is critically important to understand the various factors that contribute to the problem, in order to effectively address it. The dissertation analyses the factors that contribute to M 2 W vehicle air pollutant emissions in Delhi, and more generally, transport air pollution and energy consumption in India. This analysis can help identify policy levers to target key contributory factors, and make transport emissions and energy consumption measurement and modeling efforts more effective. The dissertation addresses proximate technological as well as underlying institutional contributory factors. Since transport air pollution is a function of per-vehicle emissions as well as overall vehicle activity, factors that contribute to both of these components of the problem are addressed. Further, the dissertation focuses on system-wide air pollution due to M 2 W vehicle activity, not merely vehicle exhaust pipe emissions. Also, it addresses important contextual characteristics that critically influence emissions in the Indian context. Thus, technological-curative as well as preventive alternatives may be identified to address the problem comprehensively and effectively over the long term.  1.2.2  The Institutional Setting  The dissertation critically examines the institutional setting in relation to the transport air pollution problem in the Indian context. The actors whose roles, responsibilities and interactions are analyzed include key government agencies at the national and local levels, vehicle and fuel manufacturers, academic and research institutions, environmental NGOs, the courts and public interest litigators, and the media. This analysis can help identify critical institutional barriers and constraints, and mechanisms to overcome them. The institutional setting is shown to be characterized by a multitude of actors with fragmented, overlapping, and conflicting roles and responsibilities, and by restricted financial, technological, and administrative resources. Actors' interactions have been characterized by conflict. Many current and proposed policies have not been thoroughly considered, in terms of the institutional support mechanisms necessary for their success, or their long-term consequences. The result is that many policies are likely to be costly and burdensome, yet  6  ineffective in addressing transport air pollution (Author's interviews 1997; C S E 1996; C S E 1997).  1.2.3  Policy A nalytic Framework  Given the institutional setting, an analytical framework is needed for systematic thinking and effective policy analysis and implementation with regard to transport air pollution in the Indian context. This dissertation attempts to fulfill this need, by building on similar attempts in the current literature (Faiz et al 1992; Carbajo 1993; Shah, Nagpal and Brandon 1997), and by addressing their shortcomings. A systematic methodology is proposed for estimating air pollutant emissions due to various policy alternatives. The methodology stresses the need to minimize system-wide emissions due to M 2 W vehicle activity, over the long term. It takes into consideration important in-use realities, urban transport issues, the inter-dependence between modes and pollutants, and vicious circles and side-effects. Issues that are typically relegated to the policy implementation phase are considered explicitly. So are data uncertainties and variabilities, based in part on informed expert judgments. Welfare economists suggest cost-benefit analysis (CBA) as an ideal methodology for comparing policy alternatives (Carbajo 1993; Pearce and Markandya in Faiz et al 1992). The dissertation argues that operationalizing C B A would be problematic in the Indian context, because of the technical, conceptual and philosophical difficulties involved in the estimation and monetization of policy benefits. Instead, the framework proposes cost-effectiveness as a basis for estimating and valuing emissions impacts due to policy alternatives, with individual pollutants weighted to reflect their contributions to health, environmental and other impacts of concern. Further, because each policy can potentially have transport impacts other than those related to air pollution, and a range of cost and welfare impacts for different actors and groups, the framework proposes, based on the work of Edwards and von Winterfeldt (1987), Hobbs and Horn (1997), Keeney (1982, 1988a, 1988b, 1990, 1992), Keeney, von Winterfeldt and Eppel (1990) and Keeney and McDaniels (1992), that policy alternatives be evaluated in terms of multiple objectives reflecting these impacts and the diverse interests and concerns of various actors and affected groups.  7  1.2.4  Vehicle User Choices and Perspectives  Transport air pollution inevitably involves technological issues. At the same time, transport air pollution, and the effectiveness of prevention and control policies, are strongly influenced by vehicle user choices, and by how users are affected by and respond to policies. Indeed, some of the policy difficulties in the Indian context have come about as a result of these issues not having been adequately considered. An investigation into these human dimensions would be useful in targeting critical user behavioural factors that contribute to transport air pollution. It would also be useful in developing policies that are attractive to users, and therefore have a greater chance of being effective in the long term. But neither the urban transport literature focused on the L I C context nor the environmental policy analytic literature related to transport air pollution (for example, Faiz et al (1992), RITES/ORG (1994), Shah, Nagpal and Brandon (1997)), pay much attention to vehicle user behavioural factors and perspectives as they specifically relate to transport air pollution and emission prevention and control policies. This dissertation attempts to address these gaps. Based on information gathered from a questionnaire survey of, and in-depth interviews with M 2 W vehicle users in Delhi, the dissertation investigates important M 2 W vehicle user and user household preferences, choices and motivations that influence M 2 W vehicle activity and air pollutant emissions in Delhi. These preferences, choices and motivations relate to vehicle ownership, mode choice, daily travel, and vehicle purchasing, operation, maintenance, disposal and replacement. The dissertation also investigates user perspectives on various technological and regulatory policy alternatives targeted at M 2 W vehicle emissions, on public transit and bicycle commuting, and on measures that would make these policy alternatives and alternative modes more attractive to them. Finally, the dissertation explores the implications of the user choices and perspectives for policy-making and implementation, particularly in light of institutional capabilities and constraints.  1.2.5  Multiple Policy Objectives  Policies to address complex public policy problems such as transport air pollution involve environmental, health, socio-economic, safety and other implications for large sections of the  g  public and for future generations. They also involve multiple stakeholders, with multiple, conflicting interests and concerns. Gaining an understanding of these impacts, interests and concerns, and developing a broad range of policy objectives that reflect them, would be useful in terms of understanding the barriers to policy-making and implementation, and in designing policy packages that represent a win-win condition for all, thus enhancing chances of longterm policy success (Edwards and von Winterfeldt 1987; Keeney 1982; Keeney 1988a; Keeney 1992; Keeney and McDaniels 1992; Keeney and McDaniels 1999). Further, given the lack of effective co-ordination, and the tenuous linkages between the policy-analytic and decision-making communities in the Indian context (Kandlikar 1998), an integrative perspective on what is a multi-dimensional problem would be particularly valuable. Unfortunately, policy objectives for transport air pollution prevention and control are not well understood  or articulated, and are therefore  not considered in policy-making and  implementation. Analytical methods to clarify and structure public policy problems characterized by multiple stakeholders with multiple conflicting objectives are well established, and have been applied to a diversity of policy situations in Western settings. These problem-structuring tools of "value-focused thinking", which directly involve stakeholders to identify key public values, enable selection of alternatives that better serve these values (Keeney 1988b; Keeney, von Winterfeldt and Eppel 1990; Keeney 1992; Keeney and McDaniels 1992; Keeney and McDaniels 1999). This dissertation applies these tools to better understand the problem of M 2 W vehicle air pollution in the Indian context, and to identify and structure multiple objectives and measures on the basis of which to systematically create, evaluate, implement and monitor policy alternatives for long-term effectiveness. Problem-structuring tools to clarify and structure public values have been used in only a few cases in the LIC context (Gregory and Keeney 1994; McDaniels and Trousdale 1999), but have not been applied, to the author's knowledge, to a highly complex situation such as the one this dissertation addresses. At any rate, this is perhaps the first attempt at developing multiple policy objectives related to transport air pollution in the Indian context. In addition to fulfilling an important policy-analytic need in relation to this problem in this context, the  9  dissertation also addresses the question of whether such tools can be applied to better understand and structure complex public policy problems in LIC contexts generally.  1.3  L E V E L OF ANALYSIS  Any research endeavour inevitably involves choices in terms of problem definition and boundaries, the various aspects of the problem that will be addressed, and policy alternatives that will be evaluated, among other issues. This dissertation is no exception, particularly since it addresses a highly complex, multi-dimensional problem like transport air pollution. The first choice related to the component of the transport air pollution problem in the Indian context on which to focus. Note in this respect that transport air pollution prevention and control is itself only one part of the much larger problem of how to achieve a resource conserving, environmentally benign, safe, and socially just urban transport system in this context. While fully recognizing the important role of other motorized modes and transport system components, the decision was made to focus on M 2 W vehicles, based on their intensity of use, contribution to pollution on a passenger-kilometre basis, and their importance to millions of users, as already discussed. Also, it was reasoned that policies targeted at M2W vehicles would have relevance for M3W vehicles as well. The second choice related to the aspects of the M 2 W vehicle emissions problem to address. It was decided that this problem would be addressed in its environmental, technological, socio-economic, human behavioural and political-institutional dimensions, in order to gain a thorough understanding of the problem, and to be better able to inform policymaking and implementation. The treatment of contributory factors, institutional setting, vehicle user choices and perspectives, and the policy-analytic framework and multiple objectives and measures as outlined above reflects this decision. The third choice related to the contributory factors and policy alternatives on which to focus. There is a wide range of factors contributing to the per-vehicle emissions and vehicle activity components of the problem. Correspondingly, there is a wide range of technological, infrastructural, economic, regulatory and transport demand reduction policies that may be applied to address the problem. This dissertation focuses on technological factors and technological and regulatory policies, but recognizes the importance of, and discusses non-  10  technological factors influencing vehicle ownership and activity, and policies to address this component of the problem. Indeed, these factors and policies are considered throughout the dissertation, in each of the research tasks outlined earlier under Research Approach. Thus, while the discussion of contributory factors and institutional setting focuses on vehicle and fuel technology, factors such as population growth, income, land use and housing location choice, sprawl and access for non-motorized modes, and transport infrastructure and public transit provision, and actors' roles and institutional constraints with respect to these factors are also addressed. Similarly, M 2 W vehicle user choices and perspectives are investigated with respect to technological and regulatory policies as well as public transit and bicycle commuting. Lastly, while the policy-analytic framework and multiple objectives and measures apply to technological and regulatory policies targeted at M 2 W vehicles, they also accommodate other policies. Further, they reflect the interdependence between modes, and between transport air pollution and other transport system impacts, including land use. In summary, the research approach attempts to reflect the complexity and multidimensional nature of the M 2 W vehicle emissions problem, and integrates environmental policy-analytic, urban transport, engineering and planning perspectives. It is policy-relevant, and sensitive to the Indian context. It explicitly considers contextual capabilities and constraints, implementation issues, and the multiple conflicting interests and concerns of various actors and affected groups, in order to more effectively achieve its objective, which is to inform policy-making and implementation.  1.4  METHODOLOGY  This dissertation draws on published literature on a range of subjects including environmental policy, engineering, urban transport, and urbanization, reflecting the multi-dimensional nature of the transport air pollution problem. Documents relating to the state of vehicle and fuel technology and the urban transport system in the Indian context were also reviewed. In order to gain an understanding of policy-making and implementation processes, a wide range of published as well as unpublished written material, including pertinent  environmental  legislation, reports and position papers prepared recently by local and national government agencies, environmental NGOs and vehicle and fuel manufacturers and industry associations,  11  and transcripts of proceedings of Supreme Court public interest cases, were critically analyzed. In addition to published literature and other secondary sources, the dissertation draws on in-depth interviews that the author conducted with various individuals interested in and/or knowledgeable about the range of issues involved, and representatives of institutions whose actions have an important bearing on transport air pollution in the Indian context. These individuals, listed in Appendix V , included decision makers in various relevant government agencies at the national and local levels, senior executives in the Indian M 2 W vehicle and fuel industries, and academics and researchers in the fields of environmental policy and urban transport. The interviews focused on a wide range of issues, including technical and institutional factors contributing to transport air pollution; actors' roles, responsibilities and interactions; institutional barriers and constraints; considerations underlying current and proposed policies; and likely impacts of policies on vehicle users and industry. In addition to sharing their insights, the interviewees made available to the author the bulk of the documents referred to in the first paragraph in this section. As indicated, M 2 W vehicle user preferences, choices, perspectives and motivations were elicited by means of a questionnaire survey and in-depth interviews with M 2 W vehicle users in Delhi, in late 1997. The survey and in-depth interviews covered a wide range of issues, including those referred to under Research Approach. The interviews, along with the information culled from the other sources indicated helped the author gain a comprehensive understanding of the transport air pollution problem in the Indian context.  1.5  STRUCTURE OF THE DISSERTATION  Each aspect of the M2W vehicle air pollution problem in the Indian context discussed earlier under Research Approach is treated separately and in depth in the following chapters. At the same time, an attempt has been made to maintain a logical flow from one chapter to the next. Chapter II describes the problem that forms the focus of this dissertation. It presents a comprehensive picture of motor vehicle activity, and air pollution and its impacts in Delhi, and stresses the important role of M2W vehicles. The chapter presents a rationale for public policy  12  attention to, and outlines the larger context that needs to be considered in addressing, the problem. Chapter III discusses the technological, institutional and vehicle user behavioural factors that contribute to M 2 W vehicle air pollutant emissions in Delhi, and more generally, transport air pollution and energy consumption in India. Chapter IV discusses actors' roles, responsibilities and interactions, and institutional constraints and barriers, in relation to the various aspects of the transport air pollution problem discussed in Chapter III. Chapter V proposes an analytic framework for systematic thinking and effective policymaking and implementation for addressing air pollution from M 2 W vehicles in the Indian context. Chapter V I describes the methodology for eliciting M 2 W vehicle user preferences, choices, perspectives and motivations by means of a questionnaire survey and in-depth interviews, and presents and discusses the results. The chapter then explores the implications of these results for policy-making and implementation. Chapter VII details the methodology for eliciting and structuring multiple policy objectives with respect to the issue at hand, and presents the results, in the form of a fundamental objectives hierarchy, measures, and a meansends objectives network. Chapter VIII summarizes the implications of the dissertation research for transport air pollution prevention and control, and for urban transport and environmental planning, in the Indian context. The chapter also makes suggestions for further research.  13  CHAPTER H T H E TRANSPORT AIR POLLUTION PROBLEM IN DELHI, INDIA  2.1  INTRODUCTION  The purpose of this chapter is to present a rationale for public policy attention to air pollution from motorized two-wheeled (M2W) vehicles in Delhi, which is the dissertation focus. In order to achieve this purpose, the chapter discusses motor vehicle activity, and air pollution and its impacts, in Delhi. It stresses the role of M 2 W vehicles, which account for an increasingly important share of motor vehicle activity as well as transport air pollution. The dissertation focus is shown to be all the more relevant in light of the fact that Delhi's transport air pollution problem is shared by many other cities in India and other less-industrialized countries (LICs). Further, the growing importance of transport in India and other LICs in terms of global issues such as climate change, acidification and energy security is stressed. Finally, the public policy challenge is highlighted by situating Delhi's transport air pollution problem in the context of broader urban transport impacts. The chapter describes the problem that forms the focus of this dissertation, argues why it is worthy of urgent public policy attention, and outlines the larger context that needs to be considered in addressing the problem. The information on which this chapter is based was culled from a wide range of secondary sources, many of which were identified during the course of meetings conducted by the author with decision makers in various relevant government agencies, representatives of vehicle and fuel manufacturing industries, and academics and researchers. These interviewees are listed in Appendix V .  2.2  URBANIZATION AND MOTOR VEHICLE ACTIVITY IN DELHI  Delhi has a history going back more than 2000 years. The city has been the capital of several kingdoms and empires. Delhi is today the third most populous city in India, after Calcutta and Greater Mumbai. The present-day union territory of Delhi, spread over 1,483 square kilometres, comprises the cities of Delhi, popularly known as Old Delhi, the capital of Muslim India from the 12th to the 19th centuries, and New Delhi, the capital of British India from  14  1912, and of the Republic of India since independence in 1947, and adjacent rural areas.  1  Delhi has been the dominant trading and commercial centre of northern India for centuries. The city is now a key transportation centre, with several national highways and railway lines passing through.  Government and administrative services, industry and  commerce,  transportation, storage, distribution, and wholesale trading are the chief economic sectors (Encyclopaedia Britannica 1998; WHO/UNEP 1992). In 1941, the number of people in the Delhi urban area was only 700,000, which is roughly the current population of several medium-sized Western cities. By 1991, Delhi's population had increased to 8.48 million, representing a 1100% increase over a mere half-century, and a doubling in less than twenty years, since 1971 (Figures 2.1 and 2.3). If present trends 2  continue, Delhi's urban population is likely to increase to around 13 million by 2001. According to the Master Plan for Delhi, Delhi's urban and rural population could together reach 15 million in the same year (DDA 1996; Faiz et al 1992; Midgley 1994; RITES/ORG 1994; Tiwari and Kale 1997; WHO/UNEP 1992). Motor vehicle sales sky-rocketed in Delhi in the 1980s. In 1990, annual sales of motor vehicles were 370% higher than in 1980. But the rate of growth in Delhi's motor vehicle sales is tapering off. Annual motor vehicle sales in 1995 were only 19.4% higher than in 1990. Even so, over 550 motor vehicles of all types are sold daily in the city. Delhi accounted for 6.2, 18.1 and 5.1% of all-India sales of motor vehicles, cars and jeeps, and M 2 W vehicles respectively in 1995. Interestingly, Delhi's annual motor vehicle sales in 1995 equaled that in many large states such as Tamil Nadu (population 60 million). Car sales in the city exceeded those in any of the other Indian states, and were in fact three times that in Uttar Pradesh, the most populous Indian state (149 million) ( A I A M 1994a and 1995; Encyclopaedia Britannica 1998).  While the area's economy and population are centred mainly in Old Delhi, government is concentrated in New Delhi. Old Delhi has nearly ten times the population density of New Delhi (Encyclopaedia Britannica 1998). This rapid surge in population was due to the massive influx of refugees during Partition (1948), and then the strong pull Delhi has exerted since the 1950s, because of the concentration of economic and political power and employment there (Misra et al 1998). 1  2  Figure 2.1  Population Growth, Delhi Urban Area 1941-2001  14  Population, millions  12 10  1941  1951  1961  1971  1981  1991  2001  Year mgb99  Sources: DDA 1996; RITES/ORG 1994  Figure 2.2  Motor Vehicle Growth in Delhi 1971-1996  3000  Vehicle Numbers, Thousands ^Buses QM3W ^ Goods ^M2W E3Car/Jeep/Taxi  2500  2000  1500  1000  500  1971  1981  1991  1996  Year M2W/M3W B Motorized two-rthree-wheeled vehicles. Sources: AIAM 1994a; C P C B 1997; Mohan et al 1997; NCTD1997.  mgb99  Figure 2.3  Population and Motorization Growth Rates in Delhi 1941-2001  25  Growth compared to Datum Year  0 ' 1941  1951  1961  1971  1981  ' 2001  1991  Year Motor vehicle population for 2001 estimated using average growth rate for 1890s. Sources: AIAM 1894a and 1995; C P C B 1997; DDA 1996; Faiz et al 1992; Mohan at al 1997; NCTD 1997; RITES/ORG 1994; WHO/UNEP 1892. mgb99  Figure 2.4  Delhi's Motor Vehicles Compared to Other Major Indian Cities 1994  All Motor Vehicles  M2W  M3W m Delhi •Calcutta + Chennai + Mumbai + Bangalore Cars + Jeeps In 1994, Delhi had: 8.1 % of India's motorized two-wheeled vehicles 5.8 % of India's buses 14.8 % of India's cars, jeeps, and taxis 7.8 % of India's motor vehicles  Buses  Trucks 600  1000  1500  2000  2500  Thousands  M2W/M3W " Motorized two-rthree- wheeled vehicles. Source: ASRTU/CIRT 1997.  