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 …---- 1985 ------ 1986- Tons % Tons % Kwanza 201,323 13.9 155,363 10.7 Soyo 1,249,993 86.1 1,296,820 89.3 Total 1,451,316 100.0 1,452,183 100.0 Source: FPA and SONANGOL. 16/ The refinery takes the entire production of the declining Kwanza field. The Soyo (Congo basin) crude is also produced from a fairly mature field. At best its production is stable in the face of an increasing refinery throughput. - 49 - 3.18 Production/Yields. Table 3.6 summarizes the production/yield balance for the refinery from 1980 through 1986. As would be expected with a simple skimming refinery running a fairly steady crude slate over the period, there is not much variation in yield of major product streams. The total kerosene/jet fuel yield has increased as this is the product in greatest demand. There has been a slight reduction in fuel oil yield, probably because of the Lightening of the average crude slate as heavier Kwanza decreased as proportion of total feed. There may also have been some improvement in fractionation over the period, since the share of after-refinery fuel and losses has decreased significantly, from an average of about 6% in the 1980-83 period to a reasonably low 4.2% in 1985-86. Table 3.6: LUANDA REFINERY PRODdCTION/YIELD BALANCE, 1980 AND 1985-86 (Tons) -1980 - 1985 -1986- Crude run (1,237,507) 100% (1,451,316) 100% (1,452,183) 100% Product Yield: LPG 11,746 0.9% 18,269 1.3% 18,690 1.3% Gasoline 89,563 7.2% 103,450 7.1% 104,675 7.2% Naphtha 0 0.0% 9,914 0.7% 21,471 1.5% Subtotal 89,563 7.2% 113,364 7.8% 126,146 8.7% Kerosene 33,007 2,7% 45,598 3.1% 35,637 2.5% Jet fuel 118,559 9.6% 172,809 11.9% 171,064 11.0% Sub total 151,556 12.2% 218,407 15.0% 206,701 14,2% Gasoil 294,728 23.8% 353,699 24.4% 347,288 23.9% Fuel oil 609,418 49.2% 679,419 46.8% 685,767 47.2% Asphalt 7,358 0.6% 7,924 0.5% 7,253 0.5% Subtotal 616,776 49.8% 687,343 47.4% 693,020 47.7% Total (excl. loss) 1,164,379 94.1% 1,391,082 95.8% 1,391,845 95.8% Fuel & loss 73,128 5,9% 60,234 4.2% 60,338 4.2% Total 1,237,507 100.0% 1,451,316 100.0% 1,452,183 100.0% Source: FPA and SONANGOL. 3.19 Operating Costs/Efficiency. The refinery gives the impression of a well managed, well maintained, neat, and tidy facility. Maintenance, housekeeping, and safety practices have been satisfactory. The fractiona- tion performance and fuel and loss figures look favorable. However, the total operating cost figure of US$2.28/bbl appears high for this type of refinery (Table 3.7). The range of total operating costs for several other hydroskimmers over the past few years (inflated to 1986 levels) is - 50 - US$1.20-1.50/bbl, excluding depreciation. This compares with US$1.88/bbl for Luanda, excluding depreciation, The main discrepancy appears to be in labor costs. The figure of almost US$1.00/bbl compares with US$0.40- 0.60/bbl in other similar operations. Luanda's high cost reflects both the high level of staffing, 450 in all, and the excessive complement of expatriates. Given the rather low salaries for nationals, the excessive dependence on expatriates is probably the main reason for the higher than normal payroll. Table 3.7: LUANDA REFINERY OPERATING COSTS AND TOTAL GROSS MARGIN - 1986 US$/bbl US$/million Salaries, wages, benefits 0.98 10.4 Materials, chemicals 0.72 7.7 Depreciation 0.40 4.2 Financial charges 0.11 1.2 Training 0.07 0.7 lotal operating cost 2.28 24.2 Allowable profit 0.33 3.5 Gross Refinery Margin 2,61 27.7 Source: FPA. The expatriates alone probably account for US$4-S million of the total US$10.4 million/y labor cost. A significant reduction in the number of expatriates would bring the total operating costs for the Luanda refinery more in line with other similar refineries. Given appropriate human resource training and development planning there is no reason why this refinery could not be run with half of its present expatriate force. The biggest problem in encouraging such cost saving may be the lack of appropriate incentives under the existing "cost-plus" refinery gate pricing arrangements. 3.20 Distribution and Marketing. SONANGOL, which has a monopoly on the distribution and marketing of petroleum products in Angola, can rely on an e:vAnsive network of storage terminals, most of which were inherited from the former private distributors. The largest concentration of storage is in Luanda, followed by other large coastal installations in Namibe, Lobito, and Porto Amboim (Annex 8, Table 3). The large Luanda, Namibe, and Lobito terminals are tied in to the three major railway lines so as co serve the smaller depots up-country, all of which are on these rail lines. Due to deterioration of the rail systems, the inland terminals are largely dry. Most products that reach the interior come by road tanker direct to customer tanks or to the rare service stations. SONANCOL operates its own fleet of some 300 road tankers wi - an average capacity of 18,700 liters each. - 51 - 3.21 Under current conditions, the coastal region is well served, with little evidence of suppressed demand. In the interior, the normal manifestations of shortages such as service station queues and elevated black-market prices are rare for products such as gasoline and gasoil since there is very little transport or commercial activity gaing on. On products like kerosene for lighting, however, there is supprossed demand and high prices but this also is the case in the outskirts of Luanda. Table 3.8 summarizes total inland sales for 1980, 1985 and 1986. Table 3,8: INLAND PETROLEUM PRODUCT CONSUMPTION (SALES) 1980-86 (Tons) 1980 1985 1986 a/ 1980-86 (S) per annum LPG (Butane) 18,688 28,695 30,200 8.3 Gasoline - t'btor 79,567 95,004 103,680 4,5 = Aviation 1_391 987 430 -17.8 Total gasoline 80,958 95,991 104,110 4.3 Kerosene 29,835 43,908 46,310 7.6 Jet fuel 114,350 235,567 260,550 14.7 Total kerosene/jet fuel 144,185 279,475 306,860 13.4 Gasoi l 268,806 376,865 344,770 4.2 Fuel oil 121,268 104,622 119,200 -0.3 Asphalt 3,171 5,593 4,200 4.8 Total all products 637,096 891,241 909,340 6.1 a/ 1986 estimate based on 9-month actuals. Source: SONANGOL. International Comparisons 3.22 A comparison of Angola's petroleum consumption per capita in 1984 with that for seven "lower middle-income" countries shows Angolan consumption is low compared with the range of values and the average for the group. The stagnation in the economy is probably offset by military consumption but the overwhelming use of hydrogeneration for power would keep Angola's consumption of petroleum products lower than other similar countries, ceteris paribus. - 52 - Table 3.9: PER CAPITA PETROLEUM CONSUMPTION INTERNATIONAL COMPARISON, 1984 Lower/Middle-Income GNP Population Consumption USS/Cap (millions) -- Kg/Cap -- Mauritania 450 1.7 100 Zambia 470 6.4 110 Bolivia 540 6.2 195 C6te d'lvoire 610 9.9 95 Zimbabwe 760 8.1 90 Peru 1,000 18.2 370 Costa Rlca 1,190 2.5 380 Total average 7 countries 764 7.6 216 Angola 485 8.9 79 Source: World Bank and Angolan authorities. Projected Petroleum Product Consumption 3.23 SONANCOL's sales department has developed a forecast of product consumption for AngoLa to 1992. This projection is summarized for each product in Table 3.10. No details on the rationale for the assumed growth rates for each category were available to the mission. It appears that the growth in consumption of products such as gasoline, kerosene, and gasoil has moderated slightly from the actual 1980-86 rate. The most striking reduction has been for jet fuel for which per annum growth is shown as falling from 14.7% in 1980-86 to an assumed 2.2% per annum growth in 1986-92. This low growth is based on a 24% reduction in jet fuel use assumed for the military in 1987, followed by a 3% per annum growth thereafter, such that 1992 military consumption is still some 12% lower than 1986 levels. Non-military jet fuel use is assumed to grow at a fast 5.6% per annum over the period. 3.24 The basic assumption underlying the SONANGOL forecast, namely that the growth of aggregate petroleum product consumption will slow down until the early 1990s, seems to be reasonable. The mission estimates that during the remaining years of the 1980s, the rate of increase will fall below 3%, but rise above 3% in the 1990s in the event of a return to peace. - 53 - Table 3.10: PROJECTED PETROLEUM PRODUCT CONSUMPTION (Tons) Actual Assumed Actual a/ per annum Forecast per annum 1986 1980-86 1992 1986-92 LPG (Butane) 30,200 8.3% 39,213 4.0% Gasoline (99,5% motor) 104,110 4.3% 131,618 4.0% Kerosene 46,310 7.6% 65,692 6.0% Jet fuel 260,550 14.7% 297,051 2,2% Total Kerosene/jet fuel 306,860 13.4% 362,743 2.8% Gasoll 344,770 4.2% 431,346 3.8% Fuel oil 119,200 -0.3% 143,912 3.2% Asphalt 4,20 4,8% 4,730 2.0% Total all products 909,340 6.1% 1,112,562 3.4% a/ Estimated from 9-month actuals. Source: SONANGOL sales department. Pricing of Petroleum Products 3.25 The final price of petroleum products to consumers in Angola is affected by price controls at three different downstream levels: (a) crude oil into the FPA refinery; (b) products ex-refinery (refinery gate); and (c) products for final sale by the distributor, SONANGOL. At present, all of the crude oil feed for the FPA refinery comes from the FPA-operated joint-venture areas. The refinery pays the joint venture crude owners a price that yields to the crude owners an equivalent revenue from a refinery sale as from an export. Since the royalty paid on crude sold to the refinery is deducted from the income tax on production, the crude owners can accept a lower crude price for refining than for export. It also means that the refinery is paying less than the economic (opportunity) value for the crude. 3.26 An illustration of how the system works is provided in Annex 9. In the case considered, the refinery crude prices work out at US$15.25 for Soyo and US$14.40 for Kwanza based on a starting reference marker price of US$20 for "Bonny Light". The forgone Government tax revenues amount to US$2.75 in the case of Soyo and US$2.58 for Kwanza. This loss of Government revenue (US$25-30 million) is a subsidy to domestic consumption of oil. - 54 - Refinery-Gate (RC) Pricing of Products 3.27 The refinery-gate prices of finished products for sale from FPA to SONANGOL are fixed in the official price structure established January 1985 (see Table 3.11). They are based on complete refinery cost recovery including a return on net investment. The relative prices of products were set in approximate accord to international/import parity relationships. At the time this official structure was established, crude prices were much higher than at present and the relative volume assumptions and refinery operating costs have changed from the original basis. Rather than have a continuous, regular refinery gate price adjustment based on changes in crude costs and other factors, the Ministry of Finance's Ceneral Budget (OGE, Or9amento Geral do Estado) acts as a balancing fund while official prices remain stable for long periods of time. 3.28 A prel minary balance on the refinery cost recovery is performed quarterly and any over-recovery (the case at present) is returned to the OGE. Under-recovery based on the official structure would be received as a subsidy from the OCE. About two months into the new year a final balance is made based on verified operating costs for the previous year and any outstanding imbalance is settled. The pricing system is, therefore, a completely "cost-plus" arrangement. There is no particular incentive on the part of FPA to save operating costs as the saving would simply revert to the OGE. 3.29 At recent lower crude :ces and higher production/sales volume there has been considerable 3t-recovery. During 1986, some Kz 2 billioa w'ere returned to Lhe OGE out of total refinery gross sales of Kz 7.5 billion for the year. 3.30 A comparison of present official refinery-gate prices with international prices is provided in Table 3.11. The current structure for all major products is compared with a hypothetical freight and related charges figure of US$20/t added to FOB Mediterranean spot as a basis for border prices for the liquid products. For LPG, a hypothetical freight charge of US$80/t was used, although current LPG imports cost more than this estimated figure. In 1985, the official refinery gate price for all products was below average border prices. Following the decline in crude and product prices in 1986 the official structure moved, on average, above border prices on all products except gasoil. But recent crude and product price increases have undoubtedly pushed international prices to a point where, on an average, the Angola structure is below border prices. - 55 - Table 3.11: OFFICIAL REFINERY GATE PRICES VS. INTERNATIONAL PRICES (Exchange Rate of Kz 29.62 = US$1) FOB Mediterranean plus US$20/t a/ -- Official Structure -- 1985 1986 1st Half 1987 US$/t US$ USS US$ LPG 7.8950 Kz/kg 266.54 304.58 204.75 223.57 Gasoline 4.9989 Kz/liter 228.06 275.50 161.25 186.50 Kerosene/Jet fuel 5.0240 Kz/liter 209.40 281.75 172.58 179.33 Gasoil 3.5488 Kz/liter 140,95 255,00 151,92 171.50 Fuel oil (Heavy) 2.6600 Kz/kg 89.80 167.25 87.42 121.00 Official price as % International LPG 88% 1301 119% Gasoline 83% 141% 122% Kerosene/Jet fuel 74% 121% 117% Gasoil 55% 93% 82% Fuel oil (Heavy) 54% 103% 74% a/ Except LPG for which USS80/t added to FOB Mediterranean. Source: SONANGOL and World Bank. Pricing of Products to Final Consumers 3.31 There is a fixed structure of final prices at which SONANGOL sells the products. This structure was established in 1985 and is still in effect except for a few relatively minor adjustments. The revised pricing scheme as per original documents received from SONANGOL is shown in Annex 9. Allowable SONANGOL profit is 10% of the refinery gate price plus SONANGOL costs and import differential. The latter element is intended to account for the differential between the landed cost of the imported product and the refinery gate cost for domestic-sourced material. The reseller margins have all been increased slightly. The difference between all these allowable costs and the final selling price is made up by a payment to, or subsidy from, the OGE. If these paymeiits/subsidies for each product are multiplied by the 1987 budget- projected volumes of sales for each product, the total gross payment owing to the General Budget for 1987 would appear to be Kz 877 million. SONANGOL is also permitted to deduct its costs of transporting Soyo crude to the refinery. These transport costs were budgeted for 1987 at Kz 111 million. The estimated total net flow to the OGE on this payment/subsidy differential for 1987 would therefore have been Kz 766 million. If we add total budget-projected revenue from product taxes at Kz 1,735 million we arrive at a total Government revenue of Kz 2,501 million on SONANGOL product sales estimated for 1987. - 56 - 3.32 Looking at the COSt recovery on individual products, based on official RG prices and the current SONANGOL cost/profit structure, kerosene, LPG, and LFO (light fuel oil) are the only products which receive a net subsidy from the General Budget, The kerosene net subsidy at Kz 1.2034/liter and LFO at Kz 1.2568/liter are fairly minor, but LPG at Kz 7.1843/kg below cost is a significant distortion. 3.33 Apart from official prices and markets, there is also a parallel market in petroleum products, mainly kerosene and LPG. Even on the outskirts of Luanda, where supplies should be relatively plentiful, kerosene trades at high prices in small quantities. In the interior of Angola, where supplies are extremely short, the parallel market price can reach several hundred Kz/liter for kerosene and similar levels for LPG. There does not seem to be a significant parallel market for products such as gasoline and gasoil, which are normally traded in bulk or through retail service station outlets. 3.34 The fiscal treatment of products is highly discriminatory to the effect that tax considerations rather than relative prices or scarcities determine the allocation of petroleum products. To remedy this problem, products should be taxed roughly equally, at the same rate as gasoline (i.e., 133% of the refinery gate price, adjusted to the level of import/export parity). Arbitrary budget contributions (surtax) could be added to round out the retail price. Also, automotive fuels could be assessed an additional fee to cover road maintenance. 3.35 In order to illustrate the impact of adjustments towards a more reasonable valuation of the Kz, the following pricing analysis will be conducted using an illustrative exchange rate of Kz 104 per US$1 rather than the official rate of 29.62 Kz/US$. The refinery gate price based on world market prices will be recalculated and it is assumed that SONANGOL costs, margins, etc., will rise by about one-third. Since the average level of refinery gate prices was close to world prices in mid-1987, and the structure of refinety-gate prices was closer to economic costs than retail prices were, the refinery-gate prices will be used as the starting point of the exercise, except for the prices of HFO (heavy fuel oil) and gasoil, whici were significantly lower than world prices. 3.36 The adoption of a price structure similar to the above would be a useful step towards a more rational price policy. Such a structure of relative product prices approximates economic costs, thus giving consumers appropriate signals as to the relative scarcity or opportunity cost of products. In addition, if these prices were in effect, the Government would have received approximately Kz 15 billion of the rz 2.5 billion estimated for that year. A more appropriate price level would have to be set in relation to the purchasing power of the Kwanza. To illustrate this, a price build-up at the hypothetical exchange rate of Kz 104/US$l is shown in Table 3.12. At these prices, Government revenues for 1987 would have exceeded Kz 50 billion (20-25 times the actual estima:ed level of Kz 2.5 billion). To the extent that this increase in - 57 - revenues would have helped to reduce the relative magnitude of the budget deficit, it would also have exerted deflationary pressure on the overall price level even though absolute petroleum product prices would have risen only three to four times in the same period. Table 3.12: ILLUSTRATIVE PETROLEUM PRODUCT PRICING-- 1987 (Kzlliter or kg; Kz 104/USS Hypothetical Exchange Rate) Gasoline Kerosena Jet A Gasoil LPG HFO ------------- (Kz/liter) ------------ --- (Kz/kg) --- Refinery gate 17.56 17.61 17.61 14.92 27.68 11.68 Tax 23.36 23.42 23.42 19.84 36.82 15.54 SONANGOL cost 4.22 2.67 1.62 2.58 16.72 0.23 SONANGOL profit 1,09 1.01 0.89 0.79 2.88 0.42 Reseller margin 0.51 1.04 - 0.24 1.62 _ Subtotal 46.74 45.75 43.54 38.37 85.72 27.88 Surtax (budget contribution) 3.26 0.25 6.46 7.63 4.28 2.12 Hypothetical selling prices 50.00 46.00 50.CO 46.00 90.00 30.00 Government revenue 26.62 23.67 29.88 27.47 41.10 17.66 Source: Mission calculations. Economics of Refining vs. Direct Product Supply 3.37 Fundamental Economics of Refining at Luanda. Based on Angola's location in relation to major crude markets and major export refining/product supply sources, and on the marketability of its crude compared to standard world "marker" gradess, there are certain fundamental advantages/disadvantages to refining Angolan crude in the country in order to supply local markets. These advantages/disadvantages are based on the following factors: (a) the opportunity cost to a crude-producing nation of using its own crude as feedstock to an indigenous refinery, which is the revenue it will lose by not exporting the crude; (b) the export opportunity value of crude, which is based on its value in major crude markets in relation to mejor world "marker" grades; and (c) the refinery-gate opportunity value of domestically refined products, which equals: (i) the cost of landing products from major world export refining sources, or (ii) revenue earned - 58 - from export of the products, netted back to the refinery gate (as is the case with heavy fuel, which is currently produced in excess of domestic demand). 3.38 Soyo/Kwanza, the major refined grade in Angola, is not a major export grade since almost all of it is used in local refining. Poor marketability for Soyo compared to the better-know.a Cabinda would make it sell at a lower price than Cabinda in spite of Soyo's slightly better quality. It is estimated that a combined discount on Soyo for freight and marketability, excluding quality, would be about US$0.50/bbl (US$3.65/t). 3.39 The major export refining sources used as basing point standards for pricing the supply of products for the West African coast are the Mediterranean and Northwestern Europe (Rotterdam). If Angola were to import all its product needs, it is reasonable to expecc that an "efficient" freight supply mechanism would cost about US$20/t or US$2.67/bbl. In addition, in order to equate this supply cost to the cost of supplying the products at a refinery gate location, the cost of a receiving and storage terminal operation would have to be included in the calculation. The cost of such an operation on a reasonably large scale would be no less than about US$0.50/bbl. In the particular case of the Luanda refinery, Soyo crude must be transported to the refinery from the same terminal from which it would be exported. An estimated "efficient" cost of such a shuttle movement is US$3/t or US$0.4/bbl. 3.40 The fundamental economic advantage of a refinery in Luanda running local crude to supply local product markets versus a European refinery running the same quality of crude to a similar yield pattern also supplying Angola's product markets is as follows: Locational Advantage to Luanda Refinery vs. European US$/bbl Crude feedstock cost (Soyo) 0.50 Product transport 2.67 Product terminalling cost 0.50 Local transport of crude to refinery (0.40) Total Advantage to Luanda 3.27 3.41 As indicated above, this only considers the fundamental crude and product freight differentials based on the location of Luanda in relation to crude markets and product supplies, as well as a "market- - 59 - ability" evaluation of the refined crude. Operating cost differentials among refineries, considering "efficient" refiners in both cases, are rarely more than US$l/bbl and would not outweigh the fundamental geographical advantage that the Luanda refinery enjoys. What matters, however, are the relative prices that the different petroleum products and crude oil capture in the world market, the output mix, and the percentage share of fuel oil that has to be exported. Based on typical plant yields and an 80% share of fuel oil exports, the refinery operating profit has been calculated for import/export parity levels that prevailed between 1985 and mid-1987. The results are summarized in Table 3.14. A more detailed analysis is provided in Annex 10. The figures show that on average, the refinery would have enjoyed a comfortable profit margin on its "efficient" operating costs during that period if the calculations had been based on border prices. Table 3,14: SUMMARY OF LUANDA REFINERY ECONOMICS LSFO a/ Soyo crude Operating Profit to export discount 1985 1986 1st half 1987 (P ---------------…US$/t -----…--------- Base 80 3.65 8.21 9.69 4.41 Sensitivity I 100 3.65 7.10 8.31 2.11 Sensitivity 11 80 0 4.56 6.04 0.76 a/ Low sulfur fuel oil, Source: Annex 3.3. 3.42 Based on these findings, it can be concluded that the simple hydroskimming refinery in Luanda is an economically justifiable means of supplying local product requizements. Operating profits would prove positive even if: (a) the total LSFO output had to be exported, thus reducing the net-back value of LSFO by the amount of forgone proceeds from domestic sales at higher import border prices (Sensitivity I); or (b) the Soyo crude discount in comparison with Cabinda crude were zero (Sensitivity II). - 60 - IV. ELECTRICrTY SUPPLY Summary and Recommendations 4.1 Angola's power subsector, which still opLrates reasonably well, has suffered from more than a decade of neglect, caused in part by continued civil war, sabotage, losses of qualified personnel and an extreme scarcity of other resources. By 1987 the firm capacity had deteriorated to 275 MW, which is less than 60% of the total installed capacity. Transmission and distribution lines have hardly received any maintenance 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). 4.2 In order to safeguai a reasonable quality of service and to gradually restore the utilities' financial viability, immediate, strong measures are required. Priority should be given to: (a) strengthening the utilities' operational and managerial autonomy and capabilities, including accounting, billing, and the revenue collection system; (b) raising tariffs to generate a cash flow sufficient to cover financial costs; (c) reorienting the investment program to favor rehabilitation of the existing physical infrastructure; and (d) decentralizing responsibility for operations and maintenance, as is largely the case at present. 4.3 A significant and sustained improvement in operations, maintenance and management requires an influx of know-how and finance. Operational support for the Central and Southern Systems as well as advisory assistance to a proposed task force (which will be in charge of all rehabilitation works) would require 75 man-years of long-term consultants plus 25 man-years of short-term consultants at a total estimated cost of US$10 million. 4.4 A minimal priority investment program should be carried out over the next four years to repair and maintain the existing facilities and create some reserve margins to cover rising demand. This would cost about US$200 million. Efforts should be undertaken to obtain financial assistance from bilateral or multilateral aid agencies. Such "soft" financing would help reduce the financial burden of the proposed rehabilitation measures. 4.5 The mission strongly recommends that the existing least-cost capacity expansion plan be updated, based on the best available demand projection. This would also require a reassessment of the future point - 61 - at which the Capanda power project should be reintroduced in the expansion plan. 4.6 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 400%. The utilities' billing and revenue collection procedures should also be improved. However, no attempts should be made to set the tariffs at uniform levels throughout the country. In the short term, tariffs need to be simplified and restructured to enable the utilities to meet simple financial targets (e.g., covering 20%-30% of investment programs or obtaining a return of 3%-4% on assets in use). The adjustments should be designed so as to bring the level and structure of the tariffs in line with Long Run Marginal Costs in the medium term (up to 1992). 4.7 In the past, ENE (the national power company) was given neither the authority nor the means of assuming its (legal) duties of managing the entire subsector in a reasonably efficient way. Also, important mergers of utilities did not take place. Therefore, all operations and maintenance and part of the proposed rehabilitation activities should be decentralized 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 to be responsible for strategic matters (demand studies, capacity planning, tariff studies, etc. and such a unit is being established at present.) Eleccricity Supply 4.8 Electricity supply in Angola is the responsibility of two companies: ENE and SONEFE (Sociedade Nacional de Estudo e Financiamento de Empreendimentos Ultramarinos). ENE is a State enterprise, created in 1980 and incended, eventually, to become the sole nacional power utility in charge of generation, transmission, and medium voltage distribution all over the country. The company is currently operating the "Central" and "Southern" Systems and several isolated systems. SONEFE is in charge of generation and transmission in the "Northern" System, the largest system in the country, and supplies about 300 clients directly at high voltage (60 kV) and medium voltage. Distribution in the area of Luanda is the responsibility of EDEL (Empresa de Electricidade de Luanda). Low voltage distribution in the rest of the country is sometimes the responsibility of ENE but is often handled by local municipal bodies (Comissariados) who may also own small captive diesel sets. Overall Generating Conditions 4.9 Total installed capacity in ENE and SONEFE plants is approximately 463 MW. Of the total capacity, 287 MW is in hydro units, 102 MW in gas turbines, and 74 MW in diesel sets. In 1987 available - 62 - capacity reached 275 MW (59% of total) but there were severe constraints on thermal units due to difficulties in fuel supply. The two gas turbines in Luanda (56.8 MW) burn Jet B fuel while the gas turbine in Cabinda (12.3 MW) runs on natural gas. The two remaining gas turbines, in Biopio (22.8 MW) and Huambo (10 MW), run on diesel oil. Annual electricity generation in Angola peaked at 1,029 CWh in 1974, with 858 GWh (83.4%) from hydro origin. After a sharp decline in the years following independence, generation resumed growth but experienced a second decline over the period 1983-85 and is still below the values of 1974. In 1986, total generation was 754 GWh with 691 (91.7%) from hydro plants. (See Annex 13 for additionial information on electricity assets.) 4.10 Electricity supply in Angola consists of three separate grids and numerous isolated systems. The three main systems are associated with the basins of three important rivers: the Kwanza for the Northern System; the Catumbela for the Central System; aid the Cunene for the Southern System. These systems supply the main load centers in Angola: Luanda (in the Northern System); Benguela, Lobito, and Huambo (in the Central System); and Lubango and Namibe (in the Southern System). The main isolated systems are those of Cabinda, Uige, and Bie. Another important system in the province of Luanda Norte belongs to the mining company ENDIAMA (Empresa Nacional de Diamantes de Angola) and was mainly used for diamond mining activities. 4.11 Hydro has always been the main supply source. Its share has remained within 80-85% of total supply, increasing to 91% in 1986 in spite of the total and partial unavailability of the Lomaum and Biopio plants. No new hydro plant has been built since 1974. From 1980 onwards, SONEFE and ENE tried to overcome the difficulties caused by sabotage and disruption of the hydro supply by installing new gas turbines in Luanda and Huambo and diesel units in Lobito and other major centers. Because of inadequate maintenance, lack of technical assistance and spare parts, and erratic fuel deliveriesp the new facilities have not resolved supply problems. Therefore, current available capacity in the Central System is limited to 47 MW out of the 111 MW installed. Table 4.1 summarizes the installed and available capacities in the various systems and compares them with peak demand. Table 4.1: INSTALLED AVAILABLE GENERATING CAPACITY, 1987 (MW) Hydro Thermal Total Peak System Insti. Avail. Instl. Avail. Instl. Avail. Demand a/ Northern 197.6 135.0 56.8 56.8 254.6 191.8 90-100 Central 49.4 7.2 61,8 39.5 111.2 46.7 30 Southern 27.2 13.6 25.3 15.1 52.5 28.7 9-10 Isolated 12,9 2.4 31.7 5.0 44.6 7.4 na.a a/ Estimated at generation; reflects varying amounts of suppressed demand. Suppressed demand in Luanda (Northern System) is estimated at about 15% of current consumption, Source: SONEFE and ENE, Annex 13. - 63 - Northern System 4.12 The Northern System is operated by SONEFE using the hydro potential of the rivers Kwanza and Donde and two gas turbines in Luanda. The system generated 606 GWh in 1986, or 80% of all the electricity produced in Angola. Peak load at generation oscillates between 90 and 100 MW. The main plant is Cambambe with an installed capacity of 180 MW (four units of 45 Me), the largest in Angola. The plant is run-of-river with daily regulating capabilities, a minimum flow around 130 m3/s (in October), giving a firm power in the critical day of 90 MW. It was generally assumed that the next demand increase in the Northern System would be met by raising the height of the dam at Cambambe by 20 meters, which would increase total installed power to 260 MW. The site was also meant to receive a second power plant, with four units of 110 MW each. 17/ All the units in Cambambe (four units installed in 1963 and two in 1973) require urgent inspection and overhaul (Unit No.1 received a general overhaul and was recommissioned in mid-1986 but work on unit No. 2 is delayed due to shortages of foreign exchange). The situation calls for urgent attention from SONEFE and the MEP as the supply of the current peak (around 100 MW) requires three units in Cambambe or the starting of one gas turbine with high operating costs. The other hydro plant in the system is Mabubas, 54 km northeast of Luanda, built in 1953 with an installed capacity of 17.8 MW. The plant was derated to half its capacity in 1974 and completely shut down in 1986 for a rehabilitation costing about US$7 million, Due to delays in transportation and payments, it will not resume operation before late 1988. Central System 4e13 Two companies operated the Central System before independence: the Hydro Electric Company of the Upper Catumbela (HEAC, Hidro Electrica do Alto Catumbela), in charge of generation and transmission, and thL Electric Company of Lobito and Benguela (CELB, Companhia Electrica do Lobito e Benguela), in charge of distribution in the province of Benguela. In 1982 HEAC was incorporated into ENE as its Central Direction. However, although the merging of CELB was firmly planned, the company is still a separate entity in both the MEP's Plan and in its Reports. The system is based on two hydro plants located on the western course of the Catumbela river: Lomaum (35 MW) and Bi6pio (14.4 MW). A gas turbine of 22.8 MW was installed in Bi6pio in 1974 and a second gas turbine in Huambo in 1981, both interconnected with the rest of the system. The system supplies energy to the provinces of Benguela 17/ The decision to initiate Capanda is currently seen as anticipating an investment included for a later stage in SONEFE expansion plans and postponing for many years (or indefinitely) the investments in Cambambe (see para. 4.100). - 64 - and Huambo, the two main districts of Angola (after Luanda) in population and industrial development. The transmission system runs more or less parallel to the Benguela railway. 4.14 In 1974, the Central System supplied 171 GWh, or nearly 17% of generation in Angola. Peak demand (at generation) was about 29 MW, with a load factor of 67Z. In 1983, the substation at Alto Catumbela and the Lomaum plant were sabotaged, resulting in the flooding of the station and the destruction of the control and protection system. A project has been negotiated for financing part of the rehabilitation of Lomaum and the installation of two additional units of 15 MW each, to increase installed capacity to 65 MW. Excluding the cost of the new sets, the total rehabilitation cost is estimated at US$55 million. 4.15 The gas turbine at Huambo was installed in 1981 and intended for stand-by duty. The turbine has a nameplate rating of 13.0 MW, but derating due to altitude (Huar-ho is at 1,800 meters) limi-s available output to 10 MW. The turbine was bought with a minimum of spares and no service contract and was run continuously but with numerous stops and starts due to fuel shortages or disruptions on the line from Bi6pio to Benguela. Keeping the unit running without the necessary overhaul caused a serious breakdown in 1985. While negotiations for the repair of this unit (and a similar one in Cabinda) go on, Huambo is supplied from Bi6pio (264 km, 150 kV line), and has 4 x 0.8 MW rehabilitated diesels and 2 x 1 MW new diesels. The situation is serious and frequently requires load shedding at peak time. In spite of this, the situation at Huambo is much better than in Cabinda where the breakdown of the gas turbine has practically stopped public electricity supply. The required backup in Huambo is more easily insured by traditional diesel sets than by the gas turbine (which could be moved to a coastal area such as Cabinda to replace diesel sets or to Soyo, where SONANGOL has a project with 15 MW of gas-fired turbines). ENE and SONANGOL should jointly evaluate the interest and feasibility of this transfer, taking into account that gas turbines are naturally advantageous for Soyo and questionable in Huarrbo. Southern System 4.16 The main facility in the Southern System is the hydro plant at Matala, on the Cunene river, with a total capacity of 27.2 MW in two units. A third unit has been stored for several years in Belgium and its installation is now foreseen for the end of 1987. The Gove dam, upstream from Matala, was completed in 1974 for irrigation but mainly to regulate the flow at Matala. Without Gove, the natural low flow at Matala would not permit significant generation during 4-5 months of the year. Firm power in the critical day is now 10.9 MW, but present operating conditions are bad, with Unit No.1 out of order since 1984 and Unit No.2 derated to 60% of its nameplate rating. The plant suffered several fires which damaged the generators, the control room, and the power cables. A complete rehabilitation program has been designed and was supposed to start in late 1987, to replace the control room and power cables. Next - 65 - is the installation of the third unit and the rehabilitation of Units No. 1 and 2. 4.17 Both the Matala and Gove dams present serious safety problems. The 29 gates of Matala cannot be operated due to fractures in the dam's structure. The same occurs with the 3 gates at Cove due to sabotage of the control equipment in 1986. Furthermore, Gove dam, which is of earth-fill Lype, is showing considerable percolation. Unless action is taken to open at least one gate, natural (or disaster) flood conditions could occur and endanger not only Gove but the downstream Matala dam and power plant and, thus, supply to the whole Southern System, let alone loss of life and property.18/ Delays in the process are partly due to the fact that Cove belongs to the Ministry of Construction. An urgent intervention by both parties is necessary to prevent a dramatic accident. The mission suggests that Gove be placed under the responsibility and management of ENE since a power plant of 40 MW is foreseen there later. 4.18 Concerning Matala, ENE consulted several organizations which agree that Matala dam presents serious safety problems resulting from structural anomalies, inoperability of its hydraulic security devices, and complete lack of periodic observations for many years. It is, therefore, risky to start extensive rehabilitation of electrical equipment and installation of a third generator without a rigorous identification (and correction) of the causes of the problems of the dam. 4.19 A diesel plant was built in Namibe in 1980, with an installed capacity of 11.5 MW in two units. Some equipment needed to complete the plant (namely one fuel storage tank) is still missing but the plant has already operated for short periods. Coordinated operation of this facility with Matals is a priority because it is the only supply alternative to the Southern System while Matala is unavailable. ENE may need assistance with the parallel operation of Namibe and Matala as the link between them is very long and weak (330 km with half the distance at 60 kV). Transmission and Distribution 4.20 Transmission and distribution is made at 220, 150, 100, 60, 30, 20, 15, 10, and 6 kV. The proliferation of voltage levels shows that not enough attention was paid to the advantages of standardization. The situation is typical of different companies, each with its own area of influence. The variety of voltage levels increases the difficulties and costs of maintenance and makes it difficult to keep an adequate stock of spares. The number of standard voltage levels 10, 20, and 100 kV should be eliminated from new investments. The need to keep 15 kV (instead of 20 kV) is lnfortunate as this type of equipment is becoming uncommon and expensive. However, since the entire Luanda network is at 15 kV, EDEL has no choice. Total length and characteristics of transmission lines, in 1987, are given in Annex 13, Table 6. 18/ Work to repair the dam started shortly after the mission's visit. - 66 - 4.21 The 220 kV voltage level exists only in the Northern System and is used to carry energy from Cambambe to Luanda (175 km) and to make the injections to Kwanza Norte (73 km, and Kwanza Sul (125 km). The 100 kV voltage level is used only in the Northern System. No further development of 100 kV lines beyond the two existing ones should take place because it is not a standard voltage and because 60 kV and 150 kV are much more extensively used in Angola. Transmission in the Central and Southern Systems is made at 150 kV, with long lines at 60 kV in the Southern System. The lines have had no preventive maintenance since 1975 and are seldom inspected. The cable and most of the towers on the 60 kV Namibe-Tombwa line (95 km long) are very badly corroded due to a mixture of salt air and desert winds. ENE acknowledges the urgent need to replace cable and wires as well as to rehabilitate the towers, but lack of staff, inefficient organization, and scarcity of foreign exchange are delaying any decision. Electricity Demand Past Situation 4.22 Until 1974, electricity generation and consumption statistics in Angola were regularly established and published, giving a rather complete and reliable picture of its situation and evolution. Because of deterioration of metering equipment in power plants and substations, frequent outages (by sabotage or poor maintenance) of generating and transmission facilities, poor record-keeping, and lack of communication between regional centers and ENE headquarters in Luanda, the quality and reliabiiity of electricity statistics have declined. Forced to fill out forms in the absence of hard information, the utilities often supply unreliable, inconsistent figures to satisfy reporting requirements. 4.23 Growing difficulties of supply due to outages or sabotage led the Government, through ENE, to install a large number of small and medium-sized captive diesel units (up to 1.5 MW of rated power) at the facilities of local authorities (Comissariados) and major consumers. Little is known about these sets, yet they may represent a non-negligible share of total generation in Angola, not only in isolated centers but also in areas supplied by one of the three main systems. In 1986, about 200 new diesel units were acquired by ENE and installed all over the country with rated power estimated between 30 and 40 MW (total cost of pounds sterling 30 million). Production from these units partially replaced unavailable supply from the grid but is not accounted for in ENE or MEP statistics. 4.24 Gross electricity generation in Angola totaled 754 GWh in 1986. Final consumption must be estimated, as figures for electricity distribution in the MEP's reports are established from utility estimates of emission to distribution grids. Assuming an overall figure of 22% for transmission and distribution losses, total consumption may be estimated at about 590 GWh in 1986. The share of total generation among the different systems in 1986 waq as follows: 80.4% in the Northern System, - 67 - 10.0% in the Central System, 7.5% in the Southern System, and 2.1% in isolated centers. The absolute values and the pattern of generation and consumption in 1986 are not typical of earlier periods, either before or after independence, as they were constrained by temporary irregularities of supply and demand whose duration is not known. The present overall situation and past trends since 1967 are shown in more detail in Annex 13, Table 2. 4.25 Between 1967 and 1973, generation and consumption increased at an average rate of about 16.3%. The bulk of growth was concentrated in the six major load centers: Luanda, in the Northern System; Huambo, Lobito, and Benguela, in the Central System; and Lubango and Namibe in the Southern System. Only one additional load center grew in relative importance during that period--the isolated system of Cabinda. After independence, in 1975, consumption declined sharply to only about half the level of 1974. Starting in 1977, a slow recovery process lasted until 1982. Annual growth rates during that period averaged 6.0% and consumption in 1982 was roughly equal to that of 1971. Between 1983 and 1985 a new period of decline occurred, due to intensification of sabotage activities. A whole plant (Lomaum, 35 MW) and a substation (Alto Catumbela) were severely damaged and put out of order, constraining the supply in the Central System. In the Northern System, supply to the whole province of Kwanza Sul, estimated at about 20 GWh, was interrupted due to the sabotage of the 20 kV Cambambe-Cabela line. Supply had not yet been restored in May 1987. The overall negative trend of the period 1983-85 was reversed in 1986 mainly because of faster growth in the Luanda area. 4.26 In 1974, over 50% of the load was already concentrated in the provinces of Luanda and Bengo, which encompass the Luanda metropolitan area. The absolute consumption of Luanda is currently slightly higher than in 1974, and its share of the total consumption of Angola has increased to nearly 76%. The most interesting years for comparison are 1974 (the last year before independenc-), 1982 (the last year before the second declining period), and 1986. In 1974, 83% of total consumption was concentrated in the five provinces with the six main load centers. In 1986 the share of the same provinces had increased to nearly 96%. Between 1974 and 1986 Luanda's share increased while that of the other four main provinces decreased. Decreases in Benguela and Huambo were not only due to a reduction in economic activity but also to supply constraints, as the Central System facilities which serve both provinces have been particularly hard hit by acts of sabotage. Generation and consumption over the last five years are summarized in Annex 13, Tables 3 and 7, showing the contribution of each system to global supply and the share of each province in total consumption. Luanda 4.27 The consumption of the Northern System is concentrated in Luanda, which represents more than 90% of the total system load. The main part of the Luanda load is EDEL's distribution network in the urban area, the rest being industrial consumers (around 300) directly supplied - 68 - at 60 kV, 30 kV, and 15 kV by SONEFE, and some local authorities (Comissariados). Consumption has been stagnant over the period 1982-86, at a level between 480 and 500 GWh. Accepting EDEL estimates of 12% technical and 9% non-technical losses in its distribution network, total sales in the Northern System, in 1986, may have reached about 440 GWh for a total generation of 605 GWh (total losses of 27%). Figures should be taken with caution as the large variations between consecutive years may be due to incorrect measurements (different time periods, for instance) or long and frequent interruption of supply (at any voltage level). A long-run average annual growth rate of 5% is plausible (see Annex 13, Table 8 for further details). Central and Southern Systems 4.28 The disruption of supply in the Central System is the main cause of decline in overall Angolan generation and consumption since 1982. The share of the Central System in total generation which represented 20% in 1982 declined to 10% in 1986, and consumption in 1986 (79 GWh) was roughly equivalent to half the value of 1982 (145 GWh). The Central System has the highest level of suppressed demand, and more than 50% of the present generation is of thermal origin (from the gas turbine in Biopio with a specific consumption of at least 0.360 kg/kWh or 0.425 liter/kWh). Average operating costs, just for fuel, amount to Kz 3.5/kWh. Rehabilitation of the Lomaum hydro plant, foreseen for early 1989, is thus a priority cost-cutting task for ENE. 4.29 The Southern System currently represents 5.5% to 7.5% of overall generation. No significant disruption of supply has occurred and consumption has slowly but regularly increased. Some potential repressed demand is caused by shortage of materials for network extensions and connections. However, the system depends on practically a single source, the Matala hydro plant. This plant currently has safety problems, partly local (in the dam) and partly related to safety problems in the upstream Cove dam. Furthermore, ENE plans to install a third unit at Matala which will require an interruption of supply. Lubango and Namibe may thus experience serious supply problems, and ENE has not yet taken any steps (e.g., completing the Namibe thermal plant) to minimize the interruptions. 