Ghana - Power Projects

Docomnm of The World Bank FOR OFFICIAL USE ONLY Report No. 5731 PROJECT PERFORMANCE AUDIT REPORT GRANA: KPONG HYDROELECTRIC AND THE THIRD POWER PROJECT (LOAN 1380-GH, 1381-GH AND CREDIT 689-GH) June 25, 1985 Operations Evaluation Department Thik dcamemt ha. a resbitd ditible ad may be used by recipluls emly in the performance Of their egMa dglles It commisb ma aet tewrwise be discleoed without Worl Bank autherlaton. ABBREVIATIONS AND ACRONYMS ABEDA = Arab Bank for Development in Africa CEB = Communaute Electrique du Benin CIDA = Canadian International Development Agency ECG = Electricity Corporation of Ghana EDF = European Development Fund EECI = Energie Electrique de la CZte d'Ivoire EIB= European Investment Bank KfW = Kreditans talt fiir Wiederaufbau (FAG) MFP = Ministry of Fuel and Power VALCO Volta Aluminium Company VRA = Volta River Authority FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT GHANA: KPONG HYDEOELECTRIC AND THE THIRD POWER PROJECTS (LOANS 1380-GM, 1381-GH AND CREDIT 689-Ga) TABLE OF CONTENTS Page No. Preface ..................................... i Basic Data Sheet ................................ ...................... iv Highlights ......................................................... vii PROJECT PERFORMANCE AUDIT MEMORANDUK I. PROJECT SUMMARY ..... . ....... ......................... 1 II. SUPPLEMENTARY COMMENTS ................................... 4 Electricity Tariffs .................................. 6 Project Justification ....................... ....... 8 Rehabilitation and Reform Measures ..................... 11 Cofinancing Experience ................................. 14 Resettlement Experience ................................ 14 Sustainability of Sector Institutions and Project Facilities ........................................... 15 III. CONCLUSIONS .......... ............................15 Annexes: Tables 1-3 ............................ 19 Appendix I: Comments from VRA, ECG and the cofinanciers .......... 22 PROJECT COMPLETION REPORTS Part 1: Kpong Hydroelectric Project (Loan 1380-GR) I. Introduction ........................................... 32 2. Project Preparation and Appraisal ..... ................. 33 3. Project Implementation Operations and Cost ....... ....... 37 4. Operating Performance ...................a . ......... 49 5. Financial Performance ..................................... 51 6. Institutional Performance .......................... .. 53 7. Economic Justification ......................... 55 8. Bank Performance ........................ ........ -. -..56 9. Conclusions ................... *.......... 57 Annexes: 1. Main International Contracts and Their Financing ...........58 2. Estimated and Actual Project Costs ... ............... 6 3. Disbursement Schedule ......................... 61 This docment has a restricted distribution and may be used by recipients only in the performance of koffid duties. Its contens may not otherwise be dicosed without Wod Bank aunthoaon. TABLE OF CONTENTS (Cont'd) Page No. 4. Disbursements by Categories .............................. 62 5. VRA Operational Characteristics ........................... 63 6. Compliance with Major Covenants ........................... 64 7. Economic Re-evaluation .................................. 65 8. Schedule of Supervision Missions .......................... 68 9. Comparative Income Statements (1976-1982) ................ 69 10. Comparative Balance Sheets (1976-1982) .................. 70 11. Comparative Flow of Funds Statements (1976-1982) .......... 71 Part II. The Third Power Project (Loan 1381-GH and Credit 689-GH) 1. Introduction ............................... 72 2. Project Preparation ..... .............. 73 3. Project Implementation, Operation and Cost .......... 75 4. Operating Performance ..................................... 82 5. Financial Performance ....................... 83 6. Institutional Performance .............. .................. 86 7. Project Justification ............................... ... 87 8. Bank Performance ...... ........ 89 9. Conclusion ................................................ 90 Annexes: 1. List of foreign Suppliers and Contractors ................ 91 2. Schedule of Disbursements ........ ........... 93 3. Allocation of Proceeds of Loan .......................... ...94 4. Operational Characteristics ............................... 96 5. Compliance with their Covenants ........................... 97 6. Return on Investment ........................... 99 7. Schedule of Supervision Mission .............. ..... 100 8. ECG Income Statement .. ... ........ ....... . 101 9. Balance Sheet .................. ............. 102 10. Sources and Applications of Funds ..... ........ ...... 103 PROJECT PERFORMANCE AUDIT REPORT GHANA: KPONG HYDROELECTRIC AND THE THIRD POWER PROJECTS (LOANS 1380-G, 1381-GS AND CREDIT 689-G) PREFACE This report presents the results of performance audit of two power projects in Ghana: the Kpong Hydroelectric Project (Loan 1380-GH), and the Third Power Project (Loan 1381-GH, and Credit 689-GM). The Bank provided a US$39.0 million loan for the Kpong project in March 1977. The loan was made to the Volta River Authority (VRA), and was fully disbursed and closed by December 31, 1982. There were important cofinanciers. The Bank financed 20% of Kpong's foreign cost, Arab agencies 46%, the Canadian development fund 15%, and European agencies 11%. In addition, the Bank Group provided a US$9.0 million loan and a US$9.0 million credit for the Third Power Project in March 1977. The loan was made to the Electricity Corporation of Ghana (ECG), which was also the beneficiary of the credit. The credit was fully disbursed by May 11, 1978 and the loan by August 10, 1982. The Kpong project provided for the construction of 160 MW of additional hydro-generating capacity, and supportive transmission lines. The Third Power focused, however, on the development and rehabilitation of the distribution system, and the extension of the network to areas where costly diesel power could be replaced by hydroelectricity. A study of Ghana's tariff system was part of the Third Power project's objectives. This report consists of a Project Performance Audit Memorandum (PPAM) prepared by the Operations Evaluation Department (OED), and two Project Completion Reports (PCR) prepared by the West Africa Regional Office. The PCRs have examined the financial and institutional developments affecting the two electricity corporations in Ghana. While reviewing these issues, the audit has also focused on the projects' economic justification and on the sector's need for rehabilitation and reform measures. More specifically the audit has reviewed: - electricity tariffs, - project justification, - rehabilitation and reform measures, - cofinancing experience, - resettlement experience, and - sustainability of sector institutions. - ii - Both the PCRs were prepared by a Bank/IDA consultant who reviewed Bank reports, project records and filesl/. The PPAM is also based on similar sources, including the PCRs, a one-week mission to Ghana, and interview with Bank Staff. The audit mission visited ECG's distribution centers in Kumasi, Mampon, Bibiani, Awaso, Wiawaso, Tema and Accra. In addition to meeting ECG field staff, several residential, commercial, and industrial consumers were interviewed. Also, ECG's training program in Tema and training needs were considered. The mission is most grateful for the courtesy, and the assistance provided to it by the VRA and ECG management. Following normal OED procedures, copies of the draft report were sent to the country and cofivanciers for comments. Comments received have been reproduced in Appendix 1, and reflected in the audit report. 1/ For preparing VRA PCR the data source included: Appraisal Report No. 1299b-GH, March 1, 1977; President's Report No. P-1971-GH, March 8, 1977; the Loan and Project Agreements, March 24, 1977; the PCR prepared by VRA consultants; correspondence with the Borrower and internal Bank memoranda on the project issues contained in the Bank files and a visit to Ghana in January 1984. For preparing ECG PCR the data source included: Appraisal Report No. 1196a-GH, March 1, 1977; President's Report No. P-1972-GH, March 8, 1977; the Loan and Credit Agreements, March 24, 1977; the PCR prepared by EOG consultants, and other sources as indicated under VRA PCR. PROJECT PERFORMANCE AUDIT BASIC DATA SHEET GHANA - VRA, KPONG HYDROELECTRIC PROJECT (LOAN 1380-HM) KEY PROJECT DATA Item Appraisal Actual (est.) Total project cost (in US$ million) 236.5 250.6 /a Cost overrun () - 6.0 Foreign exchange cost (in US$ million) 172.0 196.8 Cost overrun in foreign exchange (in %) 14.4 Loan amount (in million US$) 39.0 39.0 Disbursed 39.0 Completion of physical components - hydroplant 6/81 12/81 - other facilities 10/79 11/82 Proportion completed by above date - 100% Equalizing discount rate (compared to thermal alternative) - 8Z-13% /b Economic rate of return 7.5% 7.5% Financial performance Good Not Good Institutional performance Good Good CEMLATIVE FORECAST AND ACTUAL DISBURSEMENT (US$ million) 1977 1978 1979 1980 1981 1982 As of December 31 (i) Appraisal estimate 7.3 12.6 23.4 33.5 39.0 - (ii) Actual - 4.8 13.8 27.4 35.2 39.0 (iii) as Z of (i) - 38 59 82 90 100 OTHER PROJECT DATA Actual or Item Original Plan Estimated First mention in file 10/72 Government application 8/73 Negotiations 10/76 1/77 Board Approval 12/76 3/22/77 Loan agreement date 3/24/77 Effectiveness date 6/24/77 8/24/77 Closing date 12/31/81 12/31/82 /c /a Because of the distortion of the exchange rate, US dollar equivalent of Cedi-cost greatly understates local expenditure (see VRA's comments, App. I, p. 1). /b PPAM, para. 32. 7T Final disbursement was made on 12/08/82. OTHER PROJECT DATA (Continued) Borrower: Volta River Authority Guarantor: Government of Ghana Follow-on project: None ]MISSION DATA No.of No. of Staff Date of Item Month/Year weeks Persons week Report Project preparation 6/74 Pre-appraisal 8/74-7/75 Appraisal 3/76 Supervision I 6/77 1.5 2 3 2/15/78 Supervision II 7/78 1 1 1 10/02/78 Supervision III 2/79 1 1 1 4/19/79 Supervision IV 11/79 1 3 3 11/26/79 Supervision V 10/80 1 3 3 11/05/80 Supervision VI 7/81 1.5 3 4.5 7/29/81 Supervision VII 10/81 1 1 1 12/01/81 8.0 16.5 COUNTRY EXCHANGE RATES Name of Currency (abbreviation) - Cedi (C) Year: Appraisal Year Average (1975) US$ I = C 1.15 Intervening Years Average (1975-1982) US$ 1 = C 1.81 Completion Year Average (1982) US$ 1 = C 2.75 PROJECT PERFORMANCE AUDIT BASIC DATA SHEET GHANA: THIRD POWER PROJECT-ECG (LOLN 1381-GH AND CREDIT 689-GH) KEY PROJECT DATA Item Appraisal Actual (est.) Total project cost (in US$ million) 26.7 29.5 /a Cost overrun (Z) - 10.4 Foreign exchange cost (in US$ million) 18.0 18.8 Cost overrun in foreign exchange (in Z) - 4.4 Credit/Loan Amount IDA 9.0 9.0 IBRD 9.0 9.0 Disbursed (Credit and Loan) - 18.0 Completion of physical components 6/80 6/82 Proportion completed by above date 60 98 Economic rate of return 7.5-21.4Z positive /b Financial performance Poor Poor Institutional performance Fair Poor CUMULATIVE FORECAST AND ACTUAL DISBURSEMENT (US$ million) 1976 1977 1978 1979 1980 1981 1982 As of December 31 (i) Appraisal estimate 6.1 13.1 16.6 18.0 18.0 18.0 18.0 (ii) Actual - 6.3 12.6 15.4 16.9 17.5 18.0 (iii) as Z of (i) - 48 76 86 94 97 100 OTHER PROJECT DATA Actual or Item Original Plan Estimated First mention in file 6/72 Government application 10/74 Negotiations 4/76 11/76 Board Approval 6/76 3/22/77 Credit agreement date 3/24/77 Loan agreement date 3/24/77 Effectiveness date 6/24/77 6/10/77 Closing date (Loan 1381-GH) 12/31/80 12/31/81 /c Closing date (Credit 689-GH) 12/31/80 12/31/80 /d /a Because of the distortion of the exchange rate, US$ equivalent of Cedi-cost greatly understates local expenditures. /b After improved utilization of the new facilities (PPAM, para. 34). /c Final disbursement was made on 2/09/82 and loan closed 8/10/82 when undisbursed balance of US$0.001 million was cancelled. /d Final disbursement made on 5/11/78. ~vi - OTHER PROJECT DATA (Continued) Borrower of Credit Government lof Ghana Borrower of Loan Electricity Corporation of Ghana (ECG) Beneficiary of Credit Electricity Corporation of Ghana (ECG) Follow-on Project Name ECG Rehabilitation Project MISSION DATA No.of No. of Staff Date of Item Month/Year weeks Persons week Report Identification 8/74 Pre-appraisal 2/75, 7/75 Appraisal 11/75 Supervision 1 6/77 1 1 1 8/05/77 Supervision II 12/77 1 2 2 2/01/78 Supervision III 2/79 1 1 1 4/13/79 Supervision IV 10/79 1 2 2 12/14/79 Supervision V 10/80 1 2 2 11/11/80 Supervision VI 6/81 1 2 2 7/29/82 Total 6 10 -vii - PROJECT PERFORMANCE AUDIT REPORT GHANA: KPONG HYDROELECTRIC AND THE THIRD POWER PROJECTS (LOANS 1380-GH, 1381-GH AND CREDIT 689-Ga) HIGHLIGHTS The two projetts, under review in this report, are complementary. The Kpong Hydroelectri. Project (Loan. 1380-GK), which VRA has undertaken, provides 160 MW of capacity, and the Third Power Project (Loan 1.381-CH and Credit 689-G), executed by ECG, focused on the development and rehabilita- tion of the distribution system, and the extension of the network to areas where hydroelectricity could replace costly diesel power. The Bank provided a US$39.0 million loan for the Kpong project; the loan being fully disbursed and closed by December 31, 1982. Also the Bank Group provided a US$9 million Loan, and a US$9 million credit for the Third Power Project. These amounts were fully disbursed by August 10, 1982. Technically, the Kpong project was a success. Implementation delays were attributable to new procurement procedures, introduced in mid-1977. Cost overruns were also modest (PPAM, para. 7). The project's success is striking, given the great difficulty caused by the worsening economic situation and political turmoil. VRA and its consultants and contractors were responsible for this accomplishment (PCR, para. 3.08). In contrast, ECG's performance was poor. Deterioration in system reliability contributed to a slowdown in sales. However, facilities constructed under the distribution project helped to ease some of the operational problems of the Corporation. Most ECG project components were completed 1 to 3 years late. Disruption in the economy was the main cause of the delay. Management weaknesses, poor maintenance, and a lack of skilled and motivated staff contributed to deterioration in the quality of service (ECG PCR, para. 4.03). Moreover, the 10% estimated cost overrun substantially understates local expenditure (PPAM, para. 12). The performance of consultants, working on the electrical part of the project, was satisfactory (ECG PCR, para. 3.27). But questions continue to be raised about the competence of the local civil works consultants. Ghana has suffered from severe drought and because of it, probably much of Kpong's additional generating capacity would be utilized several years later than originally forecast (VRA PCR, para. 7.02). However, improved utilization, which depends on water inflows, cannot be foreseen at this stage. Also many facilities constructed under the Third Power project are currently underutilized (ECG PCR, para. 3.18). But improved utilization could be achieved by the extension of service to many potential consumers in the existing service areas. - viii - Because of high inflation, the two PCRs' presentation of accounts in current prices and local currency is inadequate (PPAM, para. 16). Owing to the supply interruption to VALCO2/ and other cutbacks in sales, the Authority's financial performance is currently out of line with the SAR3/ assessment (PPAM, para. 17). ECG's financial problems have, however, been partly caused by institutional weaknesses (ECG PCR, para. 5.02), and partly by "rigidity in Governmental machinery.