60514 Extractive Industries for Development Series #12 July 2009 Sub-Saharan Africa Refinery Study Report Summary EnSys Energy & Systems, ICF International, James Hammitt, Lisa Robinson World Bank Group's Oil, Gas, and Mining Policy Division Oil, Gas, Mining, and Chemicals Department A joint service of the World Bank and the International Finance Corporation The Oil, Gas, and Mining Policy Division series publishes reviews and analyses of sector experience from around the world as well as new findings from analytical work. It places particular emphasis on how the experience and knowledge gained relates to developing country policy makers, communities affected by extractive industries, extractive industry enterprises, and civil society organizations. We hope to see this series inform a wide range of interested parties on the opportunities as well as the risks presented by the sector. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors and should not be attributed in any manner to the World Bank or its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequence of their use. Extractive Industries for Development Series #12 July 2009 Sub-Saharan Africa Refinery Study Report Summary EnSys Energy & Systems, ICF International, James Hammitt, Lisa Robinson COPYRIGHT © 2009 http://www.worldbank.org/ogmc (OR /oil OR /gas OR /mining) http://www.ifc.org/ogmc (OR /oil OR /gas OR /mining) Cover Photos: Oil rig, hematite-banded ironstone, LNG tanker Table of Contents v Acknowledgements vi List of Acronyms and Definitions of terms Used in the Report 1 Introduction 3 Summary of the Main Results 10 Health Study Overview 19 Refinery Study Overview 29 SSA Refinery Project Results and Conclusions E x t ractive Industries for Development Series iii List of Figures Figure 1: Relationship of the Health Study and the Refinery Study ................................... 2 Figure 2: Sub-Saharan Africa ­ Selected Regions ................................................................ 3 Figure 3: Modeled Reductions of PM2.5 (tons/year) for Vehicle Sources in Three Cities ....12 Figure 4: Sub-Saharan Africa Refinery Regional Groupings.............................................. 22 List of Tables Table 1: Monetary Valuation of the Health Benefits derived from the improvement of fuels quality................................................................................................................... 4 Table 2: Total SSA Refinery Investments from the Eight Cases (Billions of 2007$)............ 5 Table 3: Incremental SSA Refinery Investments from Moving to AFRI-4 Fuel Specifications (Billions of 2007$)................................................................................................ 5 Table 4: Five-Year and Ten-Year Net Present Value of Refinery Investment Costs versus Health Benefits ..................................................................................................... 6 Table 5: Major Investments Proposed by the WORLD® Model for the 2020 Constrained and Open Market Cases........................................................................................ 8 Table 6: Cities Modeled in the Health Study.................................................................... 11 Table 7: Air Pollutants and Related Health Effects ........................................................... 11 Table 8: Modeled Annual Average Total Particulate Matter Concentrations for Three Cities (population density greater than 1,000/km)........................................................ 11 Table 9: Major Differences between SSA Regions Reflected in Modeling Results............. 17 Table 10: Health Study Regional Analysis Results Using Alternate 2-Stroke Motorcycle Assumptions ..................................................................................................... 18 Table 11: AFRI Standards for Sub-Saharan Africa Transportation Fuels ........................... 19 Table 12: WORLD® Model Constraints........................................................................... 24 Table 13: Modeling Case Descriptions ............................................................................. 25 Table 14: Refinery Throughputs of Regional Groupings in the Various Cases.................. 26 Table 15: Total Refinery Investments from the Eight Cases (Billions of 2007$) ............... 27 Table 16: Incremental Costs to Move to AFRI-4 Fuel Specifications (Billions of 2007$)...... 27 Table 17: Incremental Unit Costs for AFRI-4 Fuels in the Constrained and Open Market Cases................................................................................................................. 28 Table 18: Net Present Value of Refinery Supply Costs versus Health Benefits over 5 Years ... 27 Table 19: Net Present Value of Refinery Supply Costs versus Health Benefits over 10 Years.. 30 iv S u b-Saharan Africa Refinery Project Acknowledgements This report has been funded by the Energy Sector Management Assistance Program (ESMAP) administered by the World Bank. The concept, terms of reference and overall supervision of the project has been provided by a Steering Committee including representatives from the African Refiners Association (ARA), the United Nations Environment Program (UNEP), the International Petroleum Industry Environmental Conservation Association (IPIECA), and the Oil and Gas Policy Unit of the World Bank (COCPO) and from five governments: Kenya, Nigeria, Republic of South Africa, Republic of Congo and Côte d'Ivoire. The task team leader was Eleodoro Mayorga Alba, Principal Petroleum Economist of the World Bank. ICF International was selected as consultant. The ICF International team included: Zeta Rosenberg, project leader and Refinery Study project manager, Amy Rosenstein, Health Study project manager, Martin Tallett, EnSys, Refinery Study subcontractor, Lisa Robinson, independent consultant, Health Study subcontractor, valuation section, and James Hammitt, Harvard School of Public Health, Health Study subcontractor, valuation section. ICF International would like to thank two organizations and one individual who generously provided help for this report: · The African Refiners Association whose Chairman Joel Dervain and members, particularly the Sub-Saharan Africa members, responded fully and rapidly to our many requests for data, · CITAC Africa LLP who generously shared with us their extensive database on refineries and petroleum usage in Africa. We would particularly like to thank Mark Elliott of CITAC who provided us with many useful suggestions and comments based on his many years of experience in the petroleum industry in Africa, and · Jane Akumu from the Nairobi Office of UNEP, who went out of her way to find us SSA emission data and studies. Special thanks go to Esther Petrilli-Massey for managing production of the Summary Report in the Extractive Industries for Development Series. NOTE: For copies of the full report, please send a message to: ogmc@worldbank.org E x t ractive Industries for Development Series v List of Acronyms and Definitions of Terms Used in the Report ACS American Cancer Society AFRI Fuel specifications adopted by the Africa Refiners Association AFRO D Africa Sub-Region D, classified based on mortality data; WHO Global Burden of Disease project AFRO E Africa Sub-Region E, classified based on mortality data; WHO Global Burden of Disease project AP-42 Air Pollution-42, US EPA's Compilation of Air Pollutant Emission Factors ARA African Refiners Association asl Above sea level AVHRR Advanced Very High Resolution Radiometer BenMAP US EPA's Environmental Benefits Mapping and Analysis Program BTU British thermal units CI Confidence interval CO Carbon monoxide COEE Canada Office of Energy Efficiency COI Cost of illness CPS-II Cancer Prevention Study II DAAPs South Africa's Dynamic Air Pollution Prediction System DALY Disability-adjusted life-year deg Degrees °C Degrees centigrade DEM Digital elevation model DHS Demographic and Health Surveys DOC Diesel oxidation catalyst EPA California Environmental Protection Agency E-R Exposure-response EURO European Commission, Transport & Environment standards for gasoline and diesel g/kg Grams/kilogram gal Gallons GDP Gross domestic product GNI Gross national income GNP Gross national product NO3 Nitrate I/M Inspection/Maintenance program ICD International Classification of Disease vi S u b-Saharan Africa Refinery Project IMO International Maritime Association IPPS Industrial Pollution Projection System IR Incidence rate IVEM International Vehicle Emissions Model kg Kilograms km Kilometers LAMATA Lagos Metropolitan Area Transport Authority m Meters _g/m3 Micrograms per cubic meter m/s Minimum wind speed MARPOL International Convention for the Prevention of Pollution from Ships MW Megawatts n Number NO2 Nitrogen dioxide NOAA U.S. National Oceanographic and Atmospheric Administration NOx Nitrogen oxides O3 Ozone OR Odds ratio p Probability PM Particulate matter PM10 Particulate matter less than or equal to 10 microns PM2.5 Particulate matter less than or equal to 2.5 microns ppm Parts per million PPP Purchasing power parity PR Prevalence rate psi Pound per square inch QALY Quality-adjusted life year RFO Residual fuel oil RIVM Dutch National Institute for Public Health and the Environment (Rijksinstituut voor Volksgezondheid en Milieu) RR Relative risk or risk ratio RSA Republic of South Africa RVP Reid vapor pressure SCR Selective catalytic reduction SO2 Sulfur dioxide SO4 Sulfate SSA Sub-Saharan Africa t/y Metric tons/year tpy Metric tons/year TSP Total suspended particulates US EPA United States Environmental Protection Agency UNEP United Nations Environment Programme USAID U.S. Agency for International Development VKT Vehicle kilometers traveled VOC Volatile organic compounds VSL Value per statistical life VSLY Value per statistical life year WGS UTM World geodetic system universal transverse Mercator WHO World Health Organization WRAP Western Regional Air Partnership (US) WTP Willingness to pay E x t ractive Industries for Development Series vii viii S u b-Saharan Africa Refinery Project Introduction The World Bank and the African Refiners Association (ARA), with the concurrence of the United Nations Environment Program (UNEP) and International Petroleum Industry Environmental Conservation Association (IPIECA), wish to encourage the implementation of policies to promote health and improved urban air quality through better fuel quality, greater intra-regional trade of fuel products of standardized quality, and increased investments in the refining