For a Breath of Fresh Air i For a Breath of Fresh Air ii For a Breath of Fresh Air Ten Years of Progress and Challenges in Urban Air Quality Management in India 1 9 9 3 - 2 0 0 2 Environment and Social Development Unit South Asia Region THE WORLD BANK iii For a Breath of Fresh Air Published by Environment and Social Development Unit, South Asia Region, The World Bank (India Country Office), 70 Lodi Estate, New Delhi 110003 Internet: www.worldbank.org/in Published in June 2005 Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. The findings, interpretations and conclusions expressed herein are those of the author(s), and do not necessarily reflect the views of the International Bank for Reconstruction and Development/The World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgement on the part of The World Bank concerning the legal status of any territory or the endorse- ment or acceptance of such boundaries. Cover design : Gulshan Malik Cover photos : Center for Science and Environment, New Delhi (Pictures of CNG vehicles and flyover contributed by The World Bank). City Illustrations : Kumar Arts Editorial & Production Consultant : Tapan K. Ghosh Printed at Nutech Photolithographers, New Delhi iv Table of Contents Foreword ................................................................................................................xiii Acknowledgments ................................................................................................................. xv Abbreviations and Acronyms..............................................................................................xvii Executive Summary ....................................................................................................................1 Chapter 1 : Introduction......................................................................................................5 Focus on Respirable Particulate Matter .................................................6 City Selection ...................................................................................................6 Methodological Framework ........................................................................7 Structure of the Report .................................................................................8 Chapter 2 : City-specific Interventions ..........................................................................9 Delhi....................................................................................................................9 Kolkata ...............................................................................................................9 Mumbai ........................................................................................................... 11 Hyderabad ..................................................................................................... 11 Chennai........................................................................................................... 11 International Experience with Improving Urban Air Quality .... 11 Chapter 3 : Urban Air Quality : Levels and Trend Analysis .............................. 14 Air Quality Data .......................................................................................... 14 Air Quality Standards ............................................................................... 14 Levels and Overall Trends ....................................................................... 15 City-specific Analysis................................................................................ 18 Delhi ...................................................................................................... 18 Kolkata.................................................................................................. 20 Mumbai................................................................................................. 20 Hyderabad........................................................................................... 20 Chennai ................................................................................................ 21 Chapter 4 : Understanding the Determinants of Urban Air Quality ............... 22 Evidence of Relative Contribution of Sources................................... 22 Meteorological Parameters ...................................................................... 23 Chapter 5 : The Health Impacts of Air Quality Improvement ............................ 27 Chapter 6 : Concluding Remarks.................................................................................. 29 v For a Breath of Fresh Air Annexes Annex 1 : Cleaning the Air in Delhi ­ Chronology of Key Events, 1985 - 2002 ........................................... 33 Annex 2 : Air Quality Monitoring by NEERI ........................................................ 35 Annex 3 : Data Quality Issues .................................................................................... 40 Annex 3A : Comparison of NEERI, CPCB and SPCB data................................... 45 Annex 4 : National Ambient Air Quality Standards .......................................... 55 Annex 5 : City-specific Case Studies Annex 5A : Delhi................................................................................................................. 56 Annex 5B : Kolkata ............................................................................................................ 71 Annex 5C : Mumbai ........................................................................................................... 83 Annex 5D : Hyderabad ..................................................................................................... 95 Annex 5E : Chennai......................................................................................................... 106 Annex 6 : Health Cost of Exposure to RSPM Pollution ................................... 117 Annex 7 : Questionnaire Survey for Stakeholder Input................................... 120 References ............................................................................................................... 123 Notes ............................................................................................................... 126 Tables Main Text Table 1 : Dispersion conditions and RSPM concentrations in five cities ... 24 Table 2 : Three-year annual averages of RSPM and the levels corresponding to the four scenarios.......................................................27 Table 3 : Health benefits due to reduction in RSPM levels in five cities in India under four different scenarios .................................................28 Table 4 : Sector-wise summary of key actions taken in the five cities that could.have helped in improving air quality .......................................31 Annexes Annex 2 Table A2.1 : Monitoring procedures for all.pollutants ............................................35 Annex 3 Table A3.1 : Missing SPM and RSPM data in each city ............................................43 Annex 5A Table A5A.1 : Monthly average values of key meteorological parameters for 1991-2002 (Delhi) ...........................................................................................57 vi Table of Contents Annex 5B Table A5B.1 : Overall monthly average values of key meteorological parameters (Kolkata) .................................................................................. 74 Annex 5C Table A5C.1 : Overall monthly average values of key meteorological parameters (Mumbai) ..................................................................................84 Annex 5D Table A5D.1 : Overall monthly average values of key meteorological parameters (Hyderabad) ............................................................................96 Annex 5E Table A5E.1 : Overall monthly average values of key meteorological parameters(Chennai) ............................................................................... 108 Annex 6 Table A6.1 : Air pollution concentration-response function slope (bij) per 1µg/m3 change in the mean annual level of PM10 ......................... 118 Table A6.2 : Willingness-to-Pay based health .effect costs per case derived for.in Delhi ................................................................................... 119 Figures Main Text Figure 1 : Ambient annual averages of SPM in selected large cities in Asia and the Pacific in µg/m³ in 1995 ...............................................................5 Figure 2 : Ambient annual averages of SPM and RSPM in five large Indian cities in µg/m³ in 2002 .................................................................... 7 Figure 3 : Historical PM10 emissions in the U.K. ...................................................12 Figure 4 : Annual average SPM concentrations in the five cities .....................15 Figure 5 : Annual average RSPM concentrations in the five cities ..................16 Figure 6 : Annual average SO2 concentrations in the five cities .......................17 Figure 7 : Annual average NO2 concentrations in the five cities ......................17 Figure 8 : Trend in monthly average concentration of SPM and RSPM in an industrial area of Delhi .........................................................................19 Figure 9 : Trend in monthly average concentration of RSPM in an industrial area of Mumbai ...................................................................21 Figure 10 : Receptor Modeling of PM2.5 in Delhi, Kolkata, and Mumbai ........23 Figure 11 : Monthly average temperature based on the period 1993-2002 .....25 Figure 12 : Combined (for all areas) monthly averages of RSPM, 1993-2002 ... 25 vii For a Breath of Fresh Air Annexes Annex 3A Figure A3A.1 : Comparison between CPCB and NEERI RSPM data for 2000-2003 (Delhi) ...........................................................................................45 Figure A3A.2 : Linear regression between NEERI and CPCB RSPM data, 2000-2003 (Delhi) ...........................................................................................46 Figure A3A.3 : Linear regression between RSPM data from NEERI and CPCB residential areas, 2000-2003 (Delhi) ........................................................47 Figure A3A.4 : Comparison between CPCB and NEERI SPM data, 1994-2003 (Delhi)...............................................................................................................47 Figure A3A.5 : SPM and RSPM monitored by CPCB 2000-2003 (Delhi)...................48 Figure A3A.6 : Linear regression between RSPM and SPM monitored by CPCB, 2000-2003 (Delhi)..........................................................................................48 Figure A3A.7 : Ratio of RSPM to SPM monitored by CPCB, 2000-2003 (Delhi) ...48 Figure A3A.8 : Comparison between NEERI and SPCB RSPM data for 1999-2002 (Kolkata) .....................................................................................49 Figure A3A.9 : Linear regression between NEERI and SPCB RSPM data, 1999-2002 (Kolkata) ......................................................................................49 Figure A3A.10 : Ratio of RSPM to SPM monitored by NEERI,1999-2003 (Kolkata) ..........................................................................................................49 Figure A3A.11 : Comparison between NEERI and SPCB RSPM data for 2001-2003 (Hyderabad)....................................................................................................50 Figure A3A.12 : Linear regression between NEERI and SPCB RSPM data, 2001-2003 (Hyderabad) ...............................................................................50 Figure A3A.13 : Ratio of RSPM to SPM monitored by NEERI, 2001-2003 (Hyderabad)....................................................................................................51 Figure A3A.14 : Linear regression between RSPM data from NEERI and SPCB residential areas, 2001-2003 (Hyderabad).............................................51 Figure A3A.15 : Comparison between NEERI and SPCB RSPM data for 2001-2003 (Chennai) .....................................................................................51 Figure A3A.16 : Linear regression between NEERI and SPCB RSPM data, 2001-2003 (Chennai) .....................................................................................52 Figure A3A.17 : Ratio of RSPM to SPM monitored by NEERI, 2001-2003 (Chennai)..........................................................................................................52 Figure A3A.18 : SPM and RSPM monitored by NEERI 2000-2003 (Mumbai)...........53 Figure A3A.19 : Ratio of RSPM to SPM monitored by NEERI, 2000-2003 (Mumbai)............... .......................................................................................53 Annex 5A Figure A5A.1 : Delhi : Approximate location of NEERI monitoring stations in industrial, commercial and residential areas ......................................56 viii Table of Contents Figure A5A.2 : Population growth in NCT of Delhi between the last two census periods.......................................................................................57 Figure A5A.3 : Growth in commercial motor vehicles in Delhi............................... 58 Figure A5A.4 : Growth in private motor vehicles in Delhi ........................................ 58 Figure A5A.5 : Trend in RSPM levels in three areas of Delhi ................................... 60 Figure A5A.6 : Trend in SPM levels in three areas of Delhi....................................... 60 Figure A5A.7 : Trend in overall average RSPM concentration in Delhi................ 61 Figure A5A.8 : Trend in SO2 levels in Delhi..................................................................... 62 Figure A5A.9 : Trend in NO2 levels in Delhi ................................................................... 62 Figure A5A.10 :Trend in RSPM monthly concentration at the residential, commercial and industrial area monitoring locations ................. 63 Figure A5A.11 : Monthly average total rainfall and mean wind speed ................ 64 Figure A5A.12 : Monthly average total rainfall and RSPM concentration ........... 64 Figure A5A.13 : Monthly mean wind speed and RSPM concentration ................. 64 Figure A5A.14 : Variation in monthly averages of SO2 and NO2i ............................ 64 Figure A5A.15 : Overall variation in RSPM with seasons (Monsoon: July to September; Summer: February to June; Winter: October to January) .................................................................... 65 Annex 5B Figure A5B.1 : Kolkata - Approximate location of NEERI monitoring stations in industrial, commercial and residential areas ...............................71 Figure A5B.2 : Population in Kolkata metropolitan area between the last two census periods ...............................................................................................73 Figure A5B.3 : Growth in commercial motor vehicles in Kolkata ........................... 73 Figure A5B.4 : Growth in private motor vehicles in Kolkata .................................... 75 Figure A5B.5 : Trend in RSPM levels in three areas of Kolkata ............................... 75 Figure A5B.6 : Trend in SPM levels in three areas of Kolkata .................................. 76 Figure A5B.7 : Trend in SO2 levels in Kolkata ................................................................ 76 Figure A5B.8 : Trend in NO2 levels in Kolkata ............................................................... 76 Figure A5B.9 : Trend in RSPM monthly concentration at the residential, commercial and industrial area monitoring locations .................. 79 Figure A5B.10: Monthly average total rainfall and mean wind speed .................. 80 Figure A5B.11 : Monthly average total rainfall and RSPM concentration ........... 80 Figure A5B.12: Monthly mean wind speed and RSPM concentration ................... 80 Figure A5B.13: Variation in monthly averages of SO2 and NO2................................ 81 ix For a Breath of Fresh Air Figure A5B.14: Overall variation in RSPM with seasons (Monsoon: May to October; Summer: February to April; Winter: November to January) .................................................................81 Annex 5C Figure A5C.1 : Mumbai : Approximate location of monitoring stations in industrial , commercial and residential areas.....................................83 Figure A5C.2 : Population in Mumbai between the last two census periods .......84 Figure A5C.3 : Growth in commercial motor vehicles in Mumbai ............................85 Figure A5C.4 : Growth in private motor vehicles in Mumbai.....................................85 Figure A5C.5 : Trend in RSPM levels in Mumbai ...........................................................86 Figure A5C.6 : Trend in SPM levels in Mumbai ..............................................................86 Figure A5C.7 : Trend in SO2 levels in Mumbai ................................................................87 Figure A5C.8 : Trend in NO2 levels in Mumbai ...............................................................87 Figure A5C.9 : Trend in RSPM monthly concentration at the residential, commercial and industrial area monitoring locations .....................90 Figure A5C.10 : Monthly average total rainfall and mean wind speed .....................91 Figure A5C.11 : Monthly average total rainfall and RSPM concentration ................91 Figure A5C.12 : Monthly mean wind speed and RSPM concentration ......................92 Figure A5C.13 : Variation in monthly averages of SO2 and NO2 ..................................92 Figure A5C.14 : Overall variation in RSPM with seasons (Monsoon: June to September; Summer: March to May; Winter: October to February) ....................................................................93 Annex 5D Figure A5D.1 : Hyderabad - Approximate location of NEERI monitoring stations in industrial, commercial and residential areas ...............................95 Figure A5D.2 : Population in Hyderabad Development Area between 1991 and 2001 ...........................................................................................................96 Figure A5D.3 : Growth in commercial motor vehicles in Hyderabad ......................97 Figure A5D.4 : Growth in private motor vehicles in Hyderabad ...............................97 Figure A5D.5 : Trend in RSPM levels in three areas of Hyderabad ..........................98 Figure A5D.6 : Trend in SPM levels in three areas of Hyderabad.............................99 Figure A5D.7 : Trend in SO2 levels in Hyderabad ..........................................................99 Figure A5D.8 : Trend in NO2 levels in Hyderabad .......................................................101 Figure A5D.9 : Trend in RSPM monthly concentration at the residential, commercial, and industrial area monitoring locations ............... 102 x Table of Contents Figure A5D.10 : Monthly average total rainfall and mean wind speed .............. 103 Figure A5D.11 : Monthly average total rainfall and RSPM concentration ......... 103 Figure A5D.12 : Monthly mean wind speed and RSPM concentration ............... 103 Figure A5D.13 : Variation in monthly averages of SO2 and NO2 ........................... 104 Figure A5D.14 : Overall variation in RSPM with seasons (Monsoon: June to October; Summer: February to May; Winter: November to January) ............................................................. 104 Annex 5E Figure A5E.1 : Chennai - Approximate location of NEERI monitoring stations in industrial, commercial and residential areas ...............................106 Figure A5E.2 : Population growth in Chennai metropolitan area between the last two census periods .....................................................................107 Figure A5E.3 : Growth in commercial motor vehicles in Chennai ..........................109 Figure A5E.4 : Growth in private motor vehicles in Chennai ...................................109 Figure A5E.5 : Trend in RSPM levels in three areas of Chennai .............................. 110 Figure A5E.6 : Trend in SPM levels in three areas of Chennai ................................. 110 Figure A5E.7 : Trend in SO2 levels in Chennai............................................................... 111 Figure A5E.8 : Trend in NO2 levels in Chennai ............................................................. 111 Figure A5E.9 : Trend in RSPM monthly concentration at the residential, commercial and industrial area monitoring locations ................... 113 Figure A5E.10 : Monthly average total rainfall and mean wind speed ................... 114 Figure A5E.11 : Monthly average total rainfall and RSPM concentration .............. 114 Figure A5E.12 : Monthly mean wind speed and RSPM concentration .................... 115 Figure A5E.13 : Variation in monthly averages of SO2 and NO2 ............................... 115 Figure A5E.14 : Overall variation in RSPM with seasons (Monsoon: September to November; Summer: March to August; Winter: December to February) ........ 115 Boxes Box 1 : Recent Supreme Court directions on urban air pollution in India ...............................................................................................................6 Box 2 : Chronology of Key Actions implemented in Delhi ......................... 10 Box 3 : Air Quality at a busy traffic intersection in Delhi ........................... 19 Box 4 : Stakeholder feedback on the role of the World Bank in UAQM..........................................................................................................32 Box A3.1 : Structural issues with NAMP and implications for Data Quality ...................................................................................................44 xi For a Breath of Fresh Air Box A5A.1 : Supreme Court Directions for industrial closures .......................... 59 Box A5A.2 : PM2.5 in Ambient Air in Delhi ................................................................. 66 Box. A5A.3 : Major technical and policy interventions in Delhi (1991-2004) that impacted Air Quality ........................................................................... 68 Box A5B.4 : Major technical and policy interventions in Kolkata (1991-2004) that impacted Air Quality ........................................................................... 77 Box A5C.5 : Results of RSPM Source Apportionment in Mumbai .......................... 88 Box A5C.6 : Major technical and policy interventions in Mumbai (1991 -2004) that impacted Air Quality ........................................................................... 89 Box A5D.7 : Major technical and policy interventions in Hyderabad (1991-2004) that impacted Air Quality................................................... 100 Box A5E.8 : Major technical and policy interventions in Chennai (1991-2004) that impacted Air Quality.......................................................................... 112 xii Foreword With robust economic growth raising living therefore important to view this report as one of the standards in urban areas and a growing number of many contributions to a broader debate on what people moving to cities, urban air quality in India is it will take for large Indian cities to have clean air. of increasing concern for the public and policy- makers. While a bold program of measures to ease The findings of this report need to be complemented air pollution has been implemented in Delhi and by the findings of other studies, including a recently some other major metros, more remains to be done completed regional World Bank study on issues in in urban centres across India in order to bring air urban air quality management in South Asia and a quality to safe levels for public health, as set in large body of work undertaken by other institutions national and international standards. Presently, in India and internationally. This report was many cities are at the crossroads, having recently discussed at a stakeholder workshop in October been directed by the Supreme Court to develop air 2004, and presented at the regional Clear Air quality improvement action plans. Initiative ­ Asia conference at Agra in December 2004. This report, prepared in collaboration with the Central Pollution Control Board (CPCB), is a We would like to express our gratitude to the CPCB, contribution to the on-going efforts to assist cities the State Pollution Control Boards, and many stake- with developing or updating their air quality holders from government and non-government management strategies. It attempts to analyze air organizations and civil society who contributed their quality trends and interventions implemented in the time and knowledge in the preparation of this report. ten-year period between 1993 and 2002 in order to better inform future actions. Inevitably, the scope Michael F. Carter and power of the quantitative analysis presented in Country Director this report is constrained by data limitations. It is India xiii For a Breath of Fresh Air xiv Acknowledgments This study has been conducted by the South Asia Pune, Dr Ranjan Bose (Senior Fellow), Tata Energy Environmental and Social Development Unit Research Institute, New Delhi, Mr Atanu Ganguli (SASES) of the World Bank with assistance from the (Assistant Director), Society of Indian Automobile Central Pollution Control Board (CPCB), Delhi. Manufacturers, Delhi, Mr M. M. Sommaya Dr B. Sengupta (Member Secretary) and (Manager), Bharat Petroleum, Mumbai, and Dr Shantanu Dutta (Senior Project Scientist) of CPCB Dr Rita Pandey (Fellow), National Institute of Public supported the work. The other CPCB staff members Finance and Policy, Delhi. who provided access to data and information were Dr R.C. Trivedi (Additional Director), and The World Bank acknowledges the contribution of all Mr Naresh Badhwar (Asst. Environmental Engineer). the respondents of the Urban Air Quality Questionnaire Survey which was a part of this study, Information was also provided by the State Pollution and is grateful to Clean Air Initiative ­ Asia, for Control Boards, and Transport Departments of Delhi, putting up the questionnaire on their website and list Andhra Pradesh, Karnataka, Maharashtra, Tamil server. Nadu and West Bengal. Assistance was provided by Dr K.V. Ramani (Jt Chief Environmental Scientist), The World Bank team consisted of Sameer Akbar APPCB, Mr B. Ramaiah (Member Secretary), KSPCB, (Task Team Leader), Masami Kojima (Lead Mr M.N. Jayaprakash (Sr Environmental Engineer), Energy Specialist), K. Mukundan (Sr Urban KSPCB, Mr C.R. Mohd Suleman (Jt Commissioner Specialist), M. Khaliquzaman (Consultant), Transport) Karnataka, Dr D.B. Boralkar (Member Kseniya Lvovsky (Lead Environmental Economist), Secretary), MPCB, Mr S.B. Sahasrabudhe Uma Balasubramanian and Faisal Zia Siddiqui (Dy Transport Commissioner), Maharashtra, (Consultant). Valuable comments from peer Ms Girija Vaidyanathan (Chairperson), TNPCB, reviewers Maureen Cropper, Ed Dotson, Jitu Shah, Dr T. K. Gupta (Sr Environmental Engineer), and Todd Johnson, all of the World Bank, as well as WBPCB, M. D.K.Bakshi (Dy Secretary) West Bengal external reviewers from CPCB, and Gunnar Eskeland Transport Department, Mr Rajeev Talwar (Transport (Director, CICERO), Norway, are gratefully acknow- Commissioner), Delhi, Mr T.M. Balakrishnan ledged. The team is grateful to Rachid Benmessaoud (Additional Transport Commissioner), Delhi. (Operations Advisor, India Country team) and Jeffrey S. Racki (Sector Director, SASES) for support The data collection efforts were further supported by and guidance. Dr Rakesh Kumar (Senior Assistant Director), National Environmental Engineering Research This work was undertaken in 2003-04 and the final Institute (NEERI), Mumbai, Dr Animesh Kumar report prepared in December 2004. The findings and (Scientist and Head) NEERI, Nagpur, Mr R.C. Dixit conclusions of this report are entirely those of the (Scientist and Head) NEERI Zonal Lab, Delhi, authors and should not be attributed to the World Dr R. Jayabalou (Scientist and Head), Chennai Bank, its Board of Executive Directors, or to the NEERI Center, Mr A.K. Biswas (Scientist and Head), countries they represent. Kolkata NEERI Center, Mr R.C. Srinivasan (Advisor-Transport Research), Ministry of Road SameerAkbar Transport and Highway, Delhi, Mr G.S. Prakasa Rao KseniyaLvovsky (Director), Indian Meteorological Department (IMD), Masami Kojima xv For a Breath of Fresh Air xvi Abbreviations and Acronyms APPCB: Andhra Pradesh Pollution Control Board LPG: Liquified Petroleum Gas AQM: Air Quality Management LSHS: Low Sulfur Heavy Stock Art. 21: Article 21 of the constitution of India which LMV: Light Motor Vehicle guarantees Right to Life and Liberty MCD: Municipal Corporation of Delhi BMC: Brihan Mumbai Corporation MCH: Municipal Corporation of Hyderabad °C: degree centigrade Mil.: million CMC: Chennai Municipal Corporation mm: millimeters CNG: Compressed Natural Gas MMRDA: Mumbai Metropolitan Regional Commercial Vehicles: as per MORTH definition these DevelopmentAuthority include the following: multi-axled/articulated buses and trucks, light motor vehicles (goods), buses, MoRTH: Ministry of Road Transport and Highways taxies, light motor vehicles (passengers) ­ MPCB: Maharashtra Pollution Control Board autorickshaws MPD-20001: Master Plan of Delhi 2001 CPCB: Central Pollution Control Board MT:MetricTonne CSE: Center for Science and Environment NAAQS: National Ambient Air Quality Standards DDA: Delhi Development Authority NAMP: National Air Quality Monitoring Program EPCA: Environmental Protection Control Authority NCR: National Capital Region ESMAP: Energy Sector Management Assistance Program NCT: National Capital Territory FO: fuel oil NDMC: New Delhi Municipal Corporation GEMS: Global Environment Monitoring System NEERI: National Environmental Engineering Research Institute GNI: Gross National Income NGO: Non-governmental Organization GSDP: Gross State Domestic Product PIL: Public Interest Litigation GOI: Government of India PM: Particulate Matter HUDA: Hyderabad Urban Development Area PM2.5: Particulate Matter less than 2.5 microns in IMD: Indian Meteorological Department aerodynamic diameter km2: Square kilometers PM10: Particulate Matter less then 10 microns in KMA: Kolkata Metropolitan Authority aerodynamic diameter (used interchangeably with RSPM in this report) KMC: Kolkata Municipal Corporation Private Vehicles: as per MoRTH definition these KSPCB: Karnataka Pollution Control Board include the following: two-wheelers, cars, jeeps, Kmph: Kilometers per hour omni buses, tractors, trailers, and other vehicles LCV: Light Commercial Vehicle RSPM: Respirable Suspended Particulate Matter (used interchangeably with PM10 in this report) xvii For a Breath of Fresh Air RTV: Rural Transport Vehicle (a minibus with a UAQM: Urban Air Quality Management good ground clearance) UNEP: United Nations Environment Program SC: Supreme Court VOSL: Value of Statistical Life SPCB: State Pollution Control Board WBPCB: West Bengal Pollution Control Board SPM: Suspended Particulate Matter WHO: World Health Organization TD: Transport Department WTP: Willingness to Pay TERI: The Energy Research Institute (formerly Tata Energy Research Institute) Note : Currency Conversion Rate used in the study TNPCB: Tamil Nadu Pollution Control Board 1 US dollar = Rs 45 approximately. xviii Profile of Sample Cities for this Study Delhi: The national capital located in northern India with a population of approximately 9.8 million people and average population density of 7,021 people per square kilometer (km2) is part of a larger national capital territory of 12.8 million inhabitants. It is located on the banks of the river Yamuna, and has a tropical semi- arid climate with hot summers and cold winters. The mean total annual rainfall is about 750 millimeters (mm), and low wind speeds prevail in winter months. Delhi has a per capita income of about Rs 24,450 (2000- 01) which is almost twice the national average. Industry plays an important role in economic growth, though in recent years its share has gone down as compared to the services sector. The number of motor vehicles in Delhi, about 4 million, is the highest in the country, with two-wheelers constituting approximately 65 percent. It has the distinction of having all road-based public transport running on compressed natural gas (CNG) since 2002. With a past that was heavily influenced by the British Raj , New Delhi is considered a well- planned city with broad roads and green spaces. However, Old Delhi is characterized by mixed land-use and a large number of pedestrians and slow moving traffic. Kolkata: The capital of West Bengal located in north-eastern India near the Bay of Bengal with a population of approximately 4.6 million people and average population density of 24,705 people per km2 is part of a metropolitan area of 13.2 million inhabitants. It is divided into Howrah on the west and Kolkata city on the river Hoogly flowing in a north-south direction. It has a humid tropical climate with hot summers and moderate winters. The mean total annual rainfall is about 1,700 mm, and low wind speeds prevail in winter months. Kolkata has a per capita income of about Rs 10,636 (1998-99). Industry plays a very important role in economic growth. The number of motor vehicles in Kolkata is about 700,000, with two-wheelers constituting approximately 45 percent. However, the river also serves as an important waterway to the rural hinterland. With a past that was also heavily influenced by the British Raj, new Kolkata is considered a well planned city. However, the Old city is characterized by mixed land-use and a large number of pedestrians and slow moving traffic. Mumbai: The capital of Maharashtra located on the west coast by the Arabian sea with a population of approximately 12 million people and average population density of 25,449 persons per km2is part of a metropolitan area of 16.4 million inhabitants. It has a humid tropical climate with hot summers and moder- ately cold winters. It has severe monsoons with mean total annual rainfall of about 2,300 mm, with a large diurnal variation in wind speeds. Mumbai has a per capita income of about Rs 31,922 (2000-01). It has always been famous as the commercial hub of the country, though it used to have a strong textile industry that has dwindled over the years. The number of motor vehicles in Mumbai is almost 1.1 million, with two-wheelers constituting approximately 45 percent . It has one of the most efficient public transport systems in India with rail-road carrying almost 70 percent of the traffic. From a urban planning perspective, Mumbai is a well- regulated city. Hyderabad: The capital of Andhra Pradesh located on the Deccan plateau in south-central India with a population of approximately 3.4 million people and an average population density of 19,930 persons per km2 is part of a metropolitan area of 5.5 million inhabitants. It has a hot steppe climate with hot summers and moderately cold winters. The mean total annual rainfall is about 900 mm, with high wind speeds persisting for large part of the year. Hyderabad has a per capita income of about Rs 10,590 (1998-99). It has developed in recent years as a significant Information Technology center in India with associated service sector becoming developed. The number of motor vehicles in Hyderabad is 1.3 million, with two-wheelers constituting approximately 75 percent. From a urban planning perspective, Hyderabad is a well-regulated city, although the old city is quite congested. Chennai: The capital of the state of Tamil Nadu located on the east coast south India by the Bay of Bengal with a population of approximately 4.2 million people and an average population density of 24,138 persons per km2is part of a metropolitan area of 6.4 million inhabitants. It has a tropical marine climate with hot summers and moderate winters. The mean total annual rainfall is about 1,300 mm, with high wind speeds persisting for large part of the year. Chennai has a per capita income of about Rs. 36,138 (1998-99). Industry plays a very important role in economic growth. It has also developed a good service sector in recent years. The number of motor vehicles in Chennai is 1.3 million, with two-wheelers constituting approximately 75 percent. Like Hyderabad and Mumbai, Chennai has also had a well regulated growth over a period of time. xix For a Breath of Fresh Air xx Executive Summary 1. Even as India experienced robust economic and study addressed four basic concerns: urban growth in 1990s, air pollution in its major (1) The efforts that have been made to address cities became a cause of national concern. The levels urban air pollution of airborne suspended particulate matter (SPM) recorded in a number of metro-cities far exceeded (2) The effects of those efforts on air quality the ambient air quality standards that India had (3) The key factors that affect air quality, and should adopted along with many other countries. Two be considered in future action programs independent analyses estimated that urban air pollution in India could be responsible for a signifi- (4) The health impacts of changes in air quality. cant burden of ill-health (Brandon and Homman Information on various interventions that could 1995, WHO 2002), primarily due to human exposure have affected urban air quality, directly or indirectly, to elevated levels of respirable suspended particulate matter (RSPM), a finer fraction of SPM. in the period 1993- 2002, was collected from various secondary sources and through interviews of stake- 2. Taking cognizance of the situation, the govern- holders knowledgeable about the developments in ment implemented a series of policy interventions. A their cities. International experience in combating air special feature of decision making with regard to pollution in other polluted cities around the world urban air pollution management in India has been a has been provided for comparison and drawing leading role by a vocal civil society and the judiciary. possible lessons. Air quality data collected by the They have successfully directed the executive branch National Environmental Engineering Research to act far more aggressively than would have been Institute (NEERI) between 1993 and 2002 were otherwise possible. Delhi, particularly, has set an analyzed to assess spatial and temporal variation, as example in this regard, undertaking a comprehensive well as effects of policy interventions. The data were and far-reaching program of measures. Of these, the compared to those from other sources. As data were most publicized is the conversion of the city's public not always "consistent" within the same data set or transport to compressed natural gas (CNG) in 2000- between different data sources, only the data trends 2002. Following the success of getting these bold corroborated by different sources, have been used to measures implemented in Delhi, the Supreme Court the extent possible, for drawing conclusions. has directed a number of other highly polluted cities in India to prepare "action plans" to address the 5. Meteorological data acquired from the Indian issue of urban air pollution, incorporating many of Meteorological Department (IMD) have been used to the measures adopted by Delhi. assess the interaction between weather variables and air quality. To better understand sources of particu- Objectives of the Study late air pollution, the air quality data have been 3. The study objective was to assess the impacts of analyzed in each city (also by land use: residential, ten years of actions and interventions in five metro- commercial, and industrial), and the findings have cities so as to enable these and other cities in India to been compared to those from another study on fine design better-informed strategies and action plans particulate source apportionment in Delhi, Kolkata, for combating urban air pollution. The report pre- and Mumbai (ESMAP 2004). The air quality data sents a retrospective analysis of urban air pollution have also been interpreted in the light of policy data with a focus on particulate air pollution from interventions implemented between 1993 and 2002 1993 to 2002 in Delhi, Kolkata, Mumbai, Hyderabad for air quality improvement. To understand the and Chennai. benefits of air quality improvement, the impact on public health of recent and potential future changes Methodology in ambient concentrations of particulate matter have 4. The methodological framework followed in this been estimated. 