-4/ Despite significant tariff increases in July 1984, ECG's finances remain precarious. Also, the Corporation is short of foreign exchange for acquiring materials for the maintenance and rehabilitation of distribution facilities (PPAM, para. 18). Other major themes which the audit has reviewed include: (a) proj- ect justification, and (b) sector rehabilitation and reform. The Kpong project illustrates the weakness of the link binding some country economic and project works in the Bank. Ghana's economy, suffering from stagfla- tion,4 required that only the very high priority programs should receive the funds spent on Kpong (PPAM, para. 29). This is because the audit considers that Kpong was not required to supply the domestic market provided, however, that electrical energy exports and sales to VALCO could have been frozen at the 1976 level (PPAM, para. 29). Thus Kpong's justification depended on sales to VALCO and the export market, but the power rates for these outlets were not enough to justify the project (PPAM, para. 32). Further, given the weakness of the distribution system, it should have been possible to redirect -equipment and materials to maintain existing facilities rather than carry out the extensions in full- (ECG PCR, para. 9.02). Since many facilities, under the project, are currently underutilized (ECG PCR para. 3.18) and since distortions in the foreign exchange rate have also resulted in gross understatement of cost, the PCR's reassessment of the return on investment at about 36% exaggerates the project's net benefits. The audit mission has found that a significant portion of the potential con- sumers in the project area are not currently being served because of the shortages of meters, transformers, wires, poles, insulators, service vehi- cles, etc. Since ECG depends on imports for the supply of most materials, foreign aid in helping it to acquire these materials would improve the utili- zation of distribution facilities. Therefore, the Third Power's justifica- tion depends on the extension of service to potential consumers and on the speed of the Ghanaian economic recovery and expansion (PPAM, para. 34). But more needs to be done than merely await the outcome of economic recovery. The rehabilitation and reform program, presented in this report, comprises: - a '-alanced development program for the system's generation, transmission and distribution facilities; 2/ Volta Aluminum Company. 3/ Staff Appraisal Report (SAR). 4/ VRA's comments, App. I, p. 1. -ix- - foreign exchange supply for ECG so that it acquires materials for the improved utilization of distribution network, and for the efficient operation and maintenance of the system; - an effective training program for ECG' s staff; - revised tariffs securing for the sector a steady improvement of financial performance and reflecting gradually the future cost of power supply; - revised incentive package, particularly for ECG employees; and - sector coordination. The reform measures envisage the emergence of a unified sector and a strong distribution entity, capable of extending and administering an effi- cient and reliable distribution network, and of pursuing a well-balanced and well-integrated expansion program (PPAU, para. 47). Two other lessons which merit emphasis are that: - for hydro project in the Sahel, extensive drought periods must be considered in structuring the hydrological design and operation criteria (PPAM, para. 66); and - in economies where growth prospects are uncertain, more emphasis should be placed on ensuring full utilization of capacity of existing facilities, rather than investing in extensions of such systems (PPAM, para. 66). -1- PROJECT PERFORMANCE AUDIT MENORANDUM GHANA: KPONG HYDROELECTRIC AND THE THIRD POWER PROJECTS (LOANS 1380-GH, 1381-GE AND CREDIT 689-GR) I. PROJECT SUMMARY 1. Ghana's Five Year Development Plan (1975176-1979/80) envisioned the rapid expansion of generation and distribution facilities, the phasing out of diesel generators, and the extension of service to the Southwest and a few rural centers. The plan targets were largely incorporated in the two proj- ects under review in this report. The two projects were executed by the Volta River Authority (VRA), and the Electricity Corporation of Ghana (ECG). 2. VRA generates and transmits hydropower. ECG retails supplies at a voltage level of up to 33 kV, and operates a small number of diesel units in remote centers. VRA's clientele comprises: ECG, Volta Aluminum Company (VALCO), Akosombo township, several mining companies, and Commnmautf Elec- trique du Bfnin (CEB) which serves Togo and Benin. In terms of the 1982 sales, VALCO bought 63%, ECG 21%, and CEB 10% of the VRA supplies. 3. Of the two projects, Kpong executed by VRA, provided for: - 160 MW addition to the system' s installed capacity; - network extension (a) to tie in Kpong with VRA grid, (b) to extend- power lines to the Southwest, and (c) to reinforce and expand capa- city between Tema and Accra; - additional reactive power compensation for the system's 161 kV grid; and - resettlement of 7,000 residents of the reservoir area. 4. The Third Power Project, focusing on the development of the distribution system, comprised: - new subtransmission lines (240 km. of 22-33 kV network); - new substations (11 units of 33/11 kV systems); - expansion of existing substations (six units); - improvement of distribution network, particularly in Tema; - rehabilitation of the existing transmission and distribution system; - replacement and expansion of vehicle fleet; - engineering services for ECG; and -2- - a power sector study, "to derive therefrom a tariff structure designed to reflect the marginal cost of power...- - ECG SAR, para. 2.22. 5. Ghana has suffered from three successive years of severe drought (1981-83). Chiefly because of it Xpong's additional generating capacity will probably be utilized several years later than originally forecast (VRA PCR, para. 7.02). But improved utilization. which depends on water inflows, I/ cannot be foreseen at this stage. Because of power shortages VALCO, which had ceased operation, has now commenced limited production. ECG's energy supply had also been severely curtailed. Yet, against enormous difficulties, caused by a receding economy and political turmoil, the construction of Kpong was completed. VRA and its contractors and consultants were responsible for this remarkable achievement (VRA PCR, para. 3.08). The consultants together with "the main contractors.. .conceived and helped carry out imaginative measures to overcome the obstacles created by the near breakdown of the economy--VRA PCR, para. 3.29. Contractors, engaged in VRA's network extension, did not perform well, however-VRA PCR, para. 3.32. 6. The dam was completed six months behind schedule. Delays in bid preparation, analysis, and clarification were the cause (VRA PCR, para. 3.07). After this hurdle, implementation proceeded smoothly. Other delays, affecting electromechanical equipment and the connecting grid, did not impede utilization of the dam. Construction of the network extension was completed 5-28 2/ months late. This setback was caused by the Government's new contract approval procedure (VRA PCR, para. 3.11), which required the Authority to submit bid documents to a Public Agreement Review Committee. The Committee, in turn, sought the concurrence of the Supreme Military Council. 7. When expressed in US currency, Kpong's cost overrun was a modest 6% (VRA PCR, Table 3.03). Although the project's aggregate foreign cost increased by 14%, the local cost in terms of the US dollar fell below the appraisal estimate. The main reason seems to have been significant adjustments in the country's foreign exchange rate. Because of exchange rate adjustments it is not meaningful to discuss the local cost overruns.3 Over 80% of the foreign costs were paid for civil works and equipment contracts. The foreign cost of these contracts increased by 17%: 32% increase in civil works and a slight decline in equipment contracts (VRA PCR, Table 3.04). Over three-fourths of the increase in the foreign cost of civil works stemed from changes in the scope of work and price escalation. Further, economic dislocation added to the cost . For example, materials and equipment had to be imported because they were locally unavailable (VRA PCR, para. 3.22). 1/ VRA's comments, App. I, p.2. 2/ VRA's comments, App. I, p.2. 3/ VRA believes that "This could be done through indexation, i.e. comparisons made in 1976 Cedis," App. I, p.2. - 3 - 8. Because of cofinancing, five types of contract were awarded: three based on different versions of international competitive bidding (ICB), one on bids by Canadian firms, and one on bids by local contractors. Although cofinancing complicated the procurement, the implementation delay was mainly caused by the introduction of new procedures (PPAM, para. 6). 9. About 79% of Kpong's costs were in foreign exchange. The World Bank financed 20% of the foreign cost, Arab agencies 46%, the Canadian Development Fund (CIDA) 15%, and European agencies 11%. The Bank's US$39.0 million loan was fully disbursed and closed by December 31, 1982, about one year behind the SAR schedule. 10. In contrast to the Authority's satisfactory performance ECG did rather poorly. Its management weaknesses, poor maintenance practices, and lack of skilled and motivated staff undermined the quality of service (ECG PCR, para. 4.03). Retail sales fell partly because of the system reliability becoming still less reliable. 11. Disruption in the economy was the main cause of the Third Power's implementation delays. Materials were hard to acquire. Local funds were scarce and staff were generally demoralized and pessimistic (ECG PCR, para. 3.03). Except for part of the Tema extension, other ECG project components were completed 1k to 3 years late. 12. The ECG PCR states that chiefly because of gross distortions in foreign exchange rate, the 10% estimated cost overrun substantially understates local expenditure4/ (ECG PCR, para. 3.17). Two-thirds of the estimated cost overrun are due to price escalation, and one-third to a few modest changes in the scope of work (ECG PCR, para. 3.17a). 13. The Bank Group provided US$18.0 million for almost the total amount of the project's foreign exchange cost. One-half of the amount was in loan and one-half in credit. The package was fully disbursed and closed by August 10, 1982, about 1k years behind the original schedule. 14. ECG retained one foreign and several local consultants. The relationship between the two groups remained undefined. While the Corporation assumed that the foreign consultant would provide the necessary information and instruction to civil works consultants, this lack of clarity proved to be a stumbling block. The foreign consultant assisted ECG in engineering, procurement and supervision of equipment installation. But it did not initially provide any guidance to local consultants. Subsequently the Corporation requested that instructions be given to local consultants who supervised the work of local civil works contractors. The limited -technical ability of both local consultants and contractors allowed only slow progress, often associated with substandard work" which "had to be corrected or even redone" - ECG PCR, para. 3.04. The Corporation has now recognized the advantage of retaining one consultant for such works while providing for 4/ VRA believes that -shadow exchange rate could be used if the official rate is not accepted-, App. I, p.1. VRA does not provide any estimate of its ou-. others to subcontract from the main consultant. To summarize, the performance of consultants working on the electrical part of the Third Power was satisfactory (ECG PCR, para. 3.27). However, questions continue to be raised about the competence of local civil works consultants. 15. Further, while equipment suppliers delivered the goods on time, local contractors failed to complete the corresponding civil works expeditiously. ECG, therefore, decided to make the electrical contractors responsible for the associated civil works. The new arrangement worked well. 16. As for the financial performance of the two agencies (VRA and ECG), the PCR's presentation of accounts in current prices and in local currency, under a long-term hyperinflationary condition, is clearly insufficient. In response the Western Africa Region maintains that the PCRs have reviewed the accounts as -they are kept in local money and believe that this method would be more reliable because of lack of adequate statistical data in Ghana". However, at least VRA could have been requested to present its finances expressed in terms of a stable unit of account, since a considerable portion of its income is in foreign currency - retained for the servicing and the repayment of foreign loans (Akosombo and Kpong project loans). 17. Because of recent supply disruptions to VALCO (PPAM, para. 5) and cutbacks in other sales, the Authority is currently short of foreign exchange and its financial performance has diverged from the SAR assessment. Still, VRA was able to finance 37% of its 1977-82 capital investment from internally generated funds, compared to SAR's estimate that such savings would fund 32% of a significantly higher level of capital investment, including about $110 million for the Bui hydroproject (VRA PCR, para. 5.03). 18. In contrast, ECG's financial plight is partly institutional. The PCR states that the deferral of Government's approval of -tariff increases were apparently due to its reluctance to provide them to such an inefficiently run utility- - para. 5.02. An example in maladministration is the Corporation's receivables which were often more than 7 months' revenue (ECG PCR, para. 5.04). Despite ECG's weaknesses, the Government has authorized considerable increase in tariffs, under the Economic Recovery Program, in 1983 and 1984. These adjustments enabled the Corporation to earn an improved rate of return (ECG PCR, para. 5.03). ECG's finances remain, nevertheless, precarious: the Corporation totally lacks foreign exchange to maintain or rehabilitate its distribution facilities. Also its debt-equity ratio has deteriorated because of the sharp devaluation of the cedi, the local currency. II. SUPPLEMENTARY COMMENTS 19. The assessment of the two projects must be undertaken in the context of Ghana's severely impaired socio-economic condition. The long neglect of the economy aggravated by the several successive droughts5/ and a 5/ See VRA's comments, App. I, p.3. - 5 - poor outlook for the country's principal export, cocoa, could not be quickly reversed. The economy had all along needed a short, medig and long term development strategy that would have focused on stabilization, rehabilita- tion, and then economic growth. This is the type of strategy which is now embodied in the Economic Recovery Program. Already domestic savings and investment rates have hit such a low watermark that they cannot, on their own, forestall the continuing slide in per capita income; the domestic savings and investment rates each is now about 1% of the GDP6/. The high rate of inflation (40% a year during 1970-82) has also been disruptive. Other major indicators that clearly reveal the weakness of the economy, since the year 1970, include: a 30% drop in per capita income, an 80% f all in real wages, a 52% decline in real export earnings, and a 33% downturn in import volumes. About half of the country's much reduced export proceeds are also needed to finance petroleum imports. 