Mali - Power - Water Project

Document of The World Bank FOR OFFICIAL USE ONLY RepOrt NO. 13293 PERFORMANCE AUDIT REPORT REPUBLIC OF MALI POWER/WATER PROJECT (CREDIT 1282-MLI) JUNE 30, 1994 Operations Evaluation Department 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. CURRENCY EQUIVALENTS Currency Unit = Malian Franc until June 1984, thereafter FCFA 50 FCFA = 1 FF (till 1994) 100 FCFA = 1 FF (January 1994) 1984 US$1 = 9.5 FF = 475 FCFA 1994 US$1 = 5.6 FF = 562 FCFA US$/SDR US$/FCFA 1982 1.10311250 250 1983 1.04695408 408 1984 .98021479 479 1985 1.09842469 469 1986 1.22319365 365 1987 1.41866309 309 1988 1.34570285 285 1989 1.31410337 337 ABBREVIATIONS AND ACRONYMS ADB African Development Bank EIB European Investment Bank CCCE Caisse Centrale de Coop6ration Economique (the Caisse) CFD Caisse Frangaise de Developpement (formerly CCCE) CIDA Canadian International Development Agency DNHE Direction Nationale de I'Hydraulique et de llectricit6 EDF Electricit6 de France EDM Energie du Mali EIRR Economic Internal Rate of Return ESMAP Energy Sector Management Assistance Program FAC Fond d'Aide et de Coop6ration FCFA Franc de la Communaut6 Financire Africaine FED Fond Europ6en de D6veloppement FIRR Financial Internal Rate of Return GOM Government of Mali IDA International Development Association KfW Kreditanstalt fur Wiederaufbau LRMC Long Run Marginal Cost MIHE Ministre de l'Industrie de l'Hydraulique et de I'Electricit6 OERHN Office pour I'Exploitation des Ressources Hydraulique du Haut Niger (Selingu6 hydro power plant) OMVS Organisation pour la Mise en Valeur du Fleuve Senegal (Manantali dam with potential hydro power generation) OPEC Organisation of the Petroleum Exporting Countries PAR Project Audit Report PCR Project Completion Report PPF Project Preparation Facility SDR Special Drawing Rights TA Technical Assistance FISCAL YEAR Government: January 1 - December 31 FOR OFFICIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A Office of Director-General Operations Evaluation June 30, 1994 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Performance Audit Report on Mali Power/Water Project (Credit 1282-MLI) Attached is the "Performance Audit Report on Mali - Power/Water Project (Credit 1282- MLI)". The IDA Credit catalyzed contributions from five donors (BEI, CFD (CCCE), FAC, KfW and OPEC) to finance the relatively successful physical rehabilitation and expansion of the Bamako power and water distribution systems, plus technical assistance services aimed at improving the technical, financial and institutional performance of the sector. The technical assistance which was provided to Electricit6 du Mali (EDM), the state owned utility responsible for power and water services, failed to achieve its objectives. This component was designed to conform with IDA's vision of EDM as a commercially oriented operation; it was implemented by the Borrower to conform with its vision of EDM as a provider of power and water as a public service with social connotation. Based on the relatively good physical achievements and the poor results of the costly technical assistance, the audits rates the overall outcome of the project as marginally unsatisfactory, its sustainability as unlikely and its institutional impact as negligible. Some important lessons can be drawn from this lending operation. IDA needs to deepen its dialogue with its Borrower to achieve project ownership by the Borrower, particularly when it is the first IDA project in the sector. Also the Borrower and the co-financiers need more familiarity with the Bank guidelines and procedures to be followed throughout the lending cycle, particularly in the procurement of goods and hiring of consultants. The follow-up project, Power II (Credit 1998-MLI), has benefitted from these lessons. The sector's institutional reform and the power sub-sector's reorganization, which were financed by the Power/Water Credit, are currently being implemented. Robert Picciotto by H. Eberhard K6pp Attachment 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.  PERFORMANCE AUDIT REPORT FOR OFFICIAL USE ONLY REPUBLIC OF MALI POWER/ WATER PROJECT (CREDIT 1282-MLI) TABLE OF CONTENTS Page No. PREFACE ....... BASIC DATA SHEETS .. ......................................U EVALUATION SUMMARY ..................................... v 1. BACKGROUND .......................................... 1 The Context ........................................... 1 Project Objectives ....................................... 2 The PCR ............................................. 2 2. PROJECT PREPARATION AND APPRAISAL .................. 3 Project Preparation Phase I (1977 till 1980) ................... 4 Project Preparation Issues ................................. 4 Project Preparation Phase I (1981 till 1983) ................... 5 Studies Realized During Project Preparation ................... 7 Water Work Project ..................................... 7 3. IMPLEMENTATION AND PROJECT OUTCOME ............... 7 Project Costs ....7 EIRR and FIRR of the Project.............................. 8 Project Outcome ......................................... 9 Engineering Studies...................................... 9 Technical Assistance Component ............................ 10 Rehabilitation of the Bamako Distribution Network.............. 10 Mopti-Sevard Biomass Power Plant........................... 11 Bamako-Segou Transmission Line ............................ 12 Water Work Project................... ................. 12 4. BROAD ISSUES AND FINDINGS........ ................. 14 Country and Sector Framework............................. 14 Project Preparation................... .................. 14 Donors Coordination.................. ................ 15 Dialogue Between IDA and GOM........................... 15 Project Ownership.......................... 15 Loss Reduction and Financial Viability .............16 Project Supervision................... .................. 16 Performance of IDA ..................................... 16 Performance of the Borrower .............................. 16 This report was prepared by Alvaro Covarrubias, Task Manager, and Alain Gertsen Briand, (Consultant), who audited the Project in January 1994. Mrs. Ma.yvonne Mauprivez provided administrative assistance. 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. Table of Contents (Continued) Page No. 5. CONCLUSIONS AND LESSONS ...................................... 17 Project Overall Ratings ............................................. 17 Lessons Learned .................................................. 17 ANNEXES 1. Procurement of Consultant Services .................................. 19 2. Comments from the Borrower ...................................... 21 i PERFORMANCE AUDIT REPORT REPUBLIC OF MALI POWER/WATER PROJECT (Credit 1282-MLI) PREFACE 1. This report presents the results of the Performance Audit of the Credit 1282-MLI made by the IDA to the Republic of Mali to finance the Power/Water Project. The Credit of SDR 20.400.000, signed on October 14, 1982, benefitted the Government of Mali, which on-lent it to Electricit6 du Mali (EDM) as the executing agency of the project. The Credit became effective in May 1983, was closed in December 1988, and the disbursements were completed in June 1989. 2. The IDA's Power/Water Project was part of a larger government rehabilitation program, financed by the CCCE, FAC, KfW and OPEC. The project was appraised twice and negotiated three times, the second appraisal focussed mostly on a new TA's approach to alleviate EDM staff concern with the nature and scope of the assistance. The project included various engineering studies, some of them for the preparation of the Power II Project (Credit 1998-MLI). The rehabilitation reinforcement and expansion of the power distribution network of Bamako were activities urgently needed to absorb the energy coming from the Selingu6 hydro power plant (1980). The Water Work Project was focussed on the suppression of the worst bottlenecks in the Bamako water distribution system 3. The present Performance Audit Report (PAR) is based on the analysis of the PCR No. 10245, dated December 27, 1991, already distributed to the Executive Directors; the review of pertinent documents in the project file, the President's Report and the Staff Appraisal Report and valuable information provided by Bank staff. Part I of the PCR was prepared by the Region, and part II was produced for EDM by a consultant. In addition, the audit reviewed the project file and benefitted from discussions with IDA staff. In January 1994 the audit visited Mali, interviewed Government officers and EDM management and staff and obtained their view of the difficulties and constraints they encountered during the preparation and implementation of the project. The audit also obtained their suggestions to improve similar operations in the future. 4. The PAR analyzes aspects not fully discussed in the PCR as follows: (i) the reasons it took six years to prepare the project (1977-83); (ii) the lack of coordination among various donor activities; (iii) the causes of the poor results of the costly technical assistance component; (iv) the reasons for excluding the power sector reform from the project; (v) the correlation between the delay of the power sector reform and the failure to strengthen EDM; and (vi) the difficulties encountered in procuring consultant services for carrying out the studies. 5. Following OED's standard procedures, copies of the draft PAR were sent to the GOM and the beneficiary for comments. The co-financiers have also received a copy of the draft report for their comments. Comments from the Borrower are contained in the Annex I. No comments were received from the co-financiers.  ii PERFORMANCE AUDIT REPORT REPUBLIC OF MALI POWER/WATER PROJECT (Credit 1282-MLI) BASIC DATA SHEET KEY PROJECT DAA Aporaisal Actual IDA Credit (SDR) 20.400.000 20.393.964 Total Program Cost (foreign exch.) US$ (1982) 41.250.000 -- At appraisal time the EIRR of some projects were estimated as follows: Rehabilitation Power Distribution Network 21% Biomass Project' at Mopti 11% Bamako Water Supply System 11% No EIRR was calculated for the complete program in the SAR or the PCR. CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS (US$ million) EM EKH EM EYM EM EM FY89 TOTAL Appraisal Estimate 2.5 5.3 5.7 5.6 4.9 24.0 Actual .67 1.77 3.61 6.56 5.05 3.13 2.57 23.4 Date of Final Disbursement: June 1989 Due to the variation of the US$ value during project implementation the exact value of all contributions in US$ is not readily available. 