Burundi - Energy Sector Rehabilitation Project

Document of The World Bank FOR OFFICIAL USE ONLY ' /Z- -Z1 Report No.9068-BU STAFF APPRAISAL REPORT REPUBLIC OF BURUNDI ENERGY SECTOR REHABILITATION PROJECT FEBRUARY 14, 1991 Industry and Energy Operations Division South Central and Indian Ocean Department Africa Region This document has a restitcted distribution and may be used by recipients only In the perfonane of their official duties. Its contents may not otherwise be disclosed without World Bank authorluudon.  CURRENCY EQUIVALENTS Currency Unit - Burundi Franc (FBu) US$1.00 = FBu 175 WEIGHTS AND MEASURES Gwh gigawatt hour = 1 million kilowatt hours kgoe kilos of oil equivalent - 10,500 kilocalories kv kilovolt = 1,000 volts kw kilowatt = 1,000 watts kva kilovolt ampere - 1,000 volt ampere kwh kilowatt hour = 1,000 watts hour Mw megawatt - 1,000 kilowatt Mwh megawatt hour = 1,000 kilowatt hour toe ton of oil equivalent = 10,500,000 kilocalories ton metric ton = 1.1 US tons ABBREVIATIONS BRB Banque de la Rdpublique du Burundi CCCE Caisse Centrale de Cooperation Economique CEBEA Centre Burundais des Energies Alternatives CIDA Canadian International Development Agency DF Direction des F8rets DGC Direction Generale de Commerce DGE Direction Generale de l'Energie DGHER Direction Generale de l'Hydraulique et des Energies Rurales EDFI Electricite de France International EGL Energie des Pays des Grands Lacs ESMAP Energy Sector Management Assistance Program FY Fiscal Year GTZ Deutsche Gesellschaft fUr Technische Zusammenar beit RFW Kreditanstalt fUr Wiedereufbeu MDRA Ministere du Developpement Rural et de l'Artisanat MEM Ministere de l'Energie et des Mines OBK Organization for the Development of the Kagera River Basin ONATOUR Office Fational de la Tourbe PIU Project Implementation Unit PMU Project Management Unit REGIDESO Regie de distribution d'eau et d'electricite SAL Structural adjustment loan SCEP Service Charge des Entreprises Publiques SEP Societe d'entreposage de petrole SINELAC Societe internationale d'electricite des Pays des Grands Lacs SNEL Societe Nationale d'Electricite-Zaire UNDP United Nations Development Program GOVERNMENT OF BURUNDI FISCAL YEAR January 1 to December 31  FOR OFICAL USE ONLY REPUBLIC OF BURUNDI STAFF APPRAISAL REPORT ENERGY SECTOR REHABILITATION PROJECT Table of Contents Page No. CREDIT AND PROJECT SUMMARY . . . . . . . . . . . . . . . . . . . i-i I. THE ENERGY SECTOR . . . . . . . . . . . . . . . . . . . . . 1 Country Background .... . . . . .... . . . . . . . 1 Energy Consumption .... . . . . . . . . . .... . 1 Energy Resources . . . . . . . . . . . . . . . . . . . . . 2 Biomass Resources .... ....... . . . . . . . . 2 Hydropower .... . . . . . . . . . . . . . . .... 4 Peat ..... . . . . . . . . . . . . . . . . . .... 4 Petroleum .... . . . . . . . . . . . . . . . .... 5 New and Renewable Energies . . . . . . . . . . . . . . 6 Energy Institutions.... 6 Government Objectives.... 8 Bank Participation in the Sector .... 8 II. THE POWER SUBSECTOR .9 Demand for Electricity .9 Generating Capacity ..11 Transmission and Distribution . . . . . . . . . . . . . . 12 Further Developments in Power Generation . . . . . . . . . 12 Operations and Maintenance ..13 Electricity Tariffs ..15 Rural Electrification ..16 III. THE IMPLEMENTING AGENCIES ..16 REGIDESO ..16 Organization and Management . . . . . . . . . . . . 16 Staffing and Personnel . . . . . . . . . . . . . . . . 17 Accounting, Audit and Insurance . . . . . . . . . . . . 18 Management Information Systems . . . . . . . . . . . . 18 Assets Revaluation .... . . . .... . . . . . . . 19 Performance Contract .... . . . ... . . . . . . . 19 This report Is boasd on the findings of an appraisal mission to Burundi in Juno/July 1990, which consisted of Messrs. Potrieto Millon, Senior Economist (Task Manager); Richard Senou, Financial Analyst; Robert van dtr Plas, Biomass and Household Enorgy Specialist; Roland Brilot, Power Consultant; and Michael Patou, Potroleum Consultant. Mr. Socar Thia- was tho peer reviewer for the operation. Ms. AlIno Cayouette provided secretarial support In the preparation of the report. Mr. Michaol Sarri. and Mr. Francisco Aguirre-Sacass are the managing Division Chief and the Department Director, respectively, for the operation. This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  Page No. III. (continued) The DGHER .... . . . . . . . . . . . . . . . . . . 20 Organization and Management . . . . . . . . . . . . . 20 Staffing and Personnel . . . . . . . . . . . . . . . . 21 Accounting and Auditing . . . . . . . . . . . . . . . . 21 IV. THE PROJECT.. . . . 23 Project Objectives . . . . . . . . . . . . . . . 23 Project Description . . . . . . . . .23 Management Assistance Partnership Program . . . . . 24 Institutional Strengthening of the DGE . . . . . 25 Institutional Strengthening of the DGHER . . . . . . . 25 Energy Pricing Policies . . . . . . . . . . .25 Hydroelectric Power Plant Studies . . . . . . . . . . . 25 Charcoal Efficiency Program . . . . . . . . . . . . . . 25 Improved Charcoal Stoves Program . . . . . . . . . . . 26 Household Energy Consumption Survey . . . . . . . . . . 26 Feasibility Study of Peat Substitution . . . . . . . . 26 Rural Electrification Master Plan . . . . . . . . . . . 26 Rural Distribution Projects . . . .27 Doubling of the Bubanza-Bujumbura Transmission Line. 27 Connection Program . . . . . . . . . . . . . . . . . . 27 Project Costs ..27 Project Implementation . . . . . . . . . . . . . . . . . . 28 Implementation of the Biomass and Household Energy Component . . . . . . . . . . . . . . 29 Implementation of the Rural Electrification Component . 29 Implementation of the Power Subsector Component . . . . 30 Consulting Services ..30 Procurement ..30 Financing Plan ..31 Disbursements ..32 Monitoring and Reporting ..33 Environmental Considerations . . . . . . . . . . . . . . . 33 V. FINANCIAL ANALYSIS .33 Past Financial Performance and Present Position . . . . . 33 Financial Restructuring .35 Arrears and Collection .37 Future Financial Position .38 VI. PROJECT JUSTIFICATION AND RISKS . . . . . . . . . . . . . . 39 Long-Run Marginal Cost of Electricity . . . . . . . . . . 40 Distribution and Connection Program . . . . . . . . . . . 40 Rural Electrification Program . . . . . . . . . . . . . . 41 Reinforcement of the Bubanza-Bujumbura Transmission Line . 43 Project Risks . . . . . . . . . . . . . . . . . . . . . . 43 VII. AGREEMENTS AND RECOMMENDATIONS . . . . . . . . . . . . . . 44 Conditions of Credit Effectiveness . . . . . . . . . . . . 46 Recommendation . . . . . . . . . . . . . . . . . . . . . . 46  Page No. TEXT TABLES: 1.1 Energy Balance: 1988 . . . . . . . . . . . . . . . . . . 2 2.1 REGIDESO: Future Demand of Electricity . . . . . . . . . 10 2.2 Actual Power Rates vs. Long-Run Marginal Costs . . . . . 4.1 Estimate of Project Cost . . . . . . . . . . . . . . . . 28 4.2 Procurement Method ..31 4.3 Financing Plan ..32 5.1 REGIDESO Operating Results for FY87-89 . . . . . . . . . 34 5.2 REGIDESO - Balance Sheets ..36 5.3 REGIDESO - Receivable Ratios . . . . . . . . . . . . . 37 5.4 REGIDESO - Operating Results for 1990-95 . . . . . . . . 38 6.1 Long-Run Marginal Cost of Electricity . . . . . . . . . . 40 6.2 Economic Viability of Rural Electrification Sub-Projects 42 ANNEXES: 1.1 Burundi Energy Supply - Demand Balance - 1988. . . . . . . . 47 1.2 Consumption of Petroleum Products 1981 - 1988. . . . . . . . 48 1.3 Official Price Structure of Gasoline . . . . . . . . . . . . 49 2.1 Urban and Rural Centers with Electric Service. . . . . . . . 51 2.2 Energy and Capacity Balances of the REGIDESO Electric System 52 2.3 Characteristics of existing hydroelectric power plants . . . 53 2.4 Future Hydroelectric Projects . . . . . . . . . . . . . . 54 3.1 DGHER - Finance and Accounting Expert . . . . . . . . . . 55 3.2 Performance Contract Between the Government and REGIDESO . 56 4.1 Estimate of Project Cost ..61 4.2 Implementation Schedule ..62 4.3 Organization and Functions of the PNU . . . . . . . . . . 63 4.4 Financing Plan . . . . . . . . . . . . . . . . . . . . . . 66 4.5 Disbursement Schedule . . . . . . . . . . . . . . . . . . 67 4.6 Supervision Plan ..68 5.1 Assumptions for Financial Analysis . . . . . . . . . . . . 70 5.2 Income Statements . . . . . . . . . . . . . . . . . . . . 74 5.3 Balance Sheets . . . . . . . . . . . . . . . . . . . . . . 75 5.4 Funds Flow Statement . . . . . . . . . . . . . . . . . . . 76 5.5 Income Statement Before and After Revaluation . . . . . . 77 6.1 Economic Analysis of the Program of New Connections and Extension of the Distribution System . . . . . . . . 78 6.2 Criteria for Project Acceptability . . . . . . . . . . . . 81 6.3 Economic Analysis of the MUGARA Rural Electrification Project . . . . . . . . . . . . . . . . . . . . . . . . 86 6.4 Economic Analysis of the CITEGA-RUTEGAMA Rural Electrification Project ..90 6.5 Economic Analysis of the KIRIYAMA-RUMEZA Rural Electrification Project .95 6.6 Economic Analysis of the GAZORWE Rural Electrification Project.. 99 6.7 Economic Analysis of the Reinforcement of the Transmission Line Between Bubanza and Bujumbura . . . . 103 7.1 Selected Documents Available in the Project File . . . . 106  - i - REPUBLIC OF BURUNDI ENERGY SECTOR REHABILITATION PROJECT Credit and Project Summary Borrower: Republic of Burundi Beneficiaries: Ministry of Energy and Mines (MEM), Ministry of Rural Development and Artisanat (MDRA), REGIDESO and ONATOUR. IDA Credit: SDR 16.3 million (US$22.8 million equivalent) Terms: Standard IDA, 40 years Onlending Terms: The government of Burundi will onlend a maximum of SDR 4.4 million (US$6.2 million equivalent) from the proceeds of the Credit to REGIDESO at 8.72 percent interest rate for a period of 25 years (including a 5-year grace period). The government will bear the foreign exchange risk. Proiect Oblectives: The broad objectives of the project are to promote rational energy policies and to strengthen the efficient management of energy resources. Specifically, the project aims to: (i) develop efficient institutions in the sector and improve the quality of public investment; (ii) improve the efficiency in the use of energy resources through reforms in the pricing structure of electricity, petroleum products and woodfuels; (iii) expand the access of the population to electricity; and (iv) reduce negative environmental effects of the use of energy through the execution of charcoal efficiency and improved stoves programs. Proiect Description: The project has two major components: institutional rerorm and small investments in biomass, household energy and power. The IDA Credit proceeds will support: (a) institutional strengthening of the Directorate of Energy and the Directorate of Water Resources and Rural Energies (US$0.8 million); (b) a rehabilitation program for REGIDESO, including a management assistance partnership program with a foreign operator, training and materials (US$5.1 million); (c) pre-feasibility and feasibility studies for selected hydropower plants and an update of the Power Sector Master Plan (US$3.6 million); (d) a charcoal efficiency program to disseminate improved techniques of production and sensitize charcoalers to the issue of deforestation (US$0.7 million); (e) an improved charcoal stoves program (US$0.4 million); (f) a household energy survey and a study of peat substi- tution (US$0.4 million); (g) a Master Plan for Rural Electri- fication (US$0.3 million); (h) extensions of the distribution  - ii - system and new connections in rural and urban areas (US$5.9 million); and (i) improvements in the transmission system (US$3.26 million). Project Benefits and Risks: The project will lead to increased efficiency in the economy and to an improved allocation of resources because of the proposed pricing policies for petroleum products, electricity and woodfuels. It will strengthen institutions in a key sector, thereby improving the quality of the public investment program and efficiency in the use of energy resources. The project will alco improve efficiency in woodfuels corsumption and contribute to the design of a strategy for the management of wood resources, thereby reducing negative environmental effects. The investments supported by the project are expected to yield substantial benefits and high rates of return. The risks of slippages in policy and institutional measures are considerably mitigated by substantial up-front actions and a broad consensus. No major technical risks are foreseen. Estimated Costs: FBu billions US*illions L F T L F T A. Energy Sector Institution Buildina RECIDESO Rehabilitation Prognrm Managemnt assistance 0.18 0.60 0.68 1.03 2.87 5.89 Building and material 0.84 0.82 0.66 1.94 1.84 8.79 Training 0.01 0.08 0.09 0.04 0.47 0.62 StrengthenIng of DCE 0.00 0.09 0.09 0.00 0.60 0.50 Strengthening of DCHER 0.04 0.00 0.04 0.20 0.00 0.20 