1 For a Breath of Fresh Air Key findings 9. This report investigates the main factors affecting air quality in each city as well as significant 6. A potentially significant and encouraging differences across the cities. Comparison of air finding of this study is that ambient concentrations quality trends with intervention measures imple- of RSPM, the main pollutant of public health con- mented, indicates that reductions in RSPM concen- cern, appeared to fall between 1993 and 2002 in all trations have been achieved through a combination the five cities. This decline in RSPM levels might of measures, targeting industry, transport, and urban have led to nearly 13,000 fewer cases of premature planning/development. Further, it is also possible deaths and much greater reductions in the number of that some developments, such as significant penetra- cases of respiratory illness in these cities on an tion of clean modern fuels in the household sector annual basis by 2002, compared to the early 1990's. over the same period, and opening up of the The levels of sulfur dioxide (SO2) also declined economy to international automobile manufacturers during the same period. However, SPM levels did not who brought in more advanced technology vehicles, fall in proportion to RSPM, implying--against the reinforced the benefits of these efforts. The table that backdrop of generally falling RSPM levels--increas- follows (p.3) summarizes by sector, the key actions ing concentrations of coarse particulate matter. This taken in the five cities that could possibly have seems at first to be surprising. However, some factors helped improve air quality. While it is not possible to have been identified which could perhaps explain say (based on the available information) which of this trend. the interventions were relatively more effective, 7. Despite substantial past progress and efforts to there seems to be some evidence linking measures curb air pollution, RSPM pollution and the associ- targeted at industrial emissions with declining RSPM ated damage to health, remain elevated in all five levels, particularly in industrial areas. However, cities. The levels of RSPM are the highest and signifi- beyond this the data analyzed do not provide evi- cantly above the national standards in the northern dence linking specific measures in any particular cities of Delhi and Kolkata, especially in winter. sector/subsector with measurable improvements in Delhi continues to record the highest levels, notwith- annual average air quality levels on a city-wide scale standing the implementation of the most extensive (although measurable improvements at the micro- program of air quality improvement. Alarmingly, level were likely, such as the impact of switching to data from the most recent years shows the reversal CNG on air quality along traffic corridors). The need of a declining trend. Ambient concentrations of to target different sources, including domestic fuel nitrogen dioxide (NO2) have exhibited an increase in use for cooking and space heating, is further sup- recent years, although still at relatively low levels. ported by the recent source apportionment exercise This calls for careful monitoring and, if the trend is for fine particulate matter (PM2.5 particles smaller not arrested, NO2 control measures may need to be than 2.5 microns in diameter), discussed in this added to the city-level air quality management report, which has shown no single dominant source programs in the future. in Delhi, Kolkata, and Mumbai (ESMAP 2004). 8. Given that there is no established threshold for 10. The nature and magnitude of emission sources the health impacts from exposure to RSPM, all cities may vary between cities, and within a city with will gain substantial health benefits from further land-use. Mumbai offers a good example of the effect reductions to or even below the current national of attrition of industry and its consequent impact by annual standard of 60 micrograms per cubic meter way of reduction in RSPM and SO2 levels in an (µg/m³) for residential areas, which seems a realistic industrial area. However, if land-use development is target for Chennai and Hyderabad in the short to mixed, as in Delhi and Kolkata, then the effect of medium term. However, bringing the RSPM levels land-use on air quality is difficult to discern. Urban down to this standard is clearly a long-term target development with mixed land-use also challenges for Delhi and Kolkata. The potential benefits, how- the current practice of having different National ever, are very large: about 10,000 lives can then be Ambient Air Quality Standards (NAAQS) within a saved every year in these two cities alone. city. Having differential standards is meaningful only if land-use regulation can be implemented 2 Executive Summary strictly. Otherwise, it could even lead to a perverse variation in monthly averages between summer/ application of the regulation, whereby densely monsoon and winter months. Further, since the populated residential areas that exist in proximity to temperature in southern cities does not drop as low an industrial area are also classified as industrial, as it does in the north during winter, there is unlikely resulting in "legal" or "permissible" exposure of a to be much domestic and space heating. As solid fuel large number of individuals to unhealthy levels of use for domestic and space heating is common in air pollution. India, it is likely that the absence of the need for heating, helps to keep RSPM levels low in Hyderabad 11. In addition to sources of air pollution, meteoro- and Chennai in winters. As such, meteorological logical conditions play an important role in influenc- factors and RSPM sources also have an interactive ing ambient air quality. Meteorological conditions, effect on ambient air quality. Cities with a clear which affect pollutant dispersion potential, differ pattern of pollution peaking in winter, like in Delhi, significantly across the five cities. This study shows may need to consider developing special programs that the southern cities of Hyderabad and Chennai to target "winter pollution". have much better dispersion characteristics all year round than Delhi, Kolkata, and Mumbai. This may A Way Forward. . . partly account for lower RSPM levels recorded in Hyderabad and Chennai, with no significant varia- 12. Clearly, apart from there being a strong case, tion in monthly averages. In contrast, Delhi, Kolkata, there is substantial scope for further improving and to some extent Mumbai, see a three- to four-fold urban air quality in India. Recently, the Supreme Sector-wise summary of key actions taken in the five cities that could have helped in improving air quality Intervention type Industry Urban Transport Clean fuels Switching to cleaner Increasing share of Use of cleaner fuels fuels (reduction in domestic and (gasoline lead Sulfur, gaseous commercial users of elimination, Sulfur alternatives) cleaner fuels (gas and reduction in liquid kerosene for cooking, fuels, use of gaseous electricity for heating) fuels) Better lubricant quality and only pre- mixed 2T oil for two- and three-wheelers Improved More efficient and Better road Scrappage of old technology cleaner combustion infrastructure (road commercial vehicles technology widening, traffic and their replacement management, with a new fleet new flyovers) Stronger and Tightened and better Enforcement of land- Introduction and better enforced enforced emissions use zoning regulations enforcement of new regulation norms leading to (closure and and more stringent installation of relocation of industry emission norms for pollution control from non-conforming new and in-use vehicles* devices areas, development of green belts/areas) * One consequence of opening up of the automobile market to international competition coupled with the introduction of increasingly tighter emission standards is a shift from two-stroke to four-stroke engines among two- and three- wheelers. According to the new vehicle sale figures, the sale of four-stroke engine two-wheelers increased from 21 percent in 1997-98 to 79 percent in 2000-04, with a corresponding decrease in the sale of two-stroke engine two- wheelers (SIAM 2004). 3 For a Breath of Fresh Air Court directed cities to formulate their own action 15. The importance of strengthening data collec- plans, and drafts of action plans have been submit- tion, management, and dissemination cannot be ted by a number of cities. Given that as per the Air over-emphasized. Information on the effectiveness (Prevention and Control) Act 1981 section 19(1), (or lack thereof) of interventions already undertaken State Pollution Control Boards (SPCBs) have the across several cities in India can guide prioritizing right to declare air pollution control area within interventions and establishing effective combina- their jurisdiction, the Central Pollution Control tions in the future. For this, a stronger framework for Board (CPCB) has recommended that many more monitoring and analyzing impacts of the interven- cities should take up the action planning exercise, tions needs to develop. It should be noted that and not wait for directives from the Supreme Court monitoring the "right" pollutants using correct to initiate it. protocols would not be informative unless accompa- nied by periodic analysis. Having such a framework 13. One of the key lessons from this analysis is that becomes all the more important in a setting where action plans, to be effective, should cut across the civil society is very active, and actions driven by urban, transport and industry sectors. Furthermore, judicial directives have to be implemented, often at international experience shows a consistent pattern short notice. across all cities with very high particulate pollution (similar to that in Delhi) of significant pollution 16. The analysis points to a number of possible reduction, achieved by targeting solid fuel use by areas that cities can start working on to address their household and small establishments, a category of urban air quality concerns. These range from pollution sources largely ignored in India thus far. strengthening monitoring and information collection This category of sources warrants greater attention to developing and adopting a common framework in future programs, especially in the northern cities for urban air quality management, to targeting a mix where heating is needed in winter. of relevant sectors (urban, transport, industry). Given the city-specific nature of the problem and the 14. Another important lesson is that implementa- solution, undertaking action planning and interven- tion of the same interventions in two cities with tions at the city rather than at the state or central different meteorological conditions and mix of government levels appears to be the best approach. sources is unlikely to produce the same results. The This observation is also supported by feedback from choice of priority interventions to control urban air various stakeholders involved in air quality issues in pollution needs to be city-specific, based on adequate India who responded to a questionnaire survey local information. One of the critical requirements conducted in parallel with this study. for formulating an effective strategy is that policymakers have access to adequate data and information. 4 Chapter 1 Introduction 1.01 In the 1990s, as India experienced robust concerned, has been the leading role played by a economic and urban growth, air pollution in its vocal civil society and the judiciary in directing the major cities became a cause of national concern and executive branch to act far more aggressively than it generated worldwide attention. The levels of air- would have otherwise done (Greenspan and others borne suspended particulate matter (SPM) recorded 2004). Delhi has set an example in this regard, by in the larger metro-cities, especially Delhi but also undertaking a comprehensive and far-reaching Kolkata and Mumbai, far exceeded the ambient air program of measures, of which the best known is the quality standards adopted by India and many other mandatory conversion of the city's public transport countries (see Figure 1. Also see Annex 4 for stan- to compressed natural gas (CNG) in 2000-2002 (see dards adopted by India). Two independent analyses Box 2 and Annex 1 for the chronology and a detailed estimated that urban air pollution in India could be description of the events in Delhi). The success with responsible for about 40,000 premature deaths having these bold measures implemented in Delhi annually (Brandon and Homman 1995, WHO 2002), has led the Supreme Court to direct a number of primarily due to human exposure to elevated levels other highly polluted cities to prepare action plans of particulate matter. Delhi was identified as the city for addressing urban air pollution, incorporating having the highest mortality figure of about 7,500 many of the measures adopted by Delhi (see Box 1). deaths per annum. 1.03 While a number of measures have already 1.02 This situation clearly required urgent action, been implemented in Indian cities in the last few and a series of policy interventions followed. A years, and new cities are being directed by the courts special feature of decision-making insofar as urban to follow suit, there is a debate among air quality air quality management (UAQM) in India is experts in India and abroad about what the exact Figure 1: Ambient Annual Averages of Suspended Particulate Matter in selected large cities in Asia and the Pacific in 1995 (UNESCAP 2000) µg/m³ 5 For a Breath of Fresh Air Box 1: Recent Supreme Court directions on Urban Air Pollution in India I). The Honorable Supreme Court in a case [writ petition(civil) No 13029 of 1985, M C Mehta Vs Union of India and others] passed an order on 5 April 2002 for formulation and implementation of action plans for control of air pollution in critically polluted cities. Critically polluted cities were identified based on CPCB's report on air quality trends in India (CPCB 2000) as those that exceeded the annual National Ambient Air Quality Standard for SPM by 50 percent or more. The following nine cities were identified by the court: Agra Faridabad Jharia Jodhpur Kanpur Lucknow Patna Pune Varanasi II). On 9 May 2002, the Honorable Supreme Court in continuation of the case mentioned above, passed an order for preparing a scheme for compulsory switchover to CNG/LPG as automotive fuel in cities which are equally or more polluted than Delhi. Based on RSPM, NO and SO annual average data of 2001 from CPCB, the following 2 2 four cities were identified: Ahmedabad Kanpur Kolkata Pune III). On 14 August 2003, the Honorable Supreme Court, in continuation of the same case above, directed the Union of India and respective states to draw an action plan for lowering the RSPM levels in the following nine cities, based on CPCB data from 2002, and submit it to the Environmental Protection Control Authority (EPCA): Ahmedabad Bangalore Chennai Hyderabad Kanpur Kolkata Lucknow Mumbai Solapur impacts of specific measures have been and just how and infirm (Schwartz 1994, Pope and Dockery 1994). much the urban air quality has improved. It is in this In India, a commonly used measure of finer fractions context that this study was initiated by the World of SPM is respirable suspended particulate matter Bank in collaboration with the Central Pollution (RSPM), which in fact refers to PM10. In this analysis Control Board (CPCB). It attempts to take stock of RSPM has been used. the progress to date, both in terms of measures taken and changes in air quality, in order to better inform 1.06 As Figure 2 shows, the concentrations of both future actions. SPM and RSPM are still high in Indian cities, so that PM is clearly the key pollutant of public health 1.04 This report presents a retrospective analysis concern.1 Gaseous pollutants--such as sulfur of urban air quality interventions and trends for five dioxide (SO2), nitrogen dioxide (NO2), carbon major metros over ten years (1993 to 2002). It at- monoxide (CO), and ozone (O3)--also cause some tempts to improve, within the limitations of the adverse health effects. However, these impacts are available data and analytical methods, the under- less significant than those of RSPM, and the concen- standing of factors influencing ambient air quality in trations of these pollutants in Indian cities have different cities so as to assist in the process of formu- normally been well below the prescribed national lating future city-level strategies and action plans for standards as well as health-based guidelines of the addressing urban air pollution concerns in India. World Health Organization (WHO). SO2 and NO2, however, contribute to secondary particulate forma- Focus on Respirable Particulate Matter tion, which can be significant in some cities. 1.05 Based on a large body of research, finer City Selection fractions of particulate matter, typically measured as PM10 (particles smaller than 10 microns in diameter) 1.07 This study covers five major cities in India-- or PM2.5 (particles smaller than 2.5 microns), are the Delhi, Mumbai, Kolkata, Chennai and major air pollutants of concern from the health Hyderabad2--all with a population of more than 5 viewpoint. The impacts on health manifest them- million inhabitants in the metropolitan area. In selves in both increased morbidity (such as acute addition to the fact that these cities represent major respiratory infections and chronic obstructive lung metros with millions of people exposed to harmful disease) and premature death, especially of the weak levels of air pollution, the selection of these cities 6 Introduction was also based on the following considerations: (a) interventions. More specifically, an attempt has been (a) these cities cover a wide range in the levels of made to address the following concerns: PM pollution (see Figure 2); (b) geographically, they cover different locations (north, south, east Efforts that have been made to address urban air and west), diverse climatic conditions, and repre- pollution sent both coastal and inland cities; (c) these cities 1.09 Information on various interventions that have relatively more extensive data available; and could have affected urban air quality, directly or (d) they represent a sample that has had policy indirectly, in the ten years between 1993 and 2002 interventions over the years, that could impact air was collected from various secondary sources as well quality. These cities are also on the list of the cities as by interviewing stakeholders knowledgeable that have been asked by the Supreme Court to about the developments in their cities. This included develop action plans to tackle PM pollution (except information on all interventions and initiatives that Delhi, which is treated as a special case on account could possibly affect air quality, particularly with of actions already taken; see Annex 1). respect to RSPM levels. Experience of other (for- merly) polluted cities around the world has also been Methodological Framework summarized for comparison. 1.08 The methodological framework employed is The effect of those efforts on air quality a variation of the commonly used diagnostic frame- work of state-pressure-impact-response, used for 1.10 Data on air quality for a period of ten years assessing the "State of the Environment" in many between 1993 and 2002 were analyzed for each of countries (EEA 1999). Given its retrospective nature, the five cities to assess the current status and the analysis starts with response as a "given" set of changes over time, in order to determine the effects interventions that have taken place, and tries to of programs that were undertaken to tackle urban air assess whether the state of ambient air quality pollution. The data included the main regulated air changed as a result of the interventions. It then tries pollutants--SPM, RSPM, SO2 and NO2--presented as to assess the health impacts stemming from the time-series, as well as by land-use. Additional (current and possibly future) changes in the state of datasets for 2000-2003 were used for consistency urban air quality. Furthermore, it presents an analysis check. It is important to stress that some of the data of the pressures that affect the state of ambient air showed erratic patterns and were not necessarily quality and that can also modify the response of consistent within the same data set or between Figure 2: Ambient Annual Averages of Suspended Particulate Matter (SPM) and Respirable Suspended Particulate Matter (RSPM) in five large Indian cities in 2002 (NEERI 2002) 7 For a Breath of Fresh Air different data sources. Hence, only the air quality economic cost of these impacts using common trends corroborated by different sources, to the valuation techniques. In order to compare health extent possible, were used for drawing conclusions. benefits from air quality improvements across cities, the benefits of 10 percent and 20 percent reductions The key factors that affect air quality and need to in RSPM pollution were calculated for each of the be considered in future action programs five cities. Further, the health benefits of city-specific 1.11 The analysis of determinants of air quality changes in RSPM between the first three years (1993- included the impacts of meteorological factors and 1995) and last three years (1999-2002) of the study different pollution sources. Meteorological data from period were estimated, as were the health benefits of 1991 to 2002 were collected from the Indian Meteo- declines in RSPM levels to meet the annual National rological Department (IMD) and analyzed to assess Ambient Air Quality Standards (NAAQS). the interaction between weather variables and air quality. A number of information sources were used Structure of the Report to improve the understanding of the relative signifi- 1.13 The report is broadly structured to follow the cance of different sources3 of air pollution, including methodological framework discussed above. Chapter a recent PM2.5 source apportionment study con- 2 presents the set of interventions that have been ducted in Delhi, Kolkata, and Mumbai (ESMAP implemented in the five cities, followed by Chapter 3 2004). These findings are used for suggesting areas on the level of air quality including analysis of for future action, that have not yet been given trends in the ten-year period between 1993 and 2002. sufficient attention. Chapter 4 presents the role of determinants or variables that affect air quality, followed by Chapter The health impacts of changes in air quality 5 which estimates health impacts. Chapter 6 gives 1.12 The changes in urban air quality in the 1993- the conclusions and recommendations to strengthen 2002 period were used to assess the health impacts in future actions. terms of incidence of illness and death, as well as the 8 Chapter 2 City-specific Interventions 2.01 All five cities implemented a significant program in 1998, which resulted in 19 flyovers being number of interventions between 1993 and 2002 to constructed by 2003, out of which nine were com- address urban air quality concerns. In some cases pleted in 2001. In addition to industry, there are these interventions were driven by the judiciary, three coal-based thermal power plants and one gas- while in other cases they were initiatives of the city based power plant in Delhi (see Annex 5A, section 5). and/or state governments. Some of the interventions From the year 2000, much more environment- initiated by the central government and/or man- friendly beneficiated coal is being used in the dated by the Supreme Court were common to a large thermal power plants. number of cities. Annex 5 gives detailed notes on each city, provides data on the growth of key sec- 2.04 The years 1999 and 2000 saw actions prima- tors--transport, industry and urban development-- rily targeted at the transport sector with introduc- and lists key interventions that have been imple- tion of pre-mixed 2T engine oil for 2-stroke engine mented and that may have had a positive impact on vehicles, further reduction in diesel sulfur to 0.25 air quality. percent, and the phase-out of commercial/transport vehicles over 12 and 15 years of age, as the key Delhi interventions with potential to effect air quality. 2.02 Amongst all cities, by far the largest number 2.05 Since 2000, the development that has domi- of programs to address urban air pollution have been nated any debate on air quality in Delhi is the implemented in Delhi. Box 2 provides the chrono- conversion of all public transport buses, taxis, and logical sequence of the key policy interventions that three-wheelers to CNG in 2000­2002. The conversion have taken place in Delhi since the early 1990s program targeted two most important sources of PM (details in Annex 5A, Box A5A.3). It can be seen that emissions in the transport sector: high annual while the majority of interventions are related to the mileage diesel vehicles and two-stroke engine transport sector, there have been significant inter- gasoline vehicles. It was in response to a Public ventions that also focused on industrial and urban Interest Litigation against the government for not sectors. controlling air pollution adequately, that the Supreme Court directed the Delhi Government to 2.03 One of the major interventions targeting convert the entire city bus fleet, autorikshaws, and industry is the closure and relocation of more than taxis from liquid fuels to CNG or other gaseous fuels. 1,300 polluting industries, classified in the hazard- Because liquefied petroleum gas (LPG) had not yet ous category in 1996-1997, including industries that been authorized as an automotive fuel at the time, emit particulate matter such as stone crushers, arc this led to the introduction of CNG starting in 1996 induction furnace units, hot mix plants, and brick when an unsuccessful attempt was made to convert kilns. The impact of industrial closures and reloca- government vehicles to CNG. By the end of 2002, all tion in 1996­1997 was seen on the annual index of public buses, taxis, and three-wheelers were running industrial production compiled by the Delhi govern- on CNG (see Annex 1 for details). ment, which showed a drop in 1996­97 (Delhi Government 2002). In addition, the 1996­1997 Kolkata period coincided with the reduction in sulfur in diesel from 1 percent to 0.5 percent, and notification 2.06 In Kolkata, significant interventions related of sulfur content in fuel oil (FO) and low sulfur to both the transport and industrial sectors (see heavy stock (LSHS) at 1.8 percent and coal at Annex 5B, Box A5B.4). The major interventions that 0.4 percent.4 On the other hand, one activity that could have affected RSPM levels began to be imple- possibly contributed to a temporary increase in PM mented in 1997. An industrial siting policy for red levels was the start of a major flyover construction (hazardous) category industries was introduced in 9 For a Breath of Fresh Air Box 2: Chronology of Key Actions Implemented in Delhi 1994-95: Transport Introduction of catalytic converters and unleaded petrol. 1996: Transport and Industry Fuel quality: 0.5% sulfur diesel introduced CNG vehicles and catalytic converters for government petrol vehicles, excluding public transport introduced (but unsuccessful) Closure of 168 hazardous industries, including stone crushers completed Approved fuel notification issued for all activities (e.g. fuel oil [FO] and lower sulfur content in coal (LSHS) with 1.8% S; coal with 0.4% S). 1997: Industry Relocation of 513 industries 337 hazardous category industries shifted (total of 1160 industries closed or relocated including hot mix plants, arc induction furnaces, brick kilns). 1998: Transport Pre-mix 2T engine oil mandated for 2-stroke engine vehicles; ban on supply of loose 2T oils Phasing out of old commercial/transport vehicles (older than 15 years) Start of Delhi Metro construction. 1999: Transport Diesel sulfur reduced to 0.25% Registration of only EURO II three-wheelers and diesel taxis permitted Restricting the plying of goods vehicles during the day Taxis more than 12 year old phased out. 2000: Transport, Industry, and Urban Bharat Stage-II (Euro-II) emission norms for all private vehicles Diesel and gasoline with 0.05% sulfur content made available in NCT (mandated for private vehicles) Replacement of all pre-1990 three-wheelers and taxis with new vehicles on clean fuels (CNG in this case) Buses more than 8 years old phased out or to ply on CNG or other clean fuel The three coal-based power plants switched over to beneficiated coal Piped natural gas by March 2000 to 1,311 domestic, 9 small, and 3 large commercial establishments. 2001: Transport, Industry, and Urban Bharat Stage-II (Euro-II) emission norms for all commercial vehicles by October The sulfur content of diesel supplied to NCT Delhi further reduced to 0.05% with effect from October Replacement of all post-1990 three-wheelers and taxis with new vehicles on clean fuels Number of CNG vehicles as follows: 14,000 three-wheelers; 2,200 taxis; 400 buses; 250 Rural Transport Vehicles (RTVs); 9,500 private (26,350 total) Piped natural gas by March to 2,821 domestic, 15 small , and 5 large commercial establishments Hazardous industry closure/relocation continues: total of 2,210 closed/relocated between 1998-2001 Construction of 9 new flyovers completed. 2002: Transport and Urban 94 CNG stations established by March All diesel buses phased out/converted to CNG. Number of CNG vehicles as follows: 35,678 three-wheelers; 4,816 taxis; 4,231 buses; 2,165 RTVs; 10,350 private (57,240 total) Piped natural gas by March to 4,111 domestic, 37 small, and 5 large commercial establishments 16,340 non-destined good vehicles turned away from entering Delhi between July and November. 10 City-specific Interventions 1997. Low smoke 2T oil for two-stroke engine gaso- Chennai line vehicles was mandated in 1998, and pre-mixed 2T oil for two-wheelers was mandated in 1999. In 2.09 As in other cities, most interventions in 1996, 0.5 percent sulfur diesel was made mandatory, Chennai related to the transport sector, including and the sulfur level was further reduced to 0.25 traffic management (see Annex 5E, Box A5E.8). The percent in 2000, and to 0.05 percent in 2001. In 2001, major interventions implemented were the lowering the government also made the use of cleaner fuels of sulfur in diesel to 0.5 percent in 1996, to 0.25 mandatory in industrial boilers, and by 2003 a large percent in 2000, and to 0.05 percent in 2001, and the number of fuel conversions for industrial boilers had mandating of pre-mixed 2T oil for two-wheelers in taken place. 1999. In January 2002, government prohibited the entry of old buses into the city center and diverted Mumbai them to a new terminal on the outskirts of the city. 2.07 The majority of interventions in Mumbai International Experience with Improving focused on the transport sector (see Annex 5C, Urban Air Quality BoxA5C.6). However, the industrial sector in Mumbai was affected by a shift in the city's economy 2.10 It is useful to compare the actions taken by towards the commercial and service sectors. The the five cities with the once highly polluted cities in decline of the textile industry in Mumbai started in other parts of the world, which have managed to 1996, which led to the closure/relocation of a large significantly improve air quality. number of textile mills and their ancillary units. The 2.11 Historically, early efforts to combat urban air sulfur content of diesel in Mumbai was reduced pollution (dating back to the middle of the 20th following the same schedule as that in Kolkata: 0.5 century) focused on the use of solid fuels (coal, percent sulfur in 1996, 0.25 percent in 2000, and 0.05 wood) or high-sulfur content heavy fuel oil in percent in 2001. The conversion of old taxis to CNG household stoves and small municipal boilers for started in 1998, but picked up only in 2002 and the heating, and on industrial sources. Switching from entire fleet of old and highly polluting taxis, as well coal to natural gas and/or developing district as old three-wheelers, was phased out or converted heating networks, successfully led to a significant to CNG/LPG by 2003. It is also worth noting that reduction in PM and SO2 air pollution in once highly- between 1995 and 2002, the number of natural gas polluted cities such as London, Prague, Krakow, consumers increased from about 100 to more than Seoul, and Beijing. Programs to address industrial 100,000 in the domestic, commercial, and industrial pollution (using zoning, relocation, and mandated sectors. However, as in Delhi, one activity that control technologies) also made a major contribution possibly contributed to a temporary increase in PM to improving air quality in Pittsburg in the United levels was the construction of 20-odd flyovers during States, Manchester in the U.K., and many other the period 1997­2000. highly industrialized cities and towns. All these cities made significant progress in reducing the levels of Hyderabad particulate matter in their air without doing much to 2.08 While the majority of interventions in deal with transport pollution, which was not consid- Hyderabad targeted the transport sector, action was ered an important source of urban air pollution until also taken against industries (see Annex 5D, Box the 1970s­1980s. A5D.7). Important interventions implemented included the reduction of sulfur in diesel to 0.5 2.12 The importance attached by policymakers to percent in 1996 and to 0.25 percent in 2000, mandat- reducing emissions from commercial, residential, ing of pre-mixed 2T oil for two-wheelers in 1999, and and institutional fixed sources in the U.K. can be actions against air polluting industries in 2000. In seen in Figure 3 which shows the historical progres- 2001 and 2002, other than a restriction on the sion of PM10 emissions. registration of three-wheelers, government actions 2.13 Against the backdrop of tightening emission focused primarily on road construction and mainte- controls for large industries and power plants and nance, and better traffic planning and management. households switching to cleaner fuels, motorization 11 For a Breath of Fresh Air Figure 3: Historical PM Emissions in the U.K. 10 Source: National Atmospheric Emissions Laboratory, U.K. grew and transport emissions became a key concern controlling emissions from diesel vehicles. Between in urban air quality management. However, many 1979 and 1983, both entry to the market and fares cities, particularly in North America, undertook were deregulated. The public sector operator was aggressive measures through the 1960s and 1970s to driven out of the market and total capacity more tackle transport emissions mainly in the context of than doubled. But by 1985, regular bus fares had ground-level ozone pollution, on account of signifi- tripled, and the average age of buses had increased cantly higher consumption of gasoline in transport from 7 to 12 years. Competition concentrated on as compared to India. Los Angeles and Mexico City routes to the center of the city, which became con- are two notable examples. In Mexico City, gasoline gested and polluted by buses with too few passen- fuel specifications were tightened much more than in gers. Initial attempts to rectify the situation included the rest of the country, and a number of measures banning 20 percent of the bus fleet from operation on targeting gasoline vehicles were implemented. This each day of the week and banning buses more than included an extensive program of vehicle emissions 22 years old. But these measures gave little relief. In inspection involving loaded dynamometer tests, the early 1990s, the government introduced a system progressive tightening of emission standards for in- of competitive tendering for franchises to operate use vehicles to drive out old-technology vehicles, buses on routes entering the city center, limiting the and fewer restrictions on the use of cleaner vehicles capacity. The fare offered was one main criterion in as an incentive to encourage the replacement of old selecting franchisees; another was the environmental vehicles with new. While these measures were taken characteristics of the vehicles offered. Congestion, to combat ozone air pollution, there are some useful air pollution, and fares all fell dramatically. By the lessons from the Mexico City experience. These mid-1990s, improved service, an important benefit of mainly relate to effectively monitoring and competition, had been retained, while the drawbacks controling emissions from in-use vehicles, such as associated with competition had been largely eradi- the decision to move from static to loaded tests for cated. In the 10-year period between 1993 and 2002, the purpose of identifying gross polluters. the number of buses fell from 14,000 to 7,500, and the average age of buses came down from 15 to 5 2.14 On the policy front, the experience of Chile in years. public road passenger transport is informative for 12 City-specific Interventions 2.15 In Asia, transport is believed to be a larger management to minimize car delays, segregated bus contributor to urban PM levels, as compared to cities systems can produce both efficiency gains and in some other parts of the world, due to a special environmental benefits. While no single set of vehicle mix: a much larger share of two- and three- solutions are applicable across Indian metro cities, it wheelers with highly polluting two-stroke engines, may be worthwhile for some city authorities to take and a much higher proportion of diesel vehicles, a closer look at the experience of the South American many old and ill-maintained, compared to the countries. Americas or the former Soviet Union republics. When two-stroke engine three-wheelers, which once 2.16 An important observation from international numbered more than 50,000 were banned in Dhaka experience is a consistent pattern across all cities in December 2002, there was an immediate measur- with very high particulate pollution (similar to that able reduction in PM10 levels. Contribution of emis- in Delhi) of significant pollution reduction achieved sions by transport are further exacerbated by road by targeting solid fuel use by household and small congestion and poor traffic management. Because establishments, a category of pollution sources particulate emissions occur predominantly during largely ignored in India thus far. Another observa- times of transient operation--during acceleration-- tion is that most successful programs to control smoothing traffic speed is a key consideration in transport emissions have focused on ozone, rather reducing PM emissions. One very effective approach than particulate pollution, so that those programs in this regard is to provide public transport priorities are not directly relevant to India (although certain in the form of dedicated lanes or totally segregated principles of how to design and maintain an effec- busways. Experience in Curitiba, Brazil, and Bogotá, tive system for monitoring and enforcement of Colombia has demonstrated that, with good traffic mandated emission limits apply). 