2 The Biomass project was not found technically feasible and was cancelled and replaced by a diceol plant. PROJECTIDATES Original Actual Identification 06/77 06/77 Preparation (PPF) 02/79 02179 Appraisal (1) 06/80 06/80 Negotiations (1) 03/81 03/81 Appraisal (2) - 3 10/81 Negotiations (2) - 12/81 Negotiations (3) - 06/82 Board Approval - 07/82 Credit Signature - 10/82 Credit Effectiveness - 05/83 Credit Closing (1) - 12/87 Credit Closing (2) - 12/88 Loan final Disbursement 12/87 06/89 Project Completion - 06/89 lIM lM Kii Us U flM IM l M4 l EM l r ff IFU6Z "S F769 F79 I FY9 2 TOTAL Proappraleal 20.0 27.0 .2 47.2 Appraisal 13.4 42.3 9.0 64.7 Negotiations 14.4 .6 15.0 Supervision 20.0 29.5 28.7 41.0 13.4 10.2 4.3 10.1 1.2 156.4 Other 4.7 .4 4.1 15.6 .2 .9 .4 29.2 Total 24.7 40.8 46.5 42.1 20.6 29.5 26.7 41.9 13.4 10.2 4.6 10.1 1.2 312.6 After the setback of the firs negotiations, the original date had little meaning. Project preparation was difficult; it was appraised twice and negotiated three times. All together, frn project identification to effectiveness, it took six years. At SAR tie. iv SUPERVISION MISSIONS DATA Stage of Date of Staff-days Project Cycle Reg in field Status 1. Supervision 01/83 14 2 2. Supervision 05/83 02 2 3. Supervision 08/83 10 2 4. Supervision 01/84 16 2 5. Supervision 05/84 21 2 6. Supervision 07/84 18 2 7. Supervision 07/85 18 2 & Supervision 01/86 18 3 9. Supervision 07/86 56 3 10. Supervision 02/88 02 2 11. Supervision 06/89 12 2 Total 187 Staff-days or 37.4 Staff-weeks ' The PCR mentions 15 supervision missions; the audit could only find 11 supervision mission reports with form no. 590, plus a mention of 2 missions of a training specialist, without reports. For the Water Work, no supervision mission reports with form no. 590 were in the project file. The PCR mentions 5 supervision missions with back to office reports. OTHER PROJECTS Borrower: Republic of Mali Executing Agency: EDM - tlectricit6 du Mali Follow-up Project: Power I - Credit 1998-MLI Audit's findings in the project file (form no. 590).  V PERFORMANCE AUDIT REPORT REPUBLIC OF MALI POWER/WATER PROJECT (CREDIT 1282-MLI) EVALUATION SUMMARY Background 1. The audited project was the first IDA distribution system's bottlenecks; (vii) the operation in the Mali power sub-sector. The construction of three administrative and Power/Water Project (Credit 1282-MLI) was operational buildings in Bamako for the power approved on October 14, 1982, made effective in and water departments of EDM. May 1983 and was closed in December 1989, twelve months later than expected. The credit Project Objectives was fully disbursed. The credit signature and effectiveness' delay was due to a combination of 4. The overall objectives of the project were to the Government of Mali's difficulty in meeting extend and improve the supply of electricity and the credit conditions of one co-financier and water to the population of Bamako, Segou and depletion of the IDA fund. The Power project Mopti, and manage the sub-sector efficiently. was part of a co-financed rehabilitation and Specifically, the objectives were: (i) to strengthen expansion program of the Malian power sub- EDM's managerial and operational capability sector's infrastructures. through a substantial TA project; (ii) to rehabilitate and extend the power and water 2. The Borrower was the Government of Mali distribution system of Bamako; (iii) to build a (GOM), which on-lent it to Energie du Mali power generation plant in Mopti; and (iv) to (EDM), the national power utility with a thirty study and design a transmission line to supply year concession for the generation, and electricity to the Segou area. The sector's distribution of electricity to Bamako, the national institutional reform and EDM reorganization capital, and to some regional cities. were to be studied. 3. The Power/Water Project included: (i) the Project Preparation Experience extension, reinforcement and rehabilitation of the distribution network of Bamako; (ii) a 5. The project was appraised twice and substantial technical assistance (TA) project to negotiated three times, from identification to alleviate the managerial and operational project effectiveness the preparation process difficulties of the power utility; (iii) a set of lasted six years. A few studies were needed to engineering studies to justify and design the establish the two sectors' priorities and facilitate infrastructures to be built during the project appraisal. The issues of; (i) institutional Power/Water Project and to prepare Power reform, (ii) tariff increase versus the scope of the (Credit 1998-MLI) the follow-up IDA operation project (iii) the TA in advisory or hierarchical in the sub-sector; (iv) the construction of a function were all factors, which influenced the biomass power plant in Mopti in a rural area; (v) duration of project preparation. Finally, the the studies of the energy sector's institutional issue of institutional reform was postponed to reform and the reorganization of the power sub- the subsequent credit, and a compromise on the sector; (vi) the elimination of the Bamako water project's scope of work and tariff level was vi reached. The TA component exact nature and consultants, the commercial aspects of EDM did scope were difficult to establish. It was finally not improve and the energy losses of the system settled by mostly accepting the Malian point of stayed over 20% until the end of the project. view. 10. EDM had difficulty in meeting the financial Implementation Experience and commercial conditions of the covenant. The internal generation of 10% of the financial 6. The Power/Water Project largely achieved resources needed for future capital expenditure physical objectives (i.e., rehabilitation and was not complied with. The GOM and its extension of the power and water facilities), but agencies never succeeded in paying their failed to achieve the second part (i.e., the electricity within the three months period as had efficient management of the operations). The been agreed. Improvement of EDM's power and water sectors performance indicators commercial operations (billing and collection) for the project are considerably low&er than was not achieved during the project's life, in expected, especially in view of the TA resources spite of a costly consultancy contract to assist expended to strengthen EDM. The TA mainly EDM. The non-technical losses in the EDM in advisory function within a power utility network were about the same at the end of the without autonomy from Government control project as they were at the beginning. could not perform efficiently in the Malian context of the early 1980s. 11. Based on the above, the audit rates the project outcome as marginally unsatisfactory and Project Results its institutional development impact as negligible. The PCR rated the project outcome as 7. The cost of the studies combined with the unsatisfactory and the institutional development technical assistance project represent around impact as negligible. 51.2% of the total cost of the program. Some of the studies were intended to prepare the follow- Sustainability up project Power II i.e., the Power Expansion Master Plan, the Institutional Reform Study and 12. The sustainability of the benefits coming EDM Reorganization Study. The benefits of from the physical infrastructures built during the these studies will be reaped in the next project. Power/Water Project will mainly depend on EDM's capability to generate enough revenues 8. The EIRR and FIRR for the overall project to procure the spare parts needed to maintain its was not calculated at appraisal time, the audit power network and water system, and to obtain estimates the EIRR of the power project at the necessary expertise to repair, maintain and around 6.0% and the FIRR at -6.4% The operate its infrastructures adequately. The audit financial return of the water supply and the concludes that the sustainability of the benefits distribution system is negative. In spite of the derived from the infrastructures is unlikely as substantial and costly TA project the power sub- long as EDM is unable of generating revenues sector's managerial and operational performance and developing an appropriate skill base. The did not increase much. At best, the problems PCR sees this as a live possibility in the longer were alleviated. run and rated it as unclear. 9. The project succeeded in extending the Conclusion and Lessons Learned Bamako power distribution network, improving the service reliability and helping absorb the 13. The main conclusion of the audit is that the surplus power generated at the Selingu6 power lack of success of the audited project is to a large dam. In spite of efforts from the TA and extent the result of the following circumstances: vii (i) a difficult political and economic environment; ascertained carefully: the benefits of using large (ii) an institutional set-up that was not conducive firms with high overhead and little flexibility in to efficient operation of the power sub-sector; their approach to the TA should be compared to (iii) a general attitude among the governmental the alternative of firms to identify and select authorities that the power sub-sector and the specialists on an individual basis, especially when water sector are more social services than flexibility of recruitment and wider choice of commercial entities; and (iv) a belief by IDA that specialists are important criteria. the TA could somehow solve all the sub-sector problems and achieve institution building. This (v) Project preparation should not take six difference of view and objective between the years when time is of the essence. If the IDA and GOM resulted in a compromise which Borrower is not cognizant with ways and means weakened project ownership by the country. of preparing a project to comply with all the IDA's norms and standards, supplying external 14. The main lessons of the project are: assistance to the Borrower to speed up the process might be cost effective. (i) A careful assessment needs to be made during appraisal of the minimal institutional (vi) IDA could