Senegal - Issues and options in the energy sector

Report No. 4182-SE Senegval Issues and pttions in the Energy Sector 4182 July 1983 Report of the joint UNDP/World Bank Energy Sector Assessment IF[rogram This document has a restricted distribution. Its contents may not be disclosed without authorization from the Government, the UNDP or the World Bank.   FOR OFFICIAL USE ONLY Report No. 4182-SE SENEGAL ISSUES AND OPTIONS IN THE ENERGY SECTOR July 1983 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 UNDP Energy Account, 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, the UNDP or the World Bank.   ABSTRACT: Senegal faces two main energy problems: its dependence on imported oil and rapid deforestation due to the overexploitation of its natural forests. There are significant indigenous sources of energy to substitute for imported oil, including hydropower, limited petroleum resources still to be defined, a small natural gas deposit, peat, and lignite possibilities. Various factors restrict the development of these indigenous resources in the short term, but a concerted conservation program could have substantial and early payoffs in terms of reduced oil consumption, particularly in the industrial sector. By 199U, the combination of substitution and conservation could make it feasible to reduce projected oil imports by 4O)X. With respect to deforestation, which is especially severe in the overpopulated western regions, supply could match demand if more efficient cooking stoves were used, the eastern forests were exploited, forest management was improved, and the appropriate afforestation programs were developed. In theory, Senegal's national energy planning capability is well organized, but in practice the effectiveness of the decision- making process needs to be increased to enable the Government to cope with the much heavier claim that the energy sector will pose on human and financial resources in the coming decade. The report provides a preliminary analysis of investment priorities in the energy sector, and, to help improve coordination of energy sector assistance in Senegal, it also lists the priority areas requiring technical assistance.   ABBREVIATIONS AND ACRONYMS BNDS Banque Nationale de Developpement du Senegal CERER Centre d'Etudes et de Recherches sur les Energies Renouvelables cm Centimeter CSS Compagnie Sucriere Senegalaise DMG Department of Mines and Geology EDF Electricite de France EDS Electricite du Senegal g; Gram GDP Gross Domestic Product GPP Groupement Professionel de l'Industrie du Petrole de l'Afrique Occidentale GWh Gigawatt hour = 1,000,000 kilowatt hours ha Hectare HT high Tension kcal Kilocalorie kg Kilogram kgoe Kilogram Oil Equivalent kWh Kilowatt hour LPG Liquified petroleum gas LT Low Tension m3 Cubic meter mm Millimeter MIDC Ministry of Industrial Development and Crafts MT Medium Tension MW Megawatt = 1,000 kilowatts OMVG Organisation pour la Mise en Valeur du Fleuve Gambie OMVS Organisation pour la Mise en Valeur du Fleuve Senegal O)PEC Organization of Petroleum Exporting Countries PETROSEN Societe des Petroles du Senegal RENES Redeploiement Energetique du Senegal SAR Societe Africaine de Raffinage SEIB Societe Electrique et Industrielle du Baol SENELEC Societe Senegalaise de Distribution d'Energie Electrique SERST- Secretariat d'Etat a la Recherche Scientifique et Technique SINAES Societe Industrielle pour l'Application de l'Energie Solaire SODEVA Societe de Developpement et de Vulgarisation Agricole TAIBA CoTpagnie Senegalaise des Phosphates de Taiba toe Metric ton oil equivalent   CURRENCY EQUIVALENTS Currency Unit - CFA Franc (CFAF) US$ 1 - CFAF 330 1/ 1/ Exchange rate at time of mission. This is the rate used in the report, unless otherwise stated. ENERGY CONVERSION FACTORS CALORIFIC VALUE Fuel (million kcal/ton) toe Crude Oil 10.2 1 LPG 10.8 1.059 Gasoline 10.5 1.029 Jet Fuel 10.4 1.020 Kerosene 10.3 1.007 Gas Oil 10.2 1.01 Diesel Oil 10.2 1 Fue]L Oil 1000 9.9 0.971 Fuel Oil 1500 9.8 0.961 Fuel Oil 2500 9.7 0.951 Fuel Oil 3500 9.6 0.941 Imported Coal 6.9 0.676 Peat 3.9 0.382 Fuelwood 4.5 0.441 Charcoal 7.8 0.765 Crop Residues 3.3 - 4.2 0.324 - 0.412 Bagasse (50% moisture) 1.8 0.176 This report is based on the findings of an energy assessment mission which visited Senegal in June, 1982. The members were B. Chadenet (Consultant, Mission Chief), Masood Ahmed (Economist), T. B. Russell (Electric Power Consultant), M. Farhandi (Petroleum Engineer), N. L. Brown (Renewable Energy Consultant), C. Garrigues (Energy Conservation Consultant), J. Gorse (Forestry Specialist), P. Meier (Manpower and Training Consultant) and R. J. Rodger (Peat and Lignite Consultant). A draft of this report was discussed with the Government of Senegal in May 1983 by a follow-up mission comprising Messrs. Chadenet and Ahmed. While the report generally reflects the situation and data available as of June 1982, it has been updated to take account of a number of institutional and other changes in the subsequent year. Ms. S. Baile provided research assistance for this report and secretarial assistance was provided by Mrs. A. Fernandes, Ms. C. Abunassar, and Mrs. L. Walker-Adigwe.   SENEGAL ISSUES AND OPTIONS IN THE ENERGY SECTOR TABLE OF CONTENTS Page No. MAIN FINDINGS AND RECOMMENDATIONS ............................ i-xi I. ENERGY IN THE ECONOMY .................................. 1 The Economy ........................................ 1 The Two Principal Energy Problems .................. 2 Import Dependence .............................. 2 Deforestation ......................., 2 Energy Consumption Trends .......................... 3 Overview and Sectoral Consumption .............. 3 Petroleum Product Supplies ......................... 5 Energy Resources ................................... 6 Hydropower ..................................... 6 Petroleum ...................................... 7 Natural Gas .................................... 7 Peat ........................................... 7 Lignite ........................................ 7 Fuelwood ....................................... 7 Crop Residues .................................. 8 Solar and Wind Energy ....................... . 8 II. OPTIONS FOR REDUCING THE OIL IMPORT BILL ............... 9 Substitution Possibilities ......................... 9 Petroleum ...................................... 9 Natural Gas .................................... 10 Peat ........................................... 10 Hydropower ..................................... 10 Fuelwood ....................................... 11 Bagasse and Molasses ........................... 12 Groundnut Shells ............................... 12 Solar Energy ................................... 12 Wind Energy .................................... 14 Imported Coal, Fuelwood and Charcoal ........... 14 Energy Conservation ................................ 15 Industry ....................................... 15 Transport ...................................... 16 Buildings ...................................... 17 Conservation Strategy .......................... 17 Projected Oil Consumption .......................... 17 Petroleum Refining and Procurement Issues .......... 19  Page No III. OPTIONS FOR MEETING HOUSEHOLD ENERGY NEEDS AND THE ROLE OF FORESTRY ....................................... 22 Fuelwood Resources ................................. 22 Fuelwood Consumption ............................... 22 Main Forestry Sector Constraints ................... 24 Fuelwood Strategy .................................. 24 Forestry Master Plan ............................... 26 Petroleum Products ................................. 28 Agricultural Residues .............................. 28 Solar Energy ....................................... 29 Electricity ........................................ 29 IV. ELECTRIC POWER SUBSECTOR DEVELOPMENT ................... 30 Growth of Demand ................................... 30 Demand Issues ...................................... 31 Electricity Supply ................................. 32 Existing Facilities ............................ 33 Supply Trends ..............................,.33 Development Program ............................ 35 Distribution of Electricity ......................... 36 Supply Issues ...................................... 36 V. ENERGY PRICES AND TAXES ................................. 40 Petroleum Products ...............................,.40 Electricity ........................................ 42 Fuelwood and Charcoal ...................... I 46 Comparative Energy Prices .......................... 47 VI. INSTITUTIONS, MANPOWER AND TRAINING ..................... 49 Institutional Arrangements ......................... 49 Overall Structure .............................. 49 Petroleum ...................................... 50 Electric Power ................................. 50 Forestry ....................................... 52 Peat ........................................... 52 Solar and Wind Energy and Biomass .............. 52 Conservation ................................... 52 Manpower Issues .................................... 53 Training Needs ..................................... 53 VII. ENERGY INVESTMENT AND TECHNICAL ASSISTANCE ............. 55 Past Investment .................................... 55 Investment Requirements 1982-1990 .................. 55 Technical Assistance ............................... 58  ANNEXES 1.1 Energy Balance, 1981 ................................ 62 1.2 Consumption of Petroleum Products, 1966-1981. 63 1.3 Production of Petroleum Products, 1966-1981 .64 1.4 Hydropower Potential ................................. 65 1.5 Energy Potential of Crop Residues .67 1.6 Energy Potential of Animal Manure ................... 69 2.1 Potential Energy Savings in Industry .70 2.2 Potential Energy Savings in Transportation and Buidings .73 2.3 Projected Consumption of Petroleum Products, 1982-1990 .74 2.4 Potential Oil Savings, 1986 and 1990 .75 3.1 Projected Fuelwood Production and Consumption, 191-2016. 76 4.1 Electricity Production and Consumption, 1970-1981 77 4.2 Electricity Demand Forecasts, Interconnected Network, 1982-2005 ................................. 78 4.3 Power Station Characteristics .84 4.4 Electricity Generation by Station Type, 1973-1981 85 4.5 Generating Plants in Secondary Centers .86 4.6 Main Autoproducers ................................... 87 4.7 Employment and Productivity in Power Subsector, 1971-1981 .8 4.8 Fuel Consumption for Electricity Generation, 1973-1981 . 89 4.9 Alternative Development Programs, Interconnected Network, 1982-1990 ................................ 90 4.10 Characteristics and Costs of Hydropower Projects.... 92 5.1 Electricity Supply Tariff Effective August 1, 1981............ . 93 5.2 Trend of Electricity Prices, 1972-1981 .96 5.3 Theoretical Marginal - Cost Based Electricity Tariff .97 5.4 Fuelwood Prices .98 5.5 Evolution of Retail Petroleum Product Prices .100 6.1 Membership of Energy Commissions .101 6.2 Organization of Ministry of Industrial Development and Crafts .102 7.1 Proposed Energy Investment, 1982-1986 .103 7.2 Technical Assistance Activities in Progress .105 MAP IBRD 16654 Electricity Supply System and Hydropower Potential  TABLES 1.1 Energy and the Balance of Trade, 1972-1981 ...... ..... 2 1.2 Sectoral Consumption Patterns ........................ 3 1.3 Petroleum Consumption and GDP, 1966-1981 ............ 4 1.4 Consumption of Petroleum Products, 1973-1981........ 5 1.5 Balance of Petroleum Supply and Demand, 1981 ........ 6 2.1 Potential Annual Oil Savings by Large Industrial.... Consumers ................................ ......... 16 2.2 Internal Market for Petroleum Products, 1981-1990... 18 2.3 Potential Oil Savings, 1986 and 1990 ................. 19 3.1 Theoretical Fuelwood Supply and Unconstrained Demand, 1981-2016 ....................... ........ 23 3.2 Projected Fuelwood Consumption, 1981-2016 ........... 27 3.3 Projected Fuelwood Production Capacity, 1981-2016... 28 4.1 Electricity Sales by Consumer Category, 1970-1981... 30 4.2 Forecasts of Sales and Maximum Demand, 1981-1990 .... 31 4.3 Fuel Consumption for Electricity Generation, 1973-1981 .34 4.4 Proposed SENELEC Program Interconnected System, 1981-1990 ......................................... 36 4.5 Supply in the Interconnected Network, 1981--1990..... 39 5.1 Petroleum Product Price Structure May, 1983 ......... 41 5.2 Existing Electricity Rates and Estimated Long-run Marginal Costs .44 5.3 Permit Fees and Retail Prices for Fuelwood and Charcoal .46 5.4 Retail Prices of Fuels and Prices per Useful kWh (June 1982) .48 7.1 Energy Sector Investment, 1982-1990 - Mission Projection .58  MAIN FINDINGS AND RECOMMENDATIONS Main Problems 1. Senegal with some six million inhabitants, faces two main energy problems. First, its almost entire dependence on imported oil for its "commercial" energy places a growing burden on the balance of payments. In 1961 net oil imports absorbed over 50X of merchandise export earnings. Second, overexploitation of the natural forests, which account for over half of the country's energy supplies, is causing rapid deforestation, leading to growing scarcity and rising prices of fuelwood a-nd charcoal. Government Strategy 2. The Government's policy for solving these problems, known as "RENES", 1/ aims to halve internal consumption of petroleum products by 1990. The Government strategy is: (i) to substitute indigenous energy such as peat, or other, less costly, imported fuels such as coal for imported petroleum; and (ii) -to restrain energy demand by increasing the efficiency of utilization. Translation of RENES policy into specific operational programs and projects will require the resolution of some important issues, as summarized below. It will also entail a major increase in the allocation of investable resources to the energy sector as the country moves from an essentially recurrent cost-based commercial energy supply system to one where a progressively larger share of energy supply will be produced locally or from imported fuels which have a lower total cost but entail higher capital expenditures. Supply Options 3. Senegal's indigenous energy resources are significant in relation to its energy needs and also reasonably diversified. They comprise a share of the hydropower potential of the Senegal and Gambia rivers (about 1,400 MW, with an average production capability of 7,500 GWh [1.9 million toe]); limited petroleum resources which have not been well defined; a small natural gas deposit (40,000 toe); about 10 million dry tons (3.7 million toe) of peat, lignite possibilities; and signifi- cant fuelwood potential (1.3 million toe supplied in 1981). The long- term energy potential of biomass (crop and animal residues), and of solar and wind resources, also appears to be significant. The development of 1/ "Redeploiement Energetique du Senegal".  - ii - all these resources could ultimately supply a good portion of Senegal's energy requirements, but in the short term their development will be restricted by considerations of location, heavy investment requirements and institutional and political factors. 4. Hydropower; Exploitation of the hydropower potential of the Senegal and Gambia rivers is the responsibility of two multinational organizations. I/ Individual projects can proceed only with the agreement of the member countries which can cause delays. Furthermore, for most of the projects the irrigation aspect is determining the economic justification, while the electricity production aspect is providing, at best, a marginal improvement of the overall rate of return. This low return is largely due to the fact that the sites are 500-760 km from the main load centers in western Senegal, requiring heavy transmission investment. 5. The Manantali dam on a tributary of the Senegal river in Mali is already being developed as an irrigation project, with navigation benefits, but it could also be used to generate 240 MW of power, of which Senegal's agreed share is 40%. Its distance from the main load center at Dakar would involve high transmission costs. To be economic, probably at least half the output (120 MW) would have to be supplied to Senegal. The current estimate of the power station capital cost treats the dam as a "sunk" cost, but political negotiations have not proceeded with the other participants in the project who may wish, in any case, to charge Senegal a "rent for use of the dam to generate power. Another problem is that the project would make the Senegalese power system heavily dependent on a source of electricity located in another country. Detailed feasibility studies remain to be carried out both tor the power station and the transmission line. Kekreti (75 MW) and Sambangalou (135 MW) are the two most promising projects on the Gambia River. They are still at the preliminary study stage and are unlikely to be in operation before the early 1990s. 6. Oil and Gas: The petroleum resources have not been well defined. The Dome Flore field, with probable light oil reserves of 1.8 million toe, is the only discovery to date. The Goveirnment is trying to interest foreign companies in a possible joint exploration venture with the national oil company, PETROSEN. However, there is a jurisdictional dispute over the area with Guinea-Bissau. The identification of further reserves will depend on the success of the exploration efforts which the Government is now trying to promote. A small natural gas deposit is already being used as fuel for the baseload operation of the combustion turbine at Cap des Biches, and should suffice to generate 90 GWh. However, it is not clear whether this is a better use than for peak operation. 