4.30 The sbare of isolated centers, which in 1974 accounted for a total consumption of 115 CWh, (12.5% of the overall consumption of the country) fell noticeably, to 30 CWh (or 2%), after the breakdown of the Cabinda gas turbine. Aging of units, lack of maintenance and spares, difficulties of fuel supply, the complete cessation of mining activities in Luanda province (representing between 50-60 GWh), and a marked decrease in commercial and industrial activities were the main cause of the substantial reduction of that share. However, part of the already existing demand is likely to be supplied by captive diesel sets, not accounted for in global statistics. - 69 - Load Curve 4.31 Information on the daily and seasonal pattern of demand is only available for the Luanda area. Present coincident peak demand is about 90 MW, of which 70 MW corresponds to the EDEL network. For the whole Northern System, annual peak demand measured in Cambambe oscillated between 80 and 100 MW over the period 1978-86. 4.32 For the main ioad center of Luanda, yearly load factors vary in range between 0.63 and 0.70. Working day load factors are about 0.80. The daily load curve normally features two peaks: a morning peak, lasting from 10.00 to 13.00 and a higher evening peak, between 19.00 and 21.00. Might load remains at about 65% of peak load, indicating a fairly high permanent base load. Night load is mainly due to air conditioning and refrigerators together with public lighting. ENE statistics do not state the peak demand associated with the Central or Southern systems. Present peak demand in Huambo is around 9 MW, for a total of 13,000 customers, but load shedding at evening peak may be quite extensive and an unknown portion of this peak may be covered by captive diesel sets. A 9 MW value can be considered conservative, and could increase 20% to 30% in normal supply conditions from the grid. Peak demand in the Southern System is about 9.5 MW, measured at the Matala power plant, with a peak demand of 5 MW in Lubango, the main load center. Demand Projections 4.33 In the present context there is no firm basis for making demand projections in Angola with any reasonable degree of confidence. The difficult economic and security conditions of the country, even if considered transient, impede any demand forecasts. Application of standard forecasting methodologies (either the use of statistical data to define approximate laws of probable evolution for already electrified regions, or comparison with areas with similar geographical, demographic, and economic conditions for regions to be electrified) is hindered by an exceptionally high degree of uncertainty about future economic activity, which strongly depends on the return of peace to the country. Past records of power system performance are inadequate and unreliable and do not correspond to historical trends in a stable economic framework, 4.34 The utilities have no idea of the potential number of clients and potential consumption. It is thus understandable that ENE and SONEFE have given up making any medium- or long-term forecasts. Figures sent to MEP and included in the Plan Report of the Ministry cover only 1987 and 1988, and are considered perfunctory by the utilities. Forecasts are not based on a breakdown of total consumption by customer class or load center but only by province, and bear a rough relationship to the values of the two previous years. - 70 - SONEFE: Northern System 4.35 SONEFE estimates an annual growth rate of 2% for the Northern System 19/ under the joint effect of two opposite trends: an increase in residential consumption, mainly due to the connection of new customers, and a stagnation (or even decline) of industrial consumption. Global projected generation is distributed among the existing power plants, assuming a certain availability of units and pre-established minimum generation from thermal origin. "Distribution" is calculated using generation forecasts, assuming a fixed value of 10.6% for transmission losses and internal consumption (Cambambe village and installations). This percentage is a traditional value which SONEFE has no means to verify or correct. The share among the four provinces supplied by SONEFE is calculated on the basis of recent figures (1985 and 1986), taking into consideration the actual unavailability and expected repair times of major transmission lines. According to SONEFE, the consumption of the greater Luanda area in 1987 and 1988 is supposed to represent 94.1% and 93.2%, respectively, of the total consumption of the Northern System. ENE: Central and Souchern Systems 4.36 Generation and consumption projections in ENE are even more primitive than in SONEFE. The staff assigned to prepare projections is cxtremely reduced and ill-prepared. It faces two additional difficulties in comparison with SONEFE: the need to deal with several systems of variable extension and characteristics distributed all over the territory and the lack of information, at ENE headquarters in Luanda, on reliable historical consumption figures and realistic constraints on insta!led capacity. )3NE is currently assuming that all the systems are supply- constrained (which is essentially correct) so that primary forecasting concerns generation according to expected available supply facilities. Forecasts for 1987 and 1988 are optimistic concerning the speed of rehabilitation. Projections of "distribution" were established assuming a certain percentage of losses, arbitrarily set at 4% for the Central System and 9% for the Southern System. "Distribution" should be understood as deliveries to the distribution network at the high voltage busbars of HV/MV substations, although 4% of losses is too low for losses in a long transmission system like the Central System. ENE and SONEFE projections are summarized in Table 4.2. 19/ In October 1988, SONEFE surmised that a repressed demand of 15 '1W might exist in Luanda, and 10 MW in the rest of the Northurn System. It also thought that peak demand could increase by 6% in 1988 and that average annual growth in the period 1988-1998 could be 6%. The mission's estimates tor the same period vary between 6.7% (base case) and 9.3% (high case). - 71 - Table 4.2: ELECTRICITY GENERATION AND CONSUMPTION PROJECTIONS (GWh) Projections 1986 (Actual) .987 1988 System Generation Consumption Generation Consumption Generation Consumption Northern 605.9 506.1 615.9 551.0 628.3 561.6 (Lianda Metro shiare) (491.9) (518.4) (523.3) Cenitral 75.2 72.6 71.6 68,7 83.8 80.5 Southern 56.3 43.3 57.4 51.7 58.0 52.2 Isolated 16.0 14.2 25.5 24.4 27.0 25.8 Total 753.4 636.2 770.3 695.8 797.1 720.1 SONEFE 605,9 506.1 615.9 551.0 628.3 561.6 ENE 147,5 130.1 154.4 144.8 168.8 158.5 Total 753.4 636.2 770.3 695.8 797.1 720.1 Average Growth Rate, 1986-88 (U p.a.): System Generation Consumption SONEFE 1.83 5.34 ENE 7.00 10.40 Total 2.86 6.39 Source: ENE, SONEFE, MEP, and mission estimates. 4037 ENE and SONEFE face a number of serious and difficalt planning issues for the development of the three main systems as well as their possible interconnection. Serious thought about expansion plans and investment options requires some knowledge of future (long-term) electricity demand. The only possible approach 's to construct some alternative scenarios for the evolution of demand, which cannot be considered forecasts, but could be used to test policies for supplying load centers and to test alternative timings and options for major development of new power plants and transmission facilities and acceptable delays and cost implementations for major investment decisions. Alternative Demand Projections 4.38 SONEFE and ENE do not prepare medium- or long-term demand projections. They only compile short-term (2 years) perfunctory figures to satisfy MEP reporting requirements. However, over the years many - 72 - studies have been done about various aspects of Angola's power subsectors. This report will limit itself to a review of the two most recent ones 20/ and will also present its own set of demand projections based on past studies and on specific hypothesis about the pace of economic rehabilitation in Angola. The various projections are summarized in Table 4.3 and compared with present installed and available capacity. The main assumptions are listed after the table while details and specifics are described in Annex 14. Table 4.3: ENERGY AND DEMAND PROJECTIONS, 1986-2000 Comparison f Generation Requirements (Base Cases) Energy (GWh) Peak Demand (MW) Capacity (MW) BEP THEMAG Mission BEP THEMAG Mission installed Available Northern System 1987 554 693 618 102.3 123.7 112.0 254.6 191.8 1990 732 943 656 141.1 167.6 118.8 1995 1,278 1,514 859 244.0 274.3 155.7 2000 1,876 2,225 1,461 355.0 403.0 264.7 Central System 1987 217 175 98 43.2 37.1 18.8 111.2 46.7 1990 263 244 115 52.8 52.8 21.8 1995 528 375 196 103.3 81.3 35.7 2000 743 551 329 144.0 119.5 60.1 Southern System 1987 57 81 55 1.8 17.4 10.8 44.6 28.7 1990 67 102 58 14.0 22.2 11.7 1995 132 157 85 26.8 34.1 16.9 2000 193 231 124 38.6 50.1 24.9 UNDP: The United Nations Development Program. Source: BEP, THEMAG, mission estimates, and Annex 14. 20 Belgian Engineering Promotion: "Etude d'un Plan Directeur du Reseau Electrique National d'Angola", Bruxelles (financed by the African Development Bank); and Themag Engenharia: "Interligacao dos Sistemas Norte/Centro/Sul em Angola, Possibilidades de Interligacao com a Namibia", Estudo Preliminar, Sao Paulo, Outubro 1986 (done within the SDACC framework). - 73 - 4.39 Table 4.3 shows that this report's projections, based on relatively sanguine hypotheses, are much lower than the base cases of other studies, which must be considered excessively optimistic. If 1BRD/UNDP projection figures are even only broadly indicative of future needs, present investment plans will overshoot actual demand substantially, especially if a tariff reform were to take place. These projections further show that demand can be met substantially until the end of the 1990s by rehabilitating presently installed capacity, except in the Northern System where some new capacity will be needed after 1995. Electricity Tariffs and Utility Finances 4.40 Electricity tariffs have not changed since the 1960s except in newly electrified areas where rates have remained constant since service started. As a result, wide variations in tariffs can be observed in areas served by the same interconnected system. Variations of the order of 3 to 1 are not uncommon. However, absolute tariff levels are so low that consumers perceive electricity to be essentially free. Consumption has therefore increased, with limits set only by supply and appliance constraints. 4.41 A complex and widely varying declining block structure prevails for low voltage tariffs (often with a minimum monthly bill), with higher rates for residential consumers and lower rates for industrial consumers. In the EDEL distribution area (i.e., greater Luanda) LV tariffs discriminate among residential, commercial, industrial, and public lighting uses. Residential and commercial tariffs have a very complex declining block structure, with block sizes related to house sizes or floor area and minimum monthly bills related to meter caliber. The LV industrial tariff, covering only 2% of sales, is a three-period time-of-day tariff with energy rates declining with load factor and meter caliber. Public lighting is charged at a flat rate, which is about one- half of the average price of other LV sales. 4.42 HV tariffs usually include block rates whose size is a function of (non-coincident) peak demand, leading to average prices declining with load factor. Large consumers benefit from contracts negotiated long ago, at times of excess supply capacity. High voltage (HV) tariffs take as billing variables non-coincident peak demand and active and reactive energy. Energy rates decline with the load factor, that is, utilization of peak demand. Demand rates increase with the load factor. In Luanda, the marginal price for active energy declines from Kz 1.1/kWh (load factor of 12.5%) to Kz 0.85/kWh (load factor of 25%) and Kz 0.505/kWh (load factor of 100%). Reactive energy is charged through a monthly bill multiplier: it is free up to 60% of active energy (cos 0 = 0.8) and is charged at a price rising to 63% of the normal rate when it reaches 92% of active energy (cos 0 = 0.4). - 74 - 4.43 This tariff structure cannot be adjusted to the pattern of margindl costs in the Northern System. These are determined by the generation and transmission capacity needed to meet peak demand (at present, only 3 to 4 hours in the evening, and, in the future, also a second industrial peak period during the morning) and by local grid capacity requirements determined by maximum demand. Tariff reform would therefore require significant changes. 4.44 In the Luanda area, tariffs were set in 1962. Residential tariffs consist of three declining blocks varying with house size (12 classes) and meter size (9 classes). The rates vary from Kz 2.50/kWh (first block) to Kz 0.70/kWh (for the third block). The average price of electricity sold is about Kz 0.83/kWh. Industrial and HV tariffs are even lower, as low as Kz 0.55/kWh. 4.45 A rough estimate of the cost of producing electricity in Angola shows that energy costs alone in the ENE systems amount to about Kz 5-6/Kz ($US0.17-0.19/kWh). SONEFE's energy costs (almost 100% hydro) are about Kz 1.0/kWh. 21/ This superficial comparison of tariffs with costs points to one cause of the poor financial performance and condition of the power utilities. Other reasons are that costs are not known accurately, that the utilities are overstaffed, and that billing and collection are extremely inefficient. Accounting 4.46 The present state of financial reporting makes it almost impossible to assess the financial condition of the utilities. ENE has never produced statements of accounts since its establishment in 1980. A list of ENE's inherited fixed assets and inventories, financial claims, and liabilities was prepared for the approval of its initial net worth, but so far no accounting records have even been made. 4.47 EDEL's legal status as a public enterprise was unclear at the time of the mission.22/ The utility presented a report dated August 1984 showing a balance sheet as of December 31, 1982 based on asset values submitted for Government approval, and is preparing the statement of accounts fur 1983. These documents were never approved by the Government. 4.48 SONEFE has closed accounts for 1981 and 1982. Assets are valued at historical book value. Capital and liabilites do not reflect the de facto legal status of the utility, i.e., a State enterprise under 21/ Distribution costs alone add another Kz 0.5/kWh. This would give total average costs (excluding capacity) of Kz 1.3 to Kz 2.5/kWh in the Luanda System. 22/ EDEL's status has been legalized since. - 75 - the guise of a private corporation. Poor enforcement of accounting principles and the mixing of traditional and new accounting practices, partially influenced by new national standards (set in "Plano Nacional de Contas") produce statements of accounts that require extensive revision to come closer to international standards. 4.49 No accounting records for CELB are known. Although the transformation of CELB into ENE's Central Region is firmly planned, CELB is still considered a separate unit in plan execution and forecasting reports. 4.50 In conclusion, those power utilities that have accounts at all have extremely weak accounting procedures. While both the utilities and the Government are aware of this, the remedies being proposed are probably not the most effective. The utiUities are planning to prepare accounts for the last several years (ENE since its conception in 1980, EDEL since 1982) and the Government is proposing a new, uniform accounting plan for State enterprises. The relative merits of these proposals are questioned and a more modest suggestion, better matched to the abilities of the utilities, will be made in the following paragraphs. Financial Situation of the Power Utilities 4.51 The power utilities together incurred a global cash deficit of about Kz 1.4 billion (US$46.7 million) in 1985. Total sales revenue that year was about Kz 650 million (US$21.7 million) or 46.5% of the cash deficit. In 1986, thanks to lower investment expenditures, the total cash deficit fell to Kz 1.1 billion or 176% of consolidated sales revenues. This is a significant share of the overall budget deficit which was estimated at Kz 13 billion in 1986. Given the sfiort-term plans of the utilities, cash deficits are expected to increase in 1987 and 1988 (to about Kz 1.4 billion and Kz 1.5 billion, respectively). These cash shortfalls include about 20% of capital expenditure (however difficult that is to estimate). 4.52 Total fixed assets of the power utilities in (about) 1982 were estimated at about Kz 4.5 billion (US$150 million). A net return of 3% to 4% on these assets would require the utilities to produce a consolidated profit of Kz 150-180 million (in addition to covering costs). This is far from the present situation. However, since the cash deficit figures quoted above included some investment expenditures, covering those cash deficits (as presently estimated) would be a reasonable short-term goal for the power utilities. This would require multiplying the present tariffs by an average of 3 to 4 (excluding any exchange rate consideration--more than 50% of operating expenditures of the utilities are in foreign exchange). 4.53 Table 4.4 summarizes the cash flow situation of the power subsector as a whole. It is interesting to note that although ENE sells only one-third as much energy as EDEL (or one-fourth as much as SONEFE), - 76 - its total costs, on a cash flow basis, are 250% higher. 23/ This is due mainly to the much larger proportion of thermal generation by ENE (even though diesel oil is sold at the low price of Kz 7/liter). Table 4.4: ANGOLA: ELECTRIC POWER UTILITIES, SALES AND CASH FLOW, 1986 (Million Kz) ENE CELB SONEFE EDEL TOTAL SALES (GWh) 123.0 n.a. 506.1 365.9 995.0 (intrasectoral) (417.2) (417.2) Net Consolidated Sales 123.0 88.9 365.9 577.8 CASH INFLOWS (million Kz) 259.7 118.4 346.2 329.2 1,053.5 (Intrasectoral) (35.0) (207.8) (242.8) Net Consolidated Cash Inflows 224.7 118.4 138.4 329.2 810.7 CASH CJUTFLOWS (million Kz) 1,022,1 196.7 523.5 449.1 2,191.4 (Intrasectoral) ____ (35.0) (207.9) (242.9) Net Consolidated Casn Outflows 1,022.1 161.7 523.5 241.2 1,948.5 CASH DEFICIT (million Kz) 762.4 78.3 177.3 119.9 1,137.9 FINANCING OF CASH DEFICIT Net Budget Inflows 366.6 43.3 -38.0 123.8 495.7 Other Transf'rs a/ 395.8 35.0 215.3 -3.8 642.2 and Variations in Cash Balances a/ Including payment arrears. Source: ENE, EDEL, SONEFE, the MEP, and Mission estimates. External Debt of the Power Subsector 4.54 In 1987, the electric power subsector, excluding GAMEK, had a debt of about US$101 million, of which about US$60 million was disbursed and outstanding. The average maturity was about 8.5 years including a grace period of 2.5 years, and the average interest rate was about 8.5%. These conditions are typical of export financing schemes in the advanced market economies where most of these loans originate. In 1987, the amount needed to service this debt was about $US17 million. It is expected to rise to about US$20 million in the years 1988 to 1989, then to decline to US$15-20 million in the early 1990s. Annex 13, Table 9 gives a more detailed breakdown of available information. GAMEK's 23/ Correcting for capital expenditures (roughly estimated), ENE's costs would still be nearly twice those of EDEL. - 77 - currently disbursed debt was about US$250 million as of mid-1987 out of commitments of about US$600 million from various external creditors. It is expected to increase at about US$200 million/y in the next three years. Debt service is currently US$20-25 million/y, or a little more than the debt service of the rest of the power subsector. However, if the Capanda project were to go ahead, GAMEK's debt service would reach US$150 million by 1990. Some part of this amount may be repaid in petroleum. Billing and Collection 4.55 Billing and collection difficulties are significant in Luanda. Electricity consumption for the last four years has not been recovered. EDEL considers one of the four years (still unbilled) to be recoverable. Consumption for 1986 was expected to be billed all at once, although six-monthly bills were used in 1985. It is obvious that such lengthy billing periods are only possible when electricity tariffs are negligible. This is the case in Luanda at present. 24/ The problem of billing is about to be resolved through computerization and because EDEL management is giving priority to this problem. 4.56 In Luanda, LV collections are effected door-to-door. EDEL employees report serious difficulties in collecting payments. HV and MV consumers pay their bills through bank account transfers ordered by the utilities. The problems are therefore essentially limited to LV residential customers. 4.57 Outside Luanda, LV customers pay their bills at banks or utility offices and recoveries are a high percentage of billings. Severe penalties and disconnections are actively imposed to enforce punctual payments. HV and MV consumers also pay through bank transfers. Manpower, Staffing, and Technical Assistance 4.58 In mid-1987 the power utilities (EDEL, ENE, SONEFE, CELB) employed 3,830 persons, of whom 30 were senior-level staff (engineering, managerial, and technical grades), 200 were skilled workers, less than 1,000 were semi-skilled workers, and the remaining 2,600 were unskilled workers. The clerical staff constituted 23% of the total labor force. 4.59 To make up for the shortage of qualified personnel, the utilities recruited 104 expatriates (21 senior-level staff, 30 skilled workers, and 53 semi-skilled). Except for some skilled workers and all the semi-skilled workers (most of whom ate Angolan residents even if they are classified as technical assistants), the utilities recruited the 24/ For example, one year's average consumption of electricity in Luanda is equivalent to the parallel market price of two eggs (i.e., about Kz 1,000). - 78 - technicians either under bilateral agreements (U.S.S.R., 9; Cuba, 11) or under individual contracts (Portugal, 18). Recently, however, the Government passed a law withdrawing some contractual advantages for expatriates (who are also long-term residents). These expatriates are therefore unlikely to stay on and will probably seek work elsewhere, or in other sectors, where perquisites are better. Misallocation of Qualified Manpower 4.60 To tackle this difficult situation, the power sector should design a manpower policy targeting the operational manpower requirements. This policy should combine: (i) effective allocation of Angolan skills; (ii) technical assistance; and (iii) training. The present allocation of skilled manpower resources reflects recent hiring (of mainly students) which causes low productivity and the lopsided organization of the power utilities (76%o of senior-level staff perform non-operational duties and only 50% of the skilled workers are directly involved in operations). Furthermore, skilled personnel are relatively more scarce in the Central and Southern systems (in comparison with the Northern region), and in all parts of the country, the distribution function is least favored with respect to qualified manpower. Typical utility staffing profiles would be roughly as shown in Table 4.5, Column B. The actual staffing profiles of the utilities in Angola are shown in column A. Table 4.5: ANGOLA: ELECTRIC POWER SUBSECTOR Theoretical and Actual Staff Profiles (1987) (Percentage) A B Actual Staff Distribution Theoretical Staff Distribution Generation and Transmission 31 35 Distribution 26 50 Functional Services 43 15 Totals 100 100 Source: Angolan authorities and mission estimates. In addition to this inappropriate deployment, the utilities are excessively overstaffed with unskilled workers and low-skill clerical workers. Although salaries have been frozen since 1977, the utiLities' labor costs contribute to financial losses. Labor costs amount to 35% of billed sales in EDEL, 60% in CELB, 29% in SONEFE, and 118% in ENE. Since revenue collection (except for SONEFE and CELB) is low, labor costs are - 79 - an even higher proportion of sales revenues and generally require subsidization through the Budget. Technical Assistance 4.61 At present, the recruitment of expatriates for technical assistence does not follow any consistent policy. State-to-State arrangements (U.S.S.R. and Cuba) or individual contracts are entered into on an ad hoc basis. These expatriates generally work in isolation and qualified Angolan counterparts are extremAly scarce. Furthermore the poor management of these skills, the lack of a structure into which they could be integrated, and the absence of performance monitoring threaten the cost--effectiveness of this assistance. Consi.iering the level of staff shortages and the weak institutions of the power subsector, tech- nical assistance deserves careful attention. In the Angolan context, individual contracts, particularly those shorter than two years, are seldom effective. Main Issues and Recommendations 4.62 As is evident in the preceding description, the main issues in the electric power subsector are: (a) organization; (b) management; (c) financial situation, tariffs, billing, and collections; (d) qualified manpower, training and technical assistance; and (e) investment planning and priorities. Organization 4.63 A major issue in the organization of the subsector is the determination of an appropriate role and structure for ENE. Formally established in 1980, ENE has not yet been able to assume the role of national electricity utility it was intended to play. The various utilities that should have been integrated into the ENE structure still retain autonomous management and operating procedures. A strong sense of autonomy is felt at the regional level. The only important link with the center is the dependence of the regional departments on headquarters for purchases of imported materials/supplies. Meanwhile, the current Angolan unrest exacerbates transportation and communication difficulties. Also, the fact that 80% of total generation and consumption are outside ENE complicates the establishment of a national power company. Since SONEFE and EDEL were kept out of ENE, what emerged from its creation was an entity responsible for most of the country but without control over the - 80 - main generating system and the main load center, and with its headquarters and most of its qualified staff in Luanda, far from any of its operating installations. 4.64 ENE has also been acting as a service company, importing supplies and electric materials and selling electrical appliances. 25/ The fact that ENE often supports local authorities (Commissariados) would suggest that it be in charge of LV distribution in the Central and Southern areas which are supplied by its interconnected systems. The rationale behind such an organization is the vertical integration of functions (generation, transmission, distribution) within each regional interconnected system supported by the physical infrastructure of the network. 4.65 The importance of the tasks to be carried out together with the deeply troubled external environment and the day-to-day pressures on the reduced number of managers and skilled personnel of ENE does not permit headquarters to effectively supervise the Regional Directorates. It would seem more appropriate to decentralize all operations and maintenance as well as most of the rehabilitation activities to the Regional Directorates. At the central level, a small central unit should be created, close to the Director General. The unit would have planning and standardization functions and should give priority to demand studies (which ENE lar`ks), expansion alternatives, and tariff studies. The unit would also be responsible for standardizing accounting and budgeting procedures, standardizing technical characteristics of equipment and maintenance and safety procedures, auditing, evaluating consultants' reports or project proposals, negotiating external credits, and setting guidelines for system expansion. 4.66 In fact, given ENE's lack of success, the question arises whether the initial idea of centralizing everything is appropriate. After thorough review, it would seem that, under present conditions, centralization is not the preferred option. ENE has become a rather top- heavy organization with almost all qualified staff assigned at headquarters while most of the activities (generation, distribution) are in the Central and Southern Directions. Furthermore, communications between the "frontlines", as it were, and headquarters are deficient, so that headquarters can contribute but little to solving problems at the operations level. For all these reasons, of which the Angolan authorities are well aware, the optimal amount of centralization in the electric power subsector is less, probably much less, than was initially attempted by the creation of ENE. While not wanting to go into unnecessary detail, this Report suggests much greater decentralization, with most functions centered on the physical systems themselves. 25/ The importing and selling of materials and electrical appliances could be better handled by a separate company. It is recommended that the Government of Angola set up such a company, thereby freeing ENE from activities marginal to its main role. - 81 - 4.67 Whether the Southern and Central systems should be separate companies or very autonomous divisions of a reorganized ENE should be decided by the Government after careful analysis. Some centralized functions would need to be carried out for the whole power system, whether in a modified ENE or an "Electricity Commission" (CNE) between the operating utilities and the MEP. The main tasks in question are: (a) system planning and the standardization of equipment, safety procedures, procurement, and auditing; (b) evaluation of consultant studies; (c) coordination/execution/evaluation of demand studies; (d) setting up accounting and budgeting procedures for all utilities; and (e) preparation/updating of a least-cost system expansion program for the grids. 4.68 Part of the staff needed for these functions could come from ENE's present nucleus of staff, and part could come from the MEP's Technical Department. Other qualified staff should be assigned in priority to the operating utilities or divisions. 4.69 As for the Northern System, the merger of SONEFE and ENE should be postponed at least until: (a) final decisions have been taken on the organization of the subsector; (b) accounting and budgeting procedures have been defined and put into practice; and (c) the assets and liabilities of both companies have been ascertained and evaluated properly. SONEFE would either become a very autonomous division of a reorganized ENE or a separate entity, but subject to the centralized functions enumerated earlier. 4.70 This report suggests that distribution in the Central and Southern systems be integrated with the generation and transmission activities of the new utilities or autonomous divisions of ENE. In the Northern System, the size of the distribution task would argue in favor of retaining EDEL as a separate entity. 4.71 In management terms, the subsector would thus be made up of the following managerial units: (a) a central ENE or national electricity commission (CNE, Comissao Nacional de Electricidade) in charge of tasks that can or must effectively be carried out in a centralized fashion); (b) a Northern Division or Utility, i.e., SONEFE, in charge of generation in the northern grid; (c) a Northern Distribution Division or Company, i.e., EDEL; - 82 - (d) a Central Division or Utility in charge of generation, transmission, and distribution in the central grid; (e) a Southern Division or Utility, in charge of generation, transmission, and distribution in the southern grid; and (f) a number of isolated systems which could either be directly managed by ENE (e.g., the Cabinda or Soyo groups) or by one of the regional utilities, whichever option is the most convenient. Management 4.72 The utilities provide the MEP with plans and annual reports that could be valuable tools for the coordination of sector/subsector policy and investment. However, these plans and reports have major weaknesses which seriously hamper their usefulness both to utility management and to MEP coordination. The main managerial problems concern the MEP, the utilities, and the relations of both with the rest of the economy. They are briefly described in the following paragraphs. (Further information on institutional arrangements and problems is given in Annex 2): (a) inadequate accounting, budgeting, and financing systems in the utilities, which give a poor, incorrect, and frequently distorted picture of their situation and performance. Within the utilities, the departments which prepare the Plan (Budget) sometimes do not fully comprehend the official forms. Their contributions are not consistent with each other. No department seems able to consolidate the information received from the various sources. (b) poor intersectoral coordination. Coordination should take place at the Ministry level and be communicated to the utilities. There is no reliable information on new industrial loads. Major investments may be decided and carried out without regard to energy availability. On the other hand, utilities may be building network extensiotns for projects that have no priority within their own ministries. Coordination is poor between the utilities and the Ministry, and between the utilities and other Government agencies (such as the National Bank, whose control--or lack of control--over the foreign exchange budget exerts a major influence over the performance of the utilities); (c) lack of a decision-making process based on or supported by sound technical, economic, and financial appraisal of the projects. Investment plans are shopping lists of works shifted around from year to year as foreign exchange becomes scarcer; - 83 - (d) lack of planning capabilities to prioritize investments within budget constraints and allocate resources according to preestablished criteria. Neither the utilities nor the Ministry are able to judge the relative merits of different investments either in terms of their economic return or their impact on preventing future disruptions or meeting demand requirements. Investments are often decided not on economic grounds but based on offers of credit or political pressures; (e) total lack of information on major ongoing investments in the power subsector. This ignorance extends even to entities responsible for preparing expansion plans or formulating energy policies and strategy, including the Planning Office of the Ministry. For example, without reliable economic and financial data available on the Capanda project, it is impossible for ENE, SONEFE, or the Ministry (assuming they had the necessary manpower resources and technical skills) to evaluate the financial and economic costs of supply and to design a consistent tariff policy; and (f) the utilities incur chronic financial losses which are covered by the national budget. These losses are mostly due to poor billing and collection procedures and inadequate tariffs. So far, however, these losses do not seem to have spurred much activity to correct the basic causes of financial losses. Financial Situation and Tariffs 4.73 The financial situation of the utilities is precarious. As described in paragraphs 4.51 to 4.53, the utilities sustain financial losses which are covered by Budget subsidies or other transfers, or by payment arrears. If the Budget had no pressing revenue needs and if public resources were plentiful, this situation could possibly be sustainable. However, this is not the case. High, persistent budget deficits are causing many problems in the economy. Furthermore, with many more critical expenditures (e.g., health, nutrition) drastically curtailed, a subsidy to electricity consumption (by the better-off sectors of the population, by definition) can hardly be justified in distributional terms. 4.74 This report therefore proposes that the electric power subsector cover its costs (and a part of its investment needs--this could also be seen as a return on public capital invested in electric power) with its own revenues. 26/ Covering the present subsector deficits at present price and exchange rate levels would require a multiplication of 26/ Given the weakness of accounting procedures in the utilities, the figures presented here are only rough orders of magnitude. - 84 - tariffs by 3 or 4. At the same time, tariffs should be greatly simplified, and inefficient features, such as the "declining block" structure should be removed. However, this report also suggests a general increase in petroleum product prices. Obviously, such price increases (diesel could be increased to Kz 25 or Kz 40/liter) would have to be taken into account in setting tariffs. 27/ In addition, since almost 50% of the operating costs of the utilities are in foreign exchange, any change in the exchange rate would affect these costs and should also be reflected in a revised tariff. 4.75 Since Luanda is the largest load center, it should be dealt with first. In broad terms, EDEL's proposal for tariff reform should be accepted. It should be modified immediately t result in an average residential LV tariff of about Kz 3/kWh (up from the present Kz 0.80/kWh). Given the low absolute level of tariffs, this increase can probably be effected in a single step. A roughly similar increase should be enacted in other areas of the Northern System and also for MV and HV consumers. 4.76 In other areas, similar basic policies should be pursued. No attempt, however, should be made to set uniform tariffs in all systems, as this would constrain the design and implementation of a more efficient tariff policy in the future. The following steps are suggested as a sequence to tariff reform: (a) simplify LV tariffs (eliminating declining blocks) and raising the average price level to cover the utilities' financial deficit; (b) study and introduce changes to bring the tariff system closer to a second-best efficient system (e.g., avoid discrimination by end-use: choose appropriate billing parameters and variables such as supply voltage, contracted peak demand for HV consumers, energy demand during system peak hours, etc.); (c) set prices to meet specified financial targets (e.g., setting tariffs to cover all costs of operation plus about 20% of investment spending or, more simply, to attain a modest return of 3% to 4% on subsectoral assets in use); and (d) adjust the tariff to the level and structure of Long Run Marginal Costs. (This is a medium-term goal towards which tariff reform should aim; it should be possible to reach it by 1992, when the present priority investment rehabilitation program will be completed). 27/ The full cost of generation would be about Kz 7.5/kWh with diesel costing Kz 25/liter, and Kz 12/kWh with diesel costing Kz4O/liter. - 85 - 4.77 Setting tariffs on the basis of the utilities' financial requirements is not as simple as it appears. A comprehensive assessment of the utilities' revenue needs would require considerable efforts and time because of the scarcity of accounting/financial skills and the unsatisfactory present state of accounting, budgeting, billing, and collection procedures. Setting tariffs to cover part of investment opending could require that a realistic investment program be prepared. However, if the goal were to achieve a certain return on assets, a better definition/evaluation of assets and overall balance sheets would be required. The initial step to improve accounting information should be the adoption, for present and future operations, of a much simpler accounting system than the existing National Accounting Plan (Piano Nacional de Contas). Also, the idea of re-creating accounts for past years should be abandoned for the time being. Billing and Revenue Collection 4.78 The main problems in billing and collection are in Luanda. In other systems and load centers, billing and collection are generally satisfactory. For one thing, outside Luanda the problem is on a much smaller scale. Local managers have usually succeeded in resolving the major difficulties. 28/ Although billing and collection are generally adequate, tariffs are extremely low. Some regional managers have decided to extend the higher-priced blocks in an effort to increase the average price per kWh. 4.79 Billing and collection difficulties in Luanda are due to several factors such as the lack of effective penalties for non-payment of bills (EDEL staff are apparently unable to disconnect delinquent customers), the large number of illegal connections (possibly 10,000- 15,000 in addition to about 75,000 legal connections), and the lack of a system for the easy payment of electricity bills. In addition to these reasons, EDEL has been unable to bill ics customers expeditiously for a variety of reasons (lack of staff, lack of computers, breakdowns). Delays in billing have not helped customers maintain their payment discipline. 4.80 Broad recommendations to EDEL to improve billing and collection should be: (a) correct the internal (computer) problems which prevent timely issuance of bill+. Given the low tariffs, annual or semi- annual bills may be adequate. However, tariffs will eventually rise and then gradually more frequent billing will probably be needed. The MEP should immediately grant any assistance 28/ In Huambo after independence, power company workers resumed service and worked several months without salary. Salary payment was resumed only once successful billing and collection permitted it. - 86 - required by EDEL to recreate a credible billing/collection system; (b) obtain permission from Government authorities to disconnect delinquent customers as a routine measure in order to enforce payment of electricity bills; and (c) improve the qbility to satisfy consumer requests for service premptly (afte: the increase in tariffs and improvements in billing and collection). A campaign to legalize or rectify illegal connections should then be conducted. Heavy sanctions should thereafter be imposed on illegal connections or re-connections. 4.81 Naturally, efforts to improve billing and collection would be more rewarding if tariffs were more appropriate. Therefore, the efforts mentioned in the preceding paragraphs should proceed simultaneously with an increase in tariffs and reform of the tariff structure. Tariff reform and tariff increases, and improved billing and collection are essential ingredients in a strategy of improving the financial condition and prospects of the electric power subsector. Once these changes have been made, the utilities will be better able to supply power efficiently and reliably. Qualified Manpower 4.82 In Angola, qualified manpower is extremely scarce. A large number of qualified people left Angola at independence or during the troubled times that followed. Since then, high defense needs for qualified manpower have reduced the supply to all civilian users, whether public or private. The power subsector has suffered from the shortage. Within the subsector, distribution companies have been relatively worse off. Technical, financial/economic, and managerial skills are all extremely scarce. This situation may well explain many of the difficulties the power utilities currently face. 4.83 The organization of the power subsector (i.e., the attempt to create a national power company) and the rather ineffective deployment of what little qualified manpower there is (too many of the few qualified persons available are assigned to Luanda headquarters) compound the utilities' difficulties. Until recently, a problem of lack of incentives (inability to get access to official shops for their qualified people) prevented ENE and EDEL from recruiting qualified people. This obstacle has since been partially overcome but the problem of motivating higher level staff is a persistent one. Training 4.84 Training programs are costly and time-consuming but necessary. Therefore the most urgent training needs should be identified and acted - 87 - upon. Most probably, the priority is for on-the-job or "near-the-job" training in operation and maintenance of power pla.its and distribution lines. Administrative and accounting skills are probably almost as important. Global needs should be arrived at by asking the Regional Directions to estimate their minimum requirements (and, possibly, choose the candidates for training). In this way, training will be concentrated on the highest pciorities, and the new electricity training school will be able to make an immediate useful contribution, which would be complemented by training-oriented technical assistance. In addition, the cost effectiveness of training programs/institutions needs to be examined critically as there are currently several cases of extremely high costs per trainee. Technical Assistance 4.85 The primary objective of technical assistance should be the institutional strengthening of the utilities, especially in operations and maintenance. Because of the length and complexity of this task, an integrated and lasting assistance will be needed. This can probably best be supplied by other utilities. In all likelihood, the best vehicle for assistance would be a contract with another utility to provide management and supervision of expatriates, especially for the Central and Southern systems. Even this approach will only work if Angolan qualified manpower is deployed so that the expatriates have trainable counterparts working with them. Operational support to the Central and Southern systems would require approximately 20 persons for a period of three years. 4.86 Technical assistance will also be necessary to execute the investment program of the electric power subsector. In order to focus on technical assistance and ensure that it is used effectively, a task force in charge of aAl rehabilitation should be formed with staff contributed by SONEFE, ENE, and MEP. If the Capanda dam were delayed or postponed, some of the staff of CAMEK could also be added to this task force. This task force would be supported by a substantial infusion of external technical assistance: five persons (2 power engineers, 2 power economists, and 1 financial analyst) would be required over a period of three years. At least an equal number of qualified Angolans should work with them. This strengthened task force would prepare and supervise the execution of the various priority rehabilitation tasks. This task force would also help coordinate capital assistance to the projects under its responsibility. The technical assistance considered in the two previous paragraphs would consist of about 75 man-years of long-term consultants and about 25 man-years of short-term consultants and studies. Approximate costs would be US$10 million for the three-year period. Investment and Expansion Planning 4.87 At the time of the mission's visit, no systematic expansion planning was taking place in Angola. An approximate least-cost expansion plan for the Northern System was prepared before independence and still - 88 - survives within SONEFE, but it has not been recently updated. This may be due to lack of resources and personnel at a time when the utilities are hard-pressed to cover routine operations which are complicated by the war situation. Investment programming has also suffered, with investments generally conditioned by the availability of finance. Sometimes, when financing is available, it is limited to the foreign exchange share, with no provision for installation or civil engineering. 