find it useful to assure dynamic reform and power sub-sector reorganization follow-up on the supervision missions' findings by required to assure reasonable chance for a securing monitoring services for the most crucial capital investment and a TA project to be parts of the project; implemented efficiently and derive sustainable benefits; (vii) Suspension of disbursements and even cancellation of a project should not be a (ii) Review of sector and sub-sector problems potentially distant threat but a real option to be and constraints by experienced specialists is perceived as such by the Borrower when needed during project preparation to identify the systematic violation of major covenants becomes problems and their solutions before they a recurrent situation; and seriously impact on project implementation; (viii) Bank guidelines on procurement of (iii) Donor meetings during preparation and professional services must be respected, even implementation are necessary for the efficient when unfair competition is inconvenient to the realization of a co-financed program and by Borrower, consultants and/or contractors. doing so, to support the sense of project Disregard of the IDA guidelines, specially when ownership by the Borrower and the executing unfair competition is involved, creates precedents agency; inciting discord between consultants and co- financiers, and the studies' results are generally (iv) The cost effectiveness of various lacking credibility. approaches to supply the TA need to be  PERFORMANCE AUDIT REPORT REPUBLIC OF MALI POWER/WATER PROJECT (Credit 1282-MLI) 1. BACKGROUND The Context 1. Mali is a Sahelian country of 1.2 million km' with a population estimated at seven million, growing at an annual rate of approximately 3%. It is a poor country with a GNP per capita of about US $ 230 (1988). The commercial balance suffers from chronic deficit, mitigated by massive injection of foreign assistance. The growth of net income per capita over the last twenty years was approximately 1 % per year. 2. Mali is totally dependent on imported petroleum products for transportation and power generation for cities not connected to the main network. The fuel-wood represents 90 % of the private consumption of domestic energy. The small hydro electric potential of Mali is estimated at 5,000 GWh/year, with a firm power of about 800 MW. Three hydro electric sites have been developed with a total peak capacity of 50 MW and around 250 GWh/year. For comparison purpose in 1991 the peak demand of the Bamako network was 39 MW and the generation was 207 GWh. The peak power demand happens during the hot and dry season, when the power capacity of the hydro plants is lowest. Hydro power and energy generation is supplemented by the diesel generating plant at Bamako (Dar es Salam). 3. In spite of the Power I investment in the sub-sector, less population in percentage had access to the electricity at the end of the Power I project than at its beginning. In 1982, about 5% of the Mali population had access to the electricity i.e., 28.000 consumers of electricity were connected to the power grid or supplied by thermal generating units in ten smaller cities. In 1989, at the end of the Power/Water and Water Project (hereafter referred to as Power I) about 4% of the population had access to electricity, or about 37.000 customers connected to the EDM system, resulting in a relative decrease of electricity coverage on account of the population increase. 4. At the time of Power I project's preparation, the sub-sector organization was complex, not oriented towards efficiency nor towards accountability of the participants. In early 1980, the DNHE and the OERHN (the Selingu6 power dam) reported to one Ministry responsible for the sub-sector planning and the construction of the infrastructures; EDM reported to another Ministry and managed the sub- sector operations. One of the IDA's original goals was to improve the power sub-sector organization. Since this particular goal was judged difficult to achieve in the short run, it was decided to study the re- organization options as part of the Power/Water Project. 5. EDM, the national power utility, was created in 1961 as a joint venture of the government (55% of the shares) and foreign investors (CCCS 39% and EDF 6%) to manage the operations of the power 2 sub-sector under a 30 year "concession" system'. The GOM approved the tariffs and appointed the EDM personnel. This concession system had some advantages (access to foreign expertise, technical support etc..) and a few shortcomings for the efficient management of the operations (tariff setting, personnel selection etc...). In the late 1970s the IDA suggested the analysis and review of the advantages and disadvantages of the system and discussed the situation with the GOM during the course of first project but nothing changed. The financial performance of EDM from 1977 until 1981, based on unaudited accounts, shows a deficit in operating income beginning in the mid 1970s: the water distribution sector was responsible for the poor results; the power sub-sector showed a small positive operating income with the exception of 1979. 6. In mid 1977, the GOM requested the IDA's assistance to strengthen sub-sector operations and help finance the expansion of the Bamako distribution network. The Selingud hydro power dam was in construction and scheduled to deliver energy to Bamako towards the end of 1980. This energy had to be distributed as soon as possible to assure the Selingud project's financial viability. The Bank was not a co- financier of the Selingu6 project and had not originally supported it. The rationale behind the GOM request was EDM's disarray due to lack of qualified managerial staff, weak leadership and serious financial problems, exacerbated by the oil price increases of the mid 1970s. The lack of financial resources forced EDM to ignore preventive maintenance, and repair was difficult due to scarcity of spare parts and tools. EDM was assisted and advised at the technical level by a few EDF seconded technicians; this TA in advisory function was deemed to be ineffective. Project Objectives 7. The original objectives were: (i) ensuring improvements and increased water and power supply to the Bamako area, this was realized to a large extent; (ii) extending the service area to be supplied by the hydro-electricity coming from Selingu6, this was also realized to a large extent; (iii) supply power to the rural area around Segou, this was not realized during Power I due to delay in the studies and in the project's design. The project was initiated at the end of Power I and completed in the beginning of Power II; however, the rural distribution part has not yet been implemented; and (iv) power generation in Mopti using agricultural residues, this component was not realized due to the technical non-feasibility of the engineering concept proposed by the consultant. However, generation in Mopti was achieved with some delay using a diesel plant. The PCR 8. The PCR does not indicate how critical the situation of the Malian power sub-sector in the late 1970s was and how urgent strong, remedial actions were needed' to redress EDM's overall situation: (i) EDM was only responsible for the management of the operations and the general administration of part of the sub-sector; (ii) until 1983, when Power I became effective, a ministry was responsible for policy and strategy formulation, and the planning and construction of new investments; (iii) the managerial capability of EDM was very weak. In 1980, a Caisse Mission Report mentions: "EDM has 20 MW installed in ' A type of leamchold contract. 7 SAR, Power I, Report No. 3227-MLI, page 35. * PCR, page 1, paragraph 2. 3 Bamako, however, only 4 to 6 MW are available at any one time. The demand is estimated at 18 MW. EDM cannot supply financial results credible enough to obtain an opinion on its situation. EDM does not know how many customers it has, billing is not done regularly. The personnel are not motivated. The lack of financial resources preclude preventive maintenance and repair; over the years EDM has benefitted from a substantial technical assistance, in advisory function only. This assistance has been wholly inefficient." The PCR, part I and II, does not cover the project's preparation problems and their causes. 9. The audit notices that in 1980 EDM would not have had a power capacity supply problem, provided its operations had been well managed. The installed generating capacity of 20 MW at Bamako could have met the estimated 18MW peak load provided no suppressed load existed. This indicates that the hydro power plant of 44 MW at Selingu6, implemented without Bank participation, and coming on line in 1980, was perhaps built a bit prematurely, however it was a protection against more oil price increases. In early 1981, EDM had a surplus of energy coming from the Selingud hydro plant, the problem was how to connect enough new customers to absorb the energy. Expanding the Bamako distribution network to supply the economically justifiable demand of Bamako was therefore a priority. The alternative was to spill water or to find customers in the neighboring countries. The above situation highlights the sub-sector's priorities: (i) Reorganize the power sub-sector's institutional set-up to make it efficient by endowing EDM with full autonomy and accountability; (ii) Reorganize EDM and inject managerial expertise and experience to improve management of the power and water sectors (billing and collection) through a cost effective TA program; (iii) Rehabilitate, reinforce and expand the distribution system of Bamako to serve all customers willing and able to pay the economic cost for the electricity and water services. 