1/ The Organization for the Development of the Senegal River (OMVS), whose members are Mali, Mauritania and Senegal; and the Organization for the Development of the Gambia River (OMVG), whose members are Gambia, Guinea, Guinea-Bissau and Senegal.  - iii - 7. Petroleum Refinery: The decision of the shareholders to revamp and expand the capacity of the petroleum refinery from 900,000 to 1.2 million tons seems surprising in the light of the RENES program to halve petroleum product consumption by 1990, but it was apparently based on the export rather than the domestic prospects. Although the market outlook has worsened since the original decision was made, it now seems better to complete the project rather than to abort it, since over 40% of the total cost has been committed. Nevertheless, the appropriate level and mix of operation still needs to be determined. 8. Peat: Various studies are under way to evaluate the feasibility of developing peat deposits found in depressions between dunes north of Dakar, called "Niayes". Potential applications include electricity generation, industrial or household fuel, or as an organic material for agriculture. The Government tavors electricity generation, for which purpose it could supply two 20 MW thermal units over their lives. however, no final decisions can be taken until completion of all the present peat studies in late 1983. Further work will probably still be required before deciding on specific projects. It is unlikely that peat could be available as a fuel before 1988. 9. Lignite: Lignite occurrences have been reported from drillings for oil, water and phosphates, but virtuaLly no samples are available to test the quality. A proposed reconnaissance program with French financing will determine whether further work is justified. Even if the results of the reconnaissance are favorable, lignite could not be commercially available before the early 1990s, given the lead time for defi-ning the reserves and mounting a mining project. 10. Fuelwood. The natural forest cover has been reduced by 30% in the last thirty years, and on present trends would decline another 20% by the end of the century. The remaining fragile forest cover of the understocked and overpopulated western regions will soon be destroyed unless urgent remedial action is taken. The important cbnclusion emerging from the mission's review is that supply could match demand provided effective measures to reduce per capita consumption (efficient cooking stoves) are combined with: (i) systematic exploitation of the largely untapped forestry potential of eastern Senegal, especially Casamance; (ii) an improvement in management of natural forest in overexploited areas; and (iii) development of rural aind state-managed afforestation programs. However, forestry institutions would need strengthening, and a more rational pricing and taxation policy for fuelwood would be necessary to finance the cost (US$6 million/year in thie early 1980s, rising to US$10 million/year in the later 1980s).  - iv - 11. Solar and Wind Energy, and Biomass: The long term potential for solar and wind energy applications appears significant and small-scale use in isolated locations may soon be feasible if expected cost- reductions and technical improvements are secured. The allocation of investments for solar and wind energy development in the Fifth Plan (1977-1981) was excessive (US$27 million) in the Senegalese context. Moreover, donors have used Senegal to carry out experiments, resulting in an onerous dispersion of effort over too many projects, particularly in solar energy, with unsatisfactory results. Insufficient effort has been devoted to the collection of data on insolation and wind-speeds. The energy potential of crop residues and animal manure has been virtually ignored. The proposed Sixth Plan program is more modest (US$6 million), but suffers from similar defects in that there is little distinction between projects with good near-term potential and those that may only be viable in the longer term. 12. Imported Coal; A recent study concluded that coal could be imported at a landed cost of about US$100/ton, or about 20% cheaper than petroleum of equivalent heat content. Even at this relatively high price, coal-fired electricity generation would be cheaper than oil- fired. Coal could also be substituted for oil in the cement and phosphate industries, depending on the costs of conversion, which are being studied. 13. Electricity Supply. SENELEC's latest development program would lead to premature and excessive investments. It proposes the addition of 285 MW of capacity by 1990 for a system with a current capacity of 175 MW and maximum demand of 108 MW. According to the mission's projection, based on a more realistic load forecast, a less lavish reserve capacity criterion and the retention, rather than the retirement, of old plant, the addition of 165 MW would probably suffice, with the first unit (a 15 MW combustion turbine) not required until 1986, instead of 1984. Since the mission of June 1982, SENELEC has prepared a revised and somewhat reduced investment program which is being reviewed by the Government. Nevertheless, the mobilization of adequate financing for power development remains a critical issue. Conservation 14. The mission's review of existing studies and its own investigations confirmed that oil consumption in industry could be substantially reduced by conservation measures. Potential annual savings of 63,000 toe were identified in thirteen major industrial enterprises consuming about 178,000 toe in 1980 (30% of total internal oil consumption). Half the identified savings could be achieved by investments totalling US$16 million with payback periods of less than three years. Savings of nearly 5,000 toe could be secured in transport, at relatively little cost, through introduction of the continuous working day and improved driving techniques in public transport. Other significant potential savings were identified in transportation, but these would require substantial investment (e.g. in public transport), and further study would be needed to determine whether this would be  v justified. There is also room for significant energy savings through improved building design and more efficient use of energy equipment and appliances, including air conditioners, in commercial buildings and hiomes. Energy Pricing 15. Internal prices of petroleum prioducts have kept pace generally with international prices until 1982, when retail prices were not increased despite increases in costs. This led to a serious reduction in the Government's net revenues from petroleum product taxes in that year, since the Government guarantees the refinery a minimum of 12% return. 16. The present subsidies on certain petroleum products, mainly diesel oil for the fishing industry and liquified petroleum gas (LPG) stimulate the demand for these products and are inconsistent with the RENES objective of reducing oil consumption. They should be phased out, including the subsidy on butane. This has failed to achieve its objective of reducing household charcoal consumption, and there is no evidence that its benefits flow to the lower income families. The butane subsidy (which cost US$470,0U0 per month in mid-1982) could be more usefully allocated to reforestation and acceleration of the "Ban ak Suuf" wood stove program. 17. Failure to apply the indexing formula in the electricity tariff has kept the average electricity price lower than it should have been (by about 8% in 1981). Non-payment of arrears by Government departments (equivalent to about one year's electricity supply to these consumers) also creates financial problems for SENELEC. Despite recent increases, the present tariff structure and rates do not tully reflect the long-run marginal costs of supply and need to be adjusted in accordance with a tariff study recently completed. 18. Fuelwood retail prices do not reflect the opportunity cost of wood to the economy. The fee for taking wood from the natural forests is less than US$2/m3, and needs to be brought more in line, over a suitable period, with the fee for wood from offic:ial plantations (US$11/m3). This would also help to meet the rising cost of the forestry development prog_ram. Investment Strategy 19. Despite the scaling down of the electricity development program and the mare gradual increase in reforestation efforts that are recommended by the mission, the energy sector will still place a much heavier claim on Government resources in the 1980s than has been the case in the past. The mission's estimates indicate that energy investments during the Sixth Plan period (1982-86) would amount to 16% of total fixed investment and 2.5% of estimated GDP, or more than double the figures for the Fifth Plan period (1977-81). In the latter part of the decade, the mission identified project proposals that together may surpass 20% of fixed investment and 3% of GDP.  - vi - 20. Given the competing claims of other sectors and the country's overall resource position, there is a real possibility that some of these investments may have to be deferred even if that implies a continued high level of oil imports. This underlines the importance of developing a clear ranking of the various investment projects in terms of both the magnitude and timing of their expected contribution. An associated financing plan also needs to be drawn up outlining the role of public and private and domestic and external sources of funds. This report provides a preliminary analysis of investment priorities and identifies the additional information required for a more exact definition. In particular, once the relative costs of peat and hydropower development have been better defined, the sequence of investments for electric power development will need to be carefully formulated because this subsector will account for over two-thirds of the proposed new investments in energy. Outside of the power subsector the energy conservation and fuelwood development programs have a priority but will require relatively modest investment. In the petroleum subsector considerable resources may be required in the mid-1980's if the proposed exploration program results in commercial discoveries of oil and gas. However, the drain on Government resources may be limited by the mobilization of private capital sources in developing these resources. Sector Organization 21. The main institutional issue concerns the effectiveness of the decision-making process. The formal framework is good, with a National Energy Commission, under the President, taking policy decisions after a review of options by a National Energy Committee, under the Minister of Industrial Development and Crafts (MIDC), for which the Department of Energy in MIDC provides the technical secretariat. In practice, the arrangements are not effective because the Energy Department does not deal with the essential petroleum matters. These are handled by the Department of Mines and Geology (also in MIDC), and coordination between these two departments is poor. Neither department, nor the National Energy Commission and Committee, appears to have been effectively involved in some major decisions, such as the expansion of the refinery and the changed procedure for crude oil procurement. The present arrangements therefore need to be strengthened, for which an appropriate program of technical assistance will be required. Recommendations Reducing Oil Consumption 22. In the absence of savings through conservation and substitution, internal oil consumption 1/ (5b9,000 toe in 1981) is projected at 718,000 toe in 1986 and 747,000 toe in 1990. Nearly all the identified savings 1/ Excluding bunker sales of fuel oil, diesel and jet fuel to international transport companies refueling in Senegal.  - vii - through conservation could be achieved by 1986, by which date the mission estimates that consumption could be reduced by 68,000 toe (nearly 10%). This could rise to 206,000 toe (28% of estimated requirements) by 1990, mainly through the substitution of coal or peat, and hydropower, for oil in electricity generation. Given the scope for further substitution and conservation, a 40% reduction in oil imports by 1990 should be feasible; a figure which is close to the 50% reduction envisaged under the RENES project. 23. In addition to embarking on a program of petroleum conser- vation/substitution, a number of immediate actions can be taken by the Government to alleviate some of the problems described above. These include: Demand Management (i) Accelerate as a matter of high priority, the program to encourage the adoption of the Ban ak Suuf improved wood stove (para 3.12); act to expand the utilization of more efficient charcoal production techniques. (para 3.07). (ii) Establish a demand management program for electricity, including publicity, technical advice, service and technical audits of major consumers (para 4.05). Pricing (iii) Reorient retail energy prices towards their economic costs by phasing out the present subsidies for petroleum products (para. 5.04). (iv) Arrange for Government departments and other public bodies to pay off, over an agreed period (e.g. one year), the large arrears due for electricity supplied by SENELEC (para 5.13). (v) Allow SENELEC to apply the price variation formulae automatically and address the anomalies in the existing electricity tariff structure to bring them more in line Mwith the structure of the cost of electricity supply (para 5.13). (vi) Raise the cost of pernits to cut wood in natural forests, increasing, in accordance, the retail prices of wood and charcoal (para 5.15). Institutional 24. A number of steps will need to be taken if the Energy Department is to function effectively as an energy policy formulation and coordination unit in the Government as is its official mandate. First,  - viii - there must be stronger commitment at the national policy level to involve this Department in all important energy sector issues. Second, the coordination between this Department and the Department of Mines and Geology and other concerned agencies should be improved. In particular all relevant documents should be routinely circulated to this Depart- ment. Third, the technical capabilities of the Energy Department will need to be strengthened through training and technical assistance. With some exceptions, noted below, the organizational structure of the Department is fine but its staff needs strengthening, particularly if it is to embark on a national program of energy conservation. An adequate training program needs to be developed, including provision for a possible country specific training course in Dakar using short-term consultants in the relevant fields. Finally, training and technical assistance must be complemented with a budget for providing adequate office supplies, library and documentation and logistical support. In parallel, the salary structure must allow for the recruitment and retention of qualified and experienced staff. A number of other institutional issues which need to be addressed are listed below: (i) The Hydrocarbons Division of the Department of Mines and Geology (in the MIDC) should be transferred to the Department of Energy (also in the MIDC) to improve the supervision of multinational and Government petroleum companies and the coordination of energy pricing policy. An associated recommendation is that the National Hydrocarbons Commission. (responsible for petroleum pricing) should be incorporated into the National Energy Commission (which is responsible for all energy pricing) (para 6.06). (ii) The Government should allow the reorganized SENELEC to manage its day to day affairs, and should ensure its financial self-sufficiency. To this end, it is essen- tial to restructure SENELEC's balance sheet. The program contract ("contrat-plan") should include a plan for SENELEC's financial recovery (para. 6.10). (iii) The Chairman and Director-General of SENELEC should be appointed for a fixed term with the option of renewal (para b.11). (iv) The staffing levels in the forestry services are too low to carry out the forestry programs and should be increased (para 6.13). (v) The coordinating role of the Energy Department in renewable energy development should be strengthened, and the role of the State Secretariat (SERST) primarily responsible for research in renewable energy should be clarified (para 6.15).  - ix - (vi) The proposed renewable energy program should be revised to (a) concentrate on applications with maximum near- term potential for fuel displacement, (b) provide for the evaluation of biomass energy potential, and (c) establish a network to monitor systematically solar and wind energy resources (para 7.07). (vii) The Office of Energy Conservation in the Energy Department should be given the status of a separate division. (para 6.16). (viii) Each energy agency should be required to prepare an annual long-term manpower plan, for consolidation by MIDC into a global plan for the sector (para 6.18). (ix) Non-French foreign u-niversity degrees should be recognized in the Civil Service Regulations, and salaries should be related to functions rather than diplomas (para 6.19). (x) The forestry education curricula should be modified to fit the need for an eventual shift towards reafforesta- tion (para b.23). (xi) The Government shoul,d assign to one unit the responsiblity for coordinating the technical assistance activities of multilateral and bilateral aid agencies in energy (para 7.15). 25. The Government should also continue - if necessary with external financial or technical assistance - a series of studies to help further delineate the longer term strategy for the energy sector. The most important of these studies are: Development of Energy Supply (i) Review all of the various peat studies on completion to determine the optimum allocation of the peat between electricity generation, household, industrial fuel and agricultural use, subject to any further work that may be needed, including possible additional testing (para 2.06). (ii) Analyze the possibility of utilizing surplus bagasse to generate electricity for public supply (2.10). (iii) Review the merits of utilizing surplus molasses for alcohol production compared with other options, following completion of a Bank-financed feasibility study (para 2.10).  x (iv) Evaluate the economics and market potential for solar water heaters and, if justified, embark on a program to commercialize them (para 2.12). (v) Study the possibility, and cost, of discharging coal in Dakar port using the facilities for handling phosphate exports (para 2.14). (vi) Review the scope for substituting coal for oil in the cement and phosphate industries folLowing completion of the second phase of the coal import study (para 2.16). (vii) Ensure that the proposed power distribution master plan for rural areas examine the merits of rural electrification within the context of the overall stategy for meeting rural energy requirements in the most effective manner (para 4.06). (viii) Update the generation and transmission master plan for electricity supply, in the meant:ime deferring any decision on the 15-MW combustion turbine proposed by SENELEC for 1984 (para 4.21). (ix) Update the cost estimates for the Manantali hydropower project (including an estimate of the cost of trans- mission to Dakar); study the relative demand for elec- tricity in the three countries that would share the output of this project (para 4.21). (x) Review the relative merits of the Kekreti and Sambangalou hydropower projects on the Gambia river (para 4.21). (xi) Study the scope for using groundnut shells as fuel (para 2.11). A second stage would be to evaluate the energy potential and application of other agricultural residues (para 3.19). (xii) Undertake an adequate program of research and investigation to develop forestry technical packages suitable for arid and semi-arid zones (para 3.11). Demand Management and Conservation (xiii) Study the feasibility of introducing a continuous working day in the Dakar area (para 2.21). (xiv) Study the scope for energy savings through improved building design and more efficient use of energy equipment and appliances (para 2.22).  - xi - Technical Assistance 2b. Most of the studies listed above will require some external technical assistance for their execution. As discussed in this report, there is an urgent need to improve the coordination of energy sector assistance in Senegal to avoid potential overlap among the programs of the various donors and to ensure that. all the priority activities requiring technical assistance are undertaken as rapidly as possible. To assist in this effort, Chapter VII of this report includes a detailed list of the priority areas where technical assistance is likely to be required. This list has been discussed with the Government of Senegal and should assist interested donor agencies in programing their energy sector activities in Senegal.   I. ENERGY IN THE ECONOMY The Economy 1.01 Senegal is a poor country of some six million inhabitants, lying three quarters in the Sahel zone, characterized by low rainfall and periodic drought. Apart from groundnuts and excellent fishing potential, its natural resources are limited to some phosphate and iron ore deposits, together with an irrigation potential which, while substantial, will be costly to develop. Despite the rapid growth of urban centers, the economy remains rural. Three quarters of the population, which is growing at 2.8% p.a., live in rural areas, and, in a normal year, agriculture exports account for over half of total export earnings. Although the direct contribution of agriculture to GDP is only about 25%, most other sectors depend heavily on the agricultural sector. The mainstays of the traditional economy arE. millet cultivation and nomadic cattle raising for domestic consumption, and groundnut cultivation for export. The modern sector of the economy is concentrated in the area of Dakar, the capital, which has about one! million inhabitants, excellent port facilities, an important industrial sector and a small but fast- growing tourism industry. 1.02 Since independence in 1960, the economy has grown at only about 2% p.a., while population has grown at 2.5% p.a., so that GDP per capita has declined. The economy is highly vulnerable to fluctuations in the volume and prices of its main exports, groundnuts and phosphates. During the last decade there has been a series of bad crops, reflected in a fall in groundnut production from 1.4 million tons in 1975/76 to 0.5 million tons in 1980/81. Export prices for groundnuts and phosphates also declined sharply after 1975. The adverse effects on the balance of payments have been aggravated by the increases in the price of imported petroleum, resulting in a loss in income since 1974 of roughly 5% of GDP. Although industrial output (excluding groundnut processing) grew at a fairly even rate of about 5% p.a. in the 1970s, this was insufficient to offset these negative factors. GNP per capita in 1980 was US$450, close to the bottom end of the range for all middle-income developing countries of US$420-4500. 1.03 To correct the situation the Government has embarked on a five- year stabilization and rehabilitation program with Bank and IMF support. The aim is to stabilize the economy in the first two years of the period and to achieve an economic grcwth rate of about 4% p.a. in the subsequent three years. To improve the efficiency of the public sector and reduce its excessive claims on public finances, the program provides for medium-term program contracts ("contrats - plans") between the Government and individual public sector enterprises to set sector objectives and reduce the present levels of budgetary support.  - 2- The Two Principal Energy Problems Import Dependence 1.04 The success of the economic rehabilitation program will depend to a significant extent on developments in the energy sector. The two major energy problems in Senegal both have serious implications for the economy as a whole. The first is the virtually complete dependence on imported oil for meeting the country's "commercial" energy requirements. As a result of the sharp increases in the price of oil since 1973, the net 1/ import bill for crude oil and petroleum products rose tenfold from CFAF 2036 million (US$17 million) in 1972 to CFAF 47,485 million (US$175 million) in 1981, as shown in Table 1.1, although consumption of oil products increased only 64%. Over the same period, net oil imports, as a proportion of all merchandise imports, rose from under 3% to 19% and their share of merchandise exports (excluding energy exports) from 4% to 51%. Table 1.1: Energy and the Balance of Trade, 1972-1981 (Millions of Current CFAF) 1972 1973 1975 1978 1981 1. Petroleum Imports 4,238 5,254 14,840 20,900 71,585 2. Petroleum Re-Exports a/ 2,202 2,537 6,948 7,700 24,100 3. Net Petroleum Imports 2,036 2,717 7,892 13,200 47,485 4. Merchandise Imports 79,544 92,683 145,620 190,000 250,000 5. Non-Energy Merchandise Exports 54,601 45,152 100,854 97,300 93,700 (3) as % of (4) 2.6 2.9 5.4 6.9 19.0 (3) as % of (5) 3.7 6.0 7.8 13.6 50.7 a/ Including bunker sales. Source: Bank economic reports. Deforestation 1.05 The second main energy problem is the over-exploitation and inadequate replenishment of the natural forest-cover, particularly in the western part of the country centered around Dakar, resulting in a reduction of 30% in the forest area over the last 30 years. As a result, 1/ i.e. net of petroleum product exports, mainly to Mali and Mauritania.  -3 fuelwood and charcoal, which are used not only in rural areas but also in the cities, are becoming increasingly scarce and expensive, with pervasive effects throughout the economy. Energy Consumption Trends Overview and Sectoral Consumption 1.06 Data on historical and sectoral energy consumption in Senegal are scarce and unreliable. An overview of total and sectoral consumption in 1981 is presented in Table 1.2 (and Annex 1.1). Total energy consumed in 1981 was of the order of 1,670,000 toe consisting of 64.5% traditional woodfuels, 11.5% electric energy, and 24% petroleum fuels. Woodfuels account for the greater share of consumption in Senegal's predominantly rural economy. The industrial sector consumes about 39% of petroleum and 69% of electric energy. Transport takes up 7.3% of the total energy and 30% of petroleum consumption. The 'butanization' program has increased household consumption by over 8,000 tonnes of LPG (para. 3.17), but kerosene and LPG still account for less than 3% each of the internal consumption of all petroleum products. About 11.5% of total energy consumption (33% of petroleum consumption) is used for power generation in diesel or thermal plants. Table 1.2: 1981 Sectoral Consumption Patterns (percent) Sector Petroleum Electricity 1/ Total Energy 2/ Industry 39.0 69.0 17.2 Transport 2/ 30.7 - 7.3 Fisheries 12.8 - 3.0 Public Sector/Other 4/ 9.3 8.0 5.1 Households/Residential 8.2 23.0 67.4 TOTAL ('000 toe) 396.5 192.3 1669.8 1/ Based on diesel/fuel oil consumed for generation; not included under petroleum. 2/ Including fuelwood and charcoal estimated at 961,000 and 120,000 toe respectively. 3/ Excludes international bunker sales of jet fuel and marine diesel/fuel oil. 4/ Includes commercial consumption of electricity. Source: Annex 1.1  - 4 - 1.07 The total internal consumption of petroleum products (i.e., excluding bunker sales and re-exports) more than doubled between 1966 and 1981, rising from 281,000 to 589,000 toe (Annex 1.2). This represented an average growth rate over the period of 5.1% p.a., but there was a marked slow-down in the growth rate in the latter half of the period, as implied by Table 1.3. Since the movement in GDP was in the opposite direction, with an average growth rate of 1.6% p.a. in 1973-1981 compared with 0.6% p.a. in 1966-1973, the result was a sharp decline in the commercial energy elasticity of GDP growth between the two periods. It would be a mistake to attach any special significance to the actual values of the elasticity for the two periods, since quite different values can be obtained according to the starting and finishing years selected, reflecting the wide year-to-year fluctuations in GDP according to the vagaries of the groundnut crop. Nevertheless, it is clear that there has been some decline in the growth rate of commercial energy consumption between the two periods, partly due to the fact that the prices of petroleum products in Senegal have moved upward in line with world petroleum prices since 1973. Table 1.3: Petroleum Consumption and GDP, 1966-1981 Growth Rate % pa 1966 1973 1981 1966-73 1973-81 Total Internal Consumption, 'ooo toe a! 281.3 411.9 588.6 5.6 4.6 Consumption per capita, kgoe 70.0 87.0 101.0 3.2 1.9 GDP (billions of 1979 CFAF) 450.9 471.7 534.7 0.7 1.6 Energy per CFAF million, toe 0.62 0.87 1.10 5.0 3.0 a/ Excluding re-exports and bunker sales. Source: SAR and Bank economic reports. 1.08 As showni in Annex 1.2, consumption continued to increase after the price shock of 1973 up to 1979 but has since declined. As a result the average growth rate in the latter half of this period was only 2.0%, compared with 7.2% in 1973-77. The pattern of consumption of the various petroleum products has also undergone some changes. The main features are: (a) the steady decline in the share of gasoline, from 23% in 1973 to 19% in 1981, to which the greater increase in its price relative to the prices of other products has presumably contributed;  (b) the increased share of fuel oil, from 47% to 51%, mainly reflecting the increasedi requirements for electricity generation; (c) the steady increase in the share of LPG, in response to the Government campaign to encourage the use of butane in households in substitution for charcoal (para 2.09). Table 1.4: Consumption of Petroleum Products, 1973-1981 (% shares of totaL consumption in toe) Growth Rate % p.a. 1973 1977 197'3 1981 1973 1977 '000 X '0U00 l '00 0 XO % -77 -81 toe share toe tole toe share LPG 3.2 0.8 6.5 10.5 12.1 2.0 19.4 16.8 Gasoline 96.1 23.3 121.5 131.8 112.2 19.1 6.0 (2.0) Kerosene 10.0 2.4 12.9 13.7 11.7 2.0 6.6 (2.4) Gas/Diesel Oil 107.7 26.2 136.2 163.0 154.2 26.2 6.1 3.1 Fuel uil 194.9 47.3 267.6 295.4 298.4 50.7 8.6 2.8 411.9 100.0 544.7 614.4 588.6 100.0 7.2 2.0 Memo Item: Jet Fuel 114.9 - 155.9 184.8 154.0 - 8.0 (0.3) for International Airlines Source: Bank staff calculations from SAR data. Petroleum Product Supplies 1.09 Up to 1977 the total output of petroleum products from the refinery of Societe Africaine de Raffinage (SAR) (net of exports) exceeded the total internal demand, alt:hough there were deficiencies in individual products which had to be made good by imports. Since that year total internal demand has exceeded the output of the refinery and supplementary direct imports are currently required for all products with the exception of kerosene and regular gasoline.  - 6 - Table 1.5: Balance of Petroleum Supply and Demand 1981 1/ '000 toe Refinery Product Consumption Exports Production Imports LPG 12.1 0.8 6.8 6.3 Regular Gasoline 30.5 27.3 62.7 Super Gasoline 81.7 9.7 75.9 21.1 Kerosene 11.7 8.5 18.1 Jet Fuel - 158.9 102.4 40.1 Gas/Diesel Oil 154.2 34.1 133.5 47.5 Fuel Oil 298.4 1.5 226.1 82.2 588.6 240.8 625.5 197.2 1/ Excluding bunker sales of gas/diesel oil (85,000 toe) and fuel oil (81,000 toe), which are met through offshore trade by the distribution companies. Jet fuel bunker sales are included in exports. Source: SAR and Bank Staff calculations Energy Resources 1.10 Although small by world standards, indigenous energy resources are significant in relation to Senegal's own needs and are also fairly diversified. They include hydropower, petroleum, natural gas, peat, fuelwood, biomass (in the forn of crop residues and animal manure), solar and wind energy and, possibly, lignite. Hydropower 1.11 Hydropower is the only energy resource which has been reasonably well defined, although much of the information is based on purely preliminary studies. Fifteen potential projects have been identified on the Senegal and Gambia rivers (Annex 1.4) with a combined installed capacity of nearly 1400 MW and an average annual energy capability of 7500 GWh (1.9 million toe). However, only part of this potential would be available for Senegal, since it has to be shared with neighbouring countries (Mali and Mauritania in the case of the Senegal river, and Gambia, Guinea and Guinea-Bissau in the case of the Gambia). Senegal's share will be a matter of separate negotiation for each project, and has been agreed so far only for the Manantali project (40%, or 400 GWh - see para 2.08).  -7- Petroleum 1.12 The petroleum resources have not. been well defined. The only discovery to date is in the Dome Flore area, about 60 km. offshore from Casamance in southwest Senegal at water depths of 40-60m. Recent studies have indicated the possible existence of a light oil field with reserves estimated at up to 3 million tons, having a most likely value of 1.8 million tons. Earlier exploration had revealed a heavy oil deposit in the same area, with estimated reserves of 70 million tons, but its development would be uneconomic because of the offshore location and pumping and transport problems relating to the high viscosity. The identification of further exploitable oil reserves will depend on the success of the exploration efforts which the Government is now trying to promote, particularly in the Casamance offshore area and the Cap Vert peninsula (see para 2.03). Natural Gas 1.13 A small natural gas deposit has been discovered at Diam Niadio, near Dakar, with estimated reserves of 50 million m3 (40,000 toe). Peat 1.14 Peat deposits formed in depressions between dunes known as "Niayes', 1/ in an easily accessible location not too far from Dakar, are a potential fuel source. Estimated known resources are 46 million m3, equivalent to 9.8 million dry tons (3.7 million toe). The moisture content is high (80-90%) but falls to less than 10% after 5-6 days of air drying. The estimated net calorific value of the dried peat of 3900 kcal,/kg compares favorably with fuel peats in Finland and Ireland but the ash content, mainly sand, is high (35%). Lignite 1.15 Lignite occurrences have been reported from drillings for oil, water and phosphates at depths of up to 50m for younger deposits and 200- 600m for older deposits, but no samples are available to test the quality. A proposed reconnaissance program with French financing, scheduled in 1983, should confirm whether detailed exploration to identify commercially exploitable reserves is justified. Fuelwood 1.16 In terms of gross heat content, fuelwood is the dominant energy resource in Senegal. Estimated production capacity in 1981 was 7.2 million m3 (about 2 million toe), compared with estimated consumption of 1/ A second type of peat, consisting of the mangrove deposits in the river deltas of western Senegal, appears unsuitable for use as fuel, because of the high ash and salt content and generally poor quality.  - 8 - 4.6 million m3. According to mission projectiont depending on the regio- nal distribution of fuelwood exploitation, the production capacity could be increased by about 10% by the end of the century, which would still leave a reasonable margin over projected consumption (see para 3.16). Crop Residues 1.17 Data on the production and consumption of crop residues are lacking. Estimates based on crop production figures suggest that the theoretical maximum energy potential of crop residues in 1981/82 was about 900,000 toe (Annex 1.5). Because of technical, socio-economic and social constraints, probably only a small fraction can be used for energy conversion. Solar and Wind Energy 1.18 As with biomass, reliable data are lacking on the solar and wind energy potential. Insolation measurements have been confined to the Dakar area, where they show an average incidence of over 5.5 kWh/m2/day, equivalent to about 0.17 toe/m2/year; the direct component of insolation is high (over 65X of the total insolation). The solar resource is therefore suitable for a variety of solar-thermal (using flat plate or concentrating collectors) and photovoltaic applications. 1.19 Available data on wind speeds indicate that the most favorable possibilities are in the Dakar-Saint Louis coastal area, where there are mean annual speeds of 4-5 m/second (23-45 watts/m2). In the coastal area south of Dakar, the data indicate a mean speed of 3 m/second (10 watts/m2). These wind speeds indicate in principle a potential for water pumping, especially in the Dakar-Saint Louis area. The indicated average wind speed in the rest of the country is only 1.5 m/second, too small for water pumping. From these data, and pending information from the proposed wind resource measurement network, the wind speeds do not generally appear suitable for electricity generation either at dispersed locations or for grid supply.  - 9- II. OPTIONS FOR REDUCING ThIE OIL IMPORT BILL 2.01 Faced with the heavy and growirLg burden of oil imports on the balance of payments (para 1.04), the Government in 1981 announced an ambitious policy known as RENES, 1/ with the declared objective of reducing internal consumption of petrolLeum products by 50% over the period 1981-1990. The feasible options for achieving this can be conveniently discussed under the headings of substitution and conservation. Substitution covers the replacement of imported oil by indigenous energy sources or by other, less costly imported fuels, such as coal, fuelwood or charcoal. Conservation aims to reduce oil consumption by improving the efficiency of fuel utilization. Achievement of any significant reduction in oil imports will require effective action on both these fronts. Substitution Possibilities 2.02 The RENES program focusses primarily on the development of the country's indigenous energy resources (para 1.10) in order to reduce the demand for imported oil. The potential role of these resources in achieving the projected reduction in imported oil consumption by 1990 is discussed below. Petroleum 2.03 Although most of Senegal's sedimentary area (175,000 km2, of which 110,000 km2 offer hydrocarbon prospects) is under permit, the pace of exploration has been slow. With IDA's assistance, the Government has now devised a strategy to promote petroleum exploration in the Casamance offshore and other areas where permits are due to be relinquished starting in June 1983. The strategy is based on a complete compilation, review and interpretation of all existing geological and geophysical data on the sedimentary basins, and a revision of the existing legal and contractual framework for petroleum exploration. Consultants have already carried out a preliminary compilation of existing data and prepared technical assistance and training programs for the Department of Mines and Geology (DMG), the department within the Ministry of Industrial Development and Crafts (MIDC) responsible for the petroleum subsector, and the Societe des Petroles du Senegal (PETROSEN). Their findings provide the basis for the basin studies and technical assistance to be financed under a proposed IDA petroleum exploration promotion project. The Government has also commissioned a seismic survey (with Canadian financing 2/) of the Casamance offshore to promote this area. Finally, it has engaged consultants to revise the Petroleum Code and prepare model contracts. 1/ "Redeploiement Energetique au Senegal" 2/ By Petro-Canada International Assistance Corporation, the Canadian aid agency for the petroleum subsector.  - 10 - 2.04 In parallel with these promotional efforts, the Government intends to carry out exploration, and subsequent development, if justified, in two specific small areas which it holds - Dome Flore, where it hopes to attract a foreign partner, and Diam Niadio (near Dakar), where it proposes to use external financing if no private partner can be found. Dome Flore is the most studied project but a dispute between Senegal and Guinea-Bissau over their continental shelf boundary has caused IDA to withdraw from participation in this project. The proposed IDA credit and a possible Canadian grant does include provision for geophysical work in the Diam Nadio area. If successful, this work would be followed by drilling, possibly with a foreign partner. Natural Gas 2.05 Until a recent accident stopped gas production at Diam Niadio, this resource was being used to operate the combustion turbine at Cap des Biches at base load in order to save fuel oil by displacing steam generation. An alternative mode of operation would be to utilize the turbine for peaking purposes only thereby substituting for more valuable gas oil rather than fuel oil. While this would result in a slower utilization of the gas, whose total volume is in any event rather small (sufficient to generate about 90 GWh of electricity), the discounted value of the stream of benefits from this mode of operation could be higher because of the higher value of the substituted fuel. The mission recommends that an analysis be made of the relative benefits of both alternatives. Peat 2.06 Various studies are under way to evaluate the potential uses of the "Niayes" peat (para 1.14), either in solid form, including briquets, as an industrial or household fuel, or for electricity generation, or as an organic material, either alone or combined with phosphates and other fertilizers. The results of these studies will not be available until end 1983. An issue is likely to arise, therefore, concerning the optimum utilization of this resource. There appears to be a prevailing assumption in Senegal that SENELEC should have first call on peat for use in electricity generation (see para 4.16). However, no decisions should be taken before the present studies are completed. Even then further work may be required before deciding on specific projects and the final decisions regarding the utilization of this resource should be based on a careful economic evaluation of the options so as to maximize the net benefits. Whatever is decided, it is unlikely that peat could be available as a fuel before 1988. Hydropower 2.07 Development of the hydropower potential (para 1.11) presents some problems for the following reasons:  - 11 - (a) Development is the resporLsibility of two multinational organizations. 1/ Irrespective of their location, individual projects require the agreement of all the member countries, which may not be easily forthcoming. Failure to secure Guinea's agreement, for example, has caused delays in consideration of the Sambangalou project on the Gambia river in Senegal, since it would cause extensive flooding in Guinea, although Guinea has recently indicated its readiness to cooperate in this project within the OMVG framework. (b) All the identified projects, except the smallest one, are multi-purpose. The irrigation aspect invariably determines the feasibility of these projects which leaves the hydroelectric component: sometimes in the background of the discussions. This problem is aggravated by institutional shortcomings (see para 6.12). (c) The sites are a long way (500-750 km) from the main load centers in western Senegal, requiring heavy investment in extra-high voltage transmission lines. 2.08 The Manantali project on the Senegal river in Mali seems likely to be the earliest which could supply electricity. This is already being developed as an irrigation project, with navigation benefits, but it could also have an installed power capacity of 200 MW, and supply annually 780 GWh of firm energy (1010 GWh average). Assuming an allocation of about half of the output, an average annual supply of 500 GWh (125,000 toe) could be available to Senegal. The estimated cost of the project is US$ 700 million. The dam is scheduled for completion about mid-1988, but it is unlikely that energy could be available from this source before 1990. Moreover, there are still some technical and economic uncertainties regarding this project which need to be resolved (see para. 4.21 below). Fuelwood 2.09 Most of the fuelwood is used for cooking and heating in the home (see Chapter III), either as wood or after conversion to charcoal, which is used particularly in the cities. As such, it is a direct substitute for petroleum products such as LPG and kerosene. However, Government concern over deforestation in recent years resulted in a reverse policy of substitution in the form of the so-called "butanization" campaign to encourage households to use butane instead of charcoal. The objective was to halve charcoal consumption by the end of the Fourth Plan (1977). This objective was not achieved, since charcoal consumption actually rose over the period (from an estimated 92,500 tons to 93,500 tons), but 1/ The Organization for the Development of the Gambia River (OMVG) and the Organization for the Development of the Senegal River (OMVS).  - 12 - consumption of subsidized butane rose sharply, from 2900 tons in 1974 to 11,000 tons in 1981, about half of which had to be imported. As discussed below (Chapter III), with an effective forestry policy to improve supply, and increased efficiency of utilization, there should be no difficulty in meeting the projected rise in fuelwood requirements resulting from population growth without jeopardizing the reforestation program. This should also help to eliminate the need for the "butanization" program (see para. 5.04 and 5.16). Bagasse and Molasses 2.10 Bagasse is used by the sugar-refining industry for the production of the steam and electricity needed. for the refining process. Improvements in the refinery process and bagasse use as a result of drying are expected to yield a growing surplus of bagasse, which would be of the order of 100,000-150,000 tons by 1984 (maximum value 24,000-36,000 toe). Proposals for using this surplus, including additional electricity production for rice irri-gation and a milk production scheme, would not absorb the surplus, and a further possibility meriting serious consideration is to use it to produce electricity, possibly for supply to the public network. Another by- product of the sugar refinery which is not utilized is molasses, production of which is expected to reach 47,000 tons by 1985. This could produce about 14 million liters of ethanol (11,000 tons) for blending with gasoline. The Bank has agreed to finance a feasibility study to evaluate the relative merits of utilizing the molasses for this purpose compared with other options such as animal feed supplement, export or industrial yeast production. Groundnut Shells 2.11 The groundnut shells produced as a by-product by the groundnut oil industry are an important renewable energy source. The quantity fluctuates with the size of the groundnut crop but is estimated at about 80,000 toe this year. The industry uses groundnut shells for electricity generation, but the indications are that there is a substantial unused surplus. In 1979, for example, it was estimated that the equivalent of 12,000 toe was available in unused collections of groundnut shells. The largest groundnut oil producer, the Societe Electrique et Industrielle du Baol (SEIB), has a groundnut-processing capacity which yields about 60,000 tons of tgroundnut shells, or perhaps 21,000 toe, but uses less than half of this for electricity generation. A study should be undertaken to determine the amount of the potential groundnut surplus and the scope for substituting it for imported oil, either to generate electricity for public supply or as a household fuel. Solar Energy 2.12 An excessive number of solar energy projects has been undertaken in Senegal, which has been used by industrial countries for experimenting. The results may be summarized as folLLows:  - 13 - (a) About a dozen solar water heaters developed at CERER and installed by the Societe Industrielle pour l'Application de l'Energie Solaire (SINAES), have been installed at hotels, private houses, health clinics, a housing project and a school. There have been maintenance problems which are due probably to inefficient designs. A study is called for to improve the technical performance, reliability and economics of these heaters, although the option of importing heaters should also be considered. The potential of water heaters for commercial and industrial process heat should also be explored. (b) Solar-thermal pumps installed at various locations all appear to be out of orde.r for lack of maintenance or spare parts and unsuitability for isolated regions. Their potential use is also limited by extremely high cost (US$36,000/kW). (c) A 25-kW solar-thermal con-version system (for generating electricity) was installetd in 1981. Again, the cost, reported at around US$45,000/kW, was exorbitant. The system is operating at a reduced capacity. (d) Photovoltaic systems offer relatively better prospects for electricity generation in specific uses. About a dozen systems have been installed in sizes ranging from 0.5-2.6 kW. They are simple and easy to maintain. Most of them are operating satisfactorily, providing water to villages, and light and refrigeration to rural health clinics. The only barrier to more widespread use has been their very high cost, which in 1982 came to US$5,000-7,000 per panel kilowatt (excluding the "balance of system" costs). However, the costs are declining, especially those of the cell panels, and photovcltaic systems are expected to become competitive with diesel generating systems perhaps by 1986 in applications such as rural electrification. Some other applications (e.g. educational TV, refrigeration and small water pumps) are competitive today in several locations. (e) A few prototype solar stills and solar fish dryers have been built and tested. Based on experiments conducted on these systems in numerous developing countries, these solar applications are generally quite cost effective. However, further analysis and experimental work, including a study of the fish drying process as a whole (of which solar heating is only one element), are needed before reaching definitive conclusions on their viability.  - 14 - Wind Energy 2.13 According to a USAID survey, there are about 30 windmills in Senegal, mostly installed between 1978 and 1981. Among the few operating satisfactorily, most are used for pumping water. The others have had problems, in some cases because of inadequate attention to siting and installation. Proper wind-resource assessment is needed before embarking on any major wind-energy project. Imported Coal, Fuelwood and Charcoal 2.14 According to a 1981 consultants' study, 1/ steam coal could be imported into Senegal at a landed cost of about US$100/ton (1981 prices), equivalent to about US$14.30/million kcal. This is almost 20% below the current international price for heavy fuel oil of US$170/ton, or US$17.70/million kcal. The coal cost is based on 30,000 ton shipments and an annual tonnage of 150,000 - 400,000 tons which, according to the study, could be handled with the present discharging facilities (used primarily for phosphates) at Dakar. An area of about 10,000 m2 would be needed to receive the coal on discharge. This might be found in the area at present used for handling phosphates exported from Dakar, or in an extension planned by 1985, but this would need to be confirmed by a detailed study in conjunction with the port authorities, which would also establish the costs of this facility. 