4.88 The main issues for the electric power subsector can be listed as follows: (a) There is an absence of a coordinated policy and systematic decision-making process for investments, and this is exacerbated by the war situation. Investment decisions are taken on a short-term basis without the help of economic or financial scrutiny and under pressures of all sorts (political, from supplier countries, etc.). (b) Financing considerations generally prevail over any integrated approach or strategy. Consistent sequences of activities, when they are identified, are almost always ignored as financing is negotiated on a case-by-case basis. If a package of equipment and civil works is financed (often with pressure-selling on the part of the suppliers), it is then called a "project" even if the financial provision for it does not extend to completion and commissioning. (c) Contracts are awarded haphazardly, without making sure that the previous and subsequent stages of work have been identified, planned, and financed. As a result of this, millions of dollars of equipment has been purchased but cannot be installed, and is deteriorating. While these assets are not contributing anything, the debt, often incurred on commercial terms, is being serviced. (d) The utilities have no control over commitments and drawing of foreign exchange. The BNA has this responsibility, and coordination between the BNA and the utilities has proven to be difficult. (e) The utilities have a tendency to add new capacity rather than relying on preventive and corrective maintenance in order to keep capacity available and operational. This is partly the result of the foreign exchange shortage, which prevents the timely purchase of spare parts and materials while exter al supDliers and credits are often more easily available for purchase of large quantitites of new equipment. (f) As a result of lower levels and reliability of service, high numbers of small diesel sets and gas turbines have been purchased and are spread all over the country. They represent - 89 - a large foreign exchange debt and occupy an inordinate share of the utilities' qualified manpower, which should be concentrated on the main installations. Servicing these scattered, small systems also requires a costly logistical system (fuel, transportation, storage), and places a heavy burden on the limited resources of the utilities. 4.89 In light of these difficulties, the next few paragraphs will suggest some priorities for investment and expansion planning. Better, more effective planning and investment decisions have as goals the satisfaction of Angola's projected electric power needs at lowest cost and the strengthening of the utilities to make them better able to meet these future needs. Minimizing investment expenditures will often be required as a means to reduce costs, and higher tariffs will be needed to provide a safe financial basis for the strengthening of the utilities. Investment Priorities 4.90 In the short term, investment priorities must, of necessity, center on repair, rehabilitation, and resumed maintenance of existing facilities. Rehabilitation should proceed simultaneously on all three systems as both security and economics relegate an interconnection to a fairly distant future. The medium-term goal should be to fully restore supply capabilities in line with installed capacities, and to create adiquate reserve margins for potential demand growth. Uncertainty in the evolution of demand and the high cost of supply options coupled with resource constraints (finance and qualified staff) militate in favor of an investmnent program that addresses the most pressing needs while meeting the most stringent economic criteria. 4.91 Existing investment programs for the electric power subsector as a whole, excluding Capanda, total about US$100 million for 1987 and 1988p 75% of it in foreign exchange. A program of this size is probably beyond the financial and technical capabilities of the utilities. A scaling down of investments is thus inevitable; this must be done, keeping in mind the key priorities. The utilities' detailed investment programs are reviewed and evaluated in Annex 16. 4.92 A tentative priority investment program is described in Table 4.6 below, based on the following considerations: (a) assign highest priority to rehabilitation of existing facilities; (b) strive for improved reliability of supply to main cities which are also the main industrial areas; (c) improve supply to Luanda by addressing the main problems in generation, transmission, transformation, and distribution; - 90 - (d) postpone most small projects in isolated systems, mainly for 'ack of managerial/technical staff, even if equipment has been purchased (only those that can actually be built should be considered); (e) postpone new rural/village electrification until hydro supply conditions have been improved and tariffs readjusted; (f) limit new connections in cities until tariffs are adjusted and (especially in Luanda) until billing and collection procedures are substantially improved; and (g) plan a substantial amount of technical assistance to support ENE task forces in big rehabilitation projects such as Lomaum, the Southern System, and the Luanda distribution grid. 4.93 A general recommendation, in addition to the above considerations, is to subject every substantial project (say, exceeding Kz 50 million) to economic and financial feasibility analysis.29/ This would alert the utilities' management to the need to redesign, scale down or reschedule the project. Accurate cost data would also facilitate the search for financing for the complete projects, rather than having many projects only partly financed as at present. A project evaluation unit should be set up in the central staff of ENE (or CNE) for this purpose. This unit should also advise the Government on the cost of "political priorities" so that the quality of "political" projects could be improved. Finally, the unit would make sure that its programs for the power subsector are consistent with programs in other sectors (industry, mining, agriculture, etc.). 4.94 A minimal priority investment program in line with the above priorities and considerations was prepared by the m.ssion. 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 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. The financial burden could be reduced somewhat if efforts were made to obtain financial assistance f-om the numerous bilateral or multilateral aid agencies. All these agencies require economic and financial analyses (and are willing to fund them) but lend at considerably softer terms than export financing agencies. The suggested five-year priority investment program is presented in Table 4.6. 29/ This would make the project attractive to concessional donors and result in lower financial costs. - 91 - Table 4.6: ELECTRIC POWER SUBSECTOR INDICATIVE SUGGESTED PRIORITY INVESTMENT PROGRAM 1987/88-1992 Project --- Million --- Timing USS Kz Generation Cambambe - Repair units 2, 3, 4 5 150 ongoing MaLubas - Rehabilitation 7 210 1989-ongoing Lomaum - General rehabilitation 60 1,8C0 1989-92-ongoing B16pio - Rehabilitation of 2 units & dam 15 450 1989-91 Southern System: Matala - Repair dams & generators 25 750 1988(89)-92 Gove - Repair dam 5 150 1989-92 Namibe - Complete thermal plant 1 30 1989 Huambo - Repair/replace gas turbine 2 60 1989-ongoing Cabinda - Repair gas turbine 1 30 1989-ongoing Ulge - Erect diesel unit 1 30 1989 Transmission Bi6pio Substation - Rehabilitation 2 60 1989-90 Viana Substation - Expand/complete 4 120 1987-89 Quinfangondo - Complete substation 3 90 Luanda Substation - Replace transformers 2 60 1988-90 Renovation of various lines 10 300 1989-94 Distribution Luanda - General rehabilitation 30 900 1989-94 Small urban networks 5 150 1988-92 Renovation of other systems (Lobito, Benguelu, etc.) 15 450 ;989-94 Other 22 660 1989-94 TOTALS 215 6,450 19fl9-94 Average per year 43 1,290 Source: ENE, EDEL, SONEFE, a,d Mission estimates. 4.95 The investment program outlined in Table 4.6 may seem modest. This is inevitably so when compared to the huge expenditures programmed for Capanda, where annual investments are expected to exceed the five- year total for all utilities. However, this priority investment program is by no means small: it provides for che rehabilitation of all - 92 - installed capacity and some additions, notably at Lomaum and Matala. It would be sufficient to meet demand reliably until the mid- to late- 1990s. The investment program is by no means easy to carry out. The mission estimates that as much as US$10 million in external technical assistance may be needed to carry it out. But this is a small amount compared to the about US$50 million of technical assistance expected to be provided to the Capanda project by FURNAS alone). This small, priority investment program will certainly do morc to improve service and reliability and create a useful reserve margin in all systems than the massive investment at Capanda, which is described in the next section. The smaller program is a better risk-averse response to uncertainty than a big investment in generation in one part of the country because it will equip the utilities to be able to provide power aiywhere it may be needed. In any case, before Capanda can help improving supplies to any grid, substantial additional investments will be needed in lines, substations, transformers, distribution grids, user installations, etc. Since the Capanda project can be seen as an issue in investment and system planning, it is discussed more at length in the next section, and in detail in Annex 16. The Capanda Hydroelectric Project Summary and Recommendations 4.96 The Government's apparent decision (which can still be reversed) to advance the construction of a dam and power plant at Capanda presents several major issues. Although the analysis done is extremely rough and conclusions should be taken as tentative, several robust conclusions emerge and are stated as follows: (a) Capanda represents a significant departure from the lowest cost expansion path, even if it is not well-known and has not been updated recently; (b) the huge capacity (4 x 130 MW) planned for Capanda will probably not be needed until well into the next century. Until the year 2000, maximum requirements to fully meet the projected energy and capacity needs of the Northern System, in the worst case ("High" scenario), with hydrogeneration alone, would only be 170 MW of additional capacity and 520 GWh of additional energy. A "High" dam in Cambambe in the early 1990s, 2 x 100 MW in the second Cambambe power plant in the mid-1990s, and a "Low" dam in Capanida in the year 2000 would be enough to meet those requirements with a high reserve margin at the time of system peak and no need for thermal support. The required investments would be significantly lower and much farther into the future than the Capanda option now being considered; (c) it is a project which, by itself, will not improve substantially the reliability of service in the Northern System - 93 - and will not mitigate the problems of the other two systems at all; (d) actual and expected low demand growth rates and the availability of substantial thermal reserve (capacity and energy) 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 market is possible;30/ (e) alternative expansion sequences in the Northern System, with different timings of 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. Such studies, if immediately initiated, could be carried out within five to six months. The delay would not create any additional risk of energy shortages in the early 1990s and could bring the benefit of redirecting more efficiently the large investments already made in support infrastructures; and (f) 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. While, at times, a state may need to take all manner of risks to finance a vital investment, the case of Capanda definitively does not fall in this category. On the contrary, going ahead with Capanda will prevent Angola from undertaking activities which are truly of vital importance. 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 and the speed at which Capanda power should be developed. Background to the Capanda Project 4.97 In 1982, Angola and the U.S.S.R. signed an agreement to develop a major hydroelectric dam and plant at Capanda on the middle course of the Kwanza River. A new organization, CAMEK, attached directly to the 30/ Satisfying repressed demand in Luanda would require expensive, large, and time-consuming investments in the medium and low tension grids. - 94 - MEP, was created to execute this task. Soon thereafter, a "master contract" was signed between the MEP and a Soviet/Brazilian engineering firm, N. ODEBRECHT. By these decisions, the hydro plant at Capanda (4 x 130 MW) was given priority over the further development of Cambambe which was then part of SONEFE's expansion plan. 4.98 TECHNOPROMEXPORT (TPE) is the team leader for the project and supervisor for the construction. Other Soviet firms or institutes are responsible for geological prospecting, dam design, and equipment supply. The Brazilian engineering/construction firm, N. ODEBRECHT, or its subcontractors, are in charge of the civil works. Available information on contractual details is scarce, and seem far from comprehensive. A mechanism has been created to regulate subsequent activities by requiring additional contracts each time an important activity or set of activities is identified and programmed. So far ten "Partial Supplements" have been signed since 1985 (one with the Soviet party for geological prospecting and dam design; nine with the Brazilian party). 4.99 GAMEK's staff is made up of senior people seconded (often part- time) from other enterprises (such as SONEFE and SONANGOL) as well as 55 middle and higher level staff seconded by contract from the Brazilian utility, FURNAS CENTRAIS ELECTRICAS. GAMEK'S complete staff currently totals about 165 persons (55 Brazilians and 110 Angolans). Salaries are similar to those of other utilities in Angola, but fringe benefits (such as access to official stores) are much better. GAMEK is thus success- fully competing with other utilities for scarce qualified manpower. Capanda and the Least Cost Expansion Plan 4,100 Developing the plant at Capanda represents a serious departure from a least-cost expansion plan. The steps in this plan were as follows: (a) increase the height of the Cambambe dam, thus increasing capacity from 180 MW to 260 MW (which might be needed in the mid- to late-1990s); (b) build a low retention dam at Capanda giving Cambambe monthly regulating capability, and build a second power plant at Cambambe (2 x 110 MW), pushing total capacity at Cambambe to 400 MW (which would only be needed around the year 2000); (c) increase the dam height at Capanda and build a second stage of the second power plant at Cambambe (2 x 100 MW) (would be needed beyond reasonable present planning horizon); and - 95 - (d) build a power plant at Capanda (4 x 110 MW assumed in old studies, 4 x 130 MW in more recent ones; only required beyond the present planning horizon). 4.101 The investment costs of the above sequence were compared with those for advancing Capanda (as estimated by BEP) and gave the following results: Cambambe, US$1,030/kW; Capanda, US$1,800/kW. The comparison, however, should not be pushed too far as firm capacities are not ac- counted for and Capanda would look even less economic if the two dams were really considered alternatives rather than complements. Contrarily, however, these figures embody costs of turbine plus generators estimated by BEP at US$500/kW for Cambambe and only US$345/kW for Capanda, without any plausible explanation of the difference. Long Run Marginal Costs 40102 The Long Run Marginal Costs (LRMC) of producing electricity from Capanda are a multiple (8 or 10 times) of present average energy costs in the Northern System served by Cambambe. Rough estimates of the investment and operating costs of Capanda and a set of assumptions on the share of demand (as estimated by the UNDP/IBRD mission) supplied by Capanda provide a rough indication of the LRMC of electricity from Capanda.31/ A fair estimate of LRMC is roughly US$0.334/kWh (or Kz 10.037ikWh) with the base UNDP/IBRD scenario of demand, assuming that generation at Cambambe is limited at 780 CWh/y (which is about its firm generation capacity) 32/, and a discount rate of 12%. A sensitivity analysis (three demand scenarios, three discount rates--10%, 12%, and 15%--and three levels of generation by Cambambe) was performed. This analysis shows that LRMC could go from a minimum of US$0.127/kWh (Kz 3.80/kWh) (Cambambe generates only 200 CWh, 10% discount rate, high demand scenario) to a maximum of US$0.547/kWh (Kz 16.41/kWh) or even a bit more if Cambambe generated 1,000 GWh (which it can generate in an average year) rather than its firm energy of 780 GWh. See Annex 16, paras. 15-16 for a detailed comparison/discription of these various cases. It is clear that even under favorable assumptions, covering the costs of Capanda would require a large increase in tariffs (excluding any exchange rate shadow pricing). If demand were to grow more slowly than projected, Capanda power could not be sold and costs could rise enormously. At best, Capanda will be a project of marginal economic interest. At worst, it will be a heavy financial burden making no contribution in the medium term to the development of Angola. 31/ Based on updated information supplied by GAMEK/MEP during the October 1988 mission. 32/ 786 GWh is actually the joint firm generation of Cambambe and Mabutas, assuming 700 GWh for Cambambe and 80 GWh for Mabutas. - 96 - Technical and Financial Information 4.103 No preinvestment studies, either technical or economic, were done to aid in the decision to develop Capanda. Some old studies existed but it was decided to proceed with const-..ction pari passu with the needed technical studies. No economic or financial studies or demand projections were made specifically for the justification of the Capanda dam. 33/ Given this fact, all statistics connected with Capanda must be considered as rough approximations. Total investment cost was not precisely known but was stated to be about US$1 billion in 1986 prices. This was revised upwards in 1987 (in 1987 prices). Total cost was then estimated at $1250 million for direct costs, that is, about $1600 million including financial costs (and interest during construction). In fact, only in 1987 were technical studies advanced enough to allow the estimation of a cost (the 1985 figures were later described as "guesses"). However, even the 1987 figures probably understate costs as they do not allow for any contingencies. A rough description of costs is shown in Table 4.7. 4.104 The financing package described in Table 4.8 is close to commercial terms (interest at 7X-8%, 2 years' grace, 7-8 year repayment period). 33/ The BEP and THEMAG studies had other specific goals, i.e., expansion plan and interconnection grids. - 97 - Table 4.7: CAPANDA HYDRO PROJECT - GAMEK INVESTMENT PROGRAM (in USS million) Firm/Component Actual 1988 1989 1990 1991 1992-4 TOTAL 1985-87 Est. Act. N. ODEBRECHT TOTAL 327.9 109.1 117.1 119.6 120.3 77.2 871.3 of which: Labor 73.7 40.1 48.0 58.0 59.6 37.3 317.4 Materials 89.1 20.5 23.9 26.8 26.8 14.1 201.2 Food 12.8 8.2 11.4 11.5 10.7 9.2 63.8 TPE TOTAL 4.2 8.4 14.3 23.1 79.8 145.4 275.0 of which: Equipment 0 0 0.2 4.8 65.6 99.2 169.8 FURNAS TOTAL 23.4 10.2 7.0 7.0 6.5 11.0 65.0 (Tech. Assistance) GAMEK 9.5 4.0 4.0 3.5 3.5 10.1 34.5 TOTAL DIRECT COSTS 365.0 131.6 142.4 153.1 210.1 243.6 1,245.8 Interest + Commission 17.1 56.1 34.4 42.9 41.7 164.5 1/ 357.1 + Exchange Losses GRAND TOTAL 382.1 187.7 176.7 196.0 251.7 408.4 1,603.0 1/ Financial charges run up to 2003. GAMEK: Gabinete de Aproveltamento do Medio Kwanza (Office for the Harnessing of the Middle Kwanza). Note: Totals may not add up because of rounding. Source: Angolan authorities and GAMEK. - 98 - Table 4.8: CAPANDA: FINANCING PLAN (in USS million) COUNTRY/INSTITUTION COMMITTED SOUGHT TOTAL TERMS CACEX/ Banco do Brazil 4/ 408 120 21 528 7%-8% 2 years grace. 7-8 years repayment period. Cash Payments 162 162 408 282 690 U.S.S.R. (for equipment) 275 -- 275 3% 10-year life. Includes 3 years grace period. Other countries 2! 8 97 105 S8 million for (cash payments until funded) 3 years at commercial rates. SUBTOTAL 765 475 1,240 SUBTOTAL (Excl. Cash & Budget)3/ 691 120 811 TOTAL COST OF PROJECT (Incl. Financial Charges) 1,600 FINANCING GAP 780 i/ These amounts are carried as cash payments until such time as they are financed, 2/ This amount was committed by Brazil in 1989. 3/ The Angolan Government (or its Budget) is the financier of last resort, i.e., it would have to make up any shortfall. This shows that the shortfall, which Angola may have to make up, is really S789 million. 4/ The Government of Angola has given a general guarantee in oil to Banco do Brazil for these loans. 5/ This Is theoretically in local currency, but local expenditures have an extremely high foreign exchange interest in Angola. Source: The MEP, GAMEKO 4.105 The total amount being sought, $1240 million, falls short of even direct costs ($1245 million) let alone adding the financial costs. In October 1988, only $691 million was firmly committed (excluding GOA and OGE) which means that there was a financing gap of $909 million (inlcuding financial costs) or $554 million (excluding financial costs). Since then, an additional $120 million has been secured from Banco do Brazil. The financing gaps are therefore US$789 million and US$434 million (respectively including and excluding financiel - 99 - charges). In addition, $17 million of debt service was paid between 1985 and 1987, while in 1988 debt service was expected to reach $56 million. Except for Soviet financing, by October 1988 funds had practically run out. Were it not for an additional $120 million supplied by Brazil, work would have had to stop. All this raises serious questions about the feasibility of building Capanda as planned and reinforces the need to seriously reexamine this investment within the context of an overall system expansion plan, i.e., most probably to abandon the idea of building Capanda in one stage and to weave it into the overall development of generation along the Kwanza River. Capanda and the Need for Interconnections 4.106 The use of Capanda to supply the three interconnected systems also deserves some comment. Under the present conditions of unrest, the construction of new transmission lines to link the three systems is impossible. The existing Cambambe-Cabela line, the first stage of a future North-Center link, has been out of service since 1984 and security conditions have not allowed a restoration of service, and the Government has therefore installed diesel sets in Porto Amboim. Similarly, Line 1 between Cambambe and Luanda cannot be properly maintained and is in a precarious situation. Major Central and Southern transmission lines cannot receive maintenance for lack of security. In the meantime, regional power supplies will be rehabilitated and expanded to meet potential demand (Lomaum, Bi6pio, Huambo, Matala, Namibe). The three systems will necessarily remain self-sufficient and will have to provide for their own (separate) reserve margins. 4.107 Prolonged conditions of insecurity will keep demand growth low and will prevent construction of interconnections. Thus a larger market for Capanda power will not be able to develop. A short-term cease-fire followed by a quick economic recovery, approaching the demand growth rates projected by BEP, might allow and possibly call for an interconnection but a least-cost expansion plan would apparently postpone Capanda itself until after the year 2000, giving earlier preference to new supply sources in the Central and Southern Systems. 4.108 In an interconnected environment, Capanda would have difficulty meeting reliability criteria (too much generation in a single pole, hundreds of kilometers away from the main consumption centers), operating problems (the interconnected systems would be of the "weak longitudinal" type which is known to present difficult voltage control and dynamic oscillation problems), requiring the opening of the interconnecting links or putting serious restrictions on power transfers 34/, and on-line dispatching. No experience exists in Angola of real-time dispatching of 34/ Low load situations require special compensating devices (var static compensators) in different points of the system and may also produce severe cavitation in the turbines of Capanda. - 100 - distant interconnected plants; this requires reliable communications and control systems together with highly-trained personnel. Final Remarks 4.109 Although information is patchy, the decision to advance Capanda rests on highly optimistic expectations regarding economic development and electricity consumption, insufficient appreciation of sound economic considerations, an underestimation of the financial burden, and underestimation of the difficulties of operating and maintaining such a huge installation. The Government's wish to take advantage of its hydro resources is understandable. However, this cannot be done without large investments in transmission and distribution in addition to the generation investments. Two theoretical alternatives might eventually lead to a justification: (a) attracting foreign investments in power-intensive industries that do not require significant investments in distribution networks; (the negative experience of Zaire with its Inga power plant would be an interesting comparison here); and (b) providing generalized access to electricity to Luanda's population and providing incentives for industrial rehabilitation. While the first alternative remains theoretically open but requires a relatively long period of stability before materializing, the second would require enormous investments in distribution networks (from high to low voltage) that the utilities and the Government simply cannot afford. The issue of appropriate tariffs will not even be mentioned here but higher tariffs would have a significant impact on both alternatives. 4.110 The whole issue of Capanda remains basically that a market of a size that could justify the investment does not exist nor can be expected to materialize within a reasonable period. By the end of 1988, almost US$500 million had been spent and heavy infrastructures built in Luanda and on-site. There are large quantities of equipment and the project is at the crucial stage just before the river is to be deviated. On the other hand, financing for the completion of the project is, most likely, unavailable. Finally, there now exists an institution, GAMEK, whose good capabilities should be utilized. For all these reasons, it would seem justified to stop building and carry out a study of the feasibility of stopping the project, of looking for the best alternatives to develop the Northern System and, finally, of the best way of introducing Capanda into this system (when to resume construction, optimizing the power of the units and their coming on-line, etc.) within a global expansion plan which would develop the hydro resources to best advantage. Brief Terms of Reference for such a study (which would also include preparation of a rehabilitation program for the existing power systems) are presented in Annex B. - 101 - V. FORESTRY, WOODFUELS AND HOUSEHOLD ENERGY Summary, Conclusions, and Recommendations 5.1 While available statistics are few and unreliable, two recent studies and mission estimates have produced a picture of the supply and demand for forestry and woodfuels in Angola which can be summarized as follows: (a) most Angolans use firewood or charcoal for cooking and heating (in the cities, however, a significant minority use LPG); (b) the aggregate consumption of firewood is in the order of 2.5 million tons/y and of charcoal about 0.5 million t/y, requiIing a total removal of 6 million tons or about 10 million m of wood; (c) 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 presently consumed in the country; (d) 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; (e) in the shortage areas, the periurban population constitutes the group hardest hit. 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; (f) 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; and (g) 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. - 102 - 5.2 At present relative prices and depending on the security situation, the priorities for action in forestry/household energy ate as shown in Table 5.1. Table 5.1: WOODFUELS - PRIORITIES FOR ACTION Increased Improved Fuel Efficiency Supply Wood Area Action Substitution of Stoves System Production Peace II IlIl I IV Coast No Peace I II III IV Peace IV III 11 I Inland No Peace IV I II III Source: Mission. 5.3 If the reduction in civil strife is accompanied by structural changes in the economy such as a major devaluation carried through to commercial energy prices, then relative prices will probably change drastically. The prices of petroleum products would rise markedly while the reduction of the risk premia and increased competition (because of easier market entry) and lower rents to truck owners and other entrepreneurs would most probably lead to (comparatively) lower prices for traditional fuels. In that case, consumption of LPG might not grow as fast as in the past (although this might not be noticed for some time because of the existence of repressed demand), and traditional fuels might once again become competitive with LPG (or kerosene). Measures to improve the supply of woodfuels to the cities might then become the highest priority, as indicated in Table 5.1. 5.4 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 of Huila and Namibe, a region where the security situation is fairly good. It includes both city-oriented activities such as the improvement of stoves, and rural-based activities such as 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. 5.5 Two national activities should be carried out in support of the regional ones. One covers the development and introduction of improved - 103 - stoves and the other concerns initial development work and trials in agroforestry. The activities listed 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; (d) 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 until economic adjustment measures increase the prices of petroleum products (LPC, kerosene); Consumption and Production of Woodfuels Consumption of Woodfuels 5.6 The consumption of woodfuels in a country is never easy to measure or estimate accurately. It takes place among hundreds of thousands of households, generally in an uncoordinated fashion, largely outside the monetized economy. Until quite recently, knowledge regarding the consumption of woodfuels in Angola was extremely poor. In the absence of any consumption survey, estimates were little more than guesses. Further, available figures on wood removals reflected only officially recorded harvests of wood for fuel, corresponding to a minor share of the total. 5.7 Material presented at the First National Seminar on Woodfuel and Charcoal, held in Luanda, June 8-10, 1987, improved the situation considerably. As a part of the preparation for the seminar, the Department of New and Renewable Sources of Energy (DNRFE) within the Ministry of Energy and Petroleum (the MEP) made a survey of household fuels in seven provinces. The survey covered 4,466 persons, 1,557 of whom resided in urban areas, 2,063 in periurban areas, and 846 in rural areas. In addition, discussions were held with producers and traders of fuelwood and charcoal and traders of LPG and kerosene in the same seven provinces. As one result of the survey, average consumption figures per person and month were obtained for four main fuels (firewood, charcoal, LPG, and kerosene) in each of the three population strata (urban, periurban, and rural). The findings are presented in Table 5.2. - 104 - Table 5.2: CONSUMPTION OF HOUSEHOLD FUELS, 1987 a/ (In kg/capita/month) Firewood Charcoal LPG Kerosene Urban 23 10 2.5 1.5 Perlurban 30 13 2.2 2.2 Rural 39 13 0 2.0 Average 32 12 2.4 1.9 a/ Figures are based on data from seven provinces. Source: DNRFE. 5.8 According to this data, the urban population consumes either 23 kg of firewood or 10 kg of charcoal or 2.5 kg of LPG or 1.5 kg of kerosene per person per month. This is probably largely correct for firewood, charcoal, and LPG if the families use either of these three fuels as a source of 3nergy for cooking. However, the figure seems unrealistic for kerosene, which is used mainly for lighting. Moreover, since the data were recorded between late 1986 and early 1987, the energy needs for heating which peak between May and August may not be included. Therefore, average annual household energy consumption is likely to be higher than Table 5.2 suggests. 5.9 In the absence of any population survey since 1970, all figures on the present population of Angola are rough estimates. While they might be fairly correct as regards the national totals, they are probably quite unreliable as to the distribution of the population between different provinces or, indeed, between urban and rural areas. Most recent reports estimate that between one-quarter and one-third of the population lives in urban areas, while observations during the survey pointed to a quite different picture, with half the population living in urban and periurban areas. For the sake of the present report, it is assumed that the population is distributed as follows: Absolute Share (millions) (%) Urban population: 2.5 28 Periurban population: 1.5 17 Rural population: 5.0 55 Total 9.0 120 5.10 Based on the results of the DNRFE survey, with the specific fuel consumption figures adjusted upwards to acount for heating and taking into consideration the estimated distribution of the overall population, the tentative picture of aggregate fuelwood and charcoal consumption may look as shown in Table 5.3. - 105 - Table 5.3: AGGREGATE USE OF FIREWOOD AND CHARCOAL IN 1987 Amount Wood Needed 1,000 t/y 1,000 t/yr Domestic Consumption as Surveyed by DNRFE - Firewood 1,920 1,920 - Charcoal 440 3,080 Domestic Heating - Firewood 360 360 - Charcoal 84 590 Industry 180 180 Total 6,130 Source: DNRFEe The figures suggest that aggregate woodfuel consumption amounts to a fuelwood equivflent of ten million solid m3/y, which corresponds to 670 kg (about 1.1 m ) of wood per person per year. 5.11 A consistency check makes the above estimate appear reasonable. Based on a set of general assumptions, 35/ the country's overall woodfuel rcquirements are likely to amount to about 7.5 million tons, as detailed 35/ First, it is assumed that the daily minimum per capita requirements of cooking are equivalent to 160 kcal delivered to the pot. Second, the gross calorifics are assumed to be: 5,000 kcal/kg at 5Z mcwb for charcoal, and 3,000 kcal/kg at 25% mcwb for wood. Third, stove efficiencies are 15% for charcoal and 10% for wood stoves. Fourth, it is assumed that 50% more energy is delivered to the pot than strictly necessary (rice boiling rather than simmering, for example). Fifth, an allowance of 50% on top of the total above is made for preparation of beverages and for water heating for other purposes. Sixth, an allowance is made for non-cooking use of firewood, for example as a source of light or as a focal point for social gathering, corresponding to 25% of the energy used for cooking and water heating. Seventh, household use of firewood or charcoal fcr space heating is assumed. It is assumed that four million persons need such heating for four months per year, three- fourths using firewood and one-fourth using charcoal. Each person is supposed to nced 1 kg of wood or 0.7 kg of charcoal per day. Eighth, an allowance is made for non-household uses of firewood in restaurants, small-scale industry, lime factories, schools, hospitals, and so on, amounting to 10% of the energy value of all household use of firewood and charcoal in urban areas and 5% in rural areas. - 106 - in Table 5.4. 36/ These figures are roughly in line with findings of the DNRFE survey. While the survey recorded a total annual consumption of 2.46 million tons of firewood plus 0.52 million tons of charcoal (with a roundwood equivalent of 3.64 million tons), Table 5.4 suggests that woodfuel consumption amounts to 2.63 million tons of firewood and 0.69 million tons of charcoal (with a roundwood equivalent of 4.83 million tons). Table 5.4: HYPOTHETICAL ANNUAL WOOD FUEL DEMAND Annual Wood Population Fuel Needed Equivalents (millions) (million tons) (million tons) Cooking - Firewood 4.3 1.55 1.55 - Charcoal 3.1 0,61 4.27 Light and Social 7.4 a/ 0.54 0,54 Heating - Firewood 3.0 0.36 0.36 - Charcoal 1.0 0.084 0.59 Sum for Households 7.4 - 6.38 Industrial Use (Firewood) - 0.18 0.18 Total - 7.49 a/ Firewood with energy value corresponding to a quarter of the total fuel used for cooking (1 kg of fIrewood per day by 365 days by 4.3 million persons plus 0.36 kg of charcoal by 365 days by 3.1 million persons). Source: Mission estimates and calculations. Wood Production 5.12 No national forest inventory has ever been carried out in Angola. Depending on how the term "forest" is interpreted, the country may possess between 50 and 75 million hectares of forests, or 40% to 60% of its land surface. Recently, however, two new estimates of forest area and wood yields in Angola have been made. One was carried out by the 36/ Not all of this amount is burned in the form of wood. At least some of the agricultural residues, such as corn stalks and cobs, end up in the domestic fire. In overall terms, however, the amount is not likely to be large, as wood and charcoal are uniformly claimed to be the dominant household fuels. - 107 - consulting company E.T.C. as part of a survey of fuelwood resources in all Southern African Development Coordination Conference (SADCC) countries, while the other was part of the preparatory work for the First National Seminar on Firewood and Charcoal, held in June 1987 in Luanda. 5.13 Remote Sensing Estimate. The E.T.C. study is based on satellite images of all SADCC countries taken between 1984 and 1986. Ground truthing was undertaken in most of the 26 vegetation classes distinguished. However, none of these checks were made in Angola. The results of the study for Angola are summarized in Table 5.5. Table 5.5: MAIN FOREST FORMATIONS Standing Mean Annual Total Woody Increment Annual Area Biomass (tons per Increment Vegetation (million per hectare hectare (million Class hectares) (tons) per year) tons) 1. Transitional Rain Rorest/ Miombo Woodlands 16.0 200 6 96 2. Dense High and Medium-High Miombo Woodlands 33,2 71 2.2 73 3. Seasonal Miombo Woodlands and Wooded Sav nnas 30.7 20 0.5 15 4. Dry Deciduous Savannas 23.0 17 0.5 12 5. Other Open Forest & Bush Formations 15.8 13 0.4 7.2 TOTAL 118,7 - - 203 AVERAGE - 57 1.7 Source: E.T.C., but standing mass of Vegetation Class I has been reduced from 440 m /ha to the more plausible 200 m3/ha by the mission, and M.A.I. reduced from 17 T/Ha to 6 T/Ha (or from 3.8% to 3% of standing mass). Production Estimate by DNRFE 5.14 During the first half of 1987, DNRFE also made an estimate of production potential for fuelwood throughout the country. DNRFE based its work on an ecological classification scheme elaborated by GRANDVAUX - 108 - BARBOSA in 1970, where the natural vegetation of the country is broken down into 32 groups. DNRFE obtained updated figures for 21 vegetation groups relevant for the production of fuelwood in the country. Table 5.6 provides an overview of the main results. 37/ Table 5.6: SELECTED VEGETATION GROUPS AND THEIR FUELWOOD PRODUCTION POTENTIAL Natural Total Vegetation Fuelwood Wood Type Number Area Production "Human Production According Vegetation (million Potential Influence Potential to Barbosa Group hectares) (m /ha/year) Factor" (million m3) Tons a/ 2-4 1. Humid Forests 2.6 2.0 0.58 3.0 1.9 7-10 2. Mixed Moist Forests 13.4 0.46 0.70 4.3 2.8 11-14 3. Mixed Dry Forests 10.1 0.31 0.75 2.3 1.5 16-19 4. Miombo Formations 44.0 0.45 0.70 13,9 9.0 20-22 5. Bushland 10.2 0.19 0.78 1.5 0.9 23-24 6. Wooded Grasslands 14.0 0.20 0.81 2.3 1.5 25-26 7. Mixed Savanna/ Open Grasslands 16.5 0.10 0.86 1.4 0.9 27 8. Steppe 3,5 0.05 0.75 0.1 - Total 114.3 - - 28.8 18.4 Average 0.34 0.75 a/ Conversion factor 0.65. Source: DNRFE. 37/ DNRFE proceeded as follows: (a) determined the area of each vegetation group in each province; (b) deternined the share of each vegetation group in each province where the natural produc- tion potential has been destroyed through human influence; and (c) estimated the fuelwood production capacity of the remainder of the various vegetation groups in the various provinces. - 109 - Production Versus Consumption 5.15 Both E.T.C. and DNRFE compiled production/consumption balances for woodfuels in Angola. As some of the data--particularly on the production potential of the Angolan forests--are widely diverging, it appears that the conclusions of the two studies are quite different. Table 5.7 compares the different findings by region. Table 5.7: PROVINCIAL FUELWOOD PRODUCTiON POTENTIAL Estimated Fuelwood Fuelwood production Production Estimated 1980 Potential pe capTta/year Population ('OOOs tons) (tons) Province ('OOs) ETC DNRFE ETC DNRFE Luanda & Bengo 1,200 650 380 0.49 0.32 Benguela 590 1,400 450 2.9 0.76 Wi6 780 14,600 1,000 19 1.3 Cabinda 100 560 390 6.0 3.9 Cuene 130 8,800 590 68 4.5 Huambo ',100 5,500 530 5.0 0.48 Huila 740 8,100 1,290 11 1.7 Kuando-Kubango 140 21,000 1,890 150 14 Kwanza Norte 390 1,200 670 3.2 1.7 Kwanza Sul 590 5,300 830 9.2 1.4 Lunda Norte 150 54,700 2,190 364 15 Lunda Sul 220 7,600 1,510 35 6.9 Malanje 690 23,200 1,440 34 2.1 Moxico 230 20,000 2,880 87 12 Namibe 60 1,200 180 21 3.0 Uige 490 27,300 1,680 56 3.4 Zaire 50 1,600 480 32 9.6 Total 7,650 202,700 18,380 Averdge 36 2.4 Source: E.T.C., DNRFE, and Mission computations. 5.16 According to DNRFE, three provinces (Luanda/Bengo, Benguela, and Huambo) show production capacities below one ton of fuelwood per capita, while five (Bie, Huila, Kwanza Norte, Kwanza Sul, and Malanje) have potentials between 1 and 3 tons per capita. The first group of provinces, where around 3 million people live, is clearly unable to supply its own population with firewood from within its own borders. The other group of provinces, where a further 3 million people live, are barely able to do so. In fact, the estimated deficit from Luanda/Bengo, - 110 - amounting to some 200,000 tons, probably further aggravates the situation in nearby areas, particularly in Kwanza Norte. On the other hand, the DNRFE figures suggest that only about one-third of the population of Angola lives in provinces with ample supplies of fuelwood. Those provinces have a combined surplus of production over consumption of some 9 million tons. 5.17 Compared to the DNRFE estimates, the remote sensing data of the E.T.C. study indicate a significantly higher potential for woodfuel production, both at the macro and micro level. On average, the E.T.C. figures exceed the DNRFE estimates by a factor of 10. This may grossly underestimate the scarcity of woodfuels, but it does not contradict the view that the urban concentrations in the coastal area, compared with other regions, are poorly endowed with woodfuel resoures. 5.18 The wide gap between the two estimates may in part be attributable to the fact that they measure different things. The E.T.C. study draws a picture of Angola's overall potential for woodfuel production, regardless of whether the resource base grows close to where people live or whether the wood is obtained in small sizes, useful for rural people with simple tools, or as large tree trunks, practically useless as a raw material for domestic fuel. Thus, while it is probably correct to conclude that the aggregate sustained yield in Angola is in excess of 100 million tons per year, the annual yield accessible to and directly useful to the population of the country is certainly much lower, maybe in the vicinity of 20 million tons or somewhat higher. 5.19 Neither study, however, succeeds in convincingly describing the situation of the people concerned; nor do they provide a statistically accurate overall picture of the country's biomass reserves. What is undisputed is that from a global point of view Angola is not running short of biomass stocks. Both studies also indicete where the problem areas in domestic woodfuel supplies lie. The situation is crucial for two kinds of population concentrations: all urban concentrations on the seaboard, particularly Luanda and Benguela/Lobito, and the major urban concentrations inland, particularly Huambo. It is obvious that any attempt to solve the woodfuel problem in any of these areas will have to involve areas far beyond the urban concentrations, often even reaching beyond the home province. Institutional Issues: Administration of the Forestry Sector 5.20 The institutions dealing with forestry and woodfuels in Angola are very weak, particularly at the field level. The DNACO (National Directorate for the Conservation of Nature) within the Ministry of Agriculture has only one professional Angolan forester. In addition, there are a number of expatriates, mostly Cubans, working mainly in industrial forest operations in Cabinda. One expatriate forester is, however, working in DNACO. In the provinces where woodfuels are the dominant forest products, DNACO is typically represented by a mid-level - 111 - forest technician who has received his training, mainly in silviculture, in Cuba. As a direct consequence of its lack of funds, DNACO has very little influence over forest operations in Angola. It issues cutting licenses for both forest industries and charcoal makers but is unable to verify whether actual cutting is carried out according to the licenses. It is, of course, even less able to carry out field operations, such as reforestation. 5.21 The other institution is the DNRFE of the MEP. This Department, created in 1981, has a small but growing staff of qualified people, although none with training in forestry. DNRFE is not represented at the provincial level. 5.22 DNRFE pursues various objectives such as research and transfer of technology in new, renewable, and non-conventional forms of energy. Although it does not have any staff trained in forestry, DNRFE has conducted, so far, the only credible study of the use of woodfuels in households. The reorganization of the MEP calls for the transformation of DNRFE into a separate "institute" whose goals would be principally research and experimentation. The splitting-off of DNRFE from the MEP would leave it without any capability either in research or policy-making in woodfuels/biomass and household energy. It has taken years for DNRFE to put together a nucleus of qualified staff, which is exceedingly scarce in AngoLa. The MEP must have some capability and, if qualified staff are not readily available, then the removal of DNRFE from the MEP should be delayed until at least some qualified persons can be attached to MEP proper. A unit or division for woodfuels/biomass for household ene-rgy should be located within the reorganized Technical Department (Cabinete Tecnico). Marketing and Pricing 5.23 In Angola, marketing and pricing mechanisms operate in a distorted economic environment. Official rules for production and marketing of firewood and charcoal, which exist mostly on paper, are seldom applied in practice. Prices paid by consumers often exceed the officially set levels by a factor of 10 to 50. 38/ Thus, while official retail prices for firewood and charcoal are Kz 1.2/kg and Kz 7.5/kg, respectively, effective retail prices are between Kz 5/kg and Kz 300/kg for firewood and between Kz 13/kg and Kz 600/kg for charcoal according to the recent survey undertaken by DNRFE (Table 5.8). 38/ No one knows why official prices for woodfuel are set at all. They are not readily available; they are obviously not meant to be enforced; and except in Namibe, which is in a special situation, they are meaningless. - 112 - Table 5.8: PRICES PAID FOR FIREWOOD AND CHARCOAL IN SEVEN PROVINCES, 1987 (Kz/kg) Firewood Charcoal Province Urban Periburban Rural Urban Periburban Rural Luanda 105 235 265 'fo 280 Bengo - 60 - 130 85 165 Cabinda 40 40 - 200 320 - Huambo 40 60 55 65 310 55 Huila 50 15 - 30 75 40 Namibe 20 15 - 35 35 35 Uige - 80 - 165 160 75 Country 30 40 120 145 220 80 (Official price) (1.2) (1.2) (1.2) (7.5) (7.5) (7.5) Source: DNRFE. 5.24 The picture on woodfuel prices provided by DNRFE constitutes a vast improvement over the previous state of affairs, when there was virtually no reliable data on prices at the consumer level. The mission has also conducted a small survey of charcoal and firewood prices in eacE of three markets in or near Luanda. Retail charcoal prices were about Kz 500/kg in all three markets, i.e., somewhat higher than found by DNRFE. Firewood was sold at Kz 150/kg in the market on the outskirts of Luanda and at Kz 50/kg at the village of Cacuaco near Luanda. This result contradicts the findings of the DNRFE study (which may have been in error) where the price of firewood in the rural areas of Luanda province was found to be much higher than in the periurban areas. 