13 For a Breath of Fresh Air Chapter 3 Urban Air Quality: Levels and Trend Analysis Air Quality Data a total of 104 days a year, for 24 hours at a time in three consecutive 8-hour intervals. The available 3.01 This study uses primarily the data collected data indicate that this requirement was seldom met by the National Environmental Engineering Re- by NEERI,6 with some cities having fewer than 50 search Institute (NEERI) between 1993 and 2002, data points a year at each monitoring site even in since it is the only data set for RSPM available over recent years. The extremely limited number of that many years. It also uses, where available, the measurements raises questions about the statistical data obtained between 2000 and 2003 at monitoring significance of the data, including the question of stations not operated by NEERI to assess their whether the few measurements made are representa- correlation with the NEERI data set.5 tive. In some cities where both the State Pollution 3.02 NEERI operates three monitoring stations in Control Board (SPCB) and NEERI measured RSPM, each of the five cities on behalf of the CPCB as part the ambient concentrations measured in the same of the network under the National Air Quality land use category area sometimes differed as much Monitoring Program (NAMP). These monitoring as two-fold. In the case of Chennai, this might have stations typically attempt to cover three types of been at least in part on account of much fewer data land use: residential, commercial, and industrial. The points collected by NEERI. There is also room for stations are meant to register air pollution concen- improving monitoring procedures. A significant trations representative of a large area rather than omission is the absence of anti-static measures-- peak concentrations at air pollution hot-spots, such static charge on filters and other surfaces is the as major traffic intersections, and are thus typically largest source of particulate measurement variability located at a significant distance from major roads. and microbalance instability. These limitations, Details of NEERI's monitoring methodology and treated in more detail in Annex 3, should be borne in city-specific monitoring setup are provided in mind when interpreting the data presented in this Annex 2. In addition to NEERI's monitoring stations, report. Annex 3A presents comparisons between city authorities and state governments are free to NEERI and SPCB/CPCB data across cities for the have their own monitoring stations. period 2000 to 2003. 3.03 As discussed earlier, PM is the main pollutant Air Quality Standards of concern in Indian cities. As industrial countries 3.05 Historically, the national ambient air quality shifted their monitoring efforts to PM10 and more standards have differed by land-use, with the most recently to PM2.5 on account of overwhelming stringent standards set for "sensitive" areas, fol- evidence pointing to adverse health impacts from lowed by "residential, rural, and other areas," and finer fractions of SPM, the CPCB started publishing the most lenient standards set for "industrial" areas. RSPM data for select cities in 1999. However, moni- The current national ambient air quality standards toring of RSPM by NEERI had been started on behalf are summarized in Annex 4. All areas in the cities are of CPCB as early as 1993 under the aegis of the classified according to these land-use categories. This NAMP and continues to date. The NEERI dataset practice carries an implicit assumption that land- presents a large and ready database to assess the use, and hence air quality, differ significantly by area urban RSPM pollution in select Indian cities. in a given city. In the most recent revision of the 3.04 A description of the monitoring network and national air quality standards in 1994, RSPM stan- quality assurance/quality control (QA/QC) proce- dards were included, and land-use categorizations dures and protocols followed by NEERI is given in were defined as follows: i) industrial, ii) residential, Annex 2. The monitoring procedures established by rural, and other use, and iii) sensitive areas.7 The CPCB stipulate that PM be sampled twice a week, or standards are most liberal for industrial areas, followed by the residential, rural and other areas, 14 Urban Air Quality: Levels and Trend Analysis and most stringent for sensitive areas. However, in national annual average particulate standards for the case of the NEERI monitoring framework, the industrial areas, which are much more lenient than land-use classifications used were i) residential, those for residential areas, Delhi is in serious non- ii) commercial, and iii) industrial, where commercial compliance with respect to SPM and RSPM, while can be considered as the "other" category used by Kolkata is in non-compliance for RSPM. CPCB. This study uses the NEERI classification of areas. 3.08 It is important to point out that while NEERI data for 2002 indicate that Hyderabad and Chennai Levels and Overall Trends are in compliance with the NAAQS for both residen- tial and industrial areas, the SPCB data for 2002 3.06 For the most part, all the five cities were not shows that industrial areas of these cities are not in in compliance with the national annual average air compliance while the residential area in Hyderabad quality standards for particulate matter during the is in compliance (in Chennai, the SPCB has no ten-year period 1993- 2002, and particulate pollution monitoring station in a residential area). One expla- remains a cause of concern for all cities. Available nation for the differences could be that the locations data show that Delhi has the highest ambient con- of the monitoring stations and the days on which the centrations of airborne particulate matter, followed data are collected by NEERI and SPCB are different. by Kolkata, and then by Mumbai. Chennai and But it is doubtful whether these alone would account Hyderabad have the lowest ambient concentrations for the difference of as much as two-fold observed in among the five cities studied. The difference in Chennai in "industrial" areas (see Annex 3A). A ambient concentrations between the most polluted closer examination of Chennai data revealed that and least polluted cities is about five-fold according NEERI carried out RSPM monitoring for less than to NEERI's data, but much smaller if the data col- half the number of days as SPCB (41 days as com- lected by the CPCB and SPCBs are compared. pared to 91 days), which raises questions about the 3.07 As mentioned above, different standards have representativeness of the data collected by NEERI. been set for different land-use areas. The degree of Given these very large differences in ambient non-compliance with the national annual average pollutant concentrations in the same land-use area particulate standards for residential area remains in the same city, it is not possible to draw conclusions most serious in Delhi and Kolkata for both SPM and about whether or not Chennai or Hyderabad might RSPM, and for SPM in Mumbai. With respect to the have met the national RSPM standards in 2002. Figure 4: Annual average SPM concentrations in the five cities Note: The lower dotted line shows the national annual ambient standards for residential, rural, and other areas of 140 µg/m³. The upper dotted line shows the national annual ambient standards for industrial areas of 360 µg/m³. The concentrations shown are averaged across three monitoring sites with varying land-use. 15 For a Breath of Fresh Air Figure 5: Annual average RSPM concentrations in the five cities Note: The lower dotted line shows the national annual ambient standards for residential, rural, and other areas of 60 mg/m³. The upper dotted line shows the national annual ambient standards for industrial areas of 120 mg/m³. The concentrations shown are averaged across three monitoring sites with varying land-use. There is a large amount of missing data, details of which are given in Annex 3; the residential site monitoring station in Delhi was moved to another location in 1994; the residential monitoring site in Kolkata was moved to another location in 1997; the residential site monitoring station in Mumbai was moved to another location in 2001; the monitoring equipment in Delhi was replaced with new equipment from a different manufacturer in between 1995 and 1997, while in Mumbai it 3.09 To examine how airborne particulate pollu- decline after reaching a maximum in 1996 and 1997, tion has changed over time in each city on a city- respectively, but SPM fell much less, if at all. While wide scale, annual concentration averages, combin- the ratio of RSPM to SPM is not static, the observed ing data from the three monitoring stations, were trend of increasing coarser particle fraction does not computed. The trends between 1993 and 2002 are have a ready explanation, and calls for more rigor- shown for SPM in Figure 4 and for RSPM in Figure 5. ous data validation and quality control. However, With the exception of Delhi, there are no indications annual average SPM levels in Delhi based on an that particulate air pollution has worsened in recent independent source of data from CPCB have shown years. SPM does not show a decline over time except a similar pattern of variation, with a marginal in Kolkata. RSPM shows a slowly declining trend decline occurring in 1997 followed by an increasing recently in Chennai, Hyderabad, and Kolkata, and a trend (CPCB 2001a). One possible reason for the lack steady decline since 1997 in Mumbai. However, the of correlation between SPM and RSPM levels, overall trends need to be interpreted with caution particularly in Delhi, could be the increase in contri- since they hide a large number of significant data bution of construction related PM (for example, 19 gaps, particularly in the case of Delhi and Kolkata odd flyovers were under construction between 1998 (see Annex 3). and 2003, and the Delhi metro has been under construction since 1998), while there was a decrease 3.10 Comparison of Figures 4 and 5 shows that in in combustion-derived RSPM. the case of Delhi, while ambient SPM concentrations showed a marginal decline between 1995 and 1997 3.11 In contrast to particulate matter, annual and began to increase in subsequent years, RSPM average SO2 concentrations were low during the concentrations fell sharply from 1996 to 1997, same time period and in compliance with the na- remained steady until 2000, and rose slightly in 2001. tional annual average standard of 60 µg/m³ for In Hyderabad and Mumbai, RSPM showed a steady residential areas. NO2 concentrations were also low 16 Urban Air Quality: Levels and Trend Analysis Figure 6: Annual average SO concentrations in the five cities 2 Note: The national ambient standard for residential areas is 60 µg/m³ for SO . The concentrations are averaged across 2 three monitoring sites with varying land-use. except in Delhi and in Kolkata. The data are shown in Kolkata and the large fluctuations in NO2 concen- in Figures 6 and 7. The sharp increase in ambient SO2 trations between 1993 and 1997 in Delhi may indi- concentrations in 1998 followed by a sharp drop in cate problems with data collection8. If these data 2000 in Kolkata is more likely to reflect problems points are excluded, Delhi, Kolkata, and Mumbai with data than a sudden increase followed by a recorded falling SO2 concentrations during this decrease in SO2 emissions in the city. Similarly, the period. No particular trend could be observed with sharp increase in NO2 concentrations in 2001­2002 respect to ambient NO2 concentrations. Figure 7: Annual average NO concentrations in the five cities 2 Note: The dotted line shows the national ambient standard for residential area of 60 µg/m³ for NO . The concentrations 2 are averaged across three monitoring sites with varying land-use. 17 For a Breath of Fresh Air 3.12 Overall, by comparison with PM, the two automotive fuel in 2000-02. One possible explanation gaseous pollutants do not seem to be a (priority) is that the lower emissions from the previously problem. However, it appears that careful monitor- diesel-fueled and two-stroke gasoline engine vehicles ing of the levels of NO2 in Delhi and Kolkata is were offset by higher emissions from elsewhere, such needed, as it is a precursor to secondary particulate as increasing overall vehicle use in Delhi, particu- formation, and also ground-level ozone, which may larly of diesel vehicles (CSE 2004), and that ambient become a problem in the future. RSPM concentrations would have worsened in the absence of the CNG conversion programs. As a City-specific Analysis result, lower particulate emissions from the con- 3.13 The observation in Figure 5 that RSPM levels verted vehicles were registered as marginally, rather do not seem to be rising, and in some cases may even than dramatically, rising ambient concentrations. be falling, suggests that targeted measures mention- 3.16 Even if city-wide averages did not show a ed in the previous chapter (and in Annex 5) may marked drop in ambient RSPM concentrations, have met with some success. It is also possible that heavy traffic intersections might be expected to this improvement in air quality is partially a "collat- show such a trend. Indeed, CPCB reported a 15 eral benefit" of actions that have been undertaken at percent reduction in RSPM levels to 180 µg/m³ at a the city level, but without the explicit objective of air busy traffic intersection (Income Tax Office crossing) quality improvement, such as implementation of in 2001 from approximately 210 µg/m³ in 2000 better town planning legislation or traffic manage- (CPCB 2001b). However, the RSPM levels at the same ment schemes to ease traffic congestion. The sensi- location increased by almost 50 percent from 2001 in tivity to developments and interventions was 2002 and then decreased by 10 percent from 2002 in assessed by superimposing them on the trend in 2003 (see Box 3). Similar observations have also been ambient RSPM concentration. While it is not possible reported by others (CSE 2003). Meteorological data to assign causality in each case, some findings give show that there were fewer calm days in 2002 than stronger evidence of linkages (or lack thereof) to in 2001. However, what is not clear is whether intervention measures than others. meteorological parameters alone can account for such a large increase in RSPM between the two years. Delhi Nor is it clear why the number of calm days would 3.14 Although there are significant data gaps and have much more impact at the Income Tax Office questions about divergent trends shown by SPM and crossing than in the rest of the city. RSPM in Delhi, it is worth asking if there are cor- roborating circumstances to support a fall in ambi- 3.17 Another possible reason for the trend in ent RSPM concentrations in Delhi between 1996 and ambient RSPM concentrations observed city-wide is 1997. When superimposed on the list of interventions that given multiple sources of air pollution in Delhi, (presented in Box 2 and Annex 5A Box A5A.3), this the impact of the measures was not large enough to decrease parallels the closure and relocation of be reflected in the ambient RSPM concentrations, polluting industries, classified in the hazardous while a likely larger impact on PM2.5 levels could not category, in 1996-1997. It also coincides with the be detected because PM2.5 is not currently being reduction in sulfur in diesel and notification of sulfur monitored. A third possibility, articulated by the content in fuel oil (FO) and low sulfur heavy stock Expert Committee on Auto Fuel Policy (GoI 2002), is (LSHS). However, it should be noted that the RSPM that the reduction in PM emissions on account of samplers were replaced with entirely new models CNG conversions was no more than 12 to 16 percent, (different manufacturer, different make) during while tighter emissions standards and better fuel 1995­1997. While cross-calibration with the prior quality were responsible for 84 to 88 percent of the samplers was carried out, instrument replacement reduction from vehicular emissions. This, however, could introduce additional data uncertainties. would not explain the large rise in ambient RSPM concentrations between 2001 and 2002 at the Income 3.15 Surprisingly, the annual ambient RSPM Tax Office crossing (see Box 3). Finally, concerns averages on a city-wide scale did not show a discern- about the quality of data discussed earlier and in ible impact of the penetration of CNG as an Annex 3 cannot be discarded. One clear message 18 Urban Air Quality: Levels and Trend Analysis Figure 8: Trend in monthly average concentration of SPM and RSPM in an industrial area of Delhi Note: The dotted line shows the time period in 1997 after which SPM started increasing again. from this review of the Delhi experience is that decline occurred in the industrial area. Figure 8 shows future programs in Delhi and other cities should the trend line fitted to monthly RSPM and SPM data make an effort to improve the air quality monitoring for the 1993­2001 time period in an industrial area program. (there was no monitoring by NEERI in the industrial area in 2002). The RSPM concentrations fell from 3.18 It is worth noting that, of the three land-use about 320 µg/m³ in 1993 to less than 180 µg/m³ in areas where air quality is measured, the maximum 1999­2001, while the SPM concentration fell from Box 3: Air Quality at a busy traffic intersection in Delhi The CPCB has been monitoring air quality at the Income Tax Office (ITO) intersection for more than four years now, almost round-the-clock. It is also considered one of the best maintained air quality monitoring stations in Delhi. As expected, the major contributor to air pollution at ITO intersection, a classic example of an urban "hot spot," is vehicular emissions. Note: Figure shows trends in RSPM and NO concentrations at ITO intersection, 2000-2003. 2 19 For a Breath of Fresh Air about 600 µg/m³ to 420 µg/m³ in 1993­1997 but in 1993, the RSPM levels fell in 2002 to 70 µg/m³-- increased again to more than 600 µg/m³ in 1997­ close to the national annual standard of 60 µg/m³ for 2001. It is quite likely that the impact of industrial residential areas. A decreasing trend is evident closures and relocations resulted in a significant between 1993 and 1996, even if the high level in 1993 drop in RSPM concentrations, while the increase in is discounted. There was a sharp increase in 1997, SPM concentrations could have been temporarily followed by a continuing decline in subsequent influenced by the large-scale (on-going) construction years. Similar trends are observed for SPM concen- activity related to nine flyovers and the Delhi metro trations except between 1998 and 2000 where SPM in the 1998-2001 time period. levels remained stable but RSPM levels steadily declined. 3.19 Comparing the NEERI data with those from CPCB showed that the SPM average from the six 3.24 Overlaying the information from the list of CPCB monitoring stations also showed a similar interventions given in Annex 5C on the RSPM pattern of variation in 1993­2001 as the NEERI data concentration data, it is likely that the large-scale (see Annex 3A). construction activities related to flyover construc- tion during 1997­2000 had an adverse impact on Kolkata RSPM and SPM. The conversion of old taxis to CNG , 3. 20 The ambient RSPM concentrations in Kolkata according to estimates made by NEERI, could have fell gradually from an average of 250 µg/m³ in 1993 contributed to a reduction in PM emissions of 385 to about 130 µg/m³ in 2002. However, as in the case tons, amounting to nearly 2.5 percent of the total PM of Delhi, because of a large amount of missing data emissions in Mumbai (NEERI 2004). (see Annex 3), the trends plotted should be inter- preted with caution. If the 1993 data are removed, 3.25 Ambient RSPM concentrations in the indus- the decline in RSPM starts after 1997. As mentioned trial area recorded the biggest reduction among the in the previous chapter (and detailed in Annex 5B), three monitoring sites, falling from about 220 µg/m³ some of the major interventions targeting the trans- in 1993 to about 75 µg/m³ in 2002. The declining port and industrial sectors that could have affected trend in Mumbai between 1993 and 2002 mirrors the RSPM levels began in 1997. changes in economic activities in the city, wherein textile mills and their ancillaries either closed down 3.21 Data show that there is a generally decreas- or moved out of the city (see Figure 9). It is also ing trend in the commercial and industrial areas worth noting that between 1995 and 2002, there was (Annex 5B). In the residential area, there was a break a large increase in the number of natural gas con- in monitoring on account of change in the location of sumers in Mumbai. the monitoring site in 1997, and higher concentra- tions were recorded subsequently. A similar trend is Hyderabad evident in the case of SPM, though it is not as pro- 3.26 RSPM concentrations exhibit a decreasing nounced. The decline in RSPM is comparable be- trend in Hyderabad from 1996 onwards (Figure 5). tween commercial and industrial. As mentioned The concentrations were more or less steady between earlier, a large number of interventions in Kolkata 1993 and 1995 and increased sharply in 1996, with have focused on the industrial sector. the monitoring site in the commercial making the greatest contribution to this rise (consistent with a 3.22 When compared with RSPM data from the large increase in SPM concentrations at the same SPCB monitoring location between 2000 and 2002 it site), as shown in Annex 5D. Increased traffic con- was found that the overall averages of NEERI and gestion near the commercial area, which had become SPCB levels correspond very well to each other a major traffic junction, could account for the (Annex 3A). sudden increase in levels. Traffic flows were rational- Mumbai ized by 2001. It is worth noting that the levels of both pollutants in the industrial area were the lowest, 3.23 Mumbai has seen a significant improvement well below the annual national ambient standard for in air quality over the years. From almost 180 µg/m³ industrial areas throughout the study period, and 20 Urban Air Quality: Levels and Trend Analysis Figure 9: Trend in monthly average concentration of RSPM in industrial area of Mumbai falling below the annual standard for residential greatest contribution (see Annex 5E). As mentioned areas in 2001 and 2002. As mentioned earlier, the earlier, the mitigation measures implemented in interventions implemented in Hyderabad targeted Chennai targeted primarily the transport sector. The primarily the transport sector. decrease in annual average RSPM to below the annual national standard for residential areas at the 3.27 However, comparison with data from the three sites in 2002 coincided with the ban on entry of SPCB industrial monitoring location (Uppal) for suburban buses into the city center. 2001 and 2002 shows that the levels at the NEERI monitoring location (Nacharam), which is not too far 3.29 It should be noted, however, that comparison from Uppal, are much lower. As per the SPCB data, with data from the SPCB industrial monitoring the industrial area recorded annual average values location for 2001 and 2002, which is located in the nearly double of the NEERI industrial monitoring same area (Thiruvottiyur) as the NEERI industrial area. The overall trends in monthly averages from monitoring station, shows that the NEERI values are the NEERI and SPCB monitoring locations, tracked much lower. The lack of consistency in data between each other reasonably well till 2002, after which SPCB and NEERI, shown in Annex 3A, may be in there seems to be a greater divergence in values part due to the number of data points obtained by (Annex 3A). NEERI being less than half of that obtained by SPCB; nevertheless, this inconsistency does not allow one Chennai to confidently conclude that residential values are 3.28 As can be seen in Figure 5, RSPM concentra- indeed as low as reported by NEERI. This highlights tions in Chennai remained steady for the most part the importance of following monitoring protocols so but there was a marked decrease in 2002, with the that proper comparison and verification of the data decrease at the commercial location making the can be made. 21 For a Breath of Fresh Air Chapter 4 Understanding the Determinants of Urban Air Quality 4.01 A number of factors affect air quality. The associated with low ambient temperature is combus- knowledge of the direction and size of all these tion of solid fuels for space heating in winter. These factors and their interaction is important for devel- observations suggest that low rainfall, low wind oping an effective mitigation strategy, particularly speed, and low ambient temperature worsen air when there is no obvious dominant source of pollu- quality. tion. These factors include: 4.02 This chapter presents the available evidence 1. Composition and characteristics of emission of the impact of these factors, based on a recent sources, including source apportionment study ­ (ESMAP 2004, the first Absolute emission levels: Everything else of its kind in India) for three of the sample cities, being equal, the higher the emissions, the higher the and the analysis of meteorological and air quality ambient concentrations. If the levels of industrial data over ten years for all five cities. and transport activities are steady throughout the year but there is additional combustion of solid fuels Evidence of Relative Contribution of in winter for space heating, absolute emission levels Sources will increase in winter, worsening air quality. 4.03 In designing a policy package to reduce air Height of emissions: The height at which pollution, one important question is the contribution emissions occur has a significant impact on ambient of different sources of emissions to the overall concentrations. Stack emissions from a very tall ambient air quality. Those sources that are contribut- chimney tend to get dispersed over a much larger ing significantly to particulate air pollution should area, thereby diluting concentrations, than emissions ideally be the primary focus of policy intervention. from vehicles, small industries, and household sources occurring near ground level. Heights less 4.04 A recent study (ESMAP 2004) carried out an than 20 meters (m) are generally considered low and analysis of ambient PM2.5 samples collected in those above 75­80 m high. March, June, October, and December in 2001 in Delhi, Kolkata, and Mumbai. The study used a Location of emissions: Emissions in the technique called chemical mass balance receptor middle of a city are likely to increase ambient con- modeling. The particles were analyzed for organic centrations inside the city more than those occurring carbon, elemental carbon, metals, and ions, and the at the edge of the city, although this depends on wind hydrocarbons found in organic carbon were further direction and speed. subjected to detailed compound identification. The 2. Meteorological parameters study represented one of the first detailed fine particulate matter source apportionment studies Rainfall "washes away" pollutants, lowering ambi- carried out in South Asia. The results indicated that ent concentrations. High winds can help disperse there was no single dominant source in Delhi, pollutants while, at the same time, winds from Kolkata, and Mumbai, but rather three principal polluting areas can increase ambient concentrations. sources of particulate air pollution: vehicle exhaust, For example, monitoring stations downstream of an re-suspended road dust, and solid fuels, especially in area with high pollution will record higher ambient cities with colder winters (see Figure 10). For ex- concentrations than those outside of such wind ample, mobile sources and biomass combustion trajectories. Thermal inversion--whereby cold air is appear to contribute substantially and in several trapped under warm air--is prevalent especially in cases approximately in equal proportions (spring winter in northern India, and increases ambient and autumn in Delhi and autumn in Mumbai). The concentrations because pollutants cannot "escape" contribution of "road dust" can also be significant from air near ground level. A human activity (summer in Delhi and spring and autumn in 22 Understanding the Determinants of Urban Air Quality Figure 10: Receptor Modeling of PM in Delhi, Kolkata, and Mumbai 2.5 Note: Insufficient sample was collected in June in Mumbai to carry out hydrocarbon speciation. Mumbai). Predictably, the combined contribution of 2002 were obtained from the Indian Meteorology biomass and coal is the highest in winter in Delhi Department. For each city, monthly rainfall and and Kolkata, presumably as a result of heating. wind speed data were averaged over the 12-year Contributions from solid fuel combustion are also period and categorized into those with good, significant in non-heating seasons: spring and moderate, and poor dispersion conditions. These autumn in Delhi and autumn in Mumbai, probably criteria for good, moderate, and poor dispersion on account of considerable use of solid fuels in small- conditions were arrived at by the authors based on scale industries and by households for cooking. existing dispersion model classifications (Lvovsky et al 2001), and following qualitative analysis of 4.05 Another study, sponsored by the Mumbai variation in RSPM levels with rainfall and wind Metropolitan Regional Development (MMRDA) speeds9 (see Annex 5 for details). under the World Bank funded Mumbai Urban Transport Project (NEERI 2004), showed that the 4.07 The analysis carried out in this study yielded contribution of vehicular exhaust to RSPM was as the following observations: high as 54 percent at kerbside locations and as low as 6 percent at an urban background location (an area Poor dispersion conditions are encountered in with little activity away from any emission hotspot). Kolkata, Mumbai and Delhi for five, four, and three Re-suspended road dust contributed 9 percent to 20 months respectively, a year. percent, while the contribution of industries varied In Hyderabad, the higher wind speed stays for from 6 percent to 42 percent, depending on the about 10 months of the year. This leads to better location (see Annex 5C for details). The overall dispersion conditions over a long period of time. results of source contribution to RSPM was in the same range as the findings of the URBAIR report for In Chennai, higher precipitation in October Greater Mumbai based on data collected in the early and November partially compensates for lower wind 1990s (World Bank 1997). speeds during these months. Meteorological Parameters 4.08 The results are summarized in Table 1. As mentioned above, Delhi, Kolkata, and Mumbai had 4.06 Monthly weather data for the period 1991- several months in winter with poor dispersion 23 For a Breath of Fresh Air Table 1: Dispersion conditions and RSPM concentrations in five cities City Dispersion Condition for Pollutants Comments Good Moderate Poor Delhi June-Sept. Jan.-May Oct. ­ Dec. Good dispersion for 4 months and poor for 3 RSPM in µg/m³ 108 213 291 months Kolkata Apr.-Oct. NA Jan.-Mar. Good dispersion for 7 Nov.-Dec. months and poor for 5 RSPM in µg/m³ 108 NA 242 months Mumbai Jan., June-Oct. Apr.-May Feb.-Mar. Good dispersion for 6 Nov.-Dec. months and poor for 4 RSPM in µg/m³ 119 102 167 months Hyderabad May-Oct. Jan.-Apr. None High average wind speed Nov.-Dec. (6-16 kmph). Good or moderate dispersion all RSPM in µg/m³ 74 87 NA round the year Chennai June-Dec. Jan. - May None High average wind speed (6-16 kmph). Good or moderate dispersion all RSPM in µg/m³ 72 71 NA the year round NA ­ not applicable. Note: Good corresponds to monthly rainfall more than 50 millimeters (mm). Moderate corresponds to rainfall more than 20mm and wind speed more than 5.4 kilometers per hour (kmph) or wind speed more than 7.2 kmph. - Poor corresponds to wind speed less than 5.4 kmph and rain less than 20mm or wind less than 3.6 kmph. Frequency of calm days could also be taken as a proxy for wind speeds. The two were found to be inversely related to each other. conditions, whereas Chennai and Hyderabad had hold. For example, if most of the rain falls on one or none. Also shown in the Table are RSPM concentra- two days in a month, monthly RSPM levels are not tions corresponding to the months in question, likely to be lowered. Similarly, a few high wind days averaged over the 10-year period covering 1993­ in a month are unlikely to lower the monthly RSPM 2002. In Delhi, Kolkata, and Mumbai, there is a level. This was seen in the case of Mumbai where strong correlation between dispersion conditions and during the month of January, the average dispersion ambient concentrations of RSPM, with highly condition can be considered good but the PM level is elevated concentrations observed in the winter high as seen in Table 1, because only a handful of months when poor dispersion conditions prevail. In rainy days manage to contribute to raising the contrast, the impact of dispersion conditions on monthly average of dispersion conditions. RSPM concentrations in Chennai and Hyderabad is small. This relationship is also reflected in the 4.10 Another important contributing factor is seasonal variation in RSPM concentrations, as shown ambient temperature. Historical data covering 1993­ in Annex 5 for all the five cities. 2002 show that temperatures in Delhi from Novem- ber to February are markedly lower than those in 4.09 However, it needs to be pointed out that the other cities (see Figure 11). Kolkata had the next inverse relationship between RSPM and wind and lowest temperature during the winter months. Cold amount of rainfall is qualitative and may not always winters in Delhi and, to a lesser extent, in Kolkata 24 Understanding the Determinants of Urban Air Quality Figure 11: Monthly average temperature based on the period 1993-2002 Figure 12: Combined (for all areas) monthly averages of RSPM, 1993-2002 lead to increased fuel usage for heating purposes. 4.11 The seasonal variation of RSPM, showing When solid fuels are combusted for domestic higher levels of RSPM in winters in Delhi, Kolkata, heating, particulate emissions increase. Biomass and and Mumbai is seen in Figure 12, which shows coal combustion indeed seem to increase in winter monthly data obtained by averaging over the 10-year months in Delhi and Kolkata (ESMAP 2004). This, period for each month. Chennai and Hyderabad, in together with poor dispersion conditions, are likely contrast, show little seasonal variation. It is worth to be key contributors to high particulate air pollu- noting that the RSPM levels are similar between tion in winter in these cities. April and September in the five cities except Delhi in 25 For a Breath of Fresh Air April and May. The higher RSPM levels in Delhi of Hyderabad and Chennai have much better during these months can be explained by the well dispersion characteristics all year round than Delhi, known phenomenon of natural dust-laden winds Kolkata, and Mumbai. This may partly account for from the desert in the neighboring state of Rajasthan lower RSPM levels recorded in Hyderabad and in the north-west.1 The monthly pattern of variation 0 Chennai. Further, since temperatures in southern seen (see Figure 12) was evident, irrespective of cities do not drop as low as in the north during land-use in all the cities, showing the strong effect of winter, there is unlikely to be much domestic- meteorological variables over any localized effect of heating. Domestic-heating commonly uses "dirty" land-use. The seasonal variation in RSPM levels solid fuel and is a significant source of air pollution. during summer, monsoon, and winters is illustrated RSPM levels are perhaps low in Hyderabad and for all cities in Annex 5. Chennai in winters as there is no need for heating. As such, meteorological factors and RSPM sources also 4.12 To summarize, meteorological conditions, have an interactive effect on ambient air quality. which affect air pollution dispersion potential, differ significantly across the five cities. The southern cities 26 Chapter 5 The Health Impacts of Air Quality Improvement 5.01 Health impacts are the main impetus for air year average concentrations for 1993­1995 and the quality improvement policies worldwide, and India reference value. is no exception. In 1995, when the first-ever estimate Scenario 2: A 10 percent reduction in RSPM from of 40,000 lives lost annually in Indian cities on the reference value. account of high particulate air pollution became available (Brandon and Homman 1995),11 it pro- Scenario 3: A 20 percent reduction in RSPM from vided the much needed stimulus to all the groups the reference value. working on clean air in India.12 Scenario 4: A reduction from reference value to 5.02 This chapter presents a quantification of the an annual average of 60 µg/m³ (NAAQS for RSPM in health benefits due to improvements in air quality in residential areas) throughout the city. the five cities studied, using annual average RSPM data obtained by NEERI from 1993 to 2002. The 5.04 Scenario 1 is meant to assess the health likely health impacts as well as economic costs of benefits that have been achieved by actions to different scenarios of improvement in RSPM levels improve air quality in the past decade, while are estimated using commonly used benefit-transfer Scenarios 2 to 4 aim to illustrate potential health techniques, as explained in Annex 6. In order to benefits from further improvement in urban air smooth out annual variations in average ambient quality in each city. Table 2 shows the reduction in concentration levels due to random fluctuations in RSPM levels to which the benefit calculations were meteorological and other factors, the scenarios were applied for each of the four scenarios. constructed using the three-year average annual concentrations of RSPM for 2000­2002 as the 5.05 The health benefits, expressed as the number reference value. of lives saved and as monetary values of both averted premature death and illnesses per annum, 5.03 The health benefits of air quality improve- are summarized in Table 3. Given data and method- ment have been calculated for the following ological limitations discussed in Annex 3 and Annex scenarios: 6, these estimates of averted deaths should be taken as leading order estimates of the magnitude of Scenario 1: The difference between the three- impacts, and not as exact numbers.13 Furthermore, Table 2: Three-year annual averages of RSPM in µg/m³ and the levels corresponding to the four scenarios Delhi Kolkata Mumbai Hyderabad Chennai Three year averages 1993­1995 255 196 142 69 73 2000­2002 180 130 83 66 63 (reference value) Reduction in RSPM for each scenario Scenario 1 75 66 59 3 10 Scenario 2 18 13 8.3 6.6 6.3 Scenario 3 36 26 16.6 13.2 12.6 Scenario 4 120 70 23 6 3 27 For a Breath of Fresh Air Table 3: Health benefits due to reduction in RSPM levels in five cities in India under four different scenarios Cities Pop. Scenario 1 Scenario 2 Scenario 3 Scenario 4 (mil.) Lives Econ Lives Econ. Lives Econ. Lives Econ saved/ value saved/ value saved/ value saved/ value year (mil.