2.15 At the costs indicated, coal would be cheaper than oil for electricity generation, even after allowing for the higher capital cost of a coal-fired station and its higher operating ancd maintenance costs. The mission's projection of electricity requirements suggests that this would justify the installation of a 30-MW coal-fired station in 1988, consuming about 80,000 tons of coal a year, once the cost of the "Niayes" peat project and other hydro based options for electricity generation has been established. 2.16 Coal could also be substituted for fuel oil in the cement and phosphate industries, currently consuming about 122,000 tons/year, equivalent to about 175,000 tons of coal. This is expected to rise to about 200,000 tons of oil (285,000 tons of coal) with the planned expansion of the cement industry. The scope for coal substitution in these industries will depend on the results of a detailed study of the costs of conversion, which is currently under way as part of the second phase of the coal import study. It will also depend on the outcome of the present studies regarding the alternative uses of peat, since this could also be substituted for fuel oil in these industries. 2.17 According to a preliminary study, charcoal from Guinea-Bissau or Liberia could be imported at costs comparable to those for domestically produced charcoal. The feasibility of fuelwood or charcoal imports from the Ivory Coast is being examined in the second phase of the coal import study. 1/ "Etude de l'Importation de Bois et de Charbon Mineral au Senegal", SEMA/ORGATEC, Interim Report, August 1981.  Energy Conservation Industry 2.18 There is a significant potential for reducing oil consumption in industry by appropriate conservation measures. This conclusion is based on the investigation of thirteen major industrial enterprises, which consumed about 178,000 toe in 1980, or some 30% of internal oil consumption. This would rise to about 255,000 toe on completion of the expansion of the cement plant now under way. As shown in Annex 2.1, the industries covered were cement, phosphates, fertilizers, textiles, petroleum refining, shoes, sugar refining and flour milling. The investigation of these industries also led to the identification of further possible savings in SENELEC, which consumed 187,000 toe in 1980 for electricity generation. 2.19 As shown in Table 2.1, the potential annual savings are estimated at about 63,000 toe, excluding an expected saving of 22,000 toe in the cement industry as a result of an extension and the installation of a precalcinator now under way. Over .50,000 toe would be realized in the sugar and cement industries, and another 10,000 toe in textiles and electricity supply (SENELEC). About half the total savings could be achieved for investments totalling CFAF 5300 million (US$16 million), with payback periods of three years or less. It is not possible without further study to estimate the investments that would be required to realize the other half of the savings, expected to be achieved by the use of materials additional to gypsum in the cement industry, and the use of surplus bagasse in the sugar industry to generate electricity for public supply. However, it is thought that they could also be achieved by investments with payback periods not exceeding three years.  - 16 - Table 2.1: Potential Annual Oil Savings by Large Industrial Consumers (a) Savings for which investment cost has been estimated (b) Savings for which investment cost has not been estimated Potential Savings Required Investment Annual Toe Saved Industry (toe/year) (CFAF million) (per CFAF million) Sugar (a) 19,421 3,205 6.1 (b) 17,800 n.a. n.a. Cement (a) 800 200 4.0 (b) 13,100 n.a. n.a. Textiles 5,292 987 6.4 Electricity Supply 4,400 287 15.3 Food and Drink 1,461 445 3.3 Phosphate 470 150 3.1 Footwear 90 17 5.3 Total (a) 31,934 5,291 6.0 (b) 30,900 n.a. n.a. Source: Bank mission estimates and Gaucher Pringle study. 2.20 The figures in Table 2.1 take no account of the scope for oil savings in other industries. According to a recent study, 1/ the potential saving in the phosphate-mining industry is about 21,000 toe, but this seems optimistic, and a more detailed study is needed to establish the correct figure. Significant savings should also be possible in the rest of industry, particularly among small and medium enterprises, provided their present lack of interest in conservation can be overcome. This is a task which the proposed conservation expert for industry (see para 6.17) will have to address. Transport 2.21 The mission identified potential savings of about 26,000 toe/year in the transport sector, as shown in Annex 2.2,. The estimated savings from introduction of the continuous working day in the Dakar area by savings on transport during the lunch break (4000 toe) and the training of bus drivers in improved driving techniques (600 toe) could be secured at relatively little financial cost although there may be 1/ "Expertise Energetique d'Entreprises," Gaucher Pringle, November 1981.  - 17 - difficulties in changing established social conventions. Achievement of the other savings through the shift from private cars to buses (8000 toe), the transfer of passenger and freight traffic from road to rail (9000 toe), the establishment of a suburban railway service between Rufisque and Dakar (4300 toe) and the shortening of the Dakar-Richard Toll road (450 toe) would require substantial investment. The size of the potential savings warrant further study to determine whether the investments would be justified. This should be the first task of the proposed conservation expert for transport, who should also identify other possible sources of savings, such as the movement of petroleum products by rail rather than by road. Buildings 2.22 There is room for significant savings in the use of energy in residential and commercial buildings, especially hotels, through improved design of the building envelope (walls, roof and windows), more efficient energy-using equipment and appliances an-d greater care in using them. Further work will be needed to determine the potential scope for savings in the Senegalese context and how to achieve them. The proposed conservation expert for buildings should be responsible for this. One possible area for improvement identified by the mission is in the way individual air conditioners are used in offices, hotels and high-income homes. 2.23 Water heating in high-income nomes is generally by electric heaters. The adoption of solar water heaters in the Cap Vert peninsula could save the equivalent of 2,700 toe/year. Accurate cost data on solar collectors in Senegal are not available but, from information on similar African countries, the payback period for solar water heating systems in displacing electricity should be less than three years. Conservation Strategy 2.24 Achievement of the potential o:Ll savings described above, and the identification of other potential savings, will call for a consider- able effort within the framework of a carefully formulated strategy. A correct pricing policy has a key role to play, as discussed in Chapter V below. Improvements in organization and staffing for conservation policy formulation and implementation will also be needed, and these institu- tional aspects are discussed in Chapter VI. Projected Oil Consumption 2.25 The mission was provided with the SAR's projection of the internal demand for petroleum products (i.e. excluding fuel oil for ships' bunkers and jet fuel for interrLational aviation) to 1986. As shown in Annex 2.3, this implies an average growth rate of total demand over the period of 1.7% p.a., which is more or less in line with the recent trend (para 1.08). However, this projection does not appear to allow for the projected expansion of the cement industry, which would increase its oil consumption by 77,000 t:oe (para 2.18). This may be on  - 18 - the assumption that the industry would be switching to the use of peat 1/ as fuel, but this assumption would be unwarranted pelding the outcome of the present studies on the peat resources (para 2.06). Table 2.2, therefore, shows the adjusted requirements in 1986 and 1990, including the projected increase in the cement industry's fuel oil consumption. This raises the 1981-1986 annual growth rate for total demand to 4.1% and for fuel oil to 6%, compared with the unadjusted rates of 1.7% and 1.5% respectively. Table 2.2: Internal Market for Petroleum Produicts, 1981-1990 1/ ('000 toe) 1981 Growth Rate, % p.a. (actual) 1986 1990 1981-86 1986-90 LPG 12.4 24.3 35.5 15.6 10.0 Gasoline 112.2 123.8 128.7 2.0 1.0 Kerosene 11.7 9.1 7.1 (4.8) (6.0) Gas/Diesel Oil 154.2 161.4 165.5 0.9 0.8 Fuel Oil 298.4 399.3 409.8 6.0 0.6 Total 588.6 717.9 746.6 4.1 1.0 1/ Allows for the additional fuel oil demand steming from the cement company's proposed expansion. However, this does not take into account the potential reduction in oil demand from the program of energy conservation which is discussed below. Source: GPP and Bank staff estimates. 2.26 This projection does not allow for the specific potential fuel savings identified by the mission and described above. Most of these could be realized by 1986, given the necessary investments and other measures required, and the rest by 1990. In addition, reductions in consumption of gas/diesel oil and fuel oil are expected as a result of substitution by coal or peat in the latter part of the decade, and the installation of the first hydropower plants by 1990 (see Chapter IV). The resulting potential savings in consumption of petroleum products amount to 60,000 toe in 1986 and 187,000 toe in 1990. It should also be possible to reduce the projected butane consumption by discontinuing the present subsidy, which is no longer justified (para 2.09). Assuming 1/ Or, possibly coal, but the feasibility of this will not be known until the study now under way is completed.  - 19 - consumption stabilizes at the projected 1982 level of about 17,000 toe as a result, this would represent a further saving of 8,000 toe in 1986, rising to 19,000 toe in 1990. Summarizing, the total potential savings are 68,000 toe in 1986, or nearly 10% of the projected consumption shown in Table 7.3, and 206,000 toe in 1990, or 28% of projected consumption. These estimates do not allow for other possible reductions in oil consumption, whether through conservation measures in industries other than those identified in Annex 2.4, or through substitution, such as the use of surplus bagasse (para 2.10) or groundnut shells as fuel, and the replacement of oil by imported coal in industry. This suggests that the RENES target of reducing oil consumption 50% by 1990, while ambitious, may be achievable. Table 2.3: Potential Oil Savings Through Conservation and Substitution 1986 and 1990 (toe) 1986 1990 Butane 8,000 19,000 Gasoline 4,000 4,000 Gas/Diesel Oil 600 600 Fuel Oil 55,0 182,000 TOTAL 68,400 205,600 Source: Mission estimates and Gaucher Pringle Study. See Annex 2.4 for details. Petroleum Refining and Procurement Issues 2.27 The decision to undertake the project now under way to modify and expand the petroleum refinery operated by SAR seems at first sight surprising in the light of the RENES program to reduce internal oil consumption, but, as with previous expansions, it appears to be based on the export rather than the domestic market. The project comprised four main components, providing for: (a) expansion of port facilities to handle larger tankers and hence lower transportation costs; (b) the addition of a special unit to process cheaper crudes while still meeting international jet fuel specifications; (c) revamping to increase the capacity from 900,000 to 1.2 million tons and increase the proportion of gas oil produced; and  - 20 - (d) the addition of a hydro desulfurization (HDS) unit so that any crude can be processed while meeting the European standard for the sulphur content of g.as oil. The decision by the private majority-owned refining company to place the contract in July 1981 was apparently taken without a detailed feasibility study. However, the mission's analysis indicates that, at the time of contract signing, the project was economically justified except possibly for the HDS unit (which has since been cancelled). If the decision were to be made today, it is doubtful if the project would be justified, given the present market outlook. Nevertheless, the mission also concluded that, despite the changed market prospects, it was now better to complete the project than to abort it, since over 40% of the total cost of about US$50 million has already been committed and cancellation would involve heavy penalties. 2.28 A second issue concerns the Government's decision in 1981 to change the arrangements for the procurement of crude oil. With the change in the world crude market, OPEC members were offering crude to some countries at prices well below the official quotation. To take advantage of this situation, the Government has entered into Government to Government crude oil supply contracts. Current contracts include 750,000 tonnes per year from Nigeria and 720,000 tonnes per year from Algeria. These quantities are more than enough to meet national needs and far greater than the total oil intake of the SAR (660,000 tonnes in 1981), of which the Government is only entitled to supply 10% under the convention governing the operation of the refinery. 1/ The private shareholders in the SAR have temporarily accepted a larger share of oil supply by the Government, but the final arrangements have yet to be decided. Having manifested its desire to gain better control of its oil supply, the Government is considering, among other things, increasing its participation in the refinery from 10% to abut 50%, which would both increase its control over refinery operations and allow for a larger share of crude supply from Government to Governnent contracts. An alternative would be to change the original convention governing the operation of the refinery to permit a disproportionate share of Government procured oil. These issues are currently being analyzed, a key factor being, of course, the cost savings that accrue from the Government to Government contracts versus procurement through commercial channels. To date, a principal source of these savings has been the extended credit terms (90 days versus 30 days) associated with the Nigerian contracts. The Government has used the resulting savings to finance a part of the local resource requirements of PETROSEN's exploration program. Another factor to be considered is the potential gains/losses that could result from the disposal of the surplus crude (i.e. over and above the refinery's requirements) on international 1/ Shareholders in the refinery are normally entitled to supply crude in proportion to their shareholdings. The private shareholders are ELF (30%), BP, Mobil, Shell, Texaco, Total (11.8% each) and Esso (1%).  - 21 - markets. The contracts with Nigeria and Algeria were based on official prices and are being implemented with the assistance of a private company (ARAFENCO up to July 1982 and ECAMI thereafter 1/). Although the terms of the agreement between the Government and this company are not available, it appears that the company is commercially responsible for the success of the operations and in particular is liable for all gains/losses resulting from disposal of any surplus oil. 2.29 Given the heavy burden of o:il imports on the balance of payments, it is clearly important that the arrangements for crude oil procurement should be such as to keep its cost to a minimum. In this context two important questions need t:o be examined: first, whether further cost savings could be affected by using an existing petroleum sector agency such as the SAR or PETROSEN to implement these procurement arrangements, which would also reduce the number of agencies involved in this area; and second, whether continued reliance on Algerian and Nigerian sources of supply is cost effective given that the main purpose of the refinery modification project is 1o permit the use of heavier and cheaper types of crude oil. The mission recommends that the Government pursue these questions as a matter of priority. 1/ ARAFENCO: Arab African Energy Co. Ltd. ECAMI: Energy and Commodities African Management Investment Corporation.  - 22 - III. OPTIONS FOR MEETING HOUSEHOLD ENERGY NEEDS AND THE ROLE OF FORESTRY 3.01 The dominant energy source for meeting household energy needs in Senegal is fuelwood from the forests, used either as wood or after conversion to charcoal. This is not confined to rural households, accounting for almost three quarters of the total population, since urban households are also large consumers, mainly of charcoal. Since households consume virtually all the fuelwood and charcoal produced, this chapter discusses the role of the forestry sector in meeting their energy requirements, and the main issues which arise. The present and prospective contributions of petroleum products, mainly butane and kerosene, agricultural residues, solar energy and electricity are briefly discussed. Fuelwood Resources 3.02 Natural forest covers about 13.E8 million ha, or 70% of the country. The theoretically available annual production (mean annual increment) of 7.2 million m3 equivalent of roundwood (2 million toe), is very unevenly distributed. Nearly 90% of the potential is in eastern Senegal and Casamance, far from the main urban consuming centers in western Senegal. This reflects depletion of natural forests in the overpopulated western areas. In the 1950s, most of the fuelwood for Dakar came from the nearby Thies area, but it now has to be brought at increasingly high transportation cost from areas 300-500 km away. 3.03 The natural forest cover has been reduced by 30k in the last thirty years. According to the 1981 Senegalese Forestry Master Plan, if no effective action is taken to reverse present trends, forest cover will be further reduced to about 11.3 million hectares by the turn of the year 2000. This corresponds to a production of 6.3 million m3 of fuelwood, which amounts to a decline of 0.9 million m3 in annual production, or 12.5%, with respect to the production in 1981. Fuelwood Consumption 3.04 Total fuelwood consumption in 1981 was estimated at 4.6 million m3 equivalent of roundwood (1.3 million toe). As shown in Table 3.1, 1.3 million m3 (28%) were consumed in urban areas and 3.3 million m3 (72%) in rural areas. Over 80% of the urban consumption was in the form of charcoal, compared with only 4% in the rural areas. Average urban consumption per capita was 0.65 m3 (186 kgoe) and rural consumption 0.85 m3 (244 kgoe). In the absence of any change in the consumption of wood as fuel, consumption would rise to 7.2 million mi in 2001 and 10.1 million m3 by 2016 if population continues to grow at 2.8% p.a. As a result of increasing urbanization, a growing proportion of consumption would be in the form of charcoal, as shown in Table 3,1.  - 23 - Table 3.1: Theoretical Fuelwood Supply and Unconstrained Demand, 1981-2016 ('000 m3 equivalerLt roundwood) DEMAND THEORE- Rural Areas Urban Areas Total Demand TICAL Fuel Char- Total Fuel- Char- Total Fuel- Char- Total SUPPLY wood coal wood coal wood coal 1981 3170 130 3300 230 1070 1300 3400 1200 4600 7200 2001 4030 170 4200 540 2460 3000 4570 2630 7200 6300 2016 4516 190 4700 920 4480 5400 5480 4640 10100 5300 Source: 1981 Forestry Master Plan 3.05 On present trends, therefore, fuelwood consumption would exceed the natural forest growth before the year 2000. Severe local and regional shortages would emerge much earlier. The understocked and overpopulated western regions are alread.y in a critical position, and their remaining fragile forest cover will soon be destroyed unless urgent remedial action is taken. 3.06 An estimated 83% 1/ of 1981 consumption was met by uncontrolled production in the rural areas, where the population is accustomed to collect wood as a free good, without any license. Only the balance of 17% 1/ is collected under license by professional loggers, who provide most of the fuelwood for urban areas and have to obtain permits from the Forestry Department and pay taxes for cutting wood. 3.07 The Senegalese charcoal makers are the best in West Africa and the yield (by weight) of their traditional earth-covered kilns approaches 20%. With minor improvements, such as those developed by the US Peace Corps in Casamance (the "Meule Casamanaise"), these yields could be raised to 20-30%. Given this potential of the traditional techniques, the promotion of more sophisticated kilns, such as the portable steel and masonry block types, is not a high priorit:y. 1/ The explanation for the apparent discrepancy between these production percentages and the rural/urban consumption percentages in para 3.04 is probably that a significant portion of the uncontrolled rural production finds its way into towns.  - 24 - 3.08 The traditional fuelwood stove (the so-called "three stone" system), on the other hand, is extremely wasteful of energy, with an efficiency of only 5-10%. As a result, the Senega:Lese consume three to five times as much energy per capita for cooking as the inhabitants of a rich country. Main Forestry Sector Constraints 3.09 A solution of the fuelwood problem will depend on overcoming the main constraints to forestry development. These are: (a) the difficulty of preventing over-exploitation of the natural forest through overharvesting of wood, excessive clearing for agricultural use and overgrazing, largely as a result of population growth; (b) the difficulty of obtaining full and sustained popular and Government commitment to policies designed to control this over-exploitation. Short-term perce!ived needs, such as for increased livestock and food production, have tended in the past to weigh more heavily than the need to protect and develop forestry resources, the benefits of which may be slow in coming; and (c) the lack of forestry technical packages well adapted for marginal lands in arid and semi-arid zones. 3.10 Action to overcome these constraints on the supply side will not suffice in itself to solve the fuelwood problem. An effective strategy must include also measures to reduce consumption per capita through greater efficiency in the use of wood as fuel. Given such a reduction, it would be possible to protect and develop the forestry resources of Senegal while producing enough fuelwood to meet and even exceed the demand. This would be based primarily on exploiting the largely untapped forestry potential of eastern Senegal and, especially, Casamance, while improving the management of the remaining fragile forest cover in areas which have been over-exploited in the past, and cleveloping rural and state-managed afforestation programs. Fuelwood Strategy 3.11 The eAsential ingredients of a strategy to achieve these objectives are, on the supply side: (a) strong and sustained popular participation and Government commitment. There are encouraging signs of a change in popular attitudes towards the protection of trees, partly in response to the Forestry Deparl-ment's support for household "mini-nurseries". As in other countries, Government commitment to effective forestry policies is also much stronger than hitherto;  - 25 - (b) further strengthening of forestry sector institutions to carry out the program effectively (see Chapter VI); (c) a more rational pricing andl taxing policy for fuelwood to ensure an adequate return of funds to the sector to finance the program (see Chapter V); (d) an adequate program of research and investigation to develop forestry technical packages more suitable for arid and semi-arid lands i.e. with annual rainfall in the range 0-1000 mm/year (the packages currently available are designed for areas with average rainfall exceeding 1000 mm/year); (e) improvement of the transportation system for fuelwood. Possible improvements include better road communications between Casamance and the Dakar area, with a bridge across the Gambia river to avoid the present long detour or delays (inadequate ferry services); more use of the railways, such as the Dakar-Niger line and the proposed railway for carrying iron ore from eastern Senegal, which would also have the merit of conserving fuel, compared with road transport (para 2.21); more use of sea and river transportation via the Cas amance and Senegal rivers (with the navigational improvements associated with construction of the Diama and Manantali. dams, the Senegal river could be used to transport wood i-rom the remote parts of eastern Senegal). In implementing new ways of road and water transportation, the Goverrmuent should use the services of the private sector (already well organized for fuelwood) rather than those of the less efficient and more expensive parastatal organizations; (f) changes in land management, including an integrated approach to the interacticn of forestry, agriculture and pastoralism. More intensive livestock and agricultural practices, promoted through well balanced and operationally integrated schemes, would help to reduce further damage to the natural forest cover. They would have the additional advantage of helping in the more effective use of crop and animal wastes as energy sources (para 1.17). Roadside tree plantations as wll as those along railroads should also be encouraged, as in China and India. These would also have the advantage of acting as windbreaks; (g) adequate investment in forestry (see Chapter VII). 3.12 On the demand side, the main emphasis should be on the acceleration of the program to prconote the "Ban ak Suuf" improved wood stove, which is by far the most important activity in biomass conversion in Senegal. It was initiated two years ago with USAID funding, and is  - 26 - being conducted by the Centre d'Etudes et de Recherches sur les Energies Renouvelables (CERER). In cooperation with many organizations and other Government agencies, CERER has held training sessions in about 140 villages and two urban communities, at which between 1400 and 2000 people have learned how to build the stoves. Over 4500 stoves have already been built in villages throughout the country. CERER is also developing stoves for use in urban areas, with peat and charcoal as fuels. 3.13 The Societe de Developpement et de Vulgarisation Agricole (SODEVA) also has a program to integrate the dissemination of Ban-ak-Suuf stoves in its general agricultural extension work. This program uses 85 extension agents, each working with 15 villages, to promote the construction and use of the stoves. In addition, the Secretariat d'Etat de la Promotion Humaine has instituted a dissemination campaign in 300 rural centers, with 300 "monitrices rurales" trained in stove-building techniques. The arrangements for coordination of these programs with the main CERER effort appear to be satisfactory. For example, CERER provided the training for the first three SODEVA agents. 3.14 The two inevitable questions are how many of the stoves are actually in use, and what is the actual saving in fuel. An unplanned, impromptu visit by the mission to one village showed almost all the stoves observed had been in use that day, with significant savings in fuelwood requirements (about 50%). CERER estimates that about 60% of the stoves constructed are in use, yielding a 30-40% saving in fuelwood. These estimates seem reasonable, but, in view of the importance of this project to reforestation efforts, further verification is required on the rate of use over longer periods and the actual fuel savings of the Ban- ak-Suuf stoves. This should be done as part of the (partly) IDA-financed forestry project. A continuing effort should also be made to improve the design and lower the cost of these stoves. The Ban-ak-Suuf stove dissemination program should be expanded and made a separate line item in the budget. The stove program should receive the highest priority. Forestry Master Plan 3.15 The 1981 Forestry Master Plan incorporates most of the elements of the strategy outlined above. The main objectives for fuelwood are: (a) to reduce the regional deficits in rural areas through the establishment of rural tree plantations; (b) to supply the fuelwood needs of urban areas through management of the natural forest cover and establishment of state-managed tree plantations; and (c) to reduce per capita consumption through improved efficiency of utilization. Table 3.2 shows the resulting consumption projections for 2001 and 2016, which are respectively 6% and 22% below the unconstrained projections (para 3.04).  - 27 - Table 3.2: Projected Fuelwood Consumption, 1981-2016 Rural Areas Urban Areas Total Total Total Year m3/capita '000 m3 m3/capita '000 m3 '000 m3 1981 (actual) 0.85 3300 0.65 1300 4600 2001 0.75 3700 0.65 3100 6800 2016 0.50 3300 0.55 4600 7900 Source: Forestry Master Plan 3.16 The projected consumption figures are reasonable but the planting and production targets of the Forestry Master Plan are over- ambitious. They assume that 100,000 ha/year of natural forest cover can be taken under management over the next 20 years, concurrently with the establishment of rural and state-managed tree plantations (annual planting rising to 60,000 ha and 25,000 ha/year respectively over the next 35 years). These targets are beyond the country's absorptive capacity, and take no account of the availability of land for planting. The mission has therefore made its own projections, which are compared with the Master Plan figures in Annex 3.1 and summarized in Table 3.3. These take into account the capacity of the forestry institutions, allowing for 50,000 ha/year for the management of the natural forest cover and planting of rural and state-managed plantations at rates rising from 2500 ha/year to 25,000 and 10,000 ha/year respectively. As shown in Table 3.3, this would still permit the consumption targets to be met while leaving a margin of production capacity that should be adequate to avoid regional shortages, provided fuelwood transportation is improved as proposed.  - 28 - Table 3.3: Projected Fuelwood Production Capacity and Consumption, 1981-2016 ('000 mi3) Production Capacity Year Master Plan Mission Consumption Balance (1) (2) (3) (2) - (3) 1981 7,200 7,200 4,600 2,600 2001 11,100 7,900 6,800 1,100 2016 13,600 9,600 7,900 1,700 Source: Forestry Master Plan and Bank Staff estimates. Petroleum Products 3.17 Household consumption of petroleum products is relatively unimportant compared with fuelwood. Detailed information is lacking but most of the butane and kerosene used internally is probably consumed by households. As shown in Annex 1.2, consumption of subsidized butane more than quadrupled between 1973 and 1981, from 2,500 tons to 11,000 tons, as a result of the "butanization" policy (para 2.09). The SAR refinery has not been able to meet all the increase in demand, and the balance, amounting to over 40% of consumption in 1981, has had to be imported. According to the projections given to the mission (para 2.25), consumption would nearly triple again by 1990. Consumption of kerosene rose only 16% between 1973 and 1981, from 10,000 tons to 11,600 tons, and is projected to decline to 7,000 tons by 1990. This is probably due partly to substitution by subsidized butane. 3.18 The main issue concerns the justification of the "butanization" campaign. This has not achieved its official objective of halving charcoal consumption and has led to a rising petroleum import bill. The butane subsidy (about US$3.5 million/in the consumption year) benefits people at the higher income levels. It could be more equitably and productively employed in either reforestation (where it could finance about 3500 ha/year) or acceleration of the Ban-ak-Suuf program or, possibly, for arranging the domestic use of peat in urban areas. Agricultural Residues 3.19 Rural households presumably burn part of the crop residues (para 1.17) as fuel, as is the practice in other developing countries, but no information is available on the quantities used for this purpose. Animal manure, as well as some crop residues, are another possible energy source  - 29 - in the form of biogas. A preliminary survey of available crop and animal residues should be undertaken to determine how best to exploit their potential as a renewable energy source. Solar Energy 3.20 The main potential contributions of solar energy to meeting household needs are likely to be in water heating and photovoltaic systems. Solar water heaters have potential for residential (as well as commercial and industrial) applications, provided more effort is devoted to their development. Photovoltaic systems are already operating successfully in special applications (para 2.12), and could serve a variety of needs in isolated locations in the future (e.g. electricity for pumping water, grinding grain, village supply, educational TV and refrigeration) provided their very high costs fall as predicted. Electricity 3.21 Electricity sales to households in 1981 represented 23% of total sales (see para 4.03), which is comparable to Liberia but below the levels in Guinea and Ivory Coast. Supply from the interconnected network is confined to the urbanized areas of western Senegal. Some rural communities are served from so-called secondary and tertiary centers (see paras 4.08-4.09), but the vast majority of rural households, which account for about three quarters of the! population, have no access to electricity supply (see para 4.06), as is the case in other West African countries. 3.22 A study is planned, with Canedian financing, of electricity distribution, with the object of formulating a master plan for rural and urban electrification (see para 4.20). It is hoped to be completed in the first-half of 1984. This should provide more information on household energy needs and give a firm basis for planning the extension of electricity supplies to meet those needs.  - 30 - IV. ELECTRIC POWER SUBSECTOR DEVELOPMENT 4.01 This chapter reviews the past and projected growth of electricity demand and the associated development of the public supply system. Issues relating to electricity pricing are considered in Chapter V and the institutional arrangements in the subsector in Chapter VI. Growth of Demand 4.02 Electricity consumption, as measured by sales to consumers connected to the public supply, rose at 5.3% p.a. on average in the period 1970-1982, from 285 GWh to 530 GWh (Annex 4.1). Year-to-year growth rates were very erratic, ranging from 12.1% in 1973 to an actual decline of about 6% in 1982. Peak demand grew at about the same rate as sales, from 54 MW to 102 MW, as did gross production (332 GWh to 632 GWh). 1/ Consumption per capita 2/ rose from 73 kWh in 1970 to 124 kWh in 1981. The latter figure compares with' an average for all developing countries of 381 kWh and for developing countries in Africa of 182 kWh, and with 1978 figures for Guinea and Ivory Coast of 107 and 154 kWh respectively. 