5.25 The system suppling charcoal and firewood to Luanda is not well known. In theory, in order to engage in the woodfuel business, persons or organizations have to apply to DNACO for a cutting license. In reality, these licenses often function as a clearance to transport firewood and charcoal through the checkpoints on the outskirts of Luanda and other cities and towns. This is indicated by the fact that the volume of woodfuel cuttings covered by official licenses amounts to no more than 150,000 mn , while total cuttings are in the general order of 10 million cubic meters. 5.26 Not all the difference between the total cuttings and those licensed are illegal, however, since fuelwood cutting and charcoal production for home consumption are permitted without license. There is evidence that a significant proportion of the charcoal traded in Luanda passes the checkpoints under the guise of "home consumption". Small- scale traders shuttle between central Luanda and Viana (the woodfuels - 113 - terminal, outside the Luanda checkpoints) bringing only two sacks of charcoal per person per trip. They pass the checkpoints claiming that the charcoal they are carrying is for their own home consumption. 5.27 Based on data gathered by DNFRE, the costs of various fuels as a source of cooking heat can be approximated as shown in Table 5.9. Table 5.9: COMPARATIVE COOKING COSTS OF FOUR HOUSEHOLD FUELS a/ Estimated Effective Cost Price Heat Value End-Use of Useful Heat Fuel Kz/kg kcal/kg Efficiency Kz/1000 kcal Firewood 45 3,500 10% 130 Charcoal 180 5,000 15% 240 LPG 60 11,000 55% 10 Kerosene 135 9,000 40% 38 a/ Fuel prices are calculated as means. Source: DNRFE and mission calculations. 5.28 On average the price structure of cooking fuels is strongly biased in favour of those households which have access to LPG or kerosene, i.e., mainly relatively well-ofZ families living in urban centers. Thus, policy measures designed to raise the relative price of LPG and kerosene will mainly affect those households which, under the present conditions, are well-off. Charcoal, on the other hand, turns out to be the most expensive urban household fuel. Clearly, the terms of trade between commercial fuels (LPG, kerosene) and woodfuels are hardly in line with their relative scarcities. Rather, they reflect the fact that the fuels are currently supplied to two different markets (official and parallel) where the purchasing power of the Kwanza varies considerably. Organizations of Woodfuel Suppliers 5.29 In most cases, owners of means of transport occupy a key position as regards exploitation of wood, its conversion to firewood and charcoal, and transportation to urban areas. Truck owners engage laborers for cutting wood and producing charcoal. In certain areas of the country, however, for example in the province of Namibe, local laborers formed cooperatives which apply for cutting licenses, produ'ce firewood and charcoal and transport it to the consumption centers. As the cooperative is formed among local inhabitants licensed to use a specific area, its members are more likely to have an interest in the - 1.4 - sustained production of wood. They might, thus, respect certain restrictions such as minimum diameters for cutting, prohibition to cultivate exploited areas, obligations to carry out reforestation, etc. A limited iumber of cooperatives, each working within a set geographical area, is also fairly easy to supervise. To improve the supply of woodfuels, cooperatives might be an option where security conditions make them feasible and where the problems (as in Namibe) are not excessively serious. Elsewhere, private truck owners using hired labor would have to continue supplying the bulk of woodfuel needs. While it is not possible to do so at present, these private entrepreneurs would eventually have to be supervised more stringently to ensure that minimal rules are followed to permit forest regeneration and, thus, sustained wood production. Issues and Recommendations 5.30 In comparison with most other countries in Sub-Saharan Africa, the wood resources of Angola are large, particularly in relation to its population. Thus, for a significant percentage of the inhabitants of Angola, there is no serious fuelwood problem. Most of the forests of Angola are exploited for fuelwood far below their rates of growth. However, all urban concentrations on the seaboard and the major inland cities constitute pockets of wood shortage. 5.31 Although reliable statistics are not available, it is estimated that 3 million people live in the comparatively dry coastal areas with scarce supplies of woodfuels. While some scattered fishing and farming communities can obtain the woodfuel they need from the local bushland, most concentrations of people have to obtain woodfuels from considerable distances. Inland, mainly in the township of Huambo and other provincial capitals, about one million persons are experiencing a shortage of fuelwood. Thus, a total of some four million people in Angola, i.e., almost half the population, live in areas with a more or less pronounced shortage of woodfuels. Unless serious action is taken, the problem will probably continue to grow as people migrate toward the coast and toward towns and cities. 5.32 The characteristics of supply and demand for woodfuels are quite different on the seaboard than inland. First, the natural con- ditions for tree growth are less favorable in the dry coastal areas than in the moister inland areas. This means that there is less wood available per unit area, both at present and as a result of possible reforestation activities. Second, the access to alternative sources of household energy, mainly LPG and kerosene, is much easier on the seaboard than inland. 5.33 In the short run, security problems severely restrict the movement of fuelwood from wood-rich to wood-poor areas of the country. When the security situation improves, the institutional weaknesses of the - 115 - forestry sector are likely to become the main bottlenecks to rapid improvement. Consequently, the discussion of possible activities aiming at an improvement will have to be made in two parts: actions that can be carried out under the present security situation and those that require marked improvements in the security situation. They are labelled "Peace" and "No Peace". The priorities for action are summarized in Table 5.10. Table 5.10: PRIORITY LISTING OF ACTIVITIES Increased Improved Fuel Efficiency Supply Wood Substitution of Stoves System Froduction No Peace Coast I 11 III IV Inland IV I 11 III Peace Coast It l1 I IV Inland IV III 11 I Note: Although not shown as a distinct activity, institutional strengthening and technical assistance would become much more urgent and important with the return of peace. Actions That Can Be Undertaken Under the Present "No Peace" Situation 5.34 As the problem is different on the coast versus inland, the two areas will be treated separately. Only activitie3 explicitly directed at the problem areas, i.e., the urban concentrations on the coastal strip and the major urban concentrations inland, will be discussed. This does not imply that all is well for the estimated four to five million people living in the countryside but only that major woodfuel problems lie elsewhere. Also, information on the overall rural energy situation is unavailable. A major survey would be needed to obtain such information. The Coast: "No Peace" 5.35 A number of alternative actions to improve the situation can be envisaged. They would comprise the following: - Increased use of fuels other than firewood and charcoal (substitution). - More efficient use of firewood and charcoal. - 116 - - Improvements in the present supply system. - Establishment of forest plantations. (a) Increased Use of Fuels Other than Firewood and Charcoal. In the coastal cities and towns, the use of LPG as a cooking fuel is already fairly widespread. For the population, this fuel represei.!.s a very attractive option. Its average cost, according to DNRFE, is only about Kz 60/kg, (although the official selling price to the public is Kz 15/kg), i.e., much less per useful energy unit than both firewood and charcoal, which sell at average prices of Kz 45/kg and Kz 180/kg respectively. LPG could become an even more important household fuel in the coastal areas, 39/ even if it were to increase considerably in price. The main problems are likely to be supply shortages of gas bottles and stoves, and, possibly, cash-flow difficulties for the initial purchase of bottles among poor families. The low cost of kerosene parallels that of LPG. It is, however, much less widespread as a cooking fuel and has a less developed distribution network. On the other hand, it poses fewer technical problems in distribution. Perhaps the most important role of kerosene would be as a source of light in rural dwellings. The economics of kerosene versus LPC in urban household energy have not been studied, (b) More Efficient Use of Firewood and Charcoal. At present, firewood is almost universally burned in three-stone hearths whose thermal efficiency is quite low, probably about 10%. Charcoal is generally burned in simple square metallic stoves with efficiencies around 15%. There is obvious scope for improvements in stove design for both firewood and charcoal. There are also local varieties of improved stoves which could be used as demonstrations or starting points for more efficient designs. (c) Improvements in the Present Supply System. The present supply system is not well known as it is organized informally, with the owners of means of transport having a key position as holders of cutting licenses. In Namibe, where the security situation is much better than average, the charcoal producers have formed cooperatives which, in turn, have obtained cutting licenses and also organized the transport of the charcoal to the consumption centers. Where possible, woodfuel cooperatives are recommended. A limited number of charcoal cooperatives constitute a more easily managed group than a much larger 39/ About 30,000 tons of LPG are currently being consumed, mostly in Luanda and other coastal cities. - 117 - number of truck owners, charcoal producing farmers, and unorganized laborers. The cooperatives could get various kinds of support (tools, etc.) as compensation for the obligation to follow certain conservation rules in their operations. Those rules would primarily consist of forestry and cutting management principles to be followed in forest operations to ensure sustained wood production. (d) Establishment of Forest Plantations. The establishment of forest plantations is an awkward proposition for the coastal areas of Angola. If they were to be established close to the consumers, in the arid coastal areas, their establishment would be technically difficult and expensive. Further, tree growth would be slow, resulting in long rotations and low yield, again causing higher costs. If the plantations were to be established in the moister inland areas, long transport distances might outweigh the more advantageous growth conditions. Further, the general security situation would make the establishment and management of forest plantations difficult and costly, and the same applies to transport of the output of the plantations. Finally, in those areas where forest plantations could be expected to yield well, there are often natural forests available, which can be exploited. By exploiting the natural forests, the initial investment and the delay of five to ten years before harvesting can be avoided. Inland Areas: "No Peace" 5.36 Inland, where about one million persons live in wood-short areas, replacing wood and charcoal by other fuels, mainly LPG, is much less attractive. Instead, the following options (in order of priority) would seem more appropriate: - More efficient use of firewood and charcoal. - Improvement in the present supply system. - Improved use of existing forest plantations or establishment of new ones. - Improved supplies of modern fuels (kerosene, LPG). (a) More Efficient Use of Firewood and Charcoal. In the inland cities (Huambo and other provincial capitals) three-stone hearths for wood and simple metal stoves for charcoal are in general use. As in the coastal areas, there is scope for improvement. - 118 - (b) Improvement in the Present Supply System. The average transport distance for woodfuels from existing forests--natural or planted--to the consumers is much shorter inland than on the coast. Although it seems that the present supply systems function better inland than on the coast, there is scope for improvement, particularly in order to safeguard the long-run productivity of the wood sources. This would require the establishment of rules guiding the exploitation of the forest areas and a system for supervision and control of the licensees. (c) Improved Management of Existing Forest Plantations or Establishment of New Ones. At present, existing forest plantations, for example in the province of Huambo, are not used in any systematic manner for the supply of woodfuels to urban centers. Furthermore, present practices (cutting, burning of stumps, cultivation of cleared areas) prevent the regeneration of the harvested areas. A program for controlling cutting methods in nearby plantations should be developed. Of lower priority is whether any additional plantations are required. Inland, the growth conditions for forest plantations are generally good. Security and the difficulties of mobilizing sufficient funds for such plantations are likely to be serious problems. Alternative Priorities in a "Peace" Situation 5.37 In a situation of peace, the order of priority of the various actions would change somewhat, so that field activities, particularly forestry management and plantations, would ascend in priority when the countryside becomes secureo Such increased field activities would, in turn, require a strengthening of the institutions concerned. Institutional strengthening is not the highest priority (although gaps in the institutional structure need to be filled even under the current security situation) but with the return of peace, institutional strengthening would become a high priority. Proposed Actions 5.38 Various solutions for the problem areas on the coast and around the urban centers inland were discussed in the previous section. In this section, four groups of activities for the major problem areas will be proposed. The first is a pilot project to be executed in Huila-Namibe. The second and third ones are activities for Luanda and Benguela/Lobito, respectively, while the fourth one concerns the township of Huambo. In addition, three national-level activities are proposed to support these four regional projects: improved stoves, agroforestry, and institutional strengthening. - 119 - Pilot Project in Huila-Namibe 5.39 In the provinces of Huila and Namibe, the military situation is much better than elsewhere in Angola, with most of the two provinces safe. Also, the economy in these provinces seems to be functioning better than elsewhere. Additionally, the provinces do not have large population concentrations in areas far from the forests. It is true that the coastal towns of Namibe and Tombwa have virtually no wood resources in their neighborhood, but the fact that the towns are fairly small makes the problem more easily managed than in the cases of Luanda and Benguela/Lobito. Taken together, these facts point to the possibility of developing a pilot project in energy forestry in these two provinces. Coast (Mainly Namibe Township and its Neighborhood) - Priority I: Fuel subsitution. - Priority II: Increased efficiency of stoves. - Priority III: Improved supply system. Inland (Mainly Lubango Town and its Surroundings) - Priority I: Increased efficiency of stoves. - Priority II: Improved supply system. - Priority III: Wood production/forestry management. 5.40 Luanda. The city of Luanda and its surroundings constitute the single largest consumption center for household fuels. The use of LPG, kerosene, and electricity in households is more widespread than elsewhere but woodfuels, mainly charcoal, are still used by almost half the population. Wood and charcoal are bought in the parallel market, at about Kz 140/kg for firewood and Kz 320/kg for charcoal. In comparison, the average price for LPG in Luanda is about Kz 100/kg. In terms of useful heat delivered to the pot, the cost of wood and charcoal is about 15 times higher than LPG (Kz 400, 400, and 27 per 1,000 kcal, respectively). A logical first priority is thus to spread the use of LPG among the urban and periurban population of Luanda. At the same time, however, improved stoves for firewood and charcoal should be made available to the large share of the urban and periurban population that will continue to use wood or charcoal for many years. - 120 - 5.41 The supply system for woodfuels is not well known, or at least not well documented. Many of the present operations are carried out by individuals operating informally outside the legal framework. Therefore, mapping of the supply system is not likely to be an easy task. Improved knowledge of that system seems, however, to be a necessary condition for any action aimed at its improvement. A survey of the present supply system for firewood and charcoal to Luanda is, thus, the first step in the design of an improved system for supply of firewood and charcoal to Luanda. 5.42 Benguela/Lobito. The coastal cities of Benguela and Lobito, in the province of Benguela, exhibit much of the same characteristics as Luanda. Generally speaking, the solutions in the medium term would also be similar, i.e., a gradual change to LPG for household cooking fuel, improved stoves for fuelwood and charcoal, and preparations for an improved system for the supply of firewood and charcoal to the urban and periurban areas. The geographical/ecological position of Benguela/Lobito is, however, somewhat more favorable than that for Luanda, as areas favorable for tree growth (i.e., the escarpment) are closer to the population centers. Further, there are plantations, mainly of eucalyptus, on tha escarpment close to the border with the province of Huambo. Those plantations were established as the raw material source for the Alto Catumbela pulp and paper mill. As the mill has been closLd for some time, the forests might be available for an alternative use. The fact that the forests are located fairly close to the railway line between Huambo and Benguela should facilitate their use for energy in Benguela and Lobito. 5.43 Huambo Township. With about half a million inhabitants using firewood and charcoal, Huambo is the largest inland fuelwood shortage area. For that area, the order of priorities is the following: (a) increased efficiency of stoves; (b) improved supply system; (c) wood production (or better forestry/plantation management); and (d) fuel substitution. A project for the development and dissemination of improved wood and charcoal stoves for Huambo would have much in common with the corresponding projects in Luanda and Benguela/Lobito. The same is basically true for the improvement of the supply system, even if the natural conditions for tree growth are much more favorable in Huambo. In fact, the province of Huambo already has extensive plantations of eucalyptus, mainly along the Benguela railway line, established to provide firewood to the steam engines. Most of the locomotives now run on diesel oil instead of wood, which means that the forests should be available for other uses. Some of the forests may have been damaged or destroyed due to excessive cutting or clearing for agriculture, but most are likely to have remained productive. In the short term, the lack of - 121 - security in the forest areas may be the more serious limitation to their increased use as woodfuel sources. 5.44 In seeking to improve the supply of woodfuels for Huambo township, an inventory should be made of the existing raw material sources, whether natural or man-made. Second, a management system should be developed for the forests. Further, it might be advantageous to establish cooperatives among the producers of firewood and charcoal. It is also necessary to consider whether longer-term needs can be met from existing forests or would require additional plantations. Proposed Supportive Action at the National Level 5.45 There are three activities of value for all the provincial projects which would be carried out most advantageously at the national level or at least with strong support from an institution at the national level. First, improved cooking stoves for firewood and charcoal need to be designed and introduced. Second, suitable forms of tree cultivation in conjunction with agriculture (agroforestry) should be developed. Third, there is a need to strengthen the forestry institutions in Angola to help define and enforce policies, to support the provincial projects, and to prepare for more activities when the security situation improves. 5.46 Improved Cooking Stoves. In its preparations for the Seminar on Firewood and Charcoal in Luanda in June 1987, DNRFE made a survey of biomass cooking stoves in the seven provinces surveyed. This is the appropriate, albeit modest, beginning of a national program for the design, production, and dissemination of improved cooking stoves in Angola--which are a priority in Luanda, Benguela/Lobito and Huambo. It seems practical that much of the development work should be carried out at the national level or at least with strong technical support from that level. It is, further, advantageous that field trials for improved stoves should initially be concentrated in Luanda, where a large number of people use inefficient stoves for woodfuels bought at quite high prices. Benguela/Lobito would get second priority in the program and Huambo, third. 5.47 Agroforestry. Planting of trees which (among other things) yield fuelwood can take various forms. At one end of the spectrum there is the dedicated energy plantation, where densely planted trees form a forest which is managed for the main purpose of producing fuelwood. At the other end of the spectrum, there are trees planted for other purposes, such as to provide shade, shelter against wind, or to produce fruit, fodder or other non-wood products. While international assistance in energy forestry during the 1970s focused on "village woodlots" and other forms of collective creation of fuelwood forests, attention has shifted towards other forms of tree cultivation during the present decade. There are several reasons for this shift. First are the technical and economic problems associated with village woodlot schemes - 122 - and other forms of collective or centrally organized fuelwood plantations, such as peri-urban plantations. Another is the realization that the planting of trees for the production of fuelwood alone does not rank high in the list of priority for most rural inhabitants. 5.48 In Angola at present, both the village woodlot and the peri- urban fuelwood plantation are concepts likely to face severe implementa- tion problems. They would face the same technical and economic problems encountered elsewhere but with much less favorable technical and economic conditions. Further, the extreme scarcity of trained people for project planning and management, together with the poor security situation, would make fuelwood plantations a risky venture. Instead, the agroforestry approach would stimulate other forms of tree cultivation, where farmers plant trees for a variety of purposes, with fuelwood as a by-product. Such plantations can take the form of windbreaks, a small grove to provide shade or shelter, fruit orchards, single trees in agricultural fields, contour plantations (or soil conservation and erosion control), possibly for fodder, and so on. In many of those cases, the trees supplement rather than compete with agriculture and animal husbandry, thereby eliminating one difficulty and reducing the opportunity cost of land, so that economic criteria are more easily satisfied. 5.49 The forms of tree planting referred to above are often collectively called "agroforestry". Such tree planting, for food, fodder, shelter, etc., would obviously be only a minor source of woodfuel for the major urban areas, but could become an important woodfuel source for rural people. An important advantage of such tree planting in Angola is that it does not need centralized project management but can be integrated into other agricultural activities. Thus, the weakness of the forestry administration in Angola is less of an obstacle for agroforestry than for the traditional forms of forestry. 5.50 In the present security situation in Angola, activities in the field of agroforestry might have to be limited to training combined with field trials. Among the activities proposed above, all but one are urban-oriented. The exception, the pilot project for Huila/Namibe, would thus provide a focal point for agroforestry centered on the agricultural school of Tchivinguiro, 35 km from Lubango. Strengthening of Institutions 5.51 In Angola, the institutions in the forestry sector, i.e., the DNACO and the DNRFE, are weak. The selection of activities proposed in this chapter has been made against this background. Nevertheless, these activities would be very much easier to carry out if the institutions concerned were strengthened. The minimal needs are for some capability in the MEP (currently provided by DNRFE), a strengthening of DNACO in the provinces where projects are proposed, some technical assistance at the central level for DNACO, and some funds for very limited studies, such as a forest inventory around Huambo, and a study of woodfuels marketing systems in Luanda, Benguela/Lobito and Lubango. - 123 - Annex 1 Page 1 of 6 KACROECOOKIIC INDICATORS Table 1; INDICATORS FOR GDP AND GNP 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 GDP at factor cost n.a n.a n.a n.a n.a 3,922 4,307 4,811 4,174 (in million USS) - of which petroleum n.a n.a n.a n.a n.a n.a 1,098 !,383 1,487 1,131 GDP at market prices (In million current US$) 2,445 2,750 3,323 3,617 3,933 3,647 4,212 4,719 4,831 4,409 Deflator (1980=100) a/ 78.2 85,2 91.3 100 92.6 84.8 77.8 71.5 65.6 60.3 GDP at 1980 prices a/ (In million USS) 3,127 3,228 3,640 3,617 3,642 3,091 3,278 3,37S 3 70 2,658 GNP at current prices b/ (In million USS) n.a n.a n,a n,a n,a 3,529 3,725 4,153 4.'99 3,863 GNP at 1980 prices a/ n.a n.a n.a n.a n.a 2,991 2,899 2,969 2.755 2,329 Exchange rate: US$1= 30 Kz. a/ Estimate for 1983-86. b/ Estimates based on balance of payments statistics. Source: Ministry of Planning, the National Bank of Angola (BNA), and mission estimates. - 124 - Annex 1 Page 2 of 6 Table 2: GOVERNT BUDGET (In million Kz) 1978 1979 1980 1981 1982 1983 1984 1985 1986 Total revenues 38,866 37,155 60,143 73,708 50,656 55,589 74,556 78,528 71,205 Pet Iu(e ) n.a. n.a. 56.4 61,3 41.5 48.0 56,7 53,1 29.7 Tax Revenues n.a. n.a. 51,556 63,340 41,573 47,136 62,193 66,646 51,945 - of which petroleum n.a. n.a. 33,917 45,185 21,046 26,672 42,267 41,667 21,132 Non-tax revenues n.a. n.a, 8,587 10,368 9,083 8,453 11,643 11,882 19,263 Total Expenditures 41,331 57,540 76,920 91,640 72,133 74,427 86,360 97,987 84,207 - of which Defense 10,270 15,100 16,821 18,505 18,257 23,295 31,943 34,306 32,630 Deficit 2,465 20.385 16,777 17,932 21,477 18,838 11,804 19,459 13,002 Other areas 5,253 5,712 13,100 13,100 2,000 2,379 2,857 4,063 4,831 Deficit 2,465 20,385 27,000 17,900 21,500 11-990 7,746 11,960 15,000 Source: Ministry of Finance. - 125 - Annex I Page 3 of 6 Table 3: BALANCE OF PAYrENTS (In million Kz) 1978 1979 1980 1983 1982 1983 1984 1985 1986 Exports a/ 29,940 35,460 48,699 38,126 44,705 47,493 58,800 59,280 38,350 - of which petroleum 16,507 26,745 41,730 40,350 b/ 37,029 45,768 52,455 56,868 34,485 imports 26,640 38,840 44,503 45,417 (33,684) (29,708) (37,950) (41,527) (31,868) Balance 3,300 (1,380) 4,196 (7,291) 11,021 17,785 20,8509 17,753 6,482 Invisibles (net) (3,720) (4,350) (5,734) (12,683) (19,007) (19,792) (23,613) (26,494) (22,009) of which: - Factor services (not) n.a n.o n.a n.a (13,050) (14,520) (16,950) (18,960) (16,410) Unrequted transfers 90 300 508 1,505 794 986 1,047 1,657 2,108 Current account (330) (5,430) (1,030) (18,469) (7,192) (1,021) (1,716) (7,084) (13,419) Medium- + long- term capital 1,950 1,950 7,016 7,110 3,026 1,628 6,103 5,335 1,567 Basic balance 1,620 (3,480) 5,986 (11,359) (4,166) 607 4,387 (1,749) (11,852) Short-term capital, errors and omissions 1,080 3,930 (8,096) 5,566 3,901 (466) (2,557) 2,709 11,103 Overall balance 2,700 450 2,110 (5,793) (265) 141 1,830 960 (749) a/ Between 1978 and 1981 taxes on oil exports paid by foreign oil companies are included in service income. b/ Figures were obtained from MEP Independent of total export figures (BNA). Source: National Bank of Angola (BNA). - 126 - Annex 1 Page 4 of 6 Table 4: DISBURSED PUBLIC EXTERNAL DEST (In million Kz) 1978 1979 1980 1981 1982 1983 1984 1985 1986 Short-term n.a n.a n.o nfa 8,138 7,551 4,888 7,525 18,434 Mkdium- Long-term 11,340 11,430 18,420 59,490 62,259 63,193 68,482 73,482 73,695 TOTAL n.o n.a n.o n.o 70,397 70,744 73,370 81,007 92,129 Total as % of GNP n.a n.u n.o n.a 73.1 63.3 58.8 64,3 79.5 Total as % of exports of goods and services n.a. n.o n.o n.a 144.7 142.1 118.7 128.3 223.9 Source: The National Bank of Angola (BNA) and own calculations. - 127 - Annex 1 Page 5 of 6 Table 5: GOVERNMENT PROJECTIONS FOR KEY ECONOMIC FIGURES (in million US$) Actual Projected ----------- 1986 1987 1988 1989 1990 Revenues from oil exports 1,150 1,653 1,841 2,018 2,137 Average price of oil 12 16 16 17 18 (In USS/bbl) Trade balance 216 602 682 756 777 Current account (448) (238) (181) (142) (135) External debt 3,024 3,163 3,289 3,335 3,393 of which: - Short-term 568 219 257 309 359 - Long-term 2,456 2,944 3,032 3,046 3,034 Potential for gross 1,123 659 1,320 1,562 1,692 domestic investment Source: The National Bank of Angola (BNA). Table 6: COWPARATIVE ECONOMIC INDICATORS External Pub- Debt Government Broad Current Snare Share of GNP per GDP lic Debt as Service Expenditure Money as Account Deficit of '^'roleum Oil Income In Country Capita Growth a/ S of GNP Ratio as S of G9P % of GDP as S of GDP Sector in GDP Goverment Revenues (1985 USS) (1980-85) (1985) (1985) (1985) (1985) (1985) (1985) (1985) Zambia 390 0.1 150.8 10.2 30.3 33.4 4.2 -- -- Zimbabwe 680 2.5 31.3 32.2 39.1 45.2 2.1 -- -- Tanzania 290 0.8 48.5 16.7 24.7 -- -- -- -- Malawi 160 2.0 75.7 -- 29.5 24.3 -- -- -- Lesotho 470 0.5 30.1 6.2 22.7 48.8 3.5 -- -- Botswana 840 12.1 47.3 5.4 48,2 29.5 16.9 -- -- Mozambique 160 -9.6 -- -_ __ 94.8 b/ __ __ _- Angola 485 -2.3 59.2 22.0 71.8 142.6 4.9 30.0 53.1 OD Tunisia 1,190 4.1 56.1 24.9 40.4 48.6 7.4 -- -- Congo 1,110 7.8 86.5 19.6 -- 16.2 -- 40.0 66.6 Gabon 3,350 4.5 a/ n.a. -- -- -- -- 45.0 66.0 Nigeria 800 -3.4 17.8 30.8 -- 34.7 1.6 23.0 -- a/ Least square estimates. b/ 1984 estimate. c/ 1980-84 Source: World Development Report, 1987 and mission estimates. o0 Ft - 129 - Annex 2 Page 1 of 8 INSTITUTIONAL ORGANIZATIONS WITHIN THE ENERGY SECTOR The Ministry of Energy and Petroleum (MEP) is the principal Government body responsible for the development and implementation of national policy within the energy sector. In practice, it deals with the major aspects of the oil, gas, and electricity power industries. The main responsibilities of the MEP include: (a) preparing the National Plan for the energy sector; (b) supervising SONANGOL and the power utilities; (c) promoting and coordinating training activities within the sector; and (d) implementing international technical cooperation. The present Ministry has its origin in the 1984 merging of the Ministry of Energy and the Ministry of Petroleum--both separate entities created in 1976. In 1987 the MEP was re-organized into four units or Gabinetes with functional responsibilities over the electricity and petroleum subsectors. The four Gabinetes are: Planning, Technology, Legal, and Human Resources. In addition to the Gabinetes, the departments of New and Renewable Energy Sources and of Budget are kept under the MEP's direct supervision, at a lower administrative level. Within the MEP, the National Department for Transformation is responsible for coordination with and monitoring of the refinery industry, and the Department of External Trade monitors external and internal trade in crude oil and petroleum products. These two departments have continued in existence in spite of the reorganization. Their functions will eventually be redistributed to the Gabinetes. The MEP itself is not involved in the day-to-day management of the sector but establishes guidelines to be used by executing agencies in the implementation of policies. An organization chart for the MEP is given on page 2. MINISTRY OF ENERGY AND PETROLEUM (MEP) Inspector General | IMinisteri Deputy Minister Deputy Minister Physical Protection for Energy for Petroleum | and State Secrets | Ottices Pianning Technical and Legal a Energy Development I Economic |.|Elecricity | t r tion Studies Studies A tnalAviv l |Finance & Petroleum Contracts Labor 0 Accounting Resources Departments investment[ Studies a Professional Project Mgmt Energy Devel. rraining Commercialization I Inspection| 8 Pricing g Inspection New and External Budget Provincial Renewable Trade Managementl I Delegation Sources Source: MEP. 0 OD - 131 - Annex 2 Page 3 of 8 PETROLEUM PRODUCTS The MEP has specific managerial responsibilities within the petroleum subsector: (a) authorizing the opening of blocks for bids; (b) approving development programs; (c) authorizing the commencement of production; (d) regulating field production levels; (e) sanctioning the flaring of natural gas in special cases; (f) setting crude oil prices for tax purposes; and (g) reviewing SONANGOL's investment programs and foreign companies' accounts. Sociedade Nacional de Combustiveis de Angola (SONANGOL) is the State-owned company established in 1976 as the business arm of the Covernment with the main responsibility of implementing policy for che petroleum sector. Its structure is based on that of ANGOL, the only Portuguese company in Angola with interests in petroleum, which was nationalized after independence. SONANGOL is managed by a Director General who reports to the MEP, and three deputy directors responsible for hydrocarbons, distri- bution, and adiainistration and finance. The Deputy Director for Hydrocarbons supervises four departments: Exploration, Production and Reserves, Administration and Finance 1/, and a fourth department which handles air transport services for the oil companies. The Director General directly supervises a unit responsible for negotiations. The main function of SONANGOL is to oversee the petroleum operations of all foreign companies. It makes recommendations to the Government in respect to areas that should be open for exploration, conducts the bidding process, and handles negotiations. SONANGOL reviews the foreign companies' proposals, and, once operations begin, discusses work programs and budgets and supervises activities. 1/ The Deputy Director for Administration and Finance deals with corporate functions. - 132 - Annex 2 Page 4 of 8 In addition to exploration and production activities, SONANGOL is responsible for domestic product distribution and marketing. SONANCOL owns and operates a network of storage installations throughout the country and maintains a small marine tanker fleet and a sizeable fleet of road tankers. SONANGOL enters into partnerships with qualified foreign companies in order to obtain the requisite financial and technical resources for exploration, development, and production. These partnerships take one of two forms: (a) Joint Ventures in which both SONANGOL and its partners share in investments and receive petroleum produced according to their percentage interests. Foreign companies pay taxes and royal- ties on their equity shares of production. SONANGOL pays taxes to the Government and operates under the same foreign exchange regulations that govern oil companies. (b) Production Sharing Agreements in which the companies serve as contractors to SONANGOL, finance the full cost of investment in both exploration and development, and are compensated with a share of the oil produced. Companies' investments costs are recouped from "cost oil", set at a fixed proportion of production (normally 50%) and are then taxed on their "profit oil". Hydrocarbon deposits declared non-commercial are the property of the State and revert to SONANGOL. SONANGOL has a joint venture arrangement with STINNES, a West German oil trading firm, in which all matters pertaining to the export/- sale and import/procurement of petroleum products for Angola are handled. As a working partner, SONANGOL finances its share of investment programs in both offshore and onshore activities with CHEVRON in Cabinda, FINA in onshore Angola, and TEXACO in Block No.2. It has supervisory rights over these ventures, but limited day-to-day operational capacity. In one instance SONANGOL indirectly acts as operator. Emprcsa dos Servicios Petroliferos de Angola (ESPA) is a mixed-economy enterprise set up to operate Block No.4. It is owned by: SONANGOL (51%), BRASPETRO (24.5%), and PETROFINA (24.5%). Each company contributes to the management and staffing of the operating company. Fina Petroleos de Angola (FPA) is primarily a privately-owned company within the petroleum industry. Shares are owned: 55.6% by PETROFINA, 11.1% by the public, and 33.3% by the Government. Government shares are described as "non-participating", and do not entitle it to any share of the profits. - 133 - Annex 2 Page 5 of 8 FPA owns interests upstream in two blocks. It acts as operator of the Kwanza and Soyo fields on behalf of its joint venture partners-- SONANGOL in Kwanza, SONANGOL and TEXACO in Soyo. FPA owns and operates the refinery in Luanda, the principal product supplier to SONANGOL for the domestic market, and also provides SONANCOL with its export volume. The refinery is a comparatively high-cost operation by international standards. It operates on a "cost plus" refinery gate pricing arrangement. All verified operating costs, depreciation, and allowable profit are recovered and there is no particular incentive for cost minimization and optimization of operations. ELECTRIC POWER MEP is responsible for the main operating organizations in the electric power subsector. These are: (a) Empresa Nacional de Electricidade (ENE). ENE was created in 1980 with the aim of becoming the sole national power utility in charge of generation, transmission, and medium-voltage distribution across the country. After independence, ENE received the assets of the Junta Provincial de Electrificacao de Angola (JPEA), represented by the Southern System. The main facility in the Southern System is the Matala hydro plant on the Cunene River. The system also includes the upstream Gove dam, and an unfinished diesel plant in Namibe. In addition to several isolated systems, ENE also currently operates the Central System. This system is based on two hydro plants located on the western part of the Catumbela River at Lomaum and Biopio, and two diesel-fired turbines at Biopio and Huambo; (b) Sociedade Nacional de Estudo e Financiamento de Empreendimentos Ultramarinos (SONEFE). SONEFE is the utility in charge of generation and transmission in the Northern System, the largest in the country. The system directly supplies about 300 industrial consumers at high and medium voltage and operates using the potential of the Kwanza and Donde Rivers and two Jet B-fueled gas turbines in Luanda. The plant in Mabubas, also part of the system, has been shut down for rehabilitation and will remain shut down until late 1988. SONEFE has no centralized department for operations. It is organized into five departments: Planning and Finance, Studies and Projects, Stocks and Purchases, Human Resources, and Administrative Services; (c) Empresa de Electricidade de Luanda (EDEL). EDEL is the utility in charge of medium- and low-voltage distribution in the urban area of Luanda. Autonomous since 1980, and with its legal - 134 - Annex 2 Page 6 of 8 status not fully clarified, it remains separate from ENE due to the significance and volume of its services; (d) Companhia de Electricidade do Lobito e Benguela (CELB). CELB is the utility in charge of distribution to Lobito and Benguela. Nationalized in 1982, it remains a distinct, sepa- rate management unit, as its assets have not yet been integrated into ENE; and (e) Cabinete de Aproveitamento do Medio Kwanza (GAMEK). GAMEK was created in 1982 with the two immediate goals of coordinating and supervising the work to be done at Capanda and defining a development plan for the Middle Kwanza. Over the long term, GAMEK will be in charge of all works related to the exploita- tion of the hydro resources of the Middle Kwanza basin and any new major hydrogeneration investments in the Northern System. NATURAL GAS A limited task force attached to the Oil Production/Reserves Management in SONANGOL deals specifically with the supply aspects of the gas industry. It monitors appraisal and feasibility studies requested by oil companies operating in Angola or presented by foreign companies and consultants. FORESTRY/WOODFUELS The Ministry of Agriculture has a Vice Minister in charge of forestry but no separate forestry department. Within the Ministry, the National Directorate for the Conservation of Nature IDNACO) is the body responsible for regulating the exploitation of firewood. It has provincial representation and issues licences to parties wishing to engage in woodfuel production. The Department of New and Renewable Sources of Energy (DNRFE) within the MEP was created in 1981 and is mainly a research body with no provincial representation at the fit.d level. Its main objectives are research and transfer of technology in new, renewable, and non- conventional forms of energy. - 135 - Annex 2 Page 7 of 8 EMPRESA NACIONAL DE ELECTRICIDADE (ENE) Organization Chart. Present Central Structure |General Director |Deputy Director ||Deputy Director |Regional . .|Planning 8 !Administration | Technic al lDirection Statistics! - North *nac Studies, Projects - South I Standardization! & Construction _ ~~~~~~~~~~I _ - ...... pStocks and | . . , ! Legal Inventories jOperation 8 Planning 8 Maintenance Project Analysis |Human Resources Lomaum a/ | Project Evaluation Project I & Control l Public Relations5 New Technologies a/ Temporary, related to rehabilitation of the Lomaum hydro plant. Source: ENE. - 136 - Annex 2 Page 8 of 8 PROPOSED ORGANIZATION OF THE POWER SUBSECTOR Organization Chart Electricity Coordinating & Planning Commission Central Units Technical Support Economics & Supervision Planning Department Operating ! _SONEFE ntral Southern Dlvisions Power Power System 50oated Systems Table 1: ANGOLA ENERGY BALANCE: 1986 (In 'OOOs toe) Primary Energy Petroleum Products Natural Hydro Crude Char- Electri- LPG Gasoline/ Kero- Aviation Diesel Fuel Line Fuelwood Gas Elect Oil coal city Naphtha sene Fuel Oil Oil Total Totals Gross Supply Production 2,074 a/ 3,418 173 14,102 0 19,765 lports 12 116 16 144 144 Primary Exports (177) (12,637) 3 (12,814) Stock Changes b/ 33 (10) (12) (11) 4 (7) (34) (48) (15) Total avail. supply 2,074 3,241 173 !,498 0 0 2 (12) (ll) 120 9 (34) 96 7,080 Conversion Petroleum refining (1,445) 20 130 36 174 344 658 1,362 (83) Non-energy use (3,24)) c/ (53) d/ 0 (3,294) Charcoal production (364) 364 0 0 I.. Electric power generation e/ (173) 188 (15) (15) 0 w Conversion losses b/ (892) a/ (125) 0 (1,017) Trhns. & distr. losses (14) 0 (14) Net supply available 816 0 0 0 364 49 22 118 47 294 338 624 1,443 2,672 Secondary exports (13) (6) (508) (527) (527) Bunker sales (29) (6) (2) (37) (37) Net domestic consumption 816 0 0 0 364 49 22 105 47 265 326 114 879 2,108 Consumption by sector (not available) Industry O Transport Households/Public Agriculture 0 Other a/ Refer to Annex 8, Table 2, for assumptions. b/ Includes statistical discrepancies, stock draw (or build). c/ Either flared or reinjected. 188 MMCFD (1,695,000 toe). d/ Includes 7,253 Tons of asphalt and 45,452 tons of crude for field use. es An unknown amount of natural gas Is used for generating electricity. Source: MEP and mission estimates - 138 - Annex 3 Page 2 of 7 COMMODITY BALANCE - FUELWOOD FIREWOOD Consumption 000s toe '000 tons/y Urban 72 24 Periurban 190 65 Rural 1,600 544 Subtotal 1,862 633 Heating a/ 360 122 Industry b/ 180 61 Total 2,402 816 Conversion Factor: 0.34 toe/ton firewood. a/ Approximately 3.0 million use I kg/d for 4 months. 3 x 1 x 4 x 30 = 360,000 tons. Conversion: 360 x 0.34 toe/ton = 122,000 toe. b/ Industrial use of firewood is equivalent to: 10% firewood in urban and periurban households and 5% firewood in rural households. POPULATION: Urban 2.5 million Periburban 1.5 million Rural 5.0 million Total 9.0 million - 139 - Annex 3 Page 3 of 7 COMMODITY BALANCE - FUELWOOD CHARCOAL Consumption '000s toe '000 tons/y Urban 115 79 Periburban 103 71 Rural 226 156 Subtotal 444 306 Heating a/ 84 58 Total 528 364 Conversion Factor: 0.69 toe/ton charcoal. a/ Approximately 1.0 million use 0.7 kg/d for 4 months 1 x 0.7 x 4 x 30 = 84,000 tons. Conversion: 84 x 0.69 toe/ton = 58,000 toe. For conversion of charcoal into toe of firewood, it is assumed that 7 kg of firewood is equivalent to 1.0 kg charcoal. CHARCOAL FIREWOOD '000s toe '000s tons '000s tons Urban 115 805 273 Periurban 103 721 245 Rural 226 1,582 538 Heating 84 558 200 Total 528 3,696 1,256 Conversion factor: 0.34 toe/ton firewood. Therefore 1,256 toe are needed to produce 364 toe charcoal. Hence, total fuelwood production: '000s toe Firewood 816 Charcoal 1,256 2,072 - 140 - Annex 3 Page 4 of 7 COMMODITY BALANCE - NATURAL GAS Production (1986): 379 MMCFD. 1 ft3 = 0.0283168 m3 Calorific content of natural gas: 8.9 x 103 kcal/m3 379 x 106 ft3 x 365 days = 138,335 x 106 ft3/y 138,335 x 106 x 0.0283168 = 3,912.20 x 106 m3 Calorific content: 8v9 x 103 x 3,912.20 x 106 kcal = 34,863.12 x 109 kcal 10.6 x 106 kcal = 1 toe 1 x 34,863.12 x 109 kcal …---------------------- = 3,417.9 x 10 toe 10.2 x 1o6 kcal Equal to 3,418,000 toe Consumption (1986): 191 MMCFD. Therefore, amount of natural gas flared: (379-191) MMCFD = 188 MMCFD 188 MMCFD x 106 x 365 x 0.0283168 x 8.9 x 103 x 1 ----…-------------------------------------------- = toe 10.2 x 106 Equal to 1,695,000 toe - 141 - Annex 3 Page 5 of 7 COMMODITY BALANCE - ELECTRICITY HYDRO Systems GWh North 595 CentraL 40 South 56 Total Generation 691 1 MWh = 859,845 kcal Calorific content: 691 x 103 x 859845 x 3 x 1 10.2 x 106 Equal to 58,250.28 toe, or 58,000 toe THERMAL Systems GWh North 21 Central 41 South 0 Total Generation 62 1 MWh = 859,845 kcal Calorific content: 62 x 103 x 859,845 x 303 x 1 10.2 x 106 Equal to 5,226.51 toe, or 5,000 toe - 142 - Annex 3 Page 6 of 7 CONVERSION INTO PRIMARY ENERGY Hydro Conversion factor: 0.25 toe/MWh. 691 x 103 x 0.25 = 172.75 x 103 toe = 173,000 toe Thermal Conversion factor: 1 MWh = 0.248 toe at 34% efficiency in thermal (oil) generation. 62 x 103 x 0.2248 = 15.3 x 103 toe = 15,000 toe Total Electric Power Generatior, (Hydro + Thermal) (173,000 + 15,000) toe = 188,000 toe Conversion Losses Conversion loss (hydro) + conversion loss (thermal): L (173,000 - 58,000) + (15,000 - 5,000) 1 toe = 125,000 toe Transmission and Distribution Losses Total generation: 753 CWh. 22% transmission and distribution losses. 753 x 0.22 = 165.66 GWh 165.66 x 103 MWh x 859,845 x 103 kcal x 1 10.2 x 106 kcal Equal to 13,964.89 toe, or 14,000 toe. - 143 - Annex 3 Page 7 of 7 COMMODITY BALANCE - PETROLEUM '000s tons Conv. '000s toe Factor PRODUCTION Crude 14,102 1.00 14,102 IMPORTS Average Fuel 114,184 1.02 116,467 Casoil 16,435 0.99 16,270 LPG 10,913 1.06 11,567 PRIMARY EXPORTS Crude (12,637) 1.00 (12,637) LPG (166,782) 1.06 (176,788) REFINING Average fuel 171,064 1.02 174,485 Crude (1,452) 1.00 (1,452) a/ Fuel oil 685,767 0.96 658,336 Gasoil 347,288 0.99 343,815 Gasoline/Naphtha 126,146 1.03 129,930 Kerosene 35,637 1.01 35,993 LPG 18,690 1.06 19,810 NON-ENERGY USE Asphalt 7,253 0.99 7,180 Crude 45,452 1.00 45,452 SECONDARY EXPORTS Casoil 5,642 0.99 5,585 Gasoline/Naphtha 11,544 1.03 11,890 Fuel oil 528,766 0.96 507,615 BUNKER SALES Average fuel 28,920 1.02 29,498 Fuel oil 2,600 0.96 2,496 Gasoil 5,850 0.99 5,791 STOCK CHANGES +/(-) Average fuel 4,222 1.02 4,306 Crude 33,277 1.00 33,277 Fuel oil (35,201) 0.96 33,792 LPG (9,621) 1.06 (10,198) Gasoil (7,461) 0.99 (7,386) Gasoline/Naphtha (10,492) 1.03 (10,806) Kerosene 10,673 1.01 10,779 a/ Figure quoted is net of asphalt, i.e., (1452 - 7) = 1,445 thousand toe. Table 1: SELECTED FIGURES ON THE PETROLEUM SECTOR a/ Annual 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 Growth Rate 1978-86 Crude oil production 163 172 94 102 128 156 164 136 128 130 179 205 233 282 7.9% (in 'OOOs bbl/d) Export revenues O from crude oil (in Z million current USS) 233 n.a n.a n.a n.a 550 902 1,391 1,357 1,234 1,526 1,749 1,896 1,150 10.1% 4-4 Deflator b/ 34.7 68.9 81.4 100.0 103.4 98.7 95.1 93.2 90.5 87.6 r .s Dettated revenues from crude oil (in million USS) n.a n.a n.a n.a n.a 798 1,108 1,391 1,403 1,218 1,451 1,583 1,661 976 c/ 3.9% Export revenues frcm crude oil as % of GDP n.a n.a n.a n.a n.a 20.0 27.1 38.5 34.5 33.8 36,2 37.1 39.2 26.1 3.6% c/ Factor income trans- ferred abroad d/ as S of revenue from exports n.a n.a n,a n.a n.a n.a n.a n.a n.a 11.3 11.1 11.3 12.1 16.8 9.2% e/ a/ Using least square estimates. b/ Import unit values, International Financial Statistics. c/ Estimates. M PD d/ Attributable to the petroleum sector. X e/ Growth rate 1982-1986. 