Urban Air Quality Management and Planning in South Africa

Urban Air Quality Management and Planning in

South Africa

Yvonne Scorgie

Student Number: 909441565

Department of Geography, Environmental Management and Energy Studies, Faculty of Science, University of Johannesburg

Supervisor: Professor H. J. Annegarn

A thesis submitted to the Faculty of Science, University of Johannesburg, in fulfilment of the requirements for the degree of Doctor of Philosophy in Environmental Science 15 September 2012

Dedication

This work is dedicated to my husband Anthony Scorgie, my mother Margaret Goosen and my children Michael and Katelyn.

iii

Acknowledgments

I extend my sincere appreciation to the following individuals and organisations that have provided support during the course of my research:

 Prof. Harold Annegarn, for sharing his extensive knowledge of South Africa‟s air quality issues, and his insight into the shaping of solutions to suit local circumstances. His unwavering support and leadership have been invaluable.  Melanie Kneen (University of Johannesburg) for her work on the NEDLAC Dirty Fuels Study, specifically in regard to extracting remote sensing burn scar data to estimate the spatial extent of biomass burning and assisting in the spatial distribution of household fuel burning and road transport emissions.  Dr Lucian Burger, Hanlie Liebenberg-Enslin and Gerrit Kornelius (Airshed Planning Professionals Pty Ltd) for supporting my interest in air quality governance and providing valuable technical input. Lucian‟s guidance on the estimation and future projection of vehicle exhaust emissions during the Dirty Fuels Project, Gerrit‟s collaboration on the work undertaken in support of national emission standard setting, and Hanlie‟s shared interest in local air quality management planning is specifically acknowledged.  Renee von Gruenewaldt (nee Thomas, Airshed Planning Professionals), for assisting in updating portions of the Highveld emissions inventory used as the basis for the modelling of

tropospheric NO2 concentrations.  Dr Andreas Richter (Institute of Environmental Physics at the University of Bremen) for providing the SCIAMACHY plots used in the study.  Dr Peter Newman (UK Environment Agency) for provision of written inputs to inform the research undertaken in support of the national emission standard setting process.  Eskom Holdings for providing meteorological, source and emissions data for use in the study and for their research support.  Department of Environmental Affairs and Tourism, specifically the Chief Directorate: Air Quality Management, for sanctioning my involvement in several projects including national state of air reporting and work in support of AQA implementation.  Rina Taviv and colleagues (CSIR) for collaboration on the inaugural National State of Air Project.  City of Johannesburg, Ekurhuleni Metropolitan Municipality and City of Cape Town personnel for providing me with insight into the challenges facing local authorities and involving me in their air quality management plan development processes.  Organisations and companies having provided air quality monitoring data for use in national air quality baseline characterisation. Their proactive approach, in the absence of legislation requiring such air quality monitoring, has significantly contributed to our understanding of the status of air quality and facilitated the identification of major air pollution challenges.  Stellenbosch Automotive Engineering for providing the national vehicle fleet data using in the NEDLAC Dirty Fuels Study.  South African Weather Services, for the provision of surface and upper air data including ETA model data.

 The South African National Research Foundation, for supporting the work by providing bursary funding.

Part of this work has been published in conference papers and technical reports during the course of the research. These are listed below in chronological order:

Scorgie, Y., M.A. Kneen, H.J. Annegarn and L.W. Burger (2003a). Air Pollution in the Vaal Triangle – Quantifying Source Contributions and Identifying Cost-effective Solutions‟, Clean Air Journal, 12(4), November 2003. Scorgie, Y., H.J. Annegarn and L. Randell (2003b). Air Quality Situation Assessment for the City of Johannesburg, Final Report, Report compiled for the City of Johannesburg, Matrix Environmental Consultants, Report No. MTX/03/JHB-01d, 23 January 2003. Scorgie, Y., H.J. Annegarn and L. Randell L (2003c). Air Quality Management Plan for the City of Johannesburg, Air Quality Management Plan compiled on behalf of and in consultation with the Department of Development Planning, Transportation and Environment and the Department of Environmental Health, City of Johannesburg, Matrix Environmental Consultants, Report No. MTX/03/JHB-01e, 23 September 2003. Scorgie, Y., L.W. Burger and H.J. Annegarn (2003d). Review of International Air Quality Guidelines and Standards for the Purpose of Informing South African Air Quality Standards, Report compiled on behalf of the South African Bureau of Standards (SABS) Technical Committee on National Air Quality Standards, Report No. SANS/03/01 Rev. 0, 5 March 2003. Scorgie, Y. (2004a). Air Quality Situation Assessment for the Vaal Triangle Region, Report compiled by Matrix Environmental Consultants on behalf of the Legal Resources Centre, Report Number MTX/02/LRC-01b, Johannesburg. Scorgie, Y. (2004b). Air Quality Management Plan Development Process for Khayelitsha and the City of Cape Town, Report compiled by Airshed Planning Professionals Pty Ltd for the City of Cape Town, Report No. APP/04/CCT-03, 24 August 2004. Scorgie, Y. and R.M. Watson (2004). Updated Air Quality Situation Analysis for the City of Cape Town, Report compiled for the City of Cape Town, Airshed Planning Professionals Pty Ltd, Report No. APP/04/CCT-02, 23 August 2004. Scorgie, Y., H.J. Annegarn and L.W. Burger (2004a). Socio-Economic Impact of Air Pollution Reduction Measures - Task 1: Definition of Air Pollutants Associated with Combustion Processes, Report compiled on behalf of National Economic Development and Labour Council (NEDLAC) under the Fund for Research into Industrial Growth and Equity (FRIDGE), Airshed Planning Professionals Pty Ltd, Midrand. Scorgie, Y., L.W. Burger and H.J. Annegarn (2004b). Socio-Economic Impact of Air Pollution Reduction Measures - Task 2: Establishment of Source Inventories, and Task 3: Identification and Prioritisation of Technology Options, Report compiled on behalf of National Economic Development and Labour Council (NEDLAC) under the Fund for Research into Industrial Growth and Equity (FRIDGE), Airshed Planning Professionals Pty Ltd, Midrand.

Scorgie, Y., L.W. Burger and H.J Annegarn (2004c). Socio-Economic Impact of Air Pollution Reduction Measures - Task 4: Quantification of Environmental Benefits Associated with Fuel Use Interventions, Report compiled on behalf of National Economic Development and Labour Council (NEDLAC) under the Fund for Research into Industrial Growth and Equity (FRIDGE), Airshed Planning Professionals Pty Ltd, Midrand. Scorgie, Y., G. Paterson, L.W. Burger, H.J. Annegarn, and M.A. Kneen (2004d). Socio-Economic Impact of Air Pollution Reduction Measures – Task 4a Supplementary Report: Quantification of Health Risks and Associated Costs Due to Fuel Burning Source Groups, Report compiled on behalf of National Economic Development and Labour Council (NEDLAC) under the Fund for Research into Industrial Growth and Equity (FRIDGE), Airshed Planning Professionals Pty Ltd, Midrand. Scorgie, Y., R. Watson, T. Fischer, and L. van der Walt (2004e). Background Information Document. Air Quality Baseline Assessment for the Ekurhuleni Metropolitan Municipality, Report compiled on behalf of Ekurhuleni Metropolitan Municipality, Report No. APP/04/EMM-01rev1, Airshed Planning Professionals Pty Ltd, Midrand. Scorgie, Y., T. Radebe, and R. Watson (2004f). Development of an Air Quality Management Plan for the Ekurhuleni Metropolitan Municipality, Paper presented at the National Association for Clean Air (NACA) Conference, October 2004, Johannesburg. Scorgie, Y. (2005). Situation Analysis and Air Quality Management Plan for the Greater Alexandra Area, Report compiled on behalf of the Gauteng Department of Housing by Airshed Planning Professionals Pty Ltd, Report No. APP/04/GDH-02, 6 June 2005. Scorgie, Y. and C. Venter (2005). National State of the Environment: Atmosphere and Climate. Report compiled on behalf of the Department of Environmental Affairs and Tourism, Pretoria. Scorgie, Y., T. Fischer, and R. Watson (2005). Air Quality Management Plan for the Ekurhuleni Metropolitan Municipality, Plan compiled on behalf of and in consultation with the Department of Environment & Tourism, Ekurhuleni Metropolitan Municipality, Report No. APP/04/EMM- 02c, Airshed Planning Professionals Pty Ltd, Midrand. Scorgie, Y. (2006). Prioritisation of Registration Certificates for Review as Part of the Atmospheric Pollution Prevention Act (APPA) Registration Certificate Review Project, Report compiled on behalf of the Department of Environmental Affairs and Tourism, Chief Directorate: Air Quality Management. Baird, M. and Y. Scorgie (2006). APPA Registration Certificate Data Base Report, Report compiled on behalf of the Department of Environmental Affairs and Tourism, Report APP/DEAT/2006, Airshed Planning Professionals Pty Ltd, Midrand. Mahema, T., M. Baird, and Y. Scorgie (2006). APPA registration certificate electronic database status and information on scheduled processes operating in South Africa, Paper presented at the South African National Association for Clean Air (NACA) Annual Conference and Workshop, 18 to 20 October 2006, East London. Scorgie, Y. and G. Kornelius (2007). Air Quality Implementation: Listed Activities and Minimum Emission Standards, Output B.1 – International Review, Report compiled on behalf of the Department of Environmental Affairs and Tourism, Chief Directorate: Air Quality Management, Pretoria.

Scorgie Y, Annegarn HJ and Burger LW (2007). Air Pollution over South Africa: Persistent Problems and Emerging Issues, 14th International Union of Air Pollution Prevention and Environmental Protection Associations (IUAPPA) World Congress, 10-14 September 2007. Scorgie, Y., H.J. Annegarn, A. Richter, K. Ross, and L.W. Burger (2007). Comparison of

SCIAMACHY NO2 Observations over Southern Africa with Air Quality Modelling Results, IEEE International Geoscience and Remote Sensing Symposium (IGARSS), Barcelona, Spain, 23 – 27 July 2007. Annegarn, H.J., L.W. Burger, J. John, N. Krause, M. Naidoo, Y. Scorgie, R. Taviv and M. Zunckel (2007). National Air Quality Management Programme (NAQMP) Output C.4, Initial State of Air Report, Department of Environmental Affairs and Tourism, May 2007, Pretoria. Scorgie, Y. (2008). Guideline Inspection Protocol for Listed Activities holding Atmospheric Emission Licenses under the National Environmental Management: Air Quality Act, Guideline compiled on behalf of the Department of Environmental Affairs and Tourism, Directorate Compliance Monitoring, February 2008. Scorgie, Y. and G. Kornelius (2009). Modelling of Acid Deposition over the South African Highveld, Paper presented at the annual South African National Association for Clean Air Conference, 14-16 October 2009, Vanderbijlpark, South Africa. Scorgie, Y., H.J. Annegarn, L.W. Burger and M.A. Kneen (2012). Quantification of Health Risks and Costs Due to Atmospheric Emissions from Anthropogenic Fuel Burning within South African Conurbations and Benefits of Air Pollution Interventions, paper submitted for publication by the South African Journal of Science. Scorgie, Y., H.J. Annegarn, A. Richter, K. Ross, and L.W. Burger (2012). Modelling of

Tropospheric NO2 Intensities over the South African Highveld and Comparison with Remote Sensing Observations, paper submitted for publication by Atmospheric Chemistry and Physics.

vii

Abstract

Fossil fuel burning within residential, industrial and power generation sectors represents a persistent source of air pollution within many parts of South Africa, with the contribution of road transport emissions becoming increasingly important. Additionally, biomass burning, including agricultural burning and wild fires, represents an intermittent but seasonally significant source of atmospheric emissions.

Effective air pollution control was historically hindered by the absence of enabling legislation and cooperative governance. The promulgation of the National Environmental Management: Air Quality Act, Act 39 of 2004 represented a major step forward in the evolution of air quality management within South Africa. The historical debate regarding the practicability of effective air quality management is however ongoing. South Africa‟s continued dependence on coal to support its energy-intensive industrial and mining sectors, continued household fuel burning for space heating and cooking purposes within a number of areas, and the dire need for employment creation and focus on rapid development continue to challenge the realisation of air quality improvements.

This study investigates the multiple factors contributing to the degradation of air quality in South Africa, and the consequent human health, environmental and economic effects of this pollution. The study critically examines legal, technical and social measures implementable within a tailored system of air quality management which is compatible with socio-economic growth. This thesis integrates and expands on pertinent components of several individual research projects completed by the author during her tenure as a doctoral candidate. The research projects were completed during the period (2002 – 2009) on behalf of various parties including national and local government, standards setting bodies and private organisations.

Quantification of health risks associated with significant anthropogenic sources within several South African conurbations, covering 40% of the national population, and the establishment of cost-optimised air pollution interventions, forms a key component of the thesis. In this externalities study, emissions were estimated and effects and associated costs quantified for household fuel burning, power generation, industrial and commercial fuel burning and road transport. Total direct health costs related to inhalation exposures to fuel burning emissions were estimated to be of the order of 3.5 billion 2002 Rands per annum across health effects, conurbations and source groupings assessed. Household fuel burning was estimated to be responsible for about 68% of the total health costs estimated across all conurbations, vehicle emissions for 13%, industrial and commercial fuel burning for 13%, and power generation for about 6%.

Emission reduction opportunities were identified and assessed for residential fuel burning, coal- fired power generation, road transport, coal-fired industrial boilers and specific individual industries. It was concluded that significant health effect reductions could cost-effectively be achieved through addressing residential fuel burning as a priority. Lower benefit-cost ratios associated with industrial and vehicular interventions are due, in part, to these fuel burning sources having been more effectively regulated historically. The need for effective management of industrial and vehicle emissions is however supported. Based on international experience and local trends in vehicle activity, the contribution of transport emissions will become increasingly

viii

significant if not adequately addressed. Industrial process emissions unrelated to fuel burning may include significant emissions of criteria pollutants, in addition to trace releases of a wide range of hazardous air pollutants.

Internationally, actions taken to address air pollution problems have met with mixed results. Failure to integrate economic considerations into air quality management planning, and to integrate air quality considerations into development planning represent key weaknesses in the strategies implemented. A contribution is made in this thesis by highlighting such lessons and proposing legal, technical and social measures which, when implemented within a rational system of air quality management, are suited to addressing complex air pollution sources without negatively affecting socio-economic prosperity and equity. Components of an effective, affordable and equitable emissions control policy proposed for adoption within South Africa include phased national standards setting, compliance promotion and self-monitoring, market-based instruments, and the implementation of risk-based enforcement and compliance monitoring strategies.

Table of Contents Affidavit...... ii Dedication ...... iii Acknowledgments ...... iv Abstract ...... viii List of Figures ...... xii List of Tables ...... xvi Abbreviations ...... xix 1 Introduction...... 21 1.1 Background and Problem Statement ...... 21 1.2 Hypothesis ...... 23 1.3 Research Aims and Objectives ...... 23 1.4 Procedure and Integration of Research ...... 23 1.5 Recent Transformations in Air Quality Practices ...... 25 1.6 Structure of Thesis ...... 25 2 Air Quality over South Africa ...... 29 2.1 Atmospheric Emission Sources ...... 29 2.2 Ambient Air Quality ...... 30 2.3 Indoor Air Quality in Fuel-Burning Houses and Related Health Risks ...... 43 2.4 Summary ...... 44 3 Air Quality Management in South Africa ...... 46 3.1 Air Quality and Sustainable Development ...... 46 3.2 Advances in Air Quality Management in South Africa and Sector-specific Interventions . 55 3.3 Effective Implementation of Sector-specific Interventions ...... 59 3.4 Summary ...... 59 4 Calculation of Externalities Due to Fuel-burning Emissions ...... 61 4.1 Introduction...... 61 4.2 Study Objective ...... 61 4.3 Overview of Externality Studies ...... 62 4.4 Methodological Overview ...... 64 4.5 Scope and Limitations of Study ...... 65 4.6 Emission Quantification ...... 68 4.7 Dispersion Modelling of Air Pollutant Concentrations ...... 92 4.8 Health and Welfare Risk Estimation ...... 106 4.9 Direct Health Cost Projection ...... 108 4.10 Selection of Interventions ...... 109 4.11 Summary ...... 110 5 Emissions and Air Quality Impacts of Fuel-burning...... 111 5.1 Baseline Emissions due to Fuel-burning Sources ...... 111 5.2 Forecast Changes in Emissions given “Business as Usual” ...... 115 5.3 Air Pollutant Concentrations due to Fuel-burning Emissions ...... 120 5.4 Source Contributions to Air Pollutant Concentrations within the Vaal Triangle ...... 124 5.5 Synopsis of Source Significance ...... 126 6 Health Effects and Costs due to Fuel-burning Sources ...... 127 6.1 Study Limitations ...... 127 6.2 Inhalation Health Effects due to Exposures to Fuel Burning Emissions ...... 127 6.3 Health Cost Predictions ...... 142 7 Cost-optimisation of Air Pollution Mitigating ...... 147 7.1 Source Prioritisation and Intervention Selection...... 147 7.2 Description of Selected Interventions ...... 148 7.3 Health Impact Reductions to due Interventions ...... 158 7.4 Cost-benefit Analysis of Interventions() ...... 164

7.5 Conclusions and Recommendations ...... 167 8 Regulation and Compliance Monitoring of Industry ...... 170 8.1 Background to Investigations ...... 170 8.2 Guidance for National Emission Standard Setting ...... 171 8.3 Compliance Monitoring ...... 189 8.4 Synthesis of Study Outcomes ...... 198 9 Phased Air Quality Management Planning ...... 203 9.1 Background ...... 203 9.2 Overview of Air Quality Management Planning ...... 203 9.3 Challenges to AQMP Development and Implementation ...... 204 9.4 Air Quality Management Policy ...... 205 9.5 Ambient Air Quality Objectives ...... 208 9.6 Air Quality Management Systems ...... 215 9.7 Capacitating Air Quality Management ...... 218 9.8 Emission Reduction Planning ...... 226 9.9 Summary and Outlook ...... 235 10 Summary, Conclusions and Outlook ...... 237 10.1 Study Hypothesis ...... 237 10.2 Integrating International Lessons ...... 238 10.3 Significant Sources, Priority Pollutants and Key Affected Areas ...... 239 10.4 Health Effects and Costs due to Fuel-Burning Sources ...... 239 10.5 Cost-optimisation of Air Pollution Interventions ...... 241 10.6 Regulation and Compliance Monitoring of Industry ...... 242 10.7 Phased Air Quality Management Planning ...... 243 10.8 Overall Conclusion ...... 244 10.9 Summary of Contribution ...... 244 10.10 Future Research ...... 246 References ...... 247 Appendix A –Overview of APPA and AQA ...... 266 Appendix B –Vehicle Emission Factors and Emission Estimates ...... 275 Appendix C – Fuel-burning Emissions per Region and Source...... 277 Appendix D – Predicted Air Pollutant Concentrations due to Fuel-burning Emissions ...... 280 Appendix E – Direct Health Cost Estimates ...... 293 Appendix F – Guideline Inspection Protocol for Listed Activities holding Atmospheric Emission Licenses under the AQA ...... 295

List of Figures

Figure 1: Contributions to the evolving system of air quality governance in South Africa (adapted DEA, 2007b)...... 28 Figure 2: South African population distribution and the relative location of the ambient air quality monitoring stations (Annegarn et al., 2007) ...... 30 -3 -3 Figure 3: „Moderate‟ (daily-average PM10 50–75 µg m ), „high‟ (75-100 µg m ) and „very high‟ (>100 µg m-3) pollution days at selected sites (2004) ...... 32

Figure 4: Seasonal variations in PM10/PM2.5 ratios recorded within a residential coal burning suburb of Soweto, Johannesburg, 1996-7 (Annegarn et al., 1999) ...... 33 -3 Figure 5: Frequency of exceedances of the hourly SO2 limit of 350 µg m at selected sites during 2004 ...... 34 -3 Figure 6: Frequencies of exceedance of the hourly NO2 limit of 200 µg m at selected sites during 2004 ...... 36

Figure 7: Diurnal variations in PM2.5 black carbon concentrations, as observed in Soweto during the period 1 to 20 June 1997 (Annegarn et al., 1999) ...... 37

Figure 8: Daily average PM10 concentrations recorded by the City of Johannesburg at the Oliver Tambo Clinic in Diepsloot during 28 June to 24 November 2004 ...... 38

Figure 9: Diurnal variations in hourly average NOx and NO2 concentrations recorded by eThekwini Metropolitan Municipality at its Warwick Station during 2004 (Annegarn et al., 2007) ...... 39 Figure 10: Number of installed dustfall monitoring sites on the Central Witwatersrand and total number of monthly dustfall values in the ACTION and ALERT bands (Annegarn et al., 2007) ...... 40 Figure 11: Number of installed dustfall monitoring sites on the Central Witwatersrand and total number of monthly dustfall values in the ALERT band (Annegarn et al., 2007) ...... 41 Figure 12: Percentage of ACTION and ALERT dustfall incidents as a fraction of the total number of installed dust monitoring sites (Annegarn et al., 2007) ...... 41 Figure 13: Number of dustfall measurements recorded on the Central Witwatersrand during the period 1985 to 2005 which are in INDUSTRIAL, ACTION and ALERT ranges (Annegarn et al., 2007) ...... 42 Figure 14: Identification of some of the regions where elevated air pollutant concentrations in excess of health thresholds have been measured to occur (author‟s own figure)...... 43 Figure 15: Schematic illustration of the EU approach to action planning and reporting which is dependent on the air quality status in relation to air quality limit values and taking into account a margin of tolerance (Scorgie et al., 2003d) ...... 51 Figure 16: The damage function approach applicable to emissions and effects related to the energy generation sector (after van Horen, 1996) ...... 64

xii

Figure 17: Location of industrial and commercial fuel burning sources (red) and large-scale power stations (green) located on the Highveld, comprising the Tshwane, Johannesburg and Ekurhuleni Metropolitan areas, the Vaal Triangle and the Mpumalanga Highveld ...... 75 Figure 18: Location of industrial and commercial fuel burning sources within Cape Town ...... 76 Figure 19: Location of industrial and commercial fuel burning sources within eThekwini ...... 77 Figure 20: Percentage household fuel use for lighting, cooking and heating requirements for 1996 and 2001 (Statistics South Africa, 2002) ...... 81 Figure 21: Number of households per km² using coal to meet their space heating and/or cooking requirements on the Highveld ...... 82 Figure 22: Contribution of specific fuels to total household fuel burning emissions across all conurbations ...... 84 Figure 23: Petrol sales per magisterial district and road network density for the Johannesburg and Ekurhuleni metropolitan areas ...... 88 Figure 24: Remote sensing burn scar data showing incidences of fires during the period 1995 – 2000 over the Plateau. The Tshwane, Johannesburg-Ekurhuleni, Vaal Triangle and Mpumalanga Highveld study areas are shown...... 89 Figure 25: Remote sensing burn scar data showing incidences of fires during the period 1995 – 2000 over Cape Town. The legend shows the number of fires recorded to occur at each location during this period; the data range being no fires to 4 fires...... 90 Figure 26: Remote sensing burn scar data showing incidences of fires during the period 1995 – 2000 over eThekwini. The legend shows the number of fires recorded to occur at each location during this period; the data range being no fires to 2 fires...... 91 Figure 27: Plateau modelling domain which includes the Mpumalanga Highveld, Vaal Triangle, Johannesburg, Ekurhuleni and Tshwane study areas...... 95 Figure 28: Spatial extent of the Cape modelling domain. The modelling domain coincides exactly with the area shown in the figure...... 96 Figure 29: Spatial extent of the eThekwini modelling domain. The modelling domain coincides exactly with the area shown in the figure...... 97 Figure 30: Intra- and inter-annual variations in the number of "heating-degree-days" per month for Johannesburg for the period 1990-1999...... 99 Figure 31: Exponential curves indicating decay times for atmospheric black carbon concentrations. The measured diurnal curve represents the geometric mean values of seven days of sampling (Annegarn et al., 1999)...... 100 Figure 32: Example of diurnal trends in traffic volumes, as recorded along the N1 in Johannesburg (between 14th and Gordon Avenue off-ramps) during the period 14 September to 13 October 1999 ...... 101 Figure 33: Monthly variations in the occurrence of fires in Cape Town, eThekwini and on the Plateau – compilation based on a five-year record of burn scar information derived from remote sensing data ...... 101 Figure 34: Location of the weather stations used as input to the CALMET model for the Cape modelling domain ...... 102

xiii

Figure 35: Wind roses generated based on hourly wind field data from various meteorological station located within the Plateau modelling domain ...... 103 Figure 36: Topography of the Plateau modelling domain ...... 104 Figure 37: Land use classification of the eThekwini modelled area represented as a raster map .. 105 Figure 38: Contribution of source groups to total fuel-burning emissions estimated across all conurbations considered ...... 113 Figure 39: Trends in liquid fuel sales for the 1994 to 2004 period, South African Petroleum Industry Association (SAPIA, 2005) ...... 117 Figure 40: Location of sites (red squares) at which source contributions to predicted total

annual PM10 concentrations are illustrated in Figure 41 ...... 124

Figure 41: Predicted relative source contributions to total annual PM10 concentrations at various locations within the Vaal Triangle. Results are given for a central point within the CBDs of various areas (Sasolburg, Meyerton, Vereeniging, Vanderbijlpark) and for a central point in selected fuel-burning residential areas (Sebokeng, Sharpville) ...... 125 Figure 42: Source contributions as a percentage of overall health effects predicted due to exposures to fuel burning emissions within Cape Town, including (a) respiratory hospital admissions and premature mortality, and (b) excess leukaemia cases, due to chronic exposures to 1,3-butadiene and benzene concentrations (Scorgie et al., 2004e)...... 130 Figure 43: Source contributions as a percentage of overall health effects predicted due to exposures to fuel burning emissions within eThekwini, including (a) respiratory hospital admissions and premature mortality, and (b) excess leukaemia cases, due to chronic exposures to 1,3-butadiene and benzene concentrations (Scorgie et al., 2004e)...... 131 Figure 44: Source contributions as a percentage of overall health effects predicted due to exposures to fuel burning emissions within Johannesburg and Ekurhuleni, including (a) respiratory hospital admissions and premature mortality, and (b) excess leukaemia cases, due to chronic exposures to 1,3-butadiene and benzene concentrations (Scorgie et al., 2004e)...... 132 Figure 45: Source contributions as a percentage of overall health effects predicted due to exposures to fuel burning emissions within Tshwane, including (a) respiratory hospital admissions and premature mortality, and (b) excess leukaemia cases, due to chronic exposures to 1,3-butadiene and benzene concentrations (Scorgie et al., 2004e)...... 133 Figure 46: Source contributions as a percentage of overall health effects predicted due to exposures to fuel burning emissions within the Vaal Triangle, including (a) respiratory hospital admissions and premature mortality, and (b) excess leukaemia cases, due to chronic exposures to 1,3-butadiene and benzene concentrations (Scorgie et al., 2004e)...... 134 Figure 47: Source contributions as a percentage of overall health effects predicted due to exposures to fuel burning emissions within the Mpumalanga Highveld study area, including (a) respiratory hospital admissions and premature mortality, and (b) excess leukaemia cases, due to chronic exposures to 1,3-butadiene and benzene concentrations (Scorgie et al., 2004e)...... 135

xiv

Figure 48: Source contributions as a percentage of overall health effects predicted due to exposures to fuel burning emissions within all conurbations, including (a) respiratory hospital admissions and premature mortality, and (b) excess leukaemia cases, due to chronic exposures to 1,3-butadiene and benzene concentrations (Scorgie et al., 2004e). (Based on the 2002 emissions inventory.)...... 139 Figure 49: Percentage of total direct health costs due to inhalation exposures to fuel burning emissions incurred per conurbation ...... 143 Figure 50: Contribution of source groupings to total direct health costs estimated to occur due to fuel use and inhalation exposures to fuel burning emissions ...... 144 Figure 51: Reductions in the annual number of respiratory hospital admissions estimated due to interventions implementable by 2007 ...... 162 Figure 52: Reductions in the annual number of respiratory hospital admissions estimated due to interventions implementable by 2011 ...... 163 Figure 53: Marginal net benefit-cost ratios calculated by Leiman et al. (2007) for each intervention ...... 167 Figure 54: The environmental governance cycle (DEAT, 2007b) ...... 190 Figure 55: Air quality management planning process (Own figure) ...... 204 Figure 56: Air quality management system recommended for EMM (Scorgie et al., 2004f) ...... 217 Figure 57: Cooperative governance structures outlined at the 2005 Air Quality Governance Conference (DEA, 2007) ...... 226 Figure 58: Contributions to the evolving system of air quality governance in South Africa (adapted DEA, 2007b)...... 245

List of Tables

Table 1: Synopsis of interventions to reduce emissions from residential fuel burning ...... 57 Table 2: Information for electricity generation facilities included in the study ...... 69 Table 3: Emission factors for coal-fired power stations ...... 71 Table 4: Estimated total annual emissions from public electricity generation ...... 72 Table 5: Annual fuel use by industries and businesses within eThekwini and Cape Town ...... 72 Table 6: General emission factors selected for estimating emissions from industrial and commercial fuel burning. Emission factors were derived from the US-US-EPA's AP-42 database. (Emission factors are given in kg of pollutant emitted as a result of unit of fuel burned.) ...... 73 Table 7: Total annual emissions from industrial, commercial and institutional fuel burning within each conurbation, excluding fuel burning for electricity generation ...... 78 Table 8: Estimated total annual household fuel consumption per conurbation ...... 82 Table 9: Emission factors for estimating household fuel combustion related releases ...... 83 Table 10: Total annual emissions due to household fuel combustion ...... 83 Table 11: Total annual emissions due to vehicle emissions (exhaust and evaporative releases) - baseline scenario (2002) ...... 87 Table 12: Extent of area burnt within each conurbation - given as a composite area for the 1995-2000 period, as a total area for the 2000 fire season and indicating average and peak burn areas over 10-day periods ...... 92 Table 13: Emission factors used to quantify atmospheric emissions from biomass burning ...... 92 Table 14: Total annual emissions estimated due to biomass burning within various conurbations ...... 92 Table 15: Dose-response functions selected for the quantification of inhalation exposures to air pollutant concentrations due to fuel combustion emissions ...... 106 Table 16: Synopsis of cancer unit risk factors(a) selected for use in the study for quantifying inhalation exposures to air pollutant concentrations due to fuel combustion emissions ...... 107 Table 17: Ratios of inpatients to outpatients calculated from the Medscheme information obtained for each health condition ...... 108 Table 18: Direct health costs of public and private inpatients and outpatients calculated from information obtained from Medscheme (costs given per patient) ...... 109 Table 19. Average length of stay associated with each health condition costed ...... 109 Table 20: Average annual percentage change in real gross value (ABSA, 2002) and Capex Projects (IDC, 2003) for industrial sub-sectors of interest in this study ...... 116 Table 21: Estimated changes in vehicle emissions given changes in vehicle activity and vehicle technology ...... 119

xvi

Table 22: Summary of total persons within each population subsector of interest in the current study (Source: Census 2001, Statistics South Africa) ...... 128 Table 23: Estimated health effects, given as number of cases or incidences per annum associated with human exposures to fuel burning emissions predicted for the base year 2002(a)...... 129 Table 24: Annual incidences in health endpoints due to all causes ...... 140 Table 25: Percentage of actual health effect incidence accounted for by the health effects predicted to be due to inhalation exposures due to fuel burning emissions ...... 141 Table 26: Total direct health costs (Million 2002 Rand) due to inhalation exposures to fuel burning emissions summed per conurbation ...... 142 Table 27: Total direct health costs (Million 2002 Rand) due to inhalation exposures to fuel burning emissions summed per source grouping ...... 142 Table 28: Contribution of source groups to total direct health spending related to fuel use and inhalation exposures to fuel burning emissions per conurbation ...... 144 Table 29: Total emissions of selected pollutants, total health spending on resultant health effects related to inhalation exposures and health spending per ton of emission per source group...... 145 Table 30: Interventions to mitigate atmospheric emissions for the residential sector ...... 151 Table 31: Interventions to mitigate atmospheric emissions from vehicles ...... 153 Table 32: Interventions to mitigate atmospheric emissions from the industrial and power generation sectors ...... 154 Table 33: Interventions selected for quantitative health risk assessment and their associated scales, timeframes and emission reductions ...... 155 Table 34: Reductions in health effects due to interventions implementable by 2007, given as the reduction in actual number of admissions, cancer cases and restricted activity days...... 159 Table 35: Reductions in health effects due to interventions implementable by 2011, given as reduction in number of admissions, cancer cases & restricted activity days...... 160 Table 36: Interventions to reduce emissions resulting in the most significant health risk reductions ...... 164 Table 37: Net present values (NPV) and economic benefit/cost (BC) ratios of interventions (after Leiman et al., 2007) ...... 165 Table 38: Synopsis of industries for which emission standards have been specified in the UK, US, NSW and India, classified according to the RSA „listed activities‟ categories ..... 176 Table 39: Industry types proposed for inclusion as Listed Activities in South Africa and summary of related industry sectors prioritised by other jurisdictions ...... 180 Table 40: Potential additional categories to be considered in South Africa as Listed Activities . 182 Table 41: Comparison of the nature of emission standards issued by the US, UK and Australian NSW for the Glass Manufacturing Industry ...... 185 Table 42: Air Quality Management Policy for the City of Johannesburg (Scorgie et al., 2003c) ...... 206

xvii

Table 43: Alert and information thresholds recommended for use by CCT (Scorgie, 2004b) ..... 211 Table 44: Recommended information sources for inhalation-related health risk thresholds ...... 213 Table 45: Structures, forums and co-ordinating mechanisms recommended for consideration by the CCT (Scorgie, 2004b) ...... 223 Table 46: Source quantification and emission reduction measures consideration for inclusion in the first EMM AQM Plan (Scorgie et al., 2005) ...... 230

xviii

Abbreviations

AEL Atmospheric Emission License APCD Air Pollution Control Division APINA Air Pollution Information Network for Africa APPA Atmospheric Pollution Prevention Act, Act 45 of 1965 AQA National Environmental Management: Air Quality Act, 39 of 2004 AQM Air Quality Management AQMP Air Quality Management Plan ATSDR US Federal Agency for Toxic Substances and Disease Registry BAT Best Available Technology BATNEEC Best Available Technology Not Exceeding Excessive Cost BC Black Carbon BNM Basa njengo Magogo BPEO Best Practicable Environmental Option BTEX Benzene Toluene Ethylbenzene Xylene CAFÉ Clean Air for Europe CAPCO Chief Air Pollution Control Officer CBA Cost-Benefit Analysis CCT City of Cape Town CO Carbon monoxide CO2 Carbon dioxide DEA Department of Environmental Affairs, South Africa (previously the DEAT) DEAT Department of Environmental Affairs and Tourism, South Africa DFA Damage Function Approach DME Department of Minerals and Energy, South Africa DOAS Differential Optical Absorption Spectroscopy DSM Demand Side Management EIA Environmental Impact Assessment EMM Ekurhuleni Metropolitan Municipality EMP Environmental Management Plan EMPR Environmental Management Programme Report ENVISAT European Space Agency‟s (ESA) Environmental Satellite ERF Exposure-Response Function GHG Greenhouse Gas GOME Global Ozone Monitoring Experiment GWP Global Warming Potential HC Hydrocarbons HDD Heating Degree Days HFO Heavy Fuel Oil ICHES Integrated Clean Household Energy Strategy IDP Integrated Development Plan IP&WM Integrated Pollution and Waste Management IPC Integrated Pollution Control IRIS Integrated Risk Information System, US-EPA LPG Liquid Petroleum Gas LSF Low Smoke Fuels

xix

MRAD Minor Restricted Activity Days MRL Minimum Risk Levels NACA National Association for Clean Air NAMSA National Automobile Manufacturers Association of South Africa NEMA National Environmental Management Act, Act 107 of 1998 NGOs Non-government Organisations NMTOC Non-methane Total Organic Compounds NO Nitrogen oxides NOx Oxides of nitrogen

NO2 Nitrogen dioxide NPV Net Present Value NSPS New Source Performance Standards NSW New South Wales, Australia

O3 Ozone OEHHA Office of Environmental Health Hazard Assessment, Canada OMI Ozone Monitoring Instrument PAH Polycyclic Aromatic Hydrocarbons PM10 Particulate matter with an aerodynamic diameter of less than 10 µm PM2.5 Particulate matter with an aerodynamic diameter of less than 2.5 µm PPC Pollution Prevention and Control POCP Photochemical Ozone Creation Potential QA Quality Assurance QC Quality Control RAD Restricted Activity Days RE Renewable Energy REL Reference Exposure Levels RHA Respiratory Hospital Admissions SAAQIS South African Air Quality Information System SABS South African Bureau of Standards SANS South African National Standard SAPIA South African Petroleum Industries Association SCIAMACHY SCanning Imaging Absorption SpectroMeter for Atmospheric ChartograpHY

SO2 Sulphur dioxide STANSA Standards South Africa TOC Total Organic Compounds TSP Total suspended particulate matter UCT University of Cape Town US-EPA United States Environmental Protection Agency VCD Vertical Column Density VKT Vehicle Kilometre Travelled VOCs Volatile organic compounds WHO World Health Organisation

1 Introduction

1.1 Background and Problem Statement South Africa faces a range of persistent air pollution problems in addition to emerging issues. High sulphur dioxide and fine particulate matter levels, due mainly to fuel burning within residential, industrial and power generation sectors, continue to be of concern (Annegarn et al., 2007; Martins et al., 2007; Zunckel et al., 2007). Biomass burning, including agricultural burning and wild fires, represents an intermittent but seasonally significant source of emissions (Helas and Pienaar, 1996; Sowden et al., 2007). The co-location of industries and communities is a cause of health risk exposure and consequent conflict, exacerbated by increasing pressure to place residential areas within former industrial buffer zones (Scorgie et al., 2005). Questions remain regarding the potential for environmental effects and transboundary pollution transportation due to elevated stack emissions from coal-fired power stations, and petrochemical and metallurgical industries (Zunckel et al., 2010).

Emerging air quality issues are primarily associated with road transport, with increases in vehicle emissions projected despite proposed national mitigation measures. Although air quality limits for nitrogen dioxide and ozone are infrequently exceeded within cities, increasing trends in concentrations are apparent (Martins et al., 2007; Zunckel et al., 2007). Volatile organic compound releases from fuel filling stations, and nitrogen oxide and hydrocarbon releases from major airports further highlight the air quality implications of transportation infrastructure developments.

Air pollution control was nationally administered under the Atmospheric Pollution Prevention Act No. 45 of 1965 (APPA) for four decades. APPA made provision for the control of certain industrial processes using a registration process, the control of emissions from diesel vehicles, dust control in proclaimed areas, and the proclamation of designated smoke-control areas. Despite significant shortcomings in the APPA and its implementation, and several attempts to reform the Act in the 1980s and 1990s, the legislation was only repealed after being replaced by the National Environmental Management: Air Quality Act (Act 39) of 2004. By this date the APPA was regarded as outdated for the following key reasons:  APPA could not accommodate the constitutional allocation of air quality control functions in respect of the role of provincial and local government as stipulated in the Constitution of the Republic of South Africa. Section 155(6)(a) and (7) of the Constitution stipulates that air pollution is a local government matter and therefore has to be managed by Municipalities.  APPA focussed on source-based air pollution controls rather than on the achievement and maintenance of ambient air quality standards and thus did not address the cumulative effects of air pollution.  APPA did not deal with all pollution sources, rather focussing on point source pollution (stack emissions) from certain industrial sources.  APPA had inadequate compliance and enforcement mechanisms.  Despite making provision for controlling smoke within designated smoke-control areas, the designation and regulation of „smokeless zones‟ under APPA was restricted fuel burning in

historically white residential areas. Exposures to household fuel burning emissions within poorer black residential areas were not addressed.  The Act fostered a lack of transparency in decision making, making no provision for stakeholder participation within the air pollution control process.

Legislative inadequacies, under resourced enforcement and the relative absence of cooperative governance structures hindered the effective air quality management needed to address persistent and emerging urban air pollution issues. Attempts at reforming the legislation in the 1980s and 1990s were unsuccessful, largely as a result of the following factors (Annegarn and Scorgie, 1995; Boegman et al., 1996; Annegarn et al., 1999; Scorgie, 2001a; Ferreira and Lloyd, 2002; Scorgie et al., 2002):  Widely held views that vigorous enforcement of emission limits and air quality control would be so expensive that it would harm the economy and result in business closures.  Systematic failure under apartheid to address the causes and consequences of township fuel burning. Such environmental apartheid was evident in the declaration of smoke-free zones in urban areas (and associated successes in reducing pollution levels) being restricted to white districts. Despite electrification of historically black residential areas, coal and wood burning persisted mainly due to the cost effectiveness and multi-functional nature of these fuels. Rapid urbanisation and the growth of informal settlements also exacerbated backlogs in electricity distribution.  Absence of air quality management policies and methods tailored to suit local circumstances and inadequate resource allocation for the development and implementation of such methods. These hurdles were evident in the manner in which the air quality management strategy proposed by Annegarn and Scorgie (1995) for adoption within the Vaal Triangle was received. Commissioned by national government, this strategy was intended to address the poor air quality within the industrialised Vaal Triangle region. Notorious for its air pollution, the Vaal Triangle had already been the subject of several intensive air pollution, epidemiological and health risk assessment studies. Responses to the strategy proposed ranged from resistance to indifference, and included sentiments that the approach was overly ambitious, not economically viable, and unsuited to local circumstances. It became apparent to the author that, in addition to enabling legislation, the economic and technical viability of air quality management needed to be demonstrated, and international air quality management methods tailored and cost-optimised to suit local circumstances.

The promulgation of the National Environmental Management: Air Quality Act, Act 39 of 2004 (hereafter, the „Air Quality Act‟, AQA) provides the legislative framework for the evolution of air quality management practices. AQA reflects a paradigm shift from primarily source-based air pollution control to receptor-based air quality management which has as its goal the attainment of acceptable air quality objectives through the addressing of all significant sources. A synopsis of the air pollution control approaches supported by APPA and AQA, and responsibilities designated by each to various tiers of government, is given in Appendix A.

Despite the promulgation of AQA, questions regarding the practicability of effective air quality management have persisted. South Africa‟s continued dependence on coal to support its energy- intensive industrial and mining sectors, continued household fuel burning for space heating and

cooking purposes within a number of areas, and the dire need for employment creation and focus on rapid development represent ongoing hurdles for realising air quality improvements. Accelerated development of air quality management strategies, within the context of the new Air Quality Act is needed, within a complex matrix of needs and constraints of contributing to sustainable socio-economic development, mitigation of greenhouse gas emissions and alignment with international „good practice‟.

1.2 Hypothesis It is possible through systematic analysis of the causes of air quality degradation; the evaluation of the health and environmental consequences thereof; and the evaluation of policies, legislation and technologies, to arrive at a rational system of air quality management that simultaneously can reduce atmospheric emissions, protect human health and the environment, promote socio-economic growth and equity, and nevertheless result in a net positive contribution to the national economy.

1.3 Research Aims and Objectives The aims of this study are to investigate the multiple factors that contribute to the degradation of air quality in South Africa, evaluate the consequent human health, environmental and economic effects of this pollution, and critically examine legal, technical and social measures that could be jointly deployed within an effective system of air quality management system for South Africa.

The specific objectives to be met by the investigation are:  To integrate lessons learned internationally into the evolution of local air quality management planning policies and processes.  To identify and quantify significant sources, priority pollutants and key affected areas within South Africa.  To evaluate the effects on human health and welfare, and the economic costs of air quality degradation due to fuel-burning sources.  To carry out a cost-optimisation evaluation of air pollution mitigation measures for significant anthropogenic fuel-burning sources.  To critically evaluate risk-based enforcement and compliance monitoring methodologies for use in the regulation of industrial activities, and to provide recommendations for national emission standards setting.  To integrate the above findings into a phased system of air quality management, compatible with socio-economic development, and suitable for implementation within South Africa.

1.4 Procedure and Integration of Research The work presented comprises the integration, and in certain instances the reinterpretation and expansion, of pertinent components of several individual research projects completed by the author during the period (2003 – 2009). For this reason, the overall procedure and structure of the thesis does not follow convention. However, the required approach of documenting theory (literature review), methods, results, discussions and conclusions, is reflected within the documentation of the individual study components making up the thesis. Furthermore, the results of such study components are synthesised to address the central hypothesis and a consolidated reference list provided.

The aforementioned projects, completed on behalf of various parties including national and local government departments, standards setting bodies and private organisations, include the following research:  Air quality baseline assessments compiled for specific conurbations (Scorgie et al., 2003a, 2003b; Scorgie, 2004a; Scorgie and Watson, 2004; Scorgie et al., 2004a, 2004b, 2004c, 2004d, 2004e, 2004f; Scorgie, 2005).  Source, emission and air quality data analysis undertaken for the Atmosphere and Climate Section of the 2006 National State of Environment Report and the inaugural National State of Air Report published in 2007 (Scorgie and Venter, 2005; Annegarn et al., 2007).  Air quality management plans developed in consultation with the City of Johannesburg (Scorgie et al., 2003c), the Ekurhuleni Metropolitan Municipality (Scorgie et al., 2004e), the Greater Alexander Area (Scorgie, 2005) and assistance provided to the City of Cape Town in the development of its air quality management plan (Scorgie, 2004b; Scorgie and Watson, 2004).  Review of international air quality guidelines and standards for the purpose of informing the setting of South African Air Quality Standards (Scorgie et al., 2003d) and work undertaken as a member of the Technical Committee on National Air Quality Standards established by Standards South Africa (STANSA), a division of the South African Bureau of Standards (SABS). Standards compiled by the committee were: SANS 69:2004 Framework for setting and implementing national ambient air quality standards, and SANS 1929:2004 Ambient Air Quality - Limits for common pollutants. The author wrote the first drafts of these standards.  Dirty Fuels Study undertaken for the National Economic Development and Labour Council (NEDLAC) under the Fund for Research into Industrial Growth and Equity (FRIDGE). This project, entitled the „Study to Examine the Potential Socio-Economic Impact of Measures to Reduce Air Pollution from Combustion‟, was completed in 2004 and has not previously been published in the open literature (Scorgie et al., 2004a, 2004b, 2004c, 2004d).  Compilation of an electronic database of all Registration Certificate holders under the Atmospheric Pollution Prevention Act, Act 45 of 1965 (APPA) and development of a procedure for the prioritisation of APPA Registration Certificates for early review (Scorgie, 2006; Baird and Scorgie, 2006). Work was completed as part of the APPA Registration Certificate Review Project commissioned by the Department of Environmental Affairs (DEA) Chief Directorate: Air Quality Management.  Review of international literature pertaining to the selection of industrial activities requiring air pollution regulation and the setting of minimum emission limits for such activities, with recommendations made for applications within South Africa. This research was conducted as part of the Air Quality Act Implementation: Listed Activities and Minimum Emission Standards Project commissioned by the DEA Chief Directorate: Air Quality Management (Scorgie and Kornelius, 2007).  Review of systems and processes developed internationally for air quality compliance monitoring and compilation of a Guideline Inspection Protocol for Listed Activities holding Atmospheric Emission Licenses under the National Environmental Management: Air Quality Act (Scorgie, 2008). This work was conducted as part of a broader Compliance Monitoring Project commissioned by the DEA Compliance Monitoring Directorate aimed at the development of national systems for monitoring compliance with prioritised environmental quality and protection legislation, regulations and authorisations.

1.5 Recent Transformations in Air Quality Practices This thesis integrates and expands on the findings of several research projects completed by the author during the period (2003 – 2009). Since completing the earlier work air quality management in South Africa has been transformed, including changes in the way in which the Air Quality Act is implemented, and partially informed by the studies documented in this thesis. Examples of the regulatory developments that have taken place in recent years include the following:

 Promulgation of the National Ambient Air Quality Standards in 2009;  National Framework for Air Quality Management in South Africa, 2007;  Promulgation of the national emission limits for listed activities, as prescribed in Section 21 of the Air Quality Act;  Declaration of national air quality priority areas (Vaal Triangle, Highveld and Waterberg), and air quality management plans being compiled for these areas;  Development of the South African Air Quality Information System; and  Publication of air quality management plan progress reports in the National Air Quality Officers‟ Reports since 2008; and  Implementation of the transition of APPA permits to air quality licences under the AQA. The studies documented in this thesis have shaped several of the recent air quality management developments. There is merit in integrating, expanding, documenting and reflecting on such studies for the following reasons:

 The methods, assumptions, limitations and overall outcomes of studies which provided the basis for air policy decision making are described and critically evaluated in terms of their scientific merit given data uncertainties and sensitivities.  A number of the recommendations supported by the studies completed in earlier years have not been implemented to date. In certain cases this may be due to regulators applying a phased approach to implementation, such as in the case of national emission standard setting. Recommended actions for possible future implementation to support continued improvements in air quality management are highlighted in the thesis.  In other cases, actions by regulators may still be pending, or may be considered not sufficiently efficient or too onerous, complex or costly for implementation within South Africa. By retrospectively evaluating such cases, the author is able to reflect on the practicability of various of the air quality management measures being investigated.

1.6 Structure of Thesis Given that the work presented comprises the integration of pertinent components of several individual research projects completed by the author, the overall procedure and structure of the thesis does not follow convention. However, each study component is comprehensively documented, covering the theory underpinning the research, data and methods, results, discussions and conclusions. The findings of the individual study components are synthsised in the final chapter to address the central hypothesis of the thesis and provide a synopsis of recommended future research requirements. A consolidated reference list is provided. The structure of the thesis is provided below.

Chapter 2: Air Quality over South Africa

An overview is given of the main air quality problems and emerging issues within South Africa. While recognising the importance of regional and global issues related to air emissions, this study focuses primarily on urban air quality. Trends in ambient air quality are assessed based on selected air quality monitoring datasets, and reference made to the literature to determine the significance of indoor air pollution within fuel-burning households. Major air emission sources, priority air pollutants and communities at risk form the focus of the study.

Chapter 3: Air Quality Management in South Africa

As the rights of citizens to an environment not harmful to health or well-being having been entrenched by the South African Constitution, authorities have a duty to intervene. However, the challenge is not simply to deliver cleaner air, but to do so without hindering social and economic development. Opportunities for meaningfully addressing air pollution challenges within the context of the emerging air quality management discipline are explored. Based on the experiences of other countries, advances such as the harmonisation of air quality and development policy and planning, and integration of market measures and regulatory approaches, are considered. Specific attention is paid to potential interventions for major anthropogenic fuel-burning sources which are associated with the greatest health risks.

Chapter 4: Calculation of Externalities due to Fuel-burning Sources

The methodological approach developed to quantify externalities related to atmospheric emissions from major anthropogenic fuel-burning sources is documented. This systematic, damage function approach comprises the estimation of emissions, and the linking of such emissions to changes in air quality and associated health effects and costs.

Chapter 5: Fuel-burning Emission Estimates for Major Conurbations

Results are presented from the inventory of fuel-burning sources within various major conurbations and regions including: Johannesburg, Tshwane, Vaal Triangle, the Mpumalanga Highveld, eThekwini and Cape Town. Pollutants estimated include various criteria pollutants such as particulate matter, sulphur dioxide, nitrogen oxide, carbon monoxide and lead, in addition to several organic compounds, viz. benzene, formaldehyde, acetaldehyde and 1,3-butadiene. Emissions were estimated for the 2002 base year, with emission projections for 2011 provided based on „business as usual‟ assumptions.

Chapter 6: Health Effects and Associated Costs due to Fuel-burning

Health effects and associated financial costs due to inhalation exposures to anthropogenic fuel- burning sources are presented for various health endpoints, including respiratory hospital admissions, cardiovascular hospital admissions, premature mortality, chronic bronchitis, restricted activity days and cancer cases. Other risks of harm associated with fuel-burning are also discussed including risks due to fires, burns and accidental poisoning. Source contributions to total health effects and costs are assessed and provide the basis for the prioritisation of sources for mitigation purposes.

Chapter 7: Costs and Benefits of Mitigating Fuel-burning Sources

An overview of sector-specific interventions being given in Chapter 3, the specific mitigation measures selected for cost-benefit analysis are presented in this chapter. Health effect reductions and associated reduced costs achievable through the implementation of such measures are analysed and serve to inform the design of cost-optimised packages of air pollution mitigation measures.

Chapter 8: Prioritisation and Compliance Monitoring of Industry

Industry undertaking so-called „Scheduled Processes‟ were historically required to hold Registration Certificates under the APPA. This classic „demand and control‟ method of regulating industries using permits issued by a regulatory authority is to be continued under the AQA with a number of improvements. Under AQA, „Listed Activities‟ will be required to hold Atmospheric Emission Licenses, with one license being issued per site and all emissions including fugitive releases being regulated. The APPA to AQA transition phase requires the prioritisation of industries for early Registration Certificate review, the setting of national emission standards to inform emission limits to be included in Atmospheric Emission Licenses and procedures for monitoring compliance with such licenses. Based on the review of international experience and knowledge of local circumstances, procedures developed for the prioritisation and compliance monitoring of industry are presented and recommendations made regarding the setting of national emission standards.

Chapter 9: Phasing of Air Quality Management Planning by Local Authorities

The Air Quality Act marks a shift from national air pollution control based on source controls to decentralised air quality management through a receiving environment approach with local authorities required to include Air Quality Management Plans (AQMPs) in their Integrated Development Plans. Lessons learned during the process of assisting the cities of Johannesburg and Cape Town and the Ekurhuleni Metropolitan Municipality in the development of their first AQMPs are discussed and guidance given for the phasing in of such planning in other urban areas.

Chapter 10: Conclusions and Outlook

Results of the preceding chapters are synthesised into an overview. An evaluation is given of the extent to which the various components of technical, legal and economic analyses of air quality management in South Africa are able to answer the central hypothesis of whether rational air quality management systems are compatible with continued socio-economic prosperity and development.

The specific contributions made by the author to the air quality governance cycle is schematically illustrated in Figure 1 with reference made to research integrated within this thesis and related work drawn upon but not documented in detail in this thesis. All components of the environmental governance cycle must be in place and operating effectively for successful air quality management to be achieved.

Figure 1: Contributions to the evolving system of air quality governance in South Africa (adapted DEA, 2007b).

2 Air Quality over South Africa

This chapter provides an overview of the main air quality problems and emerging issues within South Africa. While recognising the importance of regional and global issues related to air emissions, this thesis focuses primarily on urban air quality. Trends in ambient air quality are assessed based on selected air quality monitoring datasets, and reference made to the literature to determine the significance of indoor air pollution within fuel-burning households.

2.1 Atmospheric Emission Sources Sources of trace substances in the atmospheric include both natural and anthropogenically induced releases. Natural sources include biogenic releases, wind blown dust emissions, wild fires and lightening induced NOx formation. Common anthropogenic sources of atmospheric emissions in South Africa are as follows (Held et al., 1996; Scorgie et al., 2004a):

 Industrial and commercial activities – Scheduled Processes (i.e. processes identified in the Second Schedule of the Atmospheric Pollution Prevention Act, Act 45 of 1965 (APPA) as potentially resulting in significant atmospheric emissions) and non-domestic fuel burning appliances operated by businesses, hospitals and schools. Industrial process emissions include stack and vent emissions in addition to a range of possible non-point sources, including evaporative emissions of volatile organic compounds from chemical transport, handling and storage, and fugitive dust releases from materials handling, vehicle entrainment and wind erosion of stockpiles.  Electricity generation – specifically power stations for the national grid. Over 90% of South Africa‟s electricity is generated by large coal-fired power stations, many of which are concentrated on the Highveld (Winkler, 2006). The stacks of these power stations are typically >250 m high. While particulate matter emissions are removed with >99% efficiency, there is no mitigation of sulphur dioxide and nitrogen oxide emissions. Despite the height of the stacks, downwash during unstable atmospheric conditions leads to occasional high ground level pollution concentrations (Eskom, 2002).  Residential – comprising household combustion of coal, paraffin, LPG, dung and wood. Electricity accounted for only 38% of the total energy consumed by the residential sector during 2000, the remainder having been provided by the combustion of wood (41%), coal (35%), paraffin (13.9%), vegetable wastes (6.9%) and liquefied petroleum gas (LPG) (2.9%), and the use of solar energy (0.3%) (ERI, 2001).  Transport – including petrol and diesel driven vehicle tailpipe emissions, vehicle entrained road dust, brake and tyre wear fugitives; and rail, shipping and aviation related emissions.  Biomass burning (agricultural and wild fires) – sugarcane, crop-residue burning and general wild fires (veld fires) represent significant sources of combustion-related emissions associated with agricultural areas. Between 70% and 90% of all vegetation fires are considered to be anthropogenically induced (Helas and Pienaar, 1996).  Mining – fugitive dust releases from processing and wind erosion, gaseous and particulate matter emissions from blasting, and spontaneous combustion emissions.  Waste treatment and disposal – waste industries of interest include waste incineration, landfills and waste water treatment works.

 Agricultural (non-combustion) – includes enteric fermentation; and fertiliser and pesticide application.  Informal / miscellaneous – tyre burning, informal waste combustion, fugitive dust from construction, and wind erosion of vegetation-denuded areas.

Combustion-related emissions are estimated to be responsible for over 80% of the criteria pollutants released by anthropogenic sources within urban areas (Scorgie, 2004a; Scorgie and Watson, 2004), and result in the greatest contribution to human exposure. Hence combustion emissions, industrial and domestic, have been selected as the main focus for quantification, impact assessment and mitigation measure evaluation in the subsequent chapters of this thesis.

2.2 Ambient Air Quality

2.2.1 Overview of Monitoring and Data Selection During 2004, ambient air quality monitoring was being conducted by 35 different agencies in South Africa, each employing a variety of monitoring methodologies and approaches. The location of the various continuous air quality monitoring stations operating during this year are illustrated in relation to population densities in Figure 2. Most air quality monitoring activities are conducted by metropolitan councils or industry.

Figure 2: South African population distribution and the relative location of the ambient air quality monitoring stations (Annegarn et al., 2007)

The most commonly monitored compounds are sulphur dioxide (SO2), oxides of nitrogen (NOx) including nitrogen dioxide (NO2) and nitric oxide (NO), ozone (O3) and particulate matter (PM).

Particulate matter is measured as ambient concentrations of PM10 (particulate matter with an equivalent aerodynamic diameter of less than 10 microns), PM2.5 (particulate matter <2.5 microns) and total suspended particulate matter (TSP), and as dust deposition.

Other pollutants measured include lead, carbon monoxide (CO), volatile organic compounds (VOC) (generally as benzene, toluene, ethylbenzene and xylene, BTEX, or just benzene), hazing index, selected metals (chromium, hexavalent chromium, manganese) and reduced sulphur compounds such as hydrogen sulphide (H2S) and total reduced sulphur (TRS). A detailed description of air quality monitoring across the country is documented in a report entitled Technical Compilation to Inform the State of Air Report (DEAT, 2006).

Air quality challenges related to criteria pollutants are illustrated using data from continuous air quality monitoring stations for 2004. This is the latest year for which verified data were available during the analytical phase of this research. Monitoring stations with reliable datasets, characteristics of selected environments, such as heavy industrial areas, dense traffic sites and residential fuel burning were selected. Preference was given to complete and reliable data sets, particularly data from stations likely to continue operating to facilitate future comparisons. Unfortunately long-term monitoring data sets able to demonstrate inter-annual air quality trends for the environments selected were sparse (Annegarn et al., 2007). Where available, trends identified are noted.

In evaluating ambient criteria pollutant levels, reference is made to air quality limits published within Standards South Africa (SANS 1929:2004), which were adopted in 2009 as national air quality limits by the Department of Environmental Affairs (DEA, 2009)(1). For some pollutants, the DEA has introduced interim air quality limits to be met in the short term (prior to 2015). SANS limits are comparable to European Community (EC) limit values with the exception that higher (more lenient) threshold values are adopted for fine particulate matter concentrations and annual average SO2.

Pollutants which have historically been monitored are PM10, SO2, NO2, O3, lead and CO. Benzene and PM2.5 measurements have mainly been restricted to campaigns, although continuous monitoring of these components is increasingly being implemented as part of municipal air quality management programmes.

Air quality limit exceedances have been noted to occur due to PM10, PM2.5, SO2, NO2, O3 and benzene. Ambient lead levels have reduced due to the phasing out of leaded fuels. CO concentrations, despite being well above background levels in residential coal burning areas and near busy roadways, are within limits.

1 South African Government Gazette, No. 32816, 24 December 2009.

Assessments of non-criteria pollutants are based on information from past monitoring campaigns. Whereas such campaigns have focused on VOCs, monitoring of metals (other than lead) is limited and insufficient to adequately assess exposure potentials.

2.2.2 Suspended Fine Particulate Matter

Despite high PM10 levels having been observed across the country, and by implication posing elevated associated health risks, PM10 was only recorded at about thirty stations during 2004. The SANS daily limit of 75 µg m-3 was exceeded at almost all monitoring stations, with the annual limit of 40 µg m-3 exceeded at over 40% of stations. (These air quality limits have been adopted as national limits by the DEA (2009) with the compliance date set for 1 January 2015.)

Significantly high fine particulate matter concentrations occur within residential coal and wood -3 burning areas with peak daily-average PM10 levels of as high as 500 µg m . “High” (i.e. 75 to 100 -3 -3 µg m ) to “very high” (>100 µg m ) pollution levels occur frequently with the daily PM10 limit of 75 µg m-³ exceeded on 40% to 60% of days within coal-burning areas (Figure 3).

High PM10 levels occur near heavy industry, such as within the industrialised Vaal Triangle region. -3 Peak daily-average PM10 levels (up to 200 µg m ) and air quality limit exceedances (~20% of days) are however generally lower compared to household fuel burning areas (Scorgie et al., 2003a).

-3 -3 Figure 3: „Moderate‟ (daily-average PM10 50–75 µg m ), „high‟ (75-100 µg m ) and „very high‟ (>100 µg m-3) pollution days at selected sites (2004)

Although PM10 concentrations recorded at various traffic-sites exceed air quality limits, research conducted on the Highveld indicates that only about 20% are related to vehicle exhaust emissions, with the remainder due to other sources, such as entrainment, residential fuel burning and industry

(Burger and Thomas, 2002). PM10 episodes at the City of Johannesburg‟s Buccleuch traffic station, for example, coincide closely with household coal burning within neighbouring Alexandra. Coal

burning emissions are channelled along the Jukskei River valley towards the Buccleuch station.

Whereas NOx and benzene levels recorded at Buccleuch peak during the daytime with diurnal trends indicative of traffic flow patterns, PM10 episodes occur at night and mainly during winter.

Insufficient PM2.5 monitoring datasets are available to make a systematic assessment of health risk potentials within the environments selected. Air quality limits for this size fraction are yet to be (2) established in South Africa . Reference may however be made to PM2.5 levels recorded during campaigns within residential coal burning areas and to preliminary continuous monitoring station measurements.

Ambient PM2.5 monitoring was undertaken in a residential coal burning suburb of Soweto during -3 -3 1997. Seasonal average PM2.5 levels varied from 20 µg m during summer to 150 µg m during the -3 -3 wintertime. Daily average winter PM2.5 ranged from 50 µg m to 300 µg m . Seasonal variations in PM2.5/PM10 ratios were indicative of contributing source types (Figure 4). The strongest 2 relationship between PM2.5 and PM10 concentrations occurred during winter (R = 0.99), with 2 spring being characterised with the weakest relationship (R = 0.78). PM2.5 was estimated to represent 37%, 48% and 62% of PM10 during spring, summer and winter, respectively. Household coal burning contributed significantly to winter particulate matter loadings resulting in the increased percentage of fines. Wind-blown emissions during spring accounted for the larger coarse material fraction, with intermittent gusts responsible for the poor relationship between PM2.5 and

PM10 size fractions (Annegarn et al., 1999).

Figure 4: Seasonal variations in PM10/PM2.5 ratios recorded within a residential coal burning suburb of Soweto, Johannesburg, 1996-7 (Annegarn et al., 1999)

2 The DEA gazetted draft standards for PM2.5 in August 2011. South Africa Government Gazette, No. 34493, 5 August 2011, Proposed National Ambient Air Quality Standard for PM2.5.

PM2.5 levels were measured during 2004 to comprise about 70% of PM10 at the Johannesburg Buccleuch Station. Sources contributing at this station primarily include household coal burning and traffic. Average PM2.5/PM10 ratios of 0.6 were recorded within Cape Town during the 1995-6 Brown Haze Campaign. Significant contributions of combustion sources, specifically vehicle tailpipe emissions, were noted (Wicking-Baird et al., 1997).

Regionally-transported, aged aerosols contribute significantly to background air pollutant concentrations, particularly over the interior. Source apportionment studies have identified four major contributing source types of regional significance to the atmospheric aerosol loading over South Africa: aeolian crustal material consisting of mineral soil dust; marine aerosols from the two adjacent oceans; biomass burning particles from wild fires occurring mainly north of 20S; and secondary aerosols from industrial emissions. Emissions from these sources have been observed in the past at remote sites in South Africa (Annegarn, et al., 1992; Piketh, 1995; Piketh et al., 1996; Salma et al., 1992; Maenhaut et al., 1996).

2.2.3 Sulphur Dioxide Concentrations

SO2 is the most widely measured pollutant, with over eighty continuous monitoring stations operational in South Africa during 2004. In addition, a number of passive diffusive monitoring campaigns were undertaken.

SO2 limit exceedances were recorded at various industry-related monitoring stations (Figure 5), including stations near coal-fired power stations (Kendal 2 Station), refineries (Southern Works Station) and platinum smelting operations (Waterval Station).

-3 Figure 5: Frequency of exceedances of the hourly SO2 limit of 350 µg m at selected sites during 2004

The Kendal 2 Station, situated on the Mpumalanga Highveld, is in the area most frequently impacted by plumes from large-scale coal-fired power stations, thus representing the worst-case ground level pollution contribution of such activities. Fortunately, most power stations have historically been located at remote sites. Increased population densities within potential impact zones of existing power stations, and proposed new coal-fired power stations within the same airsheds have raised concerns, and led to discussion on the need to retrofit flue-gas desulphurisation on existing power plants.

Exposure potentials are higher in areas where large industrial emitters are located close to residential areas, e.g. Durban South (Wentworth, Southern Works and Settlers stations). In such areas more emphasis is placed on interventions aimed at reducing air quality limit exceedances to protect human health.

Figure 5 highlights SO2 issues due to petrochemical, power generation and mineral processing operations. Other industries likely to result in SO2 limit exceedances include pulp and paper and metallurgical processes.

Despite raised SO2 levels within household coal burning areas (e.g. Orange Farm Station), air quality limit exceedances are infrequent, with annual average SO2 levels only ~30% of the annual air quality limit of 50 µg m-3. This is due to the relatively low sulphur content of coals (typically

<1% S). SO2 levels in wood-burning areas such as Khayelitsha are also well within air quality limits.

Ambient SO2 levels are typically low at traffic sites and non-fuel burning residential areas where other source contributions are negligible – eThekwini City Hall, Goodwood (Figure 5).

2.2.4 Nitrogen Dioxide Concentrations

NO2 is measured at approximately forty stations within four metropolitan areas, viz. Cape Town, eThekwini, Johannesburg and Tshwane. Stations are sited to measure general urban and industrial pollution.

The highest NO2 levels occur in areas affected by high traffic densities including areas adjacent to busy highways (e.g. Buccleuch Interchange in Johannesburg), within CBDs (e.g. eThekwini and

Cape Town City Hall stations) and residential areas with significant vehicle activity. Elevated NO2 levels also occur within certain industrial areas, primarily due to fuel burning related activities (Figure 6).

NO2 limits are not generally exceeded within residential wood-burning areas (e.g. Khayelitsha).

Although insufficient recent NO2 monitoring data are available to assess compliance within residential coal-burning areas, historical monitoring indicates non-compliance is unlikely (Scorgie et al., 2001).

NO2 limit exceedances, where they do occur, are presently limited to infrequent violations of short- term thresholds (Figure 6). NO2 levels along busy traffic routes within certain metropolitan areas, such as Johannesburg and Ekurhuleni, have however been noted to have increased significantly over the past decade (Scorgie et al., 2003b).

-3 Figure 6: Frequencies of exceedance of the hourly NO2 limit of 200 µg m at selected sites during 2004

2.2.5 Volatile Organic Compounds Several VOC monitoring campaigns have been undertaken for urban, traffic, industrial and residential areas. High VOC levels were recorded within residential coal burning areas, near large filling stations, busy roadways and petrochemical and chemical operations (John et al., 1998; Sowden, 1998; Taljaard, 1998; Burger and Thomas, 2002).

On-going VOC monitoring is restricted mainly to benzene, toluene, ethylbenzene and xylene (BTEX). Continuous monitoring of such compounds was initiated at four stations during 2004. Passive monitoring networks have also been established, e.g. eThekwini. Toluene, ethylbenzene and xylene levels were recorded to be within chronic health limits (Annegarn et al., 2007).

Annual benzene levels recorded are usually in the range 3 to 9 µg m-3. Exceedances of the SANS annual limit of 5 µg m-3 were recorded close to highways and busy intersections, at filling stations, near petrochemical plants and tank farms, and within residential coal burning areas (Annegarn et al., 2007; eThekwini Health, 2007). Peak period-average benzene levels within residential coal- burning areas and near petrochemical plants of up to 30 µg m-3 have been recorded during monitoring campaigns (Scorgie et al., 2001, 2003a).

H2S levels measured near certain petrochemical plants and waste water treatment works, primarily to assess odour potentials, are recorded to have exceeded not only odour thresholds, but also acute and chronic health thresholds (Scorgie, 2004a; Scorgie and Watson, 2004).

2.2.6 Ozone Concentrations Violations of the SANS hourly average ozone limit of 200 µg m-3 occur at several of the approximately forty stations, with frequencies in the range of 0.03% to 3% of hours. The highest

O3 concentrations and most frequent exceedances recorded in 2004 occurred in Johannesburg at the Buccleuch station (Annegarn et al., 2007).

2.2.7 Diurnal and Seasonal Trends Indicative of Source Types Distinct diurnal and seasonal trends are apparent in air pollution concentrations occurring as a result of certain source types. Such trends are a function of variations in source activity and in meteorology and have significant implications for exposure potentials.

a) Household Fuel Burning To demonstrate diurnal trends in domestic coal burning emissions, reference is made to aerosol black carbon (BC) concentrations measured during a monitoring study conducted in Soweto (Annegarn et al., 1999) (Figure 7). Time-series analysis of aerosol black carbon (BC) concentrations, as a tracer of domestic coal burning emissions, revealed the occurrence of morning and evening maximum concentrations and very low afternoon concentrations. Between 12h00 and 16h00, average BC concentrations were observed to be low (2 to 3 µg m-3). A sharp increase in the black carbon concentration was found to begin at 17h00, with concentrations peaking typically in the range 30 – 50 µg m-3 by 18h00. A secondary, smaller peak was observed at 21h00 or 22h00, representing the impact of a final evening addition of coal to the stoves. A subsequent gradual decrease in concentrations until 06h00 was evident, following which concentrations increase to reach a maximum of about 20 to 30 µg m-3 at 08h00.

Figure 7: Diurnal variations in PM2.5 black carbon concentrations, as observed in Soweto during the period 1 to 20 June 1997 (Annegarn et al., 1999)

The evening peak is due to the lighting of night-time fires for the purposes of cooking and space heating. The slow clearance rate of the smoke during the night hours is due to the atmosphere being very stable, with stagnant winds (<0.2 m/s), so that emissions tend to accumulate in a shallow layer close to the ground, and the fires being left to burn for an extended time for space heating. The morning peak is due to households lighting their morning fires for cooking purposes.

The rapid reduction in atmospheric BC following the morning peak is a function of morning fires being more quickly extinguished, and of the rapid atmospheric clearing rate following the dissipation of the nocturnal inversion soon after sunrise and the onset of convective mixing.

Air pollution related to residential fuel burning peaks during winter months due primarily to the increase in fuel burning for space heating purpose, with the accumulation of air pollution enhanced due to the occurrence of more intense surface-based inversions and stable atmospheric conditions (Figure 8).

Figure 8: Daily average PM10 concentrations recorded by the City of Johannesburg at the Oliver Tambo Clinic in Diepsloot during 28 June to 24 November 2004

b) Vehicle Traffic Diurnal trends in air pollutant concentrations at sites of high traffic activity may be demonstrated with reference to eThekwini‟s Warwick monitoring station (Figure 9). Diurnal trends in hourly average NO, which is emitted as a primary pollutant, and NO2, which occurs predominantly as a secondary pollutant formed through chemical conversion of NO in the atmosphere, are illustrated in Figure 9. Sharp, distinctive morning peaks and lower and more spread evening peaks typical of rush hour patterns are evident.

Seasonal trends do not typically occur in vehicle-emission related air pollutant concentrations with the exception of marginally higher or lower concentrations during the holiday months depending on the location of the station.

Figure 9: Diurnal variations in hourly average NOx and NO2 concentrations recorded by eThekwini Metropolitan Municipality at its Warwick Station during 2004 (Annegarn et al., 2007)

c) Elevated Stack Emissions Peak ground level concentrations resulting from elevated stack releases typically occur during the late morning or early afternoon; the dissipation of the nocturnal surface-based inversion and the onset of convective mixing resulting in the plume being „brought to ground‟ at this time. The exact time of day and location of ground level maximums is a function of the height of the stack and the prevailing meteorology.

2.2.8 Central Witwatersrand Dust Deposition Trends Tends in dust deposition are illustrated through reference to dust deposition sampling information for the Central Witwatersrand where wind blown dust from mine tailings impoundments has historically been a source of concern.

Dust deposition monitoring using the ASTM 1739-70 (1970) open bucket dust deposition method commenced in 1985, with six monitoring sites. As the number of gold mine reclamation sites have increased, the monitoring network has expanded to ~250 sites around gold mine tailing reclamation sites, coal and other mines, and heavy industries. Dust deposition values were evaluated historically using the Department of Environmental Affairs dust deposition guideline categories: SLIGHT (<300 mg/m2/d), MODERATE (300 to 500 mg/m2/d), HEAVY (500 to 1,200 mg/m2/d) and VERY HEAVY (>1,200 mg/m2/d).

In terms of the Standards South Africa air quality limits (SANS 1929:2004) dust deposition rates are evaluated in four bands, with specification of permissible frequencies of exceedances and

actions to be taken. The bands are: RESIDENTIAL (<600 mg/m2/d), INDUSTRIAL (600 to 1,200 mg/m2/d), ACTION (1,200 to 2,400 mg/m2/d) and ALERT (>2,400 mg/m2/d).

Figure 10 shows the number of installed monitoring sites, and the total number of monthly dust deposition values in the ACTION and ALERT bands. The number of ACTION and ALERT dust deposition incidents peaked in the period 1994 to 1998 and has decreased significantly since then. This graph is not indicative of conditions on the West Witwatersrand and Boksburg, where dust mitigation measures and monitoring have not been carried out with similar intensity.

Figure 11 indicates the number of installed monitoring sites, and the number of monthly dust deposition values in the ALERT band. The ALERT levels during the period 1996 to 1998 were due to two well documented sites where dust control procedures were not implemented. Both instances resulted in legal actions being instituted by affected parties against the mine operating company. Following mitigation, ALERT level dust emissions did not re-occur.

The percentage of ACTION and ALERT dust deposition incidents as a fraction of the total number of installed dust monitoring sites is depicted in Figure 12. As the majority of the monitoring sites are located in the vicinity of active mining sites, the relative reduction of high dustfall incidents, indicated by the trend lines, indicates progress in managing and reducing dust impacts associated with surface miningNumber on of the Monitoring Central Witwatersrand.Sites; and Number of Action & Alert Dustfall Values per Year Central Witwatersrand 70 Number of monitoring sites (12 samples per year)

Number of ACTION and ALERT level dustfalls 60 Trend line - ACTION > 1,200 mg/m2/d

30 Number Number per year

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2002 2003 2004 2005

Figure 10: Number of installed dustfall monitoring sites on the Central Witwatersrand and total number of monthly dustfall values in the ACTION and ALERT bands (Annegarn et al., 2007)

Figure 11: Number of installed dustfall monitoring sites on the Central Witwatersrand and total number of monthly dustfall values in the ALERT band (Annegarn et al., 2007)

Figure 12: Percentage of ACTION and ALERT dustfall incidents as a fraction of the total number of installed dust monitoring sites (Annegarn et al., 2007)

Dust deposition on the Witwatersrand is strongly influenced by influences of weather acting on exposed surfaces of gold mine sand dumps and slime dams. High wind speeds occur during spring and summer, with potential to generate dust. However, from November through March, frequent rain-falls keep the surfaces moist and limit dust generation. During drier autumn and winter months

(April through mid-July), winds are calm and rarely disturb the dried out surfaces. At the onset of spring (mid-July through mid-October), higher wind speeds acting on dry surfaces generate the maximum amount of dust, as evidenced by the maximum recorded dustfall incidents in these months (Figure 13).

Figure 13: Number of dustfall measurements recorded on the Central Witwatersrand during the period 1985 to 2005 which are in INDUSTRIAL, ACTION and ALERT ranges (Annegarn et al., 2007)

2.2.9 Regions Characterised by Poor Air Quality High ambient air pollution concentrations have been noted to occur in various regions of South Africa. A comprehensive inventory of such sites is currently restricted by the availability of air quality monitoring information. Most monitoring information is for areas within the large metropolitan municipalities with a range of monitoring also having been conducted by industries such as Eskom and Sasol on the Highveld plateau (Figure 2).

Regions where air pollutant concentrations in excess of health thresholds have been observed to occur are illustrated in Figure 14 and the specific pollutants resulting in such exceedances noted (Annegarn et al., 1999; Held et al., 1996; Burger and Thomas, 2002; Scorgie, 2004a; Scorgie and Watson, 2004; Scorgie et al., 2001; Scorgie et al., 2003b; eThekwini Health Department, 2004; Scorgie et al., 2004f; Scorgie et al., 2005; Liebenberg-Enslin and Petzer, 2005; Annegarn et al., 2007). Areas and pollutants indicated should not be viewed as exhaustive but merely indicative of the nature and extent of air quality challenges faced.

Figure 14: Identification of some of the regions where elevated air pollutant concentrations in excess of health thresholds have been measured to occur (author‟s own figure).

2.3 Indoor Air Quality in Fuel-Burning Houses and Related Health Risks Household fuel burning is significant, not only as a contributor to ambient air pollutant concentrations, but due to resulting high indoor air pollution exposures and associated adverse health effects.

Indoor particulate matter levels are recorded to be well above health limits. Although outdoor SO2,

CO and NO2 levels generally do not exceed air quality limits in residential fuel burning areas, notable violations of health thresholds occur indoors (Terblanche et al., 1994; van Niekerk & Swanepoel, 1999). Indoor VOC levels range notably depending on fuel characteristics and burning practices. Compounds such as benzene, carbon tetrachloride and tetrachloroethylene have been measured to exceed health limits (Taljaard, 1998).

