Angola - Issues and options in the energy sector

Report No. 7408-ANG Angola: Issues and Options in the Energy Sector May 1989 Report 61 the Joint UNDPAVorld Bank Energy S6ctor Assessment Program ~'This documen4 has a restricted distribution. Its contents may not be disclosed without authorization from the (,-qvernmient, the UNDP or thd World,3ank., 9 'I~~  JOINT UNDP/WORLD BANK ENERGY SECTOR ASSESSMENT PROGRAM Reports Already Issued Country Date Number Indonesia November 1981 3543-IND Mauritius December 1981 3510-MAS Kenya May 1982 3800-KE Sri Lanka May 1982 8792-CE Zimbabwe June 1982 3765-ZIM Haiti June 1982 3672-HA Papua New Guinea June 1982 3882-PNG Burundi June 1982 3778-BU Rwanda June 1982 3779-RW Malawi August 1982 3903-MAL Bangladesh October 1982 3873-BD Zambia January 1983 4110-ZA Turkey March 1983 3877-TU Bolivia April 1983 4213-BO Fiji June 1983 4462-FIJ Solomon Islands June 1983 4404-SOL Senegal July 1983 4182-SE Sudan July 1983 4511-SU Uganda July 1983 4453-UG Nigeria August 1983 4440-UNI Nepal August 1983 4474-NEP The Gambia November 1983 4743-GM Peru January 1984 4677-PE Costa Rica January 1984 4655-CR Lesotho January 1984 4676-LSO Seychelles January 1984 4693-SEY Morocco March 1984 4157-MOR Portugal April 1984 4824-PO Niger May 1984 4642-NIR Ethiopia July 1984 4741-ET Cape Verde August 1984 5073-CV Guinea Bissau August 1984 5083-CUB Botswana September 1984 4998-BT St. Vincent and the Grenadines September 1984 5103-STV St. Lucia September 1984 5111-SLU Paraguay October 1984 5145-PA Tanzania November 1984 4969-TA Yemen Arab Republic December 1984 4892-YAR Liberia December 1984 5279-LBR Islamic Republic of Mauritania April 1985 5224-MAU Jamaica April 1985 5466-JM C6te d'Ivoire April 1985 5250-IVC Benin June 1985 5222-BEN Continued on inside back cover  ANGOLA ISSUES AND OPTIONS IN THE ENERGY SECTOR MAY 1989 This is one of a series of reports of the Joint UNDP/World Bank Energy Sector Assessment Program. Finance for this work has been provided, in part, by the Government of Sweden, the UNDP and the World Bank, and the work has been carried out by the World Bank. This report has a restricted distribution. Its contents may not be disclosed without authorization from the Government of Angola, the UNDP or the World Bank.  This Report was based in part on the findings of a mission which visited Angola in April/May 1987. The mission was led by Michel Del Buono (Senior Economist) and comprised of Messrs. W. Teplitz-Sembitsky (Consultant, Energy Economist and Deputy Mission Leader), J. Baptista (Power Engineer/Economist), S. Dalin (Power Engineer), M. Grimaud (Natural Gas Specialist), J. Lopes (Power Engineer/Household Energy Specialist), W. Matthews (Petroleum Specialist - Refining and Distribution), M. Paues (Researcher), R. Sergio (Utility/Financial Analyst), R. Soto (Petroleum Specialist - Expioration and Contracts), H. Warfvinge (Forestry Economist). Mr. R. Bates (Deputy Division Chief) joined the mission in its final stage and participated in the round-up meetings. Mr. Pedro V. Pinheiro reviewed the sections dealing with petroleum and product marketing procurement. The main authors of this Report are Messrs. Michel Del Buono and Witold Teplitz-Sembitsky. A preliminary version of this report was discussed with the Angolan Government in late September 1988 and in early February 1989 and was issued simultaneously in Portuguese and English.  ABSTRACT Angola has sizeable hydrocarbon reserves, a large hydro potential and ample woodfuel resources. Its ability to export annually 12-14 million tons of crude oil has sustained the economy over the past dozen years of civil war. However, the war has significantly affected the energy sector. While the State oil company, SONANGOL, and the oil refinery in Luanda run well, the power sector has steadily deteriorated over the last decade, and the utilities have run into ever increasing cash deficits. Artificially low prices for both electricity and oil products are encouraging wasteful consumption, and not permiczing recovery of costs, particularly for electric power. Some investment in natural gas is justifiable, to increase the supplies of LPG. With regard to traditional fuels, the Government's "hands off" policy should be continued in the current situation until a more active policy becomes possible with the return of security in the supply areas. The proposals made in this report are designed to increase the efficiency of the energy sector, through a number of moderate investments in infrastructure, but also through changes in pricing policy, and overall improvements in sector management. The mission recommends that the focus for the electric power subsector should be on maintaining and upgrading the existing electric power infrastructure and improving the reliability of supply. The expensive Capanda project should be postponed. A power consulting firm has reevaluated the Capanda project and the Bank has agreed to review the resulting report. To keep the oil sector prospering and avert a precipitous drop in petroleum reserves and production, it is recommended that the Government continue strengthening SONANCOL and maintaining the competitiveness of contractual terms for foreign oil companies. SONANGOL should resume full responsibility for petroleum product trading to reduce the foreign exchange costs involved. Angola's LPG imports could be replaced by refractionating LPG obtained through the LPG recovery scheme in Cabinda. Domestic distribution of oil products should be assigned to an autonomous or semi-autonomous division of SONANGOL, which could, possibly, include private, foreign or domestic capital. In household energy, a return to peace would require the development of new strategies to ensure a less wasteful utilization and more competitive trading of the country's woodfuel resources, the demand for which is bound to increase if and when more reasonable prices are are set for petroleum derivatives.  ABBREVIATIONS AND ACRONYMS AGIP Italian State Oil Company (Part of ENI Group) BEP Belgian Engineering Promotion BNA Banco Nacional de Angola (the Central Bank) BRASPETRO International Affiliate of Petrobras, (State Oil Company of Brazil) CABGOC Cabinda Gulf Oil Company (Joint Venture between SONANGOL and Gulf-Chevron) CELB Companhia Eletrica do Lobito e Benguela CEPSA Oil refining company CHEVRON U.S. oil company CIDA Canadian International Development Agency COMERINT A consulting firm belonging to the ENI Croup (Italy's State Hydrocarbons Holding Company) CONOCO U.S. oil company (Continental Oil Company) DNACO National Directorate for the Conservation of Nature DNERFE Department of New and Renewable Sources of Energy EEC European Economic Commission EDEL Empresa de Eletricidade de Luanda ELF French oil company ENDIAMA Empresa Nacional de Diamantes de Angola ENE Empresa Nacional de Eletricidade ERR Economic Rate of Return ESMAF Joint UNDP/World Bank Energy Sector Management Assistance Program ESPA Empresa de Servicios Petroliferos de Angola E.T.C. Dutch Foundation for Economic Research FPA Fina ?etroleos de Angola FURNAS Furnas Centrais Eletricas, a Brazilian utility GAMEK Gabinete de Aproveitamento do Medio Kwanza (Office for the Harnessing of the Middle Kwanza) GDI Gross Domestic Investment CDP Gross Domestic Product HEAC Hidro Eletrica do Alto Catumbela HFO Heavy Fuel Oil IBRD International Bank for Reconstruction and Development INP Instituto Nacional de Petroleo JPEA Junta Provincial de Eletrificacao de Angola (utility of Southern Angola) LFO Light Fuel Oil LPG Liquefied Petroleum Gas f LRMC Long Run Marginal Costs LSFO Low Sulfur Fuel Oil  ABBREVIATIONS AND ACRONYMS (continued) MEP Ministerio dti Energia e Petroleo (Ministry of Energy and Petroleum) MPLA PT Movimento Popular de Libertacao de Angola - Partido do Trabalho (the ruling party) OGE Orcamento Geral do Estado (General Budget) PSA Production Sharing Agreement SADCC Southern African Development Coordination Conference SEF Saneamento Economico e Financeiro (Program for Economic and Financial Restructuring) SOFRELEC French engineering firm SONANCOL Sociedade Nacional de Combustiveis de Angola (State oil company) SONEFE Sociedade Nacional de Estudo e Financiamento de Empreendimentos Ultramarinos (northern Angola utility) TAU Technical and Adm.nistrative Unit (Energy) of the SADCC TEXACO Texas Oil Corporation TOTAL French oil company TPE Technopromexport (Soviet engineering company) UINDP United Nations Development Program UNIDO United Nations Industrial Development Organization  CURRENCY EQUIVALENTS US$1 = 29.62 Kz (Kwanza) 1/ 1 Kz = 3.38 US cents ENERGY TERMS AND MEASUREMENTS BCF billion cubic feet CIF cost + insurance + freight DWT deadweight tons FOB free on board ft/y feet a year GWh Gigawatt hours HV high voltage kcal kilocalories kcal/kg kilocalories per kilogramme kgoe kilogram of oil equivalent km kilometer km' square kilometers kWh kilowatt hours kV kilovolts LY low voltage MN cubic meters MAI mean annual increment MCF thousand cubic feet mcwb moisture content, wet basis MMBTU millions of British Thermal Units MMCF millions of cubic feet MMCFD millions of cubic feet per day MV medium voltage MW megawatts TCF trillion cubic feet t/d tons a day t/y tons a year toe tons of oil equivalent 1/ Official exchange rate prevailing in Angola since 1975.  TABLE OF CONTENTS Page SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS ........................... i I. ENERGY IN THE ECONOMY ....................................... . 1 General Economic Framework. . .. ....*. . . . .. . . . . ... . ... . . ..e ooo 1 Petroleum and Public Financ-... .......................... 3 International Trade and Bal.,ce of Payments.............. 3 Policy Reforms.... ..... . o...* . . . e... . o.. .. .. ooo 4 Energy Sector Overview..oo... .... ...... . oo. o .......*. 4 Petroleum Products.ooo .o.oo..*...*o.o....oo.o...o....... 6 Crude Oil ...........................ooo.o..ooooo .o ..oo.o 7 Natural Gas.................,............................ 10 Electricityo .........................................0 10 Woodfuelso oo.,...* o.o ., o.*,,o...o...oo..oos.. 11 Energy Demand Projections.... .ooooo.ooooo.o .o.o. ....... 12 Institutional Framework....................... 13 SADCC - Energy Technical and Administrative Unit (TAU).. 14 Manpower, Technical Assistance, and Trainingo............... 15 Training for the Petroleum Subsector.........................** 15 Training for the Power Subsector......................... 16 Angolan Development Strategy in the Energy Sector ........... 17 Petroleum Developmenit Strategy and Peace..o.. .o.o...... 18 SONANGOL o..o.esooooo.oo.ooooooo.oooo..................... 18 Refining and Product Supply ..o.......................... 18 Power. wo. eooro *,eoo.. ...***,,,ooo,.,*oooo,o.oo.oo.oo..oo 19 Household Energy .............. 20 II.-A. CRUDE OIL: UPSTREAM ACTIVITIES .........o ................. 21 Summary and Recommendations.......e......o... .........o.o.o . . .oo 21 Oil Exploration and Production History ...................... 23 Oil Production and Investmento ..........................00 24 Institutional and Fiscal Framework.......................... 26 Oil Taxation ooo..**,o.o..o.oe.ooeooo.o................... 28 Marginal Oil Fed .................... 30 Prospects for Oil Field Development......................... 31 II.-B. GAS SUPPLY AND UTILIZATION...... .... oooo*..o..o. . . . . . . . . . . 36 Summary and Recommendations. ..e. . . . . o. * .o. . . . . . . . . 36 Gas Reserves and Utilization................................. 37 Market Potential of Non-Associated Gas................... 39  TABLE OF CONTENTS (Continued) III. CRUDE OIL: REFINING AND PRODUCT SUPPLY ........ 42 Summary and Conclusions... ..... *........ 42 Advantage of a Correct Pricing Policy.. .................. 42 R3fining 0................. 43 Distribution ......... of ..... 0 .... *...* 44 Procurement ...........****.... . .......... 44 Production, Supply and Consumption Considerations ........... 44 Product Trading/Import-Export ........... .... . ;.. 45 International Comparisons ........ ... .. . ....... . ..... 51 Projected Petroleum Product Consumption9. ................. 52 Pricing of Petroleum Products ............................... 53 Refinery Gate (RG) Pricing of Products ................ . 54 Pricing of Products to Final Consumers ................... 55 Economics of Refining vs. Direct Product Supply .......... 57 IV. ELECTRICITY SUPPLY ............ ......................... ....... 60 Summary and Recommendations.......o. ...... ....... ... .. ..... . 60 Electricity Supply.. .... ........... .* 61 Overall Generating Conditions. .......... o.. o.o.o.o.. .... 61 Northern Sys temi a ........... ................. 63 Central System ........................................ 63 Southern System tems................ ............ 64 Transmission and Distribution ............ ..** ........ 65 Ele ctricity Demands...... .... .. .... ... .... . ..... 66 Past Situation... .. ............ 0.... 66 Luanda.E Northern tem ............... ,0*00&....... 67 E entral and Southern Systems .............. ........ 68 Load Curve .Demand Pojecion..... .. ... .... . .... .. 69 Dem and P rojectionsand Uti ..ity..ina.ce..... ....o....... 69 SONEFEi Northern System Power Utliie.......... ..... 70 ENE: Central and South ern Systems .....O..#......** ..... 70 Alternative Demand Projections ............... .............. 71 Electricity Tariffs and Utility Financesi c.* .... ... 73 A ccounting onof.ua .if .edManp o.oo* ... .. 74 Fin ancial Si tuation of the Power Utilities ............... 75 External Debt of the Power Subsector ................o .......... 76 Billing and Collection..... ................... .. 77 Manpower, Staffing, and Technical Assistance ... ..........to............ 77 Misallocation of Qualified Manpower*............. ..............*****..... .... 78 Technical Assistance ......... .. ... ...... 79 Main Issues and Recommendationso ..o.o..................... 79 Org anizati on ....... Pla..ng........ ...,.......... 79 Management ... ......... ................,....... 82 Financial Situation and Tarif r....i f... f s.O* .. 83 Billing and Revenue Collcto l.....c0t.. i on06 .. 8 5 Qualified Manoea......p`o..w...e r0.00. .. 86 Training ........ ........... ,0 .. 86 Technical Assistance.. 87 Investment and Expansion Planig n.. n in...g .....0 87 Investment Priorities ... .**o... oo.......o.o... ,eee 89  TABLE OF CONTENTS (Continued) IV. ELECTRICITY SUPPLY (Continued) The Capanda Hydroelectric Project ........................... 92 Summary and Recommendations .. .....**.. - 92 Background to the Capanda Project........................ 93 Capanda and the Least Cost Expansion Plan.a n.0..,...... 94 Long Run Marginal COSILs......................... 95 Technical and Financial Information........................... 96 Capanda and the Need for Interconnections..... 99 Final Remarks ................................. 100 V. FORESTRY, WOODFUELS AND HOUSEHOLD ENERGY....................... 101 Summary, Conclusions, and Recommendations................... 101 Consumption and Production of Woodfuels, ................... 103 Consumption of Woodfuels.. .. .... ... ........ .... . 103 Wood Production ...... * ... .. ,. 106 Production Estimate by DNRFE.......... 107 Production Versus Consumption.s.um............... 109 Institutional Issues: Administration of the Forestry Sector ...........110 Marketing and Pricing ........ .. ... .* ........... . 111 Organizations of Woodfuels Suppliers..............p 113 Issues and Recommendations ............ ...... 114 Actions That Can Be Undertaken Under the Present "No Peace" Situation..u.,..o.... 115 The Coast: "No Pea e" .........l...S1 Inland Areas: "No Peace"............ 0.0 ...... . ....... 117 Alternative Priorities in a "Peace" Situation............... 118 Proposed Actions .... ................ ................. 118 Pilot Project in Huila-Namibe ................... 119 Proposed Supportive Action at the National Levelv.... 121 Strengthening of Institutions.......... G . ..... . 122  ANNEXES Annex 1 Macroeconomic Indicators. ...* ........................ ... 123 Annex 2 Institutional Organizations within the Energy Sector ..... 129 Annex 3 Angola Energy Balance............................... ..... 137 Annex 4 Figures on Petroleum and Case ......................... 144 Annex 5 The Petroleum Law.. ..... ..... .....*..* a.. .. . .. .. . ....... 150 Annex 6 Taxation of the Petroleum Sector ........... ............... 152 Annex 7 Financial Analybis of the Ammonia/Urea Plant ............. 158 Annex 8 Figures on the Petroleum Product Subsector........... ..... 162 Annex 9 Structure of Petroleum Product Prices*.................... 172 Annex 10 Economics of Luanda Refinery Using Actual Historical Values .................... 176 Annex 11 Human Resources in the Downstream Petroleum Sector ....... 180 Annex 12 Petroleum Product Trading . ............... ......s. 182 Annex 13 Figures on the Power Subsector .........s.**** ....... 185 Annex 14 Electricity Demand Projections - Main Assumptions, BEP, THEMAG, and Mission Studies ................ 195 Annex 15 Angola - Electricity Tariff System...9-9 ......... sees 223 Annex 16 Summary of the Capanda Hydro Project ..................... 234 Annex 17 Angola: Forestry and Household Energy--Outline of Four Priority Projects. ................ ...... 0. 253 Annex 18 Draft Terms of Reference for Preparation of a Power Subsector Rehabilitation Project, updating of LCEP for the Northern System and Costs and Benefits of Halting Work on the Capanda Project.................... 265 TABLES 1 Angola. Action Plan for the Development of the Energy Sectorxviii 1.1 ComparaLtive Economic Indicators, 1985ooe-....... -. .0.00.6.. L. 1.2 Angola: Key Economic ndicatorsn...o 3 1.3 Summary of Angola Energy Balance, 19 86 5 1.4 Summary of Trends in Commercial Energy Consumption.... 6 1.5 Domestic Petroleum Product Sales by Sector, 1985.. 8 5.*... ... 7 1.6 Trends in the Performance of the Petroleum Sector. to......sees 8 1.7 Indicators for Petroleum Sector Claims on Resources......... 9 1.8 Commercial Energy Demand Projections..............es........ 12 2.1 Oil Production and Investments in the Petroleum Sector*.....*. 24 2.2 Angola: Changes in Oil Tax Revenues and Oil Output, 1985-86.. 30 2.3 Projections for Future Exploration and Development, 1987-90... 32 2.4 Oil Production and Reserves, 1986-90. 8 6.- 9....so............. 34 2.5 Economics of Ammonia/Urea Plant............................... 40 3.1 Imports of Petroleum Products, 1980-86..0.-.........so... see. 45 3.2 Jet Fuel Supply 1980-86, Imports vs. Local Refinery-Sourced... 46 3.3 Petroleum Product Exports (Cargo) 1980-86.................. 47 3.4 Net Plant Yied. i..... elds. ...a....#...... 48 3.5 Crude Oil Feed - Luanda Refinery, 1985-86 ................. ....... 48 3.6 Luanda Refinery Production/Yield Balance, 1980 and 1985-86 .... 49 3.7 Luanda Refinery Operating Costs and Total Gross Margin, 1986.. 50 3.8 Inland Petroleum Product Consumption (Sales), 1980-86.........* 51 3.9 Per Capita Petroleum Consumption--International Comparison, 1984-853..0. P e Pt oleuPo o..nmi...on....................... 52 3.10 Projected Petroleum Product Consumption ..** ...................... 53  TABLES (continued) 3.11 Official Refinery Gate Prices vs. International Prices*........ 55 3.12 Illustrative Petroleum Product Pricing (at Kz 44/US$).......... 57 3.13 Illustrative Petroleum Product Pricing (at Kz 104/U3$)........ 58 3.14 Summary of Luanda Refinery Economics.......................... 59 4.1 Installed Available Generating Capacity, 1987................" 62 4.2 Electricity Generation and Consumption Projections............ 71 4.3 Energy and Demand Projections, 1986-2000 ........ 0.. 0....... 72 4.4 Angola: Electric Power Utilities, Sales and Cash Flow, 1986 ... 76 4.5 Angola: Electric Power Subsector, Theoretical and Actual Staff Profiles, 1987..-.. ........... 78 4.6 Electric Power Subsector: Suggested Priority Investment Program, 1987(88) - 1992. ... . .......... .. * 91 4.7 Capanda Hydro Project - GAMEK - Investment Program.......0 ..... 97 4.8 Capanda: Financing Package.. . . .. .. ....... ...... . .* ...... . 98 5.1 Woodfuels - Priorities for Action ...............G.......,....... 102 5.2 Consumption of Household Fuels, 1987......... ... 104 5.3 Aggregate Use of Firewood and Charcoal, 1987...* o .... o..,... 105 5.4 Hypothetical Annual Wood Fuel Demand ............. 106 5.5 Main Forest Formations ......................... 107 5.6 Selected Vegetation Groups and Their Fuelwood Production Potential...**e................. 108 5.7 Provincial Fuelwood Production Potential..................... 109 5.8 Prices Paid for Firewood and Charcoal in Seven Provinces, 1987................ .......... ........ 112 5.9 Comparative Cooking Costs of Four Household Fuels.o .......... 113 5.10 Priority Listing of Activities.....o..................o...... 115 MAP "Angola" IBRD 20067  SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS 1. Angola is a large, potentially rich country with energy resources far in excess of its own needs. Development of some of these resources has already made a major contribution, if not to the development, at least to the maintenance of a society facing internal instability and external aggression. Angola's ability to export annually 12-14 million tons of crude oil, worth between US$1.5 and US$2 billion, has certainly helped the population to weather the last ten years with fewer hardships than otherwise. In short, petroleum production, intelligently managed, has financed the past unavoidable and considerable defense effort and the basic consumption needs of the urban population. 2. The objective of this report is to clarify, for the benefit of Angolan policymakers, some of the main issues which should be resolved in order to meet the energy needs of the country most economically and, subsequently, to develop energy as a leading sector in economic growth and development. Needless to say, many of the recommendations made can only bear fruit if and when a reasonable degree of peace returns to Angola. Therefore, a number of these, such as the recommendation to set power tariffs on the basis of Long Run Marginal Costs (LRMC), are meant to be implemented gradually. Others, such as the recommendation to set power tariffs to cover the financial costs of the utilities, can be implemented immediately. Implementing them would have an immediate favorable impact on the energy sector, on sector enterprises, and on the efficiency of resource allocation. 1/ 3. Many of the problems addressed in this report are known to the Angolan staff working in the sector and to sector managers and technicians. The analysis and most of the recommendations presented herein have been discussed extensively with energy sector policymakers, managers, and staff. The recommendations rest mostly on economic, financial sr technical principles and analysis. No attempt has been made to deal with the political dimensions of the issues, a matter with which the Angolan authorities have to deal, and which could be much more complicated than the application of analytical principles that the Energy Assessment has performed. 4. While the report presents some urgent problems facing Angolan policymakers, it is by no means a complete treatment of all issues in the energy sector. Rather, it provides an overview of the role of the energy sector in the Angolan economy and its main problems. It also describes the main features of the Angolan economy as the framework in which the energy sector functions, but does not attempt to provide an exhaustive 1/ As of January 1, 1989, the MEP has proposed to increase electricity tariffs to an average of Kz 3/kWh, in line with the analysis of this report. This should be sufficient to cover, for the time being, the cash deficit of the utilities.  - ii - treatment of general economic issues 2/. Coverage of the petroleum and gas subsector is limited to issues such as pricing, taxation, and supervision of oil companies which the Angolan authorities suggested and which are some of the most pressing ones. The report does not give a detailed descriptive picture of the oil subsector. This is, in any case, well known both in Angola and abroad and is reported on systematically by the petroleum press. Analysis of the electric power subsector does, however, intend to cover all important issues. Through extensive analysis, the aim of this report i_. to outline a development strategy for the energy sector, indicating the broad directions which management of the sector ought to take. This is summarized below and developed somewhat more fully in paragraphs 1.41 to 1.51. The summary of sector strategy is followed by discussion of the three main issues that must be addressed immediately. The report's main recommendations are presented in the form of an action plan for development of the energy sector. This plan is outlined in matrix form in Table 1 at the conclusion of the summary. Recommended Development Strategy for the Energy Sector Oil Exploration and Production 5. In the short run, Angola has little choice but to develop and exploit its oil resources as fast as it can. Continuing and even increasing investments in exploration by international oil companies should be encouraged to prevent too rapid a drop in oil production in the mid-1990s, especially in view of the recent acceleration of oil output. In fact, if current (mid-1988) production levels are to be maintained indefinitely, Angola needs to reassess whether the current level of exploration expenditures is sufficient to avert a precipitous drop in reserves and, eventually, in production. To maintain incentives for exploration, the Government should not allow the competitiveness of contractual terms for the oil companies to fall significantly behind those of other countries. The Government also needs to continue strengthening the State-owned oil company, SONANGOL (Sociedade Nacional de Combustiveis de Angola), its effective instrument for promotion and control of oil activities. The best way to strengthen SONANGOL is to increase its managerial and financial autonomy as much as possible, while improving its capacity to analyze the economic and technical issues that arise in its activities. To strengthen SONANGOL's management and supervisory role, an incentive system should be developed that will enable SONANGOL to attract and retain specialized staff. To improve its capacity to analyze economic and technical issues, an improved management information system is needed. 2/ This is done in the UNDP/World Bank Report entitled, "Angola: An Introductory Economic Review".  - iii - Refining and Petroleum Product Supply 6. In refining and petroleum product supply, the present situation is broadly satisfactory. The FPA (Fina Petr6leos de Angola) refinery in Luanda runs economically. As a result of its recent de-bottlenecking and life-prolongation, it should be able to supply most of Angola's needs for quite some time, especially if peace returns soon. Angola can continue to satisfy its excess demand via imports but should act to reduce wasteful or low-priority consumption through a substantial increase in prices and the elimination of redundant subsidies.3/ To reduce the foreign exchange costs of petroleum product import/export arrangements, SONANGOL should resume full responsibility for procurement of imports and cargo exports and seek cheaper supply sources. For the short-to-medium term, large investments in refinery capacity can probably be postponed because of the great uncertainties in forecasting demand. At a time when both the State and SONANGOL need to devote resources to petroleum exploration and production, purchasing most or all of the stock or assets of the refinery cannot be a high priority. However, if a share of refinery ownership can be obtained without great, net outlays of Angolan public resources in the framework of a reshuffling of assets and claims between SONANGOL, FINA and the Government, then such a transaction could take place. 7. Refinery and distribution operations need to be studied to improve upon the cost-plus system that is common to both. In refining, incentives toward greater eTficiency could be built in and applied. One way to achieve this goal would be to have FPA share in any savings earned through improved productivity and be penalized for inefficiencies, e.g., by paying some share of them. To improve efficiency in distribution operations, rather than creating incentives and penalties, it might be easier to spin off distribution to a relatively autonomous affiliate of SONANGOL. Additional work would be needed to define the precise form of th_ incentive system. This is proposed as a subsequent ESMAP activity. Natural Gas 8. In general, prospects for development of non-associated fields are not promising. Rather than exploiting these fields, it is better to leave the gas in the ground until economic uses can be found for it. No investments should be made in assessment or delineation of gas fields unless major economic uses for the gas have been identified. Of the associeted gas currently produced, about 50% is being used productively. New lift and reinjection schemes and the expansion of existing ones are likely to increase the utilization rate to 70% by 1990. As regards the further use of associated gas, highest pricrity should be given to an LPG fractionation scheme in Cabinda which would replace LPG imports, and an 3/ In the absence of a reliable exchange rate, the report finds it difficult to make precise price recommendations. However, internal prices of petroleum derivatives are widely held to be negligible and in need of serious adjustment, if only to dispel the false idea that petroleum products are costless.  - iv - export-oriented LPG/condensates recovery program in Block 3. If the first project goes ahead, an LPG bottle rehabilitation plant would be required as well. SONANGOL's ability to study and supervise gas-related projects should also be strengthened by forming around the few people currently dealing with gas in SONANGOL, a small unit responsible for gas matters. While large petrochemical projects (such as ammonia urea) seem to have rather dim prospects, if foreign investors are prepared to take all risks, Angola could well accept such projects, provided a reasonable price were paid for the gas. Electric Power 9. Angola's electric power subsector still operates reasonably well but has suffered from more than a decade of neglect, and the war has further contributed to this problem. Demand has stagnated, especially in industry, and households have become the main consumers of power. Given the great uncertainties in forecasting demand, a risk-averse stance with respect to investments would seem the preferable strategy. More specifically, this report suggests the following strategic guidelines: (a) maintain a reasonable quality of service, without sizeable new investments; (b) launch a rehabilitation program for existing dams, power plants, and transmission grids; (c) strengthen key central management functions, such as system and financial planning and equipment standardization, but maintain a decentralized operations structure; and (d) set the base for future growth by strengthening management, accounting, and finance and ensuring that the utilities have the managerial autonomy which they require for efficient operation. Additions to capacity should be very low on the list of priorities until after the mid-1990s. The current centerpiece of investment in electric power, the Capanda project, should be postponed and re-examined in the mid-1990s when, hopefully, a better assessment of future demand would be possible. Household Energy 10. For household use, the cities presently receive both traditional and commercial energy. A major problem is the dramatic under-valuation of commercial fuels such as Liquefied Petroleum Gas (LPG) and kerosene. With a more balanced pricing policy in the future and with the removal of inefficiencies and risks in the supply of firewood and charcoal, consumption patterns even in the cities might not so overwhelningly favor commercial fuels. For the present, the Government  v needs to make sure that supplies to the cities are maximized. The essentially correct policy of keeping "hands off" trade in traditional fur'ls should continue in the short run. However, with the return of peace, the situation should be reviewed with the aim of: (a) establishing policies to promote a more efficient and competitive trade in fuelwood and charcoal; and (b) economizing on domestic use of some of the commercial fuels, thus creating exportable surplus. This longer-term strategy is viable because the country's forestry/biomass resources are sufficient and sufficiently, broadly distributed to satisfy the needs of the population under normal situations. No large, urgent interventions in forestry are really necessary or nossible at present. However, some useful actions are suggested in paragraphs 5.34 to 5.44. Main Issues in the Energy Sector 11. There are a number of issues and problems in the Angolan energy sector. Many are analyzed and discussed in the body of this report. The most important ones, however, are highlighted in the paragraphs below. For ease of discussion, they are organized under three headings: (a) investment programming in the electric power subsector; (b) price policies and financial problems of sector enterprises; and (c) management, qualified manpower, training, and technical assistance. 12. These issues are interrelated. They are probably of about equal importance in that they must all be resolved to allow improved efficiency and viable development of the sector. While there are some problems in the petroleum and gas subsector, none seem really grave or urgent, except for pricing of petroleum products, which is discussed in paragraphs 16 to 20. Investment Programming in the Electric Power Subsector 13. Investment programming in the electric power subsector reflects an economy-wide weakness in project analysis and selection. It also reflects a weak institutional capability in the utilities and in the supervising ministries. As a result, real needs or problems are not identified and projects to satisfy or resolve them are not carried out. An extreme case of the above is the centering of the investment program of the electric power subsector on the Capanda dam. This project, which may end up costing more than US$2 billion, is being considered despite the fact that it does not resolve the subsector's problems. Generating capacity would be added that, even allowing for a high degree of uncertainty over demand, would not be needed for many years. Moreover, this additional capacity could not be used because transmission and  - vi - distribution facilities are limited as well as run-down, and could not handle increased supply. Furthermore, an enormous share of generating capacity (about 50%) would be concentrated in one distant region of the country at a time when insecurity makes transmission iines vulnerable. This investment at Capanda would create very little additional revenue for the utilities and thus could exacerbate rather than alleviate the financial problems of the sector. Given the commercial nature of the financing plan for this investment, carrying it out would greatly inflate the sectorts external debt and even threaten availability of free foreign exchange if future oil output were offered as a guarantee. Recent loans obtained in an effort to fill a large gap in the financing plan (in excess of US$500 million) are at wholly inappropriate terms (7.5%, 2% insurance fee, two years grace, and 90 months amortization period). 14. In reality, the needs of the electric power subsector are quite different from those that a project such as Capanda would resolve. The subsector needs to catch up on major maintenance, which was neglected for more than a decade. It needs this for generating plants, transmission lines, and distribution grids. Given the u:certainties, firm forecasts are not possible, so that che subsector needs flexibility to respond to demand wherever in the country it might arise. To achieve flexibility, improvements and small additions to capacity are needed everywhere, in the Northern, Central, and Southern Systems. 1J. A project to resolve these problems is essentially the mission's main investment proposal for the electric power subsector (paras. 4.90-4.95). Briefly put, the objective of the project recom- mended in this report is to rehabilitate power subsector facilities so that they are able to operate at or near their installed capacity, with some small additions in some places, such as Matala and Lomaum. Since present available capacity is only about 59% of installed capacity (i.e., 27i MW out of 470 MW) this rehabilitation project is a substantial one. It is also a much better risk-averse response to the problems of the power subsector, and the uncertainties it faces, than a lumpy investment such as Capanda. In brief, therefore, this report recommends investments in the power subsector of about US$200 million (including technical assistance) over the next five years, as opposed to present programs exceeding US$1 billion. Pricing 16. Energy pricing is an area where immediate and radical policy changes are both imperative and feasible. Severe price distortions are an economy-wide phenomenon and must be addressed in an overall policy which seeks to revive the allocative role of markets. Under the prevailing regime, prices perform a passive, accounting function or serve as an inefficient redistributional mechanism. These prices do not provide signals to assist producers and consumers in their decisions about the rational use of scarce energy resources.  a- - vii - 17. The relative price of energy in Angola has fallen precipitously in the last several years as a result of basically fixed nominal prices for petroleum products and fixed nominal electricity tariffs. Other prices have increased at rapid rates. As a result, energy prices, including tariffs, have become negligible in real terms, resulting in wasteful consumption on the demand side, and large financial deficits for energy supplying firms, in addition to sizeable subsidies paid or revenues forgone by the State budget. Since energy prices have become negligible, subsidies have become redundant, i.e., whether LPG is sold at Kz 15 per kg or Kz 25 (at a definite cqt to the budget) is essentially immaterial at the present purchasing pL Ir of the Kwanza. Justification for these consumption subsidies is therefore non-existent. Similarly, crude oil for domestic refining costs the State budget approximately US$3 per barrel in subsidies, but results in no appreciable reduction in the cost of oil products to consumers. The pricing, taxation and subsidization system for oil products is described more fully in paragraphs 3.27 to 3.36. 18. Briefly put, this report suggests that oil products be sold at prices which reflect their opportunity costs at a suitable exchange rate. Products should also be taxed at roughly similar rates (because they are close substitutes). Automotive fuel prices should include a levy for road maintenance. In paragraphs 3.35 to 3.36 some calculations are made to determine illustrative prices for oil products. With more realistic exchange rates, and minding the revenue needs of the budget, this report calculates that prices of oil products would need to be increased three- to four-fold. As a result of these higher prices, excess low-priority consumption might be reduced somewhat, yielding a greater exportable surplus. Meanwhile, the Government would reap enough revenues to substantially reduce the current budget deficit. This would have a positive deflationary impact, even if public (defense) use of products were to be tax-exempt. 19. In the case of electric power, the immediate goal should be to restore the financial viability of the utilities. In the longer run, the aim of policy should be to base tariffs on Long Run Marginal Costs. In the meantime, electricity tariffs have become meaningless (the annual cost of electricity supply to a high income urban home is equivalent to a few cans of beer) ard the financial position of the electric power utilities, untenable. In the immediate, this report suggests that electricity tariffs be increased three- to four-fold to urgently direct some resources to the utilities. This would relieve the budget from having to supply Kz 1 - 1.5 billion in subsidies each year. In the medium term, an appropriate goal could be that tariffs should cover all financial costs as well as a share (say 20%-25%) of a reasonable investment program, or more simply, obtain a modest return on assets (say 4% or 5%) in addition to covering all costs. 20. Price measures are meant to help improve efficiency, and not to substitute for it. Thus, they need to be accompanied by other supportive actions. For example, in oil refining and distribution, the cost-plus  - viii - system governing these activities gives no incentive to cut costs, as savings are automatically transferred to the budget, and losses are auto- matically covered by the budget. A formula tying reductions in costs of refining with compensation to FPA needs to be defined and implemented. This formula could be defined in a subsequent ESMAP-assisted task. Similarly, the distribution activities now carried out by SONANGOL could be carried out more efficiently in a separate enterprise, or a very autonomous affiliate of SONANCOL, and without the cost-plus arrangement currently in force. The easiest system could be a tax regime that apportions the benefits of greater efficiency between the budget and the enterprise. For the power utilities also, increased tariffs will not help unless they are accompanied by measures to strengthen billing, collections, and technical and financial management generally, including system planning and project selection under uncertainty. Management, Manpower, Training, Technical Assistance 21. The shortage of skilled and trained manpower which Angola inherited, and which was exacerbated by the exodus of Portuguese settlers, still remains one of the most pressing problems to be resolved. Many of the shortcomings in the selection, preparation, and execution of basic policies to a large extent are attributable to the lack of qualified and experienced personnel. These constraints also apply to the energy sector though it appears that the energy subsectors, notably the oil enclave, are somewhat better off than the rest of the economy. 22. Not suprisingly, the petroleum subsector has been least affected by the country's shortage of managerial capabilities and technical skills. So far SONANGOL's high- and medium-level management positions have been staffed with comparatively experienced and competent personnel. In addition, SONANCOL has had access to, and has extensively used, the expertise of foreign oil companies and consulting firms, and there is little doubt that it should continue to do so in the future, while trying to reduce the cost by tapping potential sources of concessional technical assistance. However, other existing sources of know-how, such as on-the-job training programs provided by foreign oil companies, are biased towards technical and engineering skills. SONANGOL itself more urgently needs additional expertise in the areas of management, supervision, financial analysis and economics. 23. A unique feature of Angola's Petroleum Law is that all oil companies are required to assign US$0.15/bbl produced, to a training fund which is controlled by the Ministries of Finance, Education and the Ministry of Energy and Petroleum (the MEP, Ministerio de Energia e Petr6leos). At current production rates, these payments amount to US$16.8 million per year (Kz 500 million). However, control over these funds is unclear. A more transparent mechanism should be established to make sure that the resotrces, which are paid in hard currencies, are channelled into areas with highest educational priorities.  - ix - 24. In 1986, the power sector, excluding the Office for the Harnessing of the Middle Kwanza (GAMEK, Cabinete de Aproveitamento do Medio Kwanza) employed about 4,000 persons. About 100 of these were expatriates. While the know-how provided by the expatriate employees proved remarkably cheap (US$1,000 per man-month), it contributed little to improving the management and planning capabilities of the power utilities. With only 51 higher-level technicians and professionals (30 Angolans plus 21 expatriates), the power sector is extremely short of experienced and qualified manpower. Moreover, a disproportionate share of the higher-level staff is concentrated in the head office of the national electricity company (ENE, Empresa Nacional de Electricidade) in Luanda so that many of the operating utilities are left without any professional back-up. Given the formidable task of rehabilitating the basic infrastructure of the utilities and improving the sector's overall management capabilities, an additional and sizeable inflow of financial and human resources will be required. The technical assistance needed to improve the efficiency and financial viability of the power utilities would amount to 105 man-years (equivalent to US$10 million) over a three- year period. Much larger sums would need to be raised in order to finance the implementation of a minimum rehabilitation program centered on strengthening the operational capabilities of the power sector. 25. Compared to the immediate needs of the electricity subsector, external support to the country's forestry administration in the area of woodfuel resource management is not as pressing. Thus, priority for the assignment of qualified manpower, whether national or expatriate, should go towards strengthening operating companies in the various subsectors (power utilities, SONANGOL), the policy-making units of the MEP, and, eventually, the staff of the central power planning in a yet to be created decentralized national power company. Priority should be given to training efforts and technical assistance which: (a) meet the manpower requirements of the power subsector; and (b) keep the managerial capabilities of SONANGOL at their current high level (with small improvements where necessary, for example in examining issues of gas utilization). Other Conclusions and Recommendations 26. The following sections of the Executive Summary more systematically enumerate the conclusions and recommendations of this study.  x Oil and Gas 27. Angola depends on oil income economically and politically. General Government policies on oil development have been enlightened, and thus deservedly successful. A workable modus operandi was established between SONANGOL and the MEP (which has the overall policy and supervisory mandate over oil) in supervising oil activities in Angola. Even though its enabling legislation empowers it to explore for, produce, transport, refine and distribute oil, SONANGOL has two major practical tasks. The first is to encourage foreign investment in oil exploration and production and to negotiate advantageous contractual terms with interested oil companies. The second is to supervise and control foreign oil companies and to raise the funds required to meet its share of investment programs. Therefore, the performance of SONANGOL should be judged on its success in mobilizing and steering external resources into oil operations and in supervising and controlling foreign oil companies. For these reasons, for the time being SONANCOL should minimize its involvement in upstream operations and other extraneous activities unless these strengthen its supervisory role. Similarly, the domestic distribution and marketing operations of SONANGOL could be spun off into a relatively autonomous division or subsidiary, if not privatized. Although it is inescapable that SONANCOL remain under the political control of the State, it should be granted greater managerial and financial autonomv. SONANGOL's ability to raise the funds required to meet its financial obligations is crucial. Therefore, the Government should not routinely use oil as collateral except for oil operations. SONANGOL's ability to adequately control and supervise the activities of foreign oil companies can be improved by establishing a more efficient management information system. This could be done with ESMAP assistance. Training of qualified staff, in conjunction with the development of an incentive structure for skilled personnel, is an important task. However, a more efficient utilization of existing training opportunities and institutions should suffice, rather than the creation of new or specially designed programs. A suitable incentive system could play an important role in attracting and retaining managerial and highly specialized staff. Technical assistance would still be needed both to carry out complex tasks and to help train newer staff. 