industry. The Sub- Saharan Africa Refinery Project evaluated the change necessary to improve fuel specifications and the impacts on refining operations and costs, air quality, and health. The study compared the costs of improving the quality of fuels produced in Sub-Saharan Africa (SSA) to the potential health benefits to the people in urban areas of the region. This is one of the first studies of these issues in SSA. Although there are uncertainties associated with the data, which are further compounded by the global recession, the project provides an estimate of the potential for health benefits associated with cleaner fuels, as well as the potential costs to the SSA refining industry to upgrade its conversion capacity and ensure clean fuel supply in competitive conditions. The study consists of two parts that interact as shown in Figure 1. Health Study: To estimate the reductions in air emissions associated with improved fuel specifications and vehicle emission controls, analyze the impact of the change on human health, and estimate the health benefits in economic terms. Refinery Study: To outline the upgrades necessary in the SSA refining sector to respond to global market and clean fuels trends, and to clarify the associated costs. In order to make decisions regarding the future of the SSA refining industry, the Health Study estimates health and monetary benefits associated with improved fuel specification and vehicle emission controls, by region, which in turn are compared to the costs to the refining industry, as developed in the Refinery Study. The project methodology was presented by ICF International and approved by the World Bank and the ARA Steering Committee. In some cases, the project has been E x t ractive Industries for Development Series 1 limited by the availability and quality of data. Associated methodological uncertainties are discussed at length in the report, and must be kept in mind when reviewing the results. Figure E-1: Relationship of the Health Study and the Refinery Study Model air pollutant Health Study reductions resulting from Health and better fuels and vehicles, monetary benefits Decision and vehicle controls. Point of improved fuels Estimate health impacts Refinery Study of improved air quality. Costs for changing manufacturing Determine monetary processes to value of health benefits. Other improve fuels Considerations production Estimate future petroleum demand. Determine SSA refinery assumptions and constraints. Specify cases for 2010, 2015, and 2020. Run the World® Global Downstream Model. 2 S u b-Saharan Africa Refinery Project 1. Summary of the Main Results The Health Study started by investigating the level of air pollutants in a given city for each of the three selected SSA regions (see Figure 1.1) under a Base Case (representing current or recent historical conditions) and two scenarios: Scenario 1 evaluates the benefits of reduced sulfur content in transportation fuels alone, while Scenario 2 assumed reduced sulfur fuels, and vehicles with improved emission controls, development of an inspection and maintenance (I&M) program, and phase-out of 2-stroke engines. The reduction of pollutant emissions associated with Scenario 2 resulted in significant reductions in the predicted rates of mortality and respiratory illness associated with particulate matter, despite the domination of particulate matter emissions by road dust and domestic sources. International experience shows the synergistic effects of reducing air pollutants by combining improved fuels and improved vehicles, an I&M program, and reduction of 2- stroke engines. The estimated health benefits were then associated with estimated monetary benefits. (Figure 1.1). Exhibit 2: Sub-Saharan Africa ­ Selected Regions Source: CITAC Africa LLP E x t ractive Industries for Development Series 3 Table 1: Monetary Valuation of the Health Benefits derived from the improvement of fuels quality Total Annual 5-Year NPV 10-Year NPV Benefit Billion Billion Million $2007 $2007 $2007 SSA West Region Base Case - - - Scenario 1 640 2.6 4.5 Scenario 2 4,500 18 32 SSA East Region Base Case - - - Scenario 1 340 1.4 2.4 Scenario 2 1,300 5.3 9.0 SSA Southern Region Base Case - - - Scenario 1 0 0 0 Scenario 2 252 1.0 1.8 There are significant benefits to be obtained by reducing the causes of air pollution. As the Health study shows, improving the quality of fuels is one of the key steps. The Southern region's lower benefits reflect the fact that South African fuels already meet higher fuel standards, while the West region's higher benefits reflect the higher populations and average incomes in that region. What is also apparent from the table is that while there are benefits from only improving the quality of the transportation fuels, there are significant benefits from the synergistic impact of improved fuels, improved vehicles, an I&M program, and reductions of 2-stroke engines. Most SSA refineries are not equipped to produce ultra low sulfur fuels. The Refinery Study modeled investments and operating costs associated with the improvement of fuels quality for eight different cases reflecting the economic results both in the case of an "open market" in which the SSA refineries would be fully exposed to global competition, and in the case of a "constrained market" in which all existing SSA refineries would be protected and remain in operation. Apart from the Base Case (2010), for each projected year (2015 and 2020) two cases were run: in the first refiners kept current fuel standards and in the second refiners moved to improved fuel standards, namely the ARA recommended AFRI-4 (see Table 1.2). By comparing the two cases in each year the incremental costs of moving to more stringent fuel specifications could be identified. Because of the impact of the current global recession two economic cases were constructed in order to reflect the realities of the global economy: the cases were favorable and unfavorable growth. In the unfavorable case, GDP growth in SSA was lowered and certain other factors that influence the economic health of the SSA refineries were tightened. The details are discussed in the Refinery report. 4 S u b-Saharan Africa Refinery Project Table 1.2: Total SSA Refinery Investments from the Eight Cases (Billions of 2007$) Case Description Year AFRI SSA West Southern East Specifications Total Total Total Total 210 Base 2010 Current 0.06 0.01 0.05 - 215 Open Market 2015 Current 1.89 0.02 0.28 1.60 Unfavorable 216 Open Market 2015 AFRI-4 3.14 0.47 0.54 2.13 Unfavorable 224 Constrained 2020 Current 5.40 3.19 0.60 1.61 220 Constrained 2020 AFRI-4 8.67 5.31 1.00 2.36 221 Open Market 2020 Current 5.32 3.07 0.59 1.66 222 Open Market 2020 AFRI-4 7.65 4.51 0.90 2.25 223 Open Market 2020 AFRI-4 6.19 2.98 0.64 2.56 Unfavorable Table 1.2 shows the total investment costs facing the SSA refining industry in all cases, while Table 1.3 shows the incremental costs of moving from current fuel specifications to AFRI-4 standards in 2015 and 2020. The costs are those facing refineries. The incremental capital costs at port terminals that must expand to cope with increased exports/imports are not included and neither are the incremental costs of consumer distribution infrastructure. Costs arising from increased regulations and the setting up of an I&M program are also not included. Table 1.3: Incremental SSA Refinery Investments from Moving to AFRI-4 Fuel Specifications (Billions of 2007$) Year Description SSA West Southern East Total Total Total Total 2015 Open Market 1.25 0.45 0.26 0.53 2020 Constrained 3.27 2.12 0.40 0.75 2020 Open Market 2.33 1.44 0.31 0.59 The combined benefits estimated in the Health Study, using the Scenario 2 assumptions, and the costs estimated in the Refinery Study, using the open market scenario, Case 222, the most challenging to the SSA refinery sector, are shown in Table 4 as net present value, extrapolated over both a 5-year period and a 10-year period. E x t ractive Industries for Development Series 5 Table 1.4: Five-Year and Ten-Year Net Present Value of Refinery Investment Costs versus Health Benefits Billions 2007 dollars SSA Total West East East Africa Africa Africa 5-Year Refinery Investment Costs $2.76 B $0.47 B $2.13 B $0.59 B Health Benefits over 5 Years 1 $25 B $18 B $5.3 B $1.0 B 10-Year Refinery Investment Costs $6.14 B $4.69 B $2.48 B $0.99 B Health Benefits over 10 Years 1 $43 B $32 B $9.0 B $1.8 B 1. Central value shown for elasticity=1.5; ranges for elasticities of 1.0 and 2.0 are shown in the report. For Scenario 2 (lower sulfur fuel and pollution control equipment) and alternate 2-stroke motorcycle emissions assumptions. These results indicate that, over time, the potential health benefits from Scenario 2 outweigh the costs to the SSA refineries of improving fuel specifications. Regulatory authorities should raise the SSA fuel specifications first to current refineries' typical production specifications, and then, in a concerted way, move to AFRI-4 before the year 2015. While the answer to the cost benefit analysis indicates a clear incentive to implement AFRI-4 standards, and to do so swiftly, the answer to the question about the SSA refiners' ability to function in the competitive global refining market is more complex. The global refining scenario underlying these cases is one of slack refining capacity through much of the period to 2020. Thus the SSA refiners in an open market will be faced with considerable competition from imports from the Middle East and India, regions with growing refinery capacity, and from Europe and the United States, regions with curtailed domestic demand. The implications are that SSA refineries will have to focus on costs and efficiency to function effectively in the open market. There is, if the slackness continues throughout the period, substantial potential for refinery closures in Europe and the United States, which, should they occur, will substantially improve the outlook for the SSA refineries who have invested in upgrading and de-sulphurisation capacity. However, this cannot be relied upon to reduce the potential for large import volumes foreseen in the Refinery Study. An outcome is that SSA refineries that currently have economic problems and do not invest may well be overwhelmed by future intense competition. Among the important assumptions of the modeling exercise is that refiners will be receiving prices for products delivered to domestic markets that reflect opportunity costs. If the refiners are not remunerated at economic levels they will not be able to assume