Report No 4741-ET Ethiopia: Issues and Options 47411 in the Energy Sector July 1984 Report of the joint UNDP/World Bank Energy Sector Assessment Program This document has a restricted distribution Its contents may not be disclosed without authorization from the Government, the UNDP or the World Bank _________E E_l~~ JOINT UNDP/WORLD BANK ENERGY SECTOR ASSESSMENT PROGRAM REPORTS ALREADY ISSUED Country Date No. Indonesia November 1981 3543-IND Mauritius December 1981 3510-MAS Kenya May 1982 3800-KE Sri Lanka May 1982 3792-CE Zimbabwe June 1982 3765-ZIM Haiti June 1982 3672-HA Papua New Guinea June 1982 3882-PNG Burundi June 1982 3778-BU Rwanda June 1982 3779-RW Malawi August 1982 3903-MAL BangLadesh October 1982 3873-BD Zambia January 1983 4110-ZA Turkey February 1983 3877-TU Bol i17a Arril 1983 4213-BO Fiji June 1983 4462-FIJ Solomon Islands June 1983 4404-SOL Senegal JuLy 1983 4182-SE Sudan JuLy 1983 4511-SU Uganda July 1983 4453-UG Nigeria August 1983 4440-UNI Nepal August 1983 4474-NEP Cambia November 1983 4743-GM Peru January 1984 4677-PE Costa Rica January 1984 4655-CR Lesotho Januarv 1984 4676-LSO Seychelles January 1984 4693-SEY Morocco March 195.4 4157-MOR PGor -Ug aL April 1984 4824-PO yigOr 'ay 1984 4642-NIR FOR OFFICIAL USE ONLY Report No. 4741-ET ETHIOPIA ISSUES AND OPTIONS IN THE ENERGY SECTOR July 1984 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 Ethiopia has one of the lowest levels of per capita energy con- sumption in the world. More than 90% of this energy is derived from bio- mass fuels and is almost entirely used in household cooking. The supply of these fuels historically has resulted in massive deforestation and soil erosion which now threatens an otherwise fertile agricultural base. Major programs of aforestation covering 5% of Ethiopia's landscape are required to prevent complete deforestation, to redress the imbalance between supply and demand for woodfuels for cooking, and to restore and protect the capacity for sustained agriculture in the populous highland regions. Ethiopia has very substantial hydropower resources and some small reserves of natural gas, both of which could make an important con- tribution to future economic development. The demand for petroleum pro- ducts has grown in recent years to the point where their cost has become a major burden on the balance of payments. This report presents a stra- tegy for greatly augmenting the supply of household fuels through refor- estation, through the production of new biomass fuels, and through cook- ing fuel demand management. Measures are recommended which will help define the most economic means of exploiting hydropower and natural gas, and which will lead to a better understanding of the economic potential of other prospective energy resources. In addition, measures are out- Lined to enhance the efficient use of expensive electricity and imported petroleum. The energy sector investment needed between now and 1993 is estimated to be US$1.7 billion. Recommendations also are made on energy pricing and on strengthening institutional and manpower capabilities to implement the required investment program. I ABBREVIATIONS AND ACRONYMS Abbreviations bbl barrel bd basic density Btl British Thermal Unit ft cubic feet GDP gross domestic product GWh gigawatt hour ha hectare kg kilogram kVA kilovolt ampere kW kilowatt kWh kilowatt hour LPG Liquified petroleum gas mai mean annual increment mc, mcwb moisture content, moisture content wet basis m3 cubic meter M,m thousand MM million MVA megavolt ampere MW megawatt MWh megawatt hour OD oven dry sv solid volume tonne metric tonne toe tonnes of oil equivalent tpa, tpy tonnes per annum, per year Acronyms EELPA Ethiopia Electric Light and Power Authority ERESA Eritrea Region Electric Supply Authority ENEC Ethiopian National Energy Committee EPC Ethiopian Petroleum Corporation ERR Economic Rate of Return FAWCDA Forestry and Wildlife Community Development Authority FCMC Fuelwood and Charcoal Marketing Corporation FRR Financial Rate of Return MME Ministry of Mines and Energy MOA Ministry of Agriculture PA Peasant Association SWCD Soil and Water Conservation Department WFP World Food Program This report is based on the findings of an energy assessment mission to Ethiopia in May-June, 1983. The mission assignments were H. Wackman, Mission Leader; K. Newcombe, Deputy Mission Leader; J. Besant-Jones, power system planning; C. Killoran, power engineering; A. Oduolowu, petroleum exploration and supply; T. Nayar, refinery engineering; J. Boroumand, energy balances and projections; T. Wood (cons), stove efficiency ana- lyses; F. Werner (cons), forestry planning; P. Duiker (ILO cons), manpower planning; T. Bierschenk (ILO cons), manpower planning; B. Robinson (cons), energy conservation engineering; E. Peries, Administrative Assistance. Ken Newcombe was the principal author of the report and Linda Sullivan was the principal typist. The contribution of the ILO to evaluation of man- power training and development issues is gratefully acknowledged. CURRENCY AND FUEL EQUIVALENTS Currency US$1.00 = B2.07 1B = US$0.48 CONVERSION FACTORS 1 MJ = 948 Btu = 239 kcal = 0.278 kWh 1 kWh = 3.6 MJ 1 ft3 natural gas = 1,000 Btu = 1.055 MJ l MMCF natural gas = 1 billion Btu = 24.7 toe Major Fuels 1/ toe 2/ M3 MJ/liter per metric tonne per metric tonne Petroleum Products Aviation Gasoline 33.1 1.11 1.43 Gasoline 34.3 1.10 1.37 Kerosene/Jet Fuel 36.7 1.10 1.28 Diesel 37.8 1.07 1.20 Fuel Oil 40.1 1.01 1.08 L. P. Gas 25.4 1.04 1.77 Woodfuels MJ! kg Fuelwood (air dried) 14.3 0.335 1.50 3/ Charcoal 29.0 0.679 Fuelwood (oven dry) 20.0 0.468 Basic Density of Wood 3 (unless otherwise stated) 500 kg/m Electricity kWh per TOE Thermal-equivalent (supply) 3,206 (at 27% efficiency) 1/ Conversion factors are approximations only. 2! One tonne of oil equivalent (toe) = 10.2 million kcal = 42.74 thousand megajoules 3/ Solid Volume (SV). Table of Contents Page MAIN FINDINGS AND RECOMMENDATIONS ....................... i-xvii I. ENERGY AND THE ECONOMY OF ETHIOPIA ...................... 1 The Pattern of Energy Consumption ..................... 3 Future Energy Demand .................................. 7 II. HOUSEHOLD ENERGY ........................................ 10 Introduction .......................................... 10 Preview of the Analysis and Main Findings ............. 17 Short-term Supply Options, 1983-92 .................... 19 Short-term Demand Options, 1983-92 .................... 24 Short-term Strategies ................................. 28 Summary and Conclusions on Options for the Southern Region ..................................... 29 Potential Impact of Short Term Measures ............... 33 Medium- to Long-term Strategies (1992-2012) ........... 34 - Peri-urban plantations ............................ 34 - Existing Addis Ababa Plantations .................. 36 - New Plantations Design and Management ............. 36 - The Need for Peri-Urban Forestry .................. 36 - Achievable Peri-urban Plantation Program .......... 37 Costs of Achievable Forestry Program .................. 40 Economics of Rural Forestry ........................... 42 Implementation of the Achievable Forestry Program..... 44 Consequences of Failure to Act on Rural Afforestation ................................. 45 III. ELECTRICITY .. 47 Institutional Arrangements . . 47 The Power System .. 47 Features of Present Demand . . 51 Electricity Demand Forecasts . . 54 Mission's Demand Forecast .. 59 Medium-term Supply Options .. 60 Demand Management Options .. 62 Long-term Supply Options .. 63 Export Options .. 64 Natural Gas .. 66 Future Supply and Demand Management Options for ERESA . . 68 Financing Power Sector Investment . . 71 IV. PETROLEUM ............................................... 74 Pattern of Supply ..................................... 74 Alternative Supply Options ............................ 75 Product Transportation ................................ 77 Petroleum Exploration and Production Prospects ........ 79 Production of Ogoden Gas ............................ 80 Petroleum Conservation and Substitution ............... 81 Page V. OTHER ENERGY SOURCES ........................... 86 Crop Residues .86 State Farm Crop Residue Resources .86 Coffee Residues ..................................... 90 Bagasse .91 The Market for Crop Residues .94 Ethanol .94 Solar Water Heating .95 Other Energy Forms .96 Coal-Lignite .96 Geothermal .96 Biomass Gas Sources .97 VI. ENERGY PRICIN .98 Woodfuels .98 Petroleum Pricing ....................................... 100 Electricity Pricing ..................................... 103 Tariffs ............................................... 103 Economic Costs ................... 103 Financial Requirements ........................ 105 Implementation Strategy ........................ 106 Summary ........................ 108 VII. INVESTMENT, MANPOWER, AND INSTITUTIONAL REQUIREMENTS.... Ill Energy Sector Investment ..............................111 History and Plans ..1.................. Il Future Investment ................ 112 Manpower Issues ................ 115 Institutional Arrangements for Energy Planning and Management ................ 118 The Role of the Energy Secretariat .119 Production and Marketing of Biomass Fuels .120 Coordination in the Petroleum Sector .121 ANNEXES 1.1 Ethiopia: Energy Balances, 1982 (103 x toe) .123 1.2 Summary of Methodology for Demand Forecasting in the Ethiopian Energy Assessment .124 2.1 Forest Resources .128 2.2 Logging Waste Plus Non-commercial Species Available for Energy, 1981/82 .129 2.3 An Illustration of the Place of Carbonization in a Program of Upgraded Forest Management .131 2.4 Sawmill Production, Available Waste, 1982/83 .132 2.5 Optimum Schedule of New 3iomass Energy Production and Supply to Southern Region .134 2.6 Supply and Demand for Addis Ababa as the Prime Target Market of the Southern Region .135 2.7 Wood Equivalence .136 Page 2.8 Peri-Urban Plantation (Urban) 5000 Inhabitants Volumes in solid m3 on Bark ................ 137 2.9 Peri-Urban Plantation: Silvicultural Model I, Lowlands, Below 1800 m a.s.1 ............... 138 2.10 Rural Energy Forestry Requirement .............. 147 2.11 Simple Cost Estimate for Rural Afforestation Budget ..... 148 2.12 Estimates of Number of Cattle and of Dung Production by Region of Ethiopia . ............ 149 2.13 Fuelwood Deficits and Dung and Crop Residue Consumption Estimates by Region, Ethiopia 1981/82 ..... 150 2.14 Forecast Difficiency of Woodfuels Supply to Urban Areas ................... 151 3.1 Status and Electricity Demands of Boiler Electrification Program . .152 3.2 EELPA ICS Mission's Power Demand Forecast. 153 3.3 Existing and Committed ERESA ICS Generation. . 154 3.4 Projected Investments in the Ethiopian Power Sector ..155 4.1 Refined Petroleum Product Supply Options . . 156 4.2 Petroleum Products Transportation Analysis, Asseb-Addis Ababa . ..157 6.1 Import Parity Build-up (C.I.F. Asseb) . .158 6.2 Petroleum Price Build-up and Maximum Selling Price at Some Locations in Ethiopia . .159 6.3 Build-up of Economic Cost of Petroleum Products in Ethiopia, 1983 ................ 161 7.1 A Tentative Schedule of New Energy Sector Investments in Ethiopia 1983/84 - 1992/93 ............... 162 Figure 25403: Pattern of Deterioration in Ethiopian Agroecosystems ........................................ 15 Maps 17512: Agro-ecological Zones of the Region 17513: Ethiopia Energy Assessment 14257R1: Country Map MAIN FINDINGS AND RECOMMENDATIONS Overview 1. Ethiopia recovered well from a period of internal conflict in the mid-1970s to record a GDP growth of 5% a year between 1978 and 1980. However, despite good financial management, the economy stagnated after 1981 as the terms of trade worsened, drought persisted and foreign ex- change reserves declined. Ethiopia remains one of the poorest of nations, with a per capita GDP of US$140 in 1982. The Ethiopian economy is predominantly rural, with 84% of its 32 million people living in villages and about three-quarters of the country's farms located more than a half day's walk from an all-weather road. This rural disposition is clearly reflected in the patterns of energy consumption summarized below. Of an annual per capita energy consumption of about 245 kg of oil equivalent, 93% is in the form of firewood, charcoal, crop residues and dung, and per capita consumption of petroleum products (17 kgoe) is very low even by regional standards. Final Energy Consumption, 1982 ('000 toe) Biomass Share Fuels Electricity Petroleum Total (%) Industry 30 35 102 167 2.1 Transport 349 349 4.4 Agriculture 1 31 32 0.4 Households 7,402 16 18 7,436 92.8 Commerce/Government 11 21 32 0.4 Total 7,432 63 521 8,016 100.0 Share (%) 92.7 0.8 6.5 100.0 Final Energy Consumption, 1992 Industry 104 98 198 400 3.7 Transport 23 - 576 599 5.5 Agriculture - 2 177 179 1.6 Households 9,570 42 73 9,685 88.8 Commerce/Government 13 29 42 0.4 Total 9,697 155 1,053 10,905 100.0 Share (%) 88.9 1.4 9.7 100.0 - 11 - The GOE intends to create a surge of economic recovery and growth through the vehicle of a ten-year plan involving some US$16 billion in invest- ments geared to achieve an average of 7.5% p.a. GDP growth between 1983/84 and 1992/93. There is no doubting the sincerity and fervor with which this objective is being addressed, though it will require greatly increased energy sector investment to further social and economic devel- opment, together with more effective energy pLanning and management. This is because Ethiopia has still to systematically develop its wealth of hydropower ressources, a locally significant gas resource and a variety of other energy resources with economic potential to meet the growing needs of household and commercial consumers. Growth in GDP is more like- ly to range between 3% and 4% a year through 1992, which should call for a desirable minimum level of energy sector investment (including energy and agroforestry) of US$1.6 billion for the 1983/84-1992/93 period, im- plying a major increase in the proportion of public sector investment devoted to energy. With the implementation of the optimal energy stra- tegy devised by the mission (the accelerated investment program scenario depicted in Table 1.7), biomass fuels decrease marginally and electricity and petroleum fuels increase their relative share of total energy use by 49 and 75%, respectively. Household Energy 2. The most important issue in the energy sector, and arguably in the future of Ethiopia, is the supply of household fuels, the related massive deforestation, and the resultant and insidious depletion of agri- cultural resources on which so much economic activity depends (nearly 50% of GDP, 80% of employment and 90% of exports). The increasing scarcity and cost of household fuels, particularly firewood, threatens the ability of the country even to maintain already low incomes and the quality of life of the people, particularly in the rural areas. The Ethiopian landscape has changed dramatically over this century. The estimated cover of closed canopy forests has declined from roughly 40% around the turn of the centuiry to around 3% now, and an estimated 200,000 ha and one billion tonnes of topsoil are lost each year as land is cleared, or forests thinned, to supply new areas for crops, animals and people, and to meet the growing needs for the major fuel, firewood. The cost of firewood in the capital of Addis Ababa soared from US$9 to US$90 per tonne between 1973 and 1983 and now claims up to 20% of household income. To compensate for the worsening firewood scarcity, growing amounts of natural fertilizers in dung and crop residues are being diverted to household fireplaces, reducing crop yields by more than one million tonnes of grain a year. If no more than the present level of replanting is undertaken, in 20 to 30 years all but the least accessible pockets of forest will have gone, large parts of the north will be uninhabitable, and much of the center will be like Eritrea and Tigrai today: subject to persistent drought, crop failure, famine and outmigration. To arrest these trends and meet the estimated need for woodfuel within the next 30 years, some 6 million ha would have to be planted: in effect to reforest 5% of the Ethiopian landscape. Both the land reforms and the community - iii - organizations are in place to implement the required program and mobilizing sufficient financial resources should be possible. Neverthe- less, life-styles will have to change dramatically as tree planting and management will have to be incorporated into the very fabric of rural life. Intervention on the scale envisaged will require unprecedented political commitment. Energy and agroforestry will have to be placed among the nation's top priorities for there to be any chance of success. Although in the short-term the problem will be further aggravated due to long lead times for forest development and high population growth, there are good prospects of lessening the rate of deforestation, provided there is a major increase in the mobilization of resources to meet the problem. High Cost of Petroleum 3. Besides the household energy crisis, the other major energy- related macroeconomic problem is the massive and worsening impact on Ethiopia's trade balance and foreign exchange availability of the cost of petroleum imports (net imports of US$185 million in 1982/83, equivalent to about half of foreign exchange earnings). Unless major demand manage- ment efforts are made, the increased demand, particularly from the trans- port and industrial sectors, will add to the burden through the resulting imports and will continue to constrain economic growth. Measures to alleviate the cost of petroleum imports include serious reconsideration of the future pattern of industrial development; improved technology selection; possible petroleum product pipeline and improvement in the structure and efficiency of the transport fleet; a variety of substi- tution measures (e.g., ethanol for gasoline and possibly natural gas for diesel in transportation, and biomass and natural gas for fuel oil in heat and steam-raising); and improved energy management in industrial and commercial enterprises. Overall Energy Strategy 4. The major themes of Ethiopia's energy strategy for the coming decade should be as follows: (a) To expand the supply of household energy through a greatly in- creased allocation of financial and trained human resources, and to this end, to pay particular attention to, and take con- certed action on: - reforestation, including improved forest management; - expanded development and utilization of substitute fuels, like briquetted agricultural residues; - improvement in efficiency of cooking through the use of stoves with improved fuel economy and the use of aluminum pots and mtads. - iv - (b) The development, execution and prudent financing of a least cost power supply program commensurate with a reasonable level and structure of industrial development and expanded supply to residential and commercial consumers, taking into account the country's limited investment resources, the need for improved power system planning, the potential complimentary role that could be played by natural gas and suitably scaled hydropower developments in the long-term, and the desirability of financ- ing a reasonable proportion of new investment through revised tariEfs for increased cash generation. (c) A major program of demand management, including important changes in the pricing of electricity and some petroleum fuels, rationalization of future industrial structure, selected sub- stitution options (biomass, ethanol, solar heating and possibly natural gas), a petroleum product pipeline and improved trans- port fleets, industrial energy management and oil refining facilities. (d) Expanded exploration efforts and eventual exploitation of do- mestic hydrocarbon resources, with immediate attention to the possible utilization and development of natural gas, and hope- fully, but in the longer-term, petroleum. (e) Expanded participation by the private sector in selected areas, particularly in petroleum exploration, solar water heating, and in the supply of biomass fuels (firewood, charcoal, agricul- tural residues, briquettes) to households. Key Issues and Options Expanding Woodfuel Supply 5. Afforestation. On a large scale, this is the only realistic option for resolving the household energy supply problem in the longer- term. Peri-urban plantations must be established to make urban areas self-sufficient, and rural woodlots and agroforestry on a massive scale are essential to allowing a sustainable energy supply to meet demand in rural areas. The mission estimates the need for forestry of all kinds to be 3.3 million ha planted by 1992, but considers this to be beyond the capabilities of the country in this period due largely to the institu- tional constraints of manpower and management resources, and the massive organizational and logistical efforts involved. An ambitious but poten- tially achievable program of afforestation would be to plant by 1992: (a) 960,000 ha of rural energy-forestry with special emphasis on agroforestry, costing B403 million (US$195 million); and (b) 195,000 ha of peri-urban forestry costing 3343 million (US$166 million). A further ongoing commitment of B1.2 billion (US$0.6 billion) is required during 1993-97 to meet the 2002 demand. A balance between supply and demand could be achieved between 2010 and 2020 depending on the long-term population growth and the future of alternative fuels. The mission has reviewed options for the supply and management of household energy sepa- rately for the northern and the southern regions. The northern region comprises Eritrea, Tigrai, Gondar and Wollo provinces and the southern region the remainder. With full implementation of the short-term mea- sures proposed, only a moderate shortfall need occur in the south by 1992 between absolute need for peri-urban forestry and the achievable planting rate. The shortfall in the north would be about 150,000 ha for peri- urban forestry and almost 1 million ha in rural afforestation by 1992. In short, even with the major effort here proposed, deterioration in northern agro-ecosystems and in household energy supply will increase dramatically this decade before being slowly reversed in the next. 6. Specific priority measures to be undertaken within the achiev- able aforestation program include: (a) establishment of training and research and development facil- ities in agroforestry right down to the peasant farmer level (paras 2.39, 2.46); (b) establishment of a national forestry management information system to optimize species, provenances, and plantation design and management for each ecological zone (para 2.39); and (c) a national forest and tree resource inventory to be conducted as a matter of urgency (para 2.4). In the context of peri-urban forestry, there should be: (a) increased application of fertilizer in the new Addis Ababa plantations to boost yield and reduce rotation length (para 2.33); (b) replanting of the 10,000 ha of dead and dying portions of the existing Addis Ababa plantations by 1992 (para 2.34); and (c) planting of at least 10-15% of the area of state farms with boundary woodlots and shelter-belts (para 2.37). - vi - 7. Additional Household Fuels. By greatly accelerating refores- tation, recovering existing surpluses of woody fuels 1/ from forests and from agriculture, and by more efficient cooking, a near balance can be achieved between supply and demand for household fuels in urban areas in the south and establish the basis for the north's improvement within the decade. During the next decade, when woodfuel supplies from the p'roposed expanded forestry program are minimal, it should be possible to economic- ally supply the equivalent of about 1.7 million tonnes of wood a year by 1992 to towns in the southern region in the form of charcoal from forest clearing and logging operations, and crop residue briquettes from state farms and cash crops. This would be equivalent to about half of urban household energy demand in the south. The crop residues concerned are strictly those centralized and annually burnt off on Ethiopian State Farms, not those dispersed throughout subsistence agriculture which have much higher va]Lue as animal feed and soil conditioners. Furthermore, while the residue harvesting and briquetting equipment to take advantage of this nearly unique State farm energy resource is fully commercial, the scale of the operation proposed here is quite unprecedented and success- ful implementation depends greatly on good management, careful planning and consumer acceptance of these now briquetted fuels. Northern region towns will not have access to these resources, therefore it is recom- mended that at least 40,000 tonnes of kerosene be directed each year to northern cities including Asmara and Mekele. In the southern region, LPG and electricity from hydropower can contribute significantly to household cooking energy supply within five years. Oil-fired electricity is not economical for cooking in the north and should be strongly discouraged in favor of kerosene in the short-term. There appear to be good prospects for improving the efficiency of existing charcoal stoves and electric injera cookers, and for introducing new portable metal woodstoves for urban use, warranting immediate trial and development work for each. Huge improvements in cooking efficiency occur from using aluminum pots, thus population-wide acceptance of cooking with aluminum instead of clay pots should be evaluated immediately. 8. Priority supply measures include: (a) expanding the production of charcoal at Bubeka and other forest clearing sites to 114,000 tonnes per year by 1987 (para 2.13); (b) introducing integrated forest management at selected logging and clear-felling sites with complete conversion to charcoal from all forest residues (para 2.10); (c) programming exhaustive recovery for firewood and charcoal of all savannah forest timber cleared in state farm development (paras 2.12-2.14); 1/ Woody fuels include firewood, charcoal, and crop residues such as straw, stover, coffee skins, and bagasse. - vii - (d) beginning commercial recovery of coffee skin residues in the form of briquettes during 1984, leading to production of 150,000 tonnes per year by 1992 (paras 2.19 and 5.9-5.12); (e) beginning pilot production of fuel briquettes from cereal straw and stover, and cotton residue in 1984/85, following an immediate review of the optimum residue harvesting practices, leading to fully commercial production from all sources by 1986/87 (paras 2.19 and 5.2-5.9); (f) integrating recovery of sawmill wastes as sawdust briquettes and charcoal in the refurbishment of rural sawmills (para 2.11); and (g) improving the efficiency of charcoal production with the use of improved kiln design, including portable metal kilns (para 2.15). The combined capital cost to develop these resources, if fully imple- mented, is 8113 million (US$55 million) and the yield is the equivalent of the annual yield from 163,000 ha of new peri-urban forests. 9. Economics of Rural Afforestation. The main economic justifi- cation for the proposed national afforestation program beyond the direct value of household fuel supply is the benefit to agriculture of allowing high value organic fertilizers--dung and crop residues--to be used in crop production rather than as fuels. The value of grain production foregone each year by using cow dung as a cooking fuel is about one bil- lion B (US$500 million); and this can only be reduced substantially by increasing the supply of woodfuel, given the extremely high cost and foreign exchange requirements of chemical fertilizers. Economic rates of return for rural afforestation investments are estimated to range between 25% and 80%, depending on the local circumstances, making this the high- est value investment in the energy sector. Typical economic returns for peri-urban plantations are estimated to be 20-25%. 10. Implementation. The implementation of the energy forestry program requires an expansion of the charter and of the manpower, infra- structure and administrative resources of the existing extension service of the Ministry of Agriculture, as well as an early clarification of the required interface and cooperation with the Forestry and Wildlife Commun- ity Development Authority (FAWCDA). The mission also recommends the establishment of a stove marketing organization to implement the improved cooking efficiency program. I Power Subsector 11. With due attention to demand management, major investments in new generation systems other than the 150MW Malka Wakana hydropower plant now under construction can be delayed until about 1988, allowing a much lower cost power system development plan to be identified and implemented than is presentLy being contemplated. The rate of demand growth on the - viii - main interconnected system (ICS) of the Ethiopian Electric Light and Power Authority (EELPA) has been greatly overestimated in recently pre- pared forecasts, hence the impending investment of US$29 million to re- habilitate the Abu Samuel plant (6MW) and add a fourth unit to the Finchaa hydropower station (34MW) are not warranted prior to the commis- sioning of the Malka Wakana proposed for mid-1987. Demand management, including conservation measures, can cover a 12-18 month delay in the Malka Wakana project, though the shortfall in capacity occasioned by longer delays can be met only by access to new generating plant, not in- cluding those now under consideration. Key elements in the future demand for electricity include boiler and cooker electrification programs now under implementation, and the level and structure of tariffs which are now distorted and unreasonably low. While there is a good case for elec- tric cooking for a small section of the urban population, there is no longer any justification for the electric boiler program, which has been implemented so late that none of the originally perceived benefits now apply, given that the extended period of hydro-based power surplus will come to an end by about 1986. The annual incremental economic cost of operating electric boilers as compared with the existing oil-fired boilers would be about US$8 million by 1986. On the supply side, plan- ning and evaluation must be expanded to include several very large multi- stage hydropower projects and natural gas from the Ogaden Desert, both of which diminish the status of potentially more expensive and still pro- spective geothermal power resources. In the north, the smaller intercon- nected supply system of the Eritrea Regional Electric Supply Authority (ERESA), which for the foreseeable future will reLy solely on petroleum- fired generators, must in the short-term install more efficient genera- ting plant and review locally available power resources, e.g., geother- mal, and place heavy emphasis on demand management, as the proposed interconnection with the main EELPA ICS is not expected to become econo- mic until the mid-1990s. 12. Specific measures arising from the above include: (a) A genuine least-cost development program for the EELPA ICS should be developed based on the revised demand forecast presented here and choosing between numerous major hydropower systems and natural gas (paras 3.33-3.39). (b) Boiler electrification should not be extended beyond those al- ready installed or committed and then should only be used dur- ing periods of surplus hydropower with the fuel oil or biomass fired boilers (and in the longer-term, natural gas-fired boilers) retained as the primary means of industrial steam raising. Accordingly, EELPA should cease to supply power on concessional terms to electric boilers from 1985 (the time at which present contracts expire), charging thereafter a full economic price (paras 3.7-3.8). (c) A strategy must be devised for future power supply which would protect EELPA from supply constraints in the event of a delay in the commissioning of Malka Wakana but which would not - ix - preclude a low cost long-term generation expansion plan. The early expansion of the Tis Abai power station combined possibly with the Finchaa-Bahar Dar 230kV transmission link, and inter- connection with the Sudanese grid should be considered in this regard (para 3.28). (d) The economic viability of geothermal power should be evaluated in the context of the least cost power system development for the EELPA ICS. Emphasis should be given to exploration for geothermal power in the northern part of the country covering also the ERESA system (para. 3.40). (e) Fuel oil-fired diesel engine generation should entirely dis- place diesel generation and fuel oil-fired steam turbo-alter- nators for supply to the ERESA ICS (paras 3.43-3.44) in the medium term. In the mid to late 1990s, there is the prospect of either natural gas for power production based on the current prospect of natural gas discovery in the Red Sea or. of geother- mal power as noted in (d) above. Petroleum Supply and Transportation 13. The most important problem facing Ethiopia in this area is the burgeoning cost of oil imports. In this respect there are a range of op- tions for supplying refined products and transporting them from the coast to the central highlands market, and there is a good prospect for devel- oping natural gas. In addition, petroleum exploration is being renewed with exploration promotion assistance from the World Bank and from the Soviet Union, which has financed four wells in the Ogaden desert region some 800 km east of Addis Ababa. The petroleum exploration and promotion facility financed by the recent IDA credit must be complemented with rationalized petroleum sector management and lead to a strong effort to engage private sector participation in further petroleum exploration and development. Preliminary mission estimates indicate that the most expensive option for supplying petroleum products over the next 12-15 years is to continue operating the refinery at Asseb in its present form. The competing lower cost options are either to close down the refinery and import all refined product requirements, or invest at least US$70 million in conservation, debottlenecking and secondary conversion faci- lities, resulting in greatly improved fuel efficiency in the refinery, as well as a product mix much closer to the pattern of domestic consump- tion. While the rate of return on new investment appears attractive, the early production of natural gas for local use could welL reduce the economic benefits now foreseen by reducing the market for liquid fuels and greatly altering the required product mix. Nevertheless, further engineering and economic study of refinery expansion/modernization is recommended. Regarding the inland transport of petroleum fuels, the mission finds a proposed 850 km, US$270 million product pipeline from Asseb to Addis Ababa more economic than high-volume efficient truck transportation. However, if there is any prospect of replacing diesel with compressed or liquified natural gas in road transportation in the early- to mid-1990s, the cost of the competing alternative of restruc- tured road transport will be significantly reduced. The refined product pipeline therefore must be subjected to more detailed study in the con- text of broader transport sector issues for its economic viability is closely related to the future growth of liquid petroleum consumption. In line with the finding that natural gas is a potential competitor in the power market, the mission recommends prompt drilling of at least one additional well to better define the volume of gas in place in parallel with a preliminary gas utilization study already funded under the aforementioned IDA credit. 14. Specific measures arising from the above include: (a) a detailed refinery engineering study to establish the costs of the proposed refinery modifications and the maintenance costs for the refinery over the life of the new equipment(para 4.10); (b) detailed study of the product pipeline option with, in the short term, gradual replacement of 24 tonne tanker trucks with 35-40 tonne trucks, road quality permitting, due to the signi- ficant improvement in fuel economy entailed (paras 4.14 and 4.28); and (c) the use of either the joint Ethiopia-Soviet exploration program or an additional program to drill at least one additional well on the Calub structure to better define the existing gas re- source in the Ogaden (para 4.20). In addition, there is need for immediate modification to the refinery's LPG process train, and to refinery operation, to ensure the recovery of a further 8,000 tonnes of LPG now flared, and the production of 54,000 tonnes of kerosene a year. To increase the availability of kerosene for household use and avoid its use as an industrial fuel, kerosene blending in fuel oil to lower viscosity should be minimized and eventually discon- tinued as industrial combustion systems and fuel handling facilities are upgraded, including heated storage tanks and tanker vessels, to cope with high viscosity fuel oil. Petroleum Demand Management 15. In addition to the measures mentioned above, the mission recom- mends that low cost options to improve the economy of petroleum use be impLemented quickly, primarily in industry and transportation, which account for 20% and 67% of petroleum consumption, respectively. In in- dustry, key measures should include: (a) improved combustion control, water treatment, and new heat re- covery systems throughout the manufacturing sector, particu- larly in such energy-intensive industries as cement, textiles, glass and iron-casting (paras 4.22, 4.23); - xi - (b) industrial-scale solar water heating systems in tanneries and in hotels, hospitals and hostels (paras 4.24 and 5.20); and (c) converting cement kilns to being partly fueled with crop resi- dues (para 4.24). In the transport sector there is a serious undercapacity problem which has nearly eliminated routine maintenance, contributing to highly inefficient vehicle operation and fuel consumption rates. The measures proposed by the mission for energy savings in the transport sector will not be fully effective without first investing some B50 million in new and larger trucks and additional servicing facilities. The specific measures to be adopted include: (a) aquisition of diesel pump testing equipment and the servicing facilities for the bus and truck fleet (paras 4.27-4.28); (b) turbo-charged engines and radial tires for the truck and tanker fleet operated on the Addis Ababa-Asseb route (para 4.28); and (c) the implementation of a detailed management information system for the transport fleet and revised policies for truck and bus procurement due to the wide variation in fuel economy, main- tenance requirements, and longevity experienced between dif- ferent makes and models and the substantial economy of large (35-40 tonne) over smaller trucks, particularLy for major bulk commodities such as petroleum products (para 4.29). Other Energy Sources 16. In comparison to the need for expanded fuelwood production, conventional power, and petroleum supply, most other energy sources are of relatively minor importance and should not absorb valuable skilled staff time, administrative, or financial resources out of proportion to their potential to contribute economically to future energy supplies. This is especially true of biogas, producer-gas power generation, solar electricity and wind energy. The exceptions are the recovery of agricul- tural residues from state farms and major cash crops, production of ethanol from molasses, solar water heating and, possibly, geothermal and coal resources. Pilot production of various crop residue fuel briquettes should begin immediately, as should consideration of relocating the geo- thermal exploration effort to the north. Other priority actions in this subsector include: (a) an immediate review of the potential to create surplus bagasse by upgrading the process energy efficiency of the existing Metahara, Wonji-Shoa sugar mills, and the future Finchaa sugar mill (para 5.13); - xii - (b) proceeding with the production of about 29 million liters a year of ethanol as proposed, providing adequate arrangements are made for product storage and blending, vehicle fuel system compatibility and pricing (para 5.18); (c) undertaking a national review to determine the economic poten- tial for solar water heating and the viability of local col- lector production facilities (para 5.20); and (d) accelerating and broadening the scope of the coal resource inventory now in progress (para 5.21). Energy Pricing 17. The most serious energy pricing issues concern electricity tariffs and. petroleum pricing. Petroleum products are generally under- priced against the full economic cost of supply when the shadow price of foreign exchange is taken into account. Net of taxes, only gasoline is priced well above the border price, although with its highly regulated market and mainly non-productive consumption, this is of little conse- quence. The mission recommends that: (a) the price of diesel be increased by about 7% to encourage effi- ciency in use and to raise revenue (paras 6.6, 6.7); and (b) no subsidy be applied to kerosene and LPG, because at their full economic cost they are clearly competitive with the market price of firewood and charcoal and tend to be used relatively more by higher income households (paras 6.6, 6.7). 18. Electricity prices have been kept artificiaLly low in all power systems and huge cross-subsidies occur both within and between the EELPA ICS and the oil-fired ERESA ICS. The declining block tariff structure now in place in which the price per unit declines as consumption in- creases is now inappropriate due to its promotional nature in the face of supply constraints and the heavy cost burden placed on small consumers. Tariffs will have to be adjusted both for financial and economic reasons, as the COE is now rightly insisting that EELPA meet a certain proportion of its future investment requirements from its own revenues. Tariffs in the ERESA ICS and the EELPA SCS systems are as little as one-fifth of the true costs of supply. The mission recommends that: (a) tariffs be raised to reflect the full economic costs of supply and to make a reasonable contribution to the cost of new in- vestment (paras 6.10-6.17); (b) uniform national tariffs be removed in favor of very substan- tial differences in prices, reflecting as far as practical the real variation in costs between supply systems (paras 6.10- 6.17); - xiii - (c) the near-term price of electricity in the ERESA ICS and SCS, and the EELPA SCS, at least meet the variable costs of produc- tion which are 35-40 ec/kWh (paras 6.10-6.17); and (d) time of day charges be applied over an otherwise flat tariff structure in all systems (paras 6.10-6.17). 19. While the Government is attempting to intervene in the woodfuel and charcoal markets to bring prices down, its current impact is minor. The issue is really whether Government should be involved at all in the woodfuel market at the wholesale/retail level when it could concentrate on moderating price levels by adopting policies to enhance supplies from the forests and other sources (see para 22). There appears to be no economic distortion at official government prices for firewood and char- coal, but prices in the private market are well above long-run economic production costs, reflecting the growing scarcity in good quality woody fuels. Manpower Training and Development 20. The availability of skilled manpower and its effective deploy- ment and management is regarded as the main barrier to implementing the national program of afforestation here proposed. A total of about 2,330 new skilled staff, ranging from university graduates to short-course certificate holders, will have to be trained at a cost of about US$6.5 million during this decade, and a significant start must be made in 1984. Community and agroforestry will be prime foci of this training program, which should better utilize the existing agricultural training facilities. Training in the power and petroleum sectors needs to be reviewed and redirected. The mission finds a substantial mismatch between the type of skill training in progress and the future skill re- quirements of the power sector. Plans affecting the EELPA training institute also could burden it with responsibilities for more general education, duplicating the work of existing educational institutions. Consequently, there is a need in the power subsector for: (a) a detailed manpower assessment and development plan and, in parallel (para 7.9); (b) a review of the role and responsibilities of the EELPA manpower training unit and the EELPA training institute (para 7.9). Specific measures proposed in the petroleum subsector over and above those included in the petroleum Exploration Promotion Project include: (a) further and marked increases in incentives to work and under- take further training at the Asseb refinery (para 7.10); and (b) revitalized efforts to recruit staff for the in-house training unit and to establish a fully equipped training workshop (para 7.10). - xiv - Institutional Arrangements 21. The COE recognized at an early stage that the pressing problems of energy sector development called for a coordinated approach and established the Ethiopian National Energy Committee (ENEC) within the MME as early as 1979 to assist in policy making and setting priorities. At present, the arrangements for truly integrated energy policy analysis and investment planning in the energy sector are quite inadequate, however. Difficulties arise through the low status accorded ENEC by the major im- plementing agencies which is related, in turn, to ENEC's position of min- imal influence over energy policy formation and investment. Furthermore, there are insufficient qualified energy staff in the Central Planning Supreme Council. The result is that key decisions on project priorities, pricing and other policies, studies, and so on, are taken through differ- ent channels in an uncoordinated fashion. The need for a strong and in- dependent secretariat to review energy plans and policies is confirmed, though it must be much better equipped for the task and associated more directly with decision making at the highest level. Nevertheless, its role should be analytical and advisory only: the implementing status and management integrity of the energy parastatals must be preserved. The mission emphasizes the need for establishing clear principles and opera- tional objectives to govern the interaction between an energy secretariat and the implementating agencies and consuming sectors as the first stage in rehabilitating an energy policy and planning capability. The energy secretariat need not be part of the Ministry of Mines and Energy and could conceivably be more effective in the Ethiopian government structure if it had more direct links with the Cabinet (Council of Ministers) or, provided the CPSC retains a strong influence on policies and priorities, attached to that body. Its location in the structure, and whether or not it is advised by a group of independent individuals, is secondary to the need for it to be comprised of highly skilled analysts with considerable status who appreciate the interrelationships between the various energy supply and consuming subsectors of the economy and who are permitted to review all major energy sector decisions prior to commitment. 