28. So far, the taxation system for the oil sector has worked well. Tax legislation has allowed the Government to capture windfall profits, while oil companies have been protected against a profit squeeze in periods of declining oil prices. It would therefore be counter- productive to make fundamental changes in the fiscal terms that apply to oil companies. 29. Part of the existing legislative framework was established in the 1950s and does not match the contractual approach and taxation system embodied in the more recent joint venture and production-sharing  - xi - agreements (PSAs). There are also differences in the contractual terms for joint venture and PSA operations. Although the Government is ready to deal with these problems pragmatically, it might be preferable to make small textual adjustments to the text of the legislation and contracts. 30. Significant quantities of associated gas (currently about 50% of total production) are used for gas lift and reinjection schemes. New lift and reinjection schemes are underway, and existing ones are being expanded. A target utilization rate of 70% is the goal for late 1990, up from the present 50% or so. This is a reasonable objective. No economic large-scale projects are presently known. Thus, prospects for the development of non-associated gas fields are dim. The only large-scale project capable of using sizeable quantities of natural gas is the proposed export-oriented ammonia/urea plant which would require about 50.6 MMCFD of gas. However, in view of the depressed international fertilizer market and given that gas supply costs are relatively high, Angola would not have a substantial comparative advantage even in a well- managed plant. 31. SONANGOL's ability to study and supervise even a limited number of gas-related projects should be strengthened by building a small unit responsible for gas matters, using as a nucleus the few people currently dealing with gas in SONANGOL. This unit should be in a position to monitor ongoing gas-related activities more thoroughly and to cocrdinate plans for future projects with related activities in other subsectors. 32. Highest priority should be given to two projects presently under consideration by SONANGOL: the LPC recovery scheme in Cabinda, and the export-oriented LPC/Condensates recovery program in Block 3. Other projects which deserve further investigation in the short term are: the planned LPG bottle rehabilitation plant, the proposed dual-fuel thermal power plant in the Soyo area, and the onshore plant for recovery of LPG at Malongo costing US$3-4 million (provided that demand will be adequate at the higher LPG prices which the report proposes). 33. Pricing of petroleum products at the refinery gate and to final consumers, and pricing of crude oil for domestic refining, are areas where substantial reforms could be implemented most easily. Many countries use oil prices as a fiscal mechanism to raise public revenues and to impress on the consumer the fact that oil is a scarce, costly and exhaustible resource. Both these aspects of oil pricing could fit well with the present economic situation of Angola. Yet, specific pricing recommendations are difficult to make in the macroeconomic policy environment of Angola. However, given the extreme overvaluation of the Kwanza, the standard economic prescription of using opportunity costs as the basis for pricing would only fully make sense after the value of the Kwanza has been adjusted downwards to some sort of equilibrium level (or to a level nearer to equilibrium than is currently the case). But since final petroleum product prices in Angola are below border prices even at the highly overvalued, present, official exchange rate, and the crude oil  - xii - for local refining is subsidized, these shortcomings would need to be corrected first. A series of step adjustments in prices would probably be easiest to apply. The steps could be as follows, using hypothetical exchange rates: Step One: Eliminate all subsidies to crude and products including LPG, and immediately bring all prices to border levels at the official rate of exchange. Step Two: Adjust all petroleum product prices to an exchange rate of, say, Kz 100/US$. Step Three: By this time, the Program for Economic and rinancial Reconstruction (SEF, Saneamento Econ6mico e Financeiro) should be in progress and a more adequate exchange rate might be available to guide the MEP in the pricing of petroleum products. Should the exchange rate remain fixed in spite of notable domestic price increases, the MEP could use an index of inflation to keep real product prices stable. 34. The refining of indigenous crude in Luanda in a hydroskimmer is an economically viable product-supply strategy for Angola as compared to imports of products. The FPA refinery is a reasonably run and well- maintained facility. The Government seems intent on purchasing this refinery or a genuine controlling interest in it. This would seem a low- priority use of limited Angolan funds unless it is done in such a way as to minimize the drain on public resources. Furthermore, lack of incentive to reduce costs and possibly high use of expatriate labor are the most apparent contributors to high operating costs. The FPA refinery operates on a "cost-plus" refinery gate pricing arrangement which gives no particular incentive for cost minimization and optimization of operations. Therefore, efforts should be made to design and implement a pricing scheme which encourages the refinery to operate in a more efficient way--for example, through a tax scheme that would share productivity gains between the Government and FINA. This could be done in the context of an ESMAP activity. Electric Power 35. Angola's power subsector, which still operates reasonably well, has suffered from more than a decade of neglect. By 1987 the firm capacity had deteriorated to 275 MW, which is less than 60% of total installed capacity. Transmission and distribution lines have hardly  - xiii - received any niaintenance since 1975. Though the present state of the utilities' accounts makes it almost impossible to assess their financial performance, there was little doubt that in 1987 the global cash deficit of the sector would approach the level of US$50 million (or about Kz 1.5 billion). 36. In order to safeguard a reasonable quality of service and to gradually restore the utilities' financial viability, strong measures are required immediately. Priority should be given to: improving the financial performance of the utilities; strengthening the utilities' operational and managerial capabilities, including accounting, billing, and collection systems; reorganizing the subsector to provide more internal managerial autonomy; and, most importantly, reorienting the investment program to favor rehabilitation of the existing physical infrastructure rather than expansion of capacity. 37. A significant and sustained improvement in operations, maintenance and management requires the influx of know-how and finance. Operational support for the Central and Southern Systems as well as advisory assistance to a proposed task force would require about 35 man- years of long-term consultants plus some short-term consultants at a total estimated cost of US$10 million. 38. The financial losses of the power sector are no longer sustainable. Therefore, cost recovery is a matter of utmost concern. To ensure cost recovery, there should be immediate increases in tariffs up to 400Z. The utilities' billing and revenue collection procedures should also be improved. In the short term, tariffs need to be simplified and restructured to enable the utilities to meet simple financial targets. In the medium term the adjustments should be designed so as to bring the level and structure of the tariffs in lire with Long Run Marginal Costs. 39. In the past, ENE--the national power company--was neither given the actual means nor the authority to assume the management of the subsector in a reasonably efficient way. Therefore, measures should be instituted to decentralize all operations and maintenance and part of the proposed rehabilitation activities to the Regional Directions, as this is closer to actual practice than the theoretical centralization implicit in the formal structure of ENE. At the central level, a small planning unit should be established and be responsible for strategic matters (demand studies, capacity planning, tariff studies, etc.). Such a unit is currently being established in the MEP.  - xiv - Investment Priorities in Electric Power 40. In the short term, investment priorities must center on repair, rehabilitation, and resumed maintenance of existing facilities. Rehabilitation should proceed simultaneously on all three systems as both security and economics relegate interconnection of the systems to a fairly distant future. The medium-term goal should be to fully restore supply capabilities in line with installed capacities. The investment programs for the electric power subsector as a whole, excluding Capanda, totalled about US$100 million for 1987 and 1988, 75% of it in foreign exchange. A program of this size is beyond the financial and technical capabilities of the utilities. A scaling down of future investments is thus inevitable. A tentative priority investment program described in Chapter IV should be based on the following considerations: assign highest priority to rehabilitation of existing facilities; strive for improved reliability of supply to main cities, which are also the main industrial areas; improve supply to Luanda by addressing the main problems in generation, transmission, transformation, and distribution; postpone most small projects in isolated systems, mainly for lack of managerial/technical staff, even if equipment has been purchased; postpone new rural/village electrification until hydro supply conditions have been improved and tariffs readjusted; limit new connections in cities until tariffs are adjusted and (especially in Luanda) until billing and collection procedures are substantially improved; and plan a substantial amount of technical assistance to support ENE task forces in big rehabilitation projects such as Lomaum and the Southern System. 41. A minimal priority investment progran. in line with the above priorities and considerations was prepared by the mission in collaboration with the staff and managers of the utilities and the staff of MEP. Given the above priorities and constraints, the mission sees no useful role for additions to capacity of the scale being considered at Capanda. The priority investment program should be carried out over the next five years and would cost about US$200 million (Kz 6-7 billion). This seems to accord better with the financial and managerial/technical possibilities of the subsector. However, it would still be a heavy financial and management burden on the utilities. 42. A general recommendation, in addition to the considerations listed above, is to subject every substantial project (say, exceeding US$2 million) to e:onomic and financial feasibility analysis. Capanda 43. The Government's apparent decision to advance the construction of a dam and power plant at Capanda presents several major issues. Although the analysis done is preliminary and conclusions should be taken as tentative, several robust conclusions emerge. First, Capanda  - xv - represents a significant departure from the lowest cost expansion path, even if it has not been updated recently. Second, the huge capacity (4 x 130 MW) planned for Capanda will probably not be needed until well into the next century. Third, it is a project which, by itself, will not improve the reliability of service in the Northern System and will not mitigate the problems of the other two systems at all. Fourth, actual and expected low demand growth rates a.d the availability of substantial thermal reserve would allow the postponement of this irreversible major investment decision during this period of uncertainty and stringent financial conditions, at a very low risk, until the economic environment becomes more stable and a better perception of the potential medium- and long-term demand is possible. Fifth, making,the investment in Capanda will add substantially to the public external debt burden (commercial financing). It may also undermine Angola's ability to finance the vital petroleum development program (on which its future export earnings depend) because part of Angola's future petroleum output has been earmarked as a repayment guarantee on some of the Brazilian financing for Capanda. Sixth, only about half of the financing required for the project is firmly in place. That is, out of a total cost optimistically estimated at about US$1.5 billion* about US$528 million has been secured on commercial terms from Banco do Brazil (and these funds will essentially run out by early 1989) for civil works, and about US$275 million have been firmly committed by the Soviet Union for electromechanical equipment only. Thus, only about US$800 million has been committed. For this reason, it might be preferable to stop work deliberately rather than wait until funds run out. An alternative and better justified project would be the rehabilitation of all existing systems and small-capacity additions. In any case, it is likely that the Covernment of Angola will be unable to raise funds to complete the project. Meanwhile, a consulting firm has been retained to reevaluate the Capanda Project, and the Government has requested World Bank comments on the resulting report. 44. In the final analysis, therefore, this report recommends that the existing least cost expansion plan be updated, based on the best available demand projection, so as to confirm the stage at which Capanda power should be developed. Alternative expansion sequences in the Northern System (with different timings for Capanda and complementary works in Cambambe) should be evaluated in full detail and in the context of the entire power subsector, with all economic and financial implications reassessed in a realistic framework of demand projections and updated costs. To assist in these tasks, this report includes Terms of Reference (see Annex 18) for the carrying out of a Power Subsector * This total cost figure is approximate but excludes the cost of a transmission line to Luanda and of transformers and substations. It also excludes physical and price contingencies. With all these elements, and assuming good cost control measures--which are not now in evidence--the overall cost might well exceed US$2.0 billion.  - xvi - Investment Review which would further identify and start preparing the rehabilitation of the three power systems, review or carry out the economic financial analysis of major projects and study the feasibility and costs of stopping the work on Capanda dam, protecting the works already executed and finding uses for materials and equipment already procured. Forestry, Woodfuels and Household Energy 45. While available statistics are few and unreliable, two recent studies and mission estimates have produced a picture of the situation of supply and demand for forestry and woodfuels in Angola which can be summarized as follows. 46. Most Angolans use firewood or charcoal for cooking and heating. In the cities, however, a significant minority use LPG. The aggregate consumption of firewood is in the order of 2.5 million tons/y and of charcoal about 0.5 million t/y, requiring a total removal of 6 million tons or about 10 million m3 of wood. Angola possesses some 50 million hectares of dense forests and a further 55 million hectares of woodland and savanna. Together these forests are capable of producing much more wood on a sustained basis than is at present consumed in the country. 47. Out of Angola's nine million inhabitants, almost half live in areas with more or less pronounced fuelwood shortages, either on the dry coast or in inland cities. In the shortage areas, the group hardest hit is the periurban population. They have limited access to alternative fuels (more easily available in urban centers) and, unlike most rural people, they cannot gather their own fuelwood for free. They are, furthermore, penalized by high market prices for woodfuels: the cost per thousand useful kilocalories is only Kz 10 for LPG but 10 to 20 times as much for firewood and charcoal. 48. The institutional framework for energy forestry in Angola is weak. Exploitation of fuelwood is regulated by the National Directorate for the Conservation of Nature (DNACO), which issues cutting licenses. The DNACO, however, has no resources to ensure that the actual cutting conforms to the licenses issued. 49. The creation of new forests is not the best (cheapest) way to solve the fuelwood problem. This is primarily because the dry coastal strip of Angola, where most of the people experiencing fuelwood shortage live, is poorly suited for tree-growing. 50. This report proposes several sets of priority activities at the regional and national levels. Four are regional in character while two are national. The first regional set of activities covers the provinces  - xvii - of Huila and Namibe, a region where the security situation is fairly good. It includes both city-oriented activities like the improvement of stoves and rural-based ones like improved supply systems for firewood and charcoal. The other three regional sets of activities all cover urban areas: one for Luanda; one for Benguela/Lobito; and one for Huambo township. For Luanda and Benguela/Lobito, it is proposed that emphasis be put on increased use of LPG as a domestic fuel. In Huambo township, improved stoves should be given first priority. 51. Two national activities should be carried out in support of the regional ones. One covers the development and introduction of improved stoves and the other concerns initial development work and trials in agroforestry. The activities listed, regional and national, have been grouped into four projects. They are the following: (a) a pilot project in Huila-Namibe, to integrate the various components of energy forestry, including the development of agroforestry; (b) improved cooking stoves, mainly for the urban and periurban populations in Luanda, Benguela, Lobito, and Huambo; (c) an improved supply system for woodfuels, mainly for the cities of Luanda, Benguela, Lobito, and Huambo; and (d) continuing, partial replacement of firewood and charcoal by LPG as a domestic fuel for the urban and periurban population on the coast, at least until more peaceful conditions improve supplies and lower the prices of woodfuels, and economic adjustment measures increase the prices of petroleum products (LPG, kerosene). 52. Table 1 presents in matrix form an overview of the various actions proposed in this report to enhance the development of Angola's energy sector. Each action is assigned a priority and a time frame for completion. The time frame is divided into short term (actions to be implemented immediately), medium term (actions to be implemented over two to three years), and long term (actions to be considered over a period longer than three years). Estimated costs are given for specific projects, where these are already known or have been calculated for the purposes of this report.  - xviii - Table 1: ANGOLA : ACTION PLAN FOR THE DEVELOPMENT OF THE ENERGY SECTOR Objective Action Cost Priority Time Frame a/ A, Electricity (1) Operational Increase technical assistance FIRST ST - MT strengthening and training in preparation of utilitles. for rehabilitation program. (2) Reorientation of Plan and execute a Power USS200 million FIRST ST - LT Investment policy to- Sector Rehabilitation (of which USSIO wards repair and main- investment program. million to tech- tenance of the sector's nical assistance) existing facilities. between 1988-92. (3) Financial recovery Raise tariffs 300-400% - FIRST ST-MT of the power sector. immediately. Design and implement gradually a more economically efficient tariff policy. Improved billing and revenue collection system. (4) Administrative and Except for system planning, - SECOND ST Institutional reform decentralize operations and of the power sector. maintenance. B. Crude Oil (1) Maintenance of Provide SONANGOL with a - FIRST MT - LT level of Investment greater managerial and required to prevent financial autonomy. the crude oil produc- tion rate from declining, (2) Strengthening of Establish an Improved - SECOND MT the supervisory role management information of SONANGOL. system; provide an incen- tive system to attract and retain specialized staff. (3) Improvement In Standardize the fiscal - THIRD MT - LT the competitiveness treatment of oil companies; of contractual terms revise outmoded legislation. for oil companies. a/ ST = short-term; MT = medium-term; LT 5 long-term.  - xix - Objective Action Cost Priority Time Frame a/ C. Natural Gas (1) Reduced dependence LPG fractionating offshore USS2-3 million FIRST ST on LPG imports. Cabinda to supply 30,000 t/y to the domestic market. (2) Increase In export LPG/condensates recovery in to be determined SECOND MT - LT revenues. Block 3. (3) Improvements In LPG bottle rehabilitation US$5 million SECOND MT the supply Infrastruc- plant in Luanda. ture for domestic LPG. (4) Improved power Dual fuel 15 MW thermal USS20-25 million SECOND/ MT generation In Isolated power plant In the Soyo area TN lIFD aceas. and rehabilitation of 10 MW gas-fired turbine in Cabinda. (5) Increase In LPG Onshore plant for LPG USS3-4 million THIRD ST - MT supply to domestic recovery at Malongo. market. D. Petroleum Products (1) Removal of a) Eliminate all direct - FIRST ST distortions in subsidies and adjust petroleum product refinery gate prices to and crude oil prices. higher levels. b) Increase the price level - FIRST In accordance with adjustments in exchange rate. (2) Increase In the Modify existing cost plus - MT efficiency of arrangement. refinery operations. (3) Reductlon In the SONANGOL to resume full res- - SECOND ST - MT foreign exchange ponsibility for procurement costs of petroleum of imports and cargo exports product Import/- (Low Sulfur Fuel Oil) and seek export arrangements. cheaper supply sources. a/ ST a short-term; MT a medium-term; LT = long-term.  - xx - Objective Action Cost Priority Time Frame a/ Petroleum Products (Continued) (4) Reduction in the SONANGOL to delegate the - SECOND ST - MT cost of domestic wholesale business to an petroleum product autonomous affiliate. distribution. Transport and retail business to be privatized gradually E. Woodfuels : Scenario : No Peace Coastal Areas (1) Improved fuel Increase the supply of USS5 or 6 million FIRST M4T substitution. Liquefied Petroleum Gas. (2) Increase In Produce and disseminate USS100,000 SECOND NT - LT end-use efficiency for improved stoves. fuelwood and charcoal. (3) Improvements in Develop selected measures USS75,000 THNIRD MT - LT the woodfuel supply designed to improve the infrastructure. organizational set-up, the operational efficiency and the feed stock extraction of charcoal production. Inland (4) Reduction of Introduce Improved stoves. USS600,000 FIRST ST - MT specific woodfuel consumption In urban concentrations. (5) Improvements in Develop selected measures Included In cost SECOND LT the woodfuel supply designed to increase the use above. system. of resources with marginal ecological Importance, to upgrade harvesting techniques and to maintain thA sustainability of supply. a/ ST n short-term; MT medium-term; LT a long-term.  - xXi - Objective Action Cost Prlority Time Frame a/ Woodfuels : Scenarlo Peace Coastal Areas: (6) Improvements In Develop selected measures - FIRST MT - LT the woodfuel supply designed to Improve the system. organizational set-up, operational efficiency and feedstock extraction of charcoal production. (7) Increase in the Increase the supply of LPG. - SECOND MT scope for fuel substitution. (8) Survey of Described in Chapter V. USS200,000 SECOND ST-MT woodfuel supply systems to Luanda. (9) Improvements in Produce and disseminate - THIRD ST - MT end-use efficiency. Improved stoves. Inland: (10) Improvements In Develop selected measures - SErOND ST - MT the woodfuel supply designed to increase the use system. of resources with marginal ecological importance, to upgrade harvesting techniques and to maintain the sustainability of supply. ;11) Improvements In Introduce Improved stoves. - FIRST LT end-use efficiency. a/ ST = short-term; MT = medium-term; LT = long-term.  I. ENERGY IN THE ECONOMY General Economic Framework 1.1 The People's Republic of Angola is located on the west coast of Africa, with Namibia to the south, Zambia to the east, and Zaire and the Congo to the north. The country also includes the enclave of Cabinda, which is separated from the rest of Angola by a corridor of Zairian territory and the mouth of the Zaire River. Angola covers an area of 1.27 million km2. Its land borders measure 5,070 km and it has an Atlantic coastline exceeding 1,600 km. The climate is tropical in the north, subtropical in the south, and temperate on the high plateau. In mid-1986, Angola's population was estimated at about 9 million (1970 census: 5.6 million), and the current rate of population growth is in the vicinity of 2.8%. While the overall population density is comparatively low, there has been significant migration to urban areas in recent years. At present, urban and periurban dwellers probably account for about 30% of the total population. 1.2 After independence in November 1975, Angola was left with a significant shortage of trained professionals and skilled workers of all types needed to undertake the formidable task of rebuilding the economy which had been damaged by the war and was collapsing after a mass exodus of the Portuguese settlers. However, the country also inherited: a well- developed transport infrastructure; a relatively diversified manufactur- ing sector (above all in consumer and intermediate goods); rich agricultural areas; a healthy mining sector; and (e) a sizeable enclave petroleum sector. After independence, the continuing civil war has seriously hampered economic development and remains the most serious obstacle to economic recovery. Economic policy since independence has been based on central planning and administrative controls. This economic system has had various adverse effects, including the emergence of a parallel market, severe price distortions, wasteful investment programs, undesirable effects on the distribution of real income, and a general lack of financial and fiscal discipline.l/ 1.3 Angola can be classified as a lower middle-income petroleum exporter. With an estimated GDP per capita of US$485 (1986) 2/ it ranks slightly above the average of the SADCC (Southern African Development Coordination Conference) countries. However, compared to other African countries with average hydrocarbon endowments (Tunisia, Gabon, and the People's Republic of Congo), Angola's GDP per capita is low 1/ For a comprehensive analysis of the Angolan economy, see the UNDP/World Bank report entitled, "Angola: An Introductory Economic Review", July 1988. 2/ All of these figures are subject to large errors, but they are quoted in this report as they are the only ones available.  - 2 - (Table 1.1). Moreover, despite large and increasing oil revenues, the country's GDP fell in real terms between 1980 and 1985. Angola is similar to other African countries with medium-scale hydrocarbon endowments in the sense that oil provides the lion's share of Government revenue (53% in 1985) and accounts for a significant portion of GDP (30% in 1985). As a consequence, the dramatic drop in international oil prices in late 1985 had a disastrous impact on the Angolan economy. Table 1.1: COMPARATIVE ECONOMIC INDICATORS, 1985 Share of GDP Share of Oil Income per capita Petroelum in Government Country (1985-86 US$) in GDP Revenues .^ngola 485 30% 53.1% Congo 1,110 40% 66.6% Gabon 3,350 45% 66.0% Nigeria 800 23% Source: Angolan authorities; World Development Report 1986; Mission estimates. 1A4 As indicated in Table 1.2, the post-independence development of the Angolan economy can be subdivided into three distinct periods. Between 1977 and 1981, real GDP grew at an average annual rate of 4.3%. This upswing, however, was short-lived and came to a sudden halt in 1981-82 when internal strife intensified and oil revenues dropped because of sagging international petroleum prices. As a result, between 1981 and 1983 real GDP fell at an average annual rate of 5.1%. A series of austerity measures were enacted while oil production continued to grow. These two events cushioned the fall in GDP in the period between 1983 and 1985, but the collapse of petroleum prices in late 1985 precipitated the sharp recession of 1986. The figures presented in Table 1.2 also underscore the dominant role played by the petroleum sector. Imports and Government expenditures both strongly depend on the country's petroleum exports. In 1985 these accounted for almost 96% of total merchandise exports.  - 3 - Table 1.2: ANGOLA: KEY ECONOMIC INDICATORS a/ (Percent Annual Change) Indicator 1986 1978-81 1981-83 1983-85 1986 Absolute Amount (in US$ million) Real GDP 4,409 +4.3 b/ -5.1 -1,7 -8.7 Government expenditure 3,110 +30.7 -9.9 +14.7 -14,1 Energy exports 1,150 +25.5 +5.7 +10.9 -39.7 Merchandise imports 1,062 +19.6 -19.1 +18.2 -23.3 a/ Least square estimates for average annual rate of change (%) in the value of above variables. b/ 1977-81. Source: Mission estimates (Annex 1), Petroleum and Public Finance 1.5 The most direct linkage between the petroleum sector's value- added (which accounts for about 30% of GDP) and the rest of the economy (the non-oil sectors) is the impact which oil revenues have on the Government budget and Government spending. Between 1980 and 1986 petroleum contributed on average about 63% of total tax income (or almost 52% of total Government revenues). In fact, in every year except 1986 the rise or fall in total Government revenues was almost completely attributable to a corresponding change in the level of tax receipts from the petroleum sector. The momentum of public spending proved hard to curb when oil revenues slowed down or declined. As a consequence, large budget deficits were incurred between 1979 and 1986 (totalling Kz 140 billion). This gives the impression that Government expenditure out of oil revenues was adjusted, but with significant lags. 1.6 The medium-term impact of oil revenues on GDP growth has been minimal because most of the revenue either financed current consumption (public, mainly defense, and private) or went into low-productivity public investment projects, some of which remain incomplete. Moreover, oil income allowed the country to maintain a highly overvalued exchange rate which undermined the competitiveness of domestic tradeables by making imports artificially cheap, put an upward pressure on the prices of non-tradeables, and thus tended to distort the structure of GDP. International Trade and Balance of Payments 1.7 The most striking feature of Angola's foreign trade is the steady increase in both the size and relative importance of the petroleum  - 4 - sector. The share of crude oil in total exports has increased consistently, climbing from about 30% in 1973 to 74% in 1980, and to more than 90% in 1986. Except for 1986, when oil income fell sharply, and 1981-82, when a slight temporary decrease was registered, oil export earnings have grown every year since 1978. In contrast, income from other export commodities such as coffee and diamonds--which used to account for a sizeable share of export revenues--has dropped continuously, and is today almost negligible. Since 1978 the country has always incurred a deficit on current account since the trade surplus (goods, especially crude oil, and non-factor services) and net unrequited transfers never offset the substantial--and steadily increasing--deficit on factor services. However, since 1978 Angola has invariably achieved a surplus on its capital a count. However, more than 80% of the inflows were accounted for by ftreign loans, and the small amount of foreign direct investment went mostly into oil activities. During the liquidity crisis of 1985-86, Angola's short-term foreign indebtedness increased significantly mostly because of increasing payment arrears which are recorded as short-term capital inflows. Policy Reforms 3/ 1.8 In view of the country's steadily worsening economic situation, Government authorities are seeking to develop and implement a series of policy reforms, such as granting more managerial and financial autonomy to State firms, relaxing price controls, limiting access to credit, making capital more costly, and devaluing the Kwanza. The proposed Program for Economic and Financial Reconstruction (SEF, Plano de Saneamento Econ6mico e Financeiro) contains different ideas and proposals, but the specific objectives, design, and timing of the reform are still under discussion. The measures proposed so far are expected to help eliminate the most harmful features of Angola's economic system, i.e., the lack of fiscal and monetary discipline, the distortions created by price controls, the excess supply of money, and the overvaluation of the Kwanza. Should oil prices recuperate even partially, the balance of payments constraint would be eased and rising oil revenues and manageable budget deficits should lead to a sustained recovery, especially in the context of a better incentive framework and a winding down of the war. Energy Sector Overview 1.9 By African standards, Angola is richly endowed with energy resources. The country has sizeable oil and gas reserves, a large hydro potential, and ample woodfuel resources. Currently, the proven oil reserves amount to 1,418 million bbl (sufficient to maintain the 1986 production level for the next 12 years), while natural gas reserves are 3/ These policies are described and analyzed in greater detail in "Angola: An Introductory Economic Review".  - 5 - estimated at 5 TCF. With its substantial oil and gas reserves, Angola ranks second only to Nigeria as an oil producer among Sub-Saharan countries. 1.10 In 1986, the primary energy equivalent of Angola's total commercial energy production amounted to 17.7 million toe, or 1.96 toe per capita. Only a few African countries, e.g., Libya, Algeria, and South Africa, recorded significantly higher figures on a per capita basis. If woodfuels are included (about 2 million toe of primary energy), per capita production was almost 2.2 toe. However, since more than 70% of the primary commercial energy production leaves the country in the form of crude oil and LPG (liquefied petroleum gas) exports, while 95% of the natural gas jointly supplied with crude oil is flared or reinjected, the final per capita consumption of commercial energy proves to be moderate. In 1986 it was 103 kgoe. This compares to 602 kgoe for Cabon (1985 figures), 151 kgoe for the Congo (1985), 142 kgoe for Sa6 Tome and Principe, and 24 kgoe for Mozambique (1984). If woodfuels are included, which accounted for 56% of the net domestic energy supply, final energy consumption for 1986 worked out to 297 kgoe. Accurate figures on the sectoral breakdown of final energy consumption are not available. However, in rough terms about 50% of petroleum products, which account for 95% of final commercial energy or 42% of total final energy consumption, are used in transport and for military purposes, whereas the lion's share of woodfuels goes to households. A summary of Angola's 1986 Energy Balance is given in Table 1.3. Table 1.3: SUMMARY OF ANGOLA ENERGY BALANCE 1986 (In *OOOs toe) Woodfuels Natural Crude Hydro Electricity Petroleum Total Gas Oil Products Total Production 2,074 3,418 14,102 173 - - 19,765 Total Available Supply 2,074 3,241 1,498 173 - - 7,080 Net supply available a/ 1,180 - - - 49 1,443 2,672 Final consumption b/ 1,180 - - - 49 879 2,108 a/ Adjusted for conversion losses and non-energy uses. b/ Adjusted for secondary exports and bunker sales. Source: Annex 3.  - 6 - 1.11 Table 1.4 summarizes the trends in commercial energy consumption. The figures indicate the extent to which final consumption of petroleum products and electricity have been decoupled from the overall performance of the economy, particularly in the 1980s. While real GDP declined, the consumption of petroleum products showed a sharp upward trend. Even in the aftermath of the oil price drop of 1985-86, domestic sales of petroleum derivatives continued to increase. On the other hand, the changes in electricity consumption appear to be more in line with GDP growth. Between 1977 and 1982, consumption rose in direct proportion to CDP, and then declined as the economic situation worsened. However, this is not evidence of causality. The drop in electricity consumption may well have been due to supply constraints. Since the share of low voltage consumers (households) in total consumption ircreased significantly in the 1980s to more than 50% (a trend reinforced bh extremely low tariffs), it appears that electricity ccnsumption was more or less determined by power generation (net of transmission and distribution losses). Similar arguments apply to petroleum products. Between 1980 and 1986, consumption grew most for LPG, kerosene, and jet fuels which are demand inelastic with respect to income and prices (household and military use). Thus, lack of correlation between commercial energy consumption and overall economic growth can be explained by the fact that official tariffs for power and prices of petroleum products had become insignificant and no longer served as rationing devices. Indeed, at negligible prices, consumption may prove perfectly elastic until it meets another constraint such as appliarces or supply capacity. Table 1.4: SUMMARY OF TRENDS IN COMMERCIAL ENERGY CONSUMPTION a/ Energy Type 1977-82 1982-85 1980-86 1986 Absolute Electricity +6.6 -5.5 -0.5 49 ('000s toe) Petroleum products -- +9.0 +6.6 915 ('000s toe) Total commercial energy -- -- +6.2 965 (OOOs toe) Total per capita consumption -- -- +3,3 110 (kgoe) GDP +4,3 -1,9 -3.9 485 (USS) a/ Least square estimates of average annual rates of change. Source: Angolan authorities and mission estimates (Annex 3). Petroleum Products 1.12 As can be seen from Table 1.5, Angola's structure of petroleum product demand is heavily biased towards middle distillates, which account for more than 75% of total domestic sales. Within middle distillates, gasoil plays a less important role than it does in other  - 7 - countries at the same level of economic development as Angola, indicating that demand is artificially low in road transport, industry and agriculture. On the other hand, kerosenes account for more than one quarter of the total domestic petroleum product sales. Agriculture accounts for only 2% of domestic sales of refined products, which is hardly surprising given the low level of commercial agricultural production and the fact that many rural areas are cut off from the country's transport and distribution infrastructure. Industrial demand depends on a small number of energy-intensive operations (refinery, extraction, cement) which account for about three-quarters of the country's boiler fuel consumption. The public sector accounts for more than 50% of the aviation fuels consumed in Angola. Household consumption of petroleum products--primarily gasoline, kerosene, and LPG--is strongly biased towards the urban centers. Thus, the regional and sectoral distribution of Angola's petroleum product consumption are highly distorted, i.e., they mirror the constraints imposed by both the civil war and the depressed state of the economy. Table 1.5: DOMESTIC PETROLEUM PRODUCT SALES BY SECTOR, 1985 (in %) Share In LPG/Kerosene Gas/Oil Jet Fue!s Gasoline Fuel Oil Total Sales Industry 3.8 26,2 10.3 1.7 52.2 20.6 Agriculture 0.7 3e1 - 0.5 6.8 2.2 Transport 0.2 27.1 47.1 0.8 39.3 28.8 Construction 0.2 4.8 - 0.6 0.1 2.1 Resale 77.2 20,4 - 42.2 - 19,6 Government 1.7 4.0 0.1 2.5 0.4 2.2 Defense & Security 3.1 10.1 42.0 49.8 0,3 21.1 Others 13.1 4.3 0.5 1.9 0.9 3.4 Share In Total Sales 8.2 42.6 26.7 10.7 11.8 100.0 Source: Annex 8. Crude Oil 1.13 In contrast to the rest of the economy, Angola's petroleum sector recovered rapidly after independence. Crude oil production climbed from 94,000 bbl/d in 1975 to 164,000 in 1979. Owing to the disruption of exploration and development efforts during the second half of the 1970s, a temporary decline in crude oil production was inevitable; but the trend was quickly reversed in 1981-82, and by 1983 output had re- bounded to its 1974 level. In fact, large investments undertaken in recent years have resulted in a continuous increase in oil production  - 8 - since 1982 (while the reserve-to-production rate was kept above 12 years), and this upward trend in output is likely to continue until the end of the 1990s. Except for 1986, output growth even offset the fall in international petroleum prices to the extent that between 1982 and 1985 nominal export revenues continued to increase (Table 1.6). Table 1.6: TRENDS IN THE PERFORMANCE OF THE PETROLEUM SECTOR a/ (Percent Annual Changes) 1975-79 1979-81 1981-85 1986 Crude Oil Production +16.6 -11.7 +18.0 +21.0 1978-80 1980-82 1982-85 1986 Export Revenues fromn Crude Oil +59.0 -5.8 +15.3 -39.3 1978-80 1980-82 1982-85 1986 Export Revenues as % of GDP +38.7 -6.3 +4.8 -33.4 Deflated Export Revenue +32.0 -6.4 +10.7 -41.2 from Crude Oil a/ Least square estimates of average annual rates of change. Source: Annexes 1 and 4. 1.14 Table 1.7 provides some rough indicators as to the relative magnitude of the resources which were required to keep the petroleum sector prospering. Between 1982 and 1986 the petroleum sector claimed an average of about 85% of Angola's Gross Domestic Investment ((DI). During the same period foreign direct investments in the petroleum sector accounted for more than 70% of the medium- and long-term capital inflow, while a rising share of the petroleum revenues was spent to finance the factor services required by the petroleum sector. However, these figures do not mean that in the absence of the oil boom considerable resources would have been available for non-oil ventures. On the contrary, given the extremely low level of national savings and the limited scope for productive investments in the non-oil economy, sectors other than petroleum would neither have been able to attract nor to absorb this external capital.  - 9 - Table 1.7: INDICATORS FOR PETROLEUM SECTOR CLAIMS ON RESOURCES (In percentages) 1982 1983 1984 1985 1986 Share of GDI in GDP 15.4 11.0 10.1 12.4 11.3 Investments In the petroleum sector as a share of GDI 83.2 88,2 76.8 87.2 86.6 Direct investments in the petroleum sector as a share of medium- to long-term capital Inflow 69.3 84.4 61.5 63.7 87.5 Factor income a/ transferred abroad as a share of oil export revenues 11.3 11.1 11.3 12.1 16.8 a/ Attributable to the petroleum sector. Source: Annexes 1 and 4 and mission estimates, 1.15 However, the overall performance of Angola's economy remained highly sensitive to changes in oil revenues. With rising oil revenues the petroleum sector increased its share in GDP, resulting in an oil-led growth of the economy. On the other hand, in periods of falling oil revenues (1980-82), the petroleum sector's contribution to the country's value-added tended to decline and, thus, to adversely affect GDP growth. Moreover, since 1982 when the growth of oil revenues reflected only increases in output, the foreign exchange costs associated with an increment in revenues tended to rise. As a consequence, an increasing share of the oil revenue had to be transferred abroad (16.8% in 1986) to pay for the factor services required for the development of new fields and for the additional costs of higher output (Table 1.7). 1.16 While this drawback is likely to lose importance with rising prices, another difficulty may arise in the near future. So far, the State-owned oil company, SONANGOL (Sociedade Nacional de Combustiveis de Angola) has had no problems raising the funds required to meet the financial obligations of its investment agreements with foreign oil companies. However, SONANGOL's total debt service outlays have steadily increased and in 1986 amounted to US$56.1 million, which is equal to roughly half of the company's average annual investment expenditures for exploration and development during the period 1980-86. Given the  - 10 - country's critical balance of payments situation it may well become more difficult to finance the exploration and development needed between now and 1990 to ensure that production continues to rise in the early 1990s. Thus, attention should focus on the question of how SONANGOL's financial position can be improved and to what extent the legal and fiscal framework needs to be modified to encourage new investments. Natural Gas 1.17 Angola has only begun to exploit its natural gas reserves. To date, efforts have centered on the gas which is produced in association with oil at a rate of about 1.34 MCF/bbl. Some 50% of the associated gas is put to productive use, primarily for lift and reinjection, with the remainder being flared. There is little evidence that the picture will change considerably in the near future. Some associated gas is recovered for LPG production off the coast of Cabinda. The scheme has been operating since 1983. In 1986 output was 177,000 tons of LPG, of which 168,000 tons were exported to Brazil. Revenues amounted to US$21.3 million and contributed 1.8% of the country's earnings from energy exports (crude oil, 95%; petroleum products, 3.2%). Electricity 1.18 Basically, Angola's power subsector consists of three systems. The largest one, in the north, currently accounts for about 80% of the country's electric power production. In 1986 the country's total installed capacity was estimated at 463 MW (62% hydro) of which 275 MW were firm. At that time, peak demand was about 150 MW. While total generation reached 754 GWh, consumption accounted for only 78% (590 GWh) of the energy generated. On a per capita basis, this works out at 66.3 kWh per annum, a consumption figure which is relatively low (in Mozambique, for instance, the 1984 per capita consumption of electricity was about 44 kWh, while Gabon--a country with a high degree of urbanization--recorded 756 kWh). 