Hydropower fes-ibility studies 0.02 0.59 0.61 0.10 8.38 3.48 B. Biomass and Household Eners Charcoal efficiency program 0.06 0.05 0.11 0.8 0.29 0.61 Ioproved charcoal stovo program 0.04 0.02 0.06 0.22 0.10 0.82 Household energy consumption survoy 0.01 0.02 0.08 0.06 0.11 0.17 Poat substitution study 0.00 0.06 0.05 0.00 0.26 0.26 C. Rural Electrification Master Plan 0.00 0.06 0.05 0.01 0.29 0.80 Extension proJecte 0.18 0.89 0.57 1.04 2.21 3.26 0. Power Subshctor 110 kV Transmission Line 0.08 0.64 0.60 0.82 8.09 8.41 Connections and distribution 0.28 0.47 0.69 1.80 2.66 8.66 E. ProSoct Mlansawnt A Administration 0.04 0.00 0.04 0.20 0.00 0.20 Total Bass Cost T 11 19X1 TMC 80:T 1T4 Duties and Taxes 0.26 0.00 0.28 1.69 0.00 1.59 Contingencios Physicol 0.18 0.81 0.49 1.01 1.79 2.79 Price 0.29 0.28 0.52 1.66 1.29 2.96 TOTAL PROJECT COST TV J70 r2 T.1 82.19  - iii - Financing Sources: Local Foreign Total Z of Total ---------------USS million-------------- IDA 3.2 19.6 22.8 70.8 Government 4.6 1.0 5.6 17.4 REGIDESO 3.8 -- 3.8 11.8 TOTAL 11.6 20.6 32.2 100.0 Estimated IDA Disbursements: IDA Fiscal Year FY91 FY92 FY93 FY94 FY95 ----------------US$ million------------- Annual 0.6 7.7 8.6 4.5 1.4 Cumulative 0.6 8.3 16.9 21.4 22.8 Economic Rate of Return: Physical investments 35 percent Map: IBRD 22083  REPUBLIC OF BURUNDI ENERGY SEC ^OR REHABILITATION PROJECT I. TEE ENERGY SECTOR Country Background 1.01 Burundi is a small landlocked country of 27,820 km2 with a population estimated at about 5.3 million. The population density is estimated at 179 persons per square kilometer, the second highest in Africa. Population has grown at a rate of about 2.9 percent per annum, but despite the demographic pressure on an already constrained land, urban migration has been limited and only 6 percent of the total population is urban. After Bujumbura with some 270,000 inhabitants, the next four towns have only between 10,00O and 20,000 inhabitants. 1.02 GDP per capita, estimated at about US$230 (1989), ranks among the lowest of the continent. The GDP growth has averaged 6.2 percent in real terms from 1985 to 1987, but was 5.2 percent in 1988 and only 1.5 percent in 1989. The country has limited natural resources other than a relatively fertile agricultural land. Agriculture is the predominant activity, contributing more than half of the GDP, 93 percent of employment and 88 percent of export earnings. Burundi is one of the few African countries that are self- sufficient in food. The most important export crop is coffee, which accounts for about 80 percent of export earnings. The secondary sector remains very small, accounting for about 14 percent of GDP and 8 percent of exports. Transportation costs to and from the Indian Ocean ports are high, and as a landlocked country, Burundi is vulnerable to transportation conditions in neighboring countries. 1.03 To correct economic and financial imbalances, the government embarked in 1985 on a comprehensive economic reform program that has been supported by two SALs (FY86 and FY88). The government's program to address structural issues includes a comprehensive package to strengthen macroeconomic management, create an incentive framework to stimulate exports and enhance private sector confidence, reform public enterprises, improve public expenditure management, increase the efficiency of the financial sector, liberalize the labor market and implement well-targeted measures aimed at alleviating poverty. The project described in this report is an important element of the government's economic restructuring program. It will address pricing issues in a key sector of the economy, implement needed reforms in one of the most important public enterprises, improve the quality of the public investment program, strengthen institutions in the energy sector and improve the reliability of energy supply for domestic and industrial uses. It will also address one of Burundi's major environmental concerns through the improved charcoal stoves and charcoal efficiency programs. Energy Consumption 1.04 The estimated final consumption of energy from all sources is 940,500 tons of oil equivalent (toe), corresponding to 188 kilos of oil equivalent (kgoe) per capita. Wood and other biomass meet almost 94 percent of the country needs, while petroleum products account for 5 percent, and electricity for  - 2 - 1 percent. A national energy balance for 1988 is provided in Annex 1.1 and is summarized in Table 1.1. The per capita consumption of modern fuels (petroleum products and electricity) is only 12 kgoe, one of the lowest in Africa 1/, and reflects the low income and urbanization levels, the dominance of subsistence agriculture in the economy and the small size of the industrial sector. However, consumption of modern fuels grew at an average annual rate of about 6 percent in the period 1980-88, considerably faster than population and GDP. The consumption of petroleum products increased by about 5 percent per annum and that of electricity at about 11 percent per annum. Households, which include small-scale commercial and artisanal activities, account for 92 percent of total energy consumption. Tablo 1.1: ENERGY SALANCE: 1906 Thousand Toe Percentage 1. Suppl AjiTcultural residues 106.0 11.8 Fuelwood 742.4 78.9 Charcoal 80.0 8.2 Peat 8.4 0.8 Petroleum products 49.9 6.8 Electricity 8.0 1.0 2. Demand lus holde 066.2 92.0 Industry 22.8 2.4 Pubilc Services 15.2 1.6 Transport 87.8 4.0 Source: ESMAP, Energy Assessment, 1990. Energy Resources 1.05 Biomass Resources. The total surface covered by natural forests and tree plantations is about 200,000 ha, or 7 percent of the country's total land area. Roughly, 135,000 ha are man-made plantations. Since 1978, the government has initiated several large-scale afforestation projects, and a total of approximately 46,000 ha have been planted with external funding. While these projects have been well implemented, their impact on the supply of wood has been limited due to the absence of a long-term forestry resource development strategy and to the weaknesses of existing forestry institutions. Private plantations represent 60,000 ha and provide most of the commercial wood consumed in the country. 1/ The equivalent figures for selected African countries are: Rwanda: 42 kgoe; Ethiopia: 21 kgoe; Benin: 46 kgoe; Niger: 42 kgoe; Somalia: 81 kgoe; Zambia: 380 kgoe. (Source: World Bank, World Development Report, Washington, D.C., 1989.)  1.06 Total wood consumption amounts to about 2.4 million metric tons per year, of which 17 percent is used for charcoal production. However, in terms of end use, charcoal contributes only to 4 percent of the energy requirements covered by woodfuels due to the low production efficiency of the carbonization techniques used. Purchases of wood are not prevalent, and it is estimated that only 10 percent of the wood consumed enters the commercial circuits. The low share of commercial wood indicates that demand management using taxation and the pricing system will have a limited impact on wood use and tree cutting. However, the negative impact on the environment is much larger for commercial than collected wood due to the intensive tree felling associated with commercial charcoal production. 1.07 The total sustainable supply of wood has been estinated at 1.7 million metric tons per year, which is 70 percent of the demand. The volume of wood demand over and above sustainable supply would represent a decline of about 2.6 percent per year of the total standing stock of trees. This fairly modest depletion of wood reserves could have a potentially serious impact due to the exponential pattern of depletion developments. So far, demand appears to have been met both in rural and urban areas without price increases in real terms, which could indicate that the country is at an early stage of wood resource depletion or that the data is inaccurate. In any event, there is a clear need to develop better information on the demand and supply side of woodfuels as a basis for formulating a long-term strategy for the sub-sector. The proposed project will support the execution of a household energy consumption survey (para. 4.12), while a complementary wood resource study will be financed by the French Ministry of Cooperation and Development and made available before June 30, 1992 to the Directorate of Energy (DGE) and to the World Bank. 1.08 Localized deforestation problems are already evident around some urban centers. At the same time, industrial plantations coming to maturity are seen to yield in excess of commercial offtake (1.2 million metric tons per year on a sustainable basis vs. 0.6 million ton per year). Access to some of these plantations is difficult because of the road conditions, but there is also no proper organization to manage the commercialization of wood. 1.09 Besides improvements in information and in the management of public industrial plantations, measures should be taken to conserve woodfuels. Consumption of woodfuels in traditional stoves is quite inefficient and there is considerable scope for improvements. Wood in rural areas is consumed in 3-stone open fires, with an efficiency estimated at between 10 and 15 percent. Improved wood stoves for the rural areas are the focus of an aid program from China monitored by the Ministry of Rural Development and Artisanat. A program to improve charcoal stoves in urban areas has been financed by IDA under the Power Transmission and Distribution Project (Cr. 1593-BU) and will be continued in the proposed project (para. 4.11). A total of 15,000 stoves, each saving between 30 and 35 percent of the charcoal used in traditional stoves, have been sold in Bujumbura since 1985, which is equivalent to 10 percent of the charcoal-using households. At the present time, 400 improved stoves are sold monthly. Inefficient charcoal making is also a serious problem since much wood is wasted in the process of production. More efficient charcoal production methods were introduced in some industrial plantations in the First and Second Forestry Projects (Credits 918-BU and 1620-BU), but it would be necessary to  -4- revitalize and reorganize these efforts and introduce them to traditional charcoalers who represent the largest share of produc'ion. Charcoal use is related to modernization and urban households will eventually switch to its use. At the present time, 85 percent of the population in Bujumbura uses charcoal and total consumption is estimated at 41,000 metric tons. Without energy efficiency improvements, total consumption would reach 97,800 metric tons in year 2000. The charcoal efficiency and improved stoves program (paras. 4.10 and 4.11), address the critical issue of meeting demand while reducing environmental damage. 1.10 Hydropower. Burundi is rich in hydropower resources. The theore- tical potential has been estimated at 6,000 Gwh per year, of which 1,500 Gwh are economically exploitable. This compares quite favorably with the current consumption of about 105 Gwh per year, indicating that the limited exploitation is related not to the resource base but rather to the economic and financial problems of expanding the electric system. Some of the most attractive national hydroelectric sites are located in the northwestern region of the country and the proposed project will finance the preparation of a limited number of pre-feasibility and feasibility studies (para. 4.09). Other potential sites must be developed in cooperation with other countries, as they are found along rivers forming boundaries. A feasibility study for Ruzizi III was started in 1989 under the sponsorship of the Energy Organization of the Great Lakes Region (EGL) constituted by Burundi, Rwanda and Zaire, while the study of Rusumo Falls was completed in 1987 under the sponsorship of the Organization for the Development of the Kagera River (OBK), constituted by Burundi, Rwanda, Tanzania and Uganda. 1.11 Present domestic hydropower facilities consist of 27 power plants with a total installed capacity of 32 Mw. Two of these plants (Rwegura with 18 Mv and Mugere with 8 Mw) represent 81 percent of the total installed capacity in the country. In addition, Burundi takes power from Ruzizi I and Ruzizi II. Ruzizi I is owned by the Societe Nationale d'Electricite (SNEL) of Zaire and has an installed capacity of 28.2 Mw. Ruzizi II, with an installed capacity of 26.6 Mw, is operated by the Societe Internationale d'Electricite des Pays des Grands Lacs (SINELAC), jointly owned by Burundi, Rwanda and Zaire, and was commissioned in 1989. These imports of electricity represented about 13 percent of consumption in 1989. 