22. There are two additional issues of institutional management which deserve early attention: the role of FAWCDA in woodfuels (firewood and charcoal) marketing and the charter of the EPC. FAWCDA should not attempt to compete with the vast network of wholesale and retail trade already established by small-holder entrepreneurs. Its role is more to increase supply and to assist with quality control. The mission recom- mends: (a) maximizing the involvement of the private sector in biomass fuel distribution and marketing and, wherever possible, in production (paras 7.14-7.15); and (b) a review of the utility of a possible wholesale marketing and reprocessing facility for biomass fuels in Addis Ababa (paras 7.14-7.15). - xv - So far, the EPC has confined its role to petroleum refining although it is now clear that the benefits of investment in the existing or new re- finery facilities can be quickly negated if oil is discovered or natural gas reserves are developed. Also, petroleum exploration efforts to date have been somewhat fragmented. The need for close coordination and rational investment in this sector leads the mission to recommend that: (a) the EPC assume responsibility for all investigations into the commercial production of existing and prospective petroleum resources, including natural gas (para 7.16); and (b) its manpower and resources be upgraded by recruiting and train- ing lawyers, financial analysts and additional managers to ade- quately fulfill this new integrating role in petroleum supply and resource management (para 7.16). Future Energy Demand 23. The growth in energy demand depends largely on the level of in- vestment possible in the more energy intensive sectors of the economy, particularly industry and transport, and the degree to which fuel economy and substitution measures here proposed are implemented. With the World Bank's forecast of a maximum yearly GDP growth rate of 4% for the decade, with the implementation of fuel substitution and conservation measures proposed by the mission, and through close review of the committed program of industrialization, the mission foresees an overall growth in petroleum demand of about 7% a year if there is an early economic recovery and sustained GDP growth of 4% throughout the decade. Demand for fuel oil and kerosene will grow fastest, with diesel growing at less than 6% a year in the most likely scenario. It is not anticipated that natural gas will be in production before 1994. Growth in public power supply of 8.5% per year is expected and the demand for traditional bio- mass fuels will parallel population growth although, with the mix of supply, will change rapidly towards crop residues and dung away from firewood and charcoal as deforestation is intensified. It is likely that there will be an additional 2.5 million tonnes of organic residues a year diverted from agriculture to household cooking by the end of the decade. With the recommended expansion of reforestation and forest management as well as some degree of substitution by briquetting crop residues, more efficient charcoal production, and greater supply of petroleum fuels and electricity to households for cooking, this trend of using unprocessed residuals and dung and the rate of deforestation could be reversed by some time in the 1990s. Energy Sector Investment 24. The proportion which energy sector investment makes of total national investment must be greatly increased from recent levels if sig- nificant economic development is to ensue. The mission estimates that of the B15-20 billion (US$7-10 billion) that could conceivably be available for public sector investment in the 1983/84-1992/93 period, about B3.6 billion (US$1.7 billion) should be allocated to the energy sector. Power - XVi - development will consume more than one-half and forestry and household energy investment generally about one-quarter of the required funds, with the remainder primarily for petroleum exploration, refining and distri- bution. However, it is the capacity to implement projects and to disburse funds as well as the availability of funds that are key constraints to afforestation during the next ten years. The highest priorities for investment are outlined in tabular form at the end of this chapter. It must be reiterated that several major investments are to be made during the ten years from 1983/84-1993/94 which will not yield substantial benefits until five to ten years later, such as in the development of prospective power resources, or for fuelwood plantations in drier regions. The investment program proposed is balanced between economic options for alleviating rapidly worsening deforestation and for reducing the cost of imported oil in the short-term, and greatly expand- ing the sustainable supply of cheaper indigenous fuels in the longer term. Summary of Priority Actions 25. The foregoing describes the state of the energy economy in Ethiopia, the key issues and the subsector-by-subsector options, and the more important recommendations. The key recommendations for action are summarized under four general categories of activities below (invest- ments, policy changes, institutional improvements, and studies). Investment Planning 26. A realistic and detailed energy sector investment plan is re- quired, selecting projects and programs to be funded on the basis of their relative contribution to energy supply and their relative economic return. The most urgent need is for investment planning in energy for- estry and household fuel supply which demands close cooperation among the various primary production ministries (State Farms, Agriculture, Coffee and Tea Development), FAWCDA and the Ministry of Mines and Energy. In- vestment planning by EELPA also must be quickly rationalized and upgraded in the light of new prospective power resources and changing demands for electricity. During the next five years, while the government's own planning capability is being enhanced, there may be need to resort to external assistance to provide the requited expertise. Policy Changes 27. Pricing policies for petroleum and power must be reviewed as soon as possible because in both subsectors levying correct economic prices is the key to demand management and desired interfuel substitution measures. There is also an urgent need to reconsider the role of govern- ment and encourage private sector participation in the production, deliv- ery, and marketing of biomass fuels if woodfuel supply targets are to be met. Hence, a major review of the most profitable and efficient organi- zation of this sector should be undertaken immediately by independent management consultants. - xvii - Institutional Improvements 28. There are five immediate focal points for quickly strengthening and broadening the energy policy and planning capability: (a) upgrading ENEC and/or taking action on the institutional ar- rangements of the energy sector to ensure effective coordi- nation of energy policy planning; defining a constructive and workable relationship between the energy policymaking body and implementing agencies (EELPA, EPC, FAWCDA, etc.); (b) developing EELPA's competence in power system planning; (c) defining the respective roles of FAWCDA and the MOA in deliv- ering the rural forestry program (para 2.46); (d) broadening the role of the EPC into petroleum production, ex- panding its skill base accordingly, and streamlining petroleum exploration under one agency (para 7.16); and (e) establishing a coordinated approach to manpower training to meet the growing deficiency of appropriately skilled staff in the EPC, EELPA, FAWCDA, MOA and ENEC (paras 7.7-7.10). Studies 29. Of the very many pre-feasibility, pre-investment and basic po- licy and strategy studies recommended in this assessment, the most urgent studies by subsector include the following. (a) Petroleum: (i) refinery engineering, including LPG recovery (para 4.9); (ii) preliminary natural gas production and uti- lization study (para 4.8); (iii) fuel oil conservation in in- dustry, defining detailed engineering and costs (para 4.23); (iv) prefeasibility studies on solar heating and crop residue- fired cement kilns (paras 4.24, 5.20); and (v) diesel conser- vation in transportation, defining diesel pump facility re- quirements, costs of increased servicing and fleet capacity, and improved structure of the truck fleet (paras 4.27-4.28). (b) Electricity: (i) defining a strategy for covering supply con- straints though 1990 (paras 3.26, 3.27); (ii) major hydropower options inventory (para 3.33); (iii) undertaking mini-hydro- power and fuel oil displacement of diesel generation prefeasi- bility studies (all as inputs to a least cost development study for the EELPA ICS) (paras 3.44 and 3.48). An urgent review is required of electricity conservation in the ERESA ICS system, including energy audits of major power consuming industries, plus a review of domestic solar water heating potential in the ERESA ICS and elsewhere in Ethiopia (para 3.45). - xviii - (c) Household Energy: In forestry, special studies are needed on species selection and provenance screening and strategies for agroforestry in each major agro-ecosystem, and the design and cost of a national fuelwood resource inventory for ongoing mon- itoring. Studies to facilitate short-term supplies of house- hold fuels include (i) the least-cost method of corn stover and cotton stalk harvesting and briquetting (paras 2.28 [d], 5.8); (ii) the logistics, technology packages and investment planning for carbonizing forest residue (para 2.28 [c]); and (iii) an urgent review of potential and investment requirements to gen- erate surplus bagasse in the sugar industry (para 5.14). A special review is required of the least-cost method of meeting the cooking fuel requirements including possible improved stove prospects for Asmara, Mekele and other northern towns during the next decade (para 2.30). The combination of these initiatives and the implementation of the in- vestment program and the poLicies arising should result in: (a) measurably enhanced supply and lowered cost of household fuels within the decade; (b) a rational and efficient use of expensive imported petroleum fuels and electricity; and (c) a marked reduction in the balance of payments pressure for the supply of modern energy forms. - xix - High Priority Investments for 1982-92 Estimated Sector and Item Objective Capital Cost (US$ million 1983) Household Fuels Rural Forestry . equivalent of 960,000 ha of plantations 175 in woodlots and agroforestry Peri-urban Forestry . upgrading existing and establishing 212 195,000 ha of new plantations New biomass fuels . briquetting 750,000 tonne crop residue and 73 and cooking carbonizing 900,000 tonne wood annually efficiency by 1992, and implementing improved stove programs Power Sector New power . to supply 30-40MW in 1988-89 and 360 generation 200-300MW by 1994 Displacing diesel . replacing diesel generation in the 45 generation ERESA ICS and both SCS systems with cheaper hydropower and fuel oil. Transmission . connection of SCS diesel generation new 180 projects industrial loads and new hydropower or gas generation with existing EELPA ICS Petroleum Sector Oil and gas . continued oil and gas drilling 30 inventory and reconnaissance work, and exploration utilization study promotion Natural gas . drilling to better define the 8 resource Calub resource, and gas utilization studies Gas pipeline, and . investment in development of Calub 300 (before field development resource in the Ogaden desert and 1993) of 600 transmission to markets total I. ENERGY AND THE ECONOMY OF ETHIOPIA Background 1.1 The land area of Ethiopia is 1.22 million square kilometers, making it the tenth largest country in Africa. In 1981, its population was estimated to be 32 million, 85% of which is classified as rural. Population growth is about 2.8% overall, although urbanization is pro- ceeding at about 6.5% per year. With a GDP of US$4,430 million and a per capita income of about US$140 in 1981/82, Ethiopia is among the five poorest nations in the world. The mainstay of the country's economy is agriculture, which accounts for about 50% of GDP, followed by industry, 15%, and transportation 5%. Agriculture also accounts for about 85% of employment and 90% of exports. Coffee contributes 60% of total exports, supplemented by hides and other animal products from Africa's largest livestock population. Manufacturing contributes only 10% of GDP, being limited mainly to food processing, beverages, leather, textiles, cement and petroleum refining. 1.2 In describing recent economic performance it is useful to divide the post-revolutionary era into three periods. The period from 1974 to 1977 was marked by political upheaval, armed conflicts and little eco- nomic growth. In the second period, beginning in 1978, the economy re- covered as security improved. The gross domestic product grew at 5.2% and 5.5% in real terms during FY78/79 and FY79/80, respectively. Growth came mostly from agriculture which benefited from good weather, and from industry after the resumption of industrial activity in and around Asmera and a rapid expansion in capacity utilization in the Addis Ababa region. Improved security also permitted the resumption of normal transport and commercial services. However, this impressive performance was short- lived. During FY80/81 and FY81/82, real GDP growth slowed to 3.0% and 1.5% respectively, mainly due to persistent drought, to the exhaustion of utilized capacity in industry, and severe investment constraints. Fol- lowing a sharp increase in imports and a significant decline in coffee prices, Ethiopia's terms of trade declined by about 23% over this period. Efforts to raise non-coffee export earnings were unsuccessful for many reasons, including the continuing world economic recession. In FY81/82, Ethiopia's merchandise imports reached a level more than twice the value of merchandise exports, and the current account deficit widened substan- tially to more than 7% of GDP. Energy Consumption 1.3 Table 1.1 presents the growth of GDP against the growth in demand for modern energy forms. Petroleum consumption has increased 1.8 times, and public electricity supply at three times the rate of GDP growth during the five-year period between mid-1977 and mid-1982. Por- trayed in per capita terms the growth in demand for petroleum is smaller, -2 although per capita power consumption is still spectacular, being more than double CDP growth. The growth in electricity consumption results from an expansion of the reticulation network from a relatively small base, whereas petroleum products are available more widely. Regrettably, the growth in the volume and value of exports has not been commensurate either with the growth in demand for petroleum products or with the growth in imports as a whole. In FY81/82, Ethiopia's merchandise imports reached a level more than twice the value of merchandise imports, and the current account deficit widened significantly to more than 7% of GDP. The growing cost of imported petroleum is a major factor in this equa- tion. Ethiopia imports crude oil and refined products and exports fuel oil. The fixed pattern of refinery production is such that additional middle-distillate products must be imported to meet the demand, and 70% of the fuel oil produced must be re-exported because it cannot be con- sumed domestically. In FY77/78, the proportion of export earnings devoted to purchasing oil products (net of re-exports) was only 17%, but reached 53% in 1981/82, dramatically underscoring Ethiopia's need to maximize the economy of petroleum consumption and develop more economic energy sources. Investment in the energy sector has been below 5% of total public investment in the last three financial years and clearly will have to increase markedly to have a major impact on the average cost of supply in the next decade, and to satisfy energy demand (see Chapter 7). Table 1.1: CHANGES IN THE GDP SECTORAL GROWTH, 1978/79-1981/82 (million Birr) ------ Financial Year ------- Growth 1977/78 1978/79 1979/80 1980/81 1981/82 Rate (%) - GDP (million Birr) at Constant Factor Costs 4,009.3 4,219.2 4,451.6 4,586.2 4,653.8 3.8 Population (million) 29.0 29.8 30.7 31.5 32.3 2.8 Petroleum Consumption (M kgoe) - 453.4 512.8 557.8 580.2 603.8 7.4 Electricity Demand (M kWh) 406.0 493.0 553.0 569.0 661.0 13.0 GDP per capita (Birr) (constant factor prices) 138.0 142.0 145.0 146.0 144.0 1.1 Petroleum Consumption per capita (kgoe) 15.6 17.3 18.4 18.7 19.0 5.1 Electricity Demand C/ per capita (kWh) - 14.0 16.6 18.2 18.3 20.8 9.8 a/ Growth rates for petroieum and power consumption were high for this period due primar- ily to the redeployment of industrial capacity left idle following earlier civil disturbances. b/ Petroleum consumption data have been adjusted to Gregorian calendar year from Ethiopian fiscal year for this comparison. c/ Public supply only. Source: Bank estimates. -3- Table 1.2: PETROLEUM IMPORTS, RE-EXPORTS AND TOTAL EXTERNAL TRADE, 1978-82 (million Birr) 1977/78 1978/79 1979/80 1980/81 1981/82 A. Petroleum Imports Refined Products 9.0 34.5 80.6 94.3 105.6 Crude Oil 128.3 152.6 221.8 286.6 341.4 Total 137.3 187.1 302.4 381.1 447.0 B. Petroleum Re-exports a! 25.3 38.0 50.3 58.2 64.4 C. Net Petroleum Imports 112.0 149.1 252.1 323.0 382.6 D. Total Imports 1,270.6 1,367.9 1,652.4 1,642.1 1,824.2 E. Non-petroleum Merchandise Exports 640.8 698.8 927.3 791.0 728.3 F. Net-petroleum Imports as % of Total Imports (C of D) 8.8 10.9 15.2 19.7 21.0 G. Net-petroleum Imports as % of Total Exports (C of E) 17.5 21.3 27.2 40.8 52.5 a/ Mainly surplus fuel oil. Source: GOE and World Bank estimates. The Pattern of Energy Consumption 1.4 An energy balance for Ethiopia for 1982 is provided in Annex 1.1. The gross energy supply is about 8.8 million toe and final consump- tion is about 8.0 million toe. The configuration of final energy con- sumption is provided in Table 1.3. Biomass fuels comprise 93% of final consumption in the form of fuelwood (37%), cattle dung (32%), cereal straw (23%), charcoal (1.3%), and bagasse (less than 1%), and 7% of modern fuels in the form of imported petroleum (6%) and electricity (1%). On the basis of sectoral end-use, households dominate with 93%, followed by transport, 4% and industry, 2%. The agriculture and commerce sectors each consumed less than 1%. -4- Table 1.3: FINAL ENERGY CONSUMPTION, 1982 ('000 tonnes of oil equivalent) Crop Char- Elec- Petroleum Share Fuelwood Dung Residue Bagasse coal tricity Products Total % Industry 29 1 35 102 167 2.1 Transport 349 349 4.4 Agriculture 1 31 32 0.4 Households 2,958 2,539 1,804 101 16 18 7,436 92.8 Commerce/Gov't 11 21 32 0.4 Total 2,958 2,539 1,804 29 102 63 521 8,016 100.0 Share (%) 36.9 31.7 22.5 0.4 1.3 0.8 6.5 100.0 Source: Mission estimates. Household Energy Consumption 1.5 Household energy consumption in cooking is believed to average about 2 kg of air-dried wood (25% mcwb) per person per day; however, con- sumption levels vary greatly depending on local supply constraints (see Chapter II). There is a strong cultural preference for woodfuels, either as firewood or charcoal, but their increasing cost and scarcity has re- sulted in dung and crop residues now accounting for over half of total energy consumption, with serious implications for agricultural produc- tivity due to an adverse impact on soil fertility (see para 2.4 and 2.5). Per capita consumption of modern energy forms (petroleum and electricity) is now about 25 kgoe, which makes Ethiopia one of the least energy inten- sive countries in the world. A comparison of energy intensities is made in Table 1.4 for 1980 with countries in the region. Ethiopia is shown to have only about one-third or less of the consumption of modern fuels of its direct neighbors. Table 1.4: ENERGY INTENSITIES FOR NEIGHBORING COUNTRIES, 1980 Consumption of Consumption of GDP per Modern Energy Forms Modern Energy Forms Country capita Per Capita Per Unit GDP (US$) (toe/1,000 people) (toe/million US$) Ethiopia 140 20 177 Kenya 390 143 341 Niger 300 37 108 Somalia 260 58 225 Sudan 360 69 192 Uganda 200 25 125 Source: UNDP/World Bank Energy Assessment Reports. Table 1.5: RECENT CONSUMPTION TRENDS FOR POWER AND PETROLEUM (growth rate in percentage) 1978/79 1979/80 1980/81 1981/82 Overall Power Sales by: EELPA 12.7 10.6 1.7 19.6 10.6 ERESA 190.0 24.1 11.1 3.8 12.8 Petroleum Products: a/ LPG 18.8 18.4 8.9 6.1 12.9 Gasoline 11.2 6.5 3.1 2.5 5.8 Avgas and Jet Fuel 12.4 -12.2 9.3 17.6 1.3 Kerosene -58.0 -9.5 100.0 113.2 12.8 Diesel Oil 15.5 13.1 4.7 1.2 8.5 Fuel Oil 5.1 32.4 12.4 -5.1 10.4 a/ Original calendar year data have been adjusted here to reflect consumption in Ethiopian fiscal years. Source: Power sales from ACRES International, EELPA and mission estimates. Data on petroleum products from the Ethiopian Petroleum Company and mission estimates. Electricity 1.6 The industrial sector was the largest consumer of electricity (55.1%), followed by households (25.7%), and commerce/government (17.5%). Consumption in the agriculture sector is very small (1.7%), and electri- city is not used directly for transport. Recent trends in the consump- tion of petroleum and electricity are indicated in Table 1.3. Electri- city consumption in Ethiopia is concentrated in the central highlands, where sales by Ethiopian Electric Light and Power Authority (EELPA) were 585 GWh in FY81/82. Sales of 83 GWh were made by Eritrean Region Elec- trical Supply Agency (ERESA) in the northern region, which accounted for 12% of national power consumption, and the balance of 7% was supplied from other widely dispersed sources. EELPA's sales grew by less than 3% during the post revolutionary period up to 1978, then increased sharply to average 12% a year during the ensuing period of economic recovery. Since 1981, EELPA's sales have, for reasons explained in Chapter III, been erratic, growing by 1.7% in FY80/81 and by 19.6% in FY81/82. Be- cause of localized conflict, power consumption in FY77/78 dropped in the northern region to one-fifth of the supply in the pre-revolutionary period. After the decline in hostilities in 1978/79, sales once again grew rapidly, recovering to the previous peak by FY81/82. Approximately 70% of Ethiopia's electricity is generated in hydropower systems, 26% from imported petroleum, and the balance from bagasse within the sugar industry. -6- Petroleum 1.7 The end-use of petroleum products by sector is provided in Table 1.6. The largest component of petroleum consumption is transpor- tation (67%). The remaining one-third is consumed by industry at 19.6%, followed by agriculture, 6.0%; commerce/government, 4.0%; and households, 3.4%. Table 1.6: FINAL CONSUMPTION OF PETROLEUM PRODUCTS BY SECTOR, 1982 (103 x toe) Jet Diesel Fuel Share LPG Gasoline Avgas Fuel Kerosene Oil Oil Total (%) Industry 0.2 0.2 34.3 67.2 101.9 19.6 Transport 122.4 5.8 52.1 168.8 349.1 67.1 Agriculture 31.0 31.0 6.0 Households 5.0 11.2 1.6 17.8 3.4 Commercial 0.2 0.1 18.3 2.2 20.8 4.0 Total 5.4 122.4 5.8 52.1 11.5 254.0 69.4 520.6 100.0 Share (%) 1.0 23.5 1.1 10.0 2.2 48.8 13.3 100.0 Memo Item: Power Sector Consumption 48.8 20.1 Total 5.4 122.4 5.8 52.1 11.5 302.8 89.5 589.5 Share (%) 0.9 20.8 1.0 8.8 2.0 51.4 15.2 100.0 Source: Mission estimates. 1.8 Over half of the petroleum consumed in Ethiopia is in the form of diesel, 56% of which is used in transportation, 16% in electric power generation, and the remainder for motive power in industry (11.3%), agri- culture (10.2%), and the commercial (6.0%) sector. A minimal quantity is used in households for lighting in lieu of kerosene. Diesel consumption grew by over 14% p.a. between 1978 and 1980, during which period the an- nual growth in GDP exceeded 5%. Growth in diesel consumption was sharply curtailed between 1980/81-1981/82 in response both to the general econo- mic recession and to high industrial plant factors. The demand for transportation, per se did not, however, slow down appreciably; the num- ber of trucks and buses grew at an average annual rate of almost 7% throughout the 1978-81 period. Despite a considerable expansion of the vehicle fleet, the vehicle availability is still inadequate, causing a considerable latent demand for diesel and quite possibly inhibiting the growth in GDP. 1.9 Gasoline accounts for 21% of petroleum consumption in Ethiopia. Gasoline consumption recovered to pre-revolution levels in 1977 and grew by about 11% during FY78/79, although consumption has grown progressively -7- slower since then, declining to an annual rate of only 2.5% in 1981/82. This reflects more the impact of poLicies to restrain gasoline consump- tion than low growth rates for related economic activites. Gasoline is rationed in urban centers and taxes on gasoline-powered automobiles exceed 200% of the CIF price. Jet fuel comprises 9% of petroleum con- sumption whereas aviation gasoline, kerosene and LPG comprise no more than 2% of total consumption. Kerosene and LPG are used mainly for household cooking. The supply of jet fuel and kerosene varies directly with government policy as, in the refinery, the production of kerosene and jet fuel are interchangeable. Kerosene was diverted to the produc- tion of jet fuel in 1978/79-1979/80 and made available once again for household consumption in 1981 and 1982. Fuel oil, which accounts for 15% of petroleum consumption, is used for heat and steam raising in industry (78%) and for power generation (22%). The spectacular growth rate of 20% p.a. reported between FY78/79 and FY80/81 was again due to the use of excess capacity in the manufacturing sector. The decline (-5.1%) in fuel oil consumption in FY81/82 was most likely due to the first stage of the boiler electrification program (see Chapter III) because, during the same period, all major fuel oil consuming industries were operating at close to full capacity, and large scale manufacturing grew by more than 4% (Table 1.1). Future Energy Demand 1.10 The mission has assumed that a reasonably favorable set of eco- nomic, social and environmental conditions will apply during the next decade in defining the parameters for growth in economic demand. CDP growth is assumed to be sustained at 4% a year throughout 1983 to 1992 1/, being most influenced by growth in industrial output, and great- ly increased crop production, hence a substantial growth in exports. The methodology applied to demand forecasting and the macro-economic and energy subsector assumptions utilized is provided in Annex 1.2, along with energy balances for 1992. The mix of fuels and, by implication, the cost of energy to the economy varies under this set of economic condi- tions according to whether or not energy management and development measures proposed in the following chapters are implemented. The level of funds available to the government for total public investment under this economic growth scenario is estimated to be around US$7-10 billion over the ten years concerned; 15-20% of this directed to energy sector investment should be sufficient to cover the program and projects pro- posed by the mission. However, it must be stressed that CDP growth may not be able to be sustained at 4% a year, at least in the short term, and that both the demand for modern energy forms and the need for investment in supply measures may be reduced considerably. In Table 1.7 the pattern of final energy consumption for the major fuels is compared by sector under the two sets of conditions assumed, i.e., no change from the 1/ Preliminary Government target is 7.5% p.a. -8- present in energy management and development, or the "business as usual" scenario, and full-scale implementation of the energy development and management proposals, or the "accelerated" scenario. Within industry and transport, the main influence on demand growth is through the impLemen- tation of recommended conservation measures, and the production of ethanol, whereby the growth in diesel and gasoline consumption in road transport falls from 7% to 5.1% and from 5% to 3.7%, respectively, and fuel oil in the manufacturing sector declines from 9.2% to 7.4%. The combined savings are about 90,000 toe, comprising about 20% of the pro- jected supply of these fuels and valued at about US$30 million (1983 prices) a year. Furthermore, a saving in crude oil requirements of about 4% of refinery imports, or about 30,000 tonnes per year may be economic- ally justified through conservation measures in the GOE refinery at Asseb if the recommended refinery engineering studies confirm that continued operation of the refinery is the most economic option. By contrast, in the household sector the main impact is likely to be seen through the exploitation of new biomass fuel sources and the production of woodfuels from planned energy forestry. Without any substantial effort to augment the supply of household cooking fuels, there could be an increase of 60% within ten years in the annual use of dung and crop residues for fuel (an additional 7 million tonnes per year), greatly intensifying soil erosion and the deterioration in agricultural productivity. However, if the mission's recommendations in the household sector are adopted, the rate of expansion in the use of these natural fertilizers and soil condi- tioners as fuel will be slowed almost to zero. 1.11 In summary, the pattern of future demand can be modified during the next ten years by practical and economic measures to lessen the severe cost burden of imported oil on the balance of payments and to arrest the rate of deforestation and deterioration of agricultural eco- systems. Overall final energy demand increases by about 3.3% p.a. in the 'Business as Usual' scenario, and 3.1% in the 'accelerated' scenario. The lower growth in the latter case is due to greater efficiency in end- use of the changing mix in cooking fuels towards electricity, LPG, and kerosene, and the new biomass fuels of residue briquettes and charcoal. The growth rate in demand for petroleum products is about 7.5% under both scenarios, although the mix changes as the increased supply of kerosene and LPG to households somewhat compensates for reductions made through energy conservation and interfuel substitution in industry and transpor- tation. -9- Table 1.7: FINAL SECTORAL CONSUMPTION OF SELECTED FUELS UNDER TWO ENERGY MANAGEMENT AND DEVELOPMENT SCENARIOS Consumption Annual Growth Rate ('000 toe) (%) Sector/Energy Form Present BAU "Accelerated" BAU "Accelerated" Investment 1982 --- 1992 ---- 1982-1992 ---- Industry Fuel Oil 69 167 142 9.2 7.4 Diesel 34 58 58 5.4 5.4 * Biomass 29 88 103 11.7 13.5 Road Transport Diesel 169 332 297 7.0 5.1 Gasoline 122 199 176 5.0 3.7 * Ethanol -- -- 23 -- -- Households Existing Fuelwood & Charcoal a/ 3,060 3,060 3,060 na na Dung 2,539 3,982 2,868 4.6 1.2 Crop Residue 1,804 2,825 2,015 4.6 1.2 New Biomass Fuels b/ -- -- 547 -- na New Fuelwood c/ -- -- 1,180 -- na LPG, Kerosene Electricity & Solar 33 58 113 6.0 13.3 a/ Existing forest resources are assumed in the demand projections to be pre- served so that the net annual production is sustained at the present level (3,060 mmtoe). However, in reality the annual yield is expected to decline slightly this decade due to ongoing clearing of the forest and to fall sharply in the next two decades unless reforestation occurs (see Chapter 11). b/ New biomass is charcoal from new sources, briquetted crop residues and ba- gasse. c/ Firewood from new fuelwood plantations. na = not applicable as either there is no base year production, or "no growth" as part of scenario assumption. BAU = Business as usual. Source: Mission estimates. - 10 - II. HOUSEHOLD ENERGY Introduction 2.1 The shortage of woodfuels for household cooking is the most im- portant energy problem facing Ethiopia. For almost all of the rural pop- ulation, and the majority of the urban population, cooking fuels consti- tute the only significant energy use besides the utilization of animal power and the natural flow of solar energy. Cooking is highly energy intensive in Ethiopia due to the practice of "open-fire" baking of in- jera 2/ bread on flat clay discs ("mtads"). Although some adaptation to household energy shortages (including the use of more efficient stoves) has taken place historically in areas of extreme deforestation, such as in the Northern Highlands, there appears to be rigid adherance to energy intensive traditional cooking practice which will only make it more dif- ficult to alleviate the problem. 2.2 In estimating the present and likeLy future demand for cooking fuels, account has been taken of the strong preference of the great majo- rity of the population for woodfuels (firewood and charcoal) for cooking. The mission estimates that an average of 2 kg of air dried wood (25% moisture content wet basis: mcwb) per capita per day is required to sus- tain Ethiopian cooking traditions and to respect other uses of fire (such as heating, lighting and small household industry). There are, of course, variations around this average; for example, the people of Chefa, a small town in the forest of the Southwest, use 5 kg of wood per day each, whereas residents of Moyale in arid lands close to the border with Kenya now use 1 kg per day each. 3/ The current average level of cooking fuel consumption in Addis Ababa, the center of wealth and home for 40% of the countries urban population, slightly exceeds 2 kg per capita per day. The estimated national consumption of household cooking fuels is indi- cated in Table 2.1. Overall consumption is 7.4 million toe, about 93% of the totaL national energy supply. Animal dung and crop residues now con- stitute more than half of the cooking fuels, with fuelwood and charcoal about 41%, and petroleum fuels and electricity a mere 0.4%. 4/ This massive diversion of organic materials which used to be used as animal 2/ Injera is a thin spongy pancake-like bread a few millimeters thick and usually 40-60 cm in diameter. It is made from ground and fermented teff (Eragrostis teff) or a mixture of teff and the modern grains of wheat, corn, sorghum or barley. 3/ Survey funded by Italian aid and conducted by CESEN, an Italian government-owned company. 4/ The miniscule consumption of kerosene is due partly to allocation constraints for it is not marketed beyond Addis Ababa and it is released by the Ethiopian Petroleum Corporation to the market in strictly limited quantities. - 11 - feed and as fertilizers from agriculture to fireplaces has an economic impact which reaches well beyond the energy sector and which must be reversed if Ethiopia's extensive agricultural activities are to continue to produce adequate income for the population. Historical Transition 2.3 It has been reported that at the turn of the century tropical forests covered more than 40% of the Ethiopian landscape. Natural forest cover with closed canopy declined to 16% by the early 1950s and is now only 2.7%, or possibly less. From the description of present day forest resources provided in Annex 2.1, it is evident that, of the remaining high forest cover, already 15% is much degraded, with a standing biomass less than one quarter that of its virgin state. Moreover, the widely cited resource of 20 million ha of Acacia savannah land has in fact been increasingly diminished. Once reckoned to hold a biomass stock of 30 m3 per hectare, it is now believed that the average stocking density is only 10 m3 per ha. The gradual thinning of savannah woodlands is just as serious an erosion of the woodfuel resource base as clear felling of natural high forest even though, ironically, the area under this forest cover is deemed not to have greatly changed during the past decade. The current rate of deforestation can only be roughly estimated. There is no comprehensive data base from which to monitor change. The best data are from an inventory by sample conducted between 1974 and 1979 by the British Ministry of Overseas Development, and these refer only to com- mercial timber volumes of the natural forests. Estimates presented below are consistent with the reported activities of FAWCDA 5/ and the Ministry of State Farms, and with a recent satellite study of the Munessa state forest by SIDA: Source of Forest Loss by Clear Felling Annual Loss (ha) - natural high forest for farming largely by subsistence agriculturalists 80,000 - Acacia woodlands for charcoal and for state farms 50,000 - Woodland, thickets and brush for fuelwood 30,000 Total 160,000 ha Note: It is prudent, given the poor quality of the data, to refer instead to a range of 150,000-200,000 ha/year. Source: Consultation with FAWCDA. 5/ FAWCDA = Forestry and Wildlife Conservation Development Authority. - 12 - Table 2.1: CONSUMPTION OF HOUSEHOLD FUELS, 1982 Quantities Tonnes of Oil As Utilized Equivalent Fuel Consumption (estimated) ('000 toe) (M) Firewood al 8,829,000 te 2,958 39.80 Charcoal 150,000 te 101 1.36 Animal Dung 7,864,000 te 2,539 34.16 Crop Residues 5,138,000 te 1,803 24.26 Electricity 47 CWh 16 0.22 Kerosene 10,200 te 11 0.15 Liquified Petroleum Gas 4,800 te 5 0.07 Total 7,433 100.00 a/ Including twigs and leaves. Source: Mission Estimates. 2.4 The lack of hard data on existing forest and tree resources is a serious deficiency, making the provision of a comprehensive wood re- source inventory a matter of the highest priority. It is now possible to survey the scattered "single" tree resource (woodlots, household trees, etc.) with low-level aerial photography. The latter is Ethiopia's least known timber resource, and the one for which baseline data is most ur- gently required. While the conclusions drawn here regarding afforesta- tion and fuelwood supply are not significantly influenced by the lack of precise data on forest cover, the estimates provided of fuelwood supply and demand are nevertheless crude. In 1982, it is estimated that the demand for fuelwood was 2.5 times the sustainable supply, based on exist- ing tree cover. However, this global figure is not particularly relevant due to the enormous regional variations indicated in Tabl3 2.2. Of the overall estimated shortfall in 1981/82 of 20 million m , aLmost 60% arises in the north, in Eritrea, Tigrai, Wollo and Gondar. The middle band of provinces of Cojjam, Shoa and Hararghe are only marginally better off, though their urban centers are within economic reach of the more heavily forested areas of the southwest and, with the excepcion of the lowland areas of Hararghe, they are more fertile and have moister cli- mates. All of these provinces contrast sharply in fuelwood supply and demand balance with the heavily forested provinces of Ilubabor and Kaffa in the southwest. The table points to the need for 1.7 milLion ha in full production just to meet the 1982 deficit between rural woodfuel consumption and demand (paras 2.41 and 2.48). Conceptual Basis for Strategy Formulation 2.5 At the rate at which peasant agriculturalists are currently clearing the fringes of natural high forest, this resource will be lost in about 30 years. As in the past, during this first stage of forest clearing for the purpose of developing land for food production, local - 13 - fuelwood is abundant. At present, perhaps 20 million cubic meters of wood, the same quantity that is consumed in all the households of Ethiopia, are burnt off during agricultural clearing each year. It is only sometime later that trees begin to be harvested primarily for fuel. Beyond this point it appears that a critical transition of decline begins within subsistence agriculture whereby the growing scarcity of woodfuels is linked inextricably to falling crop and animal production. This tran- sition leads to, and is clearly exacerbated by, growing urbanization in Ethiopia as the nature and level of fuel use for household cooking for most urban dwellers closely resembles that for their rural counterparts. The demand for woodfuels, and ultimately for any combustible residue, by urban dwellers or members of any concentrated settlement without a suf- ficient independent resource base (i.e., state farms) becomes an intoler- able burden on rural productivity. A conceptualization of the perceived stages of this transition follows below and in Figure 2.1. Stage I: The rate of timber harvested locally for all purposes (fuel, construction, tools, fences) exceeds, for the first time, the average rate of production. The existing timber resource is then progressively "mined"; firewood remains the main fuel source. Nu- trient cycle No. 1 begins to decline though with imperceptible impact on food production. The general reason for the imbalance is population growth. The specific reasons include urbanization and major land clearing (e.g. state-farms) whereby firewood and charcoal become cash crops leading to overcutting relative to purely local subsistence requirements. Stage II: The great majority of timber produced on farms and on surrounding land is sold out to other rural and urban markets. Peasants begin to use cereal straw and dung for fuel: the relative proportions depend on the season. Both nutrient cycles No. 2 and No. 3 are breached for the first time and nutrient cycling dimin- ishes. Combustion of crop residues and dung leads to lower inputs of soil organic matter, poor soil structure, lower retention of available nutrients in the crop root zone and reduced protection from the erosive effect of heavy rainfall. Hence, topsoil nutrient reserves begin to decline (See "spill" in the Figure). Stage III: Almost all tree cover is removed. Now a high proportion of cow dung produced is collected; the woodier cereal stalks are systematically collected and stored; and both are sold for cash to urban markets. The yields of cereal crops and, in consequence, animal carrying capacity begins to decline. Draft animal numbers and power output is reduced, hence, the area under crop also falls. Soil erosion becomes serious. Nutrient cycle No. I ceases alto- gether. Stage IV: Dung is the only source of fuel and has become a major cash crop. All dung that can be collected is collected. All crop residues are used for animal feed, though they are not sufficient for the purpose. Nutrient cycle No. 2 is negligible and No. 3 is greatly reduced. Arable land and grazing land is bare most of the - 14 - year. Soil erosion is dramatic and nutrient-rich topsoil is much depleted. Dung and dry matter production have fallen to a small proportion of previous levels. In such a situation, extended dry periods can be devastating because the ecosystem loses its capacity to recover quickly. Stage V: There is a total collapse in organic matter production, usually catalyzed by dry periods which were previously tolerable. Peasants abandon their land in search of food and other subsistence needs. Starvation is prevalent. Animal populations are devasta- ted. Rural to urban migration swells city populations, increasing demand on the rural areas for food and fuel, and the impact of urban demand is felt deeper into the hinterland (the "urban shadow" effect). This transition from the first to the final stage is in process right across Ethiopia and has reached the terminal phase in parts of Tigrai and Eritrea. 2.6 The only way to prevent the current situation in the remaining populous and fertile areas from sliding toward the terminal state of Stage V is to develop a strategy which will: (a) remove the dependency of urban settlements on their rural hin- terlands for woody fuels, and (b) reestablish a dynamic equilibrium between supply and demand for firewood in rural areas. While the development of peri-urban fuelwood plantations is an obvious component of a strategy to serve the first objective, the time required to do this is such that, even if design work began immediately, the pro- duction of woodfuels would hardly begin to be augmented before the end of the decade. Without urban self-sufficiency it will be extremely diffi- cult to achieve the second objective, as biomass fuels will continue to drain from the rural areas to the towns and cities. In addition, the situation of Northern Ethiopia where, in many places, agricultural eco- systems have deteriorated to stages IV and V, demands special, and pos- sibly separate, consideration because of the huge scale of the problem and the implied investment, and the added complexity of local hosti- lities. - 15 - Table 2.2: RURAL FUELWOOD DEMAND AND PLANTATION REQUIREMENTS, 1982 Demand Demand Additional Needed Rural Fuelwood Satis- Supply Plantation Plantation - Region Population & Poles a' faction -/ Needed Production area (1000) ('000 m ) (%) ('000 m ) m 3/ha/year ('000 ha) Arsi 1,124 1,338 50 669 18 37.2 Bale 877 1,044 40 627 15 41.8 Eritrea 1,804 2,147 10 1,931 8 241.4 Gemo Gofa 1,003 1,194 50 597 12 49.8 Gojam 1,976 2,351 40 1,411 15 94.1 Gondar 1,993 2,372 20 1,897 12 158.1 Hararghe 3,018 3,591 40 2,155 12 179.6 Illubabor 820 976 95 49 20 2.5 Kaffa 1,574 1,873 95 94 20 4.7 Shoa 4,876 5,802 30 4,061 18 225.6 Sidamo 2,744 3,265 70 979 20 49.0 Tigrai 2,105 2,505 10 2,254 8 281.8 Wollega 2,028 2,413 60 965 15 64.3 WoIlo 2,552 3,037 10 2,734 10 273.4 Total 28,479 33,908 (40) 20,423 12 1,703.3 a/ Volume in solid m3 on bark. Allowance is made for 1.09 m3 per person per year fuelwood, plus 0.1 m3 per person per year for building poles, agricultural imple- ments, etc. Rural is defined as population outside of centers of 5,000 or above. b/ In fact the tree planting is needed in small woodlots and as part of agricultural systems, as well as plantations per se. c/ Estimated demand satisfaction is the proportion that the sustainable mean annual increment of new wood production makes of the estimated demand in the province concerned. Source: Mission estimates. PATTERN OF DETERIORATION IN ETHIOPIAN AGROECOSYSTEMS Breach: Dung Removed as Fuetwood Substitute Breach: O Tree Cover Removed / \\ ~~~~~Forage & Crop Residue to fo Irwo / \\ ~~~~~~Dung Conversion Cycle. \ / \\ ~~~~~~Nltrogen.& Mineral Nutrlent _ y / \ ~~~~~~~~~Retention.6 Y l { _= 4 Acscw . j w ~~~~~~~~~~~~~CYCLF O 1E1 L L ~~~~~~~~~~~~~~~~~~~~~~~~~Mineral Nutrlent Cycling. | . * .......v ..... "1.......................... K....... I-. X.. Eroslon of Nutrient & Humus Rich Topsoil as CYCLE NO,.2 Main Nutrient Cycles CYCIS NO. 2 ~~~~~~~~~~~~are Breached Grass & Crop Residue: Nitrogen-Fixing & Retention. .Mineral Retention & Cycling. Breach: Overgrazing, Scavenging for Fuelwood Substitute Word Bank-25403 - 17 - Preview of the Analysis and Main Findings 2.7 The remainder of this chapter deals with more detailed techni- cal and economic options for alleviating the critical shortage of appro- priate cooking fuels. The presentation is divided into options for the short-term (1983-92) and for longer-term (1992 and beyond), considering first the urban and then the rural context, and examining prospects for expanding supply and for demand management in turn, where relevant. When considering a program of afforestation a distinction is drawn with re- spect to both peri-urban and rural forestry between planting requirements to meet the identified need and a rate of planting -- invariably lower -- which is thought to be achievable. The costs of the achievable forestry program and options for supplying household fuels from new sources in the short-term are then examined and summarized, followed by a review of the economic justification for the investment program outlined. In conclu- sion, comment is made on the economic and environmental implications of failing to implement the proposed or simiLar programs of afforestation and new resource mobilization. The findings presented are placed in perspective by the following tables (2.3 and 2.4) dealing with the supply/demand situation in the short and longer-term. During the next decade it is not possible because of the long establishment phase, for energy forestry to make a significant contribution to household fuel supply. Consequently, only about 20% of the demand for high quality cooking fuels can be met by 1992. However, new and expanded fuel sources and biomass fuel, particularly charcoal and crop residue briquettes, could make a major contribution (2.1 million tonnes of wood equivalent a year) to the supply of urban cooking fuels although these are available only to the southern urban centers. As a result, more than 70% of the southern urban household fuel demand could be satisfied by 1992 if the specified production programs begin promptly. The only short-term option so far identifed for the urban north is expanded kerosene (and possibly LPG) supply from the existing refinery which could meet about 20% of the local demand. By 2002, it is possible to meet about two-thirds of the total unsatisfied national demand for household cooking with new fuel sources, and production from fuelwood plantations, woodlots and agrofor- estry systems established during the late 1980s and early to mid-1990s. During this second decade, the contribution from energy forestry is over half of the total demand, and over 80% of the new supply of cooking fuels, compared -with 15% and 60%, respectively, during the first decade. By continually expanding energy forestry at the rate prescribed here, supply and demand for cooking fuels balance sometime between 2010 and 2020 (see Table 2.13). The most important concLusions arising from the presentation which follows are that it is possible and desirable by 1992 to: (a) plant 1.16 million ha of trees around cities and throughout rural agro-ecosystems; (b) harvest 1.7 million tonnes of wood equivalent a year in the form of charcoal and woody briquettes from existing forestry and agricultural wastes; - 18 - (c) increase the supply of electricity, LPG and kerosene to. house- hold cooking by about 0.4 million tonnes of wood equivalent. Table 2.3: SUMMARY OF SUPPLY AND DEMAND PROSPECTS FOR HOUSEHOLD FUELS IN THE SHORT-TERM (up to 1992) (tonnes of wood equivalent: twe) Annual Contribution Possible by 1992 From New Sources Achievable New Biomass Plantations & Fuels, Petroleum Estimated Other Tree Planting & Electricity a/ Need in 1992 Ready for Harvest - Supply Deficit - (ha/'000) (twe/ (ha/'OOO) (twe! (twe/ (twe/ million) (million) (million) million) Northern Region Peri-urban 240 1.77 15 0.11 0.35 1.31 Rural 1,406 8.62 186 1.14 -- 7.48 Southern Region Peri-urban 270 3.34 30 0.35 2.01 0.94 (including Addis Ababa) Rural 938 14.54 124 1.92 -- 12.62 Total 2,850 28.27 355 4.09 2.36 22.35 a/ This is a deficit in fuelwoood or other high quality cooking fuel and will, in practice, be met at whatever level by a combination of dung, crop residue, twigs and leaves and by simply making do with less cooking fuel. b/ Planted by 1987. Table 2.4: SUMMARY OF SUPPLY AND DEMAND PROSPECTS FOR HOUSEHOLD FUELS IN THE LONG-TERM (1993-2002) Annual Contribution Possible From New Sources by 2002 Achievable New Biomass Plantations & Other Fuels, Petroleum Estimated Need Tree Planting Ready & Electricity in 2002 for Harvest a Supply Deficit (ha/'O00) (twe/ (ha/'OO0) (twe/ (twe/ (twe/ million) million) i Ion million) Northern Region Peri-urban 482 3.53 324 2.38 0.71 0.44 Rural 1,560 9.57 910 5.58 -- 3.99 Southern Region Peri-urban 550 6.49 206 2.50 3.12 0.87 (including Addis Ababa) Rural 1,580 16.33 920 9.51 -- 6.82 Total 4,172 35.92 2,360 19.97 3.83 12.12 a/ Planted by 1997, - 19 - Short-Term Supply Options, 1983-92 2.8 Since the lead time for afforestation is long, fuelwood produc- tion from new planting will have no significant effect on supply before 1992. The only near-term prospect, then, is to harvest fuel-value mater- ials which are already in production and accessible at a comparatively low opportunity cost. A review of potential sources indicated good pros- pects for improved recovery of: (a) logging and sawmill waste; (b) clear felled timber from state farm development; and (c) agricultural residues from state farms (wheat and barley straw, corn and sorghum stover, cotton stalks), and major commercial crops (coffee residues and sugar cane bagasse). 2.9 In addition to an absolute increase in the supply of primary energy from these sources, improved carbonization efficiency can add measurably to final energy supply for wood wastes. Furthermore, there are prospects for improving the household supply of existing fuels-- electricity, kerosene and LPG. 2.10 The woodfuel resource is comprised of the remainder of the selected sawlogs, all of the noncommercial species in the logged areas, and the wastes remaining at sawmills from processing delivered sawlogs. Despite the decimation of tropical high forest, the remaining resource is still significant in absolute terms. With few exceptions this waste is Left to rot, or is burned on the spot during clearing for replanting. An analysis of the quantities of Logging waste available at the 1981/82 logging level is provided in Annex 2.2. It is estimated that 530,000 m3 (solid volume, 500 kg_m3 basic density) a year are now accessible, of which about 50,000 m are harvested, leaving 480,000 m per year of untapped fuelwood resource which increases to 780,000 m3 in 1987 and 1.1 million m3 in 1992. Due to the low bulk density of firewood, and hence the high cost of transporting it per unit energy, it is almost always de- sirable to carbonize these wastes in order to make the delivery of this wood energy to a distant urban market economic. The recovery of logging wastes and noncommercial species for fuel should be a component of inte- grated forest management to upgrade the long-term productivity of the existing high forest resource. A schema showing the place of charcoaling in improved forest management is provided in Annex 2.3. Sawmill Wastes 2.11 An analysis of fvailable sawmill waste is provided in Annex 2.4. A total of 38,000 m of slabs and 25,000 m3 of sawdust are deemed accessible. It is proposed that slab waste be carbonized and that saw- dust be dried and briquetted. The economic costs of producing sawdust briquettes at the sawmill are estimated at B49/tonne (1983 Birr). Saw- mills are located between 350 and 500 km from Addis Ababa, thus the - 20 - delivered cost is approximately B98/tonne or roughly 30% of the present market price for firewood. Charcoal production costs for the slab waIste are provided below. Sawmill wastes are expected to grow to 130,000 m in 1987 and 215,000 m3 in 1992 as sawmills are rebuilt and relocated. Hence, it is important to design efficient waste recovery into the new sawmill and sawmill refurbishment program. Recovery of Woodfuel from Agricultural Clearing 2.12 For the short-term, the focus of recovering timber from clear- ing land for agriculture must be in respect of state farm development, for it appears that the more massive resource lost in forest clearing by peasant agriculturalists, amounting to half the total yearly supply of energy for cooking, cannot be recovered without formally recognizing this "illegal" practice. Steps to resolve this dilemma are warranted though it is acknowledged that the issue is complicated. There are two options within the state farm sector which deserve urgent attention: immediate recovery of the rich Bubeka forest resource in Kaffa province, and ex- panded recovery of mostly savannah land timber in the process of regular expansion. 2.13 The Coffee Plantation Development Corporation within the Ministry of Coffee and Tea Development will clear for coffee plantations a further 5,000 ha at Bubeka during the next four to five years. The availability of wood net of other uses is about 880,000 m3 (sv 6/) per year through 1987, equivalent to 110,000 tonnes of charcoal per year, or more than 70% of the estimated annual production for Ethiopia as a whole. At present, only 500-1,000 te charcoal per year are produced at Bubeka. The recovery of this resource using fast, efficient portable metal char- coaL kilns is of the highest priority. The charcoaling operation mounted here can be transfered to recover the logging residues specified above (para 2.09) and timber cleared on other state farms, dovetailing well with that program by 1987 when the Bubeka clearing ceases. It is essen- tial, however, that adequate incentives be provided for the full coopera- tion of the Coffee Plantation Development Authority in this program. 