mgb99  3000  17  Corresponding to the rapid growth in vehicle sales, Delhi's motor vehicle fleet grew at an annual rate of around 20% in the 1970s and 1980s (Figure 2.2). The figure shows that, astounding as Delhi's population growth has been, motor vehicle numbers have grown even more rapidly. While the city's population grew approximately 18-fold in 60 years, its motor vehicle population has grown more than twenty-fold in half the time. Further, the growth rates for M 2 W vehicles and cars ~ 24-25% in the 1980s — have exceeded the overall motor vehicle growth rate. Motor vehicle numbers are not increasing at the same pace in this decade, but they are still growing at around 8% per annum. In 1996, around 2.6 million vehicles were registered in Delhi. Of these, 1.7 million (67%) were M 2 W vehicles. Additionally, 70,000 motor vehicles from neighbouring states were estimated to ply in Delhi daily ( A I A M 1994a and 1995; CPCB 1997; Mohan et al 1997; WHO/UNEP 1992). In 1994, the latest year for which data are available for all major Indian cities, Delhi had nearly as many motor vehicles as, and more M 2 W vehicles than, Calcutta, Chennai, and Greater Mumbai put together, plus Bangalore, perhaps the fastest growing Asian metropolis, and the second most motorized Indian city after Delhi (Figure 2.4). Delhi's motor vehicle fleet also exceeded that in every Indian state except Maharashtra and Gujarat. In 1995, Delhi accounted for 8.1% of motor vehicles, 7.8% of M2W vehicles, and 16.6% of cars registered nation-wide, with only a little over 1% of India's population. If current trends persist, Delhi will likely have around 5.2 million motor vehicles by 2005. Around 3.4 millions of these will likely be M 2 W vehicles ( A I A M 1994a and 1995; ASRTU/CIRT 1997; Faiz et al 1992; Mohan et al 1997; TERI 1997). To get a real sense of motor vehicle activity and its impacts, it is necessary to consider not merely motor vehicle numbers. While Delhi's motor vehicle numbers in 1995 had increased 3.4-5.2 times since 1981, depending on motor vehicle type, average trip lengths had increased 1.2-1.4 times (RITES/ORG 1994). Effectively, M 2 W vehicle and total motor vehicle kilometres in Delhi had increased 6.7 and 5.5 times in only 14 years (1981-1995), assuming no increases in daily trips per mode in the interval.  18  2.3  AIR POLLUTION AND H E A L T H EFFECTS IN DELHI  2.3.1  Air Pollution in Delhi  One does not have to be an expert to know that Delhi's air quality is poor. It is patently obvious with every breath one takes. People complain of a sore throat, wheezing, and coughing. The newspapers regularly cover the problem, and attempts to address it. This author counted 22 news items, articles, and editorials on the subject in just three of Delhi's English language newspapers in six months in 1997. The city is blanketed in a haze for much of every day, particularly in winter. M2W vehicle riders wear masks. The leaves on the trees are covered with a thick coat of grime and soot. But if "scientific" rather than merely anecdotal proof were needed, that is there too. Figures 2.5 to 2.10 show the evolving air quality situation in Delhi, compared to World Health Organization (WHO) guideline ambient air pollutant concentration limits. Appendix I compares the Indian N A A Q S (national ambient air quality standards) with corresponding WHO, U S and Californian limits. A detailed critique of the Indian N A A Q S standards is also presented in Appendix I.  3  Figures 2.5-2.10 are based primarily on Central Pollution Control Board (CPCB) and National Environmental Engineering Research Institute (NEERI) air quality data for Delhi.  4  CPCB and N E E R I data are generally not comparable, because the measurement sites are different, measurements are carried out on different days, different methodologies are employed, and analysis is conducted in different laboratories. Discrepancies exist between  The gist of the critique is that the Indian NAAQS differentiate between "industrial", "residential" and "sensitive" areas. This differentiation is pointless in the Indian context, because supposedly industrial areas are also heavily populated, and a significant proportion of industries are located in residential areas. Further, the Indian NAAQS are either unnecessarily or unrealistically stringent in some cases, and too lenient in some others, as compared to WHO limits (see Appendix I). NEERI has been operating three United Nations GEMS (Global Environment Monitoring System) air quality monitoring stations in one industrial and two residential locations in Delhi since 1978. Sulphur dioxide (S0 ), suspended particulate matter (SPM), and nitrogen dioxide (N0 ) have been regularly monitored at these locations, except in 1988 and 1989, when monitoring was discontinued. Airborne lead levels were measured at the three NEERI sites in 1990. The CPCB initiated the National Ambient Air Quality Monitoring (NAAQM) programme collaboratively with the pollution control boards in the states in 1984. This programme currently involves 290 monitoring stations in 92 cities and towns. In Delhi, the CPCB commenced air quality monitoring at five stations in 1987. As at 1996, CPCB gathered S0 SPM and N 0 data from five stations in industrial, and four stations in residential, areas. According to Cropper et al (1997), CPCB have been directly operating six stations since 1987. CPCB added three more in 1990, and these are operated on their behalf by NEERI. 3  4  2  2  2>  2  19  N E E R I and CPCB data for the same year, and between these and reports in various other sources. Further, neither NEERI nor CPCB monitored carbon monoxide (CO), ozone (O3) or 5  PM10 (suspended particulate matter (SPM) below 10 microns diameter) on a regular basis, at least during the period covered in Figures 2.5-2.10. Even those pollutants monitored by the 6  CPCB are not monitored continuously, or even regularly, contrary to requirements stipulated in the Indian N A A Q S . Thus, while the data for six Delhi sites in 1993 in CPCB (1996) were based on measurements on 69-79 days, those for the three other sites were based on measurements on only 21-23 days (Brandon and Hammonn 1995; C P C B 1995 and 1996; Shah and Nagpal 1997; WHO/UNEP 1992). With those caveats, let us now turn our attention to the annual average sulphur dioxide (S0 ), nitrogen dioxide (N0 ) and SPM levels in Delhi from 1978, presented in Figures 2.52  2  2.7. In addition to the annual averages for all sites, the maximum annual average recorded for each year is also presented, along with the area type at which the maximum occurred. For the sake of consistency, the CPCB data shown from 1990 onward is for the same three sites as for the N E E R I data from 1978. While annual average S 0 levels appeared to be rising in the early 1980s, and even 2  exceeded the applicable WHO limit of 50 ug/m in 1984-1986, they are now below this limit. 3  Similarly, annual average N 0 levels have been generally below the Indian N A A Q S standard 2  of 60 ug/m for residential areas until the early 1990s (there is no W H O annual average 3  standard for N 0 ) . However, CPCB data for 1994 and NEERI data for 1996 appear to 2  indicate an increasing trend (and even exceedence above the Indian N A A Q S limit for industrial areas in 1996). It is therefore not possible to say anything conclusive about N 0  2  trends in Delhi. However, if N 0 levels are indeed increasing in Delhi, it would not be 2  surprising, considering the rapidly increasing motor vehicular activity.  CPCB data for different cities cannot be compared, since the monitoring is conducted by the various state pollution control boards, and a large number of personnel and equipment types are involved in sampling and chemical analysis (CPCB 1996). There is a dire need for consistent air quality measurement and reporting, not just in Delhi, but across the country, to enable reliable comparisons and effective monitoring and control action. Though NEERI had apparently commenced measuring PM, , CPCB had not (CPCB 1995 and 1996; Shah and Nagpal 1997; WHO/UNEP 1992). 5  6  0  20  Most importantly, annual average S P M levels, which are strongly correlated with respiratory and cardiovascular diseases, have been routinely exceeding the WHO limit of 6090 ug/m by as much as five times, and even the Indian N A A Q S standard of 360 ug/m for 3  3  industrial areas, since the early 1980s. Incidentally, the CPCB itself designates 210 ug/m as 3  "critical" (and 70-140 u g / m as "moderate") (CPCB 1996), on what grounds precisely is not 3  clear. The one reported annual average PMio figure, for 1992 (in Brandon and Hommann 1995), is 181 ug/m , measured by NEERI. This level is nearly 2.5 times higher than even the 3  W H O annual average limit for total SPM, let alone the W H O 24-hour limit for PMio (WHO/UNEP 1992).  7  Airborne lead levels do not appear to be measured on a regular basis in Delhi. The annual average (as measured by NEERI) was reported as being around 0.5 and 0.27 ug/m in 1990 3  and 1992, and 0.20 ug/m in 1996/97 (Brandon and Hommann 1995; M o E F 1997b; 3  W H O / U N E P 1992). Annual average lead levels are below the W H O limit, and appear to be declining, likely due to the progressive reduction of lead in Indian gasoline (Appendix IV). As already mentioned, C O levels are not monitored on a regular basis. But annual average CO was measured at the ITO crossing, one of Delhi's busiest traffic centres, from 1990 (MoEF 1997b). This data shows a rising trend, from 2.7 mg/m in 1990, to 5.6 mg/m in 3  3  1996. But these figures are below even the WHO 8-hour limit (WHO/UNEP 1992). The annual average values, serious as they are in the case of health-critical air pollutants such as SPM, hide site-to-site variations. Annual average S 0 levels at all measurement sites 2  are below the applicable WHO limit generally, but those at specific industrial and even residential sites regularly exceed it (Figures 2.5 and 2.8). While the annual average S P M levels at all sites have been routinely exceeding the WHO limit by five times since the early 1980s, as noted, those at specific industrial and even residential sites have been as high as 6-10 times the W H O limit (Figures 2.7 and 2.9). There are also important seasonal variations.  8  P M is suspended particulate matter below 10 micron diameter. It is also referred to as thoracic PM (Faiz et al 1992; GVRD 1996; WHO/UNEP 1992). Though P M , is not measured, it may be calculated because it is estimated to be 0.50-0.60 of SPM by mass for Delhi (Cropper et al 1997). The WHO 24-hour P M , limit is 70 Ug/m . The Californian 24-hour P M , standard is only 30 ug/m (WHO/UNEP 1992). Seasonal variations in air pollutant concentrations in Delhi are a function of the local climatic features there. While the city's annual mean temperature is around is 25°C, the summers are extremely hot and dry, with temperatures as high as 47°C in June. In winter (November-February), temperatures dip as low as 3°C. While the heavy monsoon rains from June to October have a scrubbing effect, strong pre-monsoon westerly winds 7  1 0  0  0  3  3  0  8  21  Finally, to gain an appreciation of the acute exposures in Delhi, let us consider the shortterm (24-hour, 8-hour and 1-hour) concentrations of various pollutants (Figures 2.8-2.10). The 24-hour average S 0 and N 0 levels were above the corresponding WHO limits for many 2  2  days in 1994 at several sites in Delhi, including at residential sites. 24-hour average S P M levels exceeded the corresponding WHO limit on a daily basis. In fact, the 24-hour WHO limit was exceeded by even the monthly average SPM levels every month in 1993 and 1994 (the years for which the author has data for all months), at all of the C P C B measurement sites, both industrial and residential. Peak 24-hour average S P M levels were 6-10 times the W H O limit at many sites (Figure 2.9). In a 1997 study by Cropper et al (1997), the authors found 9  that daily levels exceeded (by several times) the WHO limit 97% of days. The W H O limit they considered was 150-230 ug/m , not 120 ug/m , which should apply in Delhi's case, since S 0 3  3  2  is also present (WHO/UNEP 1992). According to Priti Kumar et al (1997), the highest 24hour average S P M levels recorded in the three years preceding 1994 were as high as 2338, 2340 and 1227 ug/m . 3  8-hour and 1-hour CO levels were found to be within W H O guidelines, when averaged over various locations. At the busy ITO traffic junction in 1984, 8-hour average C O levels were around 5.5 mg/m , as against the WHO limit of 10 mg/m . However, the peak hourly 3  3  concentration was 29 mg/m , as against the WHO 1-hour limit of 30 mg/m (CPCB 1995). 3  3  Short-term C O levels have very likely increased dramatically since 1984 at several sites in Delhi, as motor vehicle activity has grown rapidly, with important implications for the thousands who eke out a living daily on the street, including police personnel and street vendors.  bring in large concentrations of SPM from the Thar Desert, and re-suspend it for long periods. Winters are characterized by frequent calms and temperature inversions which restrict mixing heights and thus pollutant dispersal and dilution. S0 levels typically peak in winter, due to increased coal burning for space heating, and N 0 levels peak in October-November when maximum insolation occurs (Faiz et al 1992; WHO/UNEP 1992). Once again, it should be noted that CPCB measurements were conducted only a few days each month, and for only 16 hours each day (CPCB 1995 and 1996). 2  2  9  Figure 2.5  Annual Average Sulphur Dioxide Levels in Delhi  1978-1996  100  micrograms/cubic metre  1978  1980  1982  1984  1986  1988  1990  1992  1994  Indl. = Industrial site; Res. » Residential site. Sources: C P C B 1997; NEERI data from AIAM 1997a and WHO/UNEP 1992.  Figure 2.6  1996 m9b99  Annual Average Nitrogen Dioxide Levels in Delhi  1978-1996  100  micrograms/cubic metre  80  60  40  20  1978  1980  1982  1984  1986  1988  1990  1992  1994  Indl. ° Industrial site. Res. » Residential site. Sources: C P C B 1997; NEERI data from AIAM 1997a and WHO/UNEP 1992.  1996 mgb99  Figure 2.7  Annual Average Particulate Levels in Delhi  1978-1996  800  micrograms/cubic metre Annual Average °%»Max. Annual Average Res.  .^^**^lndl.  600  400  200  1978  1980  1982  1984  1986  1988  1990  1992  1994  Indl. = Industrial sits. Res. » Residential site. Sources: CPCB 1997; NEERI data from AIAM 1997a and WHO/UNEP 1992.  Figure 2.8  Peak 24-hour Sulphur and Nitrogen Dioxide Levels at Various Sites in Delhi, 1994  300  micrograms/cubic metre  250  3N02  200 WHO 24-hr N02 limit 150  150 W H O 24-hr S 0 2 limit 125  100 50 0  y  •  r  •  /  /  ^  /  /  S  (I) = Industrial areas. (R) ° Residential areas. Source: CPCB 1996.  mgb99  1996 mgb99  Figure 2.9  Peak 24-hour SPM Levels  at Various Sites in Delhi, 1994  1200  micrograms/cubic metre  1000 800 600 400 200 WHO 24-hr limit 120 (J  «S\  sf  <§\  >  jr  <§\  *•  <5\  *r  <§\  Jtj>-  <£\  y  <£\  /  **'  ^ <r ^* mgb99  (I) = Industrial areas; (R) ° Residential areas. Source: C P C B 1998.  Figure 2.10  8-hour Mean and 1-hour Maximum Ozone Levels in Delhi 1997  300  micrograms/cubic metre  Source: Roy Chowdhury 1997.  mgb99  25  The only recent study on ozone (O3) in Delhi was conducted by the Central Road Research Institute (CRRI) at seven sites in the winter of 1993. W H O 8-hour limits were exceeded at most locations (Figure 2.10). The peak 1-hour levels across all sites averaged 203 ug/m , with the peaks in areas with heavy traffic like Karol Bagh, Daryaganj and Parliament 3  Street 30-40% in excess of the WHO 1-hour limit. Studies conducted by the Jawaharlal Nehru University (JNU) and the Tata Energy Research Institute (TERI) in 1989-90 had shown peak levels of 256 ug/m , well above the WHO 1-hour limit (CSE 1997; Roy Chowdhury 1997). 3  Delhi's smog could simply be reduced visibility due to increased SPM, but the city has the potential for photochemical smog, because of growing transport and other emissions, and a climate favourable to O3 formation, particularly in winter, which is when N O levels are high, x  and inversions occur (Faiz et al 1992; WHO/UNEP 1992). To summarize the air quality situation in Delhi: annual average CO and SO2 levels appear to be within W H O limits. So are annual average N O 2 levels, but they may be increasing (not surprisingly, given the rapid increase in motor vehicle activity). Airborne lead levels appear to be declining generally, likely due to lowered lead levels in gasoline and the introduction of unleaded gasoline, but appear to be high at isolated sites. One cannot be sure if this is due to non-transport sources. Most importantly, annual average S P M levels, which are strongly correlated with respiratory and cardiovascular diseases, have been routinely around five times the W H O limit since the 1980s. In terms of short-term exposure, daily average S 0 and N 0 2  2  levels exceed WHO limits on several days of the year, at several sites. 24-hour average S P M levels exceed WHO limits nearly every day of the year, with peak levels as high as 6-10 times the W H O limit at many sites. Based on a limited study, O3 appears to be a major problem, especially in winter, exceeding 8-hour and 1-hour WHO limits at several sites. It is often reported in the media, as in the Toronto Star (1997), that Delhi has the world's fourth worst air pollution. This ranking is immaterial. For example, Delhi could be far more polluted than presently, and be ranked the twentieth most polluted city globally. Regardless of such rankings, one can get a sense of the seriousness of Delhi's air pollution problem by considering that, while the great historical air pollution disasters, in London (in the winters of 1955-1960), and in New York (1953-1964) were characterized by 24-hour S P M concentrations of 1200-3250 and 500-1000 ug/m respectively, and excess mortalities of 53  26  30% (Smith 1994), Delhi's annual average SPM levels are themselves routinely close to 400 ug/m , and daily averages range between 500 and over 1000 ug/m at several sites on many 3  3  days in the year. Further, Delhi's peak O3 levels of 250 ug/m at some sites (Figure 2.10) are 10  3  half the peaks in Los Angeles (Lents and Kelly 1993), which has nearly four times the motor vehicles.  2.3.2  Air Pollutants and their Health Effects  The air pollutants discussed so far - CO, N 0 , S 0 , SPM, Pb, and O3 ~ have health effects 2  2  that range from irritation of eyes and mucous membranes and aggravation of asthma and bronchitis, to damage to circulatory, kidney and nervous systems, and cancer. These health effects are discussed in detail in Appendix II. Of primary concern in terms of health effects are particulates, in particular PMio (and P M . ) . PMio and PM .s are dangerous because they n  2  5  2  remain suspended longer than larger particles (5-15 days for PMi), and penetrate deeper into, and take longer to be cleared from, the respiratory tract. Transport-generated S P M is associated primarily with uncontrolled diesels, but gasoline vehicles including two-strokes, which power the majority of Indian M 2 W vehicles, are an important source as well. The bulk of transport-generated S P M is in the PMio range. Indeed, motor vehicles contribute significantly to PMio and PM2.5 emissions, and proportionately more to these fractions than to total P M . Further, transport PMio consists of a solid carbonaceous core on which unburned HCs, sulphuric acid, sulphates, nitrates, and known carcinogens and/or mutagens such as benzo(a)pyrene and polycyclic aromatic hydrocarbons (PAHs), also transport-generated,  adsorb. Transport-generated particulates also contain respirable lead  (from leaded gasoline) and other metals (Faiz et al 1992; G V R D 1995; G V R D 1998; Walsh 1991a and 1994; Wijetilleke and Karunaratne 1997). Epidemiological studies worldwide have consistently shown that non-accidental deaths and other health end-points, including hospital admissions for chronic respiratory, cardio-vascular and acute respiratory diseases, are closely linked to ambient TSP, PMio,  PM2.5,  and particulate  Interestingly, all of the historical air pollution disasters and episodes of high air-pollution related excess mortality (Smith 1994) seem to have occurred in the winter months. PM2.5, suspended particulate matter of diameter below 2.5 um, is referred to as fine particulates (GVRD 1998). 1 0  1 1  27  sulphates, in addition to SO2 and O3. Particulates are closely linked to various health-points even when no S 0 is present. TSP levels below 100 ug/m were once considered harmless, 3  2  but linking data from the great air pollution disasters and studies relating to levels below 120 ug/m showed no sign of either any threshold or saturation in the dose-response relationship 3  over a range of more than two orders of magnitude from levels in the tens of ug/m . Infants 3  are particularly vulnerable. A study in Brazil in the late 1980s found that the risk for pneumonia deaths, a major component of infant mortality in the LICs, was substantially greater than that for adult mortality, because of the sensitivity of children, and the intimate link between particulates and this end-point. Finally,  PM10 and PM2.5  are even more strongly  correlated with health end-points than TSP. Californian and Canadian studies have found that 85% of air pollution health effects were attributable to P M i (Dockery et al 1993; Fazio 1997; 0  Smith 1994; Walsh 1994a and 1994b). Another transport air pollutant of concern in the Indian context is lead. Nearly all lead in gasoline is released to the atmosphere, mainly as inorganic lead salts (chlorides, bromides and sulphates) and oxides in aerosol form, all in the PM10 range, but mostly as fine particulates. There is a significant relationship between lead in gasoline, airborne lead, and blood-lead. While high concentrations can lead to anaemia, hypertension, heart attacks, strokes, kidney failure, massive and permanent brain damage and death in adults, even low blood-lead levels induce neurological damage and lowered learning ability in children. Infants and children under five are at high risk of excessive lead exposure, because they take in dust and soil through normal mouthing habits. In this connection, it is worth noting that about 10% of 12  transport-generated  lead is deposited within 100 metres of roadsides, where children  frequently play in the LICs. A study in Mexico City, where gasoline vehicles account for the bulk of airborne lead, revealed that children who lived in low traffic areas had a significantly lower blood-lead level than those living close to main roads (CPCB 1992; C S E 1996; Driscoll et al 1992; Faiz et al 1992; Romieu, Weitzenfeld and Finkelman 1991; Romieu et al 1992; Wijetilleke and Karunaratne 1997).  Maternal lead can be transmitted to the fetus via the placenta at blood lead levels as low as 10 (ig/dl, thus affecting gestational age, birth weight, and mental development (CPCB 1992; CSE 1996; Faiz et al 1992; Wijetilleke and Karunaratne 1997). 1 2  28  The economic and social costs of the health effects of lead can be significant. Exposure to even low lead levels can cause neurological and other adverse effects in children, and can persist irreversibly into adulthood even after exposure ends (CPCB 1992; Wijetilleke and Karunaratne 1997). This fact is pertinent in the Indian context, since gasoline lead levels have been lowered and unleaded gasoline introduced only in 1995. Lead levels prior to 1995 were high (at least 0.56 g/L since 1971) (BIS 1995a; CSE 1996; Appendix IV). Indeed, high ambient lead levels were observed even as late as 1992, in areas of heavy traffic in Delhi (CPCB 1992). Lastly, note that a lot of human activity occurs, and food is cooked and eaten, in such areas. As discussed in the previous section, Delhi appears to have an ozone problem, which is likely to intensify with time. North American, European, and Japanese studies have shown that 0 is related to mortality independently of particulates, and to hospital admissions for chronic 3  and acute respiratory and cardio-vascular diseases (Bates 1994; Brandon and Hommann 1995; CPCB 1992, 1995 and 1996; CSE 1996; Faiz et al 1992; Roy Chowdhury 1997; Wijetilleke and Karunaratne 1997; WHO/UNEP 1992). While attention is typically focused on conventional (regulated) pollutants in vehicle exhaust, there are several other transport-generated air pollutants with serious health and other effects, from other vehicular and transport system sources. Metals in motor vehicular exhaust, as well as from clutch, brake and tyre abrasion, include aluminium, arsenic, beryllium, cadmium, chromium, cobalt, manganese and nickel. Many of these metals are known or suspected mutagens and carcinogens. While asbestos-free materials are now the norm for brakes and clutches in the West, asbestos, implicated in mesothelioma, a cancer of the peritoneum with an extremely poor prognosis and long latency periods, continues to be used in these applications in Indian vehicles. Asbestos and other metal emissions from motor vehicles are commonly in the respirable range (Author's interviews 1997; Bates 1994; CSE 1996; Government of Canada 1991; Faiz et al 1992; IARC 1989; Walsh 1991a; Whitelegg 1993). These emissions may be particularly relevant in the case of M 2 W vehicles, because they tend to be used to maneuver through high traffic areas, with consequently excessive brake and clutch use.  29  Benzene, a crude oil constituent, occurs in exhaust and due to fuel evaporation. Refinery processes to compensate for lead removal increase levels of this and other high-octane HCs (such as toluene and xylene, which cause neurological, kidney and liver damage). Short-term exposure to benzene in high concentrations causes respiratory tract inflammation and lung hemorrhaging, and can even be lethal, but the main concern is its linkage to adult leukemia and lung cancer. Other effects include central nervous system damage and birth defects. W H O specifies no safe limit for benzene in air. Benzene levels in gasoline were not controlled until very recently (BIS 1995a; Appendix IV), and are even now considerably higher than those in gasoline in the OECD countries (BIS 1995a; Calvert et al 1993; Mercedes Benz 1997). Aldehydes are another important exhaust product. Though diesels produce them at a greater percentage of total H C , gasoline vehicles produce them too.  13  Formaldehyde is particularly  concerning, because of its photochemical reactivity, and possible carcinogenicity. As will be discussed in subsequent chapters, aldehyde emissions will likely increase in the Indian context, due to the addition of oxygenates to gasoline, to be allowed from 2000 (BIS 1995a; C S E 1996; Government of Canada 1991; Faiz et al 1992; Walsh 1991a and 1994; Wijetilleke and Karunaratne 1997). Fuel additives can have significant impacts of their own. Lead "scavengers" such as ethylene dichloride and dibromide added to leaded gasoline to prevent excessive lead deposits are a concern, because these are emitted in the form of inorganic lead chlorides and bromides in fine P M form. Ethylene dibromide is potentially carcinogenic (Faiz et al 1992; Walsh 1991a).  