20. However, the new government (December 31, 1981) has embarked on an Economic Recovery Program in April 1983, including: moves towards a realistic exchange rate, gradual easing of price controls, improved monetary and fiscal disciplines, and rehabilitation programs for key sectors. In the face of severe shortages of food and other essentials, the Government has continued to take a tough stand on many economic fronts. Luckily a more normal weather in 1984 has also ended several successive years of drought.5/ Food, hydroelectricity, and the production of export goods have begun to recover. With sizeable new aid commitment, the economy is expected to continue its recovery in 1985. Against this checkered background, the following themes are reviewed: - Electricity tariffs; - Project justification; - Rehabilitation and reform measures; - Cofinancing experience; - Resettlement experience; and - Sustainability of sector institutions. 6/ Gross domestic savings fell from 17% of GDP in 1960 to 1% in 1982. Concurrently, gross domestic investment sagged. It was 24% of the GDP in 1960 and 1% in 1982 (see, Towards Sustained Development in Sub-Saharan Africa, World Bank, August 1984, p. 61). While per capita GDP stagnated in the 1960s, it fell by 2.5% a year in the period 1970-82. All sectors of Ghana's economy experienced negative growth in this period. 51 See VRA's comments, App. I, p.3. -6- Electricity Tariffs 21. This section briefly reviews: VRA's power rates for VALCO and CEB, and ECG's average price for the domestic consumer. 22. VALCO. Ghana and VALCO have both gained from the original Master Agreement.7Without it the Akosombo dam would not have been built, and the country's dependence on costly sources of energy would have been greater than it is today. What is more, VRA would not have existed, and a much weaker power generation authority would probably have functioned in its place. 23. The Agreement stipulated: (1) the delivery of up to 370 MW of firm power to VALCO, (2) the payment of 0.26 US4/kWh by VALCO to VRA or a minimum annual payment of US$7-8 million, and (3) a 30-year agreement (1967-97) without any adjustment clause. The rigidity of the power rate, stipulated in the Agreement, implied that: (a) VALCO would have reaped the benefit from any erosion in the purchasing power of the dollar, the erosion being unavoidable over a 30-year period, and (b) VALCO would have captured all windfall gains from any energy price upheaval which could boost aluminum prices because of a higher electricity cost, pushing up the aluminum production cost. 24. With the expansion of the Ghanaian economy, it should have been expected that Akosombo output would increasingly cater to the needs of the domestic market. Article 5 of the Agreement had recognized this aspiration: Further, th. Authority shall not after the tenth anniversary of the Perma- nent Delivery Date be obliged to increase the said amount to an amount greater than three hundred and fifteen thousand kilowatts (but this sentence is without prejudice to any higher figure which may be in effect at that date)." 25. In the context of Kpong project and additional supply to VALCO, Ghana renegotiated, in 1975, a power rate package comprising a payment of: 0.325 US4/kWh for the period 1973-75; 0.45 US4/kWh for 1976, the rate increasing gradually to 0.5 US/kWh by 19S1. Moreover, the Authority agreed to supply VALCO 30 MW of additional firm and 15 MW of interruptible electri- cal energy. The power rate on the additional supply was to be derived from an agreed formula that would use the project's revised final cost, one year before its completion. If the ensuing rate exceeded 0.975 US 4/kWh VALCO could decrease or forego entirely its claim on the 30 MW. 26. Ghana has done well in scrupulously honoring its international legal obligations. But the power rate agreement for additional supply to 7/ The Master Agreement between the Government of Ghana and Volta Aluminum Limited, concluded on January 22, 1962, includes several scheduled documents, one of which is about the power contract. PPAM, para. 23 refers to a number of provisions in that contract, defining: - power ceiling (Article 5); - contract rate (Article 12); - minimum charge (Article 13); and - period of agreement (Article 23). -7- VALCO, over the 315 MW supply ceiling, should have recouped the incremental cost of electricity generation, including a fair rate of return on investment, fully reflecting the opportunity cost of capital to the Ghanaian economy8/ (see PPAM, paras. 32-33). Moreover, since investment has its risks the Authority should have required VALCO's participation through, for example: (a) a fixed annual payment, irrespective of the actual amount supplied; and/or (b) an equity contribution towards financing the project cost. 27. CEB. The Authority has a continuing contractual commitment to supply Communautf Electrique du Bfnin (CEB) up to 80 MW of capacity. The price CEB pays has been modest. SAR estimated it at 0.82 US 4/kWh/ (VRA SAR, para. 2.15). Subsequently it was increased to around 1.4 US4IkWh in 1978, and to 2.5 US V/kWh in April 1982. From April 1, 1985, CEB's rate became 4 US4/kWh for supply up to 500 GWh, and 41 US4/kWh for supply above 500 GWh. These rates would be adjusted for inflation. Therefore, CEB's rate, now, approximates the incremental cost of Kpong supply.Lo 28. ECG. Under the Economic Recovery Program, domestic tariffs increased significantly (Cedis 1.92/kWh, July 1984)111. Given that Ghana's exchange rate is considerably distorted, ECG's retail tariff still falls short of Kpong's incremental cost uf power supply which, as shown in Annex Table 1, is 5 US4/kWh, in 1982 dollar values. Under the project, ECG had hired a consultant to undertake a sector study, define power supply costs, and propose a tariff policy for the sector. The study, completed in 1980, suggested new tariffs based on the long-run marginal cost of power 8/ See VRA's comments, App. I, p.3. 9/ This is based on a demand charge of 2.48 US$/kW per month, an energy charge of 0.00248 4/kWh, and a construction charge of US$24,500/month. On a 70% load factor and 50 MW supply, the average rate would be 0.82 4/kWh ($2.54 million divided by 306.6 an kWh). 10/ See VRA's comments, App. I, p.3. 11/ At 38 cedis to 1 US dollar (1984), the power rate would be 5.1 US/kWh (See ECG PCR, Annex 6-footnote). The recent economic report entitled: "Managing the Transition", states that electricity "tariffs were initially adjusted by 40% in April 1983, and then in January 1984 by 500%, from an average price per kWh of 10.31 pesewas to 50 pesewas" - Report No. 5289-GH, November 7, 1984, para. 1.10, p. 6. The same report considers the official exchange rate to be distorted: "The existence of a black market and the large premiums over the official rate (three times or more the official rate) reflects the huge imbalance between the demand and supply for foreign exchange..." (para. 2.07). -8- supply.12/ Because of the country's economic difficulties these recommenda- tions were not immediately implemented (ECG PCR, para. 3.12). With the recent turnaround in the economy and significant amounts of new foreign assistance in support of the country's Economic Recovery Program, it should now be possible to gradually restructure tariffs and raise their levels (see PPAM, paras. 43-45). Project Justification 29. The Kpong project exemplifies the tenuous link that binds some country economic and project works in the Bank. The project, initially esti- mated to cost US$236.5 million, claimed a considerable portion of public sector resources. But Ghana's economy required that only the very high priority program, which could help stabilize and rehabilitate the economy's critical sectors, should have received the funds spent on Kpong. The Bank staff were aware of past economic problems. In the President's report of March 8, 1977 the staff portray a fairly bleak economic picturel3/. They also outline the principal objectives of the Bank Group's assistance (President's Report, para. 14). One, with some bearings on the project objective, concerned the improvement in "the country's essential infrastruc- ture so as to relieve constraints upon economic growth-. But Ghana's slug- gish economy did not require Kpong's additional supply provided sales to VALCO and CEB were restricted to the 1976 levelt4/ (PPAM Annex Table 2). 30. Besides, domestic market needs could hardly be foreseen. The SAR presented a somewhat bright economic prospect, however: "average expected load growth of public utility consumption corresponds approximately to a GNP growth of the order of 5.0% and a GNP/electricity consumption growth ratio of up to 2, based on limited historical data- - para. 5.03. But as indicated 11/ See VRA's comments, App. I, p.3. 13/ Report No. P-1971-GH states, "over the past decade and a half GDP growth averaged 2.5 percent per annum and failed to keep pace with the growth of the population"...-The imbalance has depressed government development expenditures to an inadequate level and has also been a persistent source of inflationary pressure"... "It is difficult at present to assess the likelihood of the government taking all the economic policy measures necessary to restore equilibrium on the external account and provide a more secure basis for accelerated growth in the domestic economy." 14/ See VRA's comments, App. I, p.3. -9- earlier (PPAM, para. 19), historical data show a sagging per capita GNP, a lackluster GNP growth performance, and a receding industrialI5/ sector. 31. Since expansion of the domestic market would not justify Kpong, the question remains whether the additional sales to VALCO and CEB could have vindicated it. This issue is relevant because Akosombol6/ supplies to VALCO and CEB could have been restricted to about 365 M7/ leaving enough capacity to cater to the Peeds of the domestic market in the medium-term. Therefore, Kpong should have been justified on possible sales to VALCO and CEB. In contrast, the SAR's economic assessment is based on domestic market sales. Benefits were calculated on 75% of the Liberian tariff (3.75 US{/WWh at 1976 prices), yielding an economic rate of return of 16%. At 1982 prices a 12% rate of return from Kpong would have required a wholesale electricity price of 5 US4/kWh (PPAM Annex Table 1). Both VALCO and CEB paid, at the time of appraisal, power rates below 1 US4/kWh (PPAM, paras. 25 and 27), and their expected payments in 1980 were substantially below the rate that ensure a reasonable return on investment. 32. It might, however, be argued that VALCO's willingness to pay a rate higher than the one in the original Master Agreement (PPAM, para. 25) was probably motivated by the expectation that it could secure additional supplies from Ghana, although VRA believes that -VALCO's willingness to pay higher rates is due mainly to pressure from the Ghanaian public and the international community".18/ If the increase in benefits, resulting from the rise in VALCO's power rate, be entirely attributed to Kpong, the project's economic rate of return would still approximate 5%. Further, SAR could have argued that since Ghana cooperated in providing additional supplies, VALCO would have been more receptive to an upward revision of power rates in the 1980s. Such an expectation was, however, too tenuous to support a project of such a magnitude. A further "-gument would have been that, notwithstanding the sluggish growth of the economy, domestic sales could expand by about 3% a year because of the possibility of extending service to new customers, mostly 15/ 2.4% a year decline during the 1970-82 period (Towards Sustained Development in Sub-Saharan Africa, World Bank, August 1984, p. 58). Also the quotation from SAR para. 5.03 should be referring to an electricity consumption/GNP (not a GNP/electricity consumption) growth catio of up to 2, since, as shown in PPAM, Annex Table 2, the electrical energy demand of the domestic market (ECG and others) was expected to grow by 8.6% a year during 1976-86. 16/ PPAM Annex Table 2 shows energy supply from Akosombo and diesel sets to be about 5,600 GWh. Supplies to VALCO and CEB, at 95% load factor, would have amounted to about 3,000 GWh. 17/ As stated in PPAM, para. 24, Article 5 of the Master Agreement required 315 MW of supply for VALCO, after the tenth anniversary of the permanent delivery date. This is also reflected in VRA SAR Annex 11, Table 1. In addition, the SAR projected up to 50 MW of supply to CEB. 18/ App. I, p.4. - 10 - in the existing service areas. As shown in PPAN Annex, Table 3, benefits that the domestic economy would derive from the project are in the distant future and their present values are limited, giving an overall economic rate of return of about 7.5%, including the incremental benefits from the rise in VALC'; s rate being entirely attributed to Kpong. 33. Kpong has had some other benefits. Although the SAR could not foresee the occurrence of successive droughts, because it has been unprecedented, 19 construction of the dam has helped mitigate the severity of power cuts: Since downstream waterflows are used for electricity generation, Kpong does not compete with Akosombo for the use of Ghana's water resources. To conclude, the VRA PCR's 16Z rate of return (PCR, para. 7.06), based on ECG tariffs in effect on July 1, 198420/, is questionable. Kpong was not needed, in the medium-term, for the domestic market, and VALCO and CEB's expected payments were not enough to justify its construction. 34. As regards the justification of the Third Power, given that the system suffered from an acute shortage of spare parts, the project ceased to remain a least-cost solution, at the implementation phase. This is because the use of funds for the rehabilitation of the existing system would have been more advantageous than their use in a less effective distribution net- work. Although the ECG PCR estimates the return on investment to be about 36Z, over three times the appraisal estimate, this high rate is the result of a 49 fold increase in retail tariffs. However, continued distortion of the foreign exchange rate means that the project's foreign cost, expressed in local currency and at July 1, 1984 prices (PCR Annex Table 6), is grossly understated. It is also uncertain whether the July 1, 1984 tariffs could be maintained in real terms under the prevailing hyperinflationary situation (PPAM, para. 41). Moreover, since many facilities, under the project, are currently underutilized (ECG PCR, para. 3.13), the PCR's high rate of return exaggerates the project's net benefit. An improved utilization rate could, however, have been secured if investment in generation, transmission, sub- transmission and distribution programs were well-balanced, and more resources were allotted for maintenance and for the additional supply of meters, wires, transformers, poles, insulators, etc. An improved utilization rate will alno result from the rapid recovery and expansion of the Ghanaian economy. Since tariffs inadequately measure benefits and since energy shortages could severely retard economic development, the Third Power's justification depends ultimately on the speed of the Ghanaian economic recovery and expansion. As indicated elsewhere, the prospect for a better economic performance has recenatly improved (PPAM, para. 2), and with that the Third Power Project might well be vindicated. 