2.14 State farms are planned to expand a further 258,000 ha between 1983 and 1993. Some of this expansion will be in heavily forested areas though most will be over savannah land. A conservative estimate of tim- ber accessible from clearing is 130,000 m3 (sv) a year based on 13,000 ha per year at 10 m3ha, equivalent to 120,000 tonnes of air-dry wood, or about 10% of the current energy supply to Addis Ababa (see Annex 2.6). Efficient Charcoal Production From Available Wood Wastes 2.15 Traditional Ethiopian charcoaling is done by above-ground earthen kilns. Evidence from the mission's own field tests suggests 6/ sv = Solid volume, denoting the important difference in energy value between a solid and a stacked cubic meter of wood. - 21 - that, in practice, average efficiencies of 20% and 15% are achieved in savannah and moist forest conditions, respectively. Metal kilns'.have three clear advantages over traditional kilns in the Ethiopian context: a higher rate of conversion, greater yield and better product quality. Charcoal production with metal kilns has about 2.5 times the labor pro- ductivity of earthen kiln production, and a 70% greater yield. Charcoal produced under controlled conditions will be of higher quality and energy value. The direct economic cost of charcoal from metal kilns is about 10% less than that from earthen kilns (B174 compared with B192 per tonne delivered) as lower labor and resource costs more than compensate for the consumption of imported steel, although the economic benefits of using metal kilns are considerably greater because wood wastes are recovered at a much higher rate than with earthen kilns, displacing household fuels of high opportunity costs to higher value uses elsewhere in the economy (see para 2.41). It is therefore recommended that portable metal kilns be used as part of all forest-based woodfuel recovery systems. The first step towards implementing this recommendation is a major study of the management, institutional and specific technical and funding requirements of a comprehensive, efficient, forest residue carbonization program ad- dressing, as well, the special transport requirements. At sawmills, the construction of cheaper fixed kilns made of local bricks or of mixed clay and metal construction should be examined. 2.16 The transportation of charcoal over such long distances (400- 700 km) on rough roads requires both special transportation measures and some reprocessing and repacking. Charcoal has a low bulk density (200- 400 kg/m3), especially that from mixed species of the tropical forest (measured in Ethiopia at 230 kg/m3). The full economic cost of moving charcoal from Bebeka to Addis Ababa is about B60/te in 22te trucks. FAWCDA is now paying B140-180/te because its trucks cannot be loaded to their weight capacity due to the low volume (bulk density) of charcoal. Trucks carrying charcoal should be equipped with four meters high mesh or wooden-slatted sides to increase the loaded volume up to 80-100 i3. In addition, after long hauls, fines production may reach 20% of total char- coal weight, especially that produced from softer forest timber. Char- coal derived from logging waste recovery and other high forest operations should, therefore, be screened and repacked, and the fines briquetted at bulk depots in the major urban centers. Fortunately, a suitably inexpen- sive briquette binder, molasses, is available in the required quantities. 2.17 Under the proposed production schedule (see Annex 2.6), 35,000 tonnes per year of the charcoal delivered to urban areas would have to be reprocessed into briquettes by 1992. The mission recommends that plans be made now for pilot production of these briquettes. Given the scale of this operation, and the need, in any case, for proper wholesaling and repackaging facilities, charcoal processing should be seen as an industry in its own right, possibly to include other higher value charcoal products. 2.18 A summary of the woodfuel resources to be delivered in the form of charcoal is provided in Table 2.5. The supply potential by 1992 is sLightly more than the present estimated national consumption of - 22 - charcoal. It is noteworthy that the projected 340,000 tonne long run annual supply is roughly 35% of the estimated 1992 demand for ene*rgy by households in Addis Ababa. Agricultural Residues From State Farms 2.19 The collection and processing of agricultural residues to pro- duce good quality household cooking fuels is unheard-of in Ethiopia and is uncommon elsewhere in Africa. However, fuel briquettes of coffee husk and groundnut shell are produced commercially in Kenya and Niger respec- tively; rice husk fuel briquettes are produced commercially in India and Thailand and, pelletized bagasse is well established at a major sugar mill in Hawaii. On Ethopian state farms wheat and barley straw, and corn and sorghum stover are burned in the fields as a means of disposal. Cotton stalks are also burnt-off both to sanitize the growing region of crop pathogens, and to clear the fields for replanting. Coffee cherry skin and husk residues for the most part rot and smoulder away behind decorticating mills scattered throughout the coffee producing provinces. Finally, there is considerable latent potential for producing surplus sugarcane bagasse in the rapidly growing sugar industry. The mission examined the general context of collection, high pressure briquetting and transport of these residues and established the quantities accessible and preliminary production costs for each resource. A brief review of the findings is provided in Chapter V. Table 2.6 provides an estimate of the quantities that could now be made available. The economic costs of these fuels delivered to Addis Ababa are provided in Table 2.10 below. A total of three-quarters of a million actual tonnes a year of agricultural resi- dues is deemed accessible with appropriate management, and a total in- vestment in harvesting equipment, briquetting presses and related faci- lities of about Birr 70 million. This is roughly equivalent to both the present day household energy use in Addis Ababa and the annual fuelwood production from a 100,000 ha plantation. The availability of each resi- due will grow substantially throughout the next decade though the target adopted here is to harvest the presently "accessible" resource by 1992. Even this will be a considerable achievement for it must be stressed that nowhere else in the world is there a crop residue recovery program geared directly to the production of fuel briquettes of the magnitude of the program here proposed. Despite the proven commercial nature of all the equipment involved, the management skill required to administer econo- mically efficient residue recovery and the need for consumer acceptance increase the risks to be faced in this vitally important program. Increased Supply of Petroleum Fuels and Electricity 2.20 Kerosene, LPG and electricity are all used to some extent for cooking in urban areas. For the petroleum fuels, supply is mostly limited to Addis Ababa and environs due either to a lack of distribution facilities or inadequate distribution incentives, or both. Electricity is used by upper income groups for cooking in Addis Ababa and other major cities with a public power supply in the central highlands, and those receiving heavily subsidized power from public grids serving Asmara, Asseb and Massawa. In the course of the mission's review of refinery - 23 - management and expansion options (see Chapter IV), it was established that kerosene supply to households could be increased about fivefold, to 54,000 tonnes per year, and that in the mission's view, LPG production could grow about threefold to 12,000 tonnes per year without refinery expansion or conversion. The new supplies of kerosene and LPG would serve an additional 130,000 and 30,000 households, respectively. For comparison, there were about 300,000 households in Addis Ababa, and 85,000 in Asmara in 1982. Table 2.5: WOODFUEL RESOURCES IN THE FORM OF CHARCOAL (1000 tonnes) 1982 1987 1992 Earthen 11or Metal a/ Earthen or Metal Earthen or Metal Forest Logging Wastes 36 62 56 101 83 143 Sawmill Wastes (slabs only) 3 5 5 9 9 16 State Farms Bebeka 66 114 66 114 -- -- Other 17 17 17 17 17 17 Total 122 144 205 241 102 176 Quantities of Charcoal Briquettes Implied b/ 24 or 40 29 or 48 21 or 35 a/ Kiln efficiencies applied: Earthen kilns: 15% in forest timber, 20% on savannah land timber; metal kilns, 26% and 30%, respectively. Only earthen kilns are used in savannah land operations. b/ 20% fines for forest charcoal, 10% for savannah charcoal. Source: Mission estimates. 2.21 Electricity can only be considered as an alternative cooking fuel where there is an existing public supply, and even then will only have some prospect of being economic in the short-term if the source is the EELPA ICS at current estimates of the long run marginal cost of power in that power system (see Table 2.10). This limits consideration of the option to the urban settlements of the central highlands, with a strong focus on Addis Ababa. Until 1987, the supply of hydropower is somewhat constrained by capacity limitations;. however, an additional demand of 25 GWh per year for household cooking can be accomodated by that time, in- creasing to 39 GWh by 1992. (see Annex 3.2). This is enough to meet all the cooking energy requirements of some 80,000 households. - 24 - Table 2.6: ANNUAL AVAILABILITY OF CROP RESIDUE FULES FOR HOUSEHOLD COOKING IN THE SHORT- TO MEDIUM-TERM 1982 Crop 1982 Crop Production Residue Estimated Accessible Wood Equiv. - Crop Residues ('000 tonnes) Residue - ('000 tonnes) Coffee 250 450 150 191 Cereals on State Farms Wheat 112 210 105 147 Barley 4 6 3 4 Corn 97 266 187 226 Sorghum 32 48 24 29 Cotton on State Farms 58 253 239 334 Bagasse c/ 146 58 43 60 Total 1,291 751 991 a/ Allowing for losses, residues left for soil production and fertilization, non- compatible processing, and prohibitive transport costs. b/ Taking into account differences in net calorific value and the efficiency of present day cooking appliances using raw wood and dense dry residue briquettes. c/ Applying 50 kg dry bagasse per tonne of cane harvested fol lowing upgrading of process energy efficiency, and subtracting 15,000 tonnes of bagasse committed to other uses. Short Term Demand Options, 1983-92 2.22 The options for alleviating the shortfall of woodfuel supply by demand managment in the short- to medium-term appear confined to improv- ing the efficiency of household cooking by increasing the efficiency of cooking stoves, introducing new stoves of high efficiency, and changing the cooking vessel. During the assessment mission, the efficiency of heat transfer was measured for wood stove types burning most fueLs in common use. A special effort also was made to establish the efficiency of cooking injera by traditional and modern methods. These data and the related field observations, along with measurements of weight and mois- ture content of retailed traditional fuels, were used to define the cost of useful energy and the relative values of presently traded and prospec- tive fuels for the household subsector. Normal Cooking 2.23 Ethiopians cook meat and vegetable soup or stew (wat) dishes to eat with injera, as well as boil water, make coffee and the like, light- ing a fire for these purposes several times a day. Many urban dwellers cook with both an open fire and a charcoal stove except when cooking injera. Table 2.7 presents the present and prospective efficiencies of Ethiopian cooking. The most significant findings are that there is con- sistently a 60-70% greater efficiency of fuel use with aluminum over tra- ditional clay pots, and that the traditional charcoal stove can be modi- fied very simply to achieve a 25% improvement in efficiency. The wick- - 25 - type kerosene stoves in use are of comparable efficiency but, as ex- pected, the kerosene pressure stove has a 10-30% efficiency advantage. The wick stoves retailed locally were found to be flimsy, short-lived and potentially unsafe, which tends to negate their moderately good efficien- cy. Since there are very few kerosene stoves in use or available for sale, any program of expanded kerosene use should be based on high qua- lity wick stoves like the Nutan 2 of India, and pressure stoves; prefer- ably the latter. The mission recommends the local production under license of quality stoves; promotion of more efficient kerosene stoves, and the implementation of tighter stove import standards. Aluminum pots (3.5 liter) currently retail for B12, whereas clay pots (4.0 liter) retail for B4.50 each. The difference in price almost certainly reflects high import duties, hence the prospect of very significant energy savings must be weighed against a probably minimal loss of revenue. There is a prima facie justification for heavily promoting the use of aluminum cook- ing pots, thus the mission recommends an early pilot program to examine consumer response and the implications of various pot pricing strategies. 2.24 In addition to modifying existing charcoal stove models, there is a possible role for the proven Upper Volta portable metal wood stove and a portable Lesotho-type dung stove. Therefore, the mission recom- mends cautious programs of trial and dissemination of these stoves in urban areas in the first instance. The context and mode of dissemination is discussed further below (see para 2.26). Due to an almost universal failure to achieve much vaunted savings through wide dissemenation of "improved" stoves in numerous countries, no attempt is made here to mod- ify fuelwood production targets to account for anticipated reductions in per capita demand. It is encouraging to note, however, that considerable financial incentives exist for consumers to adopt the improvements or new stoves here identified. The simple payback on selected stoves or pots is indicated in Table 2.8 and ranges from two weeks to two months. Injera Baking 2.25 Injera baking consumes about half of all household cooking fuels. Injera is cooked every two or three days in two to three hour sessions on a large clay plate called a "mtad" measuring about 60 cm wide and 2 cm thick. This plate usually rests on stones over an open fire, though in the northern regions the fire is partially enclosed with a clay wall (the "Tigrai type" injera stove). During the past decade electric injera cookers have become popular, particularly among upper income groups in urban areas. Here an electric heating coil is wound into a groove on the underside of the "mtad." There was a need to examine injera cooking closely not only because of its enormous impact on energy consumption in Ethiopia, but also because the government has ordered a crash program of electrifying injera cooking, demanding of the electric supply authority (EELPA) the production of 10,000 electric injera cookers per year. The results of the mission's investigations are provided in Table 2.9. To summarize, injera cooking with biomass fuels is even lower in efficiency than cooking with pots on open fires, though the northern style "Tigrai" stove is more than twice as efficient as the more common "open fire" injera cookers. There are, however, marked differences in - 26 - fire management and fuels used which no doubt highlight the good perfor- mance of Tigrai cookers. Using the same, more volatile leafy fuels in the mission's own experimental design, and with no special fire manage- ment, fuel savings of 25% were achieved with the Tigrai type stove. Electric injera cookers are quite efficient by comparison, though it appears possible to achieve even higher energy efficiencies and to reduce the power rating from around 3 kW on average to 2 kW or less by using non-stick (teflon) aluminum plates with fully integrated heating coils (see Chapter III and para 2.26; vii). Besides the great merit of elec- tric injera cooking, there is a need to promote the use of more efficient biomass-fired injera cookers. Table 2.7: STOVE EFFICIENCY BY STOVE, FUEL AND COOKING VESSEL TYPE Ratio of Low - Power to High Percentage of Primary Fuel Type Stove Type Power Efficiency Heat Utilized () A. Existing Stoves Clay Pot Aluminum Pot Firewood Open fire 1.2 8 -- 0.7 13 Charcoal Square metal 1.9 17 -- 1.5 -- 15 Round metal 0.9 14 -- 2.6 -- 23 Kerosene Daulat (wick) -- -- 35 Chinese (wick) -- -- 42 Swedish Primus (pressure) -- -- 46 Dung Open fire -- 9 -- 1.1 16 Briquettes Open fire -- 15* * (estimate assuming about same behavior as dung) B. New Stoves Firewood Portable metal b/ (and briquettes) (upper volta style) -- 28 Charcoal Traditional stove -- 23 -- High sided c/ -- -- 30 Kerosene Nutan-2 (Indian) 0.4 -- 36 (wick) Briquettes Modified Paolo type stove (Lesotho) d/ -- -- 21 Dung Modified e/ -- -- 32 a/ The low/high power ratio indicates in which phase the stove is most efficient. Stoves with values greater than 1 have higher efficiency in the low power phase. Values less than 1 have higher efficiency in the high power phase. b/ Stove recently welI established in Upper Volta which is compatible with Ethiopian cooking methods. Not tested in Ethiopia but test efficiencies verified by mission. c/ Simply adding a tubular sleeve to the traditional Ethiopian stove bringing the sides up around the pot. d/ Briquettes obtained in the US for tests were extruded wood of very high density and difficult to light and burn without other materials in the fire. e/ Dung burning proved very difficult to manage in this stove. These results are achieved with constant attention. - 27 - Table 2.8: COSTS AND BENEFITS OF IMPROVING STOVE EFFICIENCY Efficiency Payback at Existing to be Cost of Cost of Present Stove Laboratory Achieved in Existing Improved Market Efficiency Efficiency Practice Stove/Pots Stove/Pots Prices a/ (PHV) (PHV) (%) (%) (%) (Birr) (Birr) (days) Open Fire using aluminum pot compared with intro- duced metal stove 13 28 20 nil 10 61 Open Fire using clay pot compared with aluminum cooking pot 8 15 13 4.5 12 b/ 18 Charcoal Stove using aluminum pot compared with high-sided, insulated sleeve and liner 23 35 30 5 8 13 Injera Cooking on open fire compared with a closed injera cooKer 4-5 8-10 7 4 8 13-18 a/ 15 ec/kg wood and BI/kg charcoal. Consumption per famiiy is 0.7 kg charcoal/day and for wood 5 kg/day average each for injera and pot-type stove cooking. b/ Financial cost only. Economic cost is probably much less. PHV = percentage of heat utilized. Source: Testing and estimates by the mission. The mission recommends a review of the prospects of success of heavy pub- licity favoring the use of Tigrai-type cookers as part of a package of cooking efficiency measures identified in Table 2.9. Implementation of a Cooking Efficiency Program 2.26 The management of research, review and field trial activities, along with all government-sponsored stove marketing and promotion activi- ties recommended here should be the responsibility of one agency: pos- sibly a stove marketing organization. This organization should not it- self undertake any stove design, testing or production, and it should attempt to contract out much of the market analysis and specific research and review work, retaining only the roles of coordination, analysis, policy development, and communications. All aid for stove work would be directed through this agency. Finally, it is desirable to have women playing a direct role in the management of the agency, including possibly the extensive network of women's organizations in dissemination. - 28 - Table 2.9: EFFICIENCY OF INJERA COOKERS Cooking Time per Energy per Cooker Energy Form Injera Cooker Type Injera Injera Efficiency (PHV) (Minutes) (MJ) (%) Biomass Fuels Twigs, leaves: sticks Open fire with "mtad" on 5.6 12.3 4.3 (1:3) stones, height 13 cm Twigs, leaves: sticks Filipini stove: molded clay design (1:4) sully enclosed with mtad inset and 13.1 11.1 4.8 chimney, height 15 cm Wood: Tigrai stove: enclosed by clay walls with many small gaps in walls 4.2 5.3 10.0 exchaust, height 14 cm Twigs, leaves: sticks Experimental: like Tigrai with chimney, (1:3) passive damper, height 12 cm 5.0 9.2 5.7 Electricity Sample of models commonly marketed in Addis Ababa 3.0-4.0 0.9 60 Mock-up of Aluminum injera cooker not available 80 PHV = percentage of heat utilized. Source: Mission testing and analysis. Short-term Strategies 2.27 On the supply side, the energy sources that can be tapped in the short-term all have to enter the cash economy to generate financiaL returns justifying the required investment. Hence, they will be directed at urban markets and will benefit the rural areas indirectly by reducing the pressure there of urban household fuel demand. The economic merit of proceeding to mobilize the new or unutilized energy resources outlined above can be judged by comparing the economic cost of useful energy delivered from the available fuels. The relative impact of developing each resource can be judged by portraying the displacement effect of supply, in order of least cost, on the major target market in each stra- tegic area: the northern and the southern region. This methodology has been applied more thoroughly to the southern region than to the northern region which the mission was unable to visit, and in respect of which it feels much less able to pass judgement. Comparative prices and present and future economic costs of existing and new household energy resources for the Addis Ababa target market are provided in Table 2.10. Dung is not included here as it is used minimally in Addis Ababa, and the compar- ison does not include injera cooking which is dealt with separately in Table 2.11. - 29 - Summary and Conclusions on Options for the Southern Region 2.28 The long-run economic costs of kerosene and electricity are comparable to one another on a useful energy basis and are slightly lower than those for LPG. If kerosene is used only in pressure stoves it remains cheaper than electricity even in the longer-term. All three are more expensive than the new short-term biomass energy sources proposed which, for normal cooking, are ranked in Table 2.8 in order of increasing cost, starting with charcoal. With the possible exception of bagasse, cost is a reasonable indication of the degree of difficulty of supply. The relevant observations and strategies that are implied can be sum- marized as follows: (a) Peri-urban plantation fuelwood is cheaper than all other sources, verifying the priority for fuelwood development for longer term supplies. Since new supplies from this source will not be available this decade, this relative economy does not alter the need for proceeding to develop new sources for short- term supply. (b) Charcoal produced in metal kilns from logging wastes and coffee residue briquettes are the lowest cost fuels. Pilot production of coffee residue briquettes and the commercial application of metal kilns should begin immediately. (c) Cotton and cereal straw residues are estimated to be 20-25% more expensive, and production is more novel and complex. Their overall economy is good and the quantities are significant, warranting comprehensive pilot production along with coffee residues. (d) Corn and sorghum stover, sawdust and bagasse residues are more difficult to access locally though all retain good economy in relation to the market price of wood and charcoal. Complex- ities of management and technological uncertainty, and hence, priority, should be resolved by immediate review followed by pilot production. The availability of resources for the energy sector in general will determine the rate at which this work proceeds, though with adequate funding, pilot production should begin in 1985. (e) The impact of aluminumh cooking pots on the efficiency of wood burning is very significant, reinforcing the view that market research is urgently required on consumer response, pricing and dissemination strategy. (f) Consumer acceptance of more efficient new wood stoves, modified charcoal stoves, and primus kerosene stoves would not alter the supply side priorities: woody fuels simply become compara- tively much cheaper. While trial production and dissemination of the innovations specified are strongly recommended, no - 30 - impact of potential savings should be reflected yet in demand projections for supply side investment planning. Rather, care- ful monitoring and evaluation procedures should be established to record the impact of the recommended cooking efficiency program. (g) For the rest of this decade at least, electric injera cooking will be cheaper than all biomass-fueled cookers (see Table 2.11) with the possible exception of the Tigrai injera cooker using residue briquettes and "new" charcoal sources. Firewood from new peri-urban plantations will be cheaper than all other fuels, but cannot be available in less than six years. Electric injera cookers using improved aluminum mtads appear to be the most economic. A program of electrifying injera cooking is economically justified within the limits of power supply in the short-term with special emphasis to be placed on the development and dissemination of high-efficiency, low-power aluminum mtads. Given the high purchase costs of electric injera cookers (B250-B350), their appeal is limited to the middle and upper income groups, or about 20% of the urban population connected to the EELPA grid. -. 31 - Table 2.10: COPARATIVE COSTS OF SUPPLY OF HOUSEHOLD FUELS EXCLUDING INJERA COOKING IN THE SOUTHERN REGION URBAN AREAS USING ADDIS ABEBA AS THE REFERENCE MARkET (Ethiopian Cents (ec) and Birr (8)) Cost Per Unit Energy Value 4/M,J Conversion Cost of Useful New Quantity Sold Efficiency (1) Energy f/MJ Accessible ('000 Actual Tonnes) LIQUIFIED PETROLEUM Gas (Propane) 8 Official Price 12.5 kg cylinders 1.20 8/kg 50 MJ/kg 2.60 45 5.78 1.5 kg cylinders 1.50 8/kg 3.00 45 6.67 Economic Cost 88.2 ec/liter 25.4 MJ/liter 3.47 45 7.72 Long Run Cost (1995) 97.5 ec/l iter 3.84 45 8.53 KEROSENE Official Price 65.4 ec/liter 36.7 MJ/liter 1.78 36 4.95 44 Sample Market Price 85 ec/liter 2.32 36 6.44 Economic Cost with wick stove 83.1 ec/liter 2.26 36 6.28 with primus stove 45 5.02 Long Run Cost (1995) - with wick stove 98.5 ec/liter 2.68 36 7.44 - with primus stove -- 45 5.96 ELECTRICITY (EELPA-ICS) Present Tariff (av) 15 ec/kWh 3.6 MJ/kWh 4.17 60 6.95 25 GWh LRMC (minimum) 16 ec/kWh 4.44 7.40 FIREWOOD Government 9.1 ec/kg 16.6 MJ/kg 0.55 8 a/ 6.88 n/a Supply (av) (15% mcwb) 13 4.23 Market Price 13.8-18.2 cc/kg 1.04 (av) 8 a/ 13.00 Economic cost from 15.4 MJ/kg 13 8.00 per i-urban B 23/m3sv (20% mcwb) plantations 0.27 8 2.90 (Addis Ababa) (450 kg/m3bd) 13 2.08 CHARCOAL Government supply 80.39/kg in 40kg sacks 29 MJ/kg 1.34 23 5.81 Sample Market prices B1.OO/kg (av) (0.2-0.25kg lots) 3.45 23 12.83 Economic cost (metal kllns) BO.18/kg 0.61 23 2.65 Economic cost (earth kilns) 80.19/kg 0.67 23 2.91 OOFFEE RESI WE Briquettes a 66/te 15.8 MJ/kg 0.42 15 2.80 150 in 40 kg sacks BAGASSE Briquettes 890/te (est.) 17.3 MJ/kg 0.52 15 3.47 43 in 40 kg sacks COTTON RESIDUE Briquettes 892/te 17.3 MJ/te 0.52 IS 3.47 239 in 40 kg sacks CEREAL STRAW Briquettes B93/te 17.A MJ/kg 0.53 15 3.53 l08 in 40 kg sacks SAWDUST Briquettes 898/te 17.7 Ni/kg 0.55 15 3.69 11 In 40 kg sacks CEFEAL STOVER Briquettes 8103/te 15.0 M./kg 0.68 15 4.53 211 In 40 kg sacks a/ With clay cooking pots; mcwb = moisture content we+ basis; bd = basic dens;ty; sv = solid volume; av = average; est = estimate. Source: Mission estimates and field measurements. - 32 - Table 2.11: COSTS OF PRESENT AND PROSPECTIVE ENERGY FORMS FOR INJERA COOKING IN ADDIS ABABA (1983 dollars) Energy Form and Cost Per Conversion Cost of Useful Cooking Mode Unit t/MJ Efficiency (%) Energy (f/MJ) (PHV) Electricity--Present average tariff -- Existing Clay 15 ec/kWh 4.17 60 6.95 "Imtad" cookers -- Potential Aluminum 80 5.21 llmtadtl cookers Electricity--LRMC -- Existing Clay "Imtad" cookers 16 ec/kWh 4.44 60 7.40 -- Potential Aluminum "mtad" cookers 80 5.55 Woodfuels -- Open Fire Cooker 4.3 . Government Supply 9.14 ec/kg 0.55 -- 12.79 . Market Price 16 ec/kg (av) 1.04 -- 24.19 P Plantation Long-run economic cost B23/m3 (sv) 0.27 -- 4.74 -- Tigrai Type Stove . Government Supply 9.14 ec/kg 0.55 __ 10.0 Market Price 16 ec/kg (av) 1.04 -- 10.40 . Plantation Long-run economic cost: Peri-urban plantations B23/m3 (sv) 0.27 -- 2.70 Residue Briquettes B90/te (av) 0.52 -- 5.17 Charcoal Long-run economic cost (metal kiln) B5-30/30 kg 0.61 -- 6.10 sack mcwb = moisture content wet basis bd = basic density sv = solid volume av = average PHV = percentage of heat utilized. Source: Mission testing and analysis. - 33 - Potential Impact of Short-term Measures Southern Region 2.29 The impact of the supply measures proposed here can be illus- trated practically for the southern region by using Addis Ababa (popula- tion 1.3 million), with half of the southern urban population, as the target market into which all new fuels are delivered. An achievable production schedule for new biomass fuels, and a dynamic energy balance for Addis Ababa for 1983-92 also showing the expansion of supplies of LPG, kerosene and electricity for cooking, are provided in Annex 2.5 and 2.6 respectively. The production of new biomass sources for crop residue is based on harvesting the presently accessible resource (Table 2.4) by 1992; for forest residues, production is based on building up the capa- city to harvest the 1992 resource after harvesting the vital Bubeka and related forest residue resources, some of which peak at 1987; and optimal management of the AfDB funded and the existing Addis Ababa fuelwood re- sources. Being optimistic, the result, which is summarized in Table 2.3, is that: (a) new biomass fuel sources and expanded supplies of modern fuels exceed the demand for cooking fuels in Addis Ababa in 1992 by almost half a million tonnes of wood equivalent (see Annex 2.7 for wood equivalence); (b) the deforestation of Addis Ababa's hinterland, now between 30,000 and 80,000 ha a year, 7/ is eliminated by 1988; and (c) the total new househoLd fuel supply is equivalent both to 60% of the 1992 deficit in the desired woodfuels supply to urban households in the southern region and to the annual production from a fuelwood plantation of about 170,000 ha. Northern Region 2.30 In the northern region, since none of the new biomass fuels can be economically delivered there from the south-central agriculture and forestry regions, the only economic option in the short-term is to expand the supply of petroleum fuels. Table 2.12 shows that cooking with kero- sene pressure stoves is 65% of the cost of the cheapest cooking method presently in use (open fire using aluminum pots). Without resorting to large scale imports of kerosene, the limit of supply to the north from the existing refinery is about 40,000 tonnes, allowing 14,000 tonnes for Addis Ababa and other southern region markets. The impact of this sup- ply, as shown in Table 2.3, is to meet some 20% of the deficit in wood- fuels supply in the north by 1992. The option to import more kerosene, with all the attendant problems of foreign exchange availability, 7/ A hectare of savannah land normally has a stocking density of 30m3 (sv) ha but with thinning in Ethiopia, most is now only IOm (sv) ha. The estimated annual deforestation reflects this range. - 34 - distribution problems and product security, should be addressed immedi- ately by a special study recommended by the mission to establish a com- prehensive strategy to meet short-term household energy needs in the northern region. In the longer term, woodfuels are the most economic source of cooking fuels for the north despite arid growing conditions. Supply of substantial quantities of firewood could be available for new planting in the early to mid 1990s if the appropriate investments were to be made now (see para 2.32) Medium- to Long-Term Strategies (1992-2012) 2.31 From the foregoing analysis it is evident that, for the longer- term, the least cost and most desirable method of meeting household energy needs for most of the population is to grow trees. Hence, a greatly increased program of afforestation is strongly recommended in the form of peri-urban plantations to meet the urban demand, and rural wood- lots and agroforestry to meet the much larger rural demand. The scale of the planting operation implied (about 3.2 million ha) to meet the 1992 demand is so high that the mission cannot foresee it being implemented. Thus, both need-based and lower, more realistic or "achievable" programs of afforestation are here outlined, involving the equivalent of 1.12 million ha of planting during 1984-92. At the present rate of planting, only 120,000 ha of dedicated fuelwood planting would be established during the same period. The approximate cost of the expanded program would be about US$370 million. The manpower and training implications of these programs are dealt with in Chapter VIII. Work to achieve these longer-term supply side effects is of the utmost urgency, and should begin immediately. Aforementioned initiatives to achieve a short-term impact interact strongly with long-term fuelwood production targets and, together with the initiatives promoted below, should form part of a care- fully integrated national program. Again, there is merit in addressing this problem in different ways in the two regions -- northern and southern -- though here a national program is outlined with reference only where relevant to unique aspects of the northern situation. Peri-Urban Plantations 2.32 Ethiopia is unique in Africa in having developed peri-urban fuelwood plantations at the turn of the century. Although the majority have been planted in the last 20 years, there are about 20,000 ha of Eucalyptus globulus fuelwood plantations surrounding Addis Ababa and 60,000 ha in other peri-urban areas. Planting currently totals about 2,000 ha per year, though an AfDB funded project of 15,000 ha for Addis Ababa should plant 1,000 ha in 1984 and proceed at the rate of 3000 ha per year thereafter to completion. SmalLer, 500-700 ha plantations for five other cities also are expected to begin in 1984. - 35 - Table 2,12: COWARATIVE COST OF SUPPLY OF HOUSEHOLD FUELS ro THE NORTHERN REGION USING AMARA AS THE REFERENCE MARKET Ethiopian cents (ec) and Birr (8) Cost Energy A/MJ Conversion Cost of Useful Per Unit Value Efficiency Energy ( /A.J) ELECTRICITY (ERESA, ICS) Present Tariff 9ec/kWh 3.6MJ/kWh 2.50 60% 4.17 Economic Cost (SRMC) 31.5ec/kWh 8.75 14.58 KEROSENE Economic Cost 76.9ec/liter 36.7MJ/1 2.10 45% 4.67 Long Run Cost (1995) 92.3ec/1iter 2.52 5.59 (Pr imus Stoves) LIQUIFIED PETROLEUM GAS Economic Cost 81.7ec/liter 25.4MJ/f 3.22 45% 7.15 Long Run Cost (1995) 91.3ec/liter 3.59 7.99 FIREWOOD (est. Long Run Cost) B40-45/mr3 Open Fire/ Normal Cooking 15.4MJ/kg 0.49(av) - Clay Pot as sold 8S 6.49 - Aluminum Pot (20% mcwb) 13% 3.77 Portable Metal Stove 28% 1.75 Injera Cooking Tigrai Stove 10% 4.90 FIREWOOD-MARKET PRICES a/ Open Fire B167/te 17.7MJ/kg (1980/81) (10% mcwb) 0.94 - Clay Pot 8% 11.79 - Aluminum Pot 13% 7.23 Injera Cooking 10% 8.80 - Portable Metal Stove 28% 3.14 CHARCOAL (est. LRMC) 8330/tonne 1.14 Existing Stove 29MJ/kg 23% 4.96 Improved Stove 30% 3.80 CHARCOAL--MARKET PRICE a/ B840/tonne (1980/81) Existing Stove 2.90 23% 12.59 Improved Stove 30% 9.67 DUNG (est.) MARKET PRICE b/ 14.7ec/kg 14.8MJ/kg 0.99 (at 10% mcwb) Open Fire - Clay Pot 9% 11.00 - Aluminum Pot 15% 6.60 Improved Stoves using aluminum pots 30S 3.30 a/ Records of prices applying late 1980 and early 1981 from ENEC. b/ Measured at Gondar market. Asmara was not visited by Mission. Source: All stove efftciencies, except LPG stoves, as measured by mission. Other data are mission estimates. mcwb = moisture content wet basis, bd = basic density, sv = soli d volume, av = average. - 36 - Existing Addis Ababa Plantations 2.33 The productivity of these plantations was estimated to be 15-20 m3/ha mean annual increment (mai) late last decade and in 1980 a survey found the average standing stock to be 118 m3/ha, implying an acceptable five-eight year rotation. There is considerable evidence now that the rotation has declined to two-three years or less in many areas, with a high volume of 5-10 cm branch and entirely leafy shoot material flowing into the market. A survey by the mission and GOE officials of woodfuels inflow to Addis Ababa estimated that 170-190,000 tonnes of Eucalypt leaf and twig matter, about 25% of the total fuelwood supply, reaches the Addis market each year. This material is at least ten times higher in mineral nutrients than the heartwood of mature trees normalLy har- vested. Its use as a fuel, and the dramatically shortened rotation, implies greaily decreased soil cover and fertility, fast declining yields (now 5-10 m /ha mai) and greater stump death. In addition, the high volatility of leafy fine firewood reduces cooking efficiency and in- creases overall demand (see Table 2.9). Dead stumps cover at least 1,000 ha and are interspersed with at least a further 5,000 ha. The mission recommends stump removal and replanting in totally dead areas, and en- richment planting in the remainder. Coppice management should be im- proved to increase productivity and reduce nutrient Loss. Without these measures, the mission estimates that productivity will decline by 50% over the decade. New Plantation Design and Management 2.34 Several aspects of proposed plantation design and management deserve review as part of a necessary diversification to upgrade biolo- gical and economic yields. First, the Ethiopian fuelwood crisis is so severe that efforts to achieve harvestable production earlier appear economically justified. Continual rotation lengths of four-five years may be achieved with heavier fertilization, good initial Land preparation (ploughing) and close spacing. The mission estimates that the Addis Ababa fuelwood plantation, for example, could yield a 25% higher mai, and be harvested at year five, with a 10% reduction in the unit cost of pro- duction. Land requirements will be reduced accordingly. In addition, there must be a wider choice of species and provenances applied, with FAWCDA being constantly alert to the performance of high yielding candi- dates from its own and overseas plantations. Finally, the multiple role of peri-urban forestry in producing both fuel and timber for construction should be explicitly recognized as the plantation development program matures. The Need For Peri-Urban Forestry 2.35 For Addis Ababa, if new biomass fuel sources (see Annexes 2.5 & 2.6) were not available, an additional some 120,000 ha (at 20 m3/ha mai) of fuelwood plantations will be required by 1987 to meet the shortfall in household energy supply in 1992. However, the maximum achievable in addition to the existing AfDB project, is 16,000 ha by 1992. For the rest of Ethiopia a program has been developed based on existing forest - 37 - cover and yield, allowing, as a simplification, no erosion to the exist- ing productive base, and applying an estimate of plantation producti,vity based on available data, and the experience of FAWCDA and mission staff. Cost data reflect both recent experience and recent detailed cost esti- mates for a number of environments. Annex 2.8 provides estimates of peri-urban plantation needs by province using the mission's estimate of need of 2 kg air dry wood per capita per day, and excluding the needs of Addis Ababa. Annex 2.9 provides three sample silvicultural and cost models for typical ecosystems encountered: lowland (below 1,800 m), highland plateau (1,800-2,800 m) and erosion prone highlands (1,800- 2,800 m). These cost models define direct cost only, whereas overall costs include infrastructure development and will be about 100% greater. The investment program outlined for peri-urban forestry is based on the higher cost, including infrastructure, of about B 1,760/ha (see para 2.42). These costs are comparable with those experienced in other African countries (US$800-1,200/ha) for industrial scale plantation with extensive infrastructure development. The long lead time for fuelwood development and the ever expanding need due to a rapidly growing popu- lation demand that a longer than ten-year perspective be taken both to review the impact of nearer-term investment and to gauge the nature and level of the ongoing commitment. To meet the 1992, 1997, and 2002 urban demand, an estimated cumulative plantation establishment of 387,000 ha, 545,000 ha, and 764,000 ha, respectively, has to be completed. It should be emphasized that the area required to meet the 1997 demand, using a five year rotation, must be planted by 1992, and so on. It follows, then, that the funds required to meet the need for peri-urban plantation development, including for Addis Ababa, are about B480 million, 1982- 1993, and B670 million Birr, 1982-97, to meet the 1997 and 2002 demands, respectively. The four provinces of the northern zone use 63% of these funds for planting 60% of the specified area. Eritrea alone requires 39% of all peri-urban plantations and 41% of all investment in the program. Achievable Peri-urban Plantation Program 2.36 The mission finds it inconceivable that peri-urban forests could be planted at the rate required to satisfy demand during the next decade. Rather, a more realistic, though optimistic, rate of planting has been proposed, bearing in mind the key constraints of land deline- ation and manpower development and training. Table 2.13 presents the planting program required to meet the estimated need, and that considered achievable under optimum conditions with a full GOE commitment. Even with the program for peri-urban forestry, the shortfall against need will be 467,000 ha in 1987, growing to 536,000 in 1992. Thereafter, the gap narrows, closing to 296,000 ha by 2002, eventually achieving a balance between 2010 and 2020. If the program of production and expanded supply from new and existing sources outlined above (para 2.8-2.21) is imple- mented, whereby about two million tonnes of wood equivalent are produced each year by 1992, the shortfall for the whole of the southern region in peri-urban forestry will be about 80,000 ha in addition to that now funded. The shortfall in the northern region remains 191,000 ha, com- pared with a requirement of 240,000 ha, once again highlighting its special problem (see Table 2.3). Table 2.13: Need and Achievable Rate of Implementation for Energy Forestry Pern-Urban Plantations a/ Rural Forestry Achievable Target Achievable Target Need Annual Accumulated Cumulative Need Annual Accumulated Cumulative Year (Accumulated) b/ Performance Planting Deficiency (Accumulated) Performance Planting Deficiency ('000 ha) 1982 335 2 2,006 10 83 4 6 20 30 84 6 12 40 70 85 8 20 60 130 86 10 30 80 210 1987 512 15 45 467 2,344 100 310 2,034 88 20 65 110 410 89 25 90 120 540 1 90 30 120 130 670 91 35 155 140 810 co 1992 731 40 195 536 2,608 150 960 1,648 93 50 240 160 1,120 94 60 300 170 1,290 95 70 370 180 1,470 96 80 450 180 1,650 1997 1032 80 530 502 3,140 180 1,830 1,310 1998 to 2002 1226 80 930 296 4,074 180 2,730 1,344 to 2012 5,095 180 4,580 565 a/ Need includes requirement of Addis Ababa, but is not reduced to take account of the production of new biomass fuels (e.g. crop residue briquettes) or- expanded LPG, kerosene or electricity supply. b/ Indicates area to be established by that date to meet demands forecast five years hence, i.e., 512,000 ha are to be established by 1987 to meet the 1992 demand for woodfuel using a five year rotation. - 39 - State Farm Forestry 2.37 Special mention is warranted of the need to reforest state farms, the establishment of which has caused the same "urban shadow" of deforestation and fuelwood shortages as described for cities (see para 2.5). State farms are almost entirely void of tree cover as a result of wholesale clearing during their establishment. Apart from causing local- ly acute firewood shortages, the removal of this tree cover has allowed severely eroding and dessicating winds to lower productivity. In the mission's view, 10-15% of state farm areas, or 20-30,000 ha, should be afforested (between blocks, along roads and surrounding compounds). Higher yields resulting from increased soil moisture and crop production should more than repay the cost of establishing plantations and any re- duction in crop land, quite apart from the value of the timber produced for fuel and construction. The area and cost of this program are covered within the overall peri-urban forestry program. Rural Energy Forestry 2.38 The lead agencies in rural forestry are FAWCDA's department of Community Forestry and the Soil and Water Conservation Department (SWCD) of the Ministry of Agriculture. The former was created in 1980 to pro- vide and service peasant association and school nurseries, demonstration plantations, and training facilities in rural forestry, including some dedicated fuelwood production. Table 2.14 illustrates the impressive start made by this department which in three years has raised over 50 million seedlings and established 45,000 ha of plantations. The SWCD activity is not aimed at woodfuel production, though this is an inevit- able by-product so that in reality there is an overlap between FAWCDA community forestry and SWCD programs. In the past two years SWCD has also planted about 45,000 ha. The World Food Program (WFP) supports the SWCD activities and FAWCDA's community forestry department, with about two-thirds of WFP funds going to the former and one-third to the lat- ter. However, the productivity of planting for soil protection is per- haps 5m3/ha mai or lower with at least 10 years to a possible harvest, and then with only limited permissible resource recovery. Thus, from a planning standpoint, the contribution from SWCD forests to woodfuel supply can be regarded as negligible. The same applies to the even more limited activities of the Relief and Rehabilitation Agency, a special agency responsible for resettlement which had a target for protective forestry of 4,000 ha in 1981/82. Technical Performance and Requirement 2.39 By regional standards the start made by Ethiopia in all forms of rural forestry is very impressive. Regrettably, the need is far larger. To improve the efficiency and economic benefits of the required effort, several fundamental improvements should be made: emphasis on genuine agroforestry, careful selection of species and provenances, more optimal spacing, and better coppice management. It follows that this im- provement can only arise through careful performance monitoring and field trials. Eucalypt species often compete with, rather than complement, - 40 - crop production. The strong start made with Eucalypt woodlots should now be complemented by an aggressive agroforestry program selecting, testing, and promoting rapidly growing trees which fix nitrogen, which cycle mineral nutrients, and which do not have intolerably vigorous surface rooting. These trees are to be closely integrated with crop production in just the same way that existing slower growing Acacias do now. Con- siderable improvement in provenance and species selection is possible for woodlots, though this will go hand in hand with simple trial work, ample seedstocks, and adequate analysis and feedback, both to program managers and to field staff. The mission recommends the early establishment of simple experimental and recording procedures at the village level linked with the essential higher-level experimentation as part of a nationwide management information system for energy forestry. The Need for Rural Energy Forestry Program 2.40 The- same methodology that was applied to forecast the demand for urban forestry has been used to generate projections of the required rural forestry effort. The results are provided in Annex 2.10. To meet the 1992, 1997 and 2002 rural demand, the equivalent of 2.3 million ha, 2.6 million ha and 3.1 million ha would have to be in production, in 1987, 1992 and 1997, respectively, having been planted at least five years earlier. The cumulative matching expenditure by these latter dates is B211 million, B245 million and B283 million, respectively. These are direct costs only as they exclude labor and infrastructure requirements, manpower training and development, administrative salaries and wages and transportation. In the period 1982-97, 670 million man-days of labor would be required for program implementation. However, regardless of the availability of funds, it is quite inconceivable for planting to proceed at the required rate. Achievable Rate of Rural Forestry Development 2.41 The key constraints on the rate of rural afforestation are cur- rent deficiencies in manpower training and development, program planning and management infrastructure. Nevertheless, if there were a truly strong national commitment to this program, the mission is optimistic that an order of magnitude increase in planting can be achieved. Refer- ring back to Table 2.13, it is believed practical to budget on achieving a rate of 100,000 ha per year within five years and 150,000 ha per year within ten years, planting 960,000 ha by 1992 compared with a need at that time for 2.61 million ha. Extrapolated out, the achievable rate of planting does not catch up with demand until about 2020. At whatever level planting can be achieved, it is noteworthy that the northern region of four provinces requires just over half of the total projected affores- tation. Costs of Achievable Forestry Program 2.42 The cost data for rural afforestation provided in Annex 2.10 refer only to direct Peasant Association costs of nurseries and to the - 41 - required man-days. A simple cost estimate is provided in Annex 2.11 for rural afforestation based on an allocation of overheads for an ultimate planting rate of 200,000 ha per year. Here, the costs compute to about B420 per ha (in 1983 Birr) or B403 million for the 1992 rural forestry target of 960,000 ha, plus peri-urban forestry at B1760/ha, half in direct costs and half for infrastructure, or B343 million. The total cost of the program is B746 million for the 1983-92 period. In this estimate peasant labor is costed at B1.00 per man-day on the basis that labor will have to be hired (using food as payment). If this labor is genuinely donated, the financial cost drops by B202 million to B544 mil- lion for the decade. In addition, the required manpower development and training, the remodeling and expansion of existing institutional capabi- lity, and the new buildings and hardware for training will cost about B21 million. These costs and those derived for the implementation of the short-term strategy are summarized below in Table 2.15. The total re- quirement is B152 million for direct investment in the measures which will have a significant impact during the 1987-92 period, including up- grading the Addis Ababa plantations (paras 2.30 and 2.31). It is assumed that the State Farms administration bears the infrastructure and manpower training cost of the crop residue program as part of their own expansion. The total cost of energy forestry for the decade is B767 million. How- ever, it is noteworthy that in the five years 1992-97 during which time trees will be planted to meet the 2002 woodfuel demand, a further 1.2 million ha is required at an additional cost of almost 31 billion. Table 2.14: FAWCDA COMMUNITY FORESTRY PRODUCTIVITY SINCE FORMATION PA and Demonstration School Seedlings Demonstration Year Trained Farmer Nurseries Nurseries Raised Plantations (toco) (millions) ('000 ha) 1980/81 10.64 30 74 1.2 14.0 1981/82 19.70 41 72 18.6 10.6 1982/83 20.00 225 67 32.0 20.0 50.34 296 213 52.8 44.6 PA = peasant association. Source: FAWCDA and Mission Estimates Economic Justification of Energy Forestry 2.43 The program of expenditure on energy forestry which the mission believes to be achievable represents more than a tenfold increase over the present rate of energy forestry operations. This program implies a diversion of resources to fuelwood production alone far greater than ever applied to foresty as a whoLe in previous budgets, and it therefore must be supported by a sound economic rationale. Fuelwood produced from peri- - 42 - urban plantations has already been shown in Table 2.10 to be the least- cost option for supplying household fuels in the longer term, being less than half the cost of all competitors besides coffee residues and char- coal from logging wastes. Another measure of the economy of the peri- urban program is the rate of return on investment, both financial and economic. The financial rate of return (FRR) will indicate how likely it is that smallholder entrepeneurs would find it attractive to participate in fuelwood production, and the economic rate of return will serve as an input to government decision making on investment priorities in the context of overall resource limitations. Firewood produced on the fringe of the large cities will be sold in competition both with firewood from the hinterland and with kerosene. Using the cost data of the AfDB Addis Ababa fuelwood plantations, which in the mission's view is 20% or so more expensive than will be future plantations, an FRR of 29% is achieved against the current average market price of firewood (Addis Ababa, Dire Dawa, Nazareth and Gondar prices), and a 20% and 23% ERR is achieved against the present and escalated (1.8% p.a.) real cost of kerosene, respectively. Economics of Rural Forestry 2.44 The economics of rural afforestation are more complex and rest substantially on the perceived consequences of rural deforestation out- lined in paras 2.4 and 2.5. Here a major quantifiable cost is the value of the fertilizer lost to agricultural production when dung is burnt as a fuel or, where there is no conceivable alternative fertilizer, the value of foregone crop production. In Ethiopia it has been possible to aquire, from a number of sources, sufficient data to make a first order estimate of these costs (a separate note is available). Using a conservative estimate of the economic cost of fertilizer (Diammonium Phosphate: DAP) at the farm gate of US$295/tonne (cf. government selling price in distri- bution centers of US$411/tonne), dung is valued at B45/tonne for its nitrogen and phosphorus content alone. However, the constraints to im- porting the quantity of fertiLizer required to substitute for all dung burnt -- 490,000 tonnes of DAP costing US$114 million (CIF) -- must be recognized, along with an inability to distribute imported fertilizer to more than, say, 20-30% of peasant farmers. Applying data gathered on the pattern of subsistence agriculture in northern Shoa province, 8/ and with the use of appropriate fertilizer response curves drawn from Ethiopian data, incremental grain yields from the application of dung otherwise burned ranged in value from 8130-190/ha, though higher values can be generated and justified. 