Guinea - First and Second Power Projects

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 13294 PERFORMANCE AUDIT REPORT REPUBLIC OF GUINEA FIRST POWER PROJECT (CREDIT 1085-GUI) AND SECOND POWER ENGINEERING AND TECHNICAL ASSISTANCE PROJECT (CREDIT 1595-GUI) 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. EXCHANGE RATES During project identification (1977) US$1 = Sylis 20 Shadow rate was estimated at twice the official rate i.e., US$1 40 Sylis Appraisal year (1980) US$1 = Sylis 20.93 During project (1981) US$1 = Sylis 20.93 implementation (1982) US$1 - Sylis 22.36 (1983) US$1 - Sylis 23.37 (1984) US$1 = Sylis 24.29 (1985) US$1 = Sylis 25.00 Exchange rate fixed (1986) US$1 = GF 360 weekly in an auction (1987) US$1 = GF 428 (Yearly average) (1988) US$1 = GF 474 (1989) US$1 = GF 591 (1990) US$1 - GF 661 (1991) US$1 = GF 792 (1992) US$1 = GF 800 (1993) US$1 = GF 900 (1994) US$1 = GF 960 MEASURES Gram (g) = 0.0022 pounds Kilometre (km) = 0.62 mile Meter (m) = 39.37 inches Kilovolt (kV) = 1,000 volts Kilowatt (kW) = 1,000 watts Kilovolt-ampere (kVA) = 1,000 volt amperes Megawatt (MW) = 1,000 kilowatts Gigawatt hour (GWh) = 1 million kilowatt hours (1OkWh) ABBREVIATIONS AND ACRONYMS AGCD - Administration G6ndrale de Coop6ration au D6veloppement (Belgium) CCCE/CFD - Caisse Centrale de Coopdration Economique (France), currently Caisse Francaise de Ddveloppement CEFC - Committee of Economic and Financial Coordination. CIDA - Canadian International Development Agency ENELGUI - Entreprisc Nationale d'Alectricit6 de Guin6e (from 1987) ESMAP - Energy Sector Management Assistance Program G IZ Gesellschaft fur Technische Zusammenarbeit (Germany) GOG - Government of Guinea IDA - International Development Association IMF - International Monetary Fund JICA - Japan International Cooperation Agency KfW - Kreditanstalt fdr Wiederaufbau (Germany) LRMC - Long Run Marginal Cost LV - Low Voltage MINEK - Ministre de l'fnergie et du Konkour6 MRNEE - Ministre des Ressources Naturelles, de l'fnergie et de l'Environnement MIP Management Improvement Program MV - Medium Voltage OCOFI - Office de Coordination Financi6re de l'Industrie ONAH - Office National des Hydrocarbures PAR -Performance Audit Report PCR - Project Completion Report PPF - Project Preparation Facility PR - President Report SAR - Staff Appraisal Report SEE - Secr6tariat d'tat A lInergie SDR Special Drawing Rights SNE - Socit6 Nationale d'Alectricit6 (until August 1987) 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 Guinea First Power Project (Credit 1085-GUI) and the Second Engineering and Technical Assistance Project (Credit 1595-GUI) Attached is the "Performance Audit Report on Guinea - First Power Project (Credit 1085- GUI) and the Second Engineering and Technical Assistance Project (Credit 1595-GUI)". The IDA Credit for the First Power project was part of the Government of Guinea's rehabilitation program of the power network of Conakry, the nation's Capital. The rehabilitation of the distribution network was completed with significant delay, cost overruns and reduction of scope of work. The Technical Assistance (TA) component of the First Power Project did not achieve its goal of strengthening the Societ6 Nationale d'Electricit6 (SNE) operations and redress its financial situation. This component's implementation was delayed by almost two years, due to problems not identified at preparation time, as well as by lengthy negotiations with executing agencies. The TA team in advisory functions was ineffective. The outcome of this project has been rated as unsatisfactory, its sustainability as unlikely and its institutional development impact as negligible. The IDA Credit for the Second Engineering and Technical Assistance project was a continuation of the First Power Project with the same contractor responsible for the TA, which failed again to reach its goals. At the end of the project the Government, supported by IDA and the donors, decided to farm out the sub-sector operations to a foreign management firm. This decision launched the sub-sector institutional reform, albeit not in the way originally intended. The Second Engineering Project included a set of engineering studies to identify the least-cost hydro plant site in Guinea to supply the Conakry network's demand. These studies did not succeed in identifying a hydro plant site less costly than the thermal generation in Conakry. However, the studies were instrumental in paving the way for a consensus between IDA and the Government on the sub-sector strategy. From 1980 until 1992, the national power utility never met the power demand of Conakry, in spite of two major assistance programs, forcing the consumers to install their own power generating capacity. The outcome of this project is unsatisfactory, its sustainability is unlikely and the institutional development impact is modest. Robert Picciotto by H. Eberhard Kopp 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.  FOR OFFICIAL USE ONLY PERFORMANCE AUDIT REPORT REPUBLIC OF GUINEA FIRST POWER PROJECT (CREDIT 1085-GUI) AND SECOND POWER ENGINEERING AND TECHNICAL ASSISTANCE PROJECT (CREDIT 1595-GUI) TABLE OF CONTENTS Page Na PREFA CE .................................................. i BASIC DATA SHEETS ..... ...................................... iii EVALUATION SUMMARY ........................................ viii 1. BACKGROUND ............................................. 1 The Context.............................................1 Institutional Organization....................................1 Power Sub-Sector Evaluation and Trend........................2 Objectives of the Projects...................................2 Description of the Projects...................................3 The PCRs............................................... 3 2. PREPARATION AND APPRAISAL OF THE PROJECTS ............. 3 3. IMPLEMENTATION AND OUTCOME OF THE PROJECTS . . 7 Actual Project Costs ...................................... 7 Costs of Engineering H ..................................... 8 Economic Rate of Return of Power I .......................... 9 Power I Project Outcome .................................... 10 Engineering II Project Outcome .............................. 12 Strengthening of the SNE - Power I and Engineering II ............. 13 This report was prepared by Alvaro J. Covarrubias, Task Manager, and Alain Gertsen-Briand, (Consultant), who audited the project in January 1994. Mrs. Maryvonne 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 (cont.) PageNo. 4. BROAD ISSUES AND FINDINGS ........................................ 14 Institutional Reforms...............................................14 Donors Coordination...............................................14 Project Preparation................................................14 Project Ownership.................................................15 Divergent Views on System Expansion...................................15 Dialogue between IDA and the GOG...................................16 Financial Viability and Tariff Increases...................................16 IBRD and IMF Policies Coordination...................................17 Projects Supervision and Follow-up..................................... 18 5. CONCLUSIONS AND LESSONS......................................... 18 Rating of the Power I Project......................................... 18 Ratings of the Engineering II and TA Projects............................. 19 IDA and Borrower Performance.. .....................................19 Lessons Learned.............. .....................................20 Annexes: 1. Comments from the Borrower ........................................... 22 2. Comments from KfW ................................................. 24 PERFORMANCE AUDIT REPORT REPUBLIC OF GUINEA FIRST POWER PROJECT (CREDIT 1085-GUI) AND SECOND ENGINEERING AND TECHNICAL ASSISTANCE PROJECT (CREDIT 1595-GUI) PREFACE This Performance Audit Report (PAR) presents the results of IDA's Power Project (Credit 1085-GUI) and Second Engineering and Technical Assistance Project (Credit 1595-GUI)'. For Power I, the audit is based on the review of the following documents: the SAR dated November 25, 1980, the Project Completion Report (PCR) dated May 6, 1988, (Report No. 7235), the project files, the President's Report and Recommendations to the Board, dated December 1, 1980, and the Development Credit Agreement between the Revolutionary People's Republic of Guinea and IDA, dated February 10 1981. The Guinea Power I Project was identified in 1978. The project was closed in 1986. For Engineering II, the audit is based on: the President's Report and Recommendations to the Board dated April 18, 1985, the Development Agreement dated June 12, 1985, and the PCR dated December 28, 1992, (Report No. 11499). The project was closed in December 1990, and an undisbursed balance of SDR 48,545 was canceled on April 30, 1992. In January 1994, an OED mission visited Guinea and interviewed present and former officers of the Government of Guinea (GOG) and the national utility who were involved in the power sub-sector at project time. Discussions were also held with project officers still with IDA. In August 1987, the Socidtd Nationale d'Electricit6 (SNE) was replaced by the Entreprise Nationale d'Electricit6 de Guinde (ENELGUI). With a new management and status, the new utility ENELGUI inherited all SNE infrastructure and liabilities. In December 1989, the GOG gave full autonomy to ENELGUI, and in February 1990, the utility's operations were farmed out to a foreign management firm. From late 1970s to early 1990s IDA's two projects were managed by six different division chiefs. Furthermore, the audit counted around twenty power engineers, financial analysts, economists, and loan and program officers, who were involved in the two projects at one time or another. On the GOG side, there was one change of Government and regime in 1984, and up to 1991, there were four changes of ministers and five new managing directors of the SNE/ENELGUI. As a result and since the project file is not complete, it is difficult to ascertain precisely causality and correlation of the events which influenced the implementation of the two audited projects. The audit could not First Power Project (Credit 1085-GUI) hereafter in the report called Power I. Second Engineering and Technical Assistance Project (Credit 1595-GUI) hereafter in the report called Engineering II . n interview all the officers involved in the projects, and those interviewed could not always remember in detail all the events, many of which occurred ten years previously. The audit reviewed some aspects of the two projects, which were not identified or analyzed in depth by the PCR, more precisely: (i) impact of the approach selected for strengthening SNE; (ii) project identification and preparation in relation to the sub-sector environment; (iii) assumptions in relation to the benefits; (iv) contract awarding process for the network rehabilitation; (v) handling of the Technical Assistance (TA) component, when it was evident the original goals could not be achieved within the existing institutional framework; and (vi) supervision and follow-up during Power I and the first part of Engineering II. In the audit's opinion, the two PCRs would have benefitted from a more thorough analysis of the causality of the poor results. Furthermore, the absence of the Borrower's perspective, PCR, part II (Power I) gave a one-sided view of the events, especially in relation to the lack of congruence of objectives between IDA and GOG which lasted until the end of the second project. Following OED procedures, copy of the draft PAR was sent to the Borrower and co- financiers for comments. The comments received from the Borrower and Kreditanstalt fUr Wiederaufbau (KfW) are reflected in the PAR and are attached as Annexes to the report. iii PERFORMANCE AUDIT REPORT REPUBLIC OF GUINEA FIRST POWER PROJECT (CREDIT 1085-GUI) BASIC DATA SHEET KEY PROJECT DATA 1980 1986 1994 Item SAR PCR AUDIT Total project cost (in US$ millions) 32.50 30.09 Idem Foreign exchange cost (in US$ millions) 28.50 27.41 - Credit amount (in SDR millions) 22.10 22.10 - (in US$ millions) 28.50 23.68 - Cost overrun in foreign exchange (US$ millions) 3.73 - (Cost over credit amount in %) 16% - Completion of project physical components: - rehabilitation of Conakry II MV distribution 80% 100% - rehabilitation of Conakry II LV distribution 80%2 80% - training and workshop centers 100% 90% - housing facilities 100% 100% Economic rate of return 23% 20% -7% CUMULATIVE FORECAST AND ACTUAL DISBURSEMENTS (SDR millions) As of December 31 FY81 FY82 FY83 [Y84 FY8 Oct.31 1986 Appraisal 3.18 8.84 16.59 21.32 22.10 Actual 1.27 7.37 13.25 19.78 21.72 22.1 Actual as % of appraisal 40 83 80 93 98 100 2 The PCR combines MV and LV distribution. iv PROJECT DATES Item Original Plan Actual Project identification 4/7 Government request 6n7 Preliminary work - PPF 3/78 Appraisal(with CCCE, KfW) 12n9 Negotiations 10/9 11/80 Board approval of credit 10/80 12/80 Credit signing - 2/81 Beginning of implementation - 2/80 Effectiveness date 1/81 5/81 Closing date 12/85 6/86' Borrower: Revolutionary People's Republic of Guinea. After April 1984: Republic of Guinea Beneficiary: Socidtd Nationale d'Alectricitd (SNE) After August 1987: ENELGUI Follow-on projects: 1) Second Engineering and Technical Assistance Project, Credit 1595-GUI, June 1985 2) Power II Project Credit 2416-GUI, February 1993 STAFFINPUTS (Staff Weeks) FY79 EIL FY81 EYU EX EY84 FY85 FY86 FY87 FY88 TOTAL Preappraisal 15.3 12.3 1.0 28.6 Appraisal .1 30.2 20.2 50.5 Negotiation 6.0 6.0 Supervision 8.6 25.4 15.9 14.1 15.4 8.2 8.3 .8 96.7 Other .4 14.7 9.8 24.9 Total 15.8 57.2 45.5 25. 