$/ year (mil. $/ year (mil.$/ year (mil.$/ year ) year) year) year) Delhi 12.8 3629 432 871 104 1742 207 5806 691 Kolkata 13.2 3293 392 649 80 1297 154 3493 416 Mumbai 11.9 5308 409 747 57 1493 115 2069 159 Hyderabad 5.5 125 10 274 21 549 42 249 19 Chennai 6.4 484 37 305 23 610 47 145 11 TOTAL 49.8 12838 1279 3091 302 6182 603 11763 1296 Notes: 1. Economic value includes the benefits of both averted premature death and illness due to reduced exposure to RSPM. Gross national income (GNI) per capita of $470 (World Bank 1994) has been used for all cities for benefit-transfer estimates of WTP for reduced risks of death and illness. 2. Population figures are taken from the 2001 census 3. Population under 15 is assumed to be 36.9 percent (UNICEF 2004) and adults to be 63.1 percent. 4. Percentage of asthmatic population is assumed to be 5 percent. 5. Crude mortality rate per 1000 is assumed to be 9 (UNICEF 2004). 6. See Annex 6 for other details on the methodology. monetary values of the health impacts reported here carried out to examine alternative scenarios. While are based on the willingness-to-pay (WTP) approach 10 percent and 20 percent reductions seem realistic as described in Annex 6, so that these values should targets for all cities over the short and medium term, be considered upper-bound estimates of the potential Scenario 4 (bringing the RSPM levels down to the health benefits. annual national standard for residential areas) is clearly a long-term target for Delhi and Kolkata 5.06 Not surprisingly, Delhi, Kolkata and Mumbai which have the highest ambient levels among the show the largest health benefits from improvements five cities, especially in winter. The potential benefits in air quality from 1993­95 to 2000­02 (Scenario 1) of lowering city-wide annual average RSPM on account of large reductions in the RSPM levels concentrations to 60 µg/m³ are very high: as many as and largest exposed populations. Mumbai shows the nearly 10,000 lives can be saved every year in these maximum number of lives saved, despite a smaller two cities, with the upper bound monetary estimates change in the RSPM levels compared to Delhi and of the health benefits exceeding US$1.0 billion per Kolkata (see Table 1, chapter 4) because a higher annum. concentration-response coefficient was used (see Annex 6 for an explanation). 5.08 The substantial health benefits that are likely to come from further air quality improvements in all 5.07 Given the substantial scope for air quality the five cities establishes a strong case for continued improvement in all the five cities and no established efforts in urban air quality management in an threshold for adverse health impacts from accelerated and more effective way. exposure to RSPM (with effects detected at levels as low as 20 µg/m³), additional calculations were 28 Chapter 6 Concluding Remarks 6.01 This study analyzed air quality data over the management programs in the coming years. period 1993­2002 in five of the largest cities in India with a view to understanding recent trends, factors 6.05 Given that there is no established threshold that influence these trends, and the associated effects for health impacts from exposure to RSPM, all cities on public health. As a retrospective study, it helps to will gain substantial health benefits from further draw lessons as also better understand repercussion reductions to or even below the current national of actions taken for urban air quality improvement. annual standard of 60 µg/m³ for residential areas, which may be achievable in Chennai and Hyderabad 6.02 The strength of this analysis to draw conclu- in the medium term. However, bringing the RSPM sions has been constrained by the limitations of the levels down to this standard is clearly a long-term available data. Nevertheless, taking these constraints target for Delhi and perhaps Kolkata. Still, the into account, a number of useful and interesting potential benefits would be very high; it may save as conclusions that can contribute significantly to the many as 10,000 lives every year in these two cities on-going policy discussions/dialogue on improving alone. urban air quality in India emerged from the study. 6.06 There is substantial scope for further improv- 6.03 A potentially significant and encouraging ing urban air quality in India. The two key questions finding of this study is that ambient concentrations are: of RSPM, the main pollutant of public health con- cern, fell between 1993 and 2002. This decline in (i) What are the most effective strategies, that are RSPM levels might have led to 13,000 fewer cases of also politically viable and socially acceptable, that premature deaths and much greater fall in the need to be taken by each city to move forward? number of cases of respiratory illness annually in (ii) What are the main factors affecting air quality in these cities by 2002, as compared to the early 1990s. each city, as well as explaining significant differences SO2 levels also declined during the same period. across cities? However, SPM levels remained steady, implying-- against a backdrop of falling RSPM levels-- This study focuses on the second underlying question increasing concentrations of coarse particulate in order to inform the first. matter, which is surprising. It would be important to confirm this trend by more systematic measurements 6.07 Comparison of air quality trends with inter- with more rigorous data validation and quality vention measures implemented indicates that assurance/control in the future. reductions in RSPM concentrations have been achieved through a combination of measures, 6.04 Despite substantial progress and efforts in the targeting industry, transport, and better urban past, RSPM pollution remains a cause for concern in planning/development. It is also possible that some all five cities. The levels of RSPM are the highest and developments not directly aimed at improving air dangerously above the national standards in the quality, reinforced the impact of these efforts. These northern cities of Delhi and Kolkata, especially in would include significant penetration of clean winter. Delhi still displays the highest levels, not- modern fuels in the household sector over the same withstanding the implementation of the most period14 and opening of the economy to international extensive program of air quality improvement automobile manufacturers, thus bringing in more measures. Ambient NO2 concentrations have exhib- advanced technology vehicles. Table 4 summarizes ited an increase in recent years, although at still by sector the key actions taken in the five cities that relatively low levels, calling for careful monitoring. possibly helped improve air quality. While it is not If the trend is not arrested, there would be a need for possible to say, based on the available information, greater NO2 control in the city-level air quality which of the interventions were relatively more 29 For a Breath of Fresh Air effective, there seems to be some evidence linking 6.10 As cities undertake the preparation of action measures targeted at industrial emissions with plans, it is important for them to learn from their declining RSPM level in industrial areas. However, own as well as international experience. One of the beyond this the data analyzed do not provide evi- key lessons from this analysis is that such plans, to be dence linking specific measures in any particular effective, should cut across the urban, transport, and sector/subsector with measurable improvements in industrial sectors. Furthermore, international experi- annual average air quality levels on a city-wide scale ence shows a consistent pattern across all cities with (although measurable improvements at the micro- very high particulate pollution (similar to that in level in the short-term were likely, such as the impact Delhi) of significant pollution reduction achieved by of switching to CNG on air quality along traffic targeting solid fuel use by households and small corridors). The need to target different sources, establishments, a category of pollution sources including solid fuel use for domestic and space largely ignored in India thus far. This category of heating, is further supported by the recent source sources warrants greater attention in future pro- apportionment for PM2.5 discussed in this report, grams, especially in northern cities where heating is which has shown no single dominant source in Delhi, needed in winter. Kolkata, and Mumbai (ESMAP 2004). 6.11 In addition to sources of air pollution, meteo- 6.08 The nature and magnitude of emission rological factors plays an important role in influenc- sources may vary between cities, and within a city ing ambient air quality. Meteorological conditions, with land-use. Mumbai offers a good example of the which affect air pollution dispersion potential, differ effect of attrition of industry and its consequent significantly across the five cities. Cities with a clear impact by way of reduction in RSPM and SO2 levels pattern of pollution peaking in winter, like in Delhi in an industrial area. However, if land-use develop- and Kolkata, may need to consider developing ment is mixed, such as in Kolkata and Delhi, then the special programs to target "winter pollution". effect of land-use on air quality is difficult to discern. Urban development with mixed land-use also 6.12 An important lesson is that implementation challenges the current practice of having different of the same interventions in two cities with different National Ambient Air Quality Standards (NAAQS) meteorological conditions and/or mix of sources is within a city. The approach to having differential unlikely to give the same results. The choice of standards would make sense only if land-use regula- priority interventions to control urban air pollution tion could be implemented strictly. Otherwise, it needs to be city-specific, based on adequate local could even lead to a perverse application of the information. One of the requirements for formulat- regulation, whereby densely populated residential ing an effective strategy is that policymakers have areas that exist in proximity to an industrial area are access to adequate data and information. also classified as industrial, resulting in "legal" 6.13. The importance of strengthening data collec- exposure of a large number of individuals to un- tion, management, and dissemination cannot be healthy levels of air pollution. over-emphasized. Information on the effectiveness 6.09 Recently, the Supreme Court directed cities to (or lack thereof) of interventions already undertaken formulate their own actions plans, and drafts of across several cities in India can guide prioritizing action plans have been submitted by a number of interventions and establishing effective combina- cities. Given that as per the Air (Prevention and tions in the future. For this, a stronger framework for Control) Act 1981 section 19(1), SPCB has the right monitoring and analyzing impacts of the interven- to declare air pollution control areas within its tions needs to develop, so that policymakers have jurisdiction, the CPCB is of the view that many more access to adequate quality data and information. It cities should take up the action planning exercise on should be noted that monitoring the "right" pollut- their own. Hence it is recommending that many ants using correct protocols would not be informa- more cities take up the preparation of action plans in tive unless accompanied by periodic analysis. order to meet air quality objectives as notified under Having such a framework becomes all the more the Air Act 1981. important in a setting where civil society is very 30 Concluding Remarks Table 4: Sector-wise summary of key actions taken in the five cities that could have helped in improving air quality Intervention type Industry Urban Transport Clean fuels Switching to cleaner Increasing share of Use of cleaner fuels fuels (reduction in domestic and (gasoline lead sulfur, gaseous commercial users of elimination, sulfur alternatives) cleaner fuels (gas and reduction in liquid kerosene for cooking, fuels, use of gaseous fuels) electricity for heating) Better lubricant quality and only pre-mixed 2T oil for two- and three- wheelers Improved More efficient and Better road Scrappage of old technology cleaner combustion infrastructure (road commercial vehicles technology widening, traffic and their replacement management, new with a new fleet flyovers) Stronger and Tightened and better Enforcement of land Introduction and better enforced emissions -use zoning regulations enforcement of new enforced norms leading to (closure and relocation of and more stringent regulation installation of pollution industry from non- emission norms for new control devices conforming areas, devel and in-use vehicles* opment of green belts/areas) * One consequence of opening up of the automobile market to international competition coupled with the introduction of increasingly tighter emission standards is a shift from two-stroke to four-stroke engines among two- and three-wheelers. According to the new vehicle sale figures, the sale of four-stroke engine two-wheelers increased from 21 percent in 1997-1998 to 79 percent in 2000-2004, with a corresponding decrease in the sale of two-stroke engine two-wheelers (SIAM 2004). active, and judicial directions have to be imple- The analysis points to a number of possible areas mented often at short notice. that cities can start working on to address their urban air quality concerns. These areas range from 6.14 Given the city-specific nature of the air pollu- strengthening monitoring and information collection tion problem and the solutions, targeting interven- to developing and adopting a common framework tions at the city-level rather than at the state or for urban air quality management to enable target- central level appears to be the best approach. This is ing a mix of relevant sectors (urban, transport, also supported by feedback from various stakehold- industry). ers involved in air quality issues in India (see Box 4). 31 For a Breath of Fresh Air Box 4: Stakeholder Feedback on the Role of the World Bank in UAQM A questionnaire survey was conducted over the internet remaining 6 percent were not sure about the answer. In (using the Clean Air Initiative ­ Asia website and response to the question on where people saw the role listserver) in parallel with this study as a means of of agencies such as the World Bank in urban air quality soliciting stakeholder perceptions on urban air quality management, the answer is shown in the figure below as well as seeking stakeholder views on the role of (all responses ranked 1 and 2 were treated as most agencies such as the World Bank in urban air quality important; 3 and 4 were treated as very important; 5 and management (UAQM). A total of 80 people replied, 6 were treated as important; 7 and 8 were treated as not representing a wide cross-section of stakeholders as important; 9 and 10 were treated as not important). (29 percent government, 24 percent private sector, 15 percent academia, 13 percent NGOs, 6 percent It is interesting to note that of those who responded to development agencies, and 13 percent others). The each category of the questions, more than 40 percent responses to questions on the role of agencies such as respondents saw knowledge sharing and assistance the World Bank in UAQM complement the findings through other sectors (such as transport, energy, urban) presented in this report. A copy of the questionnaire as the "most important," although doing stand-alone used is given in Annex 7. A separate briefing note on projects was not far behind at 38 percent; a little over 30 the findings of the survey is available via the internet percent saw capacity building as a "very important" area (www.worldbank.org/sarurbanair). of engagement; and almost 40 percent saw policy advice as an "important" area of engagement. At an An overwhelming 97.5 percent of the respondents aggregate level, about 85 percent respondents saw an thought that agencies such as the World Bank had a important (or above) role for agencies like the World role to play in addressing urban air quality concerns in Bank in doing stand-alone projects and assisting India. In response to whom agencies like the World through other sectors. In contrast, the respondents did Bank should support, 49 percent cited the city/local not see much of a role in procuring air quality monitoring administration, 27 percent cited the state government, equipment and awareness-raising. 18 percent cited the central government, and the Response to question of where people saw the role of the World Bank in UAQM 32 Annex 1 Cleaning the Air in Delhi: Chronicle of Events, 1985 ­2002 A Public Interest Writ petition was filed in 1985 In 1994, the Supreme Court passed the order that by Mr M. C. Mehta in the Supreme Court against the unleaded petrol be made available to the entire Government of India, requesting the court's inter- country by December 1996, but in selected metros, vention to ensure the fundamental right to clean air including Delhi, by April 1995. It also said that new (under Article 21 of the Constitution of India which vehicles should be equipped with Catalytic Convert- guarantees Right to Life). ers by April 1995. These orders were implemented on During the pendency of this writ, the court the said dates. passed several orders/directions to deal with the In 1996, a World Bank internal document entitled problem of vehicular pollution, but with little effect. "Cost of Inaction", estimated upto 7,500 deaths in For example, in 1989 government said it would raise Delhi on account of particulate matter air pollution. penalty on owners of polluting vehicles; in 1990 it announced vehicular emission standards for smoke. In 1996, the Center for Science and Environment (CSE) ­ a prominent NGO - brought out Slow Mur- Meanwhile, the second master plan for Delhi was der: The Deadly Story of Vehicular Pollution in India, approved in 1990. It identified hazardous industries which made the case that Delhi's air pollution was for removal from Delhi by 1993. causing severe health impacts, and argued that the One of the court orders led to formation of a fault lay in backward vehicle technology and main- committee on vehicular pollution under a retired tenance, poor fuel quality, and non-existent traffic judge (Mr Justice K. N. Saikia). This committee planning. submitted a set of recommendations to the court in In 1996, the Supreme Court issued a notice telling 1991. It also recommended the introduction of CNG the Delhi government to submit an action plan for as a vehicular fuel. controlling the city's pollution. In response the Delhi In 1991, the government appointed a and Central governments developed their first action committee under a distinguished academician plans (in 1996 and 1997 respectively). (Prof. H. B. Mathur) to evolve mass emission stan- In 1996, Supreme court issued order for compul- dards for motor vehicles for the year 1995 and 2000. sory conversion of all petrol driven government In 1992, the Ministry of Environment and Forests vehicles to CNG or to be fitted with catalytic con- brought out a document (Development and Policy verters. Only a very limited vehicles followed the Statement for Abatement of Pollution) which order. acknowledged that SPM in metro cities were higher In 1997, there was large-scale closure of hazard- than prescribed limits especially during summer ous industries in non-conforming areas as per the time. It recommended various actions vis-à-vis the second Master Plan. About 1160 industries were transport sector, including introduction of clean closed or relocated. fuels. In 1997, policy paper on control of automotive In 1993, the Government announced the first set exhaust pollution was prepared by the CPCB. It of vehicular mass emission standards for India, talked about phasing out of 15 year old vehicles. based on the Mathur Committee's recommendations, which also recommended introduction of unleaded In October 1997, after being prodded by the petrol by 1995. However, a different version of the Supreme Court, the Delhi government announced norms were notified after extending the deadline to that 15 year-old and older vehicles commercial 1996. vehicles be phased out by March 1998. But with elections round the corner the government withdrew In 1994, revised ambient air quality standards that announcement. In July 1998 the court gave a (including for RSPM) were notified. deadline of October 1998 to implement the policy, 33 For a Breath of Fresh Air which was extended to December 1998 in response to New Inter-State Bus Terminal to be built at a plea by government. entry point in north and south west to avoid In 1997, the Delhi government announced it pollution due to entry of interstate buses by 31 would introduce new vehicular emission norms in March 2000 the capital in 1998 instead of 2000. Gas Authority to expedite and expand from 9 In 1997, the central government issued its own to 80 CNG supply outlets by 31 March 2000 "White Paper on Pollution in Delhi with an Action Proper inspection and maintenance facilities Plan". to be set up for commercial vehicles with immedi- In 1998, a massive program of flyover construc- ate effect tion started. Comprehensive inspection and maintenance In 1998, physical work on the construction of the programme to be started by transport department Delhi Metro (underground) started. and private sector by 31 March 2000 The Supreme Court used the release of the White Government to set up a few more monitoring Paper as an opportunity to direct the central govern- stations and strengthen the air quality monitoring ment to establish a committee to monitor the imple- stations for monitoring critical pollutants by 1 mentation of the White Paper and to suggest other April 2000. policies to control pollution. This led to the forma- In 1998, the EPCA suggested phasing out of diesel tion of the Environmental Protection Control private vehicles. This led to the Supreme Court order Authority (EPCA) in January 1998, with membership that private non-commercial vehicles registered of government officials, one semi-government after June 1999 should adhere to EURO II norms. automobile manufacturer, and an NGO (CSE). The That led to the process of introduction of BHARAT II EPCA met once a week and followed the implemen- norms for other categories of vehicles and in other tation of the court order and the action plans. It cities, which came into force in October 2001. reported back to the court at regular intervals. The court relied heavily on the EPCA for technical In 2000, the Supreme Court ordered closure / inputs. Based on EPCA's input the following key relocation of all, except light service industry, from directions were issued by the court in July 1998 : non-conforming areas after initiation of contempt proceedings. Augmentation of public transport to 10,000 buses by 1 April 2001 In 2000, extensions were granted to the CNG program since progress was slow, citing problems Supply of only pre-mix petrol by 31 December with supply of CNG and availability of conversion 1998 for two-stroke engines of two wheelers and kits for vehicles. autos In September 2001, the Dr Mashelkar Committee Elimination of leaded petrol from National on Auto-Fuel policy was created. It recommended Capital Territory of Delhi by 1 September 1998 that the government should not mandate type of fuel Replacement of all pre-1990 autos and taxis or technology. It should decide on emission stan- with new clean vehicles on clean fuels by 31 dards. Supreme Court rejected this recommendation March 2000 and extended the CNG deadline to March 2002. No 8-year old buses to ply except on CNG or In April 2002, frustrated by delays, the court other clean fuels by 1 April 2000 ordered immediate introduction of 1500 CNG buses and replacement of 800 diesel buses per month Entire city fleet ( Delhi Transport Corporation beginning May 2002. Fines were imposed for delays. and private) to be converted to single fuel mode on CNG by 31 March 2000 By December 2002, all public transport in Delhi was converted to CNG. 34 Annex 2 Air Quality Monitoring by NEERI Table A2.1: Monitoring procedures for all pollutants Particulars Pollutants SPM PM SO NO 10 2 2 Equipment High Volume Respirable Dust Tapping in the Tapping in the Sampler Sampler Hopper Hopper Flow measuring Pressure drop across Pressure drop across Rotameters/Orifice Rotameters/Orifice device orifice in the hopper orifice in the hopper Manometers Manometers Flow rate 0.8-1.3 m3/min 0.8-1.3 m3/min 1 l/min 1 l/min Sampling period 8 hourly (round the 8 hourly (round the 4 hourly (round the 4 hourly (round the clock) clock) clock) clock) Sampling Twice a week Twice a week Twice a week Twice a week frequency Analytical method Gravimetric Gravimetric West & Gaeke Jacobs & Hochheiser Min. detection 1 µg/m³ 1 µg/m³ 0.04 µg/ml 0.03 µg/ml limit Min. reporting 10 µg/m³ 10 µg/m³ 6 µg/m³ 3 µg/m³ value Absorption NA NA 560 nm 550 nm wavelength (max) Methodology factory, chemical works, cotton mill, rubber factory, factories manufacturing paints and chemicals, and A2.01 The primary pollutants measured are SPM, foundries. Factory workers occupied residential PM10, SO2 and NO2. The sampling frequency for all areas to its southwest. A new multi-storey shopping measurements is 24 hourly, undertaken twice a week. complex and a cinema hall came up within 1 km While for SPM and PM10 it entails three sampling radius of this site. Some major industries in the area, periods of 8-hour each using two co-located moni- namely Delhi Cloth Mills and Swatantra Bharat tors, for SO2 and NO2 it entails six sampling periods Textile Mills, were closed down by order of the of 4-hour each. The standard monitoring procedures Supreme Court during 1996-97, following which for each of the pollutants is given in Table A2.1. other industries in the vicinity were also closed City-specific Monitoring Networks15 down/relocated by 2001. This station was moved to nearby Mayapuri in 2003. NEERI has three monitoring stations each in the Commercial: cities taken up in this study. They attempt to cover the three kinds of land use: industrial, commercial A2.03 The sampling station is located at a height and residential. of about 6 m on the terrace of the Ayurvedic Dispensary building in the Town Hall. This site is Delhi surrounded by the main commercial area of Industrial: Chandni Chowk. The Old Delhi Railway Station is located at a distance of about 0.4 km to its north. A2.02 This sampling station was located at a height No appreciable change has been noticed in the of about 6 m on the top of Employees State Insur- vicinity except for an increase in traffic. ance (ESI) dispensary building in Najafgarh indus- trial area till 2001. From three sides it was sur- Residential: rounded by industries including an insecticide A2.04 In November 1994, this was moved to the 35 For a Breath of Fresh Air Delhi Location Najafgarh Townhall Sarojini Nagar (Industrial) (Commercial) (Residential) Height from ground level 6 m 6 m 10 m Major sources of Industries like an Vehicular traffic No specific source pollution in the insecticide factory, (all around sampling as this is purely a vicinity cloth mill, chemical station). residential area. works, cotton mill, rubber factory, paint factories, and chemicals, and foundries which were closed down in 1996-97. terrace of the Navayug School building, Sarojini Mumbai Nagar from the post office in Netaji Nagar. This Industrial: move had become necessary due to the construction of a multi-storey building nearby. The new location is A2.08 Major changes have taken place around the Parel sampling station, which is an industrial site. surrounded by residential quarters provided by government for its staff. There is a bus depot about There were originally 13 textile mills in the Parel area 200 m away and a railway station for local trains in 1978 which were gradually closed down over the years and at present only 6 are working and that too, about half a kilometer away. An electrified railway track for goods trains is about 500 m away from this partly. There are 823 small-scale industries in the site. area, out of which 18 have smoke stacks. Two flyovers have been constructed near this site during Kolkata the last decade. Industrial: Commercial: A2.05 This site is located at Cossipore police A2.09 In the Kalbadevi area there has been no station, on Barrakpore Trunk road. There are schools, development activity during the past ten years a railway station, a thermal power plant, and indus- with the exception of the construction of three tries manufacturing hosiery, electric goods, and guns multi-storeyed buildings. and ammunition in the vicinity. Also, there are Residential: godowns and police barracks nearby. A2.10 Till 2000, the residential sampling station Commercial: was located in Bandra and then moved to Worli in A2.06 This site is located at Lalbazar police 2001. At Bandra, the height of the sampling station headquarters, Lalbazar Street, formerly known as was 5 m and there was some light vehicular Dalhousie Square. There are commercial and activity around due to nearby Bandra reclamation, residential buildings, the Dalhousie Square bus Vinoba Bhawan, and Swami Vivekanand roads. terminus, small-scale textile printing, bakery, Construction of a Marine outfall, Bandra sewage hotels, and shopping complexes around. treatment and disposal facility, and the sewage pumping station continued till 1999. Residential: A2.07 From 1992 to 1996 this site was located A2.11 The height of the Worli station is 5 m. at Mandville Garden. It was then moved to Salt The site is about 250 m from the coast and 100 m Lake where it is located at the NEERI zonal from a wastewater treatment. Construction of one laboratory, Kasba, since 1998. flyover and on-going construction of another are 36 Annex 2 : Air Quality Monitoring by NEERI Kolkata Location Cossipore Lalbazar Street Kasba (Industrial) (Commercial) (Residential) Height from ground 5 m 14 m 3 m level Major sources of School, railway station, Commercial and By-pass road, pollution in the hosiery factory building residential buildings, engineering college, vicinity to the east. small cloth printing electric industries, factory, and small bakery, and Cossipore gun & shell bakery to the east. residential buildings factory, a thermal to the east. power station, Commercial buildings Barrackpore Trunk and offices, Hoogly river, Canal, slums, and road, godowns, and traffic junction at Dalhousie residential building Hoogly river to the west. square to the west. to the west. Tala canal, Kar hospital, Bus terminus, offices and Residential plots railway yard for goods, commercial buildings, metro and building to the offices and buildings station to south. south. to south. Office building, hotels, Kolkata State Khaitan fan factory, and shops to the north. Transport Bus Rabindra Bharati University, Depot, and Indian Statistical Institute, residential building and police barracks to the north. to the north. Mumbai Location Parel Kalbadevi Bandra (Industrial ) (Commercial) (Residential) Height from ground 15 m 20 m 5 m level Major sources of Construction site of marine Textile mills (all around Mumbai harbour (east of pollution in the outfalls (temporary till year sampling station). sampling station). vicinity end of 1999). Ambedkar Road (major Road traffic (all around At Worli from 2001, arterial road adjacent to sampling station as this construction of a flyover sampling station). is a major commercial was in progress. area). No other source as this is purely a residential area. two major development activities. Two parks have of Chemical Technology pilot plant in the Nacharam been developed in the area during the last ten years. industrial area. There are non air polluting small- scale industries such as electronics, engineering, and Hyderabad leather goods. Industrial: Commercial: A2.12 This site is located at the Indian Institute A2.13 This site is located on the terrace of the 37 For a Breath of Fresh Air Hyderabad Location Nacharam Abids Tarnaka (Industrial) (Commercial) (Residential) Height from ground 10 m 10 m 8 m level Major sources of pollu- Small-scale industries Multi-storeyed buildings, Multi-storeyed buildings, tion in the such as electronics, shopping complexes, vehicular traffic on the vicinity engineering and leather increase in 2- and 3- nearby national highway. wheelers, light vehicles goods. and buses. general post office building in Abids. A number of Chennai multi-storeyed buildings have come up with Industrial: increased commercial activities. The roads around this site were also repaired. There has been a consid- A2.15 This sampling site is located at the erable increase in public and private vehicular municipal office in Thiruvottiyur. It is at a dis- traffic constituting two- and three-wheelers, light tance of 1.7 km from the centre of the city. Air vehicles, and buses. sampling is carried out on the open terrace of the municipal council's building at a height of 8-10 Residential: meters. The Manali industrial area is at a distance A2.14 This site is located on the terrace of the of 1.5-2 km on the western side of the sampling NEERI zonal laboratory building in the Indian site. The Bay of Bengal is to the east at a distance of Institute of Chemical Technology campus at about 750 km. Ennore Thermal Power House is 2.5-3 Tarnaka. A number of multi-storeyed buildings km to the north of the sampling site. The Manali and four-storey residential flats have come up industrial area consists of large chemical and around this site. There has been a substantial petrochemical industries like Madras Refineries increase in vehicular traffic around this site. Ltd., Madras Fertilizers Ltd, Manali Petrochemicals, Chennai Location Thiruvottiyur EVR Periyar Salai Santhome (Industrial) (Commercial) (Residential) Height from ground 10 m 15 m 15 m level Major sources of Manali industrial area at a Sampling site close to During 1992-98, heavy pollution in the distance of 1.5 -2 km on Chennai central station and construction activity took vicinity the western side of the Chennai Port Trust. place in the area. sampling site consisting of large chemical and petro- During 1988-99, construc- Bay of Bengal is at a chemical industries. tion of Metropolitan Rapid distance of 750 meters to Transport System (MRTS) the east and residential Ennore thermal power took place. localities are spread in the house to the north of the remaining three sides. sampling site at 2.5-3 km Construction of multi- distance. storeyed commercial and residential buildings. During 1988-98, repairs and re-laying of storm water drain and roads, opening new container godowns along the beach road had taken place. This increased movement of heavy vehicles like trucks. 38 Annex 2 : Air Quality Monitoring by NEERI Madras Petrochemicals, Kothari Industrial chemi- Parry's Corner was moved to a new place in cal, and Chlor-alkali plant. During 1988-92, the Koyambedu in the outskirts of the city. This expansion of Madras Refineries included SO2 reduced the number of trucks, lorries, three- control. During 1993-98, three to four chemical wheelers and vans plying in the Parry's Corner and petrochemical industries came up in the area that is within 1 km of the sampling site. Manali Industrial area. During 1988-98, repairs Residential: and re-laying of storm water drain and roads, and the opening of new container godowns along A2.17 This site is located at a height of about the beach road had taken place. This increased 15 meters, on the open terrace of Zoological movement of heavy vehicles and trucks. Survey of India, Santhome High Road. This road connects south Chennai to the north along the Commercial: beach road. Regular road strengthening and A2.16 This site is situated at Madras Medical repairs are observed, as it is the road for VIPs College, right in the heart of the city. Air sampling and Governor to reach the Secretariat. During is carried out on the open terrace of the 1988-92, a number of multi-storeyed buildings pharmacology building of the Madras Medical were constructed both for residential and com- College at a height of 15 meters. During 1988-99, mercial purposes. Nearby marshy land was the construction of Metropolitan Rapid Transport reclaimed by filling with municipal solid waste. System (MRTS) on an elevated bridge structure During 1992-98, there was heavy construction took place at a distance of 7.0 km as well as activity in this residential area. The Bay of construction of multi-storeyed commercial and Bengal is 750 meters to its east and residential residential buildings. During 1992-98, the main localities are spread in the remaining three sides. vegetable market in Kothwal Chavadi near 39 For a Breath of Fresh Air Annex 3 Data Quality Issues16 Data Recording stations in Delhi. (However, for the monitoring station at the Income Tax Office traffic intersection, A3.01 Ambient air quality monitoring is carried calibration is carried out every 6 months.) CPCB out manually using high volume samplers and recommends to the SPCBs to undertake calibration respirable dust sampler with gaseous attachments. at least once a year. The frequency of calibration The site operators manually record the flow rates reported by NEERI is once every 2 to 3 months. every hour and the time period of operation at the Calibration is conducted using top loading calibrator site, and note down such weather conditions as and the recording of calibration data is not uniform dust storm and rainfall in a specified format which across the different agencies. in turn is passed on to the laboratory after the sampling. The samples are analyzed in the labora- A3.05 CPCB engaged a consultant in 1996 to tory and the 4-hourly and 8-hourly concentrations carry out service, repair, and calibration of all high are recorded in the CPCB specified format and volume samplers throughout the country. A similar sent to Pollution Assessment Monitoring and exercise is being undertaken at present. Survey (PAMS) division of CPCB. All the data recording and reporting are manual up to this Data Validation point, hence adding a human interface which increases the likelihood of error in data recording. A3.06 The data are checked at CPCB where the 4-hourly and 8-hourly values are scrutinized for A3.02 The data received by CPCB are checked outliers, parametric correlation, and other items, and outliers are removed, and data are entered into taking into account data from previous years. a computer database. In case of any questions, However, there are no reference samplers used, clarifications are sought from the monitoring agen- and there is no systematic cross-calibration or cies. The values obtained through monitoring for comparison of CPCB/SPCB and NEERI samplers less than 16 hours a day are not considered for using co-located monitors in the same city. analysis. CPCB has a software for data storage and management which compiles daily average, monthly Standard Operating Manual average, annual average, standard deviation, A3.07 Recently CPCB has compiled a report titled percentage violation of National Ambient Air Guidelines for Ambient Air Quality Monitoring (CPCB Quality Standards (24-hour average), number of 2003c) which includes monitoring methods, guide- observations, and categorizes air quality in terms of lines for locating stations, and quality control and low, moderate, high, and critical levels. quality assurance in air quality monitoring. At the A3.03 This method of data recording has been in time of sanctioning new stations in the 1990s, CPCB existence for a number of years and there have not provided various monitoring agencies with terms, been major changes other than the data collection conditions, and guidelines for monitoring and format becoming a little more detailed and introduc- measurement methods. CPCB also holds periodic tion of the software. For example, since 2001, CPCB workshops on monitoring methods for the SPCBs. has started asking for qualitative information on Handling of Filter Papers weather and other local conditions (cloudy/rainy/ dusty/smoky). Weather data are obtained from the A3.08 The new guidelines (CPCB 2003c) provide Indian Meteorological Department in most cases. detailed instructions on handling filter papers. However, these guidelines were not in use during Instrument Calibration the period that is the focus of this study; instead A3.04 The frequency of calibration varies be- measurement methods were provided to the moni- tween agencies. CPCB carries out calibration once toring agencies, which included filter handling every year or two years for the ambient air quality procedures. 