the required investments. In this case the supply costs to the respective markets will be higher. Conversely, allowing SSA refineries to recover market costs for the more advanced AFRI-4 fuels will improve refinery margins and viability. One result of the modeling which may be of importance to policy makers is that the cost of maintaining all the existing SSA refineries through 2020 and requiring 6 S u b-Saharan Africa Refinery Project the investments to produce gasoline and diesel to AFRI-4 fuel specifications was estimated at an incremental $1.02 billion (2007 $) over the free market case. Such an incremental cost will have to be supported by the economies of countries where these refineries are located. SSA refineries that do well in the competitive global market usually: · Are larger and have scale · Have invested over time and are more complex · Are more efficient · Have access to good local quality crude oil, and · Have access to larger markets. However, despite the problems arising from global competition facing the SSA refiners and despite the costs of moving to AFRI-4 standards, those refiners that do invest will see improved refinery margins and increased revenues. The consequent improvement in refinery margins from these investments, based on the model assumptions, can be seen in Exhibits 6-1 through 6-3 in the main Refinery report. What these tables show are that the three crack spreads representing the three typical types of refineries improve markedly in 2015 and 2020 cases, both constrained and open market, where the refineries move to AFRI-4 standards. Crack spreads represent the gross margins achieved by the refineries. The focus of this study was the cost benefit analysis of the impact of clean transportation fuels on human health in the urban areas of SSA. As mentioned earlier there are considerable uncertainties associated with the health data and the global recession has compounded the uncertainties associated with the demand projections. Nevertheless, the health benefits are in line with similar work undertaken in Mexico and China, and are of sufficient magnitude that even if the refinery costs were doubled, on a NPV basis the benefits would still be positive. There are two other aspects of transportation fuels and their emissions to keep in mind, apart from the potential health benefits: · Looking out beyond 2020 the trends that have been apparent in SSA will likely continue, as in other developing nations. Populations will continue to grow, GDP will continue to expand and urbanization and traffic congestion will also likely grow. The net result of this is that in the next few decades vehicle missions will overwhelm pollution from other sources such as road dust in most capital cities of Sub-Saharan Africa. · SSA imports vehicles. Global vehicles, both gasoline-powered and diesel- powered, make use of increasingly sophisticated and effective control E x t ractive Industries for Development Series 7 technology. However, this technology requires clean fuels with increasingly stringent specifications in order to optimally perform. Clean fuels are required for technological reasons as well as for health benefits. Table 1.5 summarizes the major investments proposed by the model for the constrained and open market AFRI-4 2020 cases for the existing refineries. The capacities and dollar Table 1.5: Major Investments Proposed by the WORLD® Model for the 202 Case No. Description Unit All units in MT, unless otherwise specified SGI CAM Revamping Cat ref to CCR DDS to ULSD Cost $ Millions Crude distillation 275,000 90,000 Vacuum distillation 59,000 Debottlenecking Cat cracking Hydrocracking Cost $ Millions 10 220 2020 Crude Distillation AFRI-4 Vacuum distillation 7,000 Constr Cat reforming 55,000 Naptha desulfurization 113,000 FCC gasoline desulfurization 117,000 New Units Distillate desulfurization 1,686,000 312,000 VGO / FCC feed desulfurization Hydrogen Plant Mill (SCFD) 8,641 Sulfur plant (T / day) 30,000 10,000 Aromatics recovery 97,000 26,000 Isomerization 61,000 Cost $ Millions 460 70 Total Investment $ Millions 470 70 Revamping Cat ref to CCR DDS to ULSD Cost $ Millions Crude distillation Debottlenecking Cat cracking Hydrocracking Cost $ Millions 222 2020 Crude Distillation AFRI-4 Vacuum distillation 6,000 Open Cat reforming 69,000 Market Naptha desulfurization New Units FCC gasoline desulfurization 10,000 Distillate desulfurization 1,048,000 190,000 Hydrogen Plant Mill (SCFD) 8,055 Sulfur plant (T / day) 20,000 10,000 Aromatics recovery 104,000 18,000 Isomerization 61,000 Cost $ Millions 350 50 Total Investment $ Millions 350 50 8 S u b-Saharan Africa Refinery Project amounts are shown by the refinery groupings. Similar data for all eight cases can be found in the report. The refinery groupings shown in the table are as follows: SGI: Senegal, Ghana, Cóte d'Ivoire COG: Congo, Gabon ZAM: Zambia CAM: Cameroon ANG: Angola KEN: Kenya NIG: Nigeria SAF: South Africa SUD: Sudan 20 Constrained and Open Market Cases Refinery Groupings NIG COG ANG SAF ZAM KEN SUD 1,918,000 734,000 260 925,000 229,000 199,000 23,000 30 350,000 36,000 999,000 51,000 20,000 23,000 118,000 6,602,000 214,000 362,000 126,,000 290,000 1,753,000 105,000 3,340 393 9,823 1,572 4,715 50,000 10,000 10,000 30,000 30,000 26,000 498,000 2,000 38,000 86,000 1,190 120 90 620 90 100 330 1,190 120 90 910 90 100 330 1,918,000 736,000 260 925,000 199,000 23,000 30 121,000 407,000 23,000 48,000 929,000 104,000 3,643,000 46,000 3,092,000 128,000 1,791,000 9,627 1,375 5,305 50,000 10,000 30,000 4,000 512,000 2,000 87,000 620 10 600 90 340 620 10 600 90 340 E x t ractive Industries for Development Series 9 2. Health Study Overview The Health Study evaluated a Base Case, Scenario 1 (reduced sulfur in transportation fuels), and Scenario 2 (reduced fuel sulfur, improved emission controls, and increase in vehicle activity), with the following steps: · Air Quality Modeling - Select modeling locations (representative cities in Sub-Saharan Africa) - Select air quality model - Select pollutants and emission sources to be modeled · Health Impact Assessment - Select health endpoints - Identify appropriate studies to evaluate air pollution/health relationship - Compile baseline health data - Calculate the reduction in mortality or cases of disease · Valuation of Health Impacts - Determine the appropriate approach for valuation in Sub-Saharan Africa - Identify relevant studies and assess quality and applicability - Estimate the value of mortality risk reductions - Estimate the value of reductions in the risks of chronic bronchitis and exacerbation of asthma symptoms Each step is associated with important uncertainties that are summarized below and discussed in detail in the report. Air Quality Modeling To model the air quality for SSA, based primarily on the availability of emissions data, three representative cities from each of the selected regions of SSA (West, East and South - see Figure 2) were selected for quantitative analysis. Two additional cities in each region were selected for qualitative analysis (Table 2.1). Quantitative analysis involved compilation of emissions inventories and an air quality model to estimate ambient air quality resulting from local emissions of air pollutants. Emissions inventories for vehicle, area, and point sources were compiled. Emissions data included variables such as percent of vehicle kilometers traveled (VKT) by various types of vehicles, VKT on paved and unpaved roads, dust emissions factors, and industrial mix. Qualitative analysis compared local conditions (including population, total emissions, industrial mix, vehicle fleet composition, meteorology, etc.) to conditions of the quantitatively analyzed city in the same region. 10 S u b-Saharan Africa Refinery Project Table 2.1: Cities Modeled in the Health Study Assessment Description SSA Total West Total Quantitative Cotonou, Benin Johannesburg, South Africa Kampala, Uganda Qualitative Lagos, Nigeria Cape Town, South Africa Dar Es Salaam, Tanzania Qualitative Ougadougou, Burkina Faso Maputo, Mozambique Nairobi, Kenya Modeled air pollutants were those associated with potential health impacts - particulate matter (PM), sulfur dioxide (SO2), nitrogen dioxide (NOS), and benzene (Table 2.2). Table 2.2: Air Pollutants and Related Health Effects Air Pollutant Related Health Effects Particulate matter, both PM2.5 and PM10 1 Premature mortality; chronic bronchitis; asthma symptoms Sulfur dioxide (SO2) Increased incidence of asthma attacks Nitrogen dioxide (NO2) Increased mortality and asthma attacks Benzene Increased incidence of cancer Ambient air concentrations were estimated for areas in each modeled city with population density greater than 1,000 people per square kilometer, to permit comparison between cities. The modeled regional annual average total PM (primary and secondary PM, plus sulfate and nitrate) concentrations are shown in Table 2.3. The concentrations shown in the table illustrate the change in annual average total PM2.5 and PM10 concentrations from Base Case to Scenarios 1 and 2. The results for other air pollutants are shown in the report. Although the modeled Base Case air concentrations for Kampala and Cotonou appear somewhat high relative to the limited monitoring data, the focus of the study is on the relative change between scenarios, rather than the absolute number. In Johannesburg, the Base Case concentrations and the relative changes are small because baseline sulfur content in fuels is already low, there are relatively few diesel vehicles in use, and domestic use of wood as a fuel is low. Table 2.3: Modeled Annual Average Total Particulate Matter Concentrations for Three Cities (population density greater than 1,000/km) Kampala, Uganda Cotonou, Benin Johannesburg, South Africa Annual Averages Total Total Total Total Total Total (µg/m3) PM2.5 PM10 PM2.5 PM10 PM2.5 PM10 Base Case 224 371 363 567 4.6 19.7 Scenario 1 222 369 359 563 4.6 19.7 Scenario 2 219 366 275 469 4.1 19.3 E x t ractive Industries for Development Series 11 The emissions results indicate that the dominant PM emissions are from road dust and domestic sources; still, the relative contribution of vehicles to PM is reduced with each scenario as shown graphically in Figure 2.1 for PM2.5. In addition, there is a clear reduction in benzene emissions, in particular with Scenario 2. As will be described, these incremental changes translate into health benefits. Figure 2.1: Modeled Reductions of PM2.5 (tons/year) for Vehicle Sources in Three Cities Estimated Reduction in Tons/Year of PM2.5 for Vehicle Sources 7000 6000 Kampala, Uganda 5000 Cotonou, Benin 4000 Johannesburg, 3000 RSA 2000 1000 0 Base Case Scenerio 1 Scenerio 2 Kampala, Uganda 556 460 198 Cotonou, Benin 3.591 3,470 629 Johannesburg, RSA 5,797 5,797 1,873 The large reduction in percentage of PM emissions from the Base Case to Scenario 2 in Cotonou, Benin (from 22.9% to 5% for PM10 and from 33.8% to 8.2% for PM2.5, as shown in Table 2.2) is based on vehicle activity data used for the Base Case that shows a large use of 2-stroke motorcycles in that city2 and the assumption that in Scenario 2, all 2-stroke motorcycles are banned. This large reduction was initially used to derive benefit estimates for the West region. The data for the East region city, Kampala, Uganda, indicated a lower current usage of 2-stroke motorcycles. Thus, initially, higher benefits were estimated for the West region (based on the Cotonou data) as compared to the East region (based on the Kampala data). The ARA Steering Committee expressed concern that the Health Study emissions inventory under-represented motorcycle emissions in Kampala, Uganda (extrapolated to the East Africa region) and over-represented these emissions in 1 PM2.5 is particulate matter of 2.5 microns in diameter or less; PM10 is particulate matter of 10 microns in diameter or less. 2 Clean Air Initiative, Banque Mondiale. Benin. Ministère de l'environnement de la protection de la nature (MEPN). Etude de la qualité de l'air à Cotonou. Rapport Final, November 2007. (Translation from French to English.) 