1.19 Since independence, the most pronounced decline in electricity demand has taken place in the industrial sector. In 1986, industrial consumption of electricity was only 200 GWh, (i.e., 17,000 toe, compared to more than 50,000 toe in 1974), while about 170,000 toe were used in the form of boiler fuels. The residential sector's electricity con- sumption has increased steadily since the mid-1970s. In fact, it has proved to be a stable component of the country's electricity demand. In the mid-1970s, electricity became essentially a free good, i.e., its demand was not limited in any way by the consumers' willingness or (income-dependent) ability to pay, but mostly by the utilities' ability to generate and distribute power. Thus, it would be difficult to estimate the effect on demand of the significant increases in tariffs required to restore financial viability to the utilities and to adjust the relative price of electricity to the costs of alternative sources of  - 11 - energy. At the current level of tariffs, for instance, the total revenues from electricity sales are just sufficient to finance the foreign debt service (denominated at the official rate of exchange) of the utilities. Woodfuels 1.20 Woodfuels play a significant role in the country's energy balance. While fuelwood is the dominant source of energy in rural areas, most of which are cut off from the supply of commercial fuels and electricity, charcoal is consumed by urban households, which have little or no access to LPG and kerosene or are not connected to the grid. In 1986, total woodfuel consumption was probably 6-8 million tons of fuelwood, equivalent to 2 million toe of primary energy (including the fuelwood equivalent of charcoaL) and accounted for 56% of Angola's final energy demand. While there is little doubt that the country's overall biomass resources are sufficient to meet aggregate woodfuel demand on a sustained basis, the rapid growth of cities and the supply constraints imposed b) the civil war have led to regional and local imbalances and shortfalls, particularly in urban coastal areas. However, these imbalances do not affect each and every household. In a broad sense, there are two categories of urban households: the privileged households, which have access to the official supply of cheap commercial fuels and electricity (if measured in terms of the purchasing power of parallel market income, then LPG, kerosene, and electricity could be considered free goods), and underprivileged households, which have no choice but to buy on the parallel market where charcoal is the dominant fuel. 1.21 Currently, Government can do very little to improve the energy situation of rural households, and even less to increase the flow of woodfuels supplied to urban areas. In the present situation, Government intervention is likely to exacerbate the danger of disruptions in supply. The Government should therefore continue a "hands-off" policy which would avoid additional frictions until improving security would permit the establishment of more active policies related to supply and marketing. Some small, focused projects or measures could bring some relief, especially in coastal urban areas. 1.22 Among the activities proposed for the urban areas, priority is given to fuel substitution and to stove improvement, both falling within the purview of DNRFE (Department of New and Renewable Sources of Energy). The two rural-oriented groups of activities--the pilot project for Huila and Namibe and the development of agroforestry--would on the other hand require some support from DMACO (National Directorate for the Conservation of Nature). This would require some strengthening of the field representation of DNACO in the Southern Region.  - 12 - Energy Demand Projections 1.23 Given the past and current distortions in energy consumption and production, it is difficult to forecast future energy demand in terms of historic trends, nor is it certain that demand will respond to changes in GDP or prices in a predictable way. Much depends on how long the civil strife continues and to what extent a return to peace will result in a gradual resumption of growth in the agricultural and manufacturing sectors. The hypothesis of peace in the early 1990s is the cornerstone of the most-favorable-base-case scenario. This scenario consists of an average annual rate of GDP growth of 2.5% for the period between 1990 and 1995, followed by a sustained upswing which would keep the growth rate above 5% during the second half of the 1990s. 1.24 On the energy demand side, it is assumed that until 1990 the growth of petroleum product consumption will slow down to about 2.5% per annum, while electricity sales are likely to recover, rising at an average annual rate of slightly more than 2%. During the first half of the 1990s, the return to peace will dampen petroleum product demand due to the significant decrease in military consumption. However, with the subsequent recovery of the agricultural and industrial sectors, the pattern of demand will not cnly shift to productive uses (transport, industry) but also rise at a rate (3%) exceeding that of the late 1980s. But even if the demand for petroleum derivatives continues to grow at 5% during the second half of the 1990s, demand would still not exceed the capacity of the Luanda refinery. Electricity consumption, on the other hand, is assumed to increase at an average annual rate of 6.5% (1990-95), followed by an even more pronounced growth of 11% in the late 1990s. 1.25 Table 1.8 provides a summary of the base case demand projections. As can be seen, the implicit elasticity of commercial energy demand, with respect to GDP, decreases from a high figure of 1.64 in the late-1980s to about 1.5 in the ei,rly-1990s (which is still high, reflecting a need to catch up). But .urther increases in consumption would eventually lead to an elasticity just above unity. Table 1.8: COMMERCIAL ENERGY DEMAND PROJECTIONS Average Annual Rate of Increase Level of Demand (%) (in 'OOOs toe) 1987-90 1990-95 1995-2000 1990 1995 2000 Petroleum products 2.5 3.5 5 1,000 1,190 1,500 Electricity 2.5 6.5 11 67 92 155 Total Demand 2.5 3.7 5.3 1,067 1,282 1,655 GDP Elasticity 1.67 1.48 1.06 - - - Source: Mission estimates.  - 13 - Institutional Framework 1.26 The Ministry of Energy and Petroleum (MEP, Ministerio da Energia e Petr6leo) has overall responsibility for establishing and implementing national policies in the energy sector. In the petroleum field, the MEP supervises the operations of SONANGOL, which in turn supervises the operations of the international oil companies and the refinery. The MEP also supervises the operations of the three power utilities--ENE (Empresa Nacional de Electricidade), SONEFE (Sociedade Nacional de Estudo e Financiamento de Empreendimentos Ultramarinos), and EDEL (Empresa de Electricidade de Luanda). Through the DNRFE (National Department of New and Renewable Sources of Energy) the MEP keeps abreast of developments in biomass and new and renewable sources of energy. 1.27 MEP is the result of a 1984 merger of the then separate ministries of Energy and Petroleum. Until 1987, the MEP had "central" departments which reported directly to the Minister (such as the Planning and Technical Departments) and "executive" departments (such as the National Department of [Oil] Transformation which supervised the refinery) which reported directly to one of the Vice-Ministers (for Energy or Petroleum). This organization perpetuated the split between Energy and Petroleum and is being abandoned, following the MEF's "Consultative Council" in November 1986. The new MEP will have only four National Departments, or "Gabinetes", one each for Planning, Technical, Legal, and Human Resources, and the Vice-Ministers will no longer have subsectoral responsibilities. Vice-Ministers will perform assignments at the request of the Minister (Annex 2 gives organization charts). This reorganization of the MEP entails a significant reduction in staff (from 360 to 200) and seems to be part of a Government-wide restructuring effort designed to cut administrative expenditures and to diminish sector ministry supervision over State enterprises. 3upervision by sector Ministries of the routine management of State enterprises is to cease, and only the central Ministries (Planning and Finance/Central Bank) will have supervisory functions over the finances and management of State enterprises. This, of course, is to lead to greater autonomy in the management of State enterprises, which is one of the main components of the SEF. 1.28 In addition to the MEP, the Ministries of Finance and Planning and the Central Bank play a role in the control/supervision of State enterprises. The Ministry of Planning has the last word on sector in- vestments, in the sense that only the Ministry of Planning can include such investments in the (annual or pluriannual) Plan. In fact, actual investments take place in a much more haphazard fashion and no institution seems to really control the process. The Ministry of Finance taxes and subsidizes energy sector enterprises widely. With the help of SONANGOL and the MEP, the Ministry of Finance oversees the financial operations of the international oil companies and assesses and collects the various taxes on oil operations (so-called "Special Regimes"). The Ministry of Finance also subsidizes the operations of the power utilities. Finally the Banco Nacional de Angola (BNA), the only real  - 14 - bank in Angola, has the final word on the allocation of foreign exchange. This is a difficult task, and it is not suprising that foreign exchange operations are slow, complex, and disrupt external procurement programs of the enterprises, whether for current operations or for investment. SADCC - Energy Technical and Administrative Unit (TAU) 1.29 The Energy Technical and Administrative Unit (TAU) is an entity set up by the Angolan Government, or more specifically by the MEP9 to discharge the responsibility for energy sector coordination which SADCC (the Southern Africa Development Coordinating Conference) assigned to Ang^la. The TAU reports to the Minister of Energy and Petroleum of Angola, as does any other department of the MEP, but its roles do not concern Angola specifically. According to SADCC documents, the main purpose of the Energy TAU is to develop a regional energy development, conservation and security plan. 1.30 In practice, the TAU has gone about its work by establishing a portfolio of regional energy projects which can be classified under three headings: (a) national pilot projects, the results of which are expected to be applicable to other countries; (b) projects which benefit more than one country (i.e., "regional" projects); and (c) projects which support other regional projects. Although the range and diversity of projects promoted by TAU is striking, the electricity subsector is dominant. taking up to 50% of TAU's time. 1.31 The Government of Angola provides most of the TAU's funds, personnel and physical facilities. About a dozen Angolan professionals work for TAU. In addition to sizeable contributions in kind (offices, office equipment, vehicle maintenance), the Angolan Government also provided Kz 32 million in 1986 and Kz 33 million in 1987 (i.e., a little over US$1 million per annum). In addition, the TAU has attracted considerable donor support, especially from Europe (Belgium, Norway, EEC) and also from Canada. Norway provided US$300,000-400,000 per year in 1986 and 1987 while the Canadian International Development Agency (CIDA) provided about US$100,000 in each of those years. Funding for expanded activities was under consideration in late 1987, especially by CIDA, while Norway was preparing a study to redefine its support policy toward TAU. The TAU and the Angolan MEP also requested ESMAP support to conduct an evaluation of the TAU's mandate and achievements. This evaluation is currently in preparation with its report due to be released in mid-1989.  - 15 - Manpower, Technical Assistance, and Training 1.32 The shortage of qualified manpower has been a ubiquitous problem in Angola since independence. Most shortcomings in the execution of basic policies can be ascribed to the shortage of competent staff. Similarly, poor performance by State enterprises can partly be blamed on a lack of qualified personnel to properly plan, evaluate, execute, and operate public investments. In addition, qualified Angolan staff command very high wages in the service of international oil companies. While the civil service can hardly compete with oil majors, the problem of providing qualified staff with suitable incentives is a pressing one and has yet to be satisfactorily resolved. 1.33 Responses to this problem of staff shortage have varied among different organizations, as follows: (a) the Central Government is making use of bilaterally-supplied expatriate experts (from the U.S.S.R., Cuba, Eastern Europe) and gets significant help from the U.N. system in addition to having first claim on the supply of qualified Angolans; (b) the petroleum subsector has access to the qualified manpower of the international oil companies, uses high level consultants, and invests heavily in training for its own needs anl generally; (c) the power subsector has essentially concentrated on day-to-day operations (with lower service standards) while using some expatriate assistance (from Cuba, the U.S.S.R., and Portugal). Activities that require much highly qualified manpower (planning, ta-iffs, demand studies) are simply deferred; (d) in forestry, the level of upstream activities is at a virtual standstill with almost no foresters deployed while some forestry technicians are being trained in Cuba (mainly for logging and sawmilling, areas that have attracted external capital and technical assistance from Cuba and Italy); and (e) in addition, a number of individual technical and professional personnel are hired and paid directly by the Government (the "Cooperantes") but with recent budget strictures this form of technical assistance, which is often used in administrative posts rather than in advising policymakers, is declining. Training for the Petroleum Subsector 1.34 In the petroleum subsector, the effect of qualified manpower shortages has been mitigated by recourse to the expertise of the international petroleum companies and the acquisition of staff services  - 16 - from abroad. SONANCOL has effectively and advantageously used external consultants to supplement its capabilities in all aspects of its opera- tions. There is little doubt that it should continue to do so. It might be possible, however, to reduce the cost of this option by turning to concessional sources of technical assistance, such as the United Nations Development Programme (UNDP), the European Economic Commission (EEC), the World Bank Group, or bilateral aid (Canada, Norway) for at least part of its needs. 1.35 The petroleum subsector contributes (by law) to the financing of nation-wide training programs. All petroleum companies are required to pay (US$0.15 per produced barrel) into a training fund which is controlled by the Ministries of Finance, Education, and the MEP. At current rates of production, (i.e., 400,000 bbl/d) this fund accounts for about US$21 million in hard currency. 1.36 Control over this fund appears to be diffused and no mechanism exists to ensure that the highest priorities in training are identified and then funded. In addition, the contribution is paid to the State budget in hard currency, but is not available in hard currency to potential users, who must follow normal, lengthy procedures to obtain needed foreign exchange. At times, pressing needs for foreign exchange may result in a total diversion of foreign exchange to these other needs, at the expense of training. 1.37 Other training programs for the petroleum industry include: on-the-job training wit, the petroleum companies; company-owned training centers (at Malongo for CABINDA GULF; at Soyo for TEXACO); and the Sumbe Petroleum Training Institute, which is supported by UNDP and Norway and managed by COMERINT, a consulting firm belonging to the ENI Group, Italy's State hydrocarbons holding company (this Institute also serves the needs of other SADCC countries). According to SONANGOL, training opportunities in the above programs favor engineers, technicians, and skilled laborers rather than managers, economists, or accountants and other financial staff. This presents a problem because SONANGOL itself needs mostly the latter type of qualified staff, including engineers, as it does not yet operate any oil field or other facilities. Expatriate staff resources are also used in the management and operation of support/logistical facilities such as the Kwanza Offshore Petroleum Operations Support Base in Soyo, at the mouth of the Zaire River. Training for the Power Subsector 1.38 In 1986, the electric power utilities, excluding the Office for the Harnessing of the Middle Kwanza (GAMEK, Gabinete de Aproveitamento do Medio Kwanza) employed about 4000 4/ persons, of which about 100 were expatriates. The expatriate staff cost roughly Kz 3,200/man/month 4/ Of the Angolans, 30 are higher-level technical or professional staff, 200 are technicians, and 3,600 are skilled, semi-skilled, and unskilled workers. Of the 100 expatriates, 21 are higher-level technical/professional staff, 30 are higher-level technicians, and 50 are skilled workers.  - 17 - (US$1,100/man/month). This relatively low figure reflects the fact that most of these expatriates originate from Cuba or Eastern Europe (where qualified manpower is plentiful) and that many are relatively low-level technicians or workers. 1.39 According to available information, higher-level manpower is extremely scarce and badly deployed. A disproportionate share of higher- level staff is concentrated in ENE's head office in Luanda while, for example, tne Southern Grid does not have even one Angolan engineer or professional manager. The situation is similar in other utilities, with distribution companies--such as EDEL and CELB (Companhia Electrica do Lobito e Benguela)--usually worse off. In addition, expatriate staff tend to work in a vacuum because of the shortage of Angolan counterparts. No skills/knowledge can therefore be transferred. 1.40 A major investment in a training school for the electricity subsector is about to be made with financing through a FF 60 million (about US$11 million) loan from the Caisse Centrale de Cooperation Economique. This financing covers school construction and equipment, curriculum preparation, and an initial infusion of teaching staff to train future teachers, but not housing for the teachers. This could be a problem because the school will be located near Mabubas, far from any reasonable supply of housing.5/ In addition, most of the financing will go for state-of-the-art buildings and equipment. The new training institute's main objective is to upgrade the skills of the workers of power utilities. This priority fits well with this report's recommendations. However, it might have been possible to set up a more cost effective system for the same purpose. 1.41 A sizeable influx of financial and staff resources would be required to maintain the existing assets of the power subsector. The technical assistance needed for minimal improvements in power subsector efficiency would entail roughly 35 man-years of technical assistance. This level of technical assistance would have to be maintained for at least three years during which time trained Angolan staff could be hired and receive on-the-job-training from expatriate staff. A three-year power subsector rehabilitation project would thus require US$10 million for technical assistance in addition to the need to finance pressing imports of materials, supplies, and spares. Major rehabilitation investments would require much larger sums. Angolan Development Strategy in the Energy Sector 1.42 The role of energy in the Angolan economy has been highlighted in earlier sections: energy exports provide Angola with resources to keep the economy and polity functioning. In fact, the abundance and rapid development of Angola's oil reserves can be said to have financed the political survival of Angola over the past decade. 5/ The location of the school is being reevaluated. It may, in fact, be built in Luanda. This would mitigate the housing problem.  - 18 - Petroleum Development Strategy and Peace 1.43 Angola's choices in future development of its oil resources depend, at least in part, on the evolution of internal stability. If the civil war persists, then, realistically, Angola has no choice but to continue developing and exploiting its resources as fast as it can. This has been the policy over the last decade. If civil peace returns soon, paradoxically Angola could have more leeway in the speed of development of its oil resources. In the event of immediate peace, Angola's priority should be~ to carry out a structural economic reform to set the stage for a resumption of growth in agriculture and industry rather than to continue rapid expansion of oil exports. Energy would, in that situation, play its normal role of supporting growth rather than leading it. 1.44 While expenditures on petroleum exploration and development should be maintained--lead times are lengthy in the petroleum industry-- the next burst of activity in petroleum should take place when Angola will be able to "sow" petroleum revenues most readily. That is, after substarntial reforms permit (public and private) investment to become productive once more, and when qualified manpower supplies have increased through lower military needs, then there could be a higher output of Angolan training institutions and a better framework for the effective use of external technical assistance. in essence, the Government's first priority with the return of peace should be to give petroleum second priority. SONANGOL 1.45 In the immediate, the Government should continue perfecting SONANGOL--the effective instrument it has slowly created to promote and control petroleum development--under general State guidance. SONANGOL needs continued slow growth in its capabilities in both oil and gas and increased financial and managerial autonomy. Economy-wide measures under consideration to increase the autonomy of State enterprises should be extended to SONANGOL as soon as possibie, keeping in mind that SONANGOL is in a position to benefit from these measures immediately, as its management, staffing, structure, and mandate are functional and undisputed, while many other enterprises are not in such a position yet. Refining and Product Supply 1.46 The present situation in refining and petroleum product supply is broadly satisfactory. At current relative world prices for crude oil and products, the FPA (Fina Petroleos de Luanda) refinery at Luanda appears to be economic. In addition, with recent de-bottlenecking and life prolongation, the refinery can supply most of Angola's needs for quite some time, especially if peace returns soon. Thus, Angola should  - 19 - continue satisfying its excess demand through imports while attempting to reduce wasteful or low priority consumption through a substantial increase in prices and the elimination of subsidies on kerosene, gasoil, and LPG. This would give prices a role in controlling demand and would restore a price structure more in line with economic costs (i.e., world market prices, CIF or FOB. Angola, depending on whether Angola is an imoorter or exporter of the given product). In view of the above, large new investments in expansion of refinery capacity should be postponed as they have low priority. Similarly, the expenditure of large sums to buy a majority or the totality of the stock (or assets) of the refinery cannot have much of a priority when both the State and SONANGOL are in a period of financial stringency (although if the transfer could be accomplished without the disbursement of large sums, as through a reordering of assets and liabilities between SONANCOL, FPA and the Government, then this would not cause any problem). An equivalent amount invested in petroleum exploration or field development would have a much higher economic return. Independently of who owns the refinery, an incentive framework designed to lower costs at the refinery seems to be Angola's best option to reduce the economic costs of supply of petroleum products in conjunction with the elimirnation of the subsidy on crude oil for domestic refining. A similar framework should be extended t SONANGOL's domestic distribution operations. Power 1.47 With few exceptions, Angola's power infrastructure has suffered from more than a decade of active neglect. Fortunately, the demand for power stagnated during this period, with households the major active consumers of power. This allowed the utilities to maintain a reasonable level of service. Since power has become essentially free after the monetary inflation of the mid-1970s, consumption has been supply- constrained. While the quick return of peaceful conditions might make the agenda for the power subsector somewhat easier to implement, it will remain essentially unchanged over the better part of the next decade. Assuming that power is to be sold at a price covering economic costs of supply, household demand will probably not grow very fast, and increased demand will only come about through the reactivation of industry, which will take time. 1.48 The tasks facing Angolan authorities in power would therefore be: (i) maintenance of a reasonable level of service without sizeable new investments; (ii) launching of a rehabilitation program for existing dams, power plants, and transmission Lines; (iii) reorganization of the sector in a more decentralized structure better able to improve efficiency; and (iv) setting the bases for future growth by improving management, accounting, and finances through a sizeable infusion of technical assistance, simultaneous with a major training effort and an increase in tariffs. Addition of new capacity should be very low on the list of priorities at least until the mid-1990s. Mission demand  - 20 - projections show that new capacity (above and beyond existing rehabilitable capacity) would probably not be needed until the latter part of the 1990s even under sanguine assumptions about the pace of the economic recovery (Annex 14). Household Energy 1.49 The Government can do very little to improve the energy situation of rural households, especially at times of civil strife. The cities are presently receiving both commercial (LPG, kerosene, electricity) and traditional (fuelwood, charcoal) energy products. Major hardships would be caused by interruption of either source of energy. The Government should therefore continue its "hands off" policy to ensure that supplies of traditional fuels continue to reach the cities. This is not the time to crack down on wood/charcoal truckers or merchants, nor is it the time to try to enforce cutting regulations. Efforts at managing forests, charging user fees, ensuring a competitive supply system, and establishing reserves will be needed, but this should await the return of more normal conditions, especially in the countryside. In the meantime, some improvements in the distribution of kerosene (especially in suburban and rure.l areas) and LPG (mostly in urban areas) should be carried out. 1.50 Angola's forest resources are sufficient, and sufficiently broadly distributed, to satisfy the needs of the population in normal situations. After the return of peace, the supply of woodfuels to all potential users should be relatively easy to assure. The Government should, at that time, establish basic policies that promote competitive woodfuels markets. This would ensure the lowest possible prices to users, while legislation, management, and supervision of forests will ensure that the ccsts of reforestation (i.e., value of the trees themselves) are taken into account by the market. 