Health effects related to particulate matter inhalation include exacerbation of existing pulmonary disease, oxidative stress and inflammation, changes in cardiac autonomic functions, vasculature alterations, translocation of particulate matter across internal biological barriers, reduced defence mechanisms and lung damage. Most of the effects due to particles are associated with the exacerbation of existing disease states (CEPA, 1998; CARB, 2002; Morawska et al. 2005; Pope and Dockery, 2006; WHO, 2007). Effects of exposures to organic compounds range from irritation effects on the eye and dermal system to cancer risks (van Niekerk, 1998). Health risks related to pulmonary and cardiovascular systems occur due to SO2 exposures (van Niekerk, 1998).

Ample evidence exists that household fuel burning has serious adverse effects on health, particularly acute and chronic respiratory health (Terblanche et al., 1992; Terblanche et al., 1993;

Terblanche & Pols, 1994; Ehrlich & Kalkoff, 1998; van Niekerk, 1998; Mathee et al., 2001; Mathee & von Schirnding, 2003). Indoor air pollution from coal burning has been established as an important risk factor for the development of acute respiratory infections (ARI). Epidemiological data indicate that ARI is a leading cause of death in black South African children (Terblanche et al., 1992; Terblanche et al., 1993). Lifetime cancer risks within residential coal burning areas due to exposures to carginogenic compounds released were estimated to be greater than 10-4 for haematopoietic, hepatic and pulmonary systems (Scorgie et al., 2001).

2.4 Summary Characterisation of air quality on a national basis has been historically restricted by the limited number of air quality monitoring stations, and the the unavailability of comprehensive emissions inventories and dispersion modelling studies in most areas. In this study use was made of available monitoring data and emission inventories and modelling work undertaken during individual projects to partially overcome these challenges. The use of integrated Air Quality Monitoring Systems (comprising emissions inventories, monitoring and modelling) to address critical information gaps and support ongoing air quality management is further addressed in Chapter 10.

Based on available information, including previous studies undertaken by the author, it was possible to identify major sources, priority air pollutants and significantly impacted areas to provide a focus for this study. Key findings from the national state of air quality assessment are as follows:

 Human health effects related to inhalation exposures to household coal and wood burning emissions remain the most serious and pressing national air pollution problem.  High ambient sulphur dioxide and fine particulate matter concentrations due primarily to fuel burning within the household, industrial and power generation sectors represent on-going air pollution problems in many parts of South Africa.

 Elevated fine particulate matter (PM10) concentrations occur across the country, with widespread and frequent exceedances of air quality limits. Air quality limit exceedances due to sulphur dioxide are more localized (in vicinity of significant sources) and less frequent.  Collocation of heavy industries and communities presents a continued source of health risks and consequent conflict, exacerbated by increased pressure to place residential areas within industrial and mining buffer zones.  Emerging air pollution issues are associated with the transportation sector, particularly road transportation. The growth in vehicle activity and the aging of the national vehicle fleet is projected to offset planned and proposed national emission reduction measures aimed at the regulation of fuel composition and new vehicle technology.

 Although air quality limits for NO2 and O3, protective of acute health effects, are relatively infrequently exceeded within South African cities, a notable increasing trend in the concentrations of these pollutants is apparent in certain areas. The growth in vehicle activity is anticipated to contribute significantly to this trend.

 Volatile organic compound releases from fuel filling stations and nitrogen oxide and hydrocarbon releases from major airports represent other transportation sources of potential significance.

 Questions remain regarding the potential for environmental effects and transboundary pollution transportation due to medium and elevated stack emissions from petrochemical, metallurgical, mineral processing and coal-fired power stations. Investigations are ongoing with regard to acid deposition rates and impact potentials on the Highveld.  Potential risks due to exposures to heavy metals such as mercury, hexavalent chromium and manganese, and organic species, specifically dioxins, furans and PAHs have been identified. Further monitoring and a consolidation of knowledge is however required before conclusions can be drawn. The most pressing air quality management challenges facing South Africa are noted to include:

 meeting more stringent air quality standards, particularly for particulate matter;

 understanding and addressing human health risks posed by exposure to air toxics;  responding to evidence that, for some pollutants, there may be no identifiable threshold exposure below which harmful effects cease to occur;  mitigating air pollution effects that occur disproportionately in low-income communities; and  addressing industrial emissions without negatively affecting society and the economy. Major sources, priority pollutants and communities at risk having been identified, opportunities for addressing air pollution related risks within the context of the emerging air quality management discipline are explored in Chapter 3. Health effects and costs due to major sources are quantified and a cost-benefit analysis of source-specific interventions presented in Chapters 4 to 7.

3 Air Quality Management in South Africa

As the rights of citizens to an environment not harmful to health or well-being having been entrenched by the South African Constitution, authorities have a duty to intervene. This chapter highlights the challenge of effecting improvements in air quality without negatively impacting social and economic development. Opportunities for meaningfully addressing air pollution challenges within the context of the emerging air quality management discipline are explored, with reference made to the experiences gained by countries where such management has been implemented for decades. Specific attention is paid to potential interventions for major anthropogenic fuel-burning sources.

3.1 Air Quality and Sustainable Development The challenge in air quality management internationally is not simply to deliver cleaner air but to do so whilst not adversely affecting society and the economy. South Africa has taken up this challenge as is clearly evident in the objectives of the National Environmental Management: Air Quality Act, Act 39 of 2004:  “To protect the environment by providing reasonable measures for: • The protection and enhancement of the quality of air in the country; • The prevention of air pollution and ecological degradation; and • Securing ecologically sustainable development while promoting justifiable economic and social development.  Generally to give effect to section 24(b) of the Constitution in order to enhance the quality of ambient air for the sake of securing an environment that is not harmful to the health and well- being of people.” The Air Quality Act commits the country to pollution prevention and air quality improvement and maintenance concurrent with socio-economic development and not at the expense of such development. The realisation of this vision requires the careful tailoring of the various regulations being developed and implemented under this framework Act. In support of regulatory development and roll-out, it is beneficial to consider the experience gained and lessons learned by other countries in their implementation of air quality management. In so doing, it is pertinent to consider key successes and failures at the “environment” and “economy” interface, and at the “environment” and “society” interface. Reference is made to experiences of the USA and Europe, where air quality management has been sculpted and implemented for over forty years in the case of the US and over twenty five years in the case of the European Union(3). Examples are also

3 The US Clean Air Act of 1963 and the Air Quality Act of 1967 set Air Quality Criteria, Air Quality Control Regions and made provision for the development and implementation of State Implementation Plans. Under these Acts, Federal Government coordinates efforts through the US Environmental Protection Agency (US- EPA) and sets national air quality standards and approaches to pollution mitigation so that it can provide a basic level of environmental protection to all individuals in the US. State and local governments then develop, implement and enforce specific strategies and control measures to achieve the national standards and goals. Air quality management started in 1980 in the European Union with the publication of Directive 80/779/EEC which set air quality limit values for sulphur dioxide and suspended particulate matter. The 1996 Air Quality Framework Directive and its daughter Directives were a significant step aimed at establishing a harmonized structure for assessing and managing air quality throughout the EU. Within this structure, the EU Member

drawn from Australia and other southern hemisphere countries which have significant advances in air quality management in recent years.

3.1.1 Environmental-Economic Interface In its review of key results in implementing sustainable development during the period 2001-2004, the OECD noted that more ambitious environmental objectives could have been achieved for little or no additional cost. Failure to do so was given as being due to the lack of integration of environmental and economic concerns in policy making. Noting that abatement costs could rise markedly in the future as environmental standards become stricter, and less onerous actions and gross polluters are progressively eliminated, the need to use cost-efficient options in the coming years is increasingly being emphasised (OECD, 2005).

Key mistakes made in the air quality management arena, and specific examples of such, may be summarised as follows:  Failure to integrate air quality considerations into energy, transportation, land use, housing and other development planning processes. This has increased the cost of realising air quality improvements and has impeded the management of emissions from significant sources such as vehicles, residential fuel burning and power generation.

 Failure to integrate economic concerns in air quality management and planning. This has resulted in the strengthening of the environmental pillar of sustainable development at a cost to the economic pillar, as a direct consequence of choosing relatively inefficient policies. Specific components of this weakness include: • Adoption of costly air quality monitoring systems, with emphasis on the implementation of continuous monitoring of “everything” “everywhere”, regardless of the actual state of air quality at specific localities. • Overly ambitious air quality objectives. The failure of countries to systematically analyse costs and benefits of air quality management policies and to integrate findings into decision making has largely been responsible for the adoption of such too ambitious objectives. • Regulations based on the mandated use of a particular technology to realise emission reduction. In many instances such regulations have resulted in very high costs and discouraged alternative cost-saving innovations. • Targeting of sources based on their contributions to total emissions, rather than on their contributions to ambient air pollutant concentrations and more specifically to health and broader environmental effects.

States are given considerable scope to determine the actions they will take to meet their commitment to achieve air quality standards. However, the Member States must at the same time implement the other EU- level measures that comprise the overall EU air quality management system, including stationary source emission controls, technical requirements to limit emissions from motor vehicles, fuel quality standards (etc.).

• Primary focus on „command and control‟ measures, without other mechanisms such as market instruments being formulated and implemented. (This is particularly true in the case of the USA in the earlier years.)  Failure to harmonize local air quality management planning with measures aimed at addressing regional and global air pollution-related issues such as acidification, tropospheric ozone creation, stratospheric ozone depletion and climate change.

The above failures have most successfully been addressed through the implementation of the approaches documented in subsequent subsections.

a) Sustainable Cities The “sustainable cities” concept has been developed where environmental considerations, including air quality management, are systematically integrated into development planning; specifically energy, transportation and land use planning (Deakin, 2011; Cumo et al., 2012). This approach illustrates the recognition of the inherent linkages among transport, environment, energy and spatial development issues and the need for closely coordinated policies. This broadened approach to air quality management is not just being implemented by developed countries but is increasingly being implemented by developing nations who realise that cost-effective and sustainable solutions require more than the adoption of stringent standards or the implementation of advanced air pollution control technologies.

Transport indicators are, by example, being used to demonstrate cities' overall performance in terms of energy, health, and safety. In addition to reducing transport-related air emissions, more compact urban neighbourhoods served by public transport and dedicated walking/cycling networks are more energy efficient, and safer for pedestrians/cyclists, and promote active travel with attendant health benefits (Bournay, 2012). Such findings have been noted for cities as diverse as Shanghai and Copenhagen. In countries such as Australia, the historical design of cities has been based on large houses in lower density, dispersed suburbs, which are highly car dependent. The resource and emission intensity of such cities has been raised as a key barrier to sustainable development (Rauland and Newman, 2001). Measures proposed to transform Australian cities range from the decentralized management of resources using low carbon technologies suitable for local level application, to increasing the density of new urban centres to enable improved public transport infrastructure and related reductions in transport emissions (Rauland and Newman, 2001).

The AQA provides the context for the harmonization of air quality management and development planning. Each national department or province responsible for preparing an environmental implementation plan or environmental management plan in terms of Chapter 3 of the National Environmental Management Act (NEMA, Act 107 of 1998) is required to include an air quality management plan as part of that plan.

National departments required to include an air quality management plan (AQMP) in their environmental implementation plans include: Environmental Affairs; Land Affairs; Agriculture; Housing; Trade and Industry; Water Affairs and Forestry; Transport; and Defence. National departments required to include an AQMP in their environmental management plans include: Environmental Affairs; Water Affairs and Forestry; Minerals and Energy; Land Affairs; Health; and Labour. These plans must contain a description of policies, plans and programmes that may

significantly affect air quality and a description of the manner in which the relevant national department or province will ensure that the policies, plans and programmes will comply with NEMA principles.

In the local government sphere, the AQA stipulates that each municipality must include an AQMP in its integrated development plan (IDP) required in terms of Chapter 5 of the Municipal Systems Act (Act 32 of 2000). In terms of local air quality management planning, success will be gauged in terms of whether pertinent aspects of the AQMP are reflected in the transportation, energy, housing and spatial development policies and programmes documented in other sections of the IDPs.

Although the AQA has introduced an air quality management planning regime that fits within existing planning regimes, it is recognised that much work is still required to ensure that this planning is properly implemented and fully integrated with existing plans. Building on work documented in previous chapters, this challenge is addressed in Chapter 10 of this thesis.

b) Cost-effective Air Quality Monitoring Systems A broad approach to air quality monitoring is increasingly being adopted which comprises various measurement techniques ranging from passive diffusive monitoring and biomonitoring, to continuous, automated, near real-time measurement. The use of cost-effective passive diffusive monitoring campaigns to screen areas and determine the need and most suitable locations for more costly continuous monitoring is commonplace in Europe.

In addition to screening and varied monitoring, emission inventories and atmospheric dispersion modelling are widely used in Europe and elsewhere to supplement monitoring efforts. Using the source and emissions data from inventories, in addition to meteorological, terrain and land use data, dispersion models simulate the dispersion and deposition of pollutants. Some dispersion models are also capable of simulating chemical transformation in the atmosphere. Whereas continuous monitoring is typically able to provide time resolved information on air pollution concentrations at a specific point, dispersion models are able to characterise spatial and temporal variations in air pollution concentrations for the entire domain and period for which source, emission and meteorological data are available.

Monitoring, if conducted properly, has a higher degree of accuracy with monitoring data used to verify the completeness and accuracy of emission inventories and dispersion models. The integration of emission inventories, dispersion modelling and a range of monitoring techniques into comprehensive air quality monitoring systems is an efficient and cost-effective means of baseline air quality characterisation and tracking trends. This concept is further explored within the context of South African cities in Chapter 10.

c) Aligning Air Quality Standards with Sustainable Development The setting of air quality limits at levels that are not achievable using reasonable policy or technical approaches holds the potential of undermining the credibility of the air quality management system that it aims to support.

Common responses internationally to non-compliance with mandatory air quality limits have been to allow digression from the legislation (NCSA, 2003). This has taken the form of, for example, relaxing the timeframe for compliance. For industry-related non-compliance, the most common

application of this is to allow older industrial installations to continue operating until a specified date outside the new regulations in instances where the use of pollution abatement equipment would render them uneconomic.

The EC approach to this problem has been to establish and implement a system of multiple levels of air quality objectives including: limit values, target values and alert thresholds (EU, 1996; EU, 2008). The definition of such objectives is given as follows (EU, 2008):  Limit values are to be based on scientific knowledge, with the aim of avoiding, preventing or reducing harmful effects on human health and the environment as a whole. Limit values are to be attained within a given period and are not to be exceeded once attained.

 Target values are intended to avoid harmful long term effects on human health and the environment. Target levels are to be attained where possible over a given period. Such levels represent a long term goal to be pursued through cost-effective progressive methods, and are frequently termed 'long-term acceptable thresholds'. At these levels pollutants are either harmless to health and the environment, or unlikely to be reduced through expending further reasonable cost on abatement due to background sources or other factors.  Alert thresholds refer to levels beyond which there is a risk to human health from brief exposure. The exceedance of such thresholds necessitates immediate steps.

The EC system also makes provision for margins of tolerance (EU, 2008). A margin of tolerance represents a percentage of the limit value by which this value may be exceeded subject to the conditions to be laid down in the envisaged framework for standard implementation (Figure 15).

The margin of tolerance is reduced each year to reach zero on the date by which the limit value must be met. The purpose of the margin of tolerance is to identify areas with the worst air quality, i.e. Group 1 areas in which the margin of tolerance is exceeded, and to distinguish between areas in which air pollutant concentrations exceed the limit value but are within the margin of tolerance (Group 2 areas) and areas in which concentrations are below limit values (Group 3 areas). Air quality planning and reporting requirements are stipulated per group with more stringent requirements being devised for Group 1 areas. A margin of tolerance is therefore not a deviation from a limit value, but rather provides a trigger for action in the period before the limit value must be met.

The EC system of air quality objectives thus explicitly links air pollution concentrations to appropriate air quality planning and remedial action over suitable time scales. Such a system recognises that improvements in ambient air quality are likely to be brought about over extended time periods, allowing for planning, procurement and installation of abatement equipment, and that long term planning is more effective than corrective actions aimed at instantaneous improvements in air quality, or prosecutions of industrial emitters. Nevertheless, mechanisms and penalties are provided for to allow enforcement in the event of non-compliance with approved emissions plans, or failure to make progressive improvements according to agreed timelines.

The EC approach to the setting and implementation of air quality limits was adopted by Standards South Africa, as documented in its publications SANS 69:2004 Framework for setting and implementing national ambient air quality standards (SANS 69, 2004), and SANS 1929:2004 Ambient Air Quality - Limits for common pollutants (SANS 1929,2004). The work undertaken by

Scorgie et al. (2003d), which included a detailed review of the EC air quality standard setting framework, informed the drafting of the SANS documents.

Figure 15: Schematic illustration of the EU approach to action planning and reporting which is dependent on the air quality status in relation to air quality limit values and taking into account a margin of tolerance (Scorgie et al., 2003d)

d) Prioritisation of Sources Based on Impact and Future Trends The extent and toxicity of emissions is not necessarily a concise indicator of contributions to ground level air pollution concentrations and health and environmental risks. Such contributions are also a function of the height of emission, temporal variations in releases and the proximity of the source to sensitive receptors (i.e. exposure potential).

The significance of household fuel burning emissions is enhanced due to three factors: (i) the low level of emissions; (ii) the coincidence of peak emissions, typically a factor of 10 greater than if total annual emissions were averaged, with periods of poor atmospheric dispersion (i.e. night-time, winter-time); and (iii) the release of such emissions in densely populated areas, with high contributions to both indoor and outdoor pollution concentrations. The significance of vehicle emissions in terms of the contribution to air pollutant concentrations and health risks is similarly enhanced by the low level at which emissions occur and the proximity of such releases to high population densities. Vehicle emissions also tend to peak in the early morning and evenings, at which times atmospheric dispersion potentials tend to be lower.

The significance of fuel burning within industrial and power generation sectors - in terms of their contributions to air pollutant concentrations and public health risks - is frequently lower than would

be expected given the extent of emissions. This is due to these sources generally being characterised by constant, relatively high level releases, with such industries also likely to be more remotely located from residential settlements compared to household fuel burning and vehicle emissions.

The ranking of source significance based on total emissions would, for example, result in the prioritisation of industrial emissions above household fuel burning. However, should the aim be to reduce human health effects, then household fuel burning would be targeted, as is discussed further in subsequent chapters.

Future trends in emissions represent a further consideration in the ranking of sources. Vehicle emissions, for example, were only responsible for infrequent exceedances of air quality limits at sites in proximity to very busy roadways during 2004. Prioritisation of vehicle emissions is however imperative due to significant increases in vehicle activity within many South African conurbations, and the substantial lead time required to implement transport management measures and realise emission reductions for this sector. This is clearly apparent from the experiences of various countries within Europe, North and South America and Asia (Krzyzanowski et al., 2005).

Source prioritisation based on existing and projected future impact potentials is demonstrated through the damage function approach implemented by the author, as documented in Chapters 4 to 7 of this thesis.

e) Cost-benefit Analysis of Emission Reduction Strategies Prior to adoption, air quality standards and emission reduction strategies should be subjected to reviews of their economic and social consequences, and findings integrated successfully into the decision making processes. Reviews need to take into account not only the costs of implementing air quality standards and effecting emission reductions, but also the consequent benefits arising from these actions, for example health cost reductions. This provides the rationale for the cost- benefit analysis undertaken for emission reduction options targeting fuel burning sources in South Africa, as documented in Chapter 7 of this thesis.

In addition to aiding the selection of cost-optimised mitigation measures across source sectors, cost-benefit analysis is also important in informing the types and groupings of pollutants to be targeted. Important lessons can be learned from the European and US experience in this regard. Emission reduction strategies in these countries have historically been developed for individual pollutants. It has however been realised in both Europe and in the US that the implementation of emission reduction strategies for individual pollutants are not the most effective or economically efficient method of reducing health effects due to air pollution (NCSA, 2003). On the basis of this experience it is advocated that the implementation of health assessment methodologies and emission control measures for multiple pollutants be considered in South Africa.

The use of cost-benefit analysis in the targeting of source groupings and multiple pollutants, and the alignment of local air quality management and global climate protection strategy, represent an effective means of achieving the highest level of human and environmental protection in the most cost-effective manner.

f) Flexible Approach to Air Quality Impact Reduction Historically, air pollution control in South Africa has placed emphasis on the implementation of „command and control‟ measures within the industrial sector. The shift from source-based control to receiving environment management under the AQA emphasises the targeting a wider range of sources through more flexible and varied approaches. Approaches being adopted or considered for future implementation include: regulation (e.g. use of Atmospheric Emission Licenses for Listed Activities), market instruments (such as atmospheric user charges and pollution taxes), potential for voluntary agreements, education and awareness campaigns, and emissions trading.

International experience has shown that the adoption of a mix of instruments and interventions is more effective than a single instrument in realising air quality improvements across various source types. Although direct regulation has remained an important instrument in the control of industrial sources, experience indicates that emission limits should be specified, rather than specific technology requirements, thus giving companies the flexibility to select the optimum method of achieving compliance. This approach is more cost effective and likely to stimulate technological advances in pollution control techniques and production processes. For large point sources that are concentrated and few in number, instruments such as emissions and offset trading may be a more effective means of managing pollutant emissions and reducing compliance costs than retro-fitting abatement equipment to existing infrastructure.

The dispersed nature of vehicular sources makes it difficult for monitoring and enforcement authorities to target each individual source, necessitating the use of a mix of technical and non- technical measures. Technological advancements progressively reduce emissions from the vehicle fleet through mandatory emissions limits and fuel efficiencies for new vehicles. International experience has shown that purely technology-based solutions are not adequate for addressing traffic emissions, with behavioural factors related to car ownership and use of personalised vehicles also having a major influence. Vehicle emission reduction strategies should therefore also encompass non-technical measures such as transport demand and supply management measures that reduce the incentive for using personal vehicles and promote the use of public transport.

g) Harmonisation of AQM Measures with Regional and Global Measures Alignment of local AQM measures with measures aimed at addressing regional and global issues is complicated by gaps in scientific knowledge of the linkage between such issues and by the lack of supporting mechanisms for environmental policy integration. Furthermore, methodologies for assessing the broader environmental burden of policies and interventions have been largely unavailable.

Significant progress has been made in developing methods for assessing environmental burden in a more holistic way within fuel- and life-cycle analyses (LCA). Such methods facilitate comparisons of the broader environmental risks associated with sources of atmospheric emissions, and the assessment of the environmental consequences of policies and interventions.

Environmental burden methodologies are, for example, being applied which comprise the calculation of indices for Global Warming Potential (GWP), Stratospheric Ozone Depletion, Photochemical Ozone Creation Potential (POCP) and Air Acidification (Bates et al., 2003). The

integration of such indices in the LCAs undertaken for air quality interventions reduce broader environmental risks and enhance the efficiency of such measures.

3.1.2 Environmental-Social Interface a) Addressing Air Pollution through Poverty Alleviation Poverty alleviation through the implementation of an efficient social protection system is central to maintaining conditions conducive to both economic growth and environmental sustainability. Coal, wood and paraffin are used by many South African households, including electrified households, due to the relative cost-effectiveness of such fuels for space heating and cooking purposes. Many low cost housing developments and informal settlements are established in proximity to industrial and mining operations due to the land being available and inexpensive. Poorer communities are more likely to receive poor service delivery from local councils, including intermittent waste removal services that in some instances result in waste being illegally burned on pavements or waste lots. From these examples it is clearly evident that poverty alleviation could result in air quality improvements by enabling people to choose more environmentally-friendly practices and to reside further from polluting sources.

b) Addressing Environmental Injustice Environmental hazards, including air pollution exposures, are known to disproportionately affect the health of people who already have compromised health and nutrition due to their place in society. Air quality management strategies may address such environmental injustices by, for example, prioritising sources impacting on marginalised communities. The integration of air quality considerations into land use planning also has significant benefits in this regard, e.g. ensuring that low cost residential developments are not co-located with potentially significant sources of emissions such as heavy industry and mining.

In the event of unavoidable deterioration of air quality or persistently poor air quality, it is necessary to minimize the burden of those affected by measures such as improving the medical system and compensating those impacted through the polluter-pays principal.

c) Considering the Social-acceptability of Interventions To be successful air quality interventions must not only be technically viable and economically feasible but also socially acceptable. Lessons were learned in this regard in South Africa in the 1980s and 1990s when the Department of Minerals and Energy (DME) investigated the introduction of low-smoke fuels as an alternative fuel for household coal and wood burning households. During this period social studies undertaken revealed a strong link between energy use and religious, cultural and social factors and values. Both social and economic studies indicated that a single intervention was unlikely to be successful in addressing household fuel burning but rather that the DME needed to identify the least-cost, most-desirable and most-effective energy mix options (Scorgie et al., 2001; 2002). This revised approach, in which socio-economic considerations were integrated into planning, was evident in the DME‟s Integrated Clean Household Energy Strategy adopted in 2003 (DME, 2003).

3.2 Advances in Air Quality Management in South Africa and Sector-specific Interventions The Air Quality Act shifts the focus from exclusively source-based control to a receiving environment approach. Key aspects of the Act include:  Air quality targets to drive emission reductions;  Air quality management decentralisation;  Targeting of all significant sources;  Recognition of regulatory and alternative measures e.g. market mechanisms, voluntary programmes and education;  Air quality management planning by authorities and emission reduction planning by sources;  Cost-optimisation of mitigation measures; and  Access to air quality information and public consultation.

AQA‟s success depends on regulations being set in the short-term and effectively reviewed and revised in the medium- to long-term. Furthermore, long-term resource allocation, inter- departmental and inter-governmental cooperation and support by business and civil society are required. Progress made by 2011 has included the following:  Appointment of national, provincial and local air quality officers and establishment of cooperative governance structures.  Publication of air quality governance guidelines by national government.  Initiation of projects to aid the transition from APPA to AQA, including the APPA Registration Certificate Review Project (Mahema et al., 2006; Baird and Scorgie, 2006), the National Emission Standard Setting and Listed Activity Project (Scorgie and Kornelius, 2007) and the Air Quality Management Planning Project (DEA, 2011a).  Establishment of the South African Air Quality Information System.  Declaration of Vaal Triangle and the Highveld as national priority areas. The Vaal Triangle Priority Area Air quality Management Plan has been developed. The Highveld Priority Area baseline assessment has been completed (Zunckel et al., 2010) and the draft AQMP published in April 2011 and scheduled completion by March 2012.  Publication of national ambient air quality standards and national emission standards for so- called Listed Activities (DEAT 2009; DEA, 2010a).  Improvements in air quality management courses offered by tertiary institutions.

Areas requiring further work include: cost-optimisation of air quality monitoring systems, effective integration of air quality issues into development planning, use of multi-pollutant control strategies, application of market mechanisms, and the mainstreaming of air quality management within local, provincial and national planning.

Various measures targeting emissions from priority sources are being implemented or proposed. A brief overview of progress made and implications of the legislative reform process is provided in subsequent subsections.

3.2.1 Household Fuel Burning Access to electricity and regulation of emissions from residential areas was historically perceived to be the solution to household fuel burning (van Horen, 1996). Despite extensive electrification

campaigns electricity only accounted for 38% of the total energy consumed by the residential sector during 2000, the remainder provided mainly by the combustion of wood (41%), coal (35%), paraffin (14%), vegetable wastes (7%) and LPG (3%) (ERI 2001). Persistence of coal and wood burning within the residential sector is due mainly to the cost effectiveness and multi-functional nature (supports cooking and heating) of these fuels. Furthermore, rapid urbanisation and the growth of informal settlements exacerbated backlogs in the distribution of electricity (Winkler, 2006).

Designation and regulation of „smokeless zones‟ under APPA restricted fuel burning in historically white residential areas (van Horen, 1996). These smoke control regulations were not enforced in residential areas designated for occupation by Black citizens because of the perceived economic hardship of compelling use of cleaner but more expensive energy carriers, such as electricity. Even after the 1994 political transformation, smoke control regulations have not been extended to these historically Black residential areas, despite extensive further electrification of homes.

It was increasingly recognised that an integrated approach, comprising a mixture of interventions, is needed to successfully address emissions and exposures within residential fuel burning areas. In 2003 the Department of Minerals and Energy (DME) adopted the Integrated Clean Household Energy Strategy (ICHES) (DME, 2003). This strategy proposed a range of measures including refining of combustion methods and appliances, replacement of coal with electricity, alternative fuels and renewable energy and measures aimed at reducing energy requirements of dwellings (e.g. through insulation and solar passive design). To date, implementation has been restricted to campaigns to advocate the use the top-down ignition method of lighting fires (known as „Basa njengo Magogo‟) to a few coal-burning towns and suburbs, and ongoing electrification.

The splitting of the mining and energy portfolios within the Department of Minerals and Energy, in 2009, resulted in the newly established Department of Energy (DoE) being made responsible for strategic planning within this sector. The Director General of the DoE reported the Department would draft an Integrated Energy Planning Strategy during the course of the 2010/2011 financial year, and that energy efficiency programmes would be initiated favouring demand side management interventions(4).

Low-cost and no-cost energy efficient housing measures published by the International Institute for Energy Conservation (IIEC) are being implemented in certain provinces on a project-by-project basis (Winkler, 2006). Other measures being investigated by certain cities include the roll-out of „smokeless‟ braziers (City of Johannesburg) and the integration of passive design considerations into housing policies (Scorgie et al., 2003b, 2003c).

Interventions currently under consideration to reduce residential fuel burning, or the emissions and effects of such burning, are summarised in Table 1 (DME, 2003; Winkler, 2006). Reference is made to the nature and main application of each measure.

4 http://www.energy.gov.za/files/aboutus/message%20from%20dg.pdf

Table 1: Synopsis of interventions to reduce emissions from residential fuel burning

Nature of Measure Main Application Measure Technological Fuel Energy Renewable Greenfields Brownfields Emission Switching Efficiency Energy Reduction Top-down ignition X X Stove maintenance X X Stove replacement X X Electrification X X X Alternative fuel: low-smoke X X fuel, gas, paraffin, methanol Solar passive design X X Insulation of existing homes X X Free basic energy X X Renewable energy: solar X X X (cookers; water heaters)

3.2.2 Electricity Generation Measures aimed at reducing the extent of coal-fired electricity generation may be classified into demand-side and supply-side management measures. Demand-side management (DSM) refers to changes in electricity consumer behaviours through specific energy saving programmes. Supply- side management includes measures aimed at evaluating and selecting the most suitable energy generation options. Supply-side management options being considered by Eskom, the largest electricity generator, include: implementation of improved pollution control devices for existing power stations, introduction of clean coal technologies, and power generation alternatives including renewable energy and nuclear.

Despite investigations into cleaner coal technologies and alternative fuels, it is expected that coal- fired pulverised fuel power stations will continue to supply the bulk of electricity for the foreseeable future. This is evident given recent proposals for the commissioning of three additional large-scale coal-fired power stations on the Highveld plateau or the northern Limpopo province. Measures to reduce coal-fired power generation emissions include generation efficiency improvements, demand side management, renewable energy policy and abatement technology advances.

Whereas primary particulate matter emissions have historically been very effectively controlled from major coal-fired power stations, gaseous emissions such as SO2 and CO2 are not controlled and are of concern in terms of local ground-level exposures (SO2) and global warming potential

(CO2). It is feasible that proposed power stations will be required to implement desulphurisation technologies given elevated background SO2 concentrations and the anticipated introduction of stricter regulations for industries under AQA. The feasibility of retrofitting existing power stations will depend on the cost-effectiveness of doing so, and based on preliminary considerations appears unlikely. However, under pressure from international finance organisations (e.g. World Bank), all future coal-fired generating plants are likely to be equipped with flue gas desulphurisation equipment.

3.2.3 Industry Industrial processes with significant air emissions have historically been regulated using Registration Certificates issued under APPA. Such certificates were issued for specific processes, with an industry operating at specific premises able to hold multiple certificates for its various processes. Emphasis was placed on stack emissions, with fugitive sources not having been comprehensively regulated.

Under the AQA, potentially polluting industrial activities will be regulated as „listed activities‟ requiring Atmospheric Emission Licenses to operate. Licenses, whilst still being media-specific, will cover the entire facility and all sources, including fugitive emissions.

The APPA Registration Certificate Review Project (2006-8) aimed to convert the certificates of ~70 large industries to licenses. Despite these industries comprising only ~5% of the enterprises regulated under APPA, they hold about 18% of the registration certificates and are estimated to emit nearly 80% of the country‟s criteria pollutants (Scorgie, 2006). This project, representing a collaborative effort between national, provincial and local authorities, set the precedent for future conversions of emission licences to the new licences by local and provincial governments.

Proactive emission quantification and emission reduction planning is implemented by certain industries. Motivated by social responsibility, concern for the environment, fear of litigation or of future regulation (or a combination of such factors), an increasing number of industries are seeking opportunities to clean up their operations. In addition to its regulatory approach, the Government has initiated a national driver to promote cleaner production in cooperation with business.

3.2.4 Road Transport Fuel and technology interventions to reduce vehicle emissions, which are typically introduced on a national basis, have achieved significant success internationally. Traffic management measures requiring the integration of air quality considerations into transport planning, frequently at local level, have had lower rates of success.

In South Africa, significant changes in liquid fuel specifications have included reductions in the sulphur content of fuels (diesel) and phasing out of leaded petrol. Advances in vehicle technology have largely occurred due to market forces (South Africa‟s re-integration into the global vehicle manufacturing networks) and the need for engines to be compatible with international upgraded fuel specifications. Since the introduction of unleaded petrol, catalytic converter equipped petrol vehicle sales have increased steadily, comprising almost 50% of new passenger vehicle sales in April 2003. This percentage is expected to be significantly higher currently given the phasing out of leaded fuel, which was completed at the beginning of 2006. Newer vehicles are typically at least Euro 3 technology compliant.

The Implementation Strategy for the Control of Exhaust Emissions from Road-going Vehicles in South Africa was proposed by the Department of Environmental Affairs (DEA) and DME in 2003. This strategy stipulated Euro technologies for new petrol- and diesel-driven vehicles and future reductions in sulphur, benzene and aromatics content of fuels. Although certain fuel changes and new vehicle standards have supported implementation of Euro 2 standards, the strategy has not been implemented in its entirety with the DEA intending to achieve technology changes through

regulation of vehicles as „controlled emitters‟ under the AQA. As at January 2011, no classes of vehicles had yet been declared as controlled emitters under the Act. Fuel and vehicle specifications and standards currently only support the meeting of Euro 2 standards by new vehicles.

3.3 Effective Implementation of Sector-specific Interventions To be successful, sector-specific measures need to be technically-feasible, economically viable, socially acceptable and potentially also politically desirable. Furthermore, such measures need to be integrated into development planning and everyday business, supported by cooperative governance and buy-in from business and civil society.

Successful implementation of crucial sector-specific interventions has historically almost certainly been undermined by the absence of an enabling legislative environment. Metropolitan air quality management plans developed since 2002, pre-empting the promulgation of AQA, have focused primarily on:  monitoring systems design and implementation;  source prioritisation; and  mitigation of high priority sources which local authorities were mandated and able to regulate.

Due to industry being regulated nationally under APPA, local authorities were not mandated to regulate industrial emissions occurring within their jurisdictions. The absence of mechanisms for integrating air quality considerations into transportation and land use planning impeded the management of transportation emissions. Further discussion on these challenges is provided in Section 10.

The evolution of air quality management practices and promulgation of AQA provides the common vision and legislative context necessary for developing integrated intervention strategies. AQMPs to be developed by local authorities are required under AQA to be documented in Integrated Development Plans. This promotes alignment of air quality management and development (energy, transportation, land use) planning. Metropolitan and district municipalities are also given responsibility for the licensing of „listed activities‟ by AQA and are therefore better poised to address air quality effects associated with industrial emissions (within the framework of national emission standards).

Drawing on the provisions of the NEMA (Act 107 of 1998), and the rights ensured under the Promotion of Access to Information Act (Act No. 2 of 2000), the AQA provides for stakeholder consultation and public access to air quality information as integral components of the air quality governance process. Stakeholder consultation is specifically provided for within the AQMP development process. If effectively implemented, the involvement of business and society will support the identification and successful roll-out of suitable, sustainable intervention strategies.

3.4 Summary The delivery of cleaner air, whilst not impacting negatively on society and the economy, presents a noteworthy challenge. Based on the experiences gained by other countries, the following strategies have been identified as holding specific relevance for South Africa:

 Systematic integration of air quality considerations into energy, transportation, land use, housing and other development planning processes.  Alignment of local air quality management planning with measures aimed at addressing regional and global air pollution-related issues.  Integration of economic concerns in air quality management and planning, including: cost- benefit analysis of interventions, and cost optimisation of air quality monitoring.  Air pollution effect reduction through poverty alleviation.  Addressing environmental injustice by prioritising sources impacting on marginalised communities.  Considering the social-acceptability of interventions, in addition to their technological and economic viability.

Since the promulgation of the AQA in 2004 significant progress has been made in establishing the framework required for effective air quality governance. A number of sector-specific interventions are under investigation or proposed to address major sources such as household fuel burning, road transport and power generation. Experience has shown that sector-specific measures need to be technically-feasible, economically viable, socially acceptable and potentially also politically desirable if they are to be successful. Health effects and costs due to major sources have been quantified and a cost-benefit analysis of potentially viable sector-specific interventions undertaken to identify cost-effective measures, as presented in Chapters 4 to 7.

4 Calculation of Externalities Due to Fuel-burning Emissions

The quantification of health risks due to significant anthropogenic sources within several South African conurbations, and establishment of cost-optimised air pollution interventions, forms a key component of this thesis. This externalities study is documented in chapters 4, 5, 6 and 7. This chapter documents the methodological approach developed for the quantification of externalities. This systematic, damage function approach comprises the estimation of emissions, and the linking of such emissions to changes in air quality and associated health effects and costs.

4.1 Introduction Anthropogenic fuel burning sources are associated with widespread air pollution exposures within South African urban areas. The lack of quantitative data on health effects and associated costs due to such sources, and the benefits achievable by interventions addressing such sources, represents a significant hurdle for cost-effective air quality management.

To address this knowledge gap the „Dirty Fuels Study‟ was commissioned by the National Economic Development and Labour Council (NEDLAC) under the Fund for Research into Industrial Growth and Equity (FRIDGE). The author was the principle investigator of this study, during her tenure as a doctoral candidate. The study, formally entitled the „Study to Examine the Potential Socio-Economic Impact of Measures to Reduce Air Pollution from Combustion‟, was completed in 2004 and has not previously been published in the open literature (Scorgie et al., 2004a, 2004b, 2004c, 2004d). Components of the NEDLAC Dirty Fuels Study of specific relevance to this thesis are as follows:

 establishment of a damage-function approach for the costing of externalities;  quantification of significant anthropogenic fuel burning sources;  source significance ranking and intervention selection based on human health considerations; and  cost-benefit analysis of interventions proposed for implementation for significant sources.

The objectives of the NEDLAC Dirty Fuels Study and the methodology developed and implemented in the calculation of externalities related to fuel burning activities is described in this chapter. This thesis draws on a sub-set of the data analysed during the Dirty Fuels Study, focusing specifically on the priority air pollutants, sources and health effects identified for South African cities, as discussed in Section 2. Study findings are documented in subsequent chapters, including emissions estimated (Chapter 5), direct health effects and costs calculated (Chapter 6) and recommendations regarding priority sources and cost-effective interventions (Chapter 7).

4.2 Study Objective The NEDLAC Dirty Fuels Study constituted the first attempt to comprehensively quantify health effects and costs related to fuel-burning emissions across power generation, industrial, residential,

transport and agricultural sectors within the major conurbations of South Africa(5) (Scorgie et al., 2004a, 2004b, 2004c, 2004d; University of Cape Town, 2004; Bentley West Management Consultants & Airshed Planning Professionals, 2004). As the main aim of the NEDLAC study was to quantify the benefits and costs associated with fuel use interventions, it was first necessary to provide an estimate of the health effects and associated costs related to fuel use practices.

The specific objectives of this study were as follows:  Identification and quantification of key anthropogenic fuel combustion sources including projected changes in emissions in the short- to medium-term (i.e. five to ten years) given the assumption of „business as usual‟;  Simulation of ambient air pollutant concentrations occurring due to current and future „business as usual‟ emissions;  Quantification of potential exposures and prediction of health risks to modelled ambient air pollutant concentrations;  Monetary costing of direct health risks;  Prioritisation of sources based on their associated current and projected future health risks and costs;  Identification of potentially feasible interventions to reduce effects for significant sources;  Quantification of reductions in emissions and resultant reductions in air pollutant concentrations, exposures and health risks due to the implementation of selected interventions;  Economic assessment comprising a cost-benefit analysis of interventions, taking into account costs related to intervention implementation and resultant health cost savings(6); and  Provision of recommendations regarding sources to be prioritised for early abatement and the most cost-effective interventions for such sources.

The author of this dissertation was the principle investigator on all components of the NEDLAC Dirty Fuels Study during her tenure as a doctoral candidate, with the exception of the economic assessment. Information developed provided the necessary input for the economic assessment undertaken by the University of Cape Town (2004) and recently published by Leiman et al. (2007). The economic assessment will not be documented in this thesis. Reference will however be made to certain of the findings from the economic assessment where such findings are integral to the final recommendations and conclusions drawn on the basis of the integrated study.

4.3 Overview of Externality Studies The external costs and benefits of fuel cycles are determined by identification, quantification and monetisation of the associated effects. Two methodologies are typically applied to determine the externalities associated with fuel cycles, viz. top-down and bottom-up approaches. Earlier externalities studies more commonly used the top-down approach where generic damage costs were estimated at a national level for different impact categories (e.g. damage to health), with such

5 Some work had been done on the identification and quantification of externalities associated with the existing energy generation and household energy consumption sectors (van Horen, 1996). 6 This component of the study was undertaken by the University of Cape Town (2004), with a synopsis of the study published by Leiman et al. (2007).

costs then being attributed to various emissions (e.g. SO2) based on the calculated average external cost per unit of emission (Pearce et al., 1992; Friedrich & Voss, 1993; Ott, 1994). Although the top-down approach provides a first indication of the environmental externalities of energy it does not allow for the assessment of the marginal effects of additional energy supply which are of interest for planning purposes.

The bottom-up approach is also known as the impact-pathway approach or damage-function approach (DFA). This approach allows for the calculation of marginal external costs and can be applied to various types of effects for which an impact-pathway can be characterised. The application of the DFA in the case of air pollution exposure studies begins with determining the quantity of emissions from defined sources and subsequent use of dispersion models and exposure- response functions to determine the marginal damages resulting from the emissions. The final step comprises the multiplication of the marginal damages by their estimated monetary value. DFA studies are site specific, require large quantities of data and are time intensive. More recent externality studies have tended to use this approach (OECD, 1994; CEC, 1995; Ostro, 1996; van Horen, 1996; Seethaler, 1999; Segal, 1999; Kunzli et al., 2000; Nelson, 2000; Kim & Qiang, 2002; Holland and Watkiss, 2002; Fisher et al., 2002; World Bank, 2002; Amoako & Lodh, 2003; DEFRA, 2004; Borysiewicz et al., 2006; Wang & Mauzerall, 2006).

The most exhaustive study to date on the external costs of energy is the ExternE Project which began as a collaborative effort between the EC and the US in 1991. The ExternE methodology uses a damage function approach to determine the environmental external costs of fuel cycles (CEC, 1995). The main aim of the ExternE project was to quantify effects and externalities of air emissions from conventional thermal power plants, as these are likely to cause the most significant effects in the case of conventional fossil fuel cycles. To determine the damages of atmospheric pollution, the dispersion and transformation of pollutants was modelled based on a short-range and long-range atmospheric dispersion model. The local atmospheric dispersion model calculated the pollution increments for one hundred 10 x 10 km grid cells around the emission source. The regional atmospheric dispersion model calculated the pollution increments for 100 x 100 km grid cells across Europe. The pollution increments were translated into effects via exposure-response functions (ERFs) which were selected on the basis of an extensive literature survey and primarily based on recent epidemiological studies carried out across Europe (Holland & Watkiss, 2002). The ExternE Project methodology was most recently updated in 2004 (Bickel & Friedrich, 2004).

Until 2000, few studies on the externalities of energy had been carried out outside Europe and the US. Exceptions were studies conducted in Argentina (Carnevali & Suarez, 1993), South Africa (Van Horen, 1996) and Brazil (Furtado, 1996). Carnevali and Suarez (1993) assessed implications of Argentinean energy policies on air pollution emissions and emissions control costs. Van Horen (1996) quantified full fuel cycle externalities related to coal and nuclear energy for South Africa, with a focus on large scale power generation. The valuation study conducted by Furtado (1996) was aimed at assessing the „willingness to pay‟ to avoid environmental effects from hydro, coal and nuclear power in Brazil.

In more recent years an increasing number of energy externalities studies have been conducted throughout the world including studies in Asia, South America and Australasia (Segal, 1999; Kim & Qiang, 2002; Fisher et al., 2002; World Bank, 2002; Amoako & Lodh, 2003; Borysiewicz et al., 2006; Wang & Mauzerall, 2006). Although the overall approach has remained the same,

improvements have been made in emission estimation, environmental quality quantification, dose- response assessment and financial costing of effects.

Results from previous studies indicate that externalities calculated using the damage function approach are generally lower than those calculated using top-down approaches. This is partially due to the restricted number of pollutants considered, limited consideration of synergistic effects between pollutants and the adoption of linear dose-response functions in such studies (Fouquet et al., 2001). There are also considerable differences between the values obtained by the various damage function approach studies. This is mainly due to the variety of methodological approaches used, the differences in effects considered, the specific damages attributed to emissions and whether or not climate change is considered.

4.4 Methodological Overview In assessing health risks and costs due to fuel use and potential health benefits (and related savings) which may be achieved by interventions in South Africa, a damage function approach was adopted. As discussed in the previous section, this systematic approach links emissions and resource effects related to an activity to changes in environmental quality, in this case air quality (Figure 16). Such changes are in turn associated with environmental, social and health effects.

Figure 16: The damage function approach applicable to emissions and effects related to the energy generation sector (after van Horen, 1996)

Based on the damage function approach, and the review of similar studies undertaken internationally, including the large scale ExternE Project, the following procedure was developed for the fuel combustion externalities study:  Emission inventory development, with source and emissions data collated and calculated for the identified source groups in each conurbation.  Dispersion modelling of current and projected future emissions from each source group to determine resultant ambient air pollutant concentrations for all pollutants included in the study and comparison of model results with measured air quality for model verification.  Health risk quantification based on predicted air pollutant concentrations being superimposed over spatial population data and selected dose-response functions applied to estimate mortalities and morbidities per source group.  Calculation of direct health costs per source group based on the estimated morbidities and unit monetary costs for treatment obtained from the health service sector(7).  Identification of potentially feasible interventions for significant source groups based on the review of local and international practices and discussions with government departments tasked with the regulation of such sources.  Estimation of atmospheric emission reductions achievable through the implementation of selected interventions and subsequent modelling of ambient air quality improvements and calculation of health risk and cost reductions.

4.5 Scope and Limitations of Study

4.5.1 Externalities Quantified Externalities associated with fuel usage may comprehensively be described in terms of the life cycle of each fuel. The coal fuel cycle, for example, includes: occupational health issues and air and water pollution related to coal mining, water consumption and water quality effects associated with the power generation sector, and air pollution due to power generation and its associated effects on human health, acid deposition and visibility.

Given that the focus of the investigation is primarily to support urban air quality management and planning, the scope of study was restricted to a detailed quantification of the main effects associated with fuel combustion, viz. inhalation exposures of sensitive receptors to ambient air pollutant concentrations arising from fuel burning emissions. Due to the urbanised nature of the study areas selected and the proximity of many communities to sources of fuel combustion, health risks associated with combustion exposures was identified as the most important effect to consider. In support of urban air quality management strategies aimed at reducing public health risks, the study focused on exposures of the general public rather than on occupational risks.

Effects on vegetation, including crop productivity, and material damage are expected to be substantially lower in urban areas and were therefore excluded from the impact quantification

7 Costs due to estimated mortalities and indirect costs related to productivity losses resulting from absenteeism projections were also quantified as part of the economic analysis undertaken by the University of Cape Town (2004), with the findings of this study being considered in the prioritisation of sources and recommendation of interventions.

study. Furthermore, dose-response relationships for damage to materials and effects on commercial crops, fauna and flora are relatively sparse and in many cases not available for local environments.

4.5.2 Source Types Included Considering that urban air quality management is primarily geared towards the control of anthropogenic sources, and that fuel burning activities are estimated to be responsible for over 80% of ambient criteria pollutant concentrations, the study focused on anthropogenic fuel burning activities. Emissions were estimated and effects and associated costs quantified for the following major source groupings:  Household fuel burning – including coal, wood, LPG and paraffin burning;  Power generation – primarily coal-fired power generation for the national grid;  Industrial and commercial fuel burning – including heavy fuel oil (HFO), coal, wood and gas combustion; ranging from large industrial activities to small wood-fired pizza ovens; and  Vehicle tailpipe (and evaporative) emissions – including petrol- and diesel-driven vehicles.  Biomass burning (agricultural and wild fires).

Although emissions were estimated for the biomass burning source group, effects and associated costs could not be quantified due to the unavailability of information required to accurately determine temporal variations in emissions for dispersion modelling purposes. Emission estimates for this source group are however documented in this thesis to assess the relative significance of this source, and reflect on the need for future advances in the modelling of biomass burning releases.

4.5.3 Spatial Extent of Study Impacts associated with fuel burning emissions occur at several spatial scales, ranging from household- and conurbation-level (e.g. individual and community health risks) to regional and global scale effects (e.g. transboundary pollution transportation, acid deposition and climate change). To address the study objectives emphasis was placed on effects at the level of conurbations in support of urban air quality management planning, with community health risks due to inhalation exposures representing the main focus of the study. Emission inventories were established and effects modelled for the following conurbations:

 Tshwane Metropolitan Municipality (includes Pretoria)  City of Johannesburg and Ekurhuleni Metropolitan Municipality  Vaal Triangle – comprising the towns of Sasolburg, Vereeniging, Meyerton and Vanderbijlpark  Mpumalanga Highveld region  eThekwini Metropolitan Municipality (includes Durban)  City of Cape Town The population residing within the regions included in the study comprised approximately 40% of the national population of South Africa.

4.5.4 Pollutants Considered Air pollutants were selected for inclusion in the study based on their having been identified as priority pollutants resulting in widespread exposures within South African urban areas, and on the availability of dose-response functions enabling the quantification of health effects. The following air pollutants were selected: particulate matter with an aerodynamic diameter of <10 µm, PM10), sulphur dioxide (SO2), nitrogen dioxide (NO2), benzene and 1,3-butadiene. Whereas fuel burning is associated with a range of trace organic compounds, benzene and 1,3-butadiene were selected for inclusion given that these compounds have received increased attention abroad in terms of air quality standard setting and urban air quality monitoring. Both primary and secondary particulate matter (sulphates and nitrates) were included in the study.

Various damage functions studies tend to include the quantification of effects associated with a very limited number of pollutants. In many cases, only fine particulate matter are used with the health effect of NOx and SO2 assumed to arise indirectly due to the conversion of these gases to nitrate and sulphate aerosols. This has been done primarily due to the complexities of separating out the effects of SO2, NO2 and PM10; a simplifying assumption is made that as they tend to vary together in most locations and studies, they need not be treated as individual components.

At the time of undertaking the externalities study, the position of the widely referenced ExternE Project is to use dose-response functions exclusively for particulate matter and ozone, whilst omitting the inclusion of NO2 and SO2 in health risk assessments (Bickel & Friedrich, 2004). A more recent ExternE methodology update however states that the situation is not clear and opinions could change as further evidence comes to light. Bickel and Friedrich (2004, p. 83) noted that “…in particular, the Hong Kong intervention study showed a sustained benefit in mortality reductions

Exposure-Response Functions following reductions in pollution involving mostly SO2. There could indeed be significant direct effects of SO2, contrary to the current position of ExternE”.

Particulate emissions from most of the coal-fired power stations in South Africa are very effectively controlled, with control efficiencies generally in excess of 99%. Desulphurisation and denitrification technologies are not installed in any of the existing fleet of coal-fired power stations, hence these plants constitute significant, elevated sources of SO2 and NOx emissions. It was therefore considered imperative to include the separate quantification of health risks due to SO2 and

NO2 in the study, in addition to PM10. Benzene and 1,3-butadiene were included in the study to facilitate the quantification of cancer risks arising due to exposures to fuel combustion emissions.

An important limitation of the study was that ozone, a secondary pollutant formed in the atmosphere through the conversion of NOx and VOC precursors, was not included. Its inclusion would have required complex photochemical modelling, which could not be undertaken given the spatial scale of the study and numbers of case studies considered. As a result of this limitation it is expected that health risks and costs calculated for fuel combustion sources, e.g. vehicles, would be underestimated.

4.5.5 Health Endpoints The acquisition of dose-response relationships and damage costs applicable within South Africa represents the key challenge in the application of the damage function approach. Dose-response relationships are not available locally. Reference therefore needed to be made to dose-response

relationships available in the general literature. Such relationships are published for various pollutants and human health endpoints (e.g. hospital admissions, bronchodilator use, mortality).

In assessing health risks and costs associated with exposures to fuel-burning emissions, effects were calculated for the following selected health end points:

 Respiratory hospital admissions (due to PM10, SO2 and NO2 exposures)

 Cardiovascular hospital admissions (due to PM10 exposures)

 Premature mortality (due to PM10 and SO2 exposures)

 Chronic bronchitis (due to PM10 exposures)

 Restricted activity days (RAD, due to PM10 exposures)

 Minor restricted activity days (MRAD, due to SO2 exposures)  Leukemia cases (due to 1,3-butadiene and benzene exposures)

4.5.6 Base Case and Future Projections Emissions inventories were compiled and air pollutant concentration simulations and health risk calculations undertaken for the various conurbations for the base year 2002. To facilitate the projections of changes in emissions, air pollutant concentrations and associated health risks in the short to medium term, given an absence of intervention, emissions were projected for significant sources for the years 2007 and 2011.

4.6 Emission Quantification Various approaches can be used to establish a source inventory ranging from the gross estimation method to the compilation of a detailed emissions inventory. Taking into account the objectives and scope of the study it was decided to use the rapid survey method. The rapid survey method comprises both the use of reference documents and summary data for area sources (domestic fuel burning, agricultural burning) in addition to infield data collection for major sources via questionnaire and telephonic contact.

The quantification of emissions from combustion-related activities within the agricultural, transport and domestic sectors was based on reference documents, published summary data and information obtained from central sources. Emission estimates were based on source-specific monitoring data only in instances where such data were available governmental departments and major industries. Where no such data were available, and in the estimation of emissions from non-industry sources, use was made of published emission factors.

Preference was given to locally-developed emission factors where available (e.g. for domestic coal burning and vehicle exhaust emissions) (Britton, 1998; Graham, 1997; Wong & Dutkiewicz, 1998; Wong, 1999). Where no local emission factors exist reference was made to widely used foreign data sets (e.g. United States Environmental Protection Agency's AP-42 single-valued emission factors and predictive emission factor equations; European Environment Agency emission factor data bases such as Coppert III - road transport emission factors). The US emission factors for coal combustion make provision for the input of local coal quality parameters, such as ash content and sulphur content, and are therefore applicable for local applications.

4.6.1 Public Electricity Generation There are four main types of electricity generators in South Africa, viz. the national public electricity utility (Eskom), municipal generators, Independent Power Providers (IPPs), and auto generators. The auto generators are industries which generate electricity for their own use. These include the pulp mills, sugar refineries, petrochemical refineries and metallurgical industries. Only power stations generating electricity for the national grid were included in the public electricity generation source group, with power generation by auto generators covered under the industry grouping.

The South African power generation sector is heavily dependent on coal. Based on 2001 figures in which Eskom's electricity sales totalled 187 957 GWh, coal was responsible for ~92% of Eskom's power generation (Eskom Annual Report, 2002). Coal is also used by municipalities and independent power producers for the generation of electricity within Johannesburg, Cape Town, Tshwane and Bloemfontein. Coal and nuclear power stations generally provide the bulk of the base electricity load, while the pumped storage schemes and gas turbines are used primarily to meet electricity demand at the peak and in cases of emergency. A list of the coal and gas fired electricity generation facilities located within the conurbations of interest are given in Table 2.

Table 2: Information for electricity generation facilities included in the study

Effective Average Height of Facility Fuel Fuel Control Conurbation Capacity Production Emission Name Used Quantity Technology (MW)(2000) (GWh) (m) AES Kelvin A 90 New Bag ND Coal 81.3 Ktpa 73; 74 Johannesburg AES Kelvin B 260 filters (2003) Johannesburg 102 ND Gas ND ND ND Athlone(a) 29 ND Gas ND ND ND Cape Town Athlone(a) 90 184.438 Coal 119 Ktpa 100 Fabric filter Acacia(b) 7 0.15 Gas ND 14 ND Vaal Lethabo 3 558 21 572 Coal 13.6 Mtpa 275 ESPs + FGC Triangle Pretoria West 56 ND Coal 440 Ktpa 53 ND Tshwane Rooiwal 120 ND Coal 454 Ktpa 101 ND Arnot 1 858 9675 Coal 6.1 Mtpa 195 ESPs, PJFF ESPs + FGC, Duvha 3 483 22 798 Coal 14.3 Mtpa 300 PJFF ESPs + FGC, Hendrina 1 884 24 691 Coal 7.4 Mtpa 155 PJFF Mpumalanga Highveld Kendal 4 063 17 452 Coal 15.5 Mtpa 275 ESPs + FGC Kriel 2 402 21 572 Coal 11.0 Mtpa 210 ESPs + FGC Majuba 2 465 5 170 Coal 3.2 Mtpa 250 PJFF Matla 3 518 25 199 Coal 15.8 Mtpa 213; 275 ESPs + FGC Tutuka 2 240 8 962 Coal 5.6 Mtpa 275 ESPs Information sources: Cape Town City Emission Inventory Data Base; Eskom (2000, 2002); www.eskom.co.za; National Emissions Inventory Data Base (1994); Cape Metropolitan State of Environment Report (1998); National Energy Regulator. Abbreviations: ND - no data; ESP - electro static precipitator; FGC - flue gas conditioning; PJFF - pulse jet fabric filters (a) Supplements electricity generated by Koeberg (nuclear) power generation facility during peak demand period in winter months. (b) Acts as the safety back-up for Koeberg. Should there be a break in grid power for some reason, Acacia's three 60 MW gas turbine- powered generators start up automatically. (c) Average production taken as an average over the last 3 years.

The average percentage ash and energy content of coals used by Eskom varies widely, ranging from 21% to 39% ash (average of 23.4%) and 15.2 MJ/kg to 22.5 MJ/kg energy content (average of 19.5 MJ/kg). South African coals have relatively low sulphur content compared to coals elsewhere. The sulphur content used by Eskom power stations is in the range of 0.59 to 1.41 (average of 0.9%) (Eskom, 2000).

Due to the high ash content in the coal combusted, particulate matter emissions and ash production are higher than would be the case with low-ash coals. As a consequence, Eskom's pollution control policy has been concerned primarily with the control of particulate matter emissions. Most of its coal-power stations currently in operation use electro-static precipitators (ESPs) to remove the bulk of particulate matter emissions from flue gases. Pulse Jet Fabric Filters (PJFF) are in use at Majuba Power Station. This technology has also been retrofitted to several of the older power stations. An additional control technology in use is Flue Gas Conditioning (FGC) which comprises the injection of small quantities of sulphur trioxide into the gas entering the ESPs. This alters the conductivity properties of the ash, making it easier to remove in the ESP. Due to the various control measures being implemented particulate matter emissions have been controlled by in excess of 99% (i.e. >99% control efficiency).

Information on the control technologies in use at IPP and municipal power stations is not readily available. Fabric filters are in use at the municipal Athlone Power Station in Cape Town with the control efficiency being given as 99%. Kelvin Power Station, located east of Johannesburg and operated by an IPP, installed bag filters with a design control efficiency of 99.99% for the two chimney stacks servicing the 11 boilers in the plant's A-station, reducing particulate matter concentrations from 400 mg/Nm³ to less than 50 mg/Nm³ (J Delaurentis, Engineering News, 2003/03/28). The Pretoria West power station currently uses centrifugal grit arrestors to reduce particulate matter emissions. Bag filters are in use at the Rooiwal power station in Tshwane.

Desulphurisation (DESOX) and denitrification (DENOX) technologies were not historically perceived to be warranted for use in power stations. Sulphur dioxide and nitrogen oxides are therefore not routinely removed from power station flue gas. Available technologies for DENOX and DESOX include absorption (scrubbing, dry injection), adsorbtion (water wash regeneration, thermal regeneration), catalytic conversion, irradiation and microbial conversion. The potential use of such methods has been investigated by Eskom for potential retrofit of certain existing power stations and potential implementation of future power stations.

In the estimation of emissions for the coal-fired power stations reference was made to emission factors provided by Eskom (Eskom, 2000; 2002) and US Environmental Protection Agency AP42 Emission Factors given for external combustion of bituminous coal (US-EPA, 1998). These emission factors are summarised in Table 3. Insufficient information was available for the gas turbine power plants for the estimation of emissions. Such stations are used exclusively for peak demand topping or stand-by power. Quantities of diesel being combusted at the coal-fired power during start up phases were also unavailable for use in this study.

Table 3: Emission factors for coal-fired power stations

Pollutant Eskom Emission Factors (Eskom, US-EPA Emission Factors for 2000; 2002) Uncontrolled Emissions - Range Given(a) (g KWh-1) (kg ton-1 ) (kg ton-1 )

SO2 7.95 16.31 14S - 17S

NOx 3.56 7.30 2.3 - 14.1

PM10 NA NA 2.8 - 28.1 Benzene NA NA 0.000454 Abbreviations: g/KWh - grams of pollutant per kilowatt-hours of electricity generated; kg/ton - kilograms of pollutant per ton of coal burned; NA - emission factors are not available despite these pollutants being emitted from coal-fired power stations; A - ash content of coal as a percentage; S - sulphur content of coal as a percentage. Notes: (a) The US-EPA gives a range of emission factors to account for variations in the firing configuration (US-EPA, 1998).

Eskom and US-EPA emission factors were found to be comparable for all pollutants for which the Eskom emission factors were given with the exception of particulate matter. Eskom emission factors are given for controlled emissions, whereas the US-EPA gives emission factors for uncontrolled emissions, despite providing information on the control efficiencies of various technologies which can subsequently be taken into account. Eskom emission factors, where available, were applied in the estimation of emissions from Eskom power stations. US-EPA emission factors for all pollutants except for PM10 were applied to estimate emissions from municipal and IPP power stations, in addition to being used for Eskom power stations for pollutants for which Eskom emission factors were not available. In order to provide a conservative estimate of emissions, the upper range factors were applied.

In the estimation of emissions an average sulphur content of 1% and an ash content of 23.4% were assumed for all power stations. Control efficiencies of 99% were used in the estimation of particulate matter emissions for the Athlone and Kelvin Power Stations. Although the control measures in place at the power stations in Tshwane were known, the control efficiencies were not known. Control efficiencies of 95% were therefore assumed for both power stations.

PM10 is given by the US-EPA as generally comprising >90% of the total particulate matter emitted during bituminous coal combustion when effective dust controls (e.g. multiple cyclones, bag filter,

ESPs) are in use. PM10 emissions were therefore estimated as comprising 90% of the controlled total particulate matter emissions calculated for each power station.

Total emissions calculated per conurbation are presented in Table 3. Although emissions from coal-fired electricity generation are greatest in the Mpumalanga Highveld and Vaal Triangle regions, it is notable that emissions are released at higher elevations than are emissions within Johannesburg, Cape Town and Tshwane. This is largely due to Eskom's tall stack policy implemented to reduce the effect of power station emissions, specifically SO2 on near ground air quality. The heights of currently operating Eskom power stations range from ~150 m to ~300 m, with newer power stations having taller stacks due to the implementation of this policy. Tall stacks were intended to emit above the boundary layer inversion so as to make use of the dispersion potential of the upper level jet (zone of high winds located above the boundary layer inversion) and thereby reduce the potential for plumes to fumigate down to ground. The highest ground-level pollutant concentrations associated with power station emissions occur ~1 km to 3 km from the base of the stacks during periods of enhanced atmospheric convection (typically midday during

summer). The exact location of the ground level maximum is a function of the actual stack height, gas exit velocity, gas exit temperature, atmospheric stability and ambient temperature at stack height.

Table 4: Estimated total annual emissions from public electricity generation

Johannes- Mpumalanga Cape Town Vaal Triangle Tshwane TOTAL burg Highveld Coal used (Ktpa) 81 119 13 552 894 78 966 93 611

PM10 (tpa) 77 113 8 150 4 236 47 475 60 051

SO2 (tpa) 1 382 2 023 219 868 15 198 1 280 816 1 519 288

NOX (tpa) 1 146 1 678 98 457 12 605 573 548 687 434 Benzene (tpa) 0.0 0.1 6.2 0.4 36 42 Release height (m) ~70 100 275 50 - 90 150 - 300

Small temporal variations are apparent in most coal-fired power station loads and hence emissions remain relatively constant throughout the year, except during maintenance periods. The reason for this is that peak electricity demands are catered for by pumped storage schemes and gas turbine stations. The Athlone Power Station in Cape Town represents the exception. This power station is only used to cater for peak demands which cannot be met by the nuclear power station at Koeberg. Such peaks typically occur during winter months as a result of increased power demand for space heating.

4.6.2 Industrial, Commercial & Institutional Fuel Burning Data availability on the type and quantity of fuel used by industries, businesses and institutions varied considerably between conurbations. Given the detailed fuel use data available for the Cape Town and eThekwini metros (Table 4) it was possible to calculate emissions based on the comprehensive set of emission factors published by the US-EPA in its AP42 data base. Emission factors selected for application in the current study are given in Table 5.

Table 5: Annual fuel use by industries and businesses within eThekwini and Cape Town

Quantity of Fuel use per Annum by Industries, Fuel Type Businesses and Institutions eThekwini Cape Town Coal (tpa) 580 314 310 045 Anthracite (tpa) 0 103 Coke (tpa) 0 54 HFO (kLpa) 32 873 311 585 Petrol (kLpa) 0 0 Diesel (kLpa) 15 172 11 449 Paraffin (kLpa) 9 603 71 200 Wood (bags per annum) 0 183 816 Woodwaste (tpa) 8 250 104 991 LPG (kLpa) 25 089 0 Waste (tpa) 0 288 Diesel/Paraf (kLpa) 0 2 294 Gas (not specified) 494 147 kL annum-1 34 686 103 m3 annum-1

Quantity of Fuel use per Annum by Industries, Fuel Type Businesses and Institutions eThekwini Cape Town Refinery gas (tpa) 308 517 0 Light Fuel Oil (tpa) 2 400 0

Sources of information: 1997 Emissions Inventory for Durban Metropolitan (Ecoserv, 1998); 2000 Emissions Inventory for Durban South (Ecoserv, 2000); Updated fuel figures for refineries in eThekwini (personal communication, Peter Butland, 8 May 2003); Cape Town City Emissions Inventory (as received from Ed Filby, City of Cape Town, Air Pollution Control Section, March 2003).

Table 6: General emission factors selected for estimating emissions from industrial and commercial fuel burning. Emission factors were derived from the US-US-EPA's AP-42 database. (Emission factors are given in kg of pollutant emitted as a result of unit of fuel burned.)

FUEL UNITS PM10 SO2 NOX Benzene Coal kg ton-1 6.6 19S 5.5 0.00065 Anthracite kg ton-1 0.04A 19.5S 4.5 Coke kg ton-1 6.6 25.5 9 0.00065 HFO kg kL-1 3.3712 18.84S 6.6 0.0000257 Diesel kg kL-1 0.96 8 8.49 Paraffin kg kL-1 0.24 1.22 2.4 Wood kg ton-1 2.6316 0.18275 3.5819 0.030702 Wood Waste kg ton-1 2.6316 0.18275 3.5819 0.030702 LPG kg kL-1 0.072 0.0108S 2.52 Waste kg ton-1 6.3 1.73 1.83 Diesel/paraffin kg kL-1 0.96 8 8.49 Natural gas kg 103 m-3 0.0304 0.0096 1.6 0.0000336 Abbreviations: TOC - total organic compounds; NMTOC - non-methane total organic compounds A - represents the ash content of the fuel (as %). The ash content of anthracite was taken to be 15% where not specified. S - sulphur content of the fuel (as %). The sulphur content of coal and anthracite was taken to be ~1% and the sulphur content of HFO ~3.2% where not specified.

Due to the absence of primary fuel use data for the majority of sources in the remaining conurbations, it was necessary to make reference to emission estimates undertaken previously for such sources. Sources of such estimates include the following: emission figures published by sources (e.g. Sasol Annual Environmental Reports), emission estimates made by the Chief Air Pollution Control Officer as given in the 1994 National Emissions Inventory Data Base, estimates by van Nierop (1995) for sources within the Vaal Triangle. In instances where fuel use data were obtained (e.g. Alrode and Heidelberg municipalities of Ekurhuleni; various Johannesburg sources; non-Scheduled Process sources within municipalities within the Vaal Triangle), US-EPA emission factors were applied.

The National Air Pollution Source Inventory used was compiled by the office of the Chief Air Pollution Control Officer (CAPCO) which resides within the Air Pollution Control Directorate (APCD) of the Department of Environmental Affairs (DEA). (This directorate forms part of the Chief Directorate: Air Quality Management.) The data base comprises total annual emissions data for criteria pollutants, including: SO2, NOx, and total particulate matter. Parameters included in the national emissions inventory data bases are as follows: source name, description of process, district within which process is located, region within which process is located, Scheduled Process No.

(refers to processes listed in the Schedule to the Air Pollution Prevention Act), reference to 1:50 000 topographical maps (e.g. 2526AD), indicating approximate location of sources, stack height (m), gas exit volume (m³ s-1), gas exit temperature (°C), title and name of contact person, address of contact person, telephone number of contact person and date of last emission data.

Air pollution control officers employed by DEA to administer scheduled industrial processes and local air pollution departments within the cities represented the main source of source and emissions data in many cases. Although certain of the local departments (e.g. Cape Town, Durban) have electronic emissions inventory data bases, paper-based systems are used in various other areas with data requiring to be captured electronically for use in the study.

Other sources of fuel use and/or emissions data which were reviewed included: (i) the national greenhouse gas emissions inventory compiled by the CSIR on behalf of the DEA, and (ii) the regional (southern African) emissions inventory being completed by the CSIR for the SADC Air Pollution Information Network (APINA) for use in the evaluation of transboundary air pollution issues. The emissions data generated for APINA were found to be of limited use due to the spatial resolution of these data (given for 20 km by 20 km grids across the country).

In addition to the emissions estimates, various source parameters were collated for each source for use in the atmospheric dispersion model study. Such parameters included: physical site location, height of emission above ground level, inner diameter of stack or vent, stack/vent gas exit velocity or volumetric flow and stack/vent gas exit temperature. The locations of the various sources inventoried are illustrated in Figure 17, Figure 18 and Figure 19 the Highveld, Cape Town and eThekwini respectively.

Figure 18: Location of industrial and commercial fuel burning sources within Cape Town

Figure 19: Location of industrial and commercial fuel burning sources within eThekwini

A synopsis of total annual emissions estimated for each conurbation is given in Table 7. The Vaal Triangle and the Mpumalanga Highveld regions are the largest sources of criteria gaseous pollutants and particulate matter. The extent of emissions within Cape Town and eThekwini are comparable, with eThekwini being estimated to be characterised by marginally higher particulate matter emissions and Cape Town higher sulphur dioxide emissions.

Total emissions from Johannesburg- and Tshwane-based industrial / commercial / institutional sources are relatively lower. It is, however, noted that the emission estimates for these regions include only releases from Scheduled Processes with fuel burning by smaller industries, institutions and businesses not having been accounted for.

Table 7: Total annual emissions from industrial, commercial and institutional fuel burning within each conurbation, excluding fuel burning for electricity generation

Total Annual Emissions from Industrial, Commercial and Institutional Fuel Burning (tpa) (excluding Electricity Generation for the National Grid) Pollutant Vaal Mpuma- Johannes Ekur- eThekwini Cape Town Pretoria Triangle langa burg huleni TOTAL (a) (a) (b) (a) (b) (c) (d)

PM10 3 311 2 767 5 775 ND 1 068 ND ND 12 920

SO2 13 059 24 896 231 187.6 275 665 3 143 9 462 14 447 571 860

NOX 5 087 4 495 121 436.3 151 021 895 650 4 653 288 238 Benzene 0.7 3.5 ND ND 0.1 ND ND 4.3 ND - no data - emission estimates not presented since fuel use data could not be established for a sufficient percentage of industries. The relevance of these sources will therefore need to be ranked in Task 4a based on their contribution to hospital admissions for respiratory ailments rather than on excess cancer cases. Notes: (a) Emission estimates includes fuel use by Scheduled Processes and other industries, businesses and institutions. (b) Emission estimates included only fuel use by Scheduled Processes. (c) Emission estimates included only the use of fuels such as coal, coke, HFO and diesel by Scheduled Processes. Information on gas consumption by Scheduled Processes and on fuel consumption by non-Scheduled Processes was not available to support emission estimations. (d) Emission estimates based on fuel use by Scheduled Processes within Ekurhuleni, but also accounts for fuel use by non-Scheduled sources (e.g. boiler operations at schools, offices and hospitals) within Alrode and Heidelberg municipalities.

Observations made regarding key fuels and industrial sectors contributing to the estimated total emissions within each conurbation are in subsequent subsections. a) Cape Town

Emissions from coal, HFO and wood burning predominate in Cape Town. Coal is estimated to be responsible for ~84% of the total particulate matter emissions from industrial, commercial and institutional fuel burning, HFO for 12% and wood for 3.5%. The contribution of HFO and wood is greater for finer particulate matter fractions (i.e. HFO and wood estimated to account for ~33% and

~10% of PM10 emissions). HFO is the predominant source of SO2 emissions (76%), followed by coal (~24%) with diesel and paraffin making minor contributions to total SO2 emissions. Wood burning is quantified as being responsible for over 90% of the benzene emissions from this source sector.

Significant sources include: textile industries (32% of SO2 and NOx), food and beverage industries (20% of emissions), petrochemical and chemical industries (16% of emissions) and commercial and institutional fuel burning (12% of emissions). Other industrial sectors which are of interest in terms of their contribution to atmospheric emissions within Cape Town include the pulp and paper subsector and non-metallic mineral processes which includes cement and brick manufacture. b) eThekwini

Fuel usage in this conurbation includes coal, HFO, diesel, paraffin, wood waste, light fuel oil and gas (refinery, LPG and other).

Coal consumption was reduced in the early 2000s by 20% and HFO use by 26% primarily as a result of a campaign to 'phase out dirty fuels' in the region. Coal burning was estimated to be responsible for a significant percentage of the particulate matter (over 98%), SO2 (85%), NOx

(66%) and benzene (59%) emissions associated with the industrial/commercial/institutional fuel burning sector. HFO combustion contributes to particulate matter, SO2 and NOx emissions, with wood burning representing a significant additional source of benzene emissions.

More refined HFO with lower sulphur contents are being used in the Durban South area. Of the total HFO consumed in the area, 32% of the HFO comprises 0.5% sulphur, 7% of the HFO has 2% sulphur and the remaining 60% is unrefined HFO with a sulphur content of 3.5%. Light fuel oil (LFO) with a sulphur content of 0.5% is also used in place of HFO in certain instances.

Significant industrial sources of criteria pollutant emissions include: sugar refining (26% of emissions), chemical and petrochemical industries (18% of emissions), pulp and paper industries (13% of emissions). Other significant sources include textile industries which primarily use coal and HFO and food and beverage industries which use mainly coal although other fuels (HFO, diesel, paraffin, LPG) are also used.

Bagasse is burned by the sugarcane industry for power generation purposes. Particulate matter represents the most significant pollutant emitted from bagasse-fired boilers. SO2 and NOx emissions are lower than for conventional fossil fuels due to the low levels of sulphur and nitrogen associated with bagasse. The use of bagasse by sugar mills within eThekwini, in addition to coal, places more emphasis on the contribution of this industry to particulate matter emissions. c) Vaal Triangle

Coal, coking coal and HFO use by industries within the Vaal Triangle are responsible for ~35%, ~50% and ~10% respectively of total particulate matter emissions from the industrial / institutional / commercial fuel use sector. Much of the particulate matter emissions associated with coking coal are due to the production of this fuel.

Coal represents the main fuel type used by the commercial and institutional sector although anthracite, diesel and wood are used to a lesser extent.

The largest industrial sources of particulate matter emissions within the Vaal Triangle include: iron and steel industries (50% of total particulate matter emissions) and chemical and petrochemical sector (30% of PM). Other groups include: brick manufacturers that use coal and other industries (use coal, and to a lesser extent HFO, for steam generation). d) Mpumalanga Highveld

Emissions from coal combustion by metallurgical and petrochemical industries represents the greatest contribution to total emissions from the industrial / institutional / commercial fuel use sector within the Mpumalanga Highveld. The metallurgical group is estimated to be responsible for at least 50% of the particulate matter emissions from this sector. This group includes iron and steel, ferro-chrome, ferro-alloy and stainless steel manufacturers.

Petrochemical and chemical industries are primarily situated in Secunda. The use of coal for power generation and the coal gasification process represent significant sources of sulphur dioxide emissions. Particulate emissions are controlled through the implementation of stack gas cleaning equipment.

Other groups include: brick manufacturers that use coal, wood burning and wood drying by various sawmills and other heavy industries (use coal and to a lesser extent HFO for steam generation). e) Johannesburg

Coal-fired boilers and brickworks that use coal for the firing of bricks in clamp kilns represent significant industrial fuel burning sources within Johannesburg. Given the limited extent of industrial operations within Johannesburg it is expected that fuel burning by the commercial and institutional sectors could represent a significant contribution to total non-domestic fuel combustion in this metro(8). Unfortunately fuel use data were not available for operations such as hospital and school boilers. f) Ekurhuleni

A wide range of industry types are located within this metro. Detailed fuel use figures for Alrode and Heidelburg municipalities indicate the predominant use of coal in this region.

The most significant groups contributing to fuel burning emissions from the industrial fuel burning sector within Ekurhuleni are chemical industries, and food and tobacco processes. Other source groups include: metallurgical processes (ranging from precious metal refining to foundry operations, non-metallurgical processes (including brick, cement and refractory brick manufacture), and pulp and paper operations. g) Tshwane

The large number of ceramic processes located within this conurbation is notable. Such processes include brick manufacturers, refractory operations and cement producers. Refractory operations primarily use tunnel kilns with clamp kilns being used in the manufacture of clay bricks. Cement manufacturers use rotary kilns. Fuel use data are still being obtained for industrial operations located within Tshwane. Coal is expected to be the main fuel in use by the ceramic processes operating in Tshwane.

Further sources of emission within this sector include diesel combustion by incinerator operations, and fuel combustion for steam generation within agricultural industries (e.g. chicken farms). Gas- fired boilers are used by certain industries, e.g. a large-scale glass manufacturer located in Olifantsfontein.