191 SAR has stated: -Forty years of hydrological data are available for deriving streamflow records.. .The longest continuous below average run-off period occurred from 1936 to 1944 with a nine-year mean annual flow of 1,000 m3/s."-Annex 1, paras. 3-4. The recent drought was more severe. 20/ Cedis 1.92/kWh (VRA PCR, Annex 7, p. 3); see also PPAM, para. 28. - 11 - Rehabilitation and Reform Measures 35. However, more should be done than merely await the outcome of the country's economic recovery. A package of reform measures is required for sector rehabilitation and development. The package needs to include: - a program for the balanced development of generation, transmission, and distribution facilities; - a provision for foreign exchange supply to ECG so that the Corporation could acquire materials for improving the utilization of its distribution network and for the efficient operation and maintenance of its system; - an effective training program for ECG's staff; - revised tariffs, securing for the sector a steady improvement of financial performance and reflecting gradually the future cost of power supply; - a revised incentive package, primarily for ECG's staff; and - sector coordination. 36. A Balanced Rehabilitation and Development Program. Despite the acute shortage of materials, ECG is currently serving a large number of cities and towns - some important centers (for example, Sunyani and Tamale) with diesel generators. Moreover, most Ghanaian industrial establishments depend for their electricity needs on public supplies. The underutilization ,f distribution network under the project, reported in the ECG PCR, relate primarily to 161/33 kV and 33/11 kV substations as well as in the subtrans- mission grids (33 kV and 11 kV lines). The unused capacity in these facili- ties would be utilized more fully as ECG succeeds in extending service to potential customers. The new service areas, under the project (for example, Bibiani-Wiawaso-Sefwi), have many potential consumers who could be advanta- geously served. But because of the shortage of meters, transformers, wires, poles, insulators, service vehicles, etc., they must wait their turns. Therefore, a greater emphasis on the improved maintenance of the distribution system, and on the extension of service to potential customers in the exist- ing service centers should result in a better balance between the distribu- tion service provided by ECG and the generation and transmission facilities supplied by VRA. 37. Moreover, ECG is faced with many options: (a) rehabilitation and upgrading of the quality of service versus expansion of the electricity service; (b) extension of subtransmission and distribution lines to new areas versus extension of service to potential customers in the existing service centers; (c) human resource development versus expansion of physical assets. Considering the importance of building a strong ECG, the focus needs to be, at least in the short-run, on: human resource development, rehabilitation, quality of supply, and potential customers in the existing service areas. - 12 - 38. Foreign Exchange Resources. Since ECG depends on imports for the supply of most materials, foreign exchange resources are necessary for achieving the short-term objectives noted at the end of the preceding para- graph. The Bank Group can play a key role in assisting the power sector to rehabilitate and develop by financing the foreign exchange needs of the reform program (paras. 35-47). 39. Training and Human Resource Development. Already VRA has an effective training program. But ECG's center at Tema, and a few other schemes which it sponsors, fall considerably short of the distribution system's training requirements. Three areas need priority attention. They concern: (a) the artisan class, (b) the distribution engineers, and (c) a team of planners for the preparation and the monitoring of expansion programs. 40. Over 40% of the ECG staff belong to the artisan class, numbering about 2,000. They are wiremen, cable jointers, linesmen, electrical fitters, engine fitters, switch board attendants, etc., who constitute the bulk of ECG's technical field staff. Each year, about 20% of this group need to be trained or retrained. Facilities at Tema, and the supportive staff, are woefully inadequate for this task. Lack of equipment and tools is not the only problem confronting ECG. An effective training program must also be prepared, involving: (a) the determination of training requirements of each distribution district for specific skills, reflecting the trainees' job performance and know-how; (b) the preparation of suitable curricula; (c) the training and retention of a core of skillful teachers; (d) suitable accommo- dation for trainees coming from regional distribution centers; and (e) the periodic evaluation of the program's effectiveness. 41. In the past, ECG's engineering staff were eligible for overseas training after serving the Corporation for a period of three years. The South West Electricity Board (SWEB) organized the overseas program. This type of training is desirable at least for two reasons: to improve the staff morale, and expose staff to system maintenance, management and opezation in other countries. Because of foreign exchange limitations, ECG has been forced to de-emphasize its overseas training program. 42. In addition, ECG needs a team of planners in Accra and district centers to program rehabilitation and expansion work, and to monitor and evaluate progress. The training program should teach how the trainees could prepare a least-cost expansion program, and how they should collect and ana- lyze key socio-economic information needed for arriving at sound judgment about the system's extension to new centers. 43. Revised Tariffs As noted elsewhere, substantial tariff increases have already been sanctioned (PPAM, para. 28). The objective of further tariff adjustments should be: (a) to secure a steady improvement, in the medium-term, in the sector's financial performance; and (b) to achieve, shortly after realizing the financial objective, a tariff level reflecting the future cost of power supply. - 13 - 44. This dual approach in reforming the tariff system responds suitably to the Ghanaian needs, since, in the absence of a least-cost expansion pro- gram, the calculation of the long-run marginal cost (LRHC) of power supply is bound to be arbitrary. Under the Third Power, the consultants have already studied an LRMC pricing system which, according to VRA, has been reflected in the domestic tariff structure. VRA and ECG should continue their efforts in refining, in-house, the LRMC pricing study, which needs to be based on a least-cost expansion program for an increasingly efficient electricity supply industry. These refinements should measure, with a greater accuracy than now possible, the divergence bet.reen the prevailing level/structure of rates and the LRMC tariffs. 45. However, the immediate focus must be on a substained increase in power rates, in real terms, so that an 8% financial rate of return on the sector's revalued assets is secured at the end of a four to five year period. Current price regulations2l! are insufficient, since they require that ECG's retail tariffs should increase 5% a quarter until the end of March 1986. Since the cost of capital goods will continue to rise, when expressed in cedis, faster than 5% a quarter, the electricity tariffs, governed by existing regulations, will become increasingly incapable of recovering the cost of power sector investment. Therefore, power rates need to be adjusted quarterly by combining 2/ an index that accurately measures changes in the cost of capital goods for the electricity supply industry with another compo- nent introducing a sustained rate of increase in tariffs, in real terms, so as to achieve the sector's overall financial objective within a realistic time-frame. The appropriateness of a sector-specific price index for capital goods stems from the need to recover investment cost by reflecting changes in such costs on the tariff level. Since Ghana imports the bulk of its capital goods, the cost of such imports should be expressed in local currency, reflecting the increasing devaluation of the cedis. The Bank should assist VRA and ECG to work out an appropriate index for tariff adjustments. 46. Incentive Package for Employees. The ECG and VRA staff receive modest pays. VRA is, however, able to motivate, attract and retain staff not because of high monetary rewards, but primarly because the staff perceives the Authority as providing a good opportunity for career development and as giving an incentive package focused on their basic needs, such as housing and transportation. Elsewhere the need for human resource and career development 21/ Electricity Corporation of Ghana (Power Tariffs) (Amendment) Regulations, 1984. 22/ Tariffs need to be adjusted as follows: First quarter: A price index, reflecting quarterly changes in the cost of capital goods, to be expressed in cedis, for the electricity supply industry (CGEL 'ndex) + 1.10. Second quarter: CGEL Index. Third quarter: CGEL Index + 1.10. Fourth quarter: CGEL Index. - 14 - in ECG has been emphasized (PPAM, paras. 42-45). A good training program for ECG should be supplemented by a well-balanced incentive package, including housing benefits and mass transportation facilities to work place. These measures should help the Corporation to improve its work environment. 47. Sector Coordination. The Authority has thrived because of good management, an autonomous status, and the market security, which has enabled it to earn foreign exchange. Notwithstanding Ghana's other urgent needs, the maintenance of VRA's autonomy is of paramount importance: A good institution can easily be destroyed, but to rebuild it, under adverse economic conditions, would be an insuperable task. 48. The Ghanaian economy benefits from VRA operations mainly through the distribution of electrical energy among productive and other sectors. Because of ECG's weaknesses, these benefits have not been fully realized. For Ghana to obtain a much richer return from VRA operation, ECG must be transformed into a well-managed and vigorous organization. There is also a need for well-balanced expansion program, encompassing generation, transmission and distribution facilities and tailored to the Ghanaian domestic economic requirements. Cofinancing Experience 49. The cofinancing package, exclusively from Official Development Assistance, took four years to be arranged. The package has been to Ghana's advantage, since many agencies adhered to ICB and similar procurement procedures (PPAM, para. 8) and the grant element in the loan was high. For example, the Arab agencies provided US$73 million for a period ranging from 17 to 20 years, including a 5-year grace period at an annual interest cost of 2-6Z (VRA's comments, App. I, p.5). Since considerable efforts went into putting the package together, an element of additionality in the loan amount is evident, at least from the Ghanaian perspective. Resettlement Experience 50. Ghana's previous resettlement experience (Volta Lake) has been val- uable in the design of Kpong resettlement program. The Volta resettlement, which established 52 sites for relocating 80,000 people, had emphasized the development of large land holdings and farm mechanization. In contrast, Kpong resettlement has relocated some 7,000 people iG ix new towns, and its focus has been on small farms, and appropriate farming practices, including traditional agriculture. Under the recent program,people were not removed from their natural habitat. What is more, they were initially invited to participate in the decisions made concerning the towns and neighborhoods they preferred to live in. This participation was through village chiefs. 51. As in the earlier program, the resettlement objective aimed at the establishment of viable economic communities rather than refugee camps, and at equitable compensations for the loss of property and productive assets. The PCR has not assessed the experience in depth. But the contrasting strat- egies, employed by the two resettlement programs, have important lessons for other nations, and therefore, their relative merits need to be studied. Such - 15 - an assessment cannot be undertaken now, when the recent drought has caused great economic hardship, and has probably obliterated much of the gains from the two contrasting strategies. Sustainability of Sector Institutions and Project Facilities 52. Because VRA provides electricity to VALCO and CEB, and earns foreign exchange, and because the Authority enjoys an autonomous status which allows it to retain the foreign exchange income to meet its needs, VAR is in a much stronger position than ECG to operate, relatively unperturbed, by Ghana's economic tribulations. But VAR's sustainability depends also on the Government's commitment to respect the Authority's autonomy and permit it to provide and sustain the necessary environment for employing and retaining talents. 53. In contrast, ECG is a weaker organization. It cannot secure the necessary self-reliance partly because of institutional weaknesses. But it is also weak because of the inability to provide necessary incentives for well-qualified staff. To strengthen ECG and give it a status similar to that enjoyed by VRA, an ECG-VRA merger 23/ has often been proposed. If the proposal is carried out a unified sector would emerge that could promote ECG's sustainability provided, however, that the Authority is not weekened in the process. Also with a strong ECG and the implementation of rehabilitation and reform package outlined in this report, the Third Power Project facili- ties would be sustainable. There will still remain the need for a much better access to some service centers by improving the road network and by providing ECG with foreign exchange to acquire the requisite transport facilities. III. CONCLUSIONS 54. Despite great difficulties caused by a depressed economy and political unrest, Kpong was completed without excessive delays or cost overruns (PPAM paras. 5 and 7). The Authority, its consultants, and contractors are responsible for this remarkable achievement. 55. In contrast, ECG did not do well (PPAM para. 10). Dicruption in the economy was the main cause of the Third Power's implementation delay. Besides, the undefined relationship between ECG's foreign and domestic consultants caused problems (PPAM, para. 14). The Corporation has now recognized the advantage of retaining one consultant while providing for sub-contractual arrangement with others. Also, ECG's decision to make electrical contractors responsible for the associated civil works seems to have worked well (PPAM, para. 15). 