12 S u b-Saharan Africa Refinery Project Cotonou, Benin (extrapolated to the West Africa region) for the Base Case and Scenario 1.3 These concerns were based on anecdotal observations that: (1) Motorcycle use has increased in recent years in East Africa, and (2) Motorcycle use may have decreased in West Africa because of bans on the use of 2-stroke motorcycles in some areas. Therefore, regionalizing the Cotonou analysis may be problematic because some large cities in West Africa, particularly in some states of Nigeria, have banned 2-stroke engines. While 2-stroke engines have not entirely disappeared from those cities, their prevalence has decreased so using Cotonou's inventory for these cities may not be accurate. Emissions from motorcycles, as well as other motor vehicles, are a function of the total vehicle activity that is kilometers driven. While the number of vehicle types registered is useful when vehicle activity data are not available, emissions are not directly scalable with the number of vehicles. For example, buses make up a small fraction of the number of vehicles in most cities, but they are operated nearly continuously, meaning that the emissions from buses are not in proportion to their number. Therefore, whenever available, travel fraction data, which is nearly proportional to vehicle emissions, were used as a basis for quantifying vehicle activity. Motorcycle travel fractions in Cotonou and Kampala were based on published vehicle activity data. The air modeling results were scaled using the following assumptions: Alternate 2-Stroke Motorcycle Assumptions for the Base Case and Scenario 1: · West Region: VKT from 2-stroke motorcycles was reduced from 50% to 10% · East Region: VKT from 2-stroke motorcycles was increased from 20% to 50% · South Region: no changes Health Impact Assessment Health impacts are assessed by: · Selecting health endpoints4 associated with the air pollutants of concern · Identifying appropriate studies to evaluate air pollution/health relationships · Compiling baseline health data · Calculating the reduction in mortality or cases of disease associated with improved fuels Health endpoints mean the health impact of concern, such as respiratory illness or mortality. 4 E x t ractive Industries for Development Series 13 Exposure-response data from published studies quantifies the relationship between exposure to air pollutants (exposure) and health impacts (response). The most supportable studies, with high-quality exposure and health response data, were selected. However, the selected studies are from U.S. locations with varying demographics. There are uncertainties associated with extrapolating the U.S. air pollution/health impact studies to Sub-Saharan Africa, although World Bank studies have previously concluded that, under certain circumstances, there is support for extrapolating results from cities in developed countries to cities in developing countries. Literature regarding air pollution/health impact data from Africa is mainly prevalence data, useful for determining baseline rates of the relevant diseases, but providing limited exposure measurements. Other required baseline health data were collected from international organizations and databases. Valuation of Health Impacts Estimating the monetary value of the selected health risk reductions for this study involved four steps: · Determining the appropriate approach for valuation · Identifying relevant studies and assessing their quality and applicability · Estimating the value of mortality risk reductions · Estimating the value of reductions in the risks of chronic bronchitis and increased asthma symptoms Previous studies conducted by international and national organizations have used a variety of approaches to value health risk reductions. These diverse approaches in part reflect differences in the goals of the studies, and in part reflect the limitations of the then-available research. The three approaches most often used are called willingness-to-pay, cost-of-illness, and quality-adjusted life years or disability- adjusted life-years. The preferred approach to valuation in benefit-cost analysis is to rely on estimates of individual willingness-to-pay (WTP), which describe individual preferences; i.e., the affected individuals' willingness to exchange their own income for reductions in their own risks, for example, the risk of mortality or illness. The value of the mortality risk reductions associated with pollution abatement and other policies is commonly expressed as the value per statistical life (VSL). The VSL concept is frequently misunderstood -- it is not the value of saving a "life," nor is it a measure of the moral worth or inherent value of an individual. The WTP approach is consistent with the framework for pollution abatement decisions; e.g., deciding whether to allocate funds for further upgrades to refinery operations or for other projects. The monetary benefit values and the results of the benefit-cost analysis more generally, are only some of the many factors typically weighed by decision makers. Additional considerations include, for example, the distribution of the potential 14 S u b-Saharan Africa Refinery Project health impacts across different countries and subpopulations, the potential magnitude of health effects that are not quantified, and the substantial uncertainties in the valuation estimates. In addition, decision makers may consider the equity or fairness of different approaches, or may have goals or constraints that will lead to outcomes not entirely consistent with the preferences of those affected. Decision makers will need to exercise judgment in weighing the substantial uncertainty associated with these values, as well as the inability to quantify some of the effects of air pollutants. The key conclusions of the health valuation are as follows. General Framework · Although there are several current approaches, in benefit-cost analysis, the value of risk reductions is determined by estimating the affected individuals' own willingness to exchange income for the resulting benefits, so that the values reflect their preferences for spending on improved health rather than on other goods and services such as food and shelter. · Spending for risk reductions is more constrained in Sub-Saharan Africa than in many other countries, because of the relatively low average incomes in this region. Average per capita income, based on purchasing power parity, is $1,900 annually in Sub-Saharan Africa. In contrast, per capita income averages about $46,000 annually in the U.S. Significant income disparities underline these averages both across and within countries. · Estimates of willingness to pay for risk reductions are very uncertain for residents of Sub-Saharan Africa because little is known about their preferences. Instead, values are typically extrapolated from studies conducted in much wealthier countries, with adjustment for income differences. · Variation in income across SSA countries leads to large differences in the VSL estimates, because individuals' willingness to pay for risk reductions is dependent on the amount of money they have available. · Income is only one of many factors affecting the value of risk reductions. These values may also be influenced by the characteristics of the affected populations (e.g., their age and health status) and of the risks themselves (e.g., whether they involve illness or injury or are incurred involuntarily). As a result, values will vary across individuals, communities, countries, and regions for many reasons. · The valuation section of the Health Study recommends the use of ranges to reflect uncertainty. To reduce this uncertainty, more research is needed that directly elicits or reveals the preferences of individuals residing in these countries for reductions in these types of risks. The relatively wide ranges of benefit values indicate the substantial uncertainty in the VSL and VSC estimates appropriate for countries in this region, mostly related to the lack of studies conducted among the populations of concern for this report. E x t ractive Industries for Development Series 15 Extrapolating from existing studies requires understanding the effects of large income differences, and we are uncertain about the appropriate income elasticity. In addition, differences in other population characteristics and in the characteristics of the risks themselves may lead to higher or lower values. Cultural attitudes towards risks, the age of those affected, and the quality of available medical treatment will affect WTP for health risk reductions. These factors are best addressed by conducting studies in the countries of concern. Additional considerations, such as the equity of the distribution of wealth and health and non- quantified impacts, should be considered separately. In addition to income ­ measured by the DDP per capita -, there are a number of other differences between the available research and the mortality risks associated with