159 14.1 15.4 8.2 8.3 .8 206.7 Under transfer of Credit S22.GUI. 4 Date is the date on which the funds under 1085-GUI were fully disbursed and the account closed. A moratory of six months was given to reconcile all expenditures. The next credit was already disbursing through a PPF for the continuation of the activities not yet completed under Power I, more specifically the TA and some studies. V MISSION DATA Month No. of Days Item Year Sagius Trend Persons in field Man-days Project identification 05177 Project preparation 07/77 Preappraisal 04179 Appraisal 11179 Supervision 1 04/82 3 3 2 9 18 Supervision 2 09/82 2 1 2 9 18 Supervision 3 01/83 2 1 2 9 22 Supervision 4 07/83 2 2 1 12 12 Supervision 5 03/84 3 2 2 19 38 Supervision 6 07/84 3 2 2 7 14 Supervision 7 06/85 2 2 2 9 18 Supervision 8 01/86 3 3 2 10 20 160 The above result in: 160 man-days of field supervision STATISTICAL INFORMATION ON PROJECT 1. Related IDA Credits to Guinea Credit Name Purpose Year approved Status Comments First Power Rehabilitation, reinforcement of 1980 Completed PCR states project failed in (Cr. 1085) failed in power utility. institutional & financial areas. Industrial Rehabilitation of PEs, 1982 Completed Unused funds used to Rehabilitation & promotion of private enterprises cover shortfall in (Cr. Promotion (Cr. thmu credit line, TA. 1595) for TA. 1234) Petroleum Geological survey, TAa, 1983 Completed Idem. Exploration (Cr. training for oil exploration, 1438) extension of power sector TA contract. Public Enterprise Institutional legal reform of PE 1992 Board June 92 PPF provides funds for Technical Assistance sector,privatization financial studies of regulatory restructuring of key PEs. framework. Power II Proposes innovative institutional (1993) Negotiated PPF funds ongoing TA and reform; system rehabilitation & May 1992 studies. expansion. v PERFORMANCE AUDIT REPORT REPUBLIC OF GUINEA SECOND POWER ENGINEERING AND TECHNICAL ASSISTANCE PROJECT (CREDIT 1595-GUI) BASIC DATA SHEET KEY PROJECT DATA Appraisal Actual Total Project Cost US$ 20.0 million 25.0 million Credit amount SDR 8.4 million 8.4 million Credit amount US$ 8.0 million 11.0 million Borrower: Republic of Guinea Beneficiary: Socidtd Nationale d'Alectricit6 (SNE) After August 1987: ENELGUI Follow-on project: Power II Project (Credit 2416-GUI, February 1993) PROGRAM DATES Original Plan Actual Project identification 7/83 7/83 Appraisal 10/83 10/83 Negotiation 3/85 3/85 Board approval 5/85 5/85 Credit signature 6/85 6/85 Credit effectiveness 9/85 6/86 Credit completion 6/90 6/91 Project closing 12/90 12/91 CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS FY85 FY87 FY88 FY89 FY90 FY91 Appraisal US$ 1.12 2.3 2.1 1.44 .8 .24 Actual Not available in the PCR Vii STAFF INPUT (Staff Weeks) FY80 FY81 FT82 TY83 FY84 FY85 TY86 FY87 FY88 FY89 FY90 FY91 FY92 FY93 TOTAL Proappraisal .9 1.1 4.S 8.4 15.2 Appraisal 19.0 30.6 49.6 Negotiation 8.g 8.9 Supervision 2.5 29.1 32.4 33.9 22.6 18.5 12.2 18.7 169.9 Other 2.7 .1 2.2 5.7 2.0 3.8 1.3 1.0 18.8 Total .9 1.1 2.7 4.9 29.6 47.6 29.1 32.4 35.9 22.6 18.5 16.0 19.9 1.0 262.4 MISSION DATA Month/ No. of Days in Specialization Perfm.* Types of Year Persons Field represented I/ Rating 2/ Problems 3/ Supervision I 11/85 2 7 FA/PE 3 F,M H 06/86 7 7 E/PE 4 F,M,I,P 11I 12/86 2 13 E/IS IV 02/87 1 7 IS V 05/87 2 10 PE/FA 4 F,M,I,P,T VI 10/87 3 10 E/FA/PE 3 F,M,I VII 02/88 2 10 E/FA VI 07/88 2 4 E/FA IX 11/88 1 3 H X 02/89 2 16 FA/E 3 F,M,I,P XI 04/89 2 7 FA/PE 3 F,M XII 06/89 1 5 B XIII 09/89 2 6 FA/PE 3 F,M,I XIV 02/90 1 4 E XV 04/90 3 16 FA/PE 3 F,M,I XVI 10/90 1 4 FA XVII 03/91 1 12 FA 2 F,T XVIII 06/91 1 9 FA 2 F,T Abscence of rating indicates that Form 590 not prepared. 1/ PA = Financial Analyst PH = Power Engineer E = Economist IS - Institutional Specialist 2/ 1 - Problem free 2 - Moderate problems 3 - Major problems 3/ F - Financial, M = Managerial. I - Institutional, P = Political, T - Technical  viii PERFORMANCE AUDIT REPORT REVOLUTIONARY PEOPLES REPUBLIC OF GUINEA REPUBLIC OF GUINEA FIRST POWER PROJECT CREDIT 1085-GUI AND SECOND POWER ENGINEERING AND TECH. ASSISTANCE PROJECT CREDIT 1595-GUI EVALUATION SUMMARY Operations 1. This audit reviewed two projects financed by: I; (ii) a set of engineering studies to identify the (i) Credit 1085-GUI of SDR 22.100.000 approved least-cost hydro power plant to satisfy the Conakry on February 10 1981, made effective in May 1981 demand; (iii) the credit for the repair of the and closed, fully disbursed in June 1986, seven embankment of the Ban6a dam was not originally months later than expected; and (ii) Credit 1595- included. GUI of SDR 8.400.000 approved on June 12, 1985, made effective June 1986 and closed in Objective and Goals December 1991, one year later than expected, with SDR48,545 undisbursed. The delays to make the 4. The objective of the complete program was to credits effective were due to the GOG's difficulty increase the power generation from 100 GWh in in meeting all credit conditions, in fact some of 1980 to 260 GWh in 1985 to meet the expected them had to be waived. Both projects were part demand. To achieve this objective the SNE of a larger rehabilitation program aimed at the needed to strengthen its operations through a Conakry power network. comprehensive TA project, to be financed by IDA. The two goals of IDA's first project were: 2. The Borrower was the GOG, which on-lent it (i) the strengthening of the SNE by the TA. (ii) to the Soci6t6 Nationale d'Electricit6 (SNE), and the rehabilitation of the Conakry distribution after August 1987 to Energie Electrique de Guin6e network. The goals of the project Engineering II (ENELGUI), a government owned power utility & TA were: (i) to continue strengthening the SNE with monopoly rights for the generation, operations with the TA, (ii) carry out a set of transmission and distribution of electricity in studies to achieve a major investment in hydro Guinea. electric generation by the late 1980's. 3. The Power I project included: (i) the Implementation Experience extension, reinforcement and rehabilitation of the distribution network of Conakry; (ii) a massive TA 5. The Power I project did not achieve its project to strengthen the state owned power objective or its goals. The TA component did not utility; and (iii) the building of lodging facilities for strengthen the SNE operations: all the the TA expatriates, a training center, an performance indicators are negative or instrument calibration laboratory and a workshop. considerably lower than the initial target. The The Engineering II and TA project included; (i) consultants in advisory function within a power the continuation of the TA component of Power utility without autonomy from government control, ix could not perform efficiently in the Guinean generation and the dilapidated state of the context. The rehabilitation of the Medium distribution network. Voltage (MV) network was achieved, albeit not as intended and with cost overruns. The Low 9. The Power I's PCR recalculated the EIRR Voltage (LV) rehabilitation was not completed as and confirmed the 20 % estimated at appraisal intended, the scope of work being reduced by time. The audit cannot confirm this result, and approximately 20%. The physical component, estimates it around -7%. During implementation necessary to support the TA component, was of the two projects the ENELGUI financial completed two years late and with substantial cost performance did not improve. It was only in 1992 overruns. that ENELGUI achieved for the first time a positive operational income, unfortunately by 6. The Engineering II project did not realize its shutting down its thermal generation plant to goal. Again the TA, in spite of increased avoid an IMF imposed tax on fuel, relying only on responsibility, failed to strengthen ENELGUI. The its hydro plant generation and increasing even second failure of the TA forced IDA to stop the more the load shedding. experiment at the end of the second contract and suggest a new approach. The GOG then decided 10. The failure of the TA component provoked to implement some major institutional changes, on the institutional reform at the end of the the line suggested by IDA. The set of engineering Engineering II project; albeit not as originally studies could not find a hydro plant site less costly intended: in September 1990, the GOG than a thermal plant alternative. Nevertheless, the implemented one of the recommendations of a studies were instrumental in achieving congruence 1986 ESMAP study and gave full autonomy to of objective between donors and GOG for the ENELGUL. It later farmed out its operations to future power expansion program. a foreign management firm. Project Results 11. On the above basis -- and contrary to the PCR that rated the outcome of Engineering II and 7. The two projects failed to achieve the GOG's TA project as satisfactory and the institutional objective to increase the power generation to meet development impact of both projects as negligible - the demand of Conakry. In 1985, at the end of - the audit considers that the outcome of both Power I, the generation shortfall from the initial projects was overall unsatisfactory and that the estimate was approximately 47%. During Eng. II institutional development impact rates as negligible the situation worsened, and load shedding became for Power I and mods for the Engineering and a daily planned occurrence which still afflicted TA H projects. Conakry in January 1994. As a result, from 1983 till 1992 the private sector installed for its own use Sustainability 70 MW of power generating plants and in 1993 produced around 109 GWh of electricity. By 1993 12. For Power I and Eng. II projects, no benefits the private sector had installed nearly as much were achieved until the operations of the sub- generating power capacity as ENELGUI. sector were farmed out to a management firm; then some financial progress emerged. However, 8. The failure of the two costly TA projects on account of the uncertainty of the benefits and convinced the GOG in the final stage of the cost of the selected approach to redress Engineering II (1990) to give full operational ENELGUI, the sustainabilty of both projects is autonomy to ENELGUI and farm out the rated as unlikely responsibilities of the sub-sector to a foreign management firm. The above change improved Conclusions and Lssons the financial control and strengthened ENELGUI's administration, but did not improve 13. The main conclusions of the audit are that the the operations much, due to lack of power lack of success of the two projects cannot be x attributed to the non-compliance of IDA with its Government interference in operations, or own country's strategy and broad sector policies. no guarantee of full independence, The unsatisfactory outcome is mainly the result of authority and accountability. a difficult political and economic environment and a lack of an experienced skill mix of specialists (e) The conditions included in