56. Because of the drought and supply disruptions to VALCO and others, VRA's financial performance has fallen short of the SAR assessment. ECG's 231 Shortly, the Electricity Supply Board of Ireland will undertake a "Study of Power Sector Organization and Management in Ghana". - 16 - financial problems were, however, partly institutional. Despite the weakness, considerable tariff increases have been authorized in recent years (PPAM para. 28). Still, ECG's finances remain precarious, partly because the Corporation is short of foreign exchange resources to enable it to rehabilitate its distribution facilities (PPAM, para. 18), and extend service to potential consumers in the existing service areas. Also, the uncollected bills, which often fall for more than six months behind payment dates, present a serious problem. Such bills need to be promptly collected, and penal interest rates be imposed on the delinquents. 57. Ghana has suffered from severe drought and because of it the utili- zation of Kpong's additional generating capacity will depend on water in- flows. Aside from this problem, the audit considers that Kpong was not needed for the domestic market, provided supplies to VALCO and CEB could have been frozen at the 1976 levels. This means that the project's justification depended on sales to VALCO and CEB. But power rates paid by these agencies were not enough to justify Kpong (PPAM, para. 32). 58. The Kpong project illustrates the tenuous links that bind some country economic and project works in the Bank. Ghana's stagnant economy required that only the very high priority programs should receive the funds spent on Kpong (PPAM para. 29). 59. Further, given that the distribution system suffered from an acute shortage of spare parts, the Third Power Project, a least-cost solution, was no longer least-cost when it was being undertaken. Although the ECG PCR estimates the return on investment to be about 36%, this high rate is delu- sory: Because of the continued distortion of foreign exchange rate the proj- ect's foreign cost has been grossly understated. It is also uncertain whether the tariffs, measuring the benefits, could be maintained, in real terms, under the prevailing hyperinflationary conditions. Moreover, many facilities, constructed under the project, are currently underutilized. The low utilization rate would improve if ECG could acquire critical supplies of materials to extend service in existing areas. The utilization rate could also improve as a result of the overall economic recovery and growth. Hence, the Third Power's justification depends on system rehabilitation and develop- ment, and ultimately, on the state of the Ghanaian economy (PPAM, para. 34). 60. However, more should be done than merely await the outcome of the economic recovery. The reform measures, outlined in this report, include: - A balanced development program for the system's generation, trans- mission, and distribution facilities (PPAM, paras. 36 and 37); - adequate provision for foreign exchange supply to ECG so that it could acquire materials for the improved utilization of distribu- tion network, and for the efficient operation and maintenance of the system; - an effective training program for ECG staff (PPAM, paras. 39-42); - 17 - - revised tariffs securing f or the sector a steady improvement of financial performance and reflecting gradually the futcre cost of power supply to Ghana (PPAM, paras. 43-45); - revised incentive package, particularly for the ECG staff (PPAM, paras.46); and - a unified sector approach with measures to safeguard and reinforce the sector's autonomy (PPAM, paras. 47-48). These measures should be a part of a sector rehabilitation and development package under the current Economic Recovery Program. 61. The World Bank's lack of operational flexibility is of concern. Given the weakness of the distribution system and the sluggishness of demand, it should have been possible to redirect -equipment and materials to maintain existing facilities rather than carry out the extensions in full"-ECG PCR, para. 9.02. 62. VRA's autonomy, which has been the chief reason for the project's technical success, remains so far intact. The Authority's foreign exchange earnings, and its status as the country's most important parastatal company should help to reinforce it. 63. The Ghanaian economy benefits from VRA operations mainly through the distribution of the electrical energy to the productive and other sectors. But because of ECG's weaknesses these benefits have not been fully realized. Without a unified sector, and the establishment of a strong distribution entity, capable of extending and administering an efficient and reliable distribution network, the formulation of a well-balanced and well- integrated expansion program will remain distant goals24/ (PPAM, para. 48). 64. The cofinancing package, exclusively from offical development assistance, has been to Ghana's advantage, since many agencies adhered to ICB procedures, and the grant element in the loans was high. 65. The Authority has extensive experience in resettlement programs. Some 87,000 individuals have been moved around over the past 2V, decades, 7,000 people under the current project. The contrasting strategies, employed in the two resettlement schemes, are of interest to many LDCs. They should be studied further. 66. Two other lessons which merit emphasis are that: - for hydro projects in the Sahel, extensive drought periods must be considered in structuring the hydrological design and operation criteria, and 24/ According to VRA, "A strong and balanced electricity sector can still emerge under the two separate organizations. What is necessary is to improve and enhance management skills of the existing two institutions." - App. I, p.5. - 18 - - in economies where growth prospects are uncertain, more emphasis should be placed on ensuring full utilization of capacity of existing facilities, rather than investing in extensions of such systems. - 19 - ANNEX 1 Table 1: KPONG'S INCREMENTAL COST OF ENERGY SUPPLY (IN MILLIONS OF 1982 DOLLARS) /a Present Value Investment Operating Total Sales Net (12% Discount Cost Cost (5.13 USJ/kWh) Benefit Rate) 1976 1.9 1.9 (-) 1.9 (-) 1.7 1977 2.4 2.4 (-) 2.4 (-) 1.9 1978 27.1 27.1 (-)27.1 (-)19.3 1979 67.5 67.5 (-)67.5 (-)42.9 1980 68.1 68.1 (-)68.1 (-)38.6 1981 70.2 70.2 (-)70.2 (-)35.6 1982 36.8 36.8 (-)36.8 (-)16.6 1983 - 2.7 2.7 46.2 43.5 17.6 T I I I I 2015 2.7 2.7 46.2 43.5 0.4 Total (-) 0.3 /a According to the SAR estimate on Kpong, the project was expected to cost $225.0 million (SAR, para. 4.08). This estimate excludes transmission lines related to ECG's operation, and also taxes, duties, and interest during construction, but it includes $48.7 million for price contingency and owner's cost. Without this item, Kpong project cost, at mid-1976 prices, would have been $176.3 million, or at 1982 prices (year-end prices) $274 million. The latter estimate assumes a 7% annual increase in cost from mid-1976 to end 1982 (6-1/2 years), but no cost overrun, in real terms. The expenditure pattern given in the table follows the profile shown in VRA PCR, Annex 7, p. 3. Note: The table shows that a tariff level of at least 5 US 4/kWh, in 1982 prices, is required to secure a 12% rate of return (the opportunity cost of capital) on investment in Kpong. - 20 - ANNEX 2 Table 2: PROJECTED DEMAND/SUPPLY FOR ELECTRICAL ENERGY (GWh) DEMAND 1976 1981 1986 I. SAR Scenario VALCO and CEB 2,873 3,642 3,642 ECG and others 1,309 2,042 2,993 Losses 142 216 285 Total 4,324 5,901 6,920 II. Alternative Scenario VALCO and CEB 2,873 2,873 2,873 ECG and others 1,309 2,043 2,993 Losses 142 187 252 Total 4,324 5,103 6,118 SUPPLY Akosombo and diesel sets 5,600 5,600 5,600 Shortfalls under: Scenario I - 301 1,320 II - 518 Kpong 970 Shortfall under Scenario I - - 350 Source: SAR scenario and supply estimates are based on the Appraisal Report No. 1299b-GH, Annex 11, Table 2. Note: This table shows that if sales to VALCO and CEB could have been frozen at the 1976 level, and if the domestic demand had grown as envisaged in the SAR (8.6% a year), an overall power shortfall of 350 GWh could have been foreseen for 1986 in the absence of Kpong project. But if demand was projected to grow 3-6% a year, rather than 8.6% a year, a surplus of 300-900 GWh would have been forecast for 1986. Since PPAM, para. 30 questions the very basis of domestic demand projection, Kpong justi- fication for early construction depended on sales to VALCO and the domestic market. -21- ANNEX 3 Table 3: ECONOMIC RATE OF RETURN FROM SALES TO VALCO, CEB AND ECG (in millions of 1982 dollars) Present Bene fits Value of Net Total VALCO /b CEB /c Net Benefits, at Year Cost /a 315 MW 55 MW 30 MW 20 MfECG /d Total Benefits 7.5Z discount rate 1976 1.9 10.1 10.1 8.2 7.63 1977 2.6 9.4 3.6 3.3 2.9 19.2 16.8 14.54 1978 27.1 8.8 3.4 3.1 2.8 18.1 (-) 9.0 (-) 7.24 1979 67.5 8.3 3.2 2.9 2.6 17.0 (-) 50.5 (-) 37.81 1980 68.1 7.7 2.9 2.7 2.4 15.7 (-) 52.4 (-) 36.50 1981 70.2 7.2 2.8 2.5 2.2 14.7 (-) 55.5 (-) 35.64 1982 36.8 6.7 2.6 2.4 2.1 13.8 (-) 23.0 (-) 13.86 1983 2.7 6.3 2.4 2.2 2.0 12.9 10.2 5.72 1984 2.7 6.3 2.4 2.2 2.0 12.9 10.2 5.32 I I W I I I I I 2001 2.7 ' ' ' ' 6.0 18.9 16.2 2.47 2002 2.7 6.3 2.4 2.2 2.0 12.2 25.1 22.4 3.18 2003 2.7 6.3 2.4 2.2 2.0 18.7 31.6 28.9 3.81 2004 2.7 6.3 2.4 2.2 2.0 25.4 38.3 35.6 4.37 2005 2.7 6.3 2.4 2.2 2.0 26.5 39.4 36.7 4.19 2016 2.7 6.3 2.4 2.2 2.0 26.5 39.4 36.7 1.89 TOTAL (-)0.53 /a As shown in PPAM Annex Table 1. 'b (1) 315 MW x 8760 hours x .95 load factor x .24 USS t/kWh, i.e., 2,621 GWh sales for $6,291,432 at 1982 prices; (2) 55 MW x 8760 hours x .95 load factor x .50 US t/kWh, i.e., 482 GWh sales for S2,409,000 at 1982 prices; (3) 30 MW x 8,760 hours x .86 load factor x .975 US t/kWh, i.e., 226 GWh sales for $2,203,500 at 1982 prices. In all the three cases, the figures for the period 1976-82 were increased 7% a year to convert them to 1982 dollars. Similarly, one could discount benefits beyond the year 1982 by 7% a year so as to present them in 1982 prices. The table has not done this, which means it has overstated benefits as they could have been foreseen In 1976 when the project was being finalized. Since, as explained in footnote /d to the table, domestic sales have also been given reasonable values (UST7kWh), the ensuing economic rate of return is not an underestimate. /c 20 MW x 8760 hours x .8 load factor x 1.4 USt/kWh i.e.140 GWh sales for 1.96 million dollars. The table assumes that the rate can be maintained in real terms. Also, figures for the period 1976-82 were increased 7% a year to convert them to 1982 dollar. /d It would probably have been reasonable to assume, in the Kpong/SAR, that Ghana's domestic power market could expand by about 3% a year, as the ECG service gets gradually extended to the unconnected customers. Such an assumption does not rest on any significant economic upturn. The SAR could then have projected the use of Kpong power in the domestic market in the year 2001 and beyond. The ECG column in the table tries to capture the benefit of such sales at USt5/kWh at 1982 prices. Of the 900 Gwh projected power sales from Kpong in the year 2005, the following could have been assumed: ECG sales, 530 Gwh; sales to VALCOr 230 Gwh; and sales to CEB, 140 Gwh. -22- C1IMITS FRR THE BORROWER APPENDIX I VR's C2f2M Page 1 of 6 Preface OED Notes page iv Footnote 'a' seems loaded. A shadow exchange could be used if Footnote revised tne official rate. ad,Listeu twice ouring project execution is found unsuitaole. The total US dollar cost of the project shown as US $Z5U.b million is understated. Even if the local component is converted at Q27b/uS collar it puts total project's cost at US $2ol.7 million. Page viliparagraph 4: VRA's sale statistics up to 1981 with tre exCepLtiVA of 1979 do not Indicate Loat mere had ueefn a fall in sales. Ttie last sentence is also not crue. Page ix paragrapnt 3: Full utilisation of Cpong plant depends moure on inflows tnan Revised what has been Stated. Paragrapni 4: deferral tu approve *tdriff increases' cuulu nut be attriouted Revised to inefficiency in Eu system only teven if there iS anyLning like that) out also to rigidity in the Governmental machinery. Page x: Paragrapn 1 fails to take cognisance of the oevelopment in Ghana True, rld Development Report (VM).1979 in early 1970s. The economy was not all that depressea. Sales to shown that the Ghanaian GM- grw 2- a year ECG grew at the rate of 7% p.a. between 1971 and 1978 while during 1970-76 (Table Z,p.78, Item 55). industrial sales grew at the rate of 4.4% p.a. between 1973 and The 1979 MDR Indicates however, a GDP 1978. Total generation of 5341 gwh in 1981 was 10% below the consul- 0eowth of only 0.4Z a year between 1970-78 tant's projection as per their Nay 1975 report and only 1Z below the (Table 2, p.128, Item 41). Both reports Akosombo plant's originally assessed firm energy capability level confim a declining per capita income, *4ui gwh. which gmgests need for extreme caution In rne Valco option to increase its cenand to 37U ?W could not be rejectea without maur interniational repercussions. Not according to the term of original rejeted ~ithuL mjurmester Agreement (PPAM. para. 24). bimilarly to limit LEIS to the 197b level would affect the AccordiuZ to PPAM if CEB and VALCO operation of the regiunal CIMAu project in rugo, required additional power they should have alreed to pay for the Incremental cost of sup.uly (PPAM. para. 26). The reassessed firm energy capability of the Akosombo plant and -Me SAR couldnot foresee the occurrence recent drought further justified the need for Kpong. of severe droughts, since they were un- precedented. But the substance of the comment is already reflected in PPom, para. 33. -23- APPElIX I Page 2 of 6 Paragrapn Z: OED Notes There is no statistics to confina near collapse in the 'The first two lines reflected in the PPAN. The distriDutiR systeus nor "sluggishness of oemand" at least up to EC PCR Annex 6 must present benefits and costs 1978. Furthermore to say costs have been grossly unuerstatta because in constant prices. Since it has presented bene- of exchange rate uistortions is incorrect. uvervalueu cecii will .fits in terms of 1984 tariffs (1.98 g/dih),.all inflate total costs in US S. costs and benefits most also be in constant 1984 prices. This means the aggregate investment cost The 'exchange rate" issue is being overplayea. In economic of t95.4 million (1977-82) in that Anner is about analysis or evaluation if one is not happy with a given exchange rate US1.9 million. at the official exchange rate and all one has to do is to use a shadow rate instead- much less at the 'free' market rate. In contrast, the 'Third Power has alone used up $26.7 million (ECC PCR, para. 3.17) at current prices and still more in 1984 prices. Tariffs to reflect future costs of power and sector coordination have always been sector objectives. The revised PPAZI version reflects these objectives. Page xi: Paragraph 2 is not quite clear or it is not understood. What wormiated and explained in PPAM, para. 47. is intended to be put across should he precisely stated. Project sumimary: Paragraph 1, line 4: representing a sharp break with tne past" appears loaaeu since Revised projections are oasej on historical ors. Actual kn sales to ECG from 1971 to 198 showS a gruwth r-, if about 6.4% per annum as against a forecast of 9.!