14  Finally, there are important contextual factors that would likely exacerbate pollution impacts in India. The effects of air pollutants in combination may be far more serious than Diesels emit many toxic air contaminants, such as 1,3-butadiene, dioxins and dibenzofurans, polycyclic aromatic hydrocarbons (PAHs) including benzo(a)pyrene (BaP), and PAH derivatives, such as 3-nitrobenzo(a)pyrene, a powerful mutagen. Atmospheric reactions form new species. 1,3-butadiene can react with OH radicals and ozone to form formaldehyde. In short, transport can contribute significantly to air toxics and associated health risks. In 1990, transport air toxics, including PM, 1,3-butadiene, benzene, and formaldehyde, caused 54-58% of all air toxic related US cancers (Finlayson-Pitts and Pitts 1986; Williams 1989; Walsh 1991a and 1994; Wijetilleke and Karunaratne 1997). Indian data on air toxics is sparse. BaP levels in Delhi were 30-750 ng/m in 1990 (Agrawal 1997; Biswas and Dutta 1994). This is an example of how, in solving one problem, others may be created. Additionally, ethylene dichloride and dibromide also form acids, causing engine and exhaust system rusting and reduced engine oil life. Similarly, calcium and barium additives in diesel suppress visible black smoke, but can significantly increase the PAH content and mutagenicity of SOF, and particulate sulphate emissions (Faiz et al 1992; Walsh 1991a). 1 3  3  1 4  30  singly. Particulate effects are enhanced if high PMio levels are associated with high SO2, N O 2 , and ozone levels, increasingly the situation in Delhi and other Indian cities. Further, in the Indian context, large numbers of people (including infants, the old and the infirm) live and work road-side. This situation, coupled with the fact that air pollution is typically higher in high-traffic areas than farther away, causes much larger exposures for a large number of people than for the general urban population.  15,16  Recall in this context that CO levels at traffic  junctions such as ITO reach 10 mg/m over extended periods, and peak as high as 29 mg/m , 3  3  daily, and that daily average N 0 levels are high, and are likely increasing, at several sites. 2  Recurrent exposure to high N 0 levels is more harmful than continuous exposure to lower 2  concentrations (Barde and Button 1990; Brandon and Hommann 1995; C S E 1996; Faiz et al 1992; Romieu, Weitzenfeld and Finkelman 1991; Romieu et al 1992; WHO/UNEP 1992; Wijetilleke and Karunaratne 1997). Pollution exposure and impacts also depend on activity levels. S P M deposition depends not only on particle size but also on breathing effort. Larger particles are deposited in the exothoracic part of the respiratory tract, and PMio in proximity to the fine airways with normal nasal breathing, but mouth breathing typical of physical activity increases tracheobronchial and pulmonary deposition. Finally, there are synergies between pollution, poverty 17  and nutritional deficiency. For example, diets deficient in calcium, vitamin D , iron and zinc increase lead absorption (CSE 1996; Faiz et al 1992; Romieu, Weitzenfeld and Finkelman 1991; Romieu et al 1992; Wijetilleke and Karunaratne 1997).  Ambient measurements such as those conducted at fixed stations by CPCB in Delhi - high as they are for some pollutants — may actually be under-estimating the risk for a large number of vulnerable people. This may be less of a problem with 0 , which tends to disperse broadly, as opposed to lead and SPM (WHO/UNEP 1992; Wijetilleke and Karunaratne 1997). Interestingly, though, motor vehicle occupants may be at most risk to air pollutants. Air pollution levels inside cars, particularly in terms of CO and VOCs, can be up to five times higher than background concentrations, and much higher than levels to which pedestrians and cyclists are usually exposed (Taylor and Fergusson 1998). Because of the generally poor condition of vehicles, including exhaust system fracture, and the stop-and-go traffic (common in the Indian context), very high, even lethal concentrations of CO could accumulate inside vehicles (Wijetilleke and Karunaratne 1997). Studies in Delhi show that M2W and M3W vehicle riders are exposed to much higher PM and CO levels than car and bus riders (TERI 1997b). In this regard, note that, while 240 ug/m is the USEPA limit for short-term 0 exposure, lung function is adversely affected in the young if they exercise for six hours at 160 ug/m (Wijetilleke and Karunaratne 1997). 1 5  3  1 6  1 7  3  3  3  31  2.3.3  Studies linking Air Pollution and Health and Economic Costs in Delhi  No systematic Indian epidemiological study linking air pollution and health effects exists to the author's knowledge. Brandon and Hommann (1995) of the World Bank made rough estimates of the health and economic costs of air pollution in 36 Indian cities including Delhi. They used annual average air pollutant concentrations from NEERI and CPCB, dose-response data from *  18  US studies, and various conversion factors based on GNP-, medical cost-, and wage-ratios. They concluded, based on the air quality in 1992, that there are around 7500 premature deaths, and four million hospital admissions and illnesses requiring medical treatment annually due to air pollution exceeding WHO limits in Delhi, at an economic cost of around US$ 100400 million. Airborne lead exposure in Delhi was estimated to be responsible for around 41,194 hospital admissions and medical treatments, costing US$ 267,000-667,000, plus US$ 1.5-3.7 million to remedy "children's IQ points loss". The premature deaths, hospital admissions and economic costs in Delhi constituted 19, 20, and 20-30% of the corresponding totals for all 36 cities. More than 95% of the health damage due to air pollution in the 36 cities was estimated to be due to P M . and SO2, with the remainder due to high lead levels 0  (Brandon and Hommann 1995).  19  Cropper et al (1997) recently conducted a study relating daily P M levels to daily deaths in Delhi in 1991-1994. Mortality data were obtained from the New Delhi Municipal Corporation (NDMC), which has a large number of hospitals. 25% of Delhi's mortalities occur in the NDMC.  2 0  Daily mean TSP, S 0 and N O , based on CPCB data, were found to be strongly 2  x  related to cardiac and respiratory diseases. As already indicated, average annual TSP was five times the W H O limit, and daily TSP exceeded the WHO 24-hour limit 97% of the time, during 1991-1994. Surprisingly, the percentage mortality increase in Delhi was found to be only one-third that in the US studies for the same TSP increase. Thus, while applying US The study also estimated the health and economic costs of water and industrial pollution, soil degradation, and deforestation (Brandon and Hommann 1995). Wijetilleke and Karunaratne (1997) observe, in commenting on such studies, that the actual effects of S0 may be underestimated, because typically, sulphates, formed by transformation to acid aerosols, are counted as particulates. Thus, some benefits from reducing PM should properly be attributed to S0 . This point is valid, but only if fuel sulphur is also reduced. However, there is no question that reducing S0 (and lead, which as we have seen, is in the P M i range) will also reduce PM, and more specifically, P M i . The study was based on deaths in NDMC hospitals only, because NDMC was the only jurisdiction with a detailed computerized data base (Cropper et al 1997). 1 8  1 9  2  2  2  0  2 0  0  32  PMio-non-trauma death dose-response coefficients to Delhi, as Brandon and Hommann (1995) did, predicted 7500 premature deaths, Cropper et al, based on actual Delhi data, estimated only 3430 premature deaths, for the same  PMio  levels (Brandon and Hommann  1995; Cropper et al 1997). The above discrepancies highlight the difficulties in transferring dose-response functions from North American and European epidemiological studies to the LIC context. For a variety of reasons including higher pollution levels and exposure, greater diffusion of outdoor air indoors, and a higher percentage of people in marginal health due to lower living standards and poor nutrition, one would have expected that using US or European dose-response functions would underestimate health effects in the L I C context (Brandon and Hommann 1995; Wijetilleke and Karunaratne 1997). Thus, Brandon and Hommann's predicted effects should have been lower than Cropper et al's, but the reverse is the case. There are several possible reasons why P M in Delhi has a smaller effect than in the USA. Delhi's P M may be larger in size, and a larger proportion may be natural rather than combustive in origin. But Delhi's PMi /TSP ratio is only slightly smaller than the US cities 0  average, and its annual P M i is well above even the W H O 24-hour limit, as discussed. The 0  most likely reason is that in Delhi, people die at younger ages and from different causes. While over 70% of all US deaths occur after 65, over 70% of Delhi's occur before 65, with 20% occurring before age five. Further, while 46% of all non-trauma US deaths are due to cardiovascular disease, which is strongly associated with air pollution, both cardio-vascular and respiratory deaths account for only 24% of non-trauma deaths in Delhi (infectious diseases account for as many as 20%). But P M have a smaller impact even on cardio-vascular and respiratory deaths in Delhi. This may be because pneumonia, which has a weaker association with P M than say chronic obstructive pulmonary disease (COPD), comprises a larger fraction of respiratory deaths in Delhi. Also, in the USA, the impact of P M is significant for deaths after, not before, 65. In Delhi, the peak impact occurs in the 15-44 age group, with significant effects also for the 5-14 group (Cropper et al 1997). But this does not mean that P M is not a health hazard in Delhi. The dose-response relationship is flatter than in the USA, but this is likely because other diseases claim Delhi-ites sooner. Further, if life-years lost rather than lives lost is the criterion, Delhi's situation is  33  more serious. Because the largest impacts there occur in the 15-44, as opposed to the 65 plus age group in the US, more life years will be lost in Delhi on average than in the US, for each air-pollution related death (Cropper et al 1997). In any case, medical professionals report that respiratory diseases are increasing rapidly in Delhi, and that five of its 13 millions, including 40% of its children, suffer from them. In a study of 10,000 5-16 year-old school children, Dr. Chhabra of the Patel Chest Institute found that 12% had asthma. Including those who showed symptoms in the past, the figure was closer to 17%. Asthma attacks are reportedly becoming more frequent and severe, including in children, with 5% being fatal compared to only 2% a few years ago. Not surprisingly, traffic policemen are particularly vulnerable. Studies show a greater susceptibility to lung disorders, and significantly larger proportions of decreased lung capacity and abnormal COHb among them as compared to office workers (Basu 1997; Chhabra 1997; C S E 1996; Priti Kumar et al 1997; Roy Chowdhury 1997).  2.4  21  T H E R O L E OF TRANSPORT IN DELHI'S AIR POLLUTION  Figure 2.11 traces the contribution of transport to air pollutant emissions in Delhi over time, and from different information sources (Brandon and Ramankutty 1993; C S E 1996; Faiz et al 1992; WHO/UNEP 1992). Figure 2.12 shows the contribution of transport and other sectors to emissions in the most recent (1996) CPCB emissions inventory (CSE 1996). There are serious discrepancies between the CPCB inventory and that in Faiz et al (1992). Besides, it is unclear what methodology was used in each case. Given the rapid growth in motor vehicle activity in Delhi, it is surprising that while the contribution of transport to H C appears to be increasing, its contribution to CO and N O is not (Figure 2.11). Further, while it x  is unsurprising that power generation and industry account for the bulk of S P M and SO2, it is surprising that the contribution of transport, particularly to SPM, is decreasing, as the CPCB data for 1996 appears to suggest, given that transport energy consumption is increasing rapidly, and no control actions have been taken in that sector. There is a need for a reliable  Dr. Naresh Trehan of Delhi's Escorts Heart Institute can apparently identify Delhi inhabitants easily; while patients from outside have pink lungs, those from Delhi have black ones (CSE 1996).  34  inventory of emissions from transport and other sectors.  Regardless of the discrepancies,  however, the figures show the predominance of motor vehicles in terms of CO, H C and N O . x  The role of motor vehicles, even in terms of SPM and SO2 is growing with time (Faiz et al 1992).  23  The rapid growth in Delhi's motor vehicle activity since the 1970s, particularly in terms of M 2 W vehicles, has been discussed. The majority of Delhi's M 2 W (and M3W) vehicles are powered by highly polluting two-stroke engines.  24  Many of the city's buses (all diesel-  operated), and goods vehicles (predominantly diesel-operated), are of old vintage and poorly maintained, and are gross polluters, particularly in terms of P M . Though vehicular emission standards have been progressively tightened in the 1990s (Appendix III), and those related to M 2 W vehicles proposed for 2000 are some of the strictest in the world, many in-use M2W and M3W vehicles, and also cars, particularly those manufactured prior to 1991, pollute heavily ( A I A M 1994a, 1995 and 1997; C S E 1996; Faiz et al 1992; Faiz et al 1996; Shah and Nagpal 1997; Narayana 1994). Two-stroke engines are preferred for M2W and M 3 W vehicles because they are simple in design, have high power/swept volume and power/weight ratios, and are relatively inexpensive to own, operate, maintain and service. But since exhaust and intake events occur simultaneously in two-strokes, 20-30% of the fuel-air charge escapes unburned through the exhaust. Further, lubricating oil is added to the fuel to lubricate engine parts, since the fuel-air charge is drawn through a "dry" crankcase. The oil is "lost" due to combustion, but does not burn completely. These factors in combination result in high H C and P M emissions (Faiz et al 1992).  25  WHO/UNEP (1992) also remarks on discrepancies between the NEERI and CPCB emissions inventories. Apart from many other problems that will be discussed shortly (and in subsequent chapters), the CPCB transport emissions inventory (CSE 1996) likely accounts only for emissions from vehicular exhaust, not from other vehicular and transport system sources. It is certainly interesting that air quality monitoring in Delhi should have started around the time motor vehicles began to grow rapidly. While all the M2W vehicles are gasoline powered, 97% of Delhi's M3 W vehicle fleet are. Many para-transit vehicles also employ two-stroke engines (Cervero 1997; Faiz et al 1992; MoEF 1997b). Though catalytically controlled spark-ignition (gasoline) vehicles emit 50-80 and 100-120 times less PM than light- and heavy-duty diesels respectively, spark-ignition PM can be significant where poor maintenance or engine wear lead to high oil consumption, and/or where (as in two-strokes), oil is mixed with the fuel. Further, inorganic lead salts are an important exhaust component on engines operated on leaded gasoline, as 2 3  2 4  2 5  35  Fuel and lubricating oil quality also contribute significantly to the air pollution problem. Their quality is being improved in a phased manner (BIS 1995a; Appendix I V ) ,  26  but lead  content was as high as 0.56 g/L in 87 octane (and 0.8 g/L in 93 octane) gasoline until 1995. Similarly, gasoline and diesel sulphur were as high as 0.2 and 1% by weight respectively until 1995 (BIS 1995a; BIS 1995b) Fuel and lubricating oil adulteration are also an important 27  factor, particularly in the case of M3W and commercial vehicles (Raje and Malhotra 1997). Table 2.1 compares the exhaust emission factors measured on in-use M 2 W and M 3 W vehicles, cars and urban buses in the early 1990s (IIP 1994; Shah and Nagpal 1997). These measurements are relevant even today because many vehicles from that period are likely still operational (of course, fuel and oil quality have improved since then). M 2 W and M 3 W 28  vehicles carrying just 1-2 persons produce higher CO and H C and one-fourth the P M per kilometre relative to a bus (Table 2.1), despite the bus having a much larger, more powerful engine, and carrying over 40 people. Effectively then, in terms of CO and HC, Delhi now has more than two million buses, each of which carries only 1-2 people.  already noted. Finally, the mutagenicity of gasoline soluble organic fraction (SOF) per unit mass is greater than for diesel SOF (Faiz et al 1992; Lowenthal 1994; McClellan 1986; Walsh 1991a and 1994). Governments, both at the national and local levels, have instituted or are contemplating several measures in addition to progressively stringent vehicle emission standards and fuel and lubricating oil quality improvements (Appendices III and IV). These measures include enforcement of in-use emissions standards in Delhi (and in some other Indian cities), by means of a road-side no-load pollution check, on the basis of which a "Pollution Under Control" sticker is issued, scrappage of old vehicles in a phased manner in Delhi, alternative fuels, and the introduction of a mass rapid transit system (MRTS) in Delhi and some other cities (CSE 1996 and 1997; MoEF 1997b; NCTD 1997a, b and c). Fuel sulphur is an important component of exhaust PM. While approximately 98% of diesel sulphur is emitted as S0 and 2% as particulate metal sulphates, the latter can contribute up to 14% of diesel PM mass. Further, sulphur oxides from diesels may undergo reactions in the atmosphere to form acidic sulphates and sulphuric acid (Faiz et al 1992; Lowenthal 1994; Pierson 1988; Truex et al 1980). However, there are many data gaps, uncertainties, and discrepancies in Indian emission factors (Bose 1996; IIP 1994; NCTD 1996; AIAM 1997a). Some data (NCTD 1996) simply reproduces 1973 USEPA data for US vehicles. The IIP (1994) data, perhaps the most dependable, presents actual emission measurements on used vehicles. But this is not without its problems either. These problems are detailed in Chapter III. Table 2.1 is based on IIP (1994) data, supplemented with average PM values for M2W and M3 W vehicles, and buses, from Shah and Nagpal (1997). They quote data from South and South-East Asia showing P M emission factors ranging from 0.2 g/km to as high as 2 g/km in poorly maintained M2W and M3W two-stroke vehicles using poor quality lubricating oil, and from 0.75 g/km to 8 g/km in "smoke belcher" buses. 2 6  2 7  2  2 8  Figure 2.11  Transport Share of Air Pollutant Emissions in Delhi Data from NEERI and CPCB, 1981-1996  100  Transport Share % • NEER11981 E3NEERI1987 @CPCB 1996  80  60  40  20  CO  HC  NOx  SOx  NEERI - National Environmental Engineering Research Institute. C P C B - Central Pollution Control Board. Sources: Brandon and Ramankutty 1983; C S E 1996; Faiz et al 1992; WHO/UNEP 1992.  Figure 2.12  PM  mgb99  Air Pollutant Emissions Inventory for Delhi CPCB, 1996  100  % Share  80  60  40  20  CO  HC  C P C B - Central Pollution Control Board. Source: C S E 1998  NOx  SOx  PM  mgb99  Table 2.1  In-Use Indian Motor Vehicular Emission Factors Modal Comparison  Emission Factors, g/km  M2W  M3W  Car  Bus  CO  6.S  8.6  24.9  $  HC  3.9  ipiillliiilllll  5  a.t  NOx  0.03  0.26  2.01  S02  0.013  0.029  0.053  1.44  PM  0.5 (Range 0.2-2)  0.5 (Range 0.2*2)  0.33  2 (Range 0.75-8)  M2W/M3WB Motorized twoVthree-wheeled vehicles. Sources: Bose 1996; Shah and Nagpal 1997; IIP 1994.  Figure 2.13  mgb99  Motor Vehicular Exhaust Emissions in Delhi Modal Shares  59.3%  SOX  PM  M2W/M3W « Motorized two-rthree-wheeled vehicles. Calculations based on data drawn from ASRTU/CIRT 1997; Bose 1996; C P C B 1997; IIP 1994; Shsh and Nagpal 1997. mgb99  Figure 2.14  Passenger Motor Vehicular VKM and PKM in Delhi Modal Comparison E3VKM • PKM  M3W  d  i  Car/Jeep  J  Bus  I  0  20  40  60  80  % Share  VKM - vehicle-kilometres; PKM - passenger-kilometres. M2W7M3W ° Motorized two-rthree-wheeled vehicles. Calculations based on vehicle population, daily kilometre, and occupancy data from ASRTU/CIRT 1997; Boss 1996; C P C B 1997; Gol/ESCAP 1992. mgb99  Table 2.2  Emission Factors for In-Use Indian Cars and Buses  Compared to M2W Vehicle  Bus, g/veh-km ' Bus, g/pass-km  Index: M2W= 1  Car, g/veh-km  Car, g/pass-km  CO  3.8  2.3  0.8  0.03  HC  1.3  0.8  0.5  0.02  PM  0.7  0.42  4  0.15  NOx  67  41.1  323.3  S02  4.1  2.5  110.8  I-  12.&  4.1  M2W > Motorized two-wheeled vehicle. Calculations based on data from ASRTU/CIRT 1997; Bose 1996; C P C B 1997; IIP 1994; Shah and Nagpal 1997. mgb99  39  Because of their exceedingly high emissions per vehicle-kilometre, and the fact that they are used intensively, M 2 W vehicles account for significant proportions of CO, H C and P M exhaust emissions from motor vehicle activity in Delhi (Figure 2.13). Their contribution is marginal only in terms of N O and SO , for which buses (and other diesel-powered vehicles) x  are primarily responsible.  x  29  Further, buses account for only around 10% of vehicle-kilometres, but 71% of passengerkilometres, in motorized passenger vehicles in Delhi. On the other hand, M 2 W vehicles account for 60% of vehicle-kilometres, and only 16% of passenger-kilometres, in such vehicles (Figure 2.14). Thus, on an emissions per passenger-kilometre basis, M 2 W vehicles produce roughly 33, 50 and 7 times the amount of CO, HC, and P M , and one-fourth the SO , x  that buses do (Table 2.2). To conclude this section, the share of transport is high in terms of CO, H C and N O  x  emissions in Delhi, and its role in terms of S P M and SO2 is growing with time. Further, the share of transport in high traffic areas, where large numbers of people live and work, is considerably higher than on a city-wide basis, thus causing significant exposures and health risks. M 2 W vehicles are an important transport source of CO, H C and P M , particularly on an emissions per passenger-kilometre basis. While not ignoring the key role of diesel vehicles, particularly in terms of N O , SO and P M , this fact highlights the need to direct public policy x  x  attention to M 2 W vehicle emissions in Delhi (and in the Indian context generally).  2.5  DELHI'S MOTOR VEHICLE ACTIVITY AND AIR POLLUTION IN A  WIDER CONTEXT So far, we have discussed air pollution and its local health effects in Delhi, and the increasingly important role of transport in general, and M2W vehicles in particular. In order to more strongly justify public policy attention, Delhi's transport air pollution problem is shown to be shared by many other Indian and LIC cities in the following discussion. Further,  It must be stressed that all the calculations in this section focus on exhaust emissions alone. In addition to the other problems, Indian transport emissions inventories (such as that conducted by CPCB — CSE 1996) do not appear to take into account crankcase and evaporative emissions, which are likely significant in the Indian context, nor PM due to brake, clutch and tyre wear, and re-suspended dust. Emissions from these sources would need to be included for a truly comprehensive accounting of transport emissions.  40  transport air pollution and energy consumption have many non-health welfare effects, locally as well as regionally and globally. The growing role of transport in India and the other LICs in terms of climate change, acidification and energy security is discussed. Finally, Delhi's transport air pollution problem is placed in the context of broader urban transport and urbanization impacts, as a means of highlighting the public policy challenge.  2.5.1  Motor Vehicle Activity and Urban Air Pollution in India and Other LICs  Delhi is by no means unique, either in terms of rapid growth in motor vehicles, or the predominance of M 2 W vehicles. Motor vehicle growth has been at least as rapid in other Indian cities. Between 1971 and 1995, motor vehicle numbers increased more rapidly in Ahmedabad, Bangalore and Chennai than in Delhi. In fact, Chennai's motor vehicle fleet increased from only 120,000 to 768,000 in that period. While M 2 W vehicles form 67% of Delhi's motor vehicle fleet, they represent 70-80% of the fleet in Ahmedabad, Bangalore, Chennai, and Hyderabad. Only Calcutta and Mumbai have a low percentage (43%) in this respect (ASRTU/CIRT 1997). In India as a whole, motor vehicle numbers increased from only 665,000 in 1961, and 5.4 millions as late as 1981, to over 27 million in 1994 (Figure 2.15). At 8% pa (per annum) growth, this number could touch 43 million by 2000 ( A I A M 1994a; TERI 1997a). As in Delhi, M 2 W vehicles are the most rapidly growing vehicle type country-wide (44% pa in the 1980s). M 2 W vehicles represent nearly 80% of all motor vehicles produced and sold, and 67% of those registered, nationally. India arguably has the largest population of this vehicle type of any country. Even in 1988, India accounted for only 0.04% and 1.3% of the world's cars and trucks/buses, but 8.4% of global M2W (and M3W) vehicles ( A I A M 1994a and 1995; ASRTU/CIRT 1997; Faiz et al 1992; TERI 1997a). The motor vehicle trends in India are shared by many other Asian LICs. While the growth in motor vehicles is rapid globally, it is most so in Asia, as markets become saturated in the OECD, and production moves to large untapped markets in Asia, where policies now encourage investment. Though Asia accounts for a small proportion of the global fleet, motor  41  vehicles are doubling every seven years in many of its nations.  