2. PROJECT PREPARATION AND APPRAISAL 10. The audit concludes that project preparation and appraisal was unsatisfactory, and the direct cause of the poor results of the project In 1977 the situation of the power sub-sector in Mali was so dismal that the GOM contemplated ownership of EDM to a foreign firm, and idea that was not followed up by the GOM and/or supported by the donors. The timing of the IDA's first power project was a vital for the power sub-sector's rehabilitation, and for the economic utilization of the Selingu6 hydroelectric plant at the end of 1980. The project was identified in mid-1977, a PPF of US$0.97 million approved in October 1979, and the project appraisal made in June 1980. But the project was then substantially restructured during a second appraisal to factor in the EDM staffs view on the nature and scope of the TA component, and the different perception of EDM's problems by a new IDA project team. The second preparation phase went on from 1980 until 1983, when finally the Power and Water Project became effective i.e., three years after the power from Selingud was available in Bamako. The- long preparation time of the Power I Project had a negative impact on the Selingu6 project's viability and the sub-sector performance. 4 Project Preparation Phase I 11. Objective and Goals of the Project. At the time of project preparation, phase I, the GOM had no stated objective, goals and no strategy for the power sub-sector. The donors financing the Selingud hydro power plant had insisted on having its operations clearly separated from EDM, on account of its reputation for poor management. IDA reviewed the EDM situation, the various studies realized by EDF and discussed with the GOM the strategy and goals of the first IDA water and power project. The IDA's short term strategy was to avail of hydro power from Selingu6, to rehabilitate and expand the Bamako power distribution network, to supply electricity service to extend electric service to a large segment of the population, particularly the low income households of Bamako and the surrounding area. The goal for the water distribution system component was to increase the customer connections from 18% in 1980 to 24% in 1983, and population covered by the water standpipes from 21% in 1980 to 76% in 1983. The reinforcement of the existing pumping and filtration station was financed by KfW through a credit of US$4 million. IDA did not participate in this water component. Project Preparation Issues 12. The principal issues during project preparation phase I were: Scope of Work and Availability of Concessionary Funds. During preparation phase I IDA tried to raise enough money to realize all the activities deemed necessary to redress the power and water sectors. The various studies financed through the PPF were quickly completed, and the meetings between the specialists of the IDA, Caisse, FAC and EDF appear to have been productive and the exchange of information beneficial to all parties. In order to assist the GOM in finding financing for the estimated US$75 million program, IDA contacted several potential donors such as, OPEC, CIDA, EIB, KfW to ascertain their interest in the project. However, some of the potential donors were not firmly committed to a specific disbursement timetable. IDA mentioned to the GOM that the original project would have to be divided into two phases to match the availability of funds. 13. Tariff Increase Necessary to Support EDM. During appraisal the IDA suggested a 30% increase for the water and electricity, this was implemented in March 1980. Later, in August 1980 a KfW financed study9 suggested a 200% increase for the water tariff to cover costs of maintenance and operations. During phase I negotiations it was understood that another 30% increase would be necessary before project effectiveness. In October 1981 the GOM increased the electricity tariff by another 11%, demonstrating political will, in spite of strong opposition. Missing during project appraisal (phase I and II) was the assessment of the nature of the network losses (technical or non-technical) and their potential for reduction to limit the tariff increases. The SAR (page 47) indicates that the total losses were 10%. Taking into account the dismal state of the EDM network in the 1970s, this low amount of losses might have raised suspicion. If the above-mentioned figure was accurate, the only remedy to improve the EDM revenue would have been a tariff increase. However, the SAR for Power II (page 30) indicates that in 1978 the total losses were 15.6%, and thereafter until the early 1990s they were assessed to be around 22%, even after a complete network rehabilitation. 14. Institutional Reform and Scope of Technical Assistance. At the request of EDM management, a comprehensive TA program, which suggested strong measures to completely redress EDM within three ' Not part of IDA's project. 5 yearso, was prepared by EDF with the support of the IDA. A massive dose of expertise in line of command at all levels of the utility was necessary to achieve the above goal. The previous TA delivered through advices had failed to produce tangible results, energetic measures were needed to save EDM. In October 1980, EDM's reaction to the TA proposal came in the form of a counter proposal, which reduced the nature and scope of the TA project The IDA suggested it might accept some modifications to the TA component, provided they did not imperil the strengthening of EDM. Around that time the IDA's project team changed, and EDM's Managing Director became the Managing Director of the Selingu6 hydro plant. It was only in March 1981, that EDM had a new Managing Director, acceptable to both the IDA and the Caisse. At the end of 1981, the project's preparation, appraisal and negotiation process had come to a stand still on the issue of the TA's acceptance by the EDM staff union, thereby raising the question of project ownership. The two reasons given by EDM for its refusal of the TA project were: (i) the scope of the project was too comprehensive (twenty-seven specialists) and would give effective control of the utility to the expatriates; (ii) the TA component was unsustainable for cost reasons. Project Preparation Phase H (1981 till 1983) 15. The audit concludes that the compromises made during project preparation, phase H, were the principal cause for the unsatisfactory results of the TA component and the poor situation in which EDM found itself at the end of the project. Three issues were discussed during project preparation phase II; the institutional reform was not one of them. 16. The Technical Assistance. The original proposal for the TA component was designed to redress EDM completely and quickly. The GOM's refusal to commit itself to the TA component proposed by the donors resulted in a pause in the negotiations. The IDA reached a compromise with the EDM staff on the TA question: (i) the number of specialists were reduced from twenty seven to thirteen; (ii) the project's coordinator would not be coming from EDF, but would be financed by the IDA and selected by EDM, this individual would be an adviser to the EDM's Managing Director; (iii) three other specialists would be financed by the IDA and the rest by the FAC and the Caisse; (iv) expatriates would work in hierarchical function only in the departments to be created and on a temporary basis. The existing directors of each department would obtain an adviser to assist them. Surprisingly, the budget was now US$13.59 million for thirteen TA specialists, instead of US$8.82 for the original twenty-seven. 17. Reduction of the Scope of Work. The sub-sector's priorities were reassessed to match the concessionary funds availability, and the most urgent activities were identified and the others postponed to a subsequent project. The Power and Water Project was substantially re-designed at the time when a separation of power from water projects occured in the Bank organization: (i) the rehabilitation of the thermal power plant of Bamako was included (US$3.53 million); (ii) the water work project, so urgently needed and politically important for the GOM, was reduced from US$19 million to US$1.84 million; (iii) the power distribution network investment for Bamako was reduced from US$7.83 to US$5.62; (iv) the Bamako-Segou transmission line was withdrawn from Power I, for a saving of US$28.75 million and replaced by a study for US$.82 million; (v) a budget of US$1.56 was added for engineering supervision. Including 20% provision for contingencies, the new project cost was US$43.2 million instead of the original US$75 million. The co-financiers were the IDA, Caisse, FAC, EIB and OPEC. 1o IDA suggested a "shock therapy". 6 18. Tariff increase and loss reduction. In October 1981 IDA informed GOM that the project presentation to the Board was conditional to an increase in tariff of 30% for electricity and 60% for the water, effective January 1, 1982. In February 1982 the GOM decreed an 11% increase in tariffs for electricity and water. In April 1982 the GOM informed the IDA that a 30% increase was not politically feasible and it was unable to comply with the covenant. The GOM suggested another reduction of the investments rather than increase the tariff. During a re-negotiation session in Washington the GOM supplied a new un-audited EDM account for 1981 showing much better results than expected and suggested to transfer some of EDM's debt to the government. The IDA accepted a plan for 8% increase for the electricity tariffs by July 1982 and 10% increase for the water by January 1983. Overlooked during project preparation was the benefit that a reduction of losses could have had on EDM's financial results. No operational performance indicators were established or discussed". At the end of Power I (1989) the value of the power system total losses were about 20% of generation, worth about US$9 million per year. 19. Change in Objectives. The potential impact of the compromise reached by IDA Region in relation to the TA component and its original goal to redress EDM completely within three years, was assessed by the IDA energy department: "(i) EDM is in an extremely poor shape, it lost money from 1974 until appraisal time; (ii) without suitable counterpart staff and adequate TA, the Region would not proceed with the project; (iii) the project, as proposed, will not significantly improve the operations or the financial situation of EDM; (iv) the substantial reduction of the water component does not always preclude the need for a manager of water supply, it should be financed by the project and recruited on a priority basis; the above should be made clear in the SAR." The SAR" duly included a disclaimer: "The massive TA component can go a long way to arrest further deterioration of EDM operations. Nevertheless, it must be recognized that there is still a risk of failure and further assistance would be necessary." In 1982, after two appraisals, the new TA goal was far from the original goal of 1980 i.e., to redress EDM completely within three years. 