%. rio, ., oetween 1971 anu 1978 the increase is sales was 7-.b per annum. Deviations from the forecasteo trena com.lu oe explainea in terms of socio-econuianc anu political aisturb-snces isu the country after 1978. Paragraph b lines Z an 3: Consent not completely valiu. Full utilisation of Kpony oepenos Revised on inflows. Pe:ragrapn a line 5: Revised Network extension took 5 - 28 months longer. Paragrdpn / starting from liue *:: oecause of excnange rate adjustaents it is not meaningful to Footnoted discuss tne local cost overruns" It is felt tnis coulo De none Lhrou9h inuexalui i-e. comparsons made in "fl0b" Cedi!. Oaragrapn lu last sentence. tais is not true. snuula be deleteu. Deleted Paragraph 1Z: Ceoi/US $ relationship neeas to be streamlined. At best Shdauw Footnoted excndnge rate couo be used if the official rate is not acceptea. APPENDIX I Page 3 of 6 Supplementary Comments OED Notes Paragraph 19 line 3: Footnotes to PPAN paras. 19 "Several successive droughts" and 20 refer to VRA.s Paragraph 20 line 7: ents "several successive years of drought" These two phrases look slightly misleading in terms of 1976 - 78 and 1982 - 83. The 3 consecutive drought years in the 70s. 197o - 1978 had inflows of about l hAF. The worst drought years were 1982 and 1983 with inflows of l0.5 MAF and 7.2 MAF respectively. - Paragraph 23 line 1: Revised "37b M" to be changed to "3/U hW" Paragraph Zo: Pricing of supply aoove 31bM was pre aetermined by the Valco Footnote to PPAX, para- 26 Agreement hence an introauction of a new concept sucn as incremental refers to VRA's coments. pricing of energy will never be accepted in the ls. However the events of the early Sus including me reassessed firm energy capability of Akosombo made Valco to accept a power ceiling of 315 mw hence an implied marginal cost of pricing of supply above Alb MW. Paragrapn z7: CEB's rate of 2.5 US Cents/kwh is more than half of Kpong price Revised of aoout 3.5 US Lents/kwh jV(A's estimate). From April 1, 1985 CES's rate Decomes 4 US Cents/kwn for supply up to 50 gwh and 4.5 Cents/kwh for supply aDove 5UU gwh and adjusted for annual inflation. CEB's rate therefore fully relfect the marginal cost of Kpong after April 1, 1985. Paragrapn 28: The objective of the study referred to from line 5 is to move Footnote to PRAN, pars. 28 tariffs towaros long-tern marginal cost of supply. however the refers to VRA's comments. development projects included in the study have not yet been fully implementeu or chances if impleienting some of tnem i.e. dui Project Fre far remote. Project Justification Paragrapn 29: Partly revised. The paragraph refers to the Bank's rather This paragraph is very unfair. It depicts a clear ignorance of than tle Ghanaian assessment Ghanaian economy and sector economics and statistics. Hence it neeas procedure. De deleteo or rewritten to give credence to those involved in planning and execution of the project. APPENDIX 1 Page 4 of 6 Paragrdph 31: on votes There was no Statistical Dasis to confirm nuat kpung was not See footnote 17, to PPAM, pars. 31; alan Note con- required for the utwlestic marKet. AdtiUnale to lsrait Valto au LEo cerains DPAK, Anex Table 2. to 3o* 4h is also not given. A2l costs am expressed in constant price The Tdole I states prices in 1982 levels and not 196U levels. At year 1982 han been chosen as the base period. This U5 Ltents/kufh,tnasutchoice has nothing to do with the opportunity cost discounting should start from 1983 and not 197b as per Table 1. paitr. -Paragarapnter. Paara z.: PEWZ, pars. 32 is revised to reflect VR&'s view. An e3mlained in a Note to PPAM, Annex Table 2 ,VRA Vico's wi I 3rugness to pay nighter rates is uue maily to SAX could not establish that the domestic market pressure from the baraldn public and Ln internetional communaty. needed lpour power. Also see PPAK, pars. 30. It is wrong to apply Valco rate to Kpong power. This paragraph also needs to oe rewritten to ortug anti proper perspective sector econordicS ana statistics. The argisment is similar to the one presented In Paragraph 3: OED Notes concerning VRAs coments on PPAM, pars 2 and 34. As an Illustration. VRA PCF, Annex 7 It will be worthwhile to explain why 'Ib rate of return" ile gives the distribution coat of li5.4 milion (the j as% CluestlunaOle sauce tais relates to the entire sector. se as 1o ca.t Annex 6). which must be in con- stant 1984 prices, since benefits are expressed in terms of 1984 tariffs. But the distribution cost would O2y be sout $2 million at the prevailing exchange rate, and much less at the 'free market rate. True assertion that 4pong is not needed for the domestiC inart VA has not presented a convincing argument to starting fran lisie Ij u LLbhs rateb not riign enuua Lu ijustify Its justify the project. CunstruciOn SIS not acceptable. We accept ebung power as not dteano for Valcu at leastL or. true DdSI5 Of Te earSLumn hIco tariff. 2sraalrapti P: .r9ument acing auvancea .aijst imp lementation of tue Tnira -Mcoidiv g to several reports. oC. and the audit Power Project in favour Of uiverLing resources. to procurement ofmisosfel trp Cishrtfvhce. spare Parts de onstrae at lack of adequate knOwlege of the sector in mes rs nesf Pre, e6s, htc. It m istbue general and 11n particular distributive netwoirk. Lonters ofte18 uan ers. e The r oe d istribu s eao grate, under muth led atte Pree' mk rTe syster was not Suffeing frn an acute saurEage of spare a hatlo pr ta n c au t Pans at tne time of appraisal. The Third Power Project was oesi en to replace more expensive diesel generation. Tor DVa IS Of lea tu"1 trtai on trisuen Do5f Of% ilne exsLn) 1aCo rtf T enti on q uvances retinst iementa of the Tlie ? cco At the Prevailing official exchange rate (tag:US notclaran cnfuin LntjostfdCto reviews). $1). the E=C tariff of 1-98it/kuh. ashattwn in Pu i prct i vourd of auv n mres tC15CIfoocur-en o PCi Annex 6. is about C US c/kWh which does s ri p s demonstrar ar tcOiea l acot of vqPay for e r po s , incretntal cost of co l nd i ns par tivcul d strlDutivesupply (5.3 tio c/s m) - PuAn Annex Table . At the 'free' market rate the retail tariff is 1robobpy 1.3 - 2.0 US c/kWhi retrieving at most 4Th Of EPOng' Incremental cost of supply. This meas that the existing ECC tariff does not pay for any part of the incremental cost of distri- Pararaph49:bution under the Third Power Project. IIs'argapraugrapln 49:tLRChe1{$OS"-R file tenlIS Of the prar i Aoencies rfiage l eca - ea yearst InCIL141119 a D-year grace averlUU St a) auual thrE,Cr CoSa of 1.9a/b as EPCA and PCR favor a flexible approach to under- taking projects. For examle, it should have been possible to revise Power Three and shift funds for the acquisition of spare parts Instead of proceed- ing with the extension of distribution network under the project. - 26 - APPEDIX I Pag. 5 of6 GED Notes Paragrapn 56: ECG's finances lime 6) is now more of collections than lack of Revised foreign exchange to maintain or rehaillitate the systen. Paragrdpn 57 : Full utilisation f Kpong-s additional generating capacity neeus evised not De seen in terms of "several years" out rather in terms of "inflows". Audit's assertion that Kpong was not needed of the domestic This befit could not be foieseen when the marKet is not acceptea oecause it falls to give congnisance to the project was appraised. reassessed firms energy capabilities of tie two systems nob. put at 479U guh per annum. Similarly comments on Valco and CEs are not valid. Varagrapn 58: Inferences not valid eitner now or at the time tne project was See OED rate concerning VRA's coments on PPM appraised. para. 32. Paragrapn 59: The systens never experienced shortage of spare pdrts on the EG PCR, EM, end World Bank consultants do not level ueing imagined in Las report. agree with this position. Trying to seek refuge ai exchange race levels nas been As illustrated is OED Notes concerning VRAts overdune. More positive suggestion as to now to resulve the exchange cnts on WAN paras. 2.33 and 34. the race issues will be more beneficial to the review. race significantly distorts the picture. Paragraph 60: Tariff pulicy tu reflect marginal cust of suppli has been PPAN, pares. 43-45 reflect the marginal cost accepted in principle. objective. Kevise pay scale Shuuld tean a total compensation package to - ditto - incluae housing ani transportation provisions as well as allowances reflecting responsibility, inducement, environment etc. Paragrapn 61: Loment not relevant ano not fair Lu the teami wnicn took part its See OED Notes concerning VRA's coments on PPAY. the appraisals. If tne appraisals were to ue uone in 19dk para. 29. Conclusions would have Deen different. Paragrapn 62: COMwent tiat several VWA senior officials have retirea or went deleted on prolonged leave partly due to erosion of salary scales is not correct. Furthermore accidents leading to freezing of Valca's potlines in deleted 1978 vusAot really oue to unrest amiong emp loyees. Paragrapn 6of A strong and balanced electricity sector can still emerge under Icluded in PPAN, pare. 63. th' two separate organisatiuns. at is necessary is po improve aaip eSnhance management Skills Of onie existing two cimstitutins. - - APPENDIX I Page 6 of 6 OED Notes Table I All costs are expressed in con- Costs are in 1982 prices while Giscounting started from 197b scant prices. The year 1982 has instead of 1983. been chosen as the base year. This choice has nothing to do with the opportunity cost of capital, (the 12 discount rate). APPENDIX I - 28 - COIIENTS FROM TRE BORROVER CG's comments dated Nay 1985 OED Notes 1.05 Cmrrenty, E q sugplie about 900 GUh to nom 196,000 rs ential, 35,000 o~oia, an 272 imdalurial pr~ses. PCR revised. 2.08 (b) Ipova.nt of distribution (415/24OV) n pertloar in Aoa~md%a. PCR revised. am11S IN P J! 3.02 Se projects involvrhg the extension of Oupp~is to alt Refina y Reference in the PCR ar huss, Brewery in Accra and a SteeWll in ilmasi were at deleted. 3.1 1 aIggg "Represenatives of t ~ Consultant returnd in 1983 fr the ~nergiztion of the Sefri Wiamo - Bibi~n -scheme." Dh Coiit==Uint < of the subtation was dom by EG staf. Deleted. 3.21 0e Aboso Glass l~oory was mupplied with p~.r duri the period TRA curtai3ed its supply to wa. PCR revised. 3.28 8Toih in 19E5 Govenmnt was able to grant wG a 3arEe impart liee. EOG 6ould apt a itelf of this opportunity because it was short of esh." mis stat.snt s =t oo=ret. It shOud reVC "tho1gh in 193 PCR revised. Govnt s able to grant EG a largr import lan~, EG oould at utilis the whoe ien=, beosme the ba~ could mot establih the acessary Lttes of Credt." - 29 - APPENDIX I ZCZC DIST9935 JUS 0492 COlMdNTS FROM COFINANCIER OEDOD REP: TCP MET CoWAIT FUND) JWS 492 ZJR641 IN 07/07:02 OUT 7107:07 REF: KFICEN/1681 7.5.1985 FROM: KUWAIT FUND FOR ARAB ECONOMIC DEVELOPMENT, KUWAIT TO: WORLD BANK - WASHINGTON. D.C. U.S.A. ATTN: HR. YUKINORI WATANABE, DIRECTOR OPERATIONS EV4LUATION DEPARTMENT THANK YOU FOR YOUR LETTER DATED APRIL 15.1985 AND THE ATTACHED COPY OF THE FIRST DRAFT OF THE PROJECT PERFORMANCE AUDIT REPORT ON GHANA KPONG HYDRO-ELECTRIC POWER PROJECT AND YOUR THIRD POWER PROJECT. WE HAVE REVIEWED THE SAID REPORT AND ARE PLEASED TO SEE THAT KPONC PROJECT WHICH THE FUND HAD PARTICIPATED IN ITS FINANCING, HAD BEEN COMPLETED SATIS- FACTORILY AND AT A REASONABLE COST. WE ARE ALSO PLEASED TO NOTE THAT THE CONTRIBUTION OF THE THREE ARAB FUNDS (KUWAIT FUND, SAUDI FUND AND BADEA) IN THE FINANCING OF THE FOREIGN COST OF THE CIVIL WORKS OF THE DAN AND THE CONCESSIONAL NATURE OF THIS FINANCING WAS HIGHLIGHTED, INCLUDING THE ADDITIONAL FINANCING PROVIDED ST KUWAIT AND SAUDI FUNDS. WE ALSO NOTED THAT YOU RAVE ELABORATED AT LENGTH ON ALL ASPECTS PERTAINING TO THE SUPPLY AND DEMAND FOR ENERGY AND THE GENERAL TARIFF AND THAT APPLICABLE TO VALCO. TO CONCLUDE, WE HAVE NO OTHER COMMENTS TO MAKE APART FROM HOPING THAT THE VAST EXPERIENCE GAINED DRING THE IMPLE- MENTATION OF KPONG HYDRO-ELECTRIC PROJECT WOULD BE OF BENEFIT TO BOTH BORROWERS AND DEVELOPMENT INSTITUTIONS WHEN DEALING WITH SIMILAR PROJECTS IN FUTURE. KHALED AL-SRALFAN DEPUTY DIRECTOR-GENERAL (ADMINISTRATION) COMENTS FROM COFINANCIER ZCZC DisT1338 JWS 0625 OEDOD (EUROPEAN INVESTHENT BANK) REF: TCP FCA JUS0625 ZJ1B76 IN 10/08:08 OUT 10/08:f1 LUXMIOURG. 10/05/1985 EUROPEAN INVESMENT BANK FOR THE ATTENTION OF HR. WATANABI, DIRECTOR, OPERATIONS EVALUATION DEPARDENT RE: PROJECT PERFoRMANCE AUDIT REPORT: KPONG ELECTRICITY AND THIRD POWER PROJECT THANK YOU FOR SENDING US THE REPORT. PLEASED TO INFORM YOU THAT WE FOUND THE REPORT VERY INTERESTING AND HAVE NO SPECIFIC COMMENT. BEST REGARDS N. URHES - J. NOEL MONITORING DEPARTMENT FOR OPERATIONS OUTSIDE THE COMMUNITY BNKEU LU -05100836 APPENDIX I - 30 - COMENTS FROM COFINANCIER Page 1 of 2 (CIDA) 34 Agowecvwaomede Cxwdgnhans diaftp-nit V I DeuakPrna AgWncy cafat cbma IAS moA66 June 12, 1985 400/10134 Mr. Yukinori Watanabe Director, Operations Evaluation Department The World Bank 1818 H Street, N.W. WASHINGTON. D.C. 20433 USA Dear Mr. Watanabe: Thank you for your April 15, 1985 letter forwarding copies of the first draft of the Project Performance Audit Report - Ghana: Kpong Hydroelectric Power and Third Power Projects supported by Loans 1380-GH, 1381-Cb and Credit 689-GH. MED Notes We are pleased to note that the Audit Team attributes the project's success to the VRA and its consultants and contractors. This is of particular note to us as the main consultants were Canadian (Acres) and funded by CIDA. Comments on the report, particularly at the "Preface" level are as follows: 1. It is a pity that the Team did not visit Tamale (page 11) as it Tamale Was not a part of would have been useful to CIDA to have had the Teams opinion on the two aUted reports. the EG situation in the Northern Region. 2. The mention on page ix. par. 3, that because of the recent Ghana drought, "Kpong's additional generating capacity will be utilized several years later than originally forecast" is, we believe incorrect and contradicted by para 33 on page 14, which states that "construction of the dam has helped mitigate the VR se to suggest that wmved severity of power cuts". It is CIDA's understanding that, had u i, which depends on water In- lpong not been constructed, power outs associated with the f t be foreseen at this stage. recent drought would have been even more severe than was, in he tet baa been revised to read: "m fact. the case. of Kporg's 3. The Teams comments, page 13, that the power rates to VALCO and the export market are "not enough to justify the project" are of interest. ERR's of below 5 and 7.5% are mentioned with the Team stating that "Ghana's depressed economy required only the very high priority programmes should receive the funds spent on Xpong". It would seem that some further revision to the latest VALCO agreement will be required. .../2 Canad - 31 - COIENTS FRON COPINANCIER Page 2 of 2 (CIDA) 4. It was of interest to note that the Team found (page x) when reviewing the ECG project that a "significant" portion of the potential customers in the project are not currently being served because of the shortage of meters, transformers etc.-". 