air pollution in Sub-Saharan Africa. These include population characteristics such as age, life expectancy, health status, and total mortality risk.5 The study used for mortality valuation focuses on job-related deaths from accidental injuries, while air pollution leads to deaths from illness. However, the available research is not sufficient to support quantitative adjustment of the VSL to reflect these differences, especially given that these characteristics may be viewed differently across countries and cultures. The valuation estimates cover a wide range due to the significant differences in income ­ measured by GDP per capita- across these countries. These values are only one of many factors that must be considered in related decisions. They do not incorporate concerns about the equity of the distribution of income and health. Regional Analysis The estimated health benefits for each quantitatively evaluated city must be extrapolated to a regional level, in order to compare the benefits with the costs estimated in the Refinery Study. The regional extrapolation was accomplished by, first, identifying all large cities in each region, defined as those with a population greater than 300,000. Then, the estimated decreases in the numbers of deaths or cases of bronchitis or asthma in each modeled city were used to estimate the decreases for the population of each large city in the region. Finally, the estimated monetary benefit values for each country were used to estimate the regional monetary benefits of a reduction in emissions of air pollutants.6 Scaling by population and income level assumes that the scaled cities have the same reductions in emissions (including the same vehicle mix, industry mix, etc.), and the same population characteristics (e.g., the same baseline rate of disease). This is certainly an over-simplification; however, it provides an estimate of the potential impact of fuel improvements for each Sub-Saharan African region. Due to the limitations of the available research, this report recommends using the same values for adults and children, while noting that mortality 5 risks to children are generally valued more highly. We do not adjust the valuation estimates for income differences within each country. 6 16 S u b-Saharan Africa Refinery Project Health Study Results For each region, the estimated reductions are presented in a number of cases, and the associated annual, five-year, and ten-year estimated benefits7 (Table 2.4). As described above the estimated West region benefits were initially much higher than those estimated for the East region, due to the emissions assumptions used for the percentage of total vehicle activity attributable to 2-stroke motorcycles. In addition, differences in annual benefit estimates across regions result from differences in the total populations in each region and the average income for each region. The estimated West region benefits were initially higher than those estimated for the East region, because of these differences in the characteristics of the regions (Table 2.5): Table 2.4: Major Differences between SSA Regions Reflected in Modeling Results West East South Motorcycle Emissions 2-Stroke Motorcycles: 48.4% All Motorcycles: 20% of No 2-stroke motorcycles for Modeled City of total vehicle activity in total vehicle activity in currently in use.10 Cotonou, Benin8 reduced to Kampala, Uganda9 reduced zero in Scenario 2. to zero in Scenario 2. Regional Population 16.8 million 7.5 million 5.2 million Greater than Age 30 GNI1 $1,836 US $961 US $6.980 Note: 1.West region: $1,836 is average of GNI for the 20 countries used in the analysis; East region: $961 is average of GNI for the 14 countries used in the analysis; South region: $6,980 is average of GNI for 2 countries (Angola and South Africa) used in the analysis The results shown in Table 2.5 are presented using the alternate assumptions; the benefits for the East region increase from the initial estimates and the benefits for the West region decrease from the initial estimates, bringing the results closer for the two regions, although, still, because of the higher population and reported average income levels for the West region, the West region benefits are higher. The benefits calculations for the South region do not change for the Base Case or Scenario 1, as there are already lower sulfur fuels in use in that region. 7 Calculated as net present value (NPV) with a 7% discount rate. 8 Clean Air Initiative, Banque Mondiale. Benin. Ministère de l'environnement de la protection de la nature (MEPN). Etude de la qualité de l'air à Cotonou. Rapport Final, November 2007. (Translation from French to English.) 9 Stuck in Traffic: Urban Transport in Africa. Africa Infrastructure Country Diagnostic Study. Kumar A; Barrett F. October 31, 2007. 10 Data provided by the City of Johannesburg. E x t ractive Industries for Development Series 17 Table 2.5: Health Study Regional Results Using Alternate 2- Stroke Motorcycle Assumptions -All-Cause Bronchitis Asthma Respiratory Estimated Estimated Ten- Mortality - Cases - Exacerba-tions Disease Estimated Total Five-Year Year (NPV) Reductions Reductions - Reductions in Mortality - Annual Benefit (NPV) Benefit Benefit (billion - Number in PM2.5 in PM2.5 PM10 Reductions in (million 2007 (billion 2007 2007 US of Cases of: (ages>30) (ages>30) (ages 8-13) PM10 (ages<5) US dollars) US dollars) dollars) East Region Base Case 541,823 2,293,281 3,183,871 3,183,871 -- -- -- Scenario 1 532,449 2,247,042 3,179,640 3,179,640 $340 M $1.4 B $2.4 B Scenario 2 506,516 2,118,796 3,167,322 3,167,322 $1,300 M $5.3 B $9.0 B West Region Base Case 1,325,880 7,387,399 7,464,977 69,140 -- -- -- Scenario 1 1,316,489 7,334,697 7,462,041 69,006 $640 M $2.6 B $4.5 B Scenario 2 1,249,046 6,952,170 7,444,698 68,238 $4,500 M $18 B $32 B South Region Base Case 3,636 108,256 799,209 970 -- -- -- Scenario 1 3,636 108,256 799,209 970 $0 M $0 B $0 B Scenario 2 3,261 107,576 796,852 948 $252 M $1.0 B $1.8 B Notes: Using the air model results for >1,000 population density with each city, extrapolated to the region. 1. Applying country-specific VSL or VSC, elasticity of 1.5, and GNI using PPP. 2. Net present values calculated with 7% discount rate. The dollar values presented represent the middle estimate of a range from low estimated values (using an income elasticity of 2.0) to high estimated values (using an elasticity of 1.0). The range of estimates is shown in the study report. The estimated 10-year benefits of reduced sulfur fuels modeled in Scenario 2 for all of SSA (about $43 billion) are similar to the benefits shown in other regional studies. For example, the benefits of reducing health impacts in China, including total mortality and chronic bronchitis, modeled for the years 2008-2030, are about $45 billion (in 2005 US dollars) for total mortality and $10 billion (in 2005 US dollars) for chronic bronchitis.11 A similar study conducted in Mexico predicted benefits of about $40 billion (in 2000 US dollars) modeled for the years 2006-203012. The decrease in cancer risk due to potential reductions in the exposures to benzene is low in both Johannesburg and in Kampala (less than 150 cases). The larger 11 The International Council on Clean Transportation (ICCT), 2006. Costs and Benefits of Reduced Sulfur Fuels in China. Cited in presentation by Ray Minjares of ICCT, Costs and Benefits of Lower Sulfur Fuels; Implications for Eastern Africa, Eastern Africa Sub-Regional Workshop on Better Air Quality in Cities, Nairobi, Kenya, 21-22 Oct 2008. 12 Estudio de Evaluación Socioeconómica del Proyecto Integral de Combustibles, Instituto Naciónal de Ecología. 2006. Cited in presentation by Ray Minjares of ICCT, Costs and Benefits of Lower Sulfur Fuels; Implications for Eastern Africa, Eastern Africa Sub-Regional Workshop on Better Air Quality in Cities, Nairobi, Kenya, 21-22 Oct 2008. 18 S u b-Saharan Africa Refinery Project reductions in benzene modeled for Cotonou result in a potential reduction in cancer cases due to benzene exposure of up to 250 cases in Scenario 2. Because these reductions are small relative to the other benefits estimated, the reduction of cancer risk due to benzene exposure is not included in the valuation or in the regional extrapolation. 3. Refinery Study Overview Over the past two decades, the growing awareness of the role that emissions play in human health and environmental degradation had led to a general movement in many parts of the world to control emissions to reduce the impacts. This movement has mainly taken two forms: 1) the development and subsequent required use of control devices for stationary sources and vehicle sources and, 2) changes in the specifications of transportation fuels to reduce emissions of the major pollutants. These trends originated in the industrialized countries and are now spreading, at different rates, throughout the world. As in other world regions, the first improvement in the specifications of transportation fuels in Sub-Saharan Africa was the elimination of lead. The phase- out of lead is now complete and the World Bank and its partners are looking at the next step - the reduction of sulfur in transportation fuels. The ARA has taken the lead in presenting "fuel specification bands" known as the AFRI Specifications modeled after the specifications now in force in the European Union (EU). Table 3.1 following shows the main AFRI parameters for both gasoline and diesel. The assumption was made in the study that by 2020 the AFRI-4 level would be reached Table 3.1: AFRI Standards for Sub-Saharan Africa Transportation Fuels AFRI-1 AFRI-2 AFRI-3 AFRI-4 GASOLINE RON, min* 91 91 91 91 MON, min 81 81 81 81 Lead content** Unleaded Unleaded Unleaded Unleaded Sulfur content, % mass, max 0.1 0.05 0.03 0.015 Benzene content, % vol, max To be reported To be reported 5 1 DIESEL Sulfur content, % mass, max 0.8 0.35 0.05 0.005 Density at 15 C, kg/litre (min/max) 800/890 800/890 800/890 820/880 Cetane index (calculated), min 42 45 45 45 Lubricity (HFRR @ 60 C), micron, min To be reported To be reported 460 460 *A higher grade of gasoline may be marketed if required ** "Unleaded means <0.013g of lead per litre 11 The International Council on Clean Transportation (ICCT), 2006. Costs and Benefits of Reduced Sulfur Fuels in China. Cited in presentation by Ray Minjares of ICCT, Costs and Benefits of Lower Sulfur Fuels; Implications for Eastern Africa, Eastern Africa Sub-Regional Workshop on Better Air Quality in Cities, Nairobi, Kenya, 21-22 Oct 2008. 