1.12 Peat. Peat is a non-renewable source of energy found in large quantities in the country. Highland bogs are easier to exploit and do not seem to pose significant environmental problems. Although lowland peat reserves are many times larger than highland resources, their extraction may give rise to environmental problems. The production of peat in 1989 was 14,212 metric tons, all from highland bogs. The limiting factor is demand, since studies indicate that production could be increased to some 50,000 metric tons per year without expansion into lowland bogs. 1.13 Present use of peat is basically limited to institutions (army, schools, hospitals, prisons) and some artisans (bakers, brick-makers, restaurants). It is estimated that the present consumption saves about 36,000 cubic meters of wood per year. Household consumption is insignificant because of its inconveniences relatively to fuelwood and charcoal (smoke, smell, ashes). Most of the drawbacks can be removed by carbonization and briquetting,  - 5 - but a study financed by USAID concluded that this was not financially attractive at the present prices of wood and charcoal. Peat could be used more intensively for process heat in cottage industries, bakeries, brick and tile factories, replacing wood or charcoal or as a substitute for fuel-oil or electricity in industrial boilers. The largest existing industries that could convert to peat (COTEBU-textiles, BRARUDI-brewery, VERRUNDI-bottles) have refused to consider this option, even though conversion to peat is economically and financially the best option. The proposed project will finance a feasibility study to demonstrate the benefits of increasing peat utilization in secondary industries (para. 4.13). 1.14 Petroleum. Petroleum products account for about 80 percent of commercial energy consumed ir. Burundi. There is no local production and national requirements are entirely covered by imports, representing close to 15 percent of total merchandise imports. Some exploration has been undertaken in the Ruzizi Plain, where AMOCO drilled three exploratory wells in 1987. The results were inconclusive and these activities are now on hold. Total consumption is about 50,000 metric tons and per capita consumption is only 10 kgoe, which ranks among the lowest in Africa. This is due to the modest size and low petroleum intensity of the industrial sector and the relatively high cost of petroleum products. Gasoline and diesel represent more than three- fourths of total consumption, while fuel oil used in some large industrial enterprises is only 14 percent (Annex 1.2). The use of kerosene, primarily for lighting, is very low and has barely increased during the 1980s. 1.15 Petroleum products are currently imported and distributed by five private companies. The largest is FINA/BP, which has a market share of about 35 percent. The government has a 20 percent participation in one of the companies (SICOPP, with a market share of 17 percent). It has not, however, used this to intervene in the market and confines itself only to supervision and control, using mainly price fixing, transport and stocks regulations and rules for establishing new societies. The private companies own the storage facilities near the port of Bujumbura. which have a total storage capacity of 12.000 cubic meters. For security reasons, the government built a new deposit of 20,000 cubic meters in Gitega in 1983/84, but it has not yet been used. 1.16 There are two surface routes for the imports of petroleum products. The "northern corridor" is the route Mombasa-Nairobi by pipeline (white products) or by truck (fuel oil), and Nairobi-Bujumbura through Uganda and Rwanda by truck, totalling a distance of 2,190 km. Total transport cost for gasoline is estimated at US$262 per metric ton. The "central corridor' has two alternatives: (a) Dar es Salaam-Kigoma by railroad and Rigoma-Bujumbura by barge on Lake Tanganyika, for a total distance of 1,428 km and a total transport cost for gasoline of US$107 per metric ton; and (b) Dar es Salaam- Tabora-Isaka-Bujumbura by truck, with a distance of 1,580 km and a total transport cost for gasoline of US$254 per metric ton. A comparison of the costs shows little difference between the two truck alternatives in financial terms, but some stretches of the road in Tanzania are not realistically accessible during the rainy season and the majority of transporters in Burundi refuse to go to Dar es Salaam. The railroad/barge route is by far the most attractive, but its use is limited by the railroad conditions in Tanzania and its low reliability. The private companies are free to negotiate their  purchases and to choose the supply route. It is estimated that only 8 percent of petroleum imports come through the Kigoma route. This situation may change if the railroad in Tanzania is improved. 1.17 Retail prices for gasoline, diesel and kerosene remained unchanged between December 1983 and September 1990. The price of fuel oil experienced some adjustments, but the overall increase was only 20 percent. Government revenues decreased several times to absorb increases in international prices and exchange rates adjustments. As of August 1990, the level of total taxation was 22 percent for gasoline, 19 percent for diesel, 18 percent for kerosene and 28 percent for fuel oil, about half the levels in neighboring countries that face similar high transport costs. In September 1990, petroleum prices were increased by about 30 percent, but these adjustments only partially reflected the new situation of international prices. Total taxation decreased to 20 percent for gasoline and 16 percent for kerosene, while it increased to 22 percent fcr diesel and 30 percent for fuel oil. Annex 1.3 gives the details of the pricing structure. While importers are allowed a somewhat bigger margin when they use the Dar es Salaam-Kigoma-Bujumbura route, this has not created proper incentives to increase the use of the low transport cost route. Also, the existing price structure is too complicated and regulates several specific items that are not relevant for the policy objectives of the government. Following an extensive dialogue, the government has agreed to implement a simplification of the pricing structure and adopt a policy of automatic price adjustments (para. 4.08). 1.18 New and Renewable Energies. Burundi is relatively well endowed with solar radiation and it is estimated that the installed capacity of photovoltaic systems is 45 kw, of which 40 percent is used for lighting, 40 percent for water pumping and the remainder for refrigeration and other uses. A few solar crop dryers and solar water heaters are also in operation, mainly financed by donors as demonstration projects. Prices for photovoltaic systems are high and solar energy applications are, with few exceptions, not economically viable. The expansion of its use should be left to the private sector to the extent that it is commercially justifiable, although the government should review the existing level of taxes and import duties for equipment and materials. 1.19 About 140 biogas digester plants have been built with the cooperation of external donors (Germany, China and Belgium). The two main constraints for the development of this resource are the availability of animals and the initial costs. Minimum requirements are for at least four heads of cattle which must stay in a stable or enclosure at night and few rural households meet this constraint. Secondly, a biogas digester costs about FBu 150,000 (US$857), which is too expensive for most rural families. The involvement of the government should mainly be in creating awareness about the advantages of using digesters and sensitizing potential new owners. Energy Institutions 1.20 The government agency responsible for the energy sector as a whole is the Directorate of Energy (DGE) in the Ministry of Energy and Mines, which has two departments: Projects, and Research and Statistics. The foremost task of the DGE is that of keeping an overview of the energy sector as a whole,  assessing requirements and priorities between subsectors, giving general guidelines for subsector development and ensuring coordination. This requires a competent planning staff with a strong background in economics, which does not exist at the present time. The proposed project will provide support to the DGE in these aspects (para. 4.06). Clear procedures and uniform investment criteria for evaluating projects and programs will be established and monitored by the DGE. Physical planning and implementation for each individual subsector, on the other hand, would be the responsibility of the subsector agency. 1.21 On the supply side of the biomass subsector, the central institution is the Forestry Department (DF) in the Ministry for Regional Development, Tourism and Environment. Its responsibilities include the implementation of the government's forestry policies, the management of reserved natural forests and the management of man-made industrial plantations. These industrisl plantations include four externally-financed forestry projects with a substantial technical assistance component for the DF. The Ministry of Agriculture intervenes when the exploitation of wood and biomass is integrated with farming activities. Promotion of efficiency in charcoal production has also been the responsibility of the Forestry Department, while promotion of efficiency in the end use of wood and other biomass has been undertaken by the Directorate of Water Resources and Rural Energy (DGHER) in the Ministry of Rural Development and Artisanat, by the DGE, and by the Office National de la Tourbe (ONATOUR). The DGHER is also responsible for rural electrification and for the promotion of new and renewable energy resources (biogas, solar, wind). where it overlaps with the Department of Research and Statistics of the DGE, as well as with the Burundese Research Center for Alternative Energy Resources (CEBEA). The responsibility for the production, marketing and promotion of the use of peat is placed with ONATOUR, a parastatal company created in 1977 and reporting to the Directorate of Mines. The Department of Research and Statistics of the DGE undertakes household energy surveys and its activities in this area would be strengthened through the proposed project (para. 4.12). 1.22 In the electricity sector, REGIDESO is responsible for the generation, transmission and distribution in urban areas, while rural areas fall under the authority of the DGHER. This division of responsibilities seems well justified. Urban and rural electricity supply are rather different in character and call for different organizational approaches. The DGHER, being part of the Ministry of Rural Development, has its attention focussed on the rural environment and has the motivation to search for solutions adapted to each individual rural center. Both REGIDESO and the DGHER will receive assistance from the proposed project to improve their respective performance (paras. 4.03 and 4.07). 1.23 In the petroleum subsector, the principal government agency involved is the Directorate of Commerce (DGC) within the Ministry of Commerce and Industry. This unit regulates all types of commercial activities, but it has no particular expertise in the petroleum sector and cannot ensure coordination with other energy subsectors or give guidelines for subsector development. For these purposes, the DGC needs to coordinate with the DGE, but the latter needs to develop some competence in this area and support for this would be provided under the project.  