0 Source: MEP and mission estimates. - 145 - Annex 4 Page 2 of 6 Table 2: TOTAL INVESTMENTS IN PETROLEUM SECTOR (In million US$) Year Exploration Develop Other Total SONANCOL % 1980 49.1 49.5 103.0 201.6 76.4 37.9 1981 89.0 64.6 172.7 326.3 108.9 33.4 1982 200.7 311.6 (46.5) 465.8 132.7 28.5 1983 218.8 187.7 3.2 A09.7 91.2 22.3 1984 92.3 237.9 34.6 364.8 86.5 23.7 1985 186.7 292.4 43.4 522.5 100.2 19.2 1986 104.2 329.0 - 433.2 62.7 14.5 Total 940.8 1,472.7 310.4 2,723.9 658.6 24.2 Source: SONANGOL. Table 3: TOTAL INVESTMENTS BY AREA - 1980-86 (In million US$) Exploration Development Other TotaL Cabinda Offshore 49.0 653.4 114.0 816.4 30.0 Onshore Angola 35.6 134.8 71.6 242.0 8.9 of which : A 11.8 34.2 6.9 52.9 B 23.8 100.6 64.7 189.1 Block 1 186.2 - 29.9 216.1 7.9 Block 2 195.9 270.4 26.7 493.0 18.1 Block 3 402.0 413.9 62.5 878.4 32.2 Block 4 72.1 - 5.9 78.0 2.9 Total 940.8 1,472.5 310.6 2,723.9 100.0 Source: SONANCOL - 146 - Annex 4 Page 3 of 6 Table 4: CRUDE OIL PRODUCTION BY BASIN (In '000s bbl/d) 1980 1981 1982 1983 1984 1985 1986 Cabinda 90 86 80 130 159 165 190 Kwanza 3 4 5 5 4 4 3 Congo 42 40 45 43 41 63 89 Block 2 2 5 13 12 9 7 6 Block 3 - - - - - 24 50 Onshore A 2 1 1 1 1 1 1 Onshore B 38 34 31 30 31 31 32 Total 134 130 130 179 205 232 282 Source: SONANGOL. Table 5: CRUDE OIL OPERATING COSTS a/ 1982 1983 1984 1985 1986 Total (in US$ million) Block 2 8.4 7.8 6.5 4.6 2.1 Block 3 - - - 13.036.4 Onshore A 11.1 10.8 8.1 8.9 9.8 Onshore B 20.8 24.6 18.0 17.3 14.4 Cabinda 209.0 229.6 124.6 113.6 117.1 Per barrel (US$) Block 2 1.8 1.8 1.9 1.8 0.9 Block 3 - - - 1.5 2M0 Onshore A 5.3 4.8 4.1 4.9 6.5 Onshore B 1.8 2.2 1.6 1.5 1.2 Cabinda 7.1 4.8 2.2 1.9 1.7 a/ Not including expenses for exploration and development. Source: SONANGOL. - 147 - Annex 4 Page 4 of 6 Table 6: SUMMARY OF MAIN OIL DISCOVERIES IN ANGOLA Name Discovery Bbl/d average Cumulative million bbl first 6 months to 01-JUL-86 Offshore Cabinda Takula 1982 107,350 91.3 Malongo West 1967 16,900 220.6 Malongo North 1966 14,500 203.0 Kungulo 1975 13,150 37.2 Limba 1969 12,400 79.3 Vuko 1983 10,500 2.0 Malongo South 1966 8,200 67.5 Kambala 1971 7,800 13.9 Others 2,090 6.4 TOTAL 192,890 721.1 Onshore N'Zombo 1973 13,000 81.1 Quinguila 1972 7,500 28.6 Quinfuquena 1975 5,100 21.7 Quenguela 1968 2,500 36.0 Ganda 1975 2,500 4.3 Lumueno 1977 2,300 2.2 Pambo 1982 850 1.6 Luango 1977 800 1.6 Other 1,334 48.0 TOTAL 35,884 224.8 Bbl/d average Cumulative million 1986 bbl to 31-DEC-86 Block 2 Essungo 1975 5,930 17.7 Cuntla 1978 170 2.2 Total 6,100 19.9 Block 3 Palanca 1981 39,900 23.2 Pacaca 1982 9,700 3.5 Total 49,600 26.7 Source: SONANGOL and "Oil and Gas Journal". - 148 - Annex 4 Page 5 of 6 Table 7: NATURAL GAS RESERVES AND PRODUCTION Associated Gas Production (MMCFD) 1984 1985 1986 1988 1990 Area (Forecast) (Estimate) a/ Cabinda 280 277 305 305 335-355 Congo 12 12 13 8.5 6-7 Kwanza 5 6 5.5 3 3 Block 2 15 15 11.5 41 41 Block 3 - 23 44 100 130 TOTAL 312 333 379 458 515-536 a/ Mission estimate. Source: SONANGOL. Probable and possible reserves of non-associated gas or gas associated with condensates in Blocks 2 and 3. Block 2 a/ Etele 600 BCF Sulele 210 BCF Garoupa 280 BCF Lua 320 BCF Polvo 530 BCF Block 3 b/ Punja 4,040-650 BCF a/ Plus Maleva North and Prata South b/ Plus Buffalo North (minimum, 2,340-2,590 BCF) Source: SONANGOL. - 149 - Annex 4 Page 6 of 6 Table 8: ASSOCIATED GAS UTILIZATION Associated Gas Utilized (MMCFD) 1984 1985 1986 1988 1990 Area (Forecast) (Estimate) a/ Cabinda 135 145 180 b/ 215 245-265 Congo 3.5 4.5 4 7 c/ 6-7 Kwanza 1 1 1 1 1 Block 2 0.5 0.5 0.5 31 31 Block 3 - 3 4.5 a/ 8a/ 70 d/ TOTAL 140 154 191 262 353-374 % Utilization 45 46 51 57 69-70 a/ Mission estimate. b/ Breakdown: 21 for own fuel, 11.7%; 67 for reinjection, 37.2%;, 92 for gas lift, 51.1%; and others. c/ Soyo Power Plant Project having started-up before that year. d/ Based on assumption that the reinjection scheme for Impala South East will have started up before 1990. Source: SONANGOL. - 150 - Annex 5 Page 1 of 2 THE PETROLEUM LAW On September 6, 1978 a law consisting of 30 Articles governing all petroleum activities in Angola was enacted. Petroleum Law 13/78 sets general principles rather than regulatory details and establishes that: - Ownership of hydrocarbons, underground, is vested in the State. - SONANGOL is the only company to whom exploration and production rights can be granted (existing concessions at the time of enactment of the Law were declared invalid). - The Council of Ministers has approval over future concession areas. - Exploration and production rights are granted for fixed terms on a case-by-case basis. - Mining activities are to be carried out in accordance with modern petroleum industry practices. - SONANGOL may enter into arrangements with financially and tech- nically capable oil companies to carry out exploration and production activities. - Petroleum agreements can take the form of joint ventures (asso- ciations between F"MANGUL and a foreign company with each part- ner having rights and obligations in proportion to their par- ticipating interests in the venture), production sharing agree- ments (service contracts between SONANGOL and foreign compa- nies), and commercial companies. In all cases there should be a joint management body in charge of operations, and SONANGOL will have a minimum of 51% participation except where the water depth in the contract area exceeds 150 meters. - Where no commercial discovery is made, companies will have no right to recover investments. In other cases, they will have the right to recover expenses and realize profits in accordance with their contractual terms. Beyond the wellhead, title to their share of the oil produced will pass on to the companies. - In exceptional cases, the Council of Ministers can authorize contractual terms different to those outlined above. - Disputes will be solved by arbitration, in Angola, in accord- ance with procedures agreed upon between SONANGOL and the foreign company. - 151 - Annex 5 Page 2 of 2 The Law is concise, yet general in nature, making it a flexible instrument in contract negotiations. A model production sharing agree- ment exists detailing the rights and obligations of each party, leaving only the main parameters, such as work program and financial terms, open for niegotiations. - 152 - Annex 6 Page 1 of 6 TAXATION OP THE PETROLEUM SECTOR (a) Investment Sharing Operations (Joint Ventures) The fiscal regime which applies to joint ventures between SONANGOL and foreign oil companies and thus mainly effects the conces- sions for Cabinda, was originally established by Decree No. 5/84 of March 28, 1984 and comprises: (a) a 20% tax on production ("royalty"); (b) a 65.7% tax on income; and (c) a 70% tax on "excess profits" (transaction tax). Taxes on production and petroleum transactions are deductible from the tax base for income taxation. In the case of the transaction tax, provisions are made for production and investment incentives which are deductible from net revenues (profits = gross reverues minus costs) as far as they are subject to taxation. Thus, in a simplified manner, Government's total tax income, denoted as "T", accruing from joint ventures, can be broken down into the following components: Production tax = tp px Transaction tax = tt (px cx - ix - zx) Income tax = r [(1 - t p) px - cx -, px - cx - ix - zx)] where: p = price of crude oil (US$/bbl); x = crude oil output (bbl); c = unit production costs (US$/bbl); i = production incentive (US$/bbl); z = investment incentive (US$/bbl); P = production tax rate (%); tt = transaction tax rate (%); and r - income tax rate (%). It is to be noted that Angola's total take is the sum of the taxes that accrue to the Government and SONANCOL's profits (net of taxes). - 153 - Annex 6 Page 2 of 6 Assuming that the production incentive (which, in practice, is adjusted to production costs) and the investment incentive (which is a fraction of total historic investment costs) are linear in "x", then the change in total tax revenues that is due to a marginal increase in prices and/or quantities can be calculated by applying the following formula: dT = x [(l - r) (tp + tt) + r] dp + p [t + r (l-t) - rc + t (1-r) (P c iZ )] d x p p p t p with p - (c 1 z) = 0 for (p-c-i-z) < 0 p Thus, the marginal change in tax income that results from a change in gross oil revenues can be split up into: (a) a price effect (dp >< 0, dx = 0), which is equal to a constant fraction of the price-dependent change in gross revenues (xdp); and (b) a quantity effect (dx >< 0, dp = 0), which is equal to a constant fraction of the quantity-dependent change in gross revenues (pdx). In particular, unless production costs are zero and "i" and "z" happen to be sufficiently small (relative to the price of oil) to be negligible, the price effect will be more pronounced than the quantity effect. In fact, at the prevailing tax rates one obtains: dT = 0.966 . x for dx = 0 dp Supposing that "i" = US$8/bbl; that "z" = US$1.50/bbl; and that ic" = US$5.50/bbl (which is roughly in line with the current cost and incentive structure of the Cabinda concessions) and taking into account a per barrel price of US$17 (which is the Government projection for 1988), then the quantity effect works out at follows: dT = 0.542. p for dp = 0 d-x Thus, if prices are declining, output must grow at a rate exceeding the rate of price erosion in order to keep the tax income at a given level. However, the higher the level of oil prices (relative to the level of costs), the more pronounced will be the quantity effect. For instance, at "p" = US$40/bbl, the quantity effect works out at 0.786.p. - 154 - Annex 6 Page 3 of 6 While marginal tax income is linear for output and prices, the average tax income tends to increase with oil revenues (progressive tax system). Let AT = T/px denote the average tax earnings, then d (AT) = [ [ (l-r) tt (c+i+z) + rc I p 2 1 dp Obviously, the average tax income is increasing at a decreasing rate, and the higher the level of production or the price level, the lower will be the increase in average tax income that is due to a price- dependent rise in (gross) oil revenues. For instance, in the case of the parameter configuration presented above, one obtains: d (AT) / dp = 0.025 with AT = 0.581 Thus, a doubling of the petroleum price, i.e., an increase from UJS17/bbl to US$34/bbl, will raise the Government's income share from gross revenues by 42.5%. In other words: Government's take will increase from a 58.1% share to an 82.8% share of total revenues. It can therefore be concluded that, in the case of ioint ventures the main objective of the taxation system is to capture windfall profits accruing from rising oil prices. On the other hand, if prices are decreasing or constant, the tax regime tends to protect the oil companies, i.e., lower prices result in a lower (dp < 0) or stagnant Government share (dp = 0) of income from oil revenues. (b) Production Sharing Operations Basically, .-he production-sharing type of agreement that has become fashionable in many countries during the last 10 years works as follows: (a) SONANGOL subcontracts--on behalf of the Government--the opera- tional services of a foreign oil company; (b) exploration is financed and carried out by the foreign company at its own risk, i.e., exploration costs will be reimbursed only if there is a commercial discovery. (A commercial well is defined as being capable of producing a minimum number of bbl/d for a given water depth, i.e., 1,000 bbl/d for a depth of less than 50 meters, 1,500 bbl/d for a depth between 50 and 100 metersp etc.); - 155 - Annex 6 Page 4 of 6 (c) in case of a commercial discovery, the foreign company finances and undertakes the development of the reserves as approved by the Government; (d) the subcontractor recovers his costs by taking a percentage share of total production ("cost oil"), while the remainder ("profit oil") is split between the foreign oil company and SONANGOL on a sliding scale basis linked to cumulative output; and (e) the foreign company's remuneration is taxed at special rates. The key issues to be negotiated are the work program (tasks and finance) and the production splits. In the past, the foreign companies' exploration commitments included a minimum number of wells (e.g., six) to be drilled within a predetermined period of time (e.g., three years). Any extension of the exploration period was subject to additional drilling obligations. Fields with commercial discoveries had to be developed within another three years. Thus, SONANGOL's interest in sub- contracting foreign companies is to achieve a maximum number of drillholes and discoveries within the shortest period of time, without taking any of the risk involved in the operations. As soon as production starts the contractor is allowed to recoup his expenses from total output. Cost oil may account for up to 50% of annual production and includes: (a) operating costs recovered on a recurrent basis; (b) administration expenses; (c) development costs recoverable over a period of four years with a 33.3% "uplift"; and (d) exploration costs recovered from the unused balance of cost oil. Excess recoverable costs are rolled over to the next year. What is left over after costs have been recovered (including the unused share of cost oil) constitutes the annual profit oil. Clear- ly, once all major expenses have been reimbursed, the share of profit oil tends to exceed the 5C% level. This situation generally applies to, say, the first five to six years of oil production. In Angola profit oil splits between SONANGOL and a foreign company are based on a field-specific scale of cumulative production rates. This scheme avoids the shortcomings involved in the more common approach of fixing production thresholds on a daily or monthly basis, thus providing a disincentive to increase the production rate when a - 156 - Annex 6 Page 5 of 6 lower share in profits is entailed. A representative example of Angola's sliding scale scheme is given in the following table. Table 1: SLIDING SCALE FOR PROFIT OIL SHARING Cumulative Production Profit Oil (X) (Million bbl) SONANGOL Contractor 0-25 40 60 25-50 70 30 50-100 80 20 >100 90 10 A unique feature of Angola's production sharing agreements is the "price cap" provision which acts as a 100% tax on "excess" profits, defined by the extent to which the price of oil exceeds a particular pre- established upper limit. Initially, the price limit was set at US$13/bbl (in 1978 prices) and adjustments were to be made in accordance with changes in the U.N. index for manufactured goods. Since, in the meantime, real oil prices have declined below the 1978 level, the limit has been increased to US$26/bbl. However, under the current conditions this ceiling does not affect the oil company's profits. If the profit sharing agreements (PSAs) were implemented in accordance with the terms set forth in the late 1970s, no tax would apply to foreign oil companies. However, in order to take advantage of U.S. tax credit laws TEXACO requested a change in the PSAs. Under the revised contractual setting, the foreign oil companies of Block 2 and 3 pay a 50% income tax to Angola on a nominal share that exceeds the initially agreed share in profit oil by 100%. The "grossing up" involved in this scheme is financed by SONANGOL out of its own share in profit oil. Basically, this accounting system does not change the agreed split in profit oil, i.e., Angola's take from the profit oil is invariably equal to SONANGOL's share. However, it affects the Government's tax income from PSAs. If P is the profit oil, "s" the foreign company's share in profit oil with 1>5>0, and "r" the income tax rate, then Government tax revenues from the grossing-up scheme are determined by: T = (1-r - 1) s P + r [ s) P) - (l-r) - 1) s PI = r P - 157 - Annex 6 Page 6 of 6 It there were no income tax on the foreign companies' profit oil and, thus, no grossing-up financed by SONANGOL, the tax income would amount to T = r (l-s) P Thus, under the current tax-contract system, Government captures 50% of the total profit oil, while in the absence of the grossing-up scheme, Government would only receive 50% of SONANGOL's share in profit oil. In the meantime, however, the liberalization of U.S. tax laws has erased the double taxation problem so that foreign oil companies have become less interested in the option of transferring tax payments to Angola. Moreower, SONANGOL fears that it may no longer be in the position to finance the grossing-up of the foreign oil companies' share in profit oil as soon as cost oil accounts for a significantly lower share in total production than during the initial years of operation. Therefore, SONANCOL has suggested exempting foreign oil companies operating under PSAs from the Angolan income tax. Not surprisingly, Government is not very happy with the proposal. In particular, the Ministry of Finance feels that the income tax on profit oil should even be increased to 65%, a rate that would be in line with the general tax laws applying to the industrial sector (as well as to joint ventures with foreign oil companies). In fact, the basic problem is that the very logic of PSAs does not fit well with the Government's current very pressing revenue needs. While PSAs provide an incentive framework with a progressive system of sharing that works best if the investment expenditure of foreign oil companies are reimbursed as fast as possible, Government would prefer that the PSAs work as if the initial risk in exploring and devel-,4ng the oil reserves were shared as in joint ventures. Moreover, PSAs require a "strong" treasury in the host country. In the case of Angola, however, the treasury is in a comparatively weak position, struggLing with fiscal legislation that is not adjusted to a modern tax-contract system like the PSAs, and relying exclusively on the supervisory function that SONANGOL has to assume towards foreign oil companies. This dependency on SONAdGOL, coupled with the fact that Angola's take of the oil revenues from PSAs is not yet readily available and, in addition, has to be shared between the Government and SONANGOL, creates a potential for conflict. - 158- 4nnex 7 Page 1 of 4 FINANCIAL ANALYSIS OF THE AMMONIA/URA PLANT Assumptions A. Limited absorptive capacity of the national market. In recent years not more than 10,000 t/y of nitrogen fertilizers have been consumed in Angola. B. Prices. Angolan border prices for ammonia are assumed to develop as follows: Year 1995 Year 2000 Ammonia FOB Angola 200 215 Urea FOB Angola 200 210 Source: Mission estimates. C. The economic value of gas consists of two components, the production cost and the depletion value for gas in the field. The production cost is the cost of extracting the gas and shipping it to the plant gate. In the present context, the depletion value is defined as the discounted cost of using an alternative source of energy et some future date when the currently available reserves will be depleted. The depletion value is calculated by assuming that the gas production costs of the "backstopping" fields amount to US$2.50/MMBTU. The considered range for current production cost has been allowed to vary between US$1.25 and US$2.00/MMBTU. The domestic market could possibly absorb not more than the equivalent of the production of a 100-200 t/d ammonia plant, even taking a confirmed recovery of agriculture into account. However, economies of scale only justify setting up a plant with the minimum capacity of 1,000 t/d of ammonia. The analysis is therefore based on a layout with a nominal capacity of 1,500 t/d of ammonia and 500 t/d of urea. The project will require an investment of a minimum of US$330 million (at 1987 prices) for the basic equipment and all surrounding facilities, excluding working capital requirements. At best the plant could operate at a 90% capacity utilization level (329 full operating days per year) which, by international standards, is only reached in a few very well operated plants. A maximum capacity utilization of 80% would be more likely. Producing 500 t/d of urea is equal to a yearly capacity of 182,500 tons of urea. Four hundred - 159 - Annex 7 Page 2 of 4 and forty thousand tons of excess ammonia is obtained at this production level. The Economic Rate of Return (ERR) for the project has been calculated in two ways, i.e., excluding or including the depletion value of gas. As it turns out, the depletion value has only a minor effect otl the project's viability. Therefore, the sensitivity analysis can roughly be based on figures which exclude the depletion value of gas. Assuming a low gas production cost of USS1.25/MMBTU, then the economic rate of return on investment works out as follows: 12.72% ERR in the Base Case (no depletion costs) 8.49% ERR in Case C (lower annual capacity utilization and lower product prices), which is the least favorable alternative to the Base Case. Taking into account the scarcity of capital in Angola, the target discount rate should be set at 15% or higher. The effect of lower capacity utilization on the ERR is greater than the effect of lower world market prices. Taken together the impact on the ERR is considerable. The sensitivity analysis on product prices and/or utilization of capacity shows verv low or negative economic values for gas at the wellhead (gas netback value). In particular, at discount rates of 10% and more the netback values faLl below the production cost of gas. Taking into consideration the low economic returns of the project, the probability of low capacity utilization, and the uncertainties involved in the future of product prices, the project cannot be considezed economically viable. The following cases have been analyzed: Base Case Maximum capacity utilization at 90% (equivalent to 329 full operating days/year), with or without depletion costs. Case A As Base Case with maximum capacity utilization at 80% (equivalent to 292 full operating days/year). Case B As Base Case with product prices decreased by US$15/t. Case C: Combination of Cases A and B. Table 1: AMMONIA/UREA PROJECT Financial Analysis Ist Year: 1988 Lifetime: 20 years Start up: 1992 Project capacity: 182,500 tons of urea, 440,000 tons of ammonia Year Invest- Oper. UREA AMMONIA UREA AMMONIA Total Fxd.Cost Var.Cost Net income Gas Cons ment rate Output Output Price Price Sales 22 5 a/ MM US$ (%) '000s t 'OOOs t US$/t USS/t MM US$ MM USs$ MM US$ mm UJS BCF 1988 33 0% 0.0 0.0 0.00 0.00 0.00 -33.00 0.00 1989 95 0% 0.0 0.0 0.00 0.00 0.00 -95.00 0.00 1990 123 0% 0.0 0.0 0.00 0.00 0.00 -123.00 0.00 1991 139 0% 0.0 0.0 0.00 0.00 0.00 -139.00 0.00 1992 19 65% 118.6 286.0 188 188 76.07 22.00 3.25 31.82 12.03 1993 0 85% 155.1 374.0 192 192 101,59 22.00 4.25 75.34 15.73 1994 0 90% 164.3 396.0 196 196 109.81 22.00 4.50 83.31 16.65 0 1995 0 90% 164.3 396.0 200 200 112.05 22.00 4.50 85.55 16.65 1 1996 0 90% 164.3 396.0 202 203 113.57 22.00 4.50 87.07 16.65 1997 0 90% 164.3 396.0 204 206 115.08 22.00 '.50 88.58 16.65 1998 0 90% 164.3 396.0 206 209 116.60 22.00 4.50 90.10 16.65 1999 0 90% 164.3 396.0 208 212 118.12 22.00 4.50 91.62 16.65 2000 0 90% 164.3 396.0 210 215 119.63 22.00 4.50 93.13 16.65 2001 0 90% 164.3 396.0 211 217 120.59 22.00 4.50 94.09 16.65 2002 0 90% 164.3 396.0 212 219 121.55 22.00 4.50 95.05 16.65 2003 0 90% 164.3 396.0 213 221 122.50 22.00 4.50 96.00 16.65 2004 0 90% 164.3 396.0 214 223 123.46 22.00 4.50 96.96 16.65 2005 0 90% 164.3 396.0 215 225 124.41 22.00 4.50 97.91 16.65 2006 0 90% 164.3 396.0 215 225 124.41 22.00 4.50 97.91 16.65 2007 0 90% 164.3 396.0 215 225 124.41 22.00 4.50 97.91 16.65 2008 0 90% 164.3 396.0 215 225 124.41 22.00 4.50 97.91 16.65 ,. 2009 0 90% 164.3 396.0 215 225 124.41 22.00 4.50 97.91 16.65 0 2010 0 90% 164.3 396.0 215 225 124.41 22.00 4.50 97.91 16.65 M X 2011 -39 90% 164.3 396.0 215 225 124.41 22.00 4.50 136.91 16.65 4 0 a/ Not including the economic costs of gas. - 161 - Annex 7 Page 4 of 4 Table 2: ECONOMIC RATE OF RETURN At gas Base Base Case A b/ Case B b/ Case C b/ prod. cost Case I a/ Case II b/ Luwer Lower Lower capacity (US$/MMBTU) capacity prices and prices 1.25 11.52% 12.72% 10.09% 11.09% 8.49% 1.50 10.86% 11.92% 9.21% 10.24% 7.55% 1.75 10.22% 11.10% 8.30% 9.37% 6.58% 2.00 9.61% 10.26% 7.35% 8.46% 5.55% a/ Including gross depletion costs. b/ Excluding gross depletion costs. Source: Mission calculations. Gas netback value (in MM USS) …------------------- Discount Races ------------------ 5% 10Z 12% 15% 20% Base Case 3.30 2.07 1.48 0.50 -1.32 Case A 2.58 1.27 0.68 -0.30 -2.12 Case B 2.88 1.52 0.97 0.00 -1.83 Case C 2.13 0.83 0.23 -0.74 -2.56 Source: Mission calculations. - 162 - Annex 8 Page 1 of 10 PIcURzs o0 THE PMTROLUMQ PRODUCT SUBSECTOR Tabie 1: ANGOLA PETROLEUM SUPPLY - DISPOSITION BALANCE - 1980-86 (In 'OOOs tons) 1980 1981 1982 1983 1984 1985 1986 PROOUCTION Crude 6,710.0 6,470.0 6,520.8 8,949.6 10,228.1 11,599.6 14,101.5 LPG 0.0 0.0 0.0 97.8 167.7 165.8 177.0 IMPORTS LPG 6.9 6.0 9.2 6.0 6.8 11.7 10.9 Jet fuel 0.0 7.8 39.8 24.9 62.9 102.6 114.2 Gasoil 0.0 0.0 0.0 0.0 23.7 36.7 16.4 Total Products 6.9 13.8 49.0 30.9 93.4 151.0 141.5 REFINING Crude -1,237,5 -1,240.7 -1,014.3 -i,300,6 -1,380.5 -1,451.3 -1,452.2 LPG 11,7 12.8 13.0 17.3 17.7 18.3 18.7 Gasoline 89.6 85.2 84.4 92,3 103.6 103.5 104.7 Naphtha 0.0 0.0 0.0 4.6 4.0 9.9 21,5 Kerosene 33.0 34.8 31.6 32.1 39.7 45.6 35.6 Jet fuel 118.6 127.0 99.5 143.5 161.9 172.8 171.1 Gasol I 294.7 308.6 243.4 316.9 344.6 353.7 347,3 Fuel oil 609.4 604.8 479.2 606.3 636.8 679.4 685.8 Asphalt 7.4 7.1 8.0 4.2 5.0 7,9 7.3 Total Products 1,164.4 1,180.3 959.1 1,217.2 1,313.3 1,391.1 1,391.8 fuol & loss 73.1 60.4 55.3 83,4 67.2 60.2 60,3 (in S) 5.9% 4.9% 5.4% 6.4% 4.9% 4.2% 4.2% TOTAL SUPPLY Crude 5,472.5 5,229.3 5,506.5 7,649.0 8,847.5 10,'48.3 12,649.3 LPG 18,7 18,8 22,1 121,1 192.2 195.8 206.6 Gasoline 89.6 85.2 84.4 92.3 103,6 103.5 104,7 Naphtha 0.0 0.0 0.0 4.6 4.0 9.' 21.5 Kerosene 33.0 34.8 31.6 32.1 39.7 45.6 35.6 jet fuel 118.6 134.7 139.3 8a.5 224.8 275.4 285,2 Gasoil 294,7 308,6 243.4 316.9 368.3 390.4 363.7 Fuel oil 609.4 604.8 479.2 606.3 636.8 679.4 685.8 Asphalt 7.4 7.1 8.0 4.2 5.0 7.9 7.3 Total Products 1,171.3 1,194.0 1,008.0 1,345.9 1,574.4 1,708,0 1,710,4 FIELD USE Crude 25.6 26.4 36,9 43.1 57.8 51.1 45,5 CARGO EXPORTS Crude 5,576.0 5,224.9 5,378.8 7,514.2 8,789.7 10,057.3 12,637.1 LPG 0.0 0.0 0.0 97.8 167.7 173.8 166.8 Gasoline 2.6 5.4 7.2 16.8 8.2 3,2 2.9 Naphtha 0.0 0.0 0.0 4.0 0.0 5.1 8.6 Kerosene 2.4 2.1 1.9 0.0 0.0 0.0 0.0 Jet Fuel 0.0 0.0 1.0 1.8 3.5 3,0 0.0 Gasoll 23,8 7.8 4.5 5.6 9.7 7.3 5.6 Fuel oil 481.6 495.6 346.2 496.7 511.3 585.9 528.8 Total Products 510.4 510.9 360,8 622.6 700.5 778.3 712.7 - 163 - Annex 8 Page 2 of 10 Tabie 1: ANGOLA PETROLEUM SUPPLY - DISPOSITION BALANCE ^ 1980-86 (In 'OOOs tons) (Continued) 1980 1981 1982 1983 1984 1985 1986 BUNKER EXPORTS Jet Fuel 0.0 29,1 32.5 22.3 .1.7 19.8 28.9 Gasoil 0.0 37.8 35.0 38.5 11.5 11.6 5.9 Fuel oil 0.0 2.8 4.0 21.3 4.1 3.4 2.6 Total Products 0.0 69,6 71.5 82.1 37.3 34.8 37.4 INLAND SALES LPG 18.7 18.8 22.1 23.3 26.5 28.7 30.2 Gasoline 81.0 82.9 88.9 89.5 95.1 96.0 104.1 Kerosene 29.8 32.4 36.9 39.2 40.6 43.9 46.3 Jet fuel 114,4 105.7 106.0 144.4 184.5 235.6 260.6 Gasoil 268.8 300.4 323.5 334.6 340.8 376.9 344.8 Fuel oil 121.3 120.5 109.6 83.8 88.2 104.6 119.2 Asphalt 3.2 4.5 2.6 0.4 1.7 5.6 4.2 Total Products 637.1 665.2 689.7 715.2 777.5 891.2 909.3 TOTAL DISPOSITION Crude 5,601.7 5,251.3 5,415.7 7,557.3 8,847.6 '0,108.5 12,682.6 LPG 18.7 18.8 22.1 121.1 194.1 202.5 197.0 Gasoline 83.6 88.3 96.1 106.2 103.3 99.2 107.0 Naphtha 0.0 0.0 0,0 4.0 0.0 5.1 8.6 Kerosene 32.2 34.5 38.8 39.2 40.6 43.9 46.3 Jet fuel 114.4 134.7 139.5 168.5 209.7 258.3 289.5 Gasoil 292,6 308.3 328.1 340.2 350.5 384.2 356.3 Fuel oil 602.9 Gt6.1 455.8 580.5 599.6 690.5 650.6 Asphalt 3.2 4.5 2.6 0.4 1.7 5.6 4.2 Total Products 1,147.5 1,205.2 1,083.0 1,360.0 1,499.6 1,689.4 1,659.5 STOCK DRAW Crude 129.2 22.0 -90.8 -91.7 0.0 -39.8 33.2 (BUILD) or Other Residuai LPG 0.0 0.0 0.0 0.0 1.9 6.7 -9.6 Imbalance Gasoline -6.0 3.1 11.7 13.9 -0.2 -4.3 2.3 Naphtha 0.0 0.0 0.0 -0.6 -4.0 -4.8 -12.8 Kerosene -0.8 -0.3 7.2 7.2 0.9 -1.7 10.7 Jet fuel -4.2 0.0 0.2 0.0 -15.1 -17.1 4.2 Gasoil -2.1 -0.3 84.7 23.3 -17.8 -6,2 -7.4 Fuel oil -6.5 11.3 -23.4 -25.8 -37,2 11.1 -35.2 Asphalt -4.2 -2.6 -5.3 -3.8 -3.3 -2.3 -3.1 Total Products -23.8 11.1 75.0 14.1 -74.8 -18.6 -50.9 Source: SONANGOL and mission calculations. - 164 - Annex 8 Page 3 of 10 SUMMARY OF PHYSICAL FACILITIES LUANDA REFINERY FINA PETROLEOS DE ANGOLA Nominal Capacity 1.7 Million tons/year Actual Capacity 1.6 Million tons/year (a) Process Units Capacity (bbl/cd) 1/ 3 Topping plants 32,000 Vacuum distillation unit 1,900 Reformer 1,900 Merox 1,300 Naphtha Hydroheater 3,800 Kerosene Hydrodesurgurizer 2,800 Gas Recovery Unit 500 (b) Off-sites (i) Electric Power: 12.5 MW gas turbine generator 2/ 0.8 MW diesel drive generator (back-up) 3. (ii) Tankage 3/ Crude oil - 70,460 m Slops - 2,900 m3; LPG - 1,184 m'; Naphtha - 7,918 m3 Gasoline - 15,000 m3; Jet Fuel - 33,048 m3; Kerosene - 1,700 m' 3. Gasoil - 41,915 mi Fuel Oil - 95,616 m3; Asphalt - 2,543 m3 Total: 272,284 m3 1/ Barrels per calendar day. 2/ The recently installed turbine naphtha-fired generator is normally not used since it is more economic for the refinery to draw hydro power from the grid to serve its total load of about 3 MW. Occasionally the turbine is run )n request of the Government in order to supply the grid. 3/ The tankage capacity is now back, to normal levels after the major sabotage of 1981 when 45,000 m3 of capacity were lost. - 165 - Annex 8 Page 4 of 10 (iii) Receiving Facilities: 4/ Oil Port - 44 feet draft; Maximum vessel size - 75,000 DWT. (iv) Product Pipelines No. Services 1 LPG 1 Fuel oil 1 Clean products 4/ The oil port adjacent to the refinery is owned and operated by FPA. It has a 44 foot draft and can accomodate vessels of up to 75,000 DWT. All the Soyo crude is received through this facility and finished products are shipped to SONANCOL marine terminals and to export destinations from here. A significant number of the vessels bunkering in Luanda are handled by FPA at its facilities in the oil port. - 166 - Annex 8 Page 5 of 10 Table 2: FPA LUANDA REFINERY PRODUCTION BALANCE - 1980-86 ('000 tons) 1980 1981 1982 1983 1984 1985 1986 Refinery Input Crude oil 1,237.5 1,240.7 1,014.3 1,300.6 1,380.5 1,451.3 1,452.2 Refinery Output LPG 11.7 12.8 13.0 17.3 17.7 18.3 18.7 Gasoline 89.6 85.2 84.4 92.3 103.6 103.5 104.7 Naphtha 0.0 0.0 0.0 4.6 4.0 9.9 21.5 Kerosene 33.0 34.8 31.6 32.1 39.7 45.6 35.6 Jet fuel 118.6 127.0 99.5 143.5 161.9 172.8 171.1 Gasoil 294^7 308.6 243.4 316.9 344.6 353.7 347.3 Fuel oil 609.4 604.8 479.2 606.3 636.8 679.4 685.8 Asphalt 7.4 7.1 8.0 4.2 5.0 7.9 7.3 Rerun slops 5,4 4,5 4.1 6.2 5.0 4.4 4.4 Total ex-loss 1,169.8 1,184.7 963.2 1,223.4 1,318.3 1,395.5 1,396.3 fuel & loss 73,1 60.4 55.3 83.4 67.2 60.2 60.3 TOTAL 1,242.9 1,245.2 1,018.4 1,306.8 1,385.5 1,455.8 1,456.6 Note: Annual growth rate of total output: 4% (least square estimate). Source: FPA. - 167 - Annex 8 Page 6 ol 10 Table 3: ANGOLA - PRODUCT STORAGE TERMINALS (All owned by SONANGOL) Location/Name Capacity Original Owner/- m3 % of Constructor Total Coastal Luanda IBV-1 26,000 SACOR Luanda IBV-5 51,000 MOBIL Luanda IMUL 5,990 SHELL Luanda TEMAR 22,300 SHELL Subtotal 105,290 54% Namibe 30,000 TEXACO Lobito 18,900 FINA Porto Amboim 10,750 SONANGOL (USSR-built) Tombua 1,200 FINA Cabinda 900 SONANGOL Soyo 435 FINA Total Coastal 167,475 85% Up-Country Huambo II 10,120 SONANGOL (Romanian) Huambo I 1,360 FINA Malange III 6,600 SONANGOL (USSR-tuilt) Malange II 324 FfNA Malarnge I 330 SACOR Jamba 3,000 FINA Matala 2,500 FINA Tchamutete 1,200 FINA Kuito 800 SONANGOL Lubango 450 FINA Lubango 206 SACOR Lwena 370 SACOR Lwena 200 MOBIL Kaala 250 MOBIL Lucala 250 MOBIL Cubal 125 FINA Cubal 95 MOBIL Menongue 160 TEXACO Dondo 100 MOBIL Ganda 100 FINA Total Up-Country 28,540 15% TOTAL ANGOLA 196,015 100% Source: SONANGOLO - 168 - Annex 8 Page 7 of 10 Table 4: ANGOLA - INlAND PETROLEUM PRODUCT CONSUMPTION (SA'ES) - 1980-86 (Tons) Annual 1900 1981 1982 1983 1984 1985 1986 growth rate 80-86 a/ b/ LPG (Butane) 18,688 18,822 22,138 23,288 26,459 28,695 30,200 9.2% Gasoline - Motor 79,567 81,794 87,814 88,722 94,252 95,004 103,680 4.3% Aviation 1,391 1,107 1,128 744 866 987 430 -13.4% Total Gasoline 80,958 82,901 88,942 89,466 95,118 95,991 104,110 4.1% Kerosene 29,835 32,412 36,910 39,229 40,625 43,908 46,310 7.5% Jet fuel 114,350 105,650 105,965 i44,375 184,522 235,567 260,550 18.0% Totai Kerosene/Jet fuel 144,185 138,062 142,875 183,604 225,147 279,475 306,860 15.9% Gasoil 268,806 300,438 323,542 334,588 340,799 376,865 344,770 4.6% Fuel oil 121,288 120,549 109,601 83,814 88,225 104,622 119,200 -2.0% Asphalt 3L171 4,455 2,648 426 1,708 5,593 4,2C) 3.1% TOTAL All Products 637,096 665,227 689,746 715,186 777,456 891,241 909,340 6.5% a/ 1986 estimate based on 9-months' actuals. bl Least square estimates. Source: SONANGOL. - 169 - Annex 8 Page 8 of 10 Table 5: ANGOLA - PETROLEUM PRODUCT CONSUMPTION (SALES) Sectoral Breakdown, 1985 LPG Gasoline Kerosene Jet fuel Gasoil Fuel oil Total % Industry 2,217 1,608 550 24,257 98,556 54,645 181,833 20.6X Agriculture 86 484 408 0 11,578 7,115 19,670 2.2% Transport 92 755 81 111,026 102,134 41,119 255,206 28.8% Construction 84 600 38 0 17,937 140 18,800 2.1% Resale 16,387 40,112 39,687 4 77,061 0 173,251 19.6% Government 642 2,417 563 198 14,911 423 19,154 2.2% Armed Forces 1,990 47,271 268 98,903 37,963 284 186,678 21.1% Other 7,197 1,757 2,314 1,179 16,726 895 30,068 3.4% Total 28,695 95,004 43,908 235,567 376,865 104,622 884,661 100.0% Source: SONANGOL. - 170 - Annex 8 Paae 9 of 10 Table 6: SONANGOL - SALES THRO#JGH RESELLERS BY PROVINCE (In cubic meters) Ist 9 months of 1984 1985 1986 Gasolino Gasoil Kerosene Gasoline Gasoll Kerosene Gasoline Gasoll Kerosene Coastal Provinces Luanda 36,405 37,258 20,017 40,836 51,041 23,550 37,025 44,925 19,827 Other Coastal Cabinda 814 931 110 396 719 152 649 607 265 Zaire 508 542 1,047 291 632 1,011 372 604 952 Bongo 69 244 227 34 198 287 7 95 1,359 Kwanza South 884 3,264 1,576 730 2,8(03 1,171 703 2,449 811 Benguela 6,286 13,866 9,527 6,983 16,685 12,523 4,466 10,848 8,501 Namibe 493 2,287 11.9 681 2,992 2,117 536 212 1,769 Subtotal Other 9,054 21,134 14,483 9,11' 24,029 17,261 6,733 16,715 13,657 Total Coastal 45,459 58,392 34,500 49,951 75,070 40,811 43,758 61,640 33,484 Up-Country Provinces Ulge 299 1,052 2,495 124 649 1,520 50 1,775 1,030 Kwanza North 1,279 5,598 3,042 906 4,537 2,632 490 2,695 1,479 Malange 697 1,539 563 194 366 265 25 132 38 Huambo 318 168 624 5)7 544 497 435 844 257 Ble 43 62 104 141 185 153 21 90 18 Huila 3,288 7,071 3,037 3,457 7,994 3,491 2,354 5,222 3,011 Cuneno 20 24 0 9 0 25 0 53 15 Lunda North 0 0 0 0 0 0 0 0 0 Lunda South 0 0 0 0 0 0 0 0 0 Moxico 0 0 0 0 0 0 0 0 0 Kuando Kubango _ 0 0 0 0 0 0 0 0 Total Up-Country 5,944 15,514 9,865 5,338 14,275 8,583 3,375 10,811 5,848 TOTAL Angola 51,403 73,906 44,365 55,289 89,345 49,394 47,133 72,451 39,332 AS PERCENTAGES OF TOTAL ANGOLA 1984 1985 1986 Gasoline Gasoil Kerosene Gasollne Gasoll Kerosene Gasoline Gasoil Kerosene Luanda 70.8 50.4 45.1 73.9 57.1 47.7 78.6 62.0 50.4 Other Coastal 17.6 28.6 32.6 16.5 26.9 34,9 14.3 23.1 34.7 Up-Country 11.6 21.0 22.2 9.7 16.0 17.4 7.2 14.9 14.9 Source: SONANGOL. - 171 - Annex 8 Page 10 of 10 Table 7: PROJECTED ANGOLA PETROLEUM PRODUCTS CONSUMPTiON (In tons) Actual Est.ann. -------------------Forecast------------- Assumed 1986 growti 1987 1988 1990 1992 per annum a/ 1980-86 1987-92 LPG (Butane) 30,200 9.2% 31,408 32,664 35,330 38,213 4.0% Gasoline - Motor 103,680 4.3% 107,827 112,140 121,291 131,188 4.0% Aviation 430 -13.4% 430 43u 430 430 0.0% Total Gasoline 104,110 4.1% 108,257 112,570 121,721 131,618 4.0% Kerosene 46,310 7.5% 49,089 52,034 58,465 65,692 6,0% Jet fuel 260,550 18.0% 266,193 272,005 284j158 2170,051 2.2% Total Kerosene/Jet fuei306,860 15.9% 315,282 324,039 342,623 362,743 2.8% Gasoil 344,770 4.6% 357,874 371,481 400,284 431,346 3.8% Fuel oil 119,200 -2.0% 122,984 126,896 135,120 143,912 3.2% Asphalt 4,200 3.1% 4,284 4,370 4,546 4,730 2.0% Total All Products 909,340 6.5% 940,089 972,020 1,039,624 1,112,562 3.4% a/ Estimated from 9-months' actuals. Source: SONANGOL Sales Department. - 172 - Annex 9 Page 1 of 4 STRUCTURE OF PETROLEUM PRODUCT PRICES The final price of petroleum products to consumers in Angola is affected by price controls at three different transactional levels within the downstream sector: (a) Crude Oil to the Luanda Refinery (b) Products to the distributor (SONANGOL) (c) Products to the final consumer Pricing of Crude Oil to The Luanda Refinery The prices paid by the refinery for crude oil are established by a "Protocol" agreed upon between the Government and the refinery owner, Fina Petroleos de Angola (FPA). This document guarantees that the crude owners will receive as much revenue from a refinery sale as from an export sale of the same crude grade. The crude owners do not pay royalty on crude sold to the refinery. The royalty is deducted from the production income tax paid by crude owners. As the Government waives the royalty income on crude sales to the refinery, the refinery is paying less than economic opportunity cost for the crude. At present the crude oil feed for the refinery comes from tqe following FPA-operated areas: - Onshore Kwanza - FINA/SONANGOL joint venture - Onshore Congo (Soyo Crude) two areas: (a) FINA/SONANGOL joint ventures (b) FINA/TEXACO/SONANGOL joint venture. In principle, the refinery crude price formula as defined by the "Protocol" is as follows 1/: PREF = PEXP x (l-r) + t x r = PEXP - (PEXP-t)r Where, PREF = price per barrel of crude to the refinery 1/ At a fixed transport cost factor of US$1.50 the formula simplifies to: PREF = PEXP x 0.8333 + 0.25 - 173 - Annex 9 Page 2 of 4 PEXP = export price per barrel of crude t = transport cost factor allowable under royalty calculations, presently US$1.5/bbl for Congo and Kwanza crudes. r = royalty factor (0.1667). Thus, the refinery price is equal to: the export equivalent price of Congo/Kwanza crude less the royalties, adjusted fo: tha cost of transporting the crude to the refinery. Currently the export equivalent price of Congo crude is set at $US2.00/bbl off .ne market price of "Bonny Light", while Kwanza crude is priced for export at 94.32% of Congo crude. For instance, assuming that the market price of "Bonny Light" amounts to US$20.00/bbl, then the export equivalent prices work out at $USl8.00/bbl (Congo) and US$16.98/bbl (Kwanza), respectively. Thus, with r = 0.167 and t = US$1.50/bbl the refinery prices are US$15.25/bbl for Congo crude and US$14.40/bbl for Kwanza crude, involving foregone tax revenues of US$2.75/bbl in the case of Congo crude and US$2.58/bbl for Kwanza crude. Pricing of Products to the Distributor The refinery gate prices of finished products for sale from FPA to SONANGOL are fixed by Government Decree No. 18/86 of November 1985. The prices were based on complete refinery cost recovery including a 10% return on investment. As opposed to the price of crude to the refinery, these prices are not adjusted to changing crude oil prices or operating costs. The return on investment is fixed to 10% of refinery assets. Refinery deficits are covered by subsidies from the State budget, and surpluses are repaid to the State budget. Every year a balance is made based on verified operating costs for the previous year and any outstanding imbalance is settled. The pricing system is a pure "cost-plus" arrangement, providing no incentives to the refinery to minimize operating costs, as any savings would revert to the Ministry of Finance. In recent years lower crude prices and higher production and sales volumes have increased the payments from the refinery to the Ministry of Finance. During 1986 approximately 2 billion Kz was returned to the Ministry of Finance out of a total refinery gross of 7.5 billion Kz for the year. The prices of products are set mainly to attain social goals rather than to perform allocative functions. However, since prices on the world market have declined since 1985 official refinery gate prices are roughly on par with their import parity values, as shown in - 174 - Annex 9 Page 3 of 4 Table 1. In the table, the current price structure for the .najor products is compared with a CIF cost arrived at on the basis of a hypothetical freight and relate4 charges figure of US$20.00/t added to FOB Mediterranean spot. For LPG a hypothetical freight charge of US$80/t was used in addition to the FOB price. Table 1: OFFICIAL REFINERY GATE PRICES VS INTERNATIONAL PRICES -- Official Structure -- FOB Mediterranean plus US$20/t a/ US$/t 1985 1986 Ist H 1987 LFG 7.90 Kz/kg 266.54 304,58 204.75 223.67 Gasoline 5.00 Kz/litre 228.06 275.50 161.25 186.50 Kerosene/Jet fuel 5.02 Kz/litre 209.40 281.75 172.58 179.33 Gasoil 3.55 Kz/litre 140.95 255.00 151.92 171.50 Fuel oil (Heavy) 2.66 Kz/kg 89.80 167.25 87.42 121.00 Official Price as percent of International LPG 88% 130% 119% Gasoline 83% 141% 122% Kerosene/Jet fuel 74% 121% 117% Gasoil 55% 931 82% Fuel oil (Heavy) 54% 103% 74% a/ Except LPG for which US$80/t is added to FOB Mediterranean. Source: Mission estimates. Pricing of Products to Final Consumer The Decree which stipulates the price structure of products to the distributor also sets the structure of prices to the final con- sumer. Some minor adjustments have been made since the Decree came into effect in 1985. SONANGOL is the only distributor of petroleum products and operates the distribution on a "cost-plus" basis. The price structure allows SONANGOL a 10% profit margin plus a distribution cost allowance and import differential. The latter element is intended to cover the difference between the landed cost of the imported product and the refinery gate cost for domestic products. The difference between these allowable costs and the final selling price is made up of a tax to or subsidy from the Ministry of Finance. An illustration of the price buildup for the major products is shown in Table 2. - 175 - Annex 9 Page 4 of 4 Table 2: OFFICIAL PRICE STRUCTURE INCLUDING REVISIONS Gasoline Kerosene Jet B Jet A Gasoil LPG LFO HFO --------- Kz/litre ---------------- ---------Kz/k-- Refinery Gate 5.00 5.02 3.75 5,02 3.55 7.90 3.93 2.66 Tax 6.64 0.50 0.11 0.32 1.43 1.96 0,07 0.07 SONANGOL Costs 2.83 1.79 1.01 1.09 1.72 11.22 1,58 0.15 SONANGOL Profit 0.73 0.68 0.48 0.60 0.53 1.92 0.55 0.28 Reseller Margin 0.34 0.70 0.00 0.00 0.16 1.08 0.00 0.00 Import Differentia; 0.00 0.00 0.00 (0.07) 0.00 0.07 0.00 0.00 Margin 9.46 (1.70) 0.94 0.53 (0.38) (9.14) (1.33) 0.04 Final Price 25,00 7.00 6.29 7.50 7.00 15.00 4.80 3.20 Net Margin 16.10 (1.20) 1.06 0.78 1.05 (7.25) (1.26) 0.11 Source: MEP. Looking at cost recovery on individual products, based on official prices and the current SONANGOL cost/profit structure, kerosene, LPG, and light fuel oil are the only products which receive a payment from the State budget. The total payment owing to the Ministry of Finance on the margins on petroleum product distribution was projected at Kz 877 million in 1987. SONANGOL is permitted to deduct its costs for transporting Soyo crude to the refinery. These were budgeted at Kz 111 million for 1987. Therefore the net payment to the Ministry of Finance in 1987 would have been Kz 766 million. Adding Kz 1,735 million of total projected revenue accruing from product taxes, total government revenue from SONANGOL sales can be estimated at Kz 2,501 million for 1987. - 176 - Annex 10 Page 1 of 4 ECONOMICS OF THE LUANDA REFINERY USING ACTUAL HISTORICAL VALUES In order to evaluate the economics of the Luanda refinery a simplified model was developed. The revenues and costs per ton of crude run are calculated using typical plant yields for crude multiplied by product values based on FOB Mediterranean plus USS20/t plus a charge for terminalling, in these cases assumed to be US$40/t for liquid products, and US$20/t for LPG. The model is designed to allow a selected propor- tion of the total fuel oil yield to be sold to export at low sulfur fuel oil (LSFO) prices with a deduction for freight. Cabinda crude with a discount is used as the price basis for the Soyo FOB export price. Three cases were run using actual crude and product values for three different periods: 1985, 1986, and the first six months of 1987. The model and results for each period are illustrated in Tables 1, 2, and 3. As indicated with these opportunity values assumed for crude costs and product values, the gross refining margin is sufficient in all cases to cover "efficient" cash operating costs of US$10.95/t (US$1.50 per barrel) and earn a substantial operating profit. The model was set up so that the values of a few basic para- meters could be varied in order to check the sensitivity of the economies to these variations. For example, as more LSFO is sold to export the operating profit declines since a lower netback value is obtained on this than on local sales at border prices. Also, the smaller the discount assumed for Soyo crude off Cabinda "marker" the lower the operating profit. A s';mmary of the base case operating profits and assumptions and aforementioned sensitivity results is provided in the following tables. - 177 - Annex 10 Page 2 of 4 Table 1: ECONOMICS OF THE LUANDA REFINERY - RUNNING SOYO CRUDE 1985 OPPORTUNITY VALUES Assumptions LSFO to export 80.0 Total production Terminalling Liq. Products USS 4.00 per ton Terminalling LPG USS 20.00 per ton LSFO Freight etc. USNE USS 22.00 per ton Cabinda Crude FOB USS 188.76 per ton Soyo Discount USS 3.65 per ton Soyo FOB USS 185.11 per ton Local Crude Transport USS 3.00 per ton Crude & Product Cost/Revenue per Ton Values per ton ton Crude run FOB MED NYH (USS) (USS) (US$) Crude ('0OOs) 188.11 (188.11) +20 LSFO Production LPG 0.014 324.58 4.54 304.58 Gasoline 0.103 279.50 28.79 275,50 Kerosene/jet fuel 0.110 285.75 31.43 281.75 Gasoil 0.248 259.00 64.23 255.00 Fuel oil inland 0.097 171.25 16.54 167.25 LSFO exwert 0.386 159.75 61.73 167.25 181.75 Total products 0.958 207.27 Gross Refinery Margin 19.16 "Efficient" Cash Operating Costs (10.95) Operating - Profit 8.21 Source: Mission calculations. - 178 - Annex 10 Page 3 of 4 Table 2: ECONOMICS OF THE LUANDA REFINERY - RUNNING SOYO CRUDE 1986 OPPORTUNITY VALUES Assumptions LSFO to export 80.0% Total production Terminalling Liq. Products US$ 4.00 per ton Terminalling LPG USS 20.00 oer ton LSFO Freight etc. USNE USS 22.00 Der ton Cabinda Crude FOB US$ 96.92 per ton Soyo Discount US$ 3.65 per ton Soyo FOB US$ 93.27 per ton Local Crude Transport USS 3.00 per ton Crude & Product Cost/Revenue per Ton Values per ton Ton crude run FOB MED NYH (US$) (US$) (US$) Crude ('OOs) 96.27 (96.27) +20 LSFO Production LPG 0.014 224.75 3.15 204.75 Gasoline 0.103 165.25 17.02 161.25 Kerosene/jet fuel 0.110 176.58 19.42 172.58 Gasoil 0.248 155.92 38.67 151.92 Fuel oil inland 0.097 91.42 8.83 87.42 LSFO export 0.386 77.17 29.82 99.17 Totai products 0.958 116.91 Gross Refinery Margin 20.64 "'Efficient" Cash Operating Costs (10.95) Operating - Profit 9.69 Source: Mission calculations. - 179 - Annex 10 Page 4 of 4 Table 3: ECONOMICS OF THE LUANDA REFINERY - RUNNING SOYO CRUDE Ist Half 1987 OPPORTUNITY VALUES Assumptions LSFO to export 80.0% Tota oroducTion Terminalling Liq. Products US$ 4.00 Der ton Terminalling LPG US$ 20.00 De- ton LSFO Freight etc. USNE USS 22.00 per ton Cabinda Crude FOB US$ 123.18 per ton Soyo Discount USS 3.65 per ton Soyo FOB JSS 119.53 per ton Local Crude Transport US$ 3.00 per ton Crude & Product Cost/Revenue per Ton Values Per ton ton Crude run FOB MED NYH (US$) (USS) (USS) Crude ('°°°s) 122.53 (122.53) +20 LSFO Production LPG 0.014 243.67 3.41 223.67 Gasoline 0.103 190.50 19.62 186.50 Kerosene/jet fuel 0.110 183.33 20.17 179,33 Gasoil 0.248 175.50 43.52 171.50 Fuel oil inland 0.097 125.00 12.07 121.00 LSFO export 0.386 101.18 39.10 167.25 123.18 Total products 0.958 137.89 Gross Refinery Margin 15.36 "Efficient" Cash Operating Costs (10.95) Operating - Profit 4.41 Source: Mission calculations. - 180 - Annex 11 Page 1 of 2 HUMAN hESOURCES IN THE DOWNSTREAM PETROLEuM SECTOR Refinery 1. There are presently about 450 refinery employees, about 70 of whom are expatriates. 