30  By 2040, rapidly  industrializing LICs could have as many vehicles as North America and Western Europe (Champagne 1998; Faiz et al 1992; Mohan, Tiwari and Kanungo 1997; Walsh 1991a and 1994). South and East Asia collectively had only 8.5% of the world's motor vehicles, but 36% of its M 2 W and M 3 W vehicles, in 1988. M 2 W and M3W vehicles make up more than 50% of the fleet in Bangladesh, Thailand, Malaysia, Indonesia and Taiwan. In Bangkok, M 2 W vehicles increased three-fold in just six years from 1980, and might reach three million by 2000, i f trends follow those in Taipei, where each household owns at least one such vehicle (Faiz et al 1992; Poboon et al 1994). As in the Indian context, motor vehicle activity in other LICs is characterized by restricted technological, financial and administrative resources for manufacturing and maintaining clean vehicles and fuels, providing transport infrastructure, public transit and transport system management, and for policy-making and implementation. Also, unlike in the West, motor vehicle activity in the LICs is typically concentrated in one or a few major cities. In Iran, Korea, Mexico, Philippines, and Thailand, the capital cities account for 40-50% of national automobile fleets (Faiz et al 1992; Faiz and Aloisi de Larderel 1993; Sathaye, Tyler, and Goldman 1994; Shalizi and Carbajo 1994; WHO/UNEP 1992).  31  The role of the LICs is currently not significant in terms of global motor vehicle numbers and motor vehicular CO, H C and N O emissions (10-25%), but their share of these pollutants, x  and of global motor vehicular SO2 and P M , already considerable (40-60%), is expected to grow significantly, even with progressive controls. The OECD, on the other hand, will likely experience reductions (Faiz et al 1992; Walsh 1994).  32  Motor vehicle numbers increased from a mere 40,000 in 1971 to nearly 400,000 in 1993 in Bangladesh, and from 177,000 in 1970 to 1.24 million in 1995 in Sri Lanka, thus giving an annual growth rate of 40 and 25% in the two countries (Mohan, Tiwari and Kanungo 1997). This concentration is more accentuated for cars than for other modes, reflecting the concentration of economic power. While 76% of Thailand's M2W vehicles were registered outside Bangkok, only 23% cars were. India is in a much better position. 35% of its motor vehicles operate in 23 metropolitan centres. However, Bangalore, Delhi, Calcutta, Chennai and Mumbai jointly account for 17% of the national motor vehicle fleet (ASRTU/CIRT 1997; Sathaye, Tyler, and Goldman 1994). M2W vehicles (mostly in the LICs), will likely contribute significantly to increased HC emissions globally, and trucks will be a rapidly growing contributor to global NO and PM (Walsh 1994). 3 0  31  3 2  x  42  The share of transport in air pollution in LIC cities is significant and growing. The concentration of national motor vehicular activity in these cities is certainly a factor. Air pollution in many L I C (including Asian) cities, already serious, could soon rival Mexico City's, where levels of several air pollutants already exceed WHO limits by more than a factor of two. 12 of the 15 cities globally with the highest S P M levels, and the six with the highest SO2 levels are in Asia (these include Delhi, Beijing, Calcutta, Jakarta, Shanghai, Seoul, Manila, and Bangkok). Lead levels have been high in many Indian (and Asian LIC) cities, 33  with blood lead three times that in the OECD. Respiratory diseases, the second commonest cause of death in children under five in the LICs, kill more than four million globally annually (Agrawal 1997; Brandon and Ramankutty 1993; Faiz et al 1992; Faiz and Aloisi de Larderel 1993; Smith 1988 and 1994; Walsh 1994; WHO/UNEP 1992).  2.5.2  34  Transport, Energy Security, and Climate Change  Delhi consumes nearly 10% of the nation's gasoline, 3.3% of its diesel (and 4% of its petroleum products), with only 1.4% of its population. Annual growth rates for these categories range from 5-8%. Nationally, industry continues to consume the largest share (50%) of India's commercial energy, whereas transport consumes around 24% (as opposed to 5% in China). However, energy intensity is reducing in the former, and increasing at 7% annually in the latter sector, less rapidly than in only the domestic sector. Further, as in other countries, transport is a major consumer of petroleum products (44 and 51% in 1985 and 1995) (Brandon and Ramankutty 1993; Gol/ESCAP 1991; TERI 1997a). Diesel is the principal Indian petroleum product consumed (43%, as against 6% for gasoline), the primary source of transport energy (75%, as against 25% for gasoline) and the fastest consumed petroleum product since the 1970s (11.8% annually, as against 6% for gasoline). But passenger movement by private modes (predominantly urban and gasolinepowered), currently only 20% of passenger movement by road, is likely to increase much  Monthly SPM and S0 maxima in Calcutta are 891 and 300 ug/m , and annual average SPM levels as high as 1000-1700 ug/m have been recorded at traffic junctions (Priti Kumar et al 1997). It is not merely major Indian and LIC cities that have severe air pollution. Air pollution and per capita air pollution effects are significantly higher in secondary Indian cities than in Delhi. Over 90% of CPCB stations nationwide record annual SPM levels exceeding the WHO standard, with S0 levels doing likewise in many cities (Brandon and Hommann 1995; CPCB 1992 and 1996). 3 3  3  2  3  3 4  2  43  faster (13% annually) than by public modes. Also, freight is expected to grow slowly. As a result of these trends, gasoline consumption in transport will likely increase twice as rapidly as diesel. Incidentally, M2W vehicles already consume nearly 50% of Indian gasoline. In 2009, gasoline will likely contribute one-third and diesel two-thirds, of transport  energy  (GoI/ESCAP 1991; Faiz et al 1996; TERI 1997a). If motor vehicle activity trends persist, transport energy demand will increase dramatically. This trend will likely have significant energy and political security implications. Petroleum 35  product consumption increased three times between 1970 and 1990 (Figure 2.16). The gap between local oil production and refining capacity and demand is rising rapidly, and has to be met through imports. India's vulnerability to world oil prices and crises is borne out by the fact that while imports increased 9% annually in the 1970s, their cost increased 389%. The net oil import bill as a percentage of net export earnings increased from 8% in the 1960s, to over 30% and 75% in 1973/74 and 1980/81 respectively, and currently stands at 3 2 % .  36,37  The  future is worrisome, given rapid demand growth, and the possibility that Indian oil reserves will last only 25 years at current production levels (GoI/ESCAP 1991; TERI 1997a). At the global level, transport already consumes about 25% of commercial energy, and is the fastest growing end-use category. The OECD, with 15% of world population, accounted for nearly two-thirds of global commercial energy consumption for transport in 1995, but its demand over the next two or three decades is expected to be flat or growing slowly. In contrast, transport energy demand in the LICs, .now around one-third that in the OECD countries, could increase two to three times in as many decades. This has serious security implications, since transport already consumes around 45% of the world's oil. India's and China's combined oil consumption accounts for only 6.6% of the world's, but is increasing at 6.4% annually, while world oil consumption is increasing at 1.5%.  Energy consumption is increasing rapidly in all sectors. It is currently doubling every 12 years in India (and Asia), as opposed to 28 years globally (Brandon and Ramankutty 1993). Besides, oil imports alone account for nearly a quarter of all Indian imports (Nag 1997). The predominant role of diesel in Indian transport is of concern because of its implications for air pollution health effects. Further, diesel accounts for 62% of petroleum imports, which are vulnerable to price fluctuations, and it has also been heavily subsidized (GoI/ESCAP 1991; Nag 1997; TERI 1997a). 3 5  3 6  3 7  Figure 2.15  Motor Vehicle Growth in India  1961-1994 Motor Vehicles, Millions BSBuses • Goods E3Cars+Jeeps+Taxis E3M2W ED Others  1961  1971  1981  1991  1994  Year M2W = Motorized two-wheeled vehicles. Sourcee: AIAM 1894a; TERI 1997a.  Figure 2.16  mgt>99  Petroleum Products Consumption in India 1970-1995  70  Consumption, 1000 Tonnes i  Gasoline  Source: TERI 1997a.  Diesel  Petroleum Products  mgb99  45  At this rate, two-thirds of global oil will likely have to come from the politically volatile Middle East, as against only 26% in 1991. If, as some scenarios suggest, the global fleet grows to five billion motor vehicles by the end of the 21st century, transport alone would require, even with efficiency improvements, up to 50% more than the present total oil consumption world-wide (Flavin and Lenssen 1991; Grubler 1994; Holdren 1990). The rapid growth in motor vehicle activity and transport energy demand also have important implications for climate change. Globally, motor vehicles contribute significantly to emissions of gases implicated in climate change, including C O 2 and CFCs (the two major climate change gases), tropospheric O3, nitrous oxide, methane, and CO. Transport accounts 38  for around 14% of C O 2 and 25% of CFC emissions globally, and around 15% of all global warming. The share of non-OECD countries, including the LICs, is only 33 and 10% of global motor vehicle generated C O 2 and CFCs respectively. However, with rapid growth in motor vehicle activity, C O 2 emissions from global transport will likely be two-thirds above current levels by 2030, with nearly half coming from non-OECD countries. In these countries, transport C 0 emissions could increase thrice in as many decades, even with fuel economy 2  improvements. Of course, even with this increase, their share will barely equal that of the 39  OECD.  40  2.5.3  Regional Acidification and Tropospheric Ozone  Acidification results from S 0 and N O being transported by prevailing winds and returned to 2  x  earth in wet or dry form. Typically, sulphuric acid makes up 60-70% of total acidity, and nitric acid the balance. While transport contributes insignificantly to SO2, its contribution to N O is x  CFCs are not caused by M2W vehicles, the focus of this study. Indeed, CFCs are not normally associated with transport. However, CFCs from motor vehicle air-conditioning, and other automotive systems, are a major contributor to overall CFC emissions. A significant contributor to climate change, CFCs are also predominantly responsible for stratospheric ozone depletion (Faiz et al 1992; Government of Canada 1991; Walsh 1994). Cars will likely dominate in terms of C 0 . However, heavy-duty trucks, accounting for under 10% of the global fleet but almost 20% of motor vehicle-kilometres, and 25% of C 0 from road transport, will be a rapidly growing contributor, due to minimal fuel efficiency improvements. By 2030, they will likely be the dominant vehicular C 0 source (Faiz et al 1992; MacKenzie and Walsh 1990; Walsh 1991a). Thus, as Faiz et al (1992) point out, policy in most LICs would be driven by energy conservation, rather than global environmental, considerations. 3 8  3 9  2  2  2  4 0  46  substantial. O3 and its precursors (a growing problem in Delhi and other LIC cities) also travel great distances, so that high levels can occur far from major sources. For example, while O3 levels exceed WHO limits by up to 9% within Delhi, they typically do so by 20% outside (Faiz et al 1992; Roy Chowdhury 1997; Walsh 1994). Local and regional impacts due to acidification and ground level 0 (and P M ) include 3  damage to soils, vegetation and crops, and forest and aquatic ecosystems, groundwater pollution due to toxic metal leaching, impaired visibility, metal corrosion, and structural damage to buildings and monuments (the Taj Mahal is a prominent victim). Acidification and 0  3  effects in the LICs, particularly in Asia, are increasing rapidly. India and China are  particularly prone to high emissions and depositions. While this is the case, even low O3 41  levels can seriously diminish crop yields. More importantly, O3 appears to affect tropical crops more severely than US and European ones. While damage is estimated to be 10% in the U S A (except for sensitive crops in California), it could be 40% for wheat, soybean, rice and groundnut in the North Indian bread-basket states of Punjab and Haryana (Brandon and Ramankutty 1993; Faiz et al 1992; Roychowdhury 1997; Walsh 1994).  2.5.4 Other Impacts of Motor Vehicle Activity and Urbanization This dissertation focuses on transport air pollution. However, it is important to keep in mind that while air pollution might be the most widely felt impact of motor vehicle activity, there are other important transport impacts, and that air pollution is inextricably linked with them. Delhi's (and India's) road safety record, already among of the world's worst, are deteriorating steadily. In 1993, Delhi accounted for 19% of all road accidents, and 36% of all resulting mortalities, in 12 major Indian cities (interestingly, Calcutta and Bombay had more accidents, but far fewer mortalities). Nationwide, road accidents are the primary cause of accidental deaths. They caused 54,058 and 60,595 mortalities in 1990 and 1993 respectively, as many as in North America, and with a fraction of its motor vehicle activity. While car occupants accounted for only 12% of Delhi's road accident fatalities in 1994, pedestrians, cyclists and M 2 W vehicle users accounted for 42, 14 and 27% ( A I A M 1995; ASRTU/CIRT  Asia's S 0 emissions, mainly from coal-fired power plants and industry, but also domestic sources, could easily exceed the OECD's by 2010 (Brandon and Ramankutty 1993).  4 1  2  47  1997; Mohan and Tiwari 1997; Mohan, Tiwari and Kanungo 1997; Stackhouse 1995; TERI 1997a). '  42 43  Peak-hour speeds are dropping with growing motor vehicle activity, and now average 1020 km/h in Delhi, Bombay, Calcutta and many other Asian LIC cities. Besides causing time 44  and productivity losses, congestion significantly exacerbates air pollution and energy consumption (Brandon and Ramankutty 1993; CSE 1996; Faiz et al 1992; Sathaye, Tyler, and Goldman 1994; Poboon et al 1994). Since congestion involves a large number of vehicles moving in a "stop-and-go" fashion over narrow corridors, the net result in terms of local pollutant levels is bound to be dramatic. As motor vehicle activity, sprawl, and accommodation of motor vehicles feed on each other, N M V s become less viable, and access and mobility for pedestrians, N M V users, and for those too poor to afford even the least expensive modes, suffers. Even in Delhi, around 50% of the population lives in slums, and many residential households do not possess even bicycles. The trend in the LICs has been a significant reduction in non-motorized vehicle ( N M V ) shares. In Delhi (Figure 2.17), bicycle trips dramatically reduced from 36 to 7% of trips by all mechanical modes between 1957 and 1994. In the western Indian city of Pune, households with bicycles fell from 61 to 29%, while those with M 2 W vehicles rose from 17 to 41%, between 1982 and 1989 (Hillman 1990; Whitelegg 1993; Midgley 1994; Pendakur 1987 and 1988; Replogle 1992; RITES/ORG 1994; Sathaye, Tyler, and Goldman 1994). Though generally overlooked, transport wastes include used crankcase oils, lubricants, transmission and brake fluids, coolants, automobile batteries and acids, filters, tyres and solvents generated during vehicle operation, maintenance, servicing and disposal. Many of these wastes are hazardous, contain carcinogens and/or neurotoxins, and when disposed of  Other LIC countries show a similar trend. In Bangladesh, Sri Lanka, and Thailand, 53-72% of such fatalities are pedestrians and NMV users. In the USA, by contrast, less than 20% are pedestrians or cyclists. In Sri Lanka, Malaysia and Thailand, M2W vehicle users account for 34-57% of road fatalities. The high shares of NMV and M2W vehicle users in countries like India are of course due to their high shares in travel activity. Additionally, there is little separation between motorized and non-motorized modes and sidewalk activity, and traffic regulations are hardly enforced (Mohan and Tiwari 1997; Mohan, Tiwari and Kanungo 1997). Finally, M2W vehicles are unstable, and are rendered more so by being used to carry passengers and/or goods. In addition to the disproportionate representation of pedestrians and cyclists, there is also a dramatic skew in terms of gender. More than 80% of Indian road fatalities are male (Mohan, Tiwari and Kanungo 1997). However, we must be careful when comparing average speeds in different contexts, because the share of walking and non-motorized trips, which of course lower average speeds, can vary widely across contexts. 4 2  4 3  4 4  48  indiscriminately (there are virtually no controls in the Indian context), can either persist in the air, water or soil, or be absorbed by the lower organisms, and travel up the food chain. As the number of under-ground fuel storage tanks increase with growing motor vehicle activity, so will the likelihood of leakage and contamination of ground water (California Environmental Protection Agency 1988; Government of Canada 1991).  Figure 2.17  Mechanical Mode Share Changes in Delhi 1957-1994  70  % Mode Share  60 50 40 30 20  —  10  Bicycle  M2W  M3W  Car  Taxi  M2W/M3W - Motorized two-rthree-wheeled vehicles. Source: RITES/ORG 1994.  Bus  Rail  Others  mgb 99  Traffic noise, including that due to frequent horn usage (because of the chaotic traffic), is high, and increasing. Levels of the order of 80 dB(A), typical of road traffic, can induce both physiological as well as psychological effects (Barde and Button 1990). Finally: associated with high motor vehicle activity levels is the operation and maintenance of various facilities such as traffic control, policing, and street maintenance, activities that are resource-intensive and which require transport systems of their own. Transport is by no means the only source of air pollution, and neither is outdoor urban air pollution the only air pollution problem. The burning of fuelwood and other 'traditional' fuels by the vast majority for cooking and space heating creates indoor P M levels of 6000 ug/m , 3  49  and health impacts far more severe than in even the most polluted urban environments. Not only the vast majority of rural inhabitants, but also a significant number of low-income urban households, rely on these fuels. ' 45  4 6  Thus, the urban poor receive very high P M exposures  from city-wide as well as indoor sources (Brandon and Ramankutty 1993; GoI/ESCAP 1991; Priti Kumar et al 1997; Smith 1988 and 1994; TERI 1997a). Neither air pollution nor the other transport impacts discussed are the only serious urban problems in Delhi and other Indian and LIC cities. Water pollution due to ineffective sewage and human waste disposal, and other effluents, is perhaps the most widespread urban environmental problem in India (and Asia). Coupled with inadequate and overcrowded housing and poor solid waste disposal, water pollution causes water and vector borne diseases that are responsible for millions of mortalities and morbidities annually, mainly among children. It is worth noting in this connection that water pollution alone accounts for 59% of total health impacts due to environmental pollution in India, as against only 14% for air pollution. Infectious and parasitic diseases account for 18-19% of all deaths in India, and 27% of deaths among children under five, and are increasing rapidly. Thus, the scale and range of problems in Indian (and LIC) cities is massive, and many of these do far greater, and more easily preventable, damage to human health than air pollution, important as that is (Brandon and Ramankutty 1993; Brandon and Hommann 1995; Hardoy and Satterthwaite 1991; Mohan, Tiwari and Kanungo 1997).  2.6  SUMMARY AND CONCLUSIONS  Motor vehicle activity is increasing rapidly in Delhi. Also, Delhi's air quality is poor, and deteriorating. The contribution of transport to the city's air pollution is significant and  In India, these fuels even now account for 75% of total household energy use (90 and 40% in rural and urban areas), and over one-third of final energy consumption, with coal, soft coke, and charcoal accounting for another 24%. In addition to adverse health impacts, these fuels are a major cause of deforestation. Also, for example, burning dung instead of using it as fertilizer deprives the soil of nutrients (Brandon and Ramankutty 1993; GoI/ESCAP 1991; TERI 1997a). In the Indian context, the effect of high particulate levels is exacerbated by widespread smoking, which affects non-smokers only marginally less than it does smokers. Also, the growing electrical power shortage has resulted in increased use of portable generators, many of which are powered by two-stroke gasoline and diesel engines (CSE 1996). Though generators may not be significant in terms of total pollution load, they tend to be used over prolonged periods daily in residences and other enclosed areas such as markets, and are therefore of concern in terms of total human exposure and health impacts.  4 5  4 6  50  growing. M 2 W vehicles account for the lion's share of the city's motor vehicle activity. Because of this fact, and their exceedingly high emission factors, M 2 W vehicles contribute significantly to transport air pollution, particularly on a per passenger-kilometre basis. At the same time, M 2 W vehicles provide affordable mobility to millions who have few other attractive options. Transport air pollution poses a particularly difficult public policy challenge. It is an externality that transport system users impose on non-users, as opposed to congestion, which primarily affects users, who at least benefit from motor vehicles (Hanson 1992; MacKenzie et al 1992; O E C D 1992). In Indian cities, a majority of the non-users are poor and enjoy none of the benefits of motor vehicles, while involuntarily bearing the brunt of its impacts. Chronic health and welfare effects of long term exposure to multiple transport pollutants may be far more serious than those due to acute episodes involving stationary sources, and are far more difficult to document and quantify. Besides, pollutant emissions from a large number and variety of motor vehicles are far more difficult to control than from stationary sources (Faiz et al 1992). Finally, transport is arguably the most complex sector in terms of human behavioural issues. In the case of M 2 W vehicles in India, the additional challenge is how to address their emissions while minimizing adverse policy impacts for vehicle users. It is this challenge, and the challenge that transport air pollution poses generally, that provides the rationale for this dissertation. Delhi's rapid growth in motor vehicle activity, particularly in terms of M 2 W vehicles, and its deteriorating air quality, are features shared by many other Indian and LIC cities. Transport emissions, already significant in terms of the rapidly worsening urban air pollution in the LICs, will likely become even more so. Additionally, transport energy demand in the LICs could increase as much as three times in as many decades. Addressing transport-energy-air pollution linkages in India and other LICs is therefore important, in terms of local well-being as well as regional and global issues such as energy security, acidification and climate change. Focusing on M 2 W vehicle emissions in Delhi, as this dissertation does, is all the more relevant in light of the above considerations. The effort to address transport air pollution must recognize that air pollution is by no means the only important transport impact. Neither is transport the only source of air  51  pollution. Further, neither transport air pollution nor the other transport impacts are the only serious urban problems in Delhi and other Indian and LIC cities. Lastly, the Indian and L I C contexts are characterized by multiple urgent demands on meagre technical, financial and administrative resources.  