12 Estudio de Evaluación Socioeconómica del Proyecto Integral de Combustibles, Instituto Naciónal de Ecología. 2006. Cited in presentation by Ray Minjares of ICCT, Costs and Benefits of Lower Sulfur Fuels; Implications for Eastern Africa, Eastern Africa Sub-Regional Workshop on Better Air Quality in Cities, Nairobi, Kenya, 21-22 Oct 2008. E x t ractive Industries for Development Series 19 by all SSA refineries (sulfur content for gasoline 150 ppm, for diesel 50 ppm, and benzene content for gasoline 1%). The growing complexity of the vehicle emission control technologies for both personal vehicles and commercial trucks and the concomitant need for clean fuels, in addition to the growing awareness of the human health and environmental impact of vehicle source emissions, have placed increasing requirements on refineries. Sulfur is not an additive but a natural part of crude oil. Its removal processes presents both technological and economic challenges to refiners. However, by coming later than OECD regions to ultra-low sulfur fuels, SSA refineries are in a position to benefit from the operating experience and process improvements obtained elsewhere in the refining industry. Refinery Study Methodology The methodology used by ICF and its subcontractor EnSys Energy, centered around the use of a global refining model WORLD®, which allowed us to not only calculate the impacts of stricter fuel specifications on the SSA refiners but also set them in the context of the global refining and downstream system. With agreement from the Steering Committee, ICF/EnSys modeled refining activities and investment costs for the global refining sector with horizons of 2010, 2015, and 2020. Modeling 2010 (Base Case) enabled calibration against conditions close to those of 2009. Stepping the model forward to 2015 and then 2020 captures the evolution of product demand and mix in the regions and the advance to AFRI-4 specifications in each refinery subgroup, with the attendant costs and impacts. In order to model the various cases that were developed the following data had to be generated: · Annual demand for petroleum products for every SSA country out to 2020, as well as supply · Base Year configuration of each SSA refinery with information on technologies used, capacities and the type of crude oil and other feedstocks used · Expansion projects and other refinery assumptions · Delineation of the cases for 2015 and 2020 Demand Projections Energy demand within an individual country is largely a function of gross domestic product (GDP), population, and the energy intensity/efficiency of the overall economy and specific sectors within the economy. Demand is also impacted by Government policies that influence the exploitation of domestic natural resources, control imports and/or exports, affect prices in country, and impact the efficiency of energy end-use technologies. Further impact comes from 20 S u b-Saharan Africa Refinery Project the state of the domestic infrastructure and the percentage of GDP derived from the export of raw commodities13. Using prices to estimate demand is complicated. In an unfettered market, the market clearing price would determine the level of demand by consumers in the different economic sectors. While this may be the most rigorous approach to estimating demand, substantial amounts of data are required. There is also a good deal of dissension over the correct elasticities to use. Therefore, initially, ICF concluded that an accepted publicly available projection would be the best approach. ICF has been unable to find publicly available substantive projections of energy demand by country in SSA. In the various projections, such as those from OPEC, one can find Angola and Nigeria, the major oil producers identified. Other projections will identify South Africa, usually broken out because of the size of its GDP. ICF was unable to find a projection that addressed the major countries in SSA, let alone the smaller countries. A further complication is that very few countries in SSA, with the exception of South Africa, have a mixed industrial economy - which works best with the normal analytical approach. Many of the SSA countries have very rapidly growing GDPs driven largely by resource exports during a period when global commodity prices were at an all-time high and this does not translate directly into petroleum demand. ICF has therefore developed its own methodology to project petroleum demand by country out through 2020. Total projected petroleum demand for Africa was based on the projections from the Energy Information Administration tied to the price projection discussed above. North African demand was subtracted from the totals. The primary source of data for the demand model is the CITAC Africa LLP data base which provides total consumption from 2000 to 2007 for eight petroleum products and the International Energy Agency's Energy Balances which provides data from 1992 to 2007. The latter is used to extrapolate CITAC data back to 1992 and to fill in certain items missing from the CITAC database such as refinery fuel use. Regressions were then run on the consumption time series to give the trends by product and by sector and to relate the trends to population and GDP. Given projected population and GDP from the World Bank, the United Nations, the IMF, and the U.S. Census Bureau, trends from the regressions can be then superimposed on the SSA total demand to give estimates of individual country level demand. A further step was the discounting of the GDP effect in those countries whose GDP was driven largely by raw commodity exports rather than by domestic goods and services. SSA Refineries With the agreement of the Steering Committee three large demand regions were 13 There are, of course, impacts from random events that cannot be modeled looking out to the future. These can include civil wars, and extreme climate impacts among others. E x t ractive Industries for Development Series 21 created for SSA: West, Southern, and East. These regions reflected the logistics of petroleum product supply to meet demand within SSA. Then, within these three regions, nine refinery supply sub-regions were created. These nine sub-regions varied from individual refineries, e.g., Cameroon, to groups of refineries, e.g., all the Nigerian refineries. The world outside SSA was formulated in WORLD® as 20 demand regions containing a total of 23 refining groups. Explicit refining groups were incorporated for Yanbu, Saudi Arabia and for Reliance/Essar in western India. This formulation allowed the modeling to represent the global context, the SSA context, the 3-region context, and the sub-regional context. The groupings used in the model for SSA are shown in the table and map of Figure 3.1 below. Figure 3.1: Sub-Saharan Africa Refinery Regional Groupings SGI Senegal, Ghana and Cóte d'lvoire CAM Cameroon NIG Nigeria COG Congo and Gabon ANG Angola SAF South Africa ZAM Zambia KEN Kenya SUD Sudan Source: CITAC Africa LLP 22 S u b-Saharan Africa Refinery Project Refining Assumptions One goal of the modeling was to assess the investment requirements facing the individual SSA refineries if they are to move along the AFRI standards. It became clear during the Kick Off meeting that, at least as far as gasoline and diesel are concerned, there are three sets of specifications: "Official specifications" (maximum specifications allowed by the government); and "import specifications". In countries without refineries import specifications may be the same as official specifications, but in other countries with refineries governments may specify/allow a different specification with a view to blending with local production. There is a vast difference between the official specifications and the estimated actual qualities, particularly for gasoline. Knowledge of the actual qualities is critical for the modeling. If, in fact, SSA refineries are already making, for example, gasoline at the AFRI-3 level, that changes the whole investment picture. Thus, although it was initially thought costs would be based on moving from mainly AFRI-1 to AFRI-4, based on the current specifications identified, the costs are more likely related to moving from AFRI-2 and -3 to AFRI-4. The advance of regional specifications, demand and supply, as well as the influences of broader global developments, impact the investments made in the model in both existing refineries and for potential new refining centers. Within existing refineries, the investments potentially range from the relatively inexpensive, such as revamp or debottlenecking of existing units, to high cost investment decisions such as large expansions or installation of a major secondary unit such as a hydrocracker. Wholly new refining capacity can range from a new moderate scale local refinery to large and complex new refining centers. Stepping forward to 2015 and then 2020, with associated product quality, demand, and related developments, enables ICF/EnSys and the ARA Steering Committee to examine the economics of the different refinery subgroups and to clarify whether or not they can make the investments necessary to meet advancing fuel specifications and still remain profitable and competitive on the world market. During the steering committee meetings the issue of placing constraints on the model was discussed several times as it is not realistic to allow the model to optimize regardless of the situation on the ground in terms of logistics, politics etc. A simple example of this is that it is highly unlikely that a major crude producer like Nigeria will ever import non-Nigerian crude oils. For this reason, an input was specified blocking such an operation. Table 3.2 below summarizes the main constraints. Modeling Cases In order to illustrate the various options and to generate a range of costs to evaluate the impact of the various fuel changes, ICF/EnSys developed a series of cases, as E x t ractive Industries for Development Series 23 Table 3.2: WORLD® Model Constraints Activity Constraint Refinery capacity Capacities established for 2008 are used in the study. The model adds some expansion through revamping and debottlenecking allowed for 2010 and major new investments in later cases. Refinery projects Following the model protocol, only new projects that are under construction were allowed, with the exception of the assumed Ugandan refinery (see below). None of the planned projects in SSA refineries were added to the SSA database. Where economic the model adds capacity in the future cases Refinery activity Base Case operating rates are based