4.6.3 Residential Fuel Burning Despite the intensive national electrification programme a large number of households continue to burn fuel to meet all or a portion of their energy requirements. The main fuels with air pollution potentials used by households within urban areas are coal, wood and paraffin. These fuels continue to be used for primarily two reasons: (i) rapid urbanisation and the growth of informal settlements

8 In the 1970s a total of ~1 950 non-domestic fuel burning appliances were listed as operating within Johannesburg. Of the appliances for which fuel type were specified (i.e. total of 1 543), 37% of appliances used coal, 18% diesel, 15% coke, 15% gas (various), 9% anthracite, 2% HFO/oil, 2% refuse and 2% other (including paraffin, wood, sawdust and nuts).

has exacerbated backlogs in the distribution of basic services such as electricity and waste removal, and (ii) various electrified households continue to use coal and wood due to their cost effectiveness for space heating purposes and their multi-functional nature (supports cooking, heating and lighting functions). Coal is relatively inexpensive and is easily accessible in the region due to the proximity of the region to coal mines and the well-developed local coal merchant industry. Wood burning is more predominant in coastal areas due to their distances from the main coal producing areas.

Electricity accounted for only 38% of the total energy consumed by the residential sector during 2000, despite the extent of electrification, due to the persistent use of other fuels for space heating, cooking and lighting requirements (Figure 20)

Animal dung 1996 Animal dung Coal 2001 Coal 1.2% Wood 1.0% Cooking Wood 3.5% Animal dung Cooking 2.8% Animal dung 22.9% 20.5% Coal Coal Electricity Solar Electricity Solar Gas 0.2% Gas 0.0% Electricity Other Other Electricity 47.0% Paraffin Paraffin Paraffin 51.4% Paraffin Solar 21.4% Solar 21.5% Wood Wood Other Other Gas Gas 0.2% 0.7% 3.2% 2.5%

Animal dung 1996 Animal dung 2001 Coal Coal 0.7% 0.9% Heating Wood 6.6% Animal dung Heating Wood 8.1% Animal dung 24.6% 26.7% Coal Coal Electricity Electricity Gas Solar Gas Solar Other 0.2% Other 0.0% Electricity Paraffin Paraffin Electricity Paraffin Paraffin 44.4% Solar 14.6% 49.0% Solar 14.3% Wood Wood Other Other Gas 3.1% Gas 4.3% 1.2% 1.1%

Paraffin Paraffin 2001 Solar 1996 12.7% Solar Gas6.8% Candles Other 0.2% 0.0% Candles Lighting 0.2% 22.7% Lighting Gas 0.3% 28.6% Candles 0.4% Candles Other Electricity Electricity 0.8% Gas Gas Other Other Paraffin Paraffin Solar Solar Electricity Electricity 57.5% 69.7%

Figure 20: Percentage household fuel use for lighting, cooking and heating requirements for 1996 and 2001 (Statistics South Africa, 2002)

The estimation of domestic fuel burning emissions is challenging given that the amount of fuel being consumed is not known with certainty. The average coal usage per household varies depending on:  type of house (formal house, planned shack, unplanned shack or backyard shack)  whether or not a household is electrified  the number of people living in the house  the season  the availability of fuel types

 the price of fuels and the household income.

Estimates of the numbers of households within each conurbation using various fuel types (coal, LPG, paraffin and wood) were estimated based on work undertaken previously (Wicking-Baird, 1997; Scorgie, 2003, Scorgie et al., 2003; Irurah et al., 2000) and on energy use statistics and household numbers published by the South African Institute for Race Relations (2000). Numbers of households reported to use coal for meeting their energy requirements are illustrated in Figure 21 for the Highveld.

Figure 21: Number of households per km² using coal to meet their space heating and/or cooking requirements on the Highveld

Typical monthly fuel use figures, given by Afrane-Okese (1998) for various house types, were used together with the numbers of households using the various fuel types to estimate the total quantities of fuels being consumed.

Table 8: Estimated total annual household fuel consumption per conurbation Region Number of Coal Wood Paraffin LPG Households tpa tpa tpa tpa Johannesburg(a) 733 984 205 660 13 129 23 003 3 491 Vaal Triangle(b) 236 000 145 146 23 359 18 250 737 Mpumalanga 415 826 179 020 28 811 22 509 909 Highveld(c) Tshwane(c) 431 197 136 224 8 696 15 236 2 312 Ekurhuleni(c) 543 063 117 657 23 843 40 128 1 659 eThekwini(c,d) 646 918 27 326 65 209 41 299 621 Cape Town(c,d,e) 653 076 9 732 108 492 40 428 11 126 TOTAL 3 660 064 820 766 271 539 200 854 20 856 Extrapolated based on information from Scorgie et al. (2003) and fuel use figures per household given by Afrane-Okese (1998). (a) References: van Nierop, 1995; Scorgie, 2003.

(b) Extrapolated based on household energy use data from South African Institute for Race Relations (2000) and typical individual household fuel use figures published by Afrane-Okese (1998). (c) Extrapolated based on household energy use data from South African Institute for Race Relations (2000) and Irurah et al. (2000) and typical individual household fuel use figures published by Afrane-Okese (1998). (d) Extrapolated based on household energy use data from South African Institute for Race Relations (2000) and Wicking-Baird (1997) and typical individual household fuel use figures published by Afrane-Okese (1998).

In the determination of suitable emission factors reference was made to various sources including: US-EPA emission factors for residential fire places (US-EPA, 1996), the AEC's emission factors given for coal combustion within domestic stoves and braziers (Britton, 1998), in addition to the work of Wicking-Baird (1997), and Graham and Dutkiewicz (1999). A synopsis of the emission factors selected for application in the current study is given in Table 9. Total annual emissions calculated for each conurbation are summarised in Table 10.

Table 9: Emission factors for estimating household fuel combustion related releases

units SO2 NOx PM10 Benzene Coal g/kg 11.6(b) 4.0(e) 12(g) 0.0134(h) Paraffin g/l 0.1(c) 1.5(f) 0.2(f) (a) LPG g/kg 0.01 1.4 0.07 (a) Wood g/kg 0.18(d) 1.3(d) 15.7(d) 0.9(d) (a) No local or applicable international emission factors could be obtained for use in the study. (b) Initially a higher emission factor (19 g/kg) was used. Due to predicted ambient concentrations being significantly above those measure within intensive coal burning areas, this emission factor was revised. The revised factor of 11.6 g/kg is based on an assumed sulphur content of 0.61% with 5% of the sulphur being retained in the ash and the remainder emitted to atmosphere. (c) Given in g/kg, based on the sulphur content of paraffin (<0.01% sulphur). (d) Based on US-EPA emission factor for residential wood burning (US-EPA, 1996). (e) Based on the AEC household fuel burning monitoring campaign (Britton, 1998) which indicated that an average of 150 mg/MJ of NOx was emitted during cooking and space heating. Given a calorific value of 27 Mj/kg, the emission rate was estimated to be ~4 g/kg. (f) US-EPA emission factors for kerosene usage (US-EPA, 1996). (g) Initially taken to be 4 to 6 g/kg based on 2001 synopsis of studies pertaining to emissions from household coal burning (Scorgie et al., 2001). Results from simulations using this emission factor undertaken as part of the study indicated that fine particulate matter concentrations within household coal burning areas were under predicted by a factor of two. This emission factor was therefore scaled to 12 g/kg to facilitate more accurate simulation of airborne fine particulate matter within household coal burning areas. (h) Taken from Britton (1998).

Table 10: Total annual emissions due to household fuel combustion PM10 NO SO Benzene Region X 2 tpa tpa tpa tpa Johannesburg 1 054 414 4 106 14.6 Vaal Triangle 966 363 2 917 23.0 Mpumalanga 1 191 448 3 598 28.3 Highveld Tshwane 698 274 2 719 9.7 Ekurhuleni 865 358 2 581 23.0 eThekwini 1 144 430 882 59.1 Cape Town 1 752 633 548 97.8 TOTAL 7 670 2 919.3 17 350.9 255.4

Wood and coal are responsible for the bulk of the fine particulate matter emissions estimated for to occur due to household fuel burning across the various conurbations (Figure 22). Coal and wood were also the main sources of NOx emissions, with paraffin responsible for 10% and LPG for only

1% of such emissions. Coal was also estimated to be responsible for approximately 90% of the sulphur dioxide emissions with the remainder mainly due to paraffin combustion.

Figure 22: Contribution of specific fuels to total household fuel burning emissions across all conurbations

4.6.4 Vehicle Emissions Air pollution from vehicle emissions may be grouped into primary and secondary pollutants. Primary pollutants are those emitted directly into the atmosphere, and secondary are those pollutants formed in the atmosphere as a result of chemical reactions, such as hydrolysis, oxidation, or photochemical reactions. The significant primary pollutants emitted by motor vehicle exhausts include CO2, CO, VOCs, SO2, NOx, particulate matter and lead. Secondary pollutants formed due to vehicle exhaust emissions include: NO2, photochemical oxidants (e.g. ozone), HCs, sulphuric acid, sulphates, nitric acid, sulphates, nitric acid and nitrate aerosols. Emission estimates will only be undertaken for primary pollutants. The estimation of secondary pollutant formation would require complex photochemical dispersion modelling and is not within the scope of the current study.

In the estimation of petrol-driven vehicle emissions for the baseline scenario (2002) the following steps were followed:  The petrol-driven vehicle fleets were characterised based on the 1992 technology mix and the 1995 engine capacity profiles collated for the Vehicles Emission Project (Terblanche, 1995)(9). Information is given in Terblanche (1995) for Cape Town, Johannesburg, Durban, the Vaal Triangle and Pretoria. The Johannesburg and Vaal Triangle data were taken to be representative of the technology mix and engine capacities within the Mpumalanga Highveld region.  A current national vehicle population data base was obtained from Stellenbosch Automotive Engineering to supplement the spatially-resolved 1992 technology mix and 1995 engine capacity data obtained from Terblanche (1995). The national vehicle parc data, obtained by Stellenbosch Automotive Engineering for use in the Octane Study, comprises detailed information on petrol-driven vehicles sold between 1970 and 2002 including: engine

9 More current technology mix and engine capacity data is contained within the NaTIS data base. Access to such information could not, however, be obtained from the Department of Transportation during the study.

capacity, need for lead replacement petrol, presence of fuel injection and catalytic converters. The 1995 spatially-resolved engine capacity data were found to be very similar to the more current national vehicle population information and were therefore retained for use in the emissions estimations. The current national data however provided valuable data on the percentage of vehicles within the current live population that are fitted with catalytic converters (7.3%) and on the growth rate of catalytic converter use in new vehicles (47.3% of new cars purchased in 2002 were equipped with catalytic converters, with an annual average growth rate of 3.9% noted based on the 1990-2002 period).  Annual leaded and unleaded petrol sales data, obtained from SAPIA per magisterial district for 2001, were used to estimate the total vehicle kilometres travelled using fuel consumption rates suited to each engine capacity class and general fuel type(10). (Petrol consumption rates range from 7.7 to 15.1 litres per 100 km) (Wong, 1999).  Locally developed emission factors published by Wong (1999) were applied taking into account variations in such factors for different energy capacities and altitudes (coastal, Highveld factors). Emission factors used are given in Appendix B. Emissions were calculated by multiplying the emission factors by the total vehicle kilometres travelled (VKT) estimated on the basis of the 2001 fuel sales data.

In the estimation of diesel-driven vehicle emissions for the baseline scenario the following steps were followed:  Average percentages of light commercial vehicles (LCVs) and medium and heavy commercial vehicles (M&HCVs) within the national diesel vehicle fleet were obtained from Stone (2000)(11).  Diesel consumption rates were obtained for LCVs, MCVs and HCVs for coastal and highveld applications from Stone (2000) and Wong (1999). Such rates varied from 10.5 to 24.4 litres per 100 km.  Annual diesel sales data, obtained from SAPIA per magisterial district for 2001, were used to estimate the total vehicle kilometres travelled using fuel consumption rates suited to each vehicle weight category.  Locally developed emission factors published by Stone (2000) were applied taking into account variations in vehicle weight categories and altitudes (coastal, highveld factors)(12). Emission factors used are given in Appendix B. Emissions were calculated by multiplying the emission factors by the total vehicle kilometres travelled (VKT) estimated on the basis of the 2001 fuel sales data.

In the estimation of hydrocarbon releases due to evaporative emissions reference was made to the work of Wong and Dutkiewicz (1998). They found the total daily loss (i.e. total HC emissions due

10 Although detailed petrol sales data by specific petrol type were obtained per conurbation emission factors were not available for most types. The development of tailored, scientifically sound emissions factors for use in such instances could not be undertaken within the scope of the study. 11 More detailed diesel fleet data are likely to be available within the NaTIS data base but these data was not accessible during the project. 12 The 2003 reduction in the sulphur content of diesel from 0.5% to 0.3% was accounted for and the emission factors adjusted accordingly.

to hot soak and evaporative emissions) to represent between 16% and 27% of the total daily HC emissions (including tailpipe exhaust emissions plus total daily loss).

Total annual vehicle emissions, including exhaust and evaporative emissions, estimated per conurbation are given in Table 11, with a more detailed emission estimates by conurbation and pollutant provided in Appendix C.

Spatial allocation of vehicle emissions is needed for the purpose of the atmospheric dispersion simulations undertaken. Given the absence of comprehensive traffic count data, emissions were spatially allocated on the basis of variations in fuel sales per magisterial district, road densities, projected trip numbers and any available traffic count data available. The overlay of fuel sales data over road density is illustrated for Johannesburg in Figure 23. Main roadways were modelled as line sources with the remainder of emissions allocated within CBDs and other built up areas.

Table 11: Total annual emissions due to vehicle emissions (exhaust and evaporative releases) - baseline scenario (2002)

Total Annual Emissions (tpa) per Conurbation POLLUTANT UNITS Johannesburg Cape Town Vaal Triangle eThekwini Tshwane Ekurhuleni Mpumalanga TOTAL

NOX tpa 45 939 56 461 10 639 62 457 29 573 37 359 24065 266 495

SO2 tpa 5 433 8 106 1 904 11 121 4 054 6 049 5781 42 448 1,3-butadiene tpa 367 261 60 239 215 236 81 1 460 Benzene tpa 395 447 65 404 230 254 82 1 877 Particles tpa 1 085 1 603 400 2 286 827 1 257 1247 8 704

Figure 23: Petrol sales per magisterial district and road network density for the Johannesburg and Ekurhuleni metropolitan areas

4.6.5 Biomass Burning (Agricultural and Wild Fires) To estimate the extent of biomass burning it was necessary to quantify the average area burned within each conurbation. Satellite imagery was obtained to identify and quantify burn scar areas. Burn scar images generated included five-year composite scar plots (1995-2000) and plots indicating the extent of areas burned during a single fire season. The five-year burn scar composite plots for the Plateau, Cape Town and eThekwini regions are given in Figure 24, Figure 25 and Figure 26 respectively.

Figure 24: Remote sensing burn scar data showing incidences of fires during the period 1995 – 2000 over the Plateau. The Tshwane, Johannesburg-Ekurhuleni, Vaal Triangle and Mpumalanga Highveld study areas are shown.

Figure 25: Remote sensing burn scar data showing incidences of fires during the period 1995 – 2000 over Cape Town. The legend shows the number of fires recorded to occur at each location during this period; the data range being no fires to 4 fires.

Figure 26: Remote sensing burn scar data showing incidences of fires during the period 1995 – 2000 over eThekwini. The legend shows the number of fires recorded to occur at each location during this period; the data range being no fires to 2 fires.

A synopsis of the biomass burn frequency information is presented in Table 12. The percentage of the total area within each region predicted to have been burnt during the 1995-2000 period were as follows: Johannesburg (28%), Vaal Triangle (25%), Mpumalanga Highveld (12%), Tshwane (24%), eThekwini (4%) and Cape Town (11%). Emission factors derived during SARAFI-2000 (Southern African Fire-Atmosphere Research Initiative), as published by Andreae et al. (1996), were obtained for application in the estimation of atmospheric emissions from veld fires (Table 13).

Table 12: Extent of area burnt within each conurbation - given as a composite area for the 1995- 2000 period, as a total area for the 2000 fire season and indicating average and peak burn areas over 10-day periods

Total area Area Average Average Peak km2 Average Total Area burnt km2 burnt Peak % of Conurbation / % of area km2 of of area km2 of of region over during fire area burnt Region burnt in area burnt burnt in area burnt km2 dataset season in 10 days 10 days in 10 days 10 days per year 1995-2000 2000 eThekwini 5 389 237 29 0.03% 0.07% 1.84 3.58 67.14 Cape Town 5 575 631 127 0.09% 0.21% 4.89 11.53 178.65 Johannesburg 7 560 2 112 168 0.22% 0.28% 16.37 20.97 597.69 Vaal Triangle 2 434 615 25 0.20% 0.13% 4.77 3.07 174.02 Mpumalanga 37 271 4 304 472 0.09% 0.16% 33.36 58.98 1217.75 Highveld Tshwane 4 579 1 086 127 0.18% 0.35% 8.41 15.85 307.14

Table 13: Emission factors used to quantify atmospheric emissions from biomass burning

Pollutant Emission Factor Unit -1 NOX 3.1 g kg dry matter -1 SO2 0.6 g kg dry matter TPM 10 g kg-1 dry matter -1 PM2.5 5 g kg dry matter

The quantity of "dry matter" per unit area ranges from 4.5 ton per hectare for savannah areas to 25 ton per hectare for sugar cane. Total annual emissions were estimated based on the average annual area burnt taking into account the composite 1995-2000 burn scar areas (Table 14). Peak emissions were calculated based on the maximum area burnt in any 10-day period. Sugar cane fires mainly take place within eThekwini. Emission estimates from sugar cane burning were obtained from Ecoserv (1998) for inclusion in the biomass burning emissions for this conurbation.

Table 14: Total annual emissions estimated due to biomass burning within various conurbations

Total Annual Emissions due to Biomass Burning (tpa) Region NOX SO2 TPM PM2.5 Johannesburg - Ekurhuleni 834 161 2 690 1 345 Vaal Triangle 243 47 783 392 Mpumalanga Highveld 1 699 329 5 480 2 740 Tshwane 428 83 1 382 691 eThekwini 94 18 1 215(a) 608(a) Cape Town 249 48 804 402 (a) Includes emissions from sugar cane burning.

4.7 Dispersion Modelling of Air Pollutant Concentrations

4.7.1 Atmospheric Dispersion Model Selection Atmospheric dispersion models compute ambient pollutant concentrations as a function of source configurations, emission strengths and meteorological characteristics. Such models provide a useful tool to ascertain the spatial and temporal patterns in the ground level concentrations arising from the emissions of various sources. Spatial and temporal variations in air pollutant

concentrations are required to be known for the estimation of risks associated with human exposures to such concentrations.

The simulation of pollutant concentrations due to inventoried fuel combustion processes was undertaken through the application of the United States Environmental Protection Agency (US- EPA) approved California Air Resources Board (CARB) CALPUFF model (US-EPA, 1995a). The CALPUFF model is a non-steady state puff dispersion model. Due to its puff-based formulation the model is able to account for various effects, including spatial variability of meteorological conditions and dispersion over a variety of spatially varying land surfaces. The simulation of plume fumigation and low wind speed dispersion are also facilitated.

Despite the complex and data intensive nature of the CALPUFF model, it was selected for use in the current study on the on the following grounds:  the model is applicable to modelling domains as large as ~250 km in extent;  it is able to characterise spatial variations in meteorological conditions and is therefore applicable for use in complex terrain, urban and coastal environments;  the model is able to undertake first order chemical transformation calculations and is therefore suited to the prediction of secondary pollutants (e.g. quantified conversion of sulphur oxides and nitrogen oxides to sulphates and nitrates which contribute significantly to ambient fine particulate matter concentrations); and  CALPUFF is appropriate for various source configurations including point, volume, area and line sources.

Comparisons between CALPUFF results, and results generated by the Industrial Source Complex Model Short Term version 3 (ISCST3) model, have shown that CALPUFF is generally more conservative (Strimatis et al., 1998). The ISC model typically produces predictions within a factor of 2 to 10 within complex topography with a high incidence of calm wind conditions. When applied in flat or gently rolling terrain, the USA-EPA (US-EPA, 1986) considers the range of uncertainty of the ISC to be -50% to 200%. CALPUFF predictions have been found to have a greater correlation with observations, with more predictions within a factor of 2 of the observations when compared to the ISC model (Strimatis et al., 1998). It has generally been found that the accuracy of off-the-shelf dispersion models improve with increased averaging periods. The accurate prediction of instantaneous peaks are the most difficult and are normally performed with more complicated dispersion models specifically fine-tuned and validated for the location. The duration of these short-term, peak concentrations are often only for a few minutes and on-site meteorological data are then essential for accurate predictions.

4.7.2 Meteorological modelling CALPUFF requires as a minimum the input of hourly average surface meteorological data. In order to take full advantage of the model‟s ability to simulate spatially varying meteorological conditions and dispersion within the convective boundary layer it is, however, necessary to generate a three-dimensional wind field for input to the CALPUFF model. The CALMET model may be used to generate such a three-dimensional wind field for input to the CALPUFF model.

The CALMET meteorological model contains a diagnostic wind field module that includes parameterized treatments of terrain effects, including slope flows, terrain channelling and

kinematic effects, which are responsible for highly variable wind patterns. CALMET uses a two- step procedure for computing wind fields. An initial guess wind field is adjusted for terrain effects to produce a Step 1 wind field. The user specifies the vertical layers through which the domain wind is averaged and computed, and the upper air and surface meteorological stations to be included in the interpolation to produce the spatially varying guess field. The Step 1 (initial guess) field and wind observational data are then weighted through an objective analysis procedure to produce the final (Step 2) wind field. Weighting is undertaken through assigning a radius of influence to stations, both within the surface layer and layers aloft. Observational data are excluded from the interpolation if the distance between the station and a particular grid point exceeds the maximum radius of influence specified (US-EPA, 1995b; Scire and Robe, 1997; Robe and Scire, 1998).

Domain-scale winds are used to compute a terrain-forced vertical velocity, subject to an exponential, stability-dependent decay function. The kinematic effects of terrain on the horizontal wind components are evaluated by applying a divergence-minimization scheme to the initial guess wind field. The divergence minimization scheme is applied iteratively until the three-dimensional divergence is less than a threshold value.

Slope flow is parameterised in terms of the terrain slope, terrain height, domain-scale lapse rate, and time of day. The slope flow wind components are added to the wind field adjusted for kinematic effects. The thermodynamic blocking effects of terrain on the wind flow are parameterised in terms of the local Froude number. If the Froude number at a particular grid point is less than a critical value and the wind has an uphill component, the wind direction is adjusted to be tangential to the terrain.

The CALMET model incorporates boundary layer models for application to overland and overwater areas. Over land surfaces, an energy balance method is used to compute hourly gridded fields of the sensible heat flux, surface friction velocity, Monin-Obukhov length, and convective velocity scale. Mixing heights are determined from the computed hourly surface heat fluxes and observed temperature soundings using a modified Carson (1973) method. Gridded fields of Pasquill-Gifford-Turner (PGT) stability class and optional hourly precipitation rates are also determined by the model.

By using CALMET and CALPUFF in combination it is possible to treat many important complex terrain effects, including spatial variability of the meteorological fields, curved plume trajectories, and plume-terrain interaction effects. Maximum hourly average, maximum daily average and annual average concentrations for the various pollutants identified were simulated through the application of CALPUFF, using as input the compiled emissions inventory and the three- dimensional CALMET data set.

4.7.3 Modelling Domain Definition CALMET was used to simulate the wind field within the various study areas. Three modelling domains were selected for these areas viz.:  Plateau modelling domain (Figure 27) - comprises an area of 52 251 km² (272 km east-west by 192.1 km north-south). A grid resolution of 1.7 km was used with the grid comprising 160 cells along the east-west axis by 113 cells north-south. The centre of the modelling domain was at the location 26.67°S and 28.67°E.

The Plateau modelling domain encompassed the Mpumalanga Highveld, Vaal Triangle, Tshwane, Johannesburg and Ekurhuleni conurbations. It is necessary that these areas be modelled together given the potential which exists for sources within one conurbation to affect the air quality in the adjacent areas.

 Cape modelling domain (Figure 28) - comprises an area of 14 991 km² (135 km east-west by 111 km north-south). A grid resolution of 0.925 km was used with the grid comprising 146 cells along the east-west axis by 120 cells along the north-south axis. The centre of the modelling domain was at the location 34.5°S and 18.25°E.  eThekwini modelling domain (Figure 29) - comprises an area of 6 350 km² (75.6 km east- west by 84 km north-south). A grid resolution of 0.7 km was used with the grid comprising 108 cells along the east-west axis by 120 cells north-south. The centre of the modelling domain was at the location 30.21°S and 30.5°E.

Figure 27: Plateau modelling domain which includes the Mpumalanga Highveld, Vaal Triangle, Johannesburg, Ekurhuleni and Tshwane study areas.

Figure 28: Spatial extent of the Cape modelling domain. The modelling domain coincides exactly with the area shown in the figure.

Figure 29: Spatial extent of the eThekwini modelling domain. The modelling domain coincides exactly with the area shown in the figure.

4.7.4 Source and Emissions Data Source configuration data required for input to the dispersion simulations included the following:  point sources (e.g. stacks and vents) - stack heights, stack diameters, gas exit temperatures, gas exit velocity, source location and elevation;  area sources (e.g. biomass burning; domestic fuel burning) - coordinates of area source corners, source elevation and height of release; and  line sources (e.g. major roads) - coordinates of the beginning and end points of line segments, source elevation and height of release, source width. The above data were collated during the establishment of the emission inventories for the various conurbations as documented in Section 4.6. For point sources for which source configuration data were not available approximations were made based on the type and extent of the source.

During the compilation of the emissions inventory total annual emissions were calculated/collated. In the simulation of such emissions it was important to account for temporal variations in the emissions of certain sources such as domestic fuel burning and vehicle emissions.

Seasonal and diurnal trends in domestic fuel burning practices were accounted for through the estimation of hourly emission rates occurring as a result of fuel combustion for space heating purposes. Diurnal trends were also applied for the remaining household fuel burning undertaken for lighting, cooking and water heating purposes.

The demand for residential space heating, and hence the amount of coal burning, has been found to be strongly dependent on the minimum daily temperature (Annegarn and Sithole, 1999). Seasonal trends in space heating needs, and therefore in coal burning emissions, were estimated by Annegarn and Sithole (1999) by calculating the quantity of "heating-degree-days" (HDD). The HDD quantity was calculated based on the assumption that space heating is required only if the minimum daily temperature falls below 8°C. Based on the work of Annegarn and Sithole (1999), HDD was defined by the study as follows: N HDD   Tmin )8( i1 where i = 1 to N number of days per month; N = Total number of days per month; Tmin = minimum daily temperature for day i (°C), for Tmin <8°C.

HDDs were then normalised to a 30-day month to adjust for the different month lengths. To illustrate seasonal trends in the number of HDD, variations in the number of heating degree days calculated for Johannesburg each month during the period 1990 to 1999 are presented in Figure 30. This intra-annual trend in HDDs was used to apportion coal burning emissions estimated on a total annual basis.

Figure 30: Intra- and inter-annual variations in the number of "heating-degree-days" per month for Johannesburg for the period 1990-1999

To assess diurnal trends in domestic coal burning emissions, reference was made to variations in aerosol black carbon (BC) concentrations measured during a study conducted in Soweto (Annegarn et al., 1999) as illustrated in Figure 7.

Three distinct periods of atmospheric black carbon dilution/removal were identified by Annegarn et al. (1999) based on the diurnal trends of aerosol black carbon observed. The first period was defined by the hours between 18h00 and 21h00 (Period A). Following a peak concentration at 18h00, the black carbon concentration decayed rapidly until 21h00. The second period comprises the slower decline of black carbon levels between 21h00 and 05h00 (Period B). The rapid decrease of ambient black concentration levels between 09h00 and 12h00, representing the third decay period (Period C), is very similar to that of the first identified for the 18h00 to 21h00 period (Annegarn et al., 1999).

A series of exponential curves were fitted to a time series comprising the geometric mean concentration for seven days in June 1997 in order to show the rate of decay of atmospheric black carbon concentrations at various times of the day. Exponential decay curves were fitted to each of the three decay periods noted, and the following characteristic half-lives were estimated (Annegarn et al., 1999): t(1/2)A = 2 h 20 min; t(1/2)B = 4 h 15 min; and t(1/2)C = 44 min.

The fitted curves are shown in Figure 31. The decay in period A represents a moderate clearing rate. Period B is characterised by a slow clearing rate as the air at night is most stable and the mixing depth is significantly reduced. Dilution during this period is mainly due to slow horizontal advection. Period C has the most rapid clearing rate due to the dissipation of the nocturnal

inversion about two hours after sunrise, and the onset of convective turbulence induced by the heating of the ground by insolation. Black carbon concentrations occurring between midday and 4pm are indicative of the regional air mass comprising diluted nocturnal emissions with pollutants transported into the region from sources further afield.

Figure 31: Exponential curves indicating decay times for atmospheric black carbon concentrations. The measured diurnal curve represents the geometric mean values of seven days of sampling (Annegarn et al., 1999).

The large diurnal variations in black carbon concentrations and the ventilation potentials (clearing rates) calculated reflect the regular, but intermittent contributions of coal burning emissions to ambient particulate matter concentrations. Use was made of the exponential curves generated by Annegarn et al. (1999) in the apportionment of total daily emissions on an hourly basis.

Seasonal trends in vehicle activity, and hence emissions, are not clearly apparent (except for a slight reduction in vehicle activity in certain areas during the month of December). Distinctive diurnal trends in vehicle activity are however apparent. Diurnal profiles were applied in order to calculate hourly emissions from vehicles. Readily available traffic count data were used in the characterisation of such diurnal trends. Variations in vehicle volumes along the N1 highway in Johannesburg are illustrated in Figure 32, representing a typical pattern of traffic activity. Approximately 80% of vehicle activity takes place during the day-time, with activity peaking sharply in the morning and rather more diffusely in the afternoon.

Seasonal trends in fire occurrence probabilities were taken into account in the apportionment of biomass burning emissions. Such probabilities were calculated based on a five-year record of burn scar information derived from remote sensing data (Figure 33).

100

Figure 32: Example of diurnal trends in traffic volumes, as recorded along the N1 in Johannesburg (between 14th and Gordon Avenue off-ramps) during the period 14 September to 13 October 1999

Figure 33: Monthly variations in the occurrence of fires in Cape Town, eThekwini and on the Plateau – compilation based on a five-year record of burn scar information derived from remote sensing data

Emissions due to power generation and industrial and commercial fuel combustion were assumed to remain constant throughout the year.

101

4.7.5 Meteorological Data In the simulation of vertical variations in meteorological parameters provision was made in the CALMET modelling for four vertical layers in the atmosphere with 'ceiling heights' of 20 m, 200 m, 500 m, and 1 000 m to 1 500 m above the ground. This vertical structure allows for levels completely within the valleys, transitional levels and layers above most of the terrain to be parameterised.

The CALMET meteorological model requires hourly average surface meteorological data as input, including wind speed, wind direction, mixing depth, cloud cover, temperature, relative humidity, pressure and precipitation. The mixing depth is not measured routinely and needed to be calculated, based on readily available data, viz. temperature and predicted solar radiation. The daytime mixing heights were calculated with the prognostic equations of Batchvarova and Gryning (1990), while night-time boundary layer heights were calculated from various diagnostic approaches for stable and neutral conditions. Surface meteorological data were obtained from eight weather stations for use in the CALMET meteorological modelling undertaken for the Cape modelling domain (Figure 34).

Figure 34: Location of the weather stations used as input to the CALMET model for the Cape modelling domain

102

For the eThekwini modelling domain surface meteorological data for input to CALMET were obtained for six stations, viz. Wentworth, Southern Works, SAPREF, BLUFF, Athlone Park and AECI. Sea surface temperature data and weather data from ships were also obtained for use in the meteorological modelling of the Cape and eThekwini areas in order to facilitate the parameterisation of land-sea circulations.

Meteorological data for a number of weather stations were used in the CALMET modelling undertaken for the Plateau region, including Irene (Pretoria), Witbank, Johannnesburg, Vereeniging, Sasolburg, and Ermelo. Wind roses generated for the modelling domain are illustrated in Figure 35. Although wind roses are also indicated in the figure for Eskom-owned meteorological station (e.g. Kendal, Elandsfontein, Palmer and Verkykkop), data from these stations were not used in the meteorological modelling since permission for their use was not obtained.

Figure 35: Wind roses generated based on hourly wind field data from various meteorological station located within the Plateau modelling domain

Upper air data required by CALMET includes pressure, geopotential height, temperature, wind direction and wind speed for various levels. The Weather Service's radiosonde stations at Cape Town, Durban and Irene (Pretoria) were used in the modelling. These are the only upper air stations within the modelling domains. Meteorological data, recorded twice daily for seven sounding levels, were included in the meteorological simulations.

103

4.7.6 Terrain and Landuse Data The geophysical data required as input to CALMET includes land use type, elevations and various surface parameters. The land use and elevation data are entered as gridded fields. Topography data were purchased in digital form from the Surveyor General for use in the study. An example of terrain elevations within the Plateau modelling domain is given in Figure 36.

Figure 36: Topography of the Plateau modelling domain

Land use inputs for the eThekwini modelling domain are illustrated in Figure 37. The index used in this assessment is based on the Geological Survey land use and land cover classification system with main categories (level 1) and sub-categories (level 2). Land use categories were assigned per grid-square based on the topographic map indication. The predominant categories are urban build- up areas and agricultural land.

104

Figure 37: Land use classification of the eThekwini modelled area represented as a raster map

4.7.7 Dispersion Model Verification and Model Outputs Predicted air pollutant concentrations were compared with air quality monitoring data were available and indicative of pollution levels primarily from fuel burning sources. Air pollutant concentrations predicted due to most source groups were generally within the margins of accuracy of CALPUFF (i.e. -50% to +200%), with the exception of biomass burning related emissions.

Predicted air pollutant concentrations due to biomass burning related emissions were found to be unrepresentative, significantly overestimating pollutant levels in many instances. This is likely to be due to the difficulty in modelling such emissions accurately. Due to their episodic nature and unpredictable duration, biomass burning sources could not be adequately modelled given the

105

absence of detailed temporal information on fire incidence and the dispersion modelling methodology adopted for the study. The model predictions for agricultural and wild fires were therefore not used for the health risk and costing study components.

Highest hourly, highest daily and annual average air pollutant concentrations due to other source groups were output for each grid cell within the various modelling domains for use in the health risk estimation.

4.8 Health and Welfare Risk Estimation

4.8.1 Inhalation Health and Welfare Risk Dose-response functions or coefficients provide the link between exposures to ambient air pollutant concentrations and the resultant health outcomes. Given the absence of locally generated relationships it was necessary to make reference to the international literature to identify dose- response functions which are applicable to South Africa.

A synopsis of the non-carcinogenic linear dose-response functions and cancer risk factors selected for use in the study are given in Table 15 and Table 16 respectively. The dose-response coefficients and cancer risk factors were obtained from a number of meta-health studies that summarised the findings of various epidemiological, clinical and toxicological studies (WHO, 2000; CEPA/FPAC Working Group, 1998; Koenig, 2000; Fourie et al., 2003). Several of these functions have been used in previous externality studies (van Horen, 1996; Seethaler, 1999; Kim and Qiang, 2002; Holland and Watkiss, 2002; Nelson, 2000).

Table 15: Dose-response functions selected for the quantification of inhalation exposures to air pollutant concentrations due to fuel combustion emissions

POLLU- POPULATION HEALTH END POINT FUNCTION SOURCE TANT SECTOR Respiratory hospital admissions PM All persons 1.20 x 10-5 Ostro, 1994 as cited by World Bank, 1998 - daily exposures 10 Respiratory hospital admissions SO All persons 2.01 x 10-6 Maddison, 1997 as cited by WB 1998 - daily exposures 2 Respiratory hospital admissions NO All persons 1.65 x 10-6 Maddison, 1997 as cited by WB 1998 - daily exposures 2 Cardiovascular hospital PM All persons 1.01 x 10-7 Dockery et al., 1989 admission - daily exposures 10 Daily Mortality - daily PM >=65 years 4.42 x 10-7 EXMOD - as referenced by Nelson, 2000 exposures 10 Daily Mortality - daily PM <65 years 2.35 x 10-8 EXMOD - as referenced by Nelson, 2000 exposures 10 Holland and Watkiss (2002) - functions Daily Mortality - daily SO >=65 years 1.01 x 10-8 collated for application by the European exposures 2 Commission DG Environment Holland and Watkiss (2002) - functions Daily Mortality - daily SO <65 years 1.38 x 10-9 collated for application by the European exposures 2 Commission DG Environment Chronic Bronchitis (a) - annual children PM 1.61 x 10-3 Dockery and Pope, 1994 exposures 10 (<5 years) Chronic Bronchitis (a) - annual adults PM 4.90 x 10-5 Abbey et al., 1995 exposures 10 (20 years+) Restricted activity days (RAD) - Rowe et al., 1994 as cited by van Horen PM 20 - 65 years 1.60 x 10-4 daily exposures 1996

106

POLLU- POPULATION HEALTH END POINT FUNCTION SOURCE TANT SECTOR Holland and Watkiss (2002) - functions collated Minor restricted activity days -3 SO2 20 - 65 years 9.76 x 10 for application by the European Commission (MRAD) - daily exposures DG Environment (a) Chronic bronchitis is a condition characterised by excessive mucus production in the airways (bronchi) (Department of Health, 1998). The most important risk factor for chronic bronchitis in developed countries is tobacco smoking, although genetic predisposition, early childhood respiratory infections, occupational exposures and outdoor air pollution play contributory roles. Additional risk factors in South Africa include indoor smoke pollution, the chronic effects of lung infection such as tuberculosis and the combination of sub-optimal nutrition and respiratory infection in early life.

Table 16: Synopsis of cancer unit risk factors(a) selected for use in the study for quantifying inhalation exposures to air pollutant concentrations due to fuel combustion emissions

-3 Sector of Averaging Unit Risk Factor (Risk due to 1 µg m exposure) Health Endpoint Population period Benzene(b) 1,3-butadiene

Leukemia all annual 0.000006 0.00003

(a) Unit risk factors are defined as the estimated probability of a person (60-70 kg) contracting cancer as a result of

constant exposure to an ambient concentration of 1 µg m-3 over a 70 year lifetime. Unit risk factors were obtained from the WHO (2000) and from the US-EPA IRIS data base (accessed July 2003). (b) The WHO gives the cancer risk for benzene as being in the range of 4.4 x 10-6 to 7.5 x 10-6. IRIS gives the benzene risk as being in the range of 2.2 x 10-6 to 7.8 x 10-6. For the purpose of the current study the geometric mean of the WHO's range of estimates was used (i.e. 6 x 10-6).

Dose-response relationships for PM10, SO2 and NO2 exposures are typically expressed for ailments related to the human respiratory function. Health endpoints considered included respiratory hospital admissions and incidences of chronic bronchitis and asthma. Exposures to benzene and 1,3-butadiene were quantified in terms of their being carcinogens. Cancer risks due to exposure to these pollutants were considered additive when the same target organ was affected; 1,3-butadiene and benzene both cause leukaemia.

Non-carcinogenic dose-response functions are applied by multiplying the exposure (i.e. population x pollutant concentration) with the function to obtain an indication of impact. Impacts are expressed for various health endpoints, e.g. number of hospital admissions due to respiratory ailments and cardiovascular related symptoms, number of premature deaths, number of cancer cases.

Unit risk factors are applied in the calculation of carcinogenic risks. These factors are defined as the estimated probability of a person (60 - 70 kg) contracting cancer as a result of constant exposure to an ambient concentration of 1 µg m-3 over a 70 year lifetime. Unit risk factors were obtained from the WHO (2000) and from the US-EPA IRIS data base. The WHO gives the cancer risk for benzene as being in the range of 4.4 x 10-6 to 7.5 x 10-6. IRIS gives the benzene risk as being in the range of 2.2 x 10-6 to 7.8 x 10-6. For the purpose of the current study the geometric mean of the WHO's range of estimates was used (i.e. 6 x 10-6). Cancer risk factors for formaldehyde and 1,3-butadiene are given as 1.3 x 10-2 and 3 x 10-5 respectively (IRIS, 2003).

Although the study focused primarily on direct health effects associated with inhalation exposures it was decided to also consider the potential for economic (indirect) effects associated with lost productivity due to absenteeism. To quantify the extent of absenteeism due to inhalation related illnesses reference was made to a risk factor given by Rowe et al. (1994) for restricted activity days

107

(RAD) (as cited by van Horen, 1996) (Table 15). These absenteeism-related coefficients were applied to persons in the 20 to 65 year age group; persons within this group being assumed to be potentially economically active. During the economic costing of productivity losses by UCT (2004), the rate of unemployment was taken into account in quantifying monetary effects.

4.9 Direct Health Cost Projection Costs resulting from inhalation exposures to air pollution include direct and indirect costs. Direct costs are associated with health spending, i.e. cost of hospital admissions and medication. Indirect costs include financial losses due to reduced productivity resulting from the restricted activity of economically active persons. Loss of life is costed in many externality studies, in some cases being expressed as costs for reduced life expectancies.

The scope of the current study was restricted to the costing of health spending in monetary terms to provide an estimate of direct damages. Reference was made to health costs for health effects associated with air pollution inhalation. The ratio of inpatients to outpatients, and public and private costs of treatment for both inpatients and outpatients, was taken into account.

Based on the outcomes of the modelling study, costs due to estimated mortalities and indirect costs related to productivity losses resulting from absenteeism projections were quantified as part of the economic analysis undertaken by the University of Cape Town (2004). The findings of this study are taken into account in the prioritisation of sources and recommendation of interventions.

4.9.1 Data Availability and Study Assumptions Information on the incidences and costs of respiratory illnesses within South Africa were unavailable from the Department of Health. Such data were also not available in existing literature. To obtain these data it was necessary to employ Medscheme to undertake queries on their extensive database. Medscheme, a private medical aid company, holds one of the largest databases in the country comprising information on health effect incidences and costs for patients on medical aid and patients reliant on public health care in South Africa. The costs incurred to purchase the data required for the study were covered by the NEDLAC funding received for the Dirty Fuels Study.

Average private patient costs were estimated by calculating a weighted average of the high and low cluster costs reported for Medscheme patients, including both outpatients and inpatients. To account for outpatient costs, inpatient:outpatient ratios were calculated so that the number of hospital admissions (inpatients) predicted on the basis of the dose-response coefficients could be used to infer a corresponding number of outpatients so as to obtain a total number of patients for each health condition. The ratios of inpatients to outpatients calculated from the Medscheme information are summarised in Table 17.

Table 17: Ratios of inpatients to outpatients calculated from the Medscheme information obtained for each health condition

Health Condition Ratio of Inpatients to Outpatients Respiratory illness 0.049

108

The following assumptions were made, based on consultations with Medscheme personnel:  Public health costs are approximately 70% of private health costs.  In South Africa there are 7 million people on a medical aid and 33 million people who rely on public health care. (i.e. approximately 17.5% of the population, based on national figures, was likely to receive treatment from the private health care sector with the remaining 82.5% of people receiving treatments from public hospitals and clinics).

The above assumptions permitted the apportionment of the total number of patients into four groups with appropriate costs calculated for each group, viz.:  Private inpatients  Public inpatients  Private outpatients  Public outpatients.

A synopsis of the costs for each of the above groups, calculated from the Medscheme information, is given in Table 18. Weighted average lengths of stay (LOS) for the inpatient groups, calculated from the Medscheme data, are summarised in Table 19. (The average length of stay was also provided as input to the economic assessment undertaken by UCT for estimation of absenteeism durations due to exposures.)

Table 18: Direct health costs of public and private inpatients and outpatients calculated from information obtained from Medscheme (costs given per patient)

Direct Health Cost (2002 Rands) Cost of Private Cost of Public Cost of Private Cost of Public Health Condition Inpatients Inpatients Outpatients Outpatients Respiratory illness 12 235 8 565 1 295 907 Leukaemia 233 069 163 148 NA NA

Table 19. Average length of stay associated with each health condition costed

Health Condition Average Length of Stay (LOS) (given in days) Respiratory illness 4.37 Leukaemia 9.43 admissions per annum x 8.41 average length of stay days = 79.3 days

4.10 Selection of Interventions In the identification of potential options and interventions that may be implemented to reduce atmospheric emissions from fuel combustion-related sources attention was paid to technological and other options. Other options include legislative and regulatory tools, market interventions and education and awareness programmes.

Reference was made to the large amount of work has been done locally on options and interventions to reduce domestic fuel burning (van Niekerk and Swanepoel, 1999; van Niekerk and van Niekerk, 1999; Irurah et al., 2000; PEER Africa, 1997). In the identification of emission reduction options for vehicles, attention was paid to the work conducted by the South African Petroleum Industries Association (SAPIA) on cleaner petrol and diesel, and to the National

109

Automobile Manufacturers Association South Africa (NAMSA) investigations into alternative (non-petrol powered) vehicles.

Considerable information is available locally and internationally on industry specific technological options for the replacement of 'dirty fuels'. Such information was taken into account and local experiences in the field noted, e.g. Eskom's proposed implementation of pebble-bed reactor technology, switching from HFO to natural gas use by industries in Durban, switching to use of process gas as energy source within various platinum and ferro-chromium industries. Reference was also made to international best practices in fuel substitution.

Market interventions for the purpose of reducing atmospheric emissions have not traditionally been used locally. It was therefore necessary to make reference to international experience in this regard.

Following the identification of several possible interventions for significant sources, such options were prioritised base on a qualitative assessment of their environmental benefits, technological viability and socio-economical acceptability, with potentially feasible interventions being included in the quantitative cost-benefit analysis.

4.11 Summary Anthropogenic fuel burning is conjectured to account for over 80% of ambient criteria air pollutant concentrations within South African urban areas. Following the review of methodologies applied abroad, a methodology was selected and tailored to calculate externalities relating to fuel burning activities in South Africa taking into account local circumstances and data availability.

Health effects and costs associated with anthropogenic fuel burning emissions were simulated for several conurbations, accounting for 40% of South Africa‟s population. Emissions were estimated and effects and associated costs quantified for household fuel burning, power generation, industrial and commercial fuel burning and vehicular activity. Emission reduction opportunities were identified and assessed for significance sources.

Study findings are documented in subsequent chapters, including emissions estimated (Chapter 5), direct health effects and costs calculated (Chapter 6) and recommendations regarding priority sources and cost-effective interventions (Chapter 7).

110

5 Emissions and Air Quality Impacts of Fuel-burning

The externalities study, a key component of the thesis, is documented in chapters 4, 5, 6 and 7. In this chapter results are presented from the emissions inventory of fuel-burning sources within the various major conurbations and regions assessed. Emissions estimates are presented for the 2002 base year, with emission projections for 2007 and 2011 provided based on „business as usual‟ assumptions.

5.1 Baseline Emissions due to Fuel-burning Sources Total annual emissions estimated for each conurbation per fuel burning source group for the base case year of 2002 are given in Appendix C, with results discussed in the subsections below. Total emissions per source group across all conurbations are also discussed and the significance of such emissions assessed.

5.1.1 eThekwini and Cape Town Emissions from industrial/commercial/institutional fuel burning are the most significant contributors to PM10 and SO2 emissions in eThekwini and Cape Town. This sector contributes less significantly to NOx releases. Vehicle emissions are the most significant source of NOx, benzene and 1,3-butadiene emissions. Vehicles also contribute ~30% of fine particulate matter and SO2 emissions.

Shipping contributes marginally to NOx and SO2 levels, with emissions more important in terms of their being localised within the harbour area. The contribution of aircraft and rail-related emissions (where available) was found to be small. Aircraft emissions are also expected to be more important as a localised source.

Domestic fuel burning represents a significant source of fine particles. Despite the relatively small emissions from domestic fuel burning, compared with industry, the health influences of domestic fuel burning emissions are enhanced due to the low elevation at which criteria emissions occur, the timing of peak emissions and the proximity of releases to high population areas.

Biomass burning contributes to fine particulate matter emissions, representing a potentially important localised source of episodic emissions, particularly in Cape Town.

5.1.2 Vaal Triangle

Industry and power generation are the most significant sources of total particulate matter and SO2 emissions in the Vaal Triangle. These sectors are estimated to be larger sources of NOx than road transportation. Vehicle emissions are the most significant source in terms of benzene releases, as in other conurbations. Domestic fuel burning is a significant contributor to PM10 and benzene emissions. The significance of this sector is enhanced due to the low level at which emissions occur, the timing of the peak emissions and the proximity of releases to high population areas. As in Mpumalanga and elsewhere biomass burning was estimated to contribute to fine particulate matter, and to represent a potentially important localised source of episodic emissions.

111

5.1.3 Mpumalanga Highveld Power generation is the most significant source sector in Mpumalanga in terms of particulate matter, sulphur dioxide and NOx emissions. The industrial fuel burning sector is primarily important in terms of its contribution to SO2 emissions, but also contributes to PM10 and NOx emissions. Vehicle emissions are the most significant source of benzene emissions from the fuel burning processes quantified. Domestic fuel burning contributes significantly to particulate matter, and benzene emissions. The significance of this sector is enhanced due to the low level at which emissions occur, the timing of the peak emissions and the proximity of releases to high population areas. Biomass burning contributes to fine particulate matter, representing a potentially important localised source of episodic emissions.

5.1.4 Johannesburg The industrial, biomass burning and domestic fuel burning sectors are the largest contributors of particulate matter and SO2 emissions. The domestic fuel burning sector is considered to be the most significant given the low elevation of emissions, the winter-time early morning or evening peaks in emissions and the release of emissions within high human exposure areas. Domestic fuel burning also contributes significantly benzene emissions. Biomass burning represents a potentially important localised source of episodic emissions.

Vehicles are the most significant source of NOx, benzene and 1,3-butadiene emissions. This source also contributes ~30% to total fine particulate matter and SO2 emissions from fuel burning. The significance of vehicle emissions is enhanced by the low elevation at which emissions occur and their proximity to residential areas.

5.1.5 Tshwane and Ekurhuleni The contribution of the industrial and power generation sectors are estimated to be more significant in Tshwane compared to Johannesburg. The industrial and commercial fuel burning sector is also more significant in Ekurhuleni in terms of particulate matter and SO2 emissions when compared to Johannesburg. Domestic fuel burning is still evident as an important source of low elevation particulate matter and SO2 emissions in both metros.

Aircraft emissions at the Johannesburg International Airport (actually situated in Ekurhuleni) contribute marginally to NOx and SO2 emissions. Such emissions are expected to be more important as a localised source.

5.1.6 Emissions across All Conurbations The contributions of the various source groups considered in the study to total fuel-burning emissions estimated across all conurbations are illustrated in Figure 38 for key pollutants. The extent of emissions is not a concise indicator of contributions to ground level air pollution concentrations and health and environmental risks. Such contributions are also a function of the height of emission and the distance between the source and sensitive receptors.

The significance of domestic fuel burning emissions is enhanced due to three factors: (i) the low elevation above ground level of emissions, (ii) the coincidence of peak emissions, typically a factor of 10 greater than if total annual emissions were averaged, with periods of poor atmospheric dispersion (i.e. night-time, winter-time), and (iii) the release of such emissions within high human

112

population areas. These factors lead to exacerbated contributions to both indoor and ambient pollution exposures. The significance of biomass burning is similarly enhanced as a localised source of episodic emissions due the low elevation level of release and the fact that most emissions occur concentrated during a burn season of limited duration.

Figure 38: Contribution of source groups to total fuel-burning emissions across all conurbations

113

The significance of vehicle emissions, in terms of the contribution to air pollutant concentrations and health risks, is similarly enhanced by the low elevation at which emissions occur and the proximity of such releases to high population density areas. Vehicle emissions also tend to peak in the early morning and evenings, at which time atmospheric dispersion potentials are low.

The air quality effect of fuel burning within industrial and power generation sectors in terms of their contributions to air pollutant concentrations and health risks is frequently lower than would be expected given the extent of emissions. This is due to these sources generally being characterised by constant, high level releases with such emissions also likely to be more remote from residential settlements compared to household fuel burning and vehicle emissions.

Taking the above considerations into account, the most significant sources were identified as follows (not ranked):

 Industrial and commercial fuel burning sector - significant source of particles and SO2 in all areas but particularly Cape Town, eThekwini, Vaal Triangle, Ekurhuleni and Mpumalanga. (This source is of moderate significance in Tshwane(13) and of relatively low significance in

Johannesburg.) This sector was also noted to contribute to NOx emissions.

 Vehicle emissions - significant source of NOx, benzene and 1,3-butadiene emissions in all

conurbations. This sector contributes ~30% to total fine particulate matter and SO2 emissions

from fuel burning processes. The contribution of vehicles to NOx and SO2 emissions were however noted to be lower in regions/conurbations with higher industrial and power generation emissions, e.g. Vaal Triangle and Mpumalanga. (Spatial concentrations of vehicle emissions are also less likely to occur in these two regions.)

 Domestic fuel burning - significant source of low level find particulate matter and SO2 emissions. This sector contributes significantly to benzene emissions. The contribution of this sector to fine particulate matter concentrations within Cape Town and eThekwini is

enhanced due to extensive wood burning. SO2 emissions from the sector are higher for inland areas where coal burning is more widespread.

 Electricity generation - significant source of particulate matter, SO2 and NOx emissions in Mpumalanga and Vaal Triangle and to a lesser extent Tshwane. Despite the high level at which emissions are released important contributions to local ground level concentrations are possible during unstable atmospheric conditions.  Biomass burning - significant source of localised, episodic fine particulate matter emissions.

The contributions of shipping, aircraft and railway emissions were shown to be relatively small, although it was recognised that shipping and aircraft emissions could contribute significantly to localised, low level emissions.

13 NOTE: Fuel use data were not available for non-scheduled processes; it was therefore not possible to estimate comprehensively emissions from all industrial, commercial and institutional fuel burning processes. Should this be done, the significance of this sector may be increased.

114

5.2 Forecast Changes in Emissions given “Business as Usual”

5.2.1 Electricity Generation The demand for electricity has steadily increased over at the last thirty years, with, with electricity consumption having increased by a factor of greater than two during this period. During the 1995 to 2003 period, coal consumption by the electricity generation sector increased by an annual average rate of approximately 3%. Given the aim of the Integrated National Electrification Programme of ensuring that all households have access to electricity by 2010/2011 it was assumed that this increase in coal consumption would persist in the short- to medium-term (i.e. 2004 - 2011)(14). Given a 'business as usual' scenario it is therefore anticipated that emissions from the electricity generation sector will increase accordingly, with the likely exception of particulate matter emissions, due to the controls currently in place and proposed stricter emission limits under the new Air Quality Management Act and associated standards.

5.2.2 Industrial, Commercial and Institutional Fuel Burning The iron and steel industry is the sub-sector with by far the greatest coal consumption, consuming ~30% of the coal-generated energy used by the industrial sector. This sub-sector uses large quantities of coke oven gas and coking coal, and is the largest industrial consumer of fuel oil. Other industrial sub-sectors responsible for consuming significant quantities of coal include: the chemical and petrochemical, food and tobacco, pulp and paper, and non-metallurgical sub-sectors (Sectoral fuel use information made available by the Department of Minerals and Energy, March 2004).

Fuel burning by the following industry sectors was identified as contributing significantly to emissions within one or more of the conurbations:  chemical and petrochemical  textile manufacture  pulp and paper  food and tobacco (particularly sugar refining and breweries)  iron and steel and other metallurgical processes (ferro-alloy, precious metal refining, stainless steel manufacture)  non-metallurgical or ceramic processes (brick, cement and refractory manufacture)  commercial and institutional fuel burning.

Changes in the extent of emissions from these sectors are associated with: (i) changes in demand for products or change in energy requirements, (ii) increases in the fuel efficiency of processes, (iii) fuel switching, and (iv) end-of-pipe type controls.

Reference was made to ABSA's projected average annual percentage change in real gross value for industrial sub-sectors of interest for the 2003-2007 period (Table 20) (ABSA, 2002). Such

14 The projected increase in coal-fired electricity generation to 2011 has been confirmed based on subsequently published statistics (Department of Energy, 2010; Eberhard, 2011).

115

projected changes and new Capex projects provided an indication of the potential which exists for changes in emissions during the short- to medium-term.

Table 20: Average annual percentage change in real gross value (ABSA, 2002) and Capex Projects (IDC, 2003) for industrial sub-sectors of interest in this study

Average Annual % Source Change in Real Gross New Capex Projects (IDC, 2003) Group Value (ABSA, 2002) SCI - Butanol & AAA projects (Sasolburg); Sasap; Petrochemical 1.5 Carbo-Tar Purcarb project (Mpumalanga) Chemical 3.1 - 3.8 Sasol's new polymer plastic plant (Sasolburg); Textile manufacture 0.7 Mondi-Kraft expansion (KZN); Sappi Tugela paper Pulp and paper 3.7 mill upgrade (KZN) Food, beverage & tobacco 0.4 - 0.7 Wheat-beer Brewery (E Cape) Mittal Steel (previously Iscor) Newcastle project; Iron and steel 4.9 Mintek Hot Briquette iron plant Non-ferrous metals 4.3 Non-metallurgical (brick & cement) 0.6 Alpha cement kiln modernisation (NW Province)

5.2.3 Residential Fuel Burning Various factors affect the extent of household fuel combustion including: population growth, availability of electricity, household income, degree of urbanisation, and percentage of informal (unserviced) households. Population growth, reductions in household income levels and increase in informal (unserviced) households have been noted to result in increased household fuel burning.

The following trends in key drivers associated with domestic fuel burning were used as the basis to project future trends in household fuel burning emissions:  Population growth rates were projected to increase by 1.6% in the short term 2003 - 2007, but are expected to reduce to a zero growth rate during the first half of 2010 (ABSA, 2002).  The quantity of coal consumed by the merchants and domestic sector has decreased by a factor of 2.3 over the 1989 to 1999 period (Doppegieter et al., 2000), with fuel burning shown to be persistent between 1996 and 2000.  The Integrated National Electrification Programme is on-going. In 2004 it was projected that all houses will be electrified by 2010/2011.

Based on the aforementioned trends it was assumed that residential fuel burning will persist in the short-term (2003 - 2007), but will start to decrease in the medium-term as a result of lower population growth rates and on-going electrification.

According to more recent figures, 25.1% of households (3.4 million households) were not yet electrified at the beginning of 2010 (Department of Energy, 2010), with the backlog of 2.5 to 2.9 million households estimated during 2011 (Eskom, 2011). The price of electricity is also conjectured to have reverse the switch to electricity by some households, even when such households are grid-connected. It is therefore expected that emissions and associated costs associated with household fuel burning during the 2007 to 2011 period may be understated.

116

5.2.4 Vehicle Emissions The main factors affecting vehicle transport energy demand include: economic trends, demographics, fuel accessibility and supply, spatial structure and transport infrastructure (urban sprawl), inter-modal competition, lifestyle norms and regulation.

Indicators of vehicle activity rates are frequently taken to include: extent of fuel sales, total vehicle numbers, the number of vehicles per capita, number of single occupancy vehicles and average trip length. Specific factors influencing the extent of emissions from vehicle use, given fixed vehicle activity rates, include the fuel efficiency of vehicles and fuel composition.

During the 1994 to 2004 period national petrol sales increased by 14% and diesel sales by 50% (Figure 39). During the 1995 to 2001 period, the number of cars in the national parc increased annually on average by 1.6%, with LDVs and HDVs increasing by 1.9% and 0.5% respectively. Based on the vehicle registration data from the National Transportation Information System (NaTIS) data base for the year 2001 and population statistics for 2001, the estimated car ownership rate for South Africa was 129 vehicles per 1000 people. This was slightly above the global average of 120 vehicles per 1000 people(15). Increases in the extent of single occupancy vehicles, increased average trip lengths and increases in the number of cars per capita have been quoted in cities like Cape Town as proof of the growth in vehicle activity rates (Cape Town State of Environment Report, 2003).

12000

10000

8000

6000

4000

(Millions of (Millions Litres) 2000 Liquid Fuel Consumption Fuel Liquid

0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

PETROL DIESEL

Figure 39: Trends in liquid fuel sales for the 1994 to 2004 period, South African Petroleum Industry Association (SAPIA, 2005)

15 The number of passenger vehicles per 1000 people increased to 152 in 2009, and is projected to increase to 159 in 2011, and to 172 by 2014. Annual average growth in petrol sales is projected to be 5.2% during the 2010 to 2014 period (The Economist Automotive Briefing and Forecasts, 2010).

117

Whereas increased vehicle activity rates imply an increase in emissions from this sector, improvements in the fuel efficiency of vehicles, the incorporation of emission controls on new vehicles, and past and proposed changes in fuel composition result in emission reductions. Since the introduction of unleaded petrol, catalytic converter equipped petrol vehicle sales were reported to have steadily increased in number, comprising approximately 50% of new passenger vehicle sales in 2003/4. The newer vehicles at this time were typically of Euro 3 Technology (personal communication, Stuart Rayner, National Association of Automobile Manufacturers of South Africa, April 2003). Changes to fuel composition prior to 2004 included the reduction in the sulphur content of diesel from 0.5% to 0.3%. Provision was made in 2010 for niche diesel grade with a maximum sulphur content of 50 ppm(16). Various blends of bio-diesel up to a blend of 100% bio-diesel has also been permitted.

The main measures recommended for implementation in the Implementation Strategy for the Control of Exhaust Emissions from Road-going Vehicles in South Africa (version 2, 4 March 2003) included the specification of EURO technology for new vehicles and the reduction in the sulphur, lead, benzene and aromatic content of fuels. Should the vehicle and fuel measures have been implemented, substantial changes in the nature and extent of vehicle emissions would have resulted. At the time of the Dirty Fuels Study being undertaken there was uncertainty as to whether the proposed measures would be implemented. It was therefore decided to exclude these measures from „business as usual‟ projections, but to consider them as likely interventions in the cost-benefit analysis undertaken for potential interventions.

Vehicle emission legislation was enacted which required that new passenger car and LCVs manufactured from January 2008 onwards, and new MCVs and HCVs manufactured from January 2010 onwards, meet Euro 2 standards. Fuel standards and specifications which are compatible with Euro 2 standards were introduced and continue to be the fuel standards in force in 2011(17). The basis for the vehicle emission projections undertaken during the Dirty Fuel Study for the period to 2011 therefore remains valid, with the emission reductions to be realised through the implementation of further fuel and vehicle standards representative of future opportunities.

The „business as usual‟ projections for vehicle emissions were based on the following factors:  Significant spatial variations in petrol and diesel sales trends were noted to occur during the 1995 to 2002 period. The following annual average changes were noted for use in the current study: • Cape Town - 1.8% increase in petrol sales; 5.1% increase in diesel consumption • Johannesburg - 1.4% increase in petrol sales; 6.5% increase in diesel consumption

16 Low sulphur diesel is becoming increasingly available across South Africa. The move to low sulphur diesel is imperative for the introduction of new generation clean diesel technology (Rayner, 2010). 17 The South African Department of Energy issued a draft position paper in March 2011 advocating that South Africa migrate directly from the current fuel specifications and standards (CF1) which are compatible with Euro 2 emissions standard to CF2 which is equivalent to Euro 5 emissions standard on targeted key parameters. This would involve further reducing the levels of sulphur in both petrol and diesel as well as the reduction of benzene and aromatic levels in petrol to levels equivalent to the Euro 5 emissions standards. It is therefore recommended that the sulphur content in both petrol and diesel be reduced from 500 ppm to 10 ppm; benzene from 5% to 1%, and aromatics from 50% to 35% by 2017 (Government Gazette, No. 34089, No. 204, 8 March 2011).

118

• eThekwini - 0.5% increase in petrol sales; 5.7% increase in diesel consumption • Vaal Triangle - 0.8% reduction in petrol sales; 1.3% increase in diesel consumption • Tshwane - 1.2% increase in petrol sales; 4.6% increase in diesel consumption • Mpumalanga Highveld region - 1.9% reduction in petrol sales; 1.8% increase in diesel consumption • Ekurhuleni - 0.1% decrease in petrol sales; 3.2% increase in diesel consumption  Increase in the number of vehicles fitted with catalytic converters. During the 1990 to 2002 period an annual average increase in the percentage of new cars with catalytic converters was noted to be 3.9%. The assumption was made that this trend would continue in the short- to medium-term (i.e. up until 2011). Based on this rate, 67% of new petrol-driven cars will be fitted with catalytic converters by 2007, and 82.4% by 2011.

Estimated changes in vehicle emissions in the short- and medium terms (2007 and 2011 respectively) due to the 'business as usual' scenario are given in Table 21. Lead emission projections, although projected as part of the study, are not given due to leaded fuels having been phased out 2006 following completion of the emissions inventories.

Table 21: Estimated changes in vehicle emissions given changes in vehicle activity and vehicle technology

2007 Emission Scenario 2011 Emission Scenario Pollutant % reduction compared % increase compared % reduction compared % increase compared to 2002 emissions to 2002 emissions to 2002 emissions to 2002 emissions

NOX 14.1 16.6

SO2 5.2 2.8 1,3-butadiene 5.0 11.3 Benzene 6.6(a) 11.2(a) Particles 26.8 26.8 (a) There is the potential for an increase in the extent of benzene emissions due to future changes in the composition of unleaded petrol. The predicted change in benzene emissions should be confirmed with further study once more information is available on the future specifications for unleaded petrol composition.

Despite the projected increase in the number of vehicles fitted with catalytic converters, emissions of most compounds are estimate to increase due to increased vehicle activity rates (projected on the basis of fuel sales). The margin of error of vehicle emission estimates would be substantially improved should actual vehicle kilometres travelled be linked to each vehicle class (engine capacity, control technology type and age). Such links could however not be established given the information available and the scope of the study.

5.2.5 Synopsis of Future Trends in Fuel-burning Emissions The extent of emissions from the majority of source groups were not anticipated to change substantially in the short-term (i.e. 2004 - 2007). This assumption has been confirmed by the persistence of poor air quality in areas impacted by fuel burning emissions (see Section 2.2). Projections are less certain for the medium-term, i.e. for the period 2007 - 2011.

Emissions from the power generation sector, the residential fuel burning sector, the industrial and commercial fuel burning sector and biomass burning were projected to either remain relatively constant or to increase by typically 0.5% to 3% per annum during the short- to medium-terms. The

119

only notable change foreseen for the industrial sector is the large emission reductions from Sasol‟s Sasolburg Plant due to the switching from coal to natural gas. This resulted in a reduction in the significance of the chemical and petrochemical sector to total fuel burning emissions within the Vaal Triangle.

Vehicle emissions were projected to increase, assuming an absence of future controls, with various pollutants predicted to increase by up to 27% in 2007 and by up to 44% in 2011. This overall increase is projected despite reductions due to the implementation of cleaner fuel and the gradual growth of Euro 3 compliant vehicles.

The sector growth and associated emission increase projections were undertaken prior to the economic downturn of 2008. The annual rate of growth in GDP was in the range of 4.5% to 5.5% during the 2004 to 2007 period, with a negative growth rate occurring during the downturn. Economic recovery is reflected in a return to positive growth rates, with the 2011 growth in GDP being over 3%. Given the economic downturn, emission increase projections may be overstated for some sectors, primarily industry, for the 2007 to 2011 period. Whereas the growth in emission projections for the household sector, was noted based on retrospective analysis to have been understated due to short-falls in the extent of household electrification assumed, as discussed in Section 5.2.3.

5.3 Air Pollutant Concentrations due to Fuel-burning Emissions Air pollutant concentrations were simulated for various averaging periods to facilitate comparisons with monitoring results, in addition to supporting the application of the various dose-response functions selected for the health risk estimation. Isopleth plots for the various study regions, illustrating predicted spatial variations in highest daily and annual average sulphur dioxide and fine particulate matter (PM10) concentrations, and highest hourly average nitrogen dioxide concentrations are presented in Appendix D.

It is important to note that model predictions illustrated in Appendix D and discussed in subsequent subsections included contributions by residential, industrial, power generation and vehicle-related fuel combustion emissions. The model predictions presented in this thesis exclude air pollutant concentrations due to biomass burning related emissions, which found to be unrepresentative and therefore excluded from the cumulative air pollution plots. Due to the episodic nature, unpredictable duration and unknown mass combusted, biomass burning could not be adequately modelled given the source information available and using the dispersion modelling methodology applied.

5.3.1 Cape Town Significant spatial variations in airborne particulate matter concentrations were predicted to occur over the Cape Region. Peak concentrations occur over the residential wood-burning areas of Khayelitsha, and over the suburbs Table View, Bellville South and Strand that are impacted by industrial emissions, including a petrochemical refinery in the case of Table View. PM10 measurements from the seven continuous air quality monitoring stations operated by the City of Cape Town during 2002 confirm the large spatial variations in average concentrations, and the peak concentrations over Khayelitsha, Table View and Bellville South (Scorgie and Watson, 2004). No monitoring was undertaken at Strand.

120

Sources contributing significantly to ambient particulate matter concentrations in the Cape study region included primarily domestic wood burning and coal-fired boiler operations. Commercial wood burning, including wood waste burning and small-scale operations, such as wood-fired pizza ovens, also contribute significantly to ground level fine particulate matter concentrations in some areas, e.g. Strand.

The largest contributions to sulphur dioxide concentrations were HFO-fired boilers, coal-fired boilers, a pulp and paper mill and the petrochemical refinery. Maximum sulphur dioxide concentrations were predicted to coincide with industrial areas, including Bellville South, Parrow, Table View and Goodwood.

Wood waste burning, vehicle tailpipe emissions, boiler operations, the pulp and paper mill and petrochemical refinery represented the most significant contributors to nitrogen dioxide concentrations. Peak hourly average NO2 concentrations were predicted to occur in the vicinity of the N1 and N2 highways and the CBD, illustrating the important contribution made by traffic emissions. The predicted peaks reflect monitoring results from the City of Cape Town‟s Bothasig and City Hall monitoring stations, which are located in proximity to the N1 highway and within the

CBD respectively. During 2003, highest hourly average NO2 concentrations of approximately 440 µg m-3 were recorded at these sites, with 28 exceedances of the South African National Standard (SANS) hourly air quality limit of 200 µg m-3 being recorded at the City Hall monitoring station (Scorgie and Watson, 2004).

5.3.2 eThekwini Elevated sulphur dioxide concentrations, in excess of SANS air quality limits, were predicted to occur within the South Durban Basin. Sources contributing to such concentrations include two petrochemical refineries, a paper mill, traffic emissions along the busy south coast freeway and various coal-fired boiler operations.

Maximum fine particulate matter concentrations were predicted to occur within the South Durban Basin and at Chatsworth, Umlazi and Verulam. The main sources of particulate matter concentrations in the broader region were predicted to include: domestic coal and wood burning, coal-fired boilers, diesel vehicles, the Sappi Saiccor (pulp and paper) plant and Huletts (sugar) mill.

The highest NO2 concentrations were predicted to coincide with the south coast freeway, the CBD and with industrial areas situated within the South Durban Basin.

High sulphur dioxide and particulate matter concentrations measured in the South Durban Basin and elevated nitrogen dioxide concentrations recorded at monitoring stations within the CBD serve to confirm the model predictions (eThekwini Health, 2007).

5.3.3 Johannesburg and Ekurhuleni High fine particulate matter concentrations are predicted, and have been measured to occur, across the Highveld with exceedances of SANS air quality limits at all sites for which PM10 monitoring data are available (Scorgie et al., 2003b; Scorgie et al., 2004e; Annegarn et al., 2007).

Significantly high concentrations of fine particulate matter occur within fuel burning residential areas, specifically coal and wood burning areas, with maximum daily PM10 concentrations of ~500

121

µg m-3 being both predicted and measured. Measurements indicated that health limits are exceeded frequently in such areas, ranging from 20% to 40% of days (Scorgie et al., 2003b; Scorgie et al., 2004e; Annegarn et al., 2007).

Residential coal burning and coal fired boilers were noted to be the most significant anthropogenic fuel-burning related sources of airborne particulate matter in the Johannesburg and Ekurhuleni metropolitan areas. Coal boiler operations included the operations undertaken at the Zwartkoppies pump station (south of Johannesburg CBD), Impala Refinery (Springs) and NCP (Chloorkop).

The highest sulphur dioxide concentrations were predicted to be due to emissions from residential coal burning, petrol-driven vehicles and various coal and HFO boiler operations. Sulphur dioxide concentrations within residential coal burning areas are likely to exceed short-term (10-minute, hourly) air quality limits but such exceedances are relatively infrequent.

Peak NO2 concentrations were predicted to occur in the region downwind of the Johannesburg CBD and in the vicinity of the busy N3, M1 and M2 highways. Elevated nitrogen dioxide was also simulated to occur downwind of residential coal burning areas including Soweto, Tembisa and Alexandra.

5.3.4 Tshwane The most significant fuel combustion related sources in the region are the two local power stations (Pretoria West and Rooiwal), various brickwork operations and domestic fuel burning. The Rooiwal Power Station was estimated to be one of the largest point sources. Most of the brickworks are located in proximity to each other within the Moot Valley located west of the Pretoria CBD.

Although estimates of the air quality effect of coal-fired boilers were relatively low it was noted that the emissions inventory compiled for the region did not include the many small-scale boiler operations in the region.

Peak NO2 concentrations were predicted to occur downwind of the Pretoria CBD and are primarily related to vehicle exhaust emissions.

5.3.5 Mpumalanga Highveld Region The Mpumalanga Highveld region was predicted to experience the most widespread enhanced concentrations of sulphur dioxide, occurring primarily as a result of emissions from the various large coal-fired power stations situated in the region. Peak ground level concentrations are predicted and measured to typically occur during the late morning and early afternoon, due to atmospheric thermal instability and convective mixing of emissions tall stacks.

Other fuel-combustion related sources predicted to contribute significantly to SO2 concentrations over the region include large-scale industrial operations such as the Sasol Secunda Complex and Highveld Steel and Vanadium (near Witbank) and the combined effect of various brickworks. Smouldering coal discards may contribute significantly to ground level sulphur dioxide concentrations in certain areas but could not be accurately identified, quantified and modelled in the study. The contribution of residential fuel burning to ambient sulphur dioxide concentrations was noted to be relatively small.

122

Potentially significant sources of fuel combustion related particulate matter concentrations in the Mpumalanga study region include residential coal burning, brickwork operations and coal combustion by industries such as Highveld Steel and Vanadium. The coal-fired power stations located in the region also contribute to airborne concentrations of secondary pollutants as a result of the conversion of their gaseous emissions to sulphate and nitrate aerosols.

Lower NO2 concentrations were predicted to occur over much of the Mpumalanga Highveld, compared to within and downwind of the Tshwane-Johannesburg-Ekurhuleni conurbation.

5.3.6 Vaal Triangle

Predicted maximum daily and annual average PM10 and SO2 concentrations were comparable to measured concentrations within residential fuel burning areas and heavy industrial areas (excluding Sasolburg), demonstrating the important contribution of fuel-burning emissions to ambient air pollutant concentrations in these areas(18). Sulphur dioxide concentrations in the vicinity of Sasolburg were under-predicted due to the contributions of non-fuel burning related industrial emissions not having been accounted for in the simulations.

The highest fine particulate matter concentrations on the Highveld were simulated to occur within the Vaal Triangle region, such concentrations coinciding with residential coal burning and heavy industrial areas. Elevated particulate matter have also been measured to occur in proximity to large-scale mining operations in the region. Maximum daily PM10 levels were found to range between 200 and 500 µg m-3 within heavy industrial, intensive mining and residential fuel burning areas, with annual average concentrations of 80 to 100 µg m-3 evident (Appendix D, Scorgie,

2005a). In non-fuel burning residential areas average annual PM10 concentrations were typically -3 predicted and measured to range between 60 and 70 µg m with maximum daily PM10 -3 -3 concentrations ranging from 150 to 220 µg m . The SANS daily PM10 limit of 75 µg m was predicted to be exceeded over the entire Vaal Triangle, with maximum frequencies of exceedance being in the order of ~80% of days.

Maximum hourly average sulphur dioxide concentrations of between 1200 and 1500 µg m-3 have been recorded to occur in heavy industrial areas in Vanderbijlpark and Sasolburg and within neighbouring residential areas in Sasolburg. Maximum daily average concentrations in the range of 180 to 850 µg m-3 and annual average concentrations of between 40 and 80 µg m-3 have been observed in these areas.

Household fuel combustion represents a large contributor to low level particulate matter concentrations. The coke oven operation at Mittal Steel (previously Iscor) Vanderbijlpark Works represents one of the most significant industrial fuel related operations affecting the region. The highest ground level sulphur dioxide concentrations were associated with Lethabo power station emissions, with Sasol power stations and HFO combustion by Natref also contributing to ambient

18 Monitoring data available for comparison with model predictions included recent ambient monitoring conducted by Sasol, New Vaal Colliery and Iscor Vanderbijlpark Works in addition to monitoring undertaken by Mintek in Vereeniging, Sasolburg and Vanderbijlpark during the 1994-5 period (Scorgie, 2005a).

123

SO2 levels. Vehicle emissions were predicted to result in relatively slight ambient pollutant concentrations in this region.

5.4 Source Contributions to Air Pollutant Concentrations within the Vaal Triangle Source contributions to ambient air pollutant concentrations are difficult to illustrate since such contributions vary by pollutant and averaging period, in addition to varying spatially across each conurbation.

To provide an example of this, source contributions to annual average PM10 concentrations, predicted for various locations within the Vaal Triangle (Figure 40), are illustrated in Figure 41. Selected locations include sites within the CBDs of Vanderbijlpark, Vereeniging, Sasolburg and Meyerton and sites within the centre of the residential fuel burning areas of Sebokeng and Sharpville.

Figure 40: Location of sites (red squares) at which source contributions to predicted total annual

PM10 concentrations are illustrated in Figure 41

124

125

The contributions of residential fuel burning emissions to ambient annual PM10 concentrations were predicted to be distinctly different at Sebokeng and Sharpville. Whereas household fuel burning was estimated to be responsible for 84% of the annual PM10 concentrations at Sebokeng, the contribution at Sharpville was only in the order of 45%, due to the location of Sharpville in proximity and downwind of the Vanderbijlpark industrial areas.

Emissions from industrial, commercial and mining sources were predicted to be the largest contributor to annual average PM10 concentrations at most of the sites. Excluding Sebokeng, the contribution of these sources ranged from 50% at Sharpville to 88% in Meyerton. Despite the significant emissions from power generation, the contribution of such emissions to ground level ambient annual PM10 concentrations at the points noted was estimated to be below 0.2%.

Vehicle exhaust contributions were predicted to be in the range of 0.7% to 14%. The contribution of vehicle exhaust emissions to ambient PM10 concentrations should however be cautiously interpreted. Given that vehicle emissions were estimated on the basis of magisterial fuel sales data, with such emissions having been spatially allocated on the basis of road densities, the contribution of vehicle emissions is more uncertain. To gain an accurate representation of spatial variations in vehicle emission contributions one would need to use accurate, spatially and temporally resolved vehicle activity flow data in the emission estimation and dispersion simulations. Such detailed data are not routinely collected by local authorities, and so these indirect estimates represent the best available information.