52  CHAPTER m TRANSPORT AIR POLLUTION IN INDIA: A DISCUSSION OF CONTRIBUTORY FACTORS 3.1  INTRODUCTION  In order to gain a good understanding of, and to effectively address, any problem, it is important to investigate the various factors that contribute to it. This is particularly true of complex and multi-dimensional public policy challenges such as transport air pollution. In Chapter II, reference was made to some of the critical factors contributing to air pollutant emissions from M 2 W and other vehicles in Delhi, including outdated vehicle technology, poor vehicle maintenance, poor quality fuel and oil, and fuel and oil adulteration. But there are many other contributory factors. Also, addressing the transport air pollution problem in India requires an understanding of important contextual characteristics that critically influence emissions. This chapter discusses in some detail the complex of inter-locking factors that contribute to M 2 W vehicle air pollutant emissions in Delhi. This discussion can help identify policy levers to target key contributory factors. Just as importantly, it can help identify factors that are impervious to policy interventions, and must therefore be accepted as constraints. Lastly, identifying critical contributory factors will help in making transport emissions and energy consumption measurement and modeling efforts more effective. While the discussion focuses primarily on M2W vehicles in Delhi, it is relevant to the overall transport air pollution and energy consumption problem in India. Further, while the proximate causes of the problem are technological, the more fundamental causes are nontechnological and institutional, and must be considered for long-term policy effectiveness. Institutional factors are discussed in this chapter, but are treated in greater detail in Chapter  IV, which addresses  the  institutional  setting  for  policy-making and  implementation with regard to this problem in the Indian context. Vehicle user behavioural factors play a crucially important role in influencing transport air pollutant emissions. Reference is made to some key user behavioural factors, but they are discussed in detail in Chapter V I , which focuses on M 2 W vehicle user preferences, choices and motivations,  53  and their perspectives on how they would be affected by, and would respond to various policy alternatives. Since transport air pollution is a function of both per-vehicle emissions, as well as overall vehicle activity, the chapter addresses factors that contribute to both of these key components of the problem. The advantage of this approach is that technological-curative as well as preventive alternatives may be identified, for long-term effectiveness. Indeed, the chapter focuses on air pollutant emissions transport system-wide due to M 2 W vehicle activity, rather than merely from vehicle exhaust.  3.2  METHODOLOGY  The discussion in this chapter draws on published literature on a range of subjects including environmental policy, engineering, urban transport, and urbanization, reflecting the multi-dimensional nature of the transport air pollution problem. Documents relating to vehicle and fuel technology and the urban transport system in the Indian context that the author culled during the course of his field work in late 1997 are also referred to. Additionally, the chapter draws on in-depth interviews with various individuals interested in and/or knowledgeable about the range of issues involved, and representatives of institutions whose actions have an important bearing on transport air pollution in the Indian context. These individuals included decision makers in various relevant government agencies at both the national and local levels, vehicle and fuel industry representatives, academics and researchers,  1  and last but not least, M 2 W vehicle users. Combining  information from these diverse sources helped the author gain a comprehensive understanding of the various contributory factors and their interactions, and how the situation is evolving and is likely to continue to evolve over the coming years.  A list of these interviewees is provided in Appendix V. Interviewees' informed consent was obtained prior to interviews being conducted. The interview protocol, and the Informed Consent Form (Appendix VI), were approved by the Behavioural Research Ethics Board of the UBC Office of Research Services and Administration. One of the conditions of this approval was that the identity of interview participants would be kept confidential. It is for this reason that, while their information and insights were of immense value to the author, interviewees are not explicitly acknowledged. For the questionnaire, interview protocol and Informed Consent Form relating to the survey of, and in-depth interviews with M2 W vehicle users, see Appendices VM-XI. 1  54  3.3  FACTORS CONTRIBUTING T O M2W VEHICLE AIR POLLUTANT  EMISSIONS IN T H E INDIAN CONTEXT Before we discuss factors contributing to M2W vehicle air pollutant emissions, it is useful to note that it is exposure to air pollution, not merely air pollutant emissions, that determine health and welfare impacts. In turn, exposure is a function of local pollutant concentrations, duration of exposure, and the number of people exposed (Faiz et al 1992). However, air pollution impacts are not a function of exposure only. Also important are factors such as nutrition quality, and level of physical activity during exposure, as noted in Chapter II (CSE 1996; Faiz et al 1992; Romieu, Weitzenfeld and Finkelman, 1991; Romieu et al 1992; Wijetilleke and Karunaratne 1997). While all of the above factors in combination determine susceptibility to health impacts, access to quality health care influences the ability to cope with impacts, once they occur. Factors such as nutrition quality and access to quality health care are strongly related to income. And so, one might argue, are the level of physical activity to earn a living and housing location, both of which can strongly influence exposure. In short, income may likely be an important determinant of air pollution exposure and impacts. Factors such as topography, ventilation (and in turn, wind transport, dispersion and inversions), weather (in terms of precipitation), and vegetation have an important bearing on air pollutant concentrations, apart from transport and other emissions (Faiz et al 1992). Delhi's S P M concentrations would be a lot lower, for example, if its rainfall were 2  distributed more evenly over the year. The point of the foregoing is that, while the best line of action is of course to minimize emissions in the first place, measures such as housing away from areas of high local pollutant concentrations, improved nutrition, and access to good quality health care would also help.  3.3.1  Vehicle Technology  In general terms, Indian motor vehicle technology is decades behind global practice. 1950s and 1960s vintage vehicles, with considerably lower fuel economy than O E C D vehicles,  As Faiz et al (1992) point out, though air pollutant emissions in Santiago, Chile are only 10% of those in Sao Paulo, Brazil, the severity of pollution episodes in Santiago equal those in Sao Paulo.  2  55  continue to be manufactured in India and other LICs. This situation is changing, however, as more recent model vehicles are entering the Indian market with economic liberalization (Faiz et al 1992; Champagne 1998). '  3 4  Two-stroke engines power the bulk of Indian M2W vehicles, as indicated in Chapter II. M 2 W vehicles powered by four-stroke engines are expected to grow in prominence, as industry increasingly relies on this technology, among other options, in response to the stringent emission standards scheduled to come into force in 2000 (Central Motor Vehicles (Amendment) Rules 1995; CSE 1996; MoST 1996). Until recently, however, 5  four-strokes have accounted for only 10-15% of annual M2W vehicles sales, and thus constitute only a small proportion of this fleet in India. In 1997, there were around eight M 2 W vehicle models that were powered by four-stroke engines in the Indian market (Narayana 1994).  6  Table 3.1 compares the in-use fuel economy and exhaust emissions of Indian M 2 W vehicles with their two-stroke European counterparts in the early 1990s. This comparison is relevant, because many M 2 W vehicles from that period are likely operational on Indian roads. It appears from the figure that even in the early 1990s, Indian M 2 W vehicles were  For example, fuel injection is beginning to replace carburetion in cars, as emission standards become more stringent, and three-way catalytic converters, which require fuel injection for optimum effectiveness, become necessary in order to achieve those emission standards. Catalytic converters were mandated on all new cars in the major Indian cities in 1995, along with the introduction of more stringent emission standards (Central Motor Vehicles (Amendment) Rules 1995; CSE 1996; Faiz et al 1992; MoST 1996). Fuel economy data on Indian vehicles is sketchy, particularly for vehicles other than M2W and M3W vehicles. Where fuel economy data is presented, test conditions are rarely indicated. Average fuel economy for new Indian cars (including recent models) appears to be around 13 km/1 on the Indian driving cycle, compared to around 11 km/1 in Europe for both urban and highway driving, and 17 km/1 incity in North America (AIAM 1994b and 1997b; Environment Canada 1998; Faiz et al 1996; IIP 1994). Four-stroke engines deliver one power stroke for every two revolutions of the engine crankshaft, as opposed to two-strokes, which deliver a power stroke every revolution. The problem of fuel-air charge escaping unburned through the exhaust is eliminated on four strokes. Also, four strokes employ crankcase lubrication. HC, CO and PM emissions are reduced significantly, and fuel economy is considerably improved, compared to two-strokes. However, four strokes are larger (and heavier) for the same power output, and more expensive than two strokes. This is precisely why two-strokes are preferred for small M2W vehicles (Faiz et al 1992 and 1996). M2W vehicles sold over the last 30 years, and manufactured to this day, in India and other LICs, were developed in the 1960s by the Japanese, Italians and others. At the time, there was understandably no great interest in either emissions or fuel economy, and most such vehicles under 200 cc capacity were twostrokes. Only Honda offer a variety of four-strokes with displacements below 200 cc, and Kawasaki and Yamaha have a few models (Duleep 1994). Not surprisingly, all the motorcycle models offered by HeroHonda in India are four-strokes (so is the Kawasaki-Bajaj 4s) (AIAM 1994b and 1995). 3  4  5  6  56  considerably superior to their European counterparts. However, it should be noted that 7  the Indian driving cycle, which was used to generate the data for Indian M 2 W vehicles in Table 3.1, is based on, but is not identical to, the E C E (Economic Commission for Europe) cycle, which itself reportedly underestimates emission factors by 15-25% with respect to more realistic driving at the same average speed. Further, while the E C E cycle 8  requires starting vehicles after a six hour soak at 20-30 °C, the Indian cycle was run on warm vehicles until recently, and could have underestimated C O and H C on this score alone ( A I A M 1996a; Calvert et al 1993; Faiz et al 1992; Faiz et al 1996; IIP 1994; Central Motor Vehicles (Amendment) Rules 1995; CSE 1996; MoST 1996). Finally, it is not 9  clear to what extent the Indian M2W vehicle data (or for that matter, the European data) in Table 3.1 were representative of in-use fuel and oil quality, and vehicle operation and maintenance. Actual emission factors for Indiari-M2W vehicles could therefore be higher than the numbers shown in Table 3.1.  The fuel economy of Indian two-stroke and four-stroke M2W vehicles, and two-stroke mopeds, averaged 54, 73 and 61 km/1 respectively on the Indian driving cycle, on the tests by IIP (1994). However, fuel economy figures reported during the author's user interviews averaged 35 km/1 for two-strokes. This matches very well the 35 km/1 figure quoted in Duleep (1994). The fuel economy figures for European and US M2W vehicles average 20-25 km/1, and 42 km/1 on European mopeds (Faiz et al 1996). Even the US FTP driving cycle, perhaps the most representative cycle in terms of accelerations and transients, likely underestimates CO and HC (and also NO ), because the maximum speed and acceleration levels are lower than those achieved in reality (Faiz et al 1992; Faiz et al 1996). Calvert et al (1993) estimate that this underestimation could be as much as two times. The ECE cycle is far less representative of actual driving than the US FTP, involving as it does steady state conditions linked by uniform speed changes, and much lower maximum speed and acceleration levels than the US FTP. Additional driving cycles have been developed to make both the US FTP and E C E cycles more representative of reality. In the latter case, an extra urban driving cycle has been added to account for much higher speeds outside urban areas. Further, in order to account for the enhanced CO levels at low temperatures, the US FTP procedure has recently been modified to include a CO emissions test at -7 °C in addition to the existing 20-30 °C. The ECE cycle on the other hand continues to be run with the vehicle started at 20-30 °C (Faiz et al 1992; Faiz et al 1996). The Indian driving cycle involves a lower maximum speed than even the ECE cycle, on which it is based, which is reasonable given Indian conditions, but also lower acceleration levels (AIAM 1996a). Note that the extra urban driving cycle is not included in the E C E or the Indian driving cycles in the case of M2W vehicles. Cold start was made a requirement in the Indian emission standards from April 1998 for M2W and M3W vehicles, and from April 1996 for light-duty diesel vehicles (Central Motor Vehicles (Amendment) Rules 1995; CSE 1996; MoST 1996). However, there is no clear specification of the temperature at which the emissions test is to be started, or of the soak time. Since the Indian driving cycle is based on the E C E cycle, it may be assumed that test vehicles are to be soaked for six hours at 20-30 °C before starting, which are the E C E requirements (Faiz et al 1996).  7  8  x  9  57  In addition to the above issues, it should be noted that there are no evaporative controls on the fuel distribution system, or on vehicles except cars produced from 1996 (MoST 1996). Indian gasolines have a wide volatility range, with a high maximum value (BIS 1995a), as will be discussed in a subsequent section on fuel and oil quality. The vast majority of gasoline vehicles on Indian roads are carbureted, not fuel-injected ( A I A M 1994b and 1995). These facts, along with India's high ambient temperatures, which can reach 45°C in Delhi in the summer, heighten the potential for evaporative emissions, which are rich in reactive hydrocarbons that participate in the formation of tropospheric ozone.  Table 3.1  10  In-use Indian M2W Vehicle Fuel Economy and Exhaust Emission Factors Compared to Europe, Early 1990s  Fuel Economy, km/I  CO, g/km  HC, g/km  NOx, g/km  M2W Vehicle, India  54  6.5  3.9  0.03  Europe, 2s >S0 cc, Urban  25  22  15.1  0.1  M2W Vehicle = Motorized two-wheeled vehicle; 2s = Two-stroke. Sources: Bose 1996; Falz et al 1996; IIP 1994; Shah and Nagpal 1997. mgbOO  Evaporative emissions increase exponentially with increasing volatility and ambient temperatures. These emissions can be as high as 20-32% of total H C emissions on uncontrolled vehicles, and even higher on hot days.  11  Also, evaporative losses are a  Evaporative emissions comprise diurnal or 'breathing' losses due to expansion and contraction of gasoline in the fuel tank with changes in air temperature, 'hot soak' emissions from the carburetor bowl when a wanned up engine is shut down, running loss emissions due to vapour generated in fuel tanks during vehicle operation, and re-fueling emissions. Hot soak emissions do not occur in sealed fuel systems, such as those used with fuel injection systems (Calvert et al 1993). Ambient temperature and altitude play an important role in transport energy and air pollution. Elevated ambient temperatures and altitude would tend to increase evaporative emissions. Additionally, these two factors affect engine power, and therefore fuel efficiency, as air density reduces with increasing ambient temperature and altitude. Unless equipped with electronic air-fuel ratio control, or unless adjustments 1 0  11  58  significant source of vehicular benzene emissions. Evaporative emissions per kilometre from M 2 W vehicles can be as high as 40% of that in an uncontrolled car (Calvert et al 1993; Faiz et al 1992; Faiz et al 1996). In view of the foregoing, evaporative emissions from M 2 W vehicles can be significant in the Indian context (in addition to their high exhaust emissions). Recall in this connection that in Delhi, M 2 W vehicles account for around 2.7 times the vehicle-kilometres as cars. Indeed, M 2 W vehicle-kilometres are around 1.5 times the total vehicle-kilometres by all other motorized passenger transport modes in Delhi (Figure 2.14 in Chapter II).  Table 3.2  Evolution of Indian M2W Vehicle Exhaust Emission Standards Compared to Taiwan  M2W Vehicle India 1991 (g/km)  CO  12-30  HC  8-12  HC + NOx  India 2000  1991  India 1996  3,73  4.5  1.5  2  2.4  3.6  1.6  1.5  Thailand 1694  Taiwan/  mt  0,8  M2W Vehicle = Motorized two-wheeled Vehicle. Sources: CMV (Amendment) Rules 1995; C S E 1996; Faiz et al 1996; MoST 1996. mgbOO  Engine crankcase emissions can also be substantial. This source accounts for 13-25% of total VOCs on uncontrolled gasoline powered cars (Faiz et al 1992). These emissions  were made to suit the higher altitude, engines would consume more energy and produce higher exhaust emissions for the same performance as at sea level (Calvert et al 1993; Faiz et al 1992). In terms of health impacts, increased emissions at altitude would likely be coupled with the effects of altitude on the human respiratory system. In the Indian context, this is of concern in the case of Bangalore, which is situated at an elevation, and is also the most motorized city after Delhi (ASRTU/CIRT 1997).  59  were mandated to be controlled, commencing in 1996, but only for cars, not for M 2 W and M 3 W vehicles or diesels (Central Motor Vehicles (Amendment) Rules 1995; C S E 1996; M o S T 1996). As more M2W (and M3W) vehicles fitted with four-stroke engines begin to be produced in response to more stringent emission standards, crankcase emissions will gain in prominence as far as these vehicles are concerned. Crankcase H C emissions can be as high as 20% of exhaust H C emissions on four-stroke M 2 W vehicles (Hare et al 1974).  12  All in all, in-use emissions from the vehicle as a whole could be considerably higher than would appear to be the case for Indian M2W vehicles from Table 3.1. Moreover, what makes M 2 W vehicles so important in terms of urban transport air pollution in the Indian context is the fact that these vehicles are used in large numbers for daily commuting, unlike in Europe and North America. Besides, many M 2 W vehicles in Europe and North America are large-displacement, high-powered machines. Therefore, let us compare Indian M 2 W vehicle emission factors with those in Taiwan and Thailand, countries which have similar M2W vehicle ownership and usage characteristics as India. This may be done by tracing the evolution of M2W vehicle emission standards in these countries since the early 1990s (Table 3.2). As may be seen from this table, the emission standards for Indian M 2 W vehicles have improved vastly over just nine years, but have been inferior to those for their Taiwanese and Thai counterparts, until very recently. Even the Indian standards proposed for 2000 are not as stringent as those proposed for 1997 in Taiwan and Thailand, the most stringent globally except for Swiss mopeds. Besides, there were no cold start requirements in the Indian emission standards until recently, whereas the standards in Taiwan and Thailand follow the E C E cycle, which calls for a starting temperature of 20-30 °C. And the maximum speed in the Indian driving cycle is lower than in the E C E cycle ( A I A M 1996a; Central Motor Vehicles (Amendment) Rules 1995; C S E 1996; Faiz et al 1992 and 1996; MoST 1996). Further, Conformity of Production (CoP) 13  limits, which specify the maximum allowable deviation from standards for production lots,  Even so, HC emissions from exhaust as well as crankcase on four-strokes would be considerably lower than exhaust HC emissions from two-strokes (Hare et al 1974). However, there are no M2W vehicle emission standards that are more stringent than the 2000 (and even the 1996) Indian standards, with the exception of those in Taiwan and Thailand, and for Swiss and Austrian mopeds. (Central Motor Vehicles (Amendment) Rules 1995; CSE 1996; Faiz et al 1992 and 1996; MoST 1996). Note also that the 1997 standards for Taiwan and Thailand are only proposed figures. 1 2  13  60  were as high as 25 and 33% for CO and H C from Indian M 2 W vehicles in 1991, and were brought down to 20% for CO and HC+NO* in 1996 (Appendix III). Finally, unlike for vehicles in the USA, no clear stipulations as to emissions durability appear to have been made, nor are any emissions warranty or vehicle recall requirements imposed on manufacturers  (MoST  1996).  14  Consequently, there is little if any incentive for  manufacturers to design and maintain vehicles for emissions durability. The upshot of all of the foregoing is that, while Indian M 2 W (and other vehicle) emission standards have become increasingly stringent, in-use emissions per vehiclekilometre from the vehicle as a whole on even recent model Indian M 2 W vehicles could be considerably higher than the emission standards would seem to indicate. This situation is rendered all the more likely because of in-use vehicle operation and maintenance realities, which we discuss below. Before doing so, note that while we have been focusing on the engine and exhaust system, vehicle components such as transmission, brake and tyre are also important in terms of transport air pollution and energy (Alberta Energy 1988; Duleep 1994; Faiz et al 1996). In this connection, recall from Chapter II that asbestos continues to be used for vehicle brakes and clutches in Indian vehicles, because it is 40% less expensive than alternative materials.  3.3.2  Vehicle Operation, Maintenance and Disposal  Vehicle user choices and behaviours play a key role in transport emissions, and also in efforts to prevent and control them. This issue is discussed in considerable detail in Chapter VI, so only some brief points will be made here, with regard to vehicle operation and maintenance. Driving behaviours such as "jack-rabbit" starts and stops can increase emissions dramatically (Faiz et al 1992). Many M 2 W vehicle users interviewed by the author reported frequent fuel refills, and oil-fuel ratios not always as per specification,  In the USA, clear emissions durability requirements, emissions warranties, in-use surveillance, and vehicle recall have been implemented in addition to emission standards. Emissions durability requirements have been extended from 80,000 to 160,000 km. Such requirements have not been incorporated even in the European standards. In Taiwan, M2W vehicle emissions durability requirements of 6,000 km have been in force since 1991, and were upgraded to 20,000 km from 1998 (Faiz et al 1992 and Faiz etal 1996). 1 4  61  with significant implications for evaporative and exhaust P M emissions (Author's user interviews 1997). Coupled with the high motor vehicle activity rates in the Indian (and LIC) context is the high average fleet age because of low scrappage rates, and poor maintenance. Several studies worldwide have shown that maintenance is a significant factor in vehicular emissions. Among uncontrolled vehicles, H C and CO emissions between properly and poorly adjusted engines can vary by a factor of four or more. CO emissions can increase as much as 400% due to normal drift between services. Factors such as faulty ignition which have considerable scope to occur in the Indian context, can affect emissions durability.  15  This is particularly so in the case of technologies such as catalytic converters, which are entering the Indian market in response to more stringent emission standards. A damaged catalytic converter can increase H C and CO by 20 times, and N O by 3-5 times. Data from x  South and South-East Asia show that P M emissions can increase ten-fold in poorly maintained M 2 W and M 3 W two-stroke vehicles using poor quality lubricating oil (Shah and Nagpal 1997).  16  Finally, while average emissions typically increase with age, U S  studies have shown that 20% of recent model vehicles are "super emitters" (Calvert et al 1993; Faiz et al 1992 and 1996). One would expect good vehicle maintenance, given low labour and high fuel costs in the Indian context (Duleep 1994). However, many M2W vehicle users interviewed by the author (see Chapter VI) reported maintaining their vehicles themselves, or using the services of local mechanics, and only when absolutely unavoidable. Further, spurious spares are commonly used. Quality spares in India are expensive, in part because of high sales taxes (Author's user interviews 1997; Duleep 1994). Obviously, apart from user knowledge and skills, user incomes strongly influence vehicle operation, maintenance and  Faulty ignition is common on M2W and M3W vehicles in the Indian context. Spark plugs on these vehicles are highly susceptible to malfunctioning, because of dirty operating conditions, poor air filtration, and poor maintenance. Misfiring can give rise to high HC and CO emissions (Faiz et al 1992). In diesels, damaged fuel injection systems can increase PM at least 20 times. 