on the 2007 operating rates. For the Constrained Cases all existing refineries continue to operate at Base Case efficiencies. For the Open Market Cases an upper cap has been placed on the throughput at Nigerian refineries. Nigerian refineries were capped at 43 % as their average 7-year utilization between 2001 and 2007 was 32% with the highest level being 47% in 2001. A lower floor of 29% was placed on the Indeni refinery in Zambia New refinery Africa East: a new 50 mb/d refinery is entered for all 2015 and 2020 cases. capacity (This was set to run emerging Uganda crude oil.) Africa West: a cap of 200 mb/d is imposed in all 2015 and 2020 cases to reflect the financing limits expected to prevail Process units In Nigeria it was assumed that the alkylation units would not run at any time In the 2020 Constrained Cases SSA refineries are required to use the same Crude oil slates crude oil slates as in the Base Case. In the Open Market Cases some constraints were imposed such as limiting Nigerian and Angolan refineries to the use of their indigenous crude oils only. Trade In all cases the export of high sulfur distillate is not allowed. Trade in most intermediate products is constrained particularly where storage capacities would not allow the export/import of economic cargo sizes. Exports of benzene are allowed but imports are constrained. Imports of MTBE are constrained but some imports of ethanol are allowed. Infrastructure The costs of adding infrastructure capacity at ports and in the consumer distribution system (required for increased imports) were not included in the model: incremental capacity and costs for all refinery offsites was included. shown in Table 3.3. This allowed the examination of a number of possible policy options, in particular, what would happen to SSA refineries and costs in an open market in which they were fully exposed to global competition versus what would happen if all existing SSA refineries were protected (the constrained case). From these cases, ICF/EnSys was able to identify the incremental costs of moving to the AFRI-4 standards, and to distinguish these costs from the costs of investment driven by normal organic growth. Further, the cost differentials between the open market cases and the constrained cases were identified. 24 S u b-Saharan Africa Refinery Project Data Uncertainties There are considerable uncertainties associated with the demand data which is one of the main drivers in the model. While the methodology is analytically reasonable, Table 3.3: Modeling Case Descriptions Case Description Case No. 2010 2015 Economic 2015/2020 2015/2020 Conditions AFRI- AFRI-4 in SSA Current 2010 Base Case 210 X X 2015 Open Market 215 X Unfavorable X Case. SSA Refineries in open competition 2015 Open Market 216 X Unfavorable X Case. SSA Refineries in open competition 2020 Base Case: SSA 224 Favorable X Refineries kept running: 2020 Base Case ­ SSA 220 Favorable X Refineries kept running: 2020 Open Market 221 Favorable X Case: SSA Refineries in open competition: 2020 Open Market 222 Favorable X Case: SSA Refineries in open competition: 2020 Open Market 223 Unfavorable X Case: SSA Refineries in open competition: there may be countries in SSA that are at a threshold for economic takeoff and this methodology would not indicate this. In addition, there is the issue of "pent up" demand. If consumers in a specific country are faced with government regulations that limit the availability of certain products they may consume less than they wish. Likewise if the infrastructure of a country is poor enough to impede distribution consumers will consume less than they demand. If, at some point in the future, the impediment to demand is lifted, whether by revising government regulations or by improving the distribution infrastructure, there may be a surge in actual consumption as demands are met. This will have to be evaluated by expert E x t ractive Industries for Development Series 25 judgment. The complications of the demand estimates are fully discussed in Appendix A in Volume II-B, the Refinery Study Appendices. Apart from the specific problems in generating demand projections, this study was faced with the overarching problem of the global recession. Since this was a "moving target", the ICF (with the concurrence of the World Bank), took the approach of deciding on a projection that was lower than historic trend indicated. These cases are the ones indicated in the report as having economic conditions favorable to SSA. However, to deal with the continuing fall in demand an alternative case was constructed with even lower demand; these cases were labeled as unfavorable. In these latter cases, the average annual regional economic growth originally projected in the base case as 4% was scaled down to 2% per year. To some extent these levels are arbitrary. However, all international forecasts indicate the consensus belief that the developing nations will recover in a more timely fashion than the developed world and will then proceed to start growing again. Refinery Study Results The model runs using the constraints listed in Table 3.2 above resulted in refinery throughputs as a percentage of nameplate capacity as follows: It can be seen that the refineries in the SGI (Senegal, Côte d'Ivoire, and Ghana), Cameroon, South Africa and Sudan operate at high levels in most cases. Nigeria has been capped at the typical throughput rate achieved in recent years (2001/07 average =32%) and Zambia has a floor of 29%. But in most cases for the small refineries with little or no upgrading and limited local demand the model forecasts low or zero throughput. Table 3.4: Refinery Throughputs of Regional Groupings in the Various Cases Case Year AFRI REFINERY GROUPINGS Standards FAVORABLE CASES ­ SGI CAM NIG1 COG ANG SAF ZAM1 KEN SUD CONSTRAINED 210 2010 Current 85% 91% 39% 75% 70% 90% 31% 37% 85% FAVORABLE CASES ­ OPEN MARKET 215 2015 Current 85% 91% 39% 0% 7% 90% 34% 12% 85% 216 2015 AFRI-4 62% 83% 39% 0% 5% 86% 29% 0% 85% 221 2020 Current 85% 91% 43% 49% 36% 90% 31% 7% 85% 222 2020 AFRI-4 85% 91% 43% 0% 26% 90% 29% 0% 86% UNFAVORABLE CASE ­OPEN MARKET 223 2020 AFRI-4 49% 55% 43% 0% 17% 90% 29% 0% 85% 26 S u b-Saharan Africa Refinery Project Table 3.5 summarizes the total investment costs for the various cases facing the SSA refining sector to meet both normal growing demand and the requirement of moving up the AFRI scale to AFRI-4 fuel specifications. Table 3.6 separates out the investment costs of moving to AFRI-4 specifications compared to the current AFRI specifications. This table also shows the incremental cost of moving to AFRI-4 in a constrained versus an open market and the incremental difference in moving to AFRI-4 for a favorable economic situation versus an unfavorable situation. Table 3.5: Total Refinery Investments from the Eight Cases (Billions of 2007$) Case Description Year AFRI SSA West South East Specifications Total Total Total Total 210 Base 2010 Current 0.06 0.01 0.05 - 215 Open Market 2015 Current 1.89 0.02 0.28 1.60 216 Open Market 2015 AFRI-4 3.14 0.47 0.54 2.13 224 Constrained 2020 Current 5.40 3.19 0.60 1.61 220 Constrained 2020 AFRI-4 8.67 5.31 1.00 2.36 221 Open Market 2020 Current 5.32 3.07 0.59 1.66 222 Open Market 2020 AFRI-4 7.65 4.51 0.90 2.25 223 Open Market 2020 AFRI-4 6.19 2.98 0.64 2.56 Unfavorable The modeling undertaken in this study examined the incremental costs of SSA refineries moving to AFRI-4 standards only. We are aware that the target in South Africa was EUROIV and EUROV. The specifications of these standards are more stringent than those in AFRI-4. Thus the costs that were generated for South African refineries, based on AFRI-4, substantially underestimate the costs when applied to the two EURO standards. Table 3.6: Incremental Costs to Move to AFRI-4 Fuel Specifications (Billions of 2007$) Year Description SSA West South East Total Total Total Total 2015 Open Market 1.25 0.45 0.26 0.53 2020 Constrained 3.27 2.12 0.40 0.75 2020 Open Market 2.33 1.44 0.31 0.59 2020 Constrained vs. Open Market 1.02 0.80 0.10 0.11 2020 Open Market Favorable vs. Unfavorable 1.46 1.53 0.26 (0.69) This Table shows that the incremental cost for moving to AFRI-4 in the constrained case is approximately $1.02 billion more than in the open market case The refinery cost impacts depend in part on the SSA region, and are a function of the: E x t ractive Industries for Development Series 27 · Standards of the transportation fuels already being made · Technology already installed, particularly FCC and desulfurization capacity · Types of crude oil being used, for example, West Africa, in general, has good quality sweet crude oil, compared to the imported sour Middle Eastern crude oil used in South and East Africa Table 3.7 shows the incremental unit costs for 2020 for the AFRI-4 transportation fuels for both the constrained cases and the open market cases Assuming no change in the taxation included in the fuel prices, the incremental cost of improving the transport fuels quality up to the AFRI-4 level for the SSA consumer will be 0.036 US dollars per liters, Table 3.7: Incremental Unit Costs for AFRI-4 Fuels in the Constrained and Open Market Cases Case Year Units SAA Total Africa Africa Africa West South East Constrained 2020 $/barrel $5.56 $6.39 $3.69 $7.09 $/tonne $44.26 $50.86 $29.37 $56.44 Open Market 2020 $/barrel $5.46 $6.25 $3.66 $6.99 $/tonne $43.46 $49.75 $29.13 $55.64 Looking at the SSA refineries in a global context it became clear that the small, older refineries that were already experiencing problems ran into insurmountable problems in the open market cases. On the other hand the SSA refineries that could contend with global competition were faced with high costs to move to AFRI-4 standards but would gain increased refinery margins and higher revenues. What the open market cases show is that a large number of the existing SSA refineries hold their own against the foreign export refineries. SSA refineries that do well in the global context usually: · Are larger, · Have invested over time and are more complex, · Are more efficient, · Have access to good quality local crude oils, and · Have access to a larger market or have a niche market. In addition, given the projected demand growth there appears to be the opportunity for new (or expanded) refinery capacity between 200,000 and 400,000 b/d depending on economic circumstances. Refinery expansion is focused in the Africa West region, in part because of the much higher population. ICF feels that this expansion will fall in the Ghana- Côte