8/ By the International Livestock Commission for Africa: ICLA. - 43 - Table 2.15: COSTS OF AN ACHIEVABLE HOUSEHOLD ENERGY PROGRAM Yearly Production Total Investment Source by 1992 During 1983-92 ('000 actual tonnes (B million) of fuel product) Short-term Measures Crop Residues 751 72.0 Charcoal 337 41.0 Improved Cooking Efficiency n.a. 3.0 Addis Ababa Plantation Upgrading a! Added Fertilizer AfDB Project 598 9.4 Replant existing plantations 18 26.4 Subtotal 151.8 Medium to Long-term Measures ha ('000) Peri-urban Plantations 195 403.0 Rural Energy Forestry 960 343.0 Manpower Training and Development n.a. 21.0 Subtotal 1,155 767.0 Total for 1983-92: 918.8 Additional Costs to Meet Additional Planting Total Investment Household Energy Needs of 2002 1992-97 1992-97 ('000 ha) (B million) Peri-urban Plantations 335 590 Rural Energy Forestry 870 365 Total for 1992-97 1,205 955 a/ Additional wood yield forecast in air dry tonnes (25% mcwb), the same as tonnes of wood equivalent, up to and including 1992 over yields available from proposed AfDB project design and cutting program and no infill planting for existing plantations (see Annex 2.6). Source: Mission estimates. It should be clear that there are other economic benefits of using dung as a fertilizer and many benefits of afforestation per se that cannot readily be quantified. The data presented here imply an annual loss of grain production through the use of dung as a fuel of between 1 and 1.5 million tonnes, with an economic value of between B850-1275 billion. The estimated value of various uses of dung are as follows: - 44 - Dung (B/tonne) (a) Value of dung based on the cost of pro- 6-9 duction of the equivalent fuelwood through rural forestry (including labor inputs at Bl/man-day) (10-15m /ha mai); (b) Value of dung (nitrogen and phosphorus 45 content only) equated to farm gate costs of imported fertilizer; (c) Value of dung (N&P content only) equated 128-189 to the minimum incremental production of Teff when applied as a fertilizer in typical peasant agricultural system. These data indicate that the lowest value use of dung is as a fuel, jus- tifying the investment in rural forestry to supply firewood to displace it for higher value uses. The ERR's for rural forestry at the lowest yields of lOm3/ha mai, displacing dung for use as a fertilizer and assigning to it first its imported fertilizer equivalent value, then its low and high grain production equival nt value, are 35%, 59% and 70%, re- spectively. Even with a yield of 5m /ha mai, equating dung to imported fertilizer, the ERR is 23%. Clearly, investment in rural forestry is robust even in the very marginal arid areas and must be attributed high priority in the overall energy sector investment program. ImpLementation of the Achievable Energy Forestry Program Peri-urban Forestry 2.45 The institutional arrangements for implementing the proposed program of afforestation will vary between the peri-urban and rural com- ponents. The mission sees FAWCDA as the lead agency in peri-urban fores- try in accordance with the existing government proclamation. FAWCDA has the base from which to develop the capability to implement this formid- able industrialization of commercial fuelwood forestry. However, it is clear that to plant at the required rate, FAWCDA will have to limit its role beyond the plantation establishment phase to management advisory services and monitoring. Urban kebeles and, where relevant, rural peas- ant associations (PA), will have to be brought increasingly into day-to- day management of the maintenance, harvesting and marketing phases. In the long run, kebeles should also become the prime executing agencies. It is essential for there to be adequate incentives for Kebele and PA leadership to manage plantations for high sustained yield: unlike the situation of heavy overcutting for short-term gains now a feature of the management of the Addis Ababa plantations. The mission favors the allo- cation of plantation blocks to smallholder families under the general supervision of local authorities, and within guidelines established by FAWCDA, with appropriate stumpage fees paid to FAWCDA enabling it to profitably recover its investment and to reinvest quickly in new planta- tions. - 45 - 2.46 FAWCDA will require a flexible and dynamic institutional arrangement in order to cope with the early development phase of the next decade when skilled manpower and infrastructure will certainly be lim- ited. As over half of the plantations to be established will be about 500 ha and located near small towns, the option of establishing mobile extension and support units of well-trained staff supporting the local FAWCDA plantation manager (at least one technician per 500 ha) should be seriously considered. Plantations also could be designed and established in standard 500 ha packages tailored to the needs of particular regions, facilitating budgeting, procurement arrangements, and management infor- mation services. Rural Energy Forestry 2.47 The implementation of this program must take full advantage of the existing and successful structures including both the regionalized political organization from province, through Awraja, Woreda, and PA's, and the administrative, public service vehicles of FAWCDA's Community Forestry Department and the Ministry of Agriculture's extension services. There are several options open to COE though one essential element is to establish effective contact right down at the PA level, and through the "contact farmer", with individual peasants. This approach is fundamental because of the importance of genuine agroforestry, on the one hand to augment the supply of fuelwood and on the other hand to protect and revitalize agricultural productivity. Agroforestry involves forestry intimately interwoven with, rather than on the fringes of, crop produc- tion, hence it is appropriate that the concept be delivered through the familiar agency of the agriculture extension service. Nevertheless, FAWCDA could provide the subject specialist at say, the Woreda level, to ensure that advantage is taken of its existing community forestry exper- tise. FAWCDA would continue to deal with PA's or cooperatives concerning the design and establishment of fuel and timber woodlots for the commu- nity, and with the establishment of nurseries, training in woodlot man- agement and the like. The Ministry of Agriculture can best incorporate agroforestry extension services by strengthening the basic structure pro- posed by the Banks Minimum Package Program, once again ensuring a simple and hopefully consistent line of advice to the contact farmers within PA's. The level at which an active interface between MOA and FAWCDA takes place, and the nature of the interaction required, must be clari- fied and agreed as a matter of priority. Details of the overall manpower training and development requirements of the achievable forestry program are provided in Chapter VIII. Consequences of Failure to Act on Rural Afforestation 2.48 Two interlinked examples of the present direction of change serve to illustrate the dire consequences of inaction: deforestation in Gojam and the pattern of dung supply and demand in the populous grain producing regions. Gojam, a major grain producing province with 2.1 mil- lion people, presently has an estimated sustainable production of only - 46 - 40% of the provincial demand for fuelwood, or a supply equivalent to the annual production of a forest of about 160,000 ha with 100 m3/ha standing biomass. The mission calculated the impact of population growth on the available fuel sources if there is no further afforestation. From this forest resource base, and with the estimated cow dung supply and demand pattern provided in Annex 2.12, by the turn of the century Gojam will have no trees, no surplus dung, and merely old tree roots, brush and grass for fuel. This situation will be mirrored in Hararghe, Bale and so on across the middle belt of highland provinces all reaching more or less the situation of Eritrea today within 10 to 20 years. Equally alarming are the conclusions to be drawn by matching present supply and demand patterns for dung and crop residues (Annex 2.12) with the growth in household energy demand by province (Annex 2.13). Even assuming no reduction in the growing stock of trees, the present annual cow dung pro- duction accessible is entirely utilized in Bale, Coma Gofa and Wollo by 2000, and is exhausted within ten years in Hararghe, Gondar, Tigrai. In Wollega, Gojam and Sidamo, 50-70% utilization is implied, much like the situation now in Tigrai and Gondar. Nor is it realistic to forecast a significant increase in cattle population in most of these provinces. Therefore, substantially more than double the quantity of dung now burnt will be diverted away from crop and fodder production by the mid- to late-1990s. The indications are, quite simply, that the conditions now prevailing in the northern region will progressively appear in most of the populous central and western regions during the next 10-20 years. 2.49 In summary, the mission believes that over a period of 30 years it is possible to restore the supply and demand balance for rural energy, and to stabilize and rejuvenate agricultural productivity. The effort during that period is equivalent to reforesting 5% of the Ethiopian land- scape. The required inputs of funds, manpower and other resources are large, but not prohibitive. To achieve these objectives the GOE must make an unparalled commitment to afforestation and biomass resource mobi- lization. The loss of soil, and the rate of deterioration in soil fer- tility, is now so great 9/ that there must be grave doubts as to whether declining rural productivity can be arrested. However, it is clear that if left for another decade, with only the present level of activity, the task may well become impossible. 2.50 In the event that the rate of afforestation proposed here is unmet, or that the production of the novel and apparently economic resi- due fuels prove infeasible, kerosene and LPG will become the next least cost option for much of the urban population. This option has serious balance of trade implications. For example, if half of the anticipated 1992 population of Addis Ababa were to switch to kerosene as a cooking fuel they would consume 140 million liters a year at an economic cost of US$56 million, excluding real price escalation. 9/ Soil erosion in the Ethiopian highlands is estimated to average 20 tonnes per hectare per year, exceeding one billion tonnes of soil loss each year. From a long time series of aerial photography, it has been estimated that about 1% of the surface area of Ethiopia's best soils are lost each year. - 47 - III. ELECTRICITY 3.1 Ethiopia has considerable power resources in the form of hydro- power and natural gas and is in the process of exploring the potential for geothermal power. The main issues in the power sector center on the feasible level and source of future supply and, in turn, on the rate of growth in demand, the least cost program of power sector development and the related financing and pricing requirements. Decisions to be made on the future sources of power supply have significant implications for the development of other subsectors of the energy economy, making timely and comprehensive resource assessment and thorough power systems analysis of fundamental importance to overall energy sector investment planning. Institutional Arrangements 3.2 Electricity is generated under the authority of the Ethiopia Electric Light and Power Authority (EELPA) which was formed in 1956 as an autonomous parastatal and since 1976 has been under the Ministry of Mines and Energy. While EELPA maintains overall responsibility throughout the country for public power generation, transmission and distribution, the supply of power to the northern region, including the major centers of Asmara and Massawa, is provided by its unconsolidated subsidiary, the Eritrea Region Electricity Supply Agency (ERESA). The only reason for this unconsolidated arrangement is to facilitate discrete accounting as ERESA consumers are heavily subsidized. Presently the tariff level in force in the ERESA supply region is similar to that in the main inter- connected system (ICS) of the central highlands, even though the latter is supplied primarily by fairly low cost hydropower and the former is supplied from very expensive diesel and residual fuel oil-fired genera- tion. Each of the EELPA and the ERESA supply regions are composed of an interconnected system and a series of satellite generation centers referred to as self-contained systems (SCS). With only one major excep- tion, the Tis Abai hydropower plant at the outlet of Lake Tana which serves Bahir Dar, the SCS systems are supplied from diesel generators. In addition to public power supply systems, there are a host of captive generation systems within government and private enterprises as well as municipally-owned plants in many small towns and villages. The Power System 3.3 The population of areas served by electricity was estimated in 1980 to be about 2.8 million, which is about 9% of the total population. There are approximately 290,000 households, or about 50% of the popula- tion connected with power supply within the supply areas. Average per capita consumption was 18 kWh/year in 1980/81 (21 kWh/year in 1981/82) which is extremely low even by regional standards. 1O/ The distribution - 48 - of installed capacity by type is indicated in Table 3.1. Of the 330 MW of generating capacity in public, quasi-public and private hands,.66% is hydro, 11% is oil-fired steam, and 23% is diesel engine generation. The main power plants are marked on the accompanying maps. The public systems account for 283 MW or close to 90% of total installed capacity. Electricity production so far has been concentrated in the central highlands of Ethiopia where consumption has grown steadily from 1961 to 1982 at an average annual rate of 9.0%. In the northern region, the available production data indicate a growth of 1.5% p.a. between 1971 and 1981. The much slower growth is explained mainly by the impact of military operations on the economy of the regional capital of Asmara. Power generation had fallen to one-fifth of the 1973 level by 1978 and was restored to the earlier peak only in 1982. Greatly increased oil- fired generation of power during the last four years has added substan- tially to the average cost of power production nationally. This recent growth has simply restored production levels achieved a decade ago. Table 3.2 also indicates that growth in publicly-generated power sales nationally in the 1977-81 post-revoLutionary period of economic recovery was about 13% p.a. Of 1982 electricity consumption, 57% went to industry, 17% to government and commerce, and 26% to households (see Chapter I, Table 1.2) 3.4 The EELPA ICS generates about 73% of all public power supply and about three-quarters of this is used in Addis Ababa. A detailed analysis of recent trends in generation and sales for the EELPA ICS is provided in Table 3.3. Growth in sales was stagnant in 1981 but re- covered again in 1982 due largely to interconnections with the SCS, com- missioning of the first electric boiler loads and a spurt in domestic demand. The overall growth rate for the FY77-82 period was 9.4%, includ- ing interconnections. Understandably, the sectoral distribution of the sales within EELPA in 1982/3 was similar to the estimated distribution of sales nationally, with 60% to industry, 28% to households, and 12% to government, commerce and other categories. 3.5 Sales data for the EELPA SCS for 1977-82 are given in Table 3.4. The average annual growth rate for sales in the total system was 11.7%. By far the Largest single load center in the SCS is at Bahir Dar, situated on the southern edge of Lake Tana at the source of the Blue Nile River (Tis Abai). The public supply system is obtained from the run-of- river Tis Abai power station which has a present installed capacity of 7.6 MW (two 3.8 MW units). A third unit of 3.8 MW is scheduled to be installed by 1984. In addition, a project to narrow the outlet from Lake Tana to the Tis Abai is scheduled to be completed in 1985 to regulate the outflow from the Lake to increase hydro power production from the Tis Abai station during the dry season. Present production capacity for the station is about 30 GWh/year and this will be increased to about 100 10/ Comparative per capita consumption data for 1980 for other countries in the region are (kWh/year): Kenya--92; Tanzania--44; Uganda--21; Malawi--64; Sudan--60. - 49 - GWh/year after 1985. The remaining sales in the SCS (about 63 GWh/year in 1982) are produced from a multitude (over 200) of small public and privately-owned diesel generators which average less than 300 kW in size. About 50% of SCS sales are to industrial consumers, about 30% are to domestic consumers and 20% to commercial establishments. Table 3.1: INSTALLED CAPACITY FOR POWER GENERATION, 1981 (MW) Power Source Hydropower Steam Diesel Total Public Systems EELPA - ICS 207.20 -- 6.50 213.7 - SCS 10.07 -- 20.14 30.2 Subtotal ERESA - ICS -- 15.00 26.2 41.2 - sCS -- -- 8.70 8.7 Private Generation Asseb Refinery -- 13.5 -- 13.5 Sugar Corporation - Wonji -- 1.92 -- 1.9 - Shoa -- 0.38 -- 0.4 - Metahara -- 6.00 -- 6.0 Nippon Mining -- -- 0.48 0.5 State Farms -- -- 4.78 4.8 112 Towns and Villages -- -- 9.38 9.4 Total 217.27 36.8 76.18 330.3 ICS = Interconnected system SCS = Self contained system Source: Acres-EELPA reports and energy assessment mission. Table 3.2: PUBLICLY GENERATED POWER SALES IN ETHIOPIA, FY78-82 (GWh) 1977/78 1978/79 1979/80 1980/81 1981/82 Growth (Z p.a.) EELPA ICS 318 358 402 414 485 10.6 SCS 68 77 79 75 93 -- Subtotal 3 4T5 489 57§ 9.7 ERESA ICS 15 53 67 76 79 14.2 a/ SCS 5 5 5 4 4 -- Subtotal n 5g 77 n N Total 406 49 553 569 661 13.0 a/ 1978/79-1981/82 Source: EELPA. Table 3.3: EELPA [CS SALES AND GENERATION FY77-82 Street Light Small Large Intercon- Total Peak System System Domestic Commercial & Agriculture Industry d/ Industry d/ nected Centers Sales Generation Demand Load Factor Losses financial Year (GWh) (GWh) (GWh) (GWh) (GWh) (1Wh) (GWh) S Change (GWh) (MW) (M) (3 Sales) 1977 67.48 32.74 5.87 25.61 177.57 - 309.28 -- 348.15 71.1 55.9 12.6 1978 77.71 32.66 6.17 26.77 175.15 - 318.46 +10.3 361.37 73.9 55.8 13.5 1979 89.23 36.71 11.65 29.46 189.31 1.65 358.00 +12.4 410.71 83.0 56.5 14.8 1980 99.20 36.66 6.46 29.71 219.99 9.85 401.87 *12.3 453.58 87.8 59.0 12.7 1981 107.82 37.05 4.39 30.00 219.36 15.02 -/ 413.64 -0.8 478.44 90.0 60.7 15.7 4.95 2.87 0.54 6.00 0.66 - -- -- -- -- 112.77 39.92 4.93 36.00 220.02 -- 413.64 +2.9 1982 135.36 48.30 9.14 43.71 248.25 b/ 484.76 +17.2 541.00 102,5 60.3 11.6 1982/1981 (*200%) (+21.0%) (+85.4%) (+21.4%) (+12.8%) (+17.2%) (+13.2%) (*13.2%) 48, Imputed Cat.gory Losses '. Sales) 14.6 14.6 14.6 14.6 8.7 -- 11.6 1982 Imputed Category Load 0n Factor (5) 50 55 55 60 70 -- 60.3 (Peak Load on Systems MW) 35.4 11.5 2.1 9.5 44.0 -- -- -- -- 102.5 Average Annual Growth Rate c/ (S) FY77-81 excluding interconnection 12.5 3.7 -- 4.3 6.7 -- 8.5 -- 9.0 -- FY77-82 including interconnection 14.4 7.5 -- 10.7 7.4 __ 9.4 __ 9.4 a! Load Centers connected to ICS from FY79 onwards for FY79-FY82 comprise Assela, Dila, Shashamenie, Yirgalem, Yirga Chefe, Zewar. b/ Including 20.26 GWh boiler sales. c/ computed from least-squares "best fit" curve tor ddta. d/ includes new industrial plants. Source: EELPA and mission anialysis. - 51 - Table 3.4: EELPA SCS SALES, FY77-82 FY SCS Total a/ Interconnection SCS Net (GWh) (CWh) (GWh) 1977 60.90 -- 60.90 1978 68.22 -- 68.22 1979 78.18 1.65 76.53 1980 88.83 9.85 78.98 1981 89.50 15.02 74.55 1982 109.70 16.68 93.02 a/ Before deducting sales to load centers interconnected with the ICS. Source: EELPA. 3.6 Generation and sales data for ERESA for 1977-82 are provided in Table 3.5. The average annual growth rate in sales was 13.1%. Genera- tion was estimated to be 95 GWh in FY82/83, of which 53% was from diesel and fuel oil-fired diesel engine generation sets and 47% from fuel oil- fired steam turbo-alternators. Average system losses are typical of small systems though some improvement should be achievable. The sectoral pattern of sales differs from the EELPA ICS only in respect of the domes- tic sector where artificially low tariffs encourage cooking and water heating with electricity. More economic means of serving both of these demands are available (see para 3.44). About 60% of ICS sales are to industrial consumers and 40% are to residential and commercial consumers. Features of Present Demand 3.7 During the 1960s and early 1970s, the average annual growth in demand on the EELPA ICS was about 12%, but there was virtually zero growth due to the post-revolutionary disruption of the country between 1974 and 1976. The World Bank-financed Finchaa Hydropower Project of 100 MW was commissioned in 1973, yet by 1975/76 demand had not risen as ex- pected. Hence, it was correctly perceived that there would be consider- able excess capacity and energy available at very low cost for some years ahead. To take advantage of this unexpected surpLus, EELPA promoted the use of electricity for both industrial steam raising and household cook- ing. As a first phase of a boiler electrification program, EELPA and the Ministry of Industry negotiated five-year contracts for the supply of power to electric boilers to take effect by 1980 at the latest. In all, there were 14 boiler installations planned with a total installed capa- city of 68 MW, or about 40% of the total demand in 1984. Contracted tariffs ranged from 1.35 ec/kWh to 2.5 ec/kWh (0.65-1.21 USC/kWh). By June of 1983, only two installations had been completed although all boilers had been purchased and several had been installed and were awaiting connection to power supply. Details of energy requirements, - 52 - maximum demand and the current status of boiler installations are pro- vided in Annex 3.1. By 1984, nine boilers were to be connected apnd all but one will be operating by mid-1985. Regrettably, the seven or so year delay in program implementation has meant that most of the benefits originally perceived will not materialize. Not only is peak demand rapidly approaching the power system firm supply capacity, but the long- run cost of electricity is much greater than has ever been considered (see Chapter VI). Therefore, the supply contracted for boiler consump- tion has a much higher opportunity cost than the tariff levels. Further- more, recent GOE policy edicts require EELPA to generate substantial revenues and therefore the very low boiler tariffs will impose a finan- cial burden on EELPA. In economic terms, considering only the energy costs of operation, electric boilers will incur economic costs of about B10 million per year in addition to the cost of fuel oil-fired boilers serving the same demand. Finally, there have been two major technical flaws in the design and procurement of electric boiler systems. First, the requirement for extensive water treatment was largely overlooked as was the need to provide adequate manpower training for the sophisticated management required for high quality water treatment. The equipment for effective water treatment is itself a major expense. Second, the boilers selected are not suitable for base-load operation since they have a very high rate of heat transfer per unit electrode area which makes them high- ly susceptible to scaling, even with low levels of water impurities. In all likelihood the boilers installed will have an economic working life of less than ten years. Table 3.5: ERESA ICS AND SCS: SALES AND GENERATION, FY77-82 ics …-CS-- ----- ------- Combined Year Sales Generation Demand Losses Sales Generation Losses (GWh) (GWh) (MW) (% Sales) b/ (GWh) (GWh) (% Sales) b/ 1977 45.1 49.0 -- 8.6 -- -- -- 1978 15.2 17.0 -- 11.8 4.8 5.6 16.7 1979 52.6 63.7 11 21.1 4.6 6.3 37.0 1980 67.0 76.1 17.3 13.6 4.5 6.3 40.0 1981 76.3 86.8 18.5 13.8 4.3 5.8 35.0 1982 79.2 a/ 92.2 19.7 16.4 4.2 5.6 33.3 a! Estimated by mission. b/ Believed to include considerable non-technical iosses. Source: EELPA. 3.8 For the above reasons, the mission considers that it would be uneconomic to continue the electric boiler program beyond its first phase, and that the only justification for continuing with the first phase is that the equipment already has been purchased and therefore - 53 - constitutes a sunk cost. In view of the impending shortfall of system capacity on the EELPA ICS (Table 3.7), the mission considers that 4ELPA should invoke its contractual right to cease supplying power on conces- sional terms from 1985 onwards, and to charge thereafter an economic price for power supplied to electric boilers. The mission's analysis shows that electricity is a much costlier source of energy than fuel oil in terms of economic cost for boiler operation. Provided that market prices reflect economic costs, there would be a strong incentive to re- store fuel oil in place of electricity as boiler fuel. Existing non- electric boilers should be kept in working condition for back-up opera- tion when electric boilers break down or undergo maintenance and, even- tually, for continuous use. Fortunately, in all plants except one, 11/ the existing non-electric boilers are still in place although expenditure on overhaul is required on some of them. 3.9 Electric cooking also has been heavily promoted inside and out of government as a means both of reducing fuelwood consumption and of utilizing "supposedly" cheap surplus hydropower. The program is focused on the EELPA ICS area, particularly in Addis Ababa. EELPA has contri- buted to this program since the 1970s through the manufacture and sale of electric cookers. EELPA estimates that the number of electric injera cookers in use at the end of 1982 was between 10,000 and 15,000, or about 3-5% of Addis Ababa households. Presently the market is supplied by both EELPA and private manufacturers. EELPA's present production capacity is about 2,000 cookers per year, although in the recent past actual produc- tion has been limited to about 1,000 cookers per year due to constraints on the supply of components. EELPA estimates that private manufacturers have a combined production capacity of between 2,000 and 3,000 cookers per year. Presently, the demand for electric injera cookers appears to exceed this supply capacity at market prices. 12/ EELPA is under direc- tion from the government to produce 10,000 injera cookers per year to combat the fuelwood problem. Whereas up to now the Ministry of Industry appears to be doing very little in manufacturing electric cookers, EELPA is prepared to supply a substantial proportion of the market in order to set quality standards and to prevent private manufacturers from exploit- ing the present imbalance between demand and supply by raising prices to a market clearing level. 3.10 Injera cookers currently available draw a load of about 3 kW on average, and electric cooking pots are rated at 0.50 kW though few are in use. EELPA has developed a more efficient model of injera cooker which draws a load of 2.5 kW. This development will result in a significant 11/ The exception is the United Oil and Soap plant in Addis Ababa. This boiler presently consumes 7,000-10,000 tonnes of fuelwood annually. Proposals for alternative fuels for this plant are given in Chapter V. 12/ Mid-1983 market prices in Addis Ababa: EELPA cooker -- B250; private manufacturers -- between B250 and B350. - 54 - reduction in maximum demand on the power system, and will provide some alleviation of the projected critical period for power system supply before the commissioning of the Malka Wakana hydropower station (see para 3.27). The extent of this alleviation depends upon growth in demand for these cookers, which is expected to be strong for at least the short- term, and the degree to which the use of these cookers can be timed to avoid daily peak demand periods on the power system. 3.11 The dominant market for electric cooking appliances is the middle and upper income group of urban dwellers, which is the only group that can afford the initial cost of an electric cooker. At present prices there is a substantial cost savings from using electricity which is a key factor in the strong demand for electric cookers. In economic terms, the choice between electricity and fuelwood depends upon a variety of factors and, in theory, fuelwood should be the cheaper energy form for injera cooking in the latter half of the 1980s and beyond. 13/ However, electric injera cooking is clean and convenient compared with open-fire injera cooking, which usually generates voluminous smoke from the vola- tile fine woody fuels used. In practice, it is likely that, at least among the upper income group, demand for electric cookers will remain strong (see recommendations regarding electric cooking in Chapter II). Electricity Demand Forecasts 3.12 Power demand forecasts up to 2005 have been prepared by EELPA's consultants 14/ for both the EELPA and ERESA areas. The projections are based on a number of components including basic growth in the economy, new industrial development, industrial boiler electrification, domestic electric cooking appliances, extension of the interconnected system (ICS) and rural electrification. They explicity assumed that tariff levels would not be changed in real terms. The consultants concluded that these factors would result in a significant rate of increase in power demand, especially during the period 1982-87. EELPA ICS Forecast Demand 3.13 Average forecast growth rates by the consultants for demand on the EELPA ICS are given as follows: 13/ The economic cost of energy for cooking at projected conversion efficiencies is: electricity -- 7.4 ec/MJ; fuelwood -- minimum 2.7 to 4.7 ec/MJ depending on cooking type (Table 2.11). 14/ Acres International Ltd., Canada. - 55 FY82-87 FY87-92 FY92-97 FY97-2002 Average Forecast Growth Rate (%/a) 19.2 8.6 6.3 5.6 The contribution of the demand components to the overall projected growth on the EELPA ICS is given in Table 3.6. Table 3.6: EELPA-ICS: COMPONENTS OF CONSULTANTS' FORECAST DEMAND a/ (GWh) FY80 FY82 FY87 FY92 FY97 FY2002 Demand Category (actual) --------------- (projected) ------------- Domestic-General 119 Commerce b/ 51 Small Industry 35 Subtotal 205 229 321 450 632 886 Domestic Electric Injera Cookers and Pots c/ 5 66 131 183 257 Large Industry Existing General 249 277 337 409 498 606 Boilers 0 59 342 582 821 1061 New Plants d/ 16 177 245 292 349 New Interconnected Load Centers 25 223 402 586 797 Totai Demand 454 610 1466 2219 3012 3956 a/ demand at generation stations. b/ including street lighting and agricultural demands. c/ included in Domestic General category for 1980. d/ Included in Industrial Existing category for 1980. Source: Acres "Power Planning Studyt (October 1982). 3.14 The consultants' standard forecast is for 14.9% p.a. growth in capacity demand and 16.7% a year growth in energy demand between 1982 and 1989. The underlying growth rate in demand throughout the forecast peri- od is projected to be 7% a year for domestic, commerce and small indus- tries, and 7% a year for existing large industries. The abnormally high growth rates in the short- to medium-term are the result of demands from the special categories, namely industrial boiler electrification, new industrial loads, injera cookers and new load centers connected to the ICS. Both the consultants and the mission assumed that in the short term - 56 - domestic demand would be inelastic. According to the consultants' fore- cast, demand will overtake available system power capability on the EELPA ICS in early 1984, and will exceed firm system energy generating capabi- lity in mid 1985, as demonstrated in Table 3.7. 3.15 Actual sales for the past two years on the EELPA ICS have been consistently below the consultant's forecast sales for this period, as shown below: FY80 FY81 FY82 MW CWh MW CWh MW CWh Consultant's Forecast Demand a/ 88 b/ 454 b/ 104 499 128 610 Actual Demand 88 454 90 478 103 541 Ratio Actual to Forecast 1.00 1.00 0.87 0.96 0.80 0.89 a/ at generation station. b/ actual demand for FY1980. Source: EELPA 3.16 Furthermore, the mission considers that the consultant's fore- cast growth rates for demand for the EELPA ICS are unrealistically high, especially in the short-term. Therefore, the mission has prepared its own demand forecast that is lower than the consultant's by virtue of a lower base year demand (FY1982) and lower forecast growth rates. The rationale for this view is given in the following paragraphs which are then aggregated into the mission's forecast for total demand. 3.17 Future Electric Boiler and Cooker Demands. It is assumed here that the mission's recommendation to discontinue the electric boiler pro- gram will be accepted and implemented. This has a major effect on future demand, for in the present demand forecast boilers are an expanding load through 2000. In addition, in the early years the maximum supply con- tracted (and included in the consultant's forecast) overestimated the actual consumption. The mission's projected demand from electric boilers is therefore as follows: FY84 FY85 FY86 FY87 Beyond FY87 Maximum Demand a/ 41 56 56 20 0 Energy Consumption (GWh/year) 137 179 179 80 0 a/ Unit coincidence factor for boiler maximum demands are assumed for these values and therefore this demand would be less according to the extent that the factor is less than unity. Table 3.7: EELPA ICS: ACRES AND MISSION FORECASTS OF DEMAND AND GENERATION CAPABILITY, FY83-95 - Mission Forecast ----- -----Acres Forecast----- Generating System Available System System FY Capability Reserve a/ Capacity Demand Balance Demand Balance (GWh) (MW) (MW) (GWh) (MW) (GWh) (MW) (GWh) (MW) (CWh) (MW) (GWh) (MW) 1983 1052 204 33 1052 171 618 126 434 45 732 152 320 19 84 1052 204 33 1052 171 735 164 317 7 920 192 132 -21 85 1052 204 33 1052 171 862 200 190 -19 e/ 1126 233 -74 -62 86 1052 204 33 1052 171 877 213 d/ 175 -42 d/ 1329 280 -277 -109 87 1212 b/ 264 b/ 33 1212 231 934 196 278 35 1466 311 -254 -80 88 1760 (348.5/248.5/381.5) c/ 35 1760 313.5 1052 213 708 100.5 1638 347 122 33.5 89 1760 381.5 40 1760 341.5 1145 232 615 109.5 1795 380 -35 -38.5 '4 90 1760 381.5 40 1760 341.5 1241 250 519 71.5 1932 408 -172 -66.5 91 1760 381.5 40 1760 341.5 1342 270 418 71.5 2073 438 -313 -96.5 92 1760 381.5 40 1760 341.5 1453 292 307 49.5 2219 469 -459 -127.5 93 1760 381.5 40 1760 341.5 1566 315 194 26.5 2369 501 -609 -159.5 94 1760 381.5 40 1760 341.5 1687 340 73 1.5 2523 533 -763 -191.5 95 1760 381.5 40 1760 341.5 1791 362 -31 -20.5 2681 567 -921 -225.5 a/ Criteria is greater of 10% capability or largest single unit. b/ Increase in capability (a) R. Amarti diversion + 176GWh/y from mid FY87 (b) Melka Wakana: 5 X 30MW units +560 GWh/y commissioned at 3 month intervals from mid FY87; 5,5MW/28GWh pa thermal plant retired end FY87. / Finctiaa extension (+33MW)-- penstock connection carried out during last stage of work in mid FY1988 once Malka Wakana is fully commissioned when there is sufficient surplus to close down Finchaa temporarily without incurring power shortages. d/ Maximum demand could be lower--depends on rate of disconnection of boilers. e/ Power bhorlageb could be avoided by careful maintenance scheduling. Source: Mission estimales. - 58 - 3.18 For electric cookers the mission estimates that the size of the present potential market for electric injera cookers is about 50,000 households, corresponding in number roughly to those families in the main EELPA ICS supply area with monthly incomes exceeding B300 (1980 Birr) with a similar number of electric cooking pots. It is assumed that the supply of cookers will meet demand and that the electric injera unit demand increasingly will be supplied by 2.5 kW units instead of 3 kW units. The promising aluminum mtads the mission has tested and is pro- moting for further trial by EELPA (see Table 2.9) offer the possibility of further reductions in load. 3.19 EELPA's consultants estimated the probability distribution of cooking load on the power system from a general observation of preferred cooking times, from which they concluded that the contribution to the EELPA ICS peak demand from electric cooking appliances could be derived by using a system coincidence factor of 3.2% with the evening peak, which occurs at about 7:00 p.m. Most households only use their injera cookers during daylight hours to avoid overloading the local distribution supply equipment after dark by superimposing the cooking load on the lighting load. However, the daily load curve for the EELPA ICS has two weekday peak periods, with the higher peak occuring during early evening and a slightly lower peak occuring around midday, so that present cooking schedules will tend to coincide more with the midday peak demand period than with the evening peak. Growth in the use of electric cooking appLi- ances will therefore tend to reduce the difference between the peaks and, possibly, reverse the ranking. In the absence of survey data on house- hold cooking patterns, the mission considers 15% a more reasonable coin- cidence factor for electric cookers with system daily peak load period. The mission's projections for demand from electric cooking appliances, including maximum demands coincident with system peak periods, are given in Annex 3.2. Based on this survey, household power consumption for lighting and other appliances over and above cooking is expected to grow at about 7% a year. 3.20 Future Industrial and CommerciaL Demand. Within the manufac- turing sector of industry, cement and textiles are responsible for 80% of new large industrial power demands of about 118 GWh per year by 1987. However, there is a substantial demand arising from the new sugar indus- try at Finchaa for which the opportunity for self-sufficiency through the year-round use of its potential seasonal bagasse surplus has been ne- glected. There will now be a load on the public power system of about 1 MW, with instantaneous back-up requirements of 4 MW, during the cane crushing season and about 4 MW peak outside of it for irrigation, hous- ing, and workshops in the first phase of development, occasioning an annual demand of about 20 GWh unless the plant design and the pattern of demand internal to the industry (e.g., homes and villages) are changed. Both the design of the sugar mill and the associated housing neglect ready opportunities for improving energy efficiency. Apart from the greater cost of operation implied, the Finchaa power demand arises at just the time that capacity constraints on the EELPA ICS are most criti- cal. The mission therefore recommends an urgent design review with the objective of maximizing the supplies of bagasse and achieving year-round - 59 - self-sufficiency and complete autonomy from EELPA. Even including the Finchaa project, a load demand growth of only 4% p.a. is estimated for existing industry due to an already high level of plant capacity utiliza- tion. Growth in the commercial sector is less restricted with new hotels, office blocks and general urban infrastructure expansion expected to create a demand growth of 7% p.a., slightly in excess of the projected urban population growth rate of 6.5% p.a. 3.21 Interconnection. Throughout the 1970s there were no major extensions to the EELPA ICS. However, EELPA connected six isolated load centers to the ICS between 1979 and 1982, and EELPA now has an ambitious program for extending the EELPA ICS to new load centers in the SCS and for village electrification. The rationale for this program is the need to reduce consumption of imported oil, to alleviate fuelwood shortages for households, and to promote development away from conurbations by providing power for industries, commerce and community centers. The main ICS transmission network consists of a radial system centered on Addis Ababa, based on 230, 130 and 66 kV lines. A regional 66 kV network is presently being constructed in the Lake Tana area to serve load centers with hydropower from the Tis Abai station. The major transmission proj- ects being planned are a 230-kV line between the Finchaa power station and Bahir Dar, and a 230-kV line between Melkassa and Dire Dawa. EELPA has a long-term objective of interconnecting the EELPA and ERESA ICSs, and the two 230-kV projects can be viewed as components of an eventual national transmission system. The economic justification for extending the transmission system has to be carried out case by case. However, the ultimate constraint on the rate of extension will probably be availabil- ity of resources. The mission has adopted a lower rate of investment in transmission extension than planned by EELPA that is in line with lower and more realistic estimates of likely resource availability. Neverthe- less, the mission forecast shows an incremental load (excluding system losses) from interconnected centers on the ICS of 20 MW/83 GWh/year in 1990 and 46 MW/204 GWh/year in 1995. Mission's Demand Forecast 3.22 By aggregating the estimated growth projections for each of the key determinants of future power demand, the mission's revised forecast is for energy and power demand to grow at 8.5% p.a. as opposed to the 15- 17% growth rate projected by EELPA and consultants. Nevetheless, under the mission's forecast, per capita power consumption will be about 40 kWh per year by 1992. The comparison made between these data in Table 3.7 indicates that under the mission's forecast the capacity constraint is much less severe and appears first in 1985, two years later than under the consultant's forecast, and that no energy deficit arises prior to 1987 providing there is no substantial delay in commissioning the Malka Wakana scheme. Furthermore, the level of the capacity deficit in 1985 and 1986 is below the capacity of electric boilers installed by that time and will disappear if the government adopts the mission's firm recommen- dation to relegate electric boilers to back-up status. Once again, the importance of maintaining and, where necessary, rehabilitating the exist- ing oil and wood-fired boiler capacity is emphasized. - 60 - Medium-Term Supply Options 3.23 There are only two projects to augment power supply in the medium term that have been funded and are currently under construction. These projects are the US$29 million Amarti-Finchaa diversion, providing an additional 176 GWh a year through the existing Finchaa Power Station, and the Soviet-funded and contracted Malka Wakana Hydropower Plant of 152 MW and 560 GWh average annual energy production, located 250 km south of Addis Ababa on the Wabe Shebelle River. The Amarti-Finchaa diversion is scheduled to be completed by mid-1987 and the Malka Wakana Station also by mid-1987 (first of four 38 MW units). As capacity availability is the earliest and most severe constraint to power supply, the consultant's demand forecast suggests that investments in new capacity are required immediately to alleviate the impending shortfall. EELPA and its consul- tants have foreseen this capacity being avaiLable from two main sources: the now unused Aba Samuel hydropower plant and the addition of a fourth 34 MW unit in the Finchaa power house for which some provision was made at the time of construction. 3.24 The Aba Samuel scheme could be rehabilitated as a 6 MW plant for about B13 million by mid-1985, or it could be upgraded to up to 36 MW at a cost of about B56 million for completion late in 1986 at the earli- est. In the former case the benefit is minimal and would make little difference to system capacity constraints. In the latter case, the plant only could be used for peak period supply due to limitations on storage capacity; consultants 15/ to EELPA also have expressed reservations about the feasibility of obtaining as much as 36 MW output since the high dis- charge rates involved in rapid loading could cause hydraulic problems. The consultants consider that such problems could be avoided only with a lower installed capacity of about 21 MW. 3.25 The addition of a fourth turbo-generating unit to Finchaa Power Station also was originally proposed as a means of alleviating power shortages before commissioning the Malka Wakana Station. However, the station would have to be shut down completely for at least a month so that a wye junction could be fitted at the bottom of the penstock to pro- vide a feeder to the new unit. Such a shutdown would involve removing about 50% of system capacity (100 MW out of 204 MW) and thus would create massive power shortages with an unacceptable degree of disruption to the national economy. In fact, the earliest that the new unit could be com- missioned is presently in early 1987, and therefore it would not allevi- ate the situation before the scheduled commissioning date of Malka Wakana. However, the proposal does have merit in the context of longer- term planning as a means of rectifying the present imbalance between capacity and energy constraints, which requires new capacity to be added to the system about 1-1/2 years in advance of the need for additional 15/ Lahmeyer International "Review Report on Feasibility Study for Aba Samuel HPP Rehabilitation/Redevelopment" (May 1983). - 61 - energy capability. Rectification of this balance would thus allow investment in the next major hydropower scheme to be postponed by 1-1/2 years. Furthermore, immediately after the commissioning of Malka Wakana, there will be sufficient surplus in system capability over system demand to allow a shutdown of the Finchaa station to occur without precipitating power shortages. The mission recommends that EELPA investigate the economic justification for installing the fourth unit at Finchaa station in such a manner. 3.26 The Malka Wakana project is the key to short- and medium-term planning for the EELPA ICS. At full commissioning, the project will add 75% additional capacity and 55% additional energy to the existing system hydro capability. According to the mission's demand forecast, the addi- tion of this relatively large increment will meet the increase in demand for six years. However, as with any major hydroelectric project, there are risks of serious delays in completing the project. EELPA is not involved in programming project implementation; this is being handled by a high leveL Ethiopian Committee appointed by the Government. 16/ 3.27 There are a number of options for covering the risk of a seri- ous delay to the Melka Wakana Project. EELPA have favored proceeding with preparation of the Gibe Daka project (para 3.33) to minimize the im- plementation period if a decision to proceed quickly was required. This strategy only makes sense if there is a serious possibility that the Malka Wakana project will not be completed, and at this stage there is no justification for such a view. Furthermore, the Gibe Daka project does not form part of the least-cost program, and a hasty commitment to the project runs the reverse risk of simultaneous investment in two large projects when only one is required. For example, if both the Malka Wakana and Gibe Daka projects were to be commissioned around 1990, this would add 300 MW/1300 GWh per year to system capability, equivalent to additions of 150% and 130% to system capacity and energy. In the mis- sion's view, this outcome would constitute a massive uneconomic in- vestment. 