a credit which influenced project preparation and agreement need to be realistic, and take implementation, combined with the failure of IDA into account the local economic and social over ten years to reach substantive agreement with environment and, barring a case of force the GOG regarding goals for the sub-sector. majeure, must be enforced. Too stringent conditions might result in delay in credit 14. The main lessons of the audit are: effectiveness, or be waived during project implementation, signalling to the Borrower (a) When the sub-sector institutional that if he waits long enough before framework does not allow clear, separate implementing the conditions, in the end responsibility and accountability of each IDA will be lenient. entity involved in the project, and when the executive agency does not have the (f) Donors' coordination, whether done by the absorptive capability for a large capital Borrower and/or IDA, is important for the investment project, donors might consider success of a multi-donors financed it wise as a first step to initiate a power program. Effective donors' coordination sub-sector reform and delay investment, requires from the coordinator excellent until appropriate sectorial and institutional communication and interpersonal skills as reforms have been achieved. well considerable sectoral experience to achieve the credibility needed for efficient (b) Technical assistance even with managerial coordination. authority, cannot financially redress a utility only through repeated tariff (g) Infringement of Bank guidelines for increases. Required Tariff increases procurement of professional services or quickly become prohibitive, when the laxity with conflict of interest should not distribution network is dilapidated and the be tolerated. Otherwise it is conducive to technical and non-technical losses are suspicion and discord between consultants, many times over accepted norms. donors and Borrower, with negative impacts on the project. The results of the (c) Technical assistance needs to be planned studies done by consultants with obvious by specialists with experience in each conflict of interest are generally contested discipline and with a good appreciation of and not credible. the various constraints of the utility's environment. TA projects would (h) Subsequently to supervision missions, benefit if the consultant fees were tied to vigorous follow-up, when needed, is performance and involve clear authority important to assure implementation of and accountability. remedial measures. Resident missions can be helpful in these instances. (d) A cannot substitute for a dysfunctional institutional framework, including  PERFORMANCE AUDIT REPORT REPUBLIC OF GUINEA FIRST POWER PROJECT (CREDIT 1085-GUI) AND SECOND POWER ENGINEERING AND TECHNICAL ASSISTANCE PROJECT (CREDIT 1595-GUI) I. BACKGROUND The Context 1. The projects reviewed in the present audit benefitted the Revolutionary People's Republic of Guinea (until 1984), thereafter the Republic of Guinea. The country has an area of 246.000 km' and is mostly mountainous. The population is approximately six million, growing at a rate of around 3% P/A. 2. From the country's independence in 1958 until 1984 the GOG established a command economy with all productive activities being the exclusive responsibility of state owned public enterprises under highly centralized government control. After independence the economy entered a downward trend with foreign exchange shortage, poor productivity, declining investments, until 1973 when the government established the regime of economic enclaves for the exportation of bauxite. The economy picked up and the per capita GDP grew at a rate of 3.2% in real term to reach US$292 in 1979. At that time an overture was made by the GOG to obtain economic assistance from the countries and institutions of the western countries. Institutional Organization 3. IDA had, prior to the Power I project, some experience with the economic situation in Guinea through its involvement in the financing of the Bauxite enclave at Bok6 and with other projects oriented towards exports and generation of foreign exchange. The IDA financing in Guinea through the years was not regular but nevertheless significant. From 64.5 million in 1965/69 to 42.6 million in 197579. In April 1984 the Guinean Government, after a coup d'dtat, changed, and a new regime, favorable to the liberalization of the economy, came to power. In April 1987, SNE' was replaced by ENELGUI. In March 1989, the ENELGUI management was removed and the control of ENELGUI was given to a Board of directors. Finally, in September 1990 the GOG took a radical policy measure and gave the management of ENELGUI to a foreign management firm, with full authority and autonomy to streamline ENELGUI's structures and redress its financial situation. The above series of measures taken by the GOG were motivated by the failure of SNE/ENELGUI to produce operational and financial results in spite of a costly TA project financed by IDA and CIDA. At the time of project preparation (1978) the power sub-sector was in need of rehabilitation of the The national power utility. 2 existing infrastructures and a complete review of its institutional organization from ministry departments down to lower echelon of the power utility. The power sub-sector organization was a maze of conflicting authorities and overlapping responsibilities with no accountability. Power Sub-Sector Evolution and Trend 4. The Guinean power sub-sector had expanded, from the time of independence until late 1970s, by means of supplier credits to finance turn-key projects. After completion of the various projects the contractors left the country. The projects, in which the Guineans had not actively participated and were now in charge, were poorly maintained and operated. The chronic shortage of foreign exchange did not allow for the purchase of spare parts and equipment cannibalization was the way to obtain spare parts, a decline of the power system ensued. Frequent interruptions of service became chronic and load shedding was done routinely. 5. In 1977, at the time of the Power I Project's identification the Conakry power sub-sector was the exclusive responsibility of SNE (Soci6t6 Nationale d'tlectricit6), a department of MINEI. In February 1978, IDA made an advance of US$220,000 under a PPF to finance engineering studies to prioritize the activities for rehabilitating the SNE network. In early 1979 it became obvious that urgent repairs were necessary for the Grandes Chutes Dam, which was leaking and on the verge of collapse. IDA agreed to finance the repair under an emergency engineering credit (S-22GUI - January 12, 1979) for US$1.14 million (Engineering I project). In the mid 1979s, the precarious condition of the power system was assessed by IDA staff, assisted by a consulting engineering firm. The amount of funds for the rehabilitation of the SNE infrastructure was beyond the financing capability of IDA. Therefore IDA sounded out other donors, and KfW and CCCE agreed to provide the additional financing. Engineering II and TA (credit 1595-GUI) started disbursements, through a PPF, in August 1984. The project was identified by the GOG and IDA as a necessary step to complete the Power I project. In July 1983, during discussions with IDA, the GOG asked for a PPF and a second engineering credit to finance some of the activities of Power I, which were postponed under credit 1085-GUI due to cost overrun. Objectives of the Projects 6. The projects' objectives as described in the staff appraisal reports are summarized below: (a) Power I Project Objectives were to: improve reliability of service; finance part of the 1980-84 rehabilitation and expansion of Conakry distribution system; strengthen SNE by providing technical assistance or management, financial and technical matters; and (b) Engineering and TA Project Objectives were to: strengthen the power sector's organization and management; reduce dependence on imported by developing an expansion plan based on hydroelectric power; improve system efficiency and reliability; and provide technical assistance to the Natural Hydrology Service. 3 Description of the Projects 7. Power IProject (1980-86). The project was divided among the three donors: (i) rehabilitation of the Tombo diesel power plant with related electrical transmission infrastructures, was to be the responsibility of the KfW; (ii) rehabilitation of the Grandes Chutes power plant, a tariff study and the study of the hydro power potential of the Samou River was to be the responsibility of the CCCE; (iii) the rehabilitation, reinforcement and extension of Conakry H1 distribution network, the construction of a training center, a workshop and the supply of TA to improve the performance of the power sub-sector would be financed by IDA. At preparation time, IDA was referring to parallel rather than joint co-financing. A cross effectiveness clause between donors was originally refused by IDA. 8. Engineering II and TA Project (1985-1991). The project included supplementary financing from other donors. They were: (i) KfW, supplying some spare parts for the diesel power generating units at Tombo, and the rehabilitation of the hydro power plant at Danea; (ii) CCCE, for the TA and supply of spare parts for the hydro power plant at Grandes Chutes and for financing a feasibility study for the next hydro electric power project; (iii) CIDA financed part of IDA's TA project, the construction of a warehouse at Tombo, a feasibility study for a hydro electric power generation project in competition with the CCCE financed project, and a few studies of small hydro power sites; (iv) AGCD, financed some TA for the National Department of Hydraulic. The PCRs 9. The PCRs for Power I and Engineering II' audited here describe in detail the problems IDA confronted during projects implementation. The PCRs admit the projects were not successful since they did not achieve their objectives. The PCRs ask the reader to take the broader view and consider all the difficulties the two projects had to face. The PCR, Power I, did not include a Part II, therefore the view of the counterpart is not known. The PCR, Engineering II, did not originally include a Part II, later a copy was found and it is an excellent contribution giving much information missing in part I, specially in relation to project costs. The two PCRs, Part I, could have analyzed more precisely the main cause of the two projects' unsuccessful outcome. The PCR, Engineering II, does not supply data on project costs i.e., at budget time, at contract time nor the final costs. The audit tried to cover some of these aspects in more details. 6 Power I Credit 1085-GUI - PCR No. 7235 May 1988; Eng. II Credit 1595-GUI - PCR No. 11499 December 1992. 7 The Table of Content indicates that information on costs is given on page 10 and Project Costs for Credit Category on page 11. No information on costs is given on those pages or anywhere else in the PCR. The basic data sheets, which are supposed to quickly inform the reader on fundamental data such as disbursements, respect of implementation schedule or delays, are missing. 4 2. PREPARATION AND APPRAISAL OF THE PROJECTS 10. Project Preparation Outcome. Notwithstanding the difficult context of Guinea in the late 1970s, the audit believes that Power I's preparation suffered from significant weakness: (i) the institutional set-up of the power sub-sector was complex and inefficient, serious consideration should have been given to either reform it or postpone the project; (ii) the technical assistance component in advisory function without competent counterparts was a risk which was overlooked. (iii) the lack of operational autonomy of the SNE was not perceived as a critical factor for the project's implementation; (iv) the technical requirements of the project were not clearly identified; (v) the budget for the rehabilitation of the distribution network was insufficient to achieve the intended goal. 11. Projects' Objectives. The audit could not find evidence that the GOG and IDA developed a mutual understanding regarding a common policy, strategy and objective for the power sub-sector during project preparation. IDA's policy and strategy in relation to the Guinean power sub-sector are not clearly stated in the documents related to Power I and Engineering II projects. In fact, the ESMAP report (1986)', was the first document clearly suggesting a sectoral strategy for the energy sector and the power sub-sector. 