1.24 An important new body is the National Commission for Water and Energy, created in 1989, where all the ministries involved in the energy sector are represented. The role of the Commission is consultative to the government, but once its recommendations are approved, they become official policy. The main objective of the Commission is to promote the harmonious and rational utilization of water and energy resources. The sub-commission on energy is chaired by the Minister of Energy and Mines, with the DGE serving as secretariat. This will lend considerable authority to the DGE and ensure effective communications with all the ministries involved in the energy sector. Government Objectives 1.25 The Ministry of Energy and Mines issued a basic document on energy sector policy in July 1989. This document states that the objectives of the government in the sector are to: (i) develop and rationally manage national energy resources; (ii) provide appropriate energy for the development of rural areas; (iii) provide low-cost energy for industry and craftsmanship; (iv) maintain the existing energy infrastructure and make it profitable, and (v) reduce foreign energy dependence, particularly on imported petroleum products. To implement these objectives, the government has decided to strengthen the DGE in the areas of planning, coordination and control, to promote the connection of the most important rural centers to the electric grid, to reinforce the training of charcoalers in more efficient techniques, to extend the improved charcoal stoves program and to review the pricing policies for electricity, wood and charcoal. The proposed operation would help the government in implementing its strategy for the sector by strengthening key institutions, financing selected investments and promoting pricing reforms. Bank Participation in the Sector 1.26 IDA provided funds for a project to expand the power transmission and distribution system of REGIDESO with a Credit for SDR 12.9 million (Cr. 1593-BU) signed on June 14, 1985. The project included: (a) a 110 kv transmission line from Ruzizi II to Bubanza; (b) subtransmission lines to supply five regional towns and distribution networks in these towns; (c) distribution networks in six low-income districts of Bujumbura; (d) rehabi- litation of the 70 kv transmission line from Ruzizi I to Bujumbura; (e) construction and equipping of a training center for REGIDESO; (f) a program for improved charcoal stoves in Bujumbura; and (g) consulting engineering services and technical assistance. The rehabilitation of the transmission line from Ruzizi I to Bujumbura was not undertaken under the project because of legal issues with the government of Zaire, but has been incorporated into a project under execution in Zaire. The training center is under construction, and the execution of the improved charcoal stoves program continues, but all the other project components have been completed. The project achieved its physical objectives, but the technical assistance was not successful because of its limited nature and lack of commitment from high level management. To improve the performance of REGIDESO, a comprehensive rehabilitation program has been included in the proposed project. This program is part of a Performance Contract signed between the management of REGIDESO and the government and is monitored by an interministerial monitoring committee. A foreign operator has been called to assist in managing the enterprise (paras. 3.12 and 4.03).  1.27 IDA also financed the construction of the Ruzizi II hydroelectric power plant that is jointly owned by Burundi, Rwanda, and Zaire (Credits 1419-BU, 1420-RW and 1421-ZR) and two forestry projects (Credits 918-BU and 1620-BU) which have subsequently created important new energy supply options. The Ruzizi II power plant has been successfully completed and is operational since June 1989, but the issue of pricing the energy sold has not been fully resolved. The forestry operations established about 13,000 ha of pine and eucalyptus plantations and provided some training for charcoalers. A more efficient model of charcoal stoves was developed under the Urban Development Project (Cr. 1049-BU). The Nickel Exploration/Engineering Project (Cr. 1154-BU) helped to finance the trial peat production in the Akanyaru Basin. Under the Joint UNDP/World Bank Energy Assessment Program, a complete review of the energy sector was done in 1981 and in 1990. As indicated in para. 6.01, this last study supplied the basic framework for the proposed project. II. THE POWER SUBSECTOR Demand for Electricity 2.01 The consumption of electricity in Burundi reached a level of 105.3 Gwh in 1988 and of 102.2 Gwh in 1989, representing a per capita consumption of around 20 kwh, one of the lowest values in the world. Only 1.5 percent of the total population uses electricity and 82 percent of the total consumption is concentrated in Bujumbura. Sales by REGIDESO are 92 percent of the total, while auto-producers account for about 7 percent of consumption and the DGHER for about 1 percent. The average annual rate of growth of consumption was 10.8 percent in the period 1980-88 and 6.8 percent in the period 1985-88. In 1989, there was a decline in the consumption of electricity of around 3.0 percent, which is explained by the temporary closing of the most important consumer (VERRUNDI-bottle manufacturer) and a general decline in industrial activity. Industry and commerce account for about 47 percent of the total consumption. 2.02 The number of consumers has increased from 5,408 in 1980 to 13,799 in 1989, an average annual growth rate of 10.6 percent. The DGHER has 362 users, and REGIDESO 13,437, of which 13,283 are low voltage and 154 medium voltage. The number of new clients connected to the system of REGIDESO was 645 in 1988 and 1,507 in 1989. The increase is explained by the new policy on connections established within the rehabilitation plan of the institution. The 10 largest users of electricity are VERRUNDI (bottle manufacturer), COTEBU (textiles), BRARUDI (brewery), BRAGITA (brewery), OCIBU (coffee), the University of Burundi, three hotels in Bujumbura and REGIDESO itself. In a normal year, these users account for about 45 percent of the total consumption of electricity. 2.03 Consumption per domestic household is estimated at 2,300 kwh per year and has declined systematically over the years. This phenomenon is explained by the fact that the expansion of electrification progressively reaches areas of lesser development and lower incomes. Given the low proportion of the population which has access to electricity, this trend is likely to continue in the future. Before 1980, only two towns (Bujumbura and Gitega) had electric  - 10 - service. At the end of 1989, there are 54 centers with electricity (Annex 2.1), of which 25 are located in the interconnected grid. Bujumbura has about 270,000 inhabitants and the next four towns (Gitega, Ngozi, Kayanza, Rumonge) have between 10,000 and 20,000 inhabitants. There are only 18 urban centers with a population of more than 5,000. This demographic structure is very costly to the objective of increasing the access of population to electric services, and makes rural electrification of crucial importance for the country. 2.04 Another major obstacle to expand the number of consumers has been the high cost of connections. Until July 1989 all new users had to make a lump-sum payment equivalent to the cost of the individual connection plus a participation in the extension of the grid. Since high-standard imported materials were used, the charges were quite high. Average connection charges have been estimated between US$500 and US$750. In 1989, these charges were decreased to cover only the individual connection, and term payments of up to 24 months were allowed. Low cost equipment has been imported and average connection charges have decreased to about FBu 35,000 (US$200). This largely explains why 1,507 new users (a 9.3 percent increase) were connected in 1989. Nevertheless, the electrification rate in Bujumbura is estimated to be only around 20 percent, and a substantial number of connections requests remains unfulfilled because of deficiencies in the supply of materials and in organization and planning. The new connection policy of REGIDESO will be supported by improvements in the internal organization and planning for the execution of connections, which are included in the Rehabilitation Plan of REGIDESO to be financed under the proposed project. Financing for an intensive program of new urban connections (para. 4.17) and for rural electrification (para. 4.15) is also included. 2.05 Statistical methods are not appropriate to project the future evolu- tion of electricity consumption in Burundi. Growth will depend fundamentally on actions by the government and REGIDESO to increase the number of new connections or on the connection of some new large industrial consumers, which is unlikely. An average growth rate of 6 percent per year has been applied for planning purposes. This rate is similar to the rate observed between 1985 and 1988 and has been obtained using the methodology applied by Electricite de France International (EDFI) in the Power Sector Master Plan. Production requirements are estimated on the basis of a gradual decrease of total losses from the level of 20 percent observed in 1989 to 15 percent by 1995 and to 12 percent by 2000. Peak demand is calculated on the basis of 5,100 hours of utilization. The values of consumption, production and peak demand for the period 1989-2000 are given in Table 2.1 for the electric system of REGIDESO. Table 2.1: REGIDESO - FUTURE DEMAND OF ELECTRICITY 1989 !/ 1990 1991 1992 1998 1994 1996 1996 1997 1996 19m 2000 Consumption (Owh) 98.8 99.4 106.4 111.7 118.4 126.6 188.1 141.0 149.6 158.5 168.0 178.1 Production (Cwh) 117.8 124.8 130.1 186.2 142.7 149.4 168.6 164.0 178.8 182.2 198.1 202.4 Peak demand (Mw) 28.0 24.4 26.5 26.7 28.0 29.3 80.7 82.8 84.1 85.7 87.9 89.7 */Actual figures excludes DCER and auto producers. Vource: ESMAP, Enercy Assessment, 1990.  - 11 - 2.06 Based on these demand projections, energy and capacity balances were constructed and are presented in Annex 2.2. The guaranteed energy from existing domestic hydroelectric plants is estimated at 102.0 Gwh, and the energy available to Burundi from Ruzizi I and Ruzizi II Gwh under the agree- ments already in pla-e is 59.7 Gwh. In this situation, existing plants and the available supply from Ruzizi I and II can satisfy all consumption requirements until 1995. However, REGIDESO is negotiating with SNEL an increase in the energy available from Ruzizi I, and a recent hydrology study in the region indicates a possible increase in the energy available from Ruzizi II (SINELAC, R66valuation du Productible de la Centrale Hydroelectrique Ruzizi II, February 1990). Under these circumstances, an additional power generating plant would not be needed before 1998. If the third unit of Ruzizi II is added before that date, the need for a new plant may even be postponed for another year. Generating Capacity 2.07 Total production of electricity in Burundi was 124 Gwh in 1988 and 126 Gwh in 1989, including 116.6 Gwh and 117.3 Gwh respectively from the system of REGIDESO. There are 27 hydroelectric power plants in operation in the country with a total installed capacity of 32.2 Mw. REGIDESO operates seven plants with a total installed capacity of 30.6 Mw, which represents 95 percent of the total. Two of these plants represent over 80 percent of the total installed capacity in the country (para. 1.11). The DGHER operates eight plants with a total capacity of 0.8 Mw and auto-producers, most of which are religious missions, have 12 hydroelectric power plants with a total installed capacity of 0.8 Mw. Of the total hydroelectric power plants, 17 have an installed capacity of less than 200 kw (Annex 2.3) and can be classified as micro-hydro. 2.08 Besides the domestic power plants, Burundi gets hydroelectric energy from Ruzizi I and Ruzizi II. The average energy that can be produced in Ruzizi I is estimated at 148 Gwh per year. A rehabilitation of this plant is under execution. The energy from Ruzizi I is received as payments for debts of SNEL with REGIDESO and with the Banque de la Republique du Burundi (BRB). Payments for the first debt are fixed at 12.8 Gwh per year until July 2005, and for the second one at 17.6 Gwh per year until June 1991. However, REGIDESO has not consumed all the energy that is due, and SNEL has accumulated debt over and above what is established in the agreements. At the end of 1989, the balance to be paid by SNEL was 66.8 Gwh on the convention SNEL/BRB and 255.3 Gwh on the convention SNEL/REGIDESO. Ruzizi II is jointly owned by Burundi, Zaire and Rwanda and operates since July 1989 with two units and a total installed capacity of 26.6 Mw. Average available energy is estimated at 141 Gwh per year, of which REGIDESO is assumed to use one third. Recent studies suggest that the available energy could be increased to 197 Gwh with two units and 223 Gwh with three units (para. 2.06). 