40 Annex 3 : Data Quality Issues A3.09 Filter papers are normally desiccated for 1998 and 1999. NEERI reports that the new 24 hours before being used. After being used they samplers were compared (and cross-calibrated) are desiccated for another 24-48 hours depending with the old samplers before being deployed in on the next shift at that monitoring station. All the field. CPCB stations have precision balances while NEERI and SPCBs normally have a single precision balance Location of Monitoring Stations and Days located at a central laboratory to which samples are of Monitoring transported (after being folded and kept in enve- A3.12 The monitoring stations are located in lopes). All the precision balances can record values areas with varying land use. While NEERI has up to 5 decimal places. The monitoring stations do three monitoring stations per city--one each for not have temperature or humidity controls. At industrial, commercial, and residential areas-- monitoring stations, temperature of absorbing CPCB/SPCBs have many more monitoring solutions for gaseous pollutants is maintained by stations also meant to cover different land-use using ice, and in the laboratory desiccators are used areas. However, in the case of Chennai, all SPCB for moisture control of PM samples. monitors are in industrial areas. A3.10 One serious omission is the absence of anti- A3.13 In cities where both CPCB/SPCB and static measures. Static charge on filters and other NEERI have monitoring stations, their locations surfaces is the largest source of particulate measure- are different, and they do not necessarily conduct ment variability and microbalance instability, with its monitoring on the same days. In addition, the adverse impact increasing with decreasing particulate monitoring stations are operated by different sets deposition on a given filter. Not adopting anti-static of operators contracted by each agency. measures is, therefore, likely to have greater adverse effects on PM data quality in lower-pollution cities A3.14 While currently all monitoring agencies than in Delhi or Kolkata, and also in summer much try to adhere to a protocol of gathering 8 samples more than in winter. Anti-static measures include: per month for both SPM and RSPM (on the same days), in the 1990s NEERI used to run SPM Dissipation of the static charge from the samplers for 4 days and RSPM samplers for a sample filter by passing the filter over an alpha- different set of 4 days. As a result, in addition to emitter source (such as Po-210 "microspot") prior to having only 4 samples for each of these pollutants, weighing. the two PM readings at that time could also not be Grounding of the operator of the micro- directly compared. Furthermore, on a handful of balance to prevent transfer of the electrical charge days when RSPM and SPM were measured at the to the PM sample filter (by using, for example, a same time, it was not unusual to find that RSPM grounding wrist strap similar to that used by concentrations were higher than SPM concentra- microelectronics technicians) tions, clearly indicating serious problems. Change in Equipment for RSPM A3.15 With respect to the number of days data Monitoring were captured, there are many months during the 10-year study period when no data are available A3.11 While the CPCB has been using the at a given monitoring stations. Important features same make of equipment since they started of missing data are highlighted below (full details RSPM monitoring in the late 1990s, NEERI are given in Table A3.1). replaced its equipment in the mid- to late-1990s in all cities (before that, due to non-availability In Delhi, data from June to August in of standardized indigenous equipment for RSPM 1993 are missing entirely at the three monitoring monitoring in the country, NEERI was using stations. Because RSPM and SPM concentrations equipment developed in-house). In Delhi, the are low during these months, these missing data instruments were replaced between 1995 and would artificially elevate the annual average 1997, while in Mumbai they were replaced in RSPM level for that year. In 1994, the residential 41 For a Breath of Fresh Air area had data from only November and December. Overall Uncertainty in Data Since RSPM concentrations are high during these months, the "annual average" would once again A3.17 There is no systematic estimation of be artificially high. In August 1996, a "low RSPM" uncertainty or error in the data. However, based month, all the three sites are missing data, on knowledge of various sources of error, the elevating the annual average. In November 2000, monitoring agencies report an estimated error of two monitoring stations are missing data, about ±20 percent in readings today. Given that lowering the annual average since RSPM tends to there have been gradual improvements in data be high in November. During the second half of collection over the years, errors in the past were 2001 and all of 2002, no data were collected at the most probably greater. industrial site. CPCB Initiatives to Ensure Data Quality In Kolkata, a large amount of data is A3.18 Recognizing the need to improve the missing between 1993 and 1997; at the residential quality of data, CPCB has initiated several site there are no data for the entire year in 1997. measures in recent years. These are: The first year in which some data are available for every month of the year in Kolkata was 1999. Monitoring stations are inspected by Zonal Offices of CPCB. Deficiencies found are In Mumbai, the residential area had no communicated to the monitoring agencies. data in 1993 except January, and data were also missing in six months out of twelve in 1994. Training programmes on ambient air quality monitoring are organized for field and In Hyderabad, the three monitoring laboratory staff of various monitoring agencies. stations missed a few months of data collection in 1995­1997. Analytical quality control exercises using Ring Test Facility are organized regularly every In Chennai, every monitoring station year at CPCB in which various monitoring agen- had some RSPM data in every month during the cies including NEERI have participated. relevant period but SPM data were not collected at the industrial site during the first seven months Review meetings are held at Zonal of 2003. Offices of CPCB in which problems and deficien- cies with the data are discussed with the monitor- A3.16 Missing data affect the results in two ing agencies. ways. Few data points make the data less repre- sentative under all circumstances. If RSPM concen- Training programmes on software for trations vary markedly between land-use areas, data management have been held for monitoring which monitoring stations are missing data agencies. CPCB has also developed an environmen- matters in the calculation of city-wide averages. tal data bank in which all the monitoring agencies In addition, in cities where ambient RSPM concen- are requested to enter data online on the CPCB trations exhibit strong seasonal dependence, the website and the data will be disseminated to public months in which data are missing can raise or on a daily basis. Some of the data checks on data lower annual averages artificially. validation are incorporated in the software. 42 Annex 3 : Data Quality Issues Table A3.1: Missing SPM and RSPM data in each city Year Months SPM Data missing for RSPM data missing for DELHI 1993 February Residential area June ­ August All areas All areas December Residential area 1994 January ­ October Residential area Residential area March Commercial area Industrial area December Industrial area 1995 November Residential area 1996 August All areas All areas 1998 January Commercial area 2000 August Industrial area September Industrial area Residential area November Residential Industrial and Residential areas 2001 May Industrial area Industrial area July ­ December Industrial area Industrial area 2002 January ­ December Industrial area Industrial area 2003 January ­ February KOLKATA 1993 April ­ May Industrial and Commercial areas Industrial and Commercial areas June ­ July All areas All areas August Industrial and Commercial areas Industrial and Commercial areas September Industrial area Residential area October All areas November Industrial and Residential areas December Industrial and Residential area 1994 January Commercial area Residential February ­ March Industrial and Commercial areas All areas April Industrial and Commercial areas Industrial and Commercial areas May Commercial area Commercial area July All areas 1995 March-April Residential October-December Residential 1996 January-February All areas Residential June All areas July Industrial August Industrial and Commercial areas September Commercial area Commercial and Industrial areas October Commercial area November Commercial and Residential areas December All areas Industrial and Residential area 1997 January All areas All areas February Industrial and Residential areas March ­ December Residential area August Industrial area Commercial and Residential areas 1998 January ­ February Residential area February All areas Industrial area March Industrial area July Industrial area Commercial area September ­October Residential area MUMBAI 1993 February - December Residential area Residential area December Commercial area 1994 January Commercial area Commercial and Residential February ­ April Commercial and Residential Residential area August ­ September Residential area Residential area HYDERABAD 1995 April ­ May Industrial area 1996 March Residential area August ­ November Commercial area Commercial area 1997 January Industrial area February Commercial area CHENNAI 1993 January ­ July Industrial area 43 For a Breath of Fresh Air Box A3.1: Structural Issues with NAMP and Implications for Data Quality In 2000, a "needs assessment" for procurement of air quality monitoring equipment for the CPCB was commissioned under the World Bank funded Environmental Management Capacity Building (EMCB) project. The report, delivered in 2002, highlighted three fundamental structural issues: There is inadequate financing for maintenance and operation of the instruments There is inadequate infrastructure for supporting analytical work There is a very serious shortage of skilled technical staff. The conclusion that followed was not that more equipment ought to be purchased, but that some of the fundamental structural issues need to be addressed as a first priority. In 2001, the World Bank commissioned The Energy Research Institute (TERI) to undertake a review of past and on- going work on UAQ in India. They reported that "there are no significant efforts towards quality control or harmonizing the data generated by different agencies involved in monitoring network." This is also acknowledged by CPCB which earlier stated that "the involvement of several agencies increase the probability of variations and personal biases reflecting on the data. Therefore the air quality statistics are indicative rather than absolute" (CPCB 2000). Further, the TERI review also reported the results of an unpublished NEERI report prepared for the UNEP/WHO, shown in the table below. Integrated mean error in pollutant monitoring City Overall Error Range (%) SPM SO NO 2 2 Bangalore +8 to ­25 +14 to ­20 +14 to ­19 Kolkata ­3 to ­14 +24 to ­37 +13 to ­28 Chennai +1 to +4 +22 to ­31 +28 to ­38 Delhi ­7 to ­12 +16 to ­27 +11 to ­22 Mumbai +17 to +26 +12 to ­45 ­3 to ­30 Source: Unpublished report by TERI for UNEP/WHO, "Review of GEMS air monitoring network in operation in certain cities of South East Asia." To quantify the uncertainties shown in the table a parallel sampling was done at the same place with centrally calibrated equipment and following a strict quality control protocol. Reference: TERI 2001 44 Annex 3A Comparison of NEERI and CPCB/SPCB Data A3A.01 In the absence of a formal quality if any set of data deviated substantially from the assurance/quality control (QA/QC) system, there patterns observed elsewhere. At its simplest, theory are only limited checks that can be performed to suggests that if a pollutant is generated at a constant assess the validity of the data collected in the past. rate and its variance is affected by dispersion and Such checks include comparison of data obtained scavenging, the pollutant concentration would be from different agencies in the same city, checking expected to form a normal distribution. If there is a inherent variability in the data, and inter-compari- large step change in the source, the distribution son of data collected at different sites in the same would be expected to be bimodal (being the sum of city. two normal distributions with widely separated medians). In the case of smaller changes, the distri- A3A.02 Comparison of RSPM data from differ- bution may be skewed. In the case of large and rapid ent agencies in the same city is the focus of this changes in the pollution levels, the observed distri- annex. For brevity, the data were averaged across butions may even be log normal. Large seasonal all monitoring stations for each agency. The results changes in the pollution mean that deviations from shown here need to be interpreted with caution the expected normal distribution could be common. since averaging across different monitoring stations Monthly averaged data can be used for this purpose, can lead to calculation artifacts. Only in two cases is and may in fact be better than daily average data the comparison presented for areas with similar because some of the short-range fluctuations due to land use but at different locations in the same city. changes in the inventory and dispersion conditions A3A.03 In addition, although not reported in are smoothed out, thereby magnifying equipment detail, this study also examined frequency distribu- performance and calibration problems. This exami- tion plots and correlations between two sites in the nation found that SO2 and NO2 data might be more same city using the data obtained in the year 2000. problematic than RSPM data. More specifically, Frequency distribution plots of data points indicate RSPM tended to give smoother distribution curves Figure A3A.1: Comparison between CPCB and NEERI RSPM data for 2000-2003 (Delhi) Note: The values represent monthly averages for three NEERI stations and six CPCB stations. No data available at the NEERI industrial site during the second half of 2001 and all of 2002. 45 For a Breath of Fresh Air Figure A3A.2: Linear regression between NEERI and CPCB RSPM data, 2000-2003 (Delhi) than SO2 and NO2 plots, some of which tended to indicating that there is moderate correlation in resemble randomly generated numbers. the dataset between the two sites. A3A.04 Data collected at different sites in the Relationship between SPM and RSPM data same city were also plotted to see the level of correla- tion. At any given time, it would be reasonable to A3A.07 Figure A3A.4 shows the comparison expect pollution levels at different monitoring between SPM measurements by NEERI and CPCB. stations to be correlated, although to varying degree The values shown are overall averages of values depending on dispersion conditions and proximity to recorded at the three NEERI stations and six CPCB major emission sources. The extent of correlation stations between 1994 and 2003. It can be observed would be indicated by R-squared for a scatter plot of that while there is much greater scatter in the case of data from two locations. NEERI data (perhaps on account of fewer data points compared to CPCB data), the overall SPM Delhi averages followed similar trends of a slight initial decrease followed by a slight increase. A3A.05 In Delhi, data from the two agencies track similar trends but with significant differ- A3A.08 The CPCB has been conducting ambient ences in magnitude in the winter months, as monitoring of RSPM at six locations in Delhi since shown in Figure A3A.1. When compared with June 2000, along with SPM, using co-located moni- the data monitored by CPCB between 2000 and tors. Figure A3A.5 shows the overall averages of SPM 2003 it was found that the overall monthly and RSPM based on data from the six monitoring averages of three NEERI stations [Najafgarh locations, operated at the same time. Both SPM and Road (I), Town Hall (C), Sarojini/Netaji Nagar RSPM show a slight decrease in mid-2001, after (R)] and six CPCB stations [Nizamuddin (R), which SPM shows a slight increase in 2002 while Ashok Vihar (R), Shahzada Bagh (I), Shahdara RSPM remains stable. (I), Janakpuri (R), Siri Fort (R)]17 correspond reasonably well. Regression analysis between A3A.09 Regression analysis between RSPM and the two datasets returned an R-squared values SPM returned an R-squared values of 0.38 (on of 0.52 (on forcing the linear regression line forcing the linear regression line through the origin)19 through the origin)18 as shown in Figure A3A.2. as shown in Figure A3A.6. The area-specific values ranged from a minimum of ­ 0.76 (in one of the A3A.06 Further, when the data from the two industrial areas) to a maximum of 0.50 (in one of the residential monitoring stations located close to each residential areas). These variations could indicate other were analyzed, regression analysis returned an that sources of RSPM and SPM differ significantly in R-squared values of 0.42 as shown in Figure A3A.3, different parts of a city. For example, impact of the 46 Annex 3A : Comparison of NEERI and CPCB/SPCB Data Figure A3A.3: Linear regression between RSPM data from NEERI and CPCB residential areas, 2000-2003 (Delhi) large number of flyovers being constructed since While the ratio decreased substantially in 2002, it 1998 (see Annex 5A Box A5A.3) on ambient again increased in 2003. concentrations at each monitoring station could vary markedly. Some of the variation could also be a Kolkata result of calculation artifacts: regression should be carried out on RSPM and SPM concentrations A3A.11 In Kolkata, the NEERI data correspond measured at the same site at the same time, but very well to SPCB data as shown in Figure A3A.8. monthly averages were used, introducing a source of Regression analysis between the two datasets re- error. Lastly, experimental artifacts cannot be ruled turned an R-squared value of 0.79 as shown in Figure out. A3A.9. A3A.10 Figure A3A.7 shows that the ratio of A3A.12 Figure 3.10 shows that the ratio of RSPM RSPM to SPM declined from June 2000, when it used to SPM stabilized from 2002 onwards at about 0.5 to be 0.6 to about 0.4 at the end of December 2003. after having fluctuated widely earlier. Figure A3A.4: Comparison between CPCB and NEERI SPM data, 1994-2003 (Delhi) 47 For a Breath of Fresh Air Figure A3A.5: SPM and RSPM monitored by CPCB 2000-2003 (Delhi) Note: The values represent monthly averages for six CPCB stations. Figure A3A.6: Linear regression between RSPM and SPM monitored by CPCB, 2000-2003 (Delhi) Figure A3A.7: Ratio of RSPM to SPM monitored by CPCB, 2000-2003 (Delhi) 48 Annex 3A : Comparison of NEERI and CPCB/SPCB Data Figure A3A.8: Comparison between NEERI and SPCB RSPM data for 1999-2002 (Kolkata) Figure A3A.9: Linear regression between NEERI and SPCB RSPM data, 1999-2002 (Kolkata) Figure A3A.10: Ratio of RSPM to SPM monitored by NEERI, 1999-2003 (Kolkata) 49 For a Breath of Fresh Air Hyderabad data from the residential and commercial sites, although close in location, differed significantly, with A3A.13 In Hyderabad, SPCB data consistently the data obtained at the industrial area monitoring showed higher values compared to NEERI data, as site being markedly skewed to lower concentrations shown in Figure A3A.11. However, the overall trends than those at the commercial site. The commercial in monthly averages from the NEERI and SPCB and residential sites, although much farther apart monitoring locations tracked each other reasonably than commercial and industrial, showed similar well till 2002 after which there seems to be a greater frequency distribution patterns. These observations divergence in values. Regression analysis between suggest that it may be possible that there are the two datasets returned an R-squared value of 0.19 common sources of emissions in the residential and as shown in Figure A3A.12. commercial areas (for example, vehicular traffic). A3A.14 Frequency distribution plots of data A3A.15 Figure A3A.13 shows that the ratio of obtained by NEERI at its three sites showed that the Figure A3A.11: Comparison between NEERI and SPCB RSPM data for 2001-2003 (Hyderabad) Figure A3A.12: Linear regression between NEERI and SPCB RSPM data, 2001-2003 (Hyderabad) 50 Annex 3A : Comparison of NEERI and CPCB/SPCB Data Figure A3A.13 : Ratio of RSPM to SPM monitored by NEERI, 2001-2003 (Hyderabad) Figure A3A.14: Linear regression between RSPM data from NEERI and SPCB residential areas, 2001-2003 (Hyderabad) Figure A3A.15: Comparison between NEERI and SPCB RSPM data for 2001-2003 (Chennai) 51 For a Breath of Fresh Air Figure A3A.16: Linear regression between NEERI and SPCB RSPM data, 2001-2003 (Chennai) Figure A3A.17 : Ratio of RSPM to SPM monitored by NEERI, 2001-2003 (Chennai) RSPM to SPM declined from about 0.5 in 2001-2002 Chennai to about 0.3 in 2003. A3A.17 The data from Chennai exhibit almost A3A.16 Further, when data from the two residen- divergent trends at times between SPCB and NEERI tial monitoring stations (located far from each other) as shown in Figure A3A.15. The lack of agreement were analyzed, regression analysis returned an was evident even when data from two closely R-squared values of 0.16 as shown in Figure A3A.14, located monitoring stations in an industrial area indicating that there is poor correlation in the (Thiruvottiyur) were compared. Regression analysis dataset between the two sites. Given Hyderabad between the two datasets returned an R-squared seems to have good dispersion conditions, this level value of -0.18 as shown in Figure A3A.16. The lack of of discrepancy in the data from two agencies may be consistency in data between SPCB and NEERI may on account of experimental problems with RSPM be in part due to the number of data points obtained monitoring or differences in monitoring protocols, by NEERI being less than half of that obtained by such as the number of days monitored. SPCB. This highlights the importance of following 52 Annex 3A : Comparison of NEERI and CPCB/SPCB Data Figure A3A.18: SPM and RSPM monitored by NEERI 2000-2003 (Mumbai) Figure A3A.19: Ratio of RSPM to SPM monitored by NEERI, 2000-2003 (Mumbai) monitoring protocols so that proper data compari- A3A.20 Figure A3A.19 shows that unlike other son can be made. cities there seems to be a seasonal trend in the ratio of RSPM to SPM, with the ratio between 0.4 and 0.5 A3A.18 Figure A3A.17 shows that there is a in winter months and close to 0.2 in the summer and decreasing trend in the ratio of RSPM to SPM over monsoon months. Overall, there appears to be a the years. While in 2001 it was close to 0.5, it slightly decreasing trend in the ratio of RSPM to decreased to 0.3 by the end of 2003. SPM over the years. Mumbai General Observations A3A.19 In Mumbai all monitoring is undertaken A3A.21 The higher R-squared values between by NEERI. Figure A3A.18 shows the overall aver- NEERI and SPCB data in Delhi and Kolkata ages of SPM and RSPM based on data from the three strengthen the notion about sources being more monitoring locations. The seasonal variation in uniformly dispersed across these two cities on levels of both pollutants is evident from the figure. account of mixed land-use, as compared to 53 For a Breath of Fresh Air Hyderabad where land-use regulation is better explanation for the observation that, even though implemented and hence mixed-use is less common the monitoring locations are not the same, there is (unfortunately a similar comparison can be made better correlation between the data from two with Chennai since all SPCB monitoring stations in different agencies in Delhi and Kolkata than Chennai are in industrial areas). This is one possible Hyderabad. 54 Annex 4 National Ambient Air Quality Standards National Ambient Air Quality Standards (NAAQS), 1994 Concentration in Ambient Air Pollutant Time Sensitive Industrial Residential Method of Weighted Area Area Rural & Measurement Other Areas Sulfur Annual* 15 µg/m3 80 µg/m3 60 µg/m3 Improved West and Gaeke Method Dioxide (SO ) 24 hours** 30 µg/m3 120 µg/m3 80 µg/m3 Ultraviolet Flurescence 2 Oxides of Annual*` 15 µg/m3 80 µg/m3 60 µg/m3 Jacob & Hochheiser Modified Nitrogen (Na-Arsenite) Method as NO 24 hours** 30 µg/m3 120 µg/m3 80 µg/m3 Gas Phase Chemiluminescence 2 Suspended Annual* 70 µg/m3 360 µg/m3 140 µg/m3 Particulate Minute) High Volume Sampling, (Average Matter Flow Rate not Less Than 1.1m3/minute (SPM) 24 hours** 100 µg/m3 500 µg/m3 200 µg/m3 Respirable Annual* 50 µg/m3 120 µg/m3 60 µg/m3 Particulate Matter Respirable Particulate Matter Sampler (RPM) Size Less Than 10 µm 24 hours** 75 µg/m3 150 µg/m3 100 µg/m3 Lead (Pb) Annual* 0.50 µg/m3 1.0 µg/m3 0.75 µg/m3 AAS Method After Sampling Using EPM 2000 or Equivalent Filter Paper 24 hours** 0.75 µg/m3 1.5 µg/m3 1.00 µg/m3 Carbon 8 hours** 1.0 mg/m3 5.0 mg/m3 2.0 mg/m3 Non-Dispersive Infrared Spectroscopy Monoxide (CO) 1 hour 2.0 mg/m3 10.0 mg/m3 4.0 mg/m3 Note : * Annual arithmetic mean of minimum 104 measurements in a year taken twice a week 24 hourly at uniform intervals. ** 24-hourly/8-hourly values should be met 98% of the time in a year. However, 2% of the time, it may exceed but not on two consecutive days. Source : Air Quality Status and Trends in India. CPCB, 1994 55 For a Breath of Fresh Air Annex 5A Delhi 1. Geographical Location west and south and by Uttar Pradesh on the east, A5A.01 Delhi, the National Capital of the country, separated by river Yamuna flowing in a north-south is surrounded by the state of Haryana on the north, direction. The growth of the city has been concen- trated between the river Yamuna in the east and the Figure A5A.1: Approximate location of NEERI monitoring stations in industrial, commercial and residential areas. 56 Annex 5A : Delhi Figure A5A.2: Population growth in NCT of Delhi between the last two census periods (Pandey 2003) Aravalli hills in the west. The area of National Capital metropolitan cities in India. As per 2001 census, the Territory (NCT) of Delhi is 1,397 square kilometers population of the NCT area was 12.8 million com- (km2). The urban area in 2001 was 891.1 km2 pared to 8.4 million in 1991. This shows a decennial compared to 685.3 km2in 1991 (GoNCTD 2002). growth rate of 51.9 percent between 1991 and 2001. Figure A5A.1 shows approximate location of the The annual growth rate of population of Delhi pollution monitoring stations. during 1991-2001 was 4.18 percent, almost double the national average (Maps of India 2003). The 2. Demography average population density was 7,021 persons per A5A.02 Delhi is one of the most rapidly growing km2in 2001. Figure A5A.2 shows the annual popula- tion growth rate in Delhi. Table A5A.1: Monthly average values of key meteorological parameters for 1991-2002 Mean Mean Mean No. of Relative Total Max. Min. Wind Calm Days Humidity % Rainfall Temp. Temp. Speed (mm) (0C) (0C) (kmph) 0830 hrs 1730 hrs 0830 hr 1730 hrs January 20.4 7.4 6.8 11 4 85 52 24 February 23.8 10.2 7.3 8 2 78 42 22 March 29.4 15.2 8.2 6 1 64 35 11 April 36.1 20.9 8.0 3 1 45 22 10 May 40.0 26.0 9.1 3 3 43 25 37 June 38.9 27.8 9.0 3 3 58 43 94 July 35.6 27.6 8.5 3 3 74 62 150 August 33.8 26.6 7.8 5 4 79 68 240 September 34.1 25.0 6.9 5 4 75 59 122 October 32.7 19.1 3.8 14 10 70 47 14 November 28.2 12.7 4.0 15 17 74 50 5 December 22.9 8.1 4.8 15 16 81 57 10 Total/year 91 66 740 Note: Data on calm days and relative humidity are collected twice a day. 57 For a Breath of Fresh Air Figure A5A.3: Growth in commercial motor vehicles in Delhi (MoRTH 2003) Figure A5A.4: Growth in private motor vehicles in Delhi (MoRTH 2003) 3. Climate Wind speeds are typically higher in the summer and monsoon periods; in winter, calm conditions are A5A.03 The weather profile for Delhi is shown in frequent. Table A5A.1, based on 12 years of data. The city experiences a tropical semi-arid climate. It is charac- 4. Economy terized by hot and dry summer (February to June), moderate monsoon (July to September), and a cool A5A.04 The annual per capita income of Delhi winter (October to January). The mean monthly was Rs 24,450 in 2000-01, more than twice the temperature ranges from 7.4°C in winters to 40°C in national average. The gross state domestic product summer months. Dust storms occur frequently (GSDP) was Rs 366.85 billion in 2000-01. An analysis during summer months, leading to build-up of of the sectoral composition of GSDP shows that the particulate matter in the atmosphere. The heavy share of the primary sector20 had gone down signifi- monsoon rains following summer months act as a cantly from 3.9 percent during 1993-94 to 1.4 percent "scrubber" that removes the particulate matter. during 2000-01. In contrast, the tertiary sector, 58 Annex 5A : Delhi Box A5A.1 : Supreme Court Directions for Industrial Closures (i) Hazardous/Noxious heavy and large industries it is permitted under MPD-2001. The Court further The Supreme Court vide its order dated 8/7/1996 directed the Delhi Government to render all possible directed that 168 industries falling in hazardous assistance to the concerned brick kiln owners to categories under the Master Plan of Delhi ­ 2001 changeover to the new technology and in the setting (MPD-2001), should stop functioning in the city of Delhi up of the modern plants with flyash- sand-lime from 30/11/1996. However, those industries could technology. relocate/shift themselves to any other industrial estate (v) Arc/Induction Furnaces in the National Capital Region (NCR) or outside of it. The Supreme Court vide its order dated 26/11/1996 (ii) The Supreme Court vide its order dated 6/9/1996 directed that the 21 arc/induction furnaces falling ordered that 513 industries falling under hazardous `H' under hazardous category under the MPD-2001 category under the MPD-2001, should stop functioning close down and stop operating from 31/3/1997 in and operating in the city of Delhi from 31/1/1997. Delhi. However, these arc/induction furnaces could However, those industries could relocate/shift them- relocate/shift themselves to any other industrial selves to any other industrial estate in NCR. estate in the NCR. (iii) Hot Mix Plants (vi) The Supreme Court vide its order dated The Supreme Court vide its order dated 10/10/1996 19/12/1996 directed that 337 industries falling under directed that 43 Hot Mix Plants operating in Delhi close hazardous category under the MPD­2001 close down and stop operating in Delhi from 28/2/1997, and down and stop operating from 30/6/1997 in Delhi. be relocated to any other industrial estate in the NCR. However, those industries could relocate/shift themselves to any other industrial estate in the NCR. (iv) Brick Kilns (vii) All industries located in residential area The Supreme Court vide its order dated 26/11/1996 (except light service industries) directed that 246 brick kilns operating in the Union Territory of Delhi falling under hazardous category under The Supreme Court vide its order dated 12/9/2000 the MPD­2001, should close down and stop function- directed and appointed the Ministry of Urban ing from 30/6/1997 in Delhi and relocate/shift within Development to act as the Nodal Agency for the NCR. The Supreme Court further directed the brick matter of relocating/shifting of industries as per kiln owners to indicate before 31/1/1997 in writing to MPD-2001 functioning and operating in residential the Delhi Government and Delhi Pollution Control areas of Delhi. The court on 7/12/2000 directed that Committee that the concerned brick kilns intended to under the supervision of the Nodal Agency, the shift to the new technology of manufacturing bricks by Government of National Capital Territory of Delhi, the using flyash-sand-lime technology. The Delhi Pollution Municipal Corporation of Delhi and the Delhi Devel- Control Committee was directed to monitor the setting opment Authority would close all the polluting units up of the new project of the concerned brick kilns. After functioning in non-conforming/residential areas or obtaining the consent and no objection certificate from zones within a period of four weeks from the date of the Delhi Pollution Control Committee and also from the the order. Central Pollution Control Board, the concerned brick kilns would be permitted to operate at the same site, if References: CPCB 2003a, GoNCTD 2003. Note : Dates to be read as day/month/year. which contributed 70.5 percent to state economy Delhi. The manufacturing sector provided employ- during 1993-94, increased its share to 78.4 percent ment to 1.44 million people. The Master Plan of Delhi during 2000-01. The share of the secondary sector in prohibits setting up of large and heavy industries. the GSDP of Delhi declined from 25.6 percent in There is, however, scope for expanding small scale 1993-94 to 20.2 percent during 2000-01 due to industries. As a result, industrial growth is mainly in closure of polluting industrial units on environmen- the small-scale sector. tal grounds (GoNCTD 2002). A5A.06 The Delhi government issued an indus- 5. Industry trial policy statement in 1982, which emphasized the promotion of sophisticated industries that could A5A.05 The economic census of 1998 indicates achieve optimum levels of production using less that there were 129,363 manufacturing enterprises in space and power, while generating employment. 59 For a Breath of Fresh Air Household industries were permitted and 73 types Delhi Government in 2000, out of a total of 125,000 of industries were identified that could be run in regis-tered industries, 98,000 industries were located residential houses with a maximum powerload of 5 in non-conforming areas in violation of the Master kilo-watts (kW), provided that the industry did not Plan of Delhi. Industries had been set up in unautho- cause pollution or congestion and could be operated rized colonies, urban villages, resettlement colonies, within a space of 30 square meters. However, this the walled city, and other residential pockets (TERI policy led to a proliferation of industries in non- 2002). As a result of a writ petition filed in the conforming areas. According to a survey by the Supreme Court in 1985 (M. C. Mehta versus Figure A5A.5: Trend in RSPM levels in three areas of Delhi. Note: The dotted line shows the annual national ambient standard of 60 µg/m³ for RSPM for residential areas. Figure A5A.6: Trend in SPM levels in three areas of Delhi Note: The dotted line shows the annual national ambient standard of 140µg/m³ for SPM for residential areas. There is a large amount of missing data. Examples include no data from June to August in 1993 at the three monitoring sites, no data from January to October 1994 at the residential monitoring site, no data in August 1996 at the three monitoring sites, and no data available at the industrial site in the second half of 2001 and all of 2002; the residential site monitoring station was moved to another location in 1994; the monitoring equipment was replaced with new equipment from a different manufacturer in between 1995 and 1997. 