12. Later this oversight became a source of confusion and misunderstanding. In March 1985, during a meeting between IDA and the GOG, IDA mentioned its objective was to strengthen SNE. Later, the GOG affirmed during a donors' meeting that its objective for the power sub-sector was to give the SNE the power generating capability to meet the demand of Conakry. The audit concludes that the objectives of IDA and GOG were not congruent. GOG's priority and immediate objective was to increase the power generation to satisfy the Conakry system demand, and after this had been achieved, reorganize the sub-sector' (PCR, Part II, page 28 and 30). 13. Institutional Reform. No institutional reform was intended by IDA during preparation and implementation of Power I nor during preparation of Engineering II. IDA underestimated the negative impact of the existing institutional set-up on the implementation of the rehabilitation program. In fact, the lack of institutional reform was the main cause of the Power I and Engineering II not reaching their objective and goals. Without reform of the institutional set-up, autonomy for SNE was not possible, and without operational autonomy the TA could not succeed. IDA energy department warned the Region during project preparation that they underestimated the following difficulties; (i) the institutional organization was complex; eight ministries shared the responsibilities and control of the sub-sector; (ii) the SNE was a government department (physically situated inside one of the Ministries buildings) without control even on its day-to-day operations; (iii) the supervision of new investments was done by the Ministry. IDA Region's answer was that its goal of promoting institutional reform would be done by first strengthening SNE. The rationale for this strategy is given in the SAR (page 6, paragraph 2.3). Guinea: Issues and Options in the Energy Sector. November 1986, No. 6137-GUI. The GOG tried to reorganize the sub-sector in 1982 but was stopped by IDA. 5 14. The Region recognized that the implementing agency for the project would be the Ministry (SAR page 13, paragraph 3.01) and stated that SNE answering directly to two ministries (OCOFI and MINEK) would not hamper the execution of the project. It appears that at the onset of the project the Region decided to support the utility (SNE) which had no authority on the sub-sector evolution and little operational authority. The risk was clearly identified, but underestimated. In deciding to use SNE as a channel for restructuring the sub-sector, the Region took a decision which had a profound impact on the project's outcome. 15. Nature and Scope of the TA Component. TA was identified as the project's priority by IDA, the GOG and the other donors for the success of the rehabilitation program. At appraisal time, IDA's management was concerned that the rehabilitation program would be managed by a power utility which could not even maintain its existing infrastructures. The Region believed that a massive TA component inside the SNE would be the answer to the lack of expertise and experience of local counterparts. 16. Two issues were discussed inside the Region in relation to the TA component: (i) the amount of the TA in relation to the absorptive capacity of SNE; (ii) should the consultant recruited for TA be in a hierarchical function with authority over SNE staff to be trained, or alternatively be in an advisory function under the authority of the staff to be trained? In May 1981, when the TOR for the TA were issued, the decision was made in favor of the advisory function approach. A consulting firm, specialized in TA and with experience in Africa was retained by IDA (October 24, 1980) to study in detail the needs of the TA for the GOG rehabilitation program. It recommended a reorganization of the sub-sector as a prerequisite to the TA. IDA rejected the suggestion and excused the firm on the grounds that it lacked experience on the subject. 17. The Region's position is clarified in the Decision Memorandum (February 13, 1980). A step- by-step approach to the sub-sectors reorganization was selected since a major overhaul of the sub- sector would be raising fundamental questions of ideology. However, neither the number of steps nor the timetable for their implementation were identified. The Region's decision to strengthen SNE with advisors was not accepted by IDA energy department, which expressed its concern in writing (November 27, 1979): (i) by supporting SNE in the present sub-sector set-up, IDA would achieve little more than improvement in operations and maintenance. (ii) suggest a degree of consolidation of the sub-sector and a plan for SNE's progress towards autonomy. It was recommended that the project should not be financed considering the size of the project and the nature of the components. The Region pleaded for a reconsideration of this recommendation for strategic reasons (March 25, 1980). 18. The GOG took the initiative of reorganizing the sub-sector and enacted a law to separate the generator from transmission and distribution functions. This was not implemented because IDA reacted vigorously to the Government's initiative with a threat of suspension of disbursement and obtained the status quo ante. In return the GOG extracted a pledge from IDA to finance a sub- sector reorganization study. The audit learned during the interviews in Conakry (January 1994) that two groups inside the GOG were almost equally divided on the need for institutional reform. The support of IDA for one group or the other was the decisive factor. 19. Regarding the issue of hierarchical versus advisory function for the TA, the project file does not clearly indicate IDA's position on this issue at preparation time until mid-1984 when, at a donors' meeting, the Ministry (after a change of regime and government) declared that the GOG had no 6 objection to the expatriates being in a hierarchial function with full responsibility. IDA reaction to this statement is spelled out in a memo (March 21, 1984): "Whereas in the past a Guinean had to be in charge, the GOG had recently agreed to expatriate responsibilities in the power and highway projects. But the substitution of IDA imposed discipline in the form of expatriate expertise and management under Guinean control and responsibility is a potentially dangerous move. If we accept the "island of control" concept, we run the risk, if GOG adopts the practice of handing over project execution to foreign management." 20. Rehabilitation of Conalay Distribution Network. At preparation time some important points were overlooked, and their impact on the projects' outcome was underestimated: (i) the lack of implementation capability of the Borrower; (ii) the scope of work and the budget needed to completely rehabilitate the LV network of Conakry; (iii) the technical coordination required between the three donors; (iv) the absence of foreign exchange to procure spare parts, tools, maintenance equipment; (v) the impact on SNE's financial viability by not installing meters to all customers. 21. At preparation time no technical coordination was made by the consulting engineer or IDA specialists to verify all the technical interfaces between the three projects (IDA, KfW, CCCE). Later this oversight had technical and indirect financial consequences for the following reasons: (i) a power network is more than the sum of its components. No precaution was taken to assure that all three projects would work smoothly and efficiently together. The network's protective relay system and its internal harmonization was overlooked; (ii) without coordination of work schedule the two firms doing the rehabilitation of the generating units at Tombo and at Grandes Chutes in parallel were bound to take large units out of service at overlapping times. For instance by taking two units of 5 MW out of service for months, the Conakry network would be deprived of a large percentage of its power supply. This happened and resulted in load shedding, since no reserves existed and no generation back-up had been planned. 22. During project preparation the critical problem of foreign exchange shortage was identified, but its negative impact on the project was overlooked (Memo, February 13, 1980, page 2). IDA was aware that the covenant requesting counterpart TA funds in foreign exchange was not realistic, nevertheless, no provision for spare parts was made. The impact of this oversight lasted to the end of the project. In early 1983, two years after the beginning of the project, when the TA team arrived they found no tools, no spare parts and no money to buy them; as a result the TA maintenance efforts were initially stymied. 23. At project preparation time IDA suggested the small contracts (lots) approach for project implementation with separation of engineering, procurement and construction contracts. This required a lot of time for management and supervision by the Borrower and the contractors. When all contracts were awarded, the Borrower was faced with cost overruns and delays in construction, resulting in reduction in scope of work on the most vital part of the project i.e. the rehabilitation of the LV network. 24. Tariff and Network Losses. During project appraisal the seriousness of the network losses was underestimated and the cause of the problem ignored. In fact, the SNE/ENELGUI dilapidated distribution network without proper connections and installed meters was a financial sieve. The losses, technical and non-technical, and their impact on the financial viability on SNE, were overlooked. Tariffs increases to obtain financial viability without first reducing the network losses to a reasonable level, will quickly reach the point of diminishing return. In late 1989 when the 7 operations of ENELGUI were given to a management firm, the losses were estimated at around 40%. Increasing tariffs in such conditions likely increases the rate of non-technical losses. The need for a complete rehabilitation of the LV distribution network to be able to connect meters and bill customers as well as suppress illegal connections was overlooked. The PCR, Power I, mentions (page 12): "The positive income projected during appraisal was in contrast to SNE's consistent deficit throughout 1985." In fact this situation lasted until the end of Engineering H in 1990. 25. Rationale and Preparation of Engineering H. In the course of the first project, some of the activities initially included in the project could not be financed due to large cost overruns. The missing activities were: (i) preparation of the Power II project; (ii) the study of the sub-sector's institutional reorganization; (iii) the Power Generation Expansion Master Plan. At that time it was evident that the Power I project costs had been underestimated in relation to the cost of the implementation approach suggested for: (i) the rehabilitation of the distribution network of Conakry II; (ii) the cost of giving the TA to a large foreign power utility instead of utilizing less costly alternatives. In November 1983 an IDA mission visited Guinea to prepare the Engineering II and TA project. In the meantime, the most urgent needs of the sub-sector were financed through an advance of US$1 million, made by using funds from a Petroleum Exploration credit (1438-GUI), and a PPF of US$720.000. They were used to finance; (i) the continuation of the TA project; (ii) the Power Expansion Master Plan; (iii) the sub-sector's institutional reorganization study. On April 18, 1985 after obtaining the necessary supplementary co-financing, the President's Report (PR) was presented to the Board for credit approval. IDA concluded that the project was not facing any special risks other than the lack of autonomy of SNE. The question of the TA consultants still being in an advisory function was not identified as a risk. In fact, IDA had just resisted a suggestion by the GOG of having them in a hierarchical function. 