20. Problems during Project Preparation. The issues of the sub-sector's institutional reform, EDM's autonomy and its internal reorganization were not addressed during project preparation phase II; the decision was postponed to a subsequent project. In the meantime it was decided to study the problem, and the mandate was given to a consultant. The consultant was a shareholder, and at the same time the consulting engineer doing studies, and the supplier of the TA who advised EDM on the quality of the studies. During project preparation phase II, the dialogue and coordination between the donors deteriorated. The TA compromise reached during the second appraisal was not easily accepted by the Caisse and EDF (both shareholders of EDM). At one point the Caisse even had doubts about participating in the project's financing. The IDA suspected that EDF might be reluctant to accept the new TA contract (June 25 1981), since, in essence, it was a repeat of the previous TA experiment which had failed. After a high level meeting between the IDA and the Caisse management, it was agreed that the IDA would finance the services of three TA experts and the project's coordinator. However, for reasons of team cohesiveness, EDF would recruit all the TA experts with the exception of the project's coordinator. The Caisse would transfer some of the funds originally allocated to the TA to the rehabilitation of the Bamako thermal generating plant. No indication in the project file. 12 SAR, page 30. 7 Studies Realized During Project Preparation - (PPF) US$970,000 21. The audit concludes that the seven studies realized to prepare the project's components were overall unsatisfactory. 22. Problems and Issues. The IDA and EDM opted for giving all the studies to two consultants, one for the water sector studies, and the other for the power sub-sector. Five of the seven studies developed problems of various magnitude. The project file is far from explicit on the causes of the problems; however the audit identified some of them: (i) the ToR were not clear in relation to the methodology and criteria. One economic study had to be done twice (Bamako-Segou); (ii) the studies suggested scope of work generally in excess of the fund's availability and as a result supplementary studies were needed to reduce each component to an acceptable level (Water Work and Electricity Distribution Network); (iii) one of the consultants (a power utility) had a tendency to follow methods and norms specific to their own network and prepared budgets accordingly. Substantial reviews of the studies were needed at implementation time by an engineering firm to prepare new budgets and take into account the IDA's suggestion for a new construction approach (small lots); (iv) the power utility, which was a shareholder of EDM and already supplying some TA, was asked to formulate the TA package needed for EDM. Water Work Project 23. The preparation of the water project is not well documented in the project file. During project preparation phase I, the project for the water system of Bamako was considered a priority and a budget of US$21 million was allocated to mitigate the worst bottlenecks of the system. During phase II, preparation of the water project's budget was reduced to US$2 million. The IDA had a few years before financed some work (Credit 943) for the Bamako water system. In fact during the five years preceding the Power/Water Project, approximately US$12 million had been invested in the water system. In parallel to the IDA credit, the KfW was financing a Water Tariff Study and some rehabilitation work for the Bamako water treatment plant for approximately US$4 million, as well as supplying spare parts. To supplement the rehabilitation of the water system two TA experts were allocated to assist the EDM water department, one would advise the Director and the other would supervise the rehabilitation work. 3. IMPLEMENTATION AND PROJECT OUTCOME Project Costs 24. The PCR indicates the IDA budget in SDR for the project has no cost overruns. The audit confirms this statement (see Table No. 3-1). However, the audit could not find the information needed to establish the final cost for each item in US$ because during project implementation the value of the US$ varied considerably in relation to the European currency and the SDR. 8 TABLE 3-1 IDA CONTRIBUTION TO THE PROJECT (IN SDR) Budget Cotingert Actual Variation in % 1. Civil Work 1,420,000. 1,586,992. 1,671,000. 5.3 % 2. Banako Distribution 5,170,000. 5,778,816. 6,218,000. 7.6 % 3. Water Work 1,530,000. 1,709,928. 1,869,000. 9.3 % 4. Dar Salam 370,000. 413,512. 657,000. 59.0 % 5. Technical Assistance 6,430,000. 7,186,168. 5,924,000. - 17.6 % 6. Vehicles and Office equipment 770,000. 860,552. 1,235,000. 43.5 % 7. Eng. Studies 1,820,000. 2,034,032. 2,1 00. 7.8 % 8. Sub-total 17,510,000. 19,569,176. 19,766,000. 1.0 % 9. PPF 830,000. 830,000. 629,000. -24.2 % 10. Contingencies(11.76%) 2,060,000. 0 0 Total 20,400,000. 20,400,000 20,394000 IDA financed four TA specialists for SDR 5,924,000 i.e., SDR 370,000 man/year. EIRR and FIRR of the Project 25. The SAR13 of Power I estimated the EIRR for three separate projects: (i) for the Bamako distribution network the EIRR was estimated at 21%; (ii) for the Biomass Power plant it was 11%; and (iii) for the Water Work it was 11%. The Biomass plant was canceled and replaced by a standard generating plant (diesel). The benefits from the sale of the electricity in the Bamako distribution network were based on the marginal cost of the electricity from the extension of Sotuba hydro plant. This particular project was postponed sine die, therefore the SAR estimates for the various EIRR are not relevant anymore. 26. The audit has recalculated the EIRR and the FIRR of the complete Power I project including only the TA component directly noted to the project, on the basis of incremental costs versus incremental benefits: (i) the incremental revenues and the value of the non-technical losses are taken as proxy for the complete economic benefits of the project; (ii) the useful life of the project including the implementation period is twenty years. The costs of the 0 & M are the real costs of EDM taken from the Annual Reports (1986 to 1991) and from the SAR (1983 to 1985); (iii) the investment costs are taken from the PCR and the data collected during the audit's field mission (January 1994); (iv) the Water Work project costs and benefits are not included in the calculations and no separate EIRR and FIRR have been calculated for this particular project Performance indicators for the Water Project are given in Table No. 3-3 and they speak for themselves. 27. The recalculated EIRR of the power project is approximately 6.0%, based on the above-mentioned assumptions and the available data. The FIRR is around -6.4%. The above EIRR and FIRR are an SAR 3227-MLI, June 10, 1982. 9 approximation because the financial accounts of EDM were not certified by an independent auditor. Also, the costs paid by EDM for the electricity coming from Selingu6 were perhaps not the real economic cost. The FIRR results are not due to low tariffs since they are substantially higher than the LRMC and appear to have reached the limit of the consumers' willingness to pay. Project Outcome 28. The audit concludes the physical implementation of Power I was overall satisfactory. In spite of procurement difficulties (see Annex), most of the project's components were implemented within the budget without major shortcomings, albeit with some delay. The physical components of the distribution and the water work system are of good quality, the buildings of reasonable quality, and all the infrastructures are still operational (1994). TABLE 3-2 EDM PERFORMANCE DATA - BAMAKO NETWORK 1982 1983 1994 1985 1986 1987 1988 1989 1990 1991 Generation GWh 103.7 121 131.1 147.8 159.2 172.3 181.2 185.6 199.8 206.9 Growth rate % 27 16 8 13 7.7 8 5 2.4 7.6 3.5 Billed GWh 80.1 93.4 100.2 116.4 124 134.9 140.5 145 157.5 164.8 Losses GWh 23.6 27.6 30.9 31.4 35.4 37.4 40.6 42.3 42.1 42.1 Losses % 23 23 23 21.4 22 22 22.5 22 21 20 Customers x 1000 18.5 19.8 21.7 23.5 24.6 26.6 27.5 27.9 29.5 32 Growth rate % 8 7 9.6 8.3 4.7 8.1 3.4 1.4 5.7 8.4 Tariff FCFA/kWh " 45.4 60.2 60.8 60.8 60 59.6 59.2 Demand MW 19.3 23 24.4 28.1 30.8 33.8 37.3 36.7 38.2 3&3 Growth rate % 19.9 19.17 6 15 9.6 9.7 10.3 -1.6 4 3 Collection Arrears - months 6.3 5.3 9.3 7.9 6.5 9.1 Operating Income, million FCFA: - Electricity sub-sector 776 1,734 1,223 652 163 352 - Water Sector (68) (177) (61) (176) (332) (539) - Combined electricity/water 707 1,557 1,162 476 (168) (187) Engineering Studies 29. The audit concludes that eight of the twelve studies were of poor quality. Some resulted in poor technical choice (Biomass) and the alternative design not being adequate (first design of the Mopti power plant); others had to be done twice (Bamako-Segou line) and a few studies generated controversy among the donors. The four studies without problems were all realized by engineering firms. Since it was the first IDA power project in Mali, great care should have been taken at the project's onset to explain to the GOM, EDM the rules for the selection of consultants. In 1993 (before devaluation, the LRMC of the electricity at the distribution voltage level for Barnako was approxinately 49.2 PCPA/kWh. 10 Technical Assistance Component 30. The TA component was well implemented. The TA did alleviate EDM's managerial constraints, and operations improved modestly during the first project However, the goal intended during project preparations phase I (i.e., completely and efficiently redress EDM within three years") was not met. 31. Spending about US$13 million to alleviate the EDM operations and management's constraints raises the question of TA cost effectiveness. The TA had a negligible impact on institutional development mainly because of its poor design. 