5. Additionally, of importance is the Team's mention in page xii that "for hydro projects in the Sahel, extensive drought periods must be considered in structuring hydrological design and operation criteria". This statement is relevant not only to the Sahel but throughout most of African Savannah regions. In fact, the problem is, in many cases, made worse by the fact that no standard "Code of Practice" exists for the evaluation of hydrological information in Africa. As a result designs have been subject in a number of instances to the whim of various specialists, some with little experience in the very sensitive nature of Africa's hydraulic regimes. Perhaps the Bank could give encouragement to the development of a "Code of Practice for the interpretation and use of African hydro- logical information". (It should be noted that this implies no criticism of Acres' hydrological work, which is generally undertaken on the conservative basis essential to such work in Africa). In conclusion, we found the Audit Report to be a useful document and look forward to receiving the final version of it when it is completed. Sincerely, C. Straby Project Officer Ghana/Regional Program Anglophone Africa Branch - 32 - GHANA VOLTA RIVER AUTHORITY PROJECT COMPLETION REPORT KPONG HYDROELECTRIC PROJECT - LOAN 1380-GH 1. INTRODUCTION 1.01 In Ghana public electricity supply is the responsibility of Volta River Authority (YRA) and of Electricity Corporation of Ghana (ECG), both statutory corporations owned by the Government of Ghana. Created in 1961 to build and operate the first hydropower plant on the Volta River at Akosombo, VRA supplies bulk power to ECG, to the Volta Aluminum Company (VALCO), which operates a smelter at Tema, to several gold, diamond, manganese, and bauxite mining industries, to the Akosombo township, and to CommunautC Electrique du B6nin (CEB), which supplies Togo and B6nin. ECG, established in 1967, is responsible for the distribution of electricity to all other consumers and for the generation of electricity for public supply in areas that cannot be economically connected to VRA's high voltage system. 1.02 In 1956, a commission established -he feasibility of developing the hydropower potential of the Volta river and recommended the implementation of a first stage at Akosombo, basically to supply ECG and an aluminum smelter. In the late 50's, Kaiser Aluminium and Chemical Corporation (USA) decided to sponsor the smelter and in 1962 the Bank and other lenders agreed to participate in financing the hydroplant that VRA was to construct. Thus, the Akosombo project, the establishment of VRA, and the development of VALCO's aluminum smelter are closely connected. The two most important agreements reflecting this interdependence are: (i) the Master Agreement of 1961 between Government and Kaiser which defines the general conditions under which Kaiser would develop the smelter, one of these being Government's committment to construct Akosombo and to give VALCO the special status of pioneering industry, which translated into a series of concessions; (ii) the Power Contract of 1962 between VRA and VALCO which obligates VRA to deliver up to 370 MW of firm power to VALCO and VALCO to pay VRA US mills 2.625 per kWh for the energy actually taken or a minimum charge whichever was the higher; VALCO's payments were to be in USS; the elect-icity rate to remain constant for 30 years. 1.03 The Bank assisted VRA in developing the Akosombo plant (completed in December 1981) and the related transmission facilities (completed in November 1982) with two loans, 310-GH for US$ 47 million in 1962 and 618-GH for USS 6 million in 1969. It also made two credits (118-GH for US$ 10 million and 256-GH for US$ 7.1 million) to Ghana, for relending to ECG, to help improve the utility's subtransmission and distribution facilities. - 33 - 1.04 In 1976, VRA's system comprised: the Akosombo hydropower plant, situated some 100 km northwest of Accra, and equipped with six units with total rated output of 792 MW; its continuous overload capacity is 900 MW, and its long-term dependable generation estimated at appraisal was 5,400 GWh per year; the powerhouse is located at the toe of a 90 m high rockfill dam impounding a reservoir which covers an area of 8,400 km2 and a volume of 61 billion m3 ; - a grid of 162 kV transmission lines concentrated in the southeastern part of the country, six 67 km long circuits connecting Akosombo with the heavy loads in Tema (in particular VALCO) and Accra, a 650 km long loop feeding various substations of ECG and of mines, and a 200 km double circuit line supplying energy to Togo and Benin; - various substations of which the most important is the Volta substation, located some three kilometers from the VALCO smelter at Tema; this facility also operates as load dispatching and system control center. 1.05 VRA is responsible for planning generation and high-voltage transmission, ECG for subtransmissicn and distribution. For planning, VRA normally employs consultants. In 1969-1971 it commissioned a number of studies to determine the requirements for additional generating capacity and the most desirable next addition. These studies compared several alternative sequences of development including, in particular, the Kpong plant dowastream from Akosombo and the Bui plant on the Black Volta, near the border with Upper Volta. The studies concluded that the Kpong development should be further investigated as it seemed likely to be the most economical choice. 1.06 At this stage, the Bank became actively involved in the project that was ultimately to become the main component of the complex operation asso- ciated with Loan 1380-GE. The present report discusses the preparation and implementation of the Kpong hydro project and of its financing, based on the consultant's completion report, the IBRD supervision reports and files, as well as the findings of recent missions to Ghana carried out by a Bank financial analyst and a consultant. 2. PROJECT PREPARATION AND APPRAISAL Origin and Preparation of the Project 2.01 Project preparation started -2 1973, when the Bank assisted VRA in preparing the terms of reference for the feasibility study of the Kpong hydro plant. In 1974-75 consultants carried out this study wi.th a CIDA grant. On the basis of preliminary results of the study, which confirmed the merits of Kpong as the next VRA generating station, the Bank sent its first project preparation mission to Ghana in June 1974. This mission requested additional work in the feasibility study, in particular investigation of irrigation possibilities with the hydro development. The mission also made sure that the environmental and resettlement problems were properly assessed. This point - 34 - was particularly importE 't because the implementation of the Akosombo project had serious shortcomings in the planning and execution of environmental and resettlement programs. 2.02 In 1975, consultants submitted-their feasibility report which not only demonstrated that Kpong was the next project that VRA should execute but that Ghana needed the additional energy by 1981. This deadline left very little time for the many steps that were needed until completion of the project. Therefore, VRA immediately (i.e. mid-1975) commissioned consultants to carry out the design of the plant, to assist VRA in procurement, and to supervise construction and project execution. 2.03 In 1973, the Bank started to assist VRA in contacting other lending agencies for financing, particularly as it had become clear that any project package including the Kpong development would require investments in foreign exchange exceeding what the Bank could allocate for the project. Putting together the financing package took about four years. The decisive turn was the May 1976 meeting of colenders. At this meeting, the following seven agencies confirmed their agreement in principle to provide funds for the project: - the Canadian International Development Agency (CIDA), - the Arab Bank for Economic Development in Africa (ABEDA), - the European Development Fund (EDF), - the European Investment Bank (EIB), - the Kuwait Fund, - the Saudi Fund, and - the IBRD However, it took another nine months to finalize the financing package. 2.04 The main issue which delayed progress of project preparation was undoubtedly negotiations on the VALCO power rates. Indeed, the rate of US mills 2.625/kWh which in accordance with the 1962 Power Contract between VRA and VALCO was to remain constant for 30 years, could no longer be considered equitable compensation to VRA and Ghana for the use of the country's hydropower resource and for VRA's service to VALCO. With worldwide inflation at a level never anticipated in the early 60's, compounded by the 1973 oil crisis, several agencies, foremost CIDA and the Bank, thought that a revision of the Power Contract should be a condition for considering lending at all. While VALCO did not oppose such a revision, its assessment of fair compensation was very different from that of most other institutions interested in the Kpong project, in particular Government and VRA. In late 1973, VRA and VALCO agreed to appoint consultants to study the principles for defining a fair energy price. Though the study was somewhat limited because it only suggested one approach, it helped the parties in their negotiations which, besides the rate issue, were also concerned with the supply of additional power to VALCO. The supply issue was solved in early 1976, but it took until August 1976 for VRA and VALCO to agree on the electricity rates. On the one hand, the revised agreement stipulated that VRA would provide VALCO with an additional 30 MW of firm capacity and 15 MW of interruptible service above and beyond the basic 370 MW to supply a fifth potline, which VALCO intended to commission in 1977. On the other hand, the agreement provided for: - 35 - - a retroactive increase of the basic rate from 2.625 US mills/kWh to 3.25 US mills/kWh for the period 1973-75; - further adjustments resulting in the rate from 4.5 US mills/kWh in 1976 to 5.0 US mills/kWh in 1981; - a temporary rate of 6.75 US mills/kWh for the additional 30 XW of firm and 15 MW of interruptible power; and a review of the rate for additional power about one year prior to the commissioning of the Kpong plant, i.e. -when actual costs of the facility would be reasonably well known; in case the rate resulting from this review were to exceed 9.75 US mills/kWh, VALCO would have the option to reduce or forego altogether the additional power. 2.05 In February 1975, during the above negotiations, the Bank pre- appraised the project and in February 1976, when the prospects for agreement on the rates and on the financing seemed good, it appraised the project which, in the meantime, had been expanded to include further components of VRA's investment program. 2.06 While consultants in their studies of the Kpong project Ied also designed the associated 161 kV transmission lines, other consultants had concluded that by 1980 the system would need additional reactive power compensation and that the best solution would consist in installing a 25 NVAR synchronous condenser at Prestea, the most westernly point of the grid, and of a static condenser of the same capacity at the northernmost point of the system. Italconsult also determined that by 1980 the thermal rating of the two km long 161 kV single circuit lines connecting Tema and Accra would be exceeded whenever one circuit failed and that therefore VRA had to foresee the installation in the short run of an additional circuit. Finally, several export-oriented industries, some existing (gold mines, saw mills), some developing (metal and wood industries), justified the connection of the northern area with the western part of VRA's grid, rather than continuation of the supply with isolated diesel stations. Therefore, VRA agreed to hook 'p the northern region to its grid through an 80 km long 161 kV single circuit line. ECG was to install the associated subtransmission and distribution facilities under its third extension program, which was at the centre of the Bank/IDA operation (Loan 1381-GH and Credit 689-GH) that was in preparation in parallel with the proposed Kpong project. During appraisal, the Bank included the above additional items of VRA's investment program 1976-81 into the project. 2.07 Consultants prepared the design of the Kpong plant and the tender documents so that by December 1975, before appraisal took place, VRA could start the prequalification procedure for the main contracts. By April 1976, firms were preqrilified and, in August, VRA called bids for the civil works 1/ In their 1975 study of VRA's 161 kV grid. - 36 - and turbines, generators, and powerhouse crane. It received the bids in February and March 1977, just in time to integrate the cost information into the project presentation to the Bank's Board of Directors. The Loan documents were signed on March 24, 1977. 2.05 Successful project preparation and appraisal, on a schedule which was very tight, was achieved because of the competence of VRA's management, of which several members had been involved in previous Bank operations and were therefore familiar with the most important Bank procedures. It was also facilitated by the high level of the consultant's professionalism and their familiarity with the working ways of the various agencies involved, which was a decisive factor in getting the project started on time. Project Role 2.09 Already in the 1950's, hydroelectric power was identified as Ghana's major energy resource with an estimated potential generation of some 9,000 GWh per year. AccorCingly, Government designated the development of this resource as one of its major goals. By 1972 a generating capacity of 5,400 GWh/year already existed at Akosombo. The fact that it had been possible to find a customer (VALCO) for a large part of the generation had made it possible to build a large power plant producing low cost energy. This, in turn, permitted the supply of inexpensive power to a substantial part of the country and gave Ghana an advantage over most other West African countries in overcoming the oil crisis. Hydroelectric resources are still Ghana's most valuable energy source as extensive oil exploration has not yet led to the discovery of significant oil reserves and Ghana has no known coal deposits. It was clear, however, that the next hydro plant after Akosombo (Kpong) would be mue costlier. 2.10 The Bank's specific sector objectives for the period 1976-1981 were: a) to increase the firm generating capacity to be able to supply the demand growth beyond 1978, in particular that of VALCO, which expected to start operating a fifth pot line in its aluminum smelter in 1977; b) to carry out the feasibility studies for the generating and transmission facilities required for the years 1983-84 onwards; c) to extend hydropower supply to (i) areas uneconomically served by diesel plants and (ii) rural areas not yet served; d) to extend and reinforce subtransmission and distribution systems in order to improve the quality of service, and to take care of load growth; and e) to rationalize sector tariffs. The Kpong project pursued the objectives (a), (b) and the transmission part of (c). The transmission component included facilities to supply the Sefwi- Wiawso-Bibiani area where important export-oriented industries were developing. The parallel ECG operation, which was supported by Loan 1381-GH - 37 - and Credit 689-GH, aimed at the subtransmission and distribution part of goal (c) and at objectives (d) and (e). 