26. The Engineering II Project did not include any activities for increasing the power generation. The decision for the next power generation project was on hold until the completion a Power Expansion Master Plan, which was supposed to be realized during Power I, but due to lack of funds was completed only in 1986. The distribution network's complete rehabilitation was not identified at that time as a priority and was left to be done by the SNE, assisted by the TA. The PR mentions that the rehabilitation of the Conakry generation system with 47 MW hydro'e and 14 MW diesel would be sufficient to meet Conakry's power demand through the late 1980s. At that time Conakry suffered regular and extensive load shedding and poor quality service, in part due to lack of power generation. 3. IMPLEMENTATION AND OUTCOME OF THE PROJECTS Actual Project Costs 27. Power I Project. The audit obtained some information (field mission, January 1994) from the SNE/ENELGUI files to assess the variations between the original and the final costs of the main components of Power I. The audit concludes that the principal cause of the variations between the original and final costs are due to poor project preparation and weak counterpart management during 10 This installed capacity of hydro electric power is not firn power, since it depends on hydrology. 8 implementation. The consulting engineer was asked to deviate from the least-cost solution: (i) to change the network design for a more expensive one; (ii) the turnkey approach for implementation was changed to the small contracts/lots approach. The impact of these changes on the budget were not carefully assessed. Furthermore, SNE had the responsibility of the management of a project approach (small lots) in which it had no experience. The audit compared the budget in SDR (1980) with disbursements in SDR (1983) to find which components had cost overruns. Table 1: Power I Project Costs (in SDR) Budget Available Actual Change SDR 1980 SDR 1983 1. Conakry Distribution 6.591.228 8.875.088 9.675.000 9.6 % 2. TA Buildings 2.665.570 3.589.190 4.994.000 39.0 % 3. TA Services 3.513.706 4.731.205 4.600.000 2.8 % 4. Scholarships 783.516 1.055.004 887.000 -16.0 % 5. Refund of Advances 1.293.000 1.293.000 1.293.000 NA 6. Conakry Services 1.308.552 1.761.965 2.001.000 13.5 % 7. Power II preparation 589.656 794.548" - - 8. Sub-total (1) 16.745.228 (2) 23.450.000 9. Contingencies (32%) 5.354.772 - 10. Total 22.100.000 22.100.000 23.650.000 Power II preparation funds 794.548 22.855.452 Total contingencies, including Power II preparation funds, were SDR 5.9 million or 36% of budgeted costs. In May 1983, the cost overrun was SDR 7.3 million or 33% of the initial budget. The funds to make up the difference came from Power II preparation (SDR 589.656), a Petroleum Exploration credit of US$1 million and a PPF from Engineering II for US$720.000, and approximately US$1.2 million from CIDA to assist in the financing of the TA. The TA was about one year late in the field and the contract duration was cut, in spite of that the cost was 2.8% more than original budget plus contingencies. 28. Engineering II and TA Project. The budget of the complete project including the contribution of the other donors is given in the President's Report, page ii. The costs for the IDA project can be obtained in SDR from the Development Credit Agreement (page 13) and can be approximately compared to the information supplied by the PCR, part II, GOG's contribution. The PCR, Part I, does not supply any information on disbursements and fund allocations. 11 Was used to increase contingencies. 9 Table 2: Engineering I Project Cost - IDA Component (in SDR) Budgett2 Amendments Available Actual* Change 1. Studies. 540.000 900.000 1.183.680 2.370.400 + 100 % 2. TA 3.240.000 3.240.000 4.261.428 3.996.372 - 6 % 3. Training 1.260.000 0 0 0 4. Spare parts 950.000 950.000 1.249.440 154.577 - 87.6% 5. Dam repair (emergency) - 900.000 1.183.680 1.357.577 + 14.7% 6. Special account 110.000 110.000 144.672 95.817 - 33.8% 7. Refund PPF 720.000 377,100 377.100 374.100 - 8.0% 8. Sub-total 6.820.000 6.477.100 8.400.000 &351.454 - 6.0% 9. Contingencies 1.580.00 1.922.900 0 10. Total 8.400.000 8.400.000 8.400.000 8.351.454 The difference of SDR 48.545 was canceled on December 31, 1991. - The cost of the studies does not include a substantial contribution by CCCE and CIDA. - The cost of the TA does not include a substantial contribution by CIDA. - The cost of the spare parts (-87.6%) does not reflect a contribution by CIDA of approximately US$1 million. - Some studies were financed through a PPF of Power H. 29. One item not in the budget was the US$900.000 for the emergency repair of the Banda reservoir dam, which was in danger of collapsing if not quickly repaired. This disbursement was approved by the IDA management on November 21, 1986. To the studies' cost must be added approximately US$2.37 million from CIDA for the Fomi hydro site, and around US$5.6 million from the CCCE for the Garafiri feasibility study. The total cost of the studies is around US$11 million. For the TA, the CIDA contribution, directly and indirectly through IDA, must be added an amount of around US$3 million, for a total of around US$6.8 million. To sum up the disbursement situation for Engineering II & TA, the total cost is around US$25 million. Economic Rate of Return of Power I 30. The SAR and the PCR of the Power I project give an estimate of the EIRR based on the complete rehabilitation program. The SAR supplies a list of assumptions on which the results of the EIRR depend. The PCR does not list the assumptions, however the data in the calculation sheet allow the reader to assume some of them. 12 Credit Agreement, June 12, 1985. 13 Source: PCR, Part H. 10 31. The assumptions retained by the audit for the recalculation of the EIRR are the same as those in the SAR with the following differences: (i) The methodology consists in comparing the incremental benefits versus the incremental costs. The value of the benefits coming from the existing assets prior to 1980 are not credited to the project. (ii) The economic benefits of the project are derived by taking the willingness to by the customers, which is estimated at US 18C/kwh"'. (iii) The useful life of the project after completion (considering the quality of the maintenance) is 20 years. (iv) The cost of 0 & M for the incremental costs are not hypothetical,but the real costs of ENELGUI's 0 & M ($/1988) taken from the first year the accounting results could reasonably be relied upon. The costs (1981) for 0 & M before the project are taken from the SAR (page 65), and indexed till 1988 (+32%). (v) The investment costs are taken from the PCR and the project files. One million was added in 1985 to take into account CIDA's contribution to the TA. (vi) The technical losses retained by the SAR were 15% through the years. The audit has taken the current appreciation of ENELGUI for the losses. Until 1991 the total losses of the network were on average 45% of the generation. ENELGUI believes that 30% of the losses are for the technical losses and 15% are for non-technical losses. The audit takes 20% for the technical losses as being reasonable considering the state of the network. (vii) The audit takes the real power generation of ENELGUI from 1980 onward and not the optimistic forecast retained in the SAR. 32. Based on the above assumptions the recalculated EIRR is around -7%, but this is only an order of magnitude. The statistics of ENELGUI before 1988 are not reliable in relation to the exact amount of technical versus non-technical losses, ENELGUI knows only the total (1994). Furthermore, the 0 & M could not be audited until the late 1980s due to lack of data and book keeping accuracy. Power I Project Outcome 33. The audit rates the outcome of Power I as unsatisfactory. Power I did not meet its stated objectives for the following reasons: (i) in the Guinean political, social and institutional environment of the 1980s, the absorptive capacity of the Ministry/SNE for the TA was non-existent, irrespective of the approach chosen for the TA (advisory or hierarchical); (ii) placing the TA inside SNE in an advisory instead of a hierarchical function was a recipe for minimal results; (iii) without institutional reform as a prerequisite and operational autonomy of SNE, no TA could have succeeded in the The LRMC of generation for Bamako is estimated at 1 Ic/kWh (1992). The SAR for Power II (page 25, July 1992) estimates the LRMC for Conakry at the distribution level around 15C/kWh. In January 1993, ENELGUI estimated a similar value without tax. The Audit has retained 18C/kWh for the calculation of the benefits of the incremental generation as being approximately the customer's willingness to pay, only if ENELGUI could guarantee service quality and continuity. 11 Guinean environment of the 1980's. The TA project was poorly designed and its implementation was chaotic. In the opinion of the audit and some of SNE/ENELGUI officers still with the company, it should have been stopped as soon as it was evident that all indicators were unsatisfactory and no positive results could be expected without institutional reform and a sub-sector reorganization. 34. Rehabilitation of the Distribution Network. The audit considers, after review of available documents and a field mission, that the rehabilitation of the distribution network was marginally unsatisfactory and the sustainability uncertain. 35. At the time of the project's preparation the Conakry I MV distribution network (City center) had just been rehabilitated, but not the LV system. IDA concentrated its attention on Conakry H MV/LV distribution network. The audit concludes that the outcome of the Conakry II network rehabilitation was unsatisfactory on the account of not reaching its physical goal of completely rehabilitating the LV network, cost overruns and having a doubtful sustainability of benefits, since little maintenance was possible due to lack of spare parts. 36. As a result of the reduction in scope of work on some construction lots, some procurement lots, due to poor assessment and coordination, did not match the need of other lots. For the rehabilitation of the MV overhead distribution line, the contractor decided not to rehabilitate, but demolish the existing network and build a new one; as a result some MV equipment was missing to complete the job. The contractor responsible for the rehabilitation and extension of the LV distribution network was obliged to reduce the scope of work for new connections, the surplus of meters (procurement lot based on the original plan) were installed on the old network, within months those meters had disappeared. The original goal of eliminating illegal connections was not achieved, the census and metering of all LV consumers was not done. In 1990 the consulting engineers doing the Master Plan for the complete rehabilitation of distribution network for Conakry (PPF Power II), reviewed the work done during Power I and concluded: (i) the overhead MV network is generally in good condition; however, it is operated at 15 kV instead of 20 kV due to some deficiencies in equipment rating; (ii) the underground MV system is weak and some parts are not in operation due to breakdown; (iii) most of the MV/LV sub-stations are in a sorry state and need to be completely rehabilitated. (iv) The LV network is in a terrible state and in need of complete rehabilitation. 37. After reviewing the situation during a field mission (January 1994) the audit can only confirm the above-mentioned information. For instance the audit learned that 80% of the MV/LV transformers originally installed had burned out by 1992 due to overloading and have been replaced by bigger transformers which in turn are getting overloaded, since ENELGUI does not have the money to buy the equipment to split the overloaded circuits or install switching equipment and adequate protection to match the load growth and the fault level. 38. TA Services Component. The outcome of this component is very unsatisfactory and its few results i.e., training of ENELGUI staff, unsustainable because the staff once trained either left ENELGUI or was reassigned to unrelated positions. The three stages of a technical assistance generally are: (i) trainees with proper academic education learn their specialty by working under the instruction of a specialist with experience; (ii) later, when ability has been demonstrated, the trainees are gradually given control of some of the operations under supervision of the senior specialist; (iii) when stage two is satisfactory, the trainees are given full responsibility with advice from the senior specialist. IDA recommended during preparation, after reviewing the state of SNE's counterparts and organization, a TA program based on the third stage of the above-mentioned training cycle. 