3.28 A strategy is required which involves a minimum of additional investment during the next few years over that for the Malka Wakana pro- ject and which also is consistent with long-term planning. The mission recommends that such a strategy be prepared as a matter of urgency and envisages that it will be based on demand management (paras 2.29 to 3.32) as a means of coping with a delay in the commissioning of Malka Wakana of up to one year, and on other options for increasing supply capability for a longer delay. One option is the construction of Tis Abai II power sta- tion (30 MW) at the same time as the proposed 230 kV line between Finchaa and Bahir Dar to interconnect this station with the EELPA ICS. Another option is interconnection of the Ethiopian and Sudanese main systems 16/ This arrangement reflects GOE existing policy of funding major capital projects from the central government resources and handing over the assets to the operating utility on completion. - 62 - through a transmission linkage between Finchaa and Rosieres. This option has the long-term (post 1990) attraction of taking advantage of surplus Ethiopian hydro power (after commissioning of the Malka Wakana project and subsequently other projects) to displace expensive thermal power in the Blue Nile grid in and around Khartoum and to take advantage of non- coincidence of system peak load periods. In the medium-term (to 1990), 340 MW (300 MW thermal, 40 MW hydro) are scheduled to be installed in the Blue Nile Grid based on the need to meet system demand during the season of low Blue Nile flow (March-May) in a proiected dry year. Thus, there will be considerable underutilized thermal generating capacity in the system. Advancing the timing of interconnection of the two systems to 1988 would provide the Ethiopian system with back-up capacity to cover the risk of delay to Malka Wakana. The mission recommends that the investigation of both options be given top priority. Demand Management Options 3.29 The conclusion of the preceeding paragraphs is that a vital option realistically available to alleviate short-term power and energy constraints is demand management. Electric boilers will continue to have only short-term application and electric cooking will be accessible just to middle and upper income groups, nevertheless making a significant contribution to household energy supply. Demand management measures which are justified are (a) achieving greater efficiency of electric cookers; (b) correcting power factor distortions; (c) reducing trans- mission and distribution losses; and (d) conducting a thorough review of the potential for solar water heating. The first of these measures is implicit in the mission's revised demand forecast which assumes a reduc- tion from 3 kW to at most 2.5 kW maximum demand for each electric injera cooker. 3.30 The mission's tests of aluminum "mtads" show that energy savings of 35% are achievable over clay "mtads" due to greater plate con- ductivity. The mission has received advice from cookware manufacturers that a non-stick aluminum "mtad" without electric fittings could be pro- duced in bulk in the US for US$15-17. A sealed unit with aluminum "mtad," resistance coils, thermostat and switchgear would cost at the most BlO0 and would yield a reduction in load of at Least 0.5 kW. The cost of saving 1 kW in peak system demand by this means would amount to about B1,335 at a coincidence factor for cooking demand with system peak demand of 0.15 and if the whole cost of the unit was written off against this saving. In fact, the incremental cost of a unit with an aluminum "mtad" over a unit with a clay "mtad" would be much less than B100 and thus, the cost of saving one kW in peak system demand would be much less than B1,335. In contrast, the investment required to install one kW of hydroelectric power is about B2,000. The mission recommends that EELPA pursue the development of non-stick aluminum "mtads" with demonstration units placed in households and by adopting this design for units manufac- tured by itself. EELPA also should consider making available sealed aluminum "mtads," possibly initially at a subsidized price so they are competitive with existing units, to private manufacturers of cookers to achieve an early and major impact on the average demand and efficiency of electric cookers. - 63 - 3.31 Power factors of 0.6-0.8 are common in large Ethiopian indus- trial establishments and are as low as 0.5 in small industrial plants due to the proportion of the plant load comprised by motors. Such low fac- tors cause gross underutilization of existing power system capacity and losses of power throughout the transmission and distribution system. While power factor correction would not directly benefit the industries concerned unless the tariff structure provided incentives, it would ease EELPA's capacity constraint. The mission estimates that 2 to 3 MVA could be saved by installing power factor correction equipment and by enforcing strict financial penalties for individual plant power factors below 0.95. The data in Table 3.3 and 3.4 indicate that losses, defined as the dif- ference between power generation and power sales, are substantial and, in the case of ERESA, quite serious. The relative contributions of tech- nical inefficiency and failures in consumer billing are not known, though it should be possible to contain technical losses to 10% of generation. Hence, a detailed review of the causes of inefficiency is recommended, together with a program of action to implement improvements, focusing particularly on the short-term impact, such as power factor correction. 3.32 Solar water heating is likely to be economic in all hotels, hostels and hospitals now using electric water heating. Payback periods of four to six years are likely in economic terms in industrial and com- mercial applications in the EELPA system. A major opportunity exists to install solar water heaters in several thousand high quality homes in Addis Ababa. Retrofitting of adequate solar water heating to existing electric storage heaters appears to be feasible with paybacks of four to five years. The mission recommends that a review be made of the benefits of installing these systems for households and of recovering the instal- lation costs through monthly billing over an extended period. Appropri- ate upward revision of tariffs (para 6.11) should provide the financial incentive to introduce these installations. Long-Term Supply Options Hydroelectric Power 3.33 According to the EELPA-Acres standard forecast, the next major hydroelectric scheme following Malka Wakana is required by late 1989 (Table 3.7). According to the mission's forecast, this scheme would not be required until 1994. The difference in timing is significant for planning options. Planning is sufficiently advanced at present on only one scheme at Gibe Daka (150 MW/735 GWh/y), to achieve a commissioning date of 1990. The detailed design of the Gibe Daka scheme, at an esti- mated unit cost of US$1260/kW including associated transmission, is now underway with aid funding from the Italian Government. However, the Gibe Daka scheme is not the least-cost option identified by EELPA's consul- tants. 17/ The most promising hydropower schemes are available in the Aleltu, Upper Beles and Halele Werabesa River basins, and through the 17/ Acres "Power Planning Study," October 1982. - 64 - Chemoga Yeda diversion scheme. These options need to be compared on an equal basis with the currently proposed Gibe Daka (formerly Gilgel.Gibe) scheme. The estimated capacity and average energy production from these schemes are given in Table 3.8. The most promising of these alternatives are those whose development can take place in stages with increments that are compatible with the growth in power demand. For example, the hydro potential of the Aleltu River system could be developed with four units ranging from 120 MW to 190 MW, and the Chemoga Yeda diversion in two stages of 210 MW and 100 MW each. Preliminary estimates for the develop- ment of these cascade-type resources are US$900-1100/kW including associ- ated transmission. The mission estimates that the long-run marginal cost of power from an optimized development program is about 16ec/kWh sold.18/ Pre-feasibility survey work is required at both these sites to confirm the feasibility of these options. Under the consultant's demand forecast there is insufficient time to carry out such work, whereas under the mission's forecast there is adequate time. The mission recommends that survey work for at least one and preferably both of these options be started as soon as possible to develop a range of planning options, to be followed by feasibility studies. 3.34 Beyond the early 1990s, Ethiopia has a wealth of options for developing domestic resources for power production. EELPA's consultants consider that the total hydroelectric potential of the country is in the range of 15,000 to 30,000 MW. Although a proportion of this potential would be uneconomic to develop, nevertheless there exists potential of many times the present level of demand and that could be developed rela- tively cheaply by world standards. The topography of the region permits good regulation of hydro resources to be obtained at low cost. The hydro projects described in the preceeding paragraph are the most attractive of many options presently identified. Ethiopia therefore possesses poten- tial greatly in excess of foreseeable domestic requirements and is also surrounded by countries which are relatively poorly endowed with energy resources. The scope for developing power potential for export is an attractive long-term possibility which should be explored at political and technical levels. Export Options 3.35 As noted above, Ethiopia has extensive low cost indigenous hydropower potential, and is much more favorably endowed than its neigh- bouring countries who are having to increase their importation of hydro- carbons to support development of their power systems. The most promis- ing opportunities for realizing mutually beneficial developments are with Sudan and Djibouti. 3.36 The interconnection of the EELPA ICS with the Blue Nile Grid is one option for development with Sudan that already has been discussed in 18/ The mission has estimated the average incremental cost of power as a measure of the long run marginal cost. - 65 - connection with a strategy for covering the risk of delay in commission- ing the Malka Wakana project (para. 3.28). The option capitalizes on the sunk investment in transmission capacity between Roseires and Khartoum and the low output sent from Roseires to Khartoum during the Blue Nile River dry season (at least six months per year), so that exports of power from Ethiopia would be predominantly during this season. EELPA's consul- tants (Acres) estimated the cost of the transmission link between Finchaa and Roseires to be about US$66 million (1982 prices). Given the mutual benefits to Sudan and Ethiopia discussed in para. 3.28, the mission recommends this option be investigated as a matter of priority. Table 3.8: PRODUCTION CAPABILITY OF POTENTIAL HYDROPOWER PROJECTS Installed Average Capacity Energy Project (MW) (GWh/y) 1. Malka Wakana 150 560 2. Gibe Daka 150-300 735 3. Chemoga Yeda Stage I 210 935 Stage II 100 420 Total Chemoga Yeda I & II 310 1355 4. Aleltu Stage I 120 525 Stage II 190 830 Stage III 160 700 Stage IV 140 610 610 2665 Stages V & VI 160 725 Total Aleltu 770 3390 5. Upper Beles 245 1100 6. Halele Werabesa I 55 245 II 300 1180 Total H. Werabesa 355 1425 Source: Acres: "Power Planning Study" (October 1982). 3.37 The development of hydro potential in Ethiopia for exporting firm power to Sudan is a long term option. Under this scenario, the - 66 - option of developing a major hydro scheme in the Lake Tana region becomes attractive as part of a strategy to supply hydropower to the northern regions of Ethiopia (ERESA systems), since the transmission distance to the northern region would be substantially less than would be the case with development of a site north of Addis Ababa (Aleltu, Chemoga). Eco- nomies of scale also could be obtained from developing a hydro site if the development were sized to supply a substantial load in Sudan as well as the relatively small load of the ERESA ICS. Suitable sites have been identified on the Upper Beles River (245 MW/1100 GWh per year) and the Rahad River (585 MW/2565 GWh per year). 3.38 Interconnection with Djibouti could be realized by either a transmission line from Dire Dawa or by a line from Kombolcha to serve both Djibouti and Assab. Acres judged the extension from Kombolcha to be a better option. On the basis of supplying 105 GWh/year of off-peak energy to Djibouti by 1995, Acres concluded that the value of fuel sav- ings to Djibouti would be about three times the annualized cost of trans- mission and generation supply from Ethiopia. Therefore, the mission recommends that this option be considered seriously once the availability of hydro energy in Ethiopia becomes assured. Natural Gas 3.39 Natural gas must be seriously considered as a power source for the 1990s and for early next century. Very preliminary analysis by the mission indicates an economic cost of about 15eclkWh sold for the known Ogaden Desert gas resource, which is comparable to the long-run cost of power from hydroelectric resources. 19/ This estimate is derived by weighing the entire cost of additional field development, an 800 km pipe- line, and a gas turbine farm in Addis Ababa against the incremental market for power from 1994 onwards. The incremental usable power pro- duced in the early years of a gas project would be small and some of the gas delivery capacity could be used to supply the substantial market for gas in industry, fertilizer production and transportation in the Addis Ababa area. Some of the gas delivery costs could thus be allocated against, non-power uses. The required steps in resource evaluation to enable a fair comparison to be made between natural gas and hydropower are outlined in Chapter IV. The mission recommends that the required program of work, including a preliminary gas market analysis and con- firmatory drilling, be started as quickly as possible to fit in with the timing of further investigations of major hydropower resources in the process of determining a least cost power development plan. Geothermal Power 3.40 The present work program of geothermal exploration is described in Chapter V. This Rift Valley resource is part of the same system that is being tapped in Kenya with technical and financial assistance from the 19/ Mission estimate of 16 ec/kWh. - 67 - Bank. Geothermal power is often developed in unit sizes ranging from 15- 30 MW with units in commercial operation of up to 55 MW. At the time of the mission two dry holes had been drilled in the Lakes area south of Addis Ababa. Shortly after the mission's fieldwork potentially commer- cial steam flows were tapped. The economic cost of developing geothermal power, based on the Bank's experience, is generally in the range of US$2,000-2,500 per kW installed, which is about double the estimated cost for prospective hydropower in Ethiopia for the equivalent firm capaci- ty. Preliminary analysis suggests special circumstances would have to apply for there to be a significant role for geothermal in the EELPA ICS generation expansion program, such as the relative ease of financing smaller increments to available installed capacity, or the expedience of "infilling" between large and lumpy hydropower supply increments. However, there are many factors which influence the delivered cost of geothermal power, including steam quality and maintenance costs, drilling costs and proximity to the grid and, depending on local circumstances, the cost per kWh delivered could easily range across present estimates of the long run marginal cost of power for the EELPA ICS. The GOE, with the assistance of the EEC and UNDP, is committed to explore geothermal re- sources and to undertake a feasibility study on the viability of a 30-MW geothermal power plant near Langano in the Lakes district where drilling activities are in progress, and in Corbetti, near Awasa, and Abaya, where detailed technical studies are underway. The mission believes that the prospects for producing geothermal power competitively with hydropower or natural gas are yet to be demonstrated, and endorses the Government's endeavors to undertake a detailed evaluation of the economic cost of geo- thermal power production. The mission also recommends that the economic viability of this power source be evaluated in the context of the least cost power system development for the EELPA ICS. Similarly, it is appro- priate to use the above economic analyses as the basis for determining the likely prospects for economically producing power from other poten- tial geothermal resources in order to determine the priority for allocat- ing resources available for further geothermal or other energy resource exploration, such as in the north where further evaluation of geothermal resources in the Tendamo and Eritrea region are required to determine the prospect of supplying power to the ERESA system in the medium- to long- term. EELPA SCS Forecast Sales 3.41 Forecast sales on the EELPA SCS to 1995 are given in Table 3.9. The underlying growth rates derived by Acres are 10%/year for 1982- 87, 8.5% a year for 1987-92, and 7% a year for 1992 onwards. The remain- ing load on the SCS after interconnection with the ICS is forecast to de- cline slowly. By 1995, nearly 70% of the load from the present day SCS is forecast to be interconnected with the ICS, which represents a sub- stantial impact on efforts to reduce oil-based power generation. ERESA ICS and SCS Forecast Demand 3.42 Forecast sales and demand for the ERESA ICS and SCS are provided in Table 3.10. Sales on the ICS are projected by EELPA's - 68 - consultants to grow at an average annual rate of 12.4% a year for the 1982-95 period. Peak demand is projected to roughly treble during this period, implying an aggressive program of generation expansion. Growth of sales in the ERESA SCS is expected to be 20% per year in the short- term, mainly due to taking up suppressed demand. Table 3.9: EELPA SCS FORECAST SALES, FY83-95 (GWh) Interconnection FY FY82 SC to ICS SCS Net 1982 93 -- 93 1983 102 -- 102 1984 113 -- 113 1985 124 9 115 1986 136 19 117 1987 150 32 118 1988 163 46 117 1989 179 63 116 1990 197 83 114 1991 216 106 110 1992 238 133 105 1993 255 161 94 1994 272 190 82 1995 292 204 88 Source: Mission estimates. Future Supply and Demand Management Options for ERESA 3.43 According to EELPA's present investment program, a transmission link between the EELPA ICS and ERESA ICS is planned for the early 1990s. The mission considers that the interconnection of the two ICSs would not be an economic undertaking in this time-scale, although in the longer term it may be desirable. Without a firm, proven geothermal resource, the range of options is narrowed in the medium-term to expanded use of heavy fuel oil in slow- to medium- speed diesel engines and to imported coal. In the Bank's recent experience, imported coal is not a competi- tive primary energy source with imported heavy fuel oil for generation unit sizes of less than 20 or 30 MW. For the foreseeable future, capa- city additions to the ERESA ICS should be in 5 to 10 MW units for optimum generation efficiency and security, and therefore heavy fuel oil engines will be the least-cost source of power. 3.44 The installed capacity by type and Location of engine for the ERESA ICS is provided in Annex 3.3. The total effective capacity is - 69 - 41.2 MW, of which 15 MW is in steam plants and 26.2 MW is in diesel gen- erators, and there is a further 10 MW yet to be commissioned in 5 MW diesel engines that will use heavy fuel oil. Diesel-fired engines pro- duced 46 GWh or 51% of total generation in 1981/82. Replacement of the steam plant by diesel generators is justified from fuel savings alone. A complete conversion to fuel oil-fired diesel engines would require investment in a further 15 MW (3 X 5 MW) of diesel capacity, plus the necessary fuel oil heating and cleaning equipment to convert the existing 5-MW diesel engines into an entirely fuel oil-fired mode. The mission estimates that the conversion would effectively replace about 12.8 mil- lion liters of diesel with 3.1 million liters of fuel oil after allowing for much improved efficiency in fuel oil consumption. 20/ Table 3.10: ERESA ICS AND SCS: FORECAST SALES AND DEMAND 2c I CS ----------- SCS Genera- Combined Peak Genera- Combined Peak FY Sales a/ tion b/ Demand b/ Sales c/ tion d/ Demand d/ (GWh) (GWh) (MW) (GWh) (GWh) (MW) 1983 85.8 98.7 20.1 5.0 6.5 2.1 84 92.9 106.9 21.8 6.0 7.8 2.5 85 100.6 115.8 23.6 7.2 9.3 3.0 86 108.9 125.4 25.5 8.6 '1.2 3.6 87 118.0 135.8 27.7 iO.a 13.4 4,3 88 137.5 158.2 32.3 89 160.2 184.3 37.6 Subsequent demand on SCS from existing 90 186.6 214.7 43.8 SCS load centers wiii decline due to 91 217.4 250.2 51.0 interconnection; however, assuming 92 253.2 291.4 59.4 that diesel sets are moved to new load 93 275.2 316.8 64.6 centers, the total SCS demand will 94 299.2 344.4 70.2 remain at about a constant level. 95 325.2 374.3 76.3 a/ Acres projected growth (including interconnections) for [CS -- 1982-87: 8.3%/y; 1987-92: 16.5%/y; 1992-97: 8.7%/y. b/ Assuming system losses of 15% on sales; 56% SLF on ICS. c/ Acres projected growth 1982-87: 19.8%/y (no interconnections). d/ Assuming system losses of 30% on sales; 35% SLF on SCS. Source: Mission estimates. 20/ In effect, with an increase from 20% to 35% conversion efficiency for fuel oil, present fuel oil consumption drops from 19,870 to 11,420 te. At 35% efficiency, 12,800 liters of diesel are replaced by 11,350 te of fuel oil. Thus, 22,770 te fuel oil replaces 19,870 te fuel oiL plus 12,800 liters of diesel in the present power system. - 70 - For a capital outlay of about B16 million, economic cost annual savings of B6 million will result, giving a simple payback period of less than three years. The mission believes that heavy fuel oil diesel engine generation is the least cost option for the ERESA system and strongly recommends that EELPA review this conclusion by more detailed analysis. Demand Management in the ERESA ICS 3.45 The marginal cost of producing power in the ERESA ICS is esti- mated by consultants to EELPA 21/ to be 36 and 55 ec/kWh (17-27 USC/kWh) sold off-peak and peak for low voltage supply, respectively and similar- ly, 34 and 46 ec/kWh for high voltage industrial supply. Thus, the cost of meeting increases in demand are high and make the need for demand man- agement a high priority. The mission has identified projects described in the following paragraphs both for electricity conservation and for substitution of electricity within the ERESA ICS by the cheaper energy sources. The mission recommends that the efficiency of supply and use of power in the ERESA area be specificalLy examined to identify measures to minimize the cost of energy utilization. 3.46 Domestic Loads. In the ERESA ICS supply zone, upper income households commonly cook and heat water with electricity. The cost of cooking with electricity is more than twice the cost of cooking with kerosene (see Table 2.10). The mission recommends (Chapter II) that households in Asmara and other major cities of the north now using elec- tricity be strongly encouraged, by appropriate power tariffs and public- ity, to switch to kerosene for cooking. Some 8,500 homes have electric water heaters installed. It will be possible to supply hot water needs with solar collectors for about one-third of the economic cost of elec- trical heating. ERESA should participate actively in this program by financing the installation of solar panels and arranging for repayment on commercial terms by monthly billing. With payments spread over a decade, consumers would experience a substantial reduction in the financial costs of water heating at the present tariff and a much greater reduction if electricity tariffs are raised to reflect the cost of supply, which again underlines the importance of charging for electricity at the full eco- nomic cost of production. Electrical backup should be neither necessary nor permitted. 3.47 Commercial and Industrial Loads. In the ERESA area, trans- mission losses are high and can be reduced, perhaps to about 10%. A pro- gram of loss analysis and reduction should be mounted in combination with power factor correction enforcement for large industrial loads. The use of high efficiency sodium vapor lamps should be considered for street and other public lighting. Inappropriate and inefficient use of electricity has been observed in large industrial plants such as in glass manufacture where electricity is used for a portion of the required heating. 21/ Coopers & Lybrand Associates: "Tariff Study and Asset Revaluation" Final Report, August 1983. - 71 - Detailed energy audits and least-cost analyses should be undertaken for each of the major power consumers in the ERESA system. Small-scale Alternative Power Sources 3.48 Many isolated load centers are too small and/or too distant from the nearest part of the ICS for interconnection to be economically justified. There are well-defined opportunities for using mini-hydro- power resources and some small prospect of wood-fired power generation which may economically displace diesel-fired generation. Feasibility studies for two small hydropower plants on the Hoha and Sor Rivers have been completed by a UNDP team under a program for technical assistance. The Sor River scheme is 4 MW, with an estimated final cost including transmission of B26 million, and it would displace diesel generation at the towns of Mettu, Gambella and Dembidollo. The Hoha River scheme is 650 kW, with an estimated cost of B2.54 million, and it would replace diesel power at the town of Asosa. Both schemes would supply power for about 28 USc/kWh based on the load forecast adopted. The economic rates of return computed are 12-16% and are critically dependant on the demand forecasts which must consequently be verified before commitment is made to implement. EELPA have included seven other mini-hydro schemes in their investment program. 3.49 Wood-fired power options are confined to state farms both in the forest to the southeast of Jimma at Bubeka and in the savannah land grain growing regions. A small wood- and waste-fired steam plant should be investigated for supply to the new coffee plantations, sawmill and en- virons of Bubeka, which soon will constitute a substantial load (Chapter II). Literally hundreds of thousands of tonnes of wood-waste from clear- ing operations are burnt in this region annually. Cogeneration of steam for crop processing and power in several hundred kilowatt plants should be investigated as a substitute for diesel generation. There is also the prospect of using agricultural residue briquette-fired gasifiers for diesel power generation on those state farms that begin to produce fuel briquettes for the urban household market. This option is discussed in Chapter V, though it is not considered to be significant in the short- term. Financing Power Sector Investment 3.50 Considerable investments will be required in the power sector to meet forecast demand. EELPA's programs for investing in both genera- tion and transmission capacity will be a major component of public sector investment. EELPA have prepared an investment program for both the EELPA and ERESA systems for the period 1984-93 in which total expenditure is about B3 billion (US$1.5 billion in 1983 prices). This program is based on installing capacity to meet the Acres demand forecast, and the gen- eration component alone totals about B2 billion. The investment program therefore exceeds the system requirements according to the mission's forecast. Availability of funds will be a major consideration in - 72 - determining the rate of investment, particularly in transmission pro- jects. 3.51 The mission has prepared an alternative investment program for the period 1984 to 1995 in which total expenditure is about 82 billion in 1983 prices, of which the generation component is about B1.2 billion. This program is based on the mission's demand forecast, and is set out in Table 3.11. The program recognizes the importance of extending the transmission system by retaining most of EELPA's transmission projects. The program also includes nine mini-hydro schemes designed to replace diesel generators in isolated load centers. Table 3.11: SUMMARY OF PROJECTED INVESTMENT IN THE ETHIOPIAN POWER SECTOR (B million 1983 prices) 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 Total Generation EELPA ICS 73 126 126 101 -- 48 95 119 142 71 60 60 1022 EELPA SCS 20 22 26 21 9 9 9 2 2 2 2 2 126 ERESA ICS & SCS 26 -- -- 1 -- 1 -- 1 -- 1 -- 31 Subtotal 119 148 153 122 1O 57 105 121 145 73 63 62 1179 Transmission & Distribution 230 kV/132 kV 58 67 47 50 22 30 31 23 -- -- -- -- 328 66 kV/45 kV 2 -- -- -- 19 35 19 13 9 5 5 5 112 Substation Expansion -- -- 17 32 24 11 7 7 5 5 5 5 118 Distribution Expansion a/ 12 12 12 12 12 12 12 12 12 12 12 12 144 Subtotal 72 79 76 94 77 88 69 55 26 22 22 22 702 Infrastructure Subtotal 11 24 26 24 9 2 2 2 2 2 2 2 108 Total 202 251 256 240 96 147 176 178 173 97 87 86 1989 a/ Includes village electricity. Source: Annex 3.4. By retaining most of EELPA's transmission projects, the main cost reduc- tion occurs in generation where commissioning of the Gibe Daka project is delayed by four years and that of the subsequent hydroelectric station is also delayed by four years. In other respects this program is based on minimum investments required, in the sense that no expenditure is - 73 - included for interconnecting the EELPA ICS and the ERESA ICS, a major regional transmission system in the Eritrea/Tigrai area, gas or geother- mal power development, or for developing generation and transmission capacity for power exports. The feasibility of such developments has yet to be established either technically or economically, although after 1990 this situation could change. 3.52 The most important issue concerning the means of financing the investment program is the means of local financing. About 65% of the program is composed of direct foreign exchange expenditure, and the in- clusion of indirect foreign exchange costs increases this proportion to about 75%. It is reasonable to assume that this proportion can be funded almost entirely from foreign sources, but the difficulties of mobilizing these funds are considerable. The requirement for local financing thus will be about 25% or B500 million in 1983 prices. 3.53 In the past, it has been COE policy to fund capital expenditure from the national budget and to makeover the assets to EELPA as a grant. In return, EELPA has not been required to fund capital expenditure or debt service from internal resources, which has been a strong factor in preventing the implementation of tariff increases. It appears that the present capital work-in-progress, including the Malka Wakana project, will be treated in this manner. However, COE has decided to change its policy for future capital investments and wilL require EELPA to become self-financing. This change also will have implications for tariff levels and structure, particularly the subsidization of the EELPA SCS and ERESA systems. At present, EELPA receives a subsidy from the government for this purpose, although there is some uncertainty about the amount of this subsidy before it is actually paid. EELPA recently has commissioned a tariff study, the conclusions of which are discussed in Chapter VII. - 74 - IV. PETROLEUM 4.1 Imported petroleum products comprise only 6.5% of total final energy consumption, yet their current annual cost of about B400 million is more than half the value of imports (see Table 1.2). Inland consump- tion of petroleum products now exceeds 600,000 toe, or 13,000 bbl/per day, having grown at 7.6% p.a. between FY77/78 and FY81/82 (see Table 1.1) compared with a growth in GDP of 4.1% during the same period. The current sectoral end use of petroleum products is described in Chapter 1. Transport uses about 70% of petroleum product consumption and diesel oil dominates the product mix, providing about half of all petroleum energy consumed. 4.2 The key issues in the petroleum sector are concerned with re- ducing the cost of supply and delivery to the main central highlands market, and with efficient exploration and development of known or pros- pective petroleum resources. As the benefits to flow from satisfactory resolution of these issues will not be experienced for five years or more, the mission also gave high priority to a review of the prospects, in the short-term, for increasing the efficiency of petroleum fuel use. Pattern of Supply 4.3 Ethiopia imports both crude oil and refined petroleum products. Crude oil is refined at the countryts only refinery located at the Red Sea port of Asseb, and refined products are imported to make up the defi- cit between demand and refinery production. The 800,000 tpy refinery is owned by the COE and is operated by the Ethiopian Petroleum Corporation (EPC). Since the middle of 1980, the EPC has procured its crude oil on a contract basis from the USSR. The refinery design is such that fuel oil production can range from 32-55%, depending on the crude oil supplied. The crude oil currently in use yields 43% fuel oil, hence 60-70% of fuel oil production is exported at about 70% of the price per tonne of im- ported crude. Actual petroleum product imports and exports for 1978 through 1982 are provided in Table 4.1. Imports 4.4 During the period of post-revolution economic growth, the amount of imported refined products to supplement refinery production has grown considerably, being moderated only by, in 1981, an increase in re- finery capacity from 650,000 to 800,000 tpy of crude input. Imports of refined products as a proportion of retained petroleum imports rose from 14% in 1978 to 22% in 1982, peaking at 37% in 1979. Petroleum Distribution 4.5 The distribution and marketing of petroleum products in Ethiopia is handled by four major oil companies: ACIP, Total, Mobil, and - 75 - Shell. The EPC delivers products by pipeline to the oil company storage in the vicinity of the refinery, and from there the bulk of the products are transported by road (mostly by Natracor, the national road transport agency of the Government) to demand centers in the central highlands. However, some products are moved by coastal vessel and the Djibouti-Addis Ababa railway to Dire Dawa and central highlands markets, and coastal vessels also supply product to the port of Massawa and distribution inland from there is by road. The companies retail the product through a national network of service stations and wholesale it in bulk to major consumers. The EPC also imports refined product to make up the deficit between refinery production and demand, allowing the oil companies to import product only at its discretion. Taole 4.1: PETROLEUM PRODUCT IMPORTS AND EXPORTS, 1978-82 1978 1979 1980 1981 1982 US$ per Unit ('000 metric tonnes specified product) CIF,1982 (per bbl) Crude Oil 611.4 612.4 633.1 684.0 773.3 $30.63 Refined Products (per tonne) Avgas 9.0 16.1 8.3 6.9 8.1 $679.00 Jet. A-1 7.4 15.8 2.1 6.0 4.2 $332.55 Special Jetfuel (TS-1) 36.0 14.0 20.0 12.7 47.3 $388.76 Gasoline (regular) 2.7 41.2 21.4 10.6 9.0 $337.30 Diesel 10.1 137.9 135.9 88.7 87.3 $307.85 Kerosene 0.5 0.4 0.5 $ $307.85 Total Imports 674.10 837.8 821.3 805.9 929.2 Exports (US$/tonne) Fuel Oil 196.6 222.6 160.9 168.1 211.4 $153.40 Note: Net imports do not match consumption on an annual basis. Some products are shipped irregularly, giving rise to highly variable stock inven- tories. Source: Ethiopian Petroleum Corporation. Alternative Supply Options 4.6 The refinery was built in 1967 with Soviet technology and tech- nical assistance. Presently, the refinery is operated and overhauled with the assistance of 20 Soviet technicians and skilled workers. Many of the existing facilities are obsolescent and corroded. Key problems - 76 - include an internal loss and consumption of fuel oil and other hydro- carbons of 8% to 9% of crude oil throughput; heavy boiler tube corrosion and Low boiler thermal efficiency. The mission analyzed five of the most plausible options for reducing the cost of supply, including closing down the refinery and relying entirely on imported refined products. 4.7 The five cases examined in the mission's preliminary economic analysis of supply options were as foLlows: Case 1: Business as Usual Run the refinery as at present, continuing to make up the difference between refinery production and market demand with imported refined product and to export surpluses of fuel oil. Case 2: Conservation Only Improve refinery efficiency and impLement energy conservation measures without increasing capacity. Import the refined product deficit and export surplus fuel oil. Case 3: Conservation and Debottlenecking Increase refining capacity from 0.8 to one million tpy in addition to the measures taken in Case 2. Case 4: Conservation, Debottlenecking and Secondary Conversion In addition to Case 3, install thermal cracking and hydrotreating facilities to increase distillate production at the expense of fuel oil. Case 5: Import Only Close down the refinery and import the entire refined product re- quirement at the best open market prices possible. In the more detailed analyses that should be undertaken before commitment to a stance on the future pattern of petroleum product supply, the op- tions to process spiked crude and to expand the refinery beyond one mil- lion tonnes of annual throughput capacity also should be examined along with options for achieving economies of scale in the transportation, handling and storage of crude and refined products. 4.8 The nature and level of investment required in each of these cases iS provided in Annex 4.1. Oil price projections are based on the Bank's own estimates and the investment costs are derived from recent Bank projects on design work in this field. In each case the refined product not produced by the refinery under the level of upgrading con- cerned is imported at the then estimated real cost (CIF). - 77 - 4.9 The results of these analyses are provided below in the form of the present value of the stream of net costs at 10% and 15% discount rates and the economic rate of return (ERR) on the incremental investment in the refinery should continued operation (Case 1) be the base case, or should it be, instead, closing down the refinery altogether (Case 5). Investment PV of Net Cost PV of Net Cost ERR for Cost at 10% Discount at 15% Discount Incremental ERR (US$ million) Rate (US$ million) Rate (US$ million) Investment Overall Case 1 nil 1900 1593 base case -- Case 2 22 1861 1578 32 9% Case 3 25 1866 1573 28 7% Case 4 67 1579 1501 35 26% Case 5 nil 1859 1557 -- base case 4.10 The ERR for investment in secondary conversion is attractive though the analysis is clearly sensitive both to the long-run costs of oil (here assumed to reach US$38/bbl in real terms by 1995), the final cost of modifications, and the future demand for refined petroleum pro- ducts. The mission therefore recommends a detailed inspection of the refinery facilities followed by engineering and economic studies to determine more accurately the cost basis for comparing these and other options, and to determine the final cost and practicality of refinery expansion and upgrading. Product Transportation 4.11 A key issue in the petroleum sector is the cost of transporting petroleum products from the refinery and point of import at sea level to the main market of Addis Ababa some 870 km away and at 3,000 meters elevation in the central highlands. By the present method of road trans- port, one tonne of diesel is used for every 18 to 24 tonnes of petroleum product transported, making the energy cost of petroleum transport a significant factor in the choice of transport mode and a significant por- tion of total diesel consumption. COE is considering a number of alter- natives including a pipeline for refined products from Asseb to Addis Ababa, a railway line from Asseb to Addis Ababa, and a combination of barge and rail transport between Asseb and Addis Ababa via Djibouti using the existing rail track. The mission regards the Asseb-Addis Ababa railway line as a very long-term prospect to be considered in the much wider context of the transport sector as a whole. The existing Djibouti- Addis Ababa railway is now quite incapable of carrying more than a small proportion of the specified load. Preliminary estimates indicate a - 78 - requirement for at least several trains per day, each pulling 15 rail tanks (180 tonnes per train). Currently two combined passenger and goods trains use the line. To carry a significant, additional load the line would have to be duplicated in piaces and greatly improved to accommodate bigger trains. These investments must also be considered in the context of an entire transport sector strategy, though major new investments in additional line and rolling stock are unlikely to be made this decade. Nevertheless, rail transportation is generally much less energy intensive than road transport per tonne kilometer and the results of the EEC-funded prefeasibility study on the future of the present railway should provide guidance as to the longer term prospects of an expanded and competitive railway system. 4.12 Therefore, the mission has made a preliminary analysis of only two options: continued use of road transport and the construction of a white product pipeiine from Asseb to Addis Ababa. The mission's analyses have been supported by a considerable body of data gathered during the course of the mission from Natracor, the Ministry of Transport, and var- ious private truckers and truck dealers. In reviewing this data it became clear that marked differences exist between the energy efficiency and overall economy of operation of different classes of vehicle within the national trucking fleet in general, and the tanker fleet in particu- lar. For this reason, two classes of truck transportation, 24 tonne and 36 tonne net carrying capacity, have been analyzed as separate cases in comparison with the product pipeline. There do appear to be significant differences in the longevity and maintenance costs per tonne kilometer between certain makes and models, though the most important observation is that there are very large differences in fuel consumption per ton km, with large (i.e. 40 tonne capacity) tankers consuming much less fuel per tonne of product carried between Asseb and Addis Ababa. In respect of the pipeline, besides the usual industrial indices, the mission had the benefit of recent cost data from detailed analysis of an identical faci- lity in South Yemen. Specific Variables 4.13 The pipeline case analyzed is for a 10-inch line 860 km long with a maximum throughput of 25,000 barrels per day. Taking account of the difficult terrain, the financial cost of the pipeline is estimated to be about US$160 million, plus about US$31 million for pump stations, and US$26 million for ancillaries. The total economic cost of the pipeline is about B565 million in 1983 Birr. For the road transport option, the key variables assumed are provided in Annex 4.2. All cases are compared on the basis of investment beginning in 1984 with product volumes of 416,000 tonnes moved along this corridor in 1987 and growing to 750,000 tonnes per year at the end of the period. - 79 - Table 4.2: LEAST-COST OPTION Economic Rate of Present Value of Net Return on Pipeline Transport Mode at 10% Discount Rate Investment (US$ million) (%) Medium Capacity Trucks 756 23.5 Large Capacity Trucks High Cost 586 16.5 Moderate Cost 475 11.0 Pipeline 447 base case Note: Working papers provided to the COE. 4.14 As the present fleet is comprised of medium-size 24 tonne tankers, the most expensive option is to continue the present mode of product transportation. From Table 4.2 it is clear that the ASSAB-Addis Ababa pipeline is the least-cost and most economical means of petroleum transport inland. In the event that the GOE decides to delay the imple- mentation of the pipeline project, the lower-cost large vehicle case is a reasonable representation of future road transport costs, assuming the investment is made to alter the composition of the fleet, and that the recommended axle loading Limits are not exceeded. However, a much more detailed review of the substantial investment involved should be made not only in the context of energy sector priorities, but also taking account of overall transport needs and options. Therefore, the mission recom- mends that the National Road Transport Study proposed as part of the Bank's Second Road Sector Project due to start in early 1984 address this question in greater depth. In the short term, there is a need to review the truck procurement guidelines to ensure that optimal economy is achieved in ongoing road transporation of petroleum products (see para. 4.30). Petroleum Exploration and Production Prospects 4.15 In this component of the petroleum subsector, the main issue is the focus and extent of the existing petroleum exploration activity. The Bank is already actively supporting an expansion of petroleum exploration activities in Ethiopia. An IDA credit of US$7 million recently has been approved for petroleum exploration and promotion activities arising from the expenditure of US$400,000 on project preparation during which a con- siderable resource potential was defined. Another important issue in this subsector at this time is the future role of natural gas in the national energy economy. In support of the existing IDA credit agreement between the Bank and the GOE, the mission sought to outline the potential market for natural gas and the best way to proceed with a detailed evaluation. - 80 - Previous Petroleum Exploration 4.16 International oil companies began exploring for petroleum in the Ogaden Desert and Red Sea areas more than 30 years ago. These efforts pointed to large deposits of gas but not oil. In 1975, Tennaco confirmed a resource of about 1.3 TCF from one well, Calub-l, in the Ogaden Desert, and Mobil had a gas blowout from a well drilled in a salt formation in the Red Sea, although because neither market nor local con- ditions were conducive to further development at the time both fields were abandoned and the resource was never fully proven. 4.17 Presently, two prospective sedimentary basins are being sub- jected to some level of exploratory activity. The Bank project provides for additional seismic work in the Red Sea basin, and a joint Soviet- Ethiopian exploration program (SPEE) is in progress in the Ogaden Desert. Until 1983, Chevron Oil Company had been gathering geophysical data on the Gambela Promontary in the far southwest, just over the border from their oil find in Sudan. SPEE is the most significant of the current activities. It is three years old and covers 10,000 km2 in the Ogaden Desert region near the Tenneco discovery. The agreement provides for the collection of 1,200 line km of seismic data and for drilling four wells with the objective of finding oil at relatively shallow depths. The first well was drilled on the Shilalo structure to a depth of 2,900 m and was abandoned as a gas well, and the second is to be drilled on the Hilala structure where, from the first and only well sunk, a significant oil show was recorded. Production of the Ogaden Gas 4.18 The "known" resource of 1.3 TCF is not large enough for export but is very substantial in comparison to the local demand for modern energy forms. The primary market for natural gas is for power production as a direct competitor with hydropower. In addition, natural gas could displace fuel oil in the cement and textile industries, diesel in road and rail transportation in compressed form, and LPG in larger commercial applications. Furthermore, the National Chemical Corporation has recent- ly established a good case for a 200,000 tonne per year urea industry which could become economic with the use of local natural gas. There is already a provision for a detailed gas utilization study in the IDA credit for petroleum exploration and development so the mission did not attempt to develop more than a preliminary overview of the prospective market. At this stage, it is more important to establish whether or not there is any prospect of economically competitive production and delivery of natural gas from a resource 800 km from the major market of Addis Ababa. 4.19 An indication of economic potential can be gained by assuming that the only market for natural gas is in power production. If the entire costs of the required further drilling of the pipeline and of a receiving facility and gas-turbine farm are weighed against the incre- mental market for power from 1994 onwards, the present value unit cost of electricity from gas is estimated to be roughly 7 USC/kWh. The pipeline - 81 - in the case analyzed is 20 inches in diameter and 800 km long, with an average throughput of 220 MMCFD, and the estimated financial cost of construction is US$300 million. For the first few years of the assumed 20-year project life the pipeline is used at less than one-third of its capacity and by year 20, less than half of the estimated gas resource of 1.3 TCF is utilized. As there exists an excellent prospect of serving additional markets during the first 15 years, substantially lowering the average cost of supply, the mission concludes that the prospects for the economic development of natural gas are good and should be vigorously pursued. Gas might be delivered for as little as US$3.00/MMBtu (full economic cost) at the rate of production for power supply, which is about 40% of the cost of fuel oil in Addis Ababa. A potential gas market (for cement and fertilizer) also results in Dire Dawa some 550 km from Calub 1 in the general direction of Addis Adaba, and there may be some minor de- mand in the Ogaden region much closer to the resource. There also have been natural gas shows in the north of Ethiopia offshore in the Red Sea. Further seismic work is proceeding as part of a new evaluation of the potential of this resource which, if confirmed, would have to be con- sidered as another option for supplying power to that region. Immediate Requirements for Evaluating Gas Supply 4.20 An important next step in evaluating local utilization of natural gas is to confirm the size of the resource. One well is not a sufficient basis for expending very large sums on market analysis, hence the mission recommends that, as soon as possible, at least one additional well, suited for production if justified, be drilled on the Calub struc- ture near Calub-l. In parallel, a preliminary gas market analysis cost- ing US$100-150,000 should be undertaken to confirm and rank the foreseen end uses, and to set terms of reference for detailed study if, and when, justified. GOE should make every effort to arrange drilling of the pro- posed Calub-2 within existing petroleum exploration agreements and work programs. Petroleum Conservation and Substitution 4.21 There is considerable scope for both increased fuel economy and interfuel substitution in the industrial and transport sectors. New energy sources for industry include agricultural residues and direct solar heating. In the transport sector the only alternative fuel avail- able in the short-term is ethanol (see para. 5.18) which will substitute directly for gasoline and, at the same time, expand to a comparable degree the supply of diesel from the Asseb refinery. In the longer-term there is the prospect of natural gas, in the form of CNG or LNG, as a substitute for both diesel and gasoline. - 82 - Improving Energy Economy in Industry 4.22 The key areas for energy savings are improved combustion con- trol and heat recovery. 22/ Apart from the sugar industry, none of the boilers or furnaces inspected by the mission had combustion monitoring equipment, and combustion control was quite poor. Fuel savings of 10% can be achieved through both the provision of appropriate equipment and operator training at relatively little cost. Improving boiler water treatment is in the same category and should be included as part of a training program the mission recommends for energy efficiency. 