6% of buses (and 22% of trucks and 37% of LCVs) in Delhi were found to have a free-acceleration smoke reading of 85 HSU (Hartridge Smoke Units), corresponding roughly to 8 g/km PM. 37% of buses, 62% of trucks and 57% of LCVs registered over 65 HSU (Faiz et al 1996; IIP 1994). 1 5  1 6  62  disposal choices. At any rate, these choices exacerbate the effect of vehicle (and fuel) technology.  3.3.3  17  Ineffective Monitoring and Enforcement  Poor vehicle maintenance  is also enabled by largely ineffective monitoring and  enforcement in the Indian context. A more detailed discussion is presented in Chapter IV, but briefly, this is because inspection of in-use emissions in Delhi (and other Indian cities), which is conducted in a decentralized fashion and is riddled with corruption, is burdensome for users, who therefore find ways and means of circumventing and/or subverting the testing process (CSE 1996; Priti Kumar 1997). Moreover, the testing procedure is itself problematic. Only CO emissions are measured at idle on gasolinepowered M 2 W and M 3 W vehicles and cars (MoST 1996). These tests can identify gross malfunctions on uncontrolled, carbureted vehicles, but correlate poorly with real-life emissions, particularly for H C and N O , and give rise to false passes and failures, on x  vehicles with electronic fuel injection and catalytic converters (Faiz et al 1996).  3.3.4  Congestion, and Road Availability and Condition  As noted in Chapter II, congestion is increasing rapidly, as a result of rapid motorization, in Delhi and other Indian and Asian LIC cities. Peak-hour speeds now average 10-20 km/h in many of these cities (Brandon and Ramankutty 1993; CSE 1996; Faiz et al 1992; Sathaye, Tyler and Goldman 1994; Poboon et al 1994). As far as Delhi is concerned, the average speed for all trips is reported to be only 10.7 km/h (RITES/ORG 1994). However, a careful examination of the data reveals that speeds for motorized modes, which are key in terms of transport air pollution and energy consumption, are higher 19.7, 17.8, and 12.1 km/h for cars, M2W vehicles, and buses. Further, while the average 18  Poor maintenance is not unique to M2W and other personal motorized modes. Buses and other commercial vehicles in Delhi are for the most part owned by a large number of private parties, who have little ability or incentive to invest in proper maintenance (especially given lax inspection), let alone improved vehicle technology. Even the 15% or so of Delhi's buses operated by the Delhi Transport Corporation (DTC), the state-owned public bus transit operator in Delhi, are poorly maintained (ASRTU/CIRT 1997; Chima 1997; Duleep 1994; Gambhirand Narayan 1992). The average speed for all trips is much lower than for motorized modes, because walking accounts for as many as 32% of trips in Delhi (RITES/ORG 1994). 1 7  1 8  63  speed (presumably for motorized modes) is reported to be only 10-15 km/h in the central area, it is 25-40 km/h on arterials (CSE 1996). Emissions of all gasoline engine pollutants, except for N O , increase dramatically at x  low speeds. Excessive idling and jerky "stop-and-go" operation due to congested traffic, 19  too many intersections, and poor T S M (transport system management) further aggravate emissions. A French study showed that CO and H C emissions per vehicle-kilometre increase by around 200%, and fuel consumption and C 0 emissions by around 260%, in 2  congested flow as opposed to smooth flow (Joumard 1990 in Faiz et al 1992). So, congestion further exacerbates the effects of vehicle and fuel technology, and the vehicle user behaviours discussed above. Vehicle driving behaviour is of course influenced by user choices, knowledge and skill, but also by the level of congestion, which in turn is determined by road availability in relation to vehicular activity, the level of modal separation, and the effectiveness of T S M and traffic regulations enforcement, factors over which vehicle users have no control. Another factor is the presence of travel peak periods. But travel peaks become irrelevant when road availability is the constraining factor, in which case congestion is likely to be a round-the-clock, or at least a day-long, phenomenon. While the road and highway system is inadequate nationally, Delhi does not suffer 20  from a lack of roads. The city has an extensive road network of over 21,000 kilometres (12-feet width), with 800 km of 30-metre plus width roads. Indeed, Delhi has the highest road density per square kilometre in India, and a higher road length per capita than most countries in Europe and Asia (Mohan et al 1997; TERI 1997a). Notwithstanding this fact, and the point made earlier about the speeds for motorized modes in Delhi, congestion is increasing, and is likely to continue to do so, along with increased motor vehicle activity. So far in this section, we have been discussing the importance of transport system characteristics such as road availability, modal separation, T S M , and traffic regulations  Engines (and for that matter, vehicles as a whole) are typically designed to operate over a wide speed range. If, as is reasonable to expect, the bulk of passenger vehicles in Indian cities are operated at low to medium speeds (because of congestion), and rarely if ever out of town, it may be worthwhile considering optimizing fuel efficiency and emissions selectively for the low and medium speed range. India has only around one million kilometres of surfaced roads, with the remainder being unsurfaced (TERI 1997a). 1 9  2 0  64  enforcement. As far as transport infrastructure is concerned, it is not only these factors, but also the quality of roads, that plays an important role in transport energy consumption and emissions. First, unsurfaced roads increase re-suspended P M emissions, which can contribute significantly to total transport P M (and PMio) inventories (Bhattacharyya 2000; CSE 1996; G V R D 1995). Second, and more importantly, the low load-bearing capacity of Indian roads have a profound impact on overall fleet fuel efficiency and emissions. This is particularly so for commercial vehicles, which account for the lion's share of transport energy consumption in the country, as discussed in Chapter II. Vehicle size critically influences fleet fuel efficiency. The larger the vehicle size, the higher the potential overall transport fuel efficiency on a per unit payload basis, because payload increases faster than empty vehicle weight. Because of the low load-bearing capacity of Indian roads, vehicles of no more than around 15 ton gross vehicle weight (GVW) can be used, as opposed to the 40 ton G V W vehicles commonly seen on roads in USA, Canada, and other O E C D countries. This constraint, along with the varying speed/load conditions typical of travel on Indian roads, leads to fleet fuel efficiencies half of those in the OECD on a tonkilometre basis, and much higher emissions per ton-kilometre and road and vehicle maintenance and operating costs (Duleep 1994; Faiz et al 1996).  3.3.5  21  Fuel and Lubricating Oil Quality  The health and welfare effects of fuel quality parameters such as lead, sulphur, and benzene content were discussed in Chapter II. Apart from these effects, fuel lead and sulphur degrade catalytic converter effectiveness by forming deposits on, and blocking exhaust gas access to, the catalyst. As little as a single tank of leaded gasoline can cause permanent catalyst damage. On the other hand, fuel sulphur causes reversible catalytic performance deterioration (Faiz et al 1992). Other gasoline quality parameters that have a significant bearing on exhaust and evaporative emissions and atmospheric reactivity  Further, engine and vehicle technologies such as turbochargers and aerodynamic improvements, which have the potential to provide significant fuel economy and emissions benefits, are not cost-effective in the Indian context, because these benefits are obtained at sustained high speeds, a difficult proposition on Indian roads, and also since diesel is heavily subsidized (ASRTU/CIRT 1997; Brandon and Ramankutty 1993; CSE 1996; Duleep 1994; Faiz et al 1996; TERI 1997a).  2 1  65  include fuel volatility, aromatic, olefin and oxygen content, and distillation and depositcontrol characteristics (Faiz et al 1992 and 1996; Calvert et al 1993; Raje and Malhotra 1997). Until 1995, lead content in Indian gasoline was as high as 0.56 g/L (in 87 octane), and sulphur content was as high as 0.2% (2000 ppm) by weight.  22  Benzene, a known  carcinogen, was not controlled at all (BIS 1995a). With rapidly deteriorating urban air quality, fuel quality improvements have been implemented in a phased manner, as shown in Appendix IV, and as discussed in a subsequent paragraph. In 2000, unleaded gasoline is expected to be available country-wide. Gasoline benzene content is expected to be controlled to 5% by volume country-wide, and to 3% in the four major cities. Gasoline sulphur is expected to be lowered to 0.1% (1000 ppm) by weight, for unleaded fuels. Simultaneously, improvements have been proposed in deposit control and fuel volatility (BIS 1995a).  23  The above improvements in the quality of Indian gasoline represent a significant advance over the situation that prevailed in 1995. Even so, Indian gasoline continues to be considerably inferior to that presently available and likely to come on stream shortly in Europe and the U S A (Table 3.3). In addition to the stringent controls on sulphur, benzene and fuel volatility in US and European gasolines (Table 3.3), aromatic and olefin content are also controlled (Calvert et al 1993; Faiz et al 1996; Mercedes Benz 1997). These fuel quality differences can have significant air quality implications. In terms of gasoline sulphur, for example, an increase from 100 to 900 ppm by weight can produce a 13-14% deterioration in catalytic converter performance (Faiz et al 1992). The author learned from his interviews that many vehicles with fuel injection systems being imported into India have encountered problems with fuel injector blockage because of poor deposit control on Indian gasolines.  Lead content was 0.8 g/L in 93 octane gasoline. Such high levels of lead were necessitated in part because knock susceptibility, and thus octane requirement, increases with increased engine operating temperatures and deposit accumulation (Faiz et al 1992). Both of these factors are common in the Indian context. Diesel sulphur, an important contributor to PM, was as high as 1% (10,000 ppm) by weight until recently. It was brought down to 0.25% (2500 ppm) in Delhi and its neighbourhood in 1996, and countrywide in 1999 (BIS 1995b; Appendix IV). By comparison, diesel sulphur levels are below 100 ppm in many European countries (Mercedes Benz 1997). 2 2  2 3  66  Unleaded gasoline (0.013 g/L lead) was introduced for new four-wheeled gasolinepowered vehicles with catalytic converters in Delhi and the three other major metropolitan centres in 1995, and in all other state capitals and major cities in late 1998. As already indicated, unleaded gasoline is expected to be available country-wide in 2000. Thus, only about 10-20% of gasoline sold in Delhi was unleaded from 1995 until recently. The fuel for M 2 W and all other gasoline-powered vehicles (the overwhelming majority of vehicles on the road), has been 0.15 g/L lead gasoline, which was introduced in the four major cities in 1994. Unleaded gasoline is understood to have been restricted to new fourwheeled vehicles with catalytic converters, because of concerns about the high levels of benzene in unleaded gasoline going into the atmosphere untreated in vehicles without catalytic converters. However, it is also the case that it has been a challenge for Indian refineries to produce adequate quantities of unleaded gasoline, while also controlling important fuel quality parameters such as octane rating, benzene and other aromatics, volatility, and gumming properties (CSE 1996; Faiz et al 1996; Hari 1994; Raje and Malhotra 1997). There was no fuel price differential between leaded and unleaded gasoline in India, unlike in Mexico, when unleaded fuel was introduced in 1990 (Humberto Bravo et al 1991). Mis-fueling vehicles with catalytic converters with leaded gasoline was therefore not likely to occur on this account within Delhi and the other major cities. However, misfueling was certainly a possibility on new cars with catalytic converters operating both within and outside major cities until 2000, on account of the widespread availability of leaded fuel outside the cities. This possibility raises doubts about the effectiveness of catalytic converters, particularly given the largely ineffective I & M regime. Indeed, it is reported that many vehicles with catalytic converters in Delhi have failed the in-use emissions test (CSE 1996).  24  However, as noted, no-load in-use emissions tests are unreliable on vehicles with catalytic converters. Since unleaded gasoline is expected to be widely available nation-wide effective 2000 (Appendix IV), the possibility of catalytic poisoning due to misfueling with leaded fuel will not be an issue after this date. Also, since benzene levels are going to be controlled simultaneously, the use of unleaded fuel in vehicles without catalytic converters will not be an issue either, at least as far as benzene is concerned.  2 4  67  Table 3.3  Quality of Indian Gasoline  Compared to California and Europe  California RFG 1996  ACEA Recommendation post-2000  1000  30  30  no limit  5  0.8 (1 - USA)  lllllllllllllllll  35-70  35-60  47 (summer); 54/62 (USA)  58  India pre-1995  India 2000  Lead, g/L max.  0.56/0.80 (87/93 Octane)  0.013  Sulphur, ppmw max.  2000  Benzene, % vol. max. RVP, kPa @ 38 deg C, max.  RFG - Re-formulated gasoline; A C E A - European Automobile Manufacturers' Association; ppmw - parts per million by weight; RVP - Reid Vapour Pressure; kPa - Kilopascals. Benzene level in Thai gasoline 3.5% vol. since 1993; US RVP 49.6/55.8 in areas with serious air pollution problems. Sources: BIS 1995a; Calvert et al 1993; Faiz et al 1996; Mercedes Benz (1997). mgbOO  The implementation of the phased two-gasoline policy may therefore also have contributed to transport emissions. Further, some experts believe that restricting unleaded gasoline to vehicles with catalytic converters was a wrong policy, because lead is less preferable than benzene in terms of health effects (Friedrich and Walsh 1997). According to Weaver (1995), though, a two-gasoline policy is a good transitional measure, with a ' tightly controlled low-lead fuel for older vehicles, and less tightly controlled unleaded gasoline for newer vehicles. The first 0.1 g/L lead provides the largest octane boost, with diminishing returns thereafter, and the quickest and most economical way to reduce lead emissions generally is to reduce lead content of all gasoline grades as much as possible, rather than to have vehicles without catalytic converters use unleaded fuel. This helps conserve limited unleaded gasoline supplies for vehicles with converters, and reduces the price differential between leaded and unleaded gasolines, thus reducing the incentive to misfuel such vehicles. While this is precisely the approach adopted in India, low-lead  68  gasoline for the majority of vehicles was by no means tightly controlled in terms of other fuel parameters (BIS 1995a).  25  Finally, lubricating oil quality has important implications for transport P M and H C emissions, particularly in the case of two-stroke engines fitted on M 2 W and M 3 W vehicles, in which the oil is "lost" due to combustion, but does not burn completely.  26  Additionally, phosphorus and other oil additives coat and poison catalytic converters (Faiz et al 1992 and 1996). Two-stroke lubricating oils in India are proposed to be made phosphorus-free from 2000 (BIS 1996).  3.3.6  Fuel and Oil Pricing and A dulteration  In the Indian context, the problem of poor quality fuel and oil is further exacerbated by fuel and oil adulteration. M3W vehicle operators, who typically do not own their vehicles, commonly adulterate their gasoline with as much as 30% kerosene, and even solvents. To guard against the resulting wear and tear, they mix as much as 10% of lubricating oil, the principal source of P M emissions in two-strokes. Lubricating oil, sold loose mainly for use in M 2 W and M 3 W vehicles, is also adulterated (Raje and Malhotra 1997). Adulteration is enabled principally by the fact that kerosene (and diesel) have been heavily cross-subsidized by gasoline, as a part of the administered pricing mechanism (APM). In the case of kerosene, subsidization is justified on the grounds that this fuel caters for the energy needs of lower income urban households. Kerosene retailed at INR (Indian Rupees) 2.60 per litre in Delhi in 1995, as compared to gasoline at INR 17 per litre. Given this price differential, operators find it attractive to adulterate gasoline with kerosene, rather than using straight gasoline, even though frequent piston changes are necessary. Diesel is adulterated with kerosene as well, though the diesel-kerosene price differential is not as great as that for gasoline-kerosene (in Delhi in 1995, diesel retailed at INR 7 per litre) (TERI 1997a). Adulteration is likely at least as much of a problem in 27  2 5  The two-gasoline policy also involves the added cost of duplicating storage, delivery and distribution  facilities (Weaver 1995). In diesels, lubricating oil can contribute up to 2-48% of exhaust PM mass, and up to 88% of mutagenic and/or carcinogenic diesel particulate SOF (Pierson 1988; Truex et al 1980; Lowenthal 1994). Similar disparities exist in other Asian countries, including Pakistan, Thailand, Indonesia and China, particularly for gasoline and kerosene. But the disparity for gasoline has been by far the highest in India. 2 6  2 7  69  smaller cities and towns. In Bhopal, for example, 90% of dealers reportedly sell adulterated fuel, despite random checks and punitive measures by district authorities (Kumar et al 1997). Fuel and oil adulteration, and factors such as excessive oil usage in M 3 W vehicles, have a significant impact on transport emissions, not to mention fuel economy and engine life. As noted earlier, such in-use realities are not reflected in Indian emissions inventories. Apart from its implications for transport air pollution and emissions inventories, the foregoing discussion demonstrates the interactions between fuel adulteration and other vehicle user choices, monitoring and enforcement, and fuel pricing.  3.3.7  Motor Vehicle Activity  The foregoing sections explored the various vehicle and fuel technology, transport system, climatic and vehicle user behavioural factors that contribute to air pollutant emissions from M 2 W vehicles in the Indian context, on a vehicle-kilometre basis. It would now be useful to examine the factors that contribute to motor vehicular activity, since transport emissions are a function of emissions per vehicle-kilometre and vehicle-kilometres driven. And after all, vehicle and fuel technology have, if anything, been improving over 28  the last decade, as discussed. Transport emissions are becoming a public policy issue in that context, because of technological factors, but also because motor vehicular activity, particularly on M 2 W vehicles, and congestion, are growing rapidly. Specifically in terms of the high level of M 2 W vehicle activity in India, it would be useful to understand vehicle user motivations, within the context of wider institutional factors.  Subsidies exist in terms of coal prices and electricity tariffs as well. Apart from being a barrier to longterm financial viability, energy subsidies have significant environmental effects. An example in Indian transport is inappropriate dieselization. About 20% of Mumbai's vehicles are diesel powered, though only 9% are trucks or buses. Progress is being made in eliminating energy subsidies, including in India. The ratio of Indian domestic to international prices in 1991 were 3.8, 0.8 and 0.62 for gasoline, diesel and kerosene. In late 1997, the ratios for gasoline and diesel stood at around 2 and 1 (ASRTU/CIRT 1997; Brandon and Ramankutty 1993; CSE 1996; TERI 1997a). Motor vehicular activity is particularly important in terms of re-suspended road dust, a significant source of transport PM. While other transport emissions can be controlled fairly effectively by technologies that target per-vehicle emissions, re-suspended emissions cannot, since they are largely dependent on total vehicular activity. 2 8  70  Rising incomes are certainly an important factor contributing to rapid motorization in Indian (and other LIC) cities (Faiz et al 1992). As incomes increase, the poor majority purchase bicycles, and those who own bicycles graduate to M 2 W vehicles. Second, as 29  motor vehicle production has grown rapidly over the last decade or so ( A I A M 1994a and 1995; ASRTU/CIRT 1997), supply constraints have greatly eased. Further, with economic liberalization, many automobile purchase financing institutions have started business lately, and credit has become easy to obtain (Bhardwaj 1994). Also, manufacturers are offering old vehicle buy-back schemes in order to generate new vehicle sales (Kinetic Honda 1997). While the above factors certainly play an important role, increased motor vehicle ownership and use are also responses to circumstances in which users find themselves. A key factor in this regard is the growing consumerism, driven by social pressures and aggressive marketing, in urban India. Second, the rapid pace of urbanization and motorization have contributed to sprawl, which in a vicious circle further increases motor vehicle ownership and activity. The urban area of Delhi has grown more than 15 times since 1911, and five times since just  1981 (DDA 1996; Misra et al 1998).  Correspondingly, average trip lengths have increased 79, 122 and 62% since 1969, and 37, 40, and 24% since only 1981, for M2W vehicles, cars and buses respectively (RITES/ORG 1994).  30  The effect of urban sprawl is exacerbated by rental housing  becoming increasingly unaffordable in the heart of Delhi, which is where the bulk of employment is. Many M2W vehicle users interviewed by the author reported being forced to live far from their workplaces on account of this fact. Buses are often the only affordable motorized modes for the majority, but demand far exceeds availability. Delhi has the country's largest bus fleet, numbering around 28,000.  31  Even this fleet has not been able to keep up with ridership, which increased from 22.4% of  Nearly every M2W vehicle user interviewed by the author said s(he) would buy a car when they could afford it (see Chapter VI). Sprawl likely has a greater impact on motor vehicle activity in the suburbs and outer periphery than in the inner core. In the outlying areas of Greater Mumbai, M2W vehicle activity has reportedly increased 200% in just a decade (CSE 1996). The state-owned Delhi Transport Corporation (DTC) operates around 3000 of Delhi's buses (ASRTU/CIRT 1997; Chima 1997). A large number of private bus operators run the city's balance 25,000-odd buses. 2 9  3 0  31  71  all trips in 1957 to 39.6% in 1969 and 42.3% in 1994 (RITES/ORG 1994). Buses are generally becoming more crowded, inconvenient and time-consuming. The situation is further aggravated by the shortened bus life due to heavy use, poor fleet maintenance, and poor roads (Sathaye, Tyler and Goldman 1994).  32  Because of sprawl, unaffordable housing close to workplaces, and increasingly unreliable and inconvenient public transit, people are forced to purchase and use personal motorized modes if they can afford them. As congestion increases due to motorization, walking and cycling become increasingly tedious and hazardous, and owners of personal motorized modes begin to use them, even over short distances. Children are driven to 33  and from school, to protect them from other motor vehicles. Given these effects, it is not surprising that the mode shares of personal motorized modes have increased as rapidly as they have (RITES/ORG 1994; Sathaye, Tyler and Goldman 1994). Nor is it surprising that large numbers of the not-so-poor, for whom cars are out of reach, purchase and use M 2 W vehicles in Delhi and other Indian cities. M 2 W vehicles offer door-to-door capability, require very little parking space, can be (and typically are) parked securely inside the home, and carry passengers as well as things. Though these vehicles contribute to congestion, they can cope with it as no other motorized mode can, because of their size and maneuverability. Their average speed in Delhi is 17.8 km/h, as against 19.7 km/h for cars, and 12.1 km/h for buses. The superiority of M 2 W vehicles in this regard is demonstrated in Figure 3.1, which compares door-todoor journey times for various modes by distance, computed based on data in RITES/ORG (1994). M 2 W vehicles are only marginally slower than cars, right up to distances of 25 kilometres.  