39. An IDA Mission Report (March 12, 1985) mentions: "The TA team has performed as well as could be expected, considering they came with nothing, no tools,no spare parts, no instruments, 12 nothing at all, but only money to pay their salaries...." "At appraisal time it was deliberately avoided to give them any authority. This approach was wrong on two aspects: (i) they could not obtain decisions on their recommendations. (ii) TA performance was not possible without first tackling the institutional set-up of the SNE". The same report concludes: "Nevertheless, the TA presence could be justified, since they identified the source of the disappearance of fuel-oil at Tombo." 40. TA's Physical Component. The outcome of this component is unsatisfactory, even if the buildings are still in use. Other than supplying the technical services, the TA program included three main components: (i) the construction of housing for the expatriates; (ii) construction of a training center and an instrument calibration laboratory. This comprehensive package required coordination of all the components to assure smooth implementation and respect of the timetable. The expatriate housing was identified as being on the critical path and a prerequisite to the arrival of the TA team. The implementation timetable was allowed to slip by more than twenty months. When the training center and the workshop buildings were completed, the project was already short of funds and the needed instruments and equipment could not be purchased as intended. Engineering II Project Outcome 41. The outcome of the Engineering II project is unsatisfactory. The continuation of the TA service component without changing the approach resulted in the same unsuccessful result. However, at the very end of the project IDA obtained the autonomy of ENELGUI, as well as the replacement of the TA's contractor after two consecutive unsuccessful mandates. The various engineering studies, even if they were of good quality, had limited relevance: they were directed towards the long term generation alternatives and distracted attention from short term generation problems. 42. IDA was convinced during Engineering II that the Conakry energy demand could still be met, if all the generating units at Tombo could be properly maintained. However, with an estimated power deficit of around 27 MW at the end of the project, the situation was critical and a political embarrassment for the GOG. Furthermore, at the end of the project the network losses (technical and non-technical) were around 40% of the generation. The distribution system was completely defective, interruptions of services due to equipment breakdown and load shedding was a daily programmed occurrence. At the end of 1990, after more than ten years of IDA assistance, the situation was not much different than at the start of the project At that time a memo from IDA sums up the situation in gloomy terms: (i) Power I deemed a failure; (ii) the TA component (1978- 89) is a failure; (iii) pricing, billing and collection is a failure. (iv) Existing generating plants are in poor condition in spite of ten years of donor efforts. The audit has nothing to add to the above statement, which sums up the situation until 1990 i.e., at the end of the Engineering II and TA project. 43. The TA Component. After visiting the ENELGUI infrastructures in January 1994 the audit concludes that in spite of a comprehensive and costly TA project, SNEENELGUI never achieved even a modest degree of operational efficiency. Frequent power interruptions due to chronic break down of equipment, followed by load shedding, were and still are a serious and daily recurrent problem. Under the new management (since 1990) some improvements are noticeable in relation to financial control, administrative and accounting procedures, but without more generation and a rehabilitated network, they cannot do much more. The audit concludes that the first goal i.e., efficient operations and maintenance, was not achieved during the Engineering I project". In fact, 15 Part II, Borrower's Perspective - Some performance indicators are given on page 23. See conclusions, on page 30: The goal of the operational efficiency was never achieved in spite of the TA. 13 the TA, as originally designed, was not sustainable, and its continuation was only justified on the basis that without it the situation would have been worse, which is debatable. 44. The Engineering Studies. The project's second goal was to carry out, by the end of the 1980s, a major expansion of the power generation based on hydro-electric power. After the Power Expansion Master Plan (1986)" and the other studies were completed in 1992, it became clear that the least-cost alternative for the short/medium term for power generation was to increase the (diesel) generation at Tombo. The PCR, part II (page 26) mentions that over US$7 million17 were "wasted" studying the hydro electric power potential of various sites. The audit does not support this statement. The various studies supplied the necessary data to identify the least-cost power plant alternative, and allow a Ranking Study of all alternatives. In the Ranking Study the net present value analysis of the diesel alternative versus the best hydro power plant alternative indicates a saving of around US$60 million, (provided the oil price stays within the Bank's projection). The saving between the two best alternatives is six times the cost of all the studies. However, the above quick analysis does not take into account investments' timing: the economic cost of ignoring the option of increasing thermal generation at Tombo in the early 1980s has been substantial. The private sector had to install 70 MW of generation, not counting the suppressed demand by those who could not afford to buy generating units. The audit believes the studies were not "wasted"; it was an option (short term) that was wasted by looking after the long term alternative. Strengthening of the SNE - Power I and Engineering II 45. The outcome of both projects with respect to this objective is very unsatisfactory. The goal of the TA program was the strengthening of the managerial and technical capabilities of SNE, and gradually obtaining full autonomy of its operations from government control. No specific performance indicators or timetable were fixed at the beginning of the project, nevertheless the available operational statistics indicate that the results of this program are very unsatisfactory, since the objectives of this program were not even partly met. One point epitomizes how ineffective the TA team was in relation to the strengthening of the SNE. When the TA advisors discovered that the theft of oil"' and electricity was organized with the knowledge of SNE management and some government officers, they were told to look the other way, since their mandate was strictly to advise and not to take action. 46. The purpose and objective of a power utility is to generate reliable electricity to supply the demand, determined by tariffs set at the economic efficiency level (LRMC) or fixed by the competition. From 1980 until 1990 SNE/ENELGUI failed to achieve this fundamental objective, in spite of an important physical investment program and ten years of technical assistance. 47. At the end of the Power I project, after the rehabilitation of the hydro plants of the Samou river system and Tombo I, the power capacity of ENELGUI was on average around 41 to 43 MW. 16 Confirmed by a Ranking study in January 1992. 17 In fact, it is more likely to be around US$ 11.6 million for all the studies. Is Estimated at around US$ 7.5 million in 1983 and US$ 2.5 million for the first four months of 1984. Furthermore, SNE failed to bill US$ 5.6 million of electricity in 1983. 14 Systematic load shedding of around 20 MW was necessary, since the demand at that time was estimated at around 60 MW. The above chronic power shortage forced customers to install their own private power capacity which in 1988 was estimated at 53 MW with a energy generation of 63 GWh per year. In 1993 ENELGUI estimated the private power capacity at 70 MW with an energy generation of 109 GWh per year. The private generating capacity in Conakry, not at all connected to the network for reasons of cost and quality of services, was estimated at 11 MW with an energy generation of 30 GWh per year. The above figures demonstrate the failure of ENELGUI to attain its most fundamental objective. The private sector was forced to install a power generating capacity, at their own cost, almost equal to ENELGUI. 4. BROAD ISSUES AND FINDINGS 48. The issues and findings emerging from the two projects can be regrouped in relation to the level at which they had an impact on the sub-sector Institutional Reforms 49. In spite of advice from IDA's management and the energy department, the Region did not address the institutional reform issue until mid-course of Engineering II, and at appraisal time underestimated the impact of this oversight on the two projects' outcome. Without the adequate institutional set-up and sub-sector organization to establish responsibility and accountability for each entity, the outcome of a large capital investment program with many donors, like in Power I, was highly risky. Donors Coordination 50. A comprehensive investment program, financed by many donors over a number of years, require coordination. The bilateral donors find it difficulty to play this role; IDA as a multilateral institution, is generally accepted by all parties to fill in the function of program coordinator. During Power I, other donors repetitively asked IDA to take the lead for the coordination of the sub-sector activities. Gradually during Engineering II, IDA played a leading role in organizing sub-sector round tables and donor meetings, the final decision by the GOG to reform the sub-sector institutional set-up was certainly due to the concerted efforts of the donors. At the technical level more meetings of specialists to sort out the technical interfaces, which for a long time were left unattended, were needed. Donor meetings to achieve commonality of objective and support of each other project goals could improve the success of a co-financed program and promote a general sense of ownership by all participants. The result of IDA's increased effort of coordination with other donors was instrumental in stopping a third TA contract from being signed, when another donor was still willing to support the experiment. Without IDA's good coordination the preparation of the Power II project" would have been difficult to accomplish. Project Preparation 51. The unsatisfactory outcome of the two projects could be identified with some of the decisions taken during project preparation, such as: (i) the issue of reform of the sub-sector institutional set-up which was postponed to future credits; (ii) the poor project design and weak preparation of the t Credit 2416-GUI Febmuary 1993. 15 project components resulting from inadequate skill mix and poor assessment of the difficult situation of Guinea; (iii) the optimistic belief the TA in advisory function could solve all problems: technical, operational managerial, financial etc...; (iv) the absence of congruence of objective between IDA and the GOG, which resulted in divergence of strategy for the power expansion program. The audit concludes that the project preparation was unsatisfactory and the principal cause of failure of both projects. Project Ownership 52. The audit believes the issue of project ownership was neither significant during implementation of the Power I project nor the cause of the project's outcome. During the audit mission in Guinea (January 1994) the issue was discussed with the GOG/ENELGUI officers involved in this project. During Power I's preparation and implementation the Guinean counterparts were conscious that it was their project and even today accept responsibility to a large extent for its unsatisfactory outcome, albeit not completely, some believe IDA could have been more supportive and firm. The Guinean power sub-sector environment of the late 1970s was simply not prepared to absorb such a multi- faceted program. 