5.5 Synopsis of Source Significance Based on their contribution to ambient air pollutant concentrations the following sources were flagged as being of significance:  residential coal combustion (particularly Johannesburg, Vaal Triangle)  residential wood combustion (Cape Town)  coal-fired boilers (Cape Town, eThekwini, Johannesburg, Ekurhuleni)  HFO-fired boilers (Cape Town)  brickwork operations (Tshwane and Mpumalanga)  power generation (Mpumalanga, Tshwane, Vaal Triangle)  vehicle emissions (Cape Town, eThekwini, Johannesburg & Ekurhuleni, Tshwane).

Individual operations which were found to contribute significantly to ambient pollutant concentrations included: Sappi Fine Papers and Caltex Refinery (Cape Town), Sappi Saiccor and Huglett sugar (eThekwini), Impala Platinum Refinery (Springs), NCP and Zwartkoppies pump station (Ekurhuleni), Sasol Secunda and Highveld Steel & Vanadium (Mpumalanga), and the coke oven operations at Mittal Steel (previously Iscor) Vanderbijlpark (Vaal Triangle).

The potential that exists for the occurrence of health risks due to the above mentioned sources depends on the toxicity of the pollutants release and on the extent of exposure to predicted concentrations. Predicted health risks associated with source groups, calculated on the basis of dose-response functions and unit risk factors, are presented in Chapter 6.

126

6 Health Effects and Costs due to Fuel-burning Sources

The externalities study, a key component of the thesis, is documented in chapters 4, 5, 6 and 7. This chapter addresses the quantification of potential health effects due to inhalation exposures, paraffin-poisoning and burns related to anthropogenic fuel use practices, and provides an estimation of direct health spending associated with such effects.

6.1 Study Limitations Results for inhalation-related health effects due to anthropogenic fuel burning are presented in Sections 6.2, and the monetary costs associated with such effects are presented in Section 6.3. In the interpretation and use of the study findings, the following limitations of scope should be taken into consideration:

 Only inhalation exposures due to fuel burning related atmospheric emissions were quantified. Emissions and associated effects due to other sources of emissions were not quantified. Sources of emission which are not accounted for include fugitive dust emissions, industrial process emissions and evaporative losses. Household fuel burning emissions and vehicle emissions are largely accounted for, whereas only a portion of the emissions from industrial operations are included. This should be noted in interpreting the source contribution information provided.  Exposure pathways other than inhalation, viz. ingestion and dermal contact, are not included. It is however noted that inhalation represents the main exposure pathway for the pollutants included in the study.  Exposures to ozone concentrations (and other photochemical products) were not quantified as

part of the study. Ozone precursors include NOx and VOCs. Since these pollutants are released from all fuel burning sources, it is evident that the influence of such sources is likely to be greater than is estimated in the study. The underestimation is greatest for significant

sources of NOx or VOC emissions (e.g. vehicle emissions).  Given the methodology employed in the quantification of health risks, viz. application of dose-response functions developed based on epidemiological studies, it was not possible to quantify exposures to indoor air pollutant concentrations. This is expected to have resulted in an under-prediction of the health effects associated with domestic fuel burning emissions.

6.2 Inhalation Health Effects due to Exposures to Fuel Burning Emissions In the quantification of health effects occurring as a result of inhalation exposures to fuel combustion emissions, predicted air pollutant concentrations were overlaid over spatial population data from the 2001 census (Statistics South Africa, Census 2001). The census data make it possible to distinguish between various age groups, with population statistics given in five-year age intervals. A synopsis of the total population figures for each of the conurbations of interest is given in Table 22. The study areas comprise a total of ~18.7 million people, comprising almost 40% of the country‟s total population.

For the purpose of the study, it was necessary to assume that children were in the <5 year age group, with the potentially economically active population assumed to be within the 20 to 65 year age group. Adults were defined as persons over 20 years of age.

127

Table 22: Summary of total persons within each population subsector of interest in the current study (Source: Census 2001, Statistics South Africa)

Population Johannesburg Vaal Mpumalanga Cape Town eThekwini Tshwane Total Sector & Ekurhuleni Triangle Highveld Total 4 097 161 3 690 705 6 255 616 1 780 732 1 574 266 1 272 813 18 671 292 Population Children 369 165 332 576 1 039 198 263 447 289 264 256 154 2 549 804 (<5 years) 20 – 65 2 392 786 2 106 954 3 816 116 1 058 418 871 856 655 980 10 902 110 Adult 2 579 171 2 258 781 4 033 038 1 141 359 920 050 703 096 11 635 496 (>20 years) <65 years 3 910 776 3 538 878 216 921 82 942 48 194 47 116 7 844 827 >65 years 186 385 151 827 3 816 116 1 058 418 871 856 655 980 6 740 582

The predicted air pollutant concentration was taken to be equivalent to the dose, i.e. it was assumed that pollutant concentrations predicted for a particular location were being inhaled by the persons residing at that location. In the calculation of cancer risk it was conservatively assumed that persons residing in areas for which annual concentrations of carcinogens were predicted were exposed to such concentrations for 24 hours per day over a 70-year lifetime.

Total respiratory hospital admissions, premature mortalities, excess cancer cases and restricted activity days (RADs) due to exposures to fuel combustion related emissions during the base year (2002) are summarised in Table 23 for each region.

Total respiratory hospital admissions across all conurbations due to predicted fuel combustion related exposures were calculated to be in the order of 118 900, representing 0.64% of the population. Cardiovascular hospital admissions of about 860 per annum were estimated (0.005% of population). Exposure to fuel combustion related pollutant concentrations was found to be associated with approximately 300 premature deaths, with 0.002% of the population affected. Incidence of chronic bronchitis and cancer cases were estimated to be 110 615 and 230 respectively (i.e. 0.59% and 0.001% of population affected). A total of approximately 795 000 restricted activity days was estimated, representing 67.2 days per potentially economically active person (i.e. defined as persons 20 to 65 years of age).

Source contributions to respiratory hospital admissions, daily mortality and excess leukaemia cases predicted due to fuel burning emission exposures during the base case year (2002) are illustrated in Figure 42 to Figure 47, for each of the study regions. Health risk estimates are discussed per conurbation in the subsequent subsections.

In assessing the contributions of sources, it is useful to make reference to the "all respiratory" health endpoint, given as resulting in increased hospitalisations and related health care visits. This health endpoint takes into account synergistic respiratory effects due to particulate matter, nitrogen and sulphur dioxide exposures. Reference is also made to premature deaths associated with PM10 and SO2 exposures, and excess cancer cases due to chronic exposures to 1,3-butadiene and benzene concentrations.

128

Table 23: Estimated health effects, given as number of cases or incidences per annum associated with human exposures to fuel burning emissions predicted for the base year 2002(a).

Johannesburg & Mpumalanga Health Endpoint Cape Town eThekwini Tshwane Vaal Triangle All Conurbations Ekurhuleni Highveld Respiratory hospital admissions (due to PM10, SO2 and NO2 29 482 27 072 34 021 10 205 9 440 8 685 118 905 exposures)

Cardiovascular hospital admissions 235 201 262 57 71 35 861 (due to PM10 exposures)

Premature mortality (due to PM 10 91 80 72 19 20 17 297 and SO2 exposures) Chronic bronchitis (due to PM10 28 807 18 793 38 550 8 568 9 458 6 440 110 615 exposures) Restricted activity days (RAD, due 217 563 189 118 238 326 56 064 62 547 31 543 795 161 to PM10 exposures)

Minor restricted activity days 9 320 431 7 570 322 12 396 320 5 663 333 6 128 743 32 135 642 73 214 792 (MRAD, due to SO2 exposures)

Leukemia cases (due to 1,3- 27 44.2 67.4 71.9 9.1 6.4 226 butadiene and benzene exposures)

129

a 81 Dometic wood burning 80

3 HFO-fired boilers 4

5 Other sources 4

4 Coal-fired boilers 3

2 Diesel vehicles 3

3 Other fuel-fired boilers 2

<1 Petrol vehicles 2 Daily Mortality (Premature Deaths) 2 Domestic coal burning 2 Respiratory Hospital Admissions

0 10 20 30 40 50 60 70 80 90 Percentage Contribution

Petrol vehicles 56

Diesel vehicles 35

Dometic wood burning 9 Leukaemia Cases

0 10 20 30 40 50 60 Percentage Contribution

Figure 42: Source contributions as a percentage of overall health effects predicted due to exposures to fuel burning emissions within Cape Town, including (a) respiratory hospital admissions and premature mortality, and (b) excess leukaemia cases, due to chronic exposures to 1,3-butadiene and benzene concentrations (Scorgie et al., 2004e).

130

a 67 Dometic wood burning <1

11 Diesel vehicles 12

12 Domestic coal burning 10

1 Petrol vehicles 6

3 Other sources 3

3 Huletts sugar mill 2

2 Coal-fired boilers 2 Daily Mortality (Premature Deaths) 1 Sappi Saiccor <1 Respiratory Hospital Admissions

0 10 20 30 40 50 60 70 80 Percentage Contribution

Petrol vehicles 58

Diesel vehicles 40

Dometic wood burning 2 Leukaemia Cases

0 10 20 30 40 50 60 70 Percentage Contribution

Figure 43: Source contributions as a percentage of overall health effects predicted due to exposures to fuel burning emissions within eThekwini, including (a) respiratory hospital admissions and premature mortality, and (b) excess leukaemia cases, due to chronic exposures to 1,3-butadiene and benzene concentrations (Scorgie et al., 2004e).

131

a 61 Domestic coal burning 56

23 Dometic wood burning 21

7 Coal-fired boilers 7

4 Diesel vehicles 7

1 Petrol vehicles 6

Daily Mortality (Premature Deaths) 4 Other sources 3 Respiratory Hospital Admissions

0 10 20 30 40 50 60 70 Percentage Contribution

Petrol vehicles 56

Diesel vehicles 39

Dometic wood burning 3

Domestic coal burning 2 Leukaemia Cases

0 10 20 30 40 50 60 Percentage Contribution

Figure 44: Source contributions as a percentage of overall health effects predicted due to exposures to fuel burning emissions within Johannesburg and Ekurhuleni, including (a) respiratory hospital admissions and premature mortality, and (b) excess leukaemia cases, due to chronic exposures to 1,3-butadiene and benzene concentrations (Scorgie et al., 2004e).

132

a 19 Diesel vehicles 23

27 Other sources 20

25 Domestic coal burning 20

2 Petrol vehicles 18

15 Electricity generation 11

6 Brickworks 4

Daily Mortality (Premature Deaths) 6 Dometic wood burning 4 Respiratory Hospital Admissions

0 5 10 15 20 25 30 Percentage Contribution

Petrol vehicles 57.9

Diesel vehicles 41.7

Dometic wood burning 0.2

Leukaemia Cases Domestic coal burning 0.2

0 10 20 30 40 50 60 70 Percentage Contribution

Figure 45: Source contributions as a percentage of overall health effects predicted due to exposures to fuel burning emissions within Tshwane, including (a) respiratory hospital admissions and premature mortality, and (b) excess leukaemia cases, due to chronic exposures to 1,3-butadiene and benzene concentrations (Scorgie et al., 2004e).

133

a 57 Domestic coal burning 57

21 Dometic wood burning 20

8 Electricity generation 9

7 Other sources 6

Iscor Vanderbijlpark 4 Works (Mittal) 5

2 Coal-fired boilers 2

Daily Mortality (Premature Deaths) 1 Sasol 1 Respiratory Hospital Admissions

0 10 20 30 40 50 60 Percentage Contribution

b Dometic wood burning 53.7

Domestic coal burning 36.6

Petrol vehicles 5.6

Diesel vehicles 3.9

Other sources 0.1

Leukaemia Cases Electricity generation 0.1

0 10 20 30 40 50 60 Percentage Contribution

Figure 46: Source contributions as a percentage of overall health effects predicted due to exposures to fuel burning emissions within the Vaal Triangle, including (a) respiratory hospital admissions and premature mortality, and (b) excess leukaemia cases, due to chronic exposures to 1,3-butadiene and benzene concentrations (Scorgie et al., 2004e).

134

a 51 Electricity generation 51

12 Sasol 17

15 Domestic coal burning 12

Highveld Steel & 9 Vanadium 7

8 Dometic wood burning 7

Other sources (includes 4 vehicles) 4

Daily Mortality (Premature Deaths) 1 Coal-fired boilers 2 Respiratory Hospital Admissions

0 10 20 30 40 50 60 Percentage Contribution

b Dometic wood burning 82

Domestic coal burning 7.7

Petrol vehicles 5.3

Diesel vehicles 4.2

Electricity generation 0.6 Leukaemia Cases

Sasol 0.2

0 10 20 30 40 50 60 70 80 90 Percentage Contribution

Figure 47: Source contributions as a percentage of overall health effects predicted due to exposures to fuel burning emissions within the Mpumalanga Highveld study area, including (a) respiratory hospital admissions and premature mortality, and (b) excess leukaemia cases, due to chronic exposures to 1,3-butadiene and benzene concentrations (Scorgie et al., 2004e).

135

6.2.1 Cape Town In Cape Town, 80% of the estimated fuel-combustion related respiratory ailments resulting in increased hospitalisation and health care visits were predicted to be due to residential wood burning, with domestic coal burning only responsible for 2% (Figure 42). Boiler operations represented the second largest source of respiratory hospital admissions, accounting for 9% of incidence. Vehicle exhaust emissions were estimated to be responsible for most of the remaining cases (2% due to petrol vehicles; 3% due to diesel vehicles).

Industrial fuel combustion related emissions from the Sappi, Caltex and Athlone Power Station operations were calculated to be responsible for 0.2%, 0.5% and 0.4% respectively of the total respiratory hospital admissions. Source contributions to premature mortality was similar to those for respiratory hospital admissions, with domestic fuel burning, boiler operations and vehicle emissions responsible for 83%, 10% and 2% of cases respectively.

Fuel combustion was predicted to be responsible for 27 additional cases of leukaemia, reflecting a cancer risk of about 1 in 150 000. Cancer risks quantified during the study were primarily associated with exposures to vehicle emissions, with residential wood burning responsible for much of the remaining risk.

6.2.2 eThekwini In this conurbation 75% of the estimated fuel-combustion related respiratory hospital admissions was predicted to be due to residential fuel burning (65% due to wood burning and 10% due to coal burning) (Figure 43). Vehicle emissions were estimated to be responsible for 18% of incidence (6% due to petrol vehicles; 12% due to diesel vehicles).

There are few Heavy Fuel Oil (HFO) fired boilers in eThekwini due to a programme initiated to reduce the extent of "dirty fuels" use in the city, particularly in the Durban South industrial area. Whereas HFO-fired boilers was estimated accounted for only 0.1% of all respiratory cases, coal- fired boiler emissions were estimated to be responsible for 1.9% of such cases. The most significant single point source was predicted to be coal burning at the Hulett sugar mill (~2% of respiratory cases) and Sappi Saiccor pulp and paper plant (0.2% of cases).

Fuel combustion was predicted to be responsible for 44 additional cases of leukaemia, reflecting a cancer risk of approximately 1 in 83 000. The leukaemia risks estimated were due primarily to vehicle emissions (58% due to petrol-driven vehicles, 40% due to diesel vehicles), with domestic wood burning estimated to be responsible for 2% of risks.

6.2.3 Johannesburg and Ekurhuleni Residential fuel burning represented the most significant source of respiratory hospital admissions and premature mortalities, estimated to account for 77% and 84% of such cases respectively (56% and 61% due to coal burning, remainder due to wood burning) (Figure 44). Vehicle emissions were associated with 12% of the risks of respiratory hospitalisations (7% due to diesel vehicles).

Coal fired boilers combined were estimated to be responsible for 7% of all respiratory hospitalisations and premature mortality, with the coal boiler operations at NCP, Zwartkoppies pump station and Sappi Enstra were estimated to account for 1%, 0.6% and 0.5% of respiratory

136

hospital admissions respectively. Kelvin Power Station was calculated to be responsible for about 0.7% of cases.

Domestic fuel burning was estimated to be responsible for 3.6 excess leukaemia cases, with vehicle related emissions predicted to account for a further 64 cases.

6.2.4 Tshwane Vehicle emissions were estimated to account for ~40% of estimated fuel-combustion related hospitalisations due to all respiratory conditions (petrol 18% and diesel 23% of cases) (Figure 45). Domestic fuel burning accounted for 24% of cases, primarily due to coal combustion (20%). Power generation (primarily the Pretoria West Power Station) was estimated to be responsible for 11% of all respiratory hospital cases. Brickworks combined accounted for 4% of the respiratory hospitalisation risk and 6% of the premature mortality risk. Other sources (including furnaces) accounted for 20% of hospital admission cases.

Fuel burning emissions were predicted to be responsible for 72 excess leukaemia cases, with vehicle related emissions predicted to account for much of this risk.

6.2.5 Vaal Triangle Fuel combustion related emissions were estimated to account for ~9 400 cases of respiratory hospitalisations. Approximately 77% of such cases were predicted to be due to domestic fuel burning; 57% due to coal and 20% due to wood burning (Figure 46). The contribution of vehicles was small (0.7% of cases) compared to the conurbations considered previously. The largest point sources included coal-fired boilers combined (2% of cases), Mittal Steel Vanderbijlpark Works (coke oven plant) (~5% of cases), Sasol Sasolburg (steam stations) (~1% of cases) and Lethabo Power Station (~9% of cases).

Domestic fuel burning was estimated to be responsible for 90% of the excess leukaemia cases, with vehicle related emissions predicted to account for a further 9%.

6.2.6 Mpumalanga Highveld A relatively small number of "all respiratory" cases was predicted to occur within the Mpumalanga Highveld study area (~8 700 cases) compared to the other conurbations. Of the cases predicted, the combined effect of the seven operational coal-fired (Eskom) power stations was estimated to be responsible for 51% of such cases (Figure 47). Domestic fuel burning represented the second largest source of respiratory ailments, with coal burning accounting for 12% of cases and wood for 7% of cases (19% combined). Emissions from the steam generating power stations at the Sasol Secunda plant were predicted to be responsible for 17% of cases. Other significant sources included: Highveld Steel & Vanadium (7%) and coal fired boiler operations (2%).

The risk of leukaemia due to exposure to fuel combustion related emissions was predicted to be due primarily to domestic fuel burning exposures (~90% of cases), with vehicle emissions representing the second largest source (~10%).

137

6.2.7 Synopsis of Health Effects across Conurbations Residential fuel burning was estimated to result in the largest fraction of estimated non- carcinogenic health risks across all conurbations (Figure 48). This source accounted for approximately 70% of all estimated respiratory hospital admissions (RHA) and about 75% of all premature mortalities predicted.

Vehicle emissions were associated with 12% and 6% of the estimated RHA and daily mortality cases respectively. (The lower contribution of vehicles to daily mortality is primarily due to the absence of a dose-response function linking NO2 exposures to mortality.) Electricity generation is predicted to account for 6% of the RHA and 5% of the premature deaths respectively. Coal-fired boiler operations were the most significant industrial source grouping, estimated to account for 4% of the RHA and mortality cases.

The most significant individual industrial point sources associated with non-carcinogenic health risks were Highveld Steel & Vanadium (Mpumalanga), Mittal Steel Vanderbijlpark Works (Vaal Triangle), Sasol Secunda (Mpumalanga) and Hulett (eThekwini).

Vehicle emissions are responsible for approximately 95% of the leukaemia risks estimated to be associated with exposures to fuel burning related 1,3-butadiene and benzene concentrations (Figure 48). The remaining risk is primarily due to coal-fired boilers, power generation and domestic fuel burning.

138

a 50 Dometic wood burning 43

24 Domestic coal burning 26

10 Other sources 9

6 Diesel vehicles 7

5 Electricity generation 6

1 Petrol vehicles 5

Daily Mortality (Premature Deaths) 4 Coal-fired boilers 4 Respiratory Hospital Admissions

0 10 20 30 40 50 60 Percentage Contribution

Petrol vehicles 54

Diesel vehicles 37

Dometic wood burning 7

Leukaemia Cases Domestic coal burning 2

0 10 20 30 40 50 60 Percentage Contribution

139

6.2.8 Comparison of Predicted Health Risks due to Fuel Burning with Total Health Effects due to All Causes To put the predicted health effects into perspective it is useful to show the various estimated health effects as a percentage of the actual (measured) health effects due to all causes. The annual incidences of health endpoints given in the literature as occurring in the provinces of interest are given in Table 24. Predicted health effects due to estimated inhalation exposures to ambient air pollutant concentrations occurring due to fuel burning emissions are given as a percentage of the actual incidences in Table 25.

Table 24: Annual incidences in health endpoints due to all causes

Incidence – Given as % of Population Health Endpoint Western KwaZulu Mpuma- Source Gauteng Cape Natal langa Total mortality 1.235 1.235 1.235 1.235 (a) Increased mortality Respiratory mortality 0.141 0.141 0.141 0.141 (b) Cardiovascular mortality 0.043 0.043 0.043 0.043 (b) All respiratory 3.1 3.1 3.1 3.1 (c) Increased hospitalizations Chronic bronchitis 2.7 3.0 2.4 2.55 (d) & related health care visits Asthma 0.2 0.2 0.2 0.2 (c) Increased symptoms Asthma 5.4 4.7 3.85 3.2 (d) Decreased lung function Peak expiratory flow 4.2 3.7 5.55 1.75 (d) Sources of information: (a) Based on MRC report - Initial Estimates from the South African National Burden of Disease Study, 2000 - MRC Policy Brief No. 1 March 2003 - Debbie Bradshaw et al. (2003) (b) Causes of Death in South Africa 1997 - 2001, Advanced Release of Recorded Causes of Death, Statistics South Africa (2002) (c) Based on incidences of respiratory ailments recorded for Johannesburg (Johannesburg State of Environment Report, 2000) (d) Department of Health: SA Demographic and Health Study 1998.

Inhalation exposures to fuel burning emissions were predicted to be responsible for 18% to 24% of respiratory hospital admissions and 17% to 26% of all chronic bronchitis cases. It is evident therefore that inhalation exposures to air pollution from fuel burning are estimated to be a significant risk factor in terms of these health endpoints. Inhalation exposures due to fuel burning emissions were found to be much less significant in terms of total mortality, cancer cases and heart disease, as may have been expected – there are a wide range of other risk factors that contribute to such incidences, e.g. genetics and diet.

140

Table 25: Percentage of actual health effect incidence accounted for by the health effects predicted to be due to inhalation exposures due to fuel burning emissions

% of Actual Incidence Accounted for by Inhalation Exposures to Fuel Burning Emissions: Health Endpoint Johannesburg & Vaal Mpuma- Cape Town eThekwini Tshwane All Ekurhuleni Triangle langa Respiratory hospital 23.2 23.7 17.6 18.5 19.3 22.0 20.6 admissions Persons suffering from 0.03 0.04 0.02 0.02 0.02 0.02 0.03 heart disease Total mortality 0.18 0.17 0.09 0.09 0.10 0.11 0.13 Chronic bronchitis 26.0 17.0 25.8 20.0 25.0 19.8 22.9 Persons diagnosed with 0.001 0.001 0.001 0.004 0.001 0.001 0.001 cancer

6.2.9 Projected Changes in Inhalation Health Risks due to Fuel Burning Exposures Assuming business as usual, it is estimated that health effects due to exposures to ambient pollutant concentrations resulting from burning emissions will increase during the 2003 to 2011 period in the following order:  Cape Town (3% to 22%)  eThekwini (0% to 23%)  Johannesburg and Ekurhuleni (4% to 21%)  Tshwane (14% to 26%)  Vaal Triangle (0% to 19%)  Mpumalanga (0% to 23%). The small rate of increase in the Vaal Triangle was due to the predominance of domestic fuel burning effects and the assumption made that domestic fuel burning will not increase.

The projected increase in health risks is considered a lower bound estimate due to it having been assumed that residential fuel burning would persist in the short-term (2003 - 2007) and start to decrease in the medium-term as a result of lower population growth rates and on-going electrification. According to the Integrated National Electrification Programme all houses were projected to be electrified by 2010/2011. However, an estimated 2.5 to 2.9 million households (about 20% of households) were estimated to be unconnected to the grid by mid 2011 (Eskom, 2011). The price of electricity is also conjectured to have reverse the switch to electricity by some households, even when such households are grid-connected. It is therefore expected that the increase in inhalation health risks associated with household fuel burning during the 2007 to 2011 period may be understated.

The most significant fuel burning sources, in terms of human health risk, having been identified, it is possible to consider interventions tailored to the realisation of emission reductions from such sources.

141

6.3 Health Cost Predictions Total direct health costs calculated per health condition, conurbation and source group are given in Appendix E. Summed total direct health costs per conurbation and per source grouping are given in Table 26 and Table 27 respectively. Costs are estimated in 2002 Rand values.

Table 26: Total direct health costs (Million 2002 Rand) due to inhalation exposures to fuel burning emissions summed per conurbation

Total Cost of Respiratory Region Total Cancer Costs Total Direct Health Costs Conditions

Cape Town 859 4.7 864 eThekwini 788 7.8 796 Johannesburg & Ekurhuleni 991 12.0 1 003 Tshwane 297 12.8 310 Vaal Triangle 275 1.6 277 Mpumalanga 253 1.1 254 Total 3 463 40.1 3 503

Table 27: Total direct health costs (Million 2002 Rand) due to inhalation exposures to fuel burning emissions summed per source grouping

Total Cost of Respiratory Source Group Total Cancer Costs Total Direct Health Costs Condition

Household coal burning 889 1.0 890 Household wood burning 1 499 2.8 1 502 Household 'other fuel' 7.6 - 8 burning (gas, paraffin) Petrol-driven vehicles 164 21.4 185 Diesel-driven vehicles 256 14.9 271 Industrial & commercial 447 0.04 447 sector Power generation 200 0.01 200 Total 3 463 40.1 3 503

Total direct health costs related to inhalation exposures to fuel burning emissions were estimated to be in the order of R3.5 billion (2002 Rand) per annum across health effects, conurbations and source groupings. Respiratory illnesses due to inhalation exposures to fuel burning emissions accounted for 99% of the estimated total direct health costs, with cancer costs being negligible in comparison.

The greatest health costs were estimated to be incurred in Johannesburg, Ekurhuleni, Cape Town and eThekwini, with these conurbations accounting for approximately 76% of the estimated total health spending across all conurbations considered (Figure 49). The lower costs estimated for the Vaal Triangle and Mpumalanga Highveld were in part due to the smaller populations residing in these areas in comparison with the metropolitan areas.

142

Figure 49: Percentage of total direct health costs due to inhalation exposures to fuel burning emissions incurred per conurbation

Household fuel burning was estimated to be responsible for about 68% of the total health costs estimated across all conurbations, vehicle emissions for 13%, industrial and commercial fuel burning for 13%, and power generation for ~6% (Figure 50). Over 40% of the health spending due to household fuel burning was estimated to be due specifically to wood burning.

The contribution of source groups to health costs for individual conurbations is given in Table 28. Household fuel burning was estimated to be responsible for over 75% of the total direct health costs in Cape Town, eThekwini, Johannesburg and Ekurhuleni and the Vaal Triangle. Household fuel burning was estimated to account for 24% of direct health costs in Tshwane(19) and 19% in the Mpumalanga Highveld (~50%). Health costs related to power generation emissions, industrial and commercial fuel burning releases were estimated to be significantly higher in the Tshwane, Vaal Triangle and Mpumalanga Highveld regions compared to the contributions of these source types within other conurbations. Power generation was associated with over 50% of the total direct health costs in the Mpumalanga Highveld region.

19 Unrepresentative spatial distribution of vehicle emissions within Tshwane during dispersion simulations is anticipated to have resulted in the overestimation of health impact related to this source grouping.

143

Power generation 5.7% Industrial & Household coal commercial sector burning 12.8% 25.4%

Diesel-driven vehicles 7.7%

Petrol-driven vehicles 5.3%

Household 'other fuel' burning (gas, Household wood paraffin) burning 0.2% 42.9%

Figure 50: Contribution of source groupings to total direct health costs estimated to occur due to fuel use and inhalation exposures to fuel burning emissions

Table 28: Contribution of source groups to total direct health spending related to fuel use and inhalation exposures to fuel burning emissions per conurbation

% Contribution to Total Direct Health Costs in 2002 Source Group Johannesburg Mpumalanga Cape Town eThekwini Tshwane Vaal Triangle & Ekurhuleni Highveld Household coal burning 1.8 10.4 54.5 19.4 56.3 11.7 Household wood burning 80.1 64.3 21.1 4.3 20.2 6.9 Household 'other fuel' 0.3 0.5 0.1 0.0 0.0 0.0 burning (gas, paraffin) Petrol-driven vehicles 1.8 6.3 5.8 19.5(a) 0.3 0.1 Diesel-driven vehicles 2.7 12.1 7.2 23.7(a) 0.5 1.5 Industrial & commercial 12.8 6.4 10.7 23.0 13.3 27.6 Power generation 0.4 0.0 0.7 10.1 9.4 52.1 (a) Unrepresentative spatial distributions of vehicle emissions within Tshwane during the dispersion simulations is anticipated to have resulted in the overestimation of health effect related to this source grouping.

The relationship between emissions and resultant health costs due to inhalation related health effects varies significantly between source groupings, the reason being that the human health effect of a source is dependent not only on the extent of its emissions but also on a number of other factors. Such factors include type of pollutant released, height of release, proximity of the source to areas with high human exposure potentials, and duration and frequency of emissions. To

144

demonstrate this, total estimated health costs due to inhalation exposures (i.e. respiratory illness and cancer costs) are given per ton of total emissions for each source grouping (Table 29).

Table 29: Total emissions of selected pollutants, total health spending on resultant health effects related to inhalation exposures and health spending per ton of emission per source group

Total Emissions of PM10, Total Health Costs for SO2, NO2, Benzene, Respiratory Illnesses and Cancer Total Health Costs Source Group Formaldehyde and 1.3- Incidences Due to Inhalation per Tonne of Emission Butadiene Exposures to Fuel Burning (2002 Rand) (kilotonnes annum-1) (Million 2002 Rand) Industrial, Commercial & 872 447 ~500 Institutional Fuel Burning Electricity Generation 2 267 200 ~100 Vehicles 322 456 ~1 400 Domestic Fuel Burning 28 2 400 ~ 85 000

Although domestic fuel burning is estimated to have contributed only ~1% of the emissions across all conurbations and source groupings considered, it was predicted to account for 68% of the health costs due to inhalation-related health effects. The health effect potential of domestic fuel burning emissions is enhanced due to three factors: (i) the low level of emissions; (ii) the coincidence of peak emissions, typically a factor of 10 greater than if total annual emissions were averaged, with periods of poor atmospheric dispersion (i.e. night-time, winter-time); and (iii) the release of such emissions within high human exposure areas.

The significance of vehicle emissions in terms of the contribution to air pollutant concentrations and health risks is similarly enhanced by the low level at which emissions occur and the proximity of such releases to high exposure areas. Vehicle emissions also tend to peak in the early morning and evenings at which time atmospheric dispersion potentials are reduced. Vehicles are estimated to contribute 9% of emissions but be responsible for 13% of the health costs related to inhalation- related effects.

The significance of fuel burning within the power generation sector in terms of their contributions to ground level air pollutant concentrations and public health risks is frequently lower than would be expected given the extent of emissions. This is due to these sources generally being characterised by constant, high level releases with such emissions also likely to be more remote from residential settlement compared to household fuel burning and vehicle emissions. Although power station emissions are estimated to represent 65% of the total emissions quantified, this source grouping is only predicted to be responsible for 6% of the health costs calculated for inhalation-related effects.

The industrial, commercial and institutional fuel burning source grouping comprises a diverse range of sources ranging significantly in height of release and proximity to human settlement. Overall, this source grouping is estimated to represent 25% of the total emissions, whilst being predicted to be responsible for ~13% of the health costs calculated for inhalation-related effects.

From the information provided in Table 29 it is apparent that minor emission reductions within the domestic fuel burning sector would result in relatively significant reductions in direct health spending. Substantial emission reductions would need to be realised within the power generation

145

sector to achieve equivalent decreases in health spending. This indicates that the most cost- effective interventions are likely to be within the domestic fuel burning sector. The costs of implementing such interventions and the potential for offsetting such costs through direct health savings due to avoided damages are explored in Chapter 7.

146

7 Cost-optimisation of Air Pollution Mitigating

The externalities study, a key component of the thesis, is documented in chapters 4, 5, 6 and 7. Building on the overview of sector-specific interventions provided in Chapter 3, this chapter documents the specific mitigation measures which have been subject to cost-benefit analysis as part of the externalities study. Health effect reductions and associated reduced costs achievable through the implementation of such measures are analysed with the purpose of informing the cost-optimisation of air pollution intervention strategies.

7.1 Source Prioritisation and Intervention Selection Based on the effects of air quality and human health and welfare, it was recommended by the author to the National Economic Development and Labour Council (NEDLAC) Working Group overseeing the Dirty Fuels Study, that emission reduction opportunities be identified and assessed for the following anthropogenic fuel combustion sources:  Residential fuel burning  Coal-fired power stations  Vehicle emissions  Coal fired boilers  Specific industries undertaking significant fuel combustion. Industrial sectors notable for their fuel burning emissions were identified as including: chemical and petrochemical (specifically refineries utilising coal and HFO), pulp and paper mills, food and tobacco plants (specifically sugar refineries), metallurgical processes (specifically iron and steel plants) and non-metallurgical processes (including brick and cement manufacture).

Given that it is not possible to quantitatively evaluate all measures identified for possible implementation, the following criteria were used for the qualitative assessment and prioritisation of options:  Environmental benefits - As a minimum, the measure should ensure the reduction of emissions. Options expected to achieve the following are preferred from an environmental perspective: (i) reduce ambient air pollutant concentrations; (ii) realise health risk reductions - including occupational and public exposures; and (iii) realise environmental risk reductions.  Technical viability - The option should be practical and feasible under current conditions, with the technology required for its implementation already available. (Options that require further development involve a higher degree of uncertainty.) Options that are based on proven technologies or methods are preferable to those using unproven technologies.  Degree of uncertainty - Preference should be given to measures that are associated with a high degree of certainty with regard to measureable implementation and environmental benefit realisation.  Social acceptability and desirability - The social acceptability of measures is an important indicator of the viability of measures. Preference should be given to measures that are not only socially acceptability but that are likely to be desirable to interested and affected parties, and to the general public.  Economic feasibility - The feasibility of the measure, given the capital and operating costs associated with its implementation, and the sectors responsible for covering such costs.

147

 Strategic and political desirability - Acceptability of the measure given current legislation, regulations and strategies, with emphasis placed on measures likely to be more desirable politically.  Timeframes for implementation and environmental benefit realisation - Shorter timeframes for implementation should be given preference, since lags in the management of significant sources increase the risk for public and environmental exposure.  Development of local expertise and potential for local employment – These are minor considerations that should be taken into account when comparing measures that are environmentally beneficial, technically viable and socio-economically acceptable.

The main approaches that could be adopted in securing the implementation of interventions include a prescribed (regulatory) approach, voluntary agreements, and market incentives and disincentives. The regulatory approach could take two main forms:  Prescription of methods - involves the specification of fuel types, process options and pollution control technologies for specific operations and activities; and  Prescription of objectives - involves the specification of energy efficiency targets and/or emission limits to be met by specific operations.

A combination of both approaches is being proposed by government for the regulation of vehicle emissions. The first approach is more practical in the regulation of household fuel combustion and has traditionally been implemented. The second approach is generally preferred by industry since it allows operations to be flexible and innovative in determining methods that may be implemented to ensure the achievement of energy efficiency targets and emission limits.

7.2 Description of Selected Interventions Emission reduction measures for prioritised sources identified for possible quantitative assessment are listed in Table 30, Table 31 and Table 32 for the residential, transport and industrial and power generation sectors respectively. The assumptions made regarding the spatial scale and timeframe for implementation of the various interventions are outlined in the relevant tables.

Interventions considered to reduce emissions from residential fuel burning included several of the measures highlighted for implementation by the Department of Minerals and Energy (DME) in its Integrated Clean Household Energy Strategy (ICHES) adopted in 2003 as discussed in Section 3.2.1.

Measures included in the Implementation Strategy for the Control of Exhaust Emissions from Road-going Vehicles in South Africa proposed by the Department of Environmental Affairs (DEA) and DME in 2003 were selected interventions for the transport sector. In the absence of detailed spatial traffic flow data, it was not feasible to assess accurately the implications of measures aimed at curbing vehicle activity, e.g. implementation of designated high occupancy vehicle lanes on freeways. This is noted as a limitation of the study.

Coal-fired pulverised fuel power stations will continue to supply the bulk of South Africa‟s electricity in the short- to medium-term. Measures identified to reduce the extent of coal-fired power generation emissions include the implementation of desulphurisation technologies on power stations, the replacement of a coal-fired power station with a gas reticulation network and the

148

implementation of renewable energy policies. The potential for retrofitting existing power stations with desulphurisation technology and the cost-effectiveness of doing so has been debated for a number of years, without clear resolution.

In the selection of measures targeting the industrial sector, it was decided to specify an intervention implementable across of range operations with similar fuel combustion practices (i.e. coal fired boiler operations), in addition to selecting interventions for three specific industries. Due to the complexity of identifying plant-specific measures and estimating associated implementation costs and emission reductions achievable, consideration was given to measures already identified and investigated by industries that have been highlighted as contributing significantly to health risks. The following industry-specific interventions were selected:  Replacement of coal use with carbon monoxide at the Highveld Steel & Vanadium plant near Witbank (Mpumalanga Highveld)(20)  Realisation of emission reductions from the coke oven process at the Mittal Steel (previously Iscor) plant at Vanderbijlpark (Vaal Triangle)  Desulphurisation technology implementation at the Sasol Secunda power generation plant (Mpumalanga Highveld).

Measures were quantified only if they were clearly able to result in an emission reduction. Control efficiencies for each of the measures are provided in the various tables and reasons provided for measures excluded from the quantitative analysis. Emphasis was placed on short- and medium- term interventions able to realise emission reductions and health risk improvements in the next ten to fifteen years. A synopsis of the interventions quantified and their related control efficiencies is given in Table 33.

Difficulties were experienced in the estimation of changes in health risk due to the taxi recapitalisation project (Intervention 28)(21). The number of taxis in operation in 2002 within the conurbations being investigated was estimated based on: (i) the number of taxis given as operating nationally, i.e. 126 000, and (ii) the percentage of minibuses, including taxis and private, registered within each conurbation. As a starting point it was assumed that all current petrol-driven taxis would be replaced by 38-seater diesel powered vehicles by the year 2007. (This was assumed because details were not available regarding the proportion of 18 and 38 seater diesel vehicles which would be introduced. It was anticipated that the assumption of the 38 seater vehicles would maximise any benefits since it would result in the most significant reduction in vehicle kilometres travelled.) Minibus taxis are given as being responsible for 65% of the 2.5 billion annual passenger trips in urban areas. (Minibus taxis however accommodate more passengers than the average passenger vehicle or metered taxis.) It is estimated that the taxis operational in the various conurbations of interest in the study account for 7% to 15% of the vehicle kilometres travelled by all petrol-driven vehicles.

20 Carbon monoxide introduced as a reductant in the steel making process, replacing the use of coal in this process. 21 Since the time of writing this chapter, the taxi re-capitalisation project has undergone several delays and changes, and has not been implemented to the extent or according to the schedule originally envisaged.

149

Problems were experience in accurately estimating emissions from the current petrol-driven taxis and from the proposed replacement vehicles. The reason for this is that the available local emission factors (Wong, 1999) were developed based on exhaust monitoring tests undertaken on a range of vehicle types and ages estimated to be representative of the SA vehicle fleet. Given that the majority of the taxis which are currently operational are >10 years old, the application of the emission factors would have resulted in an underestimation of current taxi emissions. No emission factors are specifically available for the proposed diesel-powered taxis. It was recommended by the Counterpart Group that reference be made to EURO 2 standards for the purpose of estimating such emissions. EURO 2 emission standards are only given for NOx, CO and particulate matter. The accuracy of using these emission factors is not known.

The implementation of the taxi recapitalisation programme was calculated to result in reductions in

NOx emissions of between 1% and 2%, reductions in CO emissions of 3% to 7% and reductions in lead emissions of 7% to 15%. Particulate emissions are estimated to increase by between 1% and 4%. Due to the uncertainties in the applicability of the emission factors, and given that emission estimates were not made for all pollutants under consideration, it was decided not to consider the health implications of this intervention.

150

Table 30: Interventions to mitigate atmospheric emissions for the residential sector

Measure Spatial Application Timeframe Assumptions Control Efficiency DME (1) DME Basa Njengo Magogo Short to (1) National rollout of the Basa Njengo Magogo technology (1) Estimated reduction of emissions of 50% for particulate 22 ICHES( ) Implementation areas, viz. medium term will occur over the 12 year period, 2003 to 2015. matter and 20% for other pollutants from 272 000 Top down • Mpumalanga Highveld - total The BMN technology measured as resulting in a reduction in households within Johannesburg and Mpumalanga (~12% ignition (Witbank/Kwa-Guqa) reductions fuel use by 20% with emissions reduced by 50%. of coal burning households in the Plateau study areas) • Mountainous area of KwaZulu- only to be Assume 80% of Johannesburg households and 60% of i.e. 6% reduction in PM emissions and 2% reduction in Natal & Free State (Volksrust to realised by Mpumalanga Highveld households are approached (i.e. emissions of other pollutants from total coal burning within Qwa-qwa) 2012 ~272 000 households). the Plateau study areas • Free State (Bloemfontein) (33% reduction in PM and 13% reduction in other • North Cape (Kimberly) Assume 50% of households retain and use the technique (as per pollutants in Johannesburg; 25% reduction in PM and 10% • Gauteng (Soweto, Alexandra, DME's Basa Magogo 50 initiative). reduction in other emissions in Mpumalanga Highveld Orange Farm) (2) National rollout of BNM technique to all plateau study area) • NW Province (Potchefstroom) areas in the 2003-2015 period. Assume 80% of all households (2) Estimated emission reduction of 20% in PM emissions (2) All household coal burning areas approached and that 50% of households retain and use the and 8% in other pollutant emissions across all study areas. technique.

DME ICHES Plateau Short to Assume 3 plants producing 300 000 tpa of LSF will be 50% reduction in PM and 0% reduction in other emissions Low smoke medium term established by 2012 and that the quantity of LSF produced will from all coal burning households on Plateau. fuels (LSF) - total be sufficient to replace total current coal consumption in the reductions Plateau region. only to be realised by Regulations will be introduced to prohibit the use of coal and 2012 necessitate the implementation of LSFs or alternatives. It will be assumed that the LSF will be required to reduce PM emissions by a minimum of 50% with no reductions in other pollutants being required(23).(The LSF standard will require LSFs to reduce particulate matter and possibly other pollutants such as sulphur dioxide by a certain percentage prior to their being classified as LSFs. At the time the work was completed these percentage reductions had not been finalised.)

22 Department of Minerals and Energy (DME) Integrated Clean Household Energy Strategy (ICHES) 23A DME sponsored investigation was undertaken into emission reductions, air quality improvements and health risk changes associated with various proposed low smoke fuels (Scorgie et al., 2001). It was noted that certain 'low smoke fuels' have the potential to increase cancer risks despite reducing the risk of respiratory risks due to particulate exposures. Other fuels reduced the total particulate matter but increased the extent of fine particulate matter. Only four out of twelve fuels tested during the study were found to successfully reduce cancer, irritation and systemic impacts. In the quantification of the low smoke fuel intervention for the purpose of the current study it is assumed that the fuel will reduce fine particulate matter by 50% and will not result in any other health risks in excess of the risks associated with coal burning. 151

Table 30 (continued): Interventions to mitigate atmospheric emissions for the residential sector

Spatial Measure Timeframe Assumptions Control Efficiency Application DME ISCHES Plateau study Medium term Although the measure is seen to have great potential by the DME it was given Emission reduction range of 2% to 8% reduction in Housing insulation areas Scheduled for as not being ready for implementation yet as suitable (efficient and safe) and the emissions of each pollutant emitted from coal implementation affordable insulation material still has to be researched/developed. and wood burning households within Plateau study by the DME in Assume various % of existing households within Plateau study areas which areas 2006 burn either coal and/or wood are insulated by 2012 (ranging from 5% to 20% of the 1.2 million households). Insulation assumed to result in a 40% control efficiency. (Control efficiency estimation informed by: DME indicates 60%; Study by Irurah et al. (2000) indicates 40% to 70% reduction due to ceilings and 30% to 85% due to wall insulation; eMbalenhle study stated 5% to 30% reduction in fuel use (Scorgie et al., 2001). DME ISCHES National Short-term Assume that 80% of households in study areas are currently electrified with Emission reduction of 55% of pollutants released Electrification ongoing remaining 20% of households (~692 000 households) to be supplied with from currently unelectrified coal and wood burning electricity in the short term (i.e. by 2007) households. According to The South African Policy Research and Training Project undertaken by the Energy for Development Research Centre (1993), Uniform increase in emissions by all power stations approximately 45% of electrified townships households and 88% of assumed. unelectrified township household use coal on a daily basis for cooking and space heating. Assume that of the newly electrified households previously using coal or wood, 45% of such households will continue to use these fuels Stove maintenance Plateau study Short- to medium Repair or replacement of stoves within 10% of coal burning households (i.e. 20% reduction within 10% of coal burning and replacement areas term - results by ~74 400 households) households within Plateau area, i.e. total of 2% 2012 Emission reduction of 15% to 30% per coal burning household with repaired reduction in emissions stove (based on eMbalenhle study findings documented in Scorgie et al., 2001) - assume 20% reduction Electrification of National Short-term – prior Electrification of all paraffin-burning households to reduce paraffin-burning Two time periods to be taken into account, viz. (i) paraffin-burning to 2007 and its related effects (paraffin poisoning, burns, fires, inhalation exposures). one year after intervention implementation and (ii) households According to the literature, 67% to 94% of households continue to use paraffin 10 years after implementation. This is necessary in the first few years after electrification. For the purpose of the study it was since electrification does not result in all assumed that 67% of households would continue to burn paraffin the year after households discontinuing their use of paraffin, but they were electrified, following which there would be a 5% per annum rather a gradual reduction in paraffin use. reduction in the number of households which burned paraffin for the first 10 years after electrification. Paraffin use and burning related risks are estimated to be reduced by 37% one year post-electrification and by 78% 10 years post- electrification.

152

Table 31: Interventions to mitigate atmospheric emissions from vehicles

Measure Spatial Timeframe Assumptions Application DME/DEAT 2003 STRATEGY National By 2008 All new passenger cars and light commercial vehicles newly manufactured to abide by Euro 2 standards Legal stipulation of new technology for new (assume fuel specification changes to make possible achievement of such standards) vehicles National By 2012 All new passenger cars and light commercial vehicles newly manufactured to abide by Euro 4 standards (assume fuel specification changes to make possible achievement of such standards) DME/DEAT 2003 STRATEGY National By 2007 Reduction of maximum S content of unleaded petrol to 500 ppm (by 2007) Reduction of sulphur content of petrol By 2012 And to 50 ppm (by 2012) DME/DEAT 2003 STRATEGY National By 2012 Ensure maximum benzene content in petrol of <1% Reduction of benzene content of petrol DME/DEAT 2003 STRATEGY National By 2012 Ensure maximum aromatic content in petrol of <35% Reduction of aromatic content of petrol DME/DEAT 2003 STRATEGY National By 2007 0% lead in petrol by 2007 (Implemented on schedule) Phasing out of lead DME/DEAT 2003 STRATEGY National (1) By 2007 (1) Ensure maximum sulphur content of diesel <500 ppm & availability of a second diesel grade with Reduction of sulphur content of diesel maximum diesel content of 50 ppm (Assume 10% of diesel-vehicle owners elect to use low-sulphur diesel) (2) By 2012 (2) Ensure maximum sulphur content of 50 ppm 24 Taxi recapitalisation project( ) National By 2007 It was assumed that all current petrol-driven taxis would be replaced by 38-seater diesel powered vehicles by the year 2007. (This was assumed because details were not available regarding the proportion of 18 and 38 seater diesel vehicles which would be introduced. It was anticipated that the assumption of the 38 seater vehicles would maximise any benefits since it would result in the most significant reduction in vehicle kilometers travelled.) Conversion of petrol-driven vehicles to LPG National By 2012 Conversion of 10% to 20% of petrol-driven vehicle fleet to LPG Stipulation that all petrol-driven vehicles are National By 2011 All current and new passenger cars and light commercial vehicles newly manufactured to abide by Euro EURO 2 compliant 2 standards (assume fuel specification changes to make possible achievement of such standards)

24 The taxi re-capitalisation project has undergone several delays and changes, and has not been implemented to the extent or according to the schedule originally envisaged. 153

Table 32: Interventions to mitigate atmospheric emissions from the industrial and power generation sectors

Measure Spatial Application Timeframe Assumptions Control Efficiency Desulphurization of power All power stations within study By 2011 6% reduction in generation capacity due to 5.3% sulphur dioxide emission reduction, station emissions prior to areas desulphurization increase in other emissions by 6% release Decommissioning of Tshwane By 2011 Electricity generated by Pretoria West power station can 100% emission reduction for Pretoria West Pretoria West power station be replaced by residential gas usage Power Station - replacement with gas reticulation network Renewable energy (RE) National, with emission By 2011 A block of energy supply of 10 000 GWh is contributed 5.2% reduction in emissions in 2011 power technology implementation reductions primarily realised by RE technologies (solar), i.e. 6.4% of Eskom's average station emissions for power stations located in through financial incentives within Mpumalanga and Vaal annual production Vaal Triangle and Mpumalanga. (investment incentives & set Triangle A block of 37 000 GWh is contributed by RE, i.e. 23.4% 19% reduction in emissions in 2011 power asides) of Eskom's average annual production station emissions for power stations located in Vaal Triangle and Mpumalanga. Emission reduction National By 2007 Emission reduction achieved through implementation of 90% reduction in emissions from all coal- requirements for coal fired one or a combination of the following: fired boilers boilers fuel switching abatement technology implementation improvements in energy efficiency Replacement of coal use Mpumalanga By 2007 Use of carbon monoxide as a reductant in place of coal 95% reduction in particulate matter emissions, with CO at Highveld Steel and 100% reduction in benzene emissions & Vanadium

Emission reduction from Vaal Triangle By 2007 Coke oven gas cleaning project, including new H2S reduced by 96%, SO2 by 74%, ammonia coke oven process at Mittal aspiration system approved and implemented by 72%, NOx by 45%, methane by 16%, Steel (previously Iscor) benzene by 5%, PM by 15% Vanderbijlpark Plant

154

Table 33: Interventions selected for quantitative health risk assessment and their associated scales, timeframes and emission reductions

% Emission Reduction Time- No. Sector Measure Spatial scale Sub-sector Location 1,3-buta- Formalde PM10 SO2 NO Benzene Lead frame x diene hyde

1 (25) Johannesburg 20 8 8 8 8 8 8 2011 Residential Top down ignition - DME ICHES DME areas only Domestic coal Mpumalanga 15 6 6 6 6 6 6 2011 Plateau roll out - all study 2 Top down ignition - plateau roll out Domestic coal Plateau 20 8 8 8 8 8 8 2011 Residential areas Plateau implementation - 3 Low smoke fuels Domestic coal Plateau 50 0 0 0 0 0 0 2011 Residential supported by legislation 4 Housing insulation - 5% of plateau fuel Domestic fuel Plateau (5% of hh) Plateau 2 2 2 2 2 2 2 2011 Residential burning households (space heating) 5 Housing insulation - 20% of plateau fuel Domestic fuel Plateau (20% of hh) Plateau 8 8 8 8 8 8 8 2011 Residential burning households (space heating) 6 Housing insulation - 5% of all fuel All study areas (5% of Domestic fuel All areas 2 2 2 2 2 2 2 2011 Residential burning households hh) (space heating) 7 Housing insulation - 20% of all fuel All study areas (20% of Domestic fuel All areas 8 8 8 8 8 8 8 2011 Residential burning households hh) (space heating) 8 Domestic fuel Electrification National implementation All areas see emission reductions per conurbation (a) 2007 Residential (space heating) 9 Stove maintenance and replacement - All study areas (5% of Domestic fuel All areas 1 1 1 1 1 1 1 2011 Residential 5% households all areas hh) (space heating) 10 Stove maintenance and replacement - All study areas (20% of Domestic fuel All areas 4 4 4 4 4 4 4 2011 Residential 20% households all areas hh) (space heating) 11 All power stations in Electricity All areas (except Electricity Desulphurization of all PS emissions (6) 94.7 (6) (6) (6) (6) (6) 2011 Generation study areas generation eThekwini) 12 Decommissioning of Pretoria West PS - Tshwane - Pretoria West Electricity Electricity Tshwane 100 100 100 100 100 100 100 2011 Generation gas use by households PS only generation 13 RE technology implementation through Vaal Triangle & Electricity Mpumalanga, Electricity 5.2 5.2 5.2 5.2 5.2 5.2 5.2 2011 Generation financial incentives (10 000 GWh block) Mpumalanga generation Vaal Triangle 14 RE technology implementation through Vaal Triangle & Electricity Mpumalanga, Electricity 19.0 19.0 19.0 19.0 19.0 19.0 19.0 2011 Generation financial incentives (37 000 GWh block) Mpumalanga generation Vaal Triangle

25 Department of Minerals and Energy (DME) Integrated Clean Household Energy Strategy (ICHES) 155

% Emission Reduction Time- No. Sector Measure Spatial scale Sub-sector Location 1,3-buta- Formalde PM10 SO2 NO Benzene Lead frame x diene hyde 15 Emission reduction requirements for Coal-fired Industrial coal fired boilers for particulate matter National implementation All areas 90 0 0 0 0 0 0 2011 (>90% control efficiency required) boilers Mittal Steel 16 Mittal Steel (Iscor) coke oven gas Industry - iron (Iscor)Vanderbijlpark Vaal Triangle Industrial cleaning project 15 74 45 5 0 0 0 2007 Works (Vaal Triangle) & steel 17 Industrial Highveld Steel & Vanadium - Highveld Steel & replace coal use with CO use Vanadium Industry - Mpumalanga 95 0 0 100 0 0 0 2007 (Mpumalanga iron & steel Highveld) 18 Industrial Desulphurization of Sasol Industry - Sasol Secunda Mpumalanga (6) 94.7 (6) (6) (6) (6) (6) 2007 Secunda Power Station emissions petrochemical 19 Transport DME Strategy - Reduction of S National Vehicles - content of petrol to 500 ppm All areas 0 0 0 0 0 0 0 2007 implementation petrol (0.05%) (b) 20 Transport DME Strategy - Reduction of S National Vehicles - content of petrol to 50 ppm All areas 0 (b) 0 0 0 0 0 2011 implementation petrol (0.005%) 21 Transport DME Strategy - Reduction of National Vehicles - All areas 0 0 0 50 0 0 0 2011 benzene content of petrol to 1% implementation petrol 22 Transport DME Strategy - Reduction of National Vehicles - aromatics content of petrol to 35% All areas 0 0 0 0 0 0 0 2011 implementation petrol (b) 23 Transport DME Strategy - Phasing out of National Vehicles - All areas 0 0 0 0 0 0 100 2007 lead implementation petrol 24 Transport DME Strategy - Reduction of S National content of diesel to <500 ppm implementation - Vehicles - (0.05%) & second grade diesel second diesel grade (50 All areas 31.9 84.8 0 0 0 0 0 2007 diesel with 50 ppm S content available ppm S content used by 10% of vehicles) 25 Transport DME Strategy - Reduction of S National Vehicles - content of diesel to <50 ppm All areas 37 98.3 0 0 0 0 0 2011 implementation diesel (0.005%) 26 Transport DME Strategy - new passenger vehicles comply with Euro 2 National Vehicles - All areas 0 30 37 37 38 37 2007 standards (assume fuel specs implementation petrol changed)

156

% Emission Reduction Time- No. Sector Measure Spatial scale Sub-sector Location 1,3-buta- Formalde PM10 SO2 NO Benzene Lead frame x diene hyde 27 Transport DME Strategy - new passenger vehicles comply with Euro 4 National Vehicles - All areas 0 38 30 39 38 38 2011 standards (assume fuel specs implementation petrol changed) 29 Transport All petrol vehicles EURO 2 National Vehicles - All areas 0 77 95.8 95.6 99 96 2011 compliant (assume fuel spec implementation petrol changes in place) 30 Transport Conversion of 10% of petrol National Vehicles - All areas 9.2 10 9.4 10 10 10 10 2011 vehicles to LPG implementation petrol 31 Transport Conversion of 20% of petrol National Vehicles - All areas 18.4 20 18.8 20 20 20 20 2011 vehicles to LPG implementation petrol 32.1 Residential Electrification of paraffin-burning National Domestic fuel All areas households (1 year post implementation (paraffin) 33 33 33 33 n.a. n.a. n.a. 2007 electrification) 32.2 Residential Electrification of paraffin-burning National Domestic fuel All areas households (10 year post implementation (paraffin) 78 78 78 78 n.a. n.a. n.a. 2007 electrification)

Notes: (a) Percentage emission reduction per conurbation: Cape Town (7.2%), eThekwini (11%), Johannesburg & Ekurhuleni (9.9%), Tshwane (11%), Vaal Triangle (12.7%), Mpumalanga (13.8%). (b) The sulphur and aromatics content of the petrols tested by Wong (1999) on which the emission estimates for the current study was based were noted to have been below 0.05% and 35% respectively. No emission reduction was therefore assumed to be achievable. (It is however noted that the aromatics content of current fuels is much higher, with maximums in the order of 48%).

157

7.3 Health Impact Reductions to due Interventions Health risk reductions calculated to be achievable by interventions are given in Table 34 and Table 35 for interventions implementable by 2007 and 2011 respectively. Source rankings based on the reductions in respiratory hospital admissions by interventions are illustrated in Figure 51 and Figure 52.

Interventions which target residential fuel combustion were associated with the most significant reductions in respiratory hospital admissions and premature mortality as was expected given the health effects associated with this source. Low smoke fuel implementation(26)(27) was associated with the most significant non-carcinogenic effect reductions. Electrification resulted in the second greatest reductions in respiratory hospital admissions (RHA) followed by large scale housing insulation implementation.

The requirement of all petrol vehicles to be compliant with EURO 2 standards was the fourth most successful measure in terms of respiratory hospital admissions reduction. The restriction of particulate matter emissions from coal-boiler operations represented the industry intervention which resulted in the greatest RHA reductions. The manner in which such particulate matter reductions were to be realised was not stipulated given that a range of methods could be implemented (e.g. fuel switching, clean coal technology implementation, fuel efficiency improvements, abatement technology). The electricity generation intervention resulting in the largest RHA reductions was the desulphurisation of all power station emissions.

Interventions targeting vehicle emissions were predicted to result in the greatest reductions in cancer risks, as was expected. Interventions involving the requirement of technologies compliant with EURO standards, large scale conversion of petrol vehicles to LPG and the restriction of the benzene content of fuels were associated with the most significant cancer risk reductions.

26 The low smoke fuel intervention comprises making sufficient fuels available for the entire plateau household coal market coincident with the passing of legislation restricting the use of coal by households. 27 A DME sponsored investigation was undertaken into the emission reductions, air quality improvements and health risk changes associated with various proposed low smoke fuels (Scorgie et al., 2001). It was noted that certain 'low smoke fuels' have the potential to increase cancer risks despite reducing the risk of respiratory risks due to particulate exposures. Other fuels reduced the total particulate matter but increased the extent of fine particulate matter. Only four out of twelve fuels tested during the study were found to successfully reduce cancer, irritation and systemic impacts. In the quantification of the low smoke fuel intervention for the purpose of the current project it was assumed that the fuel will reduce fine particulate matter by 50% and will not result in any other health risks in excess of the risks associated with coal burning.

158

Table 34: Reductions in health effects due to interventions implementable by 2007, given as the reduction in actual number of admissions, cancer cases and restricted activity days.

Respiratory Hospital Total Annual Mortality Restricted Activity Days Cardiovascular Hospital Admissions (Due to (Due to PM , SO , Chronic Bronchitis (Due (RAD), (Due to PM No. Intervention Admissions (Due to PM 10 2 10 Leukemia Cases PM , SO & NO 10 Benzene & 1,3-butadiene to PM Exposures) Exposures by 20-65 Year 10 2 2 Exposures) 10 Exposures) Exposures) Olds) 8 Electrification -7 946 -65.5 -21.5 -8 405 -61 210 -2.48 Iscor coke oven gas cleaning 16 -103 -0.51 -0.18 -53.6 -466 0.00 project Highveld Steel & Vanadium - 17 -581 -4.91 -1.5 -751 -4 586 0.00 replace coal use with CO use Desulphurization of Sasol Secunda 18 -470 0.07 -1.9 12.4 67.1 0.00 PS emissions DME Strategy - Reduction of S 19 content of petrol to 500 ppm 0.0 0.00 0.0 0.0 0.0 0.00 (0.05%) DME Strategy - Phasing out of 23 0.0 0.0 0.0 0.0 0.0 0.00 lead DME Strategy - Reduction of S content of diesel to <500 ppm 24 -2 152 -17.1 -7.6 -3 257 -16 595 0.00 (0.05%) & second grade diesel with 50 ppm S content available DME Strategy - new passenger vehicles comply with Euro 2 26 -2 123 0.0 -2.7 0.0 0.0 -45.4 standards (assume fuel specs changed) Electrification of paraffin-burning 32.1 -83.4 -0.4 -0.3 -42.1 -386 0.00 households – 1 year post Electrification of paraffin-burning 32.2 -197 -1.00 -0.7 -99.0 -911 0.00 households – 10 years post

159

Table 35: Reductions in health effects due to interventions implementable by 2011, given as reduction in number of admissions, cancer cases & restricted activity days.

Respiratory Hospital Cardiovascular Total Annual Mortality Chronic Restricted activity Minor restricted Admissions (Due to Hospital Admissions (Due to PM , SO , Bronchitis (Due days (RAD) (Due to activity days (MRAD) Leukemia No. Intervention 10 2 PM10, SO2 & NO2 (Due to PM10 Benzene & 1,3- to PM10 PM10 Exposures by (Due to SO2 Exposures Cases Exposures) Exposures) butadiene Exposures) Exposures) 20-65 Year Olds) by 20-65 Year Olds) Top down ignition - DME 1 28 -3 807 -31.6 -9.0 -4 974 -29 990 -343 470 -0.15 ICHES( ) Top down ignition - plateau roll 2 -5 298 -44 -12.3 -6 577 -41 249 -469 375 -0.44 out 3 Low smoke fuels -13 022 -110 -30.5 -16 443 -10 3122 0.00 0.00 Housing insulation - 5% of 4 -747 -6.1 -1.7 -899 -5 711 -11 7472 -0.36 plateau fuel burning households Housing insulation - 20% of 5 -2 987 -24.3 -6.9 -3 596 -22 844 -46 9890 -1.42 plateau fuel burning households Housing insulation - 5% of all fuel 6 -1 640 -13.5 -4.5 -1 740 -12 663 -156 953 -0.42 burning households Housing insulation - 20% of all 7 -6 560 -54.1 -18.0 -6 958 -50 652 -627 812 -1.69 fuel burning households Stove maintenance and 9 replacement - 5% households all -820 -6.8 -2.3 -870 -6 331 -78 476 -0.21 areas Stove maintenance and 10 replacement - 20% households all -3 280 -27.1 -9.0 -3 479 -25 326 -313 906 -0.85 areas Desulphurization of all Power 11 -1 520 1.9 -6.4 301 1 766 -3 3513 621 -0.05 Station emissions Decommissioning of Pretoria 12 -941 -6.0 -2.5 -675 -5 614 -1 227 795 0.00 West PS - gas use by households

28 Department of Minerals and Energy (DME) Integrated Clean Household Energy Strategy (ICHES)

160

Respiratory Hospital Cardiovascular Total Annual Mortality Chronic Restricted activity Minor restricted Admissions (Due to Hospital Admissions (Due to PM , SO , Bronchitis (Due days (RAD) (Due to activity days (MRAD) Leukemia No. Intervention 10 2 PM10, SO2 & NO2 (Due to PM10 Benzene & 1,3- to PM10 PM10 Exposures by (Due to SO2 Exposures Cases Exposures) Exposures) butadiene Exposures) Exposures) 20-65 Year Olds) by 20-65 Year Olds) RE technology implementation 13 through financial incentives (10 -348 -1.1 -0.7 -199 -1 008 -1 656 195 0.00 000 GWh block) RE technology implementation 14 through financial incentives (37 -1 275 -3.9 -2.40 -725 -3 681 -6 050 376 -0.01 000 GWh block) Emission reduction requirements for coal fired boilers for 15 -4 269 -36.0 -11.0 -4 199 -35 075 0.0 0.00 particulate matter (>90% control efficiency required) DME Strategy - Reduction of S 20 content of petrol to 50 ppm -2.7 0.0 -0.02 0.0 0.0 -78 163 0.00 (0.05%) DME Strategy - Reduction of 21 0.0 0.0 -0.9 0.0 0.0 0.0 -18.02 benzene content of petrol to 1% DME Strategy - Reduction of 22 0.0 0.0 0.0 0.0 0.0 0.0 0.00 aromatics content of petrol to 35% DME Strategy - Reduction of S 25 content of diesel to <50 ppm -2 888 -23.0 -10.2 -4 360 -22 272 -4 116 495 0.00 (0.05%) DME Strategy - new passenger vehicles comply with Euro 4 27 -2 270 0.0 -2.9 0.0 0.0 -1 704 580 -48.9 standards (assume fuel specs changed) All petrol vehicles EURO 2 29 compliant (assume fuel spec -5 702 0.0 -7.1 0.0 0.0 -3 806 896 -122.8 changes in place) Conversion of 10% of petrol 30 -567 0.0 -0.8 0.0 0.0 -568 193 -12.5 vehicles to LPG Conversion of 20% of petrol 31 -1 135 0.0 -1.5 0.0 0.0 -1 136 387 -25.1 vehicles to LPG

161

No. 8 - Electrification

No. 24 - DME Strategy - Reduction of S content of diesel to <500 ppm (0.05%) & second grade diesel with 50 ppm S content available

No. 26 - DME Strategy - new passenger vehicles comply with Euro 2 standards (assume fuel specs changed)

No. 17 - Highveld Steel & Vanadium - replace coal use with CO use

No. 18 - Desulphurization of Sasol Secunda PS emissions

No. 32.2 - Electrification of paraffin-burning households – 10 years post

No. 16 - Iscor coke oven gas cleaning project

No. 32.1 - Electrification of paraffin-burning households – 1 year post

No. 23 - DME Strategy - Phasing out of lead

No. 19 - DME Strategy - Reduction of S content of petrol to 500 ppm (0.05%)

0 2000 4000 6000 8000 10000

Reduction in the Annual Number of Respiratory Hospital Admission

Figure 51: Reductions in the annual number of respiratory hospital admissions estimated due to interventions implementable by 2007

162

No. 3 - Low smoke fuels No. 7 - Housing insulation - 20% of all fuel burning households No. 29 - All petrol vehicles EURO 2 compliant (assume fuel spec changes in… No. 2 - Top down ignition - plateau roll out No. 15 - Emission reduction requirements for coal fired boilers for particulate… No. 1 - Top down ignition - DME ICHES([1]) No. 10 - Stove maintenance and replacement - 20% households all areas No. 5 - Housing insulation - 20% of plateau fuel burning households No. 25 - DME Strategy - Reduction of S content of diesel to <50 ppm (0.05%) No. 27 - DME Strategy - new passenger vehicles comply with Euro 4 standards… No. 6 - Housing insulation - 5% of all fuel burning households No. 11 - Desulphurization of all Power Station emissions No. 14 - RE technology implementation through financial incentives (37 000… No. 31 - Conversion of 20% of petrol vehicles to LPG No. 12 - Decommissioning of Pretoria West PS - gas use by households No. 9 - Stove maintenance and replacement - 5% households all areas No. 4 - Housing insulation - 5% of plateau fuel burning households No. 30 - Conversion of 10% of petrol vehicles to LPG No. 13 - RE technology implementation through financial incentives (10 000… No. 20 - DME Strategy - Reduction of S content of petrol to 50 ppm (0.05%) No. 22 - DME Strategy - Reduction of aromatics content of petrol to 35% No. 21 - DME Strategy - Reduction of benzene content of petrol to 1%

0 2000 4000 6000 8000 10000 12000 14000

Reduction in the Annual Number of Respiratory Hospital Admission

Figure 52: Reductions in the annual number of respiratory hospital admissions estimated due to interventions implementable by 2011

163

Interventions which may be given priority, from a health risk reduction perspective, are outlined in the Table 36. Based purely on health risk reductions achievable, interventions that target household fuel combustion and vehicles were found to be the most beneficial. Interventions aimed at reducing household fuel combustion resulted in the most significant reductions in respiratory hospital admissions and premature mortality. Various vehicular interventions were more effective for reducing cancer risks.

Table 36: Interventions to reduce emissions resulting in the most significant health risk reductions

Health Endpoint – Most Significant Sector Intervention Risk Reduction Low smoke fuel implementation(29) Respiratory hospital admissions, Domestic fuel Electrification of all unelectrified households chronic bronchitis, premature burning Large scale housing insulation mortality Large scale top down ignition roll-out Requirement of all petrol vehicles to comply with Euro 2 standards Requirement of new petrol vehicles to comply with Vehicles Euro 4 (Euro 2) standards Cancers Large scale conversion of petrol vehicles to LPG Restriction of benzene content of petrol to 1% Respiratory hospital admissions, Power Generation Desulphurization of all power station emissions premature mortality Industry, Respiratory hospital admissions, Restriction of particulate matter emissions from commercial & chronic bronchitis, premature coal-fired boiler operations service sector mortality

7.4 Cost-benefit Analysis of Interventions(30) The cost-optimisation of abatement measures requires identifying the optimal level of pollution abatement, when the private cost of abating an incremental unit of the pollutant equals the incremental damage done by it.

A cost-benefit analysis (CBA) of the various interventions was undertaken by researchers at the University of Cape Town (UCT, 2004; Leiman et al., 2007). The CBA focussed on health benefits arising due to the implementation of the interventions, as quantified in the current study, including reductions in morbidity and mortality. Attention was paid to medical costs saved, additions to days at work and increased labour productivity (due to reduced absenteeism). The aim of the work was to support policy development by ranking interventions according to their benefit/cost ratios.

29 The low smoke fuel intervention comprises making sufficient fuels available for the entire plateau household coal market at a cost equivalent to that of coal, coincident with the passing of legislation restricting the use of coal by households. 30 An economic analysis of the interventions identified, using as input the health risk and cost reductions quantified as part of the current study, was undertaken by the University of Cape Town and is comprehensively documented in Leiman et al. (2007). The findings of the economic analysis are integral to the recommendations and conclusions flowing from the author‟s research and are therefore briefly summarised in this section.

164

The study used financial and economic cost benefit analyses extended by employment impact studies and focused on the health care costs of air pollution in the designated urban areas(31). It did not take into account the wider costs of air pollution and the associated benefits of its reduction. Five of the thirty two different interventions for which health risk reductions were estimated could not be analysed due to incomplete or unreliable financial information, viz. Interventions 16, 17, 19, 24 and 28.

A further four interventions were excluded from the marginal benefit-cost analysis, since these interventions were already in the process of being introduced and hence their costs could be considered as “sunk costs” and excluded from the CBA. These interventions included numbers 12 (decommissioning of the Pretoria West Power Station), 26 (adoption of „Euro 2‟ technology for new passenger vehicles), 27 (adoption of „Euro 3‟) and 29 (all petrol vehicles „Euro 2‟ compliant).

The economic and financial net present value (NPV)(32) and economic benefit/cost (BC) ratio calculated by Leiman et al. (2007) for each of the interventions are summarised in Table 37. The benefit-cost ratio is the ratio of the monetary benefit in terms of savings in anticipated health costs, as a result of reduced air pollution levels attributable to the effective implementation of each intervention, to the cost of implementing the intervention.

Table 37: Net present values (NPV) and economic benefit/cost (BC) ratios of interventions (after Leiman et al., 2007)

Financial NPV Economic NPV Economic BC No. Sector Measure (Million 2003 (Million 2003 Ratio Rand) Rand) 33 1 Residential Top down ignition - DME ICHES( ) 756 654 177 2 Residential Top down ignition - plateau roll out 1 123 968 120 3 Residential Low smoke fuels -3 592 -2 914 0.4 Housing insulation - 5% of plateau fuel 4 Residential 263 226 6.0 burning households Housing insulation - 20% of plateau fuel 5 Residential 1 052 904 6.0 burning households Housing insulation - 5% of all fuel burning 6 Residential 426 368 7.9 households Housing insulation - 20% of all fuel 7 Residential 1 704 1 470 7.9 burning households 8 Residential Electrification 1 044 790 1.2 Stove maintenance and replacement - 5% 9 Residential 325 277 16.5 households all areas Stove maintenance and replacement - 20% 10 Residential 1 300 1 107 16.5 households all areas

31 The difference between the financial and economic results is that the financial analysis looks at monetary costs and benefits of the alternatives, while the economic analysis looks at the costs to society. This latter analysis is done by adjusting for shadow prices and wages and removing the distortions caused by taxes and subsidies (Leiman et al., 2007). 32 Net present value (NPV) is a standard method for the financial appraisal of long-term projects. It measures the excess or shortfall or cash flows, in present value terms once financing charges are met. 33 Department of Minerals and Energy (DME) Integrated Clean Household Energy Strategy (ICHES)

165

Financial NPV Economic NPV Economic BC No. Sector Measure (Million 2003 (Million 2003 Ratio Rand) Rand) Electricity Desulphurization of all Power Station 11 -15 446 -12 769 0.0 Generation emissions Electricity Decommissioning of Pretoria West Power 12 159 138 20.8 Generation Station - gas use by households Renewable energy technology Electricity 13 implementation through financial incentives -5 429 -4 485 0.3 Generation (10 000 GWh block) Renewable energy technology Electricity 14 implementation through financial incentives -6 341 -5 211 0.3 Generation (37 000 GWh block) Emission reduction requirements for coal 15 Industrial -191 -174 0.8 fired boilers for particulate matter (>90%) 16 Industrial Iscor coke oven gas cleaning project NQ NQ NQ Highveld Steel & Vanadium - replace coal 17 Industrial NQ NQ NQ use with CO use Desulphurization of Sasol Secunda Power 18 Industrial -1934 -1 593 0.1 Station emissions DME Strategy - Reduction of S content of 19 Transport NQ NQ NQ petrol to 500 ppm (0.05%) DME Strategy - Reduction of S content of 20 Transport -1 116 -946 0 petrol to 50 ppm (0.05%) DME Strategy - Reduction of benzene 21 Transport -1 094 -927 0 content of petrol to 1% DME Strategy - Reduction of aromatics 22 Transport -1 235 -1 051 0.1 content of petrol to 35% 23 Transport DME Strategy - Phasing out of lead 0 0 1 DME Strategy - Reduction of S content of 24 Transport diesel to <500 ppm (0.05%) & second grade NQ NQ NQ diesel with 50 ppm S content available DME Strategy - Reduction of S content of 25 Transport -442 -365 0.5 diesel to <50 ppm (0.05%) DME Strategy - new passenger vehicles 26 Transport 627 540 comply with Euro 2 standards DME Strategy - new passenger vehicles 27 Transport 420 361 comply with Euro 4 standards 28 Transport Taxi recapitalization project NQ NQ NQ All petrol vehicles EURO 2 compliant 29 Transport 1 054 907 (assume fuel spec changes in place) 30 Transport Conversion of 10% of petrol vehicles to LPG -2 383 -226 1 31 Transport Conversion of 20% of petrol vehicles to LPG -4 765 -451 1 32 Residential Electrification of paraffin burning households 0 999 1.3

The marginal benefit/cost ratios of the interventions analysed are depicted in Figure 53. A benefit/cost ratio of one is depicted as the horizontal line in the figure. Interventions with marginal BC ratios of greater than one are classified as economically justifiable in reducing the health care costs of air pollution. Of these, some offered large and immediate benefits, others smaller ones. From a policy perspective, it is recommendable that emphasis be placed on the early implementation of interventions with higher benefit/cost ratios.

166

Figure 53: Marginal net benefit-cost ratios calculated by Leiman et al. (2007) for each intervention

The main findings of the CBA are as follows (Leiman et al., 2007):  Interventions 1, 2, 4, 5, 6, 7, 8, 9 10, 12, 26, 27, 29 and 32 were estimated to have positive NPVs. Intervention 7 (insulation of 20% of households) had the highest positive economic NPV, followed by interventions 10 (stove maintenance and replacement for 20% of fuel burning households), 32 (electrification of paraffin burning households), 2 (implementation of top down ignition on plateau) and 29 (all petrol vehicles Euro 2 compliant).  Interventions 3, 11, 13, 14, 18, 20, 21, 22, 23, 25, 30 and 31 had negative NPVs. Intervention 11 (desulphurization of all power station emissions) had the highest negative economic NPV followed by 14 and 13 (comprising renewable energy implementation through financial incentives for 37 000 GWh block and 10 000 GWh block respectively), 3 (low smoke fuels) and 18 (desulphurization of Sasol Secunda power station emissions).  Most of the interventions with positive NPVs are those targeting individual households, whereas those with negative NPVs are generally industry based.  Ten of the interventions are likely to enhance net societal welfare, and these should be introduced in the order of their benefit cost ratios: i.e. 1, 2, 9, 10, 6, 7, 5, 4, 32 and 8. Such prioritized adoption of interventions would include all interventions with positive NPVs, and take into account ranking in terms of their economic efficiency.

Various sensitivity analyses were undertaken during the assessment, e.g. changes made to discount rates and assumptions regarding the mortality costing, with no significant changes in the calculated benefit-cost ratios projected.

7.5 Conclusions and Recommendations The author quantified emission reductions and resultant air quality improvements and health cost reductions achievable through the implementation of selected interventions for prioritised sources.

167

This data represented the primary input for the economic assessment of interventions undertaken by the University of Cape Town (UCT, 2004) and subsequently published by Leiman et al. (2007). Findings from the earlier research components undertaken by the author, and the outcomes from the economic assessment conducted by UCT, were subsequently synthesised by the NEDLAC Dirty Fuels Study Team co-led by the author, for communication to the NEDLAC Working Group (Bentley West Management Consultants & Airshed Planning Professionals, 2004).