1.51 Kerosene and LPG are excessively undervalued in relation to woodfuels. With more reasonable relative prices for commercial and traditional sources of energy, (i.e., with the removal of security risks and premia in the supply of woodfuels and the removal of subsidies to LPG and kerosene) it is not certain whether consumption patterns, even in the cities, would favor modern commercial fuels as much as they do now. This situation should be kept under review, and excessive consumption of modern fuels (which can be exported) should perhaps not be promoted as aggressively as has been done over the past several years. The proposed LPG fractionation project (offshore Cabinda) would add about 30,000 t/y of LPG to domestic supplies. This will eliminate high-priced LPG imports (about 10,000 tons costing US$2.5 million) yet permit a 60% increase in domestic consumption over a short period. By the time this increased supply reaches the market, prices should have been increased substantially, or the amount of Government subsidy would also have to increase.  - 21 - II.A. CRUDE OIL: UPSTREAM ACTIVITIES Summary and Recommendations 2.1 Angola depends on oil income economically and politically. All matters relating to oil are therefore of extreme concein. General Government policies on oil development have been enlightened, and thus deservedly successful. The State gave its oil enterprise, SONANGOL, the most important responsibilities in oil development by making it sole concessionaire for all hydrocarbons. A successful modus operandi was worked out between SONANCOL and the MEP in supervising oil activities in Angola. As the business arm of the Government in oil, SONANGOL will play a significant role in the development of Angola as a whole. Recommendations for SONANGOL are as follows: (a) Although the legislation establishing SONANGOL empowered it to undertake virtually all activities related to oil (exploration, production, transport, refining, distribution), in addition to general supervisory responbilities, it really has two major tasks. The first is to encourage foreign investment in oil exploration and production under general Government guidance and in accordance with the existing legislation, and to negotiate advan,ageous contracts with internal oil companies. The second is to contract, supervise, and control foreign oil companies in exploring and developing the country' s oil reserves and to raise the funds required to meet its share of investment programs. Therefore, the performance of SONANGOL should be judged on its success in mobilizing and steering external resources into oil operations and in supervising and controlling foreign oil companies rather than in terms of its performance as an operator or a project manager. For the same reason, SONANCOL should minimize its involvement in upstream operations and other extraneous activities unless these strengthen its supervisory role as the 'trustee" of Angola's petroleum reserves. Similarly, the domestic distribution and marketing operations of SONANGOL could be spun-off into a relatively autonomous division or subsidiary, if not privatized; (b) SONANGOL should be granted greater managerial and financial autonomy, particularly in the area of joint venture agreements, to improve its ability to carry out its responsibilities. Although it is inescapable that SONANGOL should remain dnder the political control of the State, it should not be unduly constrained by the short-run difficulties of the Ministries of Finance and Planning. SONANGOL should be allowed to retain a reasonable part of its after-tax cash flow; (c) SONANGOL's ability to raise the funds required to meet its financial obligations will be highly sensitive to the overall  - 22 - financial health of the country. Therefore, the Government should not routinely use oil as collateral for external loans, as this could undermine the present creditworthiness and future profitability of SONANGOL; (d) SONANCOL's ability to adequately control and supervise the activities of foreign oil companies can be improved through the establishment of a more efficient management information system. Emphasis should be given to procedures which systemize and "digest" the flood of information SONANGOL receives. Uniform accounting procedures should be established for all oil operations; (e) Training of qualified staff, in conjunction with the develop- ment of an incentive structure for skilled personnel, is an important task. For technical staff, a more efficient utilization of existing training opportunities and institutions should suffice, rather than the creation of new or specially designed programs. For managerial and highly specialized technical staff, a suitable incentive system could play an important role in attracting and retaining such personnel. Technical assistance, however, would still be needed both to carry out complex tasks and to help train newer stafF. 2.2 Even though the contractual terms for oil exploration and development are not particularly soft, many international oil companies have been attracted to Angola. The geology is highly prospective, the investment-cost-to-production ratio is low, and the operating costs are moderate. In addition, the Government and SONANGOL have shown a practical, business-minded attitude. While the State owns the hydro- carbon reserves, SONANGOL, as sole concessionnaire, has been doing most of the work required to attract companies and investments. The model contract has been conceived as a framework with room for negotiation. Thus, major changes in the legal framework are not needed. 2.3 So far, the taxation system has worked well. Government tax income from joint ventures was based more on net profits than on royalties. The progressivity of the tax legislation allowed the Government to capture windfall profits, while oil companies were protected against a profit squeeze in periods of declining oil prices. So, output rose steadily from 1981 to 1986 in spite of stagnating or declining international oil prices. It would be counterproductive to make fundamental changes in the fiscal terms which apply to oil companies. Minor modifications, however, may be useful. 2.4 Part of the existing legislative framework (especially civil and commercial laws) was established in the 1950s and, therefore, does not match the contractual approach and the taxation system embodied in the more recent joint venture and production-sharing agreements (PSAs). Moreover, differences in the contractual terms for joint venture and PSA operations (eog., tax base) have led to an unequal fiscal treatment of  - 23 - production revenues. Although the Government is ready to deal with these problems pragmatically, it might be preferable to make small adjustments to the legialation and the contracts. 2.5 Some companies have requested softening of the contractual and fiscal terms for commercially unattractive discoveries. Urgent revenue needs and high rates of time preference probably argue in favor of careful, controlled development of those marginal fields. Nonetheless, the Government should be aware of the trade-off between the short term benefits (revenues) and the possibly higher (though uncertain) future revenues which it would forego by developing these fields now. In any case, the issue of permitting development of these marginal fields is not urgent and Angola should carefully weigh the pros and cons of modifying-- strictly for this purpose--a fiscal framework that has been and is working well. 2.6 Balancing the maximizing of Government revenues with attracting foreign oil companies is a difficult task. The best strategy to maintain competitiveness (which Angola follows) is to give foreign oil companies a stable and well-defined contractual framework, leaving sufficient incentives to attract risk capital. Contractua, arrangements should be reliable and flexible, rather than generally hard or liberal. Angola's framework for petroleum activities meets most of these criteria well. Oil Exploration and Production History 2.7 Intermittent oil exploration in Angola began in 1910 and concentrated on the lower Congo and Kwanza basins, but the first discovery was not made until 1955 by a subsidiary of PETROFINA. Production from the Benfica field started in 1956, while PETRANGOL--the name under which PETROFINA was reorganized in 1957--continued its onshore exploration activities. A major breakthrough came in 1966, when GULF OIL (which had entered Cabinda in 1957 through its subsidiary, CABGOC) discovered the first offshore field (Malongo), the reserves of which proved substantially larger than the delineated onshore deposits. During the 1960s, several other international oil companies initiated exploration activities, but CABGOC's Cabinda finds remained Angola's most important source of oil. Total production rose to 49,000 bbl/d in 1969, and further increased to 163,000 bbl/d in 1973. Ninety percent of the output was exported; the remainder was used as a feedstock for the Luanda refinery which came on-stream in the late 1950s. 2.8 After independence, CABGOC--by far the largest producer--ceased its operations, and crude oil production collapsed. This was reversed when CABGOC returned to Angola in 1976. However, uncertainties about future Government policy towards oil tended to reduce the company's propensity to invest in the development of existing fields and the exploration of new prospects. As a consequence, oil output ceased growing in the late 1970s. The Government quickly responded to the new situation and implemented a series of measures to improve the  - 24 - institutional and incentive framework of the petroleum sector. The measures included: (a) the establishment of SONANGOL (1976) as the business arm of the Government to coordinate and control petroleum activities; (b) the enactment of the Petroleum Law (1978) which made SONANGOL the sole concessionaire for oil exploration and production in the country (see Annex 5); (c) the renegotiation of CABCOC's and PETRANGOL's concessions (1978), giving SONANGOL a 51% share in the existing productive operations; (d) the division of the continental shelf into 13 blocks (1978) to be offered to interested oil companies under terms of PSAs; and (e) the creation of the MEP (Ministry of Petroleum) (1979). The reorganization of the petroleum sector, the new legal and fiscal framework, the comparatively low operating costs prevailing in Angola, as well as favorable prospects for new discoveries attracted numerous international oil companies, which has led to substantial investments and to continuously increasing production since 1982. This activity has taken place despite the recent decline in international oil prices. In 1988, production reached 441,800 bbl/d, 157% above the 1974 level. oil Production and Investment 2.9 As of June 1987, 7 of the 13 blocks (of about 4,000 km2 each) making up the country's offshore area (except Cabinda) had been awarded to oil companies. Table 2.1 summarizes production and investment activities in the different areas between 1980 and 1986. Table 2.1: OIL PRODUCTION AND INVESTMENTS IN THE PETROLEUM SECTOR Total Production Total Production Total Investments 1980-86 1986 1980-86 (Million tons) (%) ('OOOs bbl/d) (%) (Million USS) (%) Cabinda 45.0 69.8 190 67.4 816.4 30,0 Congo Onshore A 0.4 0.6 1 0.4 52.9 2.0 Congo Onshore B 11.3 17.5 32 11.3 189.1 6.9 Block 1 - - - - 216.1 7.9 Block 2 2.7 4.2 6 2.1 493.0 18.1 Block 3 3.7 5.7 50 17.7 878.4 32.2 Block 4 - - - - 78.0 2.9 Kwanza 1.4 2.2 3 1.1 TOTAL 64.5 100 282 100 2,723.9 100 Source: Annex 4.  - 25 - 2.10 More than 60% of total investments (US$2.7 billion) undertaken between 1980 and 1986 went into Cabinda and Block 3. The largest share was accounted for by Block 3 which started production in 1985 and, by 1986, already contributed 17.7% of the country's total petroleum output. While investments in Block 3 concentrated on exploration (until 1983) and development (since 1985), investmeLits at Cabinda focused on increasing production from proven reserves, rather than on discovering new deposits. In terms of investment outlays, Block 2 ranks third. 2.11 In spite of significant expenditures to explore and develop new fields, the performance of Block 2 has been disappointing until recently. Production fell from a peak of 13,000 bbl/d in 1982 to 6,000 bbl/d in 1986. Many of the new discoveries are considered marginal so that some companies called for a softening of fiscal and contractual terms to encourage further development. These claims, however, appear to be exaggerated. Government has already written off a considerable share of its (potential) revenues by allowing the companies to take more "Cost Oil" than stipulated in the contracts. In addition, more recent finds are reported to be more promising and may reverse the downward trend of the past. In fact, 1988 output increased to about 35,000 bbl/d. 2.12 Other problem areas are the onshore Congo and Kwanza basins. Of the nearly US$440 million spent on these areas between 1980 and 1986, almost US$150 million was used for the development of the onshore B area. Since 1982, however, the onshore B output has been stagnating (in the vicinity of 30,000 bbl/d) while total onshore production (including onshore Congo A and Kwanza) declined from 56,000 bbl/d in 1977 to 36,000 b/d in 1986. The investments made in Blocks 1 and 4 were exclusively for exploration, with results still in the future. Thcugh almost US$200 million was spent for exploration in Block 1, only marginal discoveries were made. The companies involved have already asked for fiscal incentives to develop fields which, at present prices, are unprofitable. Exploration rights in Block 4 were awarded in 1984, but no significant discoveries have been made yet. 2.13 In 1987, some 16 foreign companies were engaged in Angola's petroleum industry and others were queuing to get in. Thus, the role to be played by SONANGOL as the sole oil concessionaire is becoming more important. SONANGOL's first joint venture was formed with CABGOC in 1978, giving SONANGOL a 51% share in the Cabinda offshore. CABGOC remained the operator, and the Association successfully embarked on a five-year investment program (1980-85) to develop production from proven deposits. CHEVRON, which took over GULF OIL in 1984, continued to invest in Cabinda, but dropped GULF's plans to reduce its share in some deeper areas.6/ In 1978, SONANGOL also obtained a 51% participation in the Congo/Kwanza onshore areas formerly held by PETRANCOL. When FPA succeeded PETRANGOL and became operator, it retained a 49% share in the 6/ However, for other reasons, CHEVRON-GULF reduced its overall share in the Cabinda Joint-Venture to 39.2% by farming out a 9.8% participating interest to AGIP.  - 26 - onshore A areas and a 32% share in the onshore B area, while TEXACO kept the remaining 16.4%. As for new blocks, SONANGOL contracted several foreign companies into PSAs under the 1978 Petroleum Law. In Block 2 it acquired a 25% working interest in an association in which TEXACO is the operator, thus providing SONANGOL with a 25% share of production in addition to the share it gets as concessionaire. Since 1984 it has also kept a 20% interest (on a carried basis) in an association led by BRASPETRO (PETROBRAS) in Block 4. Also in Block 4, SONANGOL owns 51% of the mixed company Empresa de Servicios Petroliferos de Angola (ESPA) which has operational responsibilities. SONANCOL does not hold any shares (interest) in Block 1 (AGIP is operator), Block 3 (ELF is operator), and Block 5 (which was awarded to a group of companies formed by CONOCO in 1986). It must, nonetheless, as concessionaire, supervise the activities of the operators. Institutional and Fiscal Framework 2.14 As the agency responsible for the overall coordination and development of the energy sector and the implementation of national energy policies, the MEP also has responsibility over the policies and performance of the petroleum subsector. In particular, the Director General of SONANGOL reports directly to the Minister of Energy and Petroleum. The MEP receives (through SONANGOL) detailed information and itemized financial statements on all oil-related activities to control and coordinate ongoing operations. In performing these functions, it can resort to external assistance, be it from SONANGOL or foreign consultants. 2.15 The division of tasks between the MEP and SONANGOL is as follows: The MEP is the only organ competent to decide on: (a) the authorization to open blocks for bidding, the commencement of production (including the field-specific production levels), and the flaring of gas; (b) the approval of development programs; and (c) the determination of reference prices for tax purposes. The MEP's oversight responsibilities are not meant to intrude into the day-to-day management of SONANGOL. So far, the division of labor between the MEP and SONANGOL has worked well. But there could be complications since the MEP's ability to supervise petroleum operations depends, to a large extent, on SONANGOL's ability to provide the necessary information in a timely fashion.  - 27 - 2.16 SONANGOL's responsibilities include: (a) the collection and compilation of technical and geological data prior to distinct exploration activities; (b) advisory assistance to the Government; (c) the opening of blocks for bidding and the evaluation of proposed work programs; (d) the negotiation of (production-sharing) contracts; (e) commencs and suggestions on exploration activities carried out by foreign companies (i.e., participation in Exploration Advisory Committees); and (f) the approval and auditing of all activities which follow a commercial discovery. The latter function is executed through SONANCOL's participation in so- called Operating Committees (staffed with two representatives of the contractor and two SONANGOL members, with SONANGOL appointing the voting chairman) which monitor, control, and regulate the technical and financial performance of the contractors. In addition to ex ante supervision and approval faculties, the production-sharing contract also empowers SONANCOL to undertake ex post auditing and evaluation of past activities, facilitated by reporting obligations on the part of the oil companies. 2.17 However, in contrast to SONANGOL's powerful position within the petroleum subsector, its financial latitude is strongly circumscribed. About 50% of its amortizations and 95% of its profits accrue to the Treasury, so that investment decisions are subject to the Budget reallocating funds to SONANGOL. 7/ Though in practice only the balance is transferred to and fro, this makes SONANGOL subject to the priorities of the Ministries of Finance and Planning and leaves little financial autonomy. SONANGOL's dependence on the fiscal/budget authorities has led it to rely on foreign oil companies to assist it in raising the funds required to meet its financial commitments. Skillful balancing of these two options has allowed SONANCOL to acquire the resources to finance large-scale investment programs (which, between 1980 and 1986 accounted for more than US$1 billion). By these means, SONANGOL succeeded in financing almost 25% of the total expenditures for exploration and development of the country's petroleum reserves over the seven-year period (Annex 4, Table 2). However, a higher degree of financial autonomy would help SONANCOL fulfill its obligations in the future 7/ Changes to increase the autonomy of State enterprises are being considered under the SEF program of structural adjustment. If these changes were approved and enforced, SONANGOL would no longer have to transfer any of its depreciation allowance but would only be subject to a corporation profit tax.  - 28 - development of the petroleum subsector, now that the State Budget is tight and the finances of the oil companies are less buoyant because of low world prices. 