4.03 The pattern of consumption by consumer category has undergone relatively little change, as shown in Table 4.1 (figures are for the interconnected network for 1981). The main feature is the increased share of household consumption and the decline in the share of the industrial category, reflecting the higher growth rate of domestic consumption over the period (7.8% p.a. against 6% p.a.). Table 4.1: Electricity Sales by Consumer Category, Interconnected Network, 1970-1981 (% of Total Sales) 1970 1975 1981 Residential 20 18 23 Commercial 7 6 7 Industrial 71 74 69 Street Lighting 2 2 1 100 100 100 Source: SENELEC 1/ These figures relate to public supply. Information on private electricity production is incomplete, but the industries listed in Annex 4.6 generated about 48 GWh in 1980. 2/ Consumption per capita = gross production - population.  - 31 - 4.04 Several forecasts have been made during the past year of the growth of demand in the interconnected network. The resulting range of estimates is very wide (Annex 4.2). As shown in Table 4.2, they imply average annual growth rates of sales over the period to 1990 ranging from 15.5% (highest) to 6.9% (lowest), comparedl with the actual rate for 1970- 1982 of 5.3% and for 1978-1982 of 2.7%. Table 4.2: Range of Forecasts of Sales and Maximum Demand in the Interconnected Network, 1981-1990 1982 Growth rate, % p.a. (Actual) 1985 1990 1981-1990 Sales, GWh Low 530 638 906 6.9 high 530 920 1680 15.5 Maximum Demand, MW Low 102 116 175 7.0 High 102 168 277 13.3 Source: SENELEC and consultant studies Demand Issues 4.05 The main demand issues relate to: (i) Demand Management: Not enough is being done to modify the demand for electricity through demand management measures other than pricing. A publicity campaign in 1980 to discourage waste seems to have contributed to the low growth of consumption in that year (2.5%). A recent study by consultants 1/ and the mission's own investigations indicate that there is scope for quite significant reductions in electricity consumption in households and offices (e.g. in air conditioning, water heating) and industry (e.g. improving the power factor from 0.8, or even lower, to 0.9 by installation of capacitors, more efficient lighting, refurbishing of internal distribution facilities). A sustained publicity campaign should be undertaken to encourage greater efficiency in the use of electricity. This should be the 1/ "Expertise Energetique d'Entreprises` - Gaucher Pringle, November 1981.  - 32 - responsibility of the new Office of Energy Conservation (para 6.16) in the Energy Department: of MIDC, working in cooperation with SENELEC. Technical audits should be carried out of major electricity consumers to determine the scope for savings and how to achieve them. SENELEC should also establish a technical service to provide advice to consumers on how to avoid waste. (ii) Demand Forecasts: The demand forecast on which SENELEC's proposed development program (see para 4.16) is based implies an average growth rate of 7.1% p.a. to 1990, which compares with the actual rate in the period 1970- 1982 of 5.3% p.a. The recent growth rate has been even lower, with increases of only 2.5% in 1980, 6.1% in 1981 and an actual decline of around 6% in 1982. The reasons for this decline should be carefully analyzed to determine whether it is a temporary aberration or heralds a continuation of sharply lower growth rates. In the meantime, the projection used for the recent tariff study (see para 5.11), based on 4% p.a. growth through 1985 and then 6% p.a. through 1990 (Annex 4.2), seems more realistic in the light of recent trends, the prospects for the economy and the scope for reducing consumption through demand management. It has been adopted by the mission as an appropriate basis for determining the future development program. Electricity Supply 4.06 Access to electricity supply is largely ccnfined to the bigger population centers in the western part of the country. Virtually all towns with populations exceeding 5,000, numbering about 45, have supply. About 100 of the 240 centers with populations in the 1000-5000 range have also been electrified. Village electrification is limited to areas in the immediate vicinity of these large population centers and to the so-called tertiary centers with their own suppLy. This means that the vast majority of the population living in the smaller population centers (with fewer than 1000 inhabitants), numbering nearly 13,000, are without electricity supply. While the Government aims to provide increasing access to electricity in rural areas, it recognizes that the pace of rural electrification must take into account overall resource availability as well as alternative uses for these resources. In effect, rural electrification must be viewed within the context of an integrated program of energy supply to the rural population. Given that few rural households would be able to afford either the economic cost of electricity or of electrical appliances, a careful clecision must be made to distribute any financial subsidies allocated to the promotion of rural energy to those areas where they can have immediate and widespread impact, for example reforestation and improved cookstove dissemination programs.  - 33 - Existing Facilities 4.07 Public electricity supply is mainly limited to the western coast al area of Senegal (see map IBRD 16654), served by the interconnected network operated by the national power company, Societe Nationale d'Electricite 1/ (SENELEC). As shown in Annex 4.3, this comprises two steam stations totalling 148 MW, plus a 16.5 MW combustion turbine, in Dakar, together with 11 MW of diesel generating capacity at Saint Louis and Kaolack. The largest units are 30 MW. Cap des Biches, the more modern of the two steam stations, provides nearly 80% of the total supply (Annex 4.4). Transmission is by 90 kV and 30 kV lines. 4.08 Apart from the interconnected network there are about 20 secondary centers for which SENELEC is also responsible, where electricity is supplied by small diesel generators (Annex 4.5 and map). These range in size from 15 KW to 100 KW, with a total capacity of about 5 MW. They account for less than 2% of total public supplies (Annex 4.1). 4.09 Outside the main public supply system, there are also some rural communities, the so-called tertiary cent:ers, which have their own small generating facilities, financed by the communities themselves or by the state. 4.10 Some private industries have their own generating plants, either for normal supply or for emergency use in the event of an interruption of public supply. Full details of these are not available, but some of the most important, with a combined capacity of around 24 MW, are listed in Annex 4.6. About two-thirds of this capacity is steam plant in the sugar industry, using partly bagasse and partly fuel oil, while the rest consists of diesel units. Supply Trends 4.11 Supply capability has been consistently excessive in relation to demand. In the period 1970-1981, the margin of spare generating capacity never fell below 60% of peak demand, and was 62% in mid-1982 (Annex 4.1). By SENELEC's own reserve criterion (the largest unit in the system plus the diesel sets at Saint Louis and lKaolack), there is about 25 MW of excess generating plant at present, representing some 14% of the installed capacity. 4.12 System load factor 2/ has remained around the relatively high level of 70% (Annex 4.1), reflecting t:he large share of total supply taken by industries, such as phosphates, which operate fairly continu- 1/ Previously known as Societe Senegalese de Distribution d'Energie Electrique (see para 6.08). 2/ Annual production (in GWh) -? (Maximum demand [in MW] x 8760 hours).  - 34 - ously. Both main power stations are well maintained and operated. Average availability at both in 1981 was 85%. There does not appear to be any problem in procuring spare parts except in the case of the three 3 MW standby units at Bel-Air, on account of their age (42 years). Power consumption by station auxiliaries is high at Bel-Air (9%, compared with 6% at the more modern Cap des Biches station). 4.13 Productivity, as measured by kWh produced per employee, rose at about 2.4% p.a. between 1972 and 1981, from 281 to 348 kWh, although there has been some decline in the actual level since 1979 (Annex 4.7). KWh sales per employee rose at about the same rate. The number of customers served per employee fell from 86 in 1974 to 80 in 1981, but this is still higher than in any African developing country for which data are available except Mauritius. 4.14 Power station oil consumption has risen somewhat more slowly than electricity production reflecting an improvement: in average thermal efficiency with the installation of the two 30-MW steam units at Cap des Biches in 1975 and 1978. As shown in Annex 4.8 and summarized in Table 4.3, average specific consumption declined from 307 to 298 gms/kWh. The 1981 oil consumption represented about one quarter of the national total. Fuel oil has accounted for over 94% of requirements throughout the period, and gas oil for the balance (apart from a minimal contribution of natural gas in 1981). Table 4.3: Fuel Consumption for Electricity Generation, 1973-1981 ('000 toe) 1973 7 1977 % 1981 % Fuel Oil 118.9 97.1 148.1 98.5 184.5 95.9 Gas Oil 3.6 2.9 2.2 1.5 7.6 4.0 Natural Gas - - - - 0.2 0.1 122.5 100.0 150.3 100.0 192.3 100.0 Source: Calculated from SENELEC data 4.15 Transmission and distribution losses rose from 7% to 13% between 1975 and 1979, although the 1981 figures imply a fall to 10% 1/. Parts of the transmission network are in poor condition. The oldest 90-kV lines, constructed in 1959 in the Cap des Biches region, have an 1/ There are doubts about the accuracy of this figure because of the unreliability of network metering.  - 35 - estimated residual life of less than five years and are liable to serious failure within this period if the deterioration apparent in some of the pylons and conductors is allowed to continue unchecked. In the 30-kV system, where some of the lines are over 50 years old, the average power factor is low (0.82, compared with a minimum desirable level of 0.9). Some of the 30-kV lines are operating at the limit of their transfer capability. The line from Thies to Saint Louis, for example, is weak and capable of delivering only 2 MW at Saint Louis, with losses of 24%. As a result, Saint Louis now relies mainly on its own recently installed diesel sets, which are operating on base load. A similar situation is expected to arise at Kaolack, where additional diesel sets have just been installed (Annex 4.3). This means that these two centers will, in effect, be operating independently of the main network. Development Program 4.16 SENELEC's proposed generation and transmission program to 1990 is based on a master plan for the development of the power system to the year 2005 1/ financed by the Canadian International Development Agency. Based on the consultants' medium forecast (Annex 4.2), the resulting program is much too large, since it would require the addition of 285 MW of generating plant (Annex 4.9), comprising two 15-MW combustion turbines for installation in 1984 and 1986; two 30-MW peat-fired steam 2/ units in 1986 and 1988; three 40-MW hydropower unil:s, representing half the total capacity of the Manantali project, in 1988; and the 3 x 25-MW Kekreti hydropower project in 1990. 4.17 All the existing steam units at Bel-Air (60 MW), would be retired during the period, resulting in a net addition by 1990 of 225 MW. A more stringent reserve criterion is assumed than at present, requiring that the margin of spare capacity over the projected maximum demand should equal the sum of the two largest units in the interconnected system, and that the loss of load probability should not exceed two days/year initially (compared with three days/year at present), falling to one day/year by 1990. 4.18 As shown in Annex 4.9 and summarized in Table 4.4, the proposed development program would raise installed capacity in the interconnected network from 164 MW to 389 MW, implying a significantly more rapid rate of increase (10% p.a.) than that projected for maximum demand (7.1% p.a.I). As a result, the margin of spare generating capacity would represent over 100% of peak demand in 1990, which is even more than its currently high level of 62%. 1/ "Plan Directeur - Secteur de l'Energie Electrique" - Shawinigan Engineering Company Limited, July 1981. 2/ The consultants' study specified coal or peat, but the SENELEC program assumes these units will be peat-fired.  - 36 - Table 4.4: Proposed SENELEC Program Interconnected System, 1/ 1981-1990 Average Growth 1981 Rate, % p.a. (actual) 1985 1990 1981 - 1990 Maximum Demand, MW 101 127 187 7.1 Installed Capacity 2/ MW 164 179 389 10.0 1/ Excluding Saint-Louis and Kaolack (para 4.15). 2/ Net of retirements Source: SENELEC 4.19 Since the mission of June 1982, SENELEC has prepared a revised development program which results in a reduced requirement for capacity additions in the 1980s. This program is being reviewed by the Government. Distribution of Electricity 4.20 The present master plan for power development covers generation and transmission only. A further study is proposed, with Canadian financing, to prepare a distribution master plan covering both urban and rural areas for completion by early 1984. Supply Issues 4.21 The main supply issues relate to: (i) Existing Facilities There is a lack of balance between the provision for generation and for transmission and distribution. The general condition of the generating plant is good, and the margin of spare capacity is not merely adequate but excessive. Ia contrast, parts of the transmission and distribution network are at the limit of their capacity and in poor condition, resulting in high losses and additional expense (e.g. in installing standby diesel sets at Saint Louis and Kaolack). (ii) System Planning The master plan for power system development will need to be updated to take account of the results of the peat studies now under way, the latest cost information on the coal-fired and hydropower  - 37 - options and a revised load forecast (para 4.05). The proposed distribution planning study also includes provision for this updating, but SENELEC planning staff should be associated with the revision to obtain experience in the use of- the computer model so that they will be able to carry out future updatings. (iii) Hydropower Options Power supply to Dakar from the Manatali dam located in Mali would require international arrangements for the security of supply and therefore substantial cost advantages over alternative solutions before it becomes an attractive option. 1990 seems a more realistic date for the Manantali project than 1988. The capital costs assumed for this project (Annex 4.10) are also too low. They assume that the whole cost of the dam can be treated as sunk, with no allocation to power, on the grounds that the decision to construct the dam for irrigation and navigation purposes has already been taken, without any commitment at this stage to the power component. However, negotiations with OMVS have not started on this important matter. Furthermore, it is understood that the design height of the dam was chosen to allow for optimum exploitation of the hydropower potential (see para. 2.08) and that a lower dam would be sufficient for the irrigation and navigation purposes alone. The cost differential can hardly be considered as a sunk cost at this stage since construction has scarcely begun, and presumably it would still be possible to revert to a lower height for the dam if it were decided not to build the power station. The other participants maay also wish to charge Senegal a rent" for the use of the dam to generate power. The costs of the 760 km transmission line from Manantali to western Senegal may also be underestimated, since they assume a line of 220 kV, which seems too low. The costs of the other potential hydropower projects shown in Annex 4.10 are even more uncertain and will need to be reviewed to determine their relative priority for development. A specific issue concerns the relative merits of the Kekreti and Sambangalou projects on the Gambia river. At present Kekreti appears to be preferred, although the Sambangalou costs quoted are lower. 1/ This is said to be due to opposition to Sambangalou by Guinea, where it would cause extensive flooding. However, the situation should be reviewed in the light of Guinea's recently reported readiness to support the project (para 2.07). 1/ The Sambangalou costs may not include the costs of compensation for any flooding of agricultural land and population resettlement in Guinea.  - 38 - (iv) Other Supply Options At the estimated price for imported coal of US$100/ton (para 2.14) a coal-fired station would be cheaper than the thermal alternatives. The estimated generating cost for a 30-MW coal station is about US cents 7/kWh, which provides a rough benchmark by which other options should be judged. In addition to hydropower, these other options should include peat, surplus bagasse (para 2.10) and groundnut shells (para 2.11). In considering peat, it is most unlikely that a first peat-fired unit could be in operation before 1988, given the lead-time for peat development (para 2.06). (v) Rural Electrification The rural electrification compo- nent of the proposed distribution master plan study should consider not merely the conventional alternatives of supply from the interconnected network or by diesel generator but other feasible options for meeting rural energy needs, even though some of these options, such as photovoltaic systems (para 2.12), may become economic only in a few years' time. (vi) Development Program The proposed generating plant program to 1990 (para 4.16) appears much too large. The mission's preferred alternative (Annex 4.9) assumes the more realistic growth rate of demand used for the revised tariff study (4% p.a. to 1985, 6% p.a. subsequently) and the retention of the existing Bel-Air capacity rather than its retirement. It also retains the present reserve capacity criterion, since adoption at this stage of the much stricter criterion proposed could result in an even more excessive margin of capa- city, given present uncertainties about the future trend of demand. On this basis the 15 MW combustion turbine proposed for 1984 would not be required until 1986, followed by a 20-30 MW steam unit (coal or peat) in 1988 and three 40 MW units at Manantali in 1990. 4.22 This program would result in marked changes in the pattern of electricity supply, because of the introduction of the first coal/peat and hydropower plants. As shown in Table 4.5, the main feature would be a decline of nearly 100,000 toe, or about one half, in the consumption of oil for electricity generation by 1990.  - 39 - Table 4.5: Pattern of Supply in the Interconnected Network, 1981-1990 - Mission Projection 1981 1985 1990 (actual) Generation, GWh Oil-fired 656 760 373 Natural Gas-fired 7 - - Coal or Peat-fired - - 197 Hydropower - - 505 663 760 11075 Fuel Consumption, '000 toe Fuel oil 184.5 199.4 97.6 Gas oil 7.6 Natural gas 0.2 Coal/peat - _ 53-5 192.3 199.4 151.1 Source: SENELEC and Bank staff estimates.  - 40 - V. ENERGY PRICES AND TAXES Petroleum Products 5.01 Petroleum product prices are set in two stages. First, the ex- refinery prices are determined on a cost-plus basis to cover the crude input and operating costs of the refinery and to allow it the 12% return on investment guaranteed by its convention of establishment. The next step is the determination of retail prices which include distribution costs and the net tax or subsidy that the Government imposes on each refined product. Prices at both stages are approved by the Goverment on the basis of recommendations made by the National Hydrocarbons Commission. Given the nature of the convention governing the operations of the SAR, the Government has only limited discretion in varying the level of ex-refinery prices and movements in these prices consequently mirror the changes in the cost of crude imports and the cost of operating the refinery facility. However, this is not the case at the retail sales level where the Government can significantly influence both the level and structure of prices by varying the tax or subsidy element on each product. 5.02 As a rule, ex-refinery prices determined in this manner can be analyzed to provide an indicator of the relative efficiency of meeting a country's petroleum requirements by importing crude oil and refining it locally. To do this, the equilibrium level of ex-refinery prices should be compared with national "direct-import" prices wh:ich are calculated on the basis of internationally posted refined product prices with an allowance made for the cost of transport. Unfortunately, the mission could not carry out this analysis for Senegal because the ex-refinery prices prevailing at the time of its visit were not an accurate reflection of the cost of domestic refining. This is because these prices had been frozen since November 1981 even though the cost of crude imports had risen considerably, primarily as a result of currency changes. Estimates made by the SAR in June 1982 indicate that ex- refinery prices would have to increase by approximately 10 percent even if the $10-15 million losses incurred as a result of the freeze up to that date were reimbursed to the SAR as a lump sum. 5.03 Since then, both ex-refinery and retail petroleum prices have been increased although, because of the devaluation of the CFA franc versus the US dollar, the costs of importing petroleum have also gone up. Once the new ex-refinery prices have been in effect for eight to twelve months it would be useful to compare them with the average price that would have been paid for these petroleum products if they had been imported directly.  - 41 - Table 5.1: Petroleum Product Price Structure, May 1983 US$ per metric tonne 1/ Ex-Refinery Distribution Net Taxes Retail Product Price 2/ Margin (Subsidies) 3/ Price Super Gasoline 422 111 516 1,049 Regular Gasoline 401 99 510 1,010 Kerosene 350 84 58 492 Gas Oil (Regular) 354 75 133 562 Gas Oil (Marine) 354 41 (3) 392 Gas Oil (Fishing Fleet) 354 41 (137) 258 Gas Oil (SENELEC)4/ 304 48 5 357 Fuel Oil 1500 205 31 33 269 Fuel Oil 3500 190 30 31 251 Fuel Oil 3500 (SENELEC) 190 30 (2) 218 LPG (12.5 kg bottle)4/ 340 245 (393) 192 LPG (2.7 kg bottle)47/ 340 245 (493) 92 1/ Converted at US$1 = 350 FCFA 2/ Ex-refinery price is exclusive of import duties paid by the SAR and recovered in its sales price. 3/ Includes import duties, refinery tax, specific tax, contribution to price stabilization fund, and special tax on regular and premium gasoline. 4/ Prices for July 1982, converted at US$1 = 330 FCFA. Source: Staff estimates based on data from GAS, GPP, and SAR. 5.04 Table 5.1 above shows the structure of petroleum prices prevailing in May 1983. These prices are characterized by a heavy tax on gasoline and to a lesser extent kerosene and diesel, which is partly used to offset large subsidies on LPG and gas oil sales to the fishing industry. The rationale for these subsidies and their efficiency in achieving their ostensible objectives needs to be carefully reviewed. This applies particularly to the "butanization" campaign which has signally failed to halve consumption of: charcoal as planned and whose main beneficiaries appear to have been the relatively better off households. This subsidy amounted to $470,000 per month in 1982 and its continuation is difficult to justify. Secondly, there is the question of whether subsidizing fuel costs is an optimal means of supporting the  - 42 - growing and important fishing industry. Alternative and more direct support might well achieve the same objective without the administrative complications associated with ensuring that the subsidized sales of a widely used fuel indeed go to the intended segment of the market. 5.05 The need to examine the tax and subsidy structure for petroleum products is made even more important by the fact that the fiscal contribution of this sector has declined sharply in the last year. Until 1981 retail prices were generally increased in line with costs and taxes on petroleum products provided the Government with about 7% of its current revenues (around $50 million in 1981). 1/ In 1982 the situation deteriorated because of the freeze on retail and ex-refinery prices. If ex-refinery prices were increased to the level required by the SAR convention and if retail prices remained unchanged, then the Government would have to forego about US$8 million in tax revenues in the second half of 1982, which is approximately one-third of the total taxes it would have collected from the sector over this period. While the appropriate level and composition of Government revenues is beyond the scope of this report, it is difficult to justify a fall in the contribution of the petroleum sector at a time when the overall resource position of the country is heavily strained and the resource requirements of the energy sector are large and growing. Electric ity 5.06 The present electricity tariff based on long-run marginal costs of supply (LRMC) was introduced in 1977, 2/ fol]lowing a study by Electricite de France (EDF). As shown in Annex 5.1, it provides for a fixed monthly charge per kilowatt, plus an energy charge per kWh consumed, except for domestic consumers, who pay only an energy charge. The energy charge is of the declining block type. The rates vary according to the voltage of supply (HT, MT or LT), and also, for industrial customers (MT and HT), to whether supply is taken during the peak (7 pm - 11 pm) or off-peak period (all other hours). The only concessionary rate is for small domestic consumers, with a monthly consumption not exceeding 20 kWh, who currently enjoy a discount of about 17% compared with other domestic consumers. The final charges paid by the various categories of consumer include value added tax (currently 7%) and also, in certain localities, communal tax of 2.5%. 5.07 The tariff provides for rates to be adjusted according to a formula which relates the price of electricity to the cost of fuel, labor costs and material costs, as described in Annex 5.1. In practice, however, electricity rates have been periodically adjusted only for 1/ See Annex 5.5 for the historical evolution of retail petroleum prices. 2/ Some customers continue to be charged under the old tariff, based on prices per kWh which vary according to the use for wllich the electricity is required (lighting, appliances, electric drive, etc.)  - 43 - changes in the cost of fuel and not for changes in labor and material costs. As a result, the rates charged have been lower than they would have been had the price variation formulae (which differ according to the voltage of supply) been correctly applied. In 1981, for example, the average price realized by SENELEC was CFAF 38.7/kWh (about US cents 14/kWh), or about 8% below the figure of CFAF 42/kWh which would have resulted from strict application of the formulae. 5.08 In real terms, the average price charged by SENELEC rose 44% over the 1972-82 period representing an average annual rate of increase of about 4% (Annex 5.2). The main factor in the increase was the almost 400% rise in the real cost of oil for electricity generation (nearly 19% p.a.), which more than offset the scale and operating economies achieved through the introduction of larger, more efficient generating plants during the period. As a result, the cost of fuel, which represented about 30% of the average cost of generation per kWh in 1972, was 78% in 1981. 5.09 Most private consumers pay promptly for electricity consumed, the average delay in payment being only one month. Unfortunately, the same is not true of Government departments and agencies, whose unpaid bills currently amount to CFAF 2.2 billion, or the equivalent of about one year's supply of electricity to these consumers. This imposes an undue financial burden on SENELEC, which currently has to pay about 20% interest on bank overdrafts (equivalent to an extra cost of US cents 1.2/kWh). 5.10 Tariff rates are uniform throughout the country, as a matter of deliberate Government policy to encouragea decentralization of economic activities. This results in a substantial subsidy to consumers in the secondary centers, where the costs of supply (from small diesel sets) are high. 5.11 Government policy is to review electricity tariffs periodically to take account of changing patterns of consumption and costs of supply as the power system expands. The first such review of the 1977 tariff, financed under the IDA power credit, was undertaken in 1982 by EDF. This study also prepared updated estimates of LRMC for the SENELEC system. Subsequent to this study, the Government approved a small increase in the level of electricity tariffs, but the structure of tariffs was left unchanged. Consequently there are a number of significant differences between the structure of the current tariffs and of the estimated LRMC. The more important of these, as shown in Table 5.2, are (i) the much higher capacity cost, except for the main phosphate producer (TAIBA) - for LT consumers, for example, the LRMC is over CFAF 4000/kW/month, compared with the existing tariff rate of CFAF 1000/kW/month; (ii) the difference between peak and off-peak energy LRMC for LT consumers, which is not reflected in the present tariff; and (iii) energy LRMC do not decline for LT consumers with increasing consumption. The mission recommends that these anomalies in the electricity tariff structure be addressed during the next round of tariff revisions scheduled for end 1983.  - 44 - Table 5.2: Existing Electricity Rates and Estimated Long-Run Marginal Costs (LRMC) TARIFF STRUCTURE (MAY 1983) LRMC Capacity Energy Capacity Energy Charge Charge Charge Charge CFAF/kW/Month CFAF/kWh CFAF/kW/Month CFAF/kWh Peak Off Peak Peak Off Peak HT 1/ Taiba 2/ 1083 37.21 28.03 798 35.52 28.52 Other - - - 2499 38.98 29.14 MT 1/ General 750 51.89 36.70 ) 2226 51.08 38.58 Special 3/ 1140 37.39 27.80 ) LT 1/ Domestic, Special 4/ - 63.11 63.11 ) -36.59 -36.59 ) Domestic General 4/ - 75.35 75.35 ) -36.59 -36.59 ) 4093 63.91 47.14 Commercial 4/ - two-part 930 53.93 53.93 ) -45.77 -45.77 ) - single - 78.41 78.41 ) -45.77 -45.77 ) ) Street Lighting 800 55.57 55.57 ) 1/ HT = 60 and 90 kV, MT 5.5, 6.6 and 30 kV, LT = below 5.5 kV 2/ Phosphate works. 3/ Free zone 4/ Declining block tariffs - higher figure given is for first block of consumption. Source: SENELEC and EDF tariff study. 5.12 The future tariff increases will also need to address SENELEC's financial problems which are making it increasingly difficult for the company to finance an adequate share of its investment requirements from internal cash generation. In the short term, SENELEC's liquidity needs to be improved through the settlement of short term debts and the  - 45 - building up of working capital. At the same time, measures must be taken to improve the company's long term creditworthiness. These include the optimal use of available resources and the establishment of an efficient financial management system, as well as a reduction in its high debt/equity ratio. The increase in electricity tariffs must therefore constitute an element of a recovery plan which is necessary to the future of the company, a plan which would also include the recovery of arrears, an increase in equity, and a realistic assessment of medium term investments. SENELEC's financial projections 1/ for the 6th Plan (1982- 1986) imply a 1986 average price of CFAF 44.05/kWh (US cents 16/kWh) at 1981 prices,, which would represent an average rate of increase of 2.6% p.a., compared with 3.2% p.a. in 1976-1981. 5.13 The main electricity pricing issies are: (i) The non-payment by Government departments and other public bodies of arrears due for electricity supplied by SENELEC. They should be required to pay off these arrears over an agreed period of, say, one year. (ii) The failure to adjust electricity rates in accordance with the price variation formulae, which also results in a loss of revenue to SENELEC. The revised formulae in the proposed new tariff should be applied automatically at stipulated intervals. (iii) The existing tariff structure and the rates charged, could with relatively minor changes reflect the long-run marginal costs of supply, determined in the new EDF study. There will still be the need for a further adjustment to ensure that: SENELEC achieves its financial objectives. The adjustments should be made in such a way as to minimize distortion of the incentive/disincentive effects of the strict marginal-cost based tariff rates. (iv) The declining tariff rates enjoyed by the domestic consumers are inconsistent with the desirability of encouraging consumers to reduce rather than increase their electricity consumption. The revised tariff will provide an opportunity t:o end this and also the other anomalous arrangement permitting some customers to pay at the old, pre-1977 rates. 1/ Based on 25-30% self financing; 5% p.a. real increase in the price of fuel; 10% p.a. general inflation; a ratio of electricity sales to gross electricity production of 84%; average specific fuel consumption 299 grams/KWh, 12.88% p.a. increase (in current CFAF) in the average price of electricity; 4.56% p.a. increase in staff employed; 5.5% p.a. increase in operating expenses; and 4% p.a. increase in sales.  - 46 - Fuelwood and Charcoal 5.14 The Government sets retail prices for fuelwood and charcoal but is not able to fully enforce them. The Government also charges a stumpage fee for a permit to take wood from the natural forests or official plantations. The current retail prices in Dakar and fees for permits are shown in Table 5.3 (the detailed make--up of the fuelwood price is shown in Annex 5.4). When demand exceeds supply, the wholesale price in Dakar may be greater than the official retail price, especially for charcoal. The retailers then reduce the quantity sold at the official retail price by as much as half. Currently, the difference is about CFAF 20/kg of charcoal, as shown in Table 5.3. Table 5.3: Permit Fees and Retail Prices for Fuelwood and Charcoal (May 1983) Forestry Tax or Controlled Effective Stumpage Fee Retail Price Retail Price CFAF CFAF/kg CFAF/kg Fuelwood (a) 250/stere 1/ 20 20 (natural forest) (b) 2000/stere 1/ 20 20 (official plantations) Charcoal 300/quintal 2/ 30