34  In short, M2W vehicles offer excellent and affordable  The problems of inadequate and poorly maintained transit are shared by other LIC countries. Public spending on urban transport in these countries, massive as it has been, has met no more than 15% of needs. With the myriad urgent demands on scant resources, there is little room for expansion. In Bangkok, bus trips increased 85% in the 1980s, but the fleet increased only 7%. In Mumbai, almost 80% of trips are by transit ~ 44% by bus, 36% by suburban rail ~ but the World Bank has warned that unless services are improved, personal vehicle use will increase (Brown and Jacobson 1987; CSE 1996; Pendakur 1987; Sathaye, Tyler and Goldman 1994). Compromised access also affects public transit usage, since users need to walk to and from bus-stops. Additionally, it becomes increasingly difficult for low-income people who cannot afford motorized modes to walk or cycle to essential services. Interestingly, bicycles in Delhi appear to be faster than buses right up to 25 km (RITES/ORG 1994). Obviously, there is considerable potential for efficient bicycle travel, if only the facilities existed for them. 3 2  3 3  3 4  72  mobility, and easy access to employment and other essential services, in a context in which there are few other attractive options. And thanks to technology, M 2 W vehicles are becoming increasingly easy to use. The Kinetic Honda scooter, for example, is pushbutton, rather than kick-started (Kinetic Honda 1997), and has reportedly dramatically expanded mobility for women, who are entering the work-force in large numbers. Ironically though, advances such as these will only serve to increase motor vehicle ownership and activity. Figure 3.1  Door-to-door Journey Times by Different Modes in Delhi  120  Door-to-door Journey Time, min. i  0 ' <2.5  3.8  6.3  8.8  11.3  13.8  16.3  18.8  21.3  23.8  ' >25  Distance, km M2W/M3W • Motorized two-rthree-wheeled vehicles. Calculations based on data In RITES/ORG 1994. Walk and wait times included. mgb99  Before closing this section, it is worth noting the importance of vehicle trips in terms of transport emissions and energy consumption. While sprawl increases trips and trip lengths, compromised access due to motorization tends to increase motorized trips over short distances. Short distance trips are the most polluting on a per-kilometre basis, because trip-end (cold start and hot-soak evaporative) emissions form a significant proportion of total trip emissions, and are the same regardless of trip length. Indeed, the shorter the trip, the greater the trip-end emissions as a percentage of total trip emissions. While a 32 km trip on an average 1987 model car with a catalytic converter would  73  produce 34 grams HC, a mere 8 km trip would produce 25 grams (Faiz et al 1992; Kessler and Schroeer 1993). Of course, in the Indian case, cold starts would not play a major role, but hot-soak evaporative emissions would. Further, because the vast majority of Indian vehicles are not fitted with catalytic converters, running emissions would form a much larger percentage of total emissions. Nevertheless, vehicle trips, particularly over short distances, are critically important. Nearly one-third of motorized trips (and 60% of all trips), are executed over distances less than five km in Delhi, and the average speed for trips over this distance is lower than over greater distances (RITES/ORG 1994).  3.3.8  The Role of Government Policy  Travel behaviour in terms of personal motorized mode ownership and use of course depend on urbanization, incomes, and user preferences and choices. But also critical is government policy with regard to planning and infrastructure provision for private motorized modes versus public transit and non-motorized modes, full cost pricing of travel, and land use and housing. For example, planning for personal motorized modes in the form of road building to relieve congestion and maintain high speeds, and provision of parking, serves to increase their use, and render other modes less viable. As personal motorized mode use increases, the system is increasingly designed to suit these modes, and those who can afford them. Conversely, the fact that the other modes are becoming increasingly unviable serves as a justification for not providing for them, which in turn makes them even less viable than previously, and those who rely on these modes more vulnerable. A vicious circle is created in which personal motorized modes become selfperpetuating. Further, once a city becomes dependent on automobiles, automobile reliance and inefficient land use patterns tend to become self-reinforcing (Brown and Jacobson 1987; Hillman 1990; Richmond 1990; Whitelegg 1993). Brown and Jacobson (1987) put it well: "Urban planners, by assuming ever greater automobile use, build cities that make it inevitable". Variations in car dependence and transit use, and travel behaviour generally, even among countries with similar per capita incomes and urbanization, arise largely from public policy differences. The low automobile dependence of affluent Singapore, Tokyo  74  and Hong Kong and Manhattan show that wealth need not inevitably lead to heavy dependence  on private motorized modes. Policies which favour more  compact  development, promote investments in transit, and keep automobile infrastructure spending to a minimum, dampen automobile ownership and use, while at the same time strengthening transit use, even as wealth increases (Kenworthy et al 1994; Pucher 1988). Unfortunately, urban transport decision making in the LICs is biased in favour of motorized vehicles and other extremely costly modes of transportation that serve only a small section of society. Little or no restrictions are placed on cars (of course, it would be highly problematic to put in parking restrictions, for example, without also providing attractive options). While scarce resources are made available for expensive infrastructure to  accommodate  motorization, apparently  nothing can be  found  for low-cost  improvements to benefit the poor majority. Non-motorized modes are not only ignored, but actively discriminated against (Replogle 1991; Whitelegg 1993). This is despite the fact that these modes account for a significant proportion of travel activity in the LICs. In Delhi, for example, walking and cycling (and travel by cycle rickshaws) account for nearly 39% of all trips, even among residential households (RITES/ORG 1994). It is likely that these modes account for an even higher share of trips in the city as a whole, since every second person in Delhi reportedly lives in a slum or a squatter settlement (Singh 1997b). Meanwhile, policies in countries like India (and China) are providing market and investment opportunities for international automobile manufacturers. In India, 11 heavy truck and 16 car companies are forecast for the near future, involving collaborations with Daewoo, Fiat, Ford, G M , Honda, Hyundai, Mercedes-Benz, Opel, Peugeot, Renault, Suzuki, and Volvo, among others ( A I A M 1997a; Champagne 1998).  35  In this connection, a rail-based Mass Rapid Transit System (MRTS), an idea studied repeatedly since the 1950s, is at last taking shape in Delhi. When completed in the "horizon year" of 2021, the system is expected to be 198.5 km. long, comprising 111 km.  Significant excise and import duty reductions related to motor vehicles and parts have been introduced recently (Ramachandran 1994; Mohan Ram 1994). In China, the growth of private vehicles since the 1980s has been dramatic. Foreign manufacturers are already producing a million vehicles annually. By 2000, half of this production is expected to be cars. It should be noted, however, that while this trend will fuel rapid motorization, it will also (for what it is worth) improve per-vehicle fuel economy and emissions. 3 5  75  at grade and 35.5 km. underground, and a 35.5 km. elevated portion. Construction on the initial 8.3 km. surface corridor has commenced, and the first phase of 52 kilometres (comprising 11 km. underground and the balance elevated and at grade) is expected to be completed by 2005. The first phase is expected to cost Cdn $1.6 billion, 56% of which will be loaned by Japan (Delhi Metro Rail Corporation 2001; Gambhir and Narayan 1992; Iijima 1999; Mohan et al 1997; Singh 1997a). There are divergent views on the ability of M R T S to reduce personal motor vehicle activity and congestion, and its financial viability in the LIC context. Kenworthy et al (1994), Kenworthy and Laube (1999a), Kenworthy and Laube (1999b), and Poboon et al (1994) argue strongly in favour of urban rail. They point out that a high level of service including time competitiveness is required to persuade personal motor vehicle users to switch to transit, particularly as incomes grow. Buses play an important public transit role, but cannot provide the frequency or capacity to cope with heavy passenger loads. This is particularly so in dense cities such as those in the Asian LICs. Such cities constrain bus service due to limited road capacity and severe traffic congestion. Note in this regard the unfavourable door-to-door journey times for buses compared to cars and M 2 W vehicles, and even bicycles, in Figure 3.1. On the other hand, segregated rail systems, because they are not constrained by traffic, have the potential to transport large numbers of people quickly ~ up to five times the passengers per hour as buses in mixed traffic (Mohan et al 1997). This potential is particularly enhanced when rail is well integrated with land use around stations and with feeder buses, and delivers passengers into pedestrian-friendly environments. Further, urban rail provides more reliable, comfortable, safe and high profile service than do buses. As a result, rail has the potential to attract people from personal motor vehicles in addition to only captive users (Kenworthy et al 1994; Kenworthy and Laube 1999a; Kenworthy and Laube 1999b; Poboon et al 1994). Cities with a higher level of rail service within their transit systems generally have lower automobile dependence and higher transit utilization, due to the superior speed of rail systems and the other advantages alluded to above. Further, many more rail systems around the world have provided effective sites for integrating high density mixed-use development with public transit and the formation of poly-centric urban form than have  76  bus systems. This in turn is due to the ability of rail systems to transport large numbers of people to high density nodes without compromising the pedestrian environment at these nodes. Finally, rail is the most energy efficient motorized urban transport mode, consuming 2.5-5 times less energy per passenger-kilometre than buses (Kenworthy et al 1994; Kenworthy and Laube 1999a; Kenworthy and Laube 1999b). While highly dense cities can constrain buses, they have the potential to support high capacity rail systems on account of the presence of strong corridors of development, as exemplified by Hong Kong, Singapore and Tokyo. Despite their wealth, these cities have low automobile dependence and high transit usage, in large part due to investment in rail, and the degree to which land use is integrated with transit, particularly rail (Kenworthy et al 1994; Kenworthy and Laube 1999a; Kenworthy and Laube 1999b; Poboon et al 1994). Mohan et al (1997) have assessed urban rail with specific reference to the LIC context, based on a survey of the performance of rail systems constructed over the past 25 years in Bogota, Cairo, Calcutta, Hong Kong, Istanbul, Manila, Mexico City, Medellin, Porto Allegre, Pusan, Rio de Janeiro, Santiago, Sao Paulo, Seoul, Singapore and Tunis. This survey revealed that rail systems in only three cities, Hong Kong, Seoul and Singapore were built on time. Most of the others experienced construction delays. Calcutta, the only Indian city to have an urban rail system, stands out in this regard. The city's 16.5 km. system, originally scheduled to be completed in six years, took 23 years to complete. Construction delays have occurred in cities such as Calcutta because of lack of awareness of what lies beneath the surface. Delays have also been caused by disputes of various kinds, service diversions, material and funding shortages, and traffic disruptions (Iijima 1999; Mohan et al 1997; Rekhi 1996). Urban rail systems typically entail huge capital costs, ranging from US$ 8 million to US$ 165 million per kilometre, many times more than bus based systems (Mohan et al 1997). Further, only the systems in Hong Kong, Singapore and Porto Allegre were constructed within the projected budget. Only four systems in the Mohan et al (1997) survey, including Hong Kong, show patronage levels up to or close to expectations. Once again, Calcutta's system appears to be the one with the highest cost overrun and the lowest patronage relative to what was projected. Revenue-to-operating cost ratios are  77  greater than unity only for Hong Kong, Manila, Santiago and Seoul (this figure was not available to Mohan et al (1997) for Singapore). Thus, while rail systems require heavy levels of investment, many require continued subsidies (Brown and Jacobson 1987; Sathaye, Tyler and Goldman 1994). It is difficult to estimate the impact of urban rail systems on congestion. However, there appears to be only short-lived or no impact in the majority of cities for which information exists. Congestion relief is short lived because private traffic rapidly grows to utilize released road capacity. Experience from even well run high-capacity rail systems as Singapore's and Hong Kong's shows that while they may cause bus users to transfer to them, they attract no more than a small share of private motor vehicle users (which is key in terms of reducing congestion and emissions). Further, while passengers are mostly captured from buses, reduction in bus traffic is not proportional, and in any case represents only a small portion of overall traffic. Finally, the extent of energy savings due to (underground) urban rail systems is unclear, because of energy requirements for airconditioning and escalators in stations (Mohan et al 1997; Sathaye, Tyler and Goldman 1994). Based on their survey, Mohan et al (1997) conclude that a large population with a high per capita income is required to provide the revenue base to sustain urban rail systems. A high per capita income would ensure high rail shares of work as well as non-work trips. A report quoted by Mohan et al (1997) suggests that a minimum per capita income of US$ 1000 is necessary to justify urban rail. In this regard, it is worth noting that in their survey, Calcutta, whose rail system is the worst performer in terms of various parameters also has the lowest per capita income. The three cities with the best performance in the survey, Hong Kong, Singapore and Seoul, have the highest per capita incomes. The financial viability of urban rail systems depends critically on keeping utilization and fares as high as possible, and staffing and wage levels as low as possible. But experience from several LIC cities suggests that high fares cannot be charged without losing patronage. In order to attract patronage, the integrated bus and rail fare should ideally not be much higher than the existing bus fare; if it is, the poor will continue to use buses. And any attempt to remove bus competition will likely cause major disruptions in  78  peoples' lives, and the displacement of many small operators, as would be the case in Delhi (Mohan et al 1997). All of this means that fares in LIC cities have to be subsidized. But this would effectively drain resources from other important social sectors (and rural areas), to benefit the urban middle class. This has in fact been Calcutta's experience. Between 1972 and 1978, expenditures related to its subway consumed 48% of the city's investment for all purposes. And despite massive cost overruns and time delays, and continuing subsidies, the poor majority cannot afford to ride it (Brown and Jacobson 1987). In order to be truly effective, urban rail needs to be supported by multiple intensely mixed-use centres close to stations. This is possible to achieve in dense LIC cities, and in turn, urban rail enables such densification, as discussed. Additionally, strong economic controls to curb personal motor vehicle ownership and use, by means of high taxes, parking costs and traffic restraint, would be required. In this regard, low provision of road infrastructure (to maintain transit speeds competitive with private motor vehicle speeds), and good provision for non-motorized modes, would also be desirable. The success of the urban rail systems in Hong Kong, Singapore and Tokyo is in fact largely due to these features. Half of Hong Kong's population lives within walking distance of, and 69% of rail passengers in the city walk to and from, a Mass Transit Railway station. In Singapore, which is less dense than Hong Kong, the corresponding figures are 30% and 65%, in large part because new development is being integrated with rail and routes have been planned to service existing development (Kenworthy et al 1994; Poboon et al 1994). ' 36  37  Many  LIC cities, however, do not have such large numbers of people in proximity to rail stations. This would necessitate many passengers using buses to access urban rail (Mohan et al 1997), which in turn would call for efficient feeder bus services.  While such figures are the result of conscious planning, it is also the case that the intense pressure on scarce land, and the potential for overwhelming congestion if private vehicles remain uncontrolled, has necessitated physical planning based on highly compact nodes of mixed-use development in Hong Kong (Kenworthy et al 1994). The close integration of rail and land use in Hong Kong has in turn attracted major capital contributions from private developers, and enabled on-going non-fare revenue flows from property leases around stations. Also, the transit operator is actively involved in controlling development which will determine how effective the system will be at meeting access needs (Kenworthy et al 1994).  3 6  3 7  79  Several of the issues raised above are worthy of consideration in relation to the urban rail system under construction in Delhi. Whereas Singapore's gross regional product per capita is around US$ 12,939 (Kenworthy and Laube 1999a), the per capita income in Delhi (incidentally India's wealthiest region) is only US$ 740 at current prices (NCTD 2001). In terms of the financial viability of Delhi's MRTS, it should be noted that while the Japanese loan carries a low interest rate of 2.3%, interest payments could effectively escalate due to rupee depreciation and other factors (Iijima 1999; Rekhi 1996). While Hong Kong and Singapore have the demonstrated ability to control land use, plan and implement intense mixed-use development integrated with rail transit, and apply strong economic and traffic restraint measures to curb personal motor vehicle ownership and use, it is unclear how successful Delhi will be in accomplishing these ends, which as discussed are so necessary for the success of urban rail systems. Finally, while Singapore (for example) is a small, wealthy city state, Delhi is predominantly poor, growing in all directions, cannot control in-migration, which is occurring at a rapid rate, and attracts motor vehicle activity from the surrounding regions (NCTD 2001; RITES/ORG 1994; Singh 1997b; WHO/UNEP 1992). Last but not least in the complex of factors that contribute to motorization and transport emissions and energy consumption is urbanization itself. It is important to understand the forces that drive the phenomenon, and the role of government. The locational advantages, concentration of economic and political power, and economies of scale of large cities enable them to generate economic activity and jobs disproportionate to their share of national populations. Because economic and political power  are  concentrated in urban centres, little public investment occurs in rural areas, limiting productivity and non-farm employment. Land allocation regimes often compel the poor to cultivate marginal land, with no tenure. The resulting environmental degradation and poverty drive migration to the cities. While urbanization accelerates, the dwindling resources of the hinterland are bled further to cope with its impacts. Environmental degradation, poverty, population growth and urbanization thus spiral on inexorably (Brandon and Ramankutty 1993; Brown and Young 1990; Harrison 1988; The Economist 1995; WHO 1992).  80  Because of its geographic location, Delhi has been an important trading and commercial centre for centuries. Several national highways and railway lines converge there. Its pre-eminent position was reinforced by the concentration of government, administrative, and commercial services and employment opportunities in this century. The most recent spurt in Delhi's urbanization was caused by massive investments related to the Asian Games in 1984. Delhi now adds around 600,000 people to a population of 13 millions every year ( D D A 1996; Dhingra 1997; Encyclopaedia Britannica 1998; Singh 1997b).  3.4  38  CONCLUSIONS  The foregoing discussion highlights the fact that transport air pollution is not just a matter of vehicle, fuel and oil technology. A multitude of other technological, transport system, climatic and user behavioural factors exacerbate per-vehicle kilometre emissions due to vehicle and fuel technology. These factors include congestion, road quality, fuel system evaporative controls, vehicle maintenance, and fuel and oil adulteration. While the above factors influence per vehicle-kilometre emissions, motor vehicle activity, in terms of vehicle kilometres as well as vehicle trips, also plays a key role in transport air pollution and energy consumption. After all, vehicle and fuel technology have if anything improved over the last decade. Transport air pollution has become an increasingly important public policy issue in Delhi and other Indian cities largely due to rapidly growing motor vehicle activity, and resulting congestion. The discussion in the chapter also highlights the importance of considering system-wide emissions due to motor vehicle activity, including re-suspended dust, an important source of transport P M emissions, and evaporative emissions from the vehicle as well as the fuel distribution system.  Industry and employment have been growing rapidly in Delhi. Its industrial units increased from 8000 in 1951 to 125,000 in 1991. In 1961-71 alone, small industries grew 444%. But Delhi is by no means unique. Bangkok, Dhaka, Manila, Mexico City and Shanghai also account for economic activity and jobs disproportionate to their share of population. Mumbai, with 1% of India's population, generates 10% of its industrial jobs, and more than a quarter of its foreign trade (Brandon and Ramankutty 1993; Misra et al 1998). 3 8  81  Technological, transport system and vehicle user behavioural factors that influence transport air pollution and energy consumption have underlying institutional causes. Travel behaviour in terms of personal motorized mode ownership and use, and vehicle operation and maintenance behaviours such as fuel and oil adulteration, depend on incomes, and user preferences and choices, but are also responses to circumstances in which vehicle users find themselves, which in turn are influenced strongly by institutional factors. These factors include the quality of public transit service, provision for nonmotorized modes, and fuel and spare parts prices. In short, as Figures 3.2 and 3.3 show, a whole complex of inter-locking technological, climatic, vehicle user behavioural, institutional, and ultimately political factors, influence motor vehicle activity, and transport air pollution and energy consumption. A good example of how these various factors interact is the role of fuel pricing and monitoring and enforcement in fuel adulteration, and in turn in fuel quality. The complex of factors in Figures 3.2 and 3.3 shows the magnitude of the public policy challenge. If we are to effectively deal with transport air pollution and energy over the long term, all of these factors need to be addressed in a comprehensive and integrated manner. In this connection, note the critical role of government policy with regard to vehicle emission and fuel quality standards, monitoring and enforcement of in-use vehicle emissions and fuel and oil quality, vehicle, fuel and spares duties and taxes, full cost pricing of travel, planning and infrastructure provision for private versus public and non-motorized modes, land use and housing, and regional development. Of course, some of these factors are more easy to address than others, particularly given resource constraints. Urban form and land use are strong determinants of personal motorized vehicle ownership and use (Kenworthy et al 1994; Kenworthy and Laube 1999a), but are rather difficult to control, particularly to influence travel demand and patterns, and especially over the short term. This is true even in the West where the potential for such control exists by virtue of effective local governments, because of strong cultural and political barriers (Downs 1992). In the case of Delhi, as discussed in Chapter IV, there are serious constraints in the ability to control land use, especially at the urban fringe (Chandra 1997; Misra et al 1998; Singh 1997b).  