60 Annex 5A : Delhi Figure A5A.7: Trend in overall average RSPM concentration in Delhi Government of India), and in July 1996, the 6. Transport Supreme Court ordered the closure and shifting of industrial units in Delhi that were considered to be A5A.09 Delhi is still predominantly dependent on polluting and/or operating in non-conforming areas road transport, though the metro started functioning in violation of Delhi Master Plan. The index of on a limited stretch in 2002. Buses constitute only industrial production also showed the effect of 1.2 percent of the total vehicle population, but cater industrial closures and relocation. Box A5A.1 details to about 60 percent of the total travel demand while the Supreme Court orders for industrial closures, personal vehicles account for about 30 percent. which are also discussed in Box A5A.3. Motorcycles and scooters comprise about 65 percent of the total vehicle population, while cars and jeeps A5A.07 Delhi has 28 industrial estates where account for almost 27 percent. In addition, registered about 25,000 industrial units are located. In human- and animal-driven vehicles constitute addition, approximately 26,000 industries that almost 5 percent of the total vehicle population, out were earlier permitted in residential/non- of which about 52 percent are cycle rickshaws. The conforming areas are being rehabilitated and number of cars and jeeps increased by about 5 relocated in the conforming use zones. Government percent per year between 1991 and 2001 (MoRTH has acquired approximately 1,865 acres of land for 2003). development of industrial estates, where infra- structure development is on-going (GoNCTD A5A.10 Figures A5A.3 and A5A.4 show the 2002). increase in vehicle population in Delhi between 1991 and 2002. While trucks and lorries make up the A5A.08 In addition to industry, there are three majority of (commercial) transport vehicles, two- coal-based thermal power plants (at Indraprastha wheelers dominate non-transport (private) vehicles. and Rajghat in central Delhi, and Badarpur in the south), and one gas-based power plant (at A5A.11 Delhi's registered vehicular population Indraprastha) in Delhi, which have a combined nearly doubled to about 4 million between the two generation capacity of about 1,300 MW. As per data census periods of 1991 and 2001, registering a for the year 2000, almost 5,000 metric tons (MT) of growth rate of approximately 200,000 vehicles per coal was consumed per day in Delhi at these power annum. In terms of per capita vehicle numbers this plants, in addition to small quantities of furnace translates to about 250 vehicles for every 1000 oil/light diesel oil, in progressively decreasing residents in 2001 compared to the corresponding quantities, at two of the power plants. figure of about 190 in 1991. The percentage distribu- tion of categories of motor vehicles shows that there 61 For a Breath of Fresh Air Figure A5A.8: Trend in SO levels in Delhi 2 Note: The annual national ambient standard for SO in residential areas is 60 µg/m³. 2 Figure A5A.9: Trend in NO levels in Delhi 2 Note: The annual national ambient standard for NO in residential areas is 60 µg/m³ 2 has been a rapid rise in the number of cars and jeeps average age of the vehicle fleet is also much younger in recent years, while the increase in relative share of than the registration data suggests (GoI 2002a). two-wheelers has shown a slightly declining trend A5A.13 Delhi had 1749 km of road length per 100 (GoNCTD 2002). It is also worth noting, though not km2area compared to national average of 73 km per evident from Figure A5A.4, that while the share of 100 km2in 1995-96. Since 1971-72, the road network petrol cars in all cars registered in Delhi stagnated, has increased three-fold while the number of the share of diesel cars increased from 4 percent in vehicles has increased sixteen-fold, and the result is 1998-1999 to 16 percent in 2002-2003 (CSE 2004). evident on the roads of Delhi. One of the reasons for A5A.12 However, there is controversy about the Delhi's traffic problem is that it is the epicenter for number of vehicles on Delhi roads since a large wholesale trade in north India. It is estimated that 78 number of vehicles are from neighboring states, as is percent of the vegetables and fruits, 49 percent of obvious from their registration records and number fuel, 44 percent of iron and steel, and 47 percent of plates. Further, a recent survey indicates that the foodgrains traded in Delhi are destined for other 62 Annex 5A : Delhi Figure A5A.10: Trend in RSPM monthly concentration at the residential, commercial and industrial area monitoring location (there was no monitoring at the industrial area in 2002) 63 For a Breath of Fresh Air Figure A5A.11: Monthly average total rainfall and mean wind speed Figure A5A.12: Monthly average total rainfall and RSPM concentration Figure A5A.13: Monthly mean wind speed and RSPM concentration 64 Annex 5A : Delhi Figure A5A.14: Variation in monthly averages of SO and NO 2 2 states. The convergence of five national highways, government vehicles to CNG. By the end of 2002, all while facilitating transport, aggravates the traffic public buses, taxis, and three-wheelers were running congestion in the city (GoNCTD 2002). onCNG. A5A.14 In response to a Public Interest Litigation A5A.15 In addition to addressing the ever- against the government for not controlling air growing traffic problem, the city government is pollution adequately, the Supreme Court directed the constructing a large number of flyovers, introducing Delhi Government to convert the entire city bus high capacity buses on select corridors, speeding up fleet, autos and taxis from liquid fuel to CNG or the introduction of additional lines on the metro, and other gaseous fuels. Because LPG had not yet been discussing the need for a unified metropolitan authorized as an automotive fuel at the time, this led transport authority. to the introduction of CNG starting in 1996 when an unsuccessful attempt was made to convert Figure A5A.15: Overall variation in RSPM with seasons (Monsoon: July to September; Summer: February to June; Winter: October to January) 65 For a Breath of Fresh Air 7. Urban administrative ambit of Delhi government. This has had effects on the implementation of politically A5A.16 The city of New Delhi shows a strong unpopular decisions. influence of its colonial past. It has broad tree-lined boulevards and a large number of open green spaces. A5A.18 Delhi Government's revenue receipts On the other hand, Old Delhi pre-dates the period of comprise its own revenue receipts (tax and non-tax) the Raj and is densely populated. The post-partition and grants from the Centre. The revenue receipts influx of refugees from Pakistan altered the land- increased from Rs 19.80 billion in 1994-95 to Rs 54.43 scape of New Delhi considerably, since a large billion in 2000-01, registering an annual growth rate number of refugee housing colonies came into of 17.6 percent during 1994-2001. The revenue existence with little regard for any land use regula- expenditure increased from Rs 14.30 billion in tion or environmental sanitation. The latest Master 1994-95 to Rs 36.96 billion in 2000-01 with an Plan of Delhi, MPD 2001, made provisions for average annual growth rate of 17.2 percent in 1994- various land-use categories and zones. However, it 2001. As a result, the revenue account of Delhi could only be implemented in limited measure, and government has a significant surplus. The major in the case of non-confirming industrial develop- expenditure stream indicative of expenditure on ment, with the help of the Supreme Court, as men- capital works (such as water works, drainage / tioned earlier. Unplanned settlements are common in sewerage, solid waste management, roads, etc) more Delhi to this day. than trebled between 1994-95 and 2000-01, indicat- ing spending by the government on construction of A5A.17 In administrative terms, Delhi is a Union new infrastructure as well as operation and mainte- Territory with its own legislature and a council of nance of capital infrastructure. ministers. However, a number of agencies that directly affect the urban governance, such as Mu- 8. Urban Air Quality nicipal Corporation of Delhi (MCD), the New Delhi Municipal Corporation (NDMC), and the Delhi Trends Development Authority (DDA), are not under the A5A.19 Figures A5A.5 and A5A.6 show annual Box A5A.2 : PM in Ambient Air in Delhi 2.5 The CPCB conducted PM monitoring employing portable samplers between February and July 2002 at a 2.5 busy traffic intersection (Income Tax Office crossing). The results illustrated above showed alarmingly high concentration of PM . Monthly mean concentration of PM between 2.5 2.5 February and July 2002 ranged between 46 µg/m3 and 153 µg/m3. On average, PM constituted 50 percent of PM and 25 2.5 10 percent of SPM. 66 Annex 5A : Delhi average ambient RSPM and SPM concentrations, samplers were developed in-house by NEERI, the respectively, between 1993 and 2002 in the three second generation instruments were manufactured different land-use areas. However, because of a large by a commercial manufacturer. amount of missing data (see Annex 3), the trends plotted should be interpreted with caution. The A5A.22 When compared with the data monitored nature of missing data is such that it could affect the by the CPCB between 2000 and 2003 it was found averages shown in the figures markedly. For that the overall monthly averages of (three) NEERI example, focusing on RSPM, the averages in 1993 are stations and (six) CPCB stations correspond most likely to be artificially elevated, since the three reasonably well in non-winter months, as shown in months in which ambient concentrations are gener- Annex 3A. ally low, June to August, are excluded from annual A5A.23 average calculations. In the residential area, there The overall average levels of SO2 and NO2 were no data in December 1993, lowering the annual are shown in Figures A5A.8 and A5A.9. While SO2 has been below the annual national standard, and average but unlikely to compensate for missing data showed a decreasing trend in all three areas, the in June­August. levels of NO2showed a sharp decrease in 1997, and A5A.20 RSPM shows a generally decreasing trend then increased in all three areas to reach levels in the in the commercial and industrial areas until 1997 vicinity of the annual national air quality standard after which the levels stabilize and even show a for residential areas in 2001-02. The fluctuations in slight increase. In the case of SPM, the minimum NO2 concentrations between 1993 and 1997 in Delhi level recorded is in 1997, followed by an increasing may indicate problems with data collection. trend in subsequent years. The sharp rise in RSPM A5A.24 Figure A5A.10 shows the monthly aver- and SPM in 1994 in the residential area seems to be ages of RSPM in the three areas over the years. As an experimental artifact because only the data from can be seen, there is a wide range of RSPM levels in November and December are available and taken as all three areas, with an overall maximum just above the "annual average." It is also worth noting that the 540 µg/m³, and minimum going down to approxi- residential monitoring site was shifted from the mately 20 µg/m³. Further, the RSPM levels were not terrace of a post office in Netaji Nagar to the terrace found to be significantly different between the three of a school in the neighboring Sarojini Nagar in 1994. areas (at the 10 percent level) as indicated by analy- A5A.21 RSPM averaged over the three areas as sis of variance results, pointing to the possibility that shown in Figure A5A.7. The averages for 1996 and sources of RSPM may be similar in all three areas. 1997 were 239 µg/m³ and 168 µg/m³ respectively, However, it is important to note that there are data showing that the decline is quite large (even with gaps, and these gaps are particularly serious in the missing data in 1996 for the month of August). When residential area in 1994 and in the industrial area in the chronology of interventions listed in Box A5A.3 2001 and 2002. were matched with Figure A5A.7, it appears that the 9. Seasonality closure of about 1,300 polluting industries in 1996 and 1997 time period may have contributed signifi- A5A.25 From empirical evidence, it is known that cantly to the decrease in RSPM levels, as would have the two parameters which directly impact concen- the various other interventions mentioned in Box trations of pollutants are rainfall and wind speed, A5A.3. However, these findings need to be inter- since higher precipitation leads to lower concentra- preted with caution given the data quality issues tion of pollutants, and higher wind speeds lead to discussed in Annex 3, especially the fact that the better dispersion. 1995-1997 period also coincided with the replace- ment of RSPM samplers. Although NEERI reports A5A.26 As can be seen in Figure A5A.11, rela- cross-calibration with the previous samplers before tively low wind speeds and rainfall are encountered deploying new ones to ensure consistency in mea- in Delhi for almost 3 months in a year. This is likely surements between the old and new instruments, the to lead to poor dispersion over that time period. An changes made were significant: the first generation empirical classification21 of the dispersion conditions was attempted based on the meteorological data. 67 For a Breath of Fresh Air Box. A5A.3 Major Technical and Policy Interventions in Delhi (1991-2004) that Impacted Air Quality Year Action Year Action 1991 1) Mass emission norms for new petrol vehicles 2) Benzene in petrol reduced to 3% introduced on 1/4/1991 3) Ban on alteration of vehicles by replacing 1992 1) Mass emission norms for new diesel vehicles petrol engines with diesel engines with effect introduced on 1/4/1992 from 1/4/98 2) Start of closure of stone crushers 4) Low smoke 2T oil mandated 1993 N.A. 5) Pre-mix 2T engine oil mandated for two- 1994 1) Lead in petrol reduced to 0.15 gm/litre wheelers by 31/12/1998; ban on supply in June of loose 2T oils from 31/12/1998 1995 1) Catalytic converters for 4-wheeled petrol 6) Phasing out/ban on plying of old driven vehicles mandated with effect commercial/transport vehicles more than from April 20 years old by 2/10/1998 2) Unleaded petrol introduced to compliment 7) Phasing out/ban on plying of old commer- the introduction of catalytic converters cial/transport vehicles more than 17 years 1996 1) Mass emission norms for new vehicles made old by 15/11/1998 more stringent on 1/4/1996 8) Phasing out/ban on plying of old 2) Fuel quality: 0.5% S diesel mandated in commercial/transport vehicles more than December 15 years old by 31/12/1998 3) Benzene in petrol reduced to 5% 9) Start of construction of Delhi metro 4) CNG vehicles and catalytic converters (underground) and a major flyover (unsuccessfully) introduced for government construction program vehicles, excluding public transport. 10) Leaded petrol phased out by 1/9/1998 5) Closure of 168 hazardous industries, 1999 1) India Stage-I (Euro-I) emission norms including stone crushers completed by mandated for all category of new vehicles 30 /11/1996 2) Emission norms for tractors introduced 6) Approved fuel notification issued for all 3) Taxis older than 12 years phased out by activities (e.g. FO and LSHS with 1.8% 31/3/99 S; coal with 0.4% S) 4) Registration of Autorickshaws and diesel 1997 1) Relocation of 513 industries by 31/1/1997 taxis only if they conform to Bharat 2) Closure of 43 Hot mix plant by 28/2/1997 Stage-II (Euro II) norms with effect from 3) Closure of 21 Arc/induction furnaces by 29/4/1999 31/3/97 5) Fuel quality: 0.25% S diesel made 4) Closure and relocation of 246 Brick Kilns mandatory outside Delhi by 30/6/1997 6) Restriction on the plying of goods vehicles 5) Installation of pollution control devices by during the day from August 1999 all air polluting industries by 31/12/1997 2000 1) Taxis older than 10 years phased out by mandated March 6) Ban on registration of army disposal 2) Replacement of all pre-1990 autorickshaws vehicles, government auctioned vehicles and taxis with new vehicles on clean fuels 7) Commercial goods and passenger vehicles by 31/3/2000 (LCVs) beyond a specified life span by 3) Conversion of 3-wheelers to CNG initiated 31/12/1997 4) Expansion of the number of CNG supply 8) 337 hazardous category industries shifted on outlets from 8 to 30 by 31/3/2000 30/6/96 (total of 1160 industries closed or 5) All new private (non commercial) relocated in 1997) 4-wheeled vehicles to conform to 1998 1) Emission Norms for catalytic converter Euro II norms by 1/4/2000 fitted passenger vehicles from 1/4/1998 6) Buses more than 8 years old phased out or to ply on CNG or other clean fuels 68 Annex 5A : Delhi Year Action Year Action 7) Number of CNG vehicles as follows: 3983 9) Construction of 9 flyovers completed 3-wheelers; 877 taxis; 61 buses; 13 Rural 10) Emission norms for LPG vehicles Transport Vehicles (RTVs) (total of 4834 introduced from May 2001 minus private vehicles) 2002 1) Extension of deadline for conversion of 8) Low Sulfur petrol & diesel (0.05%) entire city bus fleet to CNG mode till mandated from 1/4/2000 for private 31/1/2002 (started but not completed by vehicles deadline) 9) The three coal based power plants 2) 94 CNG stations set up by March switched over to use of beneficiated coal 3) All diesel buses phased out by November 10) Petrol with 1% benzene mandated from 4) Fresh registration of 5000 new CNG November 2000 based 3-wheelers in Delhi from 11) Piped natural gas made available by 20/12/2002 March 2000 to 1311 domestic, 9 small, and 5) Number of CNG vehicles as follows: 3 large commercial establishments 35678 3-wheelers; 4816 taxis; 4231 12) Bharat Stage-II (Euro-II) emission buses; 2165 norms for all non-commercial vehicles Rural Transport Vehicles(RTVs); 10350 by April private (57240 total) 2001 1) Bharat Stage-II (Euro-II) emission 6) Piped NG available by March to 4111 norms for all commercial vehicles by domestic, 37 small , and 5 large October commercial establishments 2) The Sulfur content of diesel supplied to 7) 16340 non-destined good vehicles turned NCT Delhi further reduced to 0.05% away from entering Delhi between July with effect from October and November 2001 (to go with the above measure) 8) Emission norms for CNG vehicles (all 3) Replacement of all post-1990 categories) and LPG vehicles (heavy autorickshaws and taxis with new vehicles duty) with effect from May 2002 and on clean fuels (Court also allowed usage of October 2002 respectively conversion kit on such vehicles) by 2003 1) Supreme Court order for preparation of 31/3/2001 Action Plan for lowering ambient RSPM 4) Entire city bus fleet (DTC and private) to be levels on 14/8/2003 converted to single fuel mode on CNG by 2) Construction of 19 flyovers completed. 31/3/2001 (started but not completed 2004 1) Notification issued by Government of by deadline) India revising the emission norms of in 5) Augmentation of public transport to use vehicles with effect from 1/10/2004. 10,000 buses by 1/4/2001 N.A. ­ Not applicable. 6) Total of 3538 hazardous category Notes (1): Interventions common to all five cities are industries closed shown in Italics. However, all common interventions were 7) Number of CNG vehicles as follows: 14000 not implemented in the same year in all cities. (2) : Dates 3-wheelers; 2200 taxis; 400 buses; to be read as day/month/year. 250 Rural Transport Vehicles (RTVs); 9500 private (26350 total) 8) Piped NG available by March to2821 domestic, 15 small , and 5 large commercial establishments 69 For a Breath of Fresh Air It was found that good dispersion conditions exist A5A.28 The commonly used definitions of from June to September, moderate conditions from summer, winter, and monsoon seasons were used to January to May, and poor dispersion conditions exist aggregate the months, in order to further explore from October to December. Further evidence of the any seasonality in the variation of RSPM levels. effect of relatively poor dispersion from October to November can be seen in the monthly variation of A5A.29 The effect of seasons on RSPM is clearly RSPM. visible in Figure A5A.15. The averages for monsoon, summer, and winter months were 101 µg/m³, A5A.27 As can be seen in Figure A5A.12 and 188 µg/m³, and 283 µg/m³, respectively. However, A5A.13, there is a strong effect of meteorological since the summer, winter and monsoon classifica- parameters, with very high levels being recorded at tion shown is subjective in nature, the empirically the end of the year. Figure A5A.14 shows that SO2 derived classification of dispersion conditions and NO2 levels also show a strong effect of meteoro- discussed at the beginning of this section was also logical factors, with fairly high NO2 concentrations used to assess seasonality in RSPM variation. It was recorded in the winter months. However, given the found that there was a strong effect of the three cold temperatures in Delhi compared to southern dispersion classifications. The RSPM averages for states in winters, the high RSPM and gaseous pollut- the good, moderate, and poor dispersion classifica- ant levels in winters are also in part a result of tions were 108 µg/m³, 213 µg/m³, and 291 µg/m³, increased solid fuel use for domestic and space respectively. heating purposes. 70 Annex 5B Kolkata 1. Geographical Location the Bay of Bengal). The boundary of Kolkata is contiguous with Kolkata Municipal Corporation A5B.01 Kolkata and Howrah are twin cities (KMC), which has an area of 104 km2 (GoI 1999). situated on the eastern and western sides of the river Figure A5B.1 shows the approximate location of the Hoogly respectively (about 150 kms upstream from pollution monitoring stations. Figure A5B.1: Approximate location of NEERI monitoring stations in industrial commercial and residential areas 71 For a Breath of Fresh Air 2. Demography tax collection of the State (GoI 1999). A5B.02 The KMA accounts for less than 20 percent A5B.07 The per capita net state domestic product of West Bengal's population. As per 2001 census, the (2000-01 at current prices) was Rs 18,021. Annual population of KMA grew at an annual growth rate of per capita income of the city was Rs 10, 636 1.8 percent from 11.02 million in 1991 to 13.2 in 2001 (1998-99). as against 11.02 million in 1991 at an annual growth rate of 1.82 percent. During the same period, the 5. Industry population of the KMC grew from 4.39 to 4.58 million, which shows that most of the growth in the A5B.08 Kolkata forms the commercial and indus- decade between 1991 and 2001 occurred outside of trial hub of eastern India. Due to the availability of KMC. infrastructural facility which includes river fronts, a port and an abundance of water and raw materials A5B.03 Figure A5B.2 shows the annual population for industrial activity, industrial development took between the two census periods. The population place predominantly along the river (GoWB 2001). density of the municipal corporation area in 2002 was 24,705 persons per km2 making Kolkata one of the A5B.09 Kolkata's traditional industries include most densely populated cities in India. jute manufacturing, light and heavy engineering, leather products and tea trading. However, it has a 3. Climate range of other industries including textiles, paper, pharmaceuticals, chemicals, tobacco, railway wag- A5B.04 The weather profile for Kolkata is shown in ons, automobiles, heavy machinery, and steel. A large Table A5B.1, based on data from 1991 to 2002. The city number of industrial estates have come up in the experiences a humid and tropical climate. It is char- metropolitan area, and approximately 329 small- acterized by hot and dry summer from February to scale units of varying capacities are operating within April, monsoon from May to October, and a moder- the metropolitan area of Kolkata. There are approxi- ately cool winter between November and January. mately 297 small but hazardous industrial units The mean monthly temperature ranges from 12.6°C operating in densely populated residential areas. in the winters to 35.6°C in summer months. Calm Many of these industries use coal-fired boilers for conditions prevail frequently during winter months, generation of steam. Other polluting units include and are more common in the evening hours. Relative coal-fired power generating units, such as the New humidity remains quite high throughout the year. Cossipore thermal power plant, Southern power generating station (Watgung) in the city and Budge 4. Economy Budge, Titagarh, Mulajore, and Bandel thermal A5B.05 The KMA accounts for 25 percent of the power plants in the suburbs or in the nearby areas. state's income with less than 20 percent of the state's According to the State Pollution Control Board, population. In terms of employment, KMA accounts industries are major contributors of air pollutants in for about 43.5 percent and 48 percent of workers in the city. About 32 tons of small particulates are secondary and tertiary sectors respectively22. Out of released to the atmosphere by the industries in the total registered small-scale units in West Bengal, Kolkata: 16 tons from New Cossipore thermal power KMA houses around 56 percent, providing about 61 plant, 14 tons as uncontrolled emissions from small percent employment. For the KMA as a whole, the boilers, and 2 tons from Southern power generating ratio of formal to informal employees is about 10.9, station and other large industries (GoWB 2001). which shows that there is a large informal sector in Recognizing that industries contribute dispropor- Kolkata. tionately to air pollution in Kolkata, the government has taken a number of steps against polluting A5B.06 The economic base of the city contributes industries (see Box A5B.4). substantially to the total revenue collection of both the central and state governments. Kolkata city 6. Transport alone shares about 80 percent of annual sales tax A5B.10 Public transportation is the predominant collection and about 97 percent of annual income form for transport of passengers in Kolkata, 72 Annex 5B - Kolkata Figure A5B.2: Population in Kolkata metropolitan area between the last two census periods (Pandey 2003). including trams, taxis, three-wheelers, local trains, High population density also results in a large metro-rail, circular rail, and ferry service, coupled number of pedestrians per unit area, which affects with non-motorized vehicles such as cycle- traffic movement. rickshaws. Over the last decade the number of A5B.11 The vehicular growth in Kolkata as shown registered automobiles in Kolkata has increased by in Figure A5B.3 and A5B.4 shows that there has been about 52 percent. The effective road area available in a steady increase in the number of vehicles in the Kolkata is less than 5 percent of the total area, city since 1991. The number of registered vehicles in leading to very high automobile density, dispropor- 2001 was more than 750,000, out of which two- tionately low percentage of road network, congestion wheelers constituted about 44 percent, cars 37 and traffic jams. The traffic movement in the city is percent, three-wheelers 2 percent, goods carriages 9 also affected by poor standard of maintenance of the percent, and buses, taxis, jeeps constituting the rest. roads, old vehicles, and a mixture of traffic including The average growth rate of the vehicle population auto-rickshaws, trucks, hand carts, hand rickshaws, has been between 3 and 5 percent per annum in the stray animals and a large number of pedestrians. last decade (MoRTH 2003). Figure A5B.3: Growth in commercial motor vehicles in Kolkata (MoRTH 2003) 73 For a Breath of Fresh Air Table A5B.1: Overall monthly average values of key meteorological parameters Mean Mean Mean No. of Relative Total Max. Min. Wind Calm Days Humidity % Rainfall Temp. Temp. Speed (mm) (0C) (0C) (kmph) 0830 hrs 1730 hrs 0830 hrs 1730 hrs January 25.4 12.6 3.4 12 23 75 61 18 February 28.8 16.6 4.2 9 14 72 55 28 March 33.6 21.4 5.3 7 11 66 48 47 April 35.6 24.6 7.3 3 3 70 58 52 May 35.6 25.9 8.7 3 2 73 68 146 June 34.2 26.4 7.8 3 3 80 78 260 July 33.1 26.2 7.0 4 4 83 82 318 August 32.8 26.3 6.3 4 5 83 83 351 September 32.9 25.8 5.4 6 8 81 81 284 October 32.4 24.0 3.8 9 17 76 76 153 November 30.0 19.3 3.0 11 25 72 69 32 December 26.8 13.7 3.0 12 27 73 65 7 Total/year 83 142 1696 Note: Data on calm days and relative humidity are collected twice a day. A5B.12 A large number of highly polluting old tax and non-tax revenues have increased substan- vehicles (about 54 percent of total) are presently tially in recent years. In particular, water charges, plying in the city streets. Of the old vehicles, 56 amusement fees, building sanction fees, and com- percent of cars, 55 percent of trucks, 56 percent of mercial markets have seen substantial growth, with buses, 46 percent of taxis, and 25 percent of two- the result that in 2002-03, self-generated revenues wheelers are more than 15 years old (GoWB 2001). outscored the government grants. The exhaust emissions from old and slow-moving traffic contribute heavily to atmospheric pollution. A 8. Urban Air Quality number of steps, some as part of country-wide Trends measures, were undertaken in the last few years to address emissions from motor vehicles (see Box A5B.15 Figures A5B.5 and A5B.6 show annual A5B.4). average ambient RSPM and SPM concentrations, respectively, between 1993 and 2002 in the three 7. Urban different land-use areas. However, as in the case of A5B.13 The land-use breakdown in 1998 included Delhi, because of a large amount of missing data (see 45 percent residential areas, 6 percent major indus- Annex 3), the trends plotted should be interpreted trial areas, 2 percent organized open space and with caution. The nature of missing data is such that recreational areas, 8 percent transportation space, it could affect the averages shown in the figures and 33 percent vacant (including cultivable) land, markedly. RSPM shows a generally decreasing trend forest and waste land, water body and swampy land in the commercial and industrial areas. In the (GoWB 2001). residential area, there was a break in monitoring on account of change in location of the monitoring site A5B.14 The KMC managed to raise its own rev- in 1997, and higher concentrations were recorded enues by almost Rs 700 million during 2002-03, an subsequently. A similar trend is evident in the case of increase of 27 percent over the previous year. Both SPM, though it is not as pronounced. The decline in 74 Annex 5B - Kolkata Figure A5B.4: Growth in private motor vehicles in Kolkata (MoRTH 2003) RSPM is comparable between the commercial and A5B.17 The overall averages of SO2 and NO2 are industrial sites. As seen in Box A5B.4 a large shown in Figures A5B.7 and A5B.8. While the number of interventions in Kolkata have focused ambient concentrations of SO2 were broadly on the industrial sector. steady in the residential area, it showed a signifi- cant increase in the commercial and industrial A5B.16 When compared with RSPM data from the areas in 1998-99, followed by a decrease in 2000. SPCB monitoring location between 2000 and 2002, There is no obvious explanation for this trend it was found that the overall averages of NEERI between 1998 and 2000, since lower sulfur limits and SPCB levels correspond very well to each other, were being implemented during this period. NO2 as shown in Annex 3A. Regression analysis between remained largely steady in all the three areas until the two datasets returned a R-squared value of about 2000, after which it showed a significant 0.80. However, this result needs to be interpreted increase. The levels of NO2 in all three areas with caution since averaging across different moni- violated the annual national standard for toring stations can lead to calculation artifacts. residential areas in 2002. There are no obvious Figure A5B.5: Trend in RSPM levels in three areas of Kolkata Note: The dotted line shows the annual national ambient standard of 60 µg/m³ for RSPM for residential areas. 75 For a Breath of Fresh Air Figure A5B.6: Trend in SPM levels in three areas of Kolkata Notes: The dotted line shows the annual national ambient standard of 140mg/m³ for SPM for residential areas. A large amount of data are missing between 1993 and 1997, including no data at the residential site for the entire year in 1997, when the monitoring site was being moved to another location. The first year in which some data are available for every month of the year in Kolkata was 1999. Figure A5B.7: Trend in SO levels in Kolkata 2 Note: The annual national ambient standard for SO in residential area is 60 µg/m³. 2 Figure A5B.8: Trend in NO levels in Kolkata 2 Note: The annual national ambient standard for NO in residential area is 60 µg/m³. 2 76 Annex 5B - Kolkata Box A5B.4.: Major Technical and Policy Interventions in Kolkata (1991-2004) that Impacted Air Quality Year Action Year Action 1991 1) Mass emission norms for new petrol vehicles 1999 1) Euro-I (India 2000) norms in KMA with effect introduced on 1/4/1991 from 1/11/1999 for new private non 1992 N.A. commercial vehicles 1993 1) In-use emission standards for smoke density 2) Pre-mixed 2T engine oil made mandatory on 26/3/1993 for two-wheelers with effect from 1/4/1999. 2) Mass emission standards for new diesel 2000 1) India Stage-I (Euro-I) emission norms vehicles introduced on 15/9/1993 man dated for all category of new vehicles 1994 1) Lead in petrol reduced to 0.15 gm/litre in June with effect from 1/4/2000 1995 1) In reference to the court case of 2) Fuel quality: 0.25% S diesel mandated Mr. M. C. Mehta versus Government of India, from 1/4/2000 the Supreme Court directed air polluting 3) Leaded petrol phased out by February industries to install air pollution control devices 4) Reduction of Benzene content in petrol to within a stipulated time period, failing which 3% in Kolkata from 1/4/2000 such industries will be liable for closure. The 2001 1) Bharat Stage-II (Euro-II) emission norms for directions of the Court were implemented all new non-commercial vehicles by July 2) Catalytic converters for 4-wheeled petrol 2) Low Sulfur petrol and diesel (0.05%) drivenvehicles mandated with effect from April mandated from 1/7/2001 3) Unleaded petrol introduced to complement 3) Introduction of stricter emission standard for the introduction of catalytic converters boilers, ceramic kilns, foundries and rolling 1996 1) Mass emission norms for new vehicles made mills of KMA with effect from 11/5/2001 more stringent on 1/4/1996 4) Cleaner fuel made mandatory for boilers 2) Fuel Quality: 0.5% S diesel mandated in (< 2 ton/hr capacity), ceramic kilns and December rolling mills 3) Benzene in petrol reduced to 5% 5) All small boilers (steam generation capacity 1997 1) Some buses of State Transport Corporation < 2 ton) of KMC area and ceramic kilns of ordered off the road by the Calcutta High KMA were directed to convert the boiler/ Court on grounds of air and noise pollution. ceramic kilns to cleaner fuel (low Sulfur oil 2) Implementation of industrial siting policy or gas) fired ones restricting setting up of polluting industries 6) Bharat Stage II (Euro-II) emission norms (red category) in municipal areas of Kolkata mandated for all new commercial vehicles Metropolitan, Authority (KMA) with effect from 23/10/2001 in KMA 1998 1) Emission norms for catalytic converter fitted 7) Supply, distribution and selling of loose passenger vehicles from 1/4/1998 2T oil in KMA banned with effect from 2) 0.25% Sulfur diesel mandated in Kolkatta 1/10/2001 Metropolitan Area from 1/4/1998 8) Emission norms for LPG vehicles intro 3) Low smoke 2T oil mandated duced from May 2001 4) Environmental Bench of High Court ordered 2002 1) Bharat Stage-II (Euro-II) emission norms for that before starting any project environmen all commercial vehicles by October tal clearance is mandatory, and the order was 2) Regularly complying industries felicitated implemented. with the Environmental Excellence Award by SPCB and other cleaner production partners 77 For a Breath of Fresh Air Year Action Year Action 3) Emission norms for CNG vehicles (all 5) 5% ethanol blended gasoline mandated from categories) and LPG vehicles (heavy duty) January with effect from May 2002 and October 2002 6) Supreme court order for preparation of respectively Action plan for lowering ambient RSPM 2003 1) Most of the ceramic kilns and small boilers levels on 14/8/2003 have been converted from coal fired to oil/gas 2004 1) Notification issued by Government of India fired system revising the emission norms of in-use 2) Closure order has been given to the units not vehicles with effect from 1/10/2004. taken any step towards conversion from coal to cleaner fuel N.A. ­ Not applicable 3) About 67 percent of the coal fired boilers and about 78 percent of the coal fired ceramic Notes : (1) Interventions common to all five cities are kilns have already been converted to oil fired shown in italics. However, all common interventions were ones not implemented in the same year in all cities. 4) Only LPG driven three wheelers are (2) Dates to be read as day/month/year. registered in Kolkata from June 2003 78 Annex 5B - Kolkata Figure A5B.9: Trend in RSPM monthly concentration at the residential, commercial and industrial area monitoring locations. 79 For a Breath of Fresh Air Figure A5B.10: Monthly average total rainfall and mean wind speed Figure A5B.11: Monthly average total rainfall and RSPM concentration Figure A5B.12: Monthly mean wind speed and RSPM concentration 80 Annex 5B - Kolkata Figure A5B.13: Variation in monthly averages of SO and NO . 2 2 explanations for the increasing trend in NO2 after levels was large, with a maximum of about 600 2000. µg/m³ and a minimum of about 20 µg/m³. How- ever, the difference in RSPM levels across the A5B.18 Figure A5B.9 shows the monthly average three areas was not found to be significant (at the concentrations of RSPM at the three monitoring 10 percent level) as indicated by analysis of locations in Kolkata. However there are a lot of variance, indicating the possibility that sources of gaps in the data until 1998, which makes it difficult RSPM may be scattered across the city. to draw meaningful conclusions concerning trends in the 1993 to 1998 period. Focusing on the data after 9. Seasonality 1998, Figure A5B.9 shows that RSPM levels in the residential area remained steady compared to the A5B.19 From empirical evidence, it is known that commercial and industrial areas, both of which the two parameters that directly impact concen- showed a decreasing trend. The range of RSPM trations of pollutants are rainfall and wind speed, Figure A5B.14: Overall variation in RSPM with seasons (Monsoon: May to October; Summer: February to April; Winter: November to January) 81 For a Breath of Fresh Air since higher precipitation leads to lower concen- temperature compared to southern states, the high trations of pollutants, and higher wind speeds RSPM and gaseous pollutant levels are also likely to lead to better dispersion. be confounded by increased solid fuel use for domes- tic heating purposes. A5B.20 As can be seen in Figure A5B.10, rela- tively low wind speeds and rainfall are encoun- A5B.23 The commonly used definitions of summer, tered in Kolkata for almost five months in a year. winter, and monsoon seasons were used to aggregate This is likely to lead to poor dispersion over that the months, in order to further explore any seasonal- time period. An empirical classification (see ity in the variation of RSPM levels. endnote #21) of the dispersion conditions was attempted based on the meteorological data. It A5B.24 As can be seen in Figure A5B.14, effects of was found that good dispersion conditions exist seasons are clearly evident. The averages for from April to October, while poor dispersion monsoon, summer, and winter months were 107 conditions exist in the rest of the year. Further µg/m³, 179 µg/m³, and 256 µg/m³, respectively. evidence of the effects of dispersion conditions A5B.25 Since the summer, winter, and monsoon can be seen in the monthly variation of RSPM. classification shown is subjective in nature, the A5B.21 As can be seen in Figure A5B.11 and empirically derived classification of dispersion A5B.12, there is a strong effect of meteorological conditions discussed at the beginning of this section parameters, with very high levels being recorded at was also used to assess seasonality in RSPM varia- the beginning and the end of the year. tion. It was found that there was a strong effect of the two dispersion classifications. The RSPM A5B.22 As Figure A5B.13 demonstrates, SO2 averages for the good and poor dispersion classifica- and NO2 levels show a strong effect of meteorologi- tions were 108 µg/m³ and 242 µg/m³, respectively. cal factors. However, given the lower winter 82 Annex 5C Mumbai 1. Geographical location ridges of low hills on the east and the west. The western ridge terminates in Malabar hill which is the A5C.01 Mumbai is situated on the Arabian Sea, highest point in the city rising to 55 m above the sea with much of it being a low-lying plain at the mean level. Mumbai or Greater Mumbai with an area of altitude of sea level. The plain is flanked by two 437 km2, stretches longitudinally for 42 km (north to Figure A5C.1: Approximate location of NEERI monitoring stations in industrial, commercial and residential areas. 83 For a Breath of Fresh Air Figure A5C.2: Population in Mumbai between the last two census periods (Pandey 2003) Table A5C.1: Overall monthly average values of key meteorological parameters Mean Mean Mean No. of Relative Total Max. Min. Wind Calm Days Humidity % Rainfall Temp. Temp. Speed (mm) (0C) (0C) (kmph) 0830 hrs 1730 hrs 0830 hrs 1730 hrs January 31.0 16.9 5.9 20 0 71 49 66 February 31.2 17.9 6.6 18 0 70 49 0 March 32.9 21.0 7.2 19 0 71 52 0 April 32.9 24.0 7.6 12 0 69 60 0 May 33.3 26.8 9.3 10 0 70 66 42 June 32.4 26.8 11.2 6 1 79 74 438 July 30.6 25.8 12.6 4 2 85 81 777 August 30.3 25.4 11.3 7 1 86 80 484 September 30.8 24.8 7.0 17 2 85 76 354 October 33.3 23.9 5.7 22 1 76 66 115 November 33.8 21.3 5.3 15 1 63 53 8 December 32.4 18.1 5.1 18 1 64 50 3 Total/year 169 11 2287 Note: Data on calm days and relative humidity is collected twice a day. 84 Annex 5C : Mumbai Figure A5C.3: Growth in commercial motor vehicles in Mumbai (MoRTH 2003) Figure A5C.4: Growth in private motor vehicles in Mumbai (MoRTH 2003) south) and has a maximum width of only 17 km A5C.03 Figure A5C.2 shows the population (Bombay First 2004). Figure A5C.1 shows approxi- between the last two census counts. A possible reason mate location of NEERI pollution monitoring of nearly constant population growth in the last stations in Mumbai. decade could be the developments in surrounding areas of Mumbai and the declining rate of migration 2. Demography into city. The population density of urban area is A5C.02 Mumbai accounts for 12 percent of 25,449 persons per km2 , making it one of the most Maharashtra's population and for 1.1 percent of densely populated cities in India. As of 1991, the the country's population. As per 2001 census, the average number of persons per household was 4.8 population of Mumbai grew by 16.7 percent from and the average number of persons living per room 9.9 million in 1991 to 11.9 million in 2001. The was 3.4 (Bombay First 2004). annual growth rate of population during 1991- 3. Climate 2001 was 1.9 percent. A5C.04 The weather profile in Mumbai between 85 For a Breath of Fresh Air Figure A5C.5: Trend in RSPM levels in Mumbai Note: The dotted line shows the annual national ambient standard of 60 µg/m³ for RSPM for residential areas. Figure A5C.6: Trend in SPM levels in Mumbai Notes: The dotted line shows the annual national ambient standard of 140µg/m³ for SPM for residential areas. There are missing data in 1993 and 1994. In particular, there are no data at the residential site except January in 1993 and for six months out of twelve in 1994. The monitoring equipment was replaced with new equipment from a different manufacturer in between 1998 and 1999. The residential site monitoring station was moved to another location in 2001. 