2.18 In the past the main advantages of Angola's petroleum sector were: (a) the promising geological potential; (b) the low ratio between investments (for exploration and development) and output; (c) the low level of operating costs; and, complementarily, (d) the practical, business-minded attitude of SONANGOL and the Government. Thanks to these advantages, Angola succeeded in attracting foreign companies and resources necessary to maintain continuous growth of petroleum production in the face of falling world oil prices. For instance, while operating costs in Angola varied between US$1.5/bbl and US$4.9/bbl in 1985, and averaged US$1.73/bbl, operating costs in the Congo varied from US$4/bbl to US$10/bbl. Moreover, in Angola, investment per unit of output was only one-third of that of the Congo. These figures illustrate the fact that Angola's oil reserves are economically attractive (measured in terms of the net-back value of the extracted oil) and, therefore, have encouraged foreign investors even though the contractual terms are among the toughest in the world, the minimum exploration requirements are high, SONANCOL's power to intervene (e.g., through the Operating Committees) is unusually broad, and the Angolan take in any commercial production is very high. Oil Taxation 2.19 Annex 6 provides a detailed description of the tax and tax-like systems which are applied to foreign companies and SONANGOL. Their main features can be summarized as follows: (a) under joint venture arrangements (Cabinda) the Government's share in oil revenues is captured in the form of a royalty (which is essentially a sales tax), a tax on "excess profits" (= the income in excess of operating costs and some allowances for investment expenditures), and a tax on net income (= value of output less operating cost less royalty and taxes on "excess profits"); (b) in the case of PSAs, the value of total output is divided between "Cost Oil" and "Profit Oil" whereby "Cost Oil" which covers normal operating costs as well as past expenditures for exploration and development may not exceed a certain percentage of total production (50%) for a predetermined number of years. The "Profit Oil" is split between SONANGOL and the foreign company in accordance with a progressive sliding scale scheme  - 29 - (in favor of SONANGOL) related to the cumulative output of the field. While the "Profit Oil" is subject to income taxation, a price-cap provision (which is essentially a 100% tax on excess profits) would apply If oil prices were to exceed the US$20/bbl level; (c) the lion's share of Angola's revenues from oil (more than 80%) has been and is still being provided by joint ventures, due to the low output attributable to PSAs (although this is beginning to change with the explosive growth of output in Block E PSA). (d) the specific tax-contract schemes which apply to joint-ventures and PSAs provide a progressive system of revenue-sharing and are designed to capture windfall profits. While in joint- ventures (from which most of Angola's take is collected in the form of taxes) the effect of volume-induced changes on taxes is less pronounced than that of price changes, the share of "Profit Oil" (which accrues to Angola from PSAs) is dependent on: (i) the speed with which the capital expenditures of foreign companies are recovered; and (ii) cumulative output. Both mechanisms protect the interests of foreign oil companies, particularly under unfavorable market conditions, and provide Angola with returns which adjust to the relative market- dependent profitability of its oil resources; and (e) PSAs are designed to give SONANGOL a significant part of Angola's take. In joint ventures, the lion's share of Angola's take accrues directly to the Government. This accounting difference and the fact that the time profile of the net revenues from PSAs is not in line with the current financial needs of the Government, have become a minor source of discord between the Treasury and SONANGOL. Furthermore, the taxation of foreign oil companies operating under PSAs has turned into an issue that affects the distribution of Angola's take between the Treasury and SONANGOL. 2.20 As international oil prices weakened after 1981-82, the tax regime for joint-ventures reduced Angola's petroleum revenues. The same mechanism which captures windfall profits for the Government also reduces the take when oil prices fall and output remains constant. Angola's response to the decline in international oil prices was to raise its output significantly. However, in periods of falling oil prices, output must grow at a rate exceeding the rate of price erosion in order to recoup lost income. This explains why Angola's oil revenues, particularly those from joint-ventures, have fallen off sharply even though it has succeeded in boosting its outp"t, above all in 1985 and 1986. Some illustrative figures are given in Table 2.2.  - 30 - Table 2.2: ANGOLA: CHANGES IN OIL TAX REVENUES AND OIL OUTPUT 1985-86 (Percentage) 1985 1986 Crude Oil Production +13,7 +21.0 Price of Crude Oil -5.5 -51.6 Tax Revenues from Oil -1.4 -49.3 Source: The MEP. 2.21 The workings cf the fiscal regime (taxes, royalties, price cap, other levies) transfer to Angola a large share of windfall oil revenues and prevent "excess profits" from accruing to the oil companies. They also protect the oil companies against a profit squeeze in periods of declining prices by shifting some of the burden of adjustment on to Government revenues. To the extent that the tax system for joint- ventures transforms a decrease in prices into a reduction of Angola's share in oil revenues (since tax income rises faster than the company's profit), it does not erode the incentives for foreign oil companies to continue operating and investing. Thus, the adverse impact of oil price declines on the activities of the oil companies was softened by the tax legislation. This made continued investment and a steady increase in output financially rewarding to the companies, which in turn provided the Government with income partly offsetting the losses caused by the fall in oil prices. Marginal Oil Fields 2.22 Several suggestions have been made recently to soften the contractual and fiscal terms of production-sharing operations for marginal fields. Many of these suggestions come from Block 1 where more than US$180 million has been invested in exploration since 1982 without finding any commercially attractive deposits. PSAs provide that any find which is not developed within three years of discovery must be handed over to SONANGOL (when no discoveries are made, the company that took the exploration risk must absorb the costs, without any recourse). Usually companies are reluctant to develop marginal fields (i.e., less lucrative than average or even unprofitable, at prevailing prices). However, SONANGOL should be wary of setting a precedent in shading contractual terms. Basically, the tax-contract system of PSAs does not distort the decision of whether or not to develop a marginal field. Oil companies may recover their outlays for exploration and development, and there is no tax or, output (which would affect marginal revenues from declining flows or marginal fields). If some fields are not developed (while others are), it is for the reason that they appear less profitable than the average. In principle, unprcfitable oil should be left in the grourd until it becomes profitable to lift. Funds saved should be invested in  - 31 - the development of more profitable fields, either proven ones or new discoveries (assuming that the probability of new discoveries justifies the delay in investment). 2.23 However, strict economic reasoning may not be appropriate in a second-best world as the Government urgently requires additional oil income. A case could be made for fiscal incentives to encourage the companies to develop less attractive finds: a comparatively lower take from marginal fields may be preferable to no additional revenues at all. And since under prevailing economic constraints any income available today is considerably more valuable than income accruing in the future (i.e., the rate of time-preference is very high), the early exploitation of marginal reserves could be justified. Nonetheless, the Government should be aware of the trade-off between the benefits of short-term revenues and the possibly higher, though uncertain, future revenues forgone by developing these fields now rather than later. Further, the setting of a precedent may undermine future negotiations. 2.24 Also under discussion are: (i) the possibility of extending the period between the commercial discovery and field development; and (ii) the unequal fiscal treatment of joint ventures and PSAs. The latter problem is created because in offshore Cabinda the tax base is calculated by consolidating total revenues and total expenditures for exploration,8/ development, and operation, whereas in PSAs, investments in exploration and development can only be recovered from oil from the same field. As a consequence, under PSAs the pay-back period for development expenditures may prove much longer than for joint ventures, However, in both cases (i.e., the determination of the length of the exploration period and the fiscal treatment of development expenditures) the Government pragmatically intends to adapt contractual terms to field-specific conditions. Thus, the issues in question no longer pose any serious obstacle to future oil development. Prospects for Oil Field Development 2.25 As can be seen from Table 2.3, the oil activities which the MEP expects to be undertaken between 1987 and 1990 center on the development of proven fields. The average level of planned annual investment is almost twice as much as in the past seven years. Highest priority is given to Cabinda which accounts for about 40% of total planned exploration and development expenditures. More than 70% of the investments projected for Cabinda will be used to develop (i.e., bring into production) new fields. One of the largest projects is the development of the Numbi field in which US$230 million will be invested. The recoverable reserves of Takula, Angola's biggest oil field (about one-third of total current production) were expected to increase to 240 million barrels in 1988 when a US$200 million water injection 8/ This means that exploration expenditures can be used to offset current revenues of other (already developed) fields.  - 32 - program was to be completed. The second largest investment is in Block 3, where most of the projected expenditures of about US$770 million will be used to accelerate the development of proven reserves. Expenditures for development are also dominant in Block 2. And, in the onshore areas of the Congo and Kwariza basins, the main objective is to maintain the present production of about 33,000 bbl/d. No development expenditures are planned for Blocks 1, 4, 5, 6, and 8. Table 2.3: PROJECTIONS FOR FUTURE EXPLORATION AND DEVELOPMENT (1987-90) Number of Number of Total Investments in Exploration Development Exploration & Development Wells Wells US$ million S Cabinda 9 134 800 39.0 Cabinda B/C 17 14 Congo Onshore 3 6 80 3.9 Block I 1 - - _ Block 2 4 24 >180 8.8 Block 3 6 31 770 37,6 Block 4 3 - - Block 5 5 - 135 6.6 Block 6 a/ 4 - 35 1,7 Block 8 a/ 4 - 35 1.7 Cabinda Onshore a/ 5 _ 15 0.7 Total 61 209 2,050 100.0 Subtotal b/ 48 209 1,965 - a/ Estimates. Exact figures are conditional on future negotiations. b/ Not including Cabinda Onshore and Blocks 6 and 8. Source: The MEP; SONANGOL; and Mission estimates. 2.26 Forty-eight exploratory wells 9/ were firmly scheduled for the period 1987-90, which means 12/y, compared to 15/y in the period 1981- 86. Based on previous contractual commitments, the annual number of projected wells will decrease from 20 in 1987 to six in 1990, indicating that new agreements will be required to keep exploration activities at the rhythm of the early 1980s. The majority of the scheduled exploratory wells (26) will be in Cabinda, reversing recent trends. Exploration in Block 1 will stop and the companies concerned will try to obtain better contractual terms to make development of marginal discoveries feasible. Since the prospects for BRASPETRO's Block 4 are not bright, it is likely that exploration there will also stop. On the other hand, a minimum of five exploration wells are scheduled for 1988-89 for Block 5, where 9/ Not including Cabinda Onshore and Blocks 6 and 8.  - 33 - drilling started in April 1987. Negotiations between SONANCOL and TOTAL/PETROFINA are under way to define an exploration and production program for Block 8 (with TOTAL as the operator), and plans called for the re-opening of Block 6 during 1988. Offers for Block 7 were requested for the second half of 1988, and negotiations are expected to take place in 1989. Further exploration in Cabinda (over what is now planned) may well result now that another international oil company, AGIP, has been admitted to the joint venture. Offers for Cabinda Onshore are still being evaluated. Negotiations may start in 1988, but are expected to be difficult. Thus if Blocks 6 and 8 as well as Cabinda Onshore are included, the total number of exploratory wells drilled in the period 1987-90 may turn out to be 67. On an annual basis, this would equal the 1981-86 average. 2.27 Estimates of future production vary considerably depending on the source of information. The figures published by the MEP (Table 2.4) are based on forecasts prepared by SONANGOL and can be considered a conservative estimate of what is achievable in the light of past and planned development activities. The projections understate the potential for a more rapid exploitation of producing fields, 10/ but also overestimate future success in developing proven fields. On the other hand, average annual investments projected for the period 1987-90 are rather high (relative to the country's finances) and might not be carried out in toto. As a result of these countervailing biases, the forecast could end up being rather accurate over a number of years. 2.28 Angola's exploration strategy has been well thought out and successful. However, Angola has usually stuck rather closely to its oil development plans. Now, for the first time, production in Block 3 is exceeding previous, firm plans by a substantive margin with output in late 1987 at 110,000 bbl/d, well above the 70,000 bbl/day that was planned in early 1987. This is leading to a rapid reduction in the reserves-to-production ratio (from about 13 years in 1987, to as little as 8 years by 1990). While there is no magic reserves-to-production ratio, many countries feel that a suitable level is between 10 and 15 years. If output in Block 3 is not expected to return to its earlier time profile, then a reassessment of the planned level of exploration activities may be needed to arrest the decline in the reserves-to- production ratio. Any worsening of exploration performance (i.e., fewer, 10/ In late 1987, production in Block 3 was raised to 110,000 bbl/d, effectively reducing the reserves-to-production ratio to about seven years. This product'on is almost double the previous projections (Table 2.4). The reasons for this massive departure from the productive path set forth in early 1987 are not quite obvious. It may be that the Government's short-term revenue needs are in favor of a more rapid exploration, which in the case of Block 3 also "benefits" foreign oil companies since higher production rates shorten the pay-back period of exploration and development costs of PSAs.  - 34 - or smaller finds) or any slowing down of the pace of exploration could quickly lead to a dangerous situation as the cushion of reserves could become dangerously thin. Also, independently of whether the Angolan authorities consider their level of revenues adequate, international markets might not. This could cause an increase in the cost of raising capital (for any purpose) for Angolan needs. 2.29 As regards the composition of crude oil output, the MEP's scenario predicts that the share of Cabinda crude in total production will decrease from about 70% in the mid-1980s to 55% in 1990, while the share of Block 3 would rise from 18% in 1986 to about 30% in 1990. This shift in output is also reflected in the projected allocation of investment funds, the structure of which will change in favor of Block 3. Thus, the strategy in petroleum is to keep the production level in Cabinda as high as possible (base production) and to spur output growth by expanding activities in Block 3. In other words, Government policy is oriented to the short term objective of maximizing output and revenues, while long term considerations are, at present, given relatively less importance. This is the expected behavior of a country in the political situation of Angola. Table 2.4: OIL PRODUCTION AND RESERVES, 1986-90 (in 'OOOs bbl/d) Production by 1986 Projected 1990 Area Actual % 1987 1988 1989 1990 in % Cabinda 190 67.4 220 256 246 246 54.9 Block 3 50 17.7 85 113 128 136 30,4 Block 2 6 2.1 10 29 39 38 8.5 Onshore 36 12.8 36 36 32 28 6,3 Total: 282 100 351 434 445 448 100 Reserves-to-Production Ratio (in years): 13.7 12.6 9.6 8.8 7.8 Source: SONANGOL and the MEP (estimate was done in late 1987). 2.30 According to the MEP, a higher level of investment than in the past will not necessarily prevent output growth from slowing down at the end of the 1990s. Thus any shortfall in planned investments will have a severe impact on future oil output and revenues. Since a considerable share of the investment is to be carried out in Cabinda, a joint venture, the investment program will place a heavy financial burden on SONANGOL. Unlike the "pure" production-sharing onerations which, for instance, are carried out in Block 3, the Cabinda Joint venture requires a 51% financial participation by SONAI4COL. Thus, since most of Angola's oil is produced in Cabinda, increasing output reqtires the assumption of high up-front costs by SONANGOL, whose financial position is weakening, and could precipitate a serious decline in the ccuntry's creditworthiness and  - 35 - credit rating. This also explains why the less costly Block 3 (where no fitarcial commitments by SONANGOL are required) is gaining increasing importance within Angola's crude oil development strategy. But as long as Cabinda remains the country's major source of oil, SONANCOL will have to meet the sizeable financial obligations required (unless it sells part of its shares). As a consequence, SONANGOL's ability to raise the required funds at low cost will be highly sensitive to both its own financial performance and to the country's ability to keep its creditors reasonably content. In ouch a delicate situation, the Government should not embark on the risky course of offering future oil as a security to creditors (other than SONANGOL's). The Government should also avoid major delays in servicing its external debt since any deterioration in the country's debt discipline will reduce the market value of the loans which have been raised in the past and, therefore, tend to raise the costs of future borrowing. 2.31 SONANGOL has recently expressed interest in gathering practical experience as an operator. So far, SONANGOL's operational responsibili- ties have been limited to Block 4 where it owns 51% of the mixed-economy company ESPA (the Empresa de Servicios Petroliferos de Angola). Plans to increase its share to 100% and, thus, to become fully responsible for operations have been dropped because of the disappointing exploration results. Nonetheless, SONANGOL may still intend to expand its upstream activities and is currently exploring the option of assuming FINA's role as the operator in onshore Cabinda. Taking over ongoing operations is a less risky strategy than exploring and developing new fields. However, while additional experience as an operator may help SONANGOL to improve its capacity to supervise foreign oil companies, resource constraints and the scarcity of qualified personnel might complicate SONANGOL's role as controller of oil activities. While this report does not strongly favor either option (operating or not operating), on baLance SONANCOL should probably not become more heavily involved in field operations, but direct experience can't hurt, provided its costs are kept low and the operating results are acceptable.  - 36 - II.B. GAS SUPPLY AND UTILIZATION Summary and Recommendations 2.32 Significant quantities of associated gas (currently about 50% of total production) are used for gas lift and reinjection schemes. Government policy calls for a rapid increase in the gas utilization rate. New lift and reinjection schemes (where compatible with oil production and reservoir characteristics) and expansion of existing ones (replacement of Livuite gas with associated gas sources) are under way. A target utilization rate of 70% is the goal for late 1990, up from the present 50% or so; this is a reasonable objective. However, pure reinjection of associated gas (to save it for future use) may have costs which exceed the economic value of the gas in the ground. 2.33 No investments in gas field assessment or delineation should be undertaken unless major economic uses have been identified. In particular, only large-scale projects which could steadily consume a considerable amount of gas would justify the delineation and development of known, existing non-associated gas fields. No economic large-scale projects are presently known, and thus prospects for the development of non-associated gas fields are dim. 2.34 The only large-scale project capable of using sizeable quanti- ties of natural gas is the proposed export-oriented ammonia/urea plant which would require about 50.6 MMCFD of gas. However, in view of the depressed international fertilizer market and given that gas supply costs are relatively high, Angola would not have a substantial comparative advantage even in a well-managed plant. Furthermore, there are no other domestic consumers who could absorb natural gas in large quantities or would have an incentive to switch to gas. Thus, the country's non- associated gas should be left in the ground until economic uses for it can be found. 2.35 As crude production in Block 3 rises, the potential for a new large-scale LPG recovery scheme could increase. If world prices for LPC do not deteriorate, such a scheme could expand exports and provide the country with additional foreign exchange. 2.36 SONANGOL's ability to study and supervise even a limited number of gas-related projects sho d be strengthened in the small unit responsible for gas matters set up around the few people currently dealing with gas in SONANGOL. This unit should be in a position to monitor ongoing gas-related activities more thoroughly and to coordinate plans for future projects with related activities in other subsectors. It should not become a bureaucratic unit which wastes manpower in the pursuit of elusive gas projects for their own sake. 2.37 Highest priority should be given to two projects presently under consideration by SONANGOL:  - 37 - (a) the LPG recovery scheme in Cabinda, i.e., on the moored LPG tanker, Berge Sisar, (estimated investment costs: US$2-3 million). This project would increase the domestic supply of LPG by about 60% and replace costly LPG imports of 8,000-10,000 tons/year. (b) the export-oriented LPG/condensates recovery program in Block 3. This project, if economic, could produce significant foreign exchange revenues. Other projects which deserve further investigation in the short term are: (c) the planned LPG bottle rehabilitation plant, with investment costs estimated at US$5 million (essential, if the first project is to go ahead); (d) the proposed dual-fuel thermal power plant in the Soyo area, and rehabilitation of the gas-fired turbine for Cabinda; and (e) the onshore plant for recovery of LPG at Malongo costing US$3-4 million (provided that demand will be adequate at the higher LPG prices which this report proposes). Gas Reserves and Utilization 2.38 Angola's natural gas resources probably exceed 5 TCF. Non- associated gas accounts for about 3 TCF of the probable reserves, and associated gas for about 2.5 TCF, of which 1.8 TCF were proven in early 1987. At present the average associated gas-to-oil ratio is approximately 1.34 MCF/1,000 bbl of oil. Thus, a crude oil output of 282,000 bbl/d in 1986 yielded 379 MMCFD of associated gas (Annex 4, Table 7). By 1990, associated gas production is expected to reach 515-536 MMCFD, although the official forecast is about 488 MMCFD. Non- associated gas reserves are not yet being exploited except for a small offshore field in Cabinda. The main resources are located offshore of Zaire province in Blocks 2 and 3, but because of the limited number of exploratory wells, information about reserves is limited (Annex 4). 2.39 In 1986 about 51% of the associated gas output was actually put to use, primarily in gas lift and reinjection schemes which account for more than 85% of the country's gas utilization (Annex 4, Table 8). Only a minor fraction of the associated gas which is not flared is directly recovered as a fuel, either for use in oil operations or for LPG, and the picture for productive uses (other than uplift and reinjection) is not likely to change in the near future. By 1990 the gas utilization rate is expected to rise to 70%, with most of the gas still being used in lift and reinjection schemes.  - 38 - 2.40 The declared policy of the Government is to increase the utilization of associated gas; hence flaring is prohibited, while recovered gas may be used free of charge. The Government has never conducted a comprehensive assessment of whether reinjection is economically justified. From an economic point of view, reinjection to conserve associated gas for future use makes sense only if the costs per CF saved do not exceed the user costs (i.e., depletion value) of non- associated gas. Although there is a lack of precise data .n this area, it is likely that this condition is not met in Angola. While the country's non-associated gas reserves are comparatively small in absolute terms, the ratio of probable reserves to projected production, i.e., the tima horizon for resource depletion, is quite high. As a consequence, the user costs of the non-associated gas tend to be negligible (probably in the vicinity of US$0.2/MMBTU), and thus the scope for profitable reinjection is limited. 2.41 A more promising option for associated gas utilization is the production of LPG. At present, the only facility in which associated gas is being recovered for LPG production is located offshore of Cabinda on the tanker, Berge Sisar. The output, consisting of a 66:34 mixture of propane and butane, rose from less than 500,000 bbl in 1983 to about 2 million bbl in 1984 and is now stable. Output is expected to rise to 2.6 million bbl/y over the next few years. In 1986, almost the total output (170,000 t) was sold to Brazil at about US$125/t (FOB Angola) or US$2.89/MMBTU. 2.42 In 1986, sales of LPG in the national market reached 32,000 t. However, the cost of imported LPG (about 12,000 t) is much higher than the FOB value of export LPGQ The import parity costs of LPG--(c.i.f. coastal terminals) are US$250/t, while the FOB value of export LPG is US$125/ton. The costs of adjusting the composition of "'export LPG" to meet the national specifications and shipping it to coastal terminals are estimated at US$90/t. Thus, the opportunity cost of LPG could be considered either US$250/t, or US$215/t (i.e., US$125/t FOB plus US$90/t fractionation and delivery costs). 11/ 2.43 The LPG fractionation project would add about 30,000 t/y (80-85 t/d) to domestic supplies. This would suffice to eliminate high- cost imports (about 10,000 t/y costing about USt2.5 million) and add 20,000 t/y to meet domestic consumption. At present prices, there is 11/ Gas is both imported and exported. Angola is a net exporter of LPG, but export LPG (66% propane, 34% butane) is not the same as the LPG consumed domstically (produced in the Luanda refinery and imported). Thus if both LPGR can be considered one good, then the opportunity cost is the FOB value of export LPG; while if they are considered to be different goods, the opportunity cost would be the CIF cost of imported LPG, and shipping it to Luanda could be used as a guide to domestic price setting or as the opportunity cost of LPG.  - 39 - little doubt that this quantity of LPG could be absorbed (provided also that bottles and stoves are available). However, a more reasonable price for LPG (paras. 3.35 and 3.36) in the range of Kz 45 90/kg (with a 12.5-kg bottle costing Kz 563 to Kz 1,125 12/) could well dampen the growth of demand. It might therefore be prudent to increase the supply to the domestic market more gradually than is implicit in the 85 t/d fractionation project. 13/ Similarly, preparation of additional projects to expand the availability of LPG for domestic consumption should best be postponed until the effect of the higher price on demand can be evaluated. Market Potential of Non-Associated Gas 2.44 As regards the exploitation of the country's non-associated gas reserves, the short- to medium-term outlook is not particularly promising. Unless a large market can be identified, the extraction and gathering of non-associated gas will almost certainly prove to be uneconomic. So far, the only large-scale project which could use a sizeable amount of natural gas as a feedstock is an ammonia/urea plant proposed for the Soyo area. The project has been under study since the early 1980s. 2.45 indivisibilities and economies of scale require a minimum capacity of 1,000 t/d of ammonia. World-class plants typically have an installed capacity of 1,500 t/d of ammonia and 500 t/d of urea, and cost about US$330 million (at 1987 prices). Capacity utilization in these plants typically hovers in the 80-90% range. The maximum output would therefore be about 164,000 t/y of urea and 396,000 t/y of ammonia. Given the limited domestic demand for nitrogen fertilizers (about 10,000 t/y in 1987), the plant would have to sell most of its output abroad where fierce competition and a general glut of fertilizers are keeping prices low. A sustained market recovery with prices above US$200/t is not likely to occur before the mid-1990s. Moreover, since the plant will require a gas supply of about 51 MMCFD (16.65 BCF/y) which cannot be met from associated gas alone, non-associated gas would have to be developed and used. Therefore, feedstock costs will most likely exceed US$1.5/MMBTU. 12 Even at the higher price for LPG, the price of one 12.5 kg. bottle could be equal to the parallel market price of only 3-4 kg. of charcoal, i.e., LPG would still be considerably less costly than woodfuels. 13/ In some uses, fractionation may not be needed. The export LPG could simply be mixed with the refinery-produced LPG and used. This is usually acceptable in household use. However, tests should be conducted to see if this is correct for any use being contemplated.  - 40 - 2.46 A detailed appraisal of the economics of the proposed urea/ ammonia plant is given in Annex 7. Table 2.5 summarizes the main findings. Table 2.5: ECONOMICS OF AMMONIA/UREA PLANT a/ Rate of Discount Internal Rate Gas Supply Costs (10%) (12%) (15%) of Return (US$/MMBTU) --------- NPV --------- () 1.25 + + - 12.72 1.50 + - - 11.92 1.75 - - 11.10 2.00 + - - 10.26 Netback value of gas 2.07 1.48 0.50 a/ Base case, not considering depletion costs. - = negative; + = positive. Source: Annex 7. Even under the most favorable base case conditions, net returns to the project will be close to zero or negative unless the rate of discount is below 12%. The considerable financial burden of this large-scale investment, the comparatively high costs of gas supply, and the uncertainties of the international fertilizer market make the proposed ammonia/urea project economically unattractive. As better opportunities of using non-associated gas may well be identified in the future, the low-return ammonia/urea project should not be undertaken at this time. Only if a private concern were willing to incur all risks and pay a reasonable price for the gas should Angola consider such a project. 2.47 In industry, the only consumers who could theoretically switch to gas, and absorb enough gas to justify investments in gathering and transport, are the cement factory and the oil refinery. At the current output (720,000 t/y of clinker), the cement plant's consumption of fuel oil is equivalent to 6 MMCFD. A proposed expansion of the factory's capacity to about 1.5 million tons of clinker in 1990 would increase the fuel requirements to 15 MMCFD. However, the cement plant now uses surplus fuel oil costing only about US$1.8/MMBTU (export parity). In addition, the potential gas supply from the nearby Kwanza field (3 MMCFD) would not even be sufficient to meet the plant's current fuel demand so that more remote sources would need to be developed, thereby increasing gas costs to US$1.5-2.0/MMBTU. Gas is therefore not competitive with fuel oil in the medium term. The same argument applies to the Luanda refinery whose fuel oil requirements are equivalent to approximately 4 MMCFD. Other industries which at present account for only 20% of the  - 41 - country's boiler fuel consumption might demand 2-5 MMCFD of gas. Thus, the low potential demand for gas and the availability of cheap alternative fuels make the near-term development of non-associated gas reserves for domestic use uneconomic. 2.48 As for power, most of the electricity is supplied from low-cost hydro plants (with a substantial hydro potential still to be developed). The use of gas would only be needed for peak-load generation in thermal power plants. However, given the alternative of petroleum products and the fact that an optimistic peak-load scenario might require at most an average of 3-4 MMCFD of gas, then power demand cannot justify any investment in gas gathering and transport. Gas for power generation might prove economically viable only in isolated areas with low primary energy requirements and cheap gas available nearby. The most promising options of this type are the 15 MW dual-fuel power plant proposed for Soyo and the rehabilitation of a 10 MW gas turbine in Malongo-Cabinda. 2.49 If the international price of LPG does not fall below the current level, there may also be a significant potential for new export- oriented LPG recovery schemes similar to the one in offshore Cabinda. The most promising area is Block 3 where oil production is expected to increase considerably, in association with large volumes of gas rich in LPG. These schemes will only be justified by exports, but they could also provide a low-cost source for small (gradually increasing) volumes of LPG (adjusted to meet national specifications) used domestically. In this respect, the planned offshore Cabinda production of 85 t/d for the domestic market is a logical step. With estimated investment costs of US$2-3 million, the project will probably replace imported LPG economi- cally. However, small-scale LPG recovery plants will not be economically viable, unless there are isolated markets which can be supplied from nearby gas resources. The onshore Cabinda project, for instance, which is supposed to produce 8 t/d of LPG from gas recovered at the Malongo oil terminal (investment costs: US$3-4 million) is a case in point, and its economics are probably marginal. Other proposals, such as the 3-4 t/d onshore Kwanza LPG recovery plant, will probably not be justified either.  - 42 - III, CRUDE OIL: REFINING AND PRODUCT SUPPLY Summary and Conclusions 3.1 Pricing of petroleum products at the refinery gate and to final consumers, and pricing of crude oil for domestic refining, are areas where substantial reforms could be implemented most easily. Yet, specific pricing recommendations are difficult to make in the macroeconomic policy environment of Angola. The normal price recommendation would be to base domestic prices of petroleum products on opportunity costs such as the CIF cost of bringing products to Angola, or the FOB cost for net exports, or full cost recovery for those products produced in Angola (i.e., remove subsidies to crude oil and specific products). However, given the extreme overvaluation of the Kwanza, the standard economic prescription of using opportunity costs as the basis for pricing would only fully make sense after the value of the Kwanza had been adjusted downwards to some sort of equilibrium level (or to a level nearer to equilibrium than is currently the case). But since final petroleum product prices in Angola are below border prices even at the highly overvalued present official exchange rate, and the crude oil for local refining is subsidized, these shortcomings would need to be corrected first. In fact, an overvalued exchange rate doesn't mean that petroleum products should continue to be consumed wastefully by being priced excessively cheap. A series of step adjustments in prices would probably be easiest to apply. The steps could be as follows, using hypothetical exchange rates: Step One: eliminate all subsidies to crude and products, including LPG, and immediately bring all prices to border levels at the official rate of exchange. Step Two: adjust all petroleum product prices to an exchange rate of, say, Kz 100/US$. Step Three: by this time, the SEF should be in progress and a more adequate exchange rate might be available to guide the MEP in the pricing of petroleum products. Should the exchange rate remain fixed in spite of notable domestic price increases, the MEP could use an index of inflation to keep real product prices stable. Advantage of a Correct Pricing Policy 3.2 Basically, pricing policies should be based on the economic cost of supplying or using a particular energy resource. Adherence to this principle ensures efficiency in resource allocation and provides consumers with correct signals for their economic decisions (i.e., it  - 43 - tells consumers the cost of an additional unit of each resource and thereby enables them to make the best choice). However, given the large distortions and the difficulties of running the economy along optimal lines, second-best policies are more expedient for Angola in the short to medium term. 14/ While these policies would involve a gradual departure from the severe price distort,.ons described below, the medium- to long- term goal should be to adjust prices and tariffs so as to reflect the true economic cost aL the margin. 3.3 Currently, the structure of petroleum product prices looks as follows: (a) the present official refinery gate prices are now, on average, slightly below border prices at the official exchange rate and mid-1987 values. Gasoil and fuel oil are priced considerably below international parity while LPG, gasoline and kerosene/jet fuel are reasonably in line with border price levels; (b) LPC, kerosene and light fuel oil (LFO) are sold at prices which are less than full cost recovery, while gasoline, jet fuel, gasoil, and heavy fuel oil (HFO) are sold above cost-- all in Kz valued at the official exchange rate; and (c) the FPA refinery pays less than the economic opportunity cost for indigenous crude oil feedstock. The Government provides an effective subsidy to FPA by not collecting royalties on crude for local refining. Refining 3.4 The refining of indigenous crude in Luanda in a hydroskimmer is an economically viable product-supply strategy for Angola as compared to imports of products. The FPA refinery is a reasonably well run and well maintained facility. Purchasing this refinery (or a controlling interest in it) would seem a wasteful use of limited Angolan financial resources. Furthermore, lack of incentive to reduce costs and possibly high use of expatriate labor are the most apparent contributors to higi operating costs. The FPA refinery operates on a "cost-plus" refinery gate pricing arrangement. On the aggregate, all verified operating costs, depreciation and allowable profit are recovered, but there is no particular incentive for cost minimization and optimization of operations. Therefore, efforts should be made to design and implement a pricing scheme which encourages the refinery to operate in a more efficient way. This could be done in the context of an ESMAP activity. 14/ For example, the second-best pricing strategies aiming at the financial strengthening of utilities that the mission proposes for the power sector.  - 44 - Distribution 3.5 SONANGOL distribution and marketing departments have excessive staff, facilities, and overheads in relation to volume distributed. The operations are based on a "cost-plus" arrangement. All verified costs and a guaranteed profit margin are supposed to be covered either through the final selling price or through a subsidy from either a profitable product or the Government budget. In practice, however, petroleum product distribution is a loss-making business since large consumers (e.g., the army, cement plants) fail to pay their bills. 3.6 In order to improve the situation a study should be undertaken, dealing with revisions in the cost-plus pricing scheme, the enforcement of a higher financial discipline on customers and, most importantly, investigation of the options for a gradual rationalization (privatization) of SONANGOL's distribution and marketing activities. Procurement .7 SONANGOL Limited, London, with its own staff and its joint venture arrangement with the West German trader, STINNES, currently manages Angola's product import and export arrangements. However, SONANGOL itself has the knowledge and capacity to organize tenders or supply contracts either as a buyer or a seller. SONANGOL might consider selling excess fuel oil (as it has done in the past) through an international tender for a one-year contract, or through sales to end- users in the United States via a brokerage firm which could do all the work for about US$0.03/bbl. Jet fuel, kerosene, and gasoil could be procured from a nearby reliable refiner such as CEPSA or CHEVRON, both of whom have processing agreements in Abidjan, Gabon, and Moanda-Zaire (south of Cabinda). In this manner, Angola may be able to lower the cost of its product imports, especially when purchases are made from affiliates of companies already present in Angola. Production, Supply and Consumption Considerations 3.8 Angola currently consumes some 0.9 million t/y of petroleum products compared with total production of crude and gas liquids in excess of 14 million t/y. In spite of its large oil production and exports, Angola's revenue requirements are so critical that any reduction in domestic consumption of petroleum products would be advantageous. The lack of incentives to cost effectiveness under existing "cost-plus" product pricing regimes for both refining and product distribution appear to be major contributors to inflated product supply costs above "efficient" levels.  - 45 - Product Trading/Import-Export 3.9 Imports. Although the principal products supply source for inland consumption is the Luanda refinery, shortfalls in certain products are met through direct product imports. Table 3.1 provides a summary of imports for the 1980-86 period. As indicated, jet fuel (A-1) has become the predominant import, mostly because of military use. Table 3.2 illustrates the growth in jet fuel imports versus refinery supply and total supply/requirements over the 1980-86 period. Imported supply in 1986 represented 40% of the total, up from zero in 1980. 3.10 Over the past three years, Angola has also experienced a small shortfall in gasoil supply, which has been met by imports. This was expected to turn into a slight surplus in 1987 as the yield from the expanded refinery at higher throughputs more than matches growth in demand. In addition, there has been a consistent shortfall in high- butane LPG from the refinery, which has been met through imports. On the other hand, since 1983 significant amounts of high propane LPG have been recovered from Cabinda associated gas. This, however, has never been used to supply the domestic market because of the difference in specifications. Since the Cabinda export type of LPG is 70% propane vs. about 30% (maximum) propane for the refinery/imported material, it is not adapted to the storage and end-use equipment in Angola. A project is proposed (para. 2.37) to recove- part of the Cabinda LPG fox household use and, if required for technical reasons, to fractionate a portion of the Cabinda production on board the floating storage tanker, Berge Sisar, to produce a high butane/low propane LPG for the national market. This would make it possible to eliminate present imports. Table 3.1: IMPORTS OF PETROLEUM PRODUCTS 1980-86 (Tons) 1980 1985 1986 LPG 6,942 11,710 10,913 Jet fuel 0 102,633 114,184 Gasoil 0 36,693 16,435 Total products 6,492 151,036 141,532 Percentage Share of Petroleum Imports LPG 100.0% 7.8% 7.7% Jet fuel 0.0% 68.0% 80.7% Gasoil 0.0% 24.3% 11.6% Total products 100.0% 100.0% 100.0% Estimated value (USS million) -- 41,85 24.94 LPG: Liquefied Petroleum Gas (LPG). Source: SONANGOL and MEP.  - 46 - Table 3.2: JET FUEL SUPPLY 1980-86 IMPORTS VS. LOCAL REFINERY-SOURCED (Tons) 1980 1981 1982 1983 1984 1985 1986 Ex-refinery 118,559 126,960 99,493 143,527 161,527 172,809 171,064 Ex-imports 0 7,764 39,796 24,925 62,895 102,633 114,184 Total supply 118,559 134,724 139,289 168,452 224,841 275,442 285,248 Percentage Share of Total Supply Ex-refinery 100.0% 94.2% 71.4% 85.2% 72.0% 62.7% 60.0% Ex-imports 0.0% 5.8% 28.6% 14.8% 28.0% 37.3% 40.0% Total supply 100.0% 100.0% 100.05% 100.0% 100.0% 100.0% 100.0% Source: SONANGOL. 3.11 Since 1986, SONANGOL has procured jet fuel and gasoil imports through its U.K. subsidiary, SONANCOL Limited. This joint venture partnership with the West German trading company, STINNES, has established a steady import pattern of 5,000-ton parcels roughly every month from Tenerife, Canary Islands (Annex 12 gives details). 3.12 The mission believes that Angola can organize its procurement without the help of an intermediary and that there are more logical sources of supply (such as Abidjan, Moanda/Zaire or Cabon) which would cause much lower freight rates (Annex 12 provides some indicative figures). It seems feasible for SONANGOL to procure middle distillates at substantially less than the current cost of Mediterranean plus US$39/t, most probably in the range of Mediterranean plus US$19-22/t. In fact, Angola should consider the option of opening tenders for 10,000-ton parcels CIF Luanda rather than going to the trouble of chartering its own vessel. 3.13 Cargo Exports. The principal export is fuel oil, which is produced in the refinery in excess of requirements. Sporadic naphtha exports also represent a habitual surplus over Angola's requirements. The remaining small, rather sporadic volumes of other clean products are principally exported to favored countries such as Sao Tome and Principe, Guinea-Bissau or Mozambique under special government-to-government arrangements. A summary of cargo exports of finished products is shown in Table 3.3.  - 47 - Table 3.3: PETROLEUM PRODUCT EXPORTS (CARGO) 1980-86 (Tons) 1980 1985 1986 LPG at Cabinda 0 173,836 166,782 Ex-refinery: GasolIne 2,606 3,192 2,902 Naphtha 0 5,069 8,642 Kerosene 2,375 0 0 Jet fuel 0 2,952 0 Gasoil 23,839 7,322 5,642 Fuel oil 481,613 585,922 528,766 Total ex-refinery 510,433 604,457 545,952 Total al; products 510,433 778,293 712,734 Source: SONANGOL. 3.14 Excess fuel oil is sold by the refinery at official refinery gate prices to SONANCOL which then sells it to the SONANCOL-STINNES (U.K.) joint venture. With the U.S. Northeast coast being the major market outlet, SONANCOL currently receives New York harbor cargo prices less US$22/t. It should therefore consider returning to its 1981-82 practice of opening tenders in the international fuel oil market, for which the mission believes SONANGOL has the expertise. 3.15 Refining. Angola's requirements for domestic proCucts are met primarily through refining of domestic crude oil. The principal facility is a simple, 1.7 million ton/y hydroskimmer in Luanda owned and operated by FPA. In addition, there is a small 100,000 t/y topping plant owned and operated by CABCOC at its Cabinda/Malongo base. 15/ The Luanda refinery dates back to 1958 when a 100,000 t/y topping unit was built. At thdt time, it was fed with crude from the newly discovered onshore Kwanza field. 3.16 The present plant has a nominal capacity of 1.7 million t/y but actual yearly capacity considering planned maintenance shutdowns and unforeseen outages is rated at 1.6 million tons. The plant was recently expanded from 1.5 million t/y nominal capacity through the 15/ This takes a slipstream of crude from the large Cabinda crude production stream, extracts gasoil and jet fuel for local operations, and returns the remainder to the crude stream.  - 48 - de-bottlenecking of one of the topping plants (Annex 8 gives technical details). Along with minor modification to the tower internals and pumps, the total cost of the expansion was only US$2 million. 3.17 Crude oil feedstock. The present feedstock consists of Kwanza and Soyo crudes, both from FPA-operated fields. Table 3.4 summarizes the estimate of net plant yields from Luanda refinery for the two crudes. Table 3.4: NET PLANT YIELDS a/ (% By Weight on Crude) Kwanza Soyo LPG 0.8 1.4 Gasoline/Naphtha 11.8 10.3 Kerosene 8.0 11.0 Gasoll 22.5 24.8 Fuel oil 53.2 48.3 Total 95.8 95.8 a/ After refinery fuel and losses, estimated at about 4.2% by weight of crude throughput. Source: FPA and SONANGOL. The Kwanza crude is of slightly lower quality than Soyo, having a higher fuel oil and lower clean products yield. Kwanza is also of slightly higher sulfur content than Soyo, but both crudes would be classified as low sulfur crudes by international standards. 16/ The feedstock mix for 1985-86 shown in Table 3.5. Table 3.5: CRUDE OIL FEED - LUANDA REFINERY, 1985-86