32. A consensus exists inside EDM (January 1994) that the TA did not, in spite of its considerable cost, succeed in making EDM a well performing power utility. Different opinions exist about the causes for the lack of TA's impact which seems to be rooted in two issues: (i) the autonomy of EDM from Government interference in its day-to-day operations should have been resolved during project preparation; (ii) the merit of hierarchical versus advisory function for the TA- it should have been decided in favor of the hierarchical function, with clear mandate to train counterparts over a fixed period of time and performance review on a regular basis. Also, important for the utility's management is the opportunity to have access to qualified TA experts with talent in technology transfer, which implies that the choice of the experts should not have been limited to only one firm. The high cost of the TA and the lack of motivation of some specialists were the main complaints raised at another managerial level,. Rehabilitation of the Bamako Distribution Network 33. The audit rates the implementation of this component as satisfactory. Some problems developed during implementation due to the poor project preparation and wide variation in the value of the US$ compared to the FCFA and the European currencies: (i) the rehabilitation project's budget was not based on a precise scope of work and detailed engineering studies. In the early 1980s, in parallel with the IDA project preparation (PPF), the CCCE financed a rehabilitation project; this created some confusion between the two projects; (ii) in 1983, the services of a consulting engineer were retained by EDM to prepare the bid documents to satisfy the IDA guidelines, to analyze the construction bids and to supervise the implementation of the project Under EDM and the IDA's instructions the engineering consultant separated the project into 8 lots for the supply and construction. The consultant's estimate for the project was US$16.1 million based on costs valid in the early 1980s. The IDA asked for a reduction in the scope of work by 50% to match the budget of US$6.6 million; (iii) altogether the bid documents were revised twice and the bid evaluations were also done twice; (iv) the contracts' costs were in total US$3.5 million instead of the US$5.7 million estimated by the consultant. (All bids were made in European currency, at a time when the US$ was very strong). 34. In 1985, to take advantage of the "availability" of funds, the IDA approved an amendment to increase the existing contracts by US$823,914 i.e., an increase of 32% of the original bids to realize work which was considered urgent Since funds were still available, the consultant and EDM suggested to increase the scope of work of the distribution with installation of more connections and meters for FCFA 80 million (approximately US$260,000). Seven Bamako districts were equipped with connections and meters. Later, the IDA accepted to finance the purchase of equipment for FCFA 430 million (US$1.4 u Fcbnuay 19, 1981. 11 million) to allow EDM and local enterprises to extend the LV network. In 1987 EDM suggested to rehabilitate a sub-station and add 45 MT/LV stations and some MT lines for FCFA 250 million (approximately US$810,000). EDM was pleased with the IDA's flexibility in allowing continuation of the rehabilitation project through contract extensions based on unit costs. 35. The final cost for the rehabilitation and extension project was: (i) Bid time US$ 2,610,767 (ii) First Amendment 823,914 based on unit cost (iii) Second Amendment 260,000 (iv) Third Amendment 1,400,000 equipment purchasing (v) Fourth Amendment 810,000 on the basis of unit costs Total 5,904,681 36. In January 1994 an OED consultant reviewed the distribution network and found it of excellent quality, probably over designed. The technical losses of this network, if well maintained, should be quite low. Mopti-Sevard Biomass Power Plant 37. The implementation of this component was unsatisfactory. In November 1979, through a PPF financing, a foreign power utility acting as an engineering consultant investigated the possibility of utilizing the rice husk available in the Mopti agricultural area to generate power. The concept was a gas generating slow combustion process, to be enriched with small amounts of fuel-oil to be used by internal combustion engines. The study indicated the process could be economically advantageous. The budget for the foreign cost was US$6.45 million. During project preparation, Phase I, the IDA contacted the OPEC Fund in Vienna to ascertain their willingness to finance the biomass power plant. In July 1981 OPEC confirmed that US$6.46 million would be available to finance the Mopti biomass plant. It was agreed the IDA would administer the loan. 38. In May 1983 the Biomass power plant was declared technically non-feasible and the project was changed to a diesel plant. At the bids opening in March 1984, the cost of the lowest bidder was about twice the budget. In August 1984 EDM decided to cancel the bids and retained the services of a engineering consultant to redesign the plant taking into account the local requirements and prepare new bid documents. Finally, the Mopti power plant was built within the limit of the budget, US$3.6 million, but with substantial delay. 39. The lessons for this component are: (i) the experimental concept of the biomass power plant was too risky for Mali, considering the maintenance capabilities of the Malian counterparts, furthermore its technical feasibility was doubtful; (ii) EDM and the IDA did the right thing in asking a specialized consulting engineer to design a plant adapted to the local conditions and proceed with call for tenders after a detailed engineering study. 12 Bamako-Segou Transmission Line 40. This project was not intended to be physically implemented during Power I. However, the feasibility studies and the detailed design were realized during the Power I. The audit concludes the realization of the studies and detailed design were unsatisfactory. 41. In order to utilize the surplus of energy coming from the Selingud hydro power plant, EDM explored the feasibility of building a transmission line between Bamako and Segou (235 km) to supply this area with low cost hydro energy and retire the existing diesel generating plant. In May 1980 EDM asked a foreign power utility, acting as consultant, to study the feasibility of the project. At the same time KfW and CIDA expressed their interest in financing the project, if it could be economically justified. The consultant's report indicated that the transmission line was the least cost compared to the local thermal diesel alternative, the project cost was estimated at around US$25 million. The IDA was not satisfied with the original economic study and asked to have another one made. In 1983 the IDA informed KfW and CIDA that the detailed engineering contract would be financed by the IDA, and a call for proposals would be made shortly. Four firms were asked to submit proposals, two power utilities and two engineering firms. The two invited power utilities were already advising EDM through the TA component, indeed one of them was supplying the project coordinator to EDM, and the potential conflicts of interest escaped EDM and the IDA's attentions. 42. The two winning proposals for the engineering services were from the two state owned power utilities. In February 1983 the IDA was not satisfied with the project design submitted by the consultant: (i) it was over-designed and not suitable for Mali; (ii) the design was too expensive and the project was no longer economically viable. EDM wanted to accept the design suggested by the consultant but IDA reacted quickly and informed EDM they would refuse to finance the project as designed. 43. In April 1985 the consultant reviewed his original concept and submitted a new cost estimate (US$16 million). In April 1986 the EIB informed the IDA that the GOM had officially requested their participation in the financing of the project. In May 1987, KfW refused to accept the design of the EDM consultant, it not being suitable for the Malian conditions and too oriented towards North American operational requirements. The consultant, being a power utility, considered its experience with network operations as the final reference for the appropriateness of the design. Finally the project was implemented during the follow on project Power II. Water Work Project 44. The audit concludes that the overall outcome of the water system rehabilitation project component is marginally unsatisfactory. The physical goal of the system's rehabilitation was achieved as demonstrated by the growth of the water demand indicated in Table No. 3-2 (page 14), but the TA performance to failed to alleviate the EDM management problem in relation to the water operations. Water losses from 1985 to 1986 increased by 54%. From 1985 until 1991 the financial losses increased seven times. The sustainability of this component is uncertain. 45. During project preparation phase I, the cost of the project was estimated at US$19 million but at the end of the preparation, Phase II, the scope of the project was reduced to a total of US$2 million. In parallel, KfW was implementing a water works rehabilitation project for around US$4 million and a study for the water tariffs. In 1983 the OPEC Fund accepted to finance US$2.75 million of rehabilitation 13 work for the Bamako water system. The work done for the Bamako water system was divided into small lots, focussing on the elimination of the system's most urgent bottlenecks. The work financed by KfW, OPEC, CCCE and the IDA were inter-meshed to such an extent that it is somewhat difficult to identify and to separate one project from another. 46. The SAR estimated the EIRR of the IDA water project at 11%. The audit cannot recalculate the EIRR since the financial contributions of KfW, IDA, OPEC and CCCE are not available. Furthermore, the benefits are difficult to assess since technical and non-technical losses are not precisely known, and the benefits coming from the stand-pipes are economically difficult to ascertain. Table No. 3-3, indicates the results of the complete rehabilitation program of the Bamako's water system. During a field mission in January 1994 the audit observed the visible parts of the water infrastructure and concluded the work is of good quality, but that the maintenance could be improved. The financial situation of the water department of EDM is such that without cross subsidies coming from the power sub-sector, the continuation of its operation on a commercial basis would be very doubtful. From 1985 until 1989, when the rehabilitation program was completed, the production of water increased by 31%. The consumption (i.e., amount of water billed) increased only by 14.8%. The losses which were 28% in 1985, were 37% in 1989. The cost of production increased by 20% (1986-90) and the financial losses increased seven times. 