The main findings of the externalities study for significant anthropogenic fuel combustion sources, documented in Chapters 4, 5, 6 and 7, are as follows:  Ranking of source significance should be based on the health effects, in terms of costs, rather than in terms of quantity of emissions.  Varied implementation options are available for facilitating or forcing intervention, such as regulation, market mechanisms and education. A combination of these needs to be considered.  From a financial and economic perspective, low (or existing) technology interventions in the domestic (household) sector can yield significant benefit in the short to medium term. Such interventions include Basa Njengo Magogo (a stove ignition method), stove maintenance, electrification and housing insulation.  There are a sufficient number of households in the domestic sector to allow for the implementation of multiple interventions without the risk of the deterioration of benefit/cost ratio of the identified interventions.  Low smoke fuels in the domestic sector, are not attractive in the short term, unless a significantly (20 to 30%) lower cost technology can be developed, rather than those technologies that are currently being considered.  The DME‟s proposed vehicle specification changes and related fuel specifications resulted in seven interventions (number 19 to 25) being considered to reduce the pollutant content of fuels. These interventions are costly to implement in the short-term but are certain to be phased in over time through the proliferation of vehicles meeting EURO2 and EURO4 standards.  Electricity generation interventions implementing high technology solutions on the supply side, such as desulphurization of power station emissions and renewable energy, are not feasible from a financial and economic perspective in the short- and medium-terms. To pursue desulphurization, considerable cost would be incurred by both Government and households. Renewable energy is not as costly as desulphurization but, relative to the costs of other interventions, is regarded as costly. Legislation has been enacted to progressively phase in renewable energy in the longer term.  The benefit, in terms of health effect of specific industrial interventions depends on scale, location and technology factors. Boiler emission reduction from coal fired boilers was found to be only marginally unfeasible, implying that a focused approach to improving the operating technology should be considered.  The bulk of savings due to reduced pollution from combustible fuels would go to government, primarily due to reduced spending in public health care.

The main conclusion drawn from the anthropogenic fuel-combustion externalities study is that significant health effect reductions can be cost-effectively achieved through addressing residential fuel burning as a priority. It is postulated that the lower benefit-cost ratios associated with

168

industrial and vehicular interventions are due in part to these sources having been more effectively regulated historically with less onerous actions already implemented. Industrial emissions have been regulated under the Atmospheric Pollution Prevention Act, Act 45 of 1965. Various improvements have been made in the quality of both petrol and diesel. Fuel burning within low- income households has continued largely unabated and therefore presents greater opportunities for cost-effective abatement.

The externalities study confirmed the importance of adopting a range of interventions, based on a variety of regulatory approaches including prescriptive methods, voluntary initiatives and market- based measures, to ensure economically efficient abatement. Furthermore, the cost-benefit analysis provided economic justification for redistribution policies such as the provision of subsidised electricity or LPG to the urban poor.

The finding that the most cost-effective mitigation measures in the short-term are those targeting household fuel burning should not detract for the need for effective management of industrial and vehicle emissions. Based on international experience and local vehicle activity data, it is evident that transport emissions are likely to become an increasingly significant contributor to urban air pollution. It should also be noted that effects related to ozone concentrations, with vehicle emissions being a major source of ozone precursors, were not accounted for in the study.

The externalities study did not take into account non-combustion related industrial emissions, nor regional and global scale influences associated with emissions. Industrial process emissions (unrelated to fuel burning) may include significant releases of criteria pollutants, in addition to trace releases of a wide range of hazardous air pollutants, including various metals and organic compounds. Elevated stack emissions from large scale power generation and industrial activities, whilst not impacting significantly on local ground level concentrations, may be more significant in (34)(35)(36) terms of regional effects such as acid deposition and tropospheric NOx densities . Industrial emission are therefore given further consideration, with the prioritisation and regulation of industrial emissions addressed in Chapter 8.

34 Josipovic (2010) conducted a critical loads assessment of acid deposition on the South African Highveld, and concluded that air pollution from acidic gases is not a current or medium term threat to regional ecosystems (beyond the central pollution source area) at current rates of emission (page vii).

35 Scorgie and Kornelius (2009) projected historical, recent and future acid deposition rates due to South African Highveld emissions to provide the basis for assessing the implications of long-term (1920 – 2020) deposition on soils, water and biodiversity. Results from this assessment have not yet been published in the open literature (as at December 2011).

36 Scorgie, Y., H.J. Annegarn, A. Richter, K. Ross, and L.W. Burger (2007). Comparison of SCIAMACHY NO2 Observations over Southern Africa with Air Quality Modelling Results, IEEE International Geoscience and Remote Sensing Symposium (IGARSS), Barcelona, Spain, 23 – 27 July 2007.

169

8 Regulation and Compliance Monitoring of Industry

Effective management of industrial emissions is a key component of air quality management. The South African Air Quality Act (AQA) of 2004 contains far reaching requirements for emissions management and control for the industrial sector. Given such requirements and the limited resources available for their enforcement, the practicability of managing industrial emissions successfully, and without adversely affecting the national economy and international competitiveness, warrants consideration.

The author conjectures that it is possible to devise an effective, affordable, equitable and sustainable emissions control policy for South Africa by adopting a combination of regulatory steps comprising phased national standards setting, compliance promotion and structured self-monitoring, market-based instruments, and a risk based enforcement strategy. During her tenure as a doctoral candidate, the author undertook a systematic investigation of international practices pertaining to the management of industrial emissions, and collaborated as a lead agent in several government task teams to device strategies to address this sector to meet the requirements of the AQA. Key components of the investigation comprised recommending an approach for national emission standard setting, and the development of a procedure for industrial compliance monitoring.

8.1 Background to Investigations Industries operating so-called „Scheduled Processes‟, deemed significant in terms of their atmospheric emissions, were historically required to hold Registration Certificates under the Atmospheric Pollution Prevention Act (APPA), Act 45 of 1965. This classic „demand and control‟ method of regulating industries through the issuing of permits by a regulatory authority will be continued under the National Environmental Management: Air Quality Act (AQA), Act 39 of 2004, however with a number of improvements. Under AQA, „Listed Activities‟ will be required to hold Atmospheric Emission Licenses, with one license being issued per site, and all emissions, including fugitive releases, being regulated through this comprehensive site licence.

The APPA to AQA transition phase required significant preparatory work, which included the following aspects:  Risk based prioritisation of industries for early Registration Certificate review;  Setting of national emission standards to inform the emission limits to be included in Atmospheric Emission Licenses; and  Development of procedures for monitoring compliance with such licenses during the implementation phase.

Based on the review of international experience and knowledge of local circumstances, the author developed and implemented procedures for prioritisation and compliance monitoring and made recommendations in respect of national emission standard setting. This work was undertaken by the author on behalf of the Department of Environmental Affairs (previously the Department of Environmental Affairs and Tourism) under the auspices of the APPA Registration Certificate Review; the National Emissions Standards Setting Project, commissioned by the Chief Directorate

170

(Air Quality and Climate Change); and the Compliance Monitoring Project commissioned by the Chief Directorate (Regulatory Services).

The risk-based prioritization of APPA Registration Certificates for review is addressed within Scorgie (2006) and is not covered in this thesis. Guidance developed in respect of the setting of national emission standards and the development of procedures for compliance monitoring are discussed in Section 8.2 and Section 8.3 respectively.

8.2 Guidance for National Emission Standard Setting A desk-top appraisal of relevant international information relating to the identification and classification of sources for which national emission standards are established, substances for which limits are specified and the basis for nationally-set emission standards was undertaken to inform the DEA‟s National Emission Standard Setting Process. The purpose of the review was the specification of clear and unambiguous recommendations of how information gathered may be adopted and/or adapted with a view to informing and fast-tracking the work required to implement Section 21 of the AQA.

8.2.1 Scope of Study Given the depth and range of international experience, the scope of the review was tailored to focus on selected, robust case studies pertinent to South Africa‟s circumstances and AQA implementation requirements, whilst also demonstrating potentially divergent approaches. Examples of local considerations include: the stack to site approach required by the APPA to AQA transition, the need to institutionalise emission standard setting and „keep it simple‟ so as to ensure timely implementation and the participatory governance requirements.

Three main case study countries were selected for inclusion in the study, namely, the United States, the United Kingdom (within the framework of the European Community) and Australia. The USA and UK were chosen because of the advanced nature of the environmental regulatory systems they have established and their proven track record in the development of best practice in this field. The USA and Australia have a highly federalised approach to regulation and this may have some lessons on cooperative governance which will be useful in the South African situation. A further reason for selecting the US, UK and Australia for analysis is that these countries were recently evaluated to support the development of compliance monitoring capabilities within the DEA‟s Regulatory Services Directorate (ref. Section 8.3). Reference was also made to specific aspects from the experience of other countries (e.g. India, Japan, Poland and China) which either illustrate alternative approaches or reinforce mainstream practices. In reviewing the countries selected, reference was made to a wide range of literature (Bernstein, 1993; Erbes, 1996; Hersh, 1996; Blackman and Harrington, 1998; Fenger et al., 1998; NEPC, 1998, 2001, 2003, 2004; DETR, 2000a, 2000b, 2000c; Poland Ministry of the Environment, 2000, 2003; Reitze, 2001; Nethconsult – BKH Consulting Engineers, 2002; DEFRA, 2003, 2005; Glicksman et al., 2003; Larssenn et al., 2003; Marbek Resource Consultants, 2003, AEAT Environment, 2004; Jagusiewicz, 2004; NSW DEC, 2004; Sengupta, 2004; US-EPA, 2005; Corden and Ritchie, 2006; OECD, 2006; Entec, 2006; World Bank, 2006; Gray et al., 2007; EC, 2007).

171

In the review of international practices and experiences close attention was paid to the following key issues:  Basis for identifying and classifying sources to be regulated using nationally-set emission standards;  Substances for which national emission standards are typically set and rationale for their selection;  Basis for the setting of national emission standards;  Divergent approaches in the implementation of national emission standards;  Future trends in industrial regulation which may influence the manner in which sources and substances are selected and emission standards set.

8.2.2 Regulatory Regime for Implementation of Standards Standards are used within command-and-control approaches to pollution control. It is important to note that emission standards are one category of a range of standards which are applied. A standard is a legally defined regulatory instrument for limiting pollution. The most commonly applied standards for the prevention of atmospheric emissions are (Bernstein, 1993):  Ambient air quality standards – limits the concentration of a pollutant in the ambient air (e.g. micrograms m-3) (indirectly determining the permissible emissions of a facility).  Emission standards – establishes the legal ceiling on the total quantity or concentration of a pollutant discharged from a pollution sources (e.g. mg m-3 in off-gas; grams per 24-hour period; kg per ton of raw material or product). Standards are typically expressed for a particular averaging period and monitoring requirements specified.  Technology-based standards – specifies the technology that should be used. For example, a facility may be required to use a scrubber to control sulphur oxide emissions with a control efficiency of at least 99%.  Performance standards – defines a performance measure (e.g. concentration of pollutant in off-gas and percent pollution removal to be achieved) and allows sources the flexibility to select the best means to meet this standard.  Product standard – establishes the legal ceiling on the total quantity or concentration of pollutants that can be emitted per unit of product output (e.g. kg per ton of product across total production cycle).  Process standard – limits emissions associated with a specific manufacturing process, e.g. mandatory replacement of mercury cells by diaphragm cells to prevent mercury emissions from chlor-alkali manufacture.

Given the requirement of Section 21 of the AQA that emission standards be specified for listed activities, the review focused on the use of emission standards in the various case studies selected. It is however notable that several other standards influence the extent of emissions permissible in such cases, particularly air quality standards and also product and performance standards in the case of the US.

Generally, emission standards are set by central governments although in some instances central governments may establish framework regulations but require local, state or regional authorities to set the standards. Sub-national standards are typically more stringent that those of the central government.

172

The setting of emission standards presupposes the existence of a monitoring agency that oversees polluters‟ activities and has the power to impose a penalty in instances of noncompliance. In the absence of enforcement powers, the only incentive the polluter has to stay within the standard is social conscience. Noncompliance with standards is typically associated with penalties (e.g. fee, loss of license). Polluters can also be prosecuted or at least threatened with prosecution in certain cases.

8.2.3 Comparison of International Approaches and Recommendation of their Implementation within South Africa In drawing on the experience of the case study countries to inform the South African process, it is useful to compare directly some aspects of the approaches in terms of targeting sources and setting emission standards, noting the most appropriate approach given local considerations.

8.2.3.1 General Observations in Respect of Emission Standards General observations made during the review of international case studies are as follows:  Emission standards are classifiable as a direct command and control instrument. The use of emission standards for prioritised industry sectors is widespread with almost all countries having adopted such a command and control approach as a primary means of controlling pollution.  Emission standards require a central authority capable of establishing rules for the conduct of polluting sources, monitoring performance with respect to those rules and enforcement of compliance.  Emission standards represent a method which tends to become more and more complicated over time e.g. US where layers of regulation have been added to deal with new problems and new information. Changes to the regulatory regime take much time and effort.  Emission standards represent just one instrument of a complex and increasingly broadening mixture of regulatory tools. Other instruments include fuel specifications, air quality limits and economic instruments.  Emission standard setting may result in „technology forcing‟. Although emission standards may not explicitly dictate technologies to be implemented, in practice they may create strong incentives for firms to choose only officially sanctioned technologies and can therefore be regarded as „technology forcing‟ (Blackman and Harrington, 1998).  Economic considerations are of considerable significance in the emission standard setting process. Key considerations in the US‟s historical approach have, for example, included avoiding the shutting down of plants, giving breaks to small facilities, or treating existing facilities more leniently than new ones. A number of recent studies have also considered the impact of environmental regulations on competitive of industries.

8.2.3.2 Adoption of a Phased Approach to Emission Standard Setting Various countries including the US, UK and India took a phased approach to national emission standards setting, gradually introducing emission standards for selected sectors. In the case of the US, new source performance standards (NSPS) have been progressively implemented since the 1990 Clean Air Act amendment (and the process is ongoing). The UK has progressively

173

introduced sectoral guidances containing emission limit values over the period 2001 to 2007 (frequently in line with the completion of the EC IPPC BREFs). India identified a block of 17 industrial sectors on which to focus attention on as a priority(37). The NSW state of Australia similarly identified 14 industrial sectors for the establishment of emission standards in its 2005 amendment of its Clean Air Regulations.

Although the AQA comprises many of the same underpinning principles of environmental governance as is characteristic of the case study countries, the manner in which Section 21 of AQA is written limits how closely South Africa can follow the examples of such countries. Most notable is the requirement that the Minister must list activities and simultaneously publish minimum emission standards and monitoring protocols in respect of substances resulting from such listed activities.

Consider a case whereby the Minister publishes a comprehensive „list of activities‟ and then follows the same rigorous and phased approach to emission standards setting as adopted by the US and UK. Considerable time (5 to 10 years) would be required to compile the necessary technical documentation and undertake sufficient stakeholder engagement to support the establishment of emission standards, monitoring protocols and compliance timeframes. This approach would clearly not meet the requirements of Section 21 of the AQA. Alternatively, should the Minister delay the publication of the list of activities pending the completion of the emission standard setting process, regulation of industry under the AQA would be unacceptably delayed.

Mindful of the requirements of the AQA, the experience of other countries, and of preliminary work undertaken during the APPA Registration Certification Review Project, a stepped approach to emission standard setting is recommended for adoption comprising the setting of emission standards for prioritised industry sectors and pollutants prior to the subsequent expansion of standards to other industries and substances. In outlining this approach, attention is paid to the procedure followed by the UK in its transition from Integrated Pollution Control (IPC) to Pollution Prevention and Control (PPC). This procedure is in line with South Africa‟s principles of transparency and stakeholder engagement and furthermore reflects to a large extent the approach (albeit more informally conceived and implemented) adopted by the DEA in its APPA Registration Certificate Review Project(38). The following steps are recommended for implementation:  Identification of key industries and associated pollutants for which emission standards are to be set at the outset.  Establish sector teams supported by a sector coordinator to carry out the consultation/ communication with industry, trade bodies and other affected parties.

37 An interesting difference in approach is that whereas the US, UK and NSW identified industry sectors on a national basis (state-wide basis in the case of NSW), India targeted industries within 24 areas characterised by significantly poor air quality. 38 South Africa adopted a “step change” rather than a phased approach in the transition from industrial regulation using registration certificates under the APPA to atmospheric emissions licenses under AQA. This step change comprised the conversion of existing registration certificates held by the several of the largest, most complex and potentially most polluting industrial sectors through the APPA Registration Certificate Review Project (initiated in January 2006 and scheduled to be completed by December 2007).

174

 Sector teams to conduct sector scoping studies – gathering information about its structure, geographical and size distribution and preferred methods of communication(39).  Sector teams to collate sector guidance documents comprising information on best available technology including associated emission standards and monitoring requirements (using international BAT documentation and industry-specific information.  Put in place mechanisms to support: • the addition of industry types to the list of activities by DEA; • emission standard setting for such industry types by DEA in consultation with stakeholders (via sector teams); • gathering of current BAT information for use in the establishment of emission standards for additional industry types and the review of previously established emission standards(40).  Intermittent additions to the list of activities and publication of relevant national emission standards for these activities.  Periodic review of national emission standards.

Benefits of the effective adoption and implementation of this approach include more focused regulatory efforts resulting in accelerated air quality improvements, development of an experienced regulator, and knowledgeable and cooperative regulated industrial sectors.

8.2.3.3 Industry Sectors and Source Types Targeted Within the case study countries emphasis is placed on cost-effective emissions reductions. This is primarily why larger industrial plants are targeted rather than smaller plants (e.g. enforcement and administration costs for regulation of 40 and 15 MW combustion plants are similar but costs relative to emission reduction potential are likely to be higher for smaller plant. Compliance cost for larger plants will probably not be significantly greater than for a smaller plant – but the difference in emissions could be significant.) (AEAT Environment, 2004).

Furthermore, if the definition of an industry type to be regulated is too broad or the threshold for coverage (e.g. power generation plants >50 MW), too low, too many installations may be included within the regulatory regime. This will have implications for identification of relevant plants, and costs of administration and enforcement.

The US, UK, NSW, India, Poland and China represent just some of the many countries who have focussed their efforts and resources by identifying and targeting key industrial sectors as a priority in the setting of emission standards. The rationale provided as the basis for selecting such sectors varies, but is essentially similar, e.g.:  „Major source‟ defined on basis of exceedance of a specified pollutant mass threshold (US);

39 Provision for stakeholder engagement was evident in the best practice case studies considered. In all three case studies, information was obtained from the regulated industry for inclusion in the technical background studies, industry and others (government agencies, wider community) were given opportunities to review and provide comment on drafts. Such stakeholder engagement and consultation was more extensive and intensive in the US and UK due to their focus on one sector at a time, their public hearing processes and phased approaches. 40 Best available techniques will change with time, particularly in the light of technical advances; the regulatory authorities must either monitor themselves or establish mechanisms to remain informed of such progress.

175

 Historically problematic polluter (UK);  Industry sector targeted by EU directives (UK, Poland);  Significant emitter based on emissions data from national emissions inventory (NSW); and  Source responsible for significant air quality degradation and health risks (India).

A comparison of the main sectors targeted to date by various countries, in relation to the proposed sectors proposed for designation as „listed activities‟ in South Africa is given in Table 38. Many similarities are apparent in terms of the industry sectors having been prioritised in terms of the setting of emission standards. In terms of identifying key sectors, it is also of interest to consider the most important industries in terms of environmental protection expenditure internationally, viz. chemicals, rubber and plastics, metal products, food, beverages and tobacco and pulp and paper (AEAT Environment, 2004).

Table 38: Synopsis of industries for which emission standards have been specified in the UK, US, NSW and India, classified according to the RSA ‘listed activities’ categories

UK PPC NSW Clean Air India's Initial RSA - Listed Regulation US NSPS Sectors Regulation 17 Target Activities Sectors Sectors Industries 1. Combustion Combustion Coal-Fired Electric Steam Generating Units Electricity Thermal Power installations Activities (coal, Electric Utility Steam Generating Units Generation Plant – Coal gas, biomass, Fossil-Fuel-Fired Steam Generators Based liquid fuels) Industrial-Commercial-Institutional Steam Thermal Power Generating Units Plant – Gas Based Onshore Natural Gas Processing: SO Emissions 2 Small Industrial-Commercial-Institutional Steam Generating Units Stationary Combustion Turbines Stationary Compression Ignition Internal Combustion Engines Stationary Gas Turbines 2.Petroleum Gasification, Bulk Gasoline Terminals Petrochemical Oil Refineries industry Liquefaction and Equipment Leaks of VOC in Petroleum Production Refining Refineries Petroleum Activities Petroleum Refineries Refining SED (Solvent Storage Vessels for Petroleum Liquids Emission Directive) Storage Vessels for Petroleum Liquids Activities VOC Emissions From Petroleum Refinery Wastewater Systems Volatile Organic Liquid Storage Vessels (Including Petroleum Liquid Storage Vessels) 3. Carbonisation Carbon Activities Carbon Black and coal Tar and Bitumen Industries gasification Activities Stand alone Coke Oven Plants

176

UK PPC NSW Clean Air India's Initial RSA - Listed Regulation US NSPS Sectors Regulation 17 Target Activities Sectors Sectors Industries 4. Metallurgical Ferrous Metals Ferroalloy Production Facilities Primary Aluminium industry Non-Ferrous Lead-Acid Battery Manufacturing Plants Aluminium Industries Metals Primary Aluminium Reduction Plants Production Copper Surface Treating Primary Copper Smelters Secondary Smelting Metals and Aluminium Lead Smelting Primary Emissions from Basic Oxygen Process Production Plastic Materials Furnaces Zinc Smelting Primary Iron Primary Lead Smelters Integrated Iron and Steel and Steel Primary Zinc Smelters Production

Secondary Brass and Bronze Production Plants Secondary Secondary Emissions from Basic Oxygen Iron and Steel Process Steelmaking Facilities Production Secondary Lead Smelters Primary Non- Steel Plants: Electric Arc Furnaces and Argon- Ferrous Oxygen Decarburization Vessels Production Steel Plants: Electric Arc Furnaces (excluding Aluminium) Secondary Non-ferrous Production (excluding Aluminium) 5. Mineral Cement and Lime Metallic Mineral Processing Plants Cement or Asbestos processing Production Non-metallic Mineral Processing Plants Lime Products industry Asbestos Calciners and Dryers in Mineral Industries Production Cement Activities and/or Industries Coal Preparation Plants Handling Glass and Glass Glass Manufacturing Plants Coal Mines Fibre Ceramic Coal Washeries Lime Manufacturing Plants Works Manufacturing Glass Industries Production of Phosphate Rock Plants Glass Portland Cement Plants Production Oil Drilling and Other Mineral Gas Extraction Fibres Industry Other Mineral Activities Ceramic Production 6. Organic Organic Equipment Leaks of VOC in the Synthetic chemical Chemicals Organic Chemicals Manufacturing Industry industry Storage of Synthetic Fiber Production Facilities Chemicals in Volatile Organic Compound (VOC) Emissions Bulk From Synthetic Organic Chemical Manufacturing Industry (SOCMI) Distillation Operations Volatile Organic Compound (VOC) Emissions from the Polymer Manufacturing Industry Volatile Organic Compound (VOC) Emissions From the Synthetic Organic Chemical Manufacturing Industry (SOCMI) Air Oxidation Unit Processes Volatile Organic Compound Emissions From Synthetic Organic Chemical Manufacturing Industry (SOCMI) Reactor Processes Rubber Tire Manufacturing Industry

177

UK PPC NSW Clean Air India's Initial RSA - Listed Regulation US NSPS Sectors Regulation 17 Target Activities Sectors Sectors Industries 7. Inorganic Inorganic Ammonium Sulfate Manufacture Agricultural Sulphuric Acid chemical Chemicals Nitric Acid Plants Fertilizer & Plants industry Storage of Phosphate Fertilizer Industry: Diammonium Ammonium Calcium Chemicals in Phosphate Plants Nitrate Carbide Plant Production Bulk Phosphate Fertilizer Industry: Granular Triple Nitric Acid Manufacturing Superphosphate Storage Facilities Plants Activities Phosphate Fertilizer Industry: Superphosphoric Involving Carbon Acid Plants Disulphide or Ammonia Phosphate Fertilizer Industry: Triple Superphosphate Plants

Phosphate Fertilizer Industry: Wet-Process Phosphoric Acid Plants Sulfuric Acid Plants Sulfuric Acid Production Units 8. Explosives Explosives Industry Production 9. Pharmaceutical Pharmaceuticals Production production 10. Incineration Disposal of Commercial and Industrial Solid Waste processes Waste by Incineration Units including Incineration Hospital/Medical/Infectious Waste Incinerators hazardous Incinerators waste Large Municipal Waste Combustors

Other Solid Waste Incineration Units

Small Municipal Waste Combustion Units 11. The Disposal of Municipal Solid Waste Landfills disposal of Waste by Landfill hazardous and Disposal of general waste Waste other than by Incineration or Landfill 12. Wood Paper, Pulp and Kraft Pulp Mills Paper, Paper products Board Pulp or Pulp industry Manufacturing Products Activities Industries Timber Activities 13. Production Plant Health and formulation Products and of pesticides Biocides 14. Animal Treatment of matter Animal and processing Vegetable Matter and Food Industries

178

UK PPC NSW Clean Air India's Initial RSA - Listed Regulation US NSPS Sectors Regulation 17 Target Activities Sectors Sectors Industries OTHER Coating Hot Mix Asphalt Facilities Activities, Asphalt Processing and Asphalt Roofing Printing and Manufacture Textile Automobile and Light Duty Truck Surface Treatments Coating Operations Manufacture of Beverage Can Surface Coating Industry Dyestuffs, Printing Ink and Flexible Vinyl and Urethane Coating and Coating Materials Printing Grain Elevators Activities Involving Rubber Graphic Arts Industry: Publication Rotogravure Printing Intensive Farming Industrial Surface Coating Production of Industrial Surface Coating Fuel from Waste Magnetic Tape Coating Facilities Recovery of Metal Coil Surface Coating Waste New Residential Wood Heaters Food Industries Petroleum Dry Cleaners Polymeric Coating of Supporting Substrates Facilities Pressure Sensitive Tape and Label Surface Coating Operations Surface Coating of Metal Furniture Sewage Treatment Plants Equipment Leaks of VOC From Onshore Natural Gas Processing Plants. Standard of Performance for Wool Fiberglass Insulation Manufacturing Plants

Emission standards are set primarily for point sources such as stacks and vents. In instances where fugitive releases are regulated using emission standards, it is required that the activity be undertaken in an enclosure with extraction, with the emission limit being set for the extraction vent. In other instances, control or management measures may be specified for diffuse emissions to reduce the potential for emissions. This is most frequently done in the case of controlling VOC emissions from chemical handling and storage. More general requirements are given for the control of fugitive dust releases, e.g. requirement that operators prepare and implement a dust management and maintenance plan for the site. More detailed information is expected to be included in the individual permits of facilities.

Based on the experience of other countries, and mindful of South Africa‟s available resources and the nature of its industrial sector, the following recommendations are made:  Industry types which could be considered for possible inclusion in the initial list of activities requiring prioritised national emission standard setting are listed in Table 39, with industries not significantly represented within South Africa removed or noted for subsequent listing and emission standard setting. Thresholds specified by various countries are given in brackets, where available, and key activities highlighted in the table.

179

Table 39: Industry types proposed for inclusion as Listed Activities in South Africa and summary of related industry sectors prioritised by other jurisdictions

RSA –Listed Activity Categories Synopsis of Internationally Prioritised Industry Sectors (Based on UK PPC, US NSPS, NSW, India) 1. Combustion installations  Coal, gas, biomass and liquid fuel combustion installations (>50 MW – UK; >30MW – NSW; >73MW - US)  Waste or recovered oil combustion (>3 MW - UK) 2.Petroleum industry  Petrochemical production and petroleum refining (including bulk storage and handling of petroleum liquids and petroleum refinery wastewater systems) (UK – no thresholds; NSW – 2000tpa petrochemicals) 3. Carbonisation and coal  Coal gasification gasification  Gas refining (>1000tpa gas - UK)  Natural gas reforming  Mineral oil refining  Activities involving pyrolysis, carbonisation, distillation, liquefaction, partial oxidation or other heat treatment of coal, lignite, oil, other carbonaceous materials or mixtures  Tar and bitumen production (>5tpd tar, bitumen or aggregate - UK) 4. Metallurgical industry  Aluminium and aluminium alloys  Iron and steel production  Copper smelters (melting capacity >20 tpd - UK)  Lead smelters (melting capacity >4 tpd - UK)  Zinc smelters (melting capacity >20 tpd - UK)  Precious metals production  Refractory metal production  Nickel processes  Cadmium processes (melting capacity >4 tpd - UK)  Ferroalloy production (silicon, chromium, manganese)  Ferrous metals (hot rolling) (>20 tph crude steel - UK)  Bulk handling or storage of iron ore (except during mining)(>500 000t - UK)  Lead-acid battery manufacturing (>6.5 tpd lead – US)  Secondary Brass and Bronze Production Plants (Reverberatory and electric furnaces of >1 000 kg production capacity and blast (cupola) furnaces of >250 kg/h production capacity – US) 5. Mineral processing industry  Cement and lime production and/or bulk handling (kilns >50tpd; 5000tpa calcium carbonate, calcium magnesium carbonate or aggregate of both - UK)  Asbestos activities  Glass and glass fibre manufacturing (>100tpa production – UK; >5tpd - US)  Ceramic production (tiles, bricks, refractory bricks, stoneware, porcelain production by firing) (kiln >75 tpd - UK)(NSW threshold is 150tpd or 30000tpa)  Coal processing/preparation plants (500tpd coal – NSW; >200tpd - US)  Metallic mineral processing plants (crushing, screening, handling)  Non-metallic mineral processing plants (crushing, screening, handling)  Phosphate rock plants (>4tph plant capacity – US)  Other mineral activities (melting capacity >20 tpd - UK) 6. Organic chemical industry  Organic chemical production including: o hydrocarbons, o organic compounds containing oxygen, sulphur, nitrogen or phosphorus, organometallic compounds (e.g. lead alkyls) o plastic materials (polymers, synthetic fibres, cellulose-based fibres) o synthetic rubbers o dyes and pigments o surface-active agents  Polymerising or co-polymerising any unsaturated hydrocarbon or vinyl chloride (>50tpd in aggregate - UK)  Use of toluene fi-isocyanate or other di-isocyanate of comparable volatility or where partly polymerised  Flame bonding of polyurethane foams or polyurethane elastomers  Recovery or purifying of acrylic acid or any ester of acrylic acid  Tyre manufacture (>50 000 tpa - UK)  Storage of chemicals in bulk

180

RSA –Listed Activity Categories Synopsis of Internationally Prioritised Industry Sectors (Based on UK PPC, US NSPS, NSW, India) 7. Inorganic chemical industry  Production of inorganic chemicals such as: o Gases (e.g. NH3, HCl, HF, H2S, SOx, NOx) o Acids (e.g. chromic acid, hydrofluoric acid, nitric acid, sulphuric acid, oleum) o Bases (e.g. ammonium hydroxide, sodium hydroxide) o Salts (e.g. ammonium chloride, sodium carbonate) o Non-metals, metal oxides, metal carbonyls o Halogens or interhalogen compounds o Manufacturing Activities Involving  Manufacturing activity involving the use of hydrogen cyanide or hydrogen sulphide  Manufacturing activity involving the use or recovery of: antimony, arsenic, beryllium, gallium, indium, lead, palladium, platinum, selenium, tellurium, thallium  Recovery of any compound of cadmium or mercury  Chemical fertilizer production (20000tpa - NSW)  Bulk storage of chemicals  Key activities in this sector are nitric acid plants, sulphuric acid plants, agricultural fertilizer production and ammonium sulphate & ammonium nitrate production 8. Explosives Industry  Explosives production 9. Pharmaceuticals production  Pharmaceutical production using a chemical or biological process 10. Incineration processes  Commercial and industrial waste incineration including hazardous waste  Hospital/Medical/Infectious waste incineration  Municipal waste incineration 11. The disposal of hazardous  Hazardous waste disposal facilities and general waste  General waste disposal facilities (>10tpd or >25000t total capacity - UK)  Disposal of Waste other than by incineration or landfill (>10tpd for hazardous waste and waste oils; >50tpd for non-hazardous waste – UK) 12. Wood products industry  Paper, pulp and board manufacturing activities (>20tpd – UK; >30 000tpa - NSW)  Timber processing plants 13. Production and formulation  Pesticides, fungicides, herbicides, rodenticides, fumigants, miticides and related product of pesticides production (NSW – 2000tpa products) 14. Animal matter processing  Tanning plants (>12tpd finished products - UK)  Animal slaughter (>50tpd - UK)  Rendering plants - animal carcasses or waste disposing or recycling (>10tpd – UK; >5000tpa - NSW)

 Consider whether to extend the proposed categories of listed activities to include the industry sectors listed in Table 40. These categories need not be added to the initial list of activities, but can be added during subsequent phases. The above activities are regulated under national/state emission standards in more developed countries. Various of these activities are however recommended for regulation as „controlled emitters‟ under the AQA (Hietkamp and Nkhwashu, 2005). Paying attention to the experience of countries such as the US and UK, but also heeding India‟s experience with its regulation of small and medium enterprises, it is recommended that further consideration be given to whether the proposed listed activities need to be expanded. However, it is proposed that such consideration be extended once more experience has been gained with the regulation, compliance monitoring and enforcement of the initial group of „listed activities‟ and the first one or two „controlled emitters‟.

181

Table 40: Potential additional categories to be considered in South Africa as Listed Activities

Possible additional  Industrial and surface coating activities categories of listed  Textile manufacture activities  Printing works (large scale)

 Intensive farming – rearing poultry (40 000 places - UK), pigs (2000 places – UK & NSW), cattle (1000 head –NSW), sheep (4000 head – NSW), horses (400 – NSW)  Recovery of waste including fuel production from waste  Food Industries – treating and processing animal raw materials (other than milk)(>75 tpd - UK) or vegetable raw materials (>300tpd - UK) or milk (>200 tpd - UK)  Hot Mix Asphalt Facilities  Sewage Treatment Plants

 Express emission standards primarily for point sources (stacks and vents) where emission monitoring is possible. Where the control of diffuse emissions is considered significant enough to warrant inclusion in national standards (e.g. fugitive dust at bulk ore/coal handling and processing plants and certain metallurgical industries; evaporative emissions from bulk chemical storage and handling), specific best practice control measures applicable across individual industries can be stipulated (e.g. floating roof tanks), or alternatively it can be required that a comprehensive fugitive emission management plan be put in place. In the latter case, the operator of the facility would be expected to demonstrate to the relevant regulatory authority that all necessary measures had been taken to minimise fugitive releases, and undertake monitoring to demonstrate continued control of its fugitives. More detailed controls or requirements should be retained for inclusion in the individual emissions license of the activity.

Reasons for limiting the initial list of activities to specific industry types which are known to be potentially significant in terms of their atmospheric emissions include:  Selection of known significant emitters based on available information, given that a comprehensive national emission inventory is not available for source selection as is used elsewhere (e.g. US, NSW);  Reduce the workload of the regulatory authority during the learning phase;  Allow more experienced industries (in terms of previous regulation, BAT.) to go through the initial phase first;  Bring more potentially polluting industries under the new regulatory regime earlier than less polluting industrial sectors;  In the absence of a comprehensive cost-benefit analysis, target industries where the benefits of regulation are expected (based on international experience) to outweigh the costs; and  Reduced risk of the initial list of activities being contested.

8.2.3.4 Prioritisation of Pollutants Substances have been prioritised for national standard setting on various grounds, the most common of which include:

 Pollutants contributing to widespread health risk, either directly (e.g. PM, SO2) and/or due to

their being precursors of significant pollutants (NOx and VOCs as precursors of ozone)

182

 Pollutants resulting in acidification (SOx, NOx)  Persistent pollutants (mercury, dioxins/furans).

The most widely regulated substances are criteria pollutants (SO2, NOx, PM, CO, opacity as surrogate for PM), total VOCs, total organic compounds (TOC), specific metals (lead, mercury, arsenic, cadmium, antimony, chromium, beryllium, manganese, nickel, selenium, vanadium), hydrogen chloride, hydrogen fluoride, total reduced sulphur (TRS), sulphur trioxide and sulphuric acid mist, polychlorinated dibenzodioxins and polychlorinated dibenzofurans (total or TEQ), asbestos and cyanides.

The tendency within the EC and the US is to concentrate on key pollutants of concern, rather than trying to target all possible emissions. The exact number of pollutants to target however ranges significantly, e.g. from the restriction of limits within New Source Performance Standards to a handful of pollutants by the US to the significant number regulated by Poland.

National emission standards are not routinely issued for greenhouse gas emissions. This is likely to be due to the increased use of market mechanisms such as emissions trading to cost-optimise emission reductions.

It is recommended that a small number (preferably 1 to 4) of pollutants be selected for the setting of emission standards for each industry type selected (with the exception of incineration for which an extended number of substances should be regulated in line with current local and international experience). Reference can be made to the pollutants regulated under the US, UK and NSW approaches to select the most suitable pollutants to target. Where appropriate, use could be made of surrogate parameters to reduce compliance monitoring costs.

8.2.3.5 BAT as a Basis for Emission Standards It is commonplace in best practice legislative environments to require that emission standards take into account best available technologies (BAT) and ambient air quality limits. In practice, minimum nationally-set emission standards tend to be based on best available technology, with the requirement that more stringent emission standards be set at lower tiers of government taking into account air quality limits. In addition to this, the use of environmental impact assessments for informing emission standards for new and modified facilities is widely accepted. This provides a safety net in cases where minimum emission standards based on BAT are not sufficiently protective of local environments.

Given that provision is made in the AQA for the setting of more stringent emission standards by provincial and local authorities, it is recommended that the national minimum emission standards be based on best available technology.

Best available technology, despite being defined in slightly different terms (or not defined at all in regulation as in the case of NSW), is implemented in similar ways in the case studies considered. It is recommended that South Africa adopt the concise EU definition of BAT, viz. (EU, 1996):

“Best available techniques` shall mean the most effective and advanced stage in the development of activities and their methods of operation which indicate the practical suitability of particular techniques for providing in principle the basis for emission limit values

183

designed to prevent and, where that is not practicable, generally to reduce emissions and the impact on the environment as a whole:  'techniques‟ shall include both the technology used and the way in which the installation is designed, built, maintained, operated and decommissioned,  'available‟ techniques shall mean those developed on a scale which allows implementation in the relevant industrial sector, under economically and technically viable conditions, taking into consideration the costs and advantages, whether or not the techniques are used or produced inside the [country] in question, as long as they are reasonably accessible to the operator,  'best‟ shall mean most effective in achieving a high general level of protection of the environment as a whole.” (European Integrated Pollution Prevention and Control Directive 96/61/EC)”

In the application of BAT for the purpose of informing emissions standards and monitoring protocols for the prioritized industry types, reference could be made to the best practice documentation published by the IPPC, UK and US. In assessing the economic viability of technologies within local industries, the simpler approach adopted by NSW could be considered whereby use is made of previous studies undertaken and information provided by industries within the sector to be regulated.

8.2.3.6 Format of Expressing Emission Standards Emission standards should not prescribe the use of one specific technique of technology (technology forcing). This has been demonstrated to be the least cost-effective aspects of historical command-and-control systems and has been shown in numerous examples to suppress technological innovation (Bernstein, 1993; Blackman & Harrington, 1998; Glicksman et al., 2003; AEAT Environment, 2004; Entec, 2006). Emission standards can be expressed in one or more of the following formats:  Emission concentrations (e.g. mg of a pollutant per m3) including and excluding volumetric flow rates permissible (e.g. m3 h-1);  Total mass (e.g. tonnes per annum, kg per day)  Emission rates (e.g. g s-1)  As a performance standard (kg pollutant per ton of raw material; kg pollutant per ton final product).

An example of the nature of emission standards issued by the US, UK and Australian NSW for the Glass Manufacturing Industry is given in Table 41 to demonstrate the differences in the manner in which such limits are expressed internationally.

The AQA stipulates that emission standards “must include the permissible amount, volume, emission rate or concentration of that substance or mixture of substances that may be emitted and the manner in which measurements must be carried out”. This requirement in the Act largely developed as a result of the manner in which emission standards have historically been specified within APPA Registration Certificates (i.e. typically as emission concentrations without limits on volumetric flows or on total masses of emissions). The specification of a total mass as a „permissible amount‟ or a „volume‟ in a general national minimum emission standard intended to regulate a number of individual industries may not be sufficient for effective management.

184

Table 41: Comparison of the nature of emission standards issued by the US, UK and Australian NSW for the Glass Manufacturing Industry

Pollutants Processes for which Emission Standards Emission Country Threshold for Inclusion Monitoring Requirement Regulated Are Given Standards Format

US NSPS Any facility constructed or modified PM, opacity Emission standards specified according to Emission intensity (g Performance tests based on after June 15, 1979. fuel used (gas, liquid) and glass of particulate/kg of ASTM methods for PM and Does not apply to hand glass manufacturing plant segments: container glass produced) direct measurement or mass melting furnaces, glass melting glass; pressed and blown glass (including balance for glass quantity furnaces designed to produce less borosilicate recipes, soda-lime and lead estimates than 4.55 Mg (5 tons) of glass per recipes, other), wool fibreglass, flat glass day and all-electric melters.

NSW Clean Glass manufactured through a firing PM, NOx, Type 1, Emission standards given for melting furnace Emission Typically continuous for Air process Type 2, Cd, Hg, (PM emission standards also given for concentrations criteria pollutants and Regulation smoke material crushing, grinding and handling) campaign for others 2002 Emission standards specified for various ages of plant UK Sector Manufacturing glass fibre or PM, SOx, NOx, Emission standards given for furnace Emission Typically continuous for Guidance manufacturing glass frit or enamel HCl, HF, As, Co, operations and for downstream processes for concentrations criteria pollutants and annual frit (100 tpa or more) (A1 Ni, Se, Ammonia, flat glass, container glass and domestic glass for others, e.g. metals installation) VOC, total metals, sections Manufacturing glass where melting capacity of plant 20 tpd or more (A2 installation).

Notes: Type 1 substance means the elements antimony, arsenic, cadmium, lead or mercury or any compound containing one or more of those elements. Type 2 substance means the elements beryllium, chromium, cobalt, manganese, nickel, selenium, tin or vanadium or any compound containing one or more of those elements.

185

It is recommended that minimum emission standards be expressed either as an emission concentration or a performance standard or, where appropriate, a combination of both, with the actual concentration or level of performance taken from BAT. Additionally, total masses of emissions permissible should be included in the Atmospheric Emissions Licenses of Listed Activities.

8.2.3.7 Specification of General Emission Standards Certain countries (e.g. Australia-NSW, China) specify general emission standards for application to industries for which sector-specific emission standards are not applicable. Taking into account the recommendation that a select list of industry types be prioritised for the setting of specific emission standards, South Africa could consider the use of general emission standards for application to industries which are not initially listed.

8.2.3.8 Emission Monitoring Requirements The emission monitoring required depends on the nature of the source, the pollutant and the emission standard.

Emission standards expressed as emission concentrations require direct stack monitoring. The sector-specific monitoring method and frequency should be taken from the best practice documentation, e.g. EU Integrated Pollution Prevention and Control BAT Reference documents or IPPC BREFs(41). In most cases, continuous emissions monitoring is prescribed for the larger sources of criteria pollutants as is typically best practice, with periodic (e.g. annual) testing campaigns stipulated e.g. for metals and persistent organic compounds. Emission standards expressed as a performance standard (e.g. kg of pollutant per ton product) requires a combination of direct monitoring and product tonnage tracking methods.

8.2.3.9 Variation of Emission Standards within Industry Sectors The setting (retention) of less stringent emission standards for older facilities has a place in the regulatory process of most of the countries considered. (It was found to be more pronounced in the case of the US and NSW, compared to the UK). It is however notable that these emission standards are not static, but that there are timeframes within which facilities are expected to meet firmer standards. Generally, the approach adopted is to link required improvements to major plant modifications and to take advantage of industry cycles. This is most readily expressed in the NSW regulations where older plants are given five year timeframes to institute upgrades which will bring them in line with more stringent emission standards.

Whereas the US tends to include the dates of facilities within individual industry specific standards, NSW sets out clear industry facility age categories which are applicable across all industry sectors regulated. The NSW approach is simple to understand, lends itself to being more readily used to stipulate cross-sector continuous improvement requirements, and can be more easily revised. This approach is recommended for possible implementation within South Africa.

41 http://eippcb.jrc.es/reference/

186

8.2.3.10 Compliance Schedules Compliance schedules may be specified in various ways. They may be generically specified for an entire industry sector or class of facilities (class defined by industry type and age of facility). Alternatively they can be negotiated and imposed at individual facilities by provisions within permits such as the Atmospheric Emission License to be issued to Listed Activities within South Africa.

Based on international experience, an effective approach would be to set minimum timeframes for compliance nationally (taking account of industry cycles), with provision being made for more restricted compliance schedules to be specified by lower government tiers for industries within their jurisdictions and/or stricter timetables being negotiated for inclusion in permits. Typical compliance timeframes, based on the US, EC and NSW case studies would be:  2 to 3 year in the case of new or substantially modified facilities  5 to 10 years in the case of existing facilities, potentially differentiated by age.

8.2.3.11 Cost-benefit analysis The case study of India clearly demonstrated that best practice international experience must be adjusted to the structure of a nation‟s economy. The assessment of available technologies enhanced by sector-wide economic analysis is a useful instrument for establishing the techno- economic viability of the prescribed standards.

Cost-benefit analyses of the implications of introducing new or revised emission standards have routinely been undertaken by countries such as the UK, US and NSW(42). Comprehensive approaches have included the costing of externalities and assessments of the extent to which cost savings (e.g. health spending reductions) due to emission reduction could offset the costs of implementing BAT to achieve the required limits.

Given the short timeframe within which the Minister is expected to publish a „list of activities‟ so as to meet the APPA to AQA transitional phase objectives, it is unlikely that detailed sector- specific cost-benefit analysis will be completed in time to inform the initial listing of activities. It is therefore recommended that the initial list of activities comprise a restricted number of industry types that are known to be potentially significant in terms of their atmospheric emissions. The targeting of industries where the benefits of regulation are expected to outweigh the costs, based on experience from developed and developing countries would substantially reduce the risks of economic effects arising due to the emission standards set. Additional measures to reduce risk during this initial phase include: (i) restricting pollutants for which emission standards are specified to the key ones for that industry type thus reducing compliance monitoring and reporting costs; (ii) taking industry cycles into account in the setting of national minimum compliance timeframes, and (iii) making provision for industries to apply for extensions based on their having completed environmental impact assessments which demonstrate that their emissions are not causing adverse effects.

42 The Regulatory Impact Statement undertaken in 2004 by the NSW government in support of its proposed emission standards provides a good example of how prior studies and information from the industry sectors to be regulated could be used in the costing of outcomes (NSW DEC, 2004).

187

In targeting industry sectors for which information on emissions and effects is less available or conclusive, particularly those comprising small and/or older operations, it is imperative that detailed cost/benefit analyses be undertaken in selecting best available technologies and setting emission standards. Provision for such studies should be made so as to extend the list of activities and associated set of national minimum emission standards in a manner which does not lead to economic effects or mass non-compliance.

8.2.3.12 Provision for Extensions to Compliance Timeframes Given potential economic implications of emission standards, and mindful that emission standard setting in South Africa is not likely to be based on comprehensive sector-based cost-benefit analysis (at least not for the initial group of „listed activities‟), it is recommended that provision be made for specific industries to apply for possible extensions to compliance timeframes. In framing this provision reference is made to a similar condition set by the NSW DECC in its Clean Air Regulation 2002 (2005 amendment). The DECC makes it clear that it does not intend that existing plants be „unnecessarily or arbitrarily required to upgrade‟ as a result of its regulations.

Within the South African context, it is recommended that a provision be included when listing activities, for the proponent of a listed activity to apply for a postponement of the compliance date and for such a postponement to be granted based on the following conditions being met:  An air pollution impact assessment completed (in accordance of the format for Atmospheric Impact Reports, as contemplated in Section 30 of the AQA and specified by the National Air Quality Officer) and submitted to DEA at least one year before the compliance date; and  Demonstration that the industry‟s air emissions are not causing any adverse effects on the surrounding population or environment. This provision would ensure that any requirement to upgrade is informed by an understanding of any environmental effect of the affected plant. At the end of the extension period granted a further extension could be made possible subject to a repeat of the impact assessment process.

8.2.3.13 Emission Standard Implementation In implementing emission standards, best practice necessitates comprehensive compliance monitoring and enforcement functions and the regular review of such standards in line with BAT developments.

8.2.3.14 Integration of Emissions Trading with Traditional Emission Regulation Market-based mechanisms, particularly emissions trading, are widely used in the US to reduce emissions and meet air quality standards. Key emission trading schemes have included the national Acid Rain Program (ARP), regional cap and trade programs such as the Ozone Transport

Commission (OTC) NOx Budget Program (1999-2002), the NOx Budget Trading Program (NBP) under the NOx State Implementation Call (2003-2008), the Clean Air Interstate Rule (CAIR) NOx ozone season trading program (May 2009 onwards), and the Regional Clean Air Incentives Market (RECLAIM) (Los Angeles, California).

Flexibility, promotion of technological innovation and cost-optimisation of emission reduction represent some of the potential advantages of emission trading schemes. The disadvantage of emissions trading schemes is that they may result in „hot spots‟ if not effectively implemented. Based on the US experience it was demonstrated that the potential for „hot spots‟ could be

188

minimised by the implementation of air quality limits and the requirement that sources comply with best available control technology (BACT). The feasibility of hybrid command and control schemes with emission trading within the US regulatory system has been established.

The EU supports the integration of regulation across media and does not provide for criteria air pollutant emissions trading despite recognising the important role such market mechanisms play in individual member states (43). In its review of the impact of the environmental legislation within the EU and other countries on the competitiveness of European industry, AEAT Environment (2004) noted that the US‟s increased use of market-based instruments would result in a lower cost approach for US industry. The implication being that in future, diverging policies with regard to industry regulation could negatively affect the competitiveness of industries in countries where cost-optimised emission reductions were not being secured by the implementation of mixed packages of regulatory tools and market mechanisms.

During 2010 the EC Director General Environment investigated whether and how a NOx/SO2 emissions trading scheme could be applied as a replacement to the BAT-based permitting system implemented under the Industrial Emissions Directive (Directive 2010/75/EC)( 44). Studies were conducted to assess the environmental and economic effects of emissions trading, and the overall societal costs and benefits of European-wide trading (Entec, 2010a, 2010b). The findings of these studies indicated that a trading scheme for these two pollutants could be more cost-effective than the BAT-based permitting system, but also reveal a number of important drawbacks. Implementation of the 2010 IED requires prompt investment decisions by industry to comply with stricter BAT requirements. By establishing an emissions trading instrument, there would be a period of uncertainty and delay implementation of the IED. The emission reduction objectives of the EU would therefore be undermined. Furthermore, there are concerns that the trading systems would result in local level effects hindering the achievement of the EU air quality objectives. The (45) EC therefore decided not to pursue the establishment of a NOx/SO2 emissions trading scheme .

The AQA makes provision for the use of market mechanisms including emissions trading. Given the success achieved by the US emissions trading systems, the potential extension of the regulation of criteria pollutants emitted by listed activities through the marrying of emission standard setting and emission trading approaches warrants investigation. Reference could be made to the studies undertaken in 2010 for the proposed European NOx/SO2 emissions trading scheme, to identify potential challenges requiring attention.

8.3 Compliance Monitoring Compliance monitoring forms a vital component of the environmental governance cycle (Figure 54). Laws and regulations are often ineffective if they are not properly enforced, and proper enforcement can only be effective if compliance monitoring takes place. The AQA makes

43 The EU Emission Trading Scheme currently makes provision for emission trading of greenhouse gases across over 10 000 individual industries situated within member states in support of it meeting its Kyoto Protocol greenhouse gas emission reduction requirements. 44 The IED directive is the successor of the 2008 IPPC Directive. 45 http://ec.europa.eu/environment/air/pollutants/stationary/studies.htm

189

provision for various tools for compliance monitoring and provides for the use of Environmental Management Inspectors to carry out compliance monitoring.

Figure 54: The environmental governance cycle (DEAT, 2007b)

The Directorate Compliance Monitoring was established within the Regulatory Services Chief Directorate of the DEA to monitor the compliance with prioritised environmental quality and protection legislation, regulations, authorizations and applied enforcement instruments and to develop and oversee national systems for such monitoring. A project was commissioned to develop the structures, systems and capacity required by the Directorate to fulfill this role. This project commenced in October 2005 and was completed during the first quarter of 2008. During her tenure as a doctoral candidate, the author was engaged as a member of the project team specifically for the purpose of providing guidance on compliance monitoring in relation to atmospheric emission related authorisations, e.g. Atmospheric Emission Licenses. The author‟s work comprised the following main components:  Review of international approaches to compliance monitoring to identify and describe areas of „best practice‟, with focus on five case study countries selected by the project team in consultation with the DEA, namely: the US, UK, Poland, India and the NSW state of Australia.  During the review of international practice, consideration was given to the underpinning philosophy and theory, the institutional structure, design and management, intergovernmental relationships (where more than one tier of government is involved), systems, processes and procedures, and information management processes and technology employed.  Drafting of a Guideline Inspection Protocol for use in compliance monitoring of Listed Activities holding Atmospheric Emission Licenses under the AQA.

190

In assessing the suitability of international approaches, close attention was paid to local circumstances and the tailoring of methods to suite such circumstances. Key outcomes of the compliance monitoring project are presented in subsequent subsections.

8.3.1 Key Findings on Best Practice Compliance Monitoring

8.3.1.1 General findings Institutional Design  Compliance monitoring should be considered as only one element in a larger regulatory cycle or circle (as depicted in Figure 54). All elements of this cycle must work effectively if successful environmental regulation is to be achieved. If one element is weak, this will undermine the effectiveness of the others.  Institutional form tends to follow function – i.e. in most countries where the importance of the regulatory cycle as a whole is understood, the institutions tasked with its implementation carry out all of its elements, and its structures are designed to facilitate the integration, communication and feedback required to make the cycle work as a whole. Compliance monitoring is not planned or implemented in isolation from the other functions of a regulatory department or institution.

Role of the operator  There is a general trend towards the use of „self monitoring‟ by operators – which is formalised within permit conditions, regulations.  Operators increasingly can „earn the right‟ to less frequent inspections by improving their performance.

Role of the „regulator‟  Regulators recognise that resources do not stretch to carrying out inspections of all facilities with authorisations and so are developing approaches for planning inspection frequencies and prioritisation of inspection activity on particular facilities. A risk-based approach (based on the risk to the environment or safety posed by a facility) is an emerging approach and one which is successfully applied in the UK and the USA.  Compliance promotion is considered as important a role of the regulator as traditional compliance monitoring and enforcement.  Regulators are increasingly being expected to raise revenues (through regulatory charges) to cover the costs of their activities.  With increasing reliance on self-monitoring, the role of inspectorates becomes one of checking and verification of information supplied by industry,  However, self-regulation and other trends (such as the use of third party contractors to carry out inspections) requires the regulators to provide comprehensive guidance information up- front to ensure that processes are followed correctly and expectations of the regulator met.  There is an increasing demand for regulatory authorities to be clear and transparent in what they do, and to actively provide access to information they collect.  There is a trend towards insistence on „performance management systems‟ for regulators and a growing body of literature on how to design and implement these.

191

Compliance inspections – procedures and protocols  A substantial body of information is available from international and regional networks to guide inspectors in the ethics, approaches and procedures to be followed. Much of this information is generic (i.e. is applicable regardless of the media being inspected or the country the inspection is being carried out in) and is thus applicable and useful in the South African context.

Compliance monitoring procedures for specific media  Most jurisdictions use the same compliance monitoring units and procedures for monitoring compliance with all authorisations regardless of media, i.e. there is rarely a separate institution or procedure to be followed for waste and air.

8.3.1.2 Country-specific observations United States and Australia Both the United States and the Australian NSW have Environmental Agencies that set up complex monitoring frameworks that are implemented at the state level. The National Environmental Departments set the policies and the framework and provide assistance and capacity where it is needed. Each State takes the monitoring framework and develops its systems to suit individual requirements, but feed information through to the national level. The national level also spends the majority of its time involved in compliance promotion and assistance (provision of advice, guidelines, policy, procedures and so on).

Compliance Monitoring is largely inspection based, and it is the US-EPA regional offices which undertake periodic inspections. The US-EPA also has a compliance incentive scheme to those organisations that do comply, and offers compliance assistance in terms of publications or websites that will provide industry with information on how to comply with regulations.

The US has lengthy procedures which may be modified to fit the South African situation, however it is highly developed and complex.

The United Kingdom The EU has set highly structured Directives for pollution control and waste management, which the UK is adopting as is necessary.

The UK has a specific risk scoring system in place (e.g. Operator and Polluter Risk Appraisal or OPRA for Waste) with well-defined procedures. It is facility based and ranks both the level of risk of the waste type and the procedure in place to limit the risk (operator performance). OPRA is linked to a Compliance Classification Scheme that classifies the level of non-compliance.

The Environment Agency has developed an approach known as „modern regulation‟ and documented its approach. Modern Regulation is based on principles such as transparency, accountability and proportionality (risk based). The development of such a model and vision for overall environmental regulation highlights the importance of thinking of compliance monitoring in South Africa in the context of an overall system, rather than as a separate and isolated activity.

192

A major advantage of the UK approach is that it is partly self-funded, as an annual licence fee is placed on all facilities (to fund inspections) and those facilities that have the highest risk in terms of the OPRA score, pay the highest fees. Inspections are carried out by the national department. Waste operators are able to interact with the system through the website and are able to check their own OPRA scores. The major disadvantage of the system is that it only monitors registered and licensed facilities and not illegal operations.

This system would best suit the South African example, except that inspections should be carried out at provincial level. The procedures for site scoring in terms of OPRA are readily available(46).

Poland (and other emerging European Economies) Many of the lessons from these countries are related to the weaknesses and problems in environmental inspection institutions that can arise when governments lack capacity and a complete suite of sound environmental legislation. However, these countries are also interesting as they are all in a period of transition and are actively building new environmental management institutions within their governments. Organisations like the OECD have done a large amount of work in preparing documents to assist them in this process. These documents will be of great value to DEA throughout the development of its own systems and structures.

In terms of waste specifically, Poland is currently restructuring its waste management system to meet the EU standards. They have a compliance and enforcement unit that looks at general environmental compliance and does not single out waste as a separate entity. However they have ranked their top 200 polluters (industries) that are targeted for inspections and monitoring. Industries are then forced to comply and get their names de-listed off the top 200 list as it greatly affects their marketing and branding.

India India has a central pollution control board and similar structures in each of its states. States may then decide whether to further decentralise regulatory activity to local bodies.

Despite the presence of these systems and some moves towards using regulatory approaches found in the developed world (such as self-regulation by operators), India‟s regulatory system is plagued by many of the problems associated with developing countries: lack of capacity, overlapping and fragmented institutions, lack of standardised procedures, and lack of public involvement in decisions.

For example, India has yet to develop regulatory standards for waste management and therefore there has been no structured compliance monitoring for waste management. The importance of enforcement of waste management regulations has been recognised with the fast growing economy, but the development of well-engineered landfill sites has been given priority. The US Asian Environmental Programme is assisting in developing a waste management system that will ultimately culminate in a compliance monitoring system. It is also involved more generally in

46 www.defra.gov.uk/emvironment/airquality/riskam/05.htm

193

trying to improve the environmental regulatory system. Developments and literature that come from these processes may be useful to DEA when they are produced.

Inspections for waste management facilities are usually only carried out once facilities are opened and during emergency incidents. This is often the case in South Africa as well, and is a reflection of an unstructured monitoring system.

8.3.1.3 Key Best Practices in Air-related Compliance Monitoring Emphasis placed on compliance assistance and promotion Compliance promotion and assistance is considered imperative by the US-EPA, UK Environment Agency and other regulatory bodies to ensure high levels of compliance. Providing access to legislative, regulatory and technical information relevant to air pollution to a wide range of audiences, and sector specific information on clean technologies, is emphasised.

Sources of information for air pollution related compliance monitoring The most widely used sources of information for air pollution related compliance monitoring in the US, UK, Australia and elsewhere include: (i) recordkeeping and reporting requirements imposed on the source; (ii) citizen reporting; and (iii) monitoring by the government from inside and outside the regulated facility.

Experience gained by these countries has shown:  Ambient air quality monitoring plays no significant role in enforcement efforts.  Citizen reporting is sporadic and unreliable, only useful as a supplemental monitoring approach.  Information received from sources needs to be carefully checked for integrity, but represents an increasingly significant information source for compliance monitoring purposes (due to technological advances and changes in the compliance regulations of countries).  Inspection remains the backbone of compliance monitoring in most countries, but is increasingly being supplemented by information from sources (i.e. self-monitoring and reporting).

Guidance on self-monitoring, -record keeping and -reporting The trend towards self-monitoring and reporting is not only critical for cost-effective compliance monitoring but is also in line with the „polluter pays‟ principle with the onus for compliance demonstrations being for the polluter‟s budget. The following guidance is provided for self- monitoring by sources and checking by authorities based on the experience of other countries:  Authorities responsible for compliance monitoring must establish clear and robust source and emission requirements, and monitoring and compliance reporting protocols. (Provision for such is made in the AQA, but requires implementation in practice with source and emission requirements, and monitoring protocols being stipulated for „listed activities‟ and „controlled emitters‟.)  Cost-optimisation of monitoring requirements by making use of surrogate measures.  Integration of measurement uncertainties into compliance assessment decision making, ranking sources as „compliant‟, „non-compliant‟ or „borderline‟ (if within the margin of uncertainty) and taking actions accordingly (e.g. routine reviews for compliant sources,

194

voluntary improvements for borderline sources and enforcement actions for non-compliant sources).  Promotion of air quality management systems as part of environmental management systems and of environmental audit protocols.

Targeting of inspections and definition of different types of inspections Inspections are very resource intensive. In response to this, countries (including India, Poland, US and UK):  Target their inspection activities to address priority sources, and sources operating in air pollution „hot spot‟ areas; and  Define different types of inspections ranging from simple „walk through‟ visits to detailed investigations, including stack monitoring by regulators.

The classification of inspection requirements (e.g. frequency of inspection, depth of inspection or type of inspection) based on source characteristics (including the magnitude of a source‟s emissions, compliance history, contribution to ambient air pollution) would give authorities the best return for their investment.

Standardisation of procedures for compliance monitoring (including self-monitoring) For compliance monitoring to be effective and fair it is considered imperative that standard procedures be developed and applied. Including procedures for:  Inspections by the various tiers of government;  Self-monitoring, recordkeeping and reporting by sources. Furthermore, the US promotes uniformity with respect to the providing of exemptions to sources (through the Applicability Determination Index).

The above approaches would be of significant benefit in South Africa where the atmospheric emissions licensing function (and presumably associated compliance monitoring) is decentralised.

Place of national inspections of sources for compliance monitoring purposes Based on the experience of other countries, specifically those who have decentralised responsibility for compliance monitoring, it may be concluded that national inspections for compliance monitoring purposes would be beneficial in the following instances:  Ordinarily no routine inspections take place (but there is a need for such inspections given complaints, emissions or effects occurring as a result of a source);  Oversight inspections, where national authorities joint provincial or local authorities at their request;  Independent inspections as part of a National Priority Area Air Quality Management Plan implementation or part of a specific national enforcement process;  Inspections to confirm compliance with nationally issued notices;  Inspections to receive or provide field training for inspectors; and  Joint inspections with provincial and local authorities for strategic industries, industries with very poor compliance histories and/or industries impacting very significantly on the environment.

195

Given the designation of responsibilities under the AQA it would be prudent for national compliance monitoring authorities to consult and coordinate with provincial and local authorities before performing an inspection within their jurisdictions.

Under the AQA, the DEA has initiated the holding of regular national and provincial air quality officer meetings. The former comprises input from national and provincial air quality officers, whereas provincial and local air quality officers are involved in the latter. Such meetings would be a pertinent place to address the coordination of stationary source compliance monitoring from an air pollution perspective.

Fair, proportionate and cost-effective air pollution control Countries implementing „good practice‟ air quality management emphasise the importance of regulating all significant sources taking into account their contribution to ambient air pollutant concentrations, particularly in areas of high exposure. The regulation of all significant sources is perceived to be the only way to achieve and maintain air quality limits in a cost-effective, fair and proportionate way. Although this approach is supported by the South African AQA care will need to be taken to ensure that the approach is implemented in practice. The establishment of authorisation, compliance monitoring and enforcement systems suitable for the regulation of industrial operations will not address air pollution problems in areas significantly impacted by, for example, vehicle exhaust emissions or household fuel burning releases.

The US, UK, India and the Australian NSW undertake holistic air quality management planning to quantify and address all significant sources, including diffuse sources such as household-related emissions, wildfires and vehicle emissions through compliance promotion and monitoring initiatives. Furthermore, in the regulation of industrial and other sources countries try to identify and implement cost-optimised emission reduction measures. In countries such as Poland, where enterprises have small profit margins, it has become apparent that regulatory failure could occur due to compliance monitoring being made to costly.

Importance of Inspection and Maintenance Programmes for the regulation of vehicles South Africa‟s vehicle emission reduction strategy comprises emission limits for new vehicles and specifications for the improvement of fuels through the reduction in lead, sulphur, benzene and aromatics contents of fuels. Such improvements are reported by other countries to be unsuccessful unless accompanied by comprehensive inspection and maintenance programmes for vehicles. The nature of such programmes differ between countries ranging from ad hoc roadside tailpipe monitoring by authorities to routine vehicle exhaust testing at testing and inspection stations (linked to vehicle license). Such stations may be owned and operated by regulatory authorities or by a third party. The US-AID (2004) publication on international experience and best practices for vehicle inspection and maintenance programs provides useful guidance in determining the most suitable approach for South Africa.

Assessment of baseline air quality prior to emission limit and ambient air quality limit setting Ambient air quality limits typically comprise the following criteria: (i) threshold levels – concentration below which air pollutant concentration should be maintained; (ii) associated averaging periods, e.g. annual average, highest daily, highest hourly; (iii) permissible frequencies of exceedance; and (iv) timeframes within which compliance must be achieved. The thresholds

196

and associated averaging periods are typically specified based on dose-response thresholds noted for human health and/or the broader environment. Permissible frequencies of exceedance and compliance timeframes are however informed by knowledge of the existing air quality and of sources contributing to air pollutant concentrations. The trend is to gradually manage down the permissible frequencies and progressively reduce air pollutant concentrations to within thresholds.

In instances where the existing air pollutant concentrations have not been characterised, significant sources established and timeframes for emission reductions considered, air quality limits are much less likely to be achieved. The importance of baseline air quality characterisation and emissions inventory development as the basis for air quality standard setting (and action planning) is addressed in the EC‟s position papers supporting the EC air quality limits.

In setting emission limits in the UK, attention is paid to emission concentrations possible through best practice implementation in addition to ensuring that resultant air pollutant concentrations are within acceptable ambient air quality limits. The implications of this for regulators and the regulated sector are as follows:  Regulators – need to set emission limits based on BAT and air quality limits;  Sources need to demonstrate compliance with air quality limits (comprises compliance assurance and demonstrations through the use of air dispersion modelling in combination with measurement).

Benefits of emission inventories, compliance data bases and open information Publicly available and accurate information on atmospheric emissions is believed to contribute to cleaner production practices by industry. Such practices are also in line with the principles of „access to air quality information‟ and „transparency‟ upheld by the AQA.

Programmes have been initiated by the US, Canada and a number of OECD countries to assess the impact of open information on industry behaviour. In the US, the total on- and off-site releases of Toxic Release Inventory (TRI) chemicals declined by 46% between 1988 and 1996 with reductions occurring for nearly every industrial sector and chemical. The public access to facility-specific information and the industry‟s interest in reducing inefficient use and demonstrating responsible citizenship were cited as crucial factors in obtaining these reductions.

International experience also shows that companies are often not aware of their emission rates until they are required to prepare and report such estimates for emission inventory activities. This can alert them to areas of inefficient or wasteful practice resulting in the company being encouraged to cut wastage to avoid costs and consequently reducing environmental harm.

8.3.2 General Inspection Protocol The general inspection protocol developed by the author, integrating lessons learned from the various international case studies considered, is included as Appendix J. This protocol was compiled on behalf of the DEA for use in assessing the extent to which Listed Activities comply with their Atmospheric Emission Licences under the AQA.

197

8.4 Synthesis of Study Outcomes Direct regulation remains an important instrument in the control of industrial sources internationally, even given the implementation of market mechanisms and other measures. Although the South African AQA provides for direct regulation of industries, resource constraints and the practicability of implementing requirements successfully whilst not adversely impacting the national economy, pose potential obstacles.

To support the overall hypothesis of the dissertation, the author conjectures that it is possible to devise an effective, affordable, equitable and sustainable emissions control policy for South Africa by adopting a combination of regulatory steps comprising national standard setting, compliance promotion and structured self-monitoring and use of market-based instruments. To this end, key outcomes were presented from studies undertaken by the author as part of an overall systematic investigation of international practices pertaining to industrial emissions control. The focus of these studies comprised:

 recommended approach for national emission standard setting, and  development of a procedure for industrial compliance monitoring.

The extent to which the assertion is addressed by the findings of the above studies, and the extent to which study outcomes have already been adopted by government, are considered here.

8.4.1 Phased National Emission Standard Setting A review of international approaches to national emission standard setting was undertaken by the author to inform the DEA‟s National Emission Standard Setting Process. The specific aim of the review was to provide clear and unambiguous recommendations of how information gathered may be adopted and/or adapted to inform and fast-track the work required to implement Section 21 of the AQA. Detailed reviews were conducted of US, UK and Australian approaches, with reference also made to experiences from India, Japan, Poland and China.

Key recommendations arising from the review with regard to the national emission standard setting process were as follows (Scorgie and Kornelius, 2007):  Adoption of a phased approach to emission standard setting, with initial setting of emission standards for prioritised industry sectors and pollutants prior to the subsequent expansion of standards to other industries and substances.  Establishment of industry sector teams to carry out consultation/communication with industry, trade bodies and other affected parties in respect of standard setting. Sector teams to gathering sector-specific information on existing practices, best available technology (BAT), BAT-based emission standards and monitoring requirements.  Establishment of mechanisms to support the addition of industry types and associated emission standards in consultation with stakeholders.  Industry types recommended for inclusion in the initial list of activities requiring prioritised national emission standard setting as follows: • Combustion installations • Petroleum industry • Carbonisation and coal gasification • Metallurgical industry

198

• Mineral processing industry • Organic chemical industry • Inorganic chemical industry • Explosives Industry • Pharmaceuticals production • Incineration processes including hazardous waste • The disposal of hazardous and general waste • Wood products industry • Production and formulation of pesticides • Animal matter processing  Production thresholds should be included in the definition of industry types to be covered by emission standards, with smaller operations initially excluded.  Consideration of whether to extend the proposed categories of listed activities to include the following industry sectors: industrial surface coating activities, textile manufacture, printing works (large scale), intensive farming (poultry, pigs, cattle, sheep, horses), recovery of waste including fuel production from waste, food industries – treating and processing animal raw materials or vegetable raw materials, hot mix asphalt facilities, sewage treatment plants.  Control of diffuse emissions significant enough to warrant inclusion in national standards (e.g. fugitive dust at bulk ore/coal handling and processing plants and certain metallurgical industries; evaporative emissions from bulk chemical storage and handling) by specifying applicable best practice control measures or requiring a comprehensive fugitive emission management plan.  Restrict emission standard setting to key pollutants of concern, rather than trying to target all possible emissions with reference made to US, UK and NSW standards in selecting the most suitable substances to target for each industry type.  Set national emission standards based on emission levels achievable through the application of best available technology (BAT), with the requirement that more stringent emission standards be set at lower tiers of government taking into account ambient air quality standards. Reference is made to the EU definition of BAT which accounts for technical and economic viability of measures.  Express emission standards either as an emission concentration or a performance standard or, where appropriate, a combination of both. Total masses of emissions permissible should be included in the Atmospheric Emissions Licenses of Listed Activities.  Consider specifying general emission standards for application to industries for which sector- specific emission standards are not applicable.  Emission standards expressed as emission concentrations require direct stack monitoring. Sector-specific monitoring method and frequency should be taken from the best practice documentation. Continuous emissions monitoring is typically prescribed for the larger sources of criteria pollutants, with periodic (e.g. annual) testing campaigns stipulated for metals, persistent organic compounds (etc.).  Emission standards should be varied to account for the age of facilities. The setting (retention) of less stringent emission standards for older facilities is a common practice. However such emission standards are not static, with timeframes given within which facilities are expected to meet firmer standards.

199

 Compliance schedules (i.e. timeframe for meeting standards) should be informed by industry cycles and take into account industry type and age of facility. Typical compliance timeframes are 2 to 3 years for new or substantially modified facilities, and 5 to 10 years for existing facilities, potentially differentiated by age.  Apply cost-benefit analysis to inform the future listing of activities and specification of BAT- based emission standards.  Given potential economic implications of emission standards, and mindful that emission standard setting in South Africa is not likely to be based on comprehensive sector-based cost- benefit analysis (at least not for the initial group of „listed activities‟), it is recommended that provision be made for specific industries to apply for possible extensions to compliance timeframes.  Implementation of emission standards requires effective compliance monitoring and enforcement functions and the regular review of standards in line with BAT developments.  The AQA makes provision for market mechanisms including emissions trading. It is recommended that the potential for extending and enhancing the regulation of criteria pollutants emitted from listed activities through the integration of emission standard setting and emission trading approaches be investigated. The Minister of Environmental Affairs published the national List of Activities and its associated national emission standards on 31 March 2010 (DEA, 2010a). The published list of activities and standards reflect the recommendations of Scorgie and Kornelius (2007) in the following ways:  A phased approach was adopted, with standards being published for the most significant industrial activities and addressing a selected priority pollutants.  Industry types recommended for inclusion (Table 42) were adopted as listed activities with the exception of explosives, pharmaceutical and pesticide production.

 The development of emission standards was supported by consultation undertaken within working groups focusing on specific industry sectors (sector teams). Working group meeting minutes and submissions received by such groups by industry, trade bodies and other affected parties are documented on the SAAQIS website(47).

 Production thresholds were included in the definition of industry types to be covered by emission standards, with smaller operations initially excluded.  Emission standards are expressed as emission concentrations rather than total mass.  Different emission standards are published for new and existing plant. Reference is made to best practice measures in the setting of emission standards for new plant, with less stringent standards being initially set for existing plant.  Provision is made for continuous emissions monitoring for compliance demonstration by large emitters of priority pollutants.

 47 South African Air Quality Information System Website, http://www.saaqis.org.za/Links.aspx

200

 Progressive compliance timeframes are established, with existing plant being given 5 years to comply with standards issued for existing plant and 10 years to comply with standards for new plant.

 Provision for the potential postponement of the compliance timeframes on a case-by-case basis pending an Atmospheric Impact Report being completed, as per Section 30 of the AQA.  A fugitive emissions management plan is required to be included in the Atmospheric Emission Licenses for listed activities that are likely to generate such emissions. The publication of Listed Activities and associated national emission standards is regarded by the National Air Quality Officer as a milestone event that will significantly assist both the regulated community and the regulator in ensuring a „level-playing field‟ and a nationally-consistent approach to significant industrial emission regulation (DEA, 2010b, pp. 6-7).

Modelling of the potential effect of the minimum emission standards in the Highveld Priority Area indicated that these minimum standards will have important air quality benefits, and may result in full compliance with air quality standards in some areas (DEA, 2010b, pp. 6-7).

Recognising that the Listed Activities and associated national emission standards published in March 2010 require further improvement, work is on-going by the DEA. During 2011, this work involved drafting accreditation requirements for stack emission monitoring and sampling, and the review and revision of Listed Activity category descriptions where necessary to ensure that sources are clearly and unambiguously identified.

8.4.2 Compliance Monitoring of Industry Laws and regulations are often ineffective if they are not properly enforced and proper enforcement can only be effective if compliance monitoring takes place. The AQA makes provision for various tools for compliance monitoring and provides for the use of Environmental Management Inspectors (EMIs) to carry out compliance monitoring.

As part of a larger project to assist the DEA‟s Compliance Monitoring Directorate with the development of structures, systems and capacity, the author undertook a review of international „best practice‟ approaches to air-related compliance monitoring. Reference was made to compliance monitoring practices within the US, UK, Poland, India and the Australian NSW state. Key best practice approaches for compliance monitoring of industrial sources were noted to include the following:

 Compliance assistance and promotion to ensure high levels of compliance.  Inspection remains the backbone of compliance monitoring in most countries, but increasingly supplemented by information from sources (i.e. self-monitoring and reporting).  Trend towards self-monitoring and reporting is critical for cost-effective compliance monitoring and in line with the „polluter pays‟ principle.  Authorities responsible for compliance monitoring must establish clear and robust source and emission requirements and monitoring and compliance reporting protocols.  Compliance monitoring requirements may be cost-optimised by making use of surrogate measures.

201

 Measurement uncertainties must be integrated into compliance assessment decision making, ranking sources as „compliant‟, „non-compliant‟ or „borderline‟ (if within the margin of uncertainty) and taking actions accordingly (e.g. routine reviews for compliant sources, voluntary improvements for borderline sources and enforcement actions for non-compliant sources).  Air pollution management systems are increasingly promoted as part of environmental management systems and environmental audit protocols as a means of tackling and demonstrating compliance.  Inspections are very resource intensive and have resulted in the following practices: • Targeting of inspection activities to address priority sources, and sources operating in air pollution „hot spot‟ areas; and • Definition of different types of inspections ranging from simple „walk through‟ visits to detailed investigations including stack monitoring by regulators. The classification of inspection requirements (e.g. frequency of inspection, depth of inspection / type of inspection) based on source characteristics (including the magnitude of a source‟s emissions, compliance history, contribution to ambient air pollution) would give authorities the best return for their investment.

Based on the review of international practices and considerations of local circumstances a Guideline Inspection Protocol for use in compliance monitoring of Listed Activities holding Atmospheric Emission Licenses under the AQA was developed by the author.

Widespread designation of EMIs was projected to take place in 2011, resulting in a significant increase in the compliance and enforcement capacity for the AQA. The National Air Quality Officer reported that the DEA was on track to meet the following targets (DEA, 2010b):

 1200 EMIs designated by 2011/2012 (from a baseline of 900 in 2007)  200 EMIs trained in air quality compliance monitoring by 2011/2012 (244 EMIs were reported to have been trained by 2011); and  100 formal compliance monitoring inspections per year by 2011/2012. The guideline inspection protocol developed is applicable for use by all tiers of government and includes a classification scheme to prioritise industries for inspection so as to streamline and cost- optimise compliance monitoring.

8.4.3 Conclusion and Outlook Based on the systematic investigation of international practices, and experience gained in tailoring selected practices for local implementation, the feasibility of devising an effective, affordable, equitable and sustainable emission control policy for South Africa has partially been demonstrated. Despite detailed cost-benefit analyses not having been completed, recommended approaches have by and large been adopted by the DEA following inter-governmental reviews and extensive consultation with interested and affected parties.

Further evaluation of the effectiveness of the emission control policy will only be possible once emission standards, compliance monitoring and risk-based enforcement measures are fully implemented and integrated with non-regulatory measures including market-based measures.

202

9 Phased Air Quality Management Planning

In her capacity as advisor to local government on AQMP development, the author developed an in depth understanding of government structures and functions, and local capabilities in respect of air pollution control, emissions inventory development, and air quality monitoring and modelling. This understanding, together with ongoing analysis of international practices, allowed the author to devise solutions to address challenges faced in effecting a change in the governance of air quality management.