2.09 REGIDESO has ten thermal power plants with an installed capacity of about 10.0 Mw. Several of these groups are old, and it is estimated that only about 5.0 Mw are really available. The biggest thermal plant is in Bujumbura and has a nominal capacity of 7.4 Mw, but since several of its units date from 1953/54, the available power is only about 3.0 Mw. This plant is used mainly for reserve purposes. The production of electricity from the thermal plants of REGIDESO is about 2 percent of total production and is likely to decline as  - 12 - some additional isolated centers are connected to the electric grid where production is entirely hydroelectric. Thermal units installed by self- producers have an estimated capacity of 6.3 Mw, with the largest groups in some tea factories (Tora: 456 kw, Rweguras 468 kw, Tezat 375 kw). Transmission and Distribution 2.10 A map in the back of the report shows the power network of Burundi. The interconnected system includes the province of Bujumbura, Bubanza, Cibitoke, Kayanza, and Ngozi, and the localities of Tora and Mugamba in the province of Bururi, and Gisozi, Mwaro and Kibumbu in the province of Muramvya. Gitega and Muramvya will be incorporated in 1991 when the 110 kV transmission line between Bujumbura and Gitega is completed. The principal transmission lines are the 110 kV lines from Ruzizi II and Rwegura to Bujumbura (RN1), the 70 kV line from Ruzizi I to Bujumbura (SNEL), the 35 kV line from Mugere to Bujumbura (Ozone) and the 30 kV lines that feed Kayanza, Ngozi, Ijenda, Tora and other localities. The total length of the high and medium voltage line is estimated at 818 km. 2.11 While the interconnected network is being expanded radially, there are two separate sub-systems under development. The first is in the south and will interconnect Nyanza Lac, Rumonge, Makamba, Rutana, Gihofi and Bururi with the Nyemanga hydroelectric power plant (1.4 Mw) through 30 kV transmission lines. The second is in the north-east and will connect Cankuzo, Muyinga, Karuzi and Kirundo with the micro-hydropower plants of Murore (24 kw), Kayenzi (800 kw), Buhiga (240 kw), and Marangara (240 kw), also through 30 kV transmission lines. Future Developments in Power Generation 2.12 The theoretical hydroelectric potential of Burundi has been estimated at 6,000 Gwh per year (Lahmeyer International, Etude du developpement des ressources hydroelectriques du Burundi, August 1983), but in practice only around 1,500 Gwh in some 40 projects could be economically exploited. The most attractive potential national projects are located in two distinct regions of the country: (i) the northwestern region, where the Gitenge-Kagunuzi and the Kaburantwa rivers are important tributaries on the east side of the Ruzizi river and have an estimated potential of about 100 Mv, and (ii) the southern region, where the Mulembwe, Jiji, Ru7ibazi and Miyovozi rivers have a potential of about 24 Mw. As indicated above (para. 1.10), other possibilities that would need to be exploited in cooperation with other countries are found on the Ruzizi river (Zaire and Rwanda) and on the Ruvubu-Kagera rivers (Rusumo Falls with Tanzania, Rwanda and Uganda). 2.13 Of the existing power plants, Rwegura (18 Mw) is located in the northwestern region, while Nyemanga (1.4 Mw) is in the southern region. There are possibilities of expanding the capacity of Nyemanga to 2.8 Mw, but this will have to wait for the expansion of consumption in the southern grid. The characteristics of the principal possible domestic new projects are given in Annex 2.4 with their costs per installed capacity and per average production. The most interesting domestic projects are Mule 34, Jiji 03, Kabu 16 and Kabu 23. Since Mule 34 and Jiji 03 are in the southern region,  - 13 - which is not connected to the main network serving the principal consumption centers, the costs of constructing the corresponding interconnection transmission lines must be added to the total cost of these projects. 2.14 In the northwestern region, the Kaburantwa and Kagunuzi rivers can be developed independently or jointly. In the first case, four sites can be developed on the Kagunizi river (Masango, Rushiha, Kagu 10 and Kagu 06) down- stream from the Rwegura power plant and two sites can be developed on the Kaburantwa river (Kabu 16 and Kabu 23). If the rivers are jointly developed, a big reservoir would be created on the Kagunuzi river (the Kagunuzi C project) and used to irrigate the Imbo valley to the south and, at the same time, to produce electric energy. In a next stage, the waters of the Kaburantwa would be diverted through a tunnel to the reservoir and used either in an extension of the Kagunuzi power plant (the Kagunizi D power project) if there is need of additional water for irrigation or, if not, in an independent power plant (Kagunuzi A). The joint development of the rivers implies the disappearance of the Kabu 23, Kabu 16, Kagu 10 and Kagu 6 projects and their replacement by Kagunuzi C plus the siphoning of the Kaburantwa river plus either Kagunuzi A or Kagunuzi D. 2.15 A comparison between the independent and joint development of the Kaburantwa and Kagunuzi rivers was done in the Power Sector Master Plan (Electricite de France International, Plan Directeur National d'Electri- fication, December 1988). According to this study, with equal conditions of electric service and at a discount rate of 10 percent, the joint development of both rivers has a cost that is 35 percent higher than the independent development in the case of limited irrigation (Kagunuzi A) and 50 percent higher in the case of maximum irrigation (Kagunuzi D). 2.16 The only domestic project for which a feasibility study has been done is Kagunuzi C, which is not the least-cost alternative for the expansion of the electric system. The proposed project will finance pre-feasibility studies for Kabu 16, Kabu 23, Rushiha and Masango, which are some of the best sites available in the country (para. 4.09). Since the interconnection with the southern grid has not been done and is not envisaged in the near future, the studies of Mule 34 and Jiji 06 have been postponed. The studies financed under the proposed project, together with those of Kagunuzi, Ruzizi III and Rusumo Falls, would be used for the selection of the future least-cost power plant that is required for the expansion of the electric system. Operations and Maintenance 2.17 The comparison between production and consumption in 1989 (111.3 Gwh versus 93.8 Gwh) implies that total losses in the electric system of REGIDESO are of the order of 20.0 percent. The comparable figure for 1988 was 15.6 percent, which indicates a substantial increase in losses. These figures include the consumption of the auxiliary services and of the resistances in the power plants that are necessary to avoid cavitation in the machines. If these uses are excluded, total losses were 18.7 percent in 1989 and 13.9 percent in 1988. These magnitudes indicate that actions to improve performance and reduce losses are required. A detailed plan to reduce technical and non-technical losses will be presented to the Bank before September 30, 1991.  - 14 - 2.18 Losses in transmission lines and their associated transformers were of the order of 3.7 percent of net production in 1989, which is somevhat on the high side given the small network of REGIDESO. This is specially the case for the 35 kV line Mugere-Bujumbura (Ozone), where losses are calculated at 5.2 percent of net production. Losses at the distribution level are high (15.6 percent of the energy supplied in 1989 and 12.5 percent in 1988), which is explained in part by the old age of the network in Bujumbura and Gitega, which is in need of urgent rehabilitation. While there is no data to separate technical and non-technical losses, there are indications that the latter are also important, due to observed deficiencies in meters, metering and billing. An ongoing intensive program of testing and recalibrating the electric meters in Bujumbura has found deficiencies in about 20 percent of them. As part of the rehabilitation plan, REGIDESO will continue to check all electric meters in its system and adopt strict measures of control to avoid billing mistakes and the theft of electricity, thereby reducing its non-technical losses. At the same time, the rehabilitation of the distribution systems of Bujumbura and Gitega will be undertaken with financing from KfW-Germany which will also reduce losses. 2.19 The number of incidents that interrupt electric service is also quite high. For example, during March 1989, there were five recorded interruptions at the level of high/medium voltage and 27 incidents in the low-voltage distribution system of Bujumbura. In October 1989, there were eight and 25 incidents respectively. While some of these incidents are due to natural causes that cannot be avoided, the quality of maintenance of the electric network is, in general, low. The personnel in charge of maintenance lacks adequate training, the stocks of spare parts and materials (especially in the interior of the country) are inadequate and there are deficiencies in the logistical support necessary for effective maintenance (communications and transport). The planning of maintenance is too general and, with the exception of Rwegura, there are no detailed programs of preventive maintenance for each of the power plants, sub-stations and high/medium voltage transmission lines. REGIDESO will present a detailed preventive maintenance plan for all its system before September 30, 1991. 2.20 The operation of the principal newer hydroelectric power plants, where service started quite recently (Rwegura, Mugere, Nyemanga), is satisfactory, but almost all the other hydroelectric power plants of REGIDESO and DGHER have experienced operational difficulties. Giheta (1983), Ruyigi (1982) and Buhiga (1984) were not in operation at the end of 1989, Ruvyironza (1980-84) has had problems in all its turbines and Marangara (1986) was out of service several times in 1989 due to deficiencies in lubrication, regulation and electronic equipment. 2.21 Isolated centers that are supplied exclusively from thermal power plants have electric service only during some hours of the day. This is due to the high cost of operation and the low level of revenues. In some places (Rutana, Cankuzo), provincial authorities pay directly for the fuel and have even received foreign assistance for this purpose. The state of these thermal groups is unknown, but routine maintenance is not done at all or is deficient. Whenever there is a failure, the community remains without service for extended  - 15 - periods of time. Since most of these centers should be under the responsibility of the DGHER, the problem will be examined within the Master Plan for Rural Electrification to be financed under the proposed project. Electricity Tariffs 2.22 Electricity tariffs in effect in January 1990 were approved by a decree of the Minister of Commerce and Industry dated March 29, 1989. For low-voltage users, there was only a constant energy charge of FBu 15/kwh (US$0.086 per kwh). Medium-voltage users had a fixed charge for subscribed capacity of FBu 6,000/kw per year (US$34.3 per kw par year) and an energy charge that varies with the hours of utilization: (i) FBu 13/kwh (US$0.074 per kwh) for the first 150 hours of utilization of the subscribed power; (ii) FBu 12/kwh (US$0.069 per kwh) for a utilization between 150 and 450 hours; and (iii) FBu 9/kwh (US$0.051 per kwh) for utilization above 450 hours. The decree of March 29, 1989 replaced tariffs that existed since March 1, 1988. 2.23 Studies of the long-run marginal cost of electricity and the reform of the tariff structure were done by Electricite de France in April 1985 and December 1988. A more detailed and complete study was done by an independent consultant in December 1989 (Fernando Lecaros, Etude de Restructuration des Tarifs d'Electricite) as part of the preparation for the proposed operation and, based on it, the government decided to implement in April 1990 a comprehensive electric tariff reform. Medium-voltage electric rates were immediately adjusted to the estimated long-run marginal cost values, while for low-voltage users, the adjustment was phased in with an initial increase of 27 percent for all consumption above 375 kwh per month. A social tariff for consumers that use less than 75 kwh per month was established. They are at the present time about 3,600 consumers in this group (approximately 30 percent of the total), but they represent only about 1.6 percent of total consumption (or 7.4 percent of residential consumption). The tariff reform also eliminated the free electricity received by high government officials and the personnel of REGIDESO. Table 2.2 presents the electricity rates as of January 1990 and April 1990 and the long-run marginal costs. Table 2.2: ACTUAL POWER RATES VS. LONW-RUN MARGINAL COSTS January 1990 April 1990 LRMC 1. Medium vOlta9-: - Powr (Fuu/kw/yoyr) 6,000 7,525 7,625 - Energy (F9u/kwh) 0 - 160 hours 18.0 16.7 18.7 150 - 450 hours 12.0 12.0 12.0 More than 460 hours 9.0 6.8 6.8 2. Low voltas": (FBu/kwh) - Socil tarTff 16.0 12.0 - - General tariff 0 - 760 kwh/bi-monthly 16.0 16.0 22.5 More than 760 kwh/bi-monthly 16.0 19.0 22.5  - 16 - Rural Electrification 2.24 The Directorate of Water Resources and Rural Energy (DGHER) of the Ministry of Rural Development has developed several micro-hydro power plants (Annex 2.3) and built transmission and distribution lines to provide services to some rural centers. Because of a lack of trained personnel in the DGHER and the small size of the plants and their dispersion, these actions have not been very successful. The operation of the plants is always precarious and inter- ruptions of service and extended repairs are common. Administration, billing and collection are inefficient. Improvements in management, operations and maintenance of the DGHER electric system are urgently needed and are included as part of the proposed project (para. 4.07). 