d'Ivoire- Nigeria area and also possibly in Angola, even though technically Angola is included in Africa South in the model. Readers of the report will also note that there is refinery expansion in Africa East. After considerable discussion it was decided to force the model to 28 S u b-Saharan Africa Refinery Project build a new 50,000 barrel per day refinery to process the new Ugandan crude oil. 4. SSA Refinery Project Results and Conclusions The final step of the SSA Refinery Project was to compare the total refinery investment costs to meet the AFRI-4 standards with the estimated monetary value of the health benefits associated with reduced emissions from cleaner fuels and implementation of pollution control equipment for vehicles. The costs and benefits were compared on the basis of the cumulative refinery investment costs in an open market and the health benefits (Scenario 2) over 5 years to 2015 and over 10 years to 2020. The net present value (NPV) was calculated using a 7 percent discount rate. Tables 4.1 and 4.2 show the results expressed in billions of 2007 U.S. dollars. The tables show that on an NPV basis the benefits to the health of the population of SSA during the period examined far exceed the investment costs required for the refineries to move to the AFRI- 4 fuel specifications. The benefits outweigh the costs even in the south region; the south has lower investment costs because that region already has refineries with advanced configurations and largely AFRI-3 fuel standards, so the step to AFRI-4 is the smallest. As both the Health and Refinery Studies have pointed out, there are considerable uncertainties largely driven by the availability and accuracy of the data. Table 4.1: Net Present Value of Refinery Supply Costs versus Health Benefits over 5 Years Billions 2007 dollars SSA Total West Region East Region South Region Refinery Investment $2.76 B $0.47 B $2.13 B $0.59 B Costs to 2015 Health Benefit over $25 B $18 B $5.3 B $1.0 B 5 years1 1.Central value shown for elasticity=1.5; ranges for elasticities of 1.0 and 2.0 are shown in the report and for Scenario 2 (lower sulfur fuel and pollution control equipment) with alternate 2-stroke motorcycle emissions assumptions. Table 4.2: Net Present Value of Refinery Supply Costs versus Health Benefits over 10 Years Billions 2007 dollars SSA Total West Region East Region South Region Refinery Investment $6.14 B $4.96 B $2.48 B $0.99 B Costs to 2020 Health Benefit over $43 B $33 B $9.0 B $1.8 B 10 years1 1. Central value shown for elasticity=1.5; ranges for elasticities of 1.0 and 2.0 are shown in the report. For Scenario 2 (lower sulfur fuel and pollution control equipment) and alternate 2-stroke motorcycle emissions assumptions. E x t ractive Industries for Development Series 29 Nevertheless, these studies provide a methodology for assessing the costs and benefits of improving the quality of transportation fuels in SSA. Although these studies looked out only to 2020, all indications are that SSA will continue to grow and that growth is likely to be accompanied by further urbanization and vehicle use. In addition, future changes in SSA will likely affect results of modeled air emission. For example, although in some cases air pollution was dominated by domestic emissions and road dust, in the future these sources may be reduced with increased modernization, leaving vehicle pollution as the predominant source of air pollution. The Health Study results indicate that there is a potential for significant health benefits in Sub-Saharan Africa's urban areas associated with the use of improved fuels, and the Refinery Study provides evidence that: · SSA refineries are already halfway up the AFRI standards, mitigating the size of the required future investments, and · Many of the SSA refineries are capable of functioning in the larger context of the global market. The combined results of the Health Study and the Refinery Study have shown that the estimates of the potential for health benefits outweigh the costs of improving fuel quality over time in SSA. These studies provide one line of evidence for the decisions to be made by Sub-Saharan African policy makers on whether to move forward with the manufacture of improved fuels. Drawing on the conclusions of both studies ICF believes that policy makers in the SSA region should consider the following recommendations. 30 S u b-Saharan Africa Refinery Project · Governments should instigate a sustained effort to update official fuel specifications to reflect current refineries' actual production. · Governments should consider upgrading the official fuel specifications in their country to AFRI-4 by 2015. · SSA refineries that could economically produce transport fuels at AFRI-4 specifications should develop investment plans to meet a 2015 deadline. · In parallel, Governments should adopt and enforce regulatory measures to improve the vehicle fleets such as: - All imported gasoline powered cars must have functioning catalytic converters - Establish an inspection and maintenance program - Encourage the phase out of old, highly polluting vehicles - Encourage the phase out or banning of 2-stroke engines · To reduce the uncertainties confronted in this study Governments, International Agencies, and Technical experts should conduct additional research with respect to: - Data on emissions in major urban centers (including permanent installations for measuring pollution levels) - Examination of the appropriate methodology for estimating health costs and benefits in developing nations - Improvements of data on product demand · The WORLD® model output indicates that major expansion will be needed in existing refineries as well as new regional refineries in the Africa West/Angola region. Lesser, but still positive growth is expected in Africa East and refining investment will probably be tied to Ugandan crude oil. More focused technical studies should concentrate on these regions. · Open market policies should be expanded to allow refineries to improve their revenue as they move to AFRI-4 standards. · Policy makers need to decide the future of the mainly topping/reforming refiners such as those in Senegal, Congo, Gabon, and Kenya. The study indicates the magnitude of the costs of modernizing and operating these refineries to meet the AFRI-4 specifications. E x t ractive Industries for Development Series 31 THE WORLD BANK OIL, GAS, AND MINING POLICY DIVISION The World Bank Group's role in the oil, gas, and mining sectors focuses on ensuring that its current interventions facilitate the extractive industries' contribution to poverty alleviation and economic growth through the promotion of good governance and sustainable development. The Oil, Gas, and Mining Policy Division serves as the Bank's global sector management unit on extractive industries and related issues for all the regions of the world. It is part of the Oil, Gas, Mining, and Chemicals Department, a joint World Bank/International Finance Corporation department. Through loans/credits/grants, technical assistance, policy dialogue, and analytical work, the Division leads a work program with multiple activities in more than 70 countries, of which almost half are in Sub-Saharan Africa. More specifically, the Division: · Advises governments on legal, fiscal, and regulatory issues and on institutional arrangements as they relate to natural resources, as well as on good governance practices. · Assists governments in setting up environmental and social safeguards in projects in order to promote the sustainable development of extractive industries. · Helps governments formulate policies that promote private sector growth and foreign direct and domestic private sector investments. · Advises governments on how to increase the access of the poor to clean commercial energy and to assess options for protecting the poor from high fuel prices. The Oil, Gas, and Mining Policy Division serves as a global technical advi-sor that supports sustainable development by building capacity and provid-ing extractive industry sector-related advisory services to resource-rich developing country gov- ernments. The Division also carries out an advocacy role through its man-agement of the following global programs: · The Extractive Industries Transparency Initiative (EITI) Multi-Donor Trust Fund, which supports countries in implementing EITI programs. · The Global Gas Flaring Reduction (GGFR) Public-Private Partnership, which brings governments and oil companies together to reduce gas flaring. · The Communities and Small-Scale Mining (CASM) Partnership, which promotes an integrated approach to addressing issues faced by artisanal and small-scale miners. · The Gender and Extractive Industries Program, which addresses gender issues in extractive industries. · The Petroleum Governance Initiative (PGI), which promotes petroleum governance frameworks, including linkages to environmental and community issues. · The Extractive Industries Technical Advisory Facility (EI-TAF), which facilitates "rapid-response" advisory services on a demand-driven basis to build capacity for extractive industry resource policy frameworks and transactions. Other publications in the Extractive Industries for Development Series are: #1 Vulnerability to Oil Price Increases: A Decomposition Analysis of 161 Countries by Robert Bacon and Masami Kojima #2 Changes in End-User Petroleum Product Prices: A Comparison of 48 Countries by Masami Kojima #3 Extractive Industries Value Chain: A Comprehensive Integrated Approach to Developing Extractive Industries by Eleodoro Mayorga Alba #4 Mining Cadastres: Promoting Transparent Access to Mineral Resources by Enrique Ortega, Alexandra Pugachevsky, and Gotthard Walser, #5 Emerging Players in Global Mining by Dr. David Humphreys #6 Changing Patterns of Household Expenditures on Energy: A Case Study of Indonesia and Pakistan by Robert Bacon, Soma Bhattacharya, and Masami Kojima #7 Financial Surety: Guidance Notes for the Implementation of Financial Surety for Mine Closure by Meredith Sassoon #8 Mining for Equity: Gender Dimensions of the Extractive Industries #9 Guidance Note for Task Team Leaders: Mainstreaming Gender into Extractive Industries Projects #10 Government Response to Oil Price Volatility: Experience of 49 Developing Countries by Masami Kojima #11 CO2 emissions; household energy use; and oil price transmission, Masami The World Bank 1818 H Street, N.W. Washington, D.C. 20433 USA www.worldbank.org/ogmc (OR /oil OR /gas OR /mining) www.ifc.org/ogmc (OR /oil OR /gas OR /mining)