82  But in any case, Delhi, including its rural areas, already has population densities ranging from 65 persons/ha to 665 persons/ha, with an average value (estimated by the author) of 90-100 persons/ha. Even in recently developed areas at the urban fringe, population densities are 200 persons/ha (DDA 1996; Mohan et al 1997; Tiwari and Kale 1997). These figures for Delhi are higher than those for North American and European cities, and even for Asian countries, with the exception of Bangkok, Hong Kong, Jakarta, Manila and Seoul (Poboon et al 1994; Raad and Kenworthy 1998; Kenworthy and Laube 1999a; Kenworthy and Laube 1999b). It is important to note that while the densities quoted above for Delhi were computed based on the total land area including rural areas, those for the other cities were based on the urbanized area and excluded all undeveloped land, regional open space, forests, agricultural land and water bodies (Kenworthy and Laube 1999a). Further, land use in Delhi is quite mixed in both the inner and outer areas, in part due to a conscious effort to develop district centres with mixed-use zoning, but also due to pressure on scarce land and high land prices, and the inability of authorities to control the proliferation of commercial operations to serve peoples' needs. So much so that household trip rates for different purposes in the outer areas are similar to those in the inner areas (Mohan et al 1997; Tiwari and Kale 1997). More importantly, non-motorized modes and public transit account for 39 and 42% of all trips, and 22 and 53% of work trips respectively, even among residential households. Further, public transit accounts for 71% of total passenger-kilometres in motor vehicles (RITES/ORG 1994). These figures compare favourably with those quoted for the least automobile dependent Asian (and global) cities in Kenworthy and Laube (1999a), with the possible exception of Hong Kong.  39  Moreover, sprawl in Delhi is quite different from, and has been caused by different reasons than, the sprawl in North America (for example). Urban sprawl in the latter case, involving small populations spread over large areas, has in part been made possible by high incomes, abundant land, substantial and subsidized transport infrastructure, and low energy prices, as Hanson (1992) argues. In a situation like Delhi's, on the other hand, Such high non-motorized mode and public transit shares in Indian cities are enabled by high densities and mixed land use, but are also due to the fact that the majority have low incomes and cannot afford personal motorized modes. 3 9  83  sprawl has occurred due to rapidly growing population pressures, which in turn have been caused largely by the in-migration of the rural poor from the hinterland, as discussed. Further, none of the features that made sprawl in North America possible have obtained in the case of Delhi, except for the fact that several national highways pass through the region. In Delhi, the lower and middle income groups have been priced out of the land market in the inner areas and have been forced to move to areas in the periphery despite the lack of adequate transit and other services there. Indeed, land demand has been concentrated outside the urban boundaries over the past three decades (Misra et al 1998). Under these circumstances, it is hardly surprising that large numbers who can 40  afford M 2 W vehicles purchase and use them, as discussed. Finally, if land use control implies influencing where people live with respect to where they work, it should be noted that many of the millions who live in Delhi's slums and squatter settlements often move where work (in the informal sector) is available. Given all of the foregoing, land use control in Delhi is neither as feasible nor as imperative as in the West.  41  As for urban rail in Delhi: it is difficult to say when it will be fully operational, and how effective it will eventually be in reducing personal motor vehicle activity and emissions. But it is reasonable to expect that the population and personal motor vehicle activity will continue to grow in Delhi and the surrounding regions in the foreseeable future, for all of the reasons discussed earlier in this Chapter. M2W vehicles will very likely continue to be the personal motor vehicle of choice for the middle classes for a long time to come in Delhi and the rest of the country, because of the lack of other affordable and attractive options. Further, urban rail, and for that matter public transit generally, will contribute nothing to reducing air pollution from the goods sector, which is predominantly diesel based and accounts for the bulk of P M , SO2 and N O 2 emissions in Delhi, as discussed in Chapter II. Finally, it is unlikely that urban rail will significantly displace buses, which are  Additionally, resettlement of the poor in the urban fringes during the mid-1970s drastically increased trip lengths for them (Mohan et al 1997). However, much higher densities than currently obtain would be desirable, especially in some of the inner areas in South Delhi. As Hong Kong's experience shows, inner city areas can accommodate densities of 300-400 persons/ha without compromising healthy environments (Mohan et al 1997). It would also be desirable to plan for high densities and mixed land use to enable high public transit and non-motorized mode use, particularly in the newly developing areas, and in areas along the proposed MRTS route in Delhi.  4 0  41  84  also diesel operated. Indeed, feeder buses will be needed to transport passengers to rail stations. In view of the foregoing, vehicle, fuel and transport infrastructure technologies will be important issues to address, particularly since these technologies in India lag behind global standards. In this regard, note that emissions and other transport impacts in Delhi are already exceedingly high despite far lower absolute motor vehicle activity levels compared to O E C D cities, and despite significant public transit and non-motorized mode shares. At the same time however, technological solutions will not be sufficient since they can be neutralized by increases in motor vehicle activity and congestion. Further, given multiple urgent demands on scarce resources, and factors such as high population pressures, there is little scope to expand transport infrastructure to accommodate growing motor vehicle activity.  42  A wide range of technological, economic, regulatory as well as transport  demand reduction policies will therefore be needed to address the transport air pollution problem effectively over the long term. This dissertation recognizes and reflects this need, by addressing the questions of how to think systematically about policies generally, and of what issues and perspectives to consider in order to make policies more attractive and effective over the long term. And although the dissertation focuses on technological and regulatory policies targeted at M2W vehicles, it has relevance for these and other policies targeted at other modes as well. Lastly, consideration of system-wide sources of pollutant emissions due to motor vehicle activity, and of contributory factors as discussed, including in particular in-use vehicle operation and maintenance, and fuel and oil quality realities, will make emissions measurement and modeling efforts more effective, in turn allowing for more effective policy evaluation and monitoring. A framework for modeling long-term emissions and 43  In this connection, it is worth critiquing the view that because there is great scope for improving transport efficiency in the LICs, their motor vehicle usage could increase two to three times without affecting their contribution to greenhouse gases (Faiz et al 1992). This may well be true, but it does not consider other transport emissions and impacts, and how LIC cities, with their meagre resources, might cope with them. In this regard, the importance of a driving cycle that reflects local driving conditions cannot be stressed enough. In the Indian context, this is borne out by studies conducted by Gandhi et al (1983), and by Tiwari and Kale (1997). Although dated, the former study, based on tests on the driving patterns of a car in Delhi, showed that acceleration and deceleration accounted for as much as 78% of the total driving time, significantly higher than in the ECE cycle (on which the Indian driving cycle is based). The latter 4 2  4 3  85  other policy impacts that takes into consideration these in-use realities, among other things, is proposed in Chapter V .  study confirmed the predominance of accelerations and decelerations, and also showed that driving patterns varied significantly from mode to mode. Finally, it is worth noting that maximum acceleration levels 3-5 times higher than in the US FTP and ECE cycles have been recorded on M2W vehicles in Bangkok (Faiz et al 1996).  86  87  88  CHAPTER IV TRANSPORT AIR POLLUTION IN INDIA: A DISCUSSION OF THE INSTITUTIONAL SETTING  4.1  INTRODUCTION  4.1.1 The Urban Challenge in Asian LICs, and the Role of Institutional Factors The global urban challenge is greatest in the LICs, particularly in Asia. Urbanization is most pronounced in Asia, with the bulk of urban growth occurring in the region's low income countries. In 2025, South Asia's urban population will likely be 1.4 billion (out of a total global population of 8 billion). Rapid urbanization in the LICs is characterized by a proliferation of megacities. In 2015, India alone will likely have four of the world's 27 (and Asia's 15) megacities, in addition to 40 cities with over one million population. Rapid 1  urbanization is causing massive environmental and social impacts in LIC cities, including those related to air pollution and transport. And it is the poor who bear the brunt of these impacts, since they are typically the most exposed to, affected by, and least capable of coping with them. Nearly half the world's poor will likely be urban, and concentrated mainly in South Asia. The poor already account for 45-60% of Calcutta's and Chennai's populations (Brandon and Ramankutty 1993; Hardoy, Mitlin and Satterthwaite 1992; Midgley 1994; Romieu, Weitzenfeld and Finkelman 1991; Romieu et al 1992; The Economist 1995). The serious urban situation in the LICs is rendered more daunting by the fact that, even as basic urban infrastructure and services are already woefully inadequate, and the resources necessary to provide them dwindle, demands multiply rapidly. While London's population took 100 years to grow from one to seven million, Delhi's increased from 0.7 to 13 million in a mere 50 years. Moreover, LIC cities lack the resources and power of their industrialized country counterparts, which are able to keep their local environments clean and export their wastes, thus directing their impacts mostly at the global level (Brown and Jacobson 1987; Midgley 1994; The Economist 1995; W C E D 1987; White and Whitney 1992). This fact is borne out, interestingly, by the OECD accounting for the bulk of motor vehicle activity, and India's megacities in 2015 will be Mumbai (with a projected population of 27 million), Calcutta, Delhi and Hyderabad, with Bangalore and Chennai not far behind (Brandon and Ramankutty 1993; Midgley 1994; The Economist 1995). 1  89  transport energy consumption and greenhouse gas emissions, with potentially global effects, but a much smaller proportion of emissions with adverse local or regional effects, whereas the situation is reversed for the LICs, as discussed in Chapter II (Faiz et al 1992; Grubler 1994; Walsh 1994). Institutional and political factors also play a crucially important role in urban environmental problems. Urban environmental issues are complex and multi-dimensional, and highly effective institutional mechanisms are required to address them. Yet, LIC governments largely lack the  institutional capacity to formulate,  implement and enforce  urban  environmental policies. It is not so much in terms of environmental legislation or even political commitment, but implementation, and monitoring and enforcement at the local level, that LIC institutional weaknesses are greatest (Brandon and Ramankutty 1993; Douglass and Lee 1996; Hardoy, Mitlin and Satterthwaite 1992). One important reason for weak institutional capacity at the local level is the fact that in most Asian countries, governance is dominated by the central government, and municipal governments have little power. Perhaps most importantly, local governments lack the authority to raise revenues, even as their mandates are expanding. Urban environmental agencies in particular have low status, and inadequate political authority and human resources to change the behaviours of firms and individuals. Thus, environmental regulations are not taken seriously. Weak institutional capacity is exacerbated by jurisdictional complexity. Actions by agencies at the municipal as well as provincial and national levels have an important bearing on urban environmental outcomes. However, these agencies often have unclear, overlapping, and uncoordinated responsibilities. Further, while urban environmental problems are cross-sectoral, most planning and investment is sectoral (Brandon and Ramankutty 1993; Douglass and Lee 1996; Hardoy, Mitlin and Satterthwaite 1992). Environmental policy and planning are also hampered by the inability to collect reliable information and conduct quality policy analyses. National capabilities that do exist in these areas are often not tapped, because of poor interaction between the policy-analytic and decision-making communities. The inability (or unwillingness) to involve the public at large results in less support for long term operations. At the same time, ad hoc responses to specific local pressures, rather than coherent, sustained programmes, become the norm (Brandon and  90  Ramankutty 1993; Douglass and Lee 1996; Hardoy, Mitlin and Satterthwaite 1992; Kandlikar 1998).  4.1.2  Chapter Objectives and Outline  As Chapter III showed, transport-generated air pollution is a complex problem, involving a range of inter-dependent technological, vehicle user behavioural and institutional factors. By its very nature, this problem requires a range of public policy interventions co-ordinated by an equally wide range of agencies. It would be useful to critically examine the institutional setting in relation to this problem in the Indian context, in terms of the effectiveness with which various concerned agencies and actors in that context formulate and implement prevention and control policies. This chapter attempts this task.  2  The chapter addresses the following research questions: Who are the actors and what are their roles, responsibilities and interactions in terms of policy-making and implementation with respect to prevention and control of transport air pollutant emissions? What are the institutional barriers and constraints? And finally, what are the implications of all of the above for transport air pollution prevention and control in the Indian context? The discussion addresses actors' roles, responsibilities, interactions, and barriers and constraints in relation to the various aspects of the transport air pollution problem discussed in Chapter III. The actors include key government agencies at various levels, vehicle and fuel manufacturers, academic and research institutions, environmental NGOs, the courts and public interest litigators, and the media. The discussion is intended to help in identifying institutional barriers and constraints that are critical to long term policy effectiveness, and in developing institutional mechanisms and arrangements to overcome these barriers and constraints. At the same time, policies can be designed to be insensitive to the lack of these mechanisms, and therefore to have a better chance of long-term effectiveness, given contextual constraints. Finally, while the discussion focuses on prevention and control of M2W vehicle emissions prevention in Delhi, it has  A similar institutional analysis focusing on the Indian context is the work by Kandlikar and Sagar (1999), which examined climate change research and analysis in India, the factors that contribute to India's capabilities and constraints in this regard, and implications for international co-operation on climate change. 2  91  relevance for transport air pollution generally, and indeed other urban environmental issues, in the Indian context.  4.2  METHODOLOGY  The discussion in this Chapter is based on a critical analysis of a wide range of published as well as unpublished written material, including pertinent environmental legislation, reports and position papers prepared by local and national government agencies, environmental NGOs, and vehicle and fuel manufacturers and industry associations, transcripts of proceedings in Supreme Court public interest cases, excerpts from reports of the Saikia Committee, which was charged by the Court to develop emission control action plans, and media reports and commentaries. Information culled from the above sources was supplemented with that obtained in the course of in-depth interviews with various individuals interested in and/or knowledgeable about the range of issues involved, and representatives of institutions whose actions have an important bearing on transport air pollutant emissions in the Indian context. These individuals included decision makers in relevant government agencies at the national and local levels, vehicle and fuel industry representatives, and academics and researchers. A list of interviewees is provided in Appendix V . In addition to sharing their insights, the interviewees made available to the author the bulk of the documents referred to in the previous paragraph. The interviews were therefore invaluable in gaining a comprehensive understanding of policymaking and implementation processes, and institutional capabilities and constraints, in relation to transport air pollution in the Indian context. Interviewees' informed consent was obtained prior to interviews being conducted. The interview protocol, and the Informed Consent Form (Appendix VI), were approved by the Behavioural Research Ethics Board of the U B C Office  of Research Services and  Administration. One of the conditions of this approval was that the identities of interview participants would be kept confidential. It is for this reason that, while their information and insights were of immense value to the author, interviewees are not explicitly acknowledged in the following discussion.  92  4.3  ACTORS, RESPONSIBILITIES AND ROLES  The Ministry of Environment and Forests (MoEF) is the lead Government of India agency for all national environmental and forestry programmes. Its mandate includes environmental policy formulation, developing and enforcing environmental legislation, and executing, coordinating and monitoring pollution prevention and control programmes. These tasks are implemented through its various divisions and agencies, one of which is the Central Pollution Control Board (CPCB). The CPCB's functions regarding air pollution include advising the Government of India, setting national air quality standards, air quality monitoring, and recommending motor vehicle emission and fuel quality standards, based on the work of committees representing, among others, R & D institutions and other ministries. The CPCB is also mandated to conduct and sponsor research, co-ordinate and assist state pollution control boards, and disseminate information on air pollution and its prevention and control. The enabling legislation include the Air (Prevention and Control of Pollution) Act (1981), and the omnibus Environment Protection Act (1986), enacted following the 1984 Bhopal disaster (CPCB 1996; C S E 1996; G o l 1981; G o l 1986; MoEF 1991; M o E F 1997a; TERI 1997a). While local state pollution control boards implement pollution control action in the Indian states (the equivalent of the Canadian provinces), the CPCB is mandated to play the role of a state board in Delhi and other union territories. Interestingly, it is also only for the union territories that CPCB is additionally mandated to conduct studies on air pollution effects. Finally, Delhi and other union territories have been declared 'air pollution control areas" (Gol 1982; G o l 1983), thus privileging these over other regions in the country, many of which have air pollution levels that are at least as high as in Delhi, as discussed in Chapter II (Brandon and Hommann 1995; CPCB 1992; CPCB 1996). M o E F recommends vehicle emission standards, but it is the Ministry of Surface Transport (MoST) that notifies and enforces them. Other responsibilities of MoST that are pertinent in the context of the present discussion include developing standards, rules, and procedures with regard to type approval and in-use vehicle emissions testing, and vehicle licensing, registration, road-worthiness, and service life, through the Central Motor Vehicle Rules (MoST 1996; Universal Law Publishing 1995). It is worth noting that, while MoST is responsible for enforcing vehicle emission standards, the agency's primary mandate is the  93  well-being of the transport industry, which is a major economic player and employer nationally. MoST's role in transport air pollution prevention and control is inherently conflicted, since the agency has the task of regulating the industry whose interests it also has the mandate to promote. The Ministry of Petroleum and Natural Gas (MoPNG), along with the oil refineries, which are predominantly state-owned (this is changing), are responsible for exploration, production, refining, distribution, marketing, and export and import of crude oil and petroleum products. As such, M o P N G and the oil refineries determine the quality of Indian transport fuels and lubricating oils, which critically influence transport emissions. Fuel and oil quality standards are nominally developed by the Bureau of Indian Standards, an independent statutory body, through committees which include the oil refineries and vehicle manufacturers, but it is actually M o P N G that drafts and implements them. Other energy-related government agencies which influence policies and actions affecting transport emissions include the Energy Policy Unit of the Planning Commission, whose mandate is to plan for energy self-sufficiency, the Oil Co-ordination Committee (OCC), which coordinates and monitors oil imports, exports, and refining, and perhaps most importantly administers fuel pricing (through the 'administered pricing mechanism", or A P M ) ,  3  and the Petroleum Conservation Research Association  (Author's interviews 1997; BIS 1995a; BIS 1995b; BIS 1996; C S E 1996; TERI 1997a). MoEF, MoST and M o P N G are the three national government agencies whose policies and actions most critically influence transport air pollution and energy consumption. But other agencies at the national level also play a role. The Ministry of Industry (Mol), whose mandate is industrial policy and promotion, was powerful until the early 1990s, but their role has declined considerably because of economic liberalization and rising environmental concerns. The Ministry of Finance plays a crucial role in setting excise and import duties and tariffs, and vehicle, fuel and spares sales taxes. Both the Union Ministry of Health, and its Delhi government counterpart, surprisingly play a marginal role in air pollution control (Author's interviews 1997; CSE 1996; CSE 1997).  The cross-subsidization of kerosene and diesel, among other fuels, under the APM, was discussed in Chapter III. 3  94  Several academic and research institutions in the public, private and quasi-public sectors are involved in R & D and policy analysis related to urban transport, vehicle and fuel technology, and transport emissions and control. Representatives of these institutions serve on committees to set vehicle emissions and fuel and oil quality standards, and/or investigate the problem of transport air pollution and energy consumption generally (BIS 1995a; BIS 1995b; BIS 1996; C S E 1996; IIP 1994; MoEF 1991; Mohan et al 1997; Mohan and Tiwari 1997; RITES/ORG 1994; TERI 1997a).  4  In addition to the above agencies at the national level, there are several local agencies in Delhi with jurisdiction over issues that have a significant bearing on the city's transport air pollution. As indicated earlier, the CPCB is mandated to play the role of a state pollution control board in Delhi and other union territories (Gol 1982; G o l 1983). At the same time, the Delhi Pollution Control Committee (DPCC), an agency of the Government of the National Capital Territory of Delhi (NCTD) has responsibility for pollution control in the city (Prem Kumar 1997). But CPCB reportedly conducts its work in Delhi in collaboration with the D P C C (Author's interviews 1997). Nevertheless, this situation has the potential to be a recipe for jurisdictional confusion. The Transport Department of the Government of N C T D is responsible primarily for vehicle licensing, registration, and inspection and road taxation in Delhi. As will be discussed later, vehicle licensing and registration are functions with important implications for in-use transport emissions control, for which the agency is also responsible. In this connection, the Transport Department of the Government of N C T D certifies and licenses fuel dispensing and service stations to test in-use vehicle emissions and repair non-complying vehicles. The regulation of vehicle service life, and therefore vehicle scrappage, is MoST's responsibility, as already indicated, but it is the Transport Department of the Government of N C T D that enforces this regulation. Further, in the case of Delhi exclusively, MoST has delegated the  These institutions include the Automotive Research Association of India (ARAI), Central Institute of Road Transport, Central Road Research Institute (CRRI), Indian Institute of Petroleum (IIP), the Indian Institutes of Technology, Indian Oil Corporation R&D Centre, National Environmental Engineering Research Institute, Operations Research Group, Rail India Technical and Economic Services Ltd., and Tata E