2.25 At the present time, of the 115 municipalities, 32 have their main village electrified, while 83 completely lack electric service. The Ministry of Energy and Mines has recently completed a catalogue of these villages, with a preliminary indication of the potential markets and possible source of electric energy, but priorities have not been determined, and no economic analysis of individual projects has been done. New micro-hydro power plants are expensive, and the approach is now to connect new rural lines to the network of REGIDESO. Since the distances are relatively small and there is excess energy in the network, this decision could make economic sense. The responsi- bilities for the administration, operation and maintenance of the new rural systems also has to be defined. DGHER has not demonstrated effectiveness in administration and most of the rural centers are not financially attractive to REGIDESO, which is itself struggling to become a profitable commercial enterprise. An alternative would be to make the municipalities responsible for the administration and have them pay REGIDESO for the required technical works, but this possibility requires further analysis of the financial situation of municipalities and assistance to organize the management of the municipal electric systems. Another alternative is to use some local institutions (church, cooperatives) that have close links to the community. All these issues will be examined in the Rural Electrification Master Plan financed under the proposed project (para. 4.14) TII. THE IMPLEMENTING AGENCIES 3.01 The Borrower will be the government of the Republic of Burundi. The Ministry of Energy and Mines (MEM) will be responsible for the overall coordination and supervision of the program and will directly implement some of the activities in the biomass and household energy components. REGIDESO will be the implementing agency for the power subsector components and the DGHER for rural electrification. A. REGIDESO Organization and Management 3.02 REGIDESO is a government-owned public enterprise created and organized by the Decree No. 1/196 dated October 2, 1968. Its objectives are E to generate, transmit and distribute electricity, to pump, treat and distribute  - 17 - water and to execute or supervise studies and works for new electricity and water facilities. REGIDESO is under the supervision of the Ministry of Energy and Mines (MEM). Responsibility for day-to-day operations lies with the General Manager, who is assisted by three managers, all appointed by the President of the Republic. The Technical Manager is in charge of electricity and water operations, maintenance and garage, and supervises two deputy managers who are responsible for the electricity and the water departments. The Administrative and Financial Manager is in charge of four units: personnel, accounting, corporate budgeting and control, and inventories supplies. The Commercial Manager supervises three unitss billing, revenues collection and relations with clients. Recently, the utility has been reorganized into four regions, each of them headed by a regional manager directly reporting to the General Manager. Staffing and Personnel 3.03 At the end of December 1989, REGIDESO employed 899 permanent agents and about 1,100 temporary agents. Temporary agents are supposed to be used only in the execution of new projects, but in practice many work on a permanent basis. The ratio of total electricity and water customers to permanent agents was 27;1, which represents a low level of efficiency. In the Performance Contract signed with the government in July 1989, REGIDESO agreed to increase this ratio to 50:1 by 1992, which would imply a substantial reduction in the number of permanent agents (up to at least 300). There are 70 cadres (managers and university graduates), or 8 percent of total staff, and 829 skilled and unskilled workers. Staff is allocated as follows: 445 for electricity, 236 for water, and 118 for administration and finance. 3.04 REGIDESO has problems with regard to the capability and efficiency of its work force. These problems are primarily the result of difficulties in recruiting qualified personnel. To upgrade the job qualifications of existing employees, the previous IDA supported project (Cr. 1593-BU) financed the construction of a training center and the design of a training program to increase the skills of specialized workers in mechanics and electricity, both to be executed by GTZ-Germany. The construction, however, was delayed and REGIDESO has not fully implemented the training component defined in the project. The construction of the training center will be completed by March 1991 and GTZ has updated the training programs and will start implementation immediately after construction. Promotion and salaries incentives are given on the basis of seniority and hierarchy rather than on qualifications and job performance. As a result, in many cases supervisors lack the skills and knowledge required for their jobs. 3.05 The Rehabilitation Program that will be financed under the project (para. 4.03) includes an expert in human resources and personnel management, who will be in charge of introducing new personnel management policies in the areas of recruitment, performance evaluation and salary incentives. He will also supervise all training activities. The project will also include training for managers, supervisors and non-technical personnel (administration, accounting and commercial aspects), which were not included in the previous project. During negotiations, the government agreed that by September 30, 1991: (a) REGIDESO would complete a detailed manpower study that includes an  - 18 - assessment of required positions and a strategy to reduce personnel and increase productivity, and (b) REGIDESO would submit a detailed training plan for all its staff. Accounting, Audit and Insurance 3.06 REGIDESO has an acceptable accounting system and maintains its books in accordance with internationally accepted accounting practices. Accounts are kept on an historical cost basis and foreign debt and transactions are revalued annually at prevailing year-end exchange rates. All exchange gains/ losses are treated in accordance with international standards. However, the obsolete status of the mechanographical system used to maintain the accounts causes significant delays in the flow of accounting information. The Financial Department requires assistance to help strengthen management, and the accounting systems and procedures need to be computerized. Urgent actions needed include the recruitment of a new Accounting Chief, proper staffing of the newly created Internal Auditing Department, and the design and implemen- tation of an appropriate management information system. Under the proposed rehabilitation program, REGIDESO would implement, with the support of both new recruits and the management assistance team, the necessary changes and improvements. During negotiations, assurances were obtained from REGIDESO that recruitment of the chiefs of the Accounting Division and the Internal Auditing Unit would take place before Credit effectiveness. 3.07 REGIDESO's accounts for FY88 and FY89 have been audited by external auditors. The audit report for FY88 was heavily qualified, particularly regarding REGIDESO's inability to reconcile the general balance with a number of individual balances and the inventory levels with the use of spare parts. Furthermore, the reports indicated that there is no reliable accounting system for capital works and that the costs of projects are obtained from incomplete estimates. These deficiencies were partially corrected in FY89, but the shortages of qualified personnel and equipment limit what can be effectively done. With the improvements envisaged in the flow of financial information through the automation of its accounting and financial information system and the strengthening of the internal audit function, it is expected that REGIDESO would be in a position to correct the noted deficiencies. During negotiations, assurances were obtained that REGIDESO would continue to provide the Bank annually with audited financial accounts no later than six months after the end of each fiscal year. 3.08 REGIDESO insures against most common risks associated with electricity operations with the National Insurance Co. of Burundi, which is a government-owned concern. The coverage does not reflect the devaluations of the Burundi Franc, nor does it take into account the realistic values of its assets. REGIDESO would review its insurance practices to ensure adequacy and cost effectiveness, and bid for competitive insurance premium costs, as there is now a private insurance company in Burundi. Management Information Systems 3.09 REGIDESO suffers from poor internal communications and the lack of timely information necessary for an effective management. With the exception of the computerized fixed asset management system financed through Cr. 1593-BU  - 19 - and recently implemented, REGIDESO lacks major automated systems. The Rehabi- litation Program financed under the proposed project (para. 4.03) will assist REGIDESO in establishing (a) appropriate policies, practices and business procedures for improved consumer billing, collection procedures, budget management and accounting; (b) a comprehensive data processing plan; and (c) requirements for computers and software. The project will also finance the purchase of some urgently needed equipment. Assets Revaluation 3.10 The study to revaluate REGIDESO's fixed assets was financed under Cr. 1593-BU. The historical value of fixed assets in service during 1988 should increase, according to the study, from FBu 21.8 billion tL. FBu 53.4 billion. This change would increase depreciation charges to more than FBu 2 billion and substantially increase the accounting losses of REGIDESO (para. 5.04). To compensate for this situation, and as part of the financial restructuring of REGIDESO (para. 5.08), the government has agreed to take responsibility for unprofitable past investments and allow REGIDESO to deduct from its operating expenses the corresponding depreciation charges. A study to identify these investments will be conduced by a foreign consultant and is expected to be completed by June 30, 1991. At the same time, in accordance with internationally accepted accounting practices, part of the revaluation surplus will be used to absorb exchange losses, thus helping in the financial restructuration of the enterprise. Overall, the results of the revaluation exercise are positive. In addition to the possibility of absorbing the exchange losses, it allowed REGIDESO to review the accounting classification of all its fixed assets and implement adequate depreciation rates for each category, improve its inventory and reconstitute a card-index box for about 7,400 items, implement an improved management eystem for fixed assets, create a unit for this purpose dnd train its staff to efficiently operate the new system. Performance Contract 3.11 A comprehensive study of REGIDESO vas completed by a foreign management consultant in April 1989. The suggisted Rehabilitation Program was reviewed by the government and approved by a-Zl the principal donors that contribute to the financing of REGIDESO (KfW a-. GTZ from Germany, Caisse Centrale-France, and the World Bank). This pr-rram was incorporated in a Performance Contract signed 1etween the government and REGIDESO in July 1989 (Annex 3.2). To ensure effective implementation, a Monitoring Committee consisting of representatives from the Ministries ox Finance and Energy and Mines and from the Service Charge des Entreprises Publiques (SCEP) was created. 