Key challenges to effecting decentralised, receiving environment air quality management in South Africa, and solutions derived to address such challenges, are outlined in this chapter. The author also reflects on whether such solutions were successful and considers the usefulness of innovations derived for future urban air quality management under the Air Quality Act.

9.1 Background Under the Atmospheric Pollution Prevention Act of 1965, air pollution control was largely restricted to national government implementing source-based controls within the industrial sector. The Constitution of South Africa, coming into effect in 1997, introduced a new mandate by stipulating that air pollution is a local government matter and therefore has to be managed by Municipalities. Legislation supporting this new mandate came in the form of the National Environmental Management: Air Quality Act of 2004 and final repeal of the APPA in 2009. Thus for several years local government had a constitutional directive to manage air pollution but no clear legislative framework nor local precedent. Furthermore, the information, tools and experience required for effecting air quality management were largely lacking.

The City of Johannesburg set a precedent in 2002 by commissioning the development of an Air Quality Management Plan to fulfil its constitutional responsibilities and address health risks due to poor air quality. The author led the development of the City of Johannesburg AQMP and the subsequent Ekurhuleni Metropolitan Municipality AQMP, and acted as advisor to the City of Cape Town during the in-house development of its AQMP (Scorgie et al., 2003c, 2004f, 2005; Scorgie, 2004b). These initial AQMPs, which represented adaptations of plans developed by jurisdictions abroad, set the precedent for the development of further AQMPs both prior to and following the Air Quality Act coming into effect. These AQMPs also substantially influenced the Manual for Air Quality Management Planning published by DEA in 2008.

9.2 Overview of Air Quality Management Planning To provide the context for discussing air quality management planning related challenges and innovations, an overview of the planning process is presented. For brevity, international air quality management practices are not documented. Reference should be made to comprehensive reviews of such practices published elsewhere (e.g. Annegarn and Scorgie, 1997a, 1997b, 1998; Beattie et al., 2000; Scorgie et al., 2003d; Milieu Ltd, 2004; Engelbrecht, 2006; Scorgie and Kornelius, 2007). Lessons learned internationally with regard to the tailoring of AQM planning to meet sustainable development objectives are however explored in Section 0, and are fundamental to the guidance provided in this chapter for AQMP development.

203

The air quality management planning process is illustrated in Figure 55. The air quality policy provides the framework for determining the air quality objectives to be adopted, the manner in which air quality is to be assessed, the types of interventions to be considered, and manner in which such interventions are selected and implemented. Air quality objectives reflect the quality of air considered acceptable in terms of protecting human health and, in some cases, the broader environment.

Figure 55: Air quality management planning process (Own figure)

Emissions inventory, air quality and meteorological monitoring, and dispersion modelling are the main tools used to assess baseline air quality, project future air quality given „business as usual‟ and undertake source contribution analysis. Evaluation against air quality objectives forms the basis for determining whether the area needs to be brought into compliance or managed to prevent future non-compliance with air quality objectives. In cases where current and future compliance is projected, it could be desirable to manage sources progressively to achieve cleaner air.

Action planning comprises the evaluation and selection of control options or interventions for significant sources aimed at meeting air quality objectives, maintaining air quality within objectives or progressively achieving further improvements in air quality within a given timeframe. During the implementation of interventions changes in air quality is tracked to assess the efficacy of interventions.

Cooperative governance, information dissemination and availability, and public and stakeholder consultation represent critical components throughout the air quality management planning process.

9.3 Challenges to AQMP Development and Implementation The main challenges identified by the author during the inaugural AQMP development processes undertaken within metropolitan municipalities were as follows:

204

 Absence of air quality management policy frameworks and the need to strengthen executive management commitment to effective air quality management.  Absence of a clear and current documented set of national air quality standards to provide the basis for the effects-based air quality management approach.  Fragmented monitoring efforts, with no comprehensive and integrated air quality monitoring and reporting systems in place to adequately characterise baseline air quality and rank the significance of sources and pollutants.  Meeting political and public calls for early (short term) emission reduction actions while taking into account uncertainties regarding the significance of sources and pollutants.  Inadequate resources allocated for air quality management, specifically with regard to personnel.  Absence of experience in and mechanisms for cooperative governance for the purpose of effecting air quality management. Cooperative governance was particularly critical given that sources were historically regulated by various government departments at different tiers of government, and given the need to integrate air quality into development planning.  Absence of experience in and mechanisms for information dissemination and public consultation as part of air quality management processes.  Lack of experience in integrating air quality considerations into land use, energy and transportation planning.  Lack of experience in the harmonisation of urban air quality management planning policies with policies and programmes addressing regional and global air quality challenges such as regional ozone and climate change.

Despite the promulgation and roll out of the AQA, most of the challenges outlined above continue to hinder air quality management, Approaches adopted to overcome the challenges identified during the inaugural AQMP development process are described in subsequent sections and reference made to lessons drawn from the experience of other countries where applicable.

9.4 Air Quality Management Policy Air quality management planning is most effectively done within a supportive policy framework. An air quality management policy should reflect the principles underpinning key legislative instruments such as the National Environmental Management Act and local environmental policies and commitments. Such principles include accountability, environmental justice, full cost accounting, freedom of information, polluter pays, good governance, public participation and integrated planning and environmental management.

Air quality management policies comprise vision statements and strategic goals and objectives which provide the context within which AQMPs are developed and implemented. The following priorities may, for example, be reflected in the Vision Statement:  The need to protect human health and welfare, the broader biophysical environment (ecosystems and the built environment) and the global environment,  The importance of ecologically sustainable development,  A commitment to the selection and implementation of best practices and technologies, and  A commitment to integrated management and planning.

205

Specific goals and objectives to be met to achieve such priorities may then be specified and an overview given of key aspects of the air quality management approach to be adopted.

The AQM policy adopted by the City of Johannesburg within its inaugural AQMP (Scorgie et al., 2003c) is for example outlined in Table 42.

Table 42: Air Quality Management Policy for the City of Johannesburg (Scorgie et al., 2003c)

Vision Statement and Mission The City of Johannesburg's vision, mission, overarching principles and general approach to air quality management reflects the vision, principles and approach adopted in terms of national and provincial policy in addition to local goals. The vision is as follows: Clean air is essential to a healthy population, a healthy environment, and, in turn, a healthy economy. The City of Johannesburg is committed to making the air in every community healthy to breathe, to reducing ecosystem damage from air pollution, and to doing its share to address global air quality problems. Air quality will be managed through the implementation of a coordinated approach to the control of air pollution and through the sustainable development of the built environment and transportation within the City. It is intended, in the long-term, that the air be rendered odourless, tasteless, look clear and have no measurable short- or long-term adverse effects on people, animals or the environment.

In the next 10 years, the Departments charged with protecting air quality envision substantial improvements in air quality, despite countervailing trends in population, development, and transportation growth. In achieving such improvements, these Departments are committed to: - Establishing a set of shared goals and strategies for air quality improvement. - Establishment and continued implementation of a comprehensive air quality monitoring and management system - Involving and educating the public with the purpose of minimizing pollution and facilitating the effective participation of the public in air quality governance - Make greater use of innovative approaches to reducing pollution. - Conducting sound research and effectively use new information technologies. - Respond creatively and vigorously to new challenges and emerging issues. - Improve the working partnership of personnel responsible for air quality management at all levels of government.

Strategic Goals and Objectives The main goals to be achieved by the City of Johannesburg through its development, implementation, review and revision of air quality management plans are as follows: - To achieve acceptable air quality levels throughout Johannesburg. - To promote a clean and healthy environment for all citizens within Johannesburg. - To minimize the negative effects of air pollution on health and the environment. - To promote the reduction of greenhouse gases so as to support the council's climate change protection programme. Specific objectives include: - To promote cleaner production and continuous improvement in best practice as it pertains to air pollution prevention and minimisation.

206

- To promote energy efficiency within all sectors including industrial, commercial, institutional, mining, transportation and domestic energy use.

Approach to Air Quality Management A shift from end-of-pipe air pollution control through the exclusive implementation of command-and- control measures to effects-based air quality management using proactive, flexible, varied and fair measures is supported by the new policy. The key approaches which are to be implemented in order to achieve policy objectives may be individually listed as follows: - Adoption of a receiving environment approach which requires the setting of local air quality objectives - Such objectives are needed to define what constitutes satisfactory air quality to ensure human health and welfare, the protection of the natural and build environment, and finally the prevention of significant decline. - Establishment of a sound technical basis for air quality management and planning. - This would include the building of technical expertise and the development and implementation of various tools such as an emissions inventory, a meteorological and air pollution monitoring network, atmospheric dispersion model, impact assessment methodologies (etc.). - Control and management of all sources of air pollution relative to their contributions to ambient air pollutant concentrations. - This will ensure that improvements in air quality are secured in the most timely, even-handed and cost-effective manner. - Implementation of a range of tools in the prevention of air pollution including: source-based command-and-control measures, market incentives and disincentives, voluntary initiatives and self-regulation and education and awareness methods. - The integration of a wide range of emission reduction measures is required given the diversity in the nature of air pollution sources. Such an approach will ensure innovative and flexible plans of action tailored to suit specific source types and local circumstances. - Identification and implementation of emission reduction measures which are: (i) environmentally beneficial taking all media into account, (ii) technically feasible, (iii) economically viable, and (iv) socially and politically acceptable. - Provision will be made for the integration of air quality issues into the transportation, housing and land use planning process to ensure that air quality issues are addressed in the long term. - Empowerment of communities by providing easy access to ambient air quality information, including information on air pollution concentrations and environmentally harmful practices. - Facilitation of public consultation and encouragement of public participation in the air quality management and planning process.

In addition to putting in place the air quality policy framework, there is also a need to ensure executive management commitment to air quality management. This need is clearly illustrated through the comparison of experiences gained during the Johannesburg and Ekurhuleni Metropolitan Municipality (EMM) AQMP development processes. The Johannesburg AQMP, despite having involved the senior managers of the various responsible departments, was only presented to the City of Johannesburg Portfolio Committee following its finalisation. It took several months to gain approval from this committee for the City to adopt and begin implementing the plan.

207

Lessons learned by the author were integrated into the Ekurhuleni Metropolitan Municipality AQMP development project, with presentations given to the EMM Portfolio Committee at various stages during the AQMP development process. Presentations included overviews of the status of air quality and of significant sources and pollutants requiring management, and the presentation of the draft AQMP. As a result of this consultative process, the executive managers of the EMM were more informed about and committed to the AQMP process, with the AQMP being officially launched by EMM shortly after its finalisation.

9.5 Ambient Air Quality Objectives Air quality objectives are fundamental to effective air quality management, providing the basis for evaluating the acceptability of air quality and determining the need for action (Figure 55). Following the implementation of interventions, such objectives assist in determining whether the measures taken were sufficient.

Air quality objectives can take the form of legally binding standards (e.g. US National Ambient Air Quality Standards) or as goals or limits aimed at triggering planning (e.g. European Community Air Quality Limit Values). Such objectives typically comprise more than pollutant concentration thresholds and associated averaging periods. Key information linked to air quality standards include:  timeframes for achieving compliance  margins of tolerance  permissible frequencies of exceedance  monitoring and data management protocol to assess and report compliance - including reference monitoring methods, Quality Assurance / Quality Control requirements, data analysis methods, data quality objectives and sampling siting criteria.

The levels at which concentration thresholds are set is dependent on the aspects of the environment which are to be protected and the level of protection to be provided for. Air quality standards or limits could for example be set primarily to protect human health with the thresholds set for pollutants such as sulphur dioxide and ozone not necessarily offering sufficient protection for entire ecosystems. Alternatively, multiple standards or limits could be set for the protection of human health and other aspects of the broader environment. EC air quality limit values for annual average sulphur dioxide are, for example, specified for the protection of ecosystems.

9.5.1 Adoption of Air Quality Objectives within Inaugural AQMPs The AQA of 2004 makes provision for mandatory national air quality standards to be set for substances or mixtures or substances which present a threat to health, well-being or the environment. Provincial authorities may also set air quality standards. In instances where national standards are already set for a particular substance, Provincial Government may set stricter standards for the province or for any geographical area in the province. Although Local Authorities are permitted to specify local emission standards, the AQA does not provide for municipalities to set ambient air quality standards. The author considers that Local Authorities may make use of air quality objectives, comprising air quality limits which are not legally binding, as the basis for assessing local air quality and determining the need for interventions. The rational for Local Authorities adopting air quality objectives to inform air quality management planning is discussed in this section.

208

Uncertainties related to national air quality standards represented a significant challenge facing local authorities during their AQM planning processes up until the eventual publication on a set of up-to-date national ambient air quality standards in December 2009(48). During the 1990s air quality guidelines were developed by the DEA for several pollutants, viz. SO2, TSP, PM10, NO2,

NOx, NO, CO, O3 and lead. These guidelines were primarily used internally by the DEA, although they were widely known by air pollution practitioners within government and the private sector. The basis for the guidelines was not documented, nor where the guidelines gazetted or subsequently reviewed and revised (until December 2009; DEA, 2009). By 2001, at which time the need for AQM planning was receiving increased emphasis, these guidelines were perceived as being out of date, incomplete and not sufficiently protective of human health and the broader environment.

Attempts made to develop a set of national air quality standards during the 2001 to 2009 period, are briefly summarised as follows:  Standards South Africa (STANSA), a division of the South African Bureau of Standards (SABS), established a Technical Committee(49) through which the following standards were developed: • SANS 69: 2004 - Framework for Setting and Implementing National Ambient Air Quality Standards (Edition 1.0)(50), and • SANS 1929: 2004 - Ambient Air Quality – Limits for Common Pollutants (Edition 1.0). Although the DEA initially mandated the STANSA standard setting process and had representatives involved in the standard setting process, the Department did not approve nor adopt the standards developed. Reasons for this are comprehensively documented in DEAT (2007).

 The ambient air quality guidelines developed by the DEAT under APPA during the 1990s were included as air quality standards in the AQA as promulgated in 2004, with the exception of sulphur dioxide, for which criteria published following an earlier revision(51) were included. The inclusion of these guidelines as standards was given to be a transitional provision, with it being stated that “Until ambient air quality standards have been established in terms of [the Act], the ambient air quality standards contained in Schedule 2 apply”.

DEAT (2007:7) acknowledged that these standards, with the exception of the SO2 standard, “…may be outdated and no longer considered to be protective of health and well-being”.  In 2006 the DEA published for comment a set of ambient air quality standards. These standards were primarily based on the criteria specified in SANS 1929:2004 (DEAT, 2006). These standards were not finalised, with the air quality standard setting processes being incorporated in 2007 into the DEA‟s National Framework development project.

48 South African Government Gazette, No. 32816, 24 December 2009. 49 The author was a member of the Technical Committee and assisted in the development of SANS 69 and SANS 1929 standards (including writing the first drafts of these standards). 50 An amended edition of SANS 1929 was published in February 2005 (Edition 1.1) with minor editorial corrections. 51 Van Niekerk W.A. (2001): Technical Background Document for the Development of a National Ambient Air Quality Standard for Sulphur Dioxide, Department of Environmental Affairs and Tourism.

209

 Under the National Framework Project, draft air quality standards were again published for comment by the DEA at the end of 2007, with comments required in early 2008 (DEAT, 2007a).  It was not until December 2009 that an up-to-date set of national air quality standards were finally promulgated for implementation under the AQA (DEAT, 2009)(52). The national standard setting process is ongoing with national standards subsequently being proposed for (53) (54) PM2.5 (DEA, 2011b) and dust deposition (DEA, 2011c) .

The development of the City of Johannesburg, EMM and City of Cape Town AQMPs during the 2002 to 2005 period was hindered by the absence of a clearly documented set of current national air quality standards to provide the basis for effects-based air quality management. Within this context the author devised the following recommendations to support the planning processes:

 Metropolitan municipalities, despite not being mandated by the AQA to develop air quality standards, should define air quality objectives as local targets to guide their air quality management planning process. With national air quality standards adopted as a minimum (referring to the outdated standards referenced within the AQA, as published in 2004), it was recommended that more protective thresholds be defined in line with the local authorities‟ strategic goals.  In selecting pollutants for which local objectives are to be established, attention should be paid to: • commonly occurring pollutants within the local authority jurisdiction that give rise to relatively widespread exposures; • pollutants for which national air quality standards exist and for which national air quality standards are in the process of being established; and • pollutants for which guidelines/standards/goals are issued by other countries, particularly those implementing similar AQM processes.  A tiered approach was advocated for setting air quality objectives, similar to that developed by the EC and reflected in SANS 69 and 1929:2004, with objectives set based on the following threshold types: • Limit values are to be based on scientific knowledge, with the aim of avoiding, preventing or reducing harmful effects on human health and the environment as a whole. • Information and investigation thresholds are intended to highlight pollutant concentrations at which the public need be informed that the most sensitive individuals may be affected and/or at which investigations into reasons for the elevated levels need to be initiated. • Alert thresholds refer to levels beyond which there is a risk to human health from brief exposure. The exceedance of such thresholds necessitates immediate steps.

52 South African Government Gazette, No. 32816, 24 December 2009. 53 South Africa Government Gazette, No. 34493, 5 August 2011, Proposed National Ambient Air Quality Standard for PM2.5. 54 South Africa Government Gazette, No. 34307, 27 May 2011, Draft National Dust Control Regulations.

210

 Initially, local air quality objectives should reflect thresholds which are protective of human health (i.e. limit values for human health). The setting of objectives which are protective of the broader environment, and the specification of information/investigation and alert thresholds should be phased in as information is collated and research completed.  To ensure that air quality objectives are not overly ambitious, timeframes for achieving such targets should be informed by modelling work to project future air pollution concentrations, including and excluding cost-effective interventions.  In the absence of national and other guidance, a procedure should be specified for dealing with the evaluation of pollutants for which no local objectives have been established.

Local air quality objectives were included in the City of Johannesburg AQMP and the EMM AQMP and recommended for consideration by the City of Cape Town (CCT), based on the review of major meta-analyses undertaken by various countries and organisations for the purpose of setting air quality limits (Scorgie et al., 2003c, 2004f; Scorgie 2004b). The limit values and associated averaging periods recommended for adoption by CCT were primarily based on human health effect data, with reference made primarily to the lowest observed adverse effect level (LOAEL) rather than exclusively to the thresholds adopted by other countries. It was however noted that the standards more recently promulgated (e.g. limit values of the EC, UK and certain of the Australian standards) closely coincide with LOAELs.

In the case of fine particulate matter, no safe threshold is recognised, with EC, UK and the limits of several other countries being set based, not on a LOAEL, but at a level indicative of an accepted degree of risk. Due to the high background concentrations of fine particulate matter characteristic of southern Africa, efforts to meet best practice PM10 limits in the short-term would be very costly and likely to be unsuccessful. The approach locally, as adopted in the SANS process, has therefore been to set more lenient limits for PM10 and recognising that this demonstrates an acceptance of a higher level of risk. The local approach of managing down fine particulate matter concentrations over longer timeframes is supported by WHO (2005).

Alert and information thresholds put forward for consideration by CCT are outlined in Table 43. It was recommended that these thresholds be finalised at a later date following: (i) at least one year of air pollutant concentrations recorded for the pollutant for which the thresholds are to be set; (ii) source contributions to ambient air pollutant concentrations established; and (iii) possible actions assessed in terms of their socio-economic acceptability and technical feasibility. It was also specified that locally-defined alert and information thresholds be revised in the event that national standards are issued.

Table 43: Alert and information thresholds recommended for use by CCT (Scorgie, 2004b)

Information Pollutant Averaging Period Alert Threshold Basis for Threshold Threshold 532 µg m-3 1064 µg m-3 10-minute average UK 15-min bands 200 ppb 400 ppb Sulphur dioxide 350 µg m-3 3 consecutive hours EC alert threshold (130 ppb) 573 µg m-3 764 µg m-3 Nitrogen dioxide 1-hour average UK bands (300 ppb) (400 ppb)

211

Information Pollutant Averaging Period Alert Threshold Basis for Threshold Threshold 400 µg m-3 3 consecutive hours EC alert threshold (209 ppb) 17.4 mg m-3 23.2 mg m-3 Carbon monoxide 8-hour average UK bands (15 ppm) (20 ppm) 180 µg m-3 360 µg m-3 Ozone 8-hour average UK bands (90 ppb) (180 ppb)

Various of the recommended thresholds given in Table 43 are based on UK air quality bands. The information threshold level is set equivalent to the „high‟ pollution level with the alert threshold indicative of „very high‟ pollution levels. The UK defines the implications of such levels as follows (DEFRA, 2011):  High pollution levels - "Significant effects may be noticed by sensitive individuals and actions to avoid or reduce these effects may be needed (e.g. reducing exposure by spending less time in polluted areas outdoors.) Asthmatics will find that their 'reliever' inhaler is likely to reverse the effects on the lung."  'Very high' pollution levels - "The effect on sensitive individuals described for 'high' levels of pollution may worsen."

In advising the City of Cape Town, it was recommended that local air quality objectives be based initially on thresholds able to protect human health. It was however recommended that CCT encourage local research aimed at identifying thresholds suited to the protection of local ecosystems, possibly through collaboration with the Western Cape provincial government. As an interim measure, reference could be made to air quality criteria published by jurisdictions abroad for the protection of vegetation.

Air quality criteria are typically only specified for commonly occurring air pollutants that result in relatively widespread public exposures. To ensure that a sound approach is adopted in the assessment of the potential for health effects from non-criteria pollutants, the author proposed that an inhalation health risk screening procedure be considered for adoption (Scorgie, 2004b) comprising the following main steps: a) Determine ambient(55) near ground(56) air pollutant concentrations through ambient air quality monitoring and/or atmospheric dispersion modelling.

For ambient air quality monitoring use must be made of a credible monitoring device and methodology. The detection level of the instrument must be below the level at which health effects are known or suspected to occur. Monitoring must be undertaken for the averaging period for which health thresholds are available (e.g. hourly averages).

55 Ambient air is defined for the purpose of implementing this procedure as being beyond the fence lines of specific industrial and mining operations in areas where public exposures are possible. 56 It is recommended that concentrations be established at about 1.5 m above ground level. This is typically set as the receptor height for assessing human exposures.

212

For atmospheric dispersion modelling use must be made of accurate source and emissions data. Emission rates may either be measured, or calculated based on mass balance equations, engineering calculations or applicable emission factors. Site-specific meteorological and topographical data should be used in the modelling. b) Obtain inhalation-related dose-response thresholds for the air pollutant under investigation from credible, preferably refereed, sources. Recommended information sources and types of thresholds are given in Table 44.

Table 44: Recommended information sources for inhalation-related health risk thresholds

Recommended Threshold Type: Averaging Period: Website: Information Sources: United States Sub-chronic inhalation Sub-chronic – weeks to www.epa.gov/iris Environmental Protection reference concentrations months Agency Integrated Risk Information System (IRIS) Chronic inhalation Chronic – 1 year average or www.epa.gov/iris reference concentrations longer Cancer unit risk factors Chronic – 1 year average or www.epa.gov/iris longer (Exposures over 70 year lifetime assumed) California Environmental Acute Reference Acute – typically 1 hour www.oehha.ca.gov Protection Agency – Exposure Levels average ranging to 8-hourly Office of Environmental (RELs) average depending on Health Hazard Assessment pollutant Chronic Reference Chronic – 1 year average or www.oehha.ca.gov Exposure Levels (RELs) longer US federal Agency for Minimal Risk Levels http://www.atsdr.cdc.go Toxic Substances and (MRLs) v/mrls.html Disease Registry (ATSDR) World Health Guideline Values and Various averaging periods, http://www.who.int/en/ Organisation Tolerable including: 30-minutes, 1-hour, Concentrations 24-hour, annual average Cancer Unit Risks Chronic – 1 year average or http://www.who.int/en/ longer (Exposures over 70 year lifetime assumed) c) Determine the major exposure pathway for the pollutant under investigation, i.e. inhalation, ingestion or dermal contact. For pollutants for which inhalation is not the major exposure pathway recognize that a comprehensive health risk assessment in which multiple-exposure pathways are taken into account is needed. d) For pollutants for which inhalation represents the major exposure pathway, assess predicted and/or measured air pollutant concentrations based on applicable dose-response thresholds. Ensure that the averaging period for such concentrations are relevant to the exposure period for which the threshold is stipulated. e) For non-carcinogenic effects, exceedances of applicable dose-response thresholds should be taken to indicate the need for a more comprehensive quantitative health risk assessment. In instances where pollutant concentrations are within such thresholds, health risks may be considered unlikely to occur. f) For carcinogens, calculate possible maximum exposed individual (MEI) cancer risks through the application of unit risk factors. In cases where calculated cancer risks are greater than 1: 1 million (i.e. one person contracting cancer out of every million exposed) consult with decision makers and affected communities to determine the acceptability of the incremental cancer risk calculated. In

213

instances where cancer risks are considered unacceptable a comprehensive quantitative health risk assessment is required. Such health risk assessments quantify actual exposures, rather than assuming maximum possible exposures, and as such are less conservative.

9.5.2 Relevance of Guidance on Air Quality Objectives for Future AQMPs The guidance in respect of local air quality objectives outlined in the previous section was developed by the author to support progress in AQMP development in the absence of national air quality standards. Given the promulgation of revised national standards in 2009, and subsequent draft standards being published for comment in 2011, a review of the applicability of such guidance for future AQMPs is warranted.

The national air quality standards published in 2009 are substantially more protective of human health and generally in line with standards published by the US, EU and Australia. Furthermore, cooperative governance structures have been established, as documented in 9.7.2.2, which can be used by local government to influence national and provincial standard setting processes. However, the following guidance remains useful for future air quality management planning:

 Establishment of information, investigation and alert thresholds for common pollutants which are continuously monitored by the jurisdiction‟s air quality monitoring network.  Setting of local air quality objectives which are protective of local ecosystems in the event that sensitive environments occur within the jurisdiction.  Establishing timeframes for achieving air quality standards based on modelling work to project future air pollution concentrations, including and excluding cost-effective interventions.  Adoption of a procedure for dealing with the evaluation of pollutants for which no national air quality standards, provincial air quality standards or local objectives have been established.

Reflecting on the guidance provided during the development of inaugural AQMPs, the author considers that there are considerable benefits to be gained in setting multiple pollutant objectives. Whereas the US and EU have historically developed objectives and emission reduction strategies for individual pollutants, it has been increasingly realised that this is not the most effective or economically efficient method of reducing the effects of air pollution (NCSA, 2003; Chow and Watson, 2011). The multiple complex relationships between a number of criteria and unregulated pollutants and the effects of such on human health, visibility, climate, odours, material and ecosystems are highlighted within Chow and Watson (2011).

It is advocated that air quality objectives, assessment methodologies and emission control measures which address multiple pollutants be considered for implementation within South African air quality management planning processes. A range of approaches may be adopted including the use of a multi-pollutant Air Quality Index for evaluating air quality, the adoption of a risk-based approach to limit combinations of pollutants that adversely affect human health, and the setting of objectives based on economic analysis of co-benefits (DEFRA, 2011; Chow and Watson, 2011).

The use of cost-benefit analysis in the targeting of source groupings and multiple pollutants, and the alignment of local air quality management and global climate protection strategy, represent an

214

effective means of achieving the highest level of human and environmental protection in the most cost-effective manner.

9.6 Air Quality Management Systems The assessment of air quality represents the cornerstone of air quality management, with emissions inventory, monitoring and modelling forming the major tools for such assessments internationally (Figure 55). The absence of comprehensive air quality management systems (AQMS) in Johannesburg, Ekurhuleni Metro and Cape Town was identified as a significant hurdle to the development of effective AQMPs. Although components of such systems were in place, e.g. some air quality monitoring, such systems were insufficient as a basis for effective AQM planning in terms of providing the following information and functionality:  Source and emissions data - including source parameters required for dispersion modelling purposes, total emissions, temporal variations in emissions data and future „business as usual‟ projections of sources and emissions.  Spatial and temporal variations in air pollutant concentrations across the metropolitan area for baseline and projected future „business as usual‟ scenarios.  Exposure potential assessment within areas where air quality objectives are exceeded.  Ranking of sources based on their contributions to (i) total emissions and (ii) ambient air pollutant concentrations, specifically within areas of high exposure potential.  Projecting emission reductions and resultant air quality improvements and impact reductions due to specific interventions.

A phased approach to AQM planning was therefore devised for implementation, with the first phases focusing primarily on the establishment and implementation of an AQMS. The integrated AQMS recommended for development and implementation by Ekurhuleni Metro is, for example, illustrated in Figure 56. System components proposed for implementation in the short-term (1-2 years) are indicated by solid lines, with components to be added at a later stage indicated by dashed lines. Initial components of the system included: an emissions inventory, air quality and meteorological monitoring network, atmospheric dispersion modelling and routine reporting. The establishment of mechanisms for cooperative governance and public consultation, and the finalisation of local air quality objectives were also proposed to be done during this period as discussed in other subsections.

Based on the outputs of the air quality monitoring system, it was recommended that health risk assessments and damage assessments be undertaken and effects costed in the medium-term (3-5 years). Such assessment may be undertaken in the following ways: (i) in-house, through the selection and acquisition of suitable models and acquisition and preparation of locally-derived input data; (ii) in-house, though the application of manual calculations based on locally-derived data and international protocols; or (iii) externally, through the appointment of consultants on a project-by-project basis.

In recommending specific emissions inventory, monitoring and modelling methods and tools, reference was made to international best practice whilst also taking into account local circumstances, with specific emphasis on cost-optimisation. Following a careful review of assessment procedures with the US and Europe, it was recommended that the European approach to integrating modelling and monitoring be adopted. Whereas US AQM systems are typically very

215

intensive in terms of monitoring and hence expensive, European authorities generally supplement strategic monitoring with more sophisticated modelling systems. The European approach was found to be more cost-effective for implementation locally and several urban airshed models, such as AirQUIS and ADMS Urban, were evaluated for use by South African metropolitan municipalities. The author advised the City of Johannesburg, Ekurhuleni Metro and the City of Cape Town to select and implement the same model to support capacity building and knowledge sharing. ADMS Urban was favoured due to their being local consultancy support for the implementation of this urban airshed model. By 2010, this model was being applied in-house by the Johannesburg, Cape Town and Ekurhuleni with these metros representing three of only five local governments with in-house air dispersion modelling capabilities. The other two metros being eThekwini Metropolitan Municipality applying AirQUIS and the Nelson Mandela Metropolitan municipality which implements AERMOD within the Enviman system (Gwaze, 2010).

Further cost-optimisation of monitoring strategies could be achieved through the integration of a range of measurement techniques including passive diffusive monitoring, biomonitoring and continuous near real-time measurement. For an example of such an approach reference should be made to the Ekurhuleni Metro AQMP (Scorgie et al., 2004f).

In considering air quality monitoring systems more suited to smaller and/or less polluted municipalities, staged approaches to air quality assessment adopted internationally are relevant to ensure cost efficiency. More in depth and complex air quality assessment processes (and hence monitoring tools) are generally reserved for application within areas identified to be worthy of further attention, based on preliminary surveys or the use of screening tools. In the UK, for example, the government has recommended a three-stage approach for local authorities reviewing and assessing the air quality in their jurisdictions (DEFRA, 2003). The first stage comprises the use of a compilation of emissions data from various sources and background concentrations of the various criteria air pollutants. Following this initial stage, pollutants can be omitted from the process when there is little likelihood of the air quality objectives being breached. The second stage comprises a more sophisticated screening using screening models and any local air quality monitoring data available. Pollutants are again omitted from the process when, on more detailed analysis, they are found unlikely to exceed the air quality objectives. The third and final stage is a more complex study of the locations and pollutants identified by the earlier stages as potentially exceeding air quality objectives. Such studies often require more advanced monitoring and air pollution dispersion modelling for predicting specific locations of future pollutant exceedances. On completion of the third stage review and assessment, action planning is required for areas where air quality objectives are, or are predicted to be exceeded.

216

Figure 56: Air quality management system recommended for EMM (Scorgie et al., 2004f)

The 2007 National Framework for Air Quality Management in South Africa highlights the need for standardised emissions inventory, air quality monitoring and dispersion modelling by local and other authorities (DEAT, 2007b). Practical experience gained by the City of Johannesburg and Ekurhuleni Metropolitan Municipality in the development and implementation of their air quality management systems have been drawn upon in the drafting of the Manual for Air Quality Management Planning in South Africa (DEA, 2008). This manual specifies the need for the establishment of an AQMS - comprising an emissions inventory, monitoring network and dispersion modelling - for jurisdictions where air quality is determined based on a baseline assessment to be „potentially poor‟ or „poor‟(57). Generic guidance is provided on emission inventory, monitoring and dispersion modelling within the manual with more detail to be provided in subsequent publications.

More in depth guidance on air quality modelling is intended to be included within the National Regulatory Dispersion Modelling Guideline which is due for completion in March 2012 (DEA, 2011d). This guideline will recommend a suite of dispersion models for regulatory use, and provide guidance on model input requirements, protocols and procedures on regulatory applications. The guideline will be referenced within the revised National Framework for Air Quality Management in South Africa in 2012 (Gwaze, 2010).

57 Poor air quality is defined as occurring when air quality standards are exceeded.

217

The development and implementation of the South African Air Quality Information System (SAAQIS) over the 2008 to 2017 period is likely to significantly improve information available for AQMP development in future years. Acting as a repository for information, SAAQIS will incorporate air quality monitoring data, emissions inventories, listed activity information and various other data (DEA, 2011d). SAAQIS also makes available AQMPs already developed for access by other authorities.

The establishment and maintenance of comprehensive emission inventories is resource intensive, and efforts to cost-optimise this process are needed. The SAAQIS Phase II – National Atmospheric Emission Inventory Project addresses this need. Given that a piecemeal approach to the development of local authority emissions inventories by various local authorities is costly, DEA has recognised that significant opportunities exist to realise economies of scale and cost savings through the application of standardised methods country-wide. The aforementioned project aims to develop and implement comprehensive standardised emission inventories with countrywide coverage, and to integrate the resultant source and emissions data within SAAQIS. This project is due for completion in March 2014 (DEA, 2011d).

The absence of comprehensive air quality monitoring systems remains a challenge for most jurisdictions within South Africa in 2011. A key lesson learned by the author is that the successful development and implementation of such systems is substantially dependent on the availability of human resources, and specifically persons experienced in emissions inventory, dispersion modelling and continuous air quality monitoring. The increasing number of jurisdictions scheduled to deploy AQM systems, and the need for access to similar skills by the private sector to address regulatory requirements, has resulted in the demand for skills far exceeding the available resource pool. The establishment of mechanisms for cooperative governance and public-private sector collaboration in regard to capacity building and resource sharing represents an important solution. Further discussion on such mechanisms is provided in subsequent sections.

9.7 Capacitating Air Quality Management Effective air quality management and planning necessitates the close integration of source control, air quality monitoring and air quality planning functions with such functions being undertaken by skilled and experience personnel.

During investigations into several metropolitan municipalities, air quality management functions were found to be spread across various sections, departments and/or directorates (Scorgie et al., 2003b, 2003c, 2004e, 2004f; Scorgie and Watson, 2004; Scorgie, 2004b). Such departments included the Department of Health, with Environmental Health Officers frequently having been responsible for air pollution control historically, and the Department of Environment, which also frequently including planning and transportation functions. In the case of the City of Cape Town, an Air Pollution Control Section had been established with air quality monitoring being undertaken by a separate Scientific Services division.

Capacity development needed for AQM planning was determined based on such investigations to extend beyond the provision of additional financial and staff resources and the acquisition by existing staff of further skills through training and education. It required, in addition, the building of new organisational and functional structures within the municipality to facilitate inter- departmental coordination and cooperation. Skill development considerations and potential

218

mechanisms to facilitate inter-departmental and inter-governmental coordination and consultation are discussed in subsequent subsections.

9.7.1 Technical Skill Development Potential interventions which can be implemented to address the training needs in respect of air quality management include:  Development of tertiary level qualifications  Provision of short courses, both in-service and by external bodies (including universities)  Partnerships between government departments  Internships  Attendance of forums, including governmental and government-private sector forums.  Networking through professional associations such as the National Association for Clean Air (NACA)  Publication of guidance documents - various guidance documents are in the process of being developed and issued by DEA and other agencies such as NACA  Provision of internet-based information  Telephone and email help desks

The author developed and presented an in-service AQM course for Ekurhuleni Metropolitan personnel during 2005 to support the successful implementation of the AQMP by the metro, and subsequently was involved in the development and presentation of a 10-day AQM short course, entitled „Introduction to Air Quality Management‟ at the University of Johannesburg. This course was first presented during 2006, with attendance including local, provincial and national government officials tasked with various air quality management functions. The University of Johannesburg training course was subsequently expanded by NACA to include specialist courses on atmospheric dispersion modelling and ambient air quality monitoring, with these courses being offered in 2010. Plans are underway to implement the NACA courses at other universities such as the University of Pretoria and the University of the North West.

Government has supported in service training by making provision for training within the terms of reference of several of its projects, and by sending personnel to attend the air quality management training courses offered by the University of Johannesburg.

Although the public and private sector efforts documented above provided some impetus, what is needed is a sustained and multi-pronged approach to meet requirements for skills in the short-, medium- and long-terms. Collaboration between government, tertiary institutions and private sector practitioners in the development of this approach is needed. The joint conferences held by government and NACA provide an example of public-private collaboration with regard to air quality management knowledge sharing. The author, in her capacity as deputy president of NACA and advisor to national government on the APPA Registration Certificate Review Project, played a role in establishing this collaboration. The author motivated for DEA to co-host the annual NACA conference and to support the involvement of air quality officers in this professional association. On the advice of the National Air Quality Officer, the Director-General of the DEA gave approval for the Air Quality Governance Lekgotla to be held back to back with the annual NACA clean air conference, and a memorandum of understanding was signed by DEA and NACA for such events. The first Annual Air Quality Governance Lekgotla was held in East London from 16-17 October

219

2006. The Lekgotla was followed by a multi-stakeholder workshop at the same venue focused on the development of the National Framework on 18 October 2006, and subsequently by NACA‟s technical conference from 19-20 October 2006.

The joint event was considered by the National Air Quality Officer as having the following advantages (DEA, 2010b, p. 114):  “Government officials involved in air quality management only needed to make one trip to participate in two events of critical importance to them, namely a practical AQA implementation event (i.e. the Lekgotla) and a best practice technical event (i.e. the NACA Conference).

 The technical aspects of air quality governance were able to be aired at an event where the country’s top experts were available for input into discussions and debates.

 Government officials involved in air quality management were given an opportunity to interact and network with the country’s top air quality management experts.

 NACA‟s twenty years of conference organisation experience added to the success of the new governance component.” The National Air Quality Officer considered the partnership between DEA and NACA to be “highly successful”. The memorandum of understanding was reviewed in 2010 and extended to cover continued joint hosting of the Annual National Air Quality Conference (weeklong event) and the development of an annual business plan for joint activities including (DEA, 2010b):  Supporting attendance of deserving air quality officials at the conference;  Development and publication of outreach and educational materials;

 Provision of bursaries to air quality officials for approved training courses;  Joint hosting of specialist air quality events such as workshops and seminars;  Design and implementation of joint research activities aimed at improving air quality management systems, protocols and procedures; and  Development and implementation of joint public awareness campaigns. Guidance documents, forums, help desks, internships and in-service short courses provide useful methods to impart skills and to retrain existing personnel. The development of suitable qualification supported by tertiary institutions nationally, the ongoing revision and issuing of up to date guidance documents, and the involvement of air quality practitioners in professional associations is crucial to ensure capacity in the longer term.

Following the publication of the National Air Quality Strategy in the UK, local authorities there expressed concern regarding the lack of necessary tools and expertise to undertake their AQM responsibilities effectively. Resource development measures developed and implemented in the UK included development of guidance documents, internet-based information and telephone and email help desks. A significant increase in the AQM practice and capability within UK local authorities including the use of monitoring and dispersion modelling was observed by Beattie et al. (2001) following the publication of the National Air Quality Strategy in 1997.

220

Despite noting successes from the UK approach, it is also critically important to learn from mistakes and oversights. With respect to training, the only profession which received adequate training during the first five years was environmental health officers, with only 23% of planning officers and 3% of economic development officers receiving any training during this period (Beattie et al., 2001). This was recognised as an important factor in the ineffectiveness of emission reduction plans. To avoid this pitfall South Africa will need to ensure that developers and planners, including spatial development, transport, housing, infrastructure and energy planners are being targeted for training, both in-service and at tertiary institutions.

9.7.2 Cooperative Governance and Stakeholder Consultation Cooperative governance mechanisms are particularly important given that different sources are controlled (or more readily controllable) by different government departments at various tiers of government. Examples of national departments, other than the DEA, that have an interest in or responsibility in respect of atmospheric emission sources are as follows (DEAT, 2007b):  Department of Health – household fuel burning and emissions from household products and building materials (indoor air quality); hospital boilers; medical waste incinerators.  Department of Minerals and Energy – Dust from mine tailings impoundments, open-cast coal mines, mine haul roads and mining operations; emissions from fires in coal mines; emissions from fossil fuel use.  Department of Water Affairs and Forestry – Emissions from veld and forest fires; biogenic emissions and sinks; effect of emissions on water quality through acidification; historically responsible for waste disposal facilities.  Department of Transport – Emissions from various transport modes and from transport infrastructure construction.  Department of Agriculture – Dust from agricultural activities such as tilling; stubble burning; sugar cane burning; crop-spraying; burning of fire breaks; effects of emissions on soil quality through acidification.  Department of Housing – Household fuel burning; electricity consumption of households.

The examples above illustrate that AQM personnel within local government will need to liaise with various national government departments in terms of collating information and successfully controlling certain sources. Such personnel will also need to build close working relationships with other local departments responsible for activities associated with significant atmospheric emissions.

9.7.2.1 Early Steps Towards Cooperation and Consultation In developing the inaugural AQMPs during 2002 to 2004, the author realised that AQMP development necessitated consultation, including inter-departmental and inter-governmental cooperation, and public consultation with emphasis on key stakeholder engagement. Given that mechanisms for such consultation and coordination were unavailable, and considering the tight timeframes stipulated for the Ekurhuleni Metropolitan Municipality and Johannesburg City AQMP development processes, it was necessary to establish interim mechanisms as part of the plan development processes. Two groups were established during these processes namely:

 Technical Working Group - comprised persons able to contribute to the process in one of two ways: (a) through providing input into the air quality management system design, e.g. participating in the design of the air quality monitoring network; and (b) assisting with

221

emission reduction measure drafting and with assessing the feasibility and cost implications of implementing measures within the industrial, mining, domestic fuel usage and transport sectors. Representatives from the various departments within the municipality responsible for environmental management, environmental health, municipal infrastructure, housing, transport and spatial planning were invited to participate on the Technical Working Group. Representation was also invited from provincial government (Gauteng Department of Agriculture, Conservation and Environment, GDACE), national government (DEA and Department of Minerals and Energy) and from various air quality forums which undertook air quality monitoring functions (e.g. Springs Air Quality Forum and Airkem in the case of the Ekurhuleni Metro).

 Air Quality Stakeholder Group – comprising representatives identified on the basis of a consultative process. This group comprised both parties impacted by air pollution and parties who may be impacted by air pollution abatement measures. The main functions of the Air Quality Stakeholder Group included: (i) assist in the categorisation of issues raised during the broad public consultation process; (ii) assist in identifying the potential for trade-offs and compromises where conflicting views are given by various stakeholders; and (iii) acting as a „sounding board‟ to assist the project team in determining whether key issues have adequately addressed.

Based on the lessons learned from the initial AQMPs developed, the author recommended the structures, forums and co-ordinating mechanisms listed in Table 45 for consideration by the City of Cape Town during its AQMP development. Such mechanisms were intended to ensure:  regular liaison and co-ordination between local, provincial and national government departments,  inter-disciplinary technical support for the identification and evaluation of air quality effects and emission reduction opportunities, and  on-going public consultation and participation in the air quality management process.

More information on the technical advisory group, government coordination committee and key stakeholder group proposed are provided in subsequent sections.

222

Table 45: Structures, forums and co-ordinating mechanisms recommended for consideration by the CCT (Scorgie, 2004b)

Structure: Composition: Main Function: Air Quality Everyday business of air pollution control, ambient air quality and meteorological AQM section personnel Management Section monitoring and air quality management and planning Guides AQM Section management with regard to: (i) regulation, by-law and 6 members, viz. MLC supervisors (South Peninsula, Blaauwberg, CCT AQM Co- guideline development and manner of be implementation, (ii) air quality Helderberg, Oostenberg municipalities and the City of Tygerberg) and a ordinating Group management plan development and implementation representative from the City of Cape Town AQM section Guides AQM Section's budget

(~12 members) - Identify air quality improvement options, including emission reduction strategies - CCT AQM section personnel - Evaluation of strategies, including: - Representatives from: - air quality improvement to be achieved - Transport and Traffic Directorate(a) - practical feasibility of strategy implementation - Spatial Planning Dept. - cost implications of strategies AQM Technical - Housing Dept. - Provide guidance on the timeframe for implementation of measures, and on the Advisory Group - Finance Directorate(b) anticipated timeframe for air quality improvement realisation - Wastewater Dept. - Rank air quality improvement measures based on their cost-efficiency and effectiveness - Waste Management Dept. - Intermittently evaluate income generation mechanisms for additional AQM - Two (Non-governmental) technical experts in the AQM field (invited) Section mandates - DECAS (provincial environmental department) - Facilitate integrated pollution control and environmental management across - DEA representative directorates (minimum 6 members) Government - CCT AQM Section personnel Facilitate liaisons between local, provincial and national government on: (i) Coordination Scheduled Processes / Controlled Emitters, (ii) Air Quality Management Plan Committee - DECAS (provincial environmental department) development and implementation - DEA representative

(~15 members) - Act as 'sounding board' to assist the AQM Section in liaising with I&APs with - CCT AQM section personnel regard to AQM Plan development & implementation Key Stakeholder - Representatives from: - Provide guidance on the likely socio-economic acceptability of (i) local air quality Group guidelines, targets & timeframes, (ii) proposed emission reduction measures, and - interested & affected local communities (iii) planned public consultative processes - industry - Assist in identifying the potential for trade-offs and compromises where - environmental NGOs; - (etc.) conflicting views are given by various stakeholders (a) Representative able to facilitate liaison between the AQM Section and various departments including: Public Transport, Road Management, Traffic Management and Transport Planning. (b) Representative to provide guidance on methods by which funds can be generated and integrated into the AQM Section's budget in accordance with the required rules and regulations.

223

a) Technical Advisory Group Important relationships exist between air quality management, land use planning, housing and transportation planning. New land use developments can influence both travel patterns and exposure levels. The siting of a residential area in proximity to an industrial area would, for example, result in increased levels of human exposure to the emissions generated by the industry. Whereas extended transportation networks and increased traffic flows resulting from such new developments would directly influence air quality through enhanced tailpipe emissions of particulate matter and increased re-entrainment of dust on roadways. Unless this relationship is recognised and channels of communication established between local and regional agencies responsible for land use planning, air quality management and transportation planning, air quality management in unlikely to succeed.

The integration of air quality considerations into housing developments is also receiving increasing attention due to the health risks associated with domestic fuel burning emissions. The intensive housing construction programmes in place at present provide the ideal opportunity for the meaningful integration of measures aimed at improving the energy efficiency of the housing shell.

One of the major deficiencies of the AQM processes has been found internationally to be due to insufficient inter-authority collaboration. The UK experience has demonstrated that the most effective way of delivering such collaboration is through groups, both within local authorities and regionally, to address air pollution issues, share experiences and in some cases share workload and expenditure. The sharing of experiences between local authorities provides for a more consistent approach to air quality management and is critical in the addressing of cross-boundary air pollution issues as part of priority area planning.

Taking account of international experiences and the departmental participation required within the CCT it was recommended that a Technical Advisory Group be established comprising representation from the Transport and Traffic Directorate, the Housing Dept., Spatial Planning Dept. and the Water and Waste Directorate. Furthermore, that this group operate in a manner which will ensure: (i) the effective and timely collection of information required for emission quantification; and (ii) the integration of air quality effect and emission reduction considerations into transport, land use, housing and waste/wastewater management planning processes.

Wherever possible, air quality considerations should be integrated into existing policies and programmes. In 2004, examples of such policies and programmes in the CCT included the Metropolitan Spatial Development Framework (MSDF), Cape Metropolitan Transport Plan, Metropolitan Housing Policy, Mining Structure Plan, Port Development Framework, Sustainable Energy, Environment & Development Programme (SEED) and the Integrated Waste Management Project.

The author further advised that it may be necessary to establish a metropolitan transportation planning work group under the Technical Advisory Group to ensure the success of air quality management strategies. The transportation planning work group could assist in providing the transportation data required for the accurate estimation and inventory of vehicle-tailpipe, rail and aircraft related emissions by the AQM division. Such a work group would be responsible for evaluating air quality changes associated with transportation planning. Other special working groups may also be considered as and when required by the Technical Advisory Group.

224

b) Government Coordination Committee The Government Coordination Committee was designated as a formal pathway of communication between various tiers of government. Issues to be addressed by such a committee include:  standardisation of air quality management plans  auditing of AQM plans  mechanisms and procedures for reporting progress being made  mechanisms and protocols for maintaining and transferring emissions and air quality data  access to national air quality and source inventory databases.  enforcement functions which may be used for various sources  procedures for inspections, authorisations and prosecutions with regard to 'Listed Activities' and 'Controlled Emitters'  local capacity building requirements.

c) Key Stakeholder Group Public participation is integral to the development and implementation of air quality management within the context of the AQA. To avoid token participation, methods must be adopted to realise effective, equitable and early public participation with attention paid to the needs of vulnerable groups.

9.7.2.2 Establishment of Mechanisms for Cooperation and Consultation Structures designed to give effect to inter-governmental cooperative governance were established by DEA under the Air Quality Act. Such structures, conceptualised by representatives of all three tiers of government at the first air quality governance conference held in 2005, are illustrated in Figure 57. These structures have since been established with Quarterly National-Provincial and Quarterly Provincial-Municipal Air Quality Officers‟ Forums being held. Additionally, an annual air quality governance conference is held in collaboration with the National Association for Clean Air (NACA) as discussed previously, and regular Communiqués are issued by the National Air Quality Officer.

The cooperative governance structures outlined in Figure 57 would replace the need for the Government Coordination Committee proposed by the author for the City of Cape Town. The issues intended to be addressed by this committee, as discussed above, could be covered by these structures.

The cooperative governance structures illustrated in Figure 57 makes provision for Local Air Quality Interest Groups to be established by Provincial Government. In the case of the Western Cape it is expected that different interest groups are likely to be established for different regions, e.g. Saldanha Bay and CCT. The Key Stakeholder Groups established by CCT could feed into the Local Air Quality Interest Groups planned.

225

Figure 57: Cooperative governance structures outlined at the 2005 Air Quality Governance Conference (DEA, 2007)

Governmental and stakeholder structures set up for the purposes of the Vaal Triangle Priority Area Air Quality Management Plan Development Project (2007-8) were similar to those established for the City of Johannesburg and EMM. The Vaal Triangle Priority Area however stretches across two provinces (Gauteng and Free State Provinces) and several municipalities, with the AQMP development process being coordinated by national government due to the Priority Area having been declared by the Minister. Priority Area AQMP development requirements in terms of inter- departmental and inter-governmental coordination and stakeholder consultation are therefore even more extensive and complex.

9.8 Emission Reduction Planning When advising government on emission reduction measures during the AQMP development processes, the author made reference to international good practice in air quality action planning. Noting however that implementation of such planning in South Africa was hindered due to inadequate data and resource constraints, the author devised a phased approach for local emission reduction planning as documented in this section.

9.8.1 International Good Practice in Action Planning The following protocol for developing an emission reduction programme is recognised internationally as being in-line with good air quality management practices: a) Identification of pollutants to be controlled b) Identification of all sources of each pollutant - and for each source determine: • quantity of emissions (including temporal patterns in extent of emissions)

226

• percentage contribution to total emissions of a pollutant • height of emission - e.g. ground, medium elevated or high elevated source • likelihood of human exposure to emissions (exposure index) - e.g. emissions near population concentrations c) Identification and evaluation of air pollution reduction strategies: • list and description of possible strategies for each source • explanation of implementation of each measure • quantification of reduction of ambient concentrations as a result of implementation of each strategy through use of dispersion model analysis • cost-benefit analysis of controlling each source with each strategy. Cost-benefit analyses should include the consideration of: . source characteristics (i.e. percentage contribution, height of emission, exposure index) – rank source significance . reduction of ambient concentrations as a result of implementation of each strategy - identify most effective strategies for ambient pollution abatement . technical feasibility of each strategy . socio-economic effects of each strategy - determine the feasibility of strategies within the socio-economic context.

9.8.2 Action Planning within Initial AQMPs Despite acknowledging the value of the above protocol and attempting to implement it within the context of developing AQMPs for the City of Johannesburg and Ekurhuleni Metro, the implementation was restricted due to the following:  Priority pollutant identification was limited due to the absence of a comprehensive air quality monitoring network. This was partially overcome by making reference to monitoring data from various campaigns and non-governmental monitoring stations, and to monitoring results from similar regions.  Identification and quantification of sources was restricted by the absence of comprehensive emissions inventories. Although all available source and emissions data were collated, significant sources could have been omitted and emissions underestimated.  Although emission reduction measures were identified for significant sources, the reduction in air pollutant concentrations due to such sources could not be quantified due to the unavailability of accurate emissions data and an air dispersion model tailored for the region.  Rigorous assessment of the technical feasibility and socio-economic viability of emission reduction measures required additional research could not be undertaken within the accelerated AQMP development process. Close attention was however made to previously conducted studies on the feasibility of the measures being considered (e.g. the NEDLAC Dirty Fuels Study undertaken by the author, as documented in Chapters 4 to 7).

As a result of the aforementioned challenges, the Emission Reduction Programmes (Action Plans) developed for inclusion in the first AQMPs for the City of Johannesburg and Ekurhuleni Metro had the following key characteristics:  Emphasis was placed on the implementation of emission reduction measures for major sources. Given the need to focus resources on the establishment of an integrated air quality monitoring system in the short-term (as documented in Section 9.6), emission reduction

227

measures were carefully selected to target the most significant sources in terms of health effect potentials.

 Identification of sources for which the implementation of emission reduction measures in the short-term is justified. Sufficient evidence of significant effects associated with certain sources exists to justify the implementation of emission reduction measures for such sources in the short-term. Such evidence was based on an integrated assessment of preliminary emission estimates, air quality and deposition monitoring data and previous health risk studies. Examples of such sources are household fuel burning and fugitive dust from partially rehabilitated or disturbed mine tailings impoundments.

Vehicle tailpipe emissions were noted during the baseline assessments to be a significant emerging air pollution issues on the basis of: (i) preliminary emission estimates; (ii) projected increases in traffic volumes and congestion rates; (iii) upward trends in ambient

NOx concentration measurements near highways; and (iv) identification of this sector as being of primary concern by many developing and developed countries. Such information was considered to provide sufficient motivation for short-term actions to be taken aimed at addressing vehicle emissions in the medium- to long-term. Due to the need for inter- departmental collaboration to adequately reduce transportation emissions, such short-term actions focused predominantly on the establishment of coordination mechanisms as described in Section 9.7.2.

 Identification of sources for which further assessment is required to determine the need for and/or most suitable type of emission reduction measures implementable. Sources of concern in terms of air toxin and malodour releases include incinerators, landfills and waste water treatment works. Insufficient information was available to determine the effect of individual operations. Attention was therefore focused on the quantification of the effects of these sources and on the implementation of the minimum control requirements stipulated for such sources in the short-term.

Other sources that were not quantifiable in terms of emissions or effects included industrial and commercial fuel burning appliances, wild fires and fugitive releases from agricultural activities. The quantification of such sources and their effects in the short-term prior to implementing emission reduction measures was again recommended.

 Need to facilitate inter-departmental co-operation in the identification and implementation of emission reduction measures for certain sources. Local authorities were noted not to be directly responsible for the regulation of certain sources identified during the baseline assessments as potentially impacting significantly on pollution potentials within the region (e.g. mine tailings, waste incineration and disposal) (ref Section 9.7.2). For such sources, attention needed to be focused in the short-term on the establishment of inter-departmental co-operative structures or the effective utilization of existing structures to support the identification and implementation of emission controls in the medium- and longer-terms.

228

 Focus on the implementation of air quality management planning approaches by specific sources rather than on isolated individual emission reduction measures. AQM planning approaches were advocated rather than implementation of emission reduction measures in a fragmented manner (e.g. implementation of dust management planning by mines and integration of air quality issues into comprehensive environmental management assessment and planning approaches by landfill sites).

A synopsis of several of the short-term (years 1 – 2) and medium-term (years 3 – 5) source quantification and emission reduction measures considered for inclusion in the first EMM AQM Plan is given in Table 46 as an example.

229

Table 46: Source quantification and emission reduction measures consideration for inclusion in the first EMM AQM Plan (Scorgie et al., 2005)

Source Short-term Measures (Year 1 & 2): Medium-term Measures (Years 3 – 5): Sector: Household - Basa Njengo Magogo (top-down ignition approach) – Brownfield sites - Investigation of alternatives to household „dirty fuel‟ burning (low smoke fuels, renewable energy options, energy efficiency through retrofitting, energy fuel burning - Integration of energy efficient measures into new housing developments demand management) Mine tailings EMM to obtain support through existing regulatory structures for implementation of following EMM to request that Department of Minerals and Energy: measures: - Ensure that all mines have approved Environmental Management Programme - Emissions inventory compilation & reporting by mining companies to Metro Reports (EMPRs), can demonstrate and periodically report compliance to EMPR commitments related to air quality and have determined the financial - Integration of comprehensive dust management plan criteria into EMPR requirements quantum and provide for dust prevention and management - Dustfall monitoring & report to the Metro by open cast mines, mines with significant haul - Challenge applications for mineral right conversion if mines do not comply roads and/or tailings impoundments with EMPR commitments - Emissions inventory compilation & reporting by mining companies to Metro EMM will require that: - Implementation of local dustfall evaluation criteria to ensure that mitigative action is taken - Closing mines comply with closure commitments, specifically dust when alarm thresholds are exceeded management plans and rehabilitation objectives - Operating mines demonstrate compliance with dustfall guidelines and implement mitigative action when alarm thresholds are exceeded

Road - Establishment of inter-departmental transport liaison group - Establishment of inter-departmental transport & land-use planning liaison transport - Standardisation and improvement of vehicle emissions testing group - Collation of vehicle emission information for the purpose of improved emission estimation - Simulation of air pollutant concentrations associated with road transportation and impact modelling emissions - Integration of transport impact monitoring into EMM ambient air quality monitoring - Simulation of air quality implications due to implementation of selected activities transport management measures and transport projects. Findings to be - Identification of pertinent transportation management measures for possible inclusion in the communicated to Transportation Planning Dept. for consideration during EMM Integrated Transport Plan being developed decision making. - Facilitation of research into cleaner transportation technologies

Airports - EMM to require that all airports in the Metro compile an emissions inventory and report - EMM to determine need for smaller airports to conduct air quality impact source and emissions data assessments based on in-house dispersion model projections - Johannesburg International Airport (JIA), due to the extent of its operations, is required to - JIA required to undertake ambient air quality monitoring and emission conduct an air quality impact assessment management plan implementation

230

Source Short-term Measures (Year 1 & 2): Medium-term Measures (Years 3 – 5): Sector: Waste disposal Landfill operations: - Design & initiate an education & awareness campaign on waste segregation and treatment - All landfill operations within EMM required to meet Department of Water Affairs and - Commission a cost-benefit study on waste segregation and recycling strategies Forestry (DWAF) minimum requirements applicable for implementation within EMM - EMM personnel to conduct site inspections with DWAF and Gauteng Department of Agriculture, Conservation and Environment (GDACE) personnel & request representation at - Consolidate findings of investigations into alternative treatment and disposal DWAF meetings for local landfill sites options and support additional investigations where required. Integrate - Large, general and hazardous landfills (and landfills not complying with minimum findings on alternatives in EIA reviews and local waste management policies requirements) will be required to commission study comprising: emissions inventory - Collate source and emissions data for incinerator operations and undertake an compilation, monitor or model air pollutant concentrations, undertake health risk screening, air quality impact assessment, including a health risk screening study, to and if any pollutants flag – initiate quantitative health risk assessment and emission determine the acceptability incinerators for the purpose of informing the reduction planning atmospheric emission licensing process. - Investigate short-term methods of waste recycling - Conduct methane gas venting project (currently underway by EMM Municipal Infrastructure, Solid Waste) Incineration: - Investigate legal status of medical waste incinerators operation in EMM – notify DEA in instances where incinerators are unpermitted & require application for permit/license within specified time - Require proof of compliance with permit conditions including emission limits given for various gases, metals and TEQ Sewage & Waste Water Treatment Works: - Large works required to commission study comprising: emissions inventory compilation, monitor or model air pollutant concentrations, undertake health risk screening, and if any pollutants flag – initiate quantitative health risk assessment and emission reduction planning

231

Source Short-term Measures (Year 1 & 2): Medium-term Measures (Years 3 – 5): Sector: Industry & - Require concerns undertaking Scheduled Processes provide proof of registration - Review and revise atmospheric emissions licenses of Listed non-domestic under APPA and demonstrate compliance with permit conditions Activities, taking into account best practices and cumulative air fuel burning - Require industries (emitting above certain thresholds) to compile emissions pollutant concentrations. appliances inventories & report source & emissions information to the Metro - Set specifications on combustion efficiency applicable to all new - Encourage emission reduction planning in instances where operations are combustion devices anticipated to result in local air quality objective exceedances - Liaise with Eskom on demand side management measures applicable - Proponents of new development must prove compliance with local air quality to the commercial and industrial sectors. objectives (taking into account existing pollutant concentrations) and demonstrate - Investigate the potential for introducing alternative tariff structures for best practicable environmental option (BPEO) is being implemented the purpose of encouraging on-site co-generation and the introduction - EMM to collate source & emissions data for small-scale non-domestic fuel burning of renewables appliances, estimate emissions & simulate air pollutant concentrations - Investigation of the potential for introducing market incentives and - EMM to reinforce rule that all new non-domestic fuel burning appliances notify disincentives for the purpose of encouraging emission reduction by Metro and supply appliance and fuel info industrial and power generation processes.

Other sources - Identification and quantification of other sources of pollution. Attention given to: - Identify emission reduction measures for other sources predicted on vehicle entrainment from unpaved public roads, agricultural emissions, veld the basis of the quantitative emissions inventory and in-house burning and railway transport atmospheric dispersion modeling or external studies to be significant - Establish routine data retrieval mechanisms for purpose of updating emissions in terms of health risks or nuisance effects inventory - Control grass burning by municipal workers - Support national legislation aimed at controlling copper wire burning for purpose of wire stripping - Investigate the use of by-law implementation for the purpose of: (i) controlling track out from construction sites, (ii) stipulating the need for dustfall monitoring and reporting of results during large-scale construction and demolition projects.

232

9.8.3 Addressing Household Fuel Burning Human health effects related to inhalation exposures to household coal and wood burning emissions remain the most serious and pressing national air pollution problem. Addressing this issue remains a significant challenge for national and urban air quality management planning.

Achieving emission reductions and resultant health benefits for the domestic sector is not straight-forward. Complex factors contribute to the persistence of household fuel burning, including poverty, electrification backlogs and social factors as discussed in Section 3.2.1. Drawing on over three decades of experience, and referencing international developments, it is evident that a mix of instruments and interventions is required to achieve improvements in the short- and longer-terms. Interventions range from the integration of air quality considerations into housing planning processes, reducing energy requirements of dwellings through insulation and solar passive design, addressing electrification backlogs, refining combustion methods and appliances, and considering alternative fuels.

Understanding the technical and economic viability of interventions, in addition to the social- acceptability of such measures, is an imperative for effective emission reduction planning for the household sector. From a financial and economic perspective, low (or existing) technology interventions in the household was concluded based on the cost-benefit study conducted (ref Chapters 4 to 7), to yield significant benefit in the short to medium term. Such interventions include Basa Njengo Magogo (a stove ignition method), stove maintenance, electrification and housing insulation. Furthermore, it was determined that there are a sufficient number of households in the domestic sector to allow for the implementation of multiple interventions without the risk of the deterioration of benefit/cost ratio of interventions. The bulk of health cost savings due to reduced pollution from household fuel combustion would accrue to government, primarily due to reduced expenditure on public health care.

9.8.4 Emission Reduction Planning for Industrial Sources The ability of local authorities to curb emissions from industrial sources has historically been limited due to such sources being regulated through a system of permits (Registration Certificates) by national government under the APPA. Under the AQA, district and metropolitan municipalities are designated as the Emission Licensing Authority for Listed Activities operating within their jurisdictions. Local authorities are also mandated to establish emission limits for all source types on condition that such limits are more stringent than those issued by national and provincial government.

Although the AQA largely came into force in 2005, the sections of the AQA dealing with Listed Activities did not having been delayed pending the completion of various transitional phase projects including: the Listed Activities and National Emission Standard Setting Project and the APPA Registration Certificate Review Project. In the short-term, pending the entry into force of the Listed Activity sections of AQA, metropolitan municipalities were advised to ensure full involvement in the APPA Registration Certificate Review Project. This process maximised the capacity development opportunities offered and placed such municipalities on the road to successful regulation of the major industrial sources within their jurisdiction.

233

Methods developed and recommendations made in respect of emission standard setting and compliance monitoring, as documented in Chapter 8 (and formalised or to be formalised by the DEA in its official documentation), will be of considerable use to metropolitan municipalities following initiation of their Emission Licensing Authority functions.

9.8.5 Harmonisation of AQM Measures with Regional and Global Measures Air quality interventions must be aligned with measures aimed at addressing regional and global issues such as soil and water acidification, regional tropospheric ozone, peak NO2 vertical column densities, stratospheric ozone depletion and climate change. The importance of doing so is demonstrated in relation to climate change below.

Significant advances have been made in scientific understanding with regard to the linkages between climate change processes and air quality pollutants (DEFRA, 2007). By example, certain pollutants, such as NOx, SO2, ammonia and VOCs are precursors of secondary aerosols which have a negative (cooling) radiative forcing of climate and also influence the radiative properties of clouds. Black carbon aerosols, a product of incomplete combustion associated with household fuel burning and diesel vehicles, absorb solar radiation and exert a positive (warming) radiative forcing. Air pollutants may also indirectly affect the climate due to their effects on ecosystem sources and sinks which in turn affect CO2 and methane concentrations.

Although the use of climate models to investigate the effect of climate change on regional air quality requires further improvement, such models have supported certain preliminary projections. Increases in temperature due to climate change are, for example, expected to result in changes in the chemistry associated with ozone formation, largely as a result of changes in atmospheric water vapour concentrations. Changes in water vapour are conjectured to result in decreases in ozone in the background troposphere but result in increased ozone concentrations in more polluted areas where higher NOx concentrations occur (DEFRA, 2007). Increases in photochemical smog episodes during hot summers have been already been projected for various parts of Europe, the US and Australia based on current models.

Given the wide-ranging implications of climate change (including effects on local air quality) and the benefits achievable through coordinated action, the benefits of implementing „win win‟ solutions addressing air quality and climate change is evident. Cleaner production, energy efficiency, waste minimisation and resource recovery measures typically lend themselves to such solutions.

Various air quality measures including end of pipe controls (e.g. desulphurisation of power station flue gases) and fuel specifications to reduce transport emissions (e.g. lower fuel S limits) have the potential to increase CO2 emissions. Certain measures may also contribute to global warming by reducing the atmospheric loading of criteria pollutants with negative radiative forcing potentials as described above. It is therefore imperative that synergies be identified where measures to improve air quality can help to ameliorate climate change and that careful consideration be given to trade-offs where policy measures in the two areas act in opposition.

In assessing the broader environmental risks associated with sources of atmospheric emissions, and the environmental consequences of air quality interventions use can be made of Environmental Burden evaluations within the framework of fuel- and life-cycle analyses (LCA).

234

Recognition of the linkages between air quality and broader environmental issues, and specifically climate change, and commitments to maximising opportunities for concurrently addressing such issues should be included in the Air Quality Policy (ref Section 9.4). In instances where different divisions are responsible for air quality and climate change, mechanisms are needed to ensure integrated planning by such divisions.

9.9 Summary and Outlook Based on the efforts of a number of air quality practitioners during the 1980s and 1990s to effect legislative reform and establish air quality management planning in South Africa, this achievement was finally realised during the 2000s.

During her tenure as a doctoral candidate, the author contributed to the evolution of air quality management in South Africa by collaborating with local government in the development of the first air quality management plans. Addressing the need to tailor and cost-optimise international methods to suite local circumstances, the author devised several innovations to support urban air quality management. Such innovations included the following:

 Devising local policy frameworks to secure executive management commitment to air quality management.  Establishing local ambient air quality objectives to provide the basis for effects-based air quality management during 2003/2004, pending the publication of up-to-date national air quality standards.  Assisted in drafting SANS 1929:2004 – Ambient Air Quality – Limits for Common Pollutants, which informed the establishment of up-to-date national air quality standards in 2009.  Cost-optimisation of air quality management systems including: • Implementation of urban airshed models to supplement monitoring. • Integration of cost-effective monitoring techniques, such as passive diffusive and biomonitoring, to supplement continuous monitoring. • Staged air quality assessment approach for smaller/less polluted municipalities.  Contribution to capacity building including: • Development and presentation of a tailored, in-service air quality management courses for Ekurhuleni Metropolitan personnel (2005). • Collaboration with Professor Harold Annegarn on the development and presentation of a 10-day short course entitled „Introduction to Air Quality Management‟ at the University of Johannesburg (2006). This course continues to target local, provincial and national government officials tasked with air quality management functions and is being further developed and expanded by the University of Johannesburg and NACA. • Motivating and supporting the establishment of an agreement between NACA and DEA for the holding of joint annual air quality conferences (2006 – 2010).  Conceptualisation and recommendation of interim mechanisms for facilitating the inter- governmental cooperation and stakeholder consultation required to facilitate AQMP development and implementation.

235

 Recommendation of staged emission reduction planning to ensure major sources were addressed as a priority, with mechanisms and processes established in the short-term for addressing more complex, emerging issues in the medium- to long-term. Ongoing work by government, in collaboration with the private sector, which represent key developments in delivering effective air quality management include:  Provision of ore in-depth guidance on air quality modelling within the National Regulatory Dispersion Modelling Guideline due for completion in March 2012 (DEA, 2011d).  Development and implementation of the South African Air Quality Information System (SAAQIS) over the 2008 to 2017 period, which will significantly improve information available for AQMP development in future years.  Cost-optimisation of country-wide, standardised emission inventory development through the SAAQIS Phase II – National Atmospheric Emission Inventory Project due for completion in March 2014 (DEA, 2011d).  Revision and expansion of the DEA-NACA agreement to support joint conferences during the post-2010 period and on-going cooperation with regard to capacity building of air quality officials, public outreach and research and development (DEA, 2010b). Based on a critical analysis of the progress having been made to date, the author considers that further work is most needed in respect of the following:

 Collaboration between government, tertiary institutions and private sector practitioners in the development of a sustained and multi-pronged approach to meet requirements for skills in the short-, medium- and long-terms.  Expansion of training to planning and economic development officers to ensure the effectiveness of emission reduction planning. Learning from the experience of the UK, South Africa will need to ensure that developers and planners, including spatial development, transport, housing, infrastructure and energy planners are targeted for training, both in-service and at tertiary institutions.  Experience needs to be gained in the integration of air quality considerations into land use, energy and transportation planning.  Processes establishment for the harmonisation of urban air quality management planning policies with policies and programmes addressing regional and global air quality challenges such as regional ozone and climate change.

236

10 Summary, Conclusions and Outlook

The findings of individual study components are synthesised in this chapter as an overall discussion of the multiple factors contributing to air quality degradation in South Africa, the consequent human health, environmental and economic effects of this pollution, and the legal, technical and social measures implementable within a phased system of air quality management to deliver cleaner air in a manner that has positive outcomes for society and the economy. An overall concluding statement addressing the central question of the thesis is presented. The original research contribution of this work is highlighted and future research requirements to address gaps identified recommended.

10.1 Study Hypothesis The promulgation of the Air Quality Act represented a major step forward in the evolution of air quality management within South Africa. The historical debate regarding the practicability of effective air quality management is however ongoing. South Africa‟s continued dependence on coal to support its energy-intensive industrial and mining sectors, continued household fuel burning for space heating and cooking purposes within a number of areas, and the dire need for employment creation and focus on rapid development continue to challenge the realization of air quality improvements. Accelerated deployment of cost-optimised air quality management strategies and systems are needed for timely delivery of cleaner air whilst contributing to sustainable socio-economic development.

International experiences, approaches and systems can be drawn upon to fast track the development of air quality management strategies and systems for South Africa. Tailoring and cost-optimisation is however needed to meet local requirements. The evaluation of lessons learned internationally is important to avoid wasted resources on unsuccessful measures. Despite significant advances internationally in air quality management system deployment to identify and quantify sources and their effects, actions taken to address problems identified have had mixed success, with interventions for complex sources such as road transportation tending to be least successful. The main reasons for this are: (i) failure to integrate economic considerations into AQM planning; (ii) failure to integrate air quality considerations into development planning, frequently due to poor inter-departmental cooperation and lack of capacity building of economic development planners; and (iii) ineffective communication of technical information and actions required to decision makers.

Based on a detailed knowledge of international air quality management practices and an understanding of local circumstances and needs, the following hypothesis was posed:

It is possible through systematic analysis of the causes of air quality degradation; the evaluation of physical, ecological and health consequences thereof; and the evaluation of policies, legislation and technologies, to arrive at a rational system of air quality management that simultaneously can reduce atmospheric emissions, protect human health and the environment, promote socio-economic growth and

237

equity, and nevertheless result in a net positive contribution to the national economy and international competitiveness.

The aims of this study was to investigate the multiple factors that contribute to the degradation of air quality in South Africa, evaluate the consequent human health, environmental and economic effects of this pollution, and critically examine legal, technical and social measures that could be jointly deployed within an effective system of air quality management system for South Africa. An overview of the main study outcomes addressing the key objectives of the thesis is provided in subsequent subsections.

10.2 Integrating International Lessons Study Objective: Integrate lessons learned internationally into the evolution of local air quality management planning policies and processes.

International experience was drawn upon to identify air quality management approaches able to deliver cleaner air whilst realizing an overall positive outcome for society and the economy. Lessons addressing the environment-economic interface and the environmental-social interface which hold specific relevance for South Africa were identified.

Systematic integration of air quality considerations into energy, transportation, land use, housing and other development planning processes, represents a key requirement for effective and cost-efficient air quality management. A component of the „sustainable cities‟ approach, significant advances continue to be made internationally in this field. The integration of economic concerns in air quality management and planning represents a related imperative, with strategies applicable for adoption by South Africa including:  cost-optimisation of air quality management systems;  prioritization of sources based on their contributions to air pollutant concentrations and associated health and environmental effects;  phased regulation, with progressive tightening of requirements over time, taking into account the technical viability and costs of interventions;  implementation of emission reduction strategies for multiple pollutants, including management of precursors of secondary pollutants; and  adoption of a mix of instruments and interventions, including „command and control‟ measures, market instruments, voluntary and other measures.

International experience in regard to air pollution reduction through poverty alleviation was identified as being pertinent within South Africa‟s socio-economic context. In certain instances air pollution can be addressed through poverty alleviation and the provision of cost-effective options for adopting less polluting practices. Household coal and wood burning and informal waste burning may, for example, be addressed in this way. Considering the social-acceptability of interventions, in addition to their technological and economic viability, represents a further lesson learned based on international cases and reinforced by local experience.

238

10.3 Significant Sources, Priority Pollutants and Key Affected Areas Study Objective: Identify and quantify significant sources, priority pollutants and key affected areas within South Africa

High ambient SO2 and fine particulate matter concentrations due primarily to fuel burning within the household, industrial and power generation sectors represent on-going air pollution problems in many parts of South Africa. Elevated PM10 concentrations occur across the country, with widespread and frequent exceedances of health thresholds. Air quality standard exceedances due to SO2 are more localized (in vicinity of significant sources) and less frequent. Co-location of heavy industries and communities presents a continued source of health risks and consequent conflict, exacerbated by increased pressure to develop residential areas within former industrial areas or buffer zones around mine tailings facilities. Biomass burning, including agricultural burning and wild fires, represents an intermittent but seasonally significant source of fine particulate matter.

Emerging air pollution issues are associated with the transportation sector, particularly road transportation. Growth in vehicle activity and aging of the national vehicle fleet is projected to offset planned and proposed national emission reduction measures aimed at the regulation of fuel composition and new vehicle technology. Although air quality standards for NO2 and O3 protective of acute health effects are relatively infrequently exceeded within South African cities, an increasing trend in the concentrations of these pollutants is apparent in certain areas. The growth in vehicle activity is anticipated to contribute significantly to this trend.

Urban air quality management is primarily geared towards the control of anthropogenic sources. Given that anthropogenic fuel burning accounts for over 80% of ambient criteria pollutant concentrations within South African urban areas, a detailed analysis of such sources was undertaken. Source contributions to ambient air pollutant concentrations were quantified for several urban conurbations and industrial regions, with the most significant sources identified to be:  Residential coal and wood combustion;  Coal- and HFO-fired boilers;  Coal-fired power generation (particularly Mpumalanga and the Vaal Triangle); and  Vehicle emissions. Individual industrial operations found to contribute significantly to ambient pollutant concentrations in the areas investigated include pulp and paper mills, petroleum refineries, sugar mills, precious metal refineries, petrochemical plants, iron and steel works and brickwork operations.

10.4 Health Effects and Costs due to Fuel-Burning Sources Study Objective: Evaluate the effects on human health and welfare, and the economic costs of air quality degradation due to fuel-burning sources.

Adopting a damage function approach, the author quantified health effects and costs associated with anthropogenic fuel burning emissions for several conurbations which together account for 40% of South Africa‟s population. The study findings confirmed that human health effects related to inhalation exposures to household coal and wood burning emissions remain the most

239

serious and pressing national air pollution problem. Residential fuel burning was estimated to account for ~70% of all respiratory hospital admissions (RHA) and ~75% of all premature mortalities predicted to occur as a result of inhalation exposures to fuel-burning sources.

Vehicle emissions were associated with 12% of the estimated RHA due to inhalation exposures to fuel burning sources. Electricity generation was predicted to account for 6% of the RHA and 5% of the premature deaths respectively. Coal-fired boiler operations were the most significant industrial source grouping, estimated to account for 4% of the predicted RHA and mortality cases. The most significant individual point sources were identified as Highveld Steel & Vanadium (Mpumalanga), Mittal Steel Vanderbijlpark Works (Vaal Triangle), Sasol Secunda (Mpumalanga) and Hulett (eThekwini).