47. From 1985 until 1986 when the IDA contribution was completed: (i) the water production increased by 29%; (ii) the water consumption increased by 19%; (iii) the number of customers increased by 6%; (iv) The water losses increased 50%: TABLE 3-3 WATER DEPARTMENT - PERFORMANCE INDICATORS 1983 1984 1985 1986 1987 1988 1989 1990 1991 Production 1000 m3 11.866 10.527 12.033 15.505 16.070 15.601 15.800 17.093 17.230 Growth rate % 18 -11 14.3 29 3.6 -3 1.3 8 8 Consumption 1000m3 7.033 6.372 8.660 10.300 10.600 9.910 9.942 11.034 11.898 Growth rate % 16 -9 35 19 3 -6.5 .32 11 7.8 No. of Customers 10,131 11,027 11,873 12,610 12,953 13,225 13,500 14,047 14,601 Losses m' 4833 4155 3372 5205 5470 5691 5858 6059 5332 Losses % 40 39 28 34 34 36 37 35 30 Tariff FCFA/m' 55 -- 118.3 118.4 116.4 116.3 117.8 117.8 Production cost FCFA/m3 139.8 151.0 156.8 168.2 167.9 183.0 Losses, million FCFA -83 -259 -348 -532 -568 -899 14 4. BROAD ISSUES AND FINDINGS Country and Sector Framework 48. The issue of indebtedness of the Malian economy and the power sub-sector in particular, had an indirect impact on the implementation of Power I, and this trend could continue in the future. During Power I debt servicing for Selingu6 was central to the question of adequacy of tariff level for Bamako. The IDA project was delayed due to the cross effectiveness clause with one donor, following Mali's default on a previous loan. The Mopti-Sevard power plant was delayed after the OPEC Fund stopped disbursements in the middle of the project due to Mali's default on a previous loan" . The above issue will be important in the future, not only for the size and cost of investments, but also for the tariff level. 49. During the preparation, phases I and II of Power I, the reform of a complex and inefficient institutional set-up was postponed to a sub-sequent project. The PCR (paras. 5 and 6.03) acknowledges the important impact of this decision on the Power I project results and even more on the overall performance of EDM. 50. The project file indicates that IDA Region was well aware during project preparation, that the GOM had no clear policy, goals or strategy for the power sub-sector. This would explain the GOM's perception that -- as mentioned to the audit mission -- the project was imposed by IDA. It would be beneficial for future projects' preparation to establish a clear consensus on objectives, and strategy for the sub-sector through comprehensive dialogue, including seminars given by senior specialists to assure a better sense of project ownership by the country authorities. Project Preparation 51. Several important issues not addressed during project preparation had an impact on the project results. 52. Assistance to the Borrower in Project Preparation. It took six years from identification to effectiveness of the project. EDM mentioned to the audit mission that they had no prior experience in preparing an IDA project. Preparation is the responsibility of the Borrower. However, assistance is sometimes needed, as it was in the Malian case. Borrowers without experience in project preparation and negotiations might consider obtaining assistance from independent specialists" who could guide or assist the Borrower's staff in complying with the IDA norms and criteria to prepare projects for quick appraisal. 53. Technical Cohesiveness of the Program. Power I was a program composed of many projects, some of which had technical and operational ramifications. EDM did not have the internal capability of reviewing the technical integrity of each project and verify all the ramifications. Assessment and review of the various projects, technical integrity and cohesiveness at the program level needed to be reviewed by experienced and independent specialists without links to the engineering firms and power utilities interested in the design and implementation of the project. " Putting IDA as loan administrator in a legal predicament with on-going contractual obligations. " Financed by the PPF. 15 55. A thorough assessment of the network loss reductions was needed before addressing the issue of tariff increases; reduction of technical and non-technical losses are very cost effective and should have been implemented before, or in parallel with, tariff increases. The TA component was not prepared by an independent specialist in TA, but by the firm which knew it would implement this particular component. Without EDM's autonomy and the TA in a non-hierarchical function to supplement the lack of experience of the counterparts, the TA component's relevance became doubtful on account of cost effectiveness. These aspects would most likely have been identified by an independent TA specialist. Donors Coordination 56. During project preparation phase I, the IDA did a commendable effort to contact potential donors to obtain financial contribution to the Power I project. OPEC was contacted successfully, KfW, CCCS and FAC were interested in participating, and other donors expressed interest in joining at a later stage. During project preparation phase II, the dialogue among the the other donors deteriorated and coordination became almost non-existent, until a first round-table of donors (ACDI, KfW and BEI) chaired by IDA met in Frankfurt in June 1986, followed by a meeting in Luxembourg. The audit concludes that earlier donor meetings and round-tables to assist the GOM with the coordination of all the parties involved in the sub-sector, could have helped in finding solutions to the difficulties encountered in the project preparation. Dialogue Between IDA and GOM 57. The delay in the project's preparation and negotiations was to some extent due to a lack of commonality of concepts between the IDA and Government authorities. The role of the Government and of a power utility does not appear to be well understood by some Malian Government officials. The distinction between a commercial public utility and a social service is blurred. Perhaps a dialogue comprising study sessions and conferences with senior specialists could have made a difference. The audit notices that the IDA's strategy, objective and goals for the sub- sector are not well accepted or understood, even in early 1994, by some Government authorities. Project Ownership 58. The question of ownership was discussed with government officials and with the EDM during the audit's field mission in January 1994. Surprisingly, at the Government and EDM level few respondents identified themselves with the first power project and do not consider having decided on the content of the project. The audit noticed that some confusion exists among the Malian in relation to Power I, as formulated during preparation phase II, and the current Power II projects. Some Malian authorities believe the Power I project was imposed by the IDA, in spite of the fact that the IDA, during project preparation phase II went to great length to change the nature and scope of the project to give the Malian authorities, EDM management and the staff union the satisfaction of full control over the project management. During project implementation GOM and EDM had the opportunity to make the procurement of goods as services, with the IDA playing a distant reviewing function. i8 Power II, Credit 1998-MLI. 16 Loss Reduction and Financial Viability 59. During project preparation the Bamako power network losses were estimated to be approximately 10%. Shortly afterwards, with better information available, the losses (technical and non-technical) were found to be approximately 22%, and worth around US$9 million per year (1993). The fastest and fairest way to improve EDM's financial performance would have been to reduce first the non-technical losses, and then, if necessary, increase the tariff. During Power II, from August 1991 to October 1993 a vigorous and systematic campaign for reduction of non-technical losses was conducted by EDM and resulted in 5 GWh being recuperated in a few months. If this trend were sustained over a period of 12 months, it may lead to a yearly saving of about 20 GWh, approximately US$4 million per year (1993) of more revenue and a long way to redress the financial performance of EDM. Tariff increases alone without first reducing the non-technical losses, tend to be too high and increase these losses. Project Supervision 60. The PCR part III mentions that 15 supervision missions were made during Power I. The Audit could find only 11 supervision mission reports in the project file; they are informative on the problems the project was facing during implementation. The non-compliance of the GOM with the financial clauses of the covenant are well documented, and some of the reports contain suggestions to redress the situation. Supervision reports, however informative they are, cannot improve the situation unless detailed remedial activities are agreed upon and implemented by both parties. It is likely the follow-up activities may require more man-power in the field than the IDA can provide. Suspension of disbursements and eventual cancellation of project should be considered not only as a threat but as a certainty, unless remedial actions are taken within a time frame. Performance of IDA 61. The audit does not concur with the PCR"' statement that IDA performed well, since project preparation took six years and the design of the TA achieved only modest results. Performance of the Borrower 62. The audit concurs with the PCR" that the borrower performance was not satisfactory. 19 PCR, chapter 8, page 8: Bank Performance. 2 PCR, chapter 9, page 9. 