4.23 Heat recovery systems offer important savings for the textile, glass, steel and cement industries. In the textile industry, a 12% sav- ings of energy use should be achieved by extracting heat from effluent from the finishing sections of the mill. Installation of ceramic or stainless steel heat exchangers to extract otherwise wasted energy from furnace flue gases should save 25% of current energy use, and useful energy can be recovered by recycling clinker cooler air for raw mill drying in the cement industry. The mission recommends that prefeasi- bility studies be commissioned immediately to cost and rank these mea- sures as part of a program of investment in energy conservation. Interfuel Substitution in Industry 4.24 Substitution of fuel oil by bagasse and cereal straw bales or pellets in the cement industry and some boilers appears practical and economic. It is now common for cement plants to provide a proportion of their energy from waste materials (e.g., rice husks, household refuse, tires) co-fired with oil or coal. In Ethiopia, fluidized bed combustion of agricultural residues is recommended for the new Mugher plant near Addis Ababa, and for the proposed Dire Dawa plant. Displacement of as much as 50% of fuel-oil normally required should be possible if the burnt gases are fired into the low temperature end of the kiln. The existing sugar industry at Shoa and Wonji can, with certain investments, generate 35-40,000 tonnes of bagasse surplus to their annual need, which is more than the amount required, and they are ideally located to serve this market. The wood-fired boilers of the United Oil and Soap factory must be refurbished in any case as back up to, and eventual replacement of, new electric boilers installed there, and at minimal additional expense they also can be made to fire baled or briquetted crop residues. Fin- ally, solar water heating is economic in at least three major hotels and seven tanneries. The preliminary design and costing of these installa- tions should be included in the national solar water heating review project later recommended (see para. 5.20). 22/ A long and varied List of pLant-specific prospects for energy conservation was compiled and made available to the GOE, and in- cluded improved insulation, repairs to combustion equipment and the like in the textile, brewing, glass, iron and steel and other industries. - 83 - Fuel Oil Viscosity 4.25 The viscosity of the fuel oil sold to industry is kept very low (600 sec-SSU) by blending in kerosene at the refinery. In all, 3,000 tonnes of kerosene are consumed in this way. On the one hand, with minor investment most industrial plants could burn higher viscosity fuel oil (1,200-1,400 sec) and on the other hand the EPC claims to be able to easily produce and handle higher viscosity. At least 2,000 tonnes of higher value kerosene per year can be saved for use in the household sector in this way. In the longer-term, much more viscous fuel oil (3,500 sec) could be used in the cement industry, providing fuel handling facilities are designed accordingly, leading to further enhanced economic benefits. 4.26 The costs and benefits of the measures proposed to increase the efficiency of petroleum fuel use and interfuel substitution in the indus- trial sector are summarized in Table 4.2. The overall savings possible within the next five years are of the order of 22,000 tonnes or 20% of the present petroleum use in industry, mostly in the form of fuel oil. The return on investment is uniformly high with a simple payback averag- ing just over one a year. The total capital requirement of about B12 million is modest when compared with current energy sector investment and should be accorded a high priority. The Ministry of Industry should be charged with the responsibility of defining and scheduling the required investments on a plant-by-plant basis to facilitate early implementation. Table 4.3: SUMMARY OF FUEL ECONOMY OF SUBSTITUTION MEASURES IN INDUSTRY Value of the Annual Approximate Annual Fuel Fuel Oil Investment Savings Savings Required Resulting a/ (tonnes) (B '000) (B '000) Combustion Control and Heat Recovery 3,900 4,000 2,800 Firing Agricultural Residues 15,000 5,000 4-6,000 Solar Water Heating 1,400 2,500 850 Firing Heavier Fuel Oil 2,000 200 840 (kerosene) Total 22,300 11,700 8,490-10,490 a/ This is not a measure of economic performance of these interfuel substitution measures. Additional information is required to estab- lish the economic rate of return of these investment prospects. Source: Mission estimates. - 84 - Transport Energy Efficiency 4.27 The mission's efforts here focussed on the truck and bus fleet which between them are responsible for more than half of all diesel con- sumption. Fuel use is excessive in both the public and private bus fleets because of overloading, insufficient maintenance, and continued use of old vehicles. Fuel efficiency in the trucking fleet is also very low by international standards. In both-cases, one aspect of the problem is insufficient vehicles whereby routine maintenance has ceased and there is widespread, premature aging of all components of the drive-train. Efforts to improve vehicle maintenance facilities will achieve relatively low returns until the fleet is adequately expanded. Possibly US$20-30 million for additional vehicles is required to achieve the reserve capa- city required to facilitate routine maintenance. This investment should be placed in the context of the much larger investment proposal arising out of the Bank-supported transport sector study recently completed. Since the savings forecast from the specific fuel economy measures pro- moted here are to some extent predicated on routine maintenance, invest- ment planning for the road transport fleet should explicitly recognize these linkages. Specific Measures 4.28 In the short-term, proven low maintenance requirements, espe- cially for diesel pumps, should become a prime element of procurement po- licy. There is an urgent need for an investment of B4.5 million in modern workshop equipment and diesel pump testing facilities for the Government's bus fleet and, similarly, B6 million for its truck fleet. Overall, annual savings of about 6,000 tonnes of diesel, or about B6.0 million, will be achieved, quite apart from savings resulting from reduced wear and so on. Savings of this kind aLready are being achieved in certain private trucking fleets in Ethiopia which have the benefit of properly programmed servicing, and of diesel testing equipment. Trucks and tankers deployed continuously on the Addis Ababa-Asseb truck route, which is a good quaLity highway, should be fitted with turbo-charged engines. The fuel savings foreseen from this conversion are about 8,000 tonnes per year, valued at 36 million, for an incremental investment of B16 million. In the process of undertaking the vital once-off fleet expansion program here provided, and in routine vehicle replacement and fleet expansion, truck size should be minimized for the particular road environment and cargo concerned, and driver having reoriented to the efficient management of larger, more sophisticated vehicles. Management Information Systems 4.29 It is evident from the detailed review of cost data for the trucking fleet compiled during the analysis of the Addis Ababa-Asseb petroleum products pipeline that much can be learned from systematic and concise monitoring of individual vehicle performance. Large differences between vehicle make and model are apparent in respect of maintenance history, fuel economy, down-time, and longevity, all of which are impor- tant in comparitive cost analyses. Not only is the fuel economy of - 85 - vehicular transport important in the overall energy economy, but trans- portation is a vital and costly component in the supply of all wood fuels which, in all respects, dwarf the other energy forms in use. The mission strongly advises a comprehensive transport management information system be established to recover and compare on an equal footing all relevant data on the classes, makes and models as an input both to transport energy management and to overall procurement policy. - 86 - V. OTHER ENERGY SOURCES 5.1 There is little doubt that some economic justification can be found for a great many energy sources new to the Ethiopian energy eco- nomy. However, the mission has attempted to confine its deliberations to the key issues agreed with the GOE and to review in depth only those energy sources which may make a significant contribution to the energy economy in the foreseeable future. Accordingly, energy sources such as solar electricity and wind power, which are seen as having only limited applications, have not been examined further and, from field observation, it has become clear, too, that biogas and producer gas power production do not yet warrant detailed analysis. The energy sources examined in detail here are crop residues of various kinds and ethanol, with some consideration of geothermal and coal resources and biomass gasification. Crop Residues 5.2 The rationale and potential for exploiting for fuel surplus lignocellulosic fuels of low opportunity cost is provided in Chapter II as part of the evaluation of a short-term strategy for augmenting house- hold energy supply. In the discussion below, the distinction must be drawn from the outset between crop residues produced on peasant farms and those produced on huge state farms and arising from the coffee and sugar industries. The former already are used as fuels and have equal or higher value as fertilizers, soil conditioners and stock-feed. There is rarely sufficient straw produced to adequately feed cattle, let alone to serve other purposes (see schema paras 2.4 and 2.5). Therefore, the use of these crop residues as household fuels is, for the most part, undesir- able, and the prime objective of the househoLd energy strategy outlined is to displace these residues and animal dung to higher value uses. On the state farms and for the specified cash crops, vast quantities of residues are left on the fields, or at processing sites, where they are either burnt or rot away. The removal of at least 750,000 tpy (330,000 toe) of this combustible residue is possible at negligible opportunity costs, and preliminary estimates indicate that the materials can be briquetted and transported several hundred kilometers to be delivered at costs competitive with firewood and other traditional fuels. The speci- fic circumstances are outlined briefly below. State Farm Crop Residue Resources 5.3 The residues concerned here are those of the wheat, barley, maize, sorghum and cotton crops. The estimated quantities of residues generated in the 1980-81 cropping season are provided in Table 5.1. Cotton, maize and wheat residues comprise 95% of the total. Resource estimates are made by first determining total residue production (stalk, - 87 - leaves, flowering parts) using harvest indices 23/ agreed with state farm agronomists and then half of this amount is allocated as an accessible resource, provided the residue left in the field is in the range of 500- 1,000 kg/ha. Higher or lower collection rates are applied to adjust to this residual where necessary. The intention here is to leave sufficient crop organic matter to sustain soil structure and prevent serious ero- sion. The residues now remaining are mostly burnt off prior to replant- ing, voiding any significant soil protection. Indeed, in the case of cotton it is desirable to burn off old plants to limit the transmission of crop pathogens from one region to the next. Rather than being an agent of depletion, the rational collection of crop residue can form part of an integrated management strategy to maximize economic returns and to stabilize long-term productivity. Table 5.1: ACTUAL CROP AND ESTIMATED RESIDUE PRODUCTION BY REGION FOR STATE FARMS, 1980-81 ('000 tonnes) Cotton Wheat Barley Sorghum Maize Total Location Bale Residue Grain Residue Grain Residue Grain Residue Grain Residue Residue South 7 37 112 105 4 3 -- -- 43 80 225 North -- -- -- -- -- -- 7 7 2 3 10 West -- -- -- -- -- 18 16 51 101 117 Awash 51 202 -- -- -- -- 1 1 1 3 206 239 105 3 24 187 558 Source: Mission estimates. 5.4 Harvesting and Processing. The basic techniques for harvesting and processing this resource vary from crop to crop. For cereal straw and stover some form of baling is required in the field during the two- to three-month harvest to allow for easy transport to, and storage at, the processing center on each farm. Baled material is then retrieved from straw, finely ground, and briquetted at an even rate all year round. All of these crop residues are naturally dry (10-15% mcwb) by the time of baling and can be briquetted or pelleted under very high pressure in heavy duty motorized steel presses without the use of a binder. Trans- port to market is conveniently handled in bulk or bags, as the briquettes have bulk densities of 600-800 kg/m . Investment is required in addi- tional tractor and trailer capacity, baling equipment, and in briquetting facilities. Cotton stalk collection is done without baling and conse- quently the whole of the residue could be processed to briquettes during the harvest season. 23/ Harvest index here is the proportion which the grain or bole makes of the total above ground biomass yield measured at a constant common moisture content. - 88 - 5.5 Problems. State farm operations currently have poor producti- vity and financial performance and are plagued with poor quality farm equipment and insufficient maintenance staff and facilities. Long delays in receiving spare parts on order simply comFound these problems so it is not unusual to have 100% overcapacity in tractors just to have sufficient numbers operational at any time. Obviously, harvesting residues adds from one to several more mechanized field processes, and increases the maintenance burden. A major effort is being made by the Ministry of State Farms to upgrade mechanical skill and equipment operator training, and the results of this effort must be experienced before scaling up mechanized farm operations. There also should be considerable attention given to the choice of technology based on energy efficiency, maintenance experience and durability; evaluating life cycle rather than first costs. Very substantial maintenance and reserve capacity allowances have been made in reviewing the economy of residue collection to reflect the pre- sent day operational environment. However, the capacity to manage addi- tional operations, regardless of cost, is currently limited, and would have to be improved before full scale harvesting and processing of fuel value residues could proceed. 5.6 Costs of Production. A summary of the mission's preliminary estimate of the costs of crop residue briquette production and delivery is provided in Table 5.2. Costs are fairly evenly distributed between field coLlection, densification and transport, with the delivered cost averaging about US$50 per tonne. The single most important cost of production is energy in the form of electricity to operate briquetting presses and, to a lesser extent, in the form of diesel to fuel farm machinery. Where only diesel power generation exists, and power is US$0.35/kWh, energy costs comprise 35-40% of the total. Where hydropower is available, energy costs fall below 20% of total costs (detailed cost data are provided separately). It is worth noting, however, that the net energy ratio for the briquette fuel cycle is about 20:1, i.e., about 20 times more primary energy is obtained than is used, including transpor- tation to the urban market. 5.7 Future of the Resource. From the plans provided to the mission by the Ministry of State Farms, the area under cereal production is to expand from 173,000 ha in 1982 to 243,000 ha in 1993. The area planted to cotton is to expand from 31,000 ha to 98,000 ha over the same period. Although expansion planned by area is significant, planners hope to achieve the greatest benefits from increasing yields. The data provided in Table 5.3 indicate that a 330% and 530% increase in grain and cotton yields, respectively, is expected by 1993. The fuel production targeted for 1992 is based on the collection of approximately half of the residue available in 1981. In fact, with moderate growth in state farms, the mission's projections of crop residue briquette production will be be- tween only one-eighth and one-fourth of the resource available in 1992. However, the value of this production will be about US$19 million at this farm gate, incorporating an attractive rate of return on investment, as well as offering a valuable source of high quality energy for on-farm processing (heat and steam-raising and power production) in the longer term. - 89 - Table 5.2: SUMMARY OF ESTIMATED COSTS OF CROP RESIDUE BRIQUETTE PRODUCTION AND DELIVERY (US$/tonne) Field Densification Transportation Collection and Packing to Market Residue and Storage Handling (330 km) Total Wheat & Barley Straw 14.40 16.30 16.00 46.80 Corn & Sorghum Stover 25.40 16.40 16.00 57.80 Cotton Stalk 10.50 17.90 16.00 44.40 Average Cost 16.70 17.90 16.00 44.40 % 33 34 33 100 Source: Mission analyses (available to the government). Table 5.3: ACTUAL AND PROPOSED PRODUCTION OF SELECTED CEREALS AND COTTON ON STATE FARMS OVER 1982-93 1981 1983 1985 1988 1993 Crop Actual (Estimate) - Projected ----------- ('000 tonnes Grain or Bole) Cereals Barley 4 5 19 25 36 Wheat 112 154 148 206 326 Maize 147 171 195 311 491 Sorghum 26 31 23 47 97 Total 289 361 385 589 950 Cotton 58 79 83 114 308 Note: Residue production will be roughly proportional to grain and cot- ton yields. Source: Ministry of State Farms. - 90 - 5.8 Scheduling Production. The rate of exploitation of this re- source is dependent on the success of the early trial work recommended in Chapter TI (para 2.19), and is constrained by the factors identified in 5.5 above. Four criteria are used for selecting state farms for the first phase of operations: proximity to the market, high yield, EELPA ICS supply, and good management record. The production schedule outlined in Annex 2.6 provides for only 20,000 tonnes of cereal stover briquettes and 40,000 tonnes each of cereal straw and cotton residues by 1987, rising sharply thereafter. By starting early pilot production at the prime Locations, there is a period of five years during which operational and maintenance skills, and farm management practive can be upgraded on the less successful farms before being burdened with residue collection and briquetting. Coffee Residues 5.9 Coffee is a major crop in Ethiopia, with total production in the range of 200-250,000 tonnes of green bean per year. Currently all but about 13,000 tonnes of this production is as unwashed beans. In the production of unwashed coffee the entire coffee cherry is sundried and carried to a centralized decorticating station where the dry skin, pulp, and husk is removed from the bean and dumped. Dry pulp residue is rough- ly 1.7 times the weight of the green bean. When removed from the bean at the cleaning stations it is very dry (10-12% mcwb) and quite often spon- taneously ignites with the heat generated by fermentation deep down in rotting piles from previous years. Some 150,000 tonnes a year of this residue are now available at 55 processing centers which are mostly small towns scattered through the main coffee producing provinces of Kaffa, Sidamo, Wollega, and IlLabador. Although the coffee industry is expected to expand greatly during the next decade the production of fuel bri- quettes here targeted is based on the residues available in 1982. 5.10 Washed Coffee. There is a premium for washed coffee in the world market, so the COE plans to expand washed coffee production to 26,000 tonnes by 1988. In the washed coffee process, the cherry is pro- cessed fresh, leaving very wet pulp residues and a dried green bean with the husk or parchment still attached. The wet pulp is discarded into water-courses, causing considerable pollution, and the husk is removed at one final cleaning station in Addis Ababa. At present, some 2,600 tonnes per year of the excellent fuel value residue is available there, expand- ing to 5,200 te per year by 1988. The wet pulp also may be processed to a dry fuel though the mission did not find this yet to be an attractive option, except perhaps in combination with pollution control measures. 5.11 Process and Costs. It is proposed that fuel briquettes be produced from the dry coffee pulp residue in parallel with seasonal de- cortication. Production units of 2,500-7,000 te briquettes per year are envisaged, matching the size of the cleaning stations. Briquetting presses would have to be sized to process the flow of residue literally blown out the back of decorticating mills. Fortunately, many of the larger processing centers are connected with the EELPA grid, for other- wise electricity costs exceed half the total cost of production. A - 91 - sample cost analysis for producing coffee residue briquettes is provided in Table 5.3. The majority of the processing centers are between 330 km and 450 km from Addis Ababa, hence, transportation is about 50% of the delivered cost of the product. Even so, coffee residue briquettes are much cheaper than cereal or cotton crop residues due to the absence of a field collection phase in production. 5.12 Development Strategy. Coffee residue fuels are among the cheapest available, being competitive even with fuelwood production in some areas, and their production should be given high priority. The mis- sion recommends that a pilot briquetting press be installed as soon as possible at the Addis Ababa washed coffee cleaning depot to process the husks removed; that dry pulp residues are delivered for trial briquet- ting to establish design parameters for the first pilot briquetting plant at a decorticating mill; and that a preliminary selection be made of the most suitable processing site for the first commercial production. The mission supports the GOE's program to amalgamate old small mills into one large station in each processing center, as this provides the opportunity to develop a fully integrated coffee cleaning and residue briquetting operation offering considerable economies of scale. Table 5.4: COST SUMMARY FOR BRIQUETTI G AND TRANSPORTING COFFEE RESIDUES a Briquet- Transport b/ Financial Economic ting Bagging (US$) Cost FE Cost % Capital Charges 4.60 2.40 3.10 10.10 76% 12.80 30 Fuel and Power 3.70 3.70 7.40 50% 8.60 26 Maintenance 1.40 4.10 5.50 75% 6.90 21 Labor 1.20 1.00 2.90 5.10 c/ 4.40 14 Total 10.90 3.40 13.80 28.10 32.70 100 a/ Per tonne in 6,000 tonne/yr units. b/ From 365 km (reference of town of Dila in Sidamo) to Addis Ababa. c/ About 75% skilled and 25% unskilled at 0.5 shadow value. Source: Mission Annexes Report (available to the Government). Bagasse 5.13 The Ethiopian sugar industry produced 146,000 tonnes of sugar in 1982 from their Metahara, Shoa and Wonji plantations, located 110-350 km from Addis Ababa. A new industry is being developed at Finchaa about 300 km west of Addis Ababa which will produce 83,000 te (4,000 tonnes of cane per day: TCD) in the first stage, to be completed by 1986 and 125,000 te (6,000 TCD) after the second phase of deveLopment early in the - 92 - 1990s. The Metahara plantation and mill was expanded from 3,000 TCD to 5,000 TCD by 1982, and the addition of a further 2,500 TCD of capacity at Shoa will be completed during 1984. None of the present mills or the new and expanded mills are designed to maximize bagasse production through energy efficient operation, yet the new Shoa and Finchaa mills for which plant is soon to be procured could become, at a small marginal cost, major bagasse producers. Bagasse fibre dried and densified in bale, pellet or briquette form is an excellent fuel from a well proven process. A study recently commissioned by the Bank showed that sugar mills of similar design to those in Ethiopia could produce bagasse sur- pluses of 15-40 kg oven dry (kgod) per wet tonne of cane produced and that with varying levels of investment including the installation of higher pressure boilers, efficient turbines, flue gas heat recovery and improved pan evaporation designs, these surpluses could increase to 50-75 kg per tonne of cane. 5.14 The present production of surplus bagasse is 9,600 tonnes per year (50% mcwb) and will grow soon to nearly 16,000 te per year concen- trated entirely at the Wonji-Shoa mill complex. Much of this surplus will be used to displace fuelwood in lime-burning nearby. However, the mill's management has acquired valuable experience in baling and storing surplus bagasse. The present and potential bagasse surplusses are out- lined in Table 5.5. With appropriate investments, bagasse surpluses could be increased tenfold, creating a valuable fuel supply to industry, households, or in the case of the Finchaa sugar industry, a high degree of year-round self-sufficiency in power. In order to take advantage of this potential, the government must intervene in the procurement pro- cesses for both the Shoa factory expansion and the new Finchaa mill to allow for relatively small design changes to the relevant steam raising, power production, and process heat components now specified to operate well below optimum energy efficiency. Although the mission identified specific fuel saving measures in all factories, and in the Finchaa design, it is desirable to have the engineering contract managers in each case systematically upgrade and review specifications. If this is not done immediately, the opportunity to generate large bagasse surpluses will be foregone for the next 15-20 years. The estimates of future supply applied in Chapter II anticipate the availability of about one- third of the potential here cited. The costs of bagasse fuel briquette production will vary greatly, increasing with the level of bagasse sur- plus envisaged, and with the existing factory and process design. The costs should fall in the range of B50 to B70 per dry tonne ex-factory. Here, an average cost of B90 per tonne delivered has been assumed in all comparative analyses. 5.15 Cane field residues represent an even larger energy resource than potential surplus bagasse where they are not used for animal feed, such as in Wonji and Shoa, and possibly Finchaa. The fuelwood supply problem is so acute that this resource should be regarded seriously. Much of the dry leaf matter is burnt off prior to harvest and the green tops and leaves are left in the fields. The government is encouraged to review local means for collecting and processing this residue without disrupting sugar production. - 93 - Table 5.5: BAGASSE SURPLUS POTENTIAL OF THE ETHIOPIAN SUGAR INDUSTRY Sugar Cane Throughput Production Bagasse Surplus Sugar Mill 1987 1982 1987 Present Planned Potential (tonnes of cane ('000 tonnes) ('000 oven dry tonnes) per day) Shoa Line 1. 1,650 43 43 -- 7.9 17 Line 2. 2,500 -- 65 Combined 26 Wonji 1,420 38 38 4.8 -- 15 Metahara 5,500 65 140 -- -- 56 Finchaa 4,000 -- 83 -- 3.9 36 Total 14,570 146 300 4.8 11.8 150 Table 5.6: COMPARATIVE ECONOMIC COST OF INDUSTRIAL BOILER FUELS, ADDIS ABABA (US$) Economic Costs per Unit Price Per Conversion of Delivered LRMC per Unit Energy Value Unit Energy Efficiency Energy (US$) US1/MJ (%) (USA/MJ) Fuel Oil $337/te 42.5 MJ/kg 0.79 78 1.01 (incl. shadow value) Electricity Large boiler (>IMW) 7.0 USt 3.6 MJ/kWh 1.94 98 1.98 Small boiler (<1MW) per kWh 90 2.16 Cereal Straw $45/te 17.4 MJ/kg 0.26 65 0.40 Bales Cereal Straw $47.76/te 17.4 MJ/kg 0.27 70 0.39 Briquettes Cereal Stover $57.76/te 15.0 MJ/kg 0.39 70 0.56 Briquettes Cotton Stover $44.43/te 17.3 MJ/kg 0.25 70 0.36 Briquettes Coffee Residue $32.72 15.8 MJ/kg 0.21 70 0.30 Briquettes Bagasse S43.50 16.6 MJ/kg 0.26 70 0.37 Briquettes Note: Boiler capital and operating costs for new combustion equipment and fuel-handling systems for baled straw and briquetted residues will be 40% and 25% higher, respectively, than for oil-fired boilers. However, the overall costs of residue fuel combustion will remain less than half of those for fuel oil, and the return on incremental investment would be high. Source: Mission estimates. - 94 - The Market for Crop Residues 5.16 The household use of briquetted residues has been discussed and compared with other options in Chapter II, and has been found to have great economic and strategic promise in the short to medium term. Tests by the mission on the use of these briquettes in stoves have shown that, used alone, and unless they are in small pieces, they are difficult to light. Once well lit, they burn extremely hot and slowly. If they were to be used as a sole fuel source, the briquettes would have to be pro- duced as wafers instead: a process already commercially available. How- ever, in practice, these fuels would be used in combination with other raw woody fuels and with charcoal, where they would be ideal for sustain- ing a slow hot fire for the simmering phase of Ethiopian cooking, and the lighting difficulties could be eliminated or minimized. 5.17 Industrial Market. In Chapter IV, reference was made to the potential use of crop residues in the cement industry and at the Akaki Oil Mill. Residue briquettes are excellent boiler fuels and in practice it is likely that they will be used both as industrial and as household fuels. For the cement industry application and for some boilers, baled rather than briquetted residue can be accepted, thereby eliminating the capital intensive briquetting phase. The costs of baled residue de- livered to the urban market would be about the same as for briquettes due to the low bulk density of the bales and the great distances for trans- portation. Comparative estimates on fuel costs for industrial applica- tions are provided in Table 5.5. These preliminary data indicate that briquetted or baled crop residues will be less than 40% of the energy cost of a fueL oil-fired boiler and 20% or less of the energy cost of electric boiler operation. Apart from the obvious implications for future industrial energy use, it is important for the COE to ensure that the new boilers acquired for any of its existing industries in back-up to electric boilers, or for future industries, should be multi-fuel package boilers able to accept fuel oil, residue briquettes and natural gas. The mission strongly recommends that measures to ensure this are taken. Ethanol 5.18 The GOE has plans for a US$11 million, 20 million liter per year ethanol distillery at Shoa, using surplus molasses from Shoa, Wonji and Metahara and a US$8 million, 14 million liter per year distillery for Finchaa producing about 7 miLlion liters in the first phase of develop- ment. The required molasses feedstock of 72,000 tonnes will be available from the 1985 sugar industry, surplus to bakers yeast and potable alcohol production which between them consume 20,000 tonnes. Surplus molasses is now dumped on roads and in rivers. The likely long run molasses market price is US$71.50/te FOB New Orleans, and provides for no better than break even with the economic cost of transport to the port of Asseb some 800-1200 km from the point of production. The mission reviewed the capi- tal and operating coscs for a 20-25 million liter per year ethanol plant at the Wonji-Shoa plantation complex and found the likely costs of - 95 - production to be between 18-20 USc per liter compared with the present border price for gasoline at Addis Ababa of about 3OUSC/liter. The eco- nomic internal rate of return on the Wonji-Shoa investment is estimated to be about 39% at the investment cost of US$0.56 million per million liters per year of firm capacity. Doubling this capital cost will still yield over 20% ERR. The strength of this investment is based on the low opportunity cost of feedstock, the great distance of the market from the point of petroleum production or delivery, and the availability of steam and power at very low cost from the sugar mills. Despite the smaller size, the Finchaa distillery should show a similar economy. Ethanol fuels also offer greater benefits for combustion in normal petroleum en- gines at high elevation as they counteract the natural leaning of the fuel. A blend of 20% ethanol and regular gasoline will be both possible and desirable. Hence, the mission supports the proposed investment in ethanol production, providing that steps are taken immediately to avoid a number of possibly crippling constraints. Constraints to Using Ethanol Fuel 5.19 Fuel handling systems in vehicles and fuel depots, and fuel storage tanks at every point in the distribution system to the point of end-use must be both ethanol tolerant and water free to ensure optimum operating economy with ethanol fuels. Casual observations made by the mission in the field point towards significant levels of moisture in service station fuel tanks in the central highlands region. At 1-2% water, phase separation between ethanol and gasoline will occur, ren- dering the fuel unusable. Similarly, all vehicles imported from now on, and those now in use and still operating in 1986, will have to be checked for fuel system material sensitive to ethanol attack. The cost of re- placement is negligible for new vehicles but the disruption caused by faulty operation could be significant. Finally, adequate arrangements will have to be made regarding pricing, transport, blending and storage facilities for ethanol with the local oil companies before any commitment is made to distillery construction. Solar Water Heating 5.20 There are economic applications for solar energy to displace electric water heating in as many as 8,500 households in Asmara, and probably as many again in Addis Ababa. An additional but unquantified domestic market exists in the major towns and cities now supplied with electricity. In reviewing industrial energy use, markets were found for solar water heating for tanneries and hotels, hospitals, and the like across Ethiopia. The economics of solar water heating have been dis- cussed briefly in Chapters III and IV. There are potential benefits in local production of solar equipment providing the market is large enough to effectively amortize the costs of establishing manufacturing faci- lities locally. The mission supports the intent of the ENEC investment proposals for solar collectors and recommends an immediate review of the national market to define the feasibility and, if viable, the size and - 96 - nature of local production facilities. If a sizeable long-term market exists the COE should examine the prospect of a joint venture w$ith an overseas company to establish local production of solar collectors and to impart the necessary skills to design and service complex industrial and commercial installations. Other Energy Forms Coal-Lignite 5.21 Lignite has been found in more than ten areas in Ethiopia. The major deposits are in Wollega, Gondar, Eritrea, and Sidamo. Recent eval- uation of the Wollega and Gondar resource revealed poor quality discon- tinuous and deep deposits interbedded with thick basaltic flows. Further exploration is now in progress in Eritrea and Sidamo. Estimates of the total resource range as high as 20 million tonnes though the data base is minimal and the costs of production have not been carefully studied. It is noteworthy that deposits tend to occur in the more deforested energy- poor regions. Low quality lignite has been used in some countries, after suitable processing, both for power production and for household fuels. ENEC has recognized that even very small low quantity deposits (100,000 tonnes plus) could prove economic to exploit in the short- to medium- term, and a thorough resource evaluation is clearly required. The mis- sion strongly recommends upgrading the level, and increasing the scope, of the COE's resource inventory now in progress, with particular emphasis on the likely economics and proximity of potential markets. Subject to confirmation of an appropriate resource, a detailed anlaysis should proceed on the economy of producing lignite briquettes for the household and small industry market. Geothermal 5.22 The high enthalpy geothermal potential of Ethiopia has been estimated at up to 4,000 MW, higher than any so far identified in Africa. The first project to reach the exploratory drilling stage was at Aluto in the Lake district sector of the Rift Valley, 140 km from Addis Ababa. There appear to be good prospects of potentiaLly commercial steam flows following early drilling and additional wells were planned in this region. An evaluation is now required of the economics of power produc- tion from this resource (see para 3.40). US$20 million were provided for this exploratory drilling by the Commission of European Communities (CEC), the UNDP, and the COE. Additional wells will be drilled at Corbetti Caldera and the Lake Abaya area, and in Tendaho in the Afar region in the near future. Furthermore, the Bank has financed under the Petroleum Exploration Promotion Project a pre-feasibility study of geo- thermal energy production in the Fantale-Dofan area in the middle Awash Valley. Complimentary economic studies are included under that project to establish the contribution that this resource might make to the energy economy of Ethiopia. - 97 - Biomass Gas Sources 5.23 The GOE is considering proposals for an ambitious biogas pro- gram comprising the installation of 27,500 2m3 digestors over the next decade. The mission examined both the social and biological environment in which it is intended to place these digestors and a pilot plant at an agricultural training college at Awassa. First, it is clear that to gen- erate and maintain a digestor, considerable changes are necessary in a peasant family's daily activity schedule and, second, the required capi- tal outlay in materials (cement, piping, gas-holder, stoves, lights and other fixtures) of B500-1,000 is well beyond most rural peasants. How- ever, the most important constraint is the daily requirement for about 25 liters of water to dilute the required daily cow dung input of 15 kg fresh weight for a 2 m3 digestor serving a family of five. It is widely reported that only 6% of the population has regular access to a reliable water supply. In the rural areas, fetching water over long distances in clay pots of 10-15 liters capacity is a costly daily ritual. The mission believes that the program of agroforestry and rural woodlot development should be given much higher priority than a biogas program, and that the installation of digestors should be justified on a strict economic basis after carefully assessing the true cost and availability of all required inputs. It is likely that these will only be feasible at a limited num- ber of locations and will have a very small impact on the overall energy needs of the rural sector. Power Gasification 5.24 There is a prima facie case for a pilot program of producer gas power generation on state farms which become involved in the residue briquetting program. These state farms are the only locations where good quality gasifier feedstock and significant diesel power generation are found together. However, in the short term there are a number of con- straints offsetting the benefits that might result from displacing diesel power generation. In the first instance it will be up to five years before residue briquettes are produced on many state farms, should that option prove viable. Second, the power systems on state farms are almost all operating intermittently, not only being limited to 12-14 hours per day but also stopping and starting several times during the day. This makes the economic operation of gasifiers quite difficult due to both the much closer management requirement to bed down and recover the gas pro- ducer for each operating period, and due to the considerably reduced pro- portion of gas in the fuel mix that would result. State farms are cur- rently burdened with numerous mechanical and other maintenance problems and in the short-term it will be difficult enough expanding operations to produce residue briquettes let alone imposing power gasification techno- logies. The mission recommends that this program be put in abeyance pending a successful crop residue briquetting program. - 98 - VI. ENERGY PRICING 6.1 This chapter deals with several aspects of energy pricing of concern to energy management and planning. The first is the relationship between economic costs of supply and fixed or market prices for the most commonly traded fuels. Second, there is consideration of the financial as well as the social and economic determinants of price levels and price structures, and third, there is the interpretation of trends of market prices with respect to implications for future energy planning. A con- siderable volume of comparative data has been presented in Chapters II and IV on prices and costs of present and prospective energy forms. Here the discussion is focused on present day prices, pricing practices, and the implication of the existing price structure within the electric power, petroleum, and biomass fuel subsectors. Woodfuels 6.2 Hard reliable data on woodfuel prices are difficult to ob- tain. In Ethiopia the conventions for measuring retail prices of these fuels appear to have changed over the decade from volume to weight and back again. Volume measurements are inherentLy inaccurate, hence the government is advised to standardize price monitoring procedures on the basis of weight wherever possible, and to specify the size of the parcel measured in each case. The mission weighed and established the fuel value of wood, charcoal and dung in many locations during fieldwork and observed huge variability in marketing practices, and in firewood qua- lity, from one marketplace to another. The mission's own data have been added to existing data to make up a time series for wood and charcoal prices for Addis Ababa extending from 1970 to 1983. Nominal retail prices have been converted into real prices using the Addis Ababa retail price indexes as shown in TabLe 6.1. In real terms the price of firewood has risen from about B17 per tonne to B55 per tonne over the last 13 years, equivalent to 9.2% per year, with the bulk of the growth occurring in the past decade. Similarly, the real price of charcoal, which now retails for over B1,000 per tonne in small lots, has risen at 6.7% p.a. The sharpest price increases occurred between 1974 and 1978 when prices for both commodities doubled in real terms and increased fourfold in nominal terms. These data provide sound evidence of the growing scarcity of woodfuels and, in turn, of the deforestation crisis in the urban hin- terlands. 6.3 Regional Data. The need to consider separating the unique situation of the northern region is stressed in Chapter II. The only apparently reliable data comparing woodfuel prices during the same period are for the last quarter of 1980. These data are provided in Table 6.2, verifying the even more extreme situation of the northern region and in particular, Asmara and Mekele, the provincial capitals of Eritrea and Tigrai, respectively. 6.4 Government Woodfuel Prices. The GOE supplies the Addis Ababa woodfuels market with charcoal and fuelwood through retail depots - 99 - operated by the Fuelwood and Charcoal Marketing Company (FCMC) of FAWCDA (see para 7.14). FAWCDA proposes the retail prices for its products on the basis of its own costs of supply and the requirement of profitability for the FCMC. Table 6.1: WOODFUEL PRICES IN ADDIS ABABA, 1970-83 Retail Market Prices Movement in Price Real Price Year Firewood Charcoal Index Unitized at 1970 Firewood Charcoal (Birr per metric tonne) a/ (Birr per metric tonne) 1970 16.94 163 1.000 17.94 163 1971 17.60 169 1.005 17.50 168 1972 18.77 180 0,944 i9.88 191 1973 19.92 191 1.028 19.38 186 1974 28.33 272 1,117 25.36 244 1975 38.79 431 1.190 32.60 362 1976 49.08 458 1.529 32.08 300 1977 76.38 553 1.784 42.81 310 1978 81.27 768 2.039 39.86 377 1979 105.75 914 2.366 44.70 386 1980 101.92 1036 2.472 41.23 419 1981 131.40 1010 2.624 50.08 384 1982 no data no data 2.770 no data no data 1983 2/ 160.00 1136 2.895 55.27 392 (first half) b (to June only) -/ a/ Standard conversion appl ied to compute cubic meter stacked to metric tonnes assumes 600 kg/m basic density with a mixture of Acacia and Eucalypt sp.; 0.6 solid volume is stacked volume, and 25% mcwb (av). Charcoal sacks are assumed to average 45 kg. b/ Mission measurements of lowest level of retail sale i.e., for charcoal 0.2-0.25 kg lots; for wood 1-2 kg lots. c/ Assuming 4.5% half yearly increase in retail price index for 1983. Sources: Statistical Bu lletin of Ethiopia, ENEC and mission records. Table 6.2: COMPARATIVE PRICES OF WOODFUELS IN ETHIOPIA; LAST QUARTER OF 1980 Province City Firewood Charcoal (Birr Der tonne) Shoa Addis Ababa 94 670 Nazareth 74 330 Harrarghe Dire Dawa 73 310 Harrar 83 330 Wollo Dessie 67 440 Eritrea Asmara 167 840 Asseb 125 780 Tigrai Mekele 204 none available Gondar Gondar 46 380 Source: ENEC and mission estimates. - 100 - The comparison between government and market prices for firewood and charcoal is provided in Table 2.8 of Chapter II. Government prices are about 40-55% and 60% of the open market prices for charcoal and wood re- spectively, depending on the size of parcel sold. One objective of the government is to significantly lower the average market price but thus far has been unable to achieve this end, partly because of a lack of customer orientation in its marketing (restrictive pack sizes, opening hours, poor quality, rationing, etc.) and partly because it is supplying only a small part of the market. FCMC now supplies less than 10% of the charcoal and about 20% of the firewood demand of the Addis Ababa mar- ket. Financially its prices appear sound, ensuring the FCMC a margin of at least 15% over its total costs of operation. From an economic stand- point it is important to know whether the full resource costs of har- vesting the logging and forest clearing residues are covered within the consumer price. In this case, the resource costs are reasonably repre- sented by the costs of replenishing the wood resource through re-estab- lishing an equivalent standing biomass in the form of plantations, though this is a narrow site-specific determination, for elsewhere there may be higher value crops than reforestation. By allocating the relevant plan- tation development cost data to the timber and energy products concerned, a crude estimate is obtained of the resource cost of about B4.30 per solid cubic meter or B6.40 per tonne of fuelwood. For charcoal produced by traditional methods the resource cost computes to B59 per tonne and the economic costs of production and delivery become B190 per tonne. Furthermore, the long-run cost of fuelwood will be approximated by the cost of production from the most expensive peri-urban plantation develop- ment in the central highlands which is about B23 per m3 or B41 per tonne (20% mcwb) delivered. Since the government selling price for woodfuels exceeds both the economic costs of present supply and the long run costs of production and supply, the main consideration should be whether it serves the desired financial objectives. In the more commercial urban components of the woodfuel production program outlined in Chapter II, here envisaged, the government can opt either to fund FAWCDA directly from the budget to undertake peri-urban plantation development or, preferably, it can allow FAWCDA to generate sufficient cash generation from its sales to become substantially self-financing in the longer- term. The point in the supply chain (i.e., ex-forest, wholesale level, retail level) at which an interface is established for cost and capital recovery should be carefully reviewed, as the mission does not see an ongoing role for FAWCDA as a retailer of woodfuels (see Chapter VIII). Petroleum Pricing 6.5 Petroleum prices are proposed by an ad hoc committee estab- lished by the Government, and the final decisions are made by the CounciL of Ministries. The Ministry of Internal Trade is responsible for imple- mentation of agreed prices. Prices have recently adjusted from levels set in April 1981. The government policy is to peg prices to import parity in an attempt to make its refinery operations profitable. For reasons outLined in Chapter IV, the present state of the refinery obstructs fulfillment of this objective. In addition, there are deli- berate cross-subsidies attempted which in 1982 caused a net overall - 101 - subsidy to the refinery of B5.63 million. However, the notion of import parity pricing disregards the cost of foreign exchange and is, in effect, just an attempt to have receipts balance with direct costs. The "import parity" buildup applied by the EPC to achieve a price C.I.F. Asseb is shown in Annex 6.1. The buildup of retail price from the posting ex- refinery is provided in Annex 6.2 accompanied by the transport cost formula applied. Differences in the 1982 import parity and the ex- refinery postings illustrate the pattern of subsidy and cross-subsidy applied (eC/liter): Comparison of Estimated Parity Petroleum Prices with Ex-Refinery Postings Kerosene Regular Gasoline Diesel Fuel Oil Import Parity 68.3 60.6 64.5 46.8 Ex-refinery Postings 67.8 88.5 52.8 38.5 Source: EPC, Mission estimates. 6.6 The economic cost of supply and the current retail price in Addis Ababa for each of these products and for LPG is provided in Table 6.3. Retail prices in four other main towns are provided in Annex 6.2. It is evident that gasoline and fuel oil are the only products priced in excess of economic costs (with the shadow pricing of foreign exchange) and that very substantial underpricing applies to kerosene and LPG. Al- though diesel is priced only 7% below the level of economic costs, it is nevertheless an important "subsidy" due to the large volume of product concerned, the need to improve energy efficiency in road transport, and the mechanism for providing road-funding revenue that is being foregone. The small penalty on fuel oil is of little consequence and can be re- garded as both a revenue raising measure and an added incentive for boiler efficiency and interfuel substitution. The gasoline market is already highly regulated both by rationing to private and public vehicles, and by limiting the importation of spark ignition engine vehicles. While there is no economic rationale for such a high price, the measure can be justified as a luxury tax. There is a need, however, to ensure that high prices for gasoline, and comparatively very low prices for diesel, do not work against a full and economic use of locally produced ethanol which cannot be blended with gasoline at greater than 20% by volume without additional investment in vehicle fuel systems. Gasoline can readily be used in light commercial vehicles now using diesel in order to expand its market. - 102 - Table 6.3: ECONOMIC COST AND RETAIL PRICES OF PETROLEUM PRODUCTS IN ADDIS ABABA, 1983 Economic Cost a/ Retail Price Difference (ec/liter) Gasoline (regular) 80.3 117.4 +46% Diesel 82.7 78.0 -7% Kerosene 83.1 65.4 -27% Liquified Petroleum Gas 88.2 67.8 -30% Fuel Oil 55.8 59.9 +7% a/ The estimated product costs, CIF, and the relevant component of in- land transport costs are included at the shadow value for foreign exchange of 1.3 times the official exchange rate. Other costs in- cluded are part handling, distribution handling, and marketing charges. Source: EPC and mission estimates as per Annex 6.3. The Case for Increased Prices 6.7 Kerosene and LPC are well below the cost of firewood and char- coal as cooking fuels in most urban areas even when priced at full eco- nomic cost (see Table 2.8). Moreover, the economy of kerosene use can be improved by making available more efficient stoves. Both petroleum fuels offer greater convenience than their biomass counterparts for normal cooking, adding further to their attraction. Finally, the propensity of higher income households to use kerosene and LPG means that, at present price levels for various fuels, lower income groups in the cash economy tend to pay more per unit of cooking energy. The mission therefore recommends removing the subsidy on these household petroleum fuels and, in accordance with the proposed short-term strategy for household energy supply, concentrating instead on investments in production, delivery and end-use facilities to make them wideLy available on the market. The case for increasing the diesel price rests, on the one hand, with the need and the potential to improve efficiency of use and, on the other hand, with the importance of the measure in raising revenue for high priority energy sector investments, and road maintenance, which in turn has a favorable impact on fuel economy. Most of the diesel is used in road transporta- tion where small-fleet or single vehicle owner-operators contract to Natracor for business. The mission has observed that the operations of this component of the fleet tend to be more sensitive to energy costs and that with increased prices, improved fuel economy is Likely, not only through better operation of existing trucks and buses, but through the added inducement to purchase more fuel efficient vehicles in the future (see Chapter IV). Finally, the revenue base in Ethiopia is narrow and the vast majority of expenditures required for the rural forestry program is in local currency. The greater portion of the would-be beneficiaries of rural aforestation, the rural peasantry, have negligible contact with - 103 - modern fuels and would be unaffected by the price increases proposed, whereas stabilizing and stimulating peasant agriculture is fundamental to modern sector development, and the surpluses generated from higher petro- leum fuel prices would contribute to this end. It is therefore recom- mended that prices for petroleum fuel products reflect their economic costs, in particular, that kerosene and LPG prices be adjusted upwards as soon as possible. Electricity Pricing 6.8 Present Tariffs. The present electricity tariffs in Ethiopia have been in force since October 1978 for the EELPA ICS and SCS and since April 1974 for the ERESA systems. The value of tariff levels by 1982 had decreased by 26% for the EELPA system and 60% for the ERESA system in real terms since the date of the last adjustment (according to the Retail Price Index in Addis Ababa). The cost of power production in the thermal systems of the EELPA SCS and ERESA ICS and SCS has increased by a greater amount than the Retail Price Index due to increases in the cost of imported petroleum products during these periods. Although production costs on the EELPA ICS have not been directly affected by petroleum price increases since the system is supplied almost entirely from hydroelectric stations, the cost of meeting increases in demand beyond the capacity of existing hydroelectric installations will have increased substantially since 1978. Therefore, tariff levels need to be reviewed. 