To put the aforementioned health effects into perspective, predicted effects due to anthropogenic fuel burning emissions were calculated as a percentage of the actual (measured) health effects due to all causes. Inhalation exposures to fuel burning emissions were predicted to be responsible for 18% to 24% of actual (recorded) respiratory hospital admissions. Inhalation exposures to air pollution from fuel burning are therefore estimated to be a significant risk factor in terms of these health endpoints. Inhalation exposures due to fuel burning emissions were found to be much less significant in terms of total mortality, cancer cases and heart disease as may be expected.

Total direct health costs related to inhalation exposures to fuel burning emissions were estimated to be in the order of R3.5 billion (2002 Rand) per annum across health effects, conurbations and source groupings. Household fuel burning was estimated to be responsible for ~68% of the total health costs estimated across all conurbations, vehicle emissions for 13%, industrial and commercial fuel burning for 13%, and power generation for ~6%. The greatest health costs were estimated to be incurred in Johannesburg, Ekurhuleni, Cape Town and eThekwini, with these conurbations accounting for approximately 76% of the estimated total health spending across all conurbations considered. The lower costs estimated for the Vaal Triangle and Mpumalanga Highveld were in part due to the smaller populations residing in these areas in comparison with the metropolitan areas.

Regional differences in source significance have been identified, demonstrating the importance of understanding local conditions in the tailoring of interventions. Household fuel burning accounted for over 75% of the projected total direct health costs in Cape Town, eThekwini, Johannesburg and Ekurhuleni and the Vaal Triangle. Health costs related to power generation and industrial fuel burning were estimated to be significantly higher in the Tshwane, Vaal Triangle and Mpumalanga Highveld regions compared to the contributions of these source types within other conurbations. Power generation was associated with over 50% of the total direct health costs projected to occur in the Mpumalanga Highveld region due to inhalation exposures to anthropogenic fuel burning emissions.

Health effects due to anthropogenic fuel-burning emissions were projected to have increased by ~10% during the 2002 to 2011 period across all conurbations considered. The 2011 projections were undertaken in 2004 based on assumptions regarding future trends in household fuel use, power generation, industrial growth and vehicle fleet growth. A critical analysis of these assumptions was undertaken in this thesis to determine the accuracy of the projections made

240

based on information available in 2011. „Business as usual‟ projections were found to be adequately representative for most sectors. However, the extent to which residential fuel burning was assumed to be addressed in the medium-term by electrification was noted to have been overstated.

It was assumed that residential fuel burning would persist in the short-term (2003 - 2007), but start to decrease in the medium-term (2007-2011) as a result of lower population growth rates and on-going electrification. According to recent figures, about 20% of households were not yet electrified by mid-2011 (Eskom, 2011). The price of electricity is also conjectured to have reversed the switch to electricity by some households, even when such households are grid- connected. It is therefore expected that emissions and associated effects and costs associated with household fuel burning projected for 2011 may be understated. Furthermore, the health effects projected do not account for indoor exposures due to residential fuel burning within poorly ventilated houses. The health effects and costs projected are therefore concluded to represent a lower bound estimate of the effects of anthropogenic fuel burning.

10.5 Cost-optimisation of Air Pollution Interventions Study Objective: Conduct a cost-optimisation evaluation of air pollution mitigation measures for significant anthropogenic fuel-burning sources.

Interventions identified to be potentially environmentally beneficial, technically viable and socio-economically acceptable were identified for significant sources. The author quantified emission reductions and resultant air quality improvements and health cost reductions achievable through the implementation of selected interventions. Results from this work represented the primary input for the economic assessment of interventions, as published by Leiman et al. (2007). In the economic assessment, costs associated with implementing interventions were quantified to facilitate an overall cost-benefit analysis of interventions.

Significant health effect reductions can be cost-effectively achieved through addressing residential fuel burning as a priority. The lower benefit-cost ratios associated with industrial and transport related interventions are likely to be due in part to these sources having been more effectively regulated historically with less onerous actions already implemented. Industrial emissions have been regulated under the Atmospheric Pollution Prevention Act, Act 45 of 1965. Various improvements have been made in the quality of both petrol and diesel. Fuel burning within low-income households has continued largely unabated and therefore presents greater opportunities for cost-effective abatement.

In addressing household fuel burning, low or existing technology interventions were determined to yield significant financial and economic benefit in the short to medium term. Such interventions include refined combustion practices (e.g. Basa Njengo Magogo, a stove ignition method), stove maintenance, electrification and housing insulation. The importance of adopting a range of interventions, including prescriptive methods, voluntary initiatives and market-based measures, to ensure economically efficient abatement was confirmed by study findings. It was determined that there are a sufficient number of households in the domestic sector to allow for the implementation of multiple interventions without the risk of the deterioration of benefit/cost ratio of interventions. Furthermore, the cost-benefit analysis provided economic justification for redistributional policies such as the provision of subsidised electricity or LPG to the urban

241

poor. The bulk of health cost savings due to reduced pollution from household fuel combustion would accrue to government, primarily due to reduced expenditure on public health care.

The study finding that the most cost-effective mitigation measures in the short-term are those targeting household fuel burning should not detract for the need for effective management of power generation, industrial and vehicle emissions. Based on international experience and local vehicle activity data it is evident that transport emissions are likely to become an increasingly significant contributor to urban air pollution. Health, plant and materials damage effects related to ozone concentrations, with vehicle emissions being a major source of ozone precursors, were not accounted for in the externalities study. Potential regional and global effects associated with industrial releases were also not accounted for given the urban focus of the study.

Electricity generation interventions implementing high technology solutions on the supply side, such as retrofitting desulphurization technologies at power stations and renewable energy, were concluded not to be feasible from an economic perspective in the short- and medium-terms. To pursue desulphurization, considerable cost would be incurred by both Government and households. Renewable energy is not as costly as desulphurization but, relative to the costs of other interventions, is regarded as costly. Legislation has been enacted to progressively phase in renewable energy in the longer term, with desulphurization under consideration for new coal- fired power stations.

10.6 Regulation and Compliance Monitoring of Industry Study Objective: Critically evaluate risk-based enforcement and compliance monitoring methodologies for use in the regulation of industrial activities, and to provide recommendations for national emission standards setting.

Direct regulation remains an important instrument in the control of industrial sources internationally, even given the implementation of market mechanisms and voluntary measures. Although the South African AQA has far reaching requirements for direct regulation of industries, resource constraints and the practicability of implementing requirements whilst not adversely impacting the national economy and international competitiveness, represent potential hurdles.

To support the overall hypothesis of the thesis, the author conjectured that it is possible to devise an effective, affordable, equitable and sustainable emissions control policy for South Africa by adopting a combination of best practice regulatory measures; such measures having been selected for their cost-effectiveness. Measures identified and tailored for implementation within South Africa included phased national standard setting, compliance promotion and structured self-monitoring, market-based instruments, and a risk-based enforcement and compliance monitoring strategy.

Based on the systematic investigation of international practices, and experience gained in tailoring selected practices for local implementation, the feasibility of devising an effective, affordable, equitable and sustainable emission control policy for South Africa has partially been demonstrated. Despite detailed cost-benefit analyses not having been completed, recommended approaches have by and large been adopted by the DEA following inter-governmental reviews and extensive consultation with interested and affected parties. Restricting the initial „list of

242

activities‟ to large industry types with significant emissions, and for which the benefits of regulation are expected to outweigh the costs based on international experience, substantially reduces the risk of economic effects arising due to the emission standards set. Additional measures taken which reduce the risk of economic effects during this initial phase include: (i) restricting pollutants for which emission standards are specified to the key ones for that industry type, thus reducing compliance monitoring and reporting costs; (ii) taking industry cycles into account in the setting of national minimum compliance timeframes; and (iii) making provision for industries to apply for extensions based on air quality assessments being undertaken.

In targeting industry sectors for which information on emissions and effects is less available or conclusive, particularly those comprising small and/or older operations, it is imperative that detailed cost benefit analyses be undertaken when setting BAT-based emission standards. Provision for such studies should be made so as to extend the list of activities and associated set of national emission standards in a manner which does not lead to economic effects or mass non-compliance.

10.7 Phased Air Quality Management Planning Study Objective: Integrate study findings into a phased system of air quality management, compatible with socio-economic development, and suitable for implementation within South Africa.

Addressing the need to tailor and cost-optimise international methods to suite local circumstances, the author devised several innovations to support local urban air quality management. The practicability of these innovations was tested in collaboration with local government during the development of the first AQMPs.

Devising air policy frameworks to secure executive management commitment to air quality management, and establishing local air quality objectives, was required to provide the foundation for air quality management. Conceptualisation and recommendation of interim mechanisms to facilitate inter-governmental cooperation and stakeholder consultation for AQMP development and implementation was needed to make integrated planning possible.

Recognising the hurdles posed by local resource constraints, innovations were sought to cost- optimise air quality management systems. Such innovations included the promotion of emissions inventory and urban airshed modelling to supplement monitoring, and the integration of cost-effective monitoring techniques (e.g. passive diffusive sampling, biomonitoring) to supplement continuous monitoring networks.

Staged emission reduction planning was identified as a necessary means of ensuring that major sources are addressed as a priority, with mechanisms and processes established in the short-term for addressing more complex and emerging issues in the medium- to long-term. Drawing on the cost-benefit study undertaken for anthropogenic fuel-combustion sources, household fuel

243

burning was identified for prioritized management in several South African cities. With a view to integrating air quality considerations into transportation, energy, household and spatial development planning in the longer-term, it was initially necessary to focus on establishing mechanisms for on-going inter-departmental cooperation.

10.8 Overall Conclusion The conclusion is reached that it is possible through systematic analysis of the causes of air quality degradation; the evaluation of physical, ecological and health consequences thereof; and the evaluation of policies, legislation and technologies, to arrive at a rational system of air quality management that simultaneously can reduce atmospheric emissions, protect human health and the environment, promote socio-economic growth and equity, and nevertheless result in a net positive contribution to the national economy and international competitiveness.

The integration and extension of several research projects enabled a detailed analysis of the multiple factors that contribute to the degradation of air quality in South Africa, and the evaluation of the consequent human health, environmental and economic effects of this pollution. Persistent and emerging sources, priority pollutants and effects and significant areas affected were established to focus air quality management efforts. Following problem identification, a critical evaluation was undertaken of the legal, technical and social measures to could contribute to the development of a cost-optimised system of air quality management in South Africa. Measures were identified and devised, and the joint implementation of such measures within a rational system of air quality management, compatible with socio-economic prosperity and equity, conceived for South Africa.

Significant progress is being made in South Africa in the development and deployment of the cost-optimised measures proposed, with a coherent system of air quality governance emerging through continued collaboration by all tiers of government, civil society and business. The uptake of strategies and systems proposed provides partial confirmation of their compatibility with South Africa‟s socio-economic growth and equity goals. Continued diligence and commitment is needed for full delivery of the system, and specifically to ensure successful interventions for complex sources by integrating air quality considerations within socio- economic development planning.

Although progress is being made with regard to the training of environmental practitioners, the expansion of in service and tertiary training for planning and economic development practitioners to support proactive emission reduction planning has yet to be realised. Given that the entire system of air quality management conceived rests on the availability of human resources, more extensive collaboration between government, tertiary institutions and the private sector is needed to ensure the availability of skills in the short-, medium- and long- terms.

10.9 Summary of Contribution Through diligent primary data collection, primary modelling, assessment of international best practice, participation in reformulation and implementation of regulations, the author made an original contribution to the evolution of a novel system of air quality management. This unique

244

system is tailored to meet South Africa‟s sustainable development, environmental justice, and environmental protection requirements.

The specific contributions made by the author to the air quality governance cycle is schematically illustrated in Figure 58 with reference made to research integrated within this thesis and related work drawn upon but not documented in detail in this thesis. All components of the environmental governance cycle must be in place and operating effectively for successful air quality management to be achieved.

Figure 58: Contributions to the evolving system of air quality governance in South Africa (adapted DEA, 2007b).

During the execution of the research for this thesis the author played a novel actor-observer role, as an independent consultant and researcher, advisor to national government responsible for the process of delivering air quality governance, and simultaneously maintaining a role as a critical observer of the process. This resulted in a thesis of unusual reflexive structure, in which the technical report of the task accomplished is itself an outcome of the underlying theories and practices uncovered as part of the research. The unique outcome is a functioning, innovative system of air quality management informed for a large part by the outcomes of a series of studies and interventions in the air quality sector, civil society, industry and government spanning a period of fifteen years.

245

10.10 Future Research Recommended areas for future research are as follows:

 Establishment of cost-effective methodologies for vehicle activity modelling to facilitate improved estimation and spatial distribution of road transportation emissions.

 Development of cost-effective methods for modelling ozone concentrations and associated exposures and risks for routine deployment within South Africa to inform decision making. The methodology established by Yarwood et al. (2011) provides an example of how cost-effective methods could be established for use within South Africa.  Cost-optimisation of cost benefit analysis methods which can be applied in South Africa to assess the viability of applying BAT-based emission standards for smaller industrial operations. The aim of such assessments being to support the extension of the list of activities and association emission standards without resulting in negative economic effects or mass non-compliance.  Design of a sustainable and multi-pronged approach for delivery of the skills required for effective air quality governance in the short-, medium- and long-terms. The research work should draw on international and local experience, account for local conditions, and engage the collaboration of government, tertiary institution and private sector practitioners.  Tailoring of methodologies for harmonizing stategies aimed at addressing local, regional and global air pollution related issues, taking South Africa‟s goals and responsibilities in terms of air quality management and climate change into account.  Further characterisation of spatial and temporal variations in biomass burning to facilitate the successful inclusion of this source within air quality modelling. Biomass burning modelling undertaken by Pitts (2011) within Western Australia provides an example of how this work may be approached. The cost-optimisation of emission reduction strategies targeting sources of ozone precursors and fine particulate matter represents a common challenge facing regulatory agencies tasked with air quality management within both developed and developing countries. Once industrial and on- road mobile sources are subject to stringent controls, additional emission reductions become progressively more costly for such sources. Attention then turns to the benefits of controlling dispersed minor sources which, when their emissions are aggregated, contribute significantly to overall emissions. Off-road diesel equipment and widespread use of consumer products such as surface coatings represent examples of such sources.

The investigation of measures to control dispersed minor sources of ozone and fine particulate matter precursors and assessment of benefits and costs of such measures is recommended as a longer term research objective. Indoor air pollution, unrelated to residential fuel-combustion, represents a further field requiring additional study locally in the longer term. Once South Africa has successfully addressed priority sources such as residential fuel burning, coal-fired power generation, industrial emissions and on-road vehicles, such research will support continued improvement in air quality in line with international practices.

246

References

ABSA (2002). Industrial Growth Projections for South Africa, ABSA Investments, Pretoria. AEAT Environment (2004). A Comparison of EU Air Pollution Policies and Legislation with Other Countries, Study No. 1 of the Series: Environmental measures and Enterprises Policy, Study commissioned by the European Commission Directorate-General for Enterprise. Afrane-Okese, Y., (1998). Domestic energy use database for integrated energy planning. Unpublished MSc thesis, Energy and Development Research Centre. Cape Town: University of Cape Town. Altshuller, A.P., W.B. Johnson, J.S. Nader, B.L. Niemann, D.B. Turner, W.E. Wilson and G. D‟Alessio (1980). Transport and fate of gaseous pollutants associated with the National Energy Program, Environmental Health Perspectives, 36, 155-179. Amoako, J. and M. Lodh (2003). The economic consequences of the health effects of transport emissions in Australian capital cities. Prepared by the Bureau of Transport and Regional Economics, Canberra, for the 32nd Australian Economists‟ Conference, 29 September to 2 October 2003, Canberra. Andreae, M.O., E. Atlas, H. Cachier, W.R. Cofer, G.W. Harris, G. Helas, R. Koppman, J. Lacaux, and D.E. Ward (1996). Trace gas and aerosol emissions from savannah fires, in J.S. Levine (Editor), Biomass Burning and Global Change, MIT Press, Cambridge, pp. 278-294. Annegarn, H.J. and M.R. Grant (1999). Direct Source Apportionment of Particulate Pollution within a Township, Final Report submitted to the Department of Minerals and Energy, Low Smoke Coal Programme, 10 July 1999, Pretoria. Annegarn, H.J. and Y. Scorgie (1995). Air Quality Management Strategy for the Vaal Triangle, Report compiled on behalf of the Department of Environmental Affairs and Tourism, Pretoria. Annegarn, H.J. and Y. Scorgie (1997a). An Air Quality Management Strategy for the Vaal Triangle, Final Report on the Vaal Triangle Intervention Project, The Clean Air Journal, 9(7), June 1997, pp. 9 – 17. Annegarn, H.J. and Y. Scorgie (1997b). An Air Quality Management Strategy for the Vaal Triangle Part II, Final Report on the Vaal Triangle Intervention Project, The Clean Air Journal, 9(8), November 1997, pp. 11 - 20. Annegarn, H.J. and Y. Scorgie (1998c). An Air Quality Management Strategy for the Vaal Triangle Part III, Final Report on the Vaal Triangle Intervention Project, The Clean Air Journal, 10(1), May 1998, pp. 3 – 17. Annegarn, H.J. and Y. Scorgie (2002). Science and Policy of Air Quality Standards Setting Processes, Paper presented at the National Association Clean Air Workshop 'Progress regarding Air Quality Legislation in South Africa: Implications for Stakeholders - First Theme: Ambient Air Quality Standard Formulation & Implementation', 19 July 2002, Sandton, South Africa. Annegarn, H.J. and J.S. Sithole (1999). Soweto Air Monitoring – SAM Trend Analysis of Particulate Pollution 1992 – 1999, Supplementary Report to the 1998 Annual Report, Report AER99.163 S SOW, 10 January 1999.

247

Annegarn, H.J., G.M. Braga-Marcazzan, E. Cereda, M. Marchionni and A. Zucchiatti (1992), Source profiles by unique ratios (SPUR) analysis - Determination of source profiles from receptor-site streaker samples, Atmospheric Environment, 26A, pp. 333-343. Annegarn, H.J., M.R. Grant, M.A. Kneen and Y. Scorgie (1999). Direct Source Apportionment of Particulate Pollution within a Township, Report for the Department of Minerals and Energy, Low Smoke Coal Programme, Annegarn Environmental Research Report No. AER98.117 DME , July 1999, Pretoria. Annegarn, H.J., L.W. Burger, J. John, N. Krause, M. Naidoo, Y. Scorgie, R. Taviv and M. Zunckel (2007). National Air Quality Management Programme (NAQMP) Output C.4, Initial State of Air Report, Department of Environmental Affairs and Tourism, May 2007, Pretoria. ARTD Management and Research Consultants (2000). Evaluation of the National Pollutant Inventory Program, Final Report to Chemicals and the Environment Branch, Environment Australia, November 2000. ASTM D1739-70 (1970). ASTM Standard D1739-70: Standard Test Method for Collection and Measurement of Dustfall (Settleable Particulate Matter),‟ ASTM International: West Conshohocken, PA, 4 pp. Batchvarova, A.E. and S.E. Gryning (1990). Applied model for the growth of the daytime mixed layer, Boundary-Layer Meteorology, 56, pp. 261-274. Bates, J., J. Cadman and M. Holland (2003). Environmental Burden Measures for Air: Global Warming, Stratospheric Ozone Depletion, Photochemical Ozone Creation and Airborne Acidification, Research and Development Technical Report P6-015/TR2, Project undertaken by AEA Technology Environment under sub-contract to WRC plc, Environment Agency, ISBN: 1844320367. Baird, M. and Y. Scorgie (2006). APPA Registration Certificate Data Base Report, Report compiled on behalf of the Department of Environmental Affairs and Tourism, Report APP/DEAT/2006, Airshed Planning Professionals Pty Ltd, Midrand. Beattie, C.I., J.W.S. Longhurst, A. Simmons, N.K. Woodfield (2000). Implementation of air quality management in urban areas of England - the role of transport planning, in L. Sucharov (Editor), Urban Transport V, Urban Transport and the Environment for the 21st Century. WIT Press. Beattie, C.I., J.W.S Longhurst and N.K Woodfield, N.K.(2001). Air quality management: evolution of policy and practice in the UK as exemplified by the experience of English local government, Atmospheric Environment, 35, pp. 1479 – 1490. Bentley West Management Consultants and Airshed Planning Professionals (2004). Study to Facilitate the Formulation of a Strategy to Reduce the Health Impacts Associated with Air Pollution from Fuel Combustion in South Africa, Task 5a and Task 5b, Report Prepared on behalf of National Economic Development and Labour Council (NEDLAC) under the Fund for Research into Industrial Growth and Equity (FRIDGE), Bently West Management Consultants, Johannesburg. Bernstein, J.D. (1993). Alternative Approaches to Pollution Control and Waste Management, Regulatory and Economic Instruments, World Bank, Washington DC.

248

Bickel, P. and R. Friedrich (2004) (Editors). ExternE Externalities of Energy, Methodology 2004 Update, European Commission, Directorate-General Science, Research and Development. Blackman, A and W. Harrington (1998). Using Alternative Regulatory Instruments to Control Fixed Point Air Pollution in Developing Countries: Lessons from International Experience, Discussion Paper 98-21, Resources into the Future, Washington DC. Boegman N., W.E. Scott, and S.J. Lennon (1996). Air Quality Management in South Africa, in G. Held, B.J. Gore, A.D. Surridge, G.R. Tosen, C.R. Turner and R.D. Walmsley (Editors), Air Pollution and its Impacts on the South African Highveld, Environmental Scientific Association, Cleveland, 144 pp. Borysiewicz, M.J., I. Garanty, A. Kozubal and W. Kacpryzyk (2006). Health and Environmental Risk Assessment of Different Fuel Cycles in Electricity Generation, CoE MANHAZ Institute of Atomic Energy and Institute of Environmental Protection, Warsaw, 157 pp. Bournay, E. (2012). Urban and transport related energy consumption. Norway, United Nations Environment Programme/Global Resource Information Database Arendal Maps and Graphics Library, 4 June 2012. Bradshaw, D., P. Groenewald, R. Laubscher, N. Nannan, B. Nojilana, R. Norman, D. Pieterse and M. Schneider (2003). Initial Estimates from the South African National Burden of Disease Study, 2000, Medical Research Council Policy Brief, No. 1, March 2003. Bradshaw, D. and K. Steyn (Editors) (2001). Poverty and Chronic Disease in South Africa, Joint SA Medical Research Council and World Health Organisation report. Bradshaw, D., M. Schneider, R. Laubscher and B. Nojilana (2002). Cause of death profile South Africa, 1996, Burden of Disease Research Unit, Medical Research Council, May 2002. Britton, M.S.T. (1998). Low-smoke Fuel Programme: Laboratory Technical Tests, Determination of Emission Factors, Department of Minerals and Energy, Report No. ES9606, July 1998. Burger, L.W. and R.G. Thomas (2002). Ambient Air Sampling Campaign: Selected Heavy Density Vehicle Activity Sites, Report EMS/02/SASOL-01 Rev 0.2, September 2002. Burger, L.W. and Y. Scorgie (2002). Air Quality Legislation: Potential for Non-compliance, paper presented at the National Association Clean Air Workshop 'Progress regarding Air Quality Legislation in South Africa: Implications for Stakeholders - First Theme: Ambient Air Quality Standard Formulation & Implementation', 19 July 2002, Sandton, South Africa. CARB (2002). Staff Report: Public Hearing to Consider Amendments to the Ambient Air Quality Standards for Particulate Matter and Sulfates, Prepared by the Staff of the California Air Resources Board and the Office of Environmental Health Hazard Assessment, May 2002. CEPA (1998). National Ambient Air Quality Objectives for Particulate Matter. Part 1: Science Assessment Document, Report by the Canadian Environmental Protection Agency (CEPA) Federal-Provincial Advisory Committee (FPAC) on Air Quality Objectives and Guidelines. Carnevali, D. and C. Suarez (1993). Electricity and the environment: Air pollutant emissions in Argentina, Energy Policy, 21(1), as referenced in Fouquet et al. (2001).

249

Carson, D.J. (1973). The development of a dry, inversion-capped, convectively unstable boundary layer, Quarterly Journal of the Royal Meteorological Society, 99, pp. 450-467. City of Johannesburg (2000). State of Environment Report. CCT (2003). State of Environment Report for the City of Cape Town. CEC – Commission of the European Communities, DG XII (Editor) (1995). ExternE: Externalities of Energy, Vol. 1-6. Office for Official Publications of the European Communities, Luxembourg. CEPA/FPAC Working Group (1998). National Ambient Air Quality Objectives for Particulate Matter. Part 1: Science Assessment Document, A Report by the Canadian Environmental Protection Agency (CEPA) Federal-Provincial Advisory Committee (FPAC) on Air Quality Objectives and Guidelines. Chartered Institute of Environmental Health (2004). Industrial Pollution Control by Local Authorities – a Management Guide, Second Edition, Chartered Institute of Environmental Health, London, September 2004. Chow, J.C. and J.G. Watson (2011). Air quality management of multiple pollutants and multiple effects, Air Quality and Climate Change, 45(3), pp. 26-32. Corden, C. and A. Ritchie (2006). Assessment of options to streamline legislation on industrial emissions, Presentation to the Meeting of the IPPC Advisory Group, 8 December 2006. Cosijn, C., (1996). Elevated Absolutely Stable Layers. A Climatology for South Africa, Unpublished Msc Thesis. Proposal submitted to the Department of Geography and Environmental Studies, University of the Witwatersrand, Johannesburg. Cumo, F., D.A. Garcia, L. Calcagnini, F. Cumo, F. Rosa and A.S. Sferra (2012). Urban policies and sustainable energy management, Sustainable Cities and Society, Volume 4, pp 29-34. DEA (2010a). National Environmental Management: Air Quality Act, 2004 (Act No. 39 of 2004) – List of Activities which result in atmospheric emissions which have or may have a significant detrimental effect on the environment, including health, social conditions, economic conditions, ecological conditions or cultural heritage, Government Gazette No. 33064, No. 248 of 2010, 31 March 2010, Department of Environmental Affairs, Pretoria. DEA (2010b). The 2010 National Air Quality Officer‟s Annual Report on Air Quality Management in the Republic of South Africa, Initial Working Draft, January 2011, Department of Environmental Affairs, Pretoria. DEA (2011a). Department of Environmental Affairs Annual Performance Report 2010/2011, Pretoria.

DEA (2011b). Proposed National Ambient Air Quality Standard for PM2.5, Government Gazette No. 34493, 5 August 2011, Department of Environmental Affairs, Pretoria. DEA (2011c). Draft National Dust Control Regulations, South African Government Gazette, No. 34307, No. 309 of 2011, Department of Environmental Affairs, Pretoria. DEA (2011d). National Air Quality Officers News, 5(2), July-September 2011, Department of Environment and Tourism, Pretoria.

250

Deakin, M. (2011). Meeting the challenge of learning from what works in the development of sustainable communities, Sustainable Cities and Society, 1(4), December 2011, pp 244–251.

DEAT (2006). National Air Quality Management Programme, Output C.4, Technical Compilation to Inform the Initial State of Air Report, Department of Environmental Affairs and Tourism, Pretoria, May 2007. DEAT (2006). Identification of Substances in Ambient Air and Establishment of National Standards for the Permissible Amount or Concentration of Each Substance in Ambient Air, Department of Environmental Affairs and Tourism, Notice 528, Government Gazette No. 28899 of 9 June 2006. DEAT (2007a). Discussion Document, The Establishment of National Standards for Ambient Air Quality in terms of Section 9(1)(A) and (B) of the National Environmental Management: Air Quality Act 2004 (Act No. 39 of 2004) issued by the Department of Environmental Affairs and Tourism, Chief Directorate of Air Quality Management and Climate Change, 30 October 2007, Reference No. CD:AQM&CC/30/10/07/1. DEAT (2007b). The 2007 National Framework for Air Quality Management in the Republic of South Africa, 11 September 2007, Department of Environmental Affairs and Tourism, Pretoria. DEAT (2008). Manual for Air Quality Management Planning, South African Department of Environmental Affairs and Tourism, Pretoria. DEAT (2009). National Environmental Management: Air Quality Act, 2004 (Act No. 39 of 2004) – National Ambient Air Quality Standards, South African Government Gazette No. 32816, 24 December 2009, Department of Environmental Affairs and Tourism, Pretoria. DEFRA (2003). Sector Guidance Note IPPC SG6, Secretary of State‟s Guidance for the A2 Surface Treatment Using Organic Solvents Sector, Department for Environment Food and Rural Affairs, Crown Copyright, United Kingdom. DEFRA (2003). Part IV of the Environment Act 1995, Local Air Quality Management, Policy Guidance, LAQM.PG(03). Department for Environment Food and Rural Affairs, Crown Copyright, United Kingdom. DEFRA (2004). Valuation of the external costs and benefits to health and environment of waste management options, Final report for DEFRA by Enviros Consulting Limited in association with EFTEC, Department for Environment Food and Rural Affairs, Crown Copyright, United Kingdom. DEFRA (2005). Integrated Pollution Prevention and Control, A Practical Guide, Edition 4, Department for Environment Food and Rural Affairs, Crown Copyright, United Kingdom. DEFRA (2007). Air Quality and Climate Change: A UK Perspective, Air Quality Expert Group, Report Published for Department for Environment, Food and Rural Affairs; Scottish Executive; Welsh Assembly Government; and Department of the Environment in Northern Ireland, London, Crown, 270 pp. DEFRA (2011). Air Pollution Health Bandings, Department for Environment Food and Rural Affairs, United Kingdom, http://uk-air.defra.gov.uk/air-pollution/bandings, website accessed December 2011.

251

Department of Health (1998). Demographic and Health Study 1998, Pretoria. Department of Health (2001). South Africa Demographic and Health Survey 1998. Pretoria. Department of Energy (2010). South African Energy Synopsis 2010, South African Department of Energy, Pretoria. Department of Energy (2011). Discussion Document on the Review of Fuel Specifications and Standards for South Africa, South African Government Gazette No. 34089, No. 204, 8 March 2011, Pretoria. DETR (2000a). The air quality strategy for England, Scotland, Wales and Northern Ireland. Working together for clean air, Department of Environment, Transport and Region, CM 4548, The Stationary Office Ltd., London. DETR (2000b). Framework for review and assessment of air quality. Report LAQM.G1(00), Department of Environment, Transport and Region , The Stationery Office Ltd., London. DETR (2000c). Developing local air quality action plans and strategies: the principal considerations, Department of Environment, Transport and Region, Report LAQM.G2(00), The Stationery Office Ltd., London. DME (2003). Integrated household clean energy strategy prospectus, Internal document, Department of Minerals and Energy, Pretoria. Dockery, D.W. C.A. Pope (1994). Acute Respiratory Effects of Particulate Air Pollution, Annual Review of Public Health, 15, pp. 107 - 132. Dockery, D.W., C.A. Pope, X. Xiping, J. Spengler, M.F. Ware, B. Ferris and F. Speizer (1993). An Association between Air Pollution and Mortality in Six US Cities, New England Journal of Medicine 329(24), pp. 1753-59. Doppegieter, J.J, J. du Toit and E. Theron (2000). Energy Futures 2000/2001, Department of Minerals and Energy, November 2000. Eberhard, A. (2011). The Future of South African Coal: Market, Investment, and Policy Challenges, Program on Energy and Sustainable Development, Working Paper No. 100, Freeman Spogli Institute for International Studies, Standford University, Standford, California. EC (2007). Beyond Regulatory Compliance, Incentives to improve the environmental performance of IPPC installations, ENV.C.4/SER/2005/0034r, Final Report, European Commission, April 2007. The Economist Automotive Briefing and Forecasts (2010). South Africa: Automotive Report, 1 August 2010, http://www.eiu.com/index.asp?layout=ib3Article&article_id=147377999&pubtypeid=11124 62496&country_id=1660000166&category_id=775133077&rf=0. Ecoserv (1998). Durban Metro: Methodology and Results of Emissions Inventory, Report compiled on behalf of Durban Metro Wastewater Management Department, Reference no. DM1_1997, February 1998. Ecoserv (2000). South Durban Sulphur Dioxide Management Systems Steering Committee: Emissions Inventory for South Durban, Reference no. DSEI1_2002, April 2003.

252

EEA (1999). Copert III. Computer Programme to Calculate Emissions from Road Transport. Methodology and Emission Factors, European Environment Agency, July 1999. Ehrlich, C. and W-D. Kalkoff (1998). Domestic Coal Combustion - One of the Major Sources of Air Pollution, Papers of the 11th World Clean Air and Environment Congress entitled "The Interface between Developing and Developed Countries", 13-18 September 1998, Durban, South Africa. Engelbrecht, J.C. (2006). A Generic Framework for an Air Quality Management Plan for South Africa, Unpublished Doctor Technologiae thesis, Department of Environmental Health, Tshwane University of Technology, November 2006. Entec (2006). Assessment of Options to Streamline Legislation on Industrial Emissions, Draft Final Report, Report compiled by Entec UK Ltd on behalf of the European Commission, December 2006. Erbes, R.E. (1996). A Practical Guide to Air Quality Compliance, Second Edition, John Wiley & Sons, New York. ERI (2001). Preliminary Energy Outlook for South Africa, Energy Research Institute, University of Cape Town, 10 October 2001. Eskom (2000). Annual Report, Eskom Holdings, Johannesburg. Eskom (2002). Annual Report, Eskom Holdings, Johannesburg. Eskom (2011). Eskom‟s Electrification Programme, Presentation to the Energy Portfolio Committee, Parliamentary Monitoring Group, 14 June 2011, Pretoria, http://www.pmg.org.za/report/20110614-salga-implementation-intergrated-national- electrification-programme-p. eThekwini Health Department (2004). Air Quality Monitoring Network: Annual Report 2004, Durban. European Union (1996). Air Quality Framework Directive, Council Directive 96/62/EC. European Union (2008). Directive of the European Parliament and of the Council on Ambient Air Quality and Cleaner Air for Europe, PE-CONS 3696/07, Brussels, 28 March 2008. Fenger, J., O. Hertel and F. Palmgren F (Editors) (1998). Urban Air Pollution – European Aspects, Kluwer Academic Publishers, Dordrecht/Boston/London. Ferreira, I.W. and H.R. Lloyd (2002). Developmental Issues and Environmental Policy in South Africa, Africa Insight, 32(1), pp. 20-24. Fisher, G.W., K. A. Rolfe, T. Kjellstrom, A. Woodward, S. Hales, A. P. Sturman, S. Kingham, J. Petersen, R. Shrestha, and D. King (2002). Health effects due to motor vehicle air pollution in New Zealand, Report to the Ministry of Transport, 20 January 2002. Fouquet, R., R. Slade, V. Karakoussis, R. Gross, A. Bauen and D. Anderson (2001). External Cost and Environmental Policy in the United Kingdom and the European Union, Centre for Energy Policy and Technology, Occasional Paper 3, Imperial College of Science, Technology and Medicine, London, June 2001, ISBN 1903144027. Furtado, R.C. (1996). The incorporation of environmental costs into power system planning in Brazil, Unpublished PhD Thesis, Imperial College, London as referenced in Fouquet et al. (2001).

253

Friedrich, R. and A. Voss (1993). External costs of electricity generation, Energy Policy, February 1993, pp. 114-122. Gerardi, W. and Y. Scorgie (2011). Cost Benefit Analysis of Options to Manage Non-road Diesel Engine Emissions, Project undertaken on behalf of the Council of Australian Governments (COAG) Standing Council on Environment and Water, ENVIRON Australia Project Number AS121234, October 2011, Sydney. Glicksman, R.L, D.L. Markell D.R. Mandelker A.D Tarlock F.R.Anderson (Editors) (2003). Environmental Protection: Law and Policy, Aspen Publishers, New York, 4th edition. Graham, J.A.N. (1997). The determination of emissions, efficiencies and the cost-effectiveness of various domestic appliances used for cooking and heating in South Africa, Journal of Energy in Southern Africa, November 1997, pp. 105-112. Graham, J.A.N and R.K. Dutkiewicz (1999). Assessing the Emissions and Cost Effectiveness of Traditional Household Fuel Burning Appliances in South Africa, The Clean Air Journal, 10(3), pp. 13-21. Gray, J., T. James, and J. Dickson (2007). Integrated regulation – experiences of IPPC in England and Wales, Water and Environment Journal, 21, pp. 69-73. Gwaze, P. (2010). Air Quality Modelling Framework: Feedback from Working Group, Presentation to the 5th Annual Air Quality Governance Lekgotla, 11-13 October 2010, Polokwane, Limpopo Province, South Africa. Helas, G. and J.J. Pienaar (1996). The influence of vegetation fires on the chemical composition of the atmosphere, South African Journal of Science, 92, pp. 132 - 136. Held, G., B.J. Gore, A.D. Surridge, G.R. Tosen, C.R. Turner and R.D. Walmsley (Editors) (1996). Air Pollution and its Impacts on the South African Highveld, Environmental Scientific Association, Cleveland, 144 pp. Hersh, R. (1996). A Review of Integrated Pollution Control Efforts in Selected Countries, Discussion Paper 97-15, Resources for the Future, Washington DC, December 1996. Hietkamp, S. and V. Nkhwashu (2005). The National Air Quality Management Programme: The Phase II – Transition Project, Report compiled on behalf of the Department of Environmental Affairs and Tourism, 17 May 2005, Pretoria. Holland, M. and P. Watkiss (2002). BeTa - Benefits Table database: Estimates of the marginal external costs of air pollution in Europe, Report compiled for European Commission Director General of the Environment. IRIS (1998). US-EPA's Integrated Risk Information Data Base, United States Environmental Protection Agency, available from www.epa.gov/iris. Irurah, D.K. (Editor) (2000). Environmentally Sound Energy Efficient Low-Cost Housing Study: Evaluation of Performance and Affordability of Intervention Technologies, Report Submitted to USAID and the Environmentally Sound Low Cost Housing Task Team (ESLCHTT).

254

Jagusiewicz, A. (2004). Inconsistencies between different EU requirements with respect to NECs affecting the preparation of the national programme in Poland-a way forward, Presentation given at the Workshop on Plans and Programmes of Air Quality and National Emission Ceilings Directives, Brussels, Borschette building, 1-2 September 2004. Josipovic, M. (2010). Acidic Deposition Emanating from the South African Highveld – A Critical Levels and Critical Loads Assessment, PhD dissertation, University of Johannesburg, South Africa. Josipovic, M., M. Zunckel, H.J. Annegarn, M.A. Kneen, S.J. Piketh and J.J Pienaar (2007). Comparison of modelled and passive sampling ozone concentrations over southern Africa, 2007 SASAS Conference, Exploring Interfaces with the Atmosphere, 13 & 14 September 2007. Josipovic, M., H.J. Annegarn, M.A. Kneen, A.J. Piketh and J.J. Pienaar. (2007). A Regional-

Scale Passive Monitoring Study of SO2, NO2 and O3 in South Africa, IGACtivities Newsletter, Issue No. 36, July 2007. John, J., R. Oosthuizen, M. Schwab, B. Wells, E. Hagelaar and M. Gondwe (1998). Atmospheric Volatile Organic Compounds in South African Metropolitan Areas, Papers of the 11th World Clean Air and Environment Congress, 13-18 September 1998, Durban, South Africa. Kim, C. and C. Qiang (2002). Air Quality in the People's Republic of China, Korean Environment Institute. Koenig, J.Q. and J.F. Mar (2000). Sulphur dioxide: Evaluation of Current California Air Quality Standards with Respect to Protection of Children. [Online] Available: www.arb.ca.gov/ch/ceh/aqstandards/oehhaso2.contractorreport091200.pdf. [6 March 2003]. Krzyzanowski, M, B. Kuna-Dibbert, and J. Schneider (2005) (Editors). Health effects of transport-related air pollution, 318 pp., World Health Organisation, Geneva. Kunzli, N., R. Kaiser and S. Medina (2000). Public health impact of outdoor and traffic related air pollution: A European assessment, Lancet, 356, 9232, pp. 795-801. Larssen, S., Shah J., Murray F. and Suzuki K. (2003). Comparative Approaches to Air Quality Management in Asia, USA, Europe, Japan and Australia, Paper Presented at the Better Air Quality 2003 Conference in Manila, the Philippines, 17-19 December 2003. Leiman, A., B. Standish, A. Boting and H. van Zyl (2007). Reducing the healthcare costs of urban air pollution: The South African experience, Journal of Environmental Management, 84, pp. 27-37. Liebenberg-Enslin, H. And G. Petzer (2005). Background information document –air quality baseline assessment for the city of Tshwane metropolitan municipality, Report compiled by Airshed Planning Professionals Pty Ltd on behalf of the Tshwane Metropolitan Municipality, Report Number.: APP/05/CTMM-01 Rev 0. Maddison, D. (1997). A Meta-analysis of Air Pollution Epidemiological Studies, Centre for Social and Economic Research on the Global Environment, London, University College London and University of East Anglia.

255

Maenhaut, W., I. Salma, J. Cafmeyer, H.J. Annegarn and M.O. Andreae (1996). Regional atmospheric aerosol composition and sources in the eastern Transvaal, South Africa, and impacts of biomass burning, Journal of Geophysical Research, 101, pp. 23.631-23.650. Mahema, T. and Gwaze P. (2011). 2011 State of Air Report and National Air Quality Indicator, Presentation to the 6th Annual Air Quality Governance Lekgotla, October 2011, East London. Mahema, T., M. Baird, and Y. Scorgie (2006). APPA registration certificate electronic database status and information on scheduled processes operating in South Africa, Paper presented at the South African National Association for Clean Air (NACA) Annual Conference and Workshop, 18 to 20 October 2006, East London. Marbek Resource Consultants (2003). Benchmarking Regulatory Regimes of Petroleum Refineries, Final Report, Prepared for Canadian Council of Ministers of the Environment, May 2003. Martins, J.J., R.S. Dhammapala, G. Lachmann, C. Galy-Lacaux and J.J. Pienaar (2007). Long- term measurements of sulphur dioxide, nitrogen dioxide, ammonia, nitric acid and ozone in southern African using passive samplers, South African Journal of Science, 103, 336-342. Mathee, A. And Y. von Shirnding (2003). Air quality and health in Greater Johannesburg, in G. McGranahan and F. Murray (Editors), Air Pollution and Health in Developing Countries, Earthscan, London, pp 207-219. Mathee, A., H.B. Rollin, N. Bruce and Y.E.R von Schirnding (2001). Fuel use and indoor air quality in electrified and un-electrified rural South African villages, Proceedings for the National Association for Clean Air Conference, Port Elizabeth. Matzopoulos, R. and G. Carolissen (2006). Estimating the incidence of paraffin ingestion, African Safety Promotion: A Journal of Injury and Violence Prevention, 3(2), pp. 4-14. Medley, A.J., C.M. Wong, T.Q. Thach, S. Ma, T.H. Lam and H.R Anderson (2002). Cardiorespiratory and all-cause mortality after restrictions on sulfur content of fuel in Hong Kong: an intervention study, Lancet, 360(9346), pp.1646-1652. Milieu Ltd (2004). Assessment of the Effectiveness of European Air Quality Policies and Measures, Case Study 2 – Comparison of the EU and US Air Quality Standards and Planning Requirements, A Project for DG Environment carried out by Milieu Ltd, the Danish Environmental Research Institute and the Center for Clean Air Policy, 4 October 2004. Morawska, L., M.R. Moore, and Z.D. Ristovski (2005). Health Impacts of Ultrafine Particles, Desktop Literature Review and Analysis, A consultancy funded by the Australian Government Department of the Environment and Heritage, Canberra. Neary, L., J.W. Kaminski and the MAQNet Team (2006). Evaluation of GEM-Chemistry Modelling and Data Assimilation System: Five year global simulation results from GEM- AQ, in The Remote Sensing of Atmospheric Constituents from Space ACCENT- TROPOSAT-2 (AT2): An ACCENT Integration Task, Combining satellite observations to study atmospheric Composition, Report from the Fifth AT2 Workshop, University of Crete, Voutes, Heraklion Crete, 3-4 July 2006.

256

Nelson, C.D. (2000). The Public Health Impacts of Particulate Emissions from Coal-fired Power Plants in Minnesota, Unpublished M.Sc. Dissertation, Faculty of the Graduate School of the University of Minnesota, October 2000. NEPC (1998). National Environment Protection Measure for the National Pollutant Inventory, Australian National Environment Protection Council, 27 February 1999. NEPC (2001). National Environment Protection (Diesel Vehicle Emissions) Measure, Australian National Environment Protection Council, 29 June 2001. NEPC (2003). Impact Statement for the National Environment Protection (Air Toxics) Measure, Australian National Environment Protection Council, May 2003. NEPC (2004). National Environment Protection (Air Toxics) Measure – Explanatory Document, Australian National Environment Protection Council, April 2004. NEPC (2001-2). National Environment Protection Council, Annual Report 2001-2, Australia. Nethconsult – BKH Consulting Engineers (2002). Regional Environment Accession Project, Project: Requirements for Monitoring of the Pollutant Releases from Installations, IPPC Directive 96/61/EC for Poland, Final Report Part A Review of Three EU Countries, Delft, the Netherlands, 3 May 2002. NSCA (2004). Air Quality Limit Values – Is there a better way? National Society for Clean Air, United Kingdom. NSW DEC (2004). Regulatory Impact Statement – Proposed Amendment to the Protection of the Environment Operations (Clear Air) Regulation 2002, Department of Environment and Conservation, New South Wales, November 2004. OECD (1994). Internalizing the social cost of transport. Organisation for Economic Co- operation and Development, Paris. OECD (2005). Implementing Sustainable Development, Key Results 2001 – 2004, Organisation for Economic Co-operation and Development, Paris. OECD (2006). Environmental Compliance and Enforcement in India: Rapid Assessment, OECD Programme of Environmental Co-operation with Asia, December 2006.

Ojelede, M. E. (2004). NOx production by lightning over southern Africa, MSc dissertation, University of the Witwatersrand, Johannesburg, South Africa. Ojelede, M. E., H. J. Annegarn, C. Price, M. A. Kneen and P. Goyns (2008). Lightning

produced NOx budget over the Highveld region of South Africa, Atmospheric Environment, 42, pp. Ostro, B. (1994). Estimating the Health Effects of Air Pollution: A Method with an Application to Jakarta, Policy Research Working Paper 1301, World Bank Policy Research Department, Washington DC. Ostro, B. (1996). A methodology for Estimating Air Pollution health Effects. World Health Organisation, Report WHO/EHG/96.5, Geneva. Ostro, B. (2004). Outdoor air pollution: Assessing the environmental burden of disease at national and local levels, Environmental Burden of Disease Series, No. 5, World Health Organization Protection of the Human Environment, Geneva.

257

Ott, W. (1996). External Costs and External Price Addings in the Swiss Energy Sector, in O. Hohmeyer, R.L. Ottinger and K. Rennings (Editors) Social Costs and Sustainability - Valuation and Implementation in the Energy and Transport Sector, Berlin: Springer-Verlag, pp. 176-183. PDG (2006). Compliance Monitoring System and Implementation Plan, Output A1, Status Quo Report, Section B – Literature Review and Case Studies, Report compiled by Palmer Development Group on behalf of the Department of Environmental Affairs and Tourism. Pearce, D., C. Bann and S. Georgiou (1992). The Social Costs of Energy, CSERGE - Centre for Social and Economic Research on the Global Environment, University College London and University of East Anglia. Piketh, S.J. (1995). Generation and transportation characteristics of suspended particles in the eastern Transvaal, Unpublished M.Sc. Dissertation, University of the Witwatersrand, Johannesburg. Piketh, S.J., H.J. Annegarn and M.A. Kneen (1996). Regional scale impacts of biomass burning emissions over southern Africa, in J.S. Levine (Editor), Biomass Burning and Global Change, MIT Press, Cambridge, pp. 320-326. Pitts, O. (2011). Modelling Cumulative Impacts of Bushfire and Industrial Emissions in the Pilbara Region of Australia Using TAPM-CTM, Paper presented at the Annual Clean Air Society of Australia and New Zealand (CASANZ) Conference, 30 July to 2 August 2011, Auckland, New Zealand. PMG (2011). SALGA recommendations for Integrated National Electrification Programme; Roll out of INEP in partnership with municipalities and Solar Water Heaters program: Eskom briefing, Parliamentary Monitoring Group, Pretoria, http://www.pmg.org.za/report/20110614-salga-implementation-intergrated-national- electrification-programme-p. Poland Ministry of the Environment (2000). The Second National Environmental Policy, Warsaw, December 2000. Poland Ministry of the Environment (2003). Poland‟s Climate Policy: The strategies for greenhouse gas emission reductions in Poland until 2020, Warsaw, October 2003. Pope III, C.A. and D.W.C. Dockery (2006). Health Effects of Fine Particulate Air Pollution: Lines that Connect, Journal of Air & Waste Management Association, 56, pp. 709-742. Preston-Whyte, R.A. and P.D. Tyson (1988). The Atmosphere and Weather over South Africa, Oxford University Press, Cape Town, 374 pp. Rauland, V. and P. Newman (2011). Decarbonising Australian cities: A new model for creating low carbon, resilient cities, 19th International Congress on Modelling and Simulation, Perth, Australia, 12–16 December 2011. Rayner, S. (2003). National Association of Automobile Manufacturers of South Africa, personal communication, April 2003. Rayner, S. (2010). South African motor industry experience and new vehicle technology related fuel requirements, National Association of Automobile Manufacturers of South Africa, Paper presented at the UNEP Partnership for Clean Fuels and Vehicles Conference in Abidjan, August 2010.

258

Reitze, A.W. (2001). Air Pollution Control Law: Compliance & Enforcement, Environmental Law Institute, Washington DC. Robe, F.R. and J.S. Scire (1998). Combining Mesoscale Prognostic and Diagnostic Wind Models: A Practical Approach for Air Quality Applications in Complex Terrain, Preprints 10th Joint Conference on the Applications of Air Pollution Meteorology, 11-16 January 1998, Phoenix, Arizona. Salma, I., W. Maenhaut, J. Cafmeyer, H.J. Annegarn and M.O. Andreae (1992). PIXE analysis of cascade impactor samples collected at the Kruger National Park, South Africa, Nuclear Instruments and Methods in Physics Research B, 85, pp. 849-855. SANS 69:2004. South African National Standard - Framework for setting and implementing national ambient air quality standards, Standards South Africa, Pretoria. SANS 1929:2004. South African National Standard - Ambient Air Quality - Limits for common pollutants, Standards South Africa, Pretoria. SAPIA (2005). Annual petrol and diesel fuel sales data for the 1994 to 2004 period received from the South African Petroleum Industry Association for use in the NEDLAC Dirty Fuel Study. Scire, J.S. and F.R. Robe (1997). Fine-Scale Application of the CALMET Meteorological Model to a Complex Terrain Site, Paper Presented at the Air & Waste Management Association‟s 90th Annual Meeting and Exhibition, June 8-13 1997, Toronto, Ontario, Canada. Schneider, C. (2000). Related Health Benefits of Reducing Power Plant Emissions, Prepared by ICF Consulting for Clear Air Task Force, Boston, October 2000. Schulze, B.R. (1986). Climate of South Africa. Part 8: General Survey, WB 28, Weather Bureau, Department of Transport, Pretoria, 330 pp. Scorgie, Y. (2001a). Air Quality Management System Development and Implementation in South Africa, Paper presented at the African Centre for Energy and Environment Air Quality Workshop, 7-9 March 2001, Warmbaths, South Africa. Scorgie, Y. (2001b). Air Quality Management Systems and their Application within Cape Town, Paper presented at the NACA Western Cape Symposium, 18 May 2001, Gordon‟s Bay, South Africa. Scorgie, Y. (2004a). Air Quality Situation Assessment for the Vaal Triangle Region, Report compiled by Matrix Environmental Consultants on behalf of the Legal Resources Centre, Report Number MTX/02/LRC-01b, Johannesburg. Scorgie, Y. (2004b). Air Quality Management Plan Development Process for Khayelitsha and the City of Cape Town, Report compiled by Airshed Planning Professionals Pty Ltd for the City of Cape Town, Report No. APP/04/CCT-03, 24 August 2004. Scorgie, Y. (2005). Situation Analysis and Air Quality Management Plan for the Greater Alexandra Area, Report compiled on behalf of the Gauteng Department of Housing by Airshed Planning Professionals Pty Ltd, Report No. APP/04/GDH-02, 6 June 2005.

259

Scorgie, Y. (2006). Prioritisation of Registration Certificates for Review as Part of the Atmospheric Pollution Prevention Act (APPA) Registration Certificate Review Project, Report compiled on behalf of the Department of Environmental Affairs and Tourism, Chief Directorate: Air Quality Management. Scorgie, Y. (2008). Guideline Inspection Protocol for Listed Activities holding Atmospheric Emission Licenses under the National Environmental Management: Air Quality Act, Guideline compiled on behalf of the Department of Environmental Affairs and Tourism, Directorate Compliance Monitoring, February 2008. Scorgie, Y (2009a). Cleaner Non-road Diesel Engine Project – Identification and Recommendation of Measures to Support the Uptake of Cleaner Non-road Diesel Engines in Australia, Final Report, Revision 2, Project undertaken on behalf of the Australian Government Department of the Environment, Water, Heritage and the Arts and the New South Wales Department of Environment, Climate Change and Water, ENVIRON Australia Project Number AS121015, December 2009, Sydney. Scorgie, Y. (2009b). VOCs from Surface Coatings – Assessment of the Categorisation, VOC Content and Sales Volumes of Coating Products Sold in Australia, Project undertaken on behalf of the Australian Commonwealth Environment Protection Heritage Council, ENVIRON Australia Project Number AS120900, Sydney. Scorgie, Y. (2010). Review of Interstate and International Emission Standards and Options for Improving Air Quality in New South Wales, Final Report – Revision 2.0, Report compiled on behalf of the New South Wales Department of Environment, Climate Change and Water, ENVIRON Australia Project Number AS121110, Sydney. Scorgie, Y. and H.J. Annegarn (2002). The Impact of Coal on People, Paper presented at the Coal and Sustainable Development Conference, hosted by Minister of Minerals & Energy, Official Parallel Event of the Johannesburg World Summit, Eskom Conference Centre, Midrand, RSA, 29 August 2002. Scorgie, Y. and G. Kornelius (2007). Air Quality Implementation: Listed Activities and Minimum Emission Standards, Output B.1 – International Review, Report compiled on behalf of the Department of Environmental Affairs and Tourism, Chief Directorate: Air Quality Management, Pretoria. Scorgie, Y. and G. Kornelius (2009). Modelling of Acid Deposition over the South African Highveld, Paper presented at the annual South African National Association for Clean Air Conference, 14-16 October 2009, Vanderbijlpark, South Africa. Scorgie, Y. and J. Malan (2002). Proactive Air Pollution Prevention and Air Quality Management - The Responsibility of Industry under the National Air Quality Management Legislation, paper presented at the National Association for Clean Air (NACA) Conference, 23-25 October 2002, Durban, South Africa. Scorgie, Y. and C. Venter (2005). National State of the Environment: Atmosphere and Climate. Report compiled on behalf of the Department of Environmental Affairs and Tourism, Pretoria. Scorgie, Y. and R.M. Watson (2004). Updated Air Quality Situation Analysis for the City of Cape Town, Report compiled for the City of Cape Town, Airshed Planning Professionals Pty Ltd, Report No. APP/04/CCT-02, 23 August 2004.

260

Scorgie, Y. and J. Greenwood (2011). VOC Emission Reductions Achievable through National Product-based Regulation of Surface Coatings, Paper presented at the Annual Clean Air Society of Australia and New Zealand (CASANZ) Conference, 30 July to 2 August 2011, Auckland, New Zealand. Scorgie, Y., H.J. Annegarn and L.W. Burger (2004a). Socio-Economic Impact of Air Pollution Reduction Measures - Task 1: Definition of Air Pollutants Associated with Combustion Processes, Report compiled on behalf of National Economic Development and Labour Council (NEDLAC) under the Fund for Research into Industrial Growth and Equity (FRIDGE), Airshed Planning Professionals Pty Ltd, Midrand. Scorgie Y, Annegarn HJ and Burger LW (2007). Air Pollution over South Africa: Persistent Problems and Emerging Issues, 14th International Union of Air Pollution Prevention and Environmental Protection Associations (IUAPPA) World Congress, 10-14 September 2007. Scorgie, Y., H.J. Annegarn and L. Randell (2003b). Air Quality Situation Assessment for the City of Johannesburg, Final Report, Report compiled for the City of Johannesburg, Matrix Environmental Consultants, Report No. MTX/03/JHB-01d, 23 January 2003. Scorgie, Y., H.J. Annegarn and L. Randell L (2003c). Air Quality Management Plan for the City of Johannesburg, Air Quality Management Plan compiled on behalf of and in consultation with the Department of Development Planning, Transportation and Environment and the Department of Environmental Health, City of Johannesburg, Matrix Environmental Consultants, Report No. MTX/03/JHB-01e, 23 September 2003. Scorgie, Y., H.J. Annegarn, A. Richter, K. Ross, and L.W. Burger (2007). Comparison of

SCIAMACHY NO2 Observations over Southern Africa with Air Quality Modelling Results, IEEE International Geoscience and Remote Sensing Symposium (IGARSS), Barcelona, Spain, 23 – 27 July 2007. Scorgie, Y., L.W. Burger and H.J. Annegarn (2002). Integrated Air Quality Management Systems for Southern Africa, paper presented at the 3rd International Conference on Environmental Management, 27-30 August 2002, Eskom Conference Centre, Midrand, South Africa. Scorgie, Y., L.W. Burger and H.J. Annegarn (2003d). Review of International Air Quality Guidelines and Standards for the Purpose of Informing South African Air Quality Standards, Report compiled on behalf of the South African Bureau of Standards (SABS) Technical Committee on National Air Quality Standards, Report No. SANS/03/01 Rev. 0, 5 March 2003. Scorgie, Y., L.W. Burger and H.J. Annegarn (2004b). Socio-Economic Impact of Air Pollution Reduction Measures - Task 2: Establishment of Source Inventories, and Task 3: Identification and Prioritisation of Technology Options, Report compiled on behalf of National Economic Development and Labour Council (NEDLAC) under the Fund for Research into Industrial Growth and Equity (FRIDGE), Airshed Planning Professionals Pty Ltd, Midrand.

261

Scorgie, Y., L.W. Burger and H.J Annegarn (2004c). Socio-Economic Impact of Air Pollution Reduction Measures - Task 4: Quantification of Environmental Benefits Associated with Fuel Use Interventions, Report compiled on behalf of National Economic Development and Labour Council (NEDLAC) under the Fund for Research into Industrial Growth and Equity (FRIDGE), Airshed Planning Professionals Pty Ltd, Midrand. Scorgie, Y., L.W. Burger and P. Hoets (2002). Environmental Benefits, Economic Feasibility and Social Acceptability of Low-smoke Fuels: Synopsis of Results from Field Experiments, paper presented at the Coal & Environment Conference, held by the Fossil Fuel Foundation, 5 - 8 August 2002, Mintek, Johannesburg, South Africa. Scorgie, Y., L.W. Burger, and M. Sowden (1999). Application of Source-Receptor Modelling to Regional Air Quality Management, Paper presented at the National Association for Clean Air Conference, „Into the Next Millennium‟, 6-8 October 1999, Cape Town. Scorgie, Y., L.W. Burger and M. Sowden (2001). Analysis, Synthesis and Consolidation of the Qalabotjha and eMbalenhle Experiments, Including Other Investigations, Report compiled on behalf of the Department of Minerals and Energy, Environmental Management Services Report No. EMS/01/DME-01, 11 June 2001, Pretoria. Scorgie, Y., T. Fischer, and R. Watson (2005). Air Quality Management Plan for the Ekurhuleni Metropolitan Municipality, Plan compiled on behalf of and in consultation with the Department of Environment & Tourism, Ekurhuleni Metropolitan Municipality, Report No. APP/04/EMM-02c, Airshed Planning Professionals Pty Ltd, Midrand. Scorgie, Y., M.A. Kneen, H.J. Annegarn and L.W. Burger (2003a). Air Pollution in the Vaal Triangle – Quantifying Source Contributions and Identifying Cost-effective Solutions‟, Clean Air Journal, 12(4), November 2003. Scorgie, Y., G. Paterson, L.W. Burger, H.J. Annegarn, and M.A. Kneen (2004d). Socio- Economic Impact of Air Pollution Reduction Measures – Task 4a Supplementary Report: Quantification of Health Risks and Associated Costs Due to Fuel Burning Source Groups, Report compiled on behalf of National Economic Development and Labour Council (NEDLAC) under the Fund for Research into Industrial Growth and Equity (FRIDGE), Airshed Planning Professionals Pty Ltd, Midrand. Scorgie, Y., T. Radebe, and R. Watson (2004). Development of an Air Quality Management Plan for the Ekurhuleni Metropolitan Municipality, Paper presented at the National Association for Clean Air (NACA) Conference, October 2004, Johannesburg. Scorgie, Y., R. Watson, T. Fischer, and L. van der Walt (2004e). Background Information Document. Air Quality Baseline Assessment for the Ekurhuleni Metropolitan Municipality, Report compiled on behalf of Ekurhuleni Metropolitan Municipality, Report No. APP/04/EMM-01rev1, Airshed Planning Professionals Pty Ltd, Midrand. Seethaler, (1999). Health Costs due to Road Traffic-related Air Pollution. An Impact Assessment Project of Austria, France and Switzerland, Prepared for the World Health Organisation Ministerial Conference on Environment and Health, London, Geneva. Segal, L. (1999). Review of Health Costs of Road Vehicle Emissions, Centre for Health Program Evaluation, Working Paper 94, West Heidelberg, Victoria, Australia, August 1999, ISBN 876662069.

262

Sengupta, B. (2004). Experience of air pollution control in last two decades in India, Paper presented at the International Conference on Better Air Quality, held at Agra during December 6 – 8, 2004. South African Institute for Race Relations (2000). Household Energy Use Data, Pretoria. Sowden, M. (1998). Macro-Scale Experiment Environmental Sampling and Air Quality, Unpublished Report, Atomic Energy Corporation, September 1998. Sowden, M., M. Zunckel and M. van Tienhoven (2007). Assessment of the status of biogenic organic emissions and impacts on air quality in southern Africa, Tellus, 59B, 535-541. Statistics South Africa (2001). Census 2001, Pretoria. (Detailed demographic database acquired from Statistics South Africa with figures for each enumerator area.) Statistics South Africa (2002). Causes of death in South Africa 1997-2001, Advanced release of recorded causes of death, Reference No. P0309.2, Pretoria. Stone, A. (2000). South African Vehicle Emissions Project: Phase II, Final Report: Diesel Engines, Report prepared for the Department of Minerals and Energy, Pretoria, February 2000. Strimaitis, D.G., J.S. Scire, and J.C. Chang (1998). Evaluation of the CALPUFF Dispersion Model with Two Power Plant Data Sets, Preprints 10th Joint Conference on the Applications of Air Pollution Meteorology, 11-16 January 1998, Phoenix, Arizona. Taljaard, J.F. (1998). Indoor Air Quality at Qalabotjha, CSIR and Enpro Report, September 1998, Pretoria. Terblanche, P. (1995). Motor Vehicle Emissions Policy Development: Phase 1, Department of Minerals and Energy, Report No. EV9404, June 1995, Pretoria. Terblanche, P. (1996). Impacts of Removing Air Pollution: Health Aspects, Report compiled on behalf of Department of Minerals and Energy, Report No. ES 9411, May 1996, Pretoria. Terblanche, P. (1998). Vaal Triangle Air Pollution Health Study. Bibliography, Summary of Key Findings and Recommendations, Medical Research Council, July 1998, Pretoria. Terblanche, P. and A.S. Pols (1994). Characterisation of Risk Factors Associated with Household Fuel Usage in South Africa, Report compiled on behalf of Department of Minerals and Energy, Report No. EO 9303, July 1994, Pretoria. Terblanche, P., R. Nel, G. Reinach, and L. Opperman (1992). Personal Exposures to Total Suspended Particulates from Domestic Coal Burning in South Africa, NACA Clean Air Journal, 8 no. 6, pp. 15-17. Terblanche, P., I.R. Danford and C.M.E. Nel (1993). Household Energy Use in South Africa, Air Pollution and Human Health, Journal of Energy in Southern Africa, May 1993, pp. 54- 57. Terblanche, P., M.E. Nel, L. Opperman and H. Nyikos (1993). Exposure to Air Pollution from Transitional Household Fuels in a South African Population, Journal of Exposure Analysis and Environmental Epidemiology, 3, pp. 15-22. Terblanche, P., I.R. Danford, and A.S. Pols (1995). Comparative Evaluation of Human Exposures to Air Pollution from Low-Smoke and Conventional Household Coal Usage, Journal of Energy in Southern Africa, August 1995, pp. 131-136.

263

University of Cape Town (2004). Socio-Economic Impact of Air Pollution Reduction Measures - Task 5b, Socio-economic Impacts associated with Prioritised Options, Report compiled in collaboration with Bentley West Management Consultants on behalf of National Economic Development and Labour Council (NEDLAC) under the Fund for Research into Industrial Growth and Equity (FRIDGE), Cape Town. United Nations (1999). Strategies and Policies for Air Pollution Abatement, Economic Commission for Europe, Geneva. US-AID (2004). Vehicle Inspection and Maintenance Programs: International Experience and Best Practices, A Report for the Office of Energy and Information Technology, Prepared by PA Government Services Inc., Bureau for Economic Growth, Agriculture and Trade, U.S. Agency for International Development, Washington DC. US-EPA (1986). Air Pollution: Improvements Needed in Developing and Managing EPA‟s Air Quality Models, GAO/RCED-86-94, B-220184, United States Environmental Protection Agency, General Accounting Office, Washington, DC. US-EPA (1995a). A User‟s Guide for the CALPUFF Dispersion Model, EPA-454/B-95-006, United States Environmental Protection Agency, Research Triangle Park, North Carolina. US-EPA (1995b). A User‟s Guide for the CALMET Meteorological Model, EPA-454/B-95- 002, United States Environmental Protection Agency, Research Triangle Park, North Carolina. US-EPA (1996). AP42 Emission Factors for Residential Wood Burning, United States Environmental Protection Agency, Research Triangle Park, North Carolina. US-EPA (1998). AP42 Emission Factors for External Combustion of Bituminous Coal, United States Environmental Protection Agency, Research Triangle Park, North Carolina. US-EPA (2005). Report on Environmental Compliance and Enforcement in India, December 2005, United States Environmental Protection Agency, Research Triangle Park, North Carolina. van Horen, C. (1996). The Cost of Power: Externalities in South Africa's Energy Sector, Energy & Development Research Centre, University of Cape Town, Cape Town. Van Niekerk, W.C.A. (1998). Environmental Health Risk Assessment Exposure to Emissions from Low-Smoke Fuels, Infotox Unpublished Report, September 1998, Pretoria. Van Niekerk, W.A. (2001). Technical Background Document for the Development of a National Ambient Air Quality Standard for Sulphur Dioxide, Report compiled for the Department of Environmental Affairs and Tourism, Pretoria. van Niekerk, A.S. and P.A. Swanepoel (1999). Sasol Synthetic Fuels: eMbalenhle Air Quality Project, Phase Two, Indoor Air Quality and Desirability, Final Report, Report No. 98/14, 29 January 1999, Report compiled for the Department of Minerals and Energy, Pretoria. Van Nierop, P.G. (1995). An Emission Inventory of Particulate Air Pollution in the Vaal Triangle, MSc Dissertation, Faculty of Engineering, University of the Witwatersrand. Wang, X. and D.L. Mauzerall (2006). Evaluating impacts of air pollution in China on public health: Implications for future air pollution and energy policies, Atmospheric Environment, 40 (2006) pp. 1706-1721.

264

Watkiss, P., D. Forster, A. Hunt, A. Smith, and T. Taylor (2004). A Comparison of EU Air Pollution Policies and Legislation with Other Countries, Study No. 1 of the Series: Environmental measures and Enterprises Policy, Study commissioned by the European Commission Directorate-General for Enterprise, AEA Technology Environment, United Kingdom. Webb, A.H. and G.C. Hunter (2004). Managing Power Station Ambient Air Quality Compliance, in World Bank (2006). India: Strengthening Institutions for Sustainable Growth, Country Environmental Analysis, October 2006, Geneva. Wong, C.T. (1999). Vehicle Emission Project (Phase II) Final Report, Department of Minerals and Energy, February 1999, Pretoria. Wong, C.T. and R.K. Dutkiewicz (1998). Evaporative Emissions Project. Final Report, Department of Minerals and Energy, Pretoria. WHO (2000). Guidelines for Air Quality, World Health Organisation, Geneva. WHO (2005). Revised Guidelines for Air Quality, World Health Organisation, Geneva. WHO (2007). Health Relevance of Particulate Matter from Various Sources, Report on a WHO Workshop, Bonn, Germany 26-27 March 2007, World Health Organisation Regional Office for Europe. Wicking-Baird, M.C., M.G. de Villiers, and R.K. Dutkiewicz (1997). Cape Town Brown Haze Study. Energy Institute Report, Report Number. GEN 182, Cape Town. Winkler, H. (Ed) (2006). Energy Policies for Sustainable Development in South Africa, Options for the Future, Energy Research Centre, University of Cape Town, April 2006, ISBN: 0- 620-36294-4, pp 225. World Bank (1998). The Effects of Pollution on Health: The Economic Toll, Pollution Prevention and Abatement Handbook, 426 pp., The World Bank, Washington. World Bank (2002). Improving Air Quality in Metropolitan Mexico City, An Economic Valuation, Environmentally and Socially Sustainable Development Sector Unit, World Bank, Policy Research Working Paper 2785, The World Bank, Washington. World Bank (2006). India: Strengthening Institutions for Sustainable Growth, Country Environmental Analysis, October 2006, The World Bank, Washington. Yarwood, G., Y. Scorgie, N. Agapides, E. Tai, P. Karamchandani, K. Bawden, J. Spencer and T. Trieu (2011). A screening method for ozone impacts of new sources based on high-order sensitivity analysis of CAMx simulations for Sydney, Paper presented at the 2011 Annual Community Modeling and Analysis System (CMAS) Conference, 24-26 October 2011, University of North Carolina. Zunckel, M., J. John, R. Taviv and M. Naaidoo (2007). National Air Quality Management Programme, Output C.4, Technical Compilation to Inform the Initial State of Air Report, Department of Environmental Affairs & Tourism, May 2007. Zunckel, M., Y. Naiker, A. Raghunandan, T. Fischer, H. Crouse, A. Ebrahim and W. Carter (2010). Air Quality Baseline Assessment for the Highveld Priority Area, Report compiled on behalf of the Department of Environmental Affairs, Pretoria.

265

Appendix A –Overview of APPA and AQA

Transition from APPA to AQA

The National Environmental Management: Air Quality Act (AQA), Act 39 of 2004 is progressively being implemented with the intention of this act replacing the Atmospheric Pollution Prevention Act (APPA), Act 45 of 1965 in its entirety. On 11 September 2005 the Air Quality Act came into force, with the exclusion of certain sections which deal with the licensing of “listed activities”. As at June 2008 both the AQA and APPA were in force, with conflicting functions in the regulation of industrial processes being avoided through the relevant sections of the AQA dealing with „listed activities‟ not yet being in force.

The relevance of each of the Acts listed previously and regulations published under these Acts to air pollution control and/or air quality management is discussed in the following subsections.

Atmospheric Pollution Prevention Act (APPA), Act 45 of 1965

Air pollution control is administered at a national level by the DEAT according to the Atmospheric Pollution Prevention Act No. 45 of 1965 as amended. The Air Pollution Prevention Act regulates the control of noxious and offensive gases emitted by industrial processes, the control of smoke and wind borne dust pollution, and emissions from diesel vehicles. The implementation of the act is charged to the Chief Air Pollution Control Officer (CAPCO).

All industries undertaking scheduled processes are controlled by the CAPCO through Best Practicable Means (BPM) using permits. Scheduled processes, referred to in the Act, are processes which emit more than a defined quantity of pollutants per year. Such processes include large combustion sources, smelting and inherently dusty industries. BPM represents an attempt to restrict emissions while having regard to local conditions, the prevailing extent of technical knowledge, the available control options, and the cost of abatement.

Although emission limits and ambient concentration guidelines are published by the Department of Environmental Affairs and Tourism (DEAT, 1994), no provision is made in the APPA for ambient air quality standards or minimum national emission limits. The decision as to what constitutes the best practicable means for each individual case is reached following discussions with the industry. A scheduled process registration certificate, containing maximum emission limits specific to the industry, is then issued.

The CAPCO is responsible for the control of dust from industry and waste dumps. Dust control from mine dumps is the result of consultation between the Government Mining Engineer and CAPCO. The control of dust is undertaken using BPM through notice in writing. Powers for dust control have selectively been delegated to local authorities.

266

With the exception of the provisions relating to noxious and offensive gases,58 and dust59 the administration and enforcement of the air pollution prevention measures imposed by the Atmospheric Pollution Prevention Act of 1965 are entrusted to local authorities by virtue of the fact that measures relating to smoke and vehicle emissions apply only in areas in which the Minister of Environmental Affairs (by notice in the Government Gazette) declared them to be applicable. These notices also delegated the responsibility to administer and enforce the pollution prevention measures to local authorities.

The obligations imposed by APPA on local authorities are briefly listed.

Section 15: Approval of the installation of fuel burning appliances in areas where Part II has been declared to apply (installations in dwelling houses expressly being excluded);

Section 16: Criteria that must be satisfied prior to the approval of the installation of fuel burning appliances;

Section 17: Obligation on local authority to serve notice on person responsible for a nuisance situation caused by smoke or any other product of combustion;

Section 18: enables local authorities to promulgate smoke control regulations. Smoke control regulations have been promulgated in terms of this section by various municipalities. These regulations are all very similar in content and typically:

 prohibit the use of waste incinerators other than appliances that are approved in writing by the council;

 impose a smoke density limit linked to the chart shown in the first schedule to the Act (based either on visual observation or measured with a light absorption meter).

In terms of section 60 of the Air Quality Act, these regulations will remain in force until specifically repealed by the municipality of the area concerned60. It is therefore important to consider the continued relevance of these by-laws even when the APPA is repealed.

58 Part II of the Atmospheric Pollution Prevention Act of 1965 (The Control of Noxious or Offensive Gases) is administered and enforced by the National Department of Environmental Affairs and Tourism. 59 Part IV (Dust Control). Although the dust control measures also apply only in areas identified by the Minister, the administration and enforcement have not been delegated to local authorities. 60 Section 60(3) of the Air Quality Act states the following: “Anything done or deemed to have been done under a provision repealed by subsection (1) and which can be done in terms of the constitutional or statutory powers of a municipality, remain in force in the area of a municipality until repealed by the municipality of that area.”

267

Section 30: Enables the Minister of Environmental Affairs to direct local authorities to take certain measures to address dust pollution instances where special circumstances exist.

Section 36(2): Delegates the administration and enforcement of measures aimed at controlling fumes emitted by vehicles to local authorities61. Emission limits and areas of application are set out in GN R 1651 of 20 September 1974. As mentioned above, in terms of section 60 of the proposed Air Quality Act, these regulations will remain in force until specifically repealed by the municipality of the area concerned. It is therefore important to consider the continued relevance of these by-laws unless they are repealed due to their no longer being necessary or due to their being replaced.

Under the APPA, air pollution management in South Africa is based entirely on source- based controls, and is focused primarily on controlling the emissions from the industrial sector. Proposed industries are permitted to go ahead if compliance with emissions limits set by CAPCO is demonstrated. Such emission limits imposed by the CAPCO reflect the prevailing extent of technical knowledge, and the availability of control options which do not exceed “excessive costs”, rather than measures to ensure the maintenance of acceptable air quality.

A synopsis of the powers and functions allocated to national, provincial and local authorities by the APPA is given in Table 1. Those powers and functions pertaining to compliance monitoring are highlighted.

National Environmental Management: Air Quality Act, Act 39 of 2004

The Atmospheric Pollution Prevention Act of 1965 is not adequate to facilitate the implementation of the principles underpinning the National Environmental Management Act (NEMA) and the Integrated Pollution and Waste Management (IP&WM) white paper. This Act is also out of line with what is internationally considered to represent good air quality management practice. The AQA is intended to reflect the overarching principles within general environmental policy and to bring legislation in line with local and international good practices as they pertain to air quality management.

The Air Quality Act requires a shift from source-based air pollution control by national government to a receiving environment, air quality management approach. Key features of the new approach to air quality governance include:

 Decentralisation of air quality management responsibilities;

 Requirement that all significant sources be identified, quantified and addressed;

61 These provisions also apply only in areas which have been identified by notice in the Government Gazette as being subject to part V of the Act.

268

 Setting of ambient air quality targets as goals driving emission reductions;

 Recognition of source-based (command-and-control) measures in addition to alternative measures, including market incentives and disincentives, voluntary programmes, and education and awareness;

 Promotion of cost-optimised mitigation and management measures;

 Stipulation of air quality management planning by authorities and emission reduction and management planning by sources; and

 Access to information and public consultation

The new approach has significant implications for government, business and civil society.

This air quality governance approach will necessitate considerable capacity building within national, provincial and local government. National government, in the form of the DEAT is responsible for the setting of national norms and standards for emission control, air quality monitoring, air quality information management and air quality planning. DEAT is also required to develop, review and revise systems and procedures for attaining compliance with air quality standards and protocols able to give effect to South Africa‟s obligations in terms of international agreements.

Under the Air Quality Act local authorities will be responsible for monitoring air pollution and meeting nationally set ambient air quality limits. In order to manage and maintain air quality to within these limits, local authorities will be required to identify sources contributing to non-compliance and develop emission reduction programmes for such sources. Air quality management systems established for baseline characterisation and tracking progress made by emission reduction programmes will be documented in Air Quality Management Plans, which are required to be compiled and integrated into the local authorities‟ Integrated Development Plans.

The Air Quality Act designates district municipalities and metropolitan municipalities as atmospheric emissions licensing authorities. Such municipalities will be responsible for the regulation of enterprises undertaking so-called „listed activities‟, i.e. activities associated with potentially significant atmospheric emissions.

Provincial environmental departments are primarily tasked with monitoring the air quality management performance of local government, but may become responsible for the licensing of „listed activities‟ in the event that: local government is unable to fulfil the function; local government requests that the function be taken by province; or local government is undertaking the listed activity requiring licensing.

Various local authorities proactively developed air quality management plans prior to the promulgation of the Air Quality Act, including: City of Johannesburg, Ekurhuleni Metropolitan Municipality, City of Cape Town, Tshwane Metropolitan Municipality

269

and Rustenburg Local Municipality. Others are in the process of doing so, such as Ethekwini Municipality.

The „Vaal Triangle‟, which faces complex and pressing air pollution challenges, has been designated as the first „priority area‟ under the Air Quality Act. This area will consequently be the focus of concerted, mandatory baseline air pollution characterisation and air quality management efforts. The plan to be developed under the leadership of national government will coordinate air quality management efforts by all tiers of government, document emissions reduction measures and establish a committee representing relevant role players to oversee the implementation of the plan.