1991 and 2002 is shown in Table A5C.1. The city 4. Economy experiences a warm tropical climate. The mean temperature ranges from 16.9 °C in winter (October A5C.05 Mumbai's annual per capita income in to February) to 33.3 °C in the summer (March to 2000-01 was Rs 31,922, which was twice that of May). Winds are generally moderate, but increase in Maharashtra. At the same time, the GSDP was force during monsoon months. The city receives Rs 414 billion with the tertiary sector accounting for heavy rainfall during monsoon (June to September). more than 60 percent of the income generated in Calm conditions prevail generally in the morning Mumbai. Interestingly, while the share of the hours, but the winds pick up in the evening. The tertiary sector in the income of Mumbai has in- relative humidity is higher in the morning hours, creased in the last decade, the share of the secondary complementary to calm conditions. and primary sectors (see endnote #20) has remained almost static (Bombay First 2004). 86 Annex 5C : Mumbai 5. Industry a decrease from 44 percent to 34 percent in the number of industries located in Mumbai between A5C.06 Most of the industries are located in 1993 and 2000. A large part of the decline was in the the eastern and northeastern corridor of Greater textile industry (Bombay First 2004). Mumbai, with a few in the western region. The number of factories in Mumbai declined from 7,832 A5C.07 Industries in the air polluting category in 1991 to 6,986 in 1999, and so did Mumbai's share include textile mills, chemicals, pharmaceuticals, of factories in the state. As per the available fuel-use engineering, and foundries. Out of 183 air polluting data for Mumbai the use of furnace oil (which is industries in the Mumbai region, 70 are large scale, used mainly in industry) also declined from 37 are medium scale and 76 are of small scale. In 1,441,888 to 1,352,567 kiloliters between 1998-99 and addition, there are 32 stone crushers in the north- 2001-02. Among the large and medium industries in eastern parts of city, and a giant fertilizer/chemical Maharashtra, while there was a slight increase in complex and thermal power plant in Chembur. absolute numbers (1,069 to 1,129), overall there was However, industrial fuel use has been shifting to Figure A5C.7: Trend in SO levels in Mumbai 2 Note: The annual national ambient standard for SO2 in residential areas is 60 µg/m³. Figure A5C.8: Trend in NO levels in Mumbai 2 Note: The annual national ambient standard for NO2 in residential areas is 60 µg/m³. 87 For a Breath of Fresh Air Box A5C.5: Results of RSPM Source Apportionment in Mumbai As part of the World Bank funded Mumbai Urban Transport Project (MUTP), a source apportionment study based on dispersion modeling was conducted with a focus on RSPM emissions from vehicular sources. Air quality monitoring was conducted at various locations in Mumbai in 2001-2002 to cover areas with different levels of vehicular move- ment. While the methodology used for source apportionment was not state-of-the-art, and hence could not resolve all sources in detail, the results provide reasonable order-of-magnitude estimates of the contribution of different sources to RSPM levels, as shown below. While the kerbside sites illustrate dominance of vehicular emissions, the control (an urban area with little activity away from any emission hot spot) and normal activity sites show that industrial sources can be major contributors to RSPM pollution in ambient air. Resuspended road dust seems to contribute significantly in all areas. The study also used secondary data to construct an overall emissions inventory as shown below. It was estimated that out of the 16,550 tonnes of RSPM emission per year emissions from industrial sources dominated, followed by area sources, vehicles, and construction activities. Reference: NEERI 2004 natural gas over the years. The number of industrial Their composition was 45 percent two-wheelers, natural gas users increased from 0 to 40 between 30 percent cars, 6 percent taxis, 6 percent heavy 1995 and 2002, while the number of commercial vehicles, 10 percent three-wheelers, and 3 percent natural gas users increased to almost 420 in the other vehicles (jeeps, omni-buses, tractors, and same time period (Mahanagar Gas 2003). trailers). 6. Transport A5C.09 The bulk of the growth of vehicle popu- lation has been on account of two-wheelers and A5C.08 Mumbai accounts for 17 percent of the three-wheelers as can be seen in Figures A5C.3 and vehicle population of the state and about 50 percent A5C.4. Since 1995 the number of trucks and lorries of the car population. The total number of vehicles in registered in Mumbai has shown a decreasing trend. Mumbai in the year 2002 was close to 1.1 million. The increasing trend in the number of light motor 88 Annex 5C : Mumbai Box A5C.6: Major Technical and Policy Interventions in Mumbai (1991 -2004) that Impacted Air Quality Year Action Year Action 1991 1) Mass emission norms for new petrol 2) Low Sulfur diesel and petrol (0.05%) vehicles introduced on 1/4/1991 mandated 1992 1) Mass emission norms for new diesel 3) Emission norms for LPG vehicles vehicles introduced on 1/4/1992 introduced from May 2001 1993 1) Policy framed to restrict the industrial 2002 1) Bharat Stage-II (Euro-II) emission norms growth in and around Mumbai called as for all commercial vehicles mandated by "Mumbai Metropolitan Region" October 1994 1) Lead in petrol reduced to 0.15 gm/litre in 2) All taxis over the age of 15 years phased June out, unless converted to run on CNG/ LPG 1995 1) Catalytic converters for 4-wheeled petrol 3) All taxis of Premier 137D model phased out driven vehicles mandated with effect from unless converted to run on CNG/LPG with April effect from 1/3/2002 2) Unleaded petrol introduced to compliment 4) All three wheelers over the age of 10 years the introduction of catalytic converters phased out, unless converted to run on 1996 1) Mass emission norms for new vehicles CNG/ LPG with effect from 1/3/2002 made more stringent on 1/4/1996 5) 35,000 CNG vehicles plying 2) Fuel quality: 0.5% S diesel mandated in 6) Emission norms for CNG vehicles (all December categories) and LPG vehicles (heavy duty) 3) Benzene in petrol reduced to 5% with effect from May 2002 and October 4) Textile industries decline started after the 2002 respectively strikes of textile workers 2003 1) All taxis over the age of 8 years phased 1997 1) CNG conversion of taxis started out, unless converted to run on CNG/LPG 2) Construction of large number of flyovers with effect from 1/1/2003 started 2) Some 3-wheelers over the age of 8 years 1998 1) Emission Norms for catalytic converter fitted phased out or, converted to run on CNG/ passenger vehicles from 1/4/1998 LPG with effect from 1/1/2003 2) Low smoke 2T oil mandated 3) All transport vehicles over the age of 3) CNG conversion of taxis on a large scale 15 years, with the exception of BEST 4) Construction of flyovers continues buses, phased out or converted to run on 1999 1) Emission norms for tractors introduced CNG/ LPG with effect from 1/1/2003 2) Pre-mixed 2T engine oil made mandatory 4) 5% ethanol blended gasoline mandated for two-wheelers with effect from January 2003 2000 1) India Stage-I (Euro-I) emission norms 5) Supreme court order for preparation of mandated for all category of new vehicles Action Plan for lowering ambient RSPM by 1/4/2000 levels on 14/08/2003 2) Fuel quality: 0.25% S diesel mandated from 2004 1) All transport vehicles over the age of 8 01/04/2000 years, with the exception of BEST buses, 3) Leaded petrol phased out in February phased out or converted to run on CNG/ LP 4) Closure of bulk drugs and pesticides with effect from 1/2/2004 manufacturing units 2) Notification issued by Government of India 5) A number of industries were converted to revising the emission norms of in-use natural gas available from a terminal near vehicles with effect from 1/10/2004. Mumbai N.A. ­ Not applicable 6) Construction of flyovers ends (about 36 in Notes: 1). Interventions common to all five cities number) are shown in Italics. However, all common interven 2001 1) Bharat Stage-II (Euro-II) emission norms tions were not implemented in the same year in all for all new non-commercial vehicles man cities. 2). Dates to be read as day/month/year. dated by July 89 For a Breath of Fresh Air Figure A5C.9: Trend in RSPM monthly concentration at the residential, commercial and industrial area monitoring locations 90 Annex 5C : Mumbai Figure A5C.10: Monthly average total rainfall and mean wind speed vehicles, taxis and three-wheelers since 1993 seems program of flyover construction, and 36-odd to have stopped in 2001. However, two-wheelers flyovers were constructed between 1997 and 2000. and cars seem to have continued on a steep increas- ing trend. In addition to the vehicles registered A5C.11 It is acknowledged that vehicular emis- within Mumbai, it is well known that a large sions contribute significantly to the air pollution in number of commercial vehicles that ply in Mumbai Mumbai. Various initiatives, some voluntary and are registered in the neighboring district of Thane. others mandated, have been implemented in Mumbai in the last few years to address vehicular A5C.10 The road network in Mumbai com- emissions, as shown in Box A5C.6. One of the major prises three main corridors: Western Express developments was the appointment of a committee Highway, Eastern Express Highway, and the (referred to as the Lal Committee) in December 1999, Central Corridor. The total length of roads in to look into various issues pertaining to air pollu- Mumbai is 1,889 km (507 km in the city, 914 km in tion in the city. This was in response to a Public western suburbs, 468 km in eastern suburbs). Interest Litigation filed in the High Court (Smoke However, the road network has become very Affected Residents Forum Vs Municipal Corpora- congested over the years with increasing number tion of Greater Mumbai and others). The recom- of vehicles, and more people traveling to and from mendations of the Lal Committee have formed the city. As a result of the growing traffic conges- the basis of a number of interventions that have tion, Mumbai embarked on a very aggressive taken place in Mumbai since 1999 (NEERI 2004). Figure A5C.11: Monthly average total rainfall and RSPM concentration 91 For a Breath of Fresh Air Figure A5C.12: Monthly mean wind speed and RSPM concentration 7. Urban 1991-92 and 1995-96, the BMC maintained an operat- ing ratio the ratio of revenue expenditure to A5C.12 Mumbai was the first City Corporation receipts between 0.99 and 1.02, which is indicative to adopt the concept of a development plan. The of good financial health of the municipality. One of land in the industrial zone are allowed to be used the major revenue stream for BMC is octroi (tax) for residential/commercial purposes. The overall collection, levied on goods coming into the city, land use breakdown includes 44 percent residential which quadrupled from Rs 6.4 million in 1993 to Rs area, 3 percent commercial area, 12 percent indus- 25.6 million, and taxes on land and buildings also trial area, and 41 percent no development zone. The showed a steep increase in the same time period (as land-use structure of Mumbai has undergone major per the detailed financial operating plan of BMC). changes in the past. Massive housing developments Both, octroi and taxes on land and buildings are also have been undertaken in previously non-urban belts good indicators of economic growth of the city. along the western corridors. New district centers have emerged in the northern suburbs (BMC 2003). A5C.14 Capital expenditure on water works, drainage/sewage, solid waste management, roads, A5C.13 The municipal limits of Mumbai come storm water drains, and street lighting more than under the jurisdiction of the Brihan Mumbai Munici- doubled in between 1993 and 2002. The total (capi- pal Corporation (BMC). According to data between tal) expenditure on infrastructure increased Figure A5C.13: Variation in monthly averages of SO and NO 2 2 92 Annex 5C : Mumbai Figure A5C.14: Overall variation in RSPM with seasons (Monsoon: June to September; Summer: March to May; Winter: October to February) significantly between 1997 and 1999. The major instruments, the changes made were significant: capital-intensive infrastructure development that the first generation samplers were developed in- happened during that period in Mumbai was the house by NEERI while the second generation construction of 36-odd flyovers in the metropoli- instruments were manufactured by a commercial tan area. This possibly contributed significantly to manufacturer. elevated concentrations of particulate matter on account of traffic congestions and the construction A5C.16 The average levels of SO2 and NO2 are related activities. shown in Figures A5C.7 and A5C.8. SO2 tended to decline in all three areas; the levels of NO2 showed 8. Air Quality greater fluctuation. A declining trend in NO2 is not discernible until 1999 in the industrial and commer- Trends cial areas. A5C.15 Figures A5C.5 and A5C.6 show annual A5C.17 Figure A5C.9 shows the monthly aver- ambient RSPM and SPM concentrations, respectively, ages of RSPM between 1994 and 2002 in the three between 1993 and 2002 in the three different land- areas (1993 is excluded because of missing data for use areas. The RSPM levels declined between 1999 residential area). The maximum and minimum and 2002, with the levels in 2002 in all the three values vary quite a lot between the three areas. The land-use areas reaching close to the annual national overall maximum was just above 400 µg/m³, and the ambient standard of 60 µg/m³ for residential areas. minimum about 20 µg/m³. However, the RSPM The largest improvement in RSPM levels took place levels were not found to be significantly different in the industrial area, indicating a possible systematic between the three areas (at the 10 percent level) as attrition of industrial source(s). However, SPM indicated by analysis of variance results, suggesting levels did not mirror the variation in RSPM, that sources of RSPM may be similar in all three suggesting that the sources of RSPM and SPM areas. may be different in Mumbai, especially given the large flyover construction program that was A5C.18 It is important to note that the sudden undertaken in Mumbai till 2000 (see Box A5C.6), decline in the RSPM concentration in 2001 in the or possible data quality problems. However, residential area coincides with a shift in the monitor- these findings need to be interpreted with caution ing station from Bandra to Worli (see Annex 2). given the 1998-1999 period also coincided with the Further examination by overlaying the information replacement of RSPM samplers. Although NEERI in Box A5C.6 on the RSPM data indicates that the reports cross-calibration with the previous samplers relatively high values between 1997 and 2000, as before deploying new ones to ensure consistency seen in Figure A5C.5, coincides with the period in measurements between the old and new of a construction program for 36-odd flyovers. 93 For a Breath of Fresh Air 9. Seasonality A5C.22 The effects of rain and wind are evident in Figures A5C.11 and A5C.12. The same pattern A5C.19 From empirical evidence, it is known of variation is seen with gaseous pollutants in Figure that the two parameters that directly impact A5C.13, with the highs occurring in the winter concentrations of pollutants are rainfall and wind months and lows occurring in the monsoon months. speed, since higher precipitation leads to lower The effect of weather parameters was found to be concentrations of pollutants, and higher wind consistent across all three land-use areas. speeds lead to better dispersion. A5C.23 The commonly used definitions of A5C.20 As can be seen in Figure A5C.10, rela- summer, winter, and monsoon seasons were used tively low wind speeds and rainfall are encountered to aggregate the months, in order to further explore in Mumbai for almost four months in a year. This is any seasonality in the variation of RSPM levels. likely to lead to poor dispersion over that time period. An empirical classification (see endnote #21) A5C.24 Figure A5C.14 clearly shows effects of of the dispersion conditions was attempted based on seasons. The averages for monsoon, summer, and the meteorological data. It was found that good winter months were 99 µg/m³, 119 µg/m³, and 167 dispersion conditions exist from June to October and µg/m³, respectively. However, since the summer, in January, moderate dispersion conditions exist in winter, and monsoon classification is subjective in April and May, and poor dispersion conditions exist nature, the empirically derived classification of in the rest of the year. Further evidence of the effects dispersion conditions discussed at the beginning of of relatively better dispersion over longer period of this section was also used to assess seasonality in time can be seen in the monthly variation of RSPM. RSPM variation. It was found that the effects of moderate dispersion classification were not dis- A5C.21 While Mumbai gets a good strong sea tinctly different from those of good dispersion breeze since it is located by the sea, it is worth noting classification. The RSPM averages for the good, that the monsoon months are also the months that moderate, and poor dispersion classifications were experience strong winds, indicating that these two 119 µg/m³ , 102 µg/m³, and 167 µg/m³, respectively. weather variables, both of which have an inverse relationship with air pollution, are in-phase in Mumbai. 94 Annex 5D Hyderabad 1. Geographical Location India. The contour level falls gradually from west to east creating almost a trough near the Musi river A5D.01 The twin cities of Hyderabad and which runs through the city (GoAP 2003a). Figure Secunderabad are located in Hyderabad district in A5D.1 shows approximate location of NEERI moni- the heart of the Deccan plateau in south-central toring stations at Hyderabad. Figure A5D.1: Approximate location of NEERI pollution monitoring stations in industrial, commercial and residential areas 95 For a Breath of Fresh Air Figure A5D.2: Population in Hyderabad Development Area between 1991 and 2001 (Pandey 2003) 2. Demography tion of Hyderabad (MCH) governs only Hyderabad district but accounts for approxi- A5D.02 The Hyderabad Development Area mately 65 percent of the population of the includes 80 percent of the district of Hyderabad Development Area. (173 km2), 20 percent of neighboring Ranga Reddy District (1526 km2), and 2 percent of neighboring A5D.03 Figure A5D.2 shows the historical Medak District (166 km2). The Municipal Corpora- population in the Hyderabad Development Table A5D.1: Overall monthly average values of key meteorological parameters Mean Mean Mean No. of Relative Total Max. Min. Wind Calm Days Humidity % Rainfall Temp. Temp. Speed (mm) (0C) (0C) (kmph) 0830 hrs 1730 hrs 0830 hrs 1730 hrs January 29.2 15.7 6.3 16 2 73 41 11 February 32.0 18.2 7.2 10 1 64 32 10 March 35.8 21.5 7.5 10 2 53 27 8 April 37.4 24.4 8.0 5 2 53 30 22 May 39.4 26.4 11.1 2 2 51 31 36 June 35.4 24.2 15.3 1 1 70 53 102 July 32.0 22.8 1 1 16.1 79 64 160 August 30.2 22.1 1 2 13.9 83 71 237 September 31.3 22.1 5 3 10.3 79 65 111 October 30.8 20.8 12 3 7.2 76 60 125 November 12 2 29.7 17.6 7.0 69 50 29 December 19 3 28.8 14.4 5.9 69 42 8 Total/year 93 25 859 Note: Data on wind speed and relative humidity is collected twice a day. 96 Annex 5D : Hyderabad Figure A5D.3: Growth in commercial motor vehicles in Hyderabad (MoRTH 2003, GoAP 2003b) Figure A5D.4: Growth in private motor vehicles in Hyderabad (MoRTH 2003, GoAP 2003b) Area, which grew from 4.34 million in 1991 to 5.53 winter (November to January) to 39.4 °C in summer million in 2001 at an annual growth rate of 2.41 (February to May). The city receives rainfall in the percent. The decadal growth rate was more than 27 monsoon months (June to October). Calm conditions percent. At the same time, the population of MCH mostly prevail in the morning hours, but the winds grew from 3.04 million to 3.43 million at a cumula- pick up later in the day. The period of relatively high tive growth rate of 18 percent, indicating that the wind speeds lasts for 10 months, barring December development area and surrounding municipality and January. The relative humidity is higher in the absorbed a large part of the growth in population. morning hours, complementary to calm conditions. The population density of Hyderabad in 2001 was 19,930 persons per km2. 4. Economy 3. Climate A5D.05 The percentage of main workers in the population of the MCH area in 2001 was 26.9 per- A5D.04 The weather profile between 1991 and cent with a slight increase from the previous census 2002 in Hyderabad is shown in Table A5D.1. Hot of 1991 in which it was 26.6 percent. Almost 72 steppe type climate prevails over Hyderabad. The percent of the workforce in 1994 was employed in mean monthly temperature varies from 14.4 °C in the service sector and about 18 percent in the 97 For a Breath of Fresh Air Figure A5D.5: Trend in RSPM levels in three areas of Hyderabad Note: The dotted line shows the annual national ambient standard of 60 µg/m³ for RSPM for residential areas. manufacturing industry. Of the rest, 8 percent 12 percent and three-wheelers another 5.5 percent of were employed in the construction sector and the the vehicle population. The remaining population is remaining in household industries and the agricultural made up of taxis, trucks and lorries, buses, and jeeps. and allied sector. The annual per capita income However, like other cities in India, there is a signifi- (1997-98) was Rs 10,590. cant difference between the number of vehicles registered and the actual number on the road, since 5. Industry vehicles retired or moved out of the city A5D.06 The city is characterized by an economic are not de-registered. base that is now dominated by the service industry, A5D.08 Motor vehicles are considered the major primarily in the information technology (IT) sector. source of air pollution in Hyderabad, and most of the Earlier it had a highly diversified economic base with actions taken against air pollution in the city have trading and manufacturing being the most important focused on the transport sector, as illustrated in Box activities. The major industry sector besides IT A5D.7. Traffic movement at the major traffic junc- includes cotton and silk textiles, cigarettes, paper, tion in the commercial area of Abids was rational- pottery, tourism, health services, electronics, and ized in 2001 in order to ease congestion. biotechnology. Amongst the polluting industries that remain in and around Hyderabad, pharmaceutical A5D.09 A public interest litigation was filed in and bulk drugs, textiles, chemicals, and leather the Andhra Pradesh Court on vehicular pollution in dominate. As per the categorization of the Andhra Hyderabad city in 1997, following a study that Pradesh (AP) Pollution Control Board, there were showed that 29 percent of those working at high 333 air polluting industries in the larger Hyderabad traffic junctions were affected by respiratory prob- Urban Development Area in early 2004, out of which lems. According to the SPCB, there is relatively 301 were meeting the emission norms (GoAP 2003a). higher increase in the number of respiratory patients than in the increase in human population or the 6. Transport number of vehicles (GoAP 2003a). A5D.07 The number of vehicles more than 7. Urban doubled between 1992 and 2002, from .48 million to 1.08 million. The vehicular growth trends are shown A5D.10 The existing land use for MCH area in in Figures A5D.3 and A5D.4. The percentage of two- 2000-2001 was 75.2 km2 of residential and plotted wheelers in Hyderabad makes up around 75 percent area, 20.6 km2 of commercial including mixed-use of the vehicle population. Motor cars make up about area, 3.07 km2 of manufacturing area, 23.8 km2 of 98 Annex 5D : Hyderabad Figure A5D.6: Trend in SPM levels in three areas of Hyderabad Note: The dotted line shows the annual national ambient standard of 140µg/m³ for SPM for residential areas.The three monitoring stations missed a few months of data collection in 1995­1997. public and semi-public area, 7.6 km2of open area Maintenance of civil works health and sanitation including urban forest area, 20.1 km2of agricultural together accounted for about 75 percent of the and other areas, 8.6 km2of water bodies (8.63), and revenue expenditure. Overall, the MCH had revenue 13.5 km2area for transport and communications savings of Rs 623 million (MCH 2003). (HUDA 2003). Compared to 1998, the residential area increased by almost 9 percent and the commer- A5D.12 Hyderabad also won the Cleanest City cial area increased by almost 2.7 percent. award given by the Housing and Urban Develop- ment Corporation (HUDCO) and Government of A5D.11 The revenue of MCH increased from India in 1999 and 2000. A noticeable feature of Rs 1.8 billion in 1998 to Rs 2.6 billion in 2001. The Hyderabad municipality is the sweeping of all major revenue receipts included taxes, fees and user roads at night, which is likely to lower ambient charges, and income from estates and grants. concentrations of resuspended dust. Figure A5D.7: Trend in SO levels in Hyderabad 2 Note: The annual national ambient standard for SO in residential area is 60 µg/m³. 2 99 For a Breath of Fresh Air Box A5D.7: Major Technical and Policy Interventions in Hyderabad (1991-2004) that Impacted Air Quality Year Action Year Action 1991 1) Mass emission norms for new petrol 2) Widening of roads undertaken vehicles introduced on 1/4/1991 3) Emission norms for LPG vehicles 1992 1) Mass emission norms for new diesel introduced from May vehicles introduced on 1/4/1992 2002 1) Stopping of permit to new autorickshaws 1993 N.A. 2) Emission norms for CNG vehicles (all 1994 N.A. categories) and LPG vehicles (heavy 1995 N.A. duty) with effect from May 2002 and 1996 1) Mass emission norms for new vehicles October 2002 respectively. made more stringent on 1/4/1996 2003 1) Bharat Stage-II (Euro-II) emission norms 2) Fuel quality: 0.5% S diesel mandated in for both new non-commercial and December commercial vehicles mandated from 3) Benzene in petrol reduced to 5% 1/4/2003 4) Lead in petrol reduced to 0.15 gm/litre 2) Phasing out of old Government vehicles 1997 N.A. started 1998 1) Emission Norms for catalytic converter 3) Low Sulfur petrol & diesel (Sulfur 0.05%) fitted passenger vehicles from 1/4/1998 mandated 2) Low smoke 2T oil mandated 4) 5% ethanol blended gasoline mandated 3) Catalytic converters for 4-wheeled petrol from January, 2003 driven vehicles mandated with effect 5) Introduction of LPG stations from April. 6) Insistence of the local administration to 4) Unleaded petrol introduced to complement provide cover at the construction sites to the introduction of catalytic converters avoid dispersion of dust 1999 1) Pre-mixed 2T engine oil made mandatory 7) Supreme court order for preparation of for 2-wheelers with effect from 1/4/1999 Action Plan for lowering ambient RSPM 2000 1) India Stage-I (Euro-I) emissions norms levels on 14/8/2003 mandated for all category of new vehicles 2004 1) Introduction of Multi Modal Transport with effect from 1/4/2000 System 2) Fuel quality: 0.25% S diesel mandated 2) Notification issued by Government of from 01/04/2000 India revising the emission norms of in- 3) Leaded petrol phased out by February use vehicles with effect from 1/10/2004. 4) Construction of flyovers started. 5) Action against air-polluting industries N.A. ­ Not applicable initiated. Notes: (1) Interventions common to all five cities are 2001 1) Construction of by-pass roads for heavy shown in Italics. However, all common interventions were vehicles started not implemented in the same year in all cities. (2) Dates to be read as day/month/year. 100 Annex 5D : Hyderabad Figure A5D.8: Trend in NO levels in Hyderabad 2 Note: The annual national ambient standard for NO in residential areas is 60 µg/m³. 2 8. Urban Air Quality 2001 and 2002 data at the industrial site with those collected at another industrial site not too Trends far from it, SPCB's Uppal site, shows that the A5D.13 Figures A5D.5 and A5D.6 show RSPM concentrations measured by NEERI are much and SPM levels, respectively, between 1993 and lower. The overall trends in monthly averages 2002 in the three different land-use areas. The data from the NEERI and SPCB monitoring locations, collected at the monitoring stations run by NEERI shown in Figure A3A.12 in Annex 3A, tracked show that the RSPM levels in 2001 and 2002 were each other reasonably well till 2002 after which generally below the annual national ambient there seems to be a greater divergence in values. standards for residential areas, and SPM levels Regression analysis between the two datasets were also meeting the annual residential area returned a R-squared value of 0.19. However, standard in 2002 except for the commercial moni- this result needs to be interpreted with caution toring site. However, it should be noted (see Annex since averaging across different monitoring 3A) that the data collected by SPCB do not show stations can lead to calculation artifacts. this trend. On the contrary, the ambient RSPM A5D.15 The average levels of SO2 and NO2 are concentrations rose between 2001 and 2002, and shown in Figures A5D.7 and A5D.8. The values of remained above 80 µg/m³ on average at SPCB's both gaseous pollutants remained well below the three monitoring sites. national annual ambient standards for all land-use A5D.14 The sudden increase in levels in 1996 areas with no significant trends discernible. The could be explained by the increased traffic conges- commercial area showed the largest fluctuations tion near the commercial area, which had become a for both gases, whereas the other two sites major traffic junction, until the traffic flow was showed fairly steady SO2 and NO2 concentrations rationalized in 2001. It is worth noting that the over the last 10 years. levels of both pollutants in the industrial area were A5D.16 Figure A5D.9 shows the monthly the lowest, and well below the national ambient average concentration of RSPM at the three standards for industrial areas in 2001 and 2002. In monitoring locations in Hyderabad. As can be seen one sense, this is not surprising given that non- in Figure A5D.9, RSPM increased between 1993 polluting industries occupy the area surrounding and 1996 and showed a decreasing trend after the site (see Annex 2). However, comparison of the 2000 in all three areas. The overall maximum did 101 For a Breath of Fresh Air Figure A5D.9: Trend in RSPM monthly concentration at the residential, commercial, and industrial area monitoring locations (trend line shown) 102 Annex 5D : Hyderabad Figure A5D.10: Monthly average total rainfall and mean wind speed Figure A5D.11: Monthly average total rainfall and RSPM concentration Figure A5D.12: Monthly mean wind speed and RSPM concentration 103 For a Breath of Fresh Air Figure A5D.13: Variation in monthly averages of SO and NO 2 2 Figure A5D.14: Overall variation in RSPM with seasons (Monsoon: June to October; Summer: February to May; Winter: November to January) not exceed 240 µg/m³, and the minima went down 10 months of the year. This is likely to lead to to approximately 20 µg/m³. The RSPM levels were better dispersion conditions over a longer period found to be significantly different between the of time. An empirical classification (see endnote # three areas (at the 10 percent level) as indicated by 21) of the dispersion conditions was attempted analysis of variance results, suggesting that sources based on the meteorological data. It was found of RSPM may not be the same in three areas. that good dispersion conditions exist from May to October, while moderate dispersion conditions 9. Seasonality exist between January and April and November A5D.17 From empirical evidence, it is known and December. Further evidence of the effects of that the two parameters which directly impact relatively better dispersion over a longer period concentration of pollutant are rainfall and wind can be seen in the monthly variation of RSPM. speed, since higher precipitation leads to lower A5D.19 Figures A5D.11 and A5D.12 show that concentration of pollutants, and higher wind there is hardly any impact of the two meteorological speeds leads to better dispersion. parameters--total rainfall and mean wind speed-- A5D.18 Figure A5D.10 shows that relatively on ambient RSPM levels. However, if the RSPM level higher wind speed stays in Hyderabad for about in August is excluded, a slight effect of wind speed and rainfall could be discerned. From 104 Annex 5D : Hyderabad Figure A5D.13, it is seen that SO2 and NO2 show and 88 µg/m³, respectively. However, since the a small effect of meteorological factors, although summer, winter, and monsoon classification the levels remain quite low in each month. shown is subjective in nature, the empirically derived classification of dispersion conditions A5D.20 The commonly used definitions of discussed at the beginning of this section was summer, winter, and monsoon seasons were used to also used to assess seasonality in RSPM variation. aggregate the months, in order to further explore It was found that there was a small effect of the any seasonality in the variation of RSPM levels. two dispersion classifications. The RSPM A5D.21 AsFigureA5D.14shows,theeffectof averages for the good and moderate dispersion seasonsisnotstrong.Theaveragesformonsoon, classifications were 74 µg/m³ and 87 µg/m³, summer,andwintermonthswere74µg/m³,82µg/m³, respectively. 105 For a Breath of Fresh Air Annex 5E Chennai 1. Geographical Location along the coast. The city is intersected by two rivers, Coovam and Adyar, which pass through the center A5E.01 Chennai district is situated in the north and north of the city, respectively, and enter the sea. east of Tamil Nadu on the coast of Bay of Bengal. It Chennai city is a separate district, covering the stretches over an area of 174 km2 , for nearly 25.6 km Chennai Municipal Corporation area (NEERI 2003b). Figure A5E.1: Approximate location of NEERI monitoring stations in industrial, commercial and residential areas 106 Annex 5E - Chennai Figure A5E.1 shows a map of Chennai with location Chennai has a hot climate, which can be termed as of NEERI pollution monitoring stations. tropical maritime monsoon type. The mean monthly temperature ranges from 21.3°C (December to 2. Demography February) to 37.2 °C in summer months (March to A5E.02 In 1991, the total population of Chennai August), the city receives quite a lot of rain during Metropolitan area was 5.42 million which the monsoon (September to November) months. increased to 6.42 million in 2001 as a decadal Calm conditions mostly prevail in the morning growth rate of 18.4 percent. Figure A5E.2 shows the hours, but then the winds pick up and average wind historical population in the Chennai Metropolitan speeds are on the higher side. The relative humidity area between 1991 and 2001. The population in is higher in the morning hours, complementary to Chennai city increased from 3.84 to 4.21 million calm conditions. during the same period, a growth of less than 4. Industry 1 percent per annum. Most growth occurred in the periphery and small and medium towns in the A5E.05 Chennai has a mix of large, medium and Metropolitan Area. small-scale industries. There are about 35 medium and large-scale industries manufacturing a variety A5E.03 Chennai city was the most densely of products. The development of small-scale indus- populated district in the state with 24,138 persons tries has been influenced by the needs of medium per km2compared to a 22,077 persons per km2in and large industrial units. As of March 1991, there 1991. While this was mainly due to migration from were 16,326 permanently registered small-scale units other parts of the state and country, it also shows in Chennai which grew to 40,300 by March 2001. that much of the population growth took place in the These units are in the fields of automobile, hosiery larger metropolitan area which reflected a cumula- and readymade garments, paper and paper products, tive growth rate of 18.4 percent. machinery parts, metal products, food processing, rubber and plastic products, and chemicals. An 3. Climate information technology park named "Tidel Park" A5E.04 The weather profile for Chennai (Table has also been set up by the state government A5E.1), is based on 12 years of data (1991 and 2002). across eight acres of land to boost the electronic Figure A5E.2: Population growth in Chennai metropolitan area between the last two census periods (Pandey 2003) 107 For a Breath of Fresh Air telecommunications and information technology 3 million. Higher employment is generated in textile industry (GoI 2002b). and textile-based industries and auto ancillaries. The contribution of the small-scale sector to the economy A5E.06 In terms of air pollution, industry is not can be gauged from the fact that it accounts for 40 much of a concern other than the Chennai Port Trust percent of the industrial production, 35 percent of and the power generating units of the Power direct exports, and 45 percent of overall exports. The Corporation in Chennai due to good dispersion average annual per capita income was Rs 36,138 conditions. In addition, there are more than 900 (1998-99). diesel generators, which are used if and when there are power outages. The two principal operations of 6. Transport the Port Trust which lead to particulate pollution are handling of coal and export of iron ore. However, all A5E.08 Chennai has a combination of three the coal handling activities are in the process of modes of public transport within the city: Metropoli- being shifted out of the port by end of 2004. The ore tan Rapid Transport System (MRTS), suburban handling is also supposed to be shifted out by end railway, and Transport Corporation buses. The buses 2005. The Power Generating units use LSHS (low run by the Metropolitan Transport Corporation are sulfur heavy stock) and naptha as fuels, and are the predominant mode of public transport in monitored quite rigorously for air emissions Chennai. The rail and roads to a large extent run in (GoTN 2004). parallel. 