3ix of the expatriates are in top management while 65 are skilled laborers (first-class operators, senior mechanics, mechanical foremen, etc.). The total numoer of expatriates is down from about 100 three years ago and is continuing to decline through attri- tion. FPA is engaged in extensive training of Angolans. They have relied heavily on the Instituto Nacional de Petroleo (INP) at Sumbe (para. 4) for training operators and at present have about 20 trainees there. They are sufficiently encouraged by the good results obtained from the training of operators at Sumbe that they are now starting to send trainees in more specialized areas such as instrument mechanics and general mechanics. They have also regularly sent people abroad for two- month training courses at PETROFINA refineries, but this program has been constrained by Government authorities of late due to foreign exchange limitations. They also have some refinery on-site training run by an Italian firm, IDEAS. SONANGOL: Distribution and Marketing 2. There are some 3,000 employees in SONANGOL'S distribution and marketing division. This does not include service station staff who operate as independent dealers on a reseller margin. About half of SONANGOL's employees are attached to the storage installations while the rest are drivers, mechanics, accountants, sales agents, etc. There are several hundred employees at the 20-odd empty inland terminals. Even if the distribution system were running properly it is apparent that with the present work force there would be overstaffing in relation to volume distributed. Excessive staif, buildings, and facilities are the result of SONANGOL's taking over the old private companies. This led to duplication, overstaffing, and higher overheads than necessary. SONANGOL recognizes that it must rationalize inherited staff and facilities. With the strategic/military situation as it is, however, it is difficult to propose a major rationalization study and program at this time. SONAN4COL has made some progress in this regard and will continue to do so--perhaps in smaller steps and in more gradual fashion than would be ideal in a "normal" environment. 3. In summary, SONANGOL has trained a total of about 100 operators at INP, Sumbe, in the past five years. This may seem large in relation to the total staff complement but turnover at SONANGOL is extremely high. SONANGOL must train many just to have an available pool as many are lost to other enterprises. One of the causes of the high turnover or drop-out rate is the low trainee pay. Trainees receive about Kz 8,000/month during training and no food benefits except for use of the refinery cafeteria while within the plant. Even a trained operator receives no more than about Kz 12,000 to Kz 15,000/month, some of it in food. - 181 - Annex 11 Page 2 of 2 4. Petroleum Training Center (PTC) - Sumbe. The information available on the PTC 1/ at Sumbe suggests that the preparation/- feasibility stage for this center was insufficient, start-up difficulties were greater than expected (16-month delay from early 1984 to late 1985) and training objectives are not being attained. UNDP/UNIDO (the United Nations Development Programme/United Nations International Development Organization), contrary to its practice, did not assign a project manager so that decision-making and problem-solving were a slow and cumbersome process. The PTC did not have the support of its main designated beneficiaries (the oil companies) in part because training programs were not designed in consultation with or to satisfy the oil companies. Naturally enough, the oil companies preferred to use their own training facilities. More recently, however, they are showing more active interest and support, to the extent of sending their own instructors to the center together with training equipment. The Government (Ministry of Education and the MEP) and the international partners (UNDP, UNIDO, NORAD, and Italy) should perhaps reconsider the scale of this project and, possibly, restart it under the firm control of the oil industry but with assistance from any of the present partners willing to continue helping Angola train people for its most important industry. The interests of the oil companies in Angola probably do not extend to other SADCC (Southern African Development Coordination Conference) countries. If so, training of nationals of other SADCC countries should still be possible, but could probably not be the main task of the PTC. 5. The greatest shortfall in meeting objectives has been in the training of instructors. Reportedly, only ten instructors were trained and only five remain at the PTC. 2/ Obviously, this is the highest priority for future action if the PTC is ever going to become even partly self-sustaining. 6. No financial information at all was available to the mission. Therefore, it is impossible to comment on the cost-effectiveness of the PTC. However, operating the PTC is clearly very expensive. Before making a decision either to continue using or stop using the center, objectives and achievements need to be compared to costs. The PTC represents a considerable claim on the scarce resources of Angola which contributed US$2-3 million/y to the operation of the center. This amount is in addition to the many millions supplied by other donors. 1/ A member of the Energy Assessment Mission visited Sumbe but this report benefitted from consulting a UNDP report entitled: UNDP: Evaluation Mission on RAF/83/22 - Assistance to the Petroleum Training Center, UNDP/UNIDO, Luanda, Final Report, 25 April, 1987. 2/ Ten other instructors from other SADCC countries were also trained and returned home. Between late 1985 and early 1987, 311 trainees attended courses or seminars. - 182 - Anndx 12 Page 1 of 3 PETROLEUM PRODUCT TRADING Middle Distillate Imports 1. SONANGOL procures its imports of jet fuel and gasoil through its U.K. subsidiary, SONANGOL LIMITED. SONANCOL LIMITED is a joint venture with the large West German trading company, STINNES, which assists in arrangements related to external product trading--both imports and exports. Before the SONANGOL-STINNES inint venture arrangement was promulgated about a year ago, SONANGOL LIi ED tried to handle middle distillates imports on an as-needed, ad hoc cargo-by-cargo basis. The sporadic nature of the demand, long distance from major supply sources, and foreign exchange/letter of credit problems all combined to make the attempt at importing in 10,000 DWT vessels somewhat complicated. With the joint venture arrangement a steady import pattern of 5,000-ton parcels roughly every month from Tenerife, Canary Islands has been established. The joint venture partnership has a 5,000 DWT tanker on time-charter which can be out-chartered to earn revenues when not required for the Angola import chartering. For instance, at present Angola via SONANGOL pays the joint venture Platt's Mediterranean plus US$39.00/t CIF Luanda for A-1 jet fuel. This covers FOB cost of acquiring the product, marine insurance, marine loss, letter of credit arrangements, necessary financing and management of the entire acquisition, and chartering operations. One half of the profits on this and other partnership activities flow back to SONANGOL LIMITED through its joint venture share. Currently, these profits on product trading are financing the entire SONANCOL LIMITED London office, whose principal activity is crude oil trading. 2. The mission has undertaken an analysis of the middle distillates procurement arrangement. Small tanker freight quotations for service from the major supply sources of Tenerife, the Canaries, and Augusta, Italy, are illustrated in Table 1. These figures result in a calculated freight cost of US$26.91/t for Tenerife-Luanda and US$39.64/t for Augusta-Luanda. If a trading commission (for chartering, financing, and general "implementation" function of the joint venture) is included in the total, CIF-FOB differential works out at US$33.62/t for Tenerife- Luanda and US$46.81/t foe AL3usta-Luanda as shown in Table 1. 3. Since the existing arrangement incorporates a total differential of US$39.00/t above Platt's Mediterranean spot price, Angola pays an "excess margin" of about US$5/t on a Tenerife sourcing. Procur- ing in Augusta would clearly be too costly. These calculations call for two comments: (a) assuming that the 5,000-ton shipment is correct, then even a high cost (i.e., distant) source of supply such as Tenerife apparently costs less than present arrangements (by about US$5/t); - 183 - Annex 12 Page 2 of 3 Table 1: DERIVED CIF-FOB DIFFERENTIAL FOR MIDDLE DISTILLATES REPLENISHMENT-LUANDA Tenerife- Augusta- Luanda Luanda Bases USS/t US$/t FOB Cost 159.51 159.51 Average Med PIW Ist 6 months, 1987 Ocean Freight 26.91 39.64 186.42 199.15 Marine loss, Insurance, 1.12 1.20 0.6% FOB pius freight Commission 5.59 5.97 3.0% FOB plus freight CIF Luanda 193.13 206.32 CIF-FOB differential 33.62 46.81 Source: SONANGOL and Annex 9, Table 1. (b) it seems evident that a 10,000 DWT vessel is much more appro- priate given that total imports exceed 100,000 t/y and Tenerife is an unnecessarily distant source of supply. US$/t Acquisition cost FOB Luanda 108.00 Ocean freight Luanda-New York 11.25 119.25 Marine loss 0.3Z 0.36 Marine insurance 0.2Z 0.24 Trading commission 3% 3.58 Landed cost New York 123.19 4. Assuming that SONANCOL-STINNES could actually sell the fuel oil at New York Harbor spot cargo prices and get a 3% trading commission for managing all "implementation" functions, it would make an "excess margin" of about US$7/t under the current arrangement. At present export levels this amounts to some US$3.5 million/y. It appears that some profits which could go to SONANGOL (and the Government) thus accrue to SONANGOL LIMITED. Even if half the profits flow back to SONANGOL LIMITED it might be worth discussing with STINNES whether a better deal could be struck. Alternatively, 3ONANGOL would probably do better by returning to its 1981-82 practices of opening tenders in the international fuel oil markets. In the past, offers received were about New York Harbor less US$2.40/bbl or US$16.80/t. Adding other incidentals, this would give - 184 - Annex 12 Page 3 of 3 Angola a netback fuel oil price of about US$115/t rather than US$108/t which may be the case under present arrangements. Fuel Oil Imports 5. The fuel oil produced in Angola is low sulfur (0.3%), high-pour material. The refinery seLls it at official refinery gate prices to SONANGOL who then sells it to the SONANGOL-STINNES joint venture. The present contract arrangement provides for SONANGOL to receive New York Harbor cargo prices less US$22.00/t. The U.S. Northeast coast is the prime disposition for this low-sulfur product. It generally moves from Luanda in 50,000-ton shipments. 6. A typical freight rate (average spot 1st 6 months 1986) for transporting this type of cargo in 50,000 DWT vessels is about 110% WS or US$11.25/t for the Luanda-New York voyage. Recent quoted New York Harbor spot cargo prices for 0.3% high-pour LSFO have averaged about US$130/t. On this basis SONANGOL-STINNES would pay SONANGOL US$108/t, which is about US$15/t less than the landed cost at New York. International Bunker Sales 7. International bunker sales fell by about 50% in the period 1981-1986. Airline bunkers have recuperated and fuel oil bunkering is also stable but there has been a marked decline in gasoil bunkers, probably because Angola became a net importer over this period. Table 2 summarizes the trends. Table 2: INTERNATIONAL BUNKER SALES 1981-86 1981 1985 1986 Jet fuel 29,074 19,827 28,920 Gasoil 37,760 11,555 5,850 Fuel oil 2,771 3,379 2,600 Total Bunkers 69,605 34,761 37,370 Source: Angolan authorities. - 185 - Annex 13 Page 1 of 10 FIGURES ON TM POWER SUBSECTOR Table 1: INSTALLED AND AVAILABLE GENERATING CAPACITY (1987) System Name of Type Number of units Capacity (MW) Date of Date of and Plant a/ and unit power Installed/Availabie Commissioning Unavailability Province (MW) Month/year Month/year North Kuanza N. Cambambe H 2 w 45 & 2 x 45 180.0 135.0 1963,1973 Bengo Mabubas H 2 x 3 & 2 x 5.9 17.8 - 1953,1959 2/1986 Luanda Luanda, GTI GT I x 25.6 25.6 25.6 1980 it Luanda, GT2 GT I x 31.2 31.2 31.2 8/1985 Subtotal 254.6 191.8 Central Benguela Lomaum H 2 x 10 & 1 x 15 35.0 - '964-1972 3/1983 Blopio H 4 x 3.6 14.4 7.2 1957 Biopio GT I x 22.8 22.8 22.8 2/1974 of Biopio D 2 x 1.5 3.0 1.5 1982 it Lobito D 4 x 5 b/ 20.0 10.0 1986 Huambo Huambo GT 1 x 10 c/ 10.0 - 1981 2/1985 to Huambo D 5 x 0.8 & 2 x 0.85 6.0 5.2 1953-1986 Subtotal 111.2 46.7 South Huila Matala H 2 x 13.6 27.2 13.6 1959 It Lubango D 3 x 0.4 & 12 x 0.2 3.6 3.6 n.a. it Namibe D 2 x 5.75 11.5 11,5 1980 it Tombwe 0 1 x 1.6 1.6 - 1970 it Jamba D 3 x 1.9 5.7 - 1968 if Saco D 2 x 1.45 2.9 - n.a Subtotal 52.5 28.7 Isolated Systems Cabinda Malongo GT I x 12.3 12.3 - 1980 1985 to it D 3 x 1.5 4.5 - 8/1971 it it D 4 x 0.3 1.2 1.2 n.a Ulge Luquixe H 3 x 0.36 1.1 1.1 1957,1968,1971 it Uiae D 3 x 06 1.8 1.8 n.a to It D 1 x 1.5 1.5 1.5 1982 a/ H = Hydroelectric; D a Diesel; GT = Gas Turbine. b/ Railway carriage mounted. c/ ISO rating 13.5 MWo Derating of 22% due to altitude. Source: SADCC, SCNEFE, and ENE. - 186 - Annex 13 Page 2 of 10 Table : ;NSTALLED AND AVAILABLE GENERATING CAPACITY (1987) (Continued) System Name of Type Number of units Capacity (MW) Date of Date of and Plant a/ and unit power Installed/Available Commissioning Unavailability Province (MW) Month/year Month/year Lunda N Luaximo H 4 x 2.4 9.6 - 1957 of Luxilo D I x 1.5 1.5 - n.a of Lucapa D 2 x 3.2 6.4 - n.a Bie Andulo H 2 x 0.05 0.1 0.1 n.a Kunje H 3 x 0.54 1.6 1.1 1/1971 ft Coemba H 2 x 0.1 0.2 0.1 n.a of Kuito D I x 0.8 & 1 x 0.5 1.3 0.5 n.a Moxico Luena D 2 x 0.6 1.2 - 1974 Huila Kubango H 2 x 0.15 0.3 - 8/1972 Subtotal 44.6 7.4 Total Angola 462.9 274.S a/ H c Hydroelectric; D = Diesel; GT = Gas Turbine. b/ Railway carriage mounted. c/ ISO rating 13.5 MW. Derating of 22% due to altitude. Source: SADCC, SONEFE, and ENE. Table 2: TOTAL GENERATION BY TYPE OF PLANT (GWh) 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 Hydro Cabinda 255.7 303.2 370.7 502.6 552.2 421.4 338.6 394.7 419.4 465.0 489.9 506.7 545.0 674.4 565.6 544.2 590.4 Mabubas 46.7 72.5 67.4 27.7 15.2 6.3 0.3 1.3 2.9 1.3 2.4 1.6 1.9 4.1 3.6 4.7 1.3 Lomaum 81.9 97.9 109.9 120.8 125.8 92.4 63.4 72.4 90.4 85.3 112.4 99.4 112.5 2.4 0.0 0.0 0.0 Olopio 37.2 34.9 34.1 43.1 37.6 42.7 14.4 24.0 16.0 39.0 22.6 32.1 38.3 34.5 16.4 12.0 35.1 Matala 91.5 61.8 67.5 77.8 62.1 69.3 9.6 10.8 8.7 10.5 10.5 10.7 8.6 10.9 8.0 9.6 8.1 Other hydro 91.5 61.R 67.5 77.8 62.1 69.3 9.6 10.8 8.7 10.5 10.5 10.7 8.6 10.9 8.0 9.6 8.1 Total Hydro 543.7 605,0 680.3 814.1 858.1 705.1 460.7 529.4 568.1 636.4 675.3 688.1 748.4 674.3 634.7 621.0 691.1 Gas turbines Luanda - - - - - - - - - - 0.1 0.0 0.3 0.1 0.5 6.2 14.2 Bioplo - - - - 8.0 0.0 0.0 0.0 0.1 0.0 1.8 8.8 0.7 60.0 61.9 49.7 15.5 Huambo - - - - - - - - - - 1.8 1.7 0.1 11.0 7.1 2.9 0.0 Cabinda - - - - - - - - - - 6.0 9.9 18.8 14.2 13.6 18.1 5.7 Total Gas t:rbines 0.0 0.0 0.0 0.0 8.0 0.0 0.0 0.0 0.1 0.0 9.7 20.4 19.9 85.3 88.1 76.3 35.4 Diesel Generation 100.0 137.0 158.5 170.2 162.7 133.2 71.4 28.8 15.0 13.0 5.3 8.0 10.3 4.2 8.6 7.5 27.0 Total Angola 643.7 742.0 838.8 984.3 1023.8 838.3 532.1 558.2 588.2 649.4 690.3 716.5 778.6 763.8 726.4 704.8 753.5 Average growth rate S p.a. 15.3 13.0 17.3 4.5 -18.5 -36.5 4.9 4.5 11.4 6.3 3,8 8.7 -1.9 -4.9 -3.0 6.9 Percentage of Total Hydro 84.46 81.54 81.10 82.71 88.41 84.11 86.58 94.84 97,41 98.00 97.83 96.04 96.12 88.28 87.98 88.11 91.72 0 D Gas turbines 0.00 0.00 0.00 0.00 0.78 0.00 0.00 0.00 0.02 0.00 1.41 2.85 2.56 11.17 11.44 10.83 4.70 Diesel 15.54 18.46 18.90 17.29 15.81 15.89 13.42 5.16 2.57 2.00 0.77 1.12 1.32 0.56 1.18 1.06 3.58 0 w I-r Source: SADCC, SONEFE. and ENE. 0 - 188 - Annex 13 Page 4 of 10 Table 3: BREAKDOWN OF GENERATION BY SYSTEM AND POWER PLANT (GWh) System/Plant Type a/ 1982 1983 1984 1985 1986 Northern Cambambe H 544,964 574,256 555,642 544,156 590,448 Mabubas H 1,851 4,113 3,572 4,660 1,266 Luanda GT 272 55 496 6,208 14,234 Sub-total 547,087 578,424 559,710 555,024 605,948 Central Lomaum H 112,500 2,428 0 0 0 Blopio H 38,300 34,516 16,408 12,801 35,114 Bloplo GT 700 60,042 61,896 49,746 15,531 Huambo GT+D 70 11,000 7,106 2,275 2,021 slopio D 0 0 3,932 1,822 10,615 Blackstone sets D 0 0 0 505 11,948 Sub-total 151,570 107,986 89,342 67,149 75,229 Southern Matala H 42,127 48,104 51,113 49,698 56,212 Namlbe 0 1,253 0 0 3,178 118 Sub-total 43,380 48,104 51,113 52,876 56,330 Total Interconnected 742,037 734,514 700,165 675,049 737,507 Cabinda Malongo GT.D 187,681 14,240 13,618 18,029 5,690 U;Qe Luquixe H 5,000 5,323 3,360 1,533 2,891 Uige 0 787 927 3,092 502 919 Subtotal 5,787 6,250 6,452 5,035 3,810 Bie Kunje H 5,310 5,338 3,555 4,795 4,751 Andulo H 270 289 150 257 245 Coemba H 235 228 152 228 182 Kulto D 1,400 985 1,698 1,277 1,197 Chinguar D ....... ..*.... 26 98 ...... N'harea D ....... 11 . , ...... ....... Subtotal 7,215 6,851 5,581 6,655 6,375 Moxico Luena 0 4,679 1,907 590 122 90 Total Isolated 36,449 29,248 26,241 29,841 15,965 TOTAL ANGOLA 778,486 763,762 726,406 704,890 753,472 a/ H4 = Hydroelectric; D = Diesel; GT = Gas Turbine Source: ENE, SONEFE, MEP reports, and Csssion estimates. - 189 - Annex 13 Page 5 of 10 Table 4: TOTAL ELECTRICITY GENERATION, DISTRIBUTION AND LOSSES (1967-1986) Generation Distribution a/ Losses Base Growth rate Base Growth rate % of Year GWh 100 a p.a. GWh 100 % p.a. GWh Generation 1967 390.8 38.0 - 372.8 40.6 - 18.0 4.6 196d 456.8 44.4 16.9 433.2 47.1 16.2 23.6 5.2 1969 541.5 52.6 18.5 519.0 56.5 19.8 22.5 4.2 1970 643.8 62.6 18.9 618.9 67.3 19.2 24.9 3.9 1971 741.9 72.1 15.3 702.0 76.4 13.4 39.9 5.4 1972 838.8 81.5 13.0 787.2 85.6 12.1 51.6 6.2 1973 984.3 95.7 17.3 914,5 99.5 16.2 69.8 7.1 1974 1,028.8 100.0 4.5 919.2 100.0 0.5 109.6 10.7 1975 838.8 81.5 -18.5 768.6 83.6 -16.4 69.7 8.3 1976 532.1 51.7 -36.5 487.9 53.1 -36.5 44.2 8.3 1977 558.8 54.3 4.9 511.9 55.7 4.9 46.3 8.3 1978 583.2 56.7 4.5 540.6 58.8 5.6 42.6 7.3 1979 649.4 63.1 11.4 596.8 64.9 10.4 53.5 8.3 1980 690.3 67.1 6.3 626.1 68.1 4.9 63.8 9.2 1981 716.5 69.6 3,8 644.0 70.1 2.9 72.5 10,1 1982 778.6 75.7 8.7 712.6 77.5 10.7 66.0 8.5 1983 763.8 74.2 -1.9 686.6 74.7 -3.6 77.2 10.1 1984 726.4 70.6 -4.9 634.3 69.0 -7.6 92.1 12.7 1985 7804.8 68.5 -3.0 605.0 65.8 -4.6 99.8 14.2 1986 753.5 73.2 6.9 636.2 69.2 5.2 117.3 15.6 Average Annual growth rates (%) calculated using least square estimates: Generation Distribution 1967 - 1973 16.6 16.2 1977 - 1982 6.9 6.6 1977 - 1986 3.1 2.1 a/ tIDistribution", figures apparently include losses in MV (60 kV and lower) distribution networks. Source: Industrial Statistics (INE: 1967-1973); ENE, SONEFE, MEP reports, and mission estimates. - 190 - Annex 13 Page 6 of 10 Table 5: TOTAL LENGTH OF TRANSMISSION LINES - 1987 North Central South Isolated Voltage system system system systems Total kV (km) (km) (km) (km) (km) 220 548 - - - 548 150 - 281 288 - 569 100 159 - - - 159 60 223 a/ 62 b/ 496 228 c/ 1,009 a/ Includes 109 km of new line Mabubas-Quibaxe (Dembos electri- fication). b/ Currently operated at 30 kV. c/ ENDIAMA system, province of Lunda N. Source: SADCC, SONEFE, and ENE. - 191 - Annex 13 Page 7 of 10 Table 6: MAIN TRANSMISSION LINES - CHARACTERISTICS System No. of Commiss. Line Circu Its kV km Conductors Year Comments North Cambambe-Luanda I 220 175.0 ACSR Crow 1963 Cambambe-Viana 1 220 158.0 " 1984 a/ Viana-Luanda 1 220 17.0 o 1984 Currqntly at 60 kV Cambamto-N'Dalatando 1 220 73.0 o 1970 Camtdmbe-Gabela 1 220 125.0 "t 1974 Out of order since 1984 N'rialatando-Cacuso I 100 97.0 ACSR 30/7 1974 Cacuso-Malange I 100 62.0 "1 1974 Mabubas-Luanda 2 60 57.0 Cu 50mm 2 1953 Mabubas-Quibaxe 1 60 109.0 n.a. 1987 b/ Center Biopio-Quileva 1 150 17.5 ACSR Panther n.a, Biopio-Lomaum 1 150 96.5 t 1964 Lomaum-A. Catumbela 1 50 48.5 of 1964 A. Catumbela-Huambo 1 150 118.5 " 1964 Huambo-Chinguar 1 60 n.a. n.a. n.a. Operated at 30 kV South Matala-Lubango I 150 168.0 ACSR Panther 1959 Matla-Jamba 1 150 120.0 " 1973 Lubango-Namibe 2 60 162.0 (i) Cu 70mm2 1960 (ii) ASCSR Partridge 1974 Namibe-Toinbwa 1 60 94.5 ACSR Civetta 1974 Highly corroded Namibe-Saco I 60 9.0 f' 1973 Jamba-Tchamutete 1 60 68.0 ACSR 26/7 1973 Out of order a/ Inactive due to works in Viana substation. b/ Dembos electrification. To be commissioned in July 1987. Source: "Estudo da Interligacao das Redes de Angolat", EDP; Lisbon 1984. - 192 - Annex 13 Page 8 of 10 Table 7: BREAKDOWN OF DISTRIBUTION BY SYSTEM AND PROVINCE (MWh) System/Prevince 1982 1983 1984 1985 1986 Northern Luanda & Bengo 440,750 475,893 453,115 459,763 491,920 Kuanza N+Malange 35,190 23,159 22,596 16,256 13,260 Kuanza S 19,980 17,946 10,322 308 925 Subtotal 495,920 516,998 486,033 476,336 506,105 Central Benguela 113,400 73,252 68,123 52,252 61,018 Huambo 33,254 30,278 17,719 10,895 11,563 Subtotal 146,654 103,530 85,842 63,147 72,581 Southern Huila 20,392 22,697 21,957 21,292 21,501 Namibe 15,787 14,328 18,038 16,960 21,789 Subtotal 36,179 37,025 39,995 38,252 43,290 TOTAL INTERCONNECTED 678,753 657,553 611,870 577,735 621,976 Isolated Cabinda 17,401 14,225 11,516 16,943 4,744 Uige 5,281 6,249 5,386 4,064 3,458 Bie 6,808 6,639 5,007 6,116 5,920 Moxico 4,397 1,907 590 121 90 Total Isolated 33,387 29,021 22,499 27,244 14,212 TOTAL ANGOLA 712,640 686,574 634,369 604,979 636,188 Source: ENE, SONEFE, MEB reports, and mission estimates. - 193 - Annex 13 Page 9 of 10 Table 8: REGIONAL BREAKDOWN OF ELECTRICITY DISTRIBUTION a/ IN SELECTED YEARS (In GWh and Percentage) Percentage of Electricity Distribution (GWh) total (3) Province 1967 1970 1972 1974 1982 1986 1974 1986 Luanda & Bengo b/ 172.8 285.6 409.0 478.8 440.8 491,. 52.1 77.3 Benguela c/ 93.5 124.4 137.8 147.2 113,4 61.0 16.0 9.6 Huila l Namibe d/ 20.6 69.3 78.2 95.5 36.2 43.3 10.4 6.8 Huambo e/ 16.6 30.3 32.1 40.1 33.3 11.6 4.4 1.8 Sub total 303.5 509,6 657.1 761.6 623.7 607.8 82.9 95.5 Others 69.3 109.3 130.1 157.6 88.9 28.4 17.1 4.5 Total 372.8 618.9 787.2 919.2 712.6 636.2 100.0 100.0 a Distribution includes customer consumption and MV and LV distribution losses. b/ Includes the city and greater Luanda. c/ Includes the urban areas of Benguela and Lobito. d/ Includes the urban areas of Lubango and Namibe. e/ Includes the urban area of Huambo. Source: Industrial Statistics (INE), ENE, and mission estimates. Table 9: FOREIGN BORRO4INGS AND FOREIGN DEBT SERVICE OF THE POWER SUBSECTOR - EXCLUDING GAMEK (In 'OOOs units) Currency Loan amounts Av. int. Av. grace Av. repmnt Debt incurred Debt Service: repayment and interest ('OOOs) Borrowers (a) contracted (b) rate % period-year period-year 1-DEC-86 1987 1988 1989 1990 ENE USS 21,951 9.2 1.8 5.7 10,733 929 1,805 4,907 4,810 574 1,463 1,488 1,027 1,503 3,268 6,395 5,837 FRF 109,361 8.9 3.8 4.9 46,285 7,382 26,610 26,610 26,125 3,444 8,64 6,145 3,660 10,826 35,250 32,755 29,785 GBP 10,323 9.2 0.0 6.9 5,907 1,493 1,493 1,493 1,429 872 653 433 214 2,365 2,146 1,926 1,643 BEF 351,098 8.4 2.0 7.4 325,333 35,939 54,835 54,835 54,835 14,956 31,146 19,935 13,63 50,895 85,981 74,770 68,468 US3 equivalents (a); 63,279 9.0 1.9 6.0 34,935 7,999 14,149 16,277 14,673 SONEFE USS 29,237 7.4 3.6 6.6 20,603 4,242 4,125 3,298 1,941 1,696 11M93 774 577 5,938 5,218 4,072 2,518 CHF 14,581 6.9 2.0 3.0 7,401 4,860 2,541 - - 426 91 5,286 2,632 - - US$ equivalents (a): 38,277 7.3 3.2 5.7 25,192 9,215 6,850 4,072 2,518 TOTALS US$ equivalents (a); > 101,556 8.4 2.4 5.9 60,127 17,214 20,999 20,349 17,191 a rmrm #-x (a) Exchange rates: French FR = US$0.157; Sterling Pound = US$1.49; Belgian FR = US$O.025; Swiss FR = US$0.62. ° X (b) In 1981 23.4 million US$ equivalent; 1983, 1.3 million; 1984, 44.9 million; 1985, 12.0 million; 1986, 0.7 million; 1987, 19.2 0 1-h million; total, 101.5 million. Source: ENE and mi-sion estimates. - 195 - Annex 14 Page 1 of 28 ELECTRICITY DEMAND PROJECTIONS - MAIN ASSUMPTIONS BEP, THEMAG, AND MISSION STUDIES 1. Since the early 1970s international consultants have tried at least four different approaches to electricity demand projections. The first was prepared by SOFRELEC in 1971, for the Junta Provincial de Electri.ficao de Angola (JPEA). l/ In 1984 a technical study was presented by Electricidade de Portugal (EDP). Recently, two studies aiming to define a minimum cost expansion plan and interconnection facilities were undertaken. Both studies covered the planning period 1986-2005. The first was financed by the Banque Africaine de Develop- pement (BAD) and prepared by Belgian Engineering Promotion (BEP) 2/, the second was prepared within the framework of the Southern African Develop- ment Coordination Conference (SADCC) by the Brazilian consulting firm THEMAG 3/. 2. The demand forecasts of these two studtes are not immediately comparable. BEP presented its forecasts by province, using the annual reports and plans' outLine of the Ministry of Energy and Petroleum's (Ministerio de Energia e Petroleos, MEP). THEMAG defined 20 major "load- centers" (corresponding to 20 main regions) to concentrate load for the purpose of studies and simulations. Connection- criLeria of new munici- palities were also different. Angolan technicians contributed very little to the BEP study, while that of THEMAG was based on existing SONEFE and ENE approximations. Both studies assumed that either no price-elasticity measure existed or that it would have no effect in spite of the urgent need for tariff increases all over the country. BEP used a sectoral approach, analyzing the development of the t,iree main sectors: industry, services, and residential, but used outdated industrial statistics (1967-72) complemented by recent Governmental intentions. Some assumptions were also made on the rate of replacemenL of other energy sources by electricity. BEP distributed aggregate demand forecasts by province and municipality. THEMAG used a global approach, preparing demand scenarios on the basis of present demand in the main substation and applying growth rates in agreement with "optimistic" and "pessimistic" market expectations. Each main substation, or "load center" was characterized by two components: one corresponding to the 1/ "Etude de l'Interconnexion des Principaux Systemes de Production d'Energie", SOFRELEC, 1971. 2/ "Etude d'un Plan Directeur de Developpement du Reseau Electrique National d'Angola"; Brussels, August 1986. 3/ "Interligacao dos Sistemas Norte/Centro/Sul em Angola. Possibilida- des de Interligacao com a Namibia"; Preliminary Report; S. Paulo, October 1986. - 196 - Annex 14 Page 2 of 28 "natural" consumption growth of existing consumers and the other resulting from the extension of electrification to new areas. 3. To overcome the uncertainties on the rehabilitation and develop- ment of the productive system and urban infrastructure, the studies assumed that the general rehabilitation of the socio-economic framework could start by the end of 1986. "Natural" consumption would experience moderate growth untiL 1989 when suppLy conditions should be back to normal. Then five years of accelerated development would follow, re- establishing full use of productive capabilities. The second half of the planning period (1995-2005) would be characterized by stable conditions of economic development and lower growth rates for electricity consump- tion. In addition to a "Base" scenario, THEMAG used "Low" and "High" scenario projections which concerned only the "Natural" component of con- sumption and resulted from "Base" projections -8% and +8% respectively. THEMAG used the same annual growth rates within each system for the three scenarios. Resulting demand differences among scenarios for that study are an outcome of distinct "starting" values. 4. In order to compare both forecasts, BEP projections were rear- ranged into regions to match load centers used by THEMAG. Additionally, all values were converted into consumption (energy) and demand at the HV busbars of HV/MV substations. 5. Energy and demand forecasts for the three systems made by BEP, THEMAG, and IBRD/UNDP are detailed in Tables 1 through 21, - 197 Annex 14 Page 3 of 28 Tabie 1: NORTHERN SYSTEM - CASE I (BEP STUDY) Energy and demand forecasts, 1986-2000 Energy Peak demand Transmission Generation Load factor Consum Annu at Annu losses requirements at at HV Grow wv Grow Energy Demand Energy Demand Generation Year GWh % MW I % GWh MW Hours % 1986 474.8 84.5 6,0 8.0 505,1 91.8 5,499 62.8 1987 520.9 9.7 94.1 11.3 6.0 8,0 554,. :02.3 5,420 61.9 1988 571.6 9.7 104.7 11.3 6.0 8.0 608. 1'3.8 5,342 61.0 1989 627.1 9.7 116.6 11,3 6.0 8.0 6b7.i '26.7 5,264 60.' 1990 688.1 9.7 129.8 11.3 6.0 8.0 732.0 141.1 5,188 59.2 1991 769.3 11.8 144.8 11.6 6.0 8.0 818.4 157.4 5,198 59.3 1992 860.0 11.8 161,6 11.6 6.0 8.0 914.9 175.7 5,209 59.5 1993 961.5 11.8 180.3 li.6 6.0 8.0 1,022.8 196.0 5,219 59.6 1994 1,047.9 11.8 201.2 11.6 6.0 8.0 1,143.5 218.7 5,229 59.7 1995 1,201.7 11.8 224.5 11.6 6.0 8.0 1,278.4 244.0 5,239 59.8 1996 1,297.5 8.0 242.0 7.8 6.0 8.0 1,380.3 263.0 5,248 59.9 1597 1,400.9 8.0 260,8 7.8 6,0 8.0 1,490.3 283.5 5,257 60.0 1998 1,512.5 8.0 281.1 7.8 6.0 8.0 1,609.1 305.6 5,266 60.1 1999 1,633.1 8,0 303.0 7.8 6.0 8.0 1,737.3 329.4 5,275 60.2 2000 1,763.2 8.0 326.6 7.8 6.0 8.0 1,875.7 355.0 5,284 60.3 Note: Annu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: BEP. - 198 - Annex 14 Page 4 of 28 Table 2: NORTHERN SYSTEM - CASE 2 (THEMAG "LOW" SCENARIO) Energy and demand forecasts, 1986-2000 Energy Peak demand Transmission Generation Load factor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWh % MW % % % GWh MW Hours % 1986 537.0 94.4 6.0 8.0 571.3 102,6 5,568 63.6 1987 599.5 11.6 104.7 10.9 6.0 8.0 637.8 113.8 5,604 64.0 1988 669.7 11.7 116.6 11.4 6.0 8.0 712.4 126.7 5,621 64.2 1989 748.1 11.7 130.2 1:.7 6.0 8.0 795.9 141,5 5,624 64,2 1990 815.5 9.0 141.9 9.0 6.0 8.0 867.6 154.2 5,625 64.2 1991 898.7 1U.2 157.1 10.7 6.0 8.0 956.1 170.8 5,599 63.9 1992 1,011.1 12.5 179.3 14.1 6.0 8.0 1,075.6 194.9 5,519 63.0 1993 1,102.1 9.0 195.4 9.0 6.0 8.0 1,172.4 212.4 5,520 63.0 1994 1,201.2 9.0 213.0 9.0 6.0 8.0 1,277.9 231.5 5,519 63.0 1995 1,309.3 9.0 232.2 9.0 6.0 8.0 1,392.9 252.4 5,519 63.0 1996 1,414.0 8.0 250.7 8.0 6.0 8.0 1,504.3 272.5 5,520 63.0 1997 1,527.2 8.0 270.8 8.0 6.0 8.0 1,624.7 294.3 5,520 63.0 1998 1,649.4 8.0 292.5 8.0 6.0 8.0 1,754.7 317.9 5,519 63.0 1999 1,781.3 8.0 315.8 8.0 6.0 8.0 1,895.0 343.3 5,521 63.0 2000 1,923.8 8.0 341.1 8.0 6.0 8.0 2,046.6 370.8 5,520 63.0 Note: Annu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: THEMAG. - 199 - Annex 14 Page 5 of 28 Table 3: NORTHERN SYSTEM - CASE 3 (THEMAG "BASE" SCENARIO) Energy and demand forecasts, 1986-2000 Energy Peak demand Transmission Generation Load factor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWh % MW % I % GWh MW Hours % 1986 583.7 102.6 6.0 8.0 621.0 111.5 5,568 63.6 1987 651.6 11,6 113.8 10.9 6.0 8.0 693.2 123.7 5,604 64.0 1988 727.9 11.7 126.7 11.3 6.0 8.0 774.4 137.7 5,623 64.2 1989 813.2 11.7 141.5 11.7 6.0 8.0 865.1 153.8 5,625 64.2 1990 886.3 9.0 154.2 9.0 6.0 8.0 942.9 167.6 5,625 64.2 1991 976.8 10.2 170.8 10.8 6.0 8.0 1,039.1 185.7 5,597 63.9 1992 1,099.0 12.5 194.9 14.1 6.0 8.0 11,69.1 21'.8 5,519 63.0 1993 1,197.9 9.0 212.4 9.6 6.0 8.0 1,274.4 230.9 5,520 63.0 1994 1,305.7 9.0 231.5 9.0 6.0 8.0 1,389.0 251.6 5,520 63.0 1995 1,423.2 9.0 252.4 9.0 6.0 8.0 1,514.0 274.3 5,519 63.0 1996 1,537.0 8.0 272.5 8.0 6.0 8.0 1,635.1 296.2 5,520 63.0 1997 1,660.0 8.0 294.4 8.0 6.0 8.0 1,766.0 320.0 5,519 63.0 1998 1,792.8 8.0 317,9 8.0 6.0 8.0 1,907.2 345.5 5,520 63.0 1999 1,936.2 8.0 343.3 8.0 6.0 8.0 2,059.8 373.2 5.520 63.0 2000 2,091.2 8.0 370.8 8.0 6.0 8.0 2,224.7 403.0 5.520 63.0 Note: Annu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: THEMAG. - 200 - ~~~~~Annex 14 - 200 - Page 6 of 28 Table 4: NORTHERN SYSTEM - CASE 4 (THEMAG "HIGH" SCENARIO) Energy and demand forecasts, 1986-2000 Energy Peak demand Transmission Generation Load factor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWhI MW % % GWh MW Hours % 1986 630.4 110,8 6.0 8,0 670.6 120.4 5,568 63.6 1987 703.7 11.6 122.9 10,9 6.0 8.0 748.6 133.6 5,604 64.0 1988 786.1 11.7 136.8 11.3 6.0 8.0 836.3 148.7 5,624 64.2 1989 878.3 11.7 152.8 11.7 6.0 8.0 934.4 166.1 5,626 64.2 1990 957.3 9.0 166.5 9.0 6.0 8.0 1,018.4 181.0 5,627 64.2 1991 1,054.9 10.2 184.5 10.8 6.0 8.0 1,122.2 200.5 5,596 63.9 1992 1,186.9 12.5 210.5 14.1 6.0 8.0 1,262.7 228.8 5,519 63.0 1993 1,293.7 9.0 229.4 ".0 6.0 8.0 1,376,3 249.3 5,520 63.0 1994 1,410.2 9.0 250.0 9.0 6.0 8.0 1,500.2 271.7 5,521 63.0 1995 1,537.1 9.0 272.6 9.0 6.0 8.0 1,635,2 296.3 5,519 63.0 1996 1,660.0 8.0 294.3 8.0 6.0 8.0 1,766.0 319.9 5,520 63.0 1997 1,792.8 8.0 318.0 8.1 6.0 8.0 1,907.2 345.7 5,518 63.0 1998 1,936.2 8.0 343.3 8.0 6.0 8.0 2,059.8 373.2 5,520 63.0 1999 2,091.1 8.0 370,8 8.0 6.0 8.0 2,224.6 403.0 5,519 63.0 2000 2,258.4 8.0 400.5 8.0 6.0 8.0 2,402.6 435.3 5,519 63.0 Note: Annu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: THEMAG. - 201 - Annex 14 Page 7 of 28 Table 5: NORTHERN SYSTEM - CASE 5 (MISSION "BASE" SCENARIO) Energy and demand forecasts, 1986-2000 Energy Peak demand Transmission Generation Load factor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWh MW % GWh MW Hours % 1986 569.5 101.C 6.0 8.0 605.9 109.8 5,519 63.0 1987 580.9 2.0 103.0 2.0 6.0 8.0 618.0 '12.0 5,519 63.0 1988 592.5 2.0 105.1 2.0 6.0 8.0 630.3 114.2 5,519 63.0 i989 604.4 2.0 107.2 2.0 6.0 8.0 642.9 116.5 5,519 63.0 1990 616.4 2.0 109.3 2.0 6.0 8.0 655.8 118.8 5,519 63.0 1991 628.8 2.0 111.5 2.0 6.0 8.0 668.9 121.2 5,519 63.0 1992 647.6 3.0 114.9 3.0 6.0 8.0 689.0 124.8 5,519 63.0 1993 680.0 5.0 120.6 5.0 6.0 8.0 723.4 131.1 5,519 63.0 1994 734,4 8.0 130.2 8.0 6.0 8.0 781.3 141.6 5,519 63.0 1995 807.9 10.0 143,3 10.0 6.0 8.0 859.4 155.7 5,519 63.0 1996 904.8 12.0 160.5 12.0 6.0 8.0 962.6 174.4 5,519 63.0 1997 1,013.4 12.0 '79.7 12.0 6.0 8.0 1,078.1 195.3 5,519 63.0 1998 1,135.0 12.0 201.3 12.0 6.0 8.0 1,207.4 218.8 5,519 63.0 1999 1,259.8 11.0 223.4 11.0 6.0 8.0 1,340.3 242.9 5,519 63.0 2000 1,373.2 9.0 243.5 9.0 6.0 8.0 1,460.9 264.7 5,519 63.0 Note: Annu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: Mission estimates. - 202 - Annex 14 Page 8 of 28 Table 6: NORTHERN SYSTEM - CASE 6 (MISSION "INTERMEDIATE, SCENARIO) Energy and demand forecasts, 1986-2000 Energy Peek demand Transmission Generation Load tactor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWh MW % % % GWh MW Hours % 1986 569.5 101,0 6.0 8.0 605.9 109.8 5,519 63.0 1987 580.9 2.0 103.0 2.0 6.0 8.0 618.0 112.0 5,519 63.0 1988 592.5 2.0 105.1 2.0 6.0 8.0 630. 114.2 5,519 63.0 1989 610.3 3.0 108.2 3.0 6.0 8.0 649.2 117.6 5,519 63.0 1990 -540.8 5.0 113.6 5.0 6.0 8.0 681.7 123X5 5,519 63.0 1991 672.8 5.0 119.3 5.0 6.0 8.0 715.8 129.7 5,519 63.0 1992 719.9 7.0 127.7 7.0 6.0 8.0 765.9 138.8 5,519 63.0 1993 777.5 8.0 137.9 8.0 6.0 8.0 827.2 149.9 5,519 63.0 1994 855.3 10.0 151.7 10.0 6.0 8.0 909.9 164,9 5,519 63.0 1995 957.9 12.4 169.9 12.0 6.0 8.0 1,019.1 184.7 5,519 63.0 1996 1,072.9 12.0 iMO.3 12.0 6.0 8.0 1,141.3 206.8 5,519 63.0 1997 1,201.6 12.0 2'3.1 12.0 6.0 8.0 1,278.3 231.6 5,5!9 63.0 1998 1,333.8 11.0 236.5 11.0 6.0 8.0 1,418.9 257.1 5,519 63.0 1999 1,453.8 9.0 257.8 9.0 6.0 8.0 1,546.6 280.3 5,519 63.0 2000 1,570.1 8.0 278.5 8.0 6.0 8.0 1,670.4 302.7 5,519 63.0 Note: Annu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: Mission estimates. - 203 - Annex 14 Page 9 of 28 Table 7: NORTHERN SYSTEM - CASE 7 (MISSION ,HIGH SCENARIO) Energy and demand forecasts, 1986-2000 Energy PeaK demand Transmission Generation Load factor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWh I MW % % % GWh MW Hours % 1986 569.5 101.0 6.0 8.0 605,9 109.8 5,519 63.0 1987 590.9 2.0 103.0 2.0 6.0 8.0 618.0 112.0 5,520 63.0 1988 592.5 2.0 105.1 2.0 6.0 8.0 630.3 114.2 5,520 63.0 1989 610.3 3.0 108.2 3.0 6.0 8.0 649.2 117.6 5,520 63.0 1990 659.1 8.0 116.9 8.0 6.0 8.0 701.2 127.0 5,520 63.0 1991 725.0 10.0 128.6 10.0 6.0 8.0 771.3 139.7 5,520 63.0 1992 812,0 12.0 114.0 12,0 6.0 8.0 863.8 156.5 5,520 63.0 1993 909.5 12.0 161.3 12.0 6.0 8.0 967.5 175.3 5,520 63.0 1994 1,019.6 12.0 180.6 12.0 6.0 8.0 1,083.6 196,3 5,520 63.0 1995 1,130.6 11.0 200.5 1.0 6.0 8.0 1,202.8 217.9 5,520 63.0 1996 1,232.4 9.0 218.5 9.0 6.0 8.0 1,211.1 237.5 5,520 63,0 1997 1,331.0 8.0 236,0 8.0 6.0 8.0 1,415.9 256,5 5,520 63,0 1998 1,437.5 8.0 254.9 8.0 6.0 8.0 1,529.2 277.0 5,520 63.0 1999 1,552.5 8.0 275.3 8.0 6.0 8.0 1,651.6 299,2 5,220 63.0 2000 1,676.7 8.0 297,3 8.0 6.0 8.0 1,783.7 323.1 5,520 63.0 Note: Annu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: Mission estimates. - 204 - Annex 14 Page 10 of 28 Table 8: CENTRAL SYSTEM - CASE 1 (BEP STUDY) Energy and demand forecasts, 1986-2000 Energy Peak demand Transmission Generation Load factor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWh % MW % % % GWh MW Hours % 1986 191.0 37.2 6.0 8.0 203.2 40.4 5,025 57.4 1987 203.7 6.7 39.8 6.9 6.0 8.0 216.7 43.2 5,014 57.2 1988 217.3 6.7 42.5 6.9 6.0 8.0 231.2 46.2 5,003 57.1 1989 231.8 6.7 45.5 6.9 6.0 8.0 246.6 49,4 4,991 57.0 1990 247.3 6,7 48.6 6.9 6.0 8.0 263.1 52.8 4,980 56.9 1991 284.2 14.9 55e6 14.3 6.0 8.0 302.4 60.4 5,006 57.1 1992 326.7 14,9 63.5 14.3 6,0 8.0 347.6 69.1 5,032 57.4 1993 375.5 14.9 72.7 14.3 6.0 8.0 399.5 79.0 5,058 57.7 1994 431.6 14.9 83,1 14.3 6.0 8.0 459.0 90.3 5,085 58.0 1995 496.1 14.9 95.0 14.3 6.0 890 527.8 103.3 5,111 58.3 1996 531.3 7.1 101.5 6.9 6.0 8.0 565.2 110.4 5,121 58.5 1997 568.9 7,1 108,5 6.9 6.0 8,0 605.2 118.0 5,131 58.6 1998 609.3 7.1 116.0 6.9 6,0 8.0 648,1 126,1 5,141 58.7 1999 652.4 7.1 124.0 6.9 6.0 8.0 694.1 134.8 5,151 58.8 2000 698.7 7.1 132.5 6.9 6.0 8.0 743.3 144.0 5,161 58.9 Note: Annu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: Mission estimates. - 205 - Annex 14 Page 11 of 28 Table 9: CENTRAL SYSTEM - CASE 2 (THEMAG "LOW" SCENARIO) Energy and demand forecasts, 1986-2000 Energy Peak demand Transmission Generation Load factor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWh % MW % % % GWh MW Hours % 1986 143.0 29.6 6.0 8.0 152.1 32.2 4,728 54.0 1987 151.5 5.9 31.4 6.1 6.0 8.0 161.2 34.1 4,722 53.9 1968 160.6 6.0 33.3 6.1 6.0 8.0 170.9 36.2 4,720 53.9 1989 182.3 13,5 38.3 15.0 6.0 8.0 193.9 41.6 4,659 53.2 1990 210.7 15.6 44.7 16.7 6.0 8.0 224.1 48.6 4,612 52.7 1991 229.6 9.0 48.8 9.0 6.0 8.0 244.3 53.0 4,610 52.6 1992 250.3 9.0 53.1 9,0 6.0 8.0 266.3 57.7 4,612 52.6 1993 272.9 9.0 57.9 9.0 6.0 8.0 290.3 62.9 4,613 52.7 1994 297.4 9.0 63.1 9.0 6.0 8.0 316.4 68.6 4,613 52.7 1995 324.2 9.0 68.8 9.0 6.0 8.0 344.9 74.7 4,615 52.7 1996 350.2 8.0 74.2 8.0 6.0 8.0 372.6 80.7 4,617 52,7 1997 378.1 8.0 80.2 8.0 6.0 8.0 402.2 87.2 4,614 52.7 1998 408.4 8.0 86.7 8.1 6.0 8.0 434.5 94.2 4,610 52.6 1999 441.0 8.0 93.6 8.0 6.0 8.0 469.1 101.7 4,611 52.6 2000 476.4 8.0 101.1 8.0 6,0 8.0 506.8 109.9 4,612 52.6 Note: _nnu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: Mission estimates. - 206 - Annex 14 Page 12 of 28 Table 10: CENTRAL SYSTEM - CASE 3 (THEMAG "BASE" SCENARIO) Energy and demand forecasts, 1986-2000 Energy Peak demand Transmission Generation Load factor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWh % MW % % % GWh MW Hours % 1986 155.4 32.2 6.0 8.0 165.3 35.0 4,723 53.9 1987 164.7 6.0 34.1 5.9 6.0 8.0 175.2 37.1 4,727 54.0 19E8 174.6 6.0 36.2 6.2 6.0 8.0 185.7 39.3 4,721 53.9 1989 198.2 13.5 41.6 14.9 6.0 8.0 210.9 45.2 4,663 53.2 1990 228.9 15.5 48.6 16.8 6.0 8.0 243.5 52.8 4,610 52.6 1991 249.6 9.0 53.0 9.1 6.0 8.0 265.5 57.6 4,609 52.6 1992 272.1 9.0 57.7 8.9 6.0 8.0 239,5 62.8 4,613 52.7 1993 296.6 9.0 62.9 9.0 6.0 8.0 315.5 68.4 4,615 52.7 1994 323.3 9.0 68.6 9.1 6.0 8.0 343.9 74.6 4,613 52.7 1995 352.3 9.0 74.8 9.0 6.0 8.0 374.8 31.3 4,612 52.6 1996 380.6 8.0 80.7 8.0 6.0 8.0 404.9 87.8 4,614 52.7 1997 411.0 8.0 87.2 8.0 6.0 8.0 437.2 94.8 4,613 52,7 1998 443.9 8.0 94,2 8.0 6.0 8.0 472.2 102.4 4,612 52.6 1999 479.4 8.0 101.7 8.0 6.0 8.0 510.0 110.5 4,614 52.7 2000 517,8 8.0 109.9 8.1 6.0 8.0 550.9 119,5 4,611 52.6 Note: Annu: Annui3; Consum: Consumption; Grow: Growth; FV: High voltage. Source: Mission estimates. - 207 - Annex 14 Page 13 of 28 Table 11: CENTRAL SYSTEM - CASE 4 (THEMAG "HIGH" SCENARIO) Energy and demand forecasts, 1986-2000 Energy Peak demand Transmission Generation Load factor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand G*neration Year GWh % MW % % % GWh MW Hours L 1986 167.8 34.8 6.0 8.0 178.5 37.8 4,719 53.9 1987 177.9 6.0 36.8 5.7 6.0 8.0 189.3 40.0 4,731 54.0 1988 188.6 6.0 39.1 6.3 6.0 8.0 200.6 42.5 4,721 53.9 1989 214.1 13.5 44.9 14.8 6.0 8.0 227.8 48.8 4,667 53.3 1990 247.3 15.5 52.5 16.9 6.0 8.0 263.1 57.1 4,610 52.6 1991 269.6 9.0 57.2 9.0 6.0 8.0 286.8 62.2 4,613 52.7 1992 293.9 9.0 62.3 8.9 6.0 8.0 312.7 67.7 4,617 52.7 1993 320.3 9.0 67.9 9.0 6.0 8.0 34Q,7 73.8 4,617 52.7 1994 349,2 9.0 74.1 9.0 6.0 8.0 371.5 80.5 4,612 52.7 1995 380.6 9.0 80.8 9.0 6.0 8.0 404.9 87.8 4,610 52.6 1996 411,0 8.0 87.2 7.9 6.0 8.0 437.2 94.8 4,613 52.7 1997 443,9 8.0 94,2 8.0 6.0 8.0 472.2 102,4 4,612 52.6 1998 479.4 8.0 101.7 8.0 6.0 8.0 510.0 110.5 4,614 52.7 1999 517.8 8.0 1098 8.0 6.0 8.0 550.9 119.3 4,616 52.7 2000 559.2 8.0 118.7 8.1 6.0 8.0 594.9 129,0 4,611 52.6 Note: Annu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: Mission estimates. - 208 - Annex 14 Page 14 of 28 Table 12: CENTRAL SYSTEM - CASE 5 (MISSION "BASE" SCENARIO) Energy and demand forecasts, 1986-2000 Energy Peak demand Transmission Generation Load factor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWh MW % % % GWh MW Hours % 1986 90.0 17.0 6.0 8.0 95.7 18.5 5,181 59.1 1987 91.8 2.0 17.3 2.0 6.0 8.0 97.7 18.8 5,181 59.1 1988 94.6 3.0 17.9 3.0 Z.1 8.0 100.6 19,4 5,181 59,1 1989 99.3 5.0 18,6 4,0 6.0 8.0 105.6 20.z 5,231 59.7 1990 108.2 9.0 20.1 8.0 6.0 8.0 115,1 21.8 5,280 60.3 1991 118.0 9.0 21.7 8.0 6.0 8.0 125,5 23.5 5,329 60.8 1992 129.8 10.0 23.4 8.0 6.0 8.0 138,0 25.4 5,427 52.0 1993 142.7 10.0 25.7 10.0 6.0 8.0 151.8 28.0 5,427 62.0 1994 159.9 12.0 28.6 11.0 6.0 8.0 70.1 31.1 5,476 62.5 1995 183.8 15.0 32.9 15,0 6.0 8.0 '95,6 35.7 5,476 62.5 1996 211.4 15.0 37.8 15.0 6.0 8.0 224.9 41,1 5,476 62.5 1997 236.8 12.0 42.3 12.0 6.0 8.0 251.9 46,0 5,476 62.5 1998 262.8 11.0 47.0 11.0 6.0 8.0 279.6 51.1 5,476 62.5 1999 286.5 9.0 51.2 9.0 6.0 8.0 304.8 55 .7 5,476 62.5 2000 309.4 8.0 55.3 8.0 6.0 8.0 329.1 60.1 5,476 62.5 Note: Annu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: Mission estimates. - 209 - Annex 14 Page 15 of 28 Table 13: CENTRAL SYSTEM - CASE 6 (MISSION "INTERMEDIATE" SCENARIO) Energy and demand forecasts, 1986-2000 Energy Peak demand Transmission Generation Load factor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWh £ MW % S % GWh MW Hours % 1986 90.0 !7.0 6.0 8.0 95.7 !8.5 5,181 59.1 1987 91.8 2.0 17.3 2.0 6.0 8.0 97.7 18.8 5,181 59,1 1988 94.6 3.0 17.9 3.0 6.0 8.0 100,6 19,. 5.'81 59.1 1989 101.2 7.0 18.9 6.0 6.0 8.0 107.6 20.6 5,230 59.7 1990 112.3 11.0 20,8 10.0 6.0 8.0 119.5 22.6 5,278 60.3 [991 125.8 12.0 22.9 10.0 6.0 8.0 133.8 24,3 5,374 61,3 1992 142.1 13.0 25.9 13.0 6.0 8.0 151.2 28.1 5,374 61.3 1993 163.4 15.0 29.3 13.0 6.0 8.0 173.9 31.8 5,469 62.4 1994 188.0 15.0 33.6 15.0 6.0 8.0 200.0 36.6 5,469 62.4 1995 216.2 15.0 38,7 15.0 6.0 8.0 230.0 42.0 5,469 62.4 1996 242.1 12.0 43.3 12.0 6.0 8.0 257.6 47.1 5,469 52.4 1997 271.2 12.0 48.5 12.0 6.0 8.0 288.5 52.7 5,469 62.4 1998 303,7 12.0 54.3 12.0 6.0 8.0 323.1 59.1 5,469 62.4 1999 337.1 11.0 60.3 t1.0 6.0 8.0 358.6 65.6 5,469 62.4 2000 367.4 9.0 65.8 9.0 6.0 8.0 390.9 71.5 5,469 62.4 Note: Annu: Annuai; Consum: Consumption; Grow: Growth; HV: High voltage. Source: Mission estimates. - 210 - Annex 14 Page 16 of 28 Table 14: CENTRAL SYSTEM - CASE 7 (MISSION t1HIGH'. SCENARIO) Energy and demand forecasts, 1986-2000 Energy Peak demand Transmission Generation Load factor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWh % MW % % % GWri MW Hours % 1986 90.0 17.0 6.0 8.0 95.7 18.5 5,181 59.1 1987 91.8 2.0 17.3 2.0 6.0 8.0 97.7 18.8 5,181 59.1 1988 94.6 3.0 17.9 3.0 6.0 8.0 100.6 19.4 5,181 59.1 1989 105.9 12.0 19.8 11.0 6.0 8.0 112.7 21.5 5,228 59.7 1990 125.0 18.0 22.8 15.0 6.0 8.0 132.9 24.8 5,365 61.2 1991 152.5 22.0 27.4 20.0 6.0 8.0 162.2 29.7 5,454 62.3 1992 179.9 18.0 32.3 18.0 6.0 8.0 191.4 35.1 5,454 62.3 1993 206,9 15,0 37.1 15.0 6.0 8.0 220.1 40.4 5,454 6?.3 1994 237.9 15.0 42.7 15.0 6.0 8.0 253.1 46.4 5,454 62.3 1995 266.5 12.0 47.8 12.0 6.0 8.0 283.5 52.0 5,454 62.3 1996 298.4 12.0 53.6 12.0 6.0 8.0 317.5 58.2 5,454 62.3 1997 331,3 11.0 59.4 11.0 6.0 8.0 352,4 64.6 5,454 62.3 1998 361.1 9.0 64.8 9.0 6.0 8.0 384.1 70.4 5,454 62.3 1999 390,0 8.0 70.0 8.0 6.0 8.0 414.9 76.1 5,454 62.3 2000 421.2 8.0 75.6 8.0 6.0 L.O 448.0 82.2 3,454 62.3 Note: Annu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: Mis5ion estimates. 211 Annex 14 Page 17 of 28 Tabte 15: SOUTHERN SYSTEM - CASE 1 (BEP STUDY) Energy and domand forecasts, 1986-2000 Energy Peak demand Transmission Generation Load factor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWh % MW % % GWh MW Hours % 1986 50,2 10.3 6.0 8.0 53.4 11.2 4,770 54.5 1987 53,1 5.8 10.9 5.8 6.0 8.0 56.5 11.8 4,769 54.4 1968 56.1 5.8 11.5 5.8 6.0 8.0 59.7 12.5 4,767 54.4 1989 59.4 5.8 12.2 5.8 6.0 8.0 63.2 13.3 4,766 54.4 1990 62.8 5.8 12.9 5.8 6.0 8.0 66.8 14.0 4,765 54.4 1991 72.0 14.6 14.7 ;3.9 6.0 8.0 76.6 16.0 4,796 54.7 1992 82.5 14.6 16.7 13.9 6.0 8.0 87.8 18.2 4,827 55.1 1993 94.5 14.6 19.0 13.9 6.0 8.0 100.6 20.7 4,858 55.5 1994 103.4 14.6 21.7 13.9 6.0 8.0 115.3 23.6 4,890 55.8 1995 124.2 14.6 24.7 13,9 6.0 8.0 132.1 26.8 4,921 56.2 1996 133,9 7.8 26.6 7.5 6.0 8.0 142.5 28.9 4,936 56.3 1997 144.4 7.8 28.6 7,5 6.0 8.0 153.6 31.0 4,950 56.5 1998 155,7 7.8 30.7 7.5 6.0 8.0 165.7 33.4 4,964 56.7 1999 167.9 7.8 33.0 7.5 6.0 8.0 178.7 35.9 4,978 56.8 2000 181.1 7,8 35.5 7.5 6.0 8.0 192.7 38.6 4,993 57.0 Note: Annu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: Mission estimates. - 212 - Annex 14 Page 18 of 28 Table 16: SOUTHERN SYSTEM - CASE 2 (THEMAG "LOW" SCENARIO) Energy and demand forecasts, 1986-2000 Energy Peak demand Transmission Generation Load tactor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWh % MW % % % GWh MW Hours £ 1986 65.9 13,9 6.0 8.0 70.1 15,1 4,640 53.0 1987 69.8 5.9 14.7 5.8 6.0 8.0 74.3 16.0 4,647 53.1 1988 74.0 6.0 15.6 6.1 6.0 8.0 78.7 17.0 4,643 53.0 1989 78.5 6.1 16.6 6.4 6.0 8.0 83.5 18.0 4,628 52.8 1990 88.5 12.7 18.e 13.0 6.0 8.0 94.1 20.4 4,617 52.7 1991 96.4 8.9 20.5 9.0 6.0 8.0 102.6 22.2 4,614 52.7 1092 105.1 9.0 22.3 9.0 6.0 8.0 111.8 24.2 4,615 52.7 1993 114.5 8.9 24.3 9.0 6.0 8.0 121.8 26.4 4,612 52.6 1994 124.8 9.0 26.5 9.1 6.0 8.0 132.8 28.8 4,607 52.6 1995 136.1 9.1 28.9 9.0 6.0 8.0 144.8 31.4 4,609 52.6 1996 147.0 8.0 31.2 8.0 6.0 8.0 156.4 33.9 4,611 52.6 1997 158.8 8.0 33.7 8.0 6.0 8.0 168.9 36.6 4,612 52.6 1998 171.5 8.0 36.4 8.0 6.0 8.0 182.4 39.6 4,611 52.6 1999 185.2 8.0 39.3 8.0 6.0 8.0 197.0 42.7 4612 52.7 2000 200.0 8.0 42.5 8.0 6.0 8.0 212.8 46.1 4,611 52.6 Note: Annu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: Mission estimates. - 213 - Annex 14 Page 19 of 28 Table 17: SOUTHERN SYSTEM - CASE 3 (THEMAG "BASE" SCENARIO) Energy and demand forecasts, 1986-2000 Energy Peak demand Transmission Generation Load factor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWh % MW % % % GWh MW Hours % 1986 71.6 15,1 6.0 8,0 76.2 16.4 4,641 53,0 1987 75,9 6.0 16,0 6.0 6.0 8.0 80.7 17.l 4,643 53.0 1988 80,4 5.9 17,0 6.3 6.0 88,0 85.5 18.5 4,629 52.8 1989 85.3 6.1 18,0 5.9 6.0 8.0 90.7 19.6 4,638 52.9 1990 96.2 12.8 20,4 13.3 6.0 8.0 102,3 22.2 4,615 52.7 1991 104.8 8.9 22.2 9.0 6.0 8.0 111.5 24.2 4,614 52,7 1992 114.2 9.0 24,2 9.0 6.0 8.0 121.5 26.3 4,613 52.7 1993 124,5 9.0 26.4 9,0 6.0 8.0 132.4 28.7 4,616 52.7 1994 135.7 9.0 29.8 9.0 6,0 8.0 144.4 31.3 4,615 52,7 1995 147.9 9.0 31,4 9.0 6.0 8.0 157,3 34.1 4,613 52,7 1996 159.8 8.0 33,9 8.0 6.0 8.0 170.0 36.8 4,614 52,7 1997 172.6 8.0 36,6 8.0 6.0 8.0 183.6 39,8 4,616 52.7 1998 186.4 8.0 39.5 8,0 6.0 8.0 198.3 43.0 4,616 52.7 1999 201.3 8.0 42,7 8.0 6.0 8.0 214.1 46,4 4,614 52.7 2000 217.4 8,0 46.1 8.0 6.0 8.0 231.3 50.1 4,615 52.7 Note: Annu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: Mission estimates. - 214 - Annex 14 Page 20 of 28 Table 18: SOUTHERN SYSTEM - CASE 4 (THEMAG "HIGH" SCENARIO) Energv and demand forecasts, 1986-2000 Energy Peak demand Transmission Generat on Load factor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWh l MW % % % GWh MW Hours % 1986 77.3 16.3 6.0 8,0 82.2 17.7 4,614 53.0 1987 82.0 6.1 17.3 6.1 6.0 8.0 87.2 18.8 4,639 53.0 1988 86.8 5.9 18.4 6,4 6,0 8.0 92.3 20.0 4,617 52.7 1989 92.1 6.1 19.4 5.4 6.0 8.0 98.0 21.1 4,646 53.0 1990 103.9 12.8 22.0 13.4 6.0 8.0 110.5 23.9 4,622 52.8 1991 113.2 9.0 24.0 9.0 6.0 8.0 120.4 26.1 4,620 52.7 1992 123.3 8.9 26.1 9.0 6.0 8.0 131,2 28.4 4,617 52.7 1993 134.5 9.1 28.5 9.0 6.0 8.0 143.1 31.0 4,621 52.7 1994 146.6 9.0 31.1 9.1 6.0 8.0 156.0 33.8 4,615 52.7 1995 159.7 8.9 33.9 9.0 6.0 8.0 169.9 36.8 4,611 52.6 1996 172.6 8.1 36.6 8.0 6.0 8.0 183.6 39.8 4,616 52.7 1997 186.4 8.0 39.5 8.0 6.0 9.0 198.3 43.0 4,616 52.7 1998 201.3 8.0 42.7 8.0 6.0 8.0 214.1 46.4 4,614 52.7 1999 217.4 8,0 46.1 8.0 6.0 8.0 231.3 50.1 4,615 52.7 2000 234,8 8.0 49.8 8.0 6.0 8.0 249.8 54.1 4,615 52.7 Note: Annu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: Mission estimates. Annex 14 - 215 - Page 21 of 28 Table 19: SOUTHERN SYSTEM - CASE 5 (MISSION "BASE, SCENARIO) Energy and demand forecasts, 1986-2000 _ Energy Peak demand Transmission Generation Load factor Consumn Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWh I MW % % % GWh MW Hours 9 1986 50.0 9.5 8.0 10.0 54.3 10.6 5,149 58.8 1987 50.5 1.0 9.7 2.0 8.0 10.0 54.9 10.8 5,098 58.2 1988 51.0 1.0 10.0 3.0 8.0 10.0 55.4 11.1 4,999 57.' 