3.12 In compliance with the Performance Contract, the government increased electricity tariffs in April 1990, restructured the Board cf Directors of REGIDESO, paid all public debts as of December 31, 1988 and authorized a capital iucrease as compensation of past debts. Some improvements have already also been noted, such as a new policy to increase connections (from 645 in 1988 to 1507 in 1989), the reduction to 15 days in the delays for billings, the placing of cash surpluses in revenue producing bank accounts, the reduction in personnel (230 in 1989), the implementation of a program to review and control electricity meters, the application of automatic cuts of electricity for  - 20 - delinquent accounts (2156 in three months), and the creation of Planning and Internal Auditing Units. To help REGIDESO implement the Performance Contract to its full extent the project will finance a Management Assistance Partnership Program with a foreign operator experienced in the management and rehabilita- tion of public utilities (para. 4.03). The operator will have responsibility for some key activities in the financial, commercial, planning and personnel management areas of REGIDESO and his remuneration will partially depend on the overall results obtained by the company. This will be complemented by short- term assistance for some specific needs, such as computerization and management information systems. The integrated and comprehensive arrangement is given a better chance of success than the ad hoc technical assistance previously financed by many donors, including the World Bank. However, the results will depend on the commitment and willingness of REGIDESO's management to act and on a close supervision of the operation by the Bank. B. The DGHER Organization and Management 3.13 A Department of Water Resources and Rural Energies within the Ministry of Rural Development was established in February 1979 with the responsibility of improving the supply of potable water in rural areas and implementing rural electrification. By Decree No. 100/209 dated December 22, 1988, this department became DGHER, with one department for energy, and another one for water. The Decree No. 100/13 dated March 11, 1986, limited REGIDESO's operating responsibility to 22 urban centers and brought all other centers under the responsibility of DGHER. However, this distribution of responsibi- lities is not being followed at the present time and REGIDESO is increasingly providing water supply and electricity services to some secondary and tertiary centers. This issue will be examined in the Rural Electrification Master Plan financed under the project. During negotiations, assurances were obtained from the government that: (a) a proposal for a new delineation of responsibilities in rural electrification between REGIDESO and DGHER would be submitted to the Bank before June 30, 1992 and (b) the government would implement the agreed recommendations before December 31, 1992. 3.14 DGHER is headed by a Director who is appointed by the President of the Republic upon nomination of the Minister of Rural Development and Artisanat (MDRA). DGHER has been the implementing agency for a Rural Water Supply Project financed by IDA (Cr. 1625-BU), which is an advanced state of imple- mentation. A new project in this sector is programmed for FY91. In general, administration and accounting for both water supply and rural electrification are weak. The proposed project will provide technical assistance and training to improve the performance of DGHER, assisting in organizing the commercial and financial aspect of rural electrification (para. 4.07). 3.15 DGHER's financial performance has to be assessed as that of a department of a Ministry. While DGHER enjoys some independence, its financial operations are embodied in the government budgetary system. With the exception of revenues associated with installing electricity meters and billing for electricity at the same rate as REGIDESO, DGHER does not have any financial resources of its own. These resources amounted to FBu 60.7 million in 1988, while the contribution from the government was FBu 87.4 million. DGHER will  - 21 - therefore continue to rely on the government appropriations and donors' contributions to sustain its operations and investments. The new rural electrification projects to be financed under the proposed project are not an additional financial burden on the operating budget of the DGHER as, in all cases, sufficient revenues will be generated to cover the costs of operations and maintenance. Agreement was obtained during negotiations that for all new rural electrification projects: (a) the internal economic rate of return will be at least 10 percent, (b) projected revenues will cover operating and maintenance expenses; (c) an adequate billing and revenue collection system has been designed and will be in place at the end of project execution; (d) consumers will be required to pay connection charges with a down payment before execution; and (e) tariffs for rural electrification would be the same as those established for REGIDESO. Staffing and Personnel 3.16 In July 1990, DGHER's staff amounted to 100 employees, of which 55 were employed in electricity operations. Approximately 20 percent of the staff is permanent and paid according to the salary structure of the public sector, which is significantly lower than that of parastatal agencies like REGIDESO. The remaining 80 percent are contractual personnel who may enjoy higher wages, but do not have the benefits of those permanently employed and may be dismissed at the end of their contract. To achieve operational and managerial efficiency in the new rural electrification projects and improve DGHER's operations, a project manager will be named within the existing administrative structure, and a finance/accounting expert will be recruited and financed under the project. This expert will work in close cooperation with the technical adviser already in place financed by the Belgian Cooperation. A description of the scope of work of the expert is provided in Annex 3.1. During negotiations, assurances were obtained from DGHER that the project manager would be appointed and that the expert would be recruited before Credit effectiveness. Accounting and Auditing 3.17 The law establishing DGHER specifies that its accounts are to be kept on an accrual basis using the double entry system and in accordance with the national accounting system plan. In practice accounts and financial reporting systems are primarily set up to determine compliance with the authorized limits of the budgets for recurrent expenditures and investment. Neither consolidated nor separated balance sheets or profit and loss accounts exist for its water supply and electricity operations. It is therefore difficult to abstract information and statistics on annual operating costs for these services. 3.18 Consumer billing is presently limited to DGHER's electricity operations in five centers, where the same tariff as REGIDESO is currently used. New consumers payments for the installation of a connection is a lump sum. To what extent meters are regularly read, bills distributed and the revenues collected is difficult to assess at present given the dispersion in the provision of services in the country's interior. Initially, in an effort to promote electricity consumption, DGHER installed connections without up-front payments and did not require regular payments of electricity bills. This resulted in arrears of about FBu 3 million as of December 31,1989, of  - 22 - which 80 percent relate to electricity sales and 20 percent to connection charges. Whilst the major part (70 percent) is due from private sector, 30 percent is due from government agencies. One of the major tasks of the finance/accounting expert would be to improve billing and revenue collection, as well as the accounting systems. 3.19 Systematic reporting regarding DGHER's operations is weak and somewhat irregular. Though a yearly report covering DGHER's operations has been prepared in the past, its scope and depth has varied considerably over time. The finance/accounting expert financed through the project is expected to establish a monthly and quarterly management reporting system and to ensure that balance sheets, income statements and fund flow statements are prepared shortly after the end of DGHER's financial year. Further, this expert will provide procedures as to how to allocate expenditures between water supply and provision of electricity. 3.20 During negotiations, assurances were obtained from DGHER that it would: (a) provide to the Bank quarterly reports covering the progress in the execution of the project and other operations according to a system of agreed monitoring indicators; (b) starting in FY91 produce separate accounts for its water and electricity operations, and (c) prepare for FY91, and every fiscal year thereafter, yearly balance sheets, income statements and fund flow statements in sufficient detail to allow for identification and assessment of its various activities. 3.21 Under the Rural Water Supply Project, DGHER agreed to have its accounts and financial statements audited by an acceptable external auditor and to submit them to the Bank within six months of the end of each fiscal year. DGHER has not been able to meet the six month deadline because of the weakness in its finance/accounting division, as described earlier. Audited accounts for FY87 and FY88 were received with delays up to one year. The audited financial statements for FY89 are not yet available. The reports are heavily qualified and the auditors highlight the existence of a defective communications system between DGHER and the municipalities that should pay for water supply, the incomplete recording of accounting transactions, the lack of accounting procedures and deficiencies in the follow-up and monitoring of billing, collection and transfer of revenues. Because of these deficiencies, the accounting books do not provide a completely reliable and accurate picture of the financial position of DGHER. The requirement to submit audited accounts within six months of the year-end is likely to be observed because of the improvements in accounting and in the preparation of financial statements resulting from the technical assistance to be provided by the project.  - 23 _ IV. THE PROJECT Project Obiectives 4.01 The broad objectives of the project are to promote rational energy policies and to strengthen the efficient management of energy resources. Specifically, the project aims to: (i) develop efficient institutions in the sector and improve the quality of public investment; (ii) improve the efficiency in the use of energy resources through reforms in the pricing structure of electricity, petroleum products and woodfuels; (iii) expand the access of the population to electricity; and (iv) reduce negative environ- mental effects of the use of energy through the execution of charcoal efficiency and improved stoves programs. Project Description 4.02 The main features of the project are summarized as follows: (a) Energl Sector Institution Building (i) implementation of a Rehabilitation Program for REGIDESO that includes a management assistance partnership program with a foreign operator, a restructuring of the financial base of the enterprise, the construction of limited new facilities to improve operations and training for its personnel; (ii) pre-feasibility and feasibility studies for selected hydroelectric power projects and an update of the power sector Master Plan; (iii) institutional strengthening of the DGE in the Ministry of Energy and Mines so that it can assume its leadership role in the areas of planning, coordination and control of the energy sector; (iv) institutional strengthening of the DGHER in the Ministry of Rural Development so that increased access to electricity in rural areas be achieved efficiently; (v) implemen- tation of pricing policies reflecting the economic cost of supply in electricity, petroleum and woodfuels; and (vi) coordination and monitoring of the public investment program in the energy sector; (b) Biomass and Household Energy (i) execution of a Charcoal Efficiency Program to disseminate improved techniques for the production of charcoal and sensitize charcoalers to the issue of deforestation; (ii) continuation of the Improved Charcoal Stoves Program which aims at reducing charcoal and wood consumption; (iii) a household energy consumption survey to improve knowledge on energy demand and help the planning process; and (iv) a feasibility study of peat substitution in secondary industries to determine the viability of increasing the use of these resources; (c) Rural Electrification (i) Rural Electrification Master Plan to evaluate potential projects in rural areas; and (ii) execution of some economically justified extensions of the distribution network in rural areas;  - 24 - (d) Power Subsector (i) doubling of the 110 kV transmission line from Bubanza to Bujumbura and extensions in associated substations; and (ii) implementation of an intensive program of new connections and extension of the distribution network in urban areas. 4.03 Management Assistance Partnership Program. This program has been defined after a comprehensive diagnostic study executed by a foreign management consulting firm and has been included in the Performance Contract between the government and REGIDESO signed in July 1989 (para. 3.11). The defined actions will be implemented by a foreign operator experienced in the management and rehabilitation of public utilities. The selection of the operator has been completed and approved by the Bank. The initiation of activities took place in November 1990, financed by a PPF advance. The operator will provide five experts for up to three years each and for a total of 12 person-years. These experts are: (i) an economist who will be responsible for evaluating and monitoring the investment programs and for organizing the recently created planning unit; (ii) an expert in commercial management in charge of imple- menting measures to increase the number of customers, to improve billing and collection and to reduce the account receivables; (iii) an expert in finance and accounting who will be in charge of implementing measures to improve the general accounting system, the internal controls and the management information system; (iv) an expert in human resources and personnel management who will be in charge of implementing measures to reduce personnel, increase productivity and rationalize policies in the areas of recruitment, job and performance evaluation, and payment incentives; (v) an expert in purchasing activities in charge of implementing an adequate procurement and stock management system. In addition, the operator will establish, implement and follow up a master plan for informatization. The total cost for these services is estimated at US$3.7 million. 