Listed Activities

All industries undertaking processes listed in the Second Schedule of the APPA have historically been regulated by national government through the administration of registration certificates which were based on best practicable means (BPM) as previously discussed. Major shortcomings of the regulation of industries under the APPA have included:

 Absence of a system for regularly reviewing and where necessary revising registration certificate conditions in line with best practice;

 Insufficient attention being paid to the potential for cumulative impacts given the co-location of an industrial operation with other industrial and other non-industrial sources. This has resulted in instances where unacceptable ambient air pollutant concentrations occur in instances where specific industries may be operating within the conditions in their registration certificates;

 Absence of coherent and routine monitoring, by both government and businesses, to assess compliance with registration certificate conditions; and

 Ineffective regulation of non-scheduled processes resulting in pollution and fugitive emissions including dust from vehicle entrainment, wind erosion of stockpiles and material handling and diffuse and evaporative emissions from smelting operations and chemical storage.

Under the Air Quality Act, the permitting of “scheduled processes” by DEAT will be replaced by the licensing of “listed activities” by district and metropolitan municipalities and provincial environmental departments. The „list of activities‟ to require licenses will be more comprehensive and will make provision for fugitive sources. Due to the need to manage air quality to within acceptable limits, the potential for cumulative air pollutant concentrations occurring due to multiple sources will be taken into account in the setting of emission limits for an individual operations.

In order to support the transition from the regulation of „scheduled processes‟ by national government to the control of „listed activities‟ by local and provincial authorities, DEAT has embarked on the APPA Registration Certificate Review Project.

270

The aim of this project is to identify inventory and initiate the conversion of registration certificates to atmospheric licenses, and in the process capacitating provincial and various local authorities who will become responsible for licensing. Having been initiated in January 2006, the project is scheduled to take 18 months to complete.

Controlled emitters

The Air Quality Act furthermore makes provision for the regulation of small-scale operations (e.g. dry cleaners, fuel filling stations, small boilers, crematoriums) through the potential listing of such operations as „controlled emitters‟ which are required to meet published emission standards. Given the frequent proximity of such operations to communities, the control of such operations is of considerable benefit.

Enterprises, if operating „listed activities‟ will need to apply for an atmospheric emission license and periodically apply for license renewal. The onus will be on the enterprise to prove compliance with license conditions and, depending on the size and nature of the activity being undertaken, may be required to appoint an emission control officer. The bottom line is that enterprises will need to „do their homework‟ with regard to understanding their emissions and associated impacts, and will need to consider cleaner production technologies and practices to continue operating in the long term. Benefits to business include being able to actively participate in air quality management planning processes, and gaining assurances due to the establishment of a more comprehensive regulatory environment.

The public has much to gain from effective air quality governance under the Air Quality Act. Provision is made for access to air quality information and opportunities for participation in air quality management processes, and the public is promised protection from impacts on their health and well-being due to air pollution including odour, dust and noise.

Direct polluters

Given the commitment to the fair and proportional control of all significant sources of air pollution under the Air Quality Act, provision is made for the identification of sectors of society as direct polluters. Such sectors may be subject to regulations and related penalties for offences, and/or the subject of focused education and awareness campaigns aimed at achieving emission reductions through behavioural changes. Motorists and persons undertaking fuel burning at their residences are likely to receive the most attention in this regard.

A synopsis of the powers and functions allocated to national, provincial and local authorities by the AQA is given in Table 2.

271

Table 1: Powers and functions allocated to national and local authorities by the APPA

National Local Designation of controlled areas (Minister) Promulgate smoke control regulations (by laws) – may include prohibitions on smoke emissions which are darker in colour or density than that stipulated in regulations, prohibitions on installation on premises which do not comply, prohibitions on the use of solid-fuel in dwellings which do not comply with specifications (etc.)

Review applications and issue registration certificates and Promulgate various by-laws including: prohibition of atmospheric pollution generation in the form of burning and/or provisional registration certificates for Scheduled offensive odours. Processes Monitor compliance with provisional registration Monitor and enforce compliance with local smoke control laws and other by laws certificate and registration certificate specificationsa Issue regulations for fuel burning appliances, dust control Prohibit the manufacture or import of fuel burning appliances for use in dwellings which do not comply with CAPCO regulations. Monitor compliance of fuel burning appliances with regulations. Designation of dust control areas Prohibit the installation of fuel burning appliances unless it is capable of being operated continuously without emitting smoke darker than that permitted (Monitor smoke from premises within designated smoke control areas) Enforcement of best practicable means of dust control for Evaluation of the siting of fuel burning appliances and plans for chimney construction any industrial processes in a dust control area which is likely to cause a nuisance to people, and deposits exceeding 20 000 cubic tons Monitor compliance with dust control regulations Several municipalities in larger metropolitan areas scheduled (by notice under APPA) to assess compliance of diesel-driven vehicles in terms of the opacity of their emissions a Chief Air Pollution Control Officer (CAPCO) and Inspectors may without previous notice enter any premises where a scheduled process is or is suspected to be carried on and inspect any process in which any noxious or offensive gas is used or produced and any apparatus for condensing any such gas or otherwise preventing the discharge thereof into the atmosphere.

272

Table 2: Powers and functions allocated to national, provincial and local authorities by the AQA

National Provincial Local Establish & review national framework Air quality monitoring Air quality monitoring Identify pollutants posing threat Monitor municipality performance Emission monitoring Establish national air quality standards (for pollutants posing Identify pollutants posing threat Identify pollutants posing threat threat) Establish national emission standards (for pollutants posing Establish provincial air quality standards (pollutants posing Establish local emission standards (pollutants posing threat) threats / listed activities / controlled emitters) threat) Appoint national AQ officer Establish provincial emission standards (priority pollutants / Appoint AQ officer listed activities / controlled emitters) Draft of AQM plan (within EMP) Appoint provincial AQ officer AQM plan (IDP) Declare priority areas AQM plan (EMP) Report on implementation of AQMP Prepare priority area AQMPs Report on implementation of AQMP Collaborate with national & local (priority areas) Prescribe regulations for implementing & enforcing Priority Area Declare priority areas Request & review Atmospheric impact reports (air quality officer) AQMPs (funding arrangements, measures to ensure compliance, penalties, etc.) Report on implementation of AQMP Prepare priority areas AQMP Establish recognition programmes (air quality officer) List activities Prescribe regulations for implementing & enforcing Priority Area Perform emission licensing authority functions (metros, DMs) AQMPs (funding arrangements, measures to ensure compliance, penalties, etc.) Declare controlled emitters List activities Declare controlled fuels Declare controlled emitters Establish standards for controlled fuels Declare controlled fuels List priority pollutants (requiring PPPs) Establish standards for controlled fuels Specify categories requiring to submit PPPs List priority pollutants (requiring PPPs) Set requirements for Pollution Prevention Plans (PPP) Specify categories requiring to submit PPPs Request & review Atmospheric impact reports (air quality Set requirements for PPP officer) Establish recognition programmes (air quality officer) Request & review Atmospheric impact reports (air quality officer) Set regulations for dust, odour, noise Establish recognition programmes (air quality officer) Review rehabilitation plans for closing mines Perform emission licensing authority functions (if Metro or DM delegates its functions or cannot fulfil its functions, or municipality applying for license) Investigate & regulate transboundary pollution

273

National Provincial Local Investigate potential international agreement contraventions Regulations on: EMCAs, open fires, incineration, ODSs, COPs, labeling, trading schemes, powers & duties of air quality officers, appeal process, monitoring, emission reduction (etc.) Review applications for exemption from section of Act Collection and management of data necessary to assess: Collection and management of data necessary to assess: Collection and management of data necessary to assess: - Compliance with the Air Quality Act - Compliance with the Air Quality Act - Compliance with the Air Quality Act - Compliance with ambient air quality and emission standards - Compliance with ambient air quality and emission standards - Compliance with ambient air quality and emission standards - Performance of organs of state in respect of air quality - Performance of organs of state in respect of air quality - Impact of AQMPs management plans and priority area air quality management plans management plans and priority area air quality management plans - Access to information by the public - Impact of, and compliance with, AQMPs and priority area - Impact of, and compliance with, AQMPs and priority area AQMPs AQMPs - Compliance with SA‟s obligations in terms of international - Access to information by the public agreements - Access to information by the public

274

Appendix B –Vehicle Emission Factors and Emission Estimates

EMISSION FACTORS FOR PETROL-DRIVEL VEHICLES (NON-CATALYTIC CONVERTER EQUIPPED): POLLUTANT UNITS COASTAL HIGHVLED LEADED UNLEADED LEADED UNLEADED PETROL PETROL PETROL PETROL NOX g/km 3.01 2.92 1.99 2.15 SO2 g/km 0.09 0.09 0.05 0.04 1,3 Butadiene g/km 0.02 0.02 0.02 0.03 Benzene g/km 0.04 0.03 0.03 0.02 Source: Wong (1999), Coppert (2000) for lead and N2O

EMISSION FACTORS FOR PETROL-DRIVEL VEHICLES (CATALYTIC CONVERTER EQUIPPED): POLLUTANT UNITS COASTAL HIGHVLED LEADED UNLEADED LEADED UNLEADED PETROL PETROL PETROL PETROL NOX g/km 0.86 0.77 0.86 0.93 SO2 g/km 0.02 0.02 0.01 0.02 1,3 Butadiene g/km 0.00 0.00 0.00 0.00 Benzene g/km 0.01 0.01 0.02 0.02 Source: Wong (1999), Coppert (2000) for lead and N2O

EMISSION FACTORS FOR DIESEL-DRIVEL VEHICLES: Sources: Wong (1999) Source: Stone (2000) POLLUTANT UNITS COASTAL HIGHVELD(a) COASTAL HIGHVELD DIESEL - DIESEL - DIESEL - DIESEL - LCVs LCVs M&H M&H NOX g/km 1.820 1.820 11.680 11.680 SO2 g/km 0.796 0.796 1.540 1.540 1,3 Butadiene g/km 0.003 0.003 0.007 0.004 Benzene g/km 0.002 0.002 0.008 0.000 Particulates g/km 0.293 0.293 0.640 0.640 FUEL CONSUMPTION (l/km) 0.105 0.105 0.239 0.244 (a) Emission factors given by Wong (1999) for diesel-driven LCVs within coastal areas assumed to be representative of highveld areas. (b)

ESTIMATED VEHICLE EMISSIONS FROM NON-CATALYTIC CONVERTER EQUIPPED VEHICLES: POLLUTANT Johannes- Cape Vaal Ethekwini Tshwane Ekurhuleni Mpumalaga burg Town Triangle NOX tpa 29 247 32 657 4 698 28 786 17 065 18 540 5 832 SO2 tpa 2 144 3 245 693 4 196 1 549 2240 2 005 1,3 Butadiene tpa 358 244 57 216 209 227 71 Benzene tpa 372 423 61 374 218 239 76

ESTIMATED VEHICLE EMISSIONS FROM CATALYTIC CONVERTER EQUIPPED VEHICLES: POLLUTANT Johannes- Cape Vaal Ethekwini Tshwane Ekurhuleni Mpumalaga burg Town Triangle NOX tpa 1 001.9 604.7 156.4 592.2 548.0 635.1 188.7 SO2 tpa 1 659.8 2 453.3 610.5 3 491.8 1 263.8 1 920.3 1 902.4 1,3 Butadiene tpa 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Benzene tpa 20.2 6.1 3.1 6.0 11.1 12.7 3.8

275

ESTIMATED VEHICLE EMISSIONS FROM DIESEL CONSUMPTION: POLLUTA Johannes- Cape Vaal Ethekwini Tshwane Ekurhuleni Mpumalaga NT burg Town Triangle NOX tpa 15 691 23 200 5 785 33 078 11 960 18 184 18 044 SO2 tpa 1 629 2 408 601 3 433 1 242 1 888 1 874 1,3 tpa 8 17 3 24 6 10 10 Butadiene Benzene tpa 2 17 1 24 2 3 3 Particulates tpa 1 085 1 603 400 2286 827 1 257 1 247

276

Appendix C – Fuel-burning Emissions per Region and Source

Ethekwini - estimated total annual emissions from fuel burning activities (tpa) Industrial, Commercial Domestic Electricity Biomass Pollutant & Vehicles Shipping Aircraft Fuel TOTAL Generation Burning Institutional Burning Fuel Burning PM10 3 311 NA 2 286 (c) (c) 151(a) 1 144 6 891 SO2 13 059 NA 11 121 850 6 18 882 25 937 NOX 5 087 NA 62 457 1 815 160 94 430 70 042 Benzene 1 NA 404 (c) (c) (b) 59 464 1,3 Butadiene (b) NA 239 (c) (c) 5 8 252 Abbreviations - NA - not applicable Notes: (a) Given as PM2.5 (particulate matter <2.5 micron). (b) No emission factor available for source quantification for this pollutant. (c) Insufficient source and/or emission estimates available for source quantification.

Cape Town - estimated total annual emissions from fuel burning activities (tpa) Pollutant Industrial, Electricity Vehicles Shipping Aircraft Biomass Domestic TOTAL Commercial Generation Burning Fuel & Burning Institutional Fuel Burning PM10 2 767 113 1 603 (c) (c) 402(a) 1 752 6 637 SO2 24 896 2 023 8 106 1 214 46 48 548 36 882 NOX 4 495 1 678 56 461 1 321 576 249 633 65 414 Benzene 4 0 447 (c) (c) (b) 98 548 1,3 Butadiene (b) - 261 (c) (c) 13 13 287 Abbreviations - NA - not applicable Notes: (a) Given as PM2.5 (particulate matter <2.5 micron). (b) No emission factor available for source quantification for this pollutant. (c) Insufficient source and/or emission estimates available for source quantification.

Vaal Triangle - total annual emissions from fuel burning activities (tpa) Pollutant Industrial, Electricity Vehicles Shipping Aircraft Biomass Domestic TOTAL Commercial Generation (d) Burning Fuel & Burning Institutional Fuel Burning PM10 8 150 400 433 NA NA 392(a) 966 15 682 SO2 219 868 1 904 1 300 NA NA 47 2 917 455 924 NOX 98 457 10 639 10 845 NA NA 243 363 231 138 Benzene 6 65 66 NA NA (c) 23 94 1,3 Butadiene (b) 60 65 NA NA 13 3 76 Abbreviations - NA - not applicable Notes: (a) Given as PM2.5 (particulate matter <2.5 micron). (b) No emission factor available for source quantification for this pollutant. (c) Insufficient source and/or emission estimates available for source quantification. (d) Aircraft emissions only quantified for large, international airports.

277

Mpumalanga Highveld - total annual emissions from fuel burning activities (tpa) Pollutant Industrial, Electricity Vehicles Shipping Aircraft Biomass Domestic TOTAL Commercial Generation (d) Burning Fuel & Burning Institutional Fuel Burning PM10 (c) 47 475 1 247 NA NA 2 740(a) 1 191 52 653 SO2 275 665 1 280 816 5 781 NA NA 329 3 598 1 566 189 NOX 151 021 573 548 24 065 NA NA 1 699 448 750 781 Benzene (c) 36 82 NA NA (b) 28 146 1,3 Butadiene (b) (c) (b) 81 NA NA 89 3 174 Abbreviations - NA - not applicable Notes: (a) Given as PM2.5 (particulate matter <2.5 micron). (b) No emission factor available for source quantification for this pollutant. (c) Insufficient source and/or emission estimates available for source quantification. (d) Aircraft emissions only quantified for large, international airports.

Johannesburg - total annual emissions from fuel burning activities (tpa) Pollutant Industrial, Electricity Vehicles Shipping Aircraft Biomass Domestic TOTAL Commercial Generation Burning Fuel & Burning Institutional Fuel Burning PM10 (c) 77 1 085 NA (c) 785(a) 1 054 3 001 SO2 3 143 1 382 5 433 NA 167 94 4 105 14 325 Benzene (c) 0 395 NA (c) (b) 15 409 1,3 Butadiene (b) (c) 367 NA (c) 26 2 394 Abbreviations - NA - not applicable Notes: (a) Given as PM2.5 (particulate matter <2.5 micron). (b) No emission factor available for source quantification for this pollutant. (c) Insufficient source and/or emission estimates available for source quantification.

Tshwane - total annual emissions from fuel burning activities (tpa) Pollutant Industrial, Electricity Vehicles Shipping Aircraft Biomass Domestic TOTAL Commercial Generation Burning Fuel Institutional Burning Fuel Burning PM10 (c) 4 236 827 NA NA 691(a) 698 6 452 SO2 9 462 15 198 4 054 NA NA 83 2 719 31 517 NOX 650 12 605 29 573 NA NA 428 274 43 531 Benzene (c) 0 230 NA NA (b) 10 241 1,3 (b) (c) (b) 215 NA NA 23 1.0 239 Butadiene Abbreviations - NA - not applicable (a) Given as PM2.5 (particulate matter <2.5 micron). (b) No emission factor available for source quantification for this pollutant. (c) Insufficient source and/or emission estimates available for source quantification.

278

Ekurhuleni - total annual emissions from fuel burning activities (tpa) Pollutant Industrial, Electricity Vehicles Shipping Aircraft Biomass Domestic TOTAL Commercial Generation Burning Fuel Institutional Burning Fuel Burning PM10 (c) NA 1 257 NA NA 559(a) 865 2 681 SO2 14 447 NA 6 049 NA NA 67 2 581 23 144 NOX 4 653 NA 37 359 NA NA 347 358 42 718 Benzene (c) NA 254 NA NA (b) 23 277 1,3 (c) NA 236 NA NA 18 3 257 Butadiene Abbreviations - NA - not applicable (a) Given as PM2.5 (particulate matter <2.5 micron). (b) No emission factor available for source quantification for this pollutant. (c) Insufficient source and/or emission estimates available for source quantification.

Total across all conurbations - total annual emissions from fuel burning activities (tpa) Pollutant Industrial, Electricity Vehicles Shipping Aircraft Biomass Domestic TOTAL Commercial & Generation Burning Fuel Institutional Burning Fuel Burning PM10 12 920 63 887 8 704 5 720(a) 7 670 98 901 SO2 571 860 1 519 288 42 448 2 064 219 686 17 351 2 153 917 NOX 288 238 687 434 266 495 3 136 1 459 3 547 2 919 1 253 229 Benzene 4 42 1 877 255 2 179 1,3 Butadiene 1 460 186 33 1 679 (a) Given as PM2.5 (particulate matter <2.5 micron).

279

Appendix D – Predicted Air Pollutant Concentrations due to Fuel- burning Emissions

280

281

282

283

284

285

286

287

288

289

290

291

292

Appendix E – Direct Health Cost Estimates

Total direct health costs due to respiratory illnesses, non-fatal paraffin poisonings, burns and cancer cases caused by fuel use – given per conurbation and source group:

Total Cancer Total Cost of Respiratory Total Direct Health Region Source Group Costs (2002 Condition (2002 Rand) Costs (2002 Rands) Rands) Coal 15,507,904 316 15,508,220 burning Domestic fuel Wood 691,237,956 428,237 691,666,193 burning burning Other 2,311,778 2,311,778 fuel Petrol 12,908,609 2,611,232 15,519,841 Vehicles Diesel 21,965,326 1,681,616 23,646,942 Cape Town Industry & commercial 110,906,673 9,524 110,916,197

Power generation 3,776,940 21 3,776,961

TOTAL 858,615,184 4,730,946 863,346,130

Coal 82,599,828 1,144 82,600,972 burning Domestic fuel Wood 511,659,364 165,950 511,825,314 burning burning Other 4,201,302 4,201,302 fuel Petrol 45,694,664 4,568,207 50,262,871 Ethekwini Vehicles Diesel 93,294,257 3,099,012 96,393,269

Industry & commercial 50,988,501 200 50,988,701

TOTAL 788,437,916 7,834,514 796,272,430

Coal 546,248,498 267,422 546,515,920 burning Domestic fuel Wood 211,192,526 360,501 211,553,027 burning burning Other 880,404 880,404 fuel Joburg & Petrol 51,119,176 6,651,757 57,770,933 Ekurhuleni Vehicles Diesel 67,421,045 4,679,925 72,100,970 Industry & commercial 107,043,740 71 107,043,811 Power generation 6,914,095 31 6,914,126

TOTAL 990,819,484 11,959,708 1,002,779,192

293

Total Cancer Total Cost of Respiratory Total Direct Health Region Source Group Costs (2002 Condition (2002 Rand) Costs (2002 Rands) Rands) Coal 60,131,964 21,091 60,153,055 burning Domestic fuel Wood 13,337,867 20,934 13,358,801 burning burning Other 37,538 37,538 fuel Petrol 53,071,692 7,446,018 60,517,710 Vehicles Diesel 68,107,894 5,279,452 73,387,346 Tshwane Industry & commercial 71,306,741 4 71,306,745

Power generation 31,221,042 385 31,221,427

TOTAL 297,214,739 12,767,883 309,982,622

Coal 155,145,114 586,810 155,731,924 burning Domestic fuel Wood 54,943,609 860,328 55,803,937 burning burning Other 126,401 126,401 fuel Petrol 621,148 100,394 721,542 Vehicles Diesel 1,205,277 64,464 1,269,741 Vaal Triangle 36,897,372 Industry & commercial 36,886,855 10,517

Power generation 25,999,166 935 26,000,101

TOTAL 274,927,571 1,623,448 276,551,019

Coal 29,605,055 85,856 29,690,911 burning Domestic fuel Wood 16,690,790 916,244 17,607,034 burning burning Other 29,714 29,714 fuel Mpumalanga Petrol 286,765 70,267 357,032 Highveld Vehicles Diesel 3,867,851 48,361 3,916,212 Industry & commercial 70,072,799 16,324 70,089,123 Power generation 132,395,961 7,212 132,403,173

TOTAL 252,948,934 1,144,264 254,093,198

294

Appendix F – Guideline Inspection Protocol for Listed Activities holding Atmospheric Emission Licenses under the AQA

295

Department of Environmental Affairs and Tourism

Compliance Monitoring Project

Listed Activities holding Atmospheric Emission Licenses under the National Environmental Management: Air Quality Act

Guideline Inspection Protocol

February 2008

Inspection Protocol Listed Activities holding NEM:AQA Atmospheric Emission Licenses

CONTENTS

Chapter Description Page

1 INTRODUCTION TO LISTED ACTIVITY COMPLIANCE 3

2 PURPOSE OF DOCUMENT 3

2.1 Applicability of Protocol 3

3 PROTOCOL BACKGROUND INFORMATION 4

3.1 List of Acronyms and Abbreviations 4

3.2 Review of Air Quality Legislation & Regulations 5

3.3 Key Terms and Definitions 8

4 INSPECTION PROTOCOL 9

4.1 Applicability 9

4.2 Legal Mandate for undertaking inspections 9

4.3 Key Compliance Requirements 10

4.4 Typical Records to Review 11

4.5 Typical Physical Features to Inspect 13

5 TYPICAL INSPECTION PROCESS 14

6 CHECKLISTS 15

7 COMPLIANCE RATING AND INTERPRETING RESULTS 15

7.1 Overall Compliance Level (Legal Compliance with Conditions of Licence) 17

7.2 Key Non-Compliance Areas and Potential for Environmental Risk 17

7.3 Environmental Risks due to Non-Compliance 18

7.4 Results Summary 19

8 ADDITIONAL SITE INFORMATION 21

9 FIELD SHEETS 21

10 CONCLUSION 21

AQA Listed Activity Compliance Protocol (Draft) Issue 1.0 / July 07

11 REFERENCES 22

APPENDICES

Appendix A Checklists Appendix B Scoring Sheet Appendix C Field Sheets Appendix D Additional Information to be Collected at Facilities

AQA Listed Activity Compliance Protocol (Draft) Issue 1.0 / July 07

1 INTRODUCTION TO LISTED ACTIVITY COMPLIANCE

The Department of Environmental Affairs and Tourism (DEAT) is responsible for ensuring that government and private organisations comply with National Legislation that promote a healthy environment for all citizens and which is free from pollution.

Compliance monitoring represents a vital component of environmental regulation. Such monitoring comprises the collection of information on compliance ‘status’ through various means including inspection by environmental authorities or third parties, ambient or emissions monitoring and self-monitoring by industrial operators.

Activities listed in terms of Section 21 of the National Environmental Management: Air Quality Act, Act 39 of 2004 require authorisations, termed Atmospheric Emission Licences, to operate. The monitoring of compliance of Listed Activities with the terms and conditions laid out in their Atmospheric Emission Licence represents an important component of the overall Compliance Monitoring functions of environmental authorities.

Compliance monitoring of industrial facilities by environmental authorities comprises both proactive and reactive monitoring functions. Reactive monitoring may be prompted by complaints or pollution incidents. Proactive monitoring refers to planned, routine monitoring of the compliance status of facilities with the frequency of such inspections determined on a risk- based approach. A protocol for conducting routine inspections of Listed Activities to determine compliance with Atmospheric Emission Licences forms the focus of this document.

2 PURPOSE OF DOCUMENT

This document has been developed as part of the Compliance Monitoring Handbook and provides a guidance protocol for compliance inspections at facilities classified as Listed Activities in terms of the National Environmental Management: Air Quality Act (AQA), Act 39 of 2004.

This protocol, which is to be part of a set containing other facility protocols, is a tool to assist in conducting compliance audits, which should ultimately inform the user whether the facility is compliant with environmental regulations. The format for this protocol follows the United States Environmental Protection Agency Protocols for Conducting Environmental Compliance Audits.

Each protocol is intended to provide guidance on key requirements, defines regulatory terminology and gives an overview of the regulatory framework affecting environmental management of the facility. The protocol is not intended to be exclusive and limiting with respect to procedures that may be followed. Other approaches and techniques can also be exploited in order to evaluate compliance of a facility.

The objectives of this compliance protocol are to provide national consistency in compliance inspections for Listed Activities, and promote communication between national, provincial and local environmental authorities on air compliance monitoring programmes for Listed Activities.

2.1 Applicability of Protocol

At the time of drafting this protocol, the sections dealing with the listing and licencing of activities under the Air Quality Act of 2004 had not yet been brought into force. As such, the regulation of air pollution from industrial facilities is still administered at a national level by the DEAT according to the Atmospheric Pollution Prevention Act (APPA), Act 45 of 1965 (as amended). Under the APPA, all industries undertaking Scheduled Processes are controlled by the Chief Air Pollution Control Officer, within the DEAT, through ‘best practicable means’ using permits which are termed Registration Certificates. Scheduled processes, referred to in

AQA Listed Activity Compliance Protocol (Draft) 3 Issue 1.0/ July 2007

the Act, are processes which emit more than a defined quantity of pollutants per year. The APPA Registration Certificate Review Project, initiated by the DEAT in January 2006, aims to convert the Registration Certificates held by a number of large industrial facilities into facility- specific Atmospheric Emission Licences by 2008 so as to aid the transition between APPA and AQA.

This protocol has been drawn up specifically to assess compliance of Listed Activities holding Atmospheric Emission Licences under Chapter 5 of the Air Quality Act and other legal conditions related to atmospheric emissions from such activities. It assumes implementation by environmental authorities tasked with the monitoring of compliance of Listed Activities following the relevant sections of the AQA coming into force.

Compliance monitoring protocols must reflect the content of the authorisation for which the compliance status is being determined. At the time of drafting of this protocol, the template for Atmospheric Emission Licences had been drafted but not yet finalised. This protocol was therefore drafted on the basis of the Draft Atmospheric Emission Licence Template and will need to be revised in the event that significant changes to the draft template are made prior to finalisation.

It is recognised that the Protocol will not reflect the content of the Registration Certificates issued in terms of the APPA which, under circumstances made provision for in section 61 of the AQA, may remain valid during a transitional period following the repeal of APPA. The Protocol does however offer an example of how compliance with the generally more restricted conditions of Registration Certificates may be assessed.

The AQA makes provision for the regulation of small-scale operations (e.g. dry cleaners, fuel filling stations, small boilers, crematoriums) through the potential listing of such operations as ‘controlled emitters’ which are required to meet published emission standards. The Protocol outlined is not intended for monitoring the compliance of ‘controlled emitters’.

3 PROTOCOL BACKGROUND INFORMATION

3.1 List of Acronyms and Abbreviations

APPA Atmospheric Pollution Prevention Act, Act 45 of 1965, as amended

AQA National Environmental Management: Air Quality Act, Act 39 of 2004

DEAT Department of Environmental Affairs and Tourism

ECA Environment Conservation Act, 1989 (Act 73 of 1989)

EPA Environmental Protection Agency

EIA Environmental Impact Assessment

EIAR Environmental Impact Assessment Regulations

EMP Environmental Management Plan

NEMA National Environmental Management Act (Act 107 of 1998)

RoD Record of Decision

SANS South African National Standards

AQA Listed Activity Compliance Protocol (Draft) 4 Issue 1.0/ July 2007

3.2 Review of Air Quality Legislation & Regulations

The National Environmental Management: Air Quality Act (AQA), Act 39 of 2004 represents the main media-specific legislation for the regulation of air quality under the National Environmental Management Act (NEMA), Act 107 of 1998. The AQA aims generally to give effect to section 24(b) of the Constitution in order to enhance the quality of ambient air for the sake of securing an environment that is not harmful to the health and well-being of people. In protecting the environment, the AQA advocates the use of reasonable measures for protecting and enhancing air quality nationally and preventing air pollution and ecological degradation.

A brief overview is given below of components of the AQA which have relevance in terms of the regulation of air pollution from industrial facilities.

Listing of Activities Resulting in Atmospheric Emissions

AQA requires the Minister of Environmental Affairs to publish a list of activities which result in atmospheric emissions and which the Minister reasonably believes have or may have a significant detrimental effect on the environment, including health, social conditions, economic conditions, ecological conditions or cultural heritage. The Act also makes provision for Provincial MECs to publish lists of activities. A list published by the Minister applies nationally, whereas lists published by MECs apply to the relevant provinces only. Minimum emission standards in respect of a substance or mixture of substances resulting from the listed activity must be established including permissible amounts, volumes, emission rates or concentrations and the manner in which emission measurements must be carried out.

Consequences of Listing

No persons may without a provisional atmospheric emission licence or an atmospheric emission licence conduct an activity which is either (a) listed on the national list anywhere in the country, or (b) listed on the list applicable in a province.

Licensing Authority

AQA designates district and metropolitan municipalities as atmospheric emissions licensing authorities. Provincial authorities are primarily tasked with monitoring the air quality management performance of local government, but may become responsible for the licensing of ‘listed activities’ within their respective provinces in the following circumstances:

• A metropolitan or district municipality delegates its functions of licensing authority to a provincial organ of state in terms of section 238 of the Constitution;

• The MEC intervenes, in terms of section 139 of the Constitution, on the grounds that the metropolitan or district municipality cannot or does not fulfil its obligations as licensing authority; or

• A municipality applies for an atmospheric emission licence. In this event, a provincial authority, as designated by the MEC, represents the licensing authority for the purpose of that application and implementation of the AQA in relation to any licence issued to the municipality.

Listed Activity Licensing and Compliance Monitoring Process

The AQA outlines the atmospheric emission licence application process, procedures for licence applications, factors to be taken into account by licensing authorities, possible decisions of licensing authorities and contents of licences and the manner in which licences are issued. Provision is also made in the AQA for the transfer, review, variation and renewal of atmospheric emission licences. The listed activity licensing and compliance monitoring process is demonstrated for a new facility in the flow diagram overleaf. If an application for an atmospheric emission licence is granted, the licensing authority must first issue a provisional

AQA Listed Activity Compliance Protocol (Draft) 5 Issue 1.0/ July 2007

atmospheric emission licence to enable the commissioning of the listed activity. The holder of the provisional atmospheric emission licence is only entitled to an atmospheric emission licence when the commissioned facility has been in full compliance with the conditions and requirements of the provisional atmospheric emission licence for at least six months. Following the atmospheric emission licence being granted, compliance monitoring will be ongoing for the duration that the licence is held.

LICENCE LICENCING COMPLIANCE APPLICANT AUTHORITY (Relevant

(Relevant Inspection municipal or Authority) DETERMINE IF provincial ACTIVITY IS LISTED environmental authority)

LICENCE APPLICATION

PROVISIONAL ATMOSPHERIC EMISSION LICENCE (AUTHORISATION) FACILITY CONSTRUCTION

FACILITY COMMISSIONING

FACILITY OPERATION PROVISIONAL & MONITORING ATMOSPHERIC EMISSION LICENCE COMPLIANCE ATMOSPHERIC INSPECTION & EMISSION LICENCE MONITORING (AUTHORISATION)

ATMOSPHERIC FACILITY EMISSION LICENCE OPERATION COMPLIANCE INSPECTION & MONITORING

FACILITY MONITORING ENFORCEMENT Authority

AQA Listed Activity Compliance Protocol (Draft) 6 Issue 1.0/ July 2007

Although the Licensing Authority and Compliance Monitoring (Inspection) Authority functions are distinctive and are illustrated as such within the previous flow diagram, it should be noted that such functions may be undertaken by the same division.

Emission Control Officers

An air quality officer may require the holder of a provisional atmospheric emission licence or an atmospheric emission licence to designate an emission control officer depending on the size and nature of the listed activity. An emission control officer must have the necessary air quality management competence in respect of the listed activity and must fulfil the following functions:

• Work towards the development and introduction of cleaner production technologies and practices;

• Take all reasonable steps to ensure compliance with the licence conditions and requirements; and

• Promptly report any non-compliance with any licence conditions or requirements to the licensing authority through the most effective means reasonably available.

Controlled Fuels

According to section 26 of the AQA, the Minister or MECs may declare a substance or mixture of substances as a ‘controlled fuel’ on the grounds that, when used as a fuel in a combustion process, it results in atmospheric emissions which present a threat to health or the environment or is reasonably believed to pose such as threat. Once declared, the use, manufacture or sale of the ‘controlled fuels’ could be regulated or prohibited.

Pollution Prevention Plans

The Minister or MEC may, based on section 29 of the AQA, declare any substance contributing to air pollution as a priority air pollutant and require that persons falling within a specified category prepare, submit for approval and subsequently implement pollution prevention plans in respect of that priority air pollutant.

Atmospheric Impact Reports

An air quality officer may require any person to submit an atmospheric impact report in a prescribed form if the officer reasonably suspects that the person has on one or more occasions contravened or failed to comply with the Air Quality Act or any conditions of a licence and that such a contravention has, or may have, detrimentally affected the environment or contributed to degradation of ambient air quality.

Air quality officers may also require an atmospheric impact report to be undertaken to inform the review or a provisional atmospheric emission licence or an atmospheric emission licence.

Measures in Respect of Dust, Noise and Offensive Odour

Standards and control measures for dust and offensive odours, which may be prescribed by the Minister of MEC according to Part 6 of the AQA, may be applicable to Listed Activities. Provision is also made for the Minister to prescribe national standards for the control of noise which will also be applicable to Listed Activities.

Rehabilitation Plans for Mines Ceasing Operation

The owner of a mine likely to cease mining operations within a period of five years must notify the Minister of the likely cessation and of any plans in place or contemplated for rehabilitating the area and preventing dust emissions after operations have stopped.

AQA Listed Activity Compliance Protocol (Draft) 7 Issue 1.0/ July 2007

Offences

The following offences have relevance in terms of the regulation of air pollution from industrial facilities. A person is guilty if an offence if they:

• Conduct a listed activity without a provisional atmospheric emission licence or atmospheric emission licence (section 22).

• Fail to take reasonable steps to prevent the emission of any offensive odour caused by any activity for which measures for the control of offensive odours are prescribed in terms of section 35.

• Fail to submit an atmospheric impact report as required (section 30).

• Fail to notify the Minister of the likely cessation of a mining operation within a period of five years and promptly notify the Minister of plans to rehabilitate and prevent pollution (section 33).

• Contravene or fail to comply with a condition or requirement of an atmospheric emission licence.

• Supply false or misleading information in any application for an atmospheric emission licence or for the transfer, variation or renewal of such a licence.

• Supply false or misleading information to an air quality officer.

• Are performing a listed activity and air pollutants are emitted due to that activity at concentrations above the emission limits specified in an atmospheric emission licence.

3.3 Key Terms and Definitions

The following key terms and definitions originate from the National Environmental Management: Air Quality Act (Act 39 of 2004).

TERM DEFINITION Any change in the composition of the air caused by smoke, soot, dust Air pollution (including fly ash), cinders, solid particles of any kind, gases, fumes, aerosols and odorous substances An officer appointed in terms of section 14 as an air quality officer, including the national air quality officer designated by the Minister, the provincial air Air quality officer quality officers designated by the respective MECs, and air quality officers designated by each municipality.

Excludes air regulated by the Occupational Health and Safety Act, 1993 (Act Ambient air 85 of 1993)

Atmospheric Any emission or entrainment process emanating from a point, non-point or emission mobile source that results in air pollution

Department Means the Department of Environmental Affairs and Tourism

AQA Listed Activity Compliance Protocol (Draft) 8 Issue 1.0/ July 2007

TERM DEFINITION

Gaseous constituents of the atmosphere, both natural and anthropogenic, Greenhouse gas that absorb and re-emit infrared radiation including carbon dioxide, methane and nitrous oxide

Licensing An authority referred to in section 36(1), (2), (3) or (4) of the AQA responsible authority for implementing the licensing system set out in Chapter 5 of the AQA

Any activity listed in terms of section 21 of the AQA, i.e. any activity listed by the Minister or MEC as resulting in atmospheric emissions which the Minister/MEC reasonably believes have or may have a significant detrimental Listed activity effect on the environment, including health, social conditions, economic conditions, ecological conditions or cultural heritage.

Minister Refers to the Minister of Environmental Affairs and Tourism

The member of the Executive Council of a province who is responsible for air MEC quality management in the province.

Source of atmospheric emissions which cannot be identified as having Non-point source emanated from a single identifiable sources or fixed location, and includes mining activities, agricultural activities and stockpiles.

Single identifiable source and fixed location of atmospheric emission, Point source including smoke stacks and residential chimneys.

Means an area declared as such in terms of section 18 of the AQA, referring to an area declared by either the Minister or a MEC as requiring specific air Priority area quality management activities to reduce or maintain air pollution within acceptable levels.

4 INSPECTION PROTOCOL

4.1 Applicability

This protocol applies to facilities undertaking Listed Activities in terms of Section 21 of the AQA. It includes provisions for all activities occurring within the facility boundary which are covered by the provisional atmospheric emission licence or atmospheric emission licence for that facility. It excludes all related activities which occur beyond the site boundary unless specific provision for such activities is made in the relevant licence.

4.2 Legal Mandate for undertaking inspections

Chapter 7 of NEMA (as amended, Amendment Act No. 46 of 2003, published 13 February 2004) makes provision for the designation of Environmental Management Inspectors (EMIs) by the Minister of Environmental Affairs and Tourism or any other organ of state, or by an MEC within the provincial environmental department, or any other provincial organ of state or any municipality within the province. The function of EMIs is to monitor and enforce compliance with a law for which he or she has been designated, including the NEMA and all subordinate regulations and environmental authorisations. The powers of EMIs in terms of

AQA Listed Activity Compliance Protocol (Draft) 9 Issue 1.0/ July 2007

routine inspections of buildings, land or premises for the purposes of ascertaining compliance with the designated legislation or conditions of an authorisation or other instrument issued in terms of such legislation, are specifically provided for in Section 31K of NEMA.

More detailed information is provided in the Information Document, Compliance Inspections by Environmental Management Inspectors of Facilities with Environmental Authorisations issue by the Environmental Management Inspectorate of the DEAT in March 2007.

4.3 Key Compliance Requirements

Key compliance requirements are those items that are directly enforceable through legislation and set the criteria for compliance monitoring which may be of relevance at particular facilities undertaking listed activities. These compliance requirements are described below:

4.3.1 Atmospheric Emission Licences / Provisional Atmospheric Emission Licences

Minimum Emission Standards, including permissible emissions and emission monitoring procedures, are to be set for Listed Activities by the Minister of Environmental Affairs and Tourism for application nationally, and may be set by MECs for implementation within relevant provinces. In terms of new or proposed facilities, such emission standards are enforceable through the Atmospheric Emission Licence Application procedure.

In the case of an authorised facility, the conditions appearing in the Provisional Atmospheric Emission Licence or the Atmospheric Emission Licence represent enforceable requirements for that specific facility. The facility-specific requirements in such licences form the basis for which compliance monitoring can take place.

4.3.2 Record of Decisions / Environmental Authorisations

The Environmental Impact Assessment Regulations, 2006, published in terms of the National Environmental Management Act (Act 107 of 1998) lists various activities which require environmental authorisation, prior to construction. Certain activities require only basic environmental assessments whereas others require full scale Environmental Impact Assessments.

The Environmental Authority, in their approval of the Environmental Impact Assessment will issue a Record of Decision that may stipulate certain requirements pertaining to the facility. Including which requirements listed in the Environmental Impact Assessment (EIA) Report are to be implemented thus rendering parts of the EIA report enforceable.

4.3.3 Environmental Management Plans

Environmental Management Plans developed for the construction, operation and decommissioning phases of the project may be enforceable through the requirements of the Record of Decision / Environmental Authorisation as is described above.

4.3.4 Rehabilitation Plans for Mines Ceasing Operations

AQA Listed Activity Compliance Protocol (Draft) 10 Issue 1.0/ July 2007

4.3.5 Other Compliance Requirements

There may be a number of ancillary plans (i.e. Closure Plans and Monitoring Plans) which may be approved by the authorities on an ad-hoc basis which facilities would have to comply.

Other possible compliance requirements relate to the following:

• Designation of emission control officers with necessary air quality management competence (where required)

• The use, manufacture or sale of ‘controlled fuels’.

• Preparation, submission for approval and subsequent implementation of pollution prevention plans in respect of priority air pollutants (where required)

• Preparation and submission of an atmospheric impact report in the manner prescribed (where required).

4.3.6 Ambient Air Quality Limits

In June 2006 the Department of Environmental Affairs and Tourism gazetted air quality standards for public comment (Government Gazette No. 28899, 9 June 2006). These standards have however not yet been finalised.

The following South African National Standards, although not legally binding, have also been defined in respect of ambient air quality:

SANS 69:2004 Framework for setting and implementing national ambient air quality standards SANS 1929:2004 Ambient air quality – Limits for common pollutants

4.4 Typical Records to Review

The following list of documents or records should be available for review during a Listed Activity compliance inspection.

Authorisations: • Atmospheric Emission Licence (AQA, Chapter 5) • Environmental Authorisations (ECA, Section 21,22, NEMA Section 24) • Other: This may include instructions from local, provincial or national departments received on an ad-hoc basis, e.g. monitoring, mitigation, management or reporting requirements. Operational and Management Plans: • Environmental Management Plans (EMPs) – including possible construction, operational, decommissioning and closure plans. • Air Quality Management Plan (AQMP) – Listed Activities are not legally required to compile and implement air quality management plans but certain have done so on a voluntary basis. • Pollution Prevention Plans (PPP) – Certain Listed Activities may have been required to undertake pollution prevention plans (under section 29 of AQA).

AQA Listed Activity Compliance Protocol (Draft) 11 Issue 1.0/ July 2007

• Rehabilitation Plans: In the case of Listed Activities comprising mining operations which are nearing closure, Rehabilitation Plans may have been compiled in line with AQA requirements. • Emergency Response Plan (ERP): These plans typically contain a number of procedures for response to emergency incidences such as upset emissions due to abatement equipment failure. • Occupational Health and Safety Plan: These plans may contain useful information regarding site procedural requirements. Operational Records: • Plant Layout Plans - indicating locations of processes, roads, material handling points, stockpile areas, stacks, monitoring sites, etc. • Production figures - Nature and volumes of product over past year; properties (moisture content, particle size distribution, composition, etc.), storage areas, handling points, emission mitigation measures in place and associated control efficiencies. • Raw Material Records: Nature and volumes of raw materials handled on site; material properties (moisture content, particle size distribution, composition, etc.), material storage areas, handling points, emission mitigation measures in place and associated control efficiencies. • Waste Records: Nature and volumes of waste generated; material properties (moisture content, particle size distribution, composition, etc.), waste storage areas, handling points, emission mitigation measures in place and associated control efficiencies. Emissions Inventory Information: • Typically comprises information on location and configuration of sources including: source coordinates, release heights, stack/vent diameters, efflux velocities, gas exit temperatures. • Emissions for each source including description of temporal variations in emissions with maximum hourly, maximum daily and average emission rates specified. • Documentation of methods used in quantifying emissions (e.g. continuous emissions monitoring, source sampling, emission factor application or use of chemical mass balances) • Margin of accuracy of emission quantification method • Management and mitigation measures in place for each source and associated control efficiencies Air Pollution Abatement Equipment Records: • Procedures for tracking operation of abatement equipment, calculating control efficiencies and availabilities • Control efficiency and availability records Emission and Air Quality Monitoring Records: • Monitoring Protocols / Procedures / Programmes • Calibration records • Raw monitoring data • Data screening and validation procedures

AQA Listed Activity Compliance Protocol (Draft) 12 Issue 1.0/ July 2007

• Data storage mechanisms • Monitoring reports including non-compliance and annual reports Emission Control Officer Information (if applicable): • Full contact details • Qualifications Other Records: • Atmospheric Impact Report – if such a report was required by an air quality officer under the AQA • Incidence Register: Register which lists incidence which have occurred including nature of incident, date, results of root cause analysis and actions taken to prevent reoccurrence. • Complaints Register: Information typically included is date and time of complaint, contact details of person complaining, results of investigations in cause of complaint, actions taken (where necessary), follow up correspondence with person making complaint.

Audit Reports: • Internal Audits - Undertaken by the organisation operating the facility. • External Audit - Undertaken by external, independent organisations or consultants with the relevant required expertise.

4.5 Typical Physical Features to Inspect

The following facilities are common to many heavy industrial sites and may be inspected during a compliance inspection:

• Raw materials handling, storage and preparation areas (including solid materials and tank farm areas)

• Production processes (e.g. crushing, smelter, fuel burning) – noting extent of fugitive emissions, primary and secondary emission capture systems, building extraction systems, etc.

• Process control rooms

• Pollution abatement equipment (e.g. baghouses, electrostatic precipitators)

• Waste handling, storage and disposal areas

• Product handling and storage areas

• On-site road network (if significant hauling activities, particularly if roads are unpaved)

• Conveyor lines

• Stack monitoring points

• Ambient air quality and meteorological monitoring sites and facilities

• Meteorological stations

AQA Listed Activity Compliance Protocol (Draft) 13 Issue 1.0/ July 2007

5 TYPICAL INSPECTION PROCESS

Preparation for Routine (Follow Up) Compliance Inspection

Enter target premises

Obtain copy of Atmospheric Emission Licence

If Licence valid, undertake site inspection according to checklists

Rate compliance according to scoring sheet

Feed site compliance score into If no Licence Prioritisation System

Report to Licencing Record major non-compliances Authority & Enforcement Authority

Record minor

non-compliances

Issue one of following depending on nature of non-compliance & compliance history of facility: Issue Directive & formal or informal warning, pre- Penalty compliance notice, pre-directive, compliance notice or directive

Follow Up Follow Up

Process flow diagram for a typical inspection procedure. (items boxed in red are to be undertaken by the Compliance Inspection Authority.

AQA Listed Activity Compliance Protocol (Draft) 14 Issue 1.0/ July 2007

6 CHECKLISTS

The compliance checklists are structured to follow the format of the Atmospheric Emission Licence template to ensure that all aspects are covered and facilitate potential modification of checklists should requirements have been removed or added in the facility-specific licence.

The following Categories of Criteria were defined and may be assessed during a compliance inspection for Listed Activities that are currently operational:

ATMOSPHERIC EMISSION LICENCE REQUIREMENTS: • Validity of Licence • General Conditions • Nature of Process / Raw Material & Products • Atmospheric Emission Monitoring & Compliance • Appliances & Measures to Prevent Air Pollution • Abnormal Releases and Emergency Response • Ambient Air Quality Monitoring • Energy Conservation & Cleaner Production • Complaints Register • Non-compliance Response & Reporting • Annual Reporting • Investigations & Reviews • Waste Disposal (Waste generated by Air Pollution Abatement) • Issues arising from Previous Inspections

OTHER POTENTIAL COMPLIANCE REQUIREMENTS (UNDER THE AIR QUALITY ACT) • Pollution Prevention Plans • Emission Control Officer • ‘Controlled Fuels’ • Atmospheric Impact Report

The Checklists comprising the specific criteria to be verified within each of the above categories are given in Appendix A.

7 COMPLIANCE RATING AND INTERPRETING RESULTS

A Scoring Sheet (overleaf) has been prepared and uses the concept of a rating out of the maximum score which has been developed from the Checklists given in Appendix A. The scoring system facilitates: • evaluation of the overall compliance level; • identification of key areas (e.g. operations, monitoring where compliance is poor; and • assessment of the environmental risk associated with non-compliance.

The Field Sheets to be used when conducting the compliance inspection are given in Appendix C with the maximum scores for each Category of Criteria indicated. The operations of the Listed Activity is scored during the compliance inspection with scores recorded on the Field Sheets in the “Actual Score” column to facilitate the subsequent Scoring of the Listed Activity as illustrated by the Scoring Sheet (overleaf and Appendix B).

AQA Listed Activity Compliance Protocol (Draft) 15 Issue 1.0/ July 2007

Example of the Scoring of the Compliance Status of a Listed Activity: FACILITY NAME: Aspen Iron & Steel Works(a) CONTROL CATEGORY: 3 Derived From ACTUAL / Predetermined % SCORE x from MAX FIELD MAXIMUM x 5 = Critical WEIGHTING DATE OF INSPECTION: 25/05/2006 SCORE SHEETS 100 1 = Non-critical INSPECTOR NAME: Andrew Taylor SHEET

Compliance Level Calculator Risk Calculator MAXIMUM ACTUAL COMPLIANCE CRITERIA SCORE SCORE(a) % SCORE WEIGHTING SUB TOTAL ATMOSPHERIC EMISSION LICENCE VALIDITY OF LICENCE 5 5 100.00 1.0 100.00 GENERAL CONDITIONS 4 2 50.00 3.0 150.00 NATURE OF PROCESS / RAW MATERIALS & PRODUCTS 6 6 100.00 5.0 500.00 ATMOSPHERIC EMISSIONS MONITORING & COMPLIANCE 4 1 25.00 4.0 100.00 APPLIANCES AND MEASURES TO PREVENT AIR POLLUTION 6 3 50.00 5.0 250.00 ABNORMAL RELEASES AND EMERGENCY RESPONSE 2 1 50.00 3.0 150.00 AMBIENT AIR QUALITY MONITORING 5 4 80.00 5.0 400.00 ENERGY CONSERVATION & CLEANER PRODUCTION 2 0 0.00 2.0 0.00 COMPLAINTS REGISTER 3 3 100.00 2.0 200.00 NON-COMPLIANCE RESPONSE & REPORTING 4 2 50.00 5.0 250.00 ANNUAL REPORTING 2 2 100.00 1.0 100.00 INVESTIGATIONS & REVIEWS 1 1 100.00 1.0 100.00 WASTE DISPOSAL 1 1 100.00 2.0 200.00 ISSUES ARISING FROM PREVIOUS INSPECTION 2 1 50.00 5.0 250.00 OTHER COMPLIANCE REQUIREMENTS (UNDER THE AQA) 4 4 100.00 3.0 300.00 TOTALS 51 36 1055.00 47.0 3050.00

% compliance % compliance unweighted (b) 70.6 weighted (c) 64.9

(a) The facility depicted and all text in red is representative of a fictitious example and is for illustrative purposes only. (b) % Compliance Unweighted = (total actual score / total maximum score) x 100 = (36 / 51) x 100 = 71% (c) % Compliance Weighted = (total weighted score / max weighted score) x 100 = (3050 / 4700) x 100 = 65%

16 AQA Listed Activity Compliance Protocol (Draft) Issue 1.0/ July 2007

7.1 Overall Compliance Level (Legal Compliance with Conditions of Licence)

An estimation of the ‘level’ to which a site is in compliance with the permit can be roughly calculated from the % of checkboxes on the Field Sheet which are ‘ticked’ to verify compliance.

Following the inspection and entry of data into a spreadsheet, a total score (sum of total number of checkboxes ‘ticked’) is provided by the spreadsheet. This number is then expressed as a % of the possible maximum total score, again this is automatically calculated by the spreadsheet and expressed as the “% Compliance Unweighted”.

For the Aspen Iron & Steel Works example:

Total Actual Compliance Score = 36

Maximum Possible Compliance Score = 51

% compliance with elements of licence = (36/51) x 100 = 70.6%

i.e. 71% of criteria were complied with

This figure is then used to assign a ‘banding’ of compliance level from the following table:

COMPLIANCE BAND COMPLIANCE %

Compliant 100%

Generally compliant 80 - 99 %

Moderate non-compliance 60 - 79%

Significant non-compliance 40-59 %

Severe non compliance 0 - 39%

For this example, the banding is moderate non-compliance

7.2 Key Non-Compliance Areas and Potential for Environmental Risk

In cases of non-compliance being found, the inspector needs more information to understand the source of the non-compliance and to identify the urgency and potential environmental risk of the situation so as to provide the most appropriate response.

Key Areas were identified which are representative of common functions across the various Categories of Criteria to illustrate the nature of the activity responsible for the non- compliances. The Key Area of ‘Incident Reporting’ for example includes criteria from the following three Categories of Criteria: • Abnormal Releases and Emergency Response – notification of abnormal releases • Complaints Register – register, report & maintain • Non-Compliance Response & Reporting – non-compliance reporting

17 AQA Listed Activity Compliance Protocol (Draft)

Key Area scores are automatically calculated in the spreadsheet when the Field Sheet is completed and areas where the Listed Activity has scored low (against possible maximum scores) are evident as in the below example.

For the Aspen Iron & Steel Works example:

Compliance Level Calculator MAXIMUM ACTUAL KEY AREA SCORE SCORE % SCORE ATMOSPHERIC EMISSION LICENCE Administration 6 6 100 General Plant Operation 6 6 100 Air Pollution Control 14 7 50 Emissions Monitoring & Reporting 5 2 40 Ambient Air Quality Monitoring 4 4 100 Ambient Air Quality Targets 1 0 0 Incident Reporting 5 4 80 Routine Reporting 2 2 100 Response to Non-compliance 4 1 25 OTHER COMPLIANCE REQUIREMENTS (UNDER AQA) 4 4 100 TOTALS 51 36 71

Based on the above table it is evident that the lowest scores indicative of significant to moderate non-compliances are related to: • Compliance with air quality targets (0%) • Response to non-compliance (25%) • Emissions monitoring and reporting (40%) • Air pollution control (50%)

Despite these non-compliance areas, it is evident that the Listed Activity ‘generally complies’ or ‘complies’ in terms of Administration, General Plant Operation, Air Quality Monitoring, Incident Reporting and other compliance requirements under the Air Quality Act.

The response to the non-compliance should therefore focus on prompting improvements in Emissions Monitoring and Reporting and Air Pollution Control (which may also address Ambient Air Quality Target exceedances) and ensuring that the Listed Activity improves its internal responsiveness to non-compliant behaviour.

7.3 Environmental Risks due to Non-Compliance

From an environmental perspective, it is important to identify whether the activity that caused the non-compliance is having a negative impact on human health and welfare or on the broader environment. Non-compliances related to Atmospheric Emissions Monitoring and Compliance or implementation of Air Pollution Controls could, for example, have significant implications in terms of environmental risk and would prompt more timely action compared to Administrative non-compliances which are unlikely to directly impact the environment or human well-being.

To provide an indication of the level of likely risk of environmental impact related to non- compliance, the inspector can use the weighted scores shown in green on the right hand columns of the scoring sheet under the Risk Calculator columns. This enables the weighting of each Category of Criteria in terms of the environmental risk it poses. The weighting of the Categories of Criteria have been suggested by experts in the field of air quality management and have been assigned in a way in which a high weighting corresponds to a high level of potential risk to the environment whilst a low weighting corresponds to a low potential risk.

18 AQA Listed Activity Compliance Protocol (Draft)

Using the weighted score, the inspector can allocate a banding based on the table below. This banding is indicative only, i.e. it must be considered as a warning system which gives the inspector a rough estimation of the possible likely risk caused by this site on a broad scale from ‘none’ to ‘serious/significant’.

Potential Environmental Risk from Non- Weighted % score Compliance None 100 Low 76 – 99 Medium 51- 75 Significant 26 – 50 Severe/ urgent 0 - 25

If the site inspection reveals a level of risk that the inspector is concerned about (and this is reinforced by his/her professional judgement), the inspector can then use supplementary information collected during the inspection, such as that listed in Appendix D, to add further detail to the assessment of risk.

An additional summary table can also be produced which extracts the scores for those elements of the licence which are given a weighting of 5 (i.e. the sources of greatest risk at the premises if poorly managed or absent).

MAXIMUM ACTUAL CRITICAL COMPLIANCE CRITERIA SCORE SCORE % SCORE ATMOSPHERIC EMISSIONS MONITORING & COMPLIANCE 4 1 25 APPLIANCES AND MEASURES TO PREVENT AIR POLLUTION 6 3 50 ABNORMAL RELEASES AND EMERGENCY RESPONSE 2 1 50 AMBIENT AIR QUALITY MONITORING 5 4 80 NON-COMPLIANCE RESPONSE & REPORTING 4 2 50 ISSUES ARISING FROM PREVIOUS INSPECTION 2 1 50

The inspector and enforcement colleagues will then have sufficient information to decide an appropriate ‘non-compliance response’ (based on enforcement policy/guidance).

7.4 Results Summary

This section provides an example of the typical results and conclusions that could be collected once a site has been inspected and uses the example that has been scored in the Scoring Sheet.

Aspen Iron & Steel Works example:

Main findings related to Compliance Status: • Overall moderate non-compliance level based on unweighted compliance score • General compliance in terms of Administration, General Plant Operation, Air Quality Monitoring, Incident Reporting and other compliance requirements under the Air Quality Act • Significant areas contributing to the non-compliance status are in regard to: o Air Quality targets o Operator Response to non-compliances o Emission Monitoring and Reporting Systems • Inadequacies in Air Pollution Control practices also contribute to the level of non- compliance (specifically in relation to certain point source emission limits being violated).

19 AQA Listed Activity Compliance Protocol (Draft)

Main findings related to Environmental Risk Potentials: • Weighted compliance score indicates a medium potential risk to the environment. • Further scrutiny of the weighted information indicates that the facility scored particularly poorly for atmospheric emissions monitoring and compliance with emission limits, scoring “1” out of a possible score of “4” with this category being weighted as 5, i.e. high risk. • Emission limit exceedances appear to be related to certain abatement equipment not being maintained at the required control efficiencies and availabilities due to poor maintenance.

Supplementary information gathered (as per Appendix D) which has relevance: • Nearest sensitive receptors are located within 500 m of the plant boundary • No previous enforcement activity • No Environmental Management System currently in place, but one in the process of being established • Number of previous complaints received in relation to dust

Concluding notes:

Emissions in excess of limits are occurring at the facility. Sensitive receptors are situated nearby and complaints have been received in respect of dust.

Ambient air quality targets are exceeded for fine particulates but it is noted that various sources other than the facility contribute to suspended particulate concentrations at the monitoring site.

Although a system for inventorying and investigating complaints is in place, no indepth root cause analysis of non-compliance events is undertaken nor measures implemented to prevent recurrence of such events. For example, there are no apparent plans in place to improve the performance of abatement equipment so as to ensure on-going compliance with emission limits.

Positive components of the facility include: (a) valid licence, (b) general compliance with conditions related to general plant operations, (c) absence of any previous enforcement action, (d) establishment of an EMS is underway.

This information is discussed with enforcement officials and a course of non- compliance action is identified.

20 AQA Listed Activity Compliance Protocol (Draft)

8 ADDITIONAL SITE INFORMATION

Appendix D includes additional information which should be collected when the premises of a Listed Activity is inspected. The collation of such information provides the inspector with greater insight into the voluntary environmental management systems being implemented and the ambient receiving environment, particularly in respect of the proximity of sensitive receptors.

The information listed is also required to provide the additional input needed for the DEAT information management system which is used to populate the prioritisation matrix applied to determine routine inspection frequency requirements for facilities.

9 FIELD SHEETS

Appendix C includes an example of the field sheets to be completed by an inspector during a typical inspection of a Listed Activity premises. Inspectors may modify such sheets based on the specific Atmospheric Emission Licence issued for a particular Listed Activity to maximise the compliance assessment and identification of problem areas.

The following terms are noteworthy:

Major Non-Compliance Definition:

A deviation that is causing or has the potential to cause serious environmental damage or pose a significant risk to human health either within and/or beyond the boundary of the premises.

Minor Non-Compliance Definition:

A deviation that is not causing or has the potential to cause serious environmental damage or pose a significant risk to human health either within or beyond the boundary of the premises.

10 CONCLUSION

This protocol has been developed as part of the Compliance Monitoring Handbook and provides a guidance protocol for compliance inspections at facilities classified as Listed Activities in terms of the National Environmental Management: Air Quality Act (AQA), Act 39 of 2004. The format for this protocol follows the United States Environmental Protection Agency Protocols for Conducting Environmental Compliance Audits.

The use of this protocol during compliance inspections will ensure that compliance with the key aspects of Atmospheric Emission Licences are evaluated and recorded. Given that the protocol was based on the draft generic Atmospheric Emission Licence template it is cautioned that the protocol may need to be updated following finalisation of the generic template. Furthermore, that inspectors should be permitted to modify the compliance checklists on the basis of the Listed Activity specific Atmospheric Emission Licence (AEL) in instances where such licences cover requirements not included in the generic AEL template and hence in this protocol.

21 AQA Listed Activity Compliance Protocol (Draft)

be followed. Other approaches and techniques can also be exploited in order to evaluate compliance of a facility.

11 REFERENCES

Atmospheric Pollution Prevention Act, Act 45 of 1965 (as amended)

DEAT, March 2007, Information Document, Compliance Inspections by Environmental Management Inspectors of Facilities with Environmental Authorisations, Environmental Management Inspectorate, Department of Environmental Affairs and Tourism.

National Environmental Management: Air Quality Act, Act 39 of 2004.

National Environmental Management Act, Act 107 of 1998

National Environmental Management Amendment Act No. 46 of 2003.

NSW DECC, February 2006, Compliance Audit Handbook, Compliance and Assurance Section, New South Wales Department of Environment and Climate Change, Australia.

US-EPA, 1997, Environmental Audit Program Design Guidelines for Federal Agencies, Office of Enforcement and Compliance Assurance, United States Environmental Protection Agency, Report No. EPA 300-B-96-011.

US-EPA, April 2001, Clean Air Act, Stationary Source Compliance Monitoring Strategy, United States Environmental Protection Agency.

US-EPA, August 2002, Conducting Environmental Compliance Inspections, Inspector’s Field Manual, International Edition, Office of Enforcement and Compliance Assurance, United States Environmental Protection Agency.

22 AQA Listed Activity Compliance Protocol (Draft)

APPENDIX A

Maximum Score Calculation Checklists

23 AQA Listed Activity Compliance Protocol (Draft)

APPENDIX A - MAXIMUM SCORE CALCULATION CHECKLISTS

AUTHORISATION / CRITERIA - VERIFY THE FOLLOWING: REGULATION CATEGORY SCORE ATMOSPHERIC VALIDITY OF Licence exists and determine if any amendments 1 EMISSION LICENCE LICENCE Licence holder and contact details are correct 1 (IN TERMS OF AQA, Facility physical location matches licence 1 CHAPTER 5) Control class designated & appropriate 1 Validity of licence (expiry date) 1 Maximum Score 5 GENERAL Plant and apparatus used in the process and all appliances for preventing or reducing to a minimum CONDITIONS atmospheric emissions are properly maintained and operated and that the necessary measures are taken to prevent atmospheric emissions 1 Licence holder understands the composition of and has an awareness of the dangers of harmful effects on the internal and external environment caused by chemicals and raw materials used or products manufactured; including knowledge of bio-degradability, toxicity, bio-accumulation and sensitising properties of each substance. (Failure to comply with this condition is a breach of the 'duty of care' responsibility stipulated in section 28 of NEMA.) 1 No building, plant or works has been materially extended, altered or added without the necessary authorisation. 1 Measurement, calculation and/or sampling and analysis is being carried out in accordance with nationally or internationally acceptable standards. In the event that a different method is applied, has the DEAT been consulted with and agreed based on satisfactory documentation confirming the equivalence and reliability of the method. 1 Maximum Score 4 NATURE OF Material flows (including raw material, product, by-product and waste streams) are monitored and recorded 1 PROCESS / RAW No unauthorised changes in the nature of the production process or significant production increases have been MATERIALS & undertaken which may significantly alter impacts, specifically changes in: PRODUCTS Types and quantities of raw materials used 1 Types and quantities of products produced 1 Types and quantities of by-products produced 1 Types and quantities of fuels used and sulphur and ash contents of fuels (if applicable) 1 Hours of operation are as specified in the licence or of shorter duration 1 Maximum Score 6 ATMOSPHERIC Point source emission monitoring and reporting requirements (including sampling/monitoring method, frequency, EMISSIONS duration, parameters measured, parameters recorded, reporting frequency) are being met 1 MONITORING & Point sources are operating at the efflux velocity and temperature given in the licence 1 COMPLIANCE Point source emissions are within maximum and average emission limits 1 Area and/or line source emission rates have been measured/calculated in accordance with national or international standards or using an equivalent method approved by the DEAT, with emission rates recorded and available for inspection 1 Maximum Score 4

25 AQA Listed Activity Compliance Protocol (Draft) Issue 1.0/ July 2007

AUTHORISATION / CRITERIA - VERIFY THE FOLLOWING: REGULATION CATEGORY SCORE ATMOSPHERIC APPLIANCES Measures are in place to record the control efficiency and availability of abatement equipment and records in this EMISSION LICENCE AND MEASURES regard are kept and are available for inspection 1 (IN TERMS OF AQA, TO PREVENT AIR Abatement equipment is operating at the permitted minimum control efficiency 1 CHAPTER 5) POLLUTION Abatement equipment is operating at the permitted minimum utilization (availability) 1 Management and mitigation measures specified for area sources are being implemented 1 Measures are in place to measure the effectiveness of area source management and mitigation measures 1 Management and mitigation measures for area sources meet the required control efficiency 1 Maximum Score 6 ABNORMAL Emergency procedures and contingency plans have been formulated for all abnormal atmospheric release RELEASES AND related risks (e.g. accidental emissions, leakages, control technology outages) as set out in section 30 of NEMA 1 EMERGENCY Notifications of abnormal releases have been undertaken in the manner and timeframe required under section RESPONSE 30 of NEMA 1 Maximum Score 2 AMBIENT AIR Ambient air quality monitoring required is being undertaken in accordance with the monitoring location, QUALITY pollutants to be measured, monitoring/sampling method, frequency and duration and the reporting frequency MONITORING stipulated 1 Monitoring/sampling is conducted in accordance with national or international standards or using an equivalent method approved by the DEAT 1 Instrument calibration certificates are maintained and are available for inspection 1 Raw data are kept for a minimum period of five years and are available for inspection on request 1 Ambient air quality levels are within the required targets 1 Maximum Score 5 ENERGY Progress is being made to implement the required energy conservation measures by the target date 1 CONSERVATION Progress is being made to implement the required cleaner production measures by the target date 1 & CLEANER Maximum Score PRODUCTION 2 COMPLAINTS An air quality complaints register is maintained and available for inspection 1 REGISTER Systems are in place to investigate and report complaints to authorities in the manner required 1 Complaints records are kept for at least five years 1 Maximum Score 3 NON- Non-compliance with permitted emission limits is recorded and reported to the Licencing Authority on a monthly COMPLIANCE basis in the manner and formate required, within fifteen (15) working days of month end. 1 RESPONSE & Root cause analysis of non-compliance events is undertaken 1 REPORTING The extent of emissions and impacts associated with non-compliance incidents is calculated/modelled where applicable 1 Measures have been implemented or are to be implemented by a specific date to prevent recurrence of non- compliances 1 Maximum Score 4

26 AQA Listed Activity Compliance Protocol (Draft) Issue 1.0/ July 2007

AUTHORISATION / CRITERIA - VERIFY THE FOLLOWING: REGULATION CATEGORY SCORE ATMOSPHERIC ANNUAL Annual reports are completed and submitted in the manner and format required by the Licencing Authority 1 EMISSION LICENCE REPORTING Copies of annual reports are kept for a period of at least five years 1 (IN TERMS OF AQA, Maximum Score 2 CHAPTER 5) INVESTIGATIONS The investigations required have been completed by the required date 1 & REVIEWS Maximum Score 1 WASTE The disposal of other waste produced (solid waste, effluent) arising from air pollution mitigation measures DISPOSAL complies with the applicable regulations and requirements of the relevant authorities 1 Maximum Score 1 ISSUES ARISING Non-compliance issues identified at the previous inspection have been dealt with and resulted in compliance 1 FROM PREVIOUS Pollution issues identified at the previous inspection has been dealt with effectively and are no longer of concern 1 INSPECTION Maximum Score 2 OTHER POLLUTION The licence holder has prepared, submitted for approval and subsequently implement a pollution prevention plan COMPLIANCE PREVENTION in respect of a priority air pollutant, if required to do so REQUIREMENTS PLANS 1 (UNDER THE AQA) EMISSION An emission control officer with the necessary air quality management competence in respect of the listed CONTROL activity has been appointed and is undertaking the necessary functions, if required to do so OFFICER 1 CONTROLLED Any 'controlled fuels' being used, manufacturer or sold at the premises are done so in accordance with the FUELS' necessary regulations 1 ATMOSPHERIC Has an atmospheric impact report been prepared in the manner prescribed and submitted, if required to do so IMPACT REPORT 1 Maximum Score 4

27 AQA Listed Activity Compliance Protocol (Draft) Issue 1.0/ July 2007

APPENDIX B

Scoring Sheet

28 AQA Listed Activity Compliance Protocol (Draft)

% compliance % compliance unweighted (b) 70.6 weighted (c) 64.9

29 AQA Listed Activity Compliance Protocol (Draft) Issue 1.0/ July 2007

APPENDIX C

Field Sheets

30 AQA Listed Activity Compliance Protocol (Draft) Issue 1.0/ July 2007

APPENDIX C - ATMOSPHERIC EMISSION LICENCE - COMPLIANCE ASSESSMENT FIELD SHEET FACILITY NAME: CONTROL CATEGORY: DATE OF INSPECTION: INSPECTOR NAME: NON-COMPLIANCE CATEGORY CRITERIA - VERIFY THE FOLLOWING: SCORE OBSERVATION MAJOR MINOR COMMENT ATMOSPHERIC EMISSION LICENCE VALIDITY OF Licence exists and determine if any amendments 1 LICENCE Licence holder and contact details are correct 1 Facility physical location matches licence 1 Control class designated & appropriate 1 Validity of licence (expiry date) 1 Maximum Score 5 GENERAL Plant and apparatus used in the process and all CONDITIONS appliances for preventing or reducing to a minimum atmospheric emissions are properly maintained and operated and the necessary measures are taken to prevent atmospheric emissions 1 Licence holder understands the composition of and has an awareness of the dangers of harmful effects on the internal and external environment caused by chemicals and raw materials used or products manufactured; including knowledge of bio-degradability, toxicity, bio-accumulation and sensitising properties of each substance. 1 No building, plant or works has been materially extended, altered or added without the necessary authorisation. 1 Measurement, calculation and/or sampling and analysis is being carried out in accordance with nationally or internationally acceptable standards. In the event that a different method is applied, has the DEAT been consulted with and agreed based on satisfactory documentation confirming the equivalence and reliability of the method. 1 Maximum Score 4

31 AQA Listed Activity Compliance Protocol (Draft) Issue 1.0/ July 2007

NON-COMPLIANCE CATEGORY CRITERIA - VERIFY THE FOLLOWING: SCORE OBSERVATION MAJOR MINOR COMMENT ATMOSPHERIC EMISSION LICENCE NATURE OF Material flows (including raw material, product, by- PROCESS / RAW product and waste streams) are monitored and MATERIALS & recorded 1 PRODUCTS No unauthorised changes in the nature of the production process or significant production increases have been undertaken which may significantly alter impacts, specifically changes in:

Types and quantities of raw materials used 1 Types and quantities of products produced 1 Types and quantities of by-products produced 1 Types and quantities of fuels used and sulphur and ash contents of fuels (if applicable) 1 Hours of operation are as specified in the licence or of shorter duration 1 Maximum Score 6 ATMOSPHERIC Point source emission monitoring and reporting EMISSIONS requirements (including sampling/monitoring MONITORING & method, frequency, duration, parameters COMPLIANCE measured, parameters recorded, reporting frequency) are being met 1 Point sources are operating at the efflux velocity and temperature given in the licence 1 Point source emissions are within maximum and average emission limits 1 Area and/or line source emission rates have been measured/calculated in accordance with national or international standards or using an equivalent method approved by the DEAT, with emission rates recorded and available for inspection 1 Maximum Score 4

32 AQA Listed Activity Compliance Protocol (Draft) Issue 1.0/ July 2007

NON-COMPLIANCE CATEGORY CRITERIA - VERIFY THE FOLLOWING: SCORE OBSERVATION MAJOR MINOR COMMENT ATMOSPHERIC EMISSION LICENCE APPLIANCES AND Measures are in place to record the control MEASURES TO efficiency and availability of abatement equipment PREVENT AIR and records in this regard are kept and are POLLUTION available for inspection 1 Abatement equipment is operating at the permitted minimum control efficiency 1 Abatement equipment is operating at the permitted minimum utilization (availability) 1 Management and mitigation measures specified for area sources are being implemented 1 Measures are in place to measure the effectiveness of area source management and mitigation measures 1 Management and mitigation measures for area sources meet the required control efficiency 1 Maximum Score 6 ABNORMAL Emergency procedures and contingency plans RELEASES AND have been formulated for all abnormal EMERGENCY atmospheric release related risks (e.g. accidental RESPONSE emissions, leakages, control technology outages) as set out in section 30 of NEMA 1 Notifications of abnormal releases have been undertaken in the manner and timeframe required under section 30 of NEMA 1 Maximum Score 2

33 AQA Listed Activity Compliance Protocol (Draft) Issue 1.0/ July 2007

NON-COMPLIANCE CATEGORY CRITERIA - VERIFY THE FOLLOWING: SCORE OBSERVATION MAJOR MINOR COMMENT ATMOSPHERIC EMISSION LICENCE AMBIENT AIR Ambient air quality monitoring required is being QUALITY undertaken in accordance with the monitoring MONITORING location, pollutants to be measured, monitoring/sampling method, frequency and duration and the reporting frequency stipulated 1 Monitoring/sampling is conducted in accordance with national or international standards or using an equivalent method approved by the DEAT. 1 Instrument calibration certificates are maintained and are available for inspection 1 Raw data are kept for a minimum period of five years and are available for inspection on request 1 Ambient air quality levels are within the required targets 1 Maximum Score 5 ENERGY Progress is being made to implement the required CONSERVATION & energy conservation measures by the target date 1 CLEANER Progress is being made to implement the required PRODUCTION cleaner production measures by the target date 1 Maximum Score 2 COMPLAINTS An air quality complaints register is maintained REGISTER and available for inspection 1 Systems are in place to investigate and report complaints to authorities in the manner required 1 Complaints records are kept for at least five years 1 Maximum Score 3

34 AQA Listed Activity Compliance Protocol (Draft) Issue 1.0/ July 2007

NON-COMPLIANCE CATEGORY CRITERIA - VERIFY THE FOLLOWING: SCORE OBSERVATION MAJOR MINOR COMMENT ATMOSPHERIC EMISSION LICENCE NON- Non-compliance with permitted emission limits is COMPLIANCE recorded and reported to the Licencing Authority RESPONSE & on a monthly basis in the manner and formate REPORTING required, within fifteen (15) working days of month end. 1 Root cause analysis of non-compliance events is undertaken 1 The extent of emissions and impacts associated with non-compliance incidents is calculated/modelled where applicable 1 Measures have been implemented or are to be implemented by a specific date to prevent recurrence of non-compliances 1 Maximum Score 4 ANNUAL Annual reports are completed and submitted in the REPORTING manner and format required by the Licencing Authority 1 Copies of annual reports are kept for a period of at least five years 1 Maximum Score 2 INVESTIGATIONS The investigations required have been completed & REVIEWS by the required date 1 Maximum Score 1 WASTE DISPOSAL The disposal of other waste produced (solid waste, effluent) arising from air pollution mitigation measures complies with the applicable regulations and requirements of the relevant authorities 1 Maximum Score 1

35 AQA Listed Activity Compliance Protocol (Draft) Issue 1.0/ July 2007

NON-COMPLIANCE CATEGORY CRITERIA - VERIFY THE FOLLOWING: SCORE OBSERVATION MAJOR MINOR COMMENT ATMOSPHERIC EMISSION LICENCE ISSUES ARISING Non-compliance issues identified at the previous FROM PREVIOUS inspection have been dealt with and resulted in INSPECTION compliance 1 Pollution issues identified at the previous inspection has been dealt with effectively and are no longer of concern 1 Maximum Score 2 OTHER COMPLIANCE REQUIREMENTS (UNDER THE AQA) POLLUTION The licence holder has prepared, submitted for PREVENTION approval and subsequently implement a pollution PLANS prevention plan in respect of a priority air pollutant, if required to do so 1 EMISSION An emission control officer with the necessary air CONTROL quality management competence in respect of the OFFICER listed activity has been appointed and is undertaking the necessary functions (if applicable) 1 CONTROLLED Any 'controlled fuels' being used, manufacturer or FUELS' sold at the premises are done so in accordance with the necessary regulations 1 ATMOSPHERIC Has an atmospheric impact report been prepared IMPACT REPORT in the manner prescribed and submitted, if required to do so 1 Maximum Score 4

36 AQA Listed Activity Compliance Protocol (Draft) Issue 1.0/ July 2007

APPENDIX D

Additional Information to be Collected for each Listed Activity

37 AQA Listed Activity Compliance Protocol (Draft) Issue 1.0/ July 2007

Supplementary Information about the Listed Activity to be Collected during Site Inspections

The additional information which should be collected when the premises of a Listed Activity is inspected are given below. The collation of such information provides greater insight into the voluntary environmental management systems being implemented and the ambient receiving environment, particularly in respect of the proximity of sensitive receptors. The information listed is also required to provide the additional input needed for the DEAT information management system which is used to populate the prioritisation matrix applied to determine routine inspection frequency requirements for facilities.

Criteria Facility Information Classification of Industry Type Site Location Coordinates (Latitude, Longitude) of Approximate Center of Operations (Decimal Degrees) Current Operating Status Fully operational Partially operational (% of permitted production) Not operational (date at which operations are likely to commence) Closing

Proximity to Human Negligible (no residencies within 3 km) Occupation Low (<10 residencies within 3 km) Medium (10 – 500 residencies within 3 km) High (urban, >500 residencies within 3 km)

Environmental Yes Management System No

Audit Status None Internal Independent Enforcement History None Administrative Action Taken Civil Court Action Taken Criminal Charges Laid Prosecution Successful Fines allocated

Complaints Frequency of complaints: 1 - Never 2 – Infrequent (one or twice in the past 5 years) 3 – Fairly frequent (1 to 3 per year) 4 – Frequent (almost every month) 5 – Very frequent (almost continuously)

Years since last Never inspected, or inspection Number of years since last known inspection

38 AQA Listed Activity Compliance Protocol (Draft) Issue 1.0/ July 2007