1989 52.0 2.0 10.2 2.0 8.0 10.0 56.5 11.3 4,999 57. 1990 53.6 3.0 10.5 3.0 8.0 1O.O 58.2 11.7 4,999 57.1 1991 56.3 5.0 11.0 5.0 8.0 10.0 61.2 12.2 4,999 57.1 1992 59.1 5.0 11.6 5.0 8.0 10.0 64.2 12.8 4,999 57.1 1993 63.8 8.0 12.5 8.0 8,0 10.0 69.4 13.9 4,999 57,1 1994 71.5 12.0 14.0 12.0 8.0 10.0 77.7 15.5 4,999 57.1 1995 77,9 9.0 15.2 9.0 8.0 10,0 84.7 16.9 4,999 57.1 1996 84.1 8.0 16.5 8.0 8.0 10.0 91.4 18.3 4,999 57.1 1997 90.9 8.0 17.8 8.0 8.0 '0.0 98,8 19.8 4,999 57.1 1998 98.1 8.0 19.2 8.0 8.0 10.0 106.7 21.3 4,999 57.1 1999 106.0 8.0 20.7 8.0 8.0 10.0 115.2 23.0 4,999 57.1 2000 114.5 8.0 22.1, 8.0 8.0 10.0 124.4 24.9 4,999 57.1 Note: Annu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: Mission estimates. - 216 - Annex 14 Page 22 of 28 Tatle 20: SOUTHERN SYSTEM - CASE 6 (MISSION "INTERMEDIATE" SCENARIO) Energy and demand forecasts, 1986-2000 Energy Peak demand Transmission Generation Load factor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWh % MW z % % GWh MW Hours % 1986 50,0 9.5 8.0 10.0 54,3 10.6 5,149 58,8 1987 50.5 1.0 9.7 2.0 8.0 10.0 54.9 10.8 5,098 58.2 1988 51.0 1.0 10.0 3.0 8.0 10.0 55.4 '1.1 4,999 57.1 1989 52.5 3.0 10.3 3.0 8.0 10.0 57.1 11,4 4,999 57.1 1990 55.2 5.0 10.8 5.0 8.0 10.0 60.0 12.0 4,999 57.1 1991 59.6 8.0 11.7 8.0 8.0 10.0 64.8 13.0 4,999 57.1 1992 65.5 10.0 12.8 10.0 8.0 10.0 71.2 14.2 4,999 57.1 1993 73.4 12.0 14.4 12.0 8.0 10.0 79.8 16.0 4,999 57.1 1994 84.4 15.0 16.5 15.0 8.0 10.0 91.7 13.4 4,999 57.1 1995 93.7 11.0 18.3 11.0 8.0 10.0 101.8 20,4 4,999 57.1 1996 101.2 8.0 19.8 8.0 8.0 10.0 110.0 22.0 4,999 57,1 1997 109.3 8.0 21.4 8.0 8.0 10.0 118.8 23.8 4,999 57.1 1998 118.0 8.0 23.1 8.0 8.0 10,0 128.3 35.7 4,999 57.1 1999 127.5 8.0 24.9 8.0 8.0 10.0 138.5 27.7 4,999 57.1 2000 137.7 8.0 26.9 8.0 8.0 10.0 149.6 29.9 4,999 57.1 Note: Annu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: Mission estimates. - 217 - Annex 14 Page 23 of 28 Table 21: SOUTHERN SYSTEM - CASE 7 (MISSION .tHIhJ1. SCENARIO) Energy and demand forecasts, 1986-2000 Energy Peak demand Transmission Generation Load factor Consum Annu at Annu losses requirements at at HV Grow HV Grow Energy Demand Energy Demand Generation Year GWh % MW % % % GWh MW Hours % 1986 50.0 9.5 8.0 10.0 54.3 '0.6 5,149 58.8 1987 50.5 1.0 9.7 2.0 8.0 10.0 54.9 :0.8 5,098 58.2 1988 51,0 1.0 10.0 3.0 8.0 10.0 55.4 1.1 4,999 57.1 1989 54.1 6.0 10.6 6.0 8.0 10.0 58.8 1.8 4,999 57.1 1990 57.3 6.0 11.2 6.0 8.0 10.0 62.3 12.5 4,999 57.1 1991 65.9 15.0 12.9 15.0 8.0 10.0 71.6 14.3 4,999 57.1 1992 75,8 15.0 14.8 15,0 8,0 10.0 82.1 16.5 4,999 57.1 1993 87.2 15.0 17.1 15.0 8.0 10.0 94.7 19.0 4,999 57.1 1994 96.7 11.0 18.9 11.0 8.0 10.0 105.2 21.0 4,999 57.1 1995 105.5 9.0 20.6 9.0 8.0 10.0 114.6 22.9 4,999 57.1 1996 113.9 8.0 22.3 8.0 8.0 10.0 123.8 24.8 4,999 57.1 1997 123.0 8.0 24.1 8.0 8.0 10.0 133,7 26.7 4,999 57.1 1998 132.8 8.0 26.0 8.C 8.0 10.0 144.4 28.9 4,999 57.1 1999 143.5 8.0 28.1 8.0 8.0 10.0 155.9 31.2 4,999 57.1 2000 154.9 8.0 30.3 8.0 8.0 10.0 168.4 33.7 4,999 57.1 Note: Annu: Annual; Consum: Consumption; Grow: Growth; HV: High voltage. Source: Mission estimates. - 218 - Annex 14 Page 24 of 28 BEP and THEMAG Studies 6. Projections for the period 1986-2005 made by both consultant groups are based on the following assumptions: (a) BEP assigns an important consumption to the Viana substation (increasing from 9% to 16% of Luanda consumption over the 1986- 2005 period) while for THEMAG such consumption is negligible; (b) demand at the Mabubas substation is neglible for BEP and about 10% of that of Luanda in the THEMAG projection; (c) THEMAC allows for an increase in the demand connected to Malanga between 1986 and 1990, due to a requirement of 15 MW at the Capanda site during building. BEP does not consider such a requirement; (d) BEP assumes Uige is supplied by the Southern System from 1986 while THEMAG considers its connection in 1995, with a nigher demand; (e) BEP assumes a doubling of consumption by Lobito and Benguela between 1990 and 1995. THEMAG limits the increase to 55% for the same period; and (f) THEMAG assumes a significant consumption at Jamba (Matala) corresponding to a resumption of iron-mining activities in Cassinga from 1986. 7. Up to the year 2000, the mission has outlined three scenarios for the evolution of demand on the three main Systems: "Base", "High", and "Intermediate". As no substantial changes are expected in the short- term, all scenarios use MEP projections for 1987 and 1988 (with minor corrections). After that date, the "Base" scenario assumes a decrease of acts of sabotage but the continuation of economic difficulties, with economic recovery beginning in 1993. The "High" scenario assumes a cessation of sabotage in 2-3 years with improved economic conditions facilitating a quicker recovery with high growth rates of electricity consumption. The "Intermediate" scenario assumes a restoration of peace before 1990 with a slower rate of economic recovery than in the "High" scenario. 8. Different rates of growth as well as load factors were assumed for the three systems to take into account different load characteristics, the relative predominance of industrial loads and higher-income residential loads, existing trends and, absolute values in - 219 - Annex 14 Page 25 of 28 1974 (just before independence), present constraints on supply facilities and realistic schedules for their rehabilitation and, when available, potential demand of previously high industrial load whose resumption is linked to the general economic recovery by Angolan authorities. With no reliable information existing on realistic operational investment programs, the scenarios do not anticipate new, sizeable, discrete industrial loads. 9. Mission projections were based on overall general assumptions and on assumptions for each of the three main systems: General Assumptions Ci) No significant load increase, under all scenarios, until 1989. The general disarray of the economy, particularly foreign currency constraints, prevents industrial recovery on the demand side as well as rehabilitation or extension of the distribution networks. Repressed residential demand will not be met until that date. (ii) Load (Energy and Demand) is calculated and projected at HV busbars of main substations and includes losses in MV distribution networks. (iii) Losses in transmission systems (220, 150 and 150 kV and 60 kV in the Southern System) are arbitrarily set at 6% for energy and 8% for peak demand (% of generation) for all the systems. (iv) Influence of possible HV interconnections among systems is not considered. (v) Supply constraints are assumed to be only those existing in 1987. The timing of rehabilitation projects to reduce these constraints has been tentatively scheduled. No additional major supply constraint is envisaged. Northern System (i) Load factor at generation assumed to be around 0.63 (5,500 hours). (ii) Energy and Demand, at HV level, have the same annual growth rates. - 220 - Anaex 14 Page 26 of 28 (iii) "Low" scenario assumes a low capacity of ZDEL to develop the Luanda distribution network and no further HV transmission net extensions until 1992. Load grows at 2% until 1991 with substantial growth rates of 8% taking place only after 1994. (iv) "High" scenario is associated with high growth rates, starting with 8% in 1990, increasing to 12% between 1992- 94 and stabilizing at 8% from 1997 onwards. It assumes development and rehabilitation of the Luanda network, recovery of industrial loads, and the possibility of a gradual supply of repressed demand. Central System Si) Energy growth rates in initial years are higher than those of demand due to a normalization of supply from the system. Annual load factor increases from 0.59 (in 1986) to nearly 0.625 in a variable number of years depending on the scenario. (ii) Energy consumption and generation value for 1986 and 1987 are higher than those presented by MEP. A large share of captive diesel sets (about 15 CWh) is considered. (iii) "Low" scenario implies a slow recovery of consumption and longer supply constraints. Pre-independence levels of 1974 are obtained only by 1994. (iv) "High" scenario assumes supply constraints are overcome by 1990 and consumption resumes quickly with growth rates of between 18-22% for the 1990-92 period. (v) The "Intermediate" scenario assumes recovery beginning in 1990, developing at a slower pace than in the "High" scenario, and lasting longer. Southern System (i) Load factor remains constant around 0.57 (5,000 hours). (ii) Energy and Demand growth rates are similar. (iii) Different assumptions on the resumption of iron-mining activities at Cassinga account for notable distinctions - 221 - Annex 14 Page 27 of 28 among the three scenarios. In the "Base" scenario, recovery starts in 1991 with increases of 1.2 MW per year. Recovery also begins in 1991 for the "High" scenario and increases by about 2 MW per year. The "Intermediate" scenario accounts for a similar development but at a slower pace. Iron mining in Cassinga may not resume, in which case all scenar.os are over-estimated. The mines have been depleted of higher quality ores. A "resumption" would essentially mean a fresh mining venture utilizing new installations yielding a different output. 10. A comparison of Energy and Peak Demand generation requirements for typical years for the seven scenarios presented is shown in Table 22. Table 22: ENERGY DEMAND FORECASTS (1986-2000) Comparison of Generation Requirements in 7 Scenarios Energy (GWh) Peak Demand (MW) THEMAG Mission THEMAG Mission Year BEP Low Base High Base Interm High BEP Low Base High Base Interm High NORTHERN SYSTEM 1986 505 571 621 671 606 606 606 91.8 102.6 111.5 120.4 109.8 109.8 109.8 1987 554 638 693 749 618 618 618 102.3 113.8 123.7 133.6 112,0 112.0 112.0 1990 732 868 943 1,018 656 682 701 141.1 154.2 167.6 181.0 118.8 123.5 127.0 1995 1,278 1,393 1,514 1,635 859 1,019 1,203 244.0 252.4 274.3 296.3 155.7 184.7 217.9 2000 1,876 2,047 2,225 2,403 1,461 1,670 1,784 365.0 370.8 403.0 436.3 264.7 302.7 323.1 CENTRAL SYSTEM ra 1986 203 152 165 179 96 96 96 40.4 32.2 35.0 37.8 18.5 18.5 18.5 1987 217 161 175 189 98 98 98 43.2 34.1 31.1 40.0 18.8 18.8 18.8 1990 263 224 244 263 115 120 135 52.8 48.6 52.8 57.1 21.8 22.6 24.8 1995 528 345 375 405 196 230 284 103.3 74.7 81.3 87.8 35.7 42.0 52.0 2000 743 507 551 595 329 391 448 144.0 109.9 119.5 129.0 60.1 71.5 82.2 SOUTHERN SYSTEM 1986 53 70 76 82 54 54 54 11.2 15.1 16.4 17.7 10.6 10.6 10.6 1987 57 74 81 87 55 b5 55 11.8 16.0 17.4 18.8 10.8 10.8 10.8 1990 67 94 102 III 58 60 62 14.0 20.4 22.2 23.9 11.7 12.0 12.5 1995 132 145 157 170 85 192 115 26.8 31.4 34.1 36.8 16.9 20.4 22.9 2000 193 213 231 250 124 150 168 38.6 46.1 50.1 54.1 24.9 29.9 33.7 Source: BEP, THEMAG, and mission estimates. m m oQH x 0 - 223 - Annex 15 Page 1 of 11 ANGOLA - ELECTRICITY TARIFF SYSTEM 1. Since the 1960s, electricity tariffs in Angola have changed only in newly electrified urban areas. In these areas of the country rates have not aLtered since the initiation of service. As a result, variation in tariffs of the order of 3:1 can often be observed in areas served by the same interconnected system. High levels of demand from existing consumers in traditional urban centers are derived from what are perceived as low price levels. Low-voltage Tariffs 2. A complex and widely varying declining block structure prevails for low-voltage tariffs with higher rates for residential consumers than for industrial ones. In greater Luanda, low-voltage tariffs, set in 1962, discriminate between residential, commercial, industrial, and public lighting uses, as follows: (a) residential and commercial tariffs have a declining block structure with block sizes (12 classes) related to house sizes or floor area and minimum monthly bills related to meter caliber (9 classes). The average price of electricity sold is Kz 0.87/kWh; (b) low-voltage industrial tariffs cover 2% of sales and are three- period, time-of-day tariffs with energy rates declining with load factor and meter caliber. The average price of eLectri- city soLd is Kz 0.55/kWh; and (c) a flat rate of about one-half the average of other low-voltage sales is charged for public lighting. High-voltage Tariffs 3. High-voltage tariffs include block rates sized as a function of non-coincident peak demand, leading to a decline in average prices with load factor. The billing variables used are non-coincident peak demand and active and reactive energy. The tariffs function as follows: (a) demand rates increase with load factor; (b) in Luanda, the marginal price for active energy declines from Kz 1.1/kWh (load factor of 12.5%) to Kz 0.85/kWh (load factor of 25%), and Kz 0.505/kWh (load factor of 100%); and (c) reactive energy is charged through a monthly bill multiplier- free up to 60% of active energy (cos 0 = 0.8) and charged at a price rising to 63% of the normal rate upon reaching 92% of active energy. - 224 - Annex 15 Page 2 of 11 4. Marginal costs in the Northern System are determined by generation and transmission capacity requirements to meet peak demand, and by local grid capacity requirements determined by maximum demand. The present tariff structure cannot be adjusted to the pattern of marginal costs and a step-by-step process is suggested to attain a structure of tariffs more closely related to the Long-Run Marginal Cost (LRMC). 5. To illustrate the complexity of the tariff system, data collected from the Empresa Nacional de Electricidade (ENE), the Sociedade Nacional de Estudo e Financiamento de Empreendimentos Ultramarinos (SONEFE), and the Empresa de Electricidade de Luanda (EDEL) on tariffs in force in May 1987 is summarized in Tables 1 (LV Residential tariffs), 2 (LV Industrial tariffs), 3 (LV tariffs in EDEL distribution area), and 4 (HV tariffs in the Northern System). Although incomplete, the data cover more than 70% of sales and are adequate to identify the major issues. Table 1: RESIDENTIAL TARIFFS - 1987 (1) (2) (3) (4) (5) Province Cabinda Uige Malanga Kwanza Norte Luanda Municipality Cabinda Cangola Malanga Bula Atumba Luanda, Viana Interconnected system - Northern Northern Number of consumers 5,993 a/ .... .... .... 55,551 b/ Enforced since Before 1965 1973 ...X 1962 Tariff structure Declining block Decl;ning block Declining block Declining block Declining block, with minimum with minimum with minimum block size monthly bill monthly bill monthly bill related to house size (12 classes); monthly fixed rate related to meter size (up to 9 classes) Block sizes kWh/ Energy rates KzlkWh 1 100/6.00 30/5.00 30/5.00 15/6.00 8 to 54/2.50 2 100/3.00 70/4.00 70/4.00 15/5.00 14 to 82/1.50 3 /2.00 150/3.50 150/3.50 70/4.00 /0.70 4 /3.00 /3.00 150/3.50 5 /3.00 6 Minimum monthly demand kWh 15 10 15 3 to 10 Notes Applicable to Applicable to Applicable to Monthly fixed rate non-industrial non-industrial non-industrial 0.12 to 5 Kz/kVA consumers consumers consumers Average price Kz 0.83/kWh 58% of EDEL energy sales Dq X a/ Totdl nuiber of LV consumers (1986). w b/ 1983. 0 tn Source: ENE, SONEFE, EDEL. Table 2: RESIDENTIAL TARIFFS - 1987 (6) (7) (8) (9) (10) Province Kwanza Sul Benguela Benguela Huambo Bie Municipality Libolo Bolambo Cubalga Caala Interconnected system .... Central - Number of consumers *0.* .... ,.. 680 a/ 3,300 a/ Enforced since 1973 1973 1974 . .*. Before 1971 Tariff structure Declining block Declining block Declining block Declining block Declining block with minimum with minimum with minimum monthly bill monthly bill monthly bill Block sizes kWh/ Energy rates Kz/kWh 1 20/4.50 30/6.00 30/4.00 25/4.00 20/5.00 2 30/3.00 70/4.00 /3.50 25/3.50 20/4.80 3 /2.00 150/3.00 50/3.20 20/4.60 4 /1.50 /2.20 20/4.40 5 20/4.20 6 /4.00 Minimum monthly demand kWh 20 10 5 Notes Applicable to Applicable to Applicable to Applicable to Applicable to non-industrial non-industrial non-industriai non-industrial non-industrial consumers. Non- consumers consumers consumers; off-peak off-peak (18 hours) profit institu- (18 hours) available available Kz 1.50 to tions and other at 1.50 to Kw 1.20/kWh 1.20/kWh at Kz 1/kWh Advertising uses charged at Kz 1.00/kWh a! Total number of LV consumers (1986). ° X Source: ENE, SONEFE, EDEL. en o H H' Table 3: RESIDENTIAL TARIFFS - 1987 (11) (12) (13) (14) (15) Province Moxico Namibe Huila Cunene Lunda Sul Municipality Lwena Namibe ..... N'Giva Interconnected system - Southern Southern Number of consumers .... 3,074 a/ .... .... .... Enforced since 1974 .... 1975 1972 1975 Tariff structure Declining block Declining block Declining block Declining block Declining block with minimum with minimum with minimum with minimum wiith minimum monthly bill monthly bill monthly bill monthyly bill monthly bill Block sizes kWh/ Energy rates Kz/kWh 1 30/5.00 30/3.50 30/3.50 30/5.00 30/5.00 2 70/4.00 50/2.75 70/2.50 70/4.00 30/4.50 3 150/3.50 /1.25 /1.50 150/3.50 /4.00 4 /3.00 /3.00 5 6 Minimum monthly demand kWh 15 10 15 15 10 Notes Applicable to Business rate Applicable to Applicable to Applicable to non-industrial Kz 2.00/kWh; non-industrial non-industrial non-industrial consumers consumers consumers consumers consumers Agriculture Kz 0.83/kWh; Kz 2.00-0.80/kWh; Administrations Advertising and non-profit Kz 1.00/kWh organizations entitled to 20% rebate D a/ Total number of LV consumers (1986). Z Source: ENE, SONEFE, EDEL. Table 4: LOW-VOLTAGE INDUSTRIAL TARIFFS - 1987 (1) (2) (3) (4) (5) Province Cabinda Uige Malange Kwanza Norte Kwanza Norte Municipality Cabinda Cangola Malange Bula Atumba N'Dalatando Interconnected system - Northern Northern Number of consumers ,.... O." Enforced since Before 1965 1973 .... .O. Tariff structure Declining block Declining block Declining block Declining block Time of day energy with minimum with minimum with minimum rates, with declining monthly bill monthly bill monthly bill blocks related to load factor; monthly fixed rated to meter size Block sizes kWh/ Energy rates Kz/kWh 00 1 100/3.00 250/3.00 250/3.00 250/3.00 2 400/1.50 /2.50 /2.50 /2.50 3 /1.00 4 5 6 Minimum r nthly demand kWh - 100 100 100 Notes Energy rates: peak 3 hours: Kz 3.00; off- peak, 8 hours: Kz 0.80; other hours: K2.00-0.85 Source: ENE, SONEFE, EDEL. 0 ' Ph. o t; Table 5: LOW-VOLTAGE INDUSTRIAL TARIFFS - 1987 (6) (7) (8) (9) (10) Province Luanda Kwanza Sul Benguela Benguela Benguela Municipality Luanda, Viana Libolo Balombo Cubal Interconnected systcm Northern .... Central Number of consumers 333 a/ .... ... .... .... Enforced since 1962 1973 1960 1973 1974 Tariff structure Time of day energy Declining block Declining blocks Declining block Declining block rates, with declin- with minimum determined by peak with minimum ing blocks related monthly bill demand (6 classes) monthly bill to load factor and load factor monthly fixed rate (4 classes) related to meter size Block sizes kWh/ Energy rates Kz/kWh 1 - 20/4.50 *../l.40 to 0.43 250/3.00 1,000/2.50 2 30/3.00 /1.50 /1.60 3 /2.00 4 5 6 Minimum monthly demand kWh 20 100 Notes Energy rates: peak, Rates applied by 3 hours: Kz 3.00; CELB off-peak, 9 hours: Kz 0.70; other: Kz 1.80-0.75; average price Kz 1.10-1.20. °Qa 2% of EDEL energy _I X sales 0 a/ 1983. F Source: ENE, SONEFE, EDEL. Table 6: LOW-VOLTAGE INDUSTRIAL TARIFFS - 1987 (11) (12) (13) (14) (15) Province Huambo Huambo Moxico Namibe Huila Municipality Caala Caala Lwena Namibe Interconnected system Central Central Sout;iern Southern Number of consumers ...*. .... .... .... Enforced since .... .... 1974 .... 1975 Tariff structure Declining block Declining block Declining block Declining block Declining block, with minumum with minimum with minimum with ,ninimum monthly bill monthly bill monthly bill monthly bibill ° Block sizes kWh/ Energy rates Kz/kWh 1 100/2.50 500/2.00 250/3.00 200/2.00 250/1.20 2 /2.00 500/1.80 /2.50 300/1.50 /1.00 3 500/1.50 /1.00 4 /1.30 5 6 Minimum monthly demand kWh 50 100 100 100 Notes Small industry Other industry HV at Kz 0.80/kWh supplies from supplies from 7.00 to 18.00 hr 7.00 to 18.00 hr e3 : Source: ENE, SONEFE, EDEL. PQ m mn co j -I- Table 7: LOW-VOLTAGE INDUSTRIAL TARIFFS - 1987 (16) (17) Province Cunene Lunda Sul Municipality N'Giva Saurimo Interconnected system - Number of consumers * e Enforced since 1972 1975 Tariff structure Declining block Declining block with minimum with minimum w monthly bill monthly bill Block sizes kWh/ Energy rates Kz/kWh 1 250/3.00 300/4.00 2 /2.50 300/3.00 3 /2.80 4 /3.00 5 6 Minimum monthly demand kWh 100 100 Notes Source: ENE, SONEFE, EDEL. OQ 0 o X I-h F Table 8: LOW-VOLTAGE TARIFFS IN EDEL DISTRIBUTION AREA - 1987 (ENFORCED SINCE 1962) (1) (2 ) &.' (3) (4) Tariff designation General domestic Lighting and other Industrial power Public lighting uses Consumer type Residential Offices, shops Industry - Share in EDEL eaergy sales b/ 58.0% 15.0% 2.0% 2.2% Nunber of consumers b/ 55,551 6.642 333 737 Tariff structure c/ Declining block, block Decliniiig block, bloch Time of day enery rates: Flat rate Kz 0.50 size related to house size related to floor peak: 3 hr, Kz 2.00/kWh; per kWh size (12 classes); area (6 classes); off-peak: 9 hr. Kz 0.70 monthly fixed rate monthly fixes rate per kWh; other hours related to meter size related to meter size Kz 1.80-0.75/kWh (up to 9 classes) (up to 9 classes) declining with load factor and meter size Block size range kWh/ Energy rates Kz/kWh: 1 8 to 54/2.50 30 to 150/2.50 - 2 14 to 82/1.50 220 to 350/1.50 3 /0.70 /0.70 Minimum monthly demand kWh 3 to 10 Related to meter sixe .... Average price b/ in Kz/kWh 0.83 1.06 1.01 0.50 a/ Shop window lighting and advertising uses may be charged by a declining block tariff: 3 consumers, average price Kz 0.88/kWh. Water heating and cooking and air conditioning uses with dedicated meter may be charged by a tariff close to industrial power tariff: 7 consumers, average price Kz 0.88/kWh. b/ 1983. o c/ Monthly fixed rate Xz 0.12 to 5.00/kVA of meter caliber. D X Source: ENE, SONEFE, EDEL. 0 Table 9: HIGH-VOLTAGE TARIFFS IN THE NORTHERN SYSTEM (1) (2) (3) (4) (5) (6) (7) Utility SONEFE SONEF SONEFE SONEFE SONEFE SONEFE, EDEL SONEFE Consumer/s EDEL Fina Petroleos SECiL Siderurgia Siderurgia Other in Luanda, Other, Kwanza Swl Nacional Nacional Benco, and Malenga, Uige Steel Furnace Other uses Kwanza Norte Enforced since .... .... .... .... .... 1965 .... Billing variables 12 month, 15 12 month, 15- 12 month, 15- Active energy: 12 month, coin- 12 month, 15- Declining block minute peak m;nute demand: minute peak W, kWh. cident peak minute peak rates related to demand kW; active P, kW; active demand: P, kW; demand: P1, kW demand: P, kW; 12 month, 15- P, kW; active energy: W, kW; active energy: peak and off- active energy: minute peak demand energy; W, kW; reactive energy. W, kWh. peak energy W, kWh; utilisation (6 classes) and reactive energy. 7 demdnd: Wl, W2 of P: H, hr; reac- load factor (4 clas- ? kWh. tive energy ses); react;ve ener- - Cos 0 gy Cos 0 Rates (in Kz) P: 55.8 P: 62.0 P: 56.4 Pl: 0, MAY/OCT H c 30: P < 50 kW and W: 0.369 W: 0.410 W: 0.373 W: 0.163 P1: 62 NOV/APR P: 39 H < 30 hr: W1= W2:0.410 W: 0 1.40/kWh MAY/OCT H < 90: .... WI: 0.07 P. ) P > 2,000 kW and NOV/APR W: 1 H > 90 hours: 0.43/kWh WI: 0.41 H < 180: NOV/APR P: 45 W: 0.6 H > 180: P. 45 W: 0.39 Reactive energy Cos 0 - 0.8.1.000 Cos 0 = 0.8:1,000 multipliers for .... .... monthly bills Cos 0 = 0.4:1,573 Cos 0 = 0.4:1,573 Notes 25% rebate on Interruptible Demand rate increas- . energy input 4 hours November ing and energy rate x to exported to April decreasing with load 1- I- cement. factor Average price for EDEL Kz 1.2 to to 1.4, 25 to 28% of EDEL sales. Source: ENE, SONEFE, EDEL. - 234 - Annex 16 Page 1 of 19 SUMMARY OF THE CAPANDA [YDRO PROJECT 1. In 1982 the Government reached an agreement with a U.S.S.R./Brazilian consortium for the development of a hydroelectric plant at Capanda on the Kwanza River about 400 km southwest of Luanda. In December of that year, a contract was signed and CAMEK, the Office for the Harnessing of the Middle Kwanza, was created to coordinate and supervise the work. CAMEK was placed under the direct supervision of the Minister of Energy and Petroleum (MEP) and is in charge of coordinating and supervising the project. 2. The main contractors for the Project are the Soviet firm TECHNOPROMEXPORT (TPE) and the Brazilian firm, N. ODEBRECHT. TPE, the team-leader, supervises Lonstruction. It also supervises related Soviet firms and institutes responsible for geological prospecting, dam design, and equipment supply. N. ODEBRECHT is in charge of the civil works and support infrastructure. 3. Future activities to be carried out under the Project are regulated by "Partial Supplement of Services" agreements consisting of additional contracts. From 1985 to mid-1987, ten "supplements" have been signed--one with the Soviet group and nine with the BraziLian group. 4. The Brazilian utility, FURNAS CENTRAIS ELECTRICAS (FURNAS), holds a contract signed in November 1984, estimated at US$65 million for the provision (on secondment) of higher-level staff to GAMEK to deal with administrative, managerial, and technical issues. In mid-1987, FURNAS had 55 persons seconded to GAMEK, and, by end-1986, about half of the contract value was to have been completed. 5. Large infrastructures were required. With the exception of those in Luanda, those at the work sites are required to counter serious security problems. Access to the site required 75 km of paved roads from Cacuso, with a 100-meter strip cleared on either side for security reasons. Two 30 kV lines from Cacuso were also required for energy supply during construction. In addition to normal facilities, heavy investments were required for transport (airports), health and communications. GAMEK estimated that 2,500 people were working on the site in July 1988 and that 4,200 would by the end of 1989. The shortage of housing in Luanda, and the large numbers of foreign technicians, workers and family members required the construction, in the capital city, of a complete residential and administrative center. 6. Work on the project, on site, started in February 1987 and excavations (deviation tunnel) started in April 1988 with deviation of the river scheduled for July 1989. The dam itself would be built after this date and the first group would come on stream in January 1993. It should be noted that measurements show that the high flow always exceeds 1500 m3/s, which requires particular care in building the deviation - 235 - Annex 16 Page 2 of 19 tunnel (and its not being lined could prove risky). Scheduling would also need to be very careful to avoid forced interruption of work during the high flow period. When Cambambe was built, and in spite of the many precautions taken, the work site was flooded three times. 7, Until recently, feasibility studies undertaken within the framework of the old SONEFE expansion plan offered the only available data on the characteristics of the Project and its position in the development of the Northern System. The plan outlined the building of the Capanda hydroelectric plant in three stages interwoven with developments in Cambambe (150 km downstream). (a) a "low" dam only for regulatory purposes, creaying a reservoir with a total volume of 900 million m , and raising the minimum guaranteed flow at Cambambe from 130 to 250 m3/s; (b) a "high" dam (height raised roughly 30 meters) creating a reservoir with a total volume of 3,400 million m3 and guaranteeirng a minimum flow in Cambambe of around 350 m/s; and (c) a power plant with an installed capacity of 440 MW (4 x 110 MW). 8. The Government's ostensible decision to develop the middle course of the Kwanza River took priority over the further development of Cambambe. Under the present Project, the three stages are combined. A dam and power plant are to be built at the same time as the dam is being raised to its final height. Finally, and without the help of optimization studies, the capacity to be installed was raised to 520 MW (4 x 130 MW). According to information received from GAMEK in October 1988, the dam will be of 3gravity type, 110 m highs with a storage capacity of 3700 million m (live storage of 2200 m ), a regularized annual flow of 500 m3/s and an average energy output of 2400 GWh. Project Costs and Investment Program 9. When the decision to proceed with the project was taken, there existed no reasonably firm, recent estimate of the costs of the investment. A figure of about US$1000 million was often quoted for direct costs based on rough updates of old studies without even a design for the dam (which was under discussion at the timel. 10. Only in 1987 did the studies carried out by TPE permitted a decision as to the type of dam and a first estimate of costs. The total direct costs at 1987 prices are estimated at US$1250 million or about $1600 million including financial charges (interest during construction). This total also most probably understates costs as it - 236 - Annex 16 Page 3 of 19 allows neither for physical contingencies nor for the high sums needed for certain costs (food, transport, infrastructures) imposed by the difficult conditions under which the project is being executed. 11. The investment program is described in Table 1. Of the US$1246 million of direct costs, at end-1988 about US$500 million had been spent, or about 40% of the total. By the end of 1989, more than 50% of the funds would have been spent. Total financial charges (including interest during construction) amount to about $360 million and average about US$40 million p.a. in the period 1988-1992, a sum which exceeds all other investments in the power subsector over the same period. Including financial charges, total expenditures at end-1988 would have reached US$570 million, US$750 million by end-1989. 12. The project is being financed essentially from external commercial loans as described below: (a) US$408 million from Banco do Brasil (CACEX) in two tranches, the first tranche of US$308 million granted in December 1984 with a disbursement period of 6.5 years (interest rate of 8%, 7.5 years repayment period starting in June 1991) and a second tranche of $100 million granted in July 1988 to be disbursed before December 1990 (interest rate of 7.15%, a flat, up-front insurance fee of 2%, 7.5 years' repayment period starting in June 1991). Both tranches are guaranteed by petroleum exports under the terms of a contract between SONANGOL and PETROBRAS. Indications in early 1989 were that Banco do Brasil was granting an additional US$120 million and that an additional US$162 million was to be negotiated. Thus, expected financing from Brazil would reach US$690 million. (b) The U.S.S.R. approved a loan of US$275 million to cover engineering design, manufacturing and installation of Soviet-supplied electromechanical equipment. The loan carries an interest rate of 3% and a repayment period of nine years, including three years of grace. (c) Other (unidentified) countries were expected to provide US$105 million through supplies' credits. Until mid-1989, only US$8 million had been obtained, from Finland, on hard commerc;al terms (3 years). (d) The Government of Angola/State Budget was to provide US$170 million, theoretically in domestic currency--however, as financier of last resort, the Angolan State would have to make up any shortfall. 13. So, theoretically, a financing plan exists for the estimated US$1240 million of direct costs (even though there are unfunded financial - 237 - Annex 16 Page 4 of 19 charges during construction of US$360 million). However, as at the end of 1988 only US$691 million was firmly committed, which means that there is a financial gap of US$912 million (or US$555 million, excluding interest during construction). Excluding Soviet financing, cumulative expenditures at the end of 1988 exceeded firm commitments, thereby raising serious questions as to the feasibility of continuing the work on the current schedule. Alternate Expansion Programs 14. Energy production costs at Capanda cannot be assessed precisely because of insufficient and unreliable information. However, based on optimistic forecasts for demand expansion, (requiring 915 GWh in 1992 and 1280 GWh in 1995, in the Northern Grid), the BEP study recommends developing hydro resources according to either of the two following strategies: Strategy 1: Priority to Northern Region; development of Cambambe and Capanda. Strategy 2: Priority to the Central and Southern Regions; development of Lomaum (Units Nos. 4 and 5), Cacombo, Jamba-la-Mina, and Cove. 15. Simultaneous development of both strategies was not recommended as the systems were assumed to be interconnected by 1992. Strategy No. 1 would require two units in Capanda in 1995, followed by small additional capacity in the Central System by the year 2000. Strategy No. 2 would require a simultaneous development of Lomaum and Cove in the Central System and Jamba-la-Mina in the Southern System. The "high" dam in Cambambe would be required in 1995 and only the two first units of Capanda by the year 2005. Strategy No. 2 leads to a net present value of investment and operating costs below that of Strategy No. 1. The relative position of the two strategies is not necessarily robust if demand were significantly lower, especially in the Central Grid (BEP projects a demand of 528 GWh while the mission's "high" scenario is only 284 GWh). 16. The comparison of SONEFE's expansion plan with alternative options done by BEP suggests that the following sequence may be the most economic: (a) "High" dam in Cambambe. A total capacity increase from 180 to 260 MW (due to head increase). The plant is provided with weekly regulating capability while its minimum flow remains 130 m3/s and its firm energy rises from 700 GWh to 1000 GWh. - 238 - Annex 16 Page 5 of 19 (b) "Low" dam in Capanda, and first stage of second power plant at Cambambe (2 x 110 MW). Cambambe is provided with month y regulating capability and its minimum flow rises to 250 m Is. (c) "High" dam in Capanda and second-stage of second power plant at Camb*mbe (2 x 110 MW); minimum flow at Cambambe rises to 350 m Is. (d) power plant at Capanda (4 x 110 MW in old studies or 4 x 130 MW in more recent ones). The above sequencing could be modified to ensure that generation could always, simultaneously satisfy the requirements for firm power (firm energy at the critical day) and for useful power for covering the peak. 17. BEP estimated the investment costs for the above sequence at US$1,030/kW for Cambambe as compared with US$1,800 for Capanda (in 1986 prices). Firm energy capabilities are not accounted for and discounting heightens the relative costs of Capanda if projects are considered alternative rather than complementary. Thus the comparison should not be carried too far. Moreover, unit turbine plus generator costs were rated by BEP, with no explanation, at US$500/kW for Cambambe and at only US$345/kW for Capanda. Estimated Costs of Energy Production at Capanda 18. The cost of energy produced in a power plant within a system can only be estimated correctly by analyzing the whole system and optimizing its operation. In spite of this, order of magnitude estimates of LRMC at Capanda can be obtained. To do this, the investment program (of Table 1) was used, with the addition of a 10% margin on the activities of TPE and ODEBRECHT to account for physical contingencies. A useful life of 50 years was hypothesized (30 years for the equipment and 75 years for the dam might be more appropriate), together with operation and maintenance costs of 2% of baseline investment costs. The first unit is assumed to come on stream January 1, 1993 and the other three at six-month intervals (the last one, on July 1, 1994). Firm energy production of 2400 GWh was assumed (prorated as to entry on stream of units). Discount rate sensitivity analysis was done using 10%, 12% and 15% together with the three demand scenarios developed by the mission. It was assumed that the demand to be satisfied would be that of the Northern Grid alone (which reduces the market for Capanda power over some years). Finally, three levels of generation were assumed at Cambambe, namely 200 GWh/year, 500 GWh/year and 780 GWh/vear. The last figure is the sum of the firm energy produced by Cambambe and Mabubas. In the average year, this easily exceeds 1000 GWh which means that if this energy were given priority, it would postpone even further the possibility of placing Capanda's entire output. - 239 - Annex 16 Page 6 of 19 19. The values obtained are shown in Tables 2 to 7, If Cambambe were to generate 780 GWH using a discount rate of 10%, this would result in LRMC varying between $0.i76 and $0.234 per kWh, depending on the demand scenario (the lower the demand growth, the higher the LRMC, given the discount rate). Using a discount rate of 15% would increase these costs to between $0.365 and $0.547 per kWh. It is obvious that the LRMC of energy at Capanda can be lowered artificially by keeping output at Cambambe artificially low, especially in conjunction with the higher demand growth scenario. In any case, covering costs would require a substantial tariff increase, thereby slowing the growth demand. Supply-Demand Balance 20. The need for higher generation capacity in the Northern Grid-- which would be met from Capanda-- depends on the evolution of demand. Since this is a predominantly hydro system, constraints will arise first in firm energy and then in peak power. Mission demand projections were used to estimate the behavior of the present system, the need for thermal support from the two gas turbines available in Luanda, and the size of the gap between demand and supply. It was assumed that Cambambe's four units wouid be overhauled and that the rehabilitated Mabubas plant would add 10 MW of firm power, corresponding to 7.2 CWh per month of energy. 21. Studies of the Kwanza River flows show that the low flow (exceeded 95% of the time) is about 130-135 m3/s. In its present state, Cambambe allows for good daily regulation on the basis of its flow of 130 m /s on the critical day. So, average firm power on the critical day would be 90 MW and adding 10 MW from Mabubas, firm power of 100 MW is available in permanence. The critical period is really very short (at Cambambe) and the flow normally allows for much higher power than on the critical day. Therefore, a very small thermal supplement would allow the system to maintain much higher firm power 1/. 22. Assuming, conservatively, that one of Cambambe's units is always unavailable (reserve), useful power would be 3 x 45 MW = 135 MW. Thermal generation might therefore be needed for one of two reasons, either to cover a deficit of firm power (daily flow not sufficient to generate energy required that day) or to cover a deficit of useful power (useful available hydro power is not sufficient to cover the daily peak). 23. Thus, without loss of reliability, it would seem logical not to add hydro capacity until total useful power (hydro and thermal) approach system peak. Any energy shortfalls could be covered by generation from the Luanda gas turbines. As shown in Tables 8-11, additional hydro capacity would only become necessary in 1997 (Base Scenario), in 1996 1/ In fact, only in 1958, a very dry year, did the low flow fall lower, to 122 m 1s. It would not be unrealistic to use a higher figure, say, 140 m3/s for the 95% guarantee. - 240 - Annex 16 Page 7 of 19 (Intermediate Scenario) or in 1994 (High Scenario). Thermal support would be needed briefly during the four years before the commissioning of new hydro capacity. This support would vary between 80 and 120 GWh per year (depending on demand) and would cost (1987 prices, no discounting) between $6.1 million and $9.1 million. 24. Under the same assumptions, the maximum requirements for power and energy to satisfy demand to the year 2000 would be 170 MW and 520 GWh (under the high demand growth scenario). Heightening the Cambambe dam inthe early 1990's and a low dam at Capanda around 2000 would amply cover the demand with a high margin of reserve and without thermal support. The investments required would be substantially lower than the cost of Capanda and would allow greater flexibility for matching investments to the actual growth of demand (i.e., reducing the uncertainty and the possibility of making costly mistakes). On the contrary, the Capanda project makes investments long before the emergence of actual demand, with adjustments limited to the timing of commissioning of the generating units. Consequently, all this causes higher costs which will need to be reflected in higher electricity tariffs. Interconnection 25. Under present security conditions, the construction of new transmission lines between systems is impossible. The existing line between Cambambe and Gabela (125 km, 220 KV) is the first stage of a future link between the Northern and Central Systems. It has been out of service since 1984 and security conditions have allowed neither its restoration nor the maintenance of important Central and Southern lines (creating supply difficulties for Porto Amboim). In the meantime regional power supply (Lomaum, Biopio, Huambo, Matala, Namibe) will be rehabilitated and expanded to meet potential demand. The three systems would remain self-sufficient and would have to provide for their own reserve margin. If a protracted war were to continue it would suppress demand growth and prevent construction of interconnections. Thus, a larger electrical market for Capanda would not develop. A short-term cease-fire followed by quick economic recovery would allow and possibly require electrical interconnections, but the least-cost expansion plan would postpone these until near the year 2000, giving preference to previously proposed supply sources in the Central and Southern systems, whose requirements are modest in absolute terms. If an interconnected system with Capanda were to be considered in the short run, problems of reliability, operations, and on-line dispatching could arise. Conclusions 26. At the present stage of development of the power systems of Angola, the construction of the dam and power plant at Capanda would be, at best, a project of marginal economic interest and, at worst, a very heavy financial burden without contribution to revolving the problems of the subsector and the country. Several major issues arise: - 241 - Annex 16 Page 8 of 19 (a) the Project represents an important departure from the least- cost expansion path; (b) the huge capacity (4 x 130 MW) will probably not be needed until the late 1990s (10 to 12 years after project completion) and no additional capacity will be required in the Northern System until then; (c) building the Capanda dam will neither improve the reliability of supply in the Northern System, nor mitigate the problems of the other two systems; and (d) investment in Capanda will substantially add to the public external debt burden and could undermine the financing of the vital petroleum development program, on which future export earnings depend. 