53. For Engineering II, the ownership question cannot be answered without nuances. The precise role of IDA in relation to institutional reform and autonomy of SNE is not clear. IDA stopped the institutional reform attempt by the GOG (1982), and tried to strengthen a state owned power utility then asked for complete autonomy of the utility and institutional reform as a condition for supporting the Garafiri hydro plant project. The Guinean counterparts were not sure that Engineering II was their project. Divergent Views on System Expansion 54. In relation to power generation expansion, the PR states that the medium and long term strategy aims at taking maximum advantage of the country's hydro electric potential, while maintaining the existing thermal plants as stand-by capacity (PR page 10). This strategy, formulated by IDA, epitomizes the fundamental difference between IDA and the GOG regarding the issue of the power generation expansion program. 55. Before the Power Expansion Master Plan results were presented (1986), IDA staff prepared an indicative development plan, which consisted of replacing the diesel generation at Tombo by a new hydro power plant. Since the study and the construction of a medium sized hydro plant probably would require up to eight years for preliminary and detailed studies as well as construction time, the IDA document does not explain how the demand of Conakry, growing at 7.5% per year, could be met. This fact was recognized by the Power Expansion Master Plan (1986): the solution to the power deficit was to add quickly some thermal generation at Tombo. This suggestion was ignored by IDA. No decision to increase the generation was taken before the end of the Engineering II project. IDA Region never reviewed its original assumptions until the Ranking Study of 1992. The GOG, while still interested in the hydro plant alternative, was obliged to turn to Italy and Japan for financing the needed thermal power generation (part II of the PCR) in violation of one of the conditions in the covenant. The GOG signed contracts with three firms to build without consulting 16 IDA': (i) In 1985, a 5 MW hydro generating plant at Ban6a; (ii) In 1988, a 5 MW diesel unit with JICA financing; (iii) In 1990, 10 MW generating units (diesel) with Italian financing. This additional power generation capacity, decided by the GOG, was aiming at reducing the growing power deficit of Conakry. Meeting the demand was the clearly stated objective of the GOG since the onset of the rehabilitation program. 56. The PCR, part H mentions in relation to the costly studies for the hydro power sites, that IDA: (i) identified from many sites the most suitable hydro power site (Garafiri); (ii) recommended a study to feasibility grade including bidding documents; (iii) suggested, at a donors' meeting (December 1989) to go ahead with the project; (iv) then in late 1990 withdrew its support. The GOG decided (December 1989) to give full autonomy to ENELGUI with managerial responsibility to a private firm, when they became aware that IDA, supported by the other donors, would not finance the Garafiri project without major improvements of ENELGUI's performance. The Audti has ascertained that originally IDA stated that Garafiri was a good choice for the next major investment in the power sub- sector and recommended the completion of detailed engineering and preparation of bidding documents to proceed expeditiously. Around November 1990, after a study of all power alternatives carried out by consultants financed by KfW, IDA Region withdrew its support for the Garafiri project, since it was clearly proven that the project was not the least-cost alternative in the short and medium run compared to the thermal generation as indicated in the original Master Plan (1986) and confirmed by the Ranking Study (1992). This sudden change of mind by IDA was not appreciated by the GOG and the CCCE: both were now in a delicate position, the CCCE having financed the Garafiri studies up to the bidding stage (US$5.6 million). The GOG after having spent approximately US$11 million on studies for long term alternatives and reorganized the institutional power sub-sector in exchange for the Garafiri project, was disillusioned and questioned IDA's motive. The Audit found out that GOG is now also convinced that Garafiri project is an economic option in the long run power expansion plan. Dialogue Between IDA and the GOG 57. The issue of a constructive dialogue between Borrower and IDA is not specific to the two audited projects, but in the case of the Power I and Engineering II, it played a significant role during preparation and implementation. The PCR, Engineering II, Part I mentions that the dialogue between IDA and GOG was not easy. This opinion is shared by the Guinean counterparts. Notwithstanding the original divergence of objective on the sub-sector's priorities, it is difficult to comprehend that from 1980 until early 1990s IDA and the GOG could not reconcile their diverging views on the sub-sector's orientation through a frank dialogue and the support of energy specialists with good communication skills. This divergence of opinions strained relations between GOG and IDA and opened the door to more differences of view with other donors."' Financial Viability and Tariff Increases 58. To the audit the question of tariff increase, which was during the two projects a cause of much debate and repeated violations of the covenant, was too focussed on achieving financial criteria 0 In violation of covenant 3.04 (b). 21 PCR, Engineering II, page 9, paragraph 10. 17 through tariff increases. In the case of a utility such as SNE with a dilapidated distribution network and technical and non-technical losses at around 40% of gross generation, tariff increases are not the only way to assure financial viability. A set of measures could have been implemented to look at all the possibilities of decreasing the losses thereby increasing the revenues. Reduction of losses is cost effective and if done jointly with tariff increases could achieve the expected financial criteria performance. 59. The tariff policy and related financial issues were laid down in the Credit Agreements': in short, the effectiveness of the credit depended on the Ministry establishing an average kWh tariff to the equivalent of US 18C. Since the Guinean Franc was suffering from continuing devaluation, the result of the US 180/kWh clause was, if applied, tantamount to a substantial yearly if not quarterly increase of tariff in local currency. The GOG hesitated to proceed with the tariff increase condition, and the Engineering II credit effectiveness was delayed by a year. 60. The financial viability of the SNE depended on three main factors: (i) an efficient distribution network, well maintained with all customers properly connected and metered; (ii) an efficient customer management system for the reading of meters, prompt billing and efficient collection, disconnection of non-paying customers, prevention of illegal connections; (iii) adequate tariffs, set at the economic efficient level (LRMC). (In the case of a small utility like ENELGUI with around 80 MW installed capacity, tariff based on the LRMC of electricity will need to be mitigated since the LRMC will be less than the average cost for some time to come). IDA insisted on solution (iii) that is tariff increase, and hoped, until 1990, that the TA would succeed with (i) and (ii). However, without (i) as a prerequisite (ii) and (iii) could not be achieved until the end of the second project, the losses of the Conakry system were around 40% of generation, which represent million of dollars of potential revenue. IBRD and IMF Policies' Coordination 61. During Engineering II project preparation, IDA insisted that ENELGUI should be able to obtain fuel-oil for its diesel generating units at a reasonable price. At that time, the GOG had a tendency to raise the tax on fuel-oil to increase its revenues. ENELGUI, being one of the largest consumers of fuel-oil, was using its scarce cash flow to pay large sums of money up-front to the Government, even before generating one kWh of electricity. The Government, the largest client of ENELGUI, was not well known to pay its electricity bills promptly and showing the example to the rest of the population. The above situation resulted in a growing cash-flow problems for ENELGUI. To assist ENELGUI, IDA put a condition (No. 4.07) in the Credit Agreement to limit the cost of fuel-oil at the CIF cost, plus handling charge (around 10%). Later, the IMF insisted on a surtax on fuel-oil without any exception, forcing the GOG to contravene the IDA covenant. The result of the above IMF condition placed ENELGUI in a difficult position. The cost of kWh (July 1991) generated with fuel-oil at Tombo I was 139 GF (20c/kWh) compared to the average tariff of 70 GF (10C/kWh)'. To redress its treasury situation and stay financially sound, ENELGUI had no Credit Agreement CR 1595-GUI, June 12, 1985, Article IV, Section 4.05c and 6.01. Lower than the specified 18C/kWh average in the covenant. Tariffs increased by 200% in 1992. 18 alternative but to shut down the thermal generation at Tombo 14, resulting in an increase in load shedding for Conakry with severe economic consequences. This "financial" load shedding allowed ENELGUI for the first time to obtain a net benefit for the last quarter of 1991 (PCR, Part II, page 24). The issue of coordination of economic and fiscal policies with impact on the power sub-sector is beyond the scope of this audit and ought to be reviewed and discussed at another level. Projects Supervision and Follow-up 62. The audit concludes that overall the supervision of the two projects were insufficient. For Power I the PCR mentions that supervision was inadequate in relation to frequency, duration and staff continuity. The audit concurs and remarks that the follow-up was probably the weakest point during Power I and the beginning of Engineering II. In general the supervision reports are clear on the magnitude of the implementation problems, but the follow-up by IDA was not commensurate to the gravity of the situation. Supervision mission reports, however informative they are, cannot change the situation unless detailed remedial activities are agreed upon by both parties and implemented. The monitoring of the follow-up activities may require more man-power in the field than IDA can provide particularly in those cases where IDA does not have a resident mission. IDA may consider sub-contracting the project monitoring to consultants. 5. CONCLUSIONS AND LESSONS Rating of Power I Project 63. The audit rates the overall outcome of this project as unsatisfactory, its institutional development impact as negligible and the sustainability of the benefits as unlikely. However, the audit makes a distinction in his assessment among the principal components: the TA project's buildings; TA services; and, the Conakry II distribution network. 64. The TA Buildings. The audit concludes the outcome of the TA physical facilities was unsatisfactory for the following reasons: (i) they were two years late and resulted in the postponement of the TA team's arrival, which in turn had a negative impact on the complete program (CCCE, KfW, IDA); (ii) it suffered from severe cost overruns, thereby absorbing a large amount of the physical and price contingencies of the project. The resulting lack of funds forced IDA to reduce the scope of work for the distribution network with a subsequent negative impact on SNE's financial performance. The buildings are still usable but in need of restoration and re-equipment. 65. The TA services. The audit concludes that the institutional development impact of the TA services component, was very unsatisfactory for the following reasons: (i) in the Guinean context of the late 1970s the relevance of the TA component as designed was of doubtful efficacy; (ii) the TA services never achieved it original goal, i.e., the strengthening of the SNE/ENELGUI skill base and management; (iii) while TA efficacy cannot be easily measured its impact on the SNE/ENELGUI operations during Power I was negligible. Tombo I was using fuel-oil, whereas Tombo U (Italian units) was using heavy residual fuel, not subjected to the surtax. Distortions could have been avoided by taxing the kWh sold, irrespective of the source of primauy energy. 