6.9 The present tariff structure for all the power systems is based on a declining block rate, whereby the marginal unit costs less than the initial unit purchased, and does not incorporate time-of-day factors. Such a structure is promotional and reflects a situation in which there exists substantial excess hydroelectric capacity, as was the case on the EELPA ICS during the mid to late 1970s. However, this situation is not likely to remain for long according to demand forecasts (Chapter III), and therefore this structure is inappropriate for the future. The de- clining block structure was never appropriate for the thermal-based sys- tems. Between consumer categories, the tariff levels for high voltage supply to industrial consumers are low relative to the levels for other categories even with allowance for lower system losses at high voltage supply. Tariff levels on the EELPA SCS and ERESA ICS are about 25% higher than on EELPA ICS, even though costs of supply relatively are much higher. The tariff levels on the ERESA SCS are nearly three times as high as on the EELPA ICS, which is a more accurate reflection of relative supply costs. Average tariff yields for the power systems are given in Table 6.4. 6.10 Economic Costs. The economic costs of supplying long-run increases in demand for power have been estimated recently by consul- tants 24/ for EELPA. These estimates are summarized in Table 6.5. The 24/ Coopers & Lybrand Associates "Tariff Study and Asset Revaluation" -- Final Report (August 1983). - 104 - consultants have differentiated between daily peak and off-peak demand periods since new investment will be determined by the need for addi- tional capacity before the need for additional energy capability. The consultant's estimated economic costs are compared to the existing tariff schedule in Table 6.6. For the EELPA ICS, there is little difference between the existing tariff and economic costs. However, the economic costs of supplying the thermal-based systems are about three times as high as the present tariff levels for the EELPA SCS and ERESA ICS, and about twice the level for the ERESA SCS. This comparison shows starkly the present degree of heavy subsidization of consumers in the three small systems. Table 6.4: Average Electricity Tariff Yields (Eth cents/kWh sold) EELPA in FY81 ERESA in FY82 Consumer Category ICS SCS Total ICS SCS Total Domestic 14.99 18.31 15.55 Commercial 15.17 20.39 16.44 ] 18.80 Street Lighting 11.00 15.00 12.93 15.00 Small Industries 16.20 20.03 17.62 Low Voltage Large Ind. 12.43 16.52 12.82 ] 10.05 High Voltage Large Ind. 8.26 7.50 8.19 - - - Agriculture 7.82 8.82 7.84 _ - - Average Revenue 12.00 14.92 12.52 14.19 32.87 14.40 Source: EELPA 6.11 The adoption of tariff Levels established according to economic costs of supply would introduce a huge differential between tariffs on the EELPA ICS and other systems. COE is sensitive to the social issues and considerations of equity that would be involved, as well as to the effects that large tariff increases could have on its overall poLicy for prices and wages. The consultants computed a tariff schedule based on uniform rates for each consumer category throughout the country. The uniform tariff level is about 50% higher than the economic cost of supplying the EELPA ICS alone, and is between one third and one half of the tariff levels for the other systems under separate tariffs. Thus, price distortions under a uniform tariff would be considerable, and the mission strongly recommends against adopting this policy. However, the mission recognizes the difficulties in changing tariffs immediately to a structure that reflects economic costs. Such a change could be intro- duced in stages over a few years. - 105 - Table 6.5: ECONOMIC COSTS OF POWER SUPPLY a/ EELPA ICS EELPA SCS c/ ERESA ICS ERESA SCS Long Run Marginal Capacity Costs (Birr/kW pa) Generation 232 106 157 324 Transmission/HV Dist. 30 10 20 10 LV Distribution 51 51 51 51 Long Run Marginal Energy Costs (Birr/kWh) 0.050 0.3757 0.3254 0.3588 Peak b/ Economic Costs (Birr/kWh) - HV 0.332 - 0.445 - - LV 0.477 0.695 0.554 0.968 Off Peak Economic Costs (Birr/kWh) - HV 0.054 - 0.342 - - LV 0.059 0.437 0.363 0.417 HV and LV: High Voltage and Low Voltage. a/ Evaluation based on Shadow Foreign Exchange Rate = 1.33 x official rate; 10% discount rate; Shadow Wage Rate for Unskilled Labor = 0.5 x financial rate. b/ Monday - Saturday 6:30 p.m. - 9:30 p.m. c/ Excluding Bahir Dar region which is treated as part of EELPA ICS. Source: Coopers & Lybrand Associates: tTariff Study and Asset Revaluation", Final Report (August 1983). Nevertheless, tariff levels should cover at least variable costs, which are largely composed of fuel costs in the thermal-based systems, and accordingly the mission recommends that tariff levels on the EELPA SCS and ERESA systems be raised immediately to at least an average rate of 35eC/kWh, with planned regular increases to the level of full economic costs over the short-term. 25/ 6.12 Financial Requirements. The consultants also examined the tariff levels that would be required to meet certain financial objec- tives, namely that revenue should cover all operating costs and yield a certain return on revalued assets 26/ or generate funds to finance 25% of 25/ The long run marginal energy costs (Table 6.5). Ideally, the tariff level should reflect specific variable costs of each system, but a uniform rate for the thermal-based systems may be more acceptable. 26/ Under IBRD Loan 596-ET (1969) for the Finchaa Project, EELPA was required to comply with a 7% rate of return on non-revalued assets. - 106 - capital investment. The self-financing condition generally dominates the other conditions with the large investment program required. The com- puted tariff levels required to meet the stated objectives are about 29% higher than marginal cost tariffs if applied across all the power sys- tems. However, the levels for consumer related charges on the EELPA ICS would be 49% higher than marginal cost tariffs if the incremental burden of the financing requirement was carried by this system alone. The con- sultants have formulated their recommended tariff levels according to the latter financial requirements. The consultants also have incorporated an allowance for anticipated inflation between 1983 and the time of imple- menting the tariff increase (7%), and a further increase (13%) to allow for a three year interval between the next increase and the subsequent increase. The consultant's proposed financial tariffs also are shown in Table 6.6. The recommended tariff structure is according to economic costs. 6.13 The consultants have based their estimates of tariff levels on EELPA's optimistic capital investment program, which amounts to about B3 billion for 1984-93 in 1983 prices. It is unlikely that this level of resources could be made available for power sector investment. The consultants used a demand projection based on the optimistic forecast growth rate in demand for the EELPA ICS derived by EELPA's engineering consultants (Acres). The investment program required to meet the Bank's demand forecast (Chapter III) is substantially lower than the program used by the consultants, and it should be possible to gather the finan- cial resources to meet this projected demand. The consultants tested the sensitivity of their proposals to the lower demand/investment scenario, and they concluded that EELPA's position would be only slightly better in this case, and this would not warrant a significantly lower financial tariff than the one they propose. 6.14 Implementation Strategy. The consultants have recommended tariff levels in their financial analysis on the assumption of a three year interval between tariff increases according to projected rates of inflation, since EELPA consider it unrealistic to assume that tariffs will be allowed to increase at more frequent intervals. Nevertheless, in economic terms a policy of relatively frequent and minor tariff increases is more sensible than occasional large increases, since it permits the supply agency to monitor demand more accurately and thereby adjust its investment program to match demand more closely, and it results in a less dramatic impact on customers at the time of the increase. 6.15 The projected existence of substantial surplus hydro generating capacity following the commissioning of the Malka Wakana project (Chapter III) raises the issue of pricing according to short run marginal costs. According to the Bank's demand forecast, there will be a surplus of about 700 GWh/y and 100 MW in 1988 which will decline steadily to zero by 1994. There will be an opportunity to offer a low tariff for electric boiler use provided that it is clearly understood by all parties that this facility only lasts while sufficient surplus capacity is available, and that it is not taken to be the signal for extending the boiler electrification program. The particular circumstances which would - 107 - justify this policy are that electric boilers are already installed so that the investments are sunk costs, and that non-electric boilers are maintained in working condition in the same plants so that the electric boilers can be disconnected when the surplus capacity is reduced. Under these conditions, the demand from electric boilers is not relevant for system planning purposes and, therefore, it has not been included in the Bank's forecast. The economic benefit from this facility would be savings in fuel oil costs, and the electricity tariff for this special category should be set to provide an incentive to switch over to using electric boilers on the basis that capital investment is a sunk cost. The tariff should not be set sufficiently low to justify the investment in new electric boilers, due to the short-term nature of the surplus power and the long-run marginal cost of power supply which makes electric boilers less economic than alternatives. The mission considers that a tariff of about 5.5e cents/kWh in 1983 prices would give the correct signal to consumers. 6.16 A particular feature of the domestic market for electricity in the main urban centers is the existence of "meter landlords" who are consumers that unofficially supply power from the public supply system from their own metered connections. EELPA estimate that tens of thou- sands of consumers are supplied through these landlords. This system is open to abuse by the landlords through exploitation of a local monopo- listic position by charging rates that are much higher than EELPA's tariffs. However, there is no recorded information on the on-selling rates, so it is not possible to reliably assess the extent of such abuse. In contrast, the meter landlords perform a useful service by re- ducing distribution and billing costs, and by making electricity afford- able to many consumers who cannot afford EELPA's connection charges. To prevent excessive abuses, EELPA should impose a ceiling on the rate that meter landlords could charge by making individual connections more af- fordable, for example, by offering long-term credit for connection charges and by treating the cost of meters as an overall distribution cost rather than charging this to the cost of individual connections. 6.17 The possibility of introducing a "lifeline" rate for the first few kWh of monthly domestic consumption is examined by the consultants in response to GOE's desire to provide cheap electricity to the poorest con- sumers. There are strong economic arguments against such a rate, for example: (a) it will involve a heavy subsidy to cover thermal energy costs outside the main ICS, and also since consumption by these consumers is predominantly from lighting and thus coincides with the period of sys- tem peak demand, and as a result advances the timing for investment in new capacity in all systems; and (b) it could stimulate a surge in demand for new domestic connections from consumers presently served by meter landlords which would be costly to satisfy and could well be beyond EELPA's capacity to service. The economic advantage of a "lifeline" rate lies in the social benefits to the poor section of the community. The consultant's propose a compromise in which the first 7 kWh of monthly consumption would be free, but there would be a minimum charge equal to the service charge. The monthly consumption limit would only allow the use of about 60 watts of lighting for about 4 hours per evening. On - 108 - balance, the mission is sympathetic to the need for making electricity available to the poorest section of the community but considers that there are better ways of achieving this than the consultant's propos- als. Therefore, the mission recommends that alternative approaches be considered. 6.18 The mission does not consider that the case for increasing tariffs arises in any way from inefficiencies in EELPA's operations. There is scope inevitably for improvements, but the case for tariff in- creases rests on the economic costs of meeting increases in demand, and not on present operating practices. The Mission notes, however, that there is scope for improvements in operating efficiency, such as the maximization of hydroelectric energy output on the main ICS, and re- duction in fuel utilization in isolated diesel stations. 6.19 Summary. The discussion on electricity pricing can be sum- marized as folLows: (a) Tariffs on the EELPA SCS and ERESA ICS are about 25%, and those in the ERESA SCS about three times higher than tariffs on the EELPA ICS. (b) Economic costs of supply on the hydro-based EELPA ICS differ little from the present tariff level, whereas they are about three times EELPA SCS and ERESA ICS, and twice ERESA SCS, tariff levels. (c) The mission concludes that: - the present tendency towards uniform national tariffs is leading to severe distortions in energy pricing and to economic inefficiency. - the declining block tariff structure now in place is inap- propriate because of its promotional nature in the face of impending capacity constraints and because it does not accurately reflect the distribution of costs. - the GOE's intent to make EELPA financially autonomous is correct and hence EELPA's financial requirements should be a key determinant of the level of tariffs on the EELPA ICS. (d) Thus, the mission recommends that: Table 6.6: COMPARISON OF EXISTING AND PROPOSED TARIFFS a/ b/ EELPA ICS EELPA SCS ERESA ICS ERESA SCS Existing Proposed New Tariff Existing Proposed Existing Proposed Existing Proposed Consumer Category Tariff Economic f/ Financial g/ Tariff New Tariff h/ Tariff New Tariff h/ Tariff New Tariff h/ Domestic (Birr/kWh) First 25 kWh/month 0.15 ] 0.15 ] Lighting 0.24 Lighting 0.37 Next 26 to 100 kWh/m 0.17 ] 0.18 ] Power, First 500 kWh/m 0.19 ] Power 0.25 c/ 0.608 Remainder 0.12 ] 0.159 0.271 0.13 ] 0.608 (Massawa 0.17) ] 0.542 Off-Peak 0.09 ] 0.13 ] Remainder 0.15 ] Commercial (Birr/kWh) First 50 kWh/month 0.15 ] 0.19 ] (as for (as Next 250 kWh/month 0.18 ] 0.150 0.271 0.22 1 0.608 domestic) 0.542 above) 0.608 Remainder 0.13 ] 0.18 ] Street Lighting (Birr/kWh) All consumption 0.11 0.145 0.262 0.15 0.618 0.428 0.37 0.618 Small Industrial (Birr/kWh) First 1000 kWh/month 0.20 0.24 (see Next 2000 kWh/month 0.15 0.19 below) Remainder 0.10 0.14 Peak 0.474 d/ 0.863 0.913 0.661 0.913 Off-Peak 0.059 d/ 0.107 0.526 0.441 0.526 Low Voltage Large Industrial Maximum Demand Charge: First 50 kW 12 10 10 (Birr/kWh/month) Next 200 kW 10 Remainder 8 8 Energy Charge: First 200 kWh/kW 0.09 First 10000 kWh/month 0.13/kWh First 100000 kWh/m 0.09 (as (Birr/month) Next 200 kWh/kW 0.08 Next 10000 kWh/month 0.12/kWh Remainder 0.06 above) Remainder 0.07 Remainder 0.11/kWh (Birr/kWh) Peak 0.477 d/ 0.863 0.913 0.661 0.913 Off-Peak 0.059 d/ 0.107 0.526 0.441 0.526 High Voltage Large Industrial Maximum Demand Charge: First 200 kW 10 (Birr/kW/month) Next 800 kW 8 Remainder 6 Energy Charge: First 200 kWh/kW 0.075 (Birr/month) Next 200 kWh/kW 0.06 Remainder 0.05 (Birr/kWh) Peak 0.332 e/ 0.601 0.541 Off-Peak 0.054 i/ 0.098 0.415 a/ Proposed tariffs by Coopers & Lybrand in Tariff Study (August 1983). b/ Service charges, minimum charges and power factor penalty clauses are additional to these tariffs. c/ Tariffs vary considerably between ERESA SCS towns. d/ Average tariff 0.097 Birr/kWh. e/ Average tariff 0.088 Birr/kWh. f/ At mid-1983 prices yL Required rates to meet rate of return on assets and self-financing criteria, with allowances for a three-year interval between tariff increases. h/ Based on marginal economic costs in mid-1983 prices with allowances (22%) for a three-year interval between tariff increases. - 110 - - the leveL of tariff on the thermal-based ERESA ICS and SCS, and the EELPA SCS be increased immediately to the level of direct costs per unit, implying about a 250% increase, and be further increased to the level of full costs within several years. - that tariff levels on the EELPA ICS be increased by 30- 40%, both to provide revenue for power system development and to balance somewhat the sharp tariff increase required on the smaller systems. - that the flat tariff structure be applied with some time- of-day metering imposed on larger industrial consumers. The highest priority in the short-term is to increase tariff levels because of important conservation and interfuel substitution measures that are dependent on correct economic pricing, and because of the urgent need to establish investment funds to finance the required power system expansion. - Ill - VII. INVESTMENT, MANPOWER AND INSTITUTIONAL REQUIREMENTS 7.1 This chapter covers the means by which policies, plans and in- vestment programs recommended elsewhere in the report can be refined and implemented. The COE has recently drafted a preliminary ten-year invest- ment plan which will need further refining and priority setting. The mission's findings hopefully complement this preparatory phase, which should culminate in a final set of energy sector policies and plans some- time in 1984. The mission also reviewed the institutional arrangements necessary for efficient planning and resource allocation in the energy sector, as well as manpower training and development needs. Energy Sector Investment History and Plans 7.2 Since 1973, investment in the energy sector has been minimal. Total capital investment in the past three years has ranged from 9-12% of GDP, and investment that can be construed as directly relating to energy has been barely 5% of that. The main reason for such a low commitment of funds to the energy sector is that, since the commissioning of the Finchaa hydropower project in 1973, there has until now been no need to invest heavily in power generation expansion in the main EELPA ICS. Similarly, the commitment of funds to energy forestry has greatly lagged the awareness of need for investment. Recent and proposed energy sector investment levels are identified in Table 7.1. The level of commitment to the energy sector is greatly influenced by the mission's somewhat ar- bitrary definition of energy forestry, which by name was not to begin substantially until 1984. The data on future GOE plans for energy sector investment are unofficial and tentative, representing the sum of invest- ments proposed by each energy-related government agency to the Central Planning and Supreme Council rather than an approved program of expendi- ture. It does reflect, however, the interests of the administration and reveals a heavy bias towards modern sector energy supply through invest- ment in power and petroleum. The proposals advanced to the CPSC for energy amount to about B12 million for 1983/84-92/93 and may be compared with the total cost of proposals made of B42 billion, and with the CPSC proposed target for public expenditure for the decade of B32 billion. Hence, energy sector investment proposals are 28% of the total, and 37% of the proposed investment ceiling. Given less optimistic assumptions concerning GDP growth (4% vs. 7.5%) and likely investment Levels (given major constraints on both domestic and foreign capital mobilization), it is the mission's view that total public sector investment is unlikely to exceed B20 billion (US$10 billion) for the decade and hence the funds available to the energy sector will be substantially reduced from the levels proposed, perhaps being of the order of three to four billion Birr, or about 40% of the expenditures now proposed to the CPSC, and 15- 20% of the likely total domestic investment. The mission's estimate of investment requirements by subsector add up to a total that is within this range (Table 7.2). - 112 - Future Investment 7.3 A review of previous chapters indicates that the main require- ments for future investment are to meet the needs for electric power, to greatly expand household fuel supplies, and to explore, develop, and improve the economy and efficiency of existing petroleum supply. The bulk of the capital required is for electric power generation and trans- mission. However, the growth in power demand, and in turn, the demand for investment capital in the power sector, is closely related to the investments in new industrial capacity which constitute 30% of the new power demand now forecast. Similarly, household power demand is somewhat influenced by the rate of interconnection, and hence by investments made in transmission and distribution during the period. Table 7.1: Average Annual Investment by Energy Subsector (Birr million) Preliminary Likely Historical Budget for 10-Year Plan Scenario (1979/80-1981/82) (1982/83) (1983/4-1992/3) (1983-92/3) (Percent) Electric Power 32.1 42.5 528 182 51 Petroleum 12.6 9.2 126 85 22 Energy Forestry a/ 27.0 36.0 75 77 21 Other 2.1 1.5 106 20 6 Total 73.8 89.2 835 364 % of Total Public Investment b/ 4.8 5.6 28 15-20 100 a/ Based on assumption that 25% of industrial plantations, 50% of protective plantations and 10% of community forestry is fuelwood investment, all at an average of 81,000/ha. Fuelwood plantation development per se begins in 1984. b/ The 1983/84-1992/93 GDP growth rate impi city assumed is 4% average. Total capital investment as a proportion of GDP begins at 13% and rises to 20%. Source: CPSC, ENEC and mission estimates. It is clear, then, that the demand for power can be managed according to the rate at which new industrial, commercial, and residential loads are committed and that investment in power supply becomes very much a ques- tion of what the country can afford and of what is economically justifi- able. This situation is rather different from that of household energy where for rural areas, and for low-income urban dwellers, an increased supply of cooking fuels is a basic need: even a matter of survival. In addition, the dynamic interplay between afforestation and agricultural production here demonstrated makes investment in rural energy suppLy fundamental to the well-being of the rural sector and essential to sus- tained social and economic development. These factors should be taken - 113 - into account in setting priorities as well as the fact that prospective returns on rural woodlots and agroforestry are as high as 80% when the non-fuel benefits of trees are taken into account; this is higher than the returns anticipated from any other energy sector investment besides some small projects in industrial energy conservation. 7.4 The question of priority aside, the rate at which funds can be disbursed is an important determinant of the respective level of invest- ment in power and petroleum supply, as well as in rural forestry, parti- cularly in the short-term. EELPA and the EPC could, within the decade, identify and manage through to the commissioning stage major capital- intensive investment in virtual enclave projects such as pipelines, refineries and hydropower stations, disbursing two or three billion Birr in the process. However, the institutions, manpower and infrastructure do not yet exist to disburse funds at the same rate in forestry -- rang- ing from clumps of trees on peasant farms to 20,000 ha plantation blocks across the whole of Ethiopia. Recognizing this, the mission has had to draw a distinction between a "need" to invest about two billion Birr and a "capacity" to disburse, at best, 800 million Birr in afforestation in the next decade. In summary, the overall level of investment in the energy sector in Ethiopia is a function, on the one hand, of economic growth, the rate of private and public saving and the level of foreign assistance, and on the other hand, of the capacity of the administration and a much constrained private sector to cope with the required rate of growth. There is also a need for energy sector investment to be allocated a much greater share of public sector investment than ever before, and to recognize explicitly in determining this proportion the contribution to GDP to flow from stabilizing rural productivity. The mission believes it reasonable that energy sector investment should com- prise up to 20% of total public sector investment. The relative priority of investment within the energy sector is a function of the pattern of future growth by subsector; the rational mix of energy forms to meet demand; and the rate at which it is possible, taking account of short term constraints on implementation capacity, to disburse funds in pri- ority areas. The mission accords highest priority to investment in the supply of household energy of all kinds and in particular to rural agro- forestry. 7.5 The mission's estimates of energy sector investment require- ments during the decade are of the order of B360 million per year (aver- age), or approximately B3.6 billion for the 1983/4-1992/3 period, with an estimated foreign exchange component of about 60%, or US$1 billion. More than half of the total will be in the power sector and about one-fifth in various forms of energy forestry. The mission's estimates are about half those of the EELPA ten-year plan proposals for power development but accord well with proposals for forestry investment by the respective agencies. Of course, with the exception of power, investments are heavi- Ly geared to the last years of the decade as the investment program gains momentum. A year-by-year schedule of investments by energy subsector is provided in Appendix 7.1. The single most important observation is that, despite the scale of the energy forestry program -- to plant 1.2 million ha in ten years -- it does not impose an intolerable financial burden, - 114 - consuming only about 4% of the funds estimated to be available for public sector investment. A summary of the key investments foreseen by pr.iority class is provided in Table 7.2. Table 7.2: MAJOR NEW INVESTMENTS BY PRIORITY CLASS Estimated Cost Project or Program (million 1983 Birr) High Priority Short-term (1983-86) Household Fuels Rural Energy Forestry and Housing 364 Peri-urban Forestry 403 Charcoal, crop residues and cooking efficiency 152 Addis Ababa plantation upgrading 36 Power Sector Finchaa Hydropower Generation (for 1988/9) 47 Diesel Displacement (small hydro, and at Asmara by fuel oil) 54 Transmission, Phase I 174 Petroleum Natural gas reserve confirmation (CALUB 2/3) 17 Conservation and Interfuel substitution 38 Longer-term (1987-92) Power Sector Next major project (Tis Abai, Gibe Daka, AleLtu, Chemoga Yeda, Natural gas) 695 Transmission, Phase II 200 Diesel Displacement (with small hydropower) 40 Petroleum a/ Gas or product pipelines - 642 Detailed gas utilization study and delivery system design 10 Lower Priority Longer-Term (1987 and beyond) Power Sector Certain transmission projects 70 Geothermal power (Northern region) 50 State farm power systems (gasifiers) 10 Petroleum Sector Refinery modernization 135 Household Sector Rural cooking efficiency program 10 Other (Malka Wakana and numerous smaller investments, mostly power sector) 493 Total 3,640 a/ Depending on the outcome of further detailed studies on the market for natural gas from the Ogaden and transport options for the Addis Ababa-Asseb transport corridor. Source: Mission estimates. - 115 - Sources of Finance 7.6 The investment program outlined herein can only be funded by a massive mobilization of local and foreign resources. Unless local funds are generated through correct pricing and appropriate tax measures, it will be extremely difficult to provide sufficient counterpart funds to attract foreign financial resources. The onus is clearly on the Govern- ment to indicate its preparedness to embark on a self-help program by generating the required local funds through the implementation of recom- mendations made elsewhere in this report regarding price increases for petroleum products and electricity, and through the levying of adequate resource taxes in commercial energy forestry and the like. Grant aid has averaged US$62 million a year, or only $2 per capita, for the last three years, making Ethiopia one of the lowest recipients of official develop- ment assistance in the world. By comparison, in 1981/82 aid transfers amounted to US$20 per capita for the rest of Africa. During this period, aid to the energy sector has been received from the EEC, UNDP, Sweden, Italy, and the Soviet Union. The EEC and UNDP co-financed the major geothermal exploration project and the latter has funded design studies for small hydropower as well as policy and planning studies. Sweden funds training programs in forestry, implicitly including energy for- estry, and Italy has funded a number of research and resource evaluation studies including a major nationwide sample survey of household energy use, still being processed. The Soviet Union has financed the Malka Wakana Hydropower project of 150 MW (about US$150 million) and a US$20 million petroleum exploration project. The mission estimates the annual requirement for foreign exchange to be about US$220 million for the energy sector investment program alone: three times more than the total development assistance of recent years. In addition to mobilizing local financial resources, there is a need for very efficient coordination first in establishing priorities within the investment program; and, second, in project preparation and presentation. Similarly, the need for institutional strengthening to cope with the policy, analysis and coordi- nation aspects of defining and delivering the energy sector investment program highlights the discussion on institutional arrangements later in this chapter. Manpower Issues 7.7 Manpower training and development issues in the energy sector are made very complex by the diversity of energy sources and the inherent differences in skills required for their management. The mission concen- trated on the manpower issues raised in consideration of a massive increase in the rate of afforestation, and dealt to a lesser extent with the manpower needs of the petroleum and power subsectors. The special training requirements of many unique components of major programs such as charcoal and briquette production, or of small though important programs such as ethanol, have not been examined, though must now be considered by the COE as part of project or program preparation. - 116 - Manpower for Energy Forestry 7.8 Skilled manpower availability is seen as one of the key con- straints to implementing the desired forestry program. Hence, the forestry program recommended is to some extent the outcome of a review of skill-training capability. The estimated manpower requirements of the achievable energy forestry program for 1983-92 are 120 graduates, 650 diploma holders and 1,560 certificate holders. These can be supplied over the decade at a cost of 13.5 million Birr, including a foreign ex- change requirement of US$4.1 million. The target programs, agencies and timing of the increased output at each level are depicted in Table 7.2. The current source of supply for new graduates is training at overseas universities at a rate of five per year. An increased supply of candi- dates can come from existing secondary education and vocational learning in Ethiopia. At best, these graduates would be available in 1987. Di- ploma holders come now from the Wondo Genet Training College and the training programs there should be expanded. However, another 30 gradu- ates per year should be able to be produced from ongoing agricultural training courses such as in Jimma. Certificate holders are the product of special six month training courses at institutions such as the Farmers Training Centers. Considerable improvements can be made in the present organization of this training, partly by expanding the level of training, and by more fully involving the Wondo Genet training college. Present resources facilitate the training of only 200 candidates per year. In summary, immediate steps should be taken to: (a) increase graduate training overseas from 5 to 20 students per year; (b) increase training at diploma level from 60 to 90 students per year; (c) rationalize and expand the training of certificate holders from 100 to 400 per year. Apart from the quite essential Swedish (SIDA) funding for the Community Forestry College (B3.7 million), no capital outlay will be required to implement this training program, but additional training and technical assistance, especially with the agro-forestry component, will certainly be needed. In this respect, every effort is required to ensure full co- operation and ongoing coordination between the MOA and FAWCDA in respect of the rates to be assigned to them in the implementation of the rural energy forestry program (see Chapter II). Manpower for Electric Power Development 7.9 EELPA have forecast a growth rate of 5.8% p.a. for the decade in their own staff which presently number about 5,300. Skilled staff are required in all categories: managerial, supervisory, technical and op- erational, and although an EELPA training institute has been in operation for some time, the main constraint has been the establishment of a man- power management system to ensure that training is well matched to needs. - 117 - For example, 120 basic technician trainees are graduated each year, with expansion planned to provide 2,300 over the decade compared with a maximum requirement of 800-900. Similarly, there are plans to upgrade the institute to a college for electric and electronic technicians though the market for these skills is hard to identify. The mission favors, instead, reorienting the training institute to cater more to in-house training; and provide refresher and skill upgrading courses for the present staff. Rather than train from the beginning in the basic technical skills it should be possible for EELPA to select new personnel on the basis of previous basic training and experience, and to focus just on those skills needed for its own particular operations. There is a need for an in-depth manpower assessment, leading, among other things, to a detailed staff development program and a clear determination of the responsibilities of the existing EELPA Manpower Planning Unit and the EELPA Training Institute. Manpower for Petroleum Supply 7.10 The present staff of the EPC number about 930 and further growth is dependent upon the expansion and upgrading of refinery require- ments and the corporation's particular responsibilities in the petroleum sector. Since the mission does not foresee the construction of a second refinery during the next decade, the focus of manpower training must be on the existing refinery at Asseb where there are high vacancy rates for general and technical managers, specialist engineers, secretary-typists, and financial personnel. There is a special problem of living and work- ing in Asseb where temperatures frequently reach 40°C and amenities are few. The EPC has attempted to provide incentives through increased sal- aries though without noticeable success. Clearly these incentives will have to be increased and should include attractive means of upgrading professional skills with a wider and more appropriate selection of on- the-job training programs in the refineries of other countries, based possibly on an exchange scheme. Table 7.3: TARGET ACTIVITY, TIMING AND COST OF SKILL TRAINING REQUIRED FOR ACHIEVABLE ENERGY FORESTRY PROGRAMME a/ Graduate Diploma Certificate 1982-87 1987-92 Total 1982-87 1987-92 Total 1982-87 1987-92 Total Peri-urban 36 21 57 67 59 126 180 510 690 Rural 3 6 9 19 38 57 227 459 686 FAWCDA (State Forests) 15 19 34 65 89 154 69 81 150 SWCD 12 10 22 116 93 209 7 14 21 Other 6 5 11 58 46 104 4 7 11 Total 72 61 133 325 325 650 487 1,071 1,558 (million Birr) Estimated Costs 4.61 3.90 8.51 1.95 1.93 3.88 0.33 0.73 1.06 a/ "Achievable" program includes 960,000 ha equivalent in agroforestry and rural woodlots and 195,000 ha of peri-urban and state farm piantations. Source: Mission estimates. - 118 - The existing training department has no professional staff despite many attempts at recruitment. Again, incentives should be increased to the level required to fill two of these posts and a training workshop should be established. The training unit should not only conduct in-plant re- fresher and new skill training but should also organize supervising and middle-management training, and be in charge of the scholarship program. Foreign assistance could be sought to equip the workshop and to supply teaching materials. Institutional Arrangements For Energy Planning and Management 7.11 The Central Planning Supreme Council has been preparing a set of development objectives for all the economic sectors, including energy, for each of the short-term plan periods. The Council also has been setting up development strategies which include corresponding investment programs for priority projects. These planning activities are carried out in constant consultation and collaboration with the implementing agencies. In this way the Central Planning Supreme Council plays a cen- tral role in coordinating and setting priorities in the energy sector. Energy investment priorities already have been identified within the framework of the ten-year indicative plan. However, the current institu- tional arrangements are not optimal for effective energy planning. This fact is known to the GOE which has commissioned the Ethiopian National Energy Committee (ENEC) through the Ministry of Mines and Energy to propose, by mid-1983, an alternative arrangement. The present arrange- ment places the Ministry of Energy and Mines as the lead agency in policy formation and development planning for the sector with ENEC initially operating as a planning secretariat reporting to the minister. ENEC is provided with a general secretary and staff. The energy parastatals, including EELPA (with ERESA), EPC, and the agencies established for geothermaL and petroLeum exploration variously report to the Permanent Secretary of the Ministry or to the Minister himself. ENEC is technically comprised of the permanent heads of all of these agencies plus other energy-related ministries meeting under the chairmanship of the Minister or the Permanent Secretary for Mines and Energy. However, since its formation in 1978, this committee has met only a few times, which is indicative of the lack of meaningful cooperation between the implementing agencies and the ENEC. In practice, the energy production and distribution agencies plan their investment programs independently of ENEC and usually address and deliver policy guidelines for their sub- sectors directly to the Central Planning and Supreme Council (CPSC), in parallel to ENEC, taking little heed of any guidance ENEC offers. The CPSC is neither well-equiped to judge the relative merits of each subsector proposal nor to undertake independent analyses to verify either conclusions drawn or the macroeconomic impact of major projects or programs. The result is that analyses to determine the relative priorities for investment in the energy sector are minimal; that supply and demand forecasting and resource evaluation is superficial, uncoordi- nated and contradictory; and that monitoring and integration of sector- wide trends in consumption and costs of production and delivery of each fuel type are inadequate. - 119 - 7.12 There are a number of choices available to the GOE, all of which should be addressed first in relation to specified goals and objec- tives. Final decisions to do with energy pricing, other major policies, and investment will be made ultimately at cabinet level. However, the energy sector proposals before the cabinet must be well supported with analysis, and must reflect the available options and their relative costs. The matters to be resolved are to do with which agency performs this task and what kinds of tools and authority it needs to be effec- tive. Presently it is more the status and discipline of ENEC that is deficient than the statutory framework in which it is placed. There is a need for a small, powerful and talented secretariat, perhaps with direct responsibility to the cabinet, but whether this resides within the Ministry or is within an office of the cabinet, or Linked to CPSC, is an open question. Furthermore, it is more important for this secretariat to be in a position of influence in the cabinet than for it to have any direct authority over the various energy agencies and parastatals, whose responsibility to carry out operations and execute prospects should be as independent as possible within broad policy and investment guidelines. There is a need for the authority to command and to receive information of all kinds from the parastatals, and a need for the resources to gather and process such information in the required time. However, the manage- ment of parastatals must be independant, answering through their respec- tive boards to the parent ministries in relation to agreed goals, objectives and principles of operation in respect of which the energy secretariat would be regularly consulted. The status of the secretariat should develop through its position of influence, and out of respect for its independence, objectivity and analytical skill. If consultation in the form of a board or committee is seen as desirable, it should be to keep the secretariat abreast of the needs and opinions of consumer groups within the community influenced by, and influencial in determining, the sector plans and policies it formulates and proposes to the cabinet. Here, the mission favors a small group composed of independent adivsors rather than permanent secretaries or heads of parastatal authorities in respect of whom concern might only naturally arise regarding objectivity of judgment. The Role of the Secretariat 7.13 It is possible to illustrate by example the role of an energy secretariat. Take the case of preparation of demand forecasts and plan- ning in the power sector. A key function of an energy secretariat would be to provide EELPA with guidelines on the current economical and soci- ally justifiable uses of electricity in each end-use sector, as well as review power demand forecasts and least-cost development proposals pre- pared by EELPA or its consultants. For example, in the present market the secretariat would be advising on the economic and strategic place of electric boilers and electric cookers. The secretariat also should be readily able to make an objective well-documented assessment of the rela- tive priority within the energy sector of the EPC's refinery and pipeline proposals. At no time, however, should there be any confusion about the respective roles of existing agencies (such as EPC, EELPA, FAWCDA, MOE) and an energy secretariat: executing agencies should implement according to agreed priorities and plans and the secretariat should monitor, inte- grate, analyze and advise. - 120 - Production and Marketing of Biomass Fuels 7.14 While the responsibilities for electricity production and petroleum supply are clear and established, the responsibility for the production, distribution and marketing of biomass fuels is less so and the institutional arrangements are embryonic. The supply and marketing of firewood and charcoal recently have been made the responsibility of the Fuelwood and Charcoal Marketing Corporation (FCMC) of FAWCDA. At present, FCMC supplies 5-10% of the charcoal and 20-30% of the firewood to Addis Ababa. While its current activity is mostLy confined to 50km around Addis Ababa, its jurisdiction is national. The FCMC's pronounced objective is to supply the whole of the urban market within ten years. During the late 1970s the supply of woodfuels deteriorated rapidly, ac- companied by a sharp escalation in the open market price of firewood and charcoal. The government was concerned that a few entrepeneurs might illegitimately be exploiting this situation to increase their profits and that charcoal production in savannah lands was responsible for major de- forestation. Hence, in 1980, private supply of charcoal by private entrepreneurs was made illegal except if carried by donkey. Now, there are at least 2,500 donkeys carrying charcoal (besides about 10,000 carry- ing firewood) to the city on market days plus about 1,000 trucks and cars carrying from l to 15 sacks in addition to their main cargo, as well as Large numbers of individuals. Some 90% of the total Addis Ababa supply of charcoal is transported in this way. Similarily, firewood can only legally be trucked to the city by FAWCDA, though it is clear from survey results that many non-FAWCDA supplies exist. The FCMC is intent on mar- keting charcoal and firewood directly to consumers through its own depots, of which it has 34 in Addis Ababa and nine in surrounding towns and cities. In the longer-term there is an obligation under government proclamation to retail these products through Kebele stores just as are sugar, tea, matches and the like now. 7.15 While the FCMC has little difficulty marketing its firewood, its charcoal, which is sold only in 40 kg average bags, is difficult to sell, despite being one-third to one-half the price of charcoal traded on the open market. FAWCDA and the mission agreed that this is due to a combination of factors including parcel size, depot opening times, char- coal quality, perception of the difficulty of transactions, and so on. However, the real question is whether or not FAWCDA should be involved in marketing these products at any level. The appropriate response to the perceived situation in which a few private entrepreneurs hold a large proportion of the woodfuels market, and generate possibly huge profits, would have been to increase competition rather than to attempt complete government takeover and regulation. It is likely that the restrictions on private supply have worked to increase the cost of the woodfuels, which still reach the market in abundance from private sources, due to inefficient use of transportation and to premiums charged to offset the risks in illegally supplying the market. Furthermore, with the proposed huge increase in charcoal supplies, the production and supply of crop residue briquettes, and a much increased supply of wood fuel from peri- urban plantations, the roles and responsibilities of each agency con- cerned need to be carefully reevaluated. The mission sees a number of options open to the government though all have in common: - 121 - (a) the need for government agencies to keep the retail marketing of woodfuels open to, and preferably in the hands of, private entrepreneurs to ensure adequate and efficient distribution; and (b) the need to maximize the involvement of the private sector in all stages of production, transportation, and wholesaling of biomass fuels, so as to ensure the maximum rate of resource mobilization. It is acknowledged that FAWCDA and the Ministry of State Farms have a legitimate need to closely manage, and possibly to undertake the conver- sion of forest logging and crop residues respectively. However, from the edge of the forest, or from the farm gate to the market, unhindered private sector involvement is desirable. Similarly, with respect to har- vesting and marketing the product of peri-urban plantations, small-holder entrepeneurs should be included. It is noteworthy that at least 20,000 people are employed regularly in the transport phase alone of woodfuel supply to Addis Ababa. It will prove exceedingly difficult and expensive to duplicate this network of supply and delivery to every tiny market- place. The COE could consider instead, a wholesale facility for the new biomass fuels generated by either itself or major contractors, whereby traders or retailers at all levels could bid for fuels of stated quality for retailing throughout the city. Here, the emphasis should be on strengthening marketing through regulation of quality and package size so that the market is better able to set fair prices. Coordination in the Petroleum Sector 7.16 The EPC has responsibility only for petroleum refining and the regulation of additional supply through the major oil companies. Explo- ration for petroleum is ostensibly the responsibility of the Ethiopian Institute of Geological Survey (EIGS) which, in turn, is part of the Ministry of Mines and Energy. Within the EIGS there are three explora- tion projects operating quite independently: the oil and gas project which oversees the joint Ethiopian-Soviet oil exploration program (SPEE), the Bank-funded petroleum exploration and production project; and the UNDP/ECE geothermal exploration project. The SPEE coordinator reports directly to the Minister of Mines and Energy first. This rather frag- mented arrangement for exploration does not appear conducive to rational resource evaluation and development, and consideration should be given to its rationalization under one program management. Second, it is obvious that if natural gas is developed or oil is discovered, the entire plan- ning for petroleum refining and supply will be greatly altered. It is preferable to have the rather dynamic developments in the petroleum sector reviewed and coordinated by one organization. The mission con- siders that one option is to expand the charter of the EPC, giving it full responsibility for investigations towards the production of natural gas, for local use. This rearrangement would require a substantial expansion in the professional staffing of the EPC, including additional executives and economists, lawyers and production managers, none of which are being trained as part of the existing IDA petroleum exploration pro- - 122 - motion project. The implications of maintaining the present institution- al arrangements in the petroleum sector are that refinery and petroleum product transport, planning and investment would not adequately take account of the prospects for production and local use of known and pros- pective petroleum resources, leading possibly to inefficient use of scarce manpower resources and to expensive duplication of, or underutili- zation of petroleum subsector infrastructures. Annex 1.1 ETHIOPIA: ENERGY BALANCES, 1982 (103 x TOE) Primary Energy Secondary Energy Traditional Non-Traditional Petroleum Products Fuel- Crude Char- Electri- Diesel Fuel Line wood Dung Straw Bagasse Hydro Oil coal city LPG Gasoline Avgas Jetfuel Kerosene Oil Oil Others Total Total I-Gross Supply 3403.3 2539.2 1803.6 58.6 187.3 738.4 14.1 5.8 24.7 11.2 92.8 8.9 157.5 8877.9 1.1 Production 3403.3 2539.2 1803.6 58.6 187.3 1.2 Inports 773.3 9.9 9.0 52.2 11.2 93.5 175.8 1.3 Primary Exports 1.4 Stock Changes 44.9 4.2 (3.2) (27.5) 0.7 8.9 (18.3) Conversion: 2-Petroleum Refinining (702.8) 5.4 108.3 57.5 0.3 212.9 303.2 15.2 0.0 3-Charcoal Production (101.8) 101.8 0.0 4-Elect. Power Generation (11.7) (187.3) 267.9 (48.8) (20.1) 0.0 5-Conversion Losses (343.8) (17.6) (25.6) (195.5) 582.5 6-Power T and D Losses (8.5) 8.5 7-Non-Energy Use 15.2 15.2 Net Supply 2957.7 2539.2 1803.6 29.3 101.8 63.5 5.4 122.4 5.8 82.2 11.5 256.9 292.0 0.0 776.2 8271.3 8-Secondary Exports 213.5 213.5 213.5 9-Bunker Sales 30.1 2.9 9.1 42.1 42.1 10-Domestic Consumption 2957.7 2539.2 1803.6 29.3 101.8 63.5 5.4 122.4 5.8 52.1 11.5 254.0 69.4 520.6 8015.7 10.1 Industry a/ 29.3 0.8 35.0 0.2 0.2 34.3 67.2 101.9 167.0 10.2 Transport 122.4 5.8 52.1 168.8 349.1 349.1 10.3 Agriculture 1.1 31.0 31.0 32.1 10.4 Households 2957.7 2539.2 1803.6 101.0 16.3 5.0 11.2 1.6 17.8 7435.6 10.5 Commercial b/ 11.1 0.2 0.1 18.3 2.2 20.8 31.9 a/ - Including construction. b/ - Including Government. Source: Mission estimates - 124 - ANNEX 1.2 Page 1 of 4 Summary of Methodology for Demand Forecasting in the Ethiopian Energy Assessment 1. A mieresectoral, fuel by fuel, approach was used for demand forecasting whereby the greatest level of detail possible was gathered on the present and likely future patterns of energy use in each major eco- nomic activity and manipulated and compiled progressively to provide a composite picture of energy use at a given future date. Only in sub- sectors where insufficient data is available is there a reversion in forecasting either to using conventional GDP-based elasticities or to the use of more global trend data. In practice, the analysis is more rigor- ous for the ntjor petroleum and electricity consuming sectors such as industry where investments underway, committed, or proposed have been closely reviewed, leading to an independent estimate of the likely com- missioning schedule for new industrial capacity under each economic growth scenario. 