4.04 To complement the above activities, GTZ-Germany will continue to provide foreign experts in the technical areas of electricity and water operations and IDA will finance experienced local accountants for up to three years, at a total cost of US$190,000. In addition, since at the present time REGIDESO's employees are dispersed in different rundown buildings, which produces loss of time and productivity and makes efficient informatization almost impossible, improvements in the physical facilities have been included in the Rehabilitation Program. A commitment from the government to contribute to the financing of the construction of new headquarters for REGIDESO has been given. IDA will finance the imported material and equipment and a small portion of the civil works, up to a total of US$1.0 million. Detailed studies will be conducted under the responsibility of the REGIDESO and will be submitted to the Bank before September 30, 1991. 4.05 Training for REGIDESO under the project is intended to complement what was financed under the Power Transmission and Distribution Project (Cr. 1593-BU). Those programs were developed to increase the skills of specialized workers in mechanics and electricity. Additional programs for managers, supervisors and other professional staff, including those in the areas of accounting, administration and marketing, will be introduced. A detailed training plan will be presented to IDA before September 30, 1991.  - 25 - The cost of this component is estimated at US$0.5 million. The Rehabilitation Program is completed by a financial restructuring of REGIDESO that is detailed in para. 5.09. 4.06 Institutional Strengthening of the DGE. This component will support the recently created Planning Unit and the Department of Research and Statistics, that will be in charge of household energy matters. The project will cover the services of a planning expert (economist) for intermittent missions over a period of two years, advanced training in project analysis and planning techniques and equipment. The estimated cost is US$0.5 million. As a result of this assistance, clear procedures and uniform investment criteria for the sector will be established. The DGE will assume the responsibility of coordinating and monitoring the public investment program (PIP) in the energy sector, and the technical assistance will help them in this endeavour. During negotiations, it was agreed that the DGE would prepare an annual report on the PIP in the energy sector and review it with the Bank. 4.07 Institutional Strengthening of the DGHER. The DGHER needs strengthening in accounting, collection and billing to be able to effectively manage rural electrification activities. At the same time, DGHER should have the capacity to undertake itself the economic analysis of potential new projects. The project will finance an expert in commercial and financial management, as well as some training in economic analysis. The cost is estimated at US$200,000. 4.08 Energy Pricing Policies. As indicated in para. 2.23, the government has already taken actions to implement an electricity rate structure based on long-run marginal costs of supply. In April 1990, medium voltage rates were adjusted to the estimated long-run marginal cost values and substantial increases were approved for low-voltage users. During negotiations, the government agreed to a further adjustment in low-voltage tariffs before June 1991 and in yearly discussions with the Bank on electricity and water rates. Petroleum prices were adjusted by about 30 percent in September 1990 to reflect increases of prices in international markets. A study to reform the pricing structure and define a new pricing policy was prepared by the Bank. The government has agreed to adopt a policy of automatic price adjustments before Credit effectiveness and to reform the petroleum pricing structure by September 30, 1991. The project will also finance studies of charcoal and wood pricing (para. 4.10) that would lead to recommendations about changes in their structure and levels. 4.09 Hydroelectric Power Plant Studies. As indicated in para. 2.16, the proposed project will finance pre-feasibility studies on four sites: Masango, Rushiha, Kabu 16 and Kabu 23. An update of the production and transmission Master Plan would then be carried out, so as to select the least-cost alternative for the future expansion of the electric system, and a feasibility study would be completed for two of the above-indicated projects. The cost of these activities has been estimated at about US$3.5 million. 4.10 Charcoal Efficiency Program. The objectives of the program are to reduce the amount of wood used in the production of charcoal by introducing energy efficiency improvements and to improve living conditions in rural areas. Actions to be undertaken during execution include: (i) a detailed study of the  - 26 - sector and of the relation among charcoalers, wood-owners, transporters, wholesalers and retailers; (ii) the organization of training and demonstration sessions of improved charcoal producing techniques; (iii) the creation of an association of professional charcoalers; (iv) a sensitization campaign on deforestation problems and solutions; and (v) studies and recommendations to improve the charcoal/wood pricing structure. The project will finance for a period of up to three years a local team consisting of a project coordinator and three charcoal trainers, as well as short-term assistance from national and expatriate specialists, vehicles, and equipment. The total cost of this program amounts to US$615,000. 4.11 Improved Charcoal Stoves Program. The objectives of this program are to reduce the consumption of charcoal in Bujumbura and other cities, thereby reducing energy expenditures of low income urban households, and to establish a self-sustaining basis for the production, distribution and commercialization of improved stoves. The stoves will be produced by independent artisans and NGOs or private companies will be involved in the sale and distribution, eventually financing the production activities and arranging for quality control. Activities undertaken under this component include: (i) technical training for stove producing artisans; (ii) establishment of an appropriate commercialization process; (iii) diversification of the stove models to be produced; (iv) implementation of publicity campaigns, and (v) surveys to evaluate the penetration rate, the savings achieved in the consumption of charcoal and the problems of commercialization. IDA's financing will cover the cost of local personnel, national and expatriate specialists, equipment and vehicles. The total cost of these activities over three years is estimated at US$315,000. 4.12 Household Energy Consumption Survey. The proposed survey will collect statistical data per geographical zone and type of household on the level of consumption of energy, the end-use per type of energy and the potentials for substitution. Training will be provided so that this task will be continued on a regular basis without external assistance. The total cost is about US$169,000. 4.13 Feasibility Study of Peat Substitution. The study will examine the technical, financial and economic feasibility of converting the boilers of some selected secondary industries to peat, procure and test one peat boiler system, and determine the viable market for peat in the industrial sector. The study would be executed by expatriate consultants under the responsibility of ONATOUR. The estimated cost amounts to US$263,500. 4.14 Rural Electrification Master Plan. The DGHER has a list of projects that it wishes to undertake in the future, but no economic analysis has been done and no priorities have been established. The Master Plan will evaluate all potential projects on the basis of technical, economic and financial criteria and give their ranking according to economic priorities. For each project, a comparison between extending the transmission lines from the national power grid and developing local sources of energy (diesel generator, micro-hydro sites) will be made. Institutional issues, such as the relations between REGIDESO and DGHER, responsibilities of municipalities and partici- pation of local organizations, will also be examined. Terms of reference for the Master Plan have been reviewed by IDA and execution will start in May 1991  - 27 - and will be financed by a PPF advance. The study will be executed by foreign consultants at a cost estimated to be US$300,000. Local personnel will assist in data gathering and analysis. 4.15 Rural Distribution Projects. During the preparation of the project, the economic feasibility of a certain number of rural electrification projects has been established (para. 6.10), and they will be financed under the proposed operation. These are mainly short extensions from existing lines or connec- tions to the network of some isolated centers already electrified. Other projects would be identified in the Master Plan. IDA's financing will cover the cost of all imported material plus the wood poles to be purchased locally, while the government will finance the construction, engineering and supervision of the lines and substations. These activities will be executed by the DGHER assisted by Belgian experts who are already in place. The total cost for this component amounts to US$3.2 million. 4.16 Doubling of the Bubanza-Bulumbura Transmission Line. Under the existing configuration of the transmission system, a single-circuit 110 kV line between Bubanza and Bujumbura carries all the power from both the Rwegura and Ruzizi II power plants. The proposed project would finance the execution of a similar 28 km 110 kV line in parallel to the existing one, plus the necessary additions to both the Bubanza and Bujumbura (RN1) substations. The design would be the same as that of the existing line and substations. The first five towers at Bujumbura's end are now double-circuit and they will be used to install the additional line. Therefore, the entrance into the city will not require additional rights of way and the impact on the population will be almost negligible. The total cost of the line and additions to the substations, including engineering and supervision, is estimated at US$3.4 million. 4.17 Connection Program. This program aims at increasing the number of urban households with access to electricity by extending the distribution network in areas which are already electrified. The proposed project would help finance imported material for about 5,000 new connections, 100 km of medium voltage lines and 50 km of low voltage lines. As in the Rural Electrification component, the design would be as economic as possible, with single phase aerial connections and wood poles. The experts financed by GTZ-Germany will provide assistance to REGIDESO for the preparation of bidding documents, construction and supervision. The estimated cost of the program is US$4.0 million. Project Costs 4.18 Project costs are detailed in Annex 4.1 and summarized in Table 4.1. Total cost, including physical and price contingencies and duties and taxes, is estimated at US$32.2 million, of which US$21.1 million (66 percent) are foreign exchange costs. The cost estimates of the energy sector institution building, biomass and household energy and of the rural electrification Master Plan have been prepared by IDA, while the costs of the other components have been estimated by REGIDESO or the DGHER, reviewed by IDA and found to be reasonable. Physical contingencies of 10 percent were added to each component to provide an allowance for unforeseen increases in the volume of work. Price contingencies assume local inflation of 8 percent in 1990/1991, 7 percent in  - 28 - 1992/1993, and 6 percent thereafter, and international inflation rates of 2.7 percent for 1990, 2.6 percent for 1991, 4.1 percent for 1992, 5.4 percent for 1993, 4.7 percent for 1994, and 3.6 percent for 1995. Table 4.1: ESTIMATE OF PROJECT COST FBu bilions USSmiliIons L F T L F T A. Energy Sector Institution Buildina RE