22. The Government's goal of taking advantage of Angola's hydro assets cannot be reached without large additional investments in transmission and distribution (together with investments in generntion which are the only ones being consdiered under the Capanda Project). Only two alternatives seem open for the utilization of Capanda power earlier than presently foreseeable demand would justify. These are: (a) attract foreign investment in power-intensive industries that do not require significant developments in the distribution network; and (b) provide generalized access to electricity to a larger propor- tion of Luanda's population and provide incentives to industrial rehabilitation. The first alternative is practically open but a relatively long period of stability is needed for it to materialize. The second alternative would require enormous investments in distribution networks (from HV to LV) that neither the utilities nor the Government can afford. Therefore, the whole issue of Capanda remains closely tied to the absence of a market that could justify the investment and the difficulty of predicting how and where such a market could develop. - 242 - &Rex 16 Page 9 of 19 Table 1: CAPANOA HYDROELECTRIC PROJECT: INVESTMENT PROGRAM (S million) Retf escription 1985-87 88 85 90 91 92 93 94 95-2003 Total N. OOEB1ECHT I Labor 73,740 40,854 47,960 57,963 59,630 37,253 0 0 0 317,400 2 Materlals 89,064 20,530 23,934 26,841 26,754 14,123 0 0 0 201,246 3 Food 12,760 8.236 11,386 11,500 10,748 9,169 0 0 0 63,799 4 Equipment 35,698 8,898 8,900 0 0 0 0 0 0 53,496 5 Freight 41,516 13,212 5,916 5,973 5,973 4,790 0 0 0 77,380 6 Subcontracts 28,109 3,679 7,511 5,150 5,155 2,137 0 0 0 51,741 7 Travel 26,642 8,284 6,223 6,715 6,715 5,244 0 0 0 59,722 8 Othor 20,395 6,630 5,313 5,453 5,287 4,515 0 0 0 46,483 9 Subtotal 327,924 109,112 117,143 119,595 120,262 77,231 0 0 0 871,267 10 Cumulative 327,924 437,036 554,179 673,774 794,036 871,267 0 0 0 871,267 TPE 11 Eng. Design 4,156 7,464 7,460 3,240 3,180 3,160 4,090 3,450 0 36,200 12 Equipment 0 0 130 4,760 65,640 72,000 27,110 140 0 7,177 13 Installatlon 0 0 5,210 13,2AO 9,220 14,210 11,200 4,720 0 57,840 14 Supervislon 0 920 1,500 1,780 1,780 1,780 1,780 1,640 U 11,180 15 Subtotal 4,156 8,384 14,300 23,060 79,820 91,150 44,180 9,950 0 275,000 16 Cumulativo 4,156 12,540 26,840 49,900 129,720 220,810 265,050 275,000 0 275,000 FU.iAS 17 Services 18,981 8,984 6,616 6,616 6,150 5,238 5,238 0 0 57,283 18 Other Exp. 4,424 1,181 348 348 324 276 276 0 0 7,177 19 Subtotal 23,405 10,165 6,964 6,964 6,474 5,514 5,514 0 0 65,000 20 Cumulative 23,405 33,570 47,498 47,498 53,972 59,486 65,000 0 0 65,000 GAWEK 21 Insurance 6,047 3,966 3,993 3,505 3,505 3,505 3,505 3,020 0 31,046 22 Other Exp. 3,500 0 0 0 0 0 0 0 0 3,500 23 Subtotal 9,547 3,966 3,993 3,505 3,505 3,505 3,502 3,020 0 34,546 24 Cumulative 9,547 13,513 17,506 21,011 24,516 28,021 31,526 34,546 0 34,546 25 Dlrect Costs 365,032 131,627 142,400 153,124 210,061 177,400 53,199 12,970 0 1,245,813 26 Cumulative 365,032 496,659 639,059 792,183 1,002,244 1,179,644 1,232,843 1,245,813 0 1,245,813 27 Interest 17,045 37,284 34,248 42,822 41,610 38,086 34,512 29,8'8 62,406 337,861 28 Commission 75 42 109 83 57 44 38 29 0 477 29 Exch. Variation 0 18,733 0 0 0 0 0 0 0 18,733 30 Fin. Charges 17,120 56,059 34,357 42,905 41,667 38,130 34,550 29,877 62,406 357,071 31 Cumulative 17,120 73,179 107,536 150,441 192,108 230,238 264,788 294,665 357,071 357,071 32 Grand Total 382,152 187,686 176,757 196,029 251,728 215,530 87,749 42,847 62,406 1,602,884 33 Cumulative 382,152 569,838 746,595 942,624 1,194,352 1,409,882 1,497,631 1,540,478 1,602,884 1,602,884 Source: GAMEK. Annex 16 - 24 3 - Page 10 of 19 Table 2: CAPANDA: TOTAL COSTS (1987 PRICES) AND YEARLY OUTPUT (Output of Cambambe: 200 GWh) Annual Generation (GWh) Year Expenditures (million US$) Scenarios Nr Year Invest O&M Total Base Intermediate High -8 1985 64 0 64 - - - -7 1986 116 0 116 - - - -6 1987 185 0 185 - - - -5 1988 159 0 159 - - - -4 1989 155 0 155 - - - -3 i990 163 0 163 - - - -2 1991 227 0 227 - - - -1 1992 191 0 191 - - - 0 1993 58 26 84 523 (a)614 (a)684 1 1994 14 26 40 581 710 884 2 1995 0 26 26 659 819 1,103 3 1996 0 26 26 763 941 1,111 4 1997 0 26 26 876 1,078 1,216 5 1998 0 26 26 1,008 1,219 1,329 6 1999 0 26 26 1,140 1,346 1,451 7 2000 0 26 26 1,263 1,470 1,584 8 2001 0 26 26 1,384 1,594 1,706 9 2002 0 26 26 1,504 1,715 1,828 10 2003 0 26 26 1,606 1,830 1,949 11 2004 0 26 26 1,714 1,952 2,078 12 2005 0 26 26 1,829 2,081 2,215 13 2006 0 26 26 1,931 2,195 2,336 14 2007 0 26 26 2,037 2,315 2,400 15 2008 0 26 26 2,149 2,400 2,400 16 2009 0 25 26 2,266 1,400 2,400 17 2010 0 26 26 2,390 2,400 2,400 18 2011 0 26 26 2,400 2,400 2,400 19 2012 0 25 26 2,400 2,400 2,400 20 2013 0 26 26 2,400 2,400 2,400 21 2014 0 26 25 2,400 2,400 2,400 49 2042 0 26 26 2,400 2,400 2,400 50 2043 0 26 26 2,400 2,400 2,400 (a) Output of Cambambe should exceed 200 GWh in that year. Notes: (1) Ecotomic life of project : 50 years, (2) Entry on stream, 1993 (1 unit/six months). (3) 1987 prices, (4) Maximum yearly output of Capanda: 240 GWh. (5) Total Investment $1331 million (information provided GAMEK, 1988)o (6) Yearly O&M : 2% of total investment. Source: Mission estimates. - 244 - Annex 16 Page 11 of 19 Table 3: CAPANDA: TOTAL cosrs (1987 PRICES) AND YEARLY OUTPUT (Output of Cambambe: 500 GWh) Annual Generation (GWh) Year Expenditures (Million US$) Scenarios Nr Year Invest 0&M Total Base Intermediate High -8 1985 64 0 64 - - - -7 1986 116 0 116 - - - -6 1987 185 0 185 - - - -5 1988 159 0 159 - - - -4 1989 155 0 155 - - - -3 1990 163 0 163 - - - -2 1991 227 0 227 - - - -1 1992 191 0 191 - - - 0 1993 58 26 84 223 327 468 1 1994 14 26 40 281 410 584 2 1995 0 26 26 359 519 703 3 1996 0 26 26 463 641 811 4 1997 0 26 26 576 778 916 5 1998 0 26 26 708 919 1,029 6 1999 0 26 26 840 1,046 1,151 7 2000 0 26 26 963 1,170 1,284 8 2001 0 26 26 1,084 1,294 1,406 9 2002 0 26 26 1,204 1,415 1,528 10 2003 0 26 26 1,306 1,530 1,649 11 2004 0 26 26 1,414 1,652 1,778 12 2005 0 26 26 1,529 1,781 1,915 13 2006 0 26 26 1,631 1,895 2,036 14 2007 0 26 26 1,737 2,015 2,162 15 2008 0 26 26 1,849 2,140 2,295 16 2009 0 26 26 1,966 2,272 2,400 17 2010 0 26 26 2,090 2,400 2,400 18 2011 0 26 26 2,219 2,400 2,400 19 2012 0 26 26 2,355 2,400 2,400 20 2013 0 26 26 2,400 2,400 2,400 21 2014 0 26 26 2,400 2,400 2,400 22 2015 0 26 26 2,400 2,400 2,400 49 2042 0 26 26 2,400 2,400 2,400 50 2043 0 26 26 2,400 2,400 2,400 Notes: (1) Economic life of project 50 years, (2) Entry on stream, 1993 (1 unit/six months). (3) 1987 prices. (4) Maximum yearly output of Cspanda: 2400 GWh. (5) Total Investment $1331 million (Information provided GAMEK, 1988). (6) Yearly 0&M : 2% of total investment, Source: Mission estimates. - 245 - Annex 16 Page 12 of 19 Table 4: CAPANDA: TOTAL COSTS (1987 PRICES) AND YEARLY OUTPUT (Output of Cambambe: 780 GWh) Annual Generation (GWh) Year Expenditures (Million US$) Scenarios Nr Year Invest O&M Total Base Intermediate High -8 1985 64 0 64 - -7 1986 1!6 0 116 - - - -6 '987 185 0 185 - - - -5 1988 159 0 159 - - - -4 1989 155 0 155 - - - -3 1990 163 0 163 - - - -2 1991 227 0 227 - - - -1 1992 191 0 191 - - - 0 1993 58 26 84 0 47 188 1 1994 14 26 40 1 130 304 2 1995 0 26 26 79 239 423 3 1996 0 26 26 183 361 531 4 1997 0 26 26 296 498 636 5 1998 0 26 26 428 639 749 6 1999 0 26 26 560 766 871 7 2000 0 26 26 683 890 1,004 8 2001 0 26 26 804 1,014 1,126 9 2002 0 26 26 924 1,135 1,248 10 2003 0 26 26 1,026 1,250 1,369 11 2004 0 26 26 1,134 1,372 1,498 12 2005 0 26 26 1,249 1,501 1,635 13 2006 0 26 26 1,351 1,615 1,756 14 2007 0 26 26 1,457 1,735 1,882 15 2008 0 26 26 1,569 1,860 2,015 16 2009 0 26 26 1,686 1,992 2,155 17 2010 0 26 26 1,810 2,131 2,302 18 2011 0 26 26 1,939 2,276 2,400 19 2012 0 26 26 2,075 2,400 2,400 20 2013 0 26 26 2,218 2,400 2,400 21 2014 0 26 26 2,368 2,400 2,400 22 2015 0 26 26 2,400 2,400 2,400 49 2042 0 26 26 2,400 2,400 2,400 50 2043 0 26 26 2,400 2,400 2,400 Notes: (1) Economic life of project : 50 years, (2) Entry on stream, 1993 (1 unit/six months). (3) 1987 prices. (4) Maximum yearly output of Capanda: 2400 GWh. (5) Total investment $1331 million (Information provided GAMEK, 1988). (6) Yearly O&M : 2% of total investment. Source: Mission estimates. - 246 - Annex 16 Page 13 of 19 Table 5: CAPANDA: LONG RUN MARGINAL COSTS (LRMC) (Yearly output of Cambambe: 200 GWh) Discount Rate Annuities (50 years) i=10% i=12% i=15% i=10% i=12% i=15% Discounted expenditures (USS Million) Investment 1,946.4 2,102.3 2,361.1 196.3 253.2 354.5 Operation and maintenance 283.6 241.8 199.1 28,6 29.1 29.9 Total 2,230.0 2,344.2 2,560.2 224,9 282.3 384,4 Discounted energy (GWh) Base scenario 14,869.9 11,772.7 8,784.4 1,499.8 1,417.6 1,318.9 Intermediate scenario 16,514.0 13,231,9 10,023.4 1,665.6 1,593.9 1,504.9 High scenario 17,598.6 14,224.6 10,902.8 1,775,0 1,712,9 1,636,9 Long Range MargiiGa! Costs (US$/kWh) Base scenario 0,150 0,199 0.291 Intermediate scenario 0.135 0.177 0.255 High scenar;o 0,127 0.165 0.235 Long Range Marginal Costs (Kz/kWh) Base scenario 4.50 5.97 8.74 Intermediate scenario 4.05 5,31 7.66 High scenario 3.80 4.94 7.04 Notes: (1) Project economic lifetime: 50 years. (2) First year of operation: 1993. (3) Base year for discounting/compounding: 1993, (4) Discouat rates: 10%, 12%, and 15%. (5) 1987 pt tes. (6) Insta .J capacity at Capanda: 4 x 130 MW = 520 MW (First Unit 1.1.1993; other units installed every six months), (7) Calculated on the basis of Table 2. Source: Mission estimates. - 247 - Annex 16 Page 14 of 19 Table 6: CAPANDA: LONG RUN MARGINAL COSTS (LRMC) (Yearly output of Cambambe: 500 GWh) Discount Rate Annuities (50 years) 1=10% i=12% 1=15% i=10% i=12% i-15% Discounted expenditures (US$ Million) Investment 1,946.4 2,102.3 2,361.1 196.3 253.2 354,5 Operation and maintenance 283.6 241.8 199.1 28.6 29.1 29.9 Total 2,230.0 2,344.2 2,560.2 224.9 282.3 384.4 Discounted energy (GWh) Base scenario 12,123.6 9,308.1 6,652.5 1,222.8 1,120.9 998.8 Intermediate scenario 13,927.0 10,888.2 7,973.5 1,404.7 1,311,1 1,197.1 High scenario 15,164.0 12,013.8 8,965.4 1,529,4 1,446.7 1,346.6 Long Range Marginal Costs (US$/kWh) Base scenario 0.184 0,252 0.385 Intermediate scenario 0.169 0.215 0.321 High scenario 0,147 0.195 0.286 Long Range Marginal Costs (Kz/kWh) Base scenario 5.52 7.56 11,55 Intermediate scenario 4.05 6.46 9.63 High scenario 3.80 5.85 8,57 Notes: (1) Project economic lifetime: 50 years. (2) First year of operation: 1993. (3) Base year for discounting/compounding: 1993. (4) Discount rates: 10%, 12%, and 15%. (5) 1987 prices. (6) Installed capacity at Capanda: 4 x 130 MW = 520 MW (First Unit 1.1.1993; other units installed every six months). (7) Calculated on the basis of Table 3. Source: Mission estimates. - 243 Annex 16 Page 15 of 19 Table 7: CAPANDA: LONG RUN MARGINAL COSTS (LRMC) (Yearly output of Cambambe: 780 GWh) Discount Rate Annuities (50 years) i=10% i=12% i=15% i=10% 1=12% i=15% Discounted expenditures (US$ Million) Investment 1,946.4 2,102.3 2,361,1 196.3 253.2 354.5 Operation and maintenance 283.6 241.8 199.1 28.6 29.1 29.9 Total 2,230.0 2,344.2 2,560.2 224.9 282.3 384,4 Discounted energy (GWh) Base scenario 9,527.1 7,009.2 4,681.1 960.9 844.0 702,8 Intermediate scenario 11,380.9 8,608.5 5,991.3 1,147,9 1,036.6 899.5 High scenario 12,681.6 9,780.8 7,012.8 1,279.1 1,177.8 1,052,9 Long Range Marginal Costs (US$/kWh) Base scenario 0.234 10.03 16.41 Intermediate scenario 0,196 8.17 12,82 High scenario 0.176 7.19 10.95 Long Range Marginal Costs (Kz/kWh) Base scenario 4 5.97 8.74 Intermediate scenario 4.05 5.31 7.66 High scenario 3.80 4.94 7.04 Notes: (1) Project economic lifetime: 50 years. (2) First year of operation: 1993, (3) Base year for discounting/compounding: 1993. (4) Discount rates: 10%, 12%, and 15%. (5) 1987 prices, (6) Installed capacity at Capanda: 4 x 130 MW = 520 MW (First Unit 1,1,1993; other units installed every six months). (7) Calculated on the basis of Table 4. Source: Mission estimates, Annex 16 Page 16 of 19 Table 8: NORTHERN SYSTEM: ENERGY AND POWER BALANCE, 1988-2000 BASE SCENARIO 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 System Needs (1) Generation (GWh) 630.3 642.9 655.8 668.9 689.0 723.4 781.3 859.4 962.6 1078.1 1207.4 1340.3 1460.9 (2) Peak Power Equivalent Needs (MW) 114.2 116,5 118.8 121.2 124.8 131.1 141.6 155.7 174.4 195.3 218.8 242.9 264.7 (3) Critical Day Generation (313 days) (MWh) 2014 2054 2095 2137 2201 2311 2496 2746 3075 3444 3858 4282 4667 (4) Firm Power Required (7500 hours) (MW) 84.0 85.7 87.4 89.2 91.9 96.5 104.2 114.6 128.3 143.7 161.0 178.7 194.8 (5) Average Daily Flow at Cambambe (m3/s) 121 124 126 129 133 139 151 166 185 208 233 258 281 Total Capacity (Hydro and Thermal) (6) Installed (MW) 236.8 254.6 254.6 254.6 254.6 254.6 254 6 254.6 254.6 254.6 254.6 254.6 254.6 254.6 (7) Useful (MW) 191.8 209.6 209.6 209.6 209.6 209.6 209.6 209.6 209.6 209.6 209.6 209,6 209.6 (8) Firm (MW) 135.4 145.4 145,4 145,4 145.4 145.4 145.4 145.4 145.4 145.4 145.4 145.4 145.4 Hydro Capacity 9) Installed (MW) 180.0 197.8 197.8 197.8 197.8 197.8 197.8 197.8 197.8 197,8 197.8 197.8 197.8 (10) Useful (MW) 135.0 152.8 152.8 152.8 152.8 152.8 152.8 152.8 152.8 152,8 152.8 152.8 152.8 ~0 (11) Firm (MW) 90.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Energy Balance (12) Deficit of Firm Hydro Power (MW) - - - - - - 4.2 14.6 28.3 43.7 61.0 78.7 94.8 (13) Thermal Base Energy Required (GWh) - - - - - - 0.2 2.6 9.3 17.1 25.7 35.2 43.6 (14) Deficit of Useful Hydro Power (MW) -20.8 -36.3 -34.0 -31.6 -28.0 -21.7 -11.2 2.9 21.6 42.5 66.0 90.1 111.9 (15) % of Deficit (Pdef/Peak) (M) - - - - - - - 1.9 12.4 21.8 30.2 37.1 42.3 (16) % Peak Energy (M) - - - - - - _ (0) 0.9 3.4 6.9 10.6 14.9 (17) Thermal Energy Required (Peak) (GWh) - - - - - - - (0) 8.7 36.7 83.3 142.1 217.7 (18) Total Thermal Energy Required (GWh) - - - - - - 0.2 2.6 18.0 53.8 109.0 177.3 261.3 (Base + Peak) Balance of Peak Power (19) Total Deficit of Useful Capacity (MW) -77.6 -93.1 -90.8 -88.4 -84.8 -78.5 -68.0 -53.9 -35.2 -14.3 9.2 33.3 55.1 Scurce: 1973 SONEFE Study; Mission estimates. Annex 16 Page 17 of 19 Table 9: NORTHERN SYSTEM: ENERGY AND POWER BALANCE, 1988-2000 INTERMEDIATE SCENARIO 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Systom reeds (1) Generation (GWh) 630.3 649.2 681.7 715.8 765.9 827.2 909.9 1019.1 1141.3 1278.3 1418.9 1546.6 1670.4 (2) Peak Power Equivalent Needs (MW) 114.2 117.6 123.5 129.7 138.8 149.9 164.9 184.7 206.8 231.6 257.1 280.3 302.7 (3) Critical Day Generation (313 days) (MWh) 2014 2074 2178 2287 2447 2643 2907 3256 3646 4084 4533 4941 5337 (4) Firm Power Required (7500 hours) (MW) 84.0 86.6 90.9 95.4 102.1 110.3 121.3 135.9 152.2 170.4 189.2 206.2 222.7 (5) Average Daily Flow at Cambambe (m3/s) 121 125 131 138 148 159 175 196 220 246 273 298 322 Total Capacity (Hydro and Thermal) (6) Installed (WV) 236.8 254.6 254.6 254.6 254.6 254.6 254.6 254.6 254.6 254.6 254.6 254.6 254.6 254.6 (7) Useful (MW) 191.8 209.6 209.6 209.6 209.6 209.6 209.6 209.6 209.6 209.6 209.6 209.6 209.6 (8) Firm (MW) 135.4 145.4 145.4 145.4 145.4 145.4 145.4 145.4 145.4 145.4 145.4 145.4 145.4 Hydro Capacity 9) Installed (MW) 180.0 197.8 -7.8 197.8 197.8 197.8 197.8 197.8 197.8 197.8 197.8 197.8 197.8 0 (10) Useful (MH) 135.0 152.8 .;'.8 152.8 152.8 152.8 152.8 152.8 152.8 152.8 152.8 152.8 152.8 1 (11) Firm (MW) 90.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Energy Balance (12) Deficit o, Firm Hydro Power (MW) - - - - 2.1 !'.3 21.3 35.9 52.2 70.4 89.2 106.2 i22.7 (13) Thermal Base Energy Required (Gwh) - - - - (0) 1.4 5.8 13.2 21.6 30.9 40.7 49.5 58.1 (14) Deficit of Useful Hydro Power (K") -20.8 -35.2 -29.3 -23.1 -14.0 -2.9 12.1 31.9 54.0 78.8 104.3 127.5 149.9 (15) % of Deficit (Pdef/Peak) (M) - - - - - - 7.3 17.3 26.1 34.0 40.6 45.5 49.5 (16) % Peak Energy (M) - - - - - - 0.3 1.9 4.9 9.2 12.9 18.0 21.5 (17) Thermal Energy Required (Peak) (GWh) _- - 2.7 19.4 55.9 117.6 183.0 278.4 359.1 (18) Total Thermal Energy Required (GWh) - - - (0) 1.4 8.5 32.6 77.5 148.5 223.7 327.9 417.2 (Base e Peak) Balance of Peak Power (19, Total Deficit of Useful Capacity (MW) -77.6 -92.0 -86.1 -79.9 -70.8 -59.7 -44.7 -24.9 -2.8 22.0 47.5 70.7 93.1 Source: 1973 SONEFE Study; Mission estimates. Annex 16 Page 18 of 19 Table 10: NORTHERN SYSTEN: ENERGY AND POWER BALANCE, 1988-2000 HIGH SCENARIO 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Systemi Needs (1) Generation (GWh) 630.3 649,2 701.2 771.3 863.8 967.5 1083.6 1202.8 1311.1 1415.9 1529.2 1651.6 1783.7 (2) Peak Power Equivalent Needs (MW) 114.2 117.6 127.0 139.7 156.5 175.3 196.3 217.9 237.5 256.5 277,0 299.2 323.v (3) Critical Day Generation (313 days) (MWh) 2014 2071 2240 2464 2760 3091 3462 3843 4189 4524 4886 5277 5699 (4) Firm Power Rec'uired (7500 hours) (MW) 84.0 86.6 93.5 102.8 115.2 129.0 144.5 160.4 174.8 188.8 203.9 220.2 237.8 (5) Average Daily Flow at Cambambe (m3/s) 121 125 135 149 166 186 209 232 253 273 295 318 344 Total Capacity (Hydro and Thermal) (6) Installed (MW) 236.8 254.6 254.6 254.6 254.6 254.6 254.6 254.6 254.6 254.6 254.6 254.6 254.6 254.6 (7) Useful (MW) 191.8 209.6 209.6 209.6 209.6 209.6 209.6 209.6 209.6 209.6 209.6 209.6 209.6 (8) Firm (MW) 135.4 145.4 145.4 145.4 145.4 145.4 145.4 145.4 145.4 145.4 145.4 145.4 145.4 Hydro Capacity (9) Installed (MW) 180.0 197.8 197,8 197.8 197.8 197.8 197.8 197.8 197.8 197.8 197.8 197.8 197.8 F (10) Useful (MW) 135.0 152.8 152.8 152.8 152.8 152.8 152.8 152.8 152.8 152.8 152.8 152.8 152.8 (11) Firm (MW) 90.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Energy Balance (12\ Deficit of Firm Hydro Power (MW) - - - 2.8 15.2 29.0 44.5 60.4 74.8 88.8 103.9 120.2 137.8 (13) Thermal Base Energy Required (GWh) - - - (0) 2.8 9.6 17.4 25.6 33.2 40.5 48.3 56.8 66.0 (14) Deficit of Useful Hydro Power (MW) -20.8 -35.2 -25.8 -13.1 3.7 22.5 43.5 65.1 84.7 103.7 124.2 146.4 170.3 (15) % of Deficit (Pdef/Peak) (5) - - - - 2.4 12.8 22.2 29.9 35.7 40.4 44.8 48.9 52.7 (06) % Peak Energy (S) - - - - (0) 1.0 3.5 6.7 9.8 12.7 17.1 20.2 25.0 (17) Thermal Energy Required (Peak) (GWh) - - - - (0) 9.7 37.9 80.6 128.5 179.8 261.5 333.6 445.9 (18) Total Thermal Energy Required (GWh) - - - (0) 2.8 19.3 55.3 106.2 161.7 220.3 309.8 390.4 511.9 (Base + Peak) Balance of Peak Power (19) Total Deficit of Useful Capacity (MW) -77.6 -92.0 -82.6 -69.9 -53.1 -34.3 -13.3 -8.3 -27.9 -46.9 67.4 89.6 113.5 Source: 1973 SONEFE Study; Mission estimates. Annex 16 - 252 - Page 19 of 19 Table 11: NORTHERN SYSTEM : ENERGY AND POWER BALANCE, 1988-2000 Notes to Tables 8, 9, 10 (1) : Demand proj3ctions. (2) :Demand projections. (3) =(1) * 1000/313 output on critical day (working day In October). (4) = (3)/24 or (4) = (1) * 1000/7500 average power on critical day (working day in October). (5) : Average daily flow of Cambambe to guarantee critical day output. (6) * 4 * 45 MW (Cambambe) + 1 * 25.6 MW (GTI Luanda) = 1 * 31.2 MW (GT2 Luanda) from 1989) 17.8 MW (mabubas) available starting 1989. (7) = (6) - 45 MW: Reserve = 1 unit of Cambambe. (8) = 90 MW (Cambambe) + 45.4 MW (80% of GTI+GT2 Luanda) + 10 MW (Mabubas, from 1989) (9) = 4 * 45 MW (Cambambe) + 17.8 MW (Mabubas, from 1989). (10) = (9) - 45 MW: Reserve = 1 unit of Cambambe. (11) 90 MW = power corresponding to minimum yearly flow of 130 m3/s in Cambambe. 10 MW = assumed firm power for Mabubas from 1989. (12) = (4) - (11) (13) taken from 1973 Sonefe study according to (12). (14) (2) - (10). (15) = (14) * 100/ (2). (16) taken from the Energy/Power load curve in the 1973 SONEFE Study. (17) (16) * (1)/100 (18) = (13) + (17) (19) = (2) - (7) Source: 1973 Sonefe study; mission estimates. - 253 - Annex 17 Page 2 of 13 DRAFT TERMS OF REFERENCE: PREPARATION OF A PILOT PROJECT IN HUILA-NAMIBE Background The provinces of Huila and Namibe are in a more favorable position to be able to undertake forest operations than most other provinces in Angola. The military situation is much better, with most of the two provinces safe. The economy in the two provinces seems to be functioning better than in most of the country. Also, the two provinces do not have large population concentrations in areas far from the forests. It is true that the coastal towns of Namibe and Tombwa have virtually no wood resources close by, but the fact that the towns are fairly small makes the problem more easily managed than in the cases of Ludnda and Benguela/Lobito. The supply of woodfuels is also better organized than elsewhere. Considering the above facts, a pilot project in the provinces of Huila and Namiba would seem to be feasible. The pilot project would contain the following items: (a) increased efficiency of stoves (coast and inland); (b) improved woodfuel supply system (coast and inland); (c) fuel substitution (coast); and (d) wood production, mainly consisting of development work in agroforestry, supplemented by training based at the agri- cultural school at Tchivinguiro. The Project would be operated jointly by the MEP (Ministry of Energy and Petroleum) through the DNRFE (Department of New and Renewable Sources of Energy) and the Ministry of Agriculture, through the DNACO (National Directorate for the Conservation of Nature). The project is closely interrelated with and could obtain significant support from other proposed projects, particularly those on Improved Cooking Stoves (ICS) and Improved Woodfuel Supply Systems (WSS). This is true for both project preparation and implementation. The preparation of the pilot project for Huila-Namibe should follow that for the ICS and WSS projects. In that way, project leaders can benefit from the experience gained in preparing these specialized projects. Project Components A team of specialists should prepare an integrated pilot project in the field of biomass energy for the provinces of Huila and Namibe. The team should base its work on & reading of the availabLe literature (especially that on the regional project prepared under UNDP - 254 - Annex 17 Page 3 of 13 financing) and on the work done to prepare the ICS and WSS projects. The team should formulate a coherent pilot project that could be replicated in other areas wherever the security situation is favorable. In preparing the pilot project, the team should consider at least the following technical components: (a) inventory of accessible forest resources, particularly forest plantations; (b) mapping of the present system for supply of woodfuels to the main urban centers (Lubango, Namibe, and Tombwa); (c) analysis of the workings of existing charcoal production -o- operatives and investigation of the possibilities of organizing other producers now operating outside the cooperative; (d) development and implementation of a revised supply system for Lubango, Namibe, and Tombwa; (e) development, introduction and dissemination of improved cooking stoves; (f) establishment of tree nurseries for the creation of windbreaks and other forms of tree planting outside forests; (g) introduction of an elementary course in agroforestry at the agricultural school at Tchivinguiro, near Lubango; and (h) strengthening of DNACO's provincial representation for Huila and Namibe using, if possible, highLy qualified manpower available at the Lubango branch of the Agostinho Neto University. The fuel substitution component listed under "Background" above is not included in the specific Terms of Reference for the project preparation team as the issue of increasing the use of LPG in coastal cities falls somewhat outside the professional competence of a forestry team. The Project Team Areas of specialization of team members should include the following; (a) forest operations for fuel production; (b) Charcoal production; (c) agroforestry; (d) stove improvement; (e) training; and (f) project economics. - 255 - Annex 17 Page 4 of 13 Two or three persons would need to work about five man-months, including three man-months in Angola. Estimated costs for the pilot project are approximately US$100,000 for the study and about US$500,000 for conducting the proposed operations on a pilot scale. - 256 - Annex 17 Page 5 of 13 2. Improved Cooking Stoves Development and distribution of improved cooking stoves has been suggested for several areas in Chapter V. The best strategy is to develop a national project for improved cooking stoves. Initial activities could well be concentrated in Luanda, where the largest single concentration of people using traditional stoves is found. The high prices for woodfuels in Luanda provide an incentive for economizing on the use of fuelwood and charcoal. The Project would be carried out by the DNRFE within the MEP. The DNRFE has already initiated activities in this field as part of its preparations for the first national seminar on firewood and charcoal, which took place in Luanda in Jun 19d7. Terms of Reference for preparation of a project to improve the efficiency of cooking stoves are attached. -- 257 - Annex 17 Page 6 of 13 DRAFT TERMS OF REFERENCE: PREPARATION OF A PROJECT FOR IMPROVED COOKING STOVES During early 1987, partially in preparation for the first national seminar on firewood and charcoal, the DNRFE (Department of New and Renewable Sources of Energy) of the MEP (Ministry of Energy and Petroleum) studied the stoves used in the seven provinces where it had undertaken a survey of the consumption of woodfuels. Although some local variations and innovations were found, the picture that emerged from the study was rather uniform: firewood was burned in ordinary three-stone stoves placed in unsheltered positions, and charcoal was burned in simple square metal stoves. There were also cases where charcoal was burned in three-stone stoves. Three arguments are generally given to support the introduction of improved cooking stoves. They are: (a) use of improved stoves can help reduce the national consumption of fuelwood, thus reducing the pressure on available forest resources and safeguarding the environment (however, it is well-known that the effect of a stove program on national wood consumption is marginal); (b) use of improved stoves reduces a household's need for wood and thus also reduces both the cash cost of procuring woodfuels and the cost in time needed for woodfueL collection; (c) use of improved stoves couLd contribute to an improvement of the environment in the kitchen. While rhe forest administration of a country often stresses the first argument, the ultimate success or failure of a stove improvement program depends on the degree of acceptance of the stoves on the part of the population. This means that the stove improvement program must be designed so that it primarily meets the needs and aspirations of the population concerned. A large number of stove projeccs are being implemented in various countries in Africa. Being a latecomer in this field, Angola can learn from successes and failures elsewhere. For this purpose, it is proposed that one or several study tours be organized for the concerned staff at DNRFE to observe other stove projects. One stove project acknowledged to be successful is the national program in Burkina Faso for the development and introduction of improved stoves for firewood and charcoal. As regards the availability of woodfuels, the capital of Burkina Faso, Ouagadougou, shares many of the characteristics of Luanda. There are also clear differences. Thus, while ch6rcoal is the principal woodfuel used in Luanda, in Ouagadougou firewood is the principal fuel. Further, the urban and peri-urban population in Ouagadougou is thoroughly organized, which makes the task of introducing - 258 - Annex 17 Page 7 of 13 improved stoves easier. Also, the stove models used in Ouagadougou are adapted to the particular social and dietary conditions prevailing in that city and are made to fit a number of standard cooking pot sizes. However, in spite of the differences between the two cities in this regard, information useful to Angola could certainly be obtained from the program in Burkina Faso. DNRFE's 1987 survey of biomass cooking stoves in seven provinces of Angola can be viewed as an appropriate, if modest, beginning of a national program for the design, production and dissemination of improved cooking stoves in Angola. Such activities have been proposed as a matter of priority in both Luanda, Benguela/Lobito, and Huambo. It seems appropriate that much of the development work should be carried out at the national level or at least with strong technical support at that level. It is, further, believed that field activities in such a program for improved stoves should initially be concentrated in Luanda, where a large number of people use inefficient stoves to burn woodfuels that are bought at quite high prices. Benguela/Lobito would get second priority in the program and Huambo third. Stove development should also be included in the pilot project in Huila-Namibe, principally in the city of Lubango. Project Components The preparation of an enlarged project for improved cooking stoves should probably start with a study visit to an ongoing successful stove project elsewhere in Africa. As mentioned above, Ouagadougou seems to be a suitable location. However, other locations should also be considered. The study tour should be organized for both the Angolans potentially concerned by the project, mainly in DNRFE, and the expatriate specialists expected to take part in project preparation. The second step in project preparation would be a search for suitable stove models and competent stove producers (existing or potential) in Angola. As a third step, the project preparation team would design an expanded project for development and introduction of improved cooking stoves in Angola, to be centered initially in Luanda. The Project Team The expatriate member or members of the project preparation team would mainly investigate the general field of stove projects, particularly in Africa. Knowledge of wood and charcoal combustion and of the technical design of stoves for these fuels is also needed. One or two expatriates would probably be needed. The expatriates would need to spend about three man-months in project preparation, including arrangements for and participation in the study tour mentioned above. - 259 - Arnex 17 Page 8 of 13 Costs Expatri.ate experts (3 man-months) $ 35,000 Expenses 10,000 Study tour (3 Angolans, 2 expatriates) 25,000 Materials and other 15,000 85,000 Importation of 1,000 stove kits 15,000 $100,000 - 260 - Annex 17 Page 9 of 13 3. Improved Woodfuel Supply System In Chapter 5, the need for an improved system for the supply of woodfuels to the major cities both on the coast and in the inland areas of Angola was discussed. A project should be undertaken to analyse the present woodfuel supply system for Luanda, Benguela-Lobito, and Huambo and to develop ways to improve the system. The Project would be organized by the DNACO within the Ministry of Agriculture. Terms of Reference for preparations for the project are attached. - 261 - Annex 17 Page 10 of 13 DRAFT TERMS OF REFERENCE: PREPARATION OF A PROJECT FOR IMPROVED WOODFUEL SUPPLY SYSTEMS Background Improved woodfuel supply systems are needed in Angola, particularly for Luanda, Benguela-Lobito, and Huambo. An improvement of the presently fairly well-functioning system for Lubango and NamLbe would also be included in the pilot project proposed for Huila-Namibe. Initially, however, efforts should probably be concentrated on Luanda. The present system for supply of woodfuels to Luanda is tiot welh known, or at least not well documented. There are uncertainties as to the geographical origin of the firewood and charcoal being used in Luanda, the organization of production, and the distribution network. Many of the present operations are carried out by individuals operating outside the formal, legal framework. Better knowledge of the present systenm is a necessary condition for any action aimed at improving the system. While no inventory of Angola's forests exists, the general extent of the forest resources in Bengo and Kwanza Norte is known. Efforts should be devoted to finding an area of natural forest able to supply a significant part of the fuel needed in Luanda. The area should then be allocated to firewood and charcoal producers, who would be obliged to follow certain management principles, primarily to ensure a successful natural regrowth of che forest cover following cutting of the wood. The main threats to such regrowth are probably grass and bush fires. Therefore, the need for fire protection must be considered. Particular attention should be devoted to finding suitable organizational forms for the production and marketing of firewood and charcoal. At present, owners of the means of transport occupy a key position in exploitation of the wood, its conversion to firewood and charcoal, and transportation of these fuels to the urban areas. Licenses to cut wood are often issued to truck owners, who have to show them in order to pass road check points with a load of firewood or charcoal. The truck owners then engage laborers to cut the wood and produce the charcoal. DNACO (National Directorate for the Conservation of Nature) is responsible for issuing the licenses. Given DNACO's acute shortage of qualified field staff, it is difficult to ascertain whether a given load of woodfuels has been produced in accordance with a given license. In certain areas of the country, for example in the province of Namibe, local laborers have formed cooperatives, which apply for cutting licenses, produce the firewood and charcoal, and transport the fuels to the centers of consumption. Such a cooperative forms a very different management unit than that of a truck owner using hired laborers. As the cooperative is formed on a geographical basis, its members are likely to - 262 - Annex 17 Page 11 of 13 value the sustained production of wood, as this will secure them continuous employment in the area they can cover. They could thus be expected to respect certain restrictions regarding exploitation, such as minimum diameters for cutting, prohibitions against cultivating exploited areas, obligations to carry out reforestation, etc. The truck owner hiring local laborers would seem to have much less to gain from following such rules. A limited number of cooperatives, each working within a set geographical area, are also fairly easy to supervise. If this reasoning is valid, it would be advisable to study whether woodfuel supply coopera- tives could also be successfully established around Luanda. In designing the woodfuel supply system for Luanda, similar systems operating in other countries in Africa can serve as useful examples. One such case is the system for supply of charcoal to Mogadishu in Somalia. The wood used is an Acacia species. It grows quite slowly (less than half a cubic meter per hectare per year) but presently covers a large area, about a million hectares. The wood is transported to Mogadishu from an average distance of about 300 km, mainly over a rather poor asphalt road. The charcoal is produced by "icooperatives" (in reality teams of workers employed by an entrepreneur), which have licenses covering 25 square kilometers. This area lasts three to seven years for the team. As the forest stands are not cut too heavily cut at thy time of exploitation, they are able to recover in about 20 years, when they can be exploited again. The Mogadishu system is well integrated into the country's legal and fiscal framework. Fees are paid for Licenses and further fees are paid by the truck owners, based on the weight of the charcoal entering Mogadishu. Although some charcoal does enter Mogadishu outside this organized system, the amounts are generally believed to be small. While there are a feq key similarities between the supply system of Mogadishu and Luanda, such as the shortage of woodfuels close to the city, the arid conditions, the long distances from forested areas, and the existence of large natural forest stands, there are also obvious differences. For example, in Angola the institutional framework for a supply system of the Mogadishu type is lacking. Further, the present security situation both in the areas potentially suitable for exploitation and along the roads leading to Luanda makes organization of a functioning, unified supply system unfeasible. Nevertheless, it is quite possible that a study of the Mogadishu system (or another system showing key similarities to Luanda) could provide inspiration for the design of an improved system for Luanda. Project Components Project preparation would proceed as follows: Phase I: Study of present system, probably including: - mapping of the present supply routes - survey of present raw material areas - 263 - Annex 17 Page 12 of 13 - study of the generation process in cut-cover areas - mapping of the present organizational set-up. Phase II: Study of a functioning woodfuel supply system for another major city in Africa. Phase III: Development of an improved system for Luanda. Most of the work in the first phase of project preparation would probably have to be done by the Angolan organizations concerned, primarily by DNACO. During Phases II and III, however, expatriates acquainted with well-functioning systems for supply of woodfuels to major cities elsewhere in Africa might provide valuable additional knowledge. Specification of the expatriate knowledge required would probably have to await the completion of Phase I outlined above. About four man-months of expatriate inputs could, however, be put forward as the minimum for project preparation as outlined above. The cost of the above expatriate manpower, together with travel, subsistence and support would be about US$75,000. - 264 - Annex 17 Page 13 of 13 4. Domestic Fuel Substitution This Project would aim at the gradual susbtitution of LPG for firewood and charcoal for use as a domestic cooking fuel. It would be directed mainly at the major urban concentrations on the coast, namely Luanda, Benguela, and Lobito. Needed for the project is mainly increased production and distribution of LPG and gas containers. In order to make it possible for poor families to use LPG, which is a much cheaper fuel than either firewood or charcoal, a simple short-term credit scheme may prove to be essential. This Project would be carried out by the appropriate units within the MEP and SONANGOL as it falls outside the competence of organizations primarily engaged in biomass energy. - 265 - Annex 18 Page 1 of 6 JOINT UNDP/WORLD BANK ESMAP PROGRAM ANGOLA: POWER SUBSECTOR INVESTMENT REVIEW AND UPDATING OF THE LEAST COST EXPANSION PLAN FOR THE NORTHERN SYSTEM 1. The Energy Assessment recently concluded for Angola by the Joint UNDP/World Bank Energy Sector Management Assistance Program (ESMAP) identifies investment planning in the electric power subsector as a major unresolved issue in the development of the energy sector as a wl. uie. Angola needs assistance in designing an investment strategy including rehabilitation, resumption of deferred maintenance, and new facilities in the context of a country with three separate grids facing great uncertainties in the pace of power demand growth in both space and time. Electricity Supply 2. Electricity supply in Angola is the responsibility of two companies: ENE and SONEFE (Sociedade Nacional de Estudo e Financiamento de Empreendimentos Ultramarinos). ENE is a State enterprise, created in 1980 and intended to become the sole national power utility in charge of generation, transmission, and medium voltage distribution all over the country. The company is currently operating the "Central" and "Southern" Systems and several isolated systems. SONEFE is in charge of generation and transmission in the "Northern" System, the largest system in the country, and supplies about 300 clients directly at high voltage (60 kV) and medium voltage. Distribution in the area of Luanda is the responsibility of EDEL (Empresa de Electricidade de Luanda). Low voltage distribution in the rest of the country is sometimes the responsibility of ENE but is often handled by loca' municipal bodies (Comissariados) who may also own small captive diese'l sets. Overall Generating Conditions 3. Total installed capacity in ENE and SONEFE plants is approximately 463 MW. Of the total capacity, 287 MW is in hydro units, 102 MW in gas turbines, and 74 MW in diesel sets. In 1987 available capacity reached 275 MW (59% of total) but there were severe constraints on thermal units due to difficulties in fuel supply. rhe two gas turbines in Luanda (56.8 MW) burn Jet B fuel while the gas turbine in Cabinda (12.3 MW) runs on natural gas. The two remaining gas turbines, in Biopio (22.8 MW) and Huambo (10 MW), run on diesel oil. Annual - 266 - Annex 18 Page 2 of 6 electricity generation in Angola peaked at 1,029 GWh in 1974, with 858 CWh (83.4%) from hydro origin. After a sharp decline in the years following independence, generation resumed growth but experienced a second decline over the period 1983-85 and is still below the values of 1974. In 1986, total generation was 754 GWh with 691 (91.7%) from hydro plants. 4. Electricity supply in Angola consists of three separate grids and numerous isolated systems. The three main systems are associated with the basins of three important rivers: the Kwanza for the Northern System; the Catumbela for the Central System; and the Cunene for the Southern System. These systems supply the main load centers in Angola: Luanda (in the Northern System); Benguela, Lobito, and Huambo (in the Central System); and Lubango and Namibe (in the Southern System). The main isolated systems are those of Cabinda, Uige, and Bie. Another important system in the province of Lunda Norte belongs to the mining company ENDIAMA (Empresa Nacional de Diamantes de Angola) and was mainly used for diamond mining activities. 5. Hydro has always been the main supply source. Its share has remained within 80-85% of total supply, increasing to 91% in 1986 in spite of the total unavailability of Lomaum and a partial unavailability of the Biopio plants. No new hydro plant has been built since 1974. From 1980 onwards, SONEFE and ENE tried to overcome the difficulties caused by sabotage and disruption of the hydro supply by installing new gas turbines in Luanda and Huambo and diesel units in Lobito and other major centers. Because of inadequate maintenance and lack of technical assistance and spare parts, the new facilities have not resolved supply problems. Therefore, current available capacity in the Central System is limited to 47 MW out of the 111 MW installed. The table below summarizes the installed and available capacities in the various systems and compares them with peak demand. ANGOLA: INSTALLED AND AVAILABLE GENERATING CAPACITrY, 1987 (MW) Hydro Thermal Total Peak System Instl. Avail. InstI. Avail. Insti. Avail. Demand a/ Northern 197.6 135.0 56.8 56.8 254.6 191.8 90-100 Central 49.4 7.2 61.8 39.5 111.2 46.7 30 Southern 27.2 13.6 25.3 15.1 52.5 28.7 9-10 Isolated 12.9 2.4 31.7 5.0 44.6 7.4 n.a. a/ Estimated at generation; reflects varying amounts of suppressed demand. Source: SONEFE and ENE. - 267 - Annex 18 Page 3 of 6 Background 6. Tentative conclusions of the Energy Assessment are to emphasize maintenance and rehabilitation and to give lower priority to additional (new) generation capacity. Furthermore, a rehabilitation program for the entire subsector was identified. It would take 5 or 6 years to complete and require about US$200 million at 1987 prices. The Assessment further made a critical analysis of some of the new investments being considered by Angola, notably the dam and power plant at Capanda. The Review 7. The present review would be built on existing subsectoral knowledge. Its objectives would be to: (i) Identify better and in greater detail the power sector rehabilitation program; analyze its needs for qualified manpower, management and materials; to identify the main constraints to the execution of this program and to find ways to resolve them; to conduct a critical review of its feasibility; (ii) Critically review the investment program of the subsector and the various utilities (ENE, EDEL, SONEFE, CELB, GAMEK); verify the justification of the investment program and its compatibility with utility and government finances and its adequation to the needs of the grids. Conduct this analysis in the context of realistic assumptions about the growth of the economy and of power demand. Review existing feasibility and prefeasibility studies for hydropower plants in the Northern, Central and Southern Systems and suggest a priority list of which ones should be updated first, especially their technical and economic characteristics. (iii) Review the technical and economic feasibility of major investments or, if needed, perform such analysis; (iv) Update the Least Cost Expansion Plan of the Northern System. Analyse the earlier expansion plans developed by SONEFE and more recent consultant studies and compare their hypotheses and conclusions. Analyze their relevance to the present and foreseeable situation of Angola. Analyze the expansion of the Northern System both with and without an interconnection to the other two systems. (v) Study what alternative uses could be found for equipment and materials already procured for projects which, for various reasons, cannot or should not be pursued. - 268 - Annex 18 Page 4 of 6 Evaluate the condition of this equipment and make recommendations regarding its utilization, including the possibility of trade among the various utilities/enterprises. (vi) Prepare an investment program with an estimate of financial costs (including the costs of safeguarding existing works which are not now needed until they will be) and a time phasing which takes into account Angolan staff constraints and a feasible level of external technical assistance. This investment program should be presented in such detail as to permit a rapid appraisal by international financial institutions and/or other bilatetal or multilateral donoes/financiers; and (vii) Identify the most urgeL. technical assistance needs to strengthen management and acccnting of the utilities. Evaluate ongoing activities under multilateral or bilateral auspices and propose a global technical assistance program, of three to five years' duration, closely tied to the availability of Angolan counterparts. 8. This investment review would considet the demand growth scenarios prepared for the Energy Assessment Report (corrected by any new information) as a basis from which to determine the need for rehabilitation and new investment and the updating of the Least Cost Expansion Plan (LCEP) for the Northern System. A sensitivity analvsis of demand growth in Luanda should be carried out by taking into account a gradual meeting of repressed demand (through improvements in the distribution grid), the effect of substantially higher tariffs and the pace of industrial sector rehabilitation (major consumers or potential consumers should be treated individually, separate from small or residential consumers). 9. The following considerations should be kept in mind as guidelines to the carrying out of the investment review: (i) Assign highest priority to rehabilitation of existing facilities; (ii) Strive for improved reliability of supply to the main cities which are also the main industrial areas; (iii) Improve supply to Luanda by addressing the main problems in generation, transmission, transformation and distribution in the Northern Grid; - 269 - Annex 18 Page 5 of 6 (iv) Postpone most small projects in isolated systems mainly for lack of managerial/technical staff, even if equipment or materials have been purchased; (v) Postpone rural/village electrification and urban household connections until hydro supply conditions have been improved, tariffs readjusted and billing and collection procedures have been strengthened; and (vi) Estimate the need for external technical assistance to support power subsector staff ( iE and other utilities) in charge of the major rehabilitation projects (Lumaum, the Southern System, transmission and distribution grids). Capanda 10. With respect to the Capanda dam and power plant, this study will include a detailed analysis of the feasibility of continuing the project or of interrupting it, for an eventual resumption of work at a future time to be determined. The main objective of this part of the Study is to provide an exhaustive analysis of the costs and benefits of interrupting the project in comparison with other alternatives available to Angola. More specificelly, the study will: (i) Identify the best moment for (and the implications of) stopping work on the Capanda project, taking into account: (i) technical aspects (size and stage of the civil works already carried out and to be carried out); ii) economic aspects (possibility of selling the energy produced, feasibility of expanding the Northern Grid and repercussions on the costs of energy in the Northern System); iii) financial aspects (status of the financing plan and probability of obtaining the additional funds needed to complete the project); and iv) organizational aspects (structure of GAMEK and its institutional and management capability). (ii) Identify alternatives to be developed in the event that work is interrupted on the Capanda Project, tying them in with the rehabilitation of the three systems, which is proposed and is being prepared through another part of this study). Identify specific investments to be made, until such time as the Capanda project can be resumed, to meet rising demand reliably. Identify the human and material resources needed and possible constraints to carrying out several projects simultaneously. (iii) Analyze the consequences of stopping work on the Capanda project under two assumptions: i) a North-Center - 270 - Annex 18 Page 6 of 6 interconnection assumed around the year 2000; ii) such an interconnection to be constructed as soon as possible (say, early 1990s). The joint analysis of stopping work at Capanda and interconnecting the system should be rarried out while fully keeping in mind the actual conditions in Angola pertaining to security, qualified manpower, financing, managerial capabilities and the status of power (rehabilitation/expansion) projects already decided upon in the Central and Southern Systems. (iv) Analyze the financial and technical feasibility of safeguarding--over the next several years--the infrastructure and civil works already executed (excavation, tunnel). (v) Evaluate the feasibility and costs of using such equipment and materials (already acquired) for other justifiable projects including other power projects (identified in this review) or projects in other sectors (roads, ports, construction, agriculture, other infrastructure, etc.). Identify ways of preserving this equipment (or of using it) to minimize the deterioration and the costs of preservation. (vi) Review legal and contractual questions which would need to be addressed should Angola decide to stop work on the project. (vii) Identify the needs of CAMEK as an institution responsible for planning the development of the Kwanza River. 11. This Review would be carried out under the supervision of the Joint UNDP/World Bank ESMAP Program which would assume overall responsibility for the Review but parts of the work could be contracted out to a utility or to an engineering firm. 12. The overall Review would require approximately 25 manmonths of consultant time (including 10 man-months for the updating of the LCEP) and would cost between US$555,000 and US$600,000. GA'T'"vn(' -014 ZAIRE ANGOLA 0 National Copital Atlantic 0 Pro;nce Capitals e N-:-c MA 0 Selected TowrIs Ocean ZAMBIA 1--' -- Main Road, Railroads ZIMBABWE NAMIBIA Rie,s BOTSWANA Airports 12' PEO,LFS REP. OF THE-CCINGO 16' B.1 Ports 9 Pr-i- B-d-ies Ime-atio.al Bowd-ies C,Bl  A Z R E Co 2?0 AMLES 01 ?O 0 1( 760 360 4& 560 KIIOMETERS Mba... Co.go,+, 2cr, 24' ZAIRE DT-b-. G E Ulge 91 +L.cap- C.At. Z A I R E '7 LUANDA'L A C-3. 7 Atlantic imo P.ft A.bi. Ocean N, gunz 12' 0 2' Benqu to ZAMBIA Luba.g ami T + C-ing. 0 c u A. N D 16' cl u B A N G '117jiva z T. W,dh-k N A I B 2W ol 14 Continued from inside front cover Reports Already Issued Togo June 1985 5221-TO Vanuatu June 1985 5577-VP. Tonga June 1985 5498-TON Western Samoa June 1985 5497-WSO Burma June 1985 5416-BA Thailand September 1985 5793-TH Sao Tome and Principe October 1985 5803-STP Ecuador December 1985 5865-EC Somalia December 1985 5796-SO Burkina January 1986 5730-BUR Zaire May 1986 5837-ZR Syria May 1986 5822-SYR Ghana November 1986 6234-CH Guinea November 1986 6137-GUI Madagascar January 1987 5700-MAC Mozambique January 1987 6128-MOZ Swaziland February 1987 6262-SW Honduras August 1987 6476-HO Sierra Leone October 1987 6597-SL Comoros January 1988 7104-COM Congo January 1988 6420-COB Gabon July 1988 6915-GA Energy Assessment Status Reports Papua New Guinea July, 1983 Mauritius October, 1983 Sri Lanka January, 1984 Malawi January, 1984 Burundi February, 1984 Bangladesh April, 1984 Kenya May, 1984 Rwanda May, 1984 Zimbabwe August, 1984 Uganda August, 1984 Indonesia September, 1984 Senegal October, 1984 Sudan November, 1984 Nepal January, 1985 Zambia August, 1985 Peru August, 1985 Haiti August, 1985 Paraguay September, 1985 Morocco January, 1986 Niger February, 1986