19 66. The Conalay Distribution Network. The audit concludes the result of this component was marginally unsatisfactory. The goal was to distribute the power efficiently and reliably to the existing and new customers of Conakry. The MV network was realized, but not as originally intended. The new customer connections were reduced by approximately 20%. The illegal connections were not suppressed as intended, meters were not installed for all customers, a survey and a register of all connections were not made. At the end of the project the amount of technical and non-technical losses could not be separately identified, but the total was around 40%. The customer billing could not be done adequately, since the network was not sufficiently rehabilitated to install meters. Overall, the audit concludes that the sustainability of the benefits were unlikely. Ratings of Engineering II and TA Projects 67. The audit rates the overall outcome of the Engineering II as unsatisfactory; its institutional development impact as modest and the sustainability of the benefits as unlikely. It did not noticeably improve the sub-sector's performance until the very end. It was the outcome of this project that convinced the Government of Guinea to take drastic steps in the institutional set up of the sector which led to the actual privatization on the sector's operational entity under the IDA Power II Project (Cr. 2416-GUI). 68. TA Services. The audit concludes that the TA outcome of Engineering II (second contract) was very unsatisfactory. After ten years and around US$15 million spent on the TA no meaningful results in SNE/ENELGUI's operations or financial performance are evident i.e., 40% energy losses, negative operational income, more load shedding etc... The late improvements, resulting from the third TA contract with a new firm, arrived too late to change the situation. 69. Major Erpansion of Hydro Energy Resources. The audit concludes that the outcome of the engineering studies is marginally unsatisfactory because of their limited relevance or poor timing. The costly feasibility studies (1985-91) never succeeded in justifying the hydro energy resources as the least-cost alternative to the thermal diesel generation which was identified as early as 1986 as the short term solution to the Conakry power generation problem. IDA and Borrower Performance 70. IDA's Performance. The overall IDA performance for the two projects vary from very unsatisfactory for Power I to unsatisfactory for Engineering II. The audit differentiates between two periods, (i) 1980 till 1986 and, ii) 1986 till 1992. Until 1986 the supervision mission reports mention serious problems and the projects' ratings were generally poor. The sub-sector problems were discussed with the Ministry, but little remedial actions followed. The Ministry did not respect many of the conditions in the covenant without firm action from IDA. After 1986 the IDA supervision mission reports are more to the point, especially in relation to the deteriorating situation with the TA and other aspects of the project. These reports suggested stronger measures to redress the situation, plans of action were negotiated with the Ministry and the follow-up was more dynamic, with pressure to obtain changes, culminating in suspension of the disbursements in 1987. IDA's sudden pro-active management and the frank discussions of the sub-sector problems with the GOG finally resulted in positive changes. The third renewal of the TA contract with the same firm was not signed, due to dynamic representation from IDA. ENELGUI was given full autonomy (1989) with some TA specialists in hierarchical function. Then in September 1990 the TA contract was transformed by the 20 GOG into a management contract, given to a foreign management firm, after competitive international call for proposals. 71. The PCR rates IDA's performance for the Engineering I project as satisfactory. The audit disagrees, especially with the handling of the study for the Power Expansion Master Plan (1986). The Ministry decided to give the Master Plan to the executing agency responsible for the TA project. IDA informed the Ministry that its choice was inappropriate due to a risk of conflict of interests. In fact, the other donors complained. In the knowledge that the arrangement violated OMS 2.50, IDA let the Ministry's decision stand. Later the results of the Master Plan were challenged by the other donors and considered disappointing by IDA. 72. Borrower Performance. The audit concludes that the performance of the Borrower, GOG/SNE/ENELGUI, was very unsatisfactory: Power I's implementation suffered from delays, cost overruns and reduction in scope of work. The TA was not used effectively, the transfer of technology was negligible. The generation losses were around 40% until the end of Engineering II, load shedding was constant during the two projects, billing and collection only improved at the end of the second project. The PCR, part I mentions that the Engineering I project left the sub-sector in the same dismal state as it was at the start of the project, the indicators support this statement". At the end of Engineering II the improvement in ENELGUI's administration and book keeping did not compensate for ten years of substantial expenditures with little to show, especially in relation to the Conakry dilapidated distribution network. Lessons Learned 73. The outcome of the audited projects suggests lessons which are general in nature and applicable to other countries. (a) When the sub-sector institutional framework does not allow clear, separate responsibility and accountability of each entity involved in the project, and when the executive agency does not have the absorptive capability for a large capital investment project, donors might consider it wise as a first step to initiate a power sub-sector reform and delay investment, until appropriate sectorial and institutional reforms have been achieved. (b) Technical assistance even with managerial authority, cannot financially redress a utility only through repeated tariff increases. Required Tariff increases quickly become prohibitive, when the distribution network is dilapidated and the technical and non-technical losses are many times over accepted norms. (c) Technical assistance needs to be planned by specialists with experience in each discipline and with a good appreciation of the various constraints of the utility's environment. TA projects would benefit if the consultant fees were tied performance and involve clear authority and accountability. Number of interruptions and load shedding. Number of clients billed, collection etc... Operating income was negative until 1992. 21 (d) TA cannot substitute for a dysfunctional institutional framework, including Government interference in operations, or no guarantee of full independence, authority and accountability. (e) The conditions included in a credit agreement need to be realistic, and take into account the local economic and social environment and, barring a case of force majeure, must be enforced. Too stringent conditions might result in delay in credit effectiveness, or be waived during project implementation, signalling to the Borrower that if he waits long enough before implementing the conditions, in the end IDA will be lenient. (f) Donors' coordination, whether done by the Borrower and/or IDA, is important for the success of a multi-donors financed program. Effective donors' coordination requires from the coordinator excellent communication and interpersonal skills as well considerable sectorial experience to achieve the credibility needed for efficient coordination. (g) Infringement of Bank guidelines for procurement of professional services or laxity with conflict of interest should not be tolerated. Otherwise it is conducive to suspicion and discord between consultants, donors and Borrower, with negative impacts on the project. The results of the studies done by consultants with obvious conflict of interest are generally contested and not credible. (h) Subsequently to supervision missions, vigorous follow-up, when needed, is important to assure implementation of remedial measures. Resident missions can be helpful in these instances. 22 Annex I " Page 1 of 2 COMMENTS FROM THE BORROWER FACSIMILE No. 001 202 475 38575 Date: 13 June 1994 To: Mr. Y. Albouy, OED, Destination: Washington, D. C. From: Ataou Bah/Darp/ENELGUI Conakry Fax: (224) 41.18.33 Messafe: Dear Sir, Thank you for sending to me the PAR for project 1085-GUI. I regret not having received it sooner, however, I feel obliged to let you know our comments. Paragraph 2, page iii indicates an undisbursed amount of SDR 48,545 while pagraph 1, page ix indicates that both credits were fully disbursed. Overall, the PAR is very detailed and identify the real causes of the unsuccessful outcome of the project, which I think they are: Lack of Autonomy The Project Unit never enjoyed a real autonomy, it was always wielded by the power of the Ministry in charge. Failure of the Technical Assistance Evident weakness of the technical assistance, aside of the internal struggles, the technical assistance limited its role to being simple advisers, regardless of the end results. Your example presented in page 15, paragraph 45 illustrates this situation very well. Failure of the Lender IDA lacked firmness vis-a-vis the Borrower, perhaps to avoid diplomatic conflicts, exception made of the withdrawal of the 1983 decree which would have creates 33 electric companies to replace SANE. It should be pointed out that at that time IDA did not have a Resident Representative. Thus the supervision missions visiting Conakry did not have the means nor the time to verify the accuracy of the information provided to them. 2 Translation from the letter received in French. 23 Annex I Page 2 of 2 Failure of the Borrower It was attempted to redress an energy problem in a hostile environment which was left unchanged. It was equivalent to trying to desalt a glass of sea water by pouring in it a drop of clean water. The Borrower did not deem useful to rely on its local managers and did not deploy any efforts to motivate them - they worked with expatriates whose total monthly salary was enough to pay all the workers of the SANE at that time. In short, the formulae "success = motivation x competence x health" was completely ignored by the Borrower. I hope to have responded to your request and remain Very truly yours, Ataoullahi Bah Director Procurement and Physical Resources Tel. (224) 44.18.89 Ext. 10 Fax. (224) 41.18.53 24 ANNEX II COMENTS FROM KfW Page 1 of 1 KfW Kreditanstalt fOr Wiedehaufbau page Telefax of 22.06.94 TO: FROM: Mr. Yves Albouy our FAX No. +49-69-7431-2944 Chief, Infrastructure & Energy Division officer-in-ch.: M. von Alvensleben Operations Evaluation Department ref. No. Alv/enelwb.694 IBRD, Washington DC telephone, +49-69-7431-3198 date: 22.06.94 Recipient's FAX No.: 001-202-522-3125. L 11 c/2 - Financial Cooperation with Guinea First & Second Power Project Draft Performance Audit Report Dear Sir, We refer to your request for comments on the above document dated June 15, 1994 and received by us on June 17, 1994. In view of the rather short deadline (5 days) we did not have time to elaborate the draft in all details. Nonetheless we fully support the views expressed in the report. We agree with the general conclusion that Power I and Power II (which were both cofinanced by the German side) failed to reach its objectives to substantially reform SNE/ENELGUI and to redress the financial situation of the former state monopoly agency. The report gives an accurate description of the reasons for project failure, namely the difficult political and economic environment, inconsistencies in the joint formulation of a sectoral policy between donors and the GOG and the lack of performance criteria tied to the technical assistance programme. KfW has also always supported the view that investments should be delayed until appropriate sectorial and institutional reforms have been achieved. Yours Sincerely KREDI ANSTALT FUR WIE ERAUFBAU IF YOU EIPERIENCE ANY PROBLEWSIN RECEIVING THIS TR4 MSMXfISSION PLEA SE PHOrE THE VuMBER LISTED ABOVE