2. Gene'ral *acro-economic parameters assumed for the projection of energy demand are as follows (adopted from World Bank (CEM, 1983)1: (a) GDP growth: 4% p.a. (b) Agriculture sector, 3.7% p.a.; industry, 5.4% p.a.; services sector 3.7% p.a. (c) Exports are projected at 7.5% per annum in real terms. (d) Imports are projected at 4.7% p.a. 3. The development of the Business as Usual Scenario (BAU) is based on the following treatment of individual fuels: Diesel (a) Diesel use in power generation is to remain constant due to a balance between growth in the SCS and private sector systems, interconnection with the EELPA ICS, and the complete conversion to fuel oil firing in the ERESA ICS. (b) Diesel in industry will grow at the rate of industrial expan- sion forecast, about 5.4% p.a. (c) Diesel in transportation is to grow in proportion to fleet size: 7% p.a. Gasoline/Jetfuel/Avgas Gasoline is assumed to grow at 5% and c.f. 4% between 1977- 82. Jetfuel and Avgas are assumed to grow at 6% p.a. based on information provided by Ethiopian Air Lines. Kerosene an; LPG Growth is prevented due to restriction on imports of bath products and by the present operation of the refinery. - 125 - ANNEX 1.2 Page 2 of 4 Fuel Oil Growth is estimated for each major factory now committed or planned and, in the mission's view, likely to be operational in the period. Cement and textile plants comprise 93% of new demand by 1992. Annual growth rate estimated is 9.2% Electricity Demand growth and consumption are projected in Annex 3.2. Traditional biomass fuels These are projected so as to maintain the supply of 2kg of air- dry wood equivalent per capita per day, but with no growth in forest resources, hence additional fuel supplies will come from agriculture in the form of crop residues and dung. 4. The development of the "Accelerated" scenario is based for pe- troleum end use on subtracting all the interfuel substitution and conser- vation measures recommended in industry for fuel oil and in transport for diesel (see Chapter IV) from the BAU scenario total. Additional LPG and kerosene supplies (55,000 toe total) are then added as recommended to household fuel consumption. For biomass fuels for cooking, all new pro- duction from forests and new biomass fuel sources (see chapter II Table 2.3) recommended by the mission are included and act to reduce the con- sumption of the less desirable agricultural fuels of dung and crop resi- dues. Annex 1.2 Page 3 of 4 ETHIOPIA: ENERGY BALANCES UNDER BUSINESS AS USUAL SCENARIO, 1992 (103 x TOE) Primary Energy Secondary Energy Traditional Non-Traditional Petroleum Product Fuel- Crude Char- Electri- Diesel Fuel Line wood Dung Residue Bagasse Hydro oil coal city LPG Gasoline Avgas Jetfuel Kerosene Oil Oil Others Total Total 1-Gross Supply 3403.3 3981.7 2824.7 117.3 577.0 739.6 91.1 10.4 89.7 434.8 626.0 12269.6 1.1 Production 3403.3 3981.7 2824.7 117.3 577.0 1.2 Imports 739.6 91.1 10.4 89.7 434.8 626.0 1.3 Primary Exports 1.4 Stock Changes Conversion: 2-Petroleum Refinining (714.0) 5.4 108.3 57.5 11.5 212.9 303.2 15.2 0.0 3-Charcoal Production (101.8) 101.8 0.0 4-Elect. Power Generation (11.7) (577.0) 649.1 (48.8) (20.1) 0.0 5-Conversion Losses (343.8) (17.6) (25.6) 471.6 867.1 6-Power T and D Losses (22.6) 22.6 7-Non-Energy Use 15.2 15.2 Net Supply 2957.7 3981.7 2824.7 88.0 101.8 154.9 5.4 199.4 10.4 147.2 11.5 598.9 283.1 0.0 1310.9 11364.7 8-Secondary Exports 102.3 102.3 102.3 9-Bunker Sales 53.9 4.3 13.5 71.7 71.7 10-Domestic Consuaption 2957.7 3981.7 2824.7 88.0 101.8 154.9 5.4 199.4 10.4 93.3 11.5 594.6 167.3 1081.9 11190.7 10.1 Industry A/ 88.0 0.8 97.6 0.2 0.2 58.0 165.1 223.5 351.2 10.2 Transport 199.4 10.4 93.3 332.1 635.2 635.2 10.3 Agriculture 2.2 176.5 176.5 178.7 10.4 Households 2957.7 3981.7 2824.7 101.0 41.8 5.0 11.2 2.1 18.3 9925.2 10.5 Commercial b/ 13.3 0.2 0.1 26.3 2.2 28.8 42.1 a/ - Including construction. b/ - Including Government. Source: Mission estimates. ANWEEX 1. 2 Ethiopia: Energy Ralances Under Accelerated Scenario. 1992 (1I3 X TOF) Primeary Energy Secondary Energy Traditional Non-'radirtioeal Petroleum Produet New Vuels Char Electrg?li-r Ta F1wo 81iaos hao klr oIe IDung Residue Bagasse Hlydro Ve coal 1cityi LPG Gasoline Avgas Jetfuel Kerosene I%tel y Others Total Nuelvoo mass Ethanol Solar Pal i-Gross Supply 3403.3 2867.9 2014.8 132.3 577.0 794.6 67.7 10.4 89.7 388.0 55.8 117Q.5 547.3 23.3 4.1 12,n99.9 1.1 Production 3403.3 2867.9 2014.8 132.3 577.0 1179.5 547.1 23.3 4.1 1.2 Imports 794.6 67.7 10.4 89.7 388.0 555.8 1.3 Primary Exports 1.4 Stock Changes Conversion: 2-Petroleum Refinining (769.0) 12.5 108.3 57.5 59.4 212.9 303.2 15.2 n.0 3-(lurcoal Production (101.8) 101.8 0.0 4-Elect. PuYer Generation (11.7) (577.0) 649.1 (37.4) (23.n) 0.0 5-Conversion Losses (343.8) (17.6) (25.6) (471.6) 858.6 6-Pawer T and D Losses (22.6) 22.6 7l-4bnEnergy Use 15.2 15.2 Pet Supply 2957.7 2867.9 2014.8 103.0 101.8 154.9 12.5 176.0 10.4 147.2 59.4 563.5 280.2 0.0 124q.2 1179.5 547.3 23.3 4.1 11,203.5 8-Secondary Exports 125.0 125.0 125.5 9-Bunker Sales 53.9 4.3 13.5 71.7 71.7 10-omestic Conausption 2957.7 2867.9 2014.8 103.0 101.8 154.9 12.5 176.0 10.4 93.3 59.4 559.2 141.7 I152.5 1179.5 547.3 23.1 4.1 11,006.8 10.1 Industry LI 103.0 0.8 97.6 0.2 0.2 5R.n 139.5 197.9 1.4 401.7 10.2 Transport 176.0 10.4 93.3 296.7 576.4 23.3 599.7 10.3 Agriculture 2.2 176.5 176.5 178.7 10.4 Households 2957.7 2867.9 2014.8 101.0 41.8 12.1 59.1 2.1 73.3 1179.5 547.3 2.7 9,786.0 10.5 Comercial b/ 13.3 0.2 0.1 26.3 2.2 28.8 42.1 a/ - Including conxstruction. b/ - Including Government. Source: Mission estimates. - 128 - Annex 2.1 FOREST RESOURCES (natural forest) A. Natural Forest Type ha 1. Closed high forest a. Coniferous 52,300 b. Montane mixed broadleaved/coniferous 158,500 c. Montane broadleaved 1,863,200 Total closed high forest 2,345,100 2. Disturbed forest a. Coniferous 7,300 b. Montane mixed broadleaved/coniferous 48,300 c. Montane mixed broadleaved 285,300 d. Lowland Broadleaved 61,300 Total disturbed forest 402,200 3. Woodland 357,000 4. Bamboo thicket (Arundinaria) 112,700 5. Thicket and low woodland 121,500 6. Lowland bamboo (Oxytenanthera) 113,400 Subtotal 3,451,900 B. Savannahland (Acacia sp.) 20,000,000 C. Planted Forests 1. Urban plantations 80,000 (of which Addis Ababa has 20,000 ha) 2. State forest plantations 120,000 3. Community Forest Plantations 200,000 Subtotal 400,000 - 129 - Annex 2.2 Page 1 of 2 LOGGING WASTE PLUS NON-COMMERCIAL SPECIES AVAILABLE FOR ENERGY 1981/82 (all solid m3) Log Logging Available Oistance to Transport Region Forest supply Logs area Biomass for energy Addis Ababa, cost (m3 ) (m /ha) (ha) (m /ha) (m3) (Gravel/Asphalt) (Birr/m3 Arsi Munessa 14,000 40 350 250 73,500 30 220 30.20 Arba Gugu 7,000 50 140 300 35,000 70 190 26.10 Bale Adaba- 17,400 30 580 250 127,600 130 310 66.60 Oado a Hararghe Din Din 6,000 40 150 200 24,000 70 200 40.40 Chercher 2,600 30 85 200 14,500 120 230 53.80 Kaffa Tiro-Folla 24,700 50 495 - 300 86,600 FO 300 40.40 Limo-Boter 4,100 30 135 150 16,200 80 120 31.30 Bebeka a/ 17,400 - - - - - - - Shoa Jibat 2,500 40 65 250 13,700 60 140 30.30 Addis Ababa 28,000 b/ - - - - - - - Sidamo Kebre- Mengist 16,300 30 540 250 118,800 160 340 76.30 Magoda 3,800 60 65 300 15,600 10 490 69.10 Wollega Gorgo-Woto 500 40 15 250 3,200 50 330 54.60 Total/Average 144,300 d/ 40 2,620 250 528,700 98.7 202.7 53.30 a/ Clearing for coffee plantation. b/ Eucalyptus around Addis Ababa for chipboard and hardboard. c/ 50% selective cutting. d/ Excluding Addis Ababa and Bebeka 98,900 m3. - 130 - Annex 2.2 Page 2 of 2 For the future, the following scenario can be envisaged: 1. According to the FAWCDA 10-year plan, logging areas will be in- creased from the present 2,600 ha to 4,800 ha by 1987/88 and to 7,700 ha by 1992/93. 2. Utiliz tion will improve from 40m3/ha today to 60m3 by 1987/88 and to 80m /ha in 1992/93 by using more of the at present non-commercial species and also variable log lengths. 3. Selective cutting will increase from 20% today to 30% by 1987 (77 and 3to 40% in 1992/93. Available "waste" will be the same, IOOm'/ha. 4. Average total stand volume will be the same as today, 250m3/ha. With these assumptions, the following calculation can be made: 1987/88: "Waste" from selective cuttings: 3 0.30 x 4,800 x 100 = 144,000 m "Waste" from clearfellings: 3 0.70 x 4,800 x 190 = 638,400 m 3 Total 782,400 m 1992/93: "Waste" from selective cuttings: 0.40 x 7,700 x 100 = 308,000 m3 "Waste" from clearfellings: 0.60 x 7,700 x 170 = 785,400 m3 Total 1,093,400 m3 Annex 2.3 AN iLLUSTRATION OF THE PLACE OF CARBONIZATION IN A PROGRAM OF UPGRADED FOREST MANAGEMENT ------ Year 1 -- Year 2 ------ Month J F M A M J J A S O N D J F M A M J J A S O N D REMARKS Operation Selection of logging area Site inspection Decision on silviculture treatment. Selective clearfellIing Marking of trees. Calculations of volume, area, seedlings needed species etc. Planning of logging tracts, loading points, etc. Roads? Planning of labor, tools, trucks, transportation, hiring, same Building of tracks, landings, improvement of roads Logging of timber _ _ _ _ _ _ _ _ Cutting of fuelwood Charcoaling Transportation _ _ _ _ Measuring of logs Nursery work Site clearing, piting Transport of seedlings Planting Weeding Replanting, weeding July/August, year 3 ._ _ _ _ _ _ _ ___ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ I - - _ _ . - - . , _ _ _ __ _ _ _ _ _ _ _ _ _ _ - 132 - Annex 2.4 Page 1 of 2 SAWMILL PRODUCTION, AVAILABLE WASTE, 1982/83 (volumes in solid m3) Number of Capacity Log allo- Production Location milLs log input cation Lumber Slabs Dust Arsi Arba Gugu 2 8,100 7,500 4,500 1,900 1,100 Bale Adaba 1 3,600 2,000 1,000 500 400 Hararghe Asba Tefari 2 6,700 3,000 1,400 900 600 Din Din 4 20,200 10,000 4,700 3,000 2,000 Kaffa Jimma 6 32,700 24,600 12,200 6,800 4,200 Botor Becho 1 4,500 3,100 1,700 900 700 Bebeka 1 8,400 8,400 4,800 4,200 2,400 Shoa Addis Ababa 6 26,500 16,100 7,700 4,600 8,000 Ambo 1 4,500 2,500 1,000 900 500 Arsi Negelle + Shashamane 5 26,100 12,800 5,900 3,800 2,600 Sidamo Awassa 1 5,000 3,000 1,200 1,100 600 Kebre Mengist + Wodera 7 51,900 29,200 1,400 7,200 5,800 Hagare Mariam 1 11,200 5,000 2,500 1,250 1,000 Wollega Gimbi 1 4,500 2,000 1,000 500 400 Total/Average 39 213,900 129,200 63,600 37,550 25,300 - 133 - Annex 2.4 Page 2 of 2 The following scenario is envisaged for the future: 1. Demand of lumber will increase as forecaste q by Ethiopian Wood Working Corporation (EWWC) i.e. to 235,000 m in 1987/88 and to 336,000 m3 in 1992/93. 2. The sawmill industry will expand accordingly (also as per EWWC plan). 3. Log supply will increase according to the FAWCDA 10-year plan, i.e. to 288,000 m3 in 1987/88 and to 616,000 m3 in 1992/93. 4. Usilization of logs will increase to 60 m3/ha in 1987/88 and to 80 m /ha in 1992/93. 5. Recovery rates will improve to average 50% in 1987/88 and to 60% in 1992/93. Waste will decrease accordingly. With these assumptions, available waste will be: 1987/88 1992/93 Slabs 72,000 m3 123,200 m3 Sawdust 56,600 m3 92,400 m3 Total 129,600 m3 215,600 m3 Annex 2.5 OPTIMUM SCHEDULE OF NEW BIOMASS ENERGY PRODUCTION AND SUPPLY TO SOUTHERN REGION (i.e., all provinces besides Eritrea, Tigral, Wollo and Gondar) Actual Weight of - THOUSAND TONNES OF WOOD EQUIVALENT* - -the Specified Fuel in Use at 1992 ('000) 1984 1985 1986 1987 1988 1989 1990 1991 1992 A. New Biomass Sources 1. Charcoal (i) Logging residues - 25 50 50 400 400 450 500 535 143 (ii) State Farms - IBebeka 53 107 214 427 -- -- -- -- -- 17 Other 10 15 20 25 30 35 40 60 64 (iii) Sawmills - 10 20 30 40 50 50 55 60 16 SUBTOTAL 63 157 304 532 470 485 540 615 659 2. Coffee Residue - 10 20 40 80 120 160 180 191 150 3. Cotton Residues - 10 20 40 80 150 200 250 334 239 4. Cereal Straw Residues - 10 20 40 60 80 100 120 151 108 5. Cereal Stover Residues - 10 20 40 80 160 200 255 211 6. Bagasse - -- - - -- -- -- -- 60 43 7. Sawdust Briquettes - - 3 6 9 12 14 20 29 20 SUBTOTAL - 30 73 146 269 442 684 770 1020 TOTAL NEW SOURCES 63 187 377 678 739 927 1224 1385 1679 B. Addis Ababa Plantation Options Notes: 1. Addis Peri-urban Fuelwood: Plantation area 1,400ha 2,100ha 2,100ha 5,000ha 5,000ha 15,600ha additional to Production '000 twe 90 AFDB project in the period 2. Addis Enrichment Planting: Plantation area Production '000 twe 1,000ha 1,000ha 2,000ha 2,000ha 2,000ha 2,000ha - 10,000ha or half of plantation replanting 3. Upgraded Management '000 twe 113 141 181 163 - additional yield gained of AFDB Project 1,500ha 1,880ha 2,400ha 3,400ha 2,900ha 2,900ha during 1983-92 by adding fertilizer, adopting a 5 year ro ation & achiev- ing 25m mai Overall TOTAL Supply 63 187 377 678 739 1040 1365 1566 1917 * See Annex 2.6. - 135 - Annex 2.6 SUPPLY AND DEMAND FOR ADDIS ABABA AS THE PRIME TARGET MARKET OF THE SOUTHERN REGION ('000 Tonnes of Wood Equivalent -- See Conversion Chart Annex 2.7) 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 Mid-Year Population (6.5% p.a.) 1,543 1,643 1,750 1,863 1,984 2,113 2,251 2,397 2,553 2,719 ('000) Number of Households 309 329 350 373 397 423 450 479 511 544 Projected Demand 1,126 1,191 1,277 1,360 1,448 1,542 1,643 1,769 1,863 1,984 ('000 twe) Projected Supply A. Non-Woody Fuels 1. Kerosene 54 60 70 80 91 91 91 91 91 91 -i.e. addn. 5.2 Ml by 1987 2. LPG 48 55 90 120 132 132 132 132 132 132 -i.e. addn 7,000te by 1987 3. Electricity 38 45 59 84 107 130 136 137 138 138 -(see Annex 3.2) SUBTOTAL A. 140 160 219 284 330 353 359 360 361 361 B. New Woody Fuel Sources 4. Charcoal 63 157 304 532 470 485 540 615 659 5. Residue Briquette 30 73 146 269 442 684 770 1,020 SUBTOTAL B. 63 187 377 678 739 927 1224 1385 1679 C. Other 6. City Agriwastes 50 53 55 58 61 64 67 70 74 78 (factory/mill wastes; dung household trees, grass, etc. D. Plantations 7. Existing Urban Plantations 180 169 158 147 135 130 125 120 100 165 -i.e. 90+75 from 8. New Upgraded B.2. Annex 2.5 Enriched Plantations 113 141 181 163 TOTAL SUPPLY (A+B+C+D) 370 508 619 866 1204 1286 1591 1915 2101 2446 E. Deficit Met By FAWCDA and Others: (i) As woody fuels 499 500 525 483 244 256 52 (ii) As charcoal 257 183 161 11 F. Surplus 146 238 462 Deficit equivalent 27,000 24,440 24,500 17,640 8,710 9,140 1,860 in hectares of Savannah Forest (30m3/ha and 700kg/m3 basic density) per year - 136 - Annex 2.7 Wood Equivalence One tonne of wood air dried 25% moisture content on a wet basis and 20 MJ/kg oven dry is 14.3 MJ lower heating value and is equivalent to: -'- 267 kgs of charcoal * 141 litres of kerosene * 83 kgs of LPG * 200 kg of sawdust briquettes * 712 kg of cereal straw briquettes * 784 kg of coffee residue briquettes * 716 kg of cotton and bagasse briquettes - 400 kWh (substituting injera cooking at av. efficiency of 60%) * about one tonne of dung burned on an open fire * 826 kg of cereal stover (corn, sorghum) briquettes - when burned in stoves in common use in Ethiopia and taking into consideration the lower heating value of the fuels concerned. A-ses 2.8 PERI-URBAN PLANTATTONS (URBAN > 5000 TN8ABITANTS VOLUMES IN SOLID m3 ON BARK) 1982 - - … Estimated Total area denand Additional to be olanted Fuelwood satis- Supply Plantation Plantation Total Incremental Incrtal Incremental Total Cost *** Inrremental by 1997 for Total cost *** Region Population demand faction Needed, 1982 Production Needed Cost ** Cost demand 1987 demand 1992 demand 1997 1982-1992 demand 2002 2002 demand 1982-1997 ('000) ('000 m3*) (X) ('000 m3) (m3/ha/year) ('000 ha) (Bitr/ha) (Birr '000) ('000 ha) ('000 ha) ('000 ha) (R8rr '000) ('000 ha) ('000 ha) (Birr '000) Arsi 83 90 50 45 20 2.3 868 1,996 1.7 2.3 3.1 8,159 4.3 13.7 11,891 Bale 45 48 40 29 18 1.6 868 1,389 1.1 1.3 1.9 5,121 2.6 8.4 7,378 Eritrea 782 852 5 809 10 80.9 924 74,752 31.5 43.3 59.2 198,568 81.1 296.0 273,504 Gem Gofa 48 52 50 26 15 1.7 736 1,251 1.3 1.7 2.4 5,225 3.3 10.4 7,654 Gojam 165 180 30 126 18 7.0 913 6,391 3.7 5.1 6.9 20,725 9.5 32.2 29,399 Gondor 163 178 15 151 12 12.6 736 9,274 5.4 7.6 10.3 26,423 14.1 50.0 36.801 Hararghe 266 290 20 232 15 15.5 924 14,322 7.1 9.8 13.5 42,411 18.4 64.3 59,413 Illubabor 29 32 90 3 20 0.2 868 174 0.6 0.8 1.1 2,344 1.5 4.2 3,646 Kaffa 117 128 90 13 20 0.7 868 608 2.4 3.3 4.4 9,374 6.1 16.9 14,669 Shoa 490 534 25 401 18 22.3 736 16,413 10.9 15.0 18.6 49,178 28.2 95.0 69,933 (excl. Addis) Sidamo 204 222 60 89 20 4.5 868 3,906 4.2 5.6 7.8 19,183 10.2 32.3 28,037 Tigrai 165 180 5 171 10 17.1 924 15,800 6.6 8.9 12.4 41,580 17.0 62.0 57,288 Wollega 88 96 50 48 15 3.2 736 2,355 2.4 3.2 4.4 9,714 6.4 19.6 14,424 Wollo 191 208 10 187 12 15.6 924 14,414 6.4 8.8 12.1 39,639 16.5 59.4 54,885 Total 2,836 3,090 (27) 2,251 (12.2) 185.2 (880) 163,045 85.3 116.7 158.1 477,644 2,942 764.4 668,922 a At 2 kg per person and day, air-dry, 670 kg per m3 3 1.09 m3 per person and year. * Excluding infrastructare and administration. About 100% may be added. *Sa 1982 prices, excluding infrastructure and administration. - 138 - Annex 2.9 Page 1 of 9 Peri-Urban Plantations 1/ Silvicultural Model I, Lowlands, below 1800m a.s.l. Eucalyptus camaldulensis Year Silvicultural treatment 1 Site clearing (cutting of existing bush) Charcoaling (if enough material) Pitting 40 x 40 x 30 cm. sparing 2.0 x 2.5m Planting of potted seedlings Application of fertilizer, 50 gram per seedling (Application of insecticide 2/) Weeding Protection (from cattle) 2 Weeding Replanting of blanks (Application of insecticide 2/) Protection (from cattle) 3 (Weeding 2/) (ApplicatTon of fertilizer 2/) Protection (from cattle) 4-5 Minor checking only 6 Clearfelling 60 m3 Removal of excessive coppice shoots Protection (from cattle) 7-10 Minor checking only 11 Clearfelling 25 m3 Removal of excessive coppice shoots Protection (from cattle) 12-15 Minor checking only 16 Clearfelling 75 m3 Removal of excessive coppice shoots Protection (from cattle) 17-20 Minor checking only 21 Clearfelling 60 m3 1/ Note that the cost data provided in this annex are for direct costs only. Total costs include roading, administration and other infra- structure and are approximately double the direct costs. Total costs for peri-urban forestry have been assumed to average Bl,760/ha in costing the forestry sub-sector programs. 2/ When needed, not standard. - 139 - Annex 2.9 Page 2 of 9 Cost Model I (Shadow values: Labor 0.5 = 1.00 Birr per day, foreign exchange 1.33) Man-days -- Cost ----------- Year Activity required Labor Material Total (Birr) 1 Siteclearing, pitting 64 64 - 64 Tools - - 10 10 Seedlings, 2000 8 8 224 a/ 232 Planting 20 20 - 20 Fertilizer - - 160 b/ 160 Application of fertilizer 16 16 - 16 Weeding 24 24 - 24 Protection 4 4 - 4 Total 136 136 394 530 2 Weeding 24 24 - 24 Replanting 15 15 70* 85 Protection 4 4 - 4 Total 43 43 70 113 3 Protection 4 4 - 4 6 Clearfelling (60m3) 30 30 - 30 Cutting shoots 15 15 - 15 Protection 4 4 - 4 Total 49 49 - 49 11 Clearfelling (75m3) 38 38 - 38 Cutting shoots 15 15 - 15 Protection 4 4 - 4 Total 57 57 - 57 16 Clearfelling (75m3) 38 38 - 38 Cutting shoots 15 15 - 15 Protection 4 4 - 4 Total 57 57 - 57 21 Clearfelling (60m3) 30 30 - 30 a/ 50% foreign exchange b/ 100% foreign exchange - 140 - Annex 2.9 Page 3 of 9 Cost Model I Year Cost Discounted Cost a/ Felling Volume (Birr) (m3) 1 530 530 - 2 113 103 - 3 4 3 - 6 49 27 60 11 57 20 75 16 57 13 75 21 30 4 60 700 a/ At 10%. - 141 - Annex 2.9 Page 4 of 9 Peri-Urban Plantations Silvicultural Model II, Highland Plateaux 1800 - 2800 a.s.1 Eucalyptus globulus Year Silvicultural treatment 1 Site clearing (cutting of existing bush) Charcoaling (if enough material) Pitting 40 x 40 x 30 cm, spacing 2 x 2 m Planting of potted seedlings Application of fertilizer, 50 gram per seedling Weeding Protection (from cattle) 2 Weeding RepLanting of blanks Protection (from cattle) 3 (Weeding *) (Application of fertilizer, 50 gram per seedling ') Protection (from cattle) 4-5 Minor checking only 6 Clearfelling 75m3 Removal of excessive coppice shoots Protection (from cattle) 7-10 Minor checking only 11 Clearfelling 90m3 Removal of excessive coppice shoots Protection (from cattle) 12-15 Minor checking only 16 Clearfelling 90m3 Removal of excessive coppice shoots Protection (from cattle) 17-20 Minor checking only 21 Clearfelling 75m3 * When needed - 142 - Annex 2.9 Page 5 of 9 Cost ModeL II (Shadow values: Labor 0.5 = 1.00 Birr per day, foreign exchange 1.33) Man-days Cost Year Activity Required Labor Material Total (Birr) 1 Site clearing, pitting 80 80 - 80 Tools - - 10 10 Seedlings (2500) 10 10 268 a/ 278 Planting 24 24 - 24 Fertilizer - - 200 b/ 200 Application of fertilizer 20 20 - 20 Weeding 30 30 - 30 Protection 4 4 - 4 Total 168 168 478 646 2 Weeding 30 30 - 30 Replanting of blanks 20 20 58* 78 Protection 4 4 - 4 Total 54 54 58 112 3 Protection 4 4 - 4 6 Clearfelling 75m3 38 38 - 38 Cutting Shoots 18 18 - 18 Protection 4 4 - 4 Total 60 60 - 60 11 Clearfelling 90m3 45 45 - 45 Cutting Shoots 18 18 - 18 Protection 4 4 - 4 Total 67 67 - 67 16 Clearfelling 90m3 45 45 - 45 Cutting Shoots 18 18 - 18 Protection 4 4 - 4 Total 67 67 - 67 21 Clearfelling 75m3 38 38 - 38 a/ 50% foreign exchange b/ 100% foreign exchange - 143 - Annex 2.9 Page 6 of 9 Cost Model II Year Cost Discounted Cost a/ Felling Volume (Birr) (m3) 1 646 646 2 112 102 3 4 3 6 60 34 75 11 67 23 90 16 67 15 90 21 38 5 75 828 a/ At 10%. - 144 - Annex 2.9 Page 7 of 9 Peri-Urban Plantations Silvicultural Model III, Erosion Prone Highlands, 1800-2800 in a.s.l Eucalyptus globulus Year Silvicultural treatment l Pegging of bunds (50% of area) a/ Site clearing (cutting of existing "brush") Simple terracing (soil bunds on 50% of the area) Pitting 40 x 40 x 30 cm, spacing 2 x 3 m Planting of potted seedlings Application of fertilizer, 50 gram per seedling Weeding Protection (from cattle) 2 Weeding Replanting of blanks Repair of bunds Protection (from cattle) 3 (Weeding b!) (Application of fertilizer, 50 gram per seedling b/) Protection (from cattle) 4-5 Minor checking only 6 Clearfelling 40m3 Removal of excessive coppice shoots Protection (from cattle) 7-10 Minor checking only 11 Clearfelling 50m3 Removal of excessive coppice shoots Protection (from cattle) 12-15 Minor checking only 16 Clearfelling 50m3 RemovaL of excessive coppice shoots Protection (from cattle) 17-20 Minor checking only 21 Clearfelling 40m3 a/ Leaving lm between rows (bunds) b/ When needed - 145 - Annex 2.9 Page 8 of 9 Cost Model III (Shadow values: Labor 0.5 = 1.00 Birr per day, foreign exchange 1.33) Man-days Cost Year Activity required Labor Material Total (Birr) 1 Pegging of bunds 4 4 - 4 Site clearing 15 15 - 15 Terracing 175 175 - 175 Pitting 30 30 - 30 Tools - - 20 20 Seedlings (1670) 3 3 188 a/ 191 Planting 16 16 - 16 Fertilizer - - 133 b/ 133 Application of fertilizer 16 16 - 16 Weeding 20 20 - 20 Protection 4 4 - 4 Total 283 283 341 624 2 Weeding 40 40 - 40 Replanting of blanks (incl. repair of bunds) 35 35 73 a/ 108 Protection 4 4 - 4 Total 79 79 73 152 3 Protection 4 4 - 4 6 Clearfelling 40m3 20 20 - 20 Cutting shoots 15 15 - 15 Protection 4 4 - 4 Total 39 39 - 39 11 Clearfelling 50m3 25 25 - 25 Cutting shoots 15 15 - 15 Protection 4 4 - 4 Total 44 44 - 44 16 Clearfelling 50m3 25 25 - 25 Cutting shoots 15 15 - 15 Protection 4 4 - 4 Total 44 44 - 44 21 Clearfelling 40m3 20 20 - 20 a/ 50% foreign exchange b/ 100% foreign exchange - 146 - Annex 2.9 Page 9 of 9 Cost Model III Year Cost Discounted Cost a/ Felling Volume (m ) (Birr) 1 624 624 2 152 138 3 4 3 6 39 22 40 11 44 15 50 16 44 10 50 21 20 3 40 815 a/ At 10%. Anne. 2.10 RURAL ENERGY FORESTRY REQUIRtMNT (Volumes in solid m3 on bark) Estimated Area to be Demand demad Additowla Eqgiv. Man-days planted by Total area Total Area to be Total Total Fuelwood satis- supply Estimated Area of Direct Total. Incremental Increeental Incremental 1992 for 1997 cost b/ M.a-davs Incremental planted by 1997 cost b/ mandays Region PopUlatian & Poles a/ faction needed Productimn Planning Coat b/ per ha m1i. demand 1987 demand 1992 demand 1997 demand 1982-92 1982-9Z demand 2002 for 2002 demand 1982-97 1982-97 ('CO0) ('ODD n3) (Z) ('000 n3) (m3/ha/year) (Birr '000) ('000 m3) ('000 m3) ('000 m3) ('000 ha) (8irr '000)(million) ('000 m3) ('000 ha) (Sirr '000) (million) Arai 1,124 1,338 50 669 18 37.2 3,348 164 6.1 157 176 195 66.5 5,986 10.8 219 78.7 7,084 12.9 8ale 877 1,044 40 627 15 41.8 3,762 164 6.9 123 138 152 69.3 6,237 11.4 170 80.6 7,254 13.3 Eritrea 1,804 2,147 10 1,931 8 241.4 21,726 261 63.0 252 283 314 347.6 31,282 90.7 350 391.4 35,224 102.1 Ge Gofa 1,003 1,194 50 597 12 49.8 4,482 132 6.6 140 157 175 89.2 9,030 11.7 194 105.4 9,488 13.8 Gojas 1,976 2,351 40 1,411 15 94.1 8,468 164 15.4 277 308 345 43.5 14,051 25.6 382 181.6 16,343 29.8 Gondar 1,993 2,372 20 1.897 12 158.1 14,229 132 20.9 280 312 346 236.3 21,264 31.2 388 268.6 24,171 35.5 Hararghe 3,018 3,591 40 2,155 12 179.6 16,164 261 46.9 422 472 525 297.9 26,814 77.8 587 346.8 31,216 90.6 Illubabor 820 976 95 49 20 2.5 226 164 0.4 115 127 143 21.9 1,972 3.7 159 29.9 2,687 5.0 Kaffa 1,574 1,873 95 94 20 4.7 422 164 0.8 220 246 274 41.7 3,753 6.8 306 57.0 5,130 9.3 Shoa 4,876 5,802 30 4,061 18 225.6 20,304 132 29.8 683 762 848 352.9 31,763 46.6 947 405.5 36,498 53.5 Sidaso 2,744 3,265 70 979 20 49.0 4,410 164 8.0 384 429 477 113.6 10,224 18.5 533 140.3 12,623 22.9 Tigrai 2,105 2,505 10 2,254 8 281.8 25,362 261 73.5 295 330 367 405.9 36,532 105.9 408 456.9 41,122 119.2 Wollega 2,028 2,413 60 965 15 64.3 5,788 132 8.5 284 316 352 127.8 11,505 72.9 394 154.1 13,869 76.2 Wollo 2,552 3X037 10 2.734 10 273.4 24.606 261 71.4 357 400 444 393.5 35,416 102.7 496 443.1 39.880 115.6 Total 28.479 33,908 (40) 20,423 12 1,703.3 153,298 (210) 358.2 3,989 4,456 4,957 2,607.6 244,829 616.2 5,533 3,139.9 282,589 669.7 a/ 1.09 m3 per person per year for fuelwood, plus 0.1 m3 per person per year for building poles, agricultural ipleneot., etc. b/ At 90 Birr per ha for inputs in Peasant Association nurseries, 1982 price level. - 148 - Annex 2.11 SIMPLE COST ESTIMATE FOR RURAL AFFORESTATION BUDGET Per ha Established (Nearest Birr) 1. Overheads of National and Regional Administration Total Cost of: * Transportation 8 - Housing: offices and homes 4 - Office Expenditure (Annual) 10 * Staff 40 2. Direct Costs at Peasant Association Level A. Nursery Materials Seedling Production (includes seed, building, services) 90 B. Civil Works Minor roading, bridging, drainage, etc. 50 C. TooLs and Sundries 8 3. Labor Peasant labor input throughplant plantation life is: * Planting 190 man-days * Harvesting 14 man-days * Weeding 6 man-days 210 man-days (budgeted at average cost of one Birr per manday) 210 420 - 149 - Annex 2.12 ESTIMATES OF NUMBER OF CATTLE AND OF DUNG PRODUCTION BY REGION OF ETHIOPIA, 1981/82 Regions Cattle Human Ratio Annual Dung Production a/ Population Population Cattle/Human (1000 tonnes) Fresh Weight Dry Weight ('000) (1000) (13% dm)* (85% dm)** Arsi 1,644 1,207 1.4 10,134 1,550 Bale 509 922 0.6 3,138 480 Gemo Gofa 602 1,051 0.6 3,711 508 Gojam 2,316 2,141 1.1 14,277 2,184 Gonder 1,524 2,156 0.7 9,394 1,437 Hararghe 1,249 3,284 0.4 7,699 1,178 Illubador 450 849 0.5 2,724 424 Kaffa 1,323 1,691 0.8 8,155 1,247 Shoa 6,419 6,792 0.9 39,569 6,052 Sidamo 1,863 2,744 0.7 11,484 1,756 Wollega 1,815 2,116 0.9 11,188 1,711 Wollo 2,181 2,743 0.8 13,444 2,056 Tigrai* 1,735 2,270 0.8 10,695 1,636 Eritrea* 989 2,586 0.4 6,096 932 TOTAL 24,619 32,741 151,758 23,211 * Base numbers estimated in the 1980/81 census of livestock. Assump- tion is that due to drought and conditions of starvation, the live- stock population in these regions will not have grown noticeably within the last year. ** dm=dry matter. All data is in metric tonnes=te. a/ Conversion factor, cattle to dung, is 202.5 kg average live weight and 83.4 kg fresh weight of dung production per day per ton of live weight (fresh weight is at 87% m.c.w.b.). Source: CPSC, CSO. Agriculture Sample; Survey, 1981/82. Preliminary results of livestock and poultry for Ethiopia, Addis Abeba, October 1982, and ICLA survey data. - 150 - Annex 2.13 FUELWOOD DEFICITS AND DUNG AND CROP RESIDUE CONSUMPTION ESTIMATES BY REGION, 1981/82 Fuelwood Dung Straw and Demand Fuelwood Dung Cow Dung Burnt in Stover Region Satisfaction Deficit Equivalent Available Households Burnt (S) ('000 m2) ('000 te) ('000 te at ('000 te, 15%mc) ('000 te 15% mcwb) (15% m.c.) Arsi 50 613 423 1,550 85 310 Bale 40 574 396 480 198 182 Eritrea 10 1,769 1,220 932 932* 264 Gamo Gofa 50 547 377 568 75 277 Gojam 40 1,292 891 2,184 446 408 Gondar 20 1,738 1,199 1,437 1,079 328 Hararghe 40 1,974 1,361 1,178 681 624 Illubador 95 45 31 424 3 26 Kaffa 95 86 59 1,247 6 49 Shoa 30 3,720 2,566 6,052 1,283 1,177 Sidamo 70 897 619 1,756 124 454 Tigrai 10 2,065 1,424 1,636 1,282 130 Wollega 60 884 610 1,711 122 446 Wollo 10 2,504 1,727 2,056 1,554 461 Total 7,864 5,138 Note: Equivalence is established using the calorific value of dry cow dung (15% m.c. wet basis), 13.8 MJ/kg, and lm3 of wood (solid) at 25% m.c.w.b. (as sold), 500 kg/mr3 basic density, and 14.3 MJ/kg, thus 9,524 MJ/m , giving 1 tonne dung = 1.45m3 3 wood. Cereal straw is 15 MJ/kg, thus 1 tonne straw = 1.09 tonne dung or 1.57m of wood. In all, the combustion efficiency of straw/stover and dung is assumed to be the same as for wood (i.e. 8% with clay pots on open 3 stone fires). - 151 - Annex 2.14 FORECAST DEFICIENCY OF WOODFUEL SUPPLY TO URBAN AREAS '000 hectares ---- Deficit Under Achievable Planting a/ Surplus to Addis b/ Net Deficit Planted By Total North South From all Sources c/ to South 1982 for 1987 271 165 106 -- 106 1987 for 1992 342 212 130 73 57 1992 for 1997 350 218 132 93 39 1997 for 2002 234 142 92 119 (27) a/ From Table 2.11. b/ FSom Annex 2.13, equated to hectares of plaitation equivalent using 1.67 m per tonne of wood equivalent and 17.4 m /ha mai average for southern urbn areas. c/ Assuming a growth in the surplus of all new fuels to Addis Abeba require- ments of 5% per year. - 152 - Annex 3.1 STATUS AND ELECTRICITY DEMANDS OF BOILER ELECTRIFICATION PROGRAM Project Existing Current Oil Electrical Boiler Boiler Estimated Estimated Date Standby Boilers Consumption Installation Capacity Maxijum Energy/Yr Load Boiler (te=tonnes) Status Demand Demand Condition Pulp & Paper 1-14 te/hr 4500 te/y Operational 12MW 12MW 42 1982 Existing (+12 months) GWh boiler to be refurbished Dire Dawa Textiles 5-42 te/y 7035 te/y Installation 15MW 12MW 66 1987+ Existing complete-requires GWH plant electricity supply retained United Oil & Soap 4-4 te/hr Wood Civil works complete 12MW 12MW 40 1985 Some refur- no mechanical or GWh bishment of electrical boilers--some may be scrapped Awash Tannery Project Abandoned (equipment purchased) Akaki Textiles 1-6.3 5000te/y Works conplete 12MW 12MW 42.3 1983 Existing 1-4.6 commissioning GWh plant re- 1-6 required tained as 1-3.7 standbv Rubber and Canvas Steam from nil Operational 650kW 0.7MW 1.88 1983 n/a Tyre Co. (I month) GWh nil Tyre Factory 2-6 te/hr 3000 te/y Installation 6MW 3MW 26.4 1983 Boiler being proceeding GWh kept as standby St. Georges Brewery 1-42 te/hr 1460 te/y Installation 9MW 9MW 13.8 1983 1 boiler 1-53 te/hr complete except GWh requires new 1-42 te/hr feeder furnace tube Akaki Oil Mills 1-1 te/hr 150 te/y Only civil half 3MW 3MW 1.8 1985 Boiler re- completed GWh tained as standby Mokjo Textiles 2-3.5 te/hr 900 te/y Installation 3fW 3MW 8.4 1983 Boilers OK complete awaiting GWh as standby commissioning Hilton Hotel 2-3 te/hr (252 te/y) Operational 650kW 650kW 2.0 1982 Boiler OX part comm'n GWh as standby Ras Hotel none 60 te/y No installation 200kW 200kW 0.6 1983 Boilers work started GWh scrapped Wabe Hotel 1/2 - 1 te/hr 100 te/y Installation complete 380kW 380kW 1.1 1983 Boiler OK GWh as standby Yakatit School ? ? ? 140kW 140kW 0.21 ? No GWh information TOTAL 68MW 246 m4h ? - Not Known Asoe. 3.2 EELPA ICS MISSION'S POWER DEMAND FORECAST --Residential - Comsercial Snall - Large Ind..tries -- FY Cookers Lighting St. Lighting Industry Existing Boilers New Large Additlon Change System Load & Power & Agriculture Industry Subtotal to ICS Total (%/Yr) Factor (%) GWh MW GCWh MW CWh MW 1WH MW CWh MW GWh MW MWh MW GWh MW M40 1W GWdh mw GWh MW Base Year Sales 1982 12.24 123.12 57.44 43.71 227.99 20.26 -- 484.76 -- --- --- 484.76 - Losses on Sales (Z) 14.6 14.6 14.6 14.6 8.7 8.7 8.7 11.6 --- 14.6 --- 11.6 --- Base Year Sales 1982 14.03 141.10 65.83 50.09 247.93 22.02 --- 541.00 102.5 --- --- 541.00 102.5 60.3 Forecast Demand 1983 15.26 8.1 150.98 34.5 70.44 14.6 53.60 10.2 257.85 42.1 76 17.4 -- --- 618.13 125.5 --- 618.13 125.5 14.3 22.4 56.2 1984 18.18 9.6 161.55 36.9 75.37 15.6 57.35 10.9 268.16 43.8 149 44.6 5.76 2.4 735.37 163.8 --- 735.37 163.8 51.2 1985 23.57 12.3 172.85 39.5 80.64 16.7 61.37 11.7 278.89 45.5 195 60.9 40.06 11.3 852.38 197.9 9.94 2.3 862.32 200.2 49.2 11.2 11.2 1986 33.59 17.5 184.96 42.3 86.29 17.9 65.66 12.5 290.05 47.4 98 (46.9) 96.72 23.5 855.27 208.0 21.85 5.0 877.12 213.0 47.0 1987 42.88 22.0 197.90 45.2 92.33 19.1 70.26 13.4 301.65 49.3 0 0 193.07 38.7 898.09 187.7 36.05 8.2 934.14 195.9 54.4 1988 52.16 26.7 211.76 48.4 98.79 20.5 75.12 14.3 313.71 51.2 0 0 248.2 40.1 999.74 201.2 52.15 11.9 1051.89 213.1 -- -- 56.3 1989 54.51 27.8 226.58 51.8 105.71 21.9 80.44 15.3 326.36 53.3 0 0 280 45 1073.50 215.1 71.72 16.4 1145.22 231.5 56.5 1990 55.10 27.9 242.44 55.4 113.11 23.4 86.07 16.4 339.31 55.4 0 0 310 50 1146.03 228.5 94.67 21.6 1240.70 250.1 56.6 8.3 8.2 1991 55.28 27.7 259.41 59.3 121.03 25.1 92.10 17.5 352.88 57.6 0 0 340 55 1220.70 240.2 121.49 27.7 1342.79 269.9 56.8 1992 57.17 28.5 277.57 63.4 129.50 26.8 98.54 18.8 367.00 59.9 0 0 370 60 1299.78 257.4 152.73 34.9 1452.51 292.3 56.7 1993 59.25 29.4 297.00 67.9 138.56 28.7 105.44 20.1 381.68 62.3 0 0 400 65 1381.93 278.4 183.84 42.0 1565.77 315.4 56.7 1994 62.23 30.9 317.79 72.6 148.26 30.7 112.82 21.5 396.95 64.8 0 0 430 711 1468.05 290.5 218.57 49.9 1686.62 340.4 56.6 1995 65.35 32.5 340.04 77.7 158.64 32.9 120.72 23.0 412.83 67.4 0 0 460 75 1557.58 308.5 233.86 53.4 1791.44 361.9 56.5 - 154 - Annex 3.3 EXISTING AND COMMITTED ERESA ICS GENERATION Number and Type of Total Installed Year of Retirement Station Name Generation Units Capacity (MW) Commissioning Date A. Existing Belesa 1 X 10 MW steam 20.0 1966 1996 1 X 5 MW steam 1964 1994 1 X 5 MW diesel 1969 1989 Kagnew 4 X 1.5 MW diesel 6.0 1966 1981 Gajaret 10 X 0.58 MW diesel 5.8 1979 1994 Girar 1 X 1.1 MW diesel 9.4 1958 1978 1 X 1.1 MW diesel 1959 1979 1 X 1.1 MW diesel 1961 1981 1 X 1.1 MW diesel 1962 1982 1 X 5.0 MW diesel 1973 1993 Total Existing (1982 year end) 41.2 B. Committed 2 X 5.0 MW diesel 10.0 1983 2003 Total ERESA ICS (1983) 51.2 Source: ACRES International - EELPA. - 155 - Annex 3.4 PROJECTED INVESTMENTS IN THE ETHIOPIAN POWER SECTOR a/ (Birr million-1983 Prices) 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 Total Generation EELPA ICS Mlalka Wakana bl 56 94 94 7 5 - - - - - 319 Amarti Diversion 12 18 19 12… …………61 Finchaa 4th Unit 5 14 14 14 47 Gibe Daka … … … … ……- --48 95 119 142 71 - -- 475 Next Hydro Project … … … … … … …- - - - 60 60 120 EELPA SCS Lake Tana Regulation 9 6… … … … … … …15 Tis Abay 3rd Unit b/ I … … … … … … … … …I Sor minihydro -- 8 10 8 26 Hoha minihydro -- 1 1 … … …… … Dedissa minihydro -- 5 6 4… …………15 Other minihydro - 7 7 7 7 7 - - - 35 Diesel Plants t0 2 2 2 2 2 2 2 2 2 2 2 32 ERESA ICS + SCS DBiesel Plants 26 - 1 I - I -- 1 - 1 -- 31 SUBTOTAL 1-19 14r8 1534 122 10 3-7 1065 12i1- 14-5 IT -63- 621179- Transmission & Distribution 230kV/i132kV Dessie Extension-~/ 24 15 … … … … -- --39 Muger Extension LI 3 … … …… …3 Shashamene- Malka Wakana.' 7 7… ……… ……14 Shashamene- A. MinclhkL 13 12 25 Geferressa- Addi. N. / 1 1 Lake Tana Region~. b I… … … 1 Alaba-Gibe Daka 9 8… ………1 7 Finchaa-Baha DaOs… … … … …- - 23 31 23… ……77 Malkeassa- Dire Dawa - 25 34 25… … … … … … … … 84 Gibe oaka-Jima - - 7 7 … … … …14 Yirgalem-Shakisso - - 6 8 6 20 Addis N.-Cotobie 3 6… ………9 Awash-Angelie 7 10 7 …… 24 66kV/45kV aarrar Regionk b 2… … …… … …… … …2 Nekemte Region… … … … … 9 9… ………18 Jimma Region 6 6 12 Muger-Fitchie … ……4 4… ………8 M. Wakana-Robie… … … … …4 4 8 Ambo-Ghion …… 3 3… ……6 Dessie Region 9 1 2 4 -25 Debre Markos-Dejen … … … … 3 3 6 Dilla-F. Genet … ……… 3 3 6 Finchaa-Shabo 3 3 6 other 66kV exten. - -- 5 5 5 15 Substation Expansion -- -- 17 32 24 Il 7 7 5 5 5 5 118 Distribution Expansion 12 12 12 12 12 12 12 12 12 12 12 12 144 (inc. Village elec.) Subtotal 72- 7- 176 94 77- -r8 79 355 26- 22 22 Y2 7 Infrastructure Workshop 3 10 10 10… …… …… …33 Training Inst. 3 5 5 3 --…… …16 Office Bldgs - 4 6 6 4 20 Vehicles 5 5 5 5 5 2 2 2 2 2 2 2 39 Subtotal 171 24- 6 24 9 -2 -2 -2 2 -2 -2 2 10-8 Total Investment 202 251 256 240 96 147 176 178 173 97 87 86 1989 Note: a/Investments required to meet mission's demand forecast. bIcompletion of projects started before 1984. Source: EELPA - 156 - Annex 4.1 REFINED PETROLEUM PRODUCTS SUPPLY OPTIONS Investment and Operating Cost for Different Cases in 1983 US Dollars I. Investment Cost Case I Case 2 Case 3 Case 4 Case 5 A. Energy Conservation - Revamping of heat exchange train in the distillation unit -- 1.5 1.5 1.5 - Revamping of three process furnaces -- 3.5 3.5 3.5 - Revamping/replacement of TPS boilers = 4.5 4.5 4.5 Subtotal -- 9.5 9.5 9.5 B. Product-Pattern Improvement - Minor modifications in distillation unit -- 0.2 0.2 0.2 - Capacity increases in catalytic reforming and hydro treating facilities -- 1.5 !.5 1.5 - Installation of Merox treating facilities -- 2.5 2.5 2.5 - Modernization of Asphalt-blowing and filing facilities -- 1.5 1.5 1.5 Subtotal 5.7 5.7 5.7 C. Expansion of Distillation Unit by Minor Debottlenecking -- -- 2.5 2.5 D. Improvement in General Operational Efficiency - Modernization/replacement of instrumentation -- 1.5 1.5 1.5 - Repiacement of water coolers with air coolers/ trim coolers -- 1.5 1.5 1.5 - Installation of on-iine blending facilities = 0.2 0.2 0.2 Subtotal -- 3.2 3.2 3.2 E. Secondary Conversion Facilities - Thermal cracker (420,000 tons/year) -- -- -- 18.0 - Distillate hydrotreating (135,000 tons/year) -- -- -- 5.0 - Off-sites facilities (50% of on-site facilities) = -- -- 12.0 Subtotal 35.0 Note: Add Contingency of 20% to all subtotals 11. Operating Costs (Million 1983 Dollars) 3.8 4.0 4.5 6.0 - 157 - Annex 4.2 Petroleum Production Transportation Analysis, Asseb-Addis Ababa Parameters adopted on the basis of records obtained Trucks Medium Size Trucks * Carrying capacity net (average): 24 tonnes * Price (CIF) Ethiopia: US$ 122,600 fully fitted * Shadow values of true cost of foreign exchange in US$ equivalent: US$ 163,058 * Repairs and maintenance including yptes: 0.3684 BIRR/km * Write-off rate due to accidents: 15% per annum : Life of vehicle other that write-off: 5 years - Salary, trip allowances and overheads: US$12,455 (1987) per truck Large Trucks * Carrying capacity net (average): 36 tonnes * Price (CIF) Ethiopia, fully fitted High price estimate: US$ 244,778 Low price estimate: US$ 179,483 * Shadow value of true cost of foreign exchange High price estimate: US$ 324,778 Low price estimate: US$ 238,653 - Repairs and maintenance including types: 0.2989/km - Write off rate due to accident: 7% per annum - Salary, allowances and overheads: US$ 12,300 (1987) per truck Common Operating Characteristics - Round trip Asseb-Addis Ababa: 7 days • Availability (average actual): 2.2 trips/month, 26.4/year * Distance covered annually: 45,408 km per vehicle - Fuel consumption: 1.4 km/litre for both sizes of trucks * Lubricants: 10% cost of fuel * Fuel prices: US$ 0.30/litre Addis Ababa, US$ 0.24/litre Asseb in 1983. - 158 - Annex 6.1 IMPORT PARITY BUILD UP (C.l.F. Assab) Jet A-1l Mogas Mogas Gasoii Fuel Kero 93 Oct. 83 Oct. 53 D.1. Oil ---------- USUSgallon ---------- US$/bbi 1. F.O.B. * 110.00 100.80 97.00 103.00 31.25 USS/bbl 2. Freight ** 6.56 6.11 5.95 6.88 2.73 " " 3. Insurance (0.12%) 0.14 0.13 0.12 0.13 0.04 " " CIF (uSt/USgal lon) 116.70 107.04 103.07 110.01 34.02 USS/bbl CIF (e4/litre) 63.82 58.53 56.36 60.16 44.29 ejVliter 4. Bank Charges (2.25S) 1.44 1.32 1.27 1.35 0.99 5. Ocean Leakage (0.5%) 0.32 0.30 0.29 0.31 0.22 6. Stamp Charges 81.00/metric tonne 0.08 0.07 0.07 0.08 0.09 7. Service Charges B.0.30/liter 0.30 0.30 0.30 0.30 0.30 Total e4/litre 65.96 60.52 58.29 62.20 45.89 ** 68.28 62.81 60.57 64.53 46.87 * 8MS posting effective 4/19/82. ** Price build-up worked out on basis of BMS posting of 2.23/82 and related AFRA forecast and basic rates. ^** (a) AFRA for Dec. 1982 175 plus premium for W.O. 35= 210. (b) Basic rate USS6.56/English ton plus differential USS4.06= US$10.62/ton. Ocean freight in US$/English ton - (a) x (b) = 2.10 x 10.62= US$22.30/ton. Annex 6.2 Page I of 2 PETROLEUM PRICE BUILD-VUP AND MAXIMUM SELLING PRICE a/ AT SOME LOCATIONS IN ETHIOPIA Distributing Asaba Unit in Ex-Refinery Excise on Municipality Company's Retailers Retailers Price at Price at Price at Price at Price at Product Ethiopian e/Lt. Price Custom Tax Tax Margin Margin Price Addis Ahaha Ofre-Dawa Asmara Cambolcha Nazareth Gasoline (old) e/Lt 88.50 10.5 2.0 4.0 3.0 108.0 1iq.n 1q9.0 113.0 tii.n 119.0 (new) - 86.90 10.5 2.0 4.0 3.0 106.4 117.4 117.4 111.4 113.4 117.4 Diesel Oil (old) - 52.75 6.50 2.0 3.75 3.0 68.0 79.0 7q.0 73.n 7s.n 79.0 (new) - 51.76 6.50 2.0 3.75 3.0 67.01 78.01 78.nl 72.nl 74.n1 78.nl Inland (old) - 38.50 - - 3.50 - 42.0 54.n 54.0 47.0 48.7 54.n Fuel Oil (new) - 41.4 - - 3.50 - 47.9 59.9 59.9 52.9 54.6 6n.n Kerosene (old) - 47.75 - - 3.75 3.0 54.50 65.50 65.50 5q.5 61.50 65.5 (new) - 47.6 - - 3.75 3.0 54.35 65.35 65.35 59.35 61.35 65.35 LPG (old) b/ Br/ton 520.00 100.0 20.0 420.0 240.0 1300.0 (new) - 519.66 100.0 20.0 420.0 240.0 1299.66 a/ Consumer prices reflect the distances of the distributing centers from Assab. b/ Prices of LPG are the same throughout the country except for differences that may be permitted by Government. For examole, the LPG orice in Asmara is B1460/te. - 160 - Annex 6.2 Page 2 Of 2 Road & Marine Transport Rates Schedule Effective From March 27, 1981 (Abbreviated by Mission) A. Tank Truck Tariff: 1. Following is the Tank Truck Tariff: Gasolines, Diesel, Aviation Fuels Fuel Oil, Marine and and Kerosene Industrial Diesel Oil Ex-Assab/ Ex-Assab/ Djibouti Ex-Massawa Djibouti Ex-Massawa Unit Rate: ’/Litre 1.1 Assab-Addis Abeba 11.000 - 12.000 - 1.2 Massawa-Asmara - 2.000 - 2.150 Unit Rate: C/M3/km 1.3 Other routes (Truck & Trailer) 12.644 17.391 13.782 18.178 1.4 Other routes (Trucks only) 27.000 37.940 29.539 41.355 2. LPG Tank Trucks 1. LPG Trucks - Assab/Addis - Birr 199.00/Ton 2. Other routes to be paid at the rate of ’22.87/m3/km. Nonethe- less, as long as it is compulsory to transport for Asmara ex-Assab by road - Assab/Asmara rate shall be Birr 359.00/Ton. B. Marine Tariff Following tariff will apply for port-to-port transfer of products by ship: Assab - Massawa C3.00/Litre Assab - Djibouti ’2.00/Litre Aden - Assab C3.00/Litre Aden - Massawa C3.50/Litre Aden - Djibouti C3.00/Litre AlL volumes to be measured at 30°C. - 161 - Annex 6.3 BUILD-UP OF ECONOMIC COST OF PETROLEUM PRODUCTS IN ETHIOPIA, 1983 (US$ and USC) Gasoline Diesel Kerosene LPG Fuel Oil Offshore Costs US$/Tonne 297 276 297 307 153 FOB ASSEB (for export) Freight 11 11 11 50 n. appi US$/tonne Ocean Losses 1.5 1.4 1.5 1.5 n. appl. (0.5% average) US$/te CIF 309.5 288.4 309.5 358.5 -- Asseb USC/litre CIF 23.10 24.00 24.18 18.71 14.37 USC/litre CIF at Shadow value of foreign exchange (1.33) 30.72 32.00 32.16 24.88 19.11 Inland Costs Port Charges (est. c/litre) 0.15 0.15 0.15 0.15 0.15 Freight to Addis Abeba 6.00 6.00 6.00 7.00 6.00 (economic cost est.) Distribution and Marketing 1.93 1.81 1.81 10.58 1.69 Estimated Economic Cost at Addis Abeba USC/litre 38.80 39.96 40.12 42.61 26.95 e’/litre 80.31 82.72 83.05 88.20 55.79 Annex 7.1 A TENTATIVE SCHEDULE OF NEW ENERGY SECTOR INVESTMENTS IN ETHIOPIA, 1983/4-1992/3 (Birr Million) Subsector 83/84 84/85 85/86 86/87 87/88 88/89 89/90 90/91 91/92 92/93 Total 1. Power a. Generation - EELPA ICS 73 126 126 101 -- 48 95 119 142 71 901 - EELPA SCS 20 22 26 21 9 9 9 2 2 2 122 - ERESA SCS/ICS 26 -- I -- 1 -- 1 1 -- 1 31 b. Transmission/ Distribution 72 79 76 94 77 88 69 55 26 22 658 c, Infrastructure 11 24 26 24 9 2 2 2 2 2 104 Subtotal 202 251 255 240 96 147 176 179 172 98 1816 2. Petroleum a. Production -- -- -- -- (5) (5) -- (90) (221) (321) 642 b. Exploration 8 9 -- -- -- -- -- -- -- -- 17 c. Retining 3 4 -- (40) (45) (50) -- -- -- -- 142 d.Demand 7 12 8 11 -- -- -- -- -- 38 Management/Subs 'n Subtotal 18 25 8 51 50 55 -- 90 221 321 839 3. Household Energy a. New biomass - Crop residues -- I 1 2 3 5 9 14 16 21 72 - Charcoal -- 1 3 7 5 5 5 5 6 6 41 b. Improved Cooking -- 1 1 1 -- -- -- -- -- -- 3 c, Fuelwood Production -- -- -- -- -- -- -- -- -- -- -- - Peri-urban Existing I 1 4 4 4 5 5 5 5 36 *New 8 13 17 21 31 42 52 63 73 83 403 - Rural Forestry 7 14 22 29 36 40 43 47 51 54 343 d. Manpower TAD 2 3 3 2 2 2 2 2 2 1 21 Subtotal 17 34 48 66 81 98 116 136 153 170 919 4. Other Energy Sources a. Ethanol 5 15 13 -- -- -- -- -- 33 b. Coal/Geothermal 3 3 ? ? -- -- -- -- -- -- 6 c. Solar and Other 1 3 3 3 4 5 2 2 2 2 27 Subtotal 9 21 16 3 4 5 2 2 2 2 66 Total 246 331 327 360 232 306 294 407 548 591 3640 = investment is dependent on outcome of prior study. ? f further funding dependent on prior activity. ______________________________________________ ~~~~~IBRD 17512 r i s X, ~~~ ETHICOPIA -- AGRO- ECOLOGICAL ZONES / ' l ' :e/ f OF THE REGION 12 -- J 1 l \ X 0 _~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~lrgh aon3s laboo 1500 mererr) g 3 ; 0>g E o \ e t a t~~~~~~~~~~~~~~~~~~~~~~I or Pote.ntral j | = / <. . i 71/ ) -~~~~~~~~~~- I~sohyots (raunfoll), n or I meters KARTOUM / All rreathes R-no I ,rr,rrrarr,, 8'O~~~~~~e Dsy- ooesRoodr fr44,4 Cu . j-CLJOs ; ; ' t?' aX Rvers, [. Present Irrigated Areas 2 P/. . kIo. o 0_ \\< : . . -.-M. Mon Forest Ar-eo >7 22 '2 ~~~~~~~~~~~~~~~~~~~~~~~~~~~Regi-na B-jecey . r, r-- -\ t,nar onWCrr are_ <2 -- - -yE 0 0 -D D J I / U T i; hs~~~~~~~~~~~~~0Dh 2fs I'r/5 - w,./hnb l,; ;tU f' ... - o rb A L i A A R ~ ~ ~ ~ ~ ~ oaezre_neow°l=w 1rl S S( o - 3/ Yb .__________________ ro - j' 9 1 2p 84048.4,, jk o 4, A U G A N D A < ' , E r< sp A X 0, S 44^ zo ~,aat.!@ 9iS I D A .t , I _____________________________ 18W~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~IR 17513 ORRA,K SAUDI F > 3 4-4'4' OCTOBER198 A""\ 9 <7. ~~~~~~ETHIOPIA SUDAN .$)25ŽSOC j N, 0' ~~~~~~~~~~~~~~~~~~~ENERGY ASSESSMENT Cd, ~~~~~~~~~ 1'~~~~-J 44,o - * 4' C/iA '' / 4$Q$IT S~~~~~~~~~~~ ERASING~El CU-MIEYD FL/JA POWER PLANTES Sr- - AIUIA~ 7 p \A D-se, [~~~'5EEEoYA 4 is o A Isoooe~~~~~~~~~~~~~~~~~,-c--d ,ed C.-,tee ct >6 / AND~~~~~~~~~~~~~~~~~~~~~~~ F/-'' TRANSAPISS.0N LINES ZARIE %ENA i RTES sslp4 +~46 i 2D !A s,teo A UePCt%'Ps' E 2 2~ ;'':oI RopetSA C '9 e- oesb RDO/gs S U D A N ~ aA ' , bi ~ ~ < , 0 4 . 00 Ύ A..~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~A - - ?rA~~~~ / OAJiI BOUTI iI A. 0 n ~ ~ VW x< x ~7M~ Ded o 4FlN EOd10 /Plo I-FF0r2 M7 Dol/ C'YT ~~~~~~~~JLIIA-ZL '-A~~~~~~~N GAMBELA . . . . .~~5F DTA eS N A,Iloeo~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~o I ~ ~ ~ ~ ~ ~ A ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~db TN,, ,VTN/O Tt '1 ~ ~ ~ ~ This S,ep 000 Reop p,epa,eA A,,~~~~~~~~~~ Th1ft,P A '2'~'Gode 04 ~021A At - /;STeRFSSie0A o&AA >0 ETHIG- lET SERVICE SOS ISe COOORA@hCS~~~~~~~~~20 oh I., UGANDA 1 200~~~4 2'/ Roe~ A0 K,cAoAA, S 3-1 K E N Y A _j~~~~~~~~~~~~~~ARA- A TOFSRSOR PI Ar~~~~~~~~~4 K sS O tfooe i., I B R D 14257R1 ETHIOPIA d / \ j*;O pO ffi a \ PU,TIO,NDENSITf5 - MIaERSE RO 04L-NADS- ES --T-fl A 0U.. lxv >e x sx \ S X X C \ n: 2 w . S m ~~~~~~~~~~~MORE THAN 40 SECONDAPRf ROADS j S U D A N >*9u'"",- h~~~~~~~~~~~~~' SOTE 1Do <9 - - - NTRAIONALBOUUDARIE= G R1V K' i E I \ 0 /) / ' (fa ke \. ( r4> D’-'- S O /~~~~~~~~~~~~~~~~~~~~~~~~~ MA L I A U G A N D A D A N h-D ;I UGANDA K E Y, A~~~~~~~~~~~~~~V