17 5. CONCLUSIONS AND LESSONS Project Overall Ratings 63. The overall outcome of the Power I project is marginally unsatisfactory because the physical goals of the project were achieved, albeit with delay and with problems not foreseen during project preparation time. The TA component's preparation was unsatisfactory, the implementation was satisfactory but could not reverse the effect of the poor preparation. It did alleviate some of the EDM's operation and managerial problems, but the overall institutional development impact is rated as negligible. The PCR rated the project outcome as unsatisfactory and its institutional development as negligible. The audit rates the sustainability of the project's benefits as unlikely as long as EDM is unable of generating revenues and developing an appropriate skill base. The PCR rated it as unclear. The institutional reform and the sub-sector reorganization originally intended were postponed to a subsequent credit. The power sub-sector performance at the end of Power I was not much better than at the beginning, and the water sector results were alarming. Lessons Learned 64. The outcome of the audited project suggests lessons which might be applicable to other projects. (i) To achieve optimal efficiency and congruence of objectives and project ownership by the Borrower as early as possible, project preparation should identify and correlate the sub-sector problems, distortions and constraints setting the priorities for interventions, and designing the project with clear objectives. (ii) In parallel with the assistance to the Borrower to prepare the project, a dialogue between IDA and utility's management is essential to obtain congruence of policy, objective, goals and strategy between the donors and the recipient at project's onset. (iii) Donor meetings are necessary for efficient preparation and smooth implementation of the various projects. The IDA, in its capacity of a multilateral institution, should organize and call these meetings. These meetings should disseminate information among all participants to avoid duplication of work, develop a common objective, and assure cohesiveness of activities. (iv) Donor meetings should regularly review the original assumptions and premises of a program lasting many years. Changes resulting from the reviews should not be considered a weakness or admission of error, but a necessary step in a dynamic review process. (v) Tariff increases to assure financial viability of a power utility suffering from abnormally high non-technical losses in power distribution penalizes honest consumers and encourages the growth of the non-technical losses. A dynamic and permanent campaign to suppress non-technical losses should precede tariff increases, and creates a sense of fairness in the population. (vi) Continual and systematic violation of some of the covenant's conditions identified by supervision mission should be reviewed by the IDA's management and if still found pertinent, 18 suspension of disbursements and even cancellation of a project should not be a potential and distant threat but a real option. (vii) Procurement for small or even medium sized rehabilitation or construction projects made through a multiplicity of small contracts does not capture economies of scale and overwhelms the Borrower's weak managerial capacity. A cost effective alternative would be a turn-key approach with performance boni (or penalties). (viii) Disregard of the Bank guidelines for hiring consultant services, especially when conflicts of interest are involved, creates precedents which incite discord among consultants, and results in lack of credibility. (xi) A project lasting many years and developing serious problems might benefit by being audited in mid-course. 19 ANNEX I Page 1 of 2 PROCUREMENT OF CONSULTANT SERVICES The hiring by EDM of consultants faced multiple difficulties as follows: (i) The study of the institutional reform and the sub-sector reorganization. IDA prepared the TOR and suggested to limit the call for proposals to a few national power utilities. This study was delayed by the GOM and did not begin before early 1985. The results generated some controversy. (ii) The Bamako-Segou transmission line detailed design. Two power utilities were in competition with two engineering firms. After an EDMIIDA meeting in Bamako, the power utility which supplied the coordinator for the Power project obtained the contract. (iii) Review of EDM commercial department (billing and collection system). The contract was awarded to an engineering firm; after spending US$3 million with little results, the contract was canceled. (iv) The Bamako Distribution Network Rehabilitation Project. Two state owned power utilities were in competition with two engineering firms. The contract was awarded to an engineering firm. (v) The Felou Power Dam Rehabilitation Study. The contract was awarded to an engineering firm. The results appear to have been satisfactory. (vi) Study of the interconnection network of Southern Mali's principal towns. The project file does not contain information on this particular project, although this problem is discussed in the Master Plan prepared by a consultant. (vii) Sotuba Hydro Power Plant Extension. The IDA objected to power utilities being hired as consultants. Finally the project was given to an engineering firm doing the work on the Felou project. The engineering firm alleged to be the winner of the competition was not awarded the contract, and when asked, IDA declined to intervene. (viii) The Bamako Water Work. The engineering firm which did the study was retained to do the detailed engineering and prepare the bid documents. The results are deemed to be satisfactory. (ix) The Mopti-Sevar6 Power Plant. The power utility which did the study for the biomass plant was, after the cancellation of this project, retained to do the detailed design and preparation of the bid documents for a diesel power plant. The bids were canceled due to being twice the normal price. The power utility was removed from the project and replaced by an engineering firm. 20 ANNEX I Page 2 of 2 (x) Mali Power Sub-Sector Expansion Master Plan. The IDA insisted at the onset that the two power utilities supplying the TA to EDM would not be invited to submit a proposal. One of them complained through its Executive Director of unfair treatment, and the other power utility could not understand why it would be excluded, when at the same time in the neighboring country, the IDA allowed them to do exactly that.2" The contract was awarded to an engineering firm. The results were satisfactory. (xi) The local architects were responsible for the design of the three administrative buildings. One architect was removed from the project. (xii) The Dar Salam Power Plant Rehabilitation. No details on the selection process or the results. An engineering firm was selected. During a field mission in January 1994 the audit found the condition of the Dar Salam power plant unsatisfactory. 21 Guinea. Engineering II and TA (Credit 1595-GUI). 21 ANNEX II Page 1 of 1 COMMENTS FROM THE BORROWER ENERGIE DU MALI DIRECTION GENERALE BAMAXO OBSERVATIONS SUR LE RAPPORT D'EVALUATION RÉTROSPECTIVE DU PREMIER PROJET ELECTRICITE ET EAU (crédit 1282 MLI) 1) EDM signifie Energie du Mali et non électricité du Mali. 2) Lire Fonds de l'OPEP et non OPEP. 3) Page V dernière ligne : suite du Projet au lieu le suivi du Projet. 4) Page VI 2ème paragraphe Lire EDM est une société d'économie Mixte qui a la concession de la production, du transport et de la distribution de l'eau et de l'électricité sur toute l'étendue du territoire national. 5) Les études portaient sur la réforme institutionnelle du secteur de l'électricité et non du secteur de l'Energie. 6) Page VII, paragraphe 7, lire : ..... Certaines des études devaient préparer le projet suivant Electricité II, notamment l'élaboration d'un plan directeur en vue ...... 7) Page 1, paragraphe 5, lire : ... EDM a été créée le 14/10/1960. Le capital se répartissait comme suit - Gouvernement du Mali : 55% - Caisse centrale de Coopération économique : 39% - Electricité de France : 6% 8) Nous ne partageons pas le point de vue qui consiste à dire "Une concession de 30 ans décourage l'investissement à long terme", car il ya le principe de l'amortissement de caducité. 9) La réalisation de la ligne THT Bamako-Ségou était prévue dans le cadre du 2ème Projet Electricité et non du ler Projet comme indiqué à la Page 2, paragraphe 7. Ce sont les études de cette ligne qui étaient prévues dans le cadre du 1er Projet Electricité/Eau. 10) Quelle est la source de la déclaration suivante : "En 1977, la situation du réseau électrique au Mali était désastreuse et les autorités envisageaient de céder EDM à une entreprise étrangère, mais cette idée n'a pas été poursuivie même soutenue par les donateurs" ? 22 ANNEX II Page 2 of 2 11) Page 6, paragraphe 18, préciser que les 20% des pertes d'électricité sur le réseau, renferment aussi les consommations les auxiliaires. 12) Page 8, Tableau 3.1. : Lire travaux de Génie Civil au lieu de travaux publics. 13) Page 13, paragraphe 42. Le contrat pour les études d'avant- projet détaillé de la ligne Bamako-Segou a été signé en 1984. L'IDA, ne pouvait donc pas en 1983 donner un avis sur la conception du projet soumis par le consultant. 14) Quelle est la source, selon laquelle "les autorités maliennes ont indiqué en Janvier 1994 que le projet leur avait été imposé par l'IDA" ? Pour nous le projet a été financé par l'IDA sur la base d'une requête officielle du Gouvernement du Mali. 15) Page 17, paragraphe 1 : les réunions de Francfort et du Luxembourg concernaient seulement les bailleurs de fonds de la ligne THT Bamako - Ségou (ACDI, KFW, BEI). 16) Page 19, paragraphe 65 ii) : Quelle est la source de cette affirmation "on déclarait qu'une compagnie d'électricité (EDM) était un service social et non une opération commerciale" ? 17) Page 21, ii) : Le choix de la compagnie chargée des études détaillée de la ligne Bamako Ségou a été proposé par EDM suite à une analyse objective et validé par l'analyse conjointe EDM/IDA à Bamako. 18) Page 22, v) : l'étude pour la remise en état de la centrale du Félou n'est pas une composante du 1er Projet Electricité/Eau. 19) Page 21 vi) : Le Plan Directeur élaboré par TRACTEBEL dans le cadre du 1er Projet Electricité/Eau traite ce problème. 20) Page 21, vii) : Il semble avoir ici une confusion, car l'étude de factibilité de la réhabilitation de la centrale de Sotuba a été faite par la SGTE tandis que l'étude d'extension a été faite par le Cabinet Sir-Alexander GIBB. 21) Page 22, xi) : le choix des architectes locaux a été fait sur appel d'offres (consultation restreinte) et les dossiers de sélection ont été approuvés par l'IDA. 21) Page 22, xii) : L'état "désastreux" de la centrale de Dar- Salam nous semble exagéré. Fait à Bamako le 8 Juin 1994.