5. Economy A5E.9 Vehicular growth in Chennai as shown in Figure A5E.3 and A5E.4 shows that there has been A5E.07 Small-scale industries play a major role in a steady increase of vehicles in the city over the past the economy of the state. The employment generated two decades. The number of registered vehicles in directly by the registered small-scale industrial units 2002 was 1.35 million. About 75 percent of vehicles in Chennai as of March 2001 numbered almost were two-wheelers, 18 percent were cars, 3 percent Table A5E.1: Overall monthly average values of key meteorological parameters Mean Mean Mean No. of Relative Total Min. Max. Wind Calm Days Humidity % Rainfall Temp. Temp. Speed (mm) (0C) (0C) (kmph) 0830 hrs 1730 hrs 0830 hrs 1730 hrs January 29.5 21.3 4.4 14 3 83 68 22 February 31.2 22.5 4.7 10 2 81 66 23 March 33.2 24.0 5.0 6 1 78 66 3 April 34.7 26.6 5.8 5 1 74 69 21 May 37.2 28.0 6.8 3 1 67 66 30 June 36.6 27.4 7.1 2 2 64 62 81 July 35.4 26.6 5.6 4 3 68 63 99 August 34.3 25.9 5.0 3 5 72 64 109 September 34.6 25.6 5.2 5 5 76 70 142 October 32.1 24.5 3.9 11 7 83 76 297 November 30.0 23.3 4.6 9 4 85 78 374 December 28.9 21.5 5.4 9 1 81 70 139 Total/year 79 34 1342 Note: Data on wind speed and relative humidity is collected twice a day. 108 Annex 5E - Chennai Figure A5E.3: Growth in commercial motor vehicles in Chennai (MoRTH 2003) Figure A5E.4: Growth in private motor vehicles in Chennai (MoRTH 2003) were three-wheelers, 2 percent were goods the commercial monitoring location, and diverted carriages, and the rest were a mix of buses, taxis them to a new terminal on the outskirts of the city. and jeeps. The average growth rate in vehicle population has been between 7 and 9 percent per 7. Urban annum in the last decade (MoRTH, 2003). A5E.12 The geographical area of Chennai City A5E.10 The unprecedented increase in the was 174 km2 . The green cover decreased from 1.7 number of private vehicles, particularly two- percent of the geographical area in 1998-99 to 0.7 wheelers, has led to congestion on the roads and percent in 1999-2000 (Chennai Corporation 2004). is a likely cause of air pollution in the city. Most Chennai has reached a level of saturation in terms of the actions being undertaken and proposed to of growth and most changes have been in terms address urban air pollution, target vehicular of land-use or by way of densification. As outlined emissions as shown in Box A5E.8. above, most growth has been in the metropolitan area. This growth has been radial, following the A5E.11 In 2002, the government banned the major road-rail network. Unlike other cities such as entry of suburban buses into the city center, close to 109 For a Breath of Fresh Air Figure A5E.5: Trend in RSPM levels in three areas of Chennai Note: The dotted line shows the annual national ambient standard of 60 mg/m³ for RSPM for residential areas. Delhi or Bangalore, the Chennai Metropolitan Corporation 2004). Capital expenditure in 2002-03 Development Authority regulates growth rather amounted to Rs 1.6 billion with maximum expendi- than developing land. ture on roads which was 1 billion. In the late 1990s, Chennai invested substantial resources in improving A5E.13 The Chennai Corporation is the oldest vehicle circulation by implementing 12 flyover municipal corporation in the country formed in schemes apart from major road rehabilitation works. 1688. The revenue receipts in 2003-04 were Rs 5.1 billion in which the major contribution was 8. Urban Air Quality from the property tax (Rs 2.4 billion). The revenue expenditure in 2003-04 was Rs 5.1 billion. The Trends operating expenses were Rs 5.7 billion and repairs and maintenance cost was Rs 3.5 billion (Chennai A5E.14 Figures A5E.5 and A5E.6 show annual Figure A5E.6: Trend in SPM levels in three areas of Chennai Note: The dotted line shows the annual national ambient standard of 140µg/m³ for SPM for residential areas. No SPM data during the first 7 months of 1993 at the industrial site. 110 Annex 5E - Chennai Figure A5E.7: Trend in SO levels in Chennai 2 Note: The annual national ambient standard for SO in residential areas is 60 µg/m³. 2 ambient RSPM and SPM levels, respectively, between monitoring location for 2001 and 2002, which is 1993 and 2002 in the three different land-use areas. located in the same area (Thiruvottiyur) as the The data collected by NEERI indicate that the levels NEERI industrial monitoring station, it was for both pollutants in all three areas were below the found that the NEERI values were much lower. annual national ambient standards in 2002. Increased Furthermore, the number of data points collected construction activities in the city center might by SPCB was more than double NEERI's, account for the sudden increase in levels in 1997 at rendering greater statistical significance to the commercial monitoring site. The decrease in SPCB's data than NEERI's. Given the available annual average RSPM to below the national data, the observation that all the three sites by standards in 2002 coincided with the ban on entry 2002 were in compliance with the annual National of suburban buses into the city center. However, Ambient Air Quality Standard for residential when compared with data from the SPCB industrial areas does not appear to hold across the city. Figure A5E.8: Trend in NO levels in Chennai 2 Note: The annual national ambient standard for NO in residential areas is 60 µg/m³. 2 111 For a Breath of Fresh Air Box A5E.8: Major Technical and Policy Interventions in Chennai (1991-2004) that Impacted Air Quality Year Action 1991 1) Mass emission norms for new petrol vehicles introduced on 1/4/1991 1992 1) Mass emission norms for new diesel vehicles introduced on 1/4/1992 1993 N.A. 1994 1) Lead in petrol reduced to 0.15 gm/litre in June 1995 1) Catalytic converters for 4-wheeled petrol driven vehicles mandated with effect from April 2) Unleaded petrol introduced to compliment the introduction of catalytic converters 1996 1) Mass emission norms for new vehicles made more stringent on 1/4/1996 2) Fuel quality: 0.5% S diesel mandated in December 3) Benzene in petrol reduced to 5% 1997 1) Emission norms for in-use vehicles implemented from 1/1/1997 1998 1) Emission norms for catalytic converter fitted passenger vehicles from 1/4/1998 2) Low smoke 2T oil mandated 1999 1) Pre-mixed 2T engine oil made mandatory for two-wheelers with effect from 1/4/1999 2000 1) India Stage-I (Euro-I) emission norms mandated for all categories of new vehicles with effect from 1/4/2000 2) Fuel quality: 0.25% S diesel mandated from 01/04/2000 3) Leaded petrol phased out by February 2001 1) Bharat Stage-II (Euro-II) emission norms for all categories of new non-commercial vehicles mandated by July 2) Low Sulfur diesel and petrol (0.05%) mandated from 1/7/2001 3) Emission norms for LPG vehicles introduced from May 2002 1) Bharat Stage-II (Euro-II) emission norms for all new commercial vehicles introduced by October 2) Entry of old buses into the center of the city prohibited, and old buses diverted to new bus terminal at Koyambedu in the outskirts of the city from 18/1/2002. 3) Emission norms for CNG vehicles (all categories) and LPG vehicles (heavy duty) with effect from May 2002 and October 2002 respectively 2003 1) 5% ethanol blended gasoline mandated with effect from January 2) Green tax levied for vehicles more than 15 years old from August 3) Supreme court order for preparation of Action Pan for lowering ambient RSPM levels on 14/8/2003 2004 1) MRTS introduction in selected areas of Chennai since 26/1/2004 2) Notification issued by Government of India revising the emission norms of in-use vehicles with effect from 1/10/2004. N.A. ­ Not applicable Notes: (1) Interventions common to all five cities are shown in Italics. However, all common interventions were not implemented in the same year in all cities. (2) Dates to be read as day/month/year. 112 Annex 5E - Chennai Figure A5E.9: Trend in RSPM monthly concentration at the residential, commercial and industrial area monitoring locations 113 For a Breath of Fresh Air Figure A5E.10: Monthly average total rainfall and mean wind speed A5E.15 The average levels of SO2 and NO2 are 9. Seasonality shown in Figures A5E.7 and A5E.8. The levels observed were well below their respective annual A5E.17 From empirical evidence, it is known national standards and there was no particular that the two parameters which directly impact trend in their variation. The levels of NO2 were concentration of pollutant are rainfall and wind lower in the period 1996 -98 than in other years. speed, since higher precipitation leads to lower concentration of pollutants, and higher wind A5E.16 Figure A5E.9 shows the monthly speeds leads to better dispersion. averages of RSPM in the three areas over the years. The maximum and minimum values did not vary A5E.18 As Figure A5E.10 shows, higher precipi- too much between the three areas. The overall tation and higher wind speed seasons follow each maximum was just above 240 µg/m³, and the other sequentially in Chennai. Such a situation is minimum went down to approximately 20 µg/m³. likely to lead to better dispersion conditions over a Further, the RSPM levels were found to be longer period of time. An empirical classification significantly different between the three areas (see endnote # 21) of the dispersion conditions was (at the 10 percent level) as indicated by analysis attempted for all the cities based on the meteoro- of variance results, suggesting that the sources of logical data. It was found that good dispersion RSPM may not be similar across the three areas. conditions exist from June to December, while Figure A5E.11: Monthly average total rainfall and RSPM concentration 114 Annex 5E - Chennai Figure A5E.12: Monthly mean wind speed and RSPM concentration Figure A5E.13: Variation in monthly averages of SO and NO 2 2 Figure A5E.14: Overall variation in RSPM with seasons (Monsoon: September to November; Summer: March to August; Winter: December to February) 115 For a Breath of Fresh Air moderate dispersion conditions exist between any seasonality in the variation of RSPM levels. January and May. Further evidence of the effects of better dispersion over a longer period of time can be A5E.22 As Figure A5E.14 shows, there are no seen in the monthly variation ofRSPM. discernible effect of seasons. The averages for mon- soon, summer, and winter months were 75 µg/m³, A5E.19 As seen in Figures A5E.11 and A5E.12, 68 µg/m³, and 77 µg/m³, respectively. there is hardly any impact of the two meteoro- logical parameters--total rainfall and mean wind A5E.23 Since the summer, winter, and monsoon speed--on RSPM levels. classification is subjective in nature, the empiri- cally derived classification of dispersion A5E.20 The variation in monthly averages of SO2 conditions discussed at the beginning of this and NO2 is shown in Figure A5E.13. It can be seen section was used to assess seasonality in RSPM that NO2 does show a slight effect of wind speed variation. It was found that there was no effect as it decreases between January and April as wind of the two dispersion classifications. The RSPM speeds increase. averages for the good and moderate dispersion classifications were 72 µg/m³ and 71 µg/m³, A5E.21 The commonly used definitions of sum- respectively. mer, winter, and monsoon seasons were used to aggregate the months, in order to further explore 116 Annex 6 Health Cost of Exposure to RSPM Pollution A6.01 A large number of epidemiological A6.03 Similar to estimates of mortality, studies have linked exposure to PM10 concentra- concentration-response functions are also derived tions to health end-points, making it possible to for morbidity impacts such as respiratory hospital conduct a meta-analysis of the different studies. admissions, chronic bronchitis, cough, asthma, etc. Given that most of the studies have been con- However, there are much fewer concentration- ducted in industrialized countries, application of response functions available for morbidity end- their results to a developing country like India points from developing countries than for mortal- raises legitimate concerns. However, recent studies ity. The pooled estimates used for morbidity end- undertaken in Mexico city, Santiago, and Bangkok point from an earlier World Bank review (Lvovsky provide support to the case for extrapolation of et al, 2000) are used in this study. It needs to be results. Further, the fact that epidemiological borne in mind that these estimates implicitly incor- studies provide information on percent change in porate the baseline incidence of the morbidity end- mortality due to absolute change in ambient levels, points (such as chronic bronchitis) in the location makes a case for extrapolation since the change is where the studies were conducted, and as a result predicted with respect to the area-specific baseline they are likely to underestimate the morbidity mortality. Furthermore, meta-analysis enables one impacts for Indian cities (since the baseline inci- to derive "best estimates" of the relationship dence in Indian cities can be expected to be higher between air pollution and health end-points, which than in cities in industrialized countries where most can be applied with greater confidence to other of the studies have been conducted). situations/countries/cities than an estimate from an individual study. While epidemiological studies Methodology for the Estimation of Health using PM10 or RSPM do not exist in India, a study Impacts in Delhi found an association between SPM levels A6.04 Health impacts are calculated using the and changes in mortality rates (Cropper et al, equation 1997). In this assessment, both meta-analytical estimates from Lvovsky et al (2000) and estimates Hi = bij * Aj * P ....................................... (1) from Cropper et al (1997) are used, as explained below. Where: stands for change; Hi is the health impact of type i per year; bij stands for the slope Estimates for Mortality and Morbidity of the concentration response function of health effect i for exposure to pollutant j per year; Aj is A6.02 Hence, this study uses the pooled the ambient concentration of pollutant j and P is estimate of 0.84 percent change in all-cause mortal- the population exposed to the pollutants. ity relative to a 10 µg/m³ change in RSPM concen- tration, derived from a number of acute exposures The quantity Aj is defined as studied in an earlier World Bank review (Lvovsky 1 0 et al. 2000), and summarized in Table A6.1. How- Aj = max [ 0, Aj ­ max( Aj , Sj )] ...........(2) ever, this estimate is used only for Mumbai, Where: Aj 1 is the observed concentration; Aj is 0 Hyderabad, and Chennai, since they had levels the background concentration and Sj is the that were close to those recorded in cities from relevant threshold or air quality standard. which the estimates have been derived. In the case of Delhi and Mumbai, which show much higher The coefficients bij used in this study are shown ambient levels, more conservative estimates of 0.42 in Table A6.1 (except that 0.42 was used for from Cropper et at (1997) are used. 23 assessing the impact of mortality in Delhi and Mumbai). It was assumed that the dose-response 117 For a Breath of Fresh Air Table A6.1: Air pollution concentration-response function slope (b ) per 1 µg/m3 change in the mean ij annual level of PM 10 Health Effects Units b ij Mortality Percentage change 0.084 Chronic Bronchitis Per 100,000 adults 6.12 Respiratory Hospital admission Per 100,000 population 1.2 Asthma attack Per 100,000 asthmatics 3,260 Emergency room visits Per 100,000 population 23.54 Restricted day activities Per 100,000 adults 5,700 Lower respiratory illness Per 100,000 Children 169 Respiratory symptoms Per 100,000 adults 18,300 Source: Lvovsky et al. 2000 ambient RSPM concentrations were obtained from A6.07 The major uncertainty that complicates NEERI measurements. the transfer of WTP estimates from industrialized countries to a country like India is the difference A6.05 In the case of mortality, the slope refers to in income levels, since WTP estimates are known percentage change in the mortality and hence it has to rise with income. Hence, VOSL estimates were to be multiplied by the crude mortality rate to get the adjusted for income effects before applying to absolute number of cases. Similarly, the other slopes India (see Table A6.2). have to be multiplied by the adult or children fraction of the population as appropriate. In the case Scaling the Costs of Health Effects of asthma attacks, the slope has to be multiplied by the fraction of asthmatics in the population. A6.08 The WTP-based values for the cost of health effects including the VOSL have been Economic Valuation of the Health Impacts obtained through extensive research mostly in the United States. These values are of course A6.06 Economic valuation of health impacts related to economic indicators and these have to in this study is based on the approach of be scaled by per capita GNP ( Gross National Willingness-to-Pay (WTP) to reduce the risk of Product) for variation in time and space. It has getting sick or dying prematurely. In valuing the been shown that country specific variation can be mortality costs, this approach uses the Value of scaled by using the relation: Statistical Life (VOSL), which is not the same as valuing actual life, but instead involves placing a Log (Vk) = r * log (Yk / Yus) + log(Vus)...............(3) value on reduction in overall risk that people face. The VOSL used in this study was also adjusted for Where: V is the valuation parameter for given a difference between an average number of years country; k and Y is the per capita GNI. Various lost by people dying due to air pollution and an values of r in the range of 0.4 to 1.2 have been average number of years lost by people surveyed reported but it has been observed that r=1 by the labor market studies, on which the majority provides conservative estimates (Lvovsky et al., of VOSL estimates are based. Valuation of morbid- 2000). With this approximation, the relation for ity in this report is also based on WTP estimates for any given city (C) becomes linearized to: avoiding suffering and inconvenience associated with an illness of a given severity (see Lvosky VC = Vus * (YC / Yus)..........................................(4) et al 2000 for details). This simplified relation is used in the calculations here and are shown in Table A6.2 (next page) . 118 Annex 6 : Health Cost of Exposure to RSPM Pollution Table A6.2: Willingness-to-Pay based health effect costs per case derived for in Delhi Health Effects WTP US (1990) India (2003) US$ India (2003) Indian Rs. Mortality (VOSL) 1,620,000 34,943 1,572,419 Chronic Bronchitis 195,000 4,206 189,273 Respiratory Hospital admission 4,225 91.13 4,101 Asthma attack 63.00 1.36 61 Emergency room visits 126.00 2.72 122 Restricted day activities 53.00 1.14 51 Lower respiratory illness 44.00 0.95 43 Respiratory symptoms 44.00 0.95 43 Note: GNP per capita for India is $470; GNP per capita for US for 1990 is $ 21790; 1$=Rs.45 Source: Lvovsky et al, 2000 119 For a Breath of Fresh Air Annex 7 Questionnaire Survey for Stakeholder Input The World Bank has been engaged in air quality While in 1995 the World Bank helped provide management issues in India for almost ten years. For impetus to the movement for clean air in urban the first time in 1995, an assessment by the World India, it has made limited efforts to scale-up its Bank assigned monetary values24 to the health engagement on urban air quality management since impacts of urban air pollution ­ an effort that then. With a large and growing portfolio of projects contributed to the movement for clean air in Indian in India ­ across sectors which are likely to present cities. Since then the political economy of decision many challenges and opportunities for addressing making with regards to urban air quality manage- urban air quality concerns ­ it is important for the ment (UAQM) in India has evolved, with an active World Bank to take stock, and understand the key role being played by civil society and judiciary in issues and challenges. That is the purpose of the influencing policy decisions. In recent years, a attached questionnaire, which is part of a study to number of cities have started to clean-up their act! assess the status of Urban Air Quality in India. The However, cities continue to grow in size and popula- study is being conducted by undertaking an assess- tion, with a consequent increase in sources of air ment of air quality in five major cities, namely ­ pollution ­ particularly the number of motor vehicles Delhi, Mumbai, Chennai, Hyderabad and Kolkata. and the vehicle-kilometers traveled. Your time and effort in completing the questionnaire will be appreciated. 120 Questionnaire Name: ............................................................................... Respirable Particulate Matter / PM10 Address: .......................................................................... Carbon Monoxide City: .................................................................................. Sulfur Dioxide Ozone Telephone Number & Email Address: .................... Nitrous Oxides Others (specify) 1. Whom do you work for? 6. In your opinion, in the last 10 years, air quality in your city of residence: Government Private Sector Improved initially then deteriorated Academia Development agency Has been deteriorating NGO Others (specify) Deteriorated initially then improved 2. How would you describe your area of work? (can check more than one box) Has been Improving Policy formulation / analysis Has not changed Others (specify) Legal Issues Enforcement 7. Please rank (on a scale of 1 to 4) the sources of air pollution in your city of residence Information dissemination by sector: Research Air Pollution Related Transport Urban26 Advocacy Others (specify) Industry27 Others (specify) 3. In your opinion, what is the status of air 8. In your opinion, over the last 10 year which pollution in your city of residence? of the following have had positive (+), negative (-), Serious Not Serious or neutral (0) impact on air quality in your city of residence (put a mark in the box; leave blank if the Moderately activity has not occurred in your city or you do not Serious Others (specify) know about it): New roads and flyovers (exclude period of 4. What is your primary source of informa- construction) tion on air pollution in your city of residence? Closure of air polluting industries Newspaper/ Technical reports TV Construction practices in the city Friends & Public Transport Colleagues Other (specify) Relocation of air polluting industry Zoning to segregate different land-uses 5. Please rank25 (on a scale of 1 to 6) the air pollutants of concern in your city of residence: Use of cleaner transport fuels 121 For a Breath of Fresh Air Cleaner fuel use in industry be active in addressing urban air quality concerns in Indian cities? Maintenance of urban infrastructure such as roads and pavements Yes No Phase-out of old vehicles 13. If agencies like the World Bank were to support urban air quality management efforts, Overall shift in nature of industry towards service should they: sector Support the central government Use of cleaner fuels for domestic use Support the state government Newer vehicles on the road Support city /local administration Compliance of industry with regulation Others (specify) Ban on open burning of garbage etc 14. Please rank (on a scale of 1 to 10), where you Others (specify) see the role of agencies like theWorld Bank in urban air quality management : 9. Do you agree that effect on human health is the single, most important manifestation of the Knowledge sharing based on international impact of urban air pollution ? experience Yes No Stand-alone environment projects Others (specify) (e.g. cleaner public transport, industrial emissions control etc.) 10. Please rank (on a scale of 1 to 7), the in- volvement of following in efforts to address air Provision of equipment for air quality quality concerns in your city of residence: monitoring Government Academia Assistance through sectors that impact air quality (e.g. transport, energy, and NGOs Judiciary urban development) Private Sector Others (specify) Formulation of policies, standards and regulations International agencies Training and capacity building 11. Please rank (on a scale of 1 to 8), the im- pediments to effective urban air quality manage- Technical studies ment in your city of residence: Awareness generation Corruption Lack of political will Others (specify) No role Poor laws and regulations A Request (only if you have the time): As part of this Lax standards study we are trying to put together a list of all actions28 that have taken place in your city of resi- Lack of information / knowledge at the dence which may have affected air quality since 1993 decision-making level (along with the year and month). It would be appre- Poor decision-making structure ciated if you could take the time to provide us a list of relevant actions based on your knowledge/ Poor/outdated technologies memory (attach a separate sheet of paper if neces- Others (specify) sary). Your effort will be acknowledged in the study report. 12 Should donor-agencies like the World Bank Thank you ! 122 References Bombay First, 2004, also available at http:// National Ambient Air Quality Monitoring Series www.bombayfirst.org/ NAAQMS/2003-2004, Ministry of Environment and Forests, New Delhi. Brihan Mumbai Municipal Corporation (BMC), 2003, Environment Status of Brihan Mumbai Chennai Corporation, 2004, also available at 2002-2003. http://www.chennaicorporation.com/ Brandon C., and Hommann K., 1995, The Cost Cropper M., and Simon N., 1996, Valuing the of Inaction, Internal document of the South Health Effects of Air Pollution, DEC Note 7, Asia Environment Unit, The World Bank, World Bank, Washington, D.C. Washington D.C. Cropper M., Simon N., Alberini A., and Sharma P. Center for Science and Environment (CSE), K., 1997, The Health Effects of Air Pollution in 2003, Rise, Stabilise, Rise, Down to Earth, Au- Delhi, India, Policy Research Working Paper 1860, gust 31, pp. 43, New Delhi. DECRG, World Bank, Washington D.C. Center for Science and Environment (CSE), Energy Sector Management Assistance Program 2004, Dieselised, Down to Earth Extra, March (ESMAP), 2004, Urban Air Quality Management 31, pp. 64, New Delhi. in South Asia, World Bank, Washington, D.C. Central Pollution Control Board (CPCB), 2000, European Environment Agency (EEA), 1999, A Air Quality Status and Trends in India, Na- Checklist for State of the Environment Reporting, tional ambient air quality monitoring series: Technical Report No. 15, Copenhagen. NAAQMS/14/1999-2000, Ministry of Environ- ment and Forests, New Delhi. Government of Andhra Pradesh (GoAP), 2003a, Air Quality Improvement Action Plan, submitted Central Pollution Control Board (CPCB), 2001a, by the Andhra Pradesh SPCB to CPCB, Delhi. Air Quality in Delhi (1989-2000), National Ambient Air Quality Monitoring Series Government of Andhra Pradesh (GoAP), NAAQMS/17/2000-2001, Ministry of Environ- Hyderabad, 2003b, Department of Transport, ment and Forests, New Delhi. personal communication with World Bank. Central Pollution Control Board (CPCB), 2001b, Government of India (GoI), 1999, A Study on Vehicular Pollution Control in Delhi-Initiatives Planning Norms, Building Bye Laws, Tariff and Impacts, National Ambient Air Quality Structure, Land Assembly & Resource Mobiliza- Monitoring Series NAAQMS/18/2001-2002, tion (of 9 Metropolitan Cities), Town and Country Ministry of Environment and Forests, Delhi Planning Organization, Ministry of Urban Devel- opment, New Delhi. Central Pollution Control Board (CPCB), 2003a, Delhi, available at website: http:// Government of India (GoI), 2002a, Expert Com- www.cpcb.delhi.nic.in/legislation/ch8dec02a.htm mittee on Auto Fuel Policy, Study Reports Volume 1, Urban Road Traffic and Air Pollution in Major Central Pollution Control Board (CPCB), 2003b, Cities, Central Road Research Institute, New Parivesh Newsletter, Ministry of Environment Delhi, in association with National Environmental and Forests, New Delhi. Engineering Research Institute, Nagpur, and Indian Institute of Petroleum, Dehradun. Central Pollution Control Board (CPCB), 2003c, Guidelines for Ambient Air Quality Monitoring, Government of India (GoI), 2002b, Industrial 123 For a Breath of Fresh Air Potential Survey Report ­ Chennai 2001-2002, Ministry of Road Transport and Highways Ministry of Small Scale Industries, also available (MoRTH), New Delhi, 2003, personal communi- at http://www.laghu-udyog.com/publications/ cation with World Bank. traderep/chennai/chennai.htm Municipal Corporation of Hyderabad (MCH), Government of NCT of Delhi (GoNCTD), 2002, 2003, personal communication with World Bank. Economic Survey of Delhi 2001-02, Planning Department, Delhi National Environmental Engineering Research Institute (NEERI), 2002, Ambient Air Quality Government of NCT of Delhi (GoNCTD), 2003, Status for Ten Cities of India, Report No. 11 for also available at website: http:// the Central Pollution Control Board, Nagpur. delhiplanning.nic.in/Write-up/2002-03/volume- I/Industries.pdf National Environmental Engineering Research Institute (NEERI), 2003, Ambient Air Quality Government of Tamil Nadu (GoTN), 2004, Action Status for Ten Cities of India, Report No. 12 for Plan for Air Quality Management in Chennai City, the Central Pollution Control Board, Nagpur. submitted by the Tamil Nadu SPCB to CPCB, Delhi. National Environmental Engineering Research Institute (NEERI), Chennai, 2003b, personal Government of West Bengal (GoWB), 1998, communication. State of the Environment Report, also available at http://kolkata.wb.nic.in/environment/html/ National Environmental Engineering Research Status Of Environment/index.htm Institute (NEERI), 2004, Particulate Matter Reduction Action Plan for Greater Mumbai, Government of West Bengal (GoWB), 2001, Sponsored by MMRDA, Mumbai. Action Plan for Control of Air Pollution in Kolkata and Howrah, submitted by West Ben- Pandey R., 2003, Personal Communication, based gal Pollution Control Board (WBPCB) to CPCB, on census data for 1991 and 2001. Delhi. Pope C. A., and Dockery D. W., 1994, Acute Greenspan Bell R., Mathur K., Narain U., Respiratory Effects of Particulate Air Pollution, Simpson D., 2004, Clearing the Air: How Delhi Annual Review of Public Health, 15 (107-132). Broke the Logjam on Air Quality Reforms, Schwartz J., 1994, Air Pollution and Daily Mortal- Environment, Vol. 46. 3 (22-39). ity: A Review and Meta Analysis, Environmental Hyderabad Urban Development Authority Research, 64 (36-52). (HUDA), 2003, A Plan for Sustainable Develop- Society of Indian Automobile Manufacturers ment ­ Hyderabad 2020, Draft Master Plan for (SIAM), 2004, personal communication with the Hyderabad Metropolitan Area, Hyderabad World Bank. Lvovsky K., Hughes G., Maddison D., Octro B., Tata Energy Research Institute (TERI), 2001a, Pearce D., 2000, Environmental Costs of Fossil Review of past and on going work on urban air Fuels : A Rapid Assessment Methodology with quality in India, a report prepared for the Application to Six Cities, World Bank, Washing- World Bank. Available online at http:// ton D.C. www.worldbank.org/sarurbanair Mahanagar Gas Limited, 2003, also available at Tata Energy Research Institute (TERI), 2001b, http://www.mahanagargas.com/ State of Environment Report for Delhi, Prepared Maps of India, 2003, also available online at for the Department of Environment, Govern- http://www.maps of india.com/ ment of National Capital Territory of Delhi. 124 References Tata Energy Research Institute (TERI), 2002 , World Bank, 1997, Urban Air Quality Manage- Towards Cleaner Air ­ A Case Study of Delhi, ment Strategy in Asia (URBAIR): Greater Prepared for the Department of Environment, Mumbai Report, World Bank Technical Paper Government of National Capital Territory of 380, Washington D.C. Delhi. World Bank 2003, India: Access of the Poor to UNICEF, 2004, www.unicef.org/ Clean Household Fuels, ESMAP, Washington, D.C. infobycountry/india World Bank, 2004, India at a Glance (internal United Nations Economic and Social Commis- briefing note, February 2004), Washington D.C. sion for Asia and Pacific (UNESCAP), 2000, State of the Environment in Asia and the Pacific, World Health Organization (WHO), 2002, The Bangkok. World Health Report, Geneva. 125 For a Breath of Fresh Air Notes 1. The national annual ambient air quality dispersion model calculations are less than 2, 2-5, standards for residential areas are 140 µg/m³ for 5-7.5 and greater than 7.5 meters/sec (m/s) SPM and 60 µg/m³ for RSPM. Furthermore, WHO (Lvovsky et al. 2000). From the analysis of the sets no threshold for the health impacts of present data, it was observed that the dispersion exposure to RSPM, as such impacts have been situation is substantially changed when wind detected even at very low levels of exposure. speed is combined with precipitation. So, the first classification (0-2 meters/sec) has been divided 2. Bangalore was originally proposed to be included into three, less than 1, 1-1.5 and greater than 1.5 in the study, but it had to be excluded due to m/s. All the other classifications are included in inadequate data on key air pollution parameters, the >2 m/s group. These sub-divisions when particularly RSPM. combined with precipitation provided reasonable qualitative description for observed variation in 3. The sources of urban air pollution are many, and PM levels in all the cities with some exceptions can broadly be classified into transport, industrial, that arise from outliers in the data sets. domestic/municipal, and fugitive sources. Specific examples include: motor vehicles, large, medium 10. These winds are locally called the "loo." and small scale industries, thermal power plants, garbage burning, domestic cooking and heating, 11. The numbers for five cities being studied were and re-suspended road dust. 7,491 in Delhi, 5,726 in Kolkata, 4,477 in Mumbai, 768 in Hyderabad, and 863 in Chennai. 4. The notification was issued by the Delhi Pollution Control Committee as part of the list of approved 12. The CSE followed up on the study with its own fuels under the Air Act. There was no prescribed assessment of premature mortality and estimated limit for sulfur in these fuels before that. higher numbers. TERI also did a similar exercise using a modified methodology and computed 5. All the data were made available by the CPCB. even higher numbers. See Annex 3A. 13. Additional caution should be exercised in using 6. Even CPCB measurements of RSPM, started in estimates for Delhi and Kolkata for Scenario 4, 1999, seldom meet the requirement of 104 when concentration-response coefficients were measurements in a year. This is a major issue with applied over a wide range of concentrations, data collection irrespective of the agency collect- which is unusual for concentration-response ing the data. studies which typically measure the impact of marginal changes in air quality. 7. The sensitive area classification refers to protected areas and eco-sensitive zones such as the area 14. For example, LPG uptake for cooking gas rose around the Taj Mahal. nation-wide from 26 percent of urban households 8. According to CPCB the high levels of NO2 in in 1993-1994 to 45 percent in 1999-2000, while the Delhi may be attributed to two natural gas based use of wood dropped significantly (See World thermal power plants emitting high levels of NO2, Bank 2003). The consumption of LPG in the plying of non-optimized converted CNG vehicles, domestic sector almost doubled in 1998-199 as poor performance of three-way catalytic convert- compared to 1990-1991 (TERI 2001) ers in gasoline vehicles, and an overall increase in 15. Information on city-specific monitoring networks the vehicle population. for each of the five study cities has been pro- 9. The usual classifications for wind speed in vided by NEERI. 126 Notes 16. Based on site visit and input from NEERI and dispersion model calculations are <2, 2-5, 5-7.5 CPCB staff in Delhi. and >7.5 meters/sec (Lvovsky et al. 2000). From the analysis of the present data, it was observed 17. Where R, C, and I refer to residential, commercial, that the dispersion situation is substantially and industrial areas respectively. changed when wind speed is combined with 18. When the concentrations of RSPM or SPM mea- precipitation. So, the first classification (0-2 sured at the same site at the same time are plotted meters/sec) has been divided into three, <1, 1-1.5 against each other, the regression line must go and >1.5 m/s. All the other classifications are through the origin. If the monitoring stations are included in the >2 m/s group. These sub- different and /or data are not taken on the same divisions when combined with precipitation days, as in this case, then the line does not neces- provided reasonable qualitative description for sarily have to go through the origin. However, observed variation in PM levels in all the cities since RSPM data are collected in the same city, with some exceptions that arise from outliers in provided there is reasonable dispersion across the the data sets. city, it is unlikely for RSPM concentrations to be 22. The secondary sector comprises of manufactur- zero in one place and non-zero in another place, ing electricity, gas and water supply and construc- and therefore the line is forced through the origin tion; and the tertiary sector comprising of throughout this annex. trade, hotels and restaurants, transport, 19. The concentration of RSPM must be zero if SPM storage, communications, financing, insurance, concentration is zero. In theory, it is possible for real estate, business services, community and SPM to be non-zero and RSPM to be zero, but in social and personal services. practice such a situation is extremely unlikely, if 23. The value is derived taken a change in mortal- not impossible. Therefore, when plotting RSPM ity relative to a change of 10 µg/m³ in SPM against SPM, the plot is forced through the origin levels from Cropper et al (1997) and using the throughout this annex. RSPM/SPM ration of 0.55. 20. The primary sector comprises of agriculture, 24. The cost of urban air pollution was calculated livestock, forestry, fishing, mining and quarrying; to be equivalent to 0.63 % of GDP, or 14 % of the secondary sector comprises of manufacturing the cost of environmental degradation of electricity, gas and water supply and construction; US$9.7 billion. and the tertiary sector comprises of trade, hotels and restaurants, transport, storage, communica- 25. All ranking should use number 1 as the high-end tions, financing, insurance, real estate, business of the scale, with following numbers signifying services, community and social and personal reduction. services. 26. Includes construction, open burning of gar- 21. "Good dispersion" conditions correspond to bage, and household fuel use monthly rainfall more than 50 mm. "Moderate" corresponds to rainfall more than 20 mm and 27. Including power plants wind speed more than 5.4 km/hr (1.5 m/sec) or 28. These could range from technical / policy wind speed more than 7.2 km/hr (2 m/sec). interventions driven by court orders, to initia- "Poor" corresponds to wind speed less than 5.4 tives that may have taken place for reasons km/hr and rain less than 20 mm or wind less than other than air quality improvement, but ended 3.6 km/hr (1 m/sec). This classification is not up affecting air quality (e.g. change in industrial based on proper dispersion model calculations. base from manufacturing to the service sector The usual classifications for wind speed in due to economic reasons). 127