2.11 Through the project, the Bank and the other participating institutions also aimed at helping Government and VRA renegotiate the energy rates with VALCO, and preserving and reinforcing the high quality of VRA's top management, particularly in the financial field. Project Description 2.12 The project comprised: Part A: the Kpong hydroelectric power plant, 24 km downstream from Akosombo, including: - a 20 m high main dam consisting of a 150 m long concrete structure containing intake and powerhouse, a 250 m long concrete spillway (capacity 18,600 m3 corresponding to a I in 10,000 year flood) and a 300 m long embankment; - 6 km long forebay dykes flanking the main dam; - a four unit powerhouse with a total installed capacity of 160 VLW with vertical shaft fixed-blade propeller turbines under a gross head of 12 m. - a 60 km long double circuit 161 kV transmission line to Tema where it was to tie in with VRA's existing transmission grid. Part B: the installation of a 25 VAR synchronous condenser at Prestea and a 25 MVAR static condenser at Kumasi; Part C: the construction of a 80 km long 161 kV single circuit transmission line to the Sefwi-Wiawso-Bibiani area and the substations at both ends of the line; and Part D: the construction of a third 25 km long 161 kV single circuit transmission line from Tema to Accra. As the construction of the Kpong plant also involved the resettlement of the inhabitants and commercial enterprises established in the area which was to be covered by the reservoir, the Loan Agreement (Schedule 1) specifically set out this resettlement as Part E, though in fact, it is a component of Part A. The final agreement among the lending agencies stipulated that from the total estimated project costs of US$ 236.5 million equivalent, VRA would finance local costs (USS 74.2 million equivalent), and the lending agencies the foreign costs (US$ 172.0 million). Tables 3.03 through 3.05 below give details about estimated costs and financing. 3. PROJECT IMPLEMENTATION, OPERATION AND COST Conditions of Effectiveness 3.01 The Loan Agreement of March 24, 1977, specified that, before the Bank - 38 - declared the loan effective, the following conditions had to be fulfilled: - all the agreements concerning project financing by lending agencies other than the Bank should be executed; - the agreement between VRA and VALCO on electricity rates (para. 2.04) should be signed; and - the Bank should have received a detailed program, satisfactory to the Bank, for the resettlement of the people affected by the project. 3.02 While the agreement on the rates, reached in August 1977, was formalized in time and the resettlement program was completed and approved on schedule, there was a minor delay in signing some of the financing agreements. Therefore, the Bank declared the Loan Agreement effective on August 24, 1977, instead of June 24, 1977. Engineering Review Board 3.03 In mid 1977, as agreed during appraisal, VRA appointed four international experts to the Engineering Review Board. These engineers of international reputation were to meet about every six months to review the design of the facilities and the progress of construction and installation. The full Board met three times but individual members provided specialized advice on other occasions. The Board suggested a few revisions to the project; their detailed analyses contributed substantially to ensuring that design and construction were at all times under tight control as far as quality, timeliness and costs of the works were concerned. Revisions to the Project 3.04 The only project revisions of any significance resulted from the initial project review by the Engineering Review Board and consisted of: - the provision of one step-up transformer per generating unit instead of one per two units; - the installation of a double-bus scheme for the 161 kY Kpong substation instead of the ring-bus scheme Acres had originally foreseen, and - the construction of a single circuit 161 kV transmission line from Kpong to Akosombo instead of the second circuit Kpong- Tema. 3.05 The three changes improved reliability at a very low cost. Although calculations tended to show that there was an economic justification for the modifications, it would seem that the modifications essentially were the result of a judgement stressing service reliability. - 39 - Implementation Schedule 3.06 Table 3.01 below shows the construction schedule at the time of project appraisal compared with actual events. The starting month in both cases is that of the invitation for tenders. Table 3.01 Construction Schedule and Implementation of Kpong Plant Appraisal Actual No. of Months from No. of Months from Date starting date Date starting date Call for tenders (civil works) Aug. 76 - Aug. 76 - Contract award (civil works) Aug. 77 12 Aug. 77 12 Call for tenders (electrical and Sep. 76 1 Sep. 76-May 78 1-20 mechanical equipment) Contract award (electrical and Feb. 77 6 Aug. 77-Apr. 79 12-32 mechanical equipment) Start of reservoir filling Nov. 80 51 June 81 58 Commissioning of Unit 1 Dec. 80 52 July 81 59 Commissioning of Unit 4 June 81 58 Dec. 81 64 3.07 The above table shows taat although the call for tenders on the civil works was on schedule, the project was completed six months late: 2-month delay in bid preparation and a 4-month delay in bid analysis and clarifica- tion. One may also conclude that, at the time of appraisal, the consultants had not satisfactorily established the procurement schedule for the electromechanical equipment and hai also underestimated the time required for this phase of project implementation. Part of the delays was also due to the Government instituting i:. mid-1977 a lengthy procedure to review proposed contract awards. However, this had little influence on overall project completion as the critical path was in the civil works. 3.OB Once started, the project proceeded at the expected pace. This was a major accomplishment by VRA, its consultants, and contractors, particularly - 40 - considering the difficulties that they had to overcome in connection with the scarcity of fuel, construction materials and other difficulties with which Ghana was faced from 1978. Unit 1 could have become operational 3 1/2 months earlier had it not been for a modification of the generators' lower bearing bracket that proved to be necessary when its installation was underway. 3.09 Scheduling and execution of the transmission facilities connected with the Kpong plant (161 kV single circuit lines Kpong - Tema and Kpong - Akosombo) as well as the associated switchyard extensions, is shown below: Table 3.02 Construction Schedule and Actual Implementation of Transmission Facilities Appraisal Actual No. of Months from No. of Months from Date starting date 1 Date starting date 1/ Call for tenders July 77 11 Apr./May 78 20-21 Award of contracts Dec. 77 16 Feb./May 79 30-33 Completion Nov. 80 51 June 81 58 3.10 The observations made above concerning the electromechanical equipment also apply to the transmission facilities. At appraisal, the program for the transmission component, which also did not lie on the critical path, had not been studied in depth, and the time allowed between call for tenders and contract award was too short, particularly considering Government's review procedure. 3.11 Project components not directly related to the Kpong development, covered installation of reactive power compensation in the 161 kV transmission system, construction of the 161 kV line to the Sefwi-Wiawso-Bibiani area, and construction of the third 161 kV circuit between Tema and Accra. For these items, VRA contracted consultants for planning, engineering, assistance in procurement, and construction supervision as this firm had already carried out the grid expansion studies. The consultant's contract was signed only in late 1977; most of the delay was due to the Government's new contract approval procedure. This same procedure held up the award of the works contracts and, finally, the unsatisfactory performance of the contractors added to the delays to the extent that the works were completed only in late 1982 (20 to 24 months behind schedule). I/ Starting date, August 76. - 41 - The Kpong Resettlement and Compensation Program 3.12 VRA undertook its first major resettlement program when it had to remove about 80,000 people from the area now covered by the Volta Lake produced by the Akosombo dam. As planning, execution, and follow-up of the move were fraught with problems, the resettlement necessary in connection with Kpong, though it concerned only about 7,000 people, was a sensitive issue. Therefore, VRA, assisted by specialists from the Kumasi University of Science and Technology, the University of Legon, the Lands Department, and other institutions studied and planned the resettlement, taking into account among others the environmental impact study carried out by Acres. In May 1977, VRA submitted to the Bank, as required for the effectiveness of the Loan, a detailed resettlement program to which it committed itself. Ultimately VRA: - resettled 1,089 households and provided them with potable water, sanitary blocks, electricity, roads, church, and school as well as technical, public health, and agricultural advice during the settling-in period; - prepared 1,140 ha of non-irrigated land for the resettled ;eople; - compensated the former owners of the lands referred to above with 1,175 ha of non-i-rigated replacement land; and - developed 607 ha of gravity irrigated land as compensation for flooded similar land belonging to Ghana Sugar Estates Ltd., to Agriculture Development Company Ltd., and the University of Ghana Agricultural Research Station Kpong. 3.13 The resettlement operation started in 1978. In December of that year, VRA removed the people from the areas where the civil works had to start. However, progress was slowed by some changes in the concept (e.g. the evacuation of Lower Kpong in addition to the construction of a dike around the affected area), questions raised about the suitability of the envisaged agricultural developments (which the specialists from the :BRD office in Abidjan helped clarifyi, disputes over land ownwership, and the economic crisis associated with scarcity of fuel and construction material. The Bank's intervention and the main civil works contractor helped put work on track again and permitted the completion of the resettlement. Government, VRA, and the affected population and companies seem satisfied with the results. The Kpong Intake for the Accra Water Supply System 3.14 As the impounding of the VoltL river at Kpong was to increase the maximum water level at the intake of the water works, it called for modifications at this intake. In 1979 it became clear that VRA and the Ghana Water and Sewerage Corporation had designed the needed changes. The Bank's advice helped start this work as soon as possible and to complete it before the water rose for the first time over the former maximum level. - 42 - Studies of VRA's Further Development 3.15 Besides the project discussed here, VRA studied a development program extending beyond project completion. These studies concentrated on the identification of new power sources and on the definition of the associated high voltage grid extensions. The main investigations concerned: - the interconnection with the Ivory Coast network with a 215 km long 225 ky single circuit line which, together with the associated substations, the consultants estimated would cost US$ 18 million, of which VRA would have to bear about US$ 8.5 million; - the Bui hydroelectric power plant on the Black Volta with an ultimate capacity of 350 MV (of which 270 MW would be firm) which the consultant estimated to cost US$270 million/1977, of which US$ 190 million would be in foreign exchange; - alternatives to the Bui plant (i.e. a further increase of installed capacity at Akosombo, thermal plants or a hydroplant on the Oti river); a development of about 140 MW of firm capacity of the Oti river may be feasible but only as the addition to follow Bui. These studies have been successfully completed. Based on them the interconnection with Ivory Coast was completed by year-end 1983. The other projects have beer shelved due to poor economic conditions in the country and weakening finances of VRA. Procurement 3.16 The Kpong project called for five types of contract: (i) those financed by Arab agencies and submitted to international competitive bidding under the rules of these agencies; (ii) those financed by CIDA and submitted to bidding only by Canadian firms in accordance to CIDA's standard procedures; (iii) those financed by IBRD and submitted to international competitive bidding under the Bank Guidelines for Procurement; (iv) those financed jointly by IBRD, EDF and EIB submitted to international competitive bidding under agreed rules which only differred slightly from the Bank Guidelines (currency calculation strictly at the rates prevalent at bid opening), and (v) those financed locally and submitted to local bidding. 3.17 The above requirements were less complicated to handle than it would appear, as the three types of international competitive bidding were very - 43 - similar. In fact, VRA does not seem to have experienced substantial difficulties with this arrangement. Nevertheless, it called for a special effort from VRA and the engineering consultants who had to prepare, for each type of contract, the appropriate tender documents, follow the individual procedures and, with Bank assistance, co-ordinate with the respective lenders. 3.18 In all cases, except the local contracts that VRA handled directly, the engineering consultant invited prequalified contractors to bid. As discussed earlier the time allotted between call for tender and award often proved too short. The prequalification procedure for the main contracts (with the longest lead time and the largest cost) took place in the first half of 1976 to allow bidding during the second half of the y-ar. Thus, VRA was able to award the contracts for the civil works, the turbines, the generators, and the powerhouse crane in August 1977; the signing of the contract for the electro-mechanical equipment and later that of several other contracts was unduly delayed by the new procedure the government introduced in June 1977. Until then VRA's Board had had power to award all its contracts for goods and services. The new procedure required VRA to submit the documents relevant to the main contracts to a Public Agreement Review Committee which, in turn, had to refer them for approval to the Supreme Military Council. Thus, the new procedure delayed the award of several contracts related to the project, except for those concerning VRA's grid extension (Parts B, C and D), it had little effect on completion dates. Annex 1 shows the main data concerning the various contracts. Project Costs 3.19 Table 3.03 below summarizes estimated and actual costs of the project *hich are detailed in Annex 2. Local costs were calculated on the basis of the average exchange rates prevalent during the various years of the project period. As the Cedi was overvalued during the project period, the values expressed in US$ may not reflect accurately actual costs. Table 3.03 Summary of Estimated and Actual Project Costs (in 7illion) Appraisal Estimate Actual Foreign Local Total Foreign Local Total US$ USS usS uss Kpong Development, including engineering 162.0 72.5 225.0 187.7 169.1 238.9 Network extension 10.0 1.7 11.5 9.1 9.4 11.7 Total 172.0 74.2 236.5 196.8 178.5 25G.6 Percentage of Appraisal estimate 114% 241% 106% -44 - 3.20 Taking into account the qualifications set forth above, the table shows that the foreign exchange cost of the project exceeded the estimate by about 14%, whereas the actual total cost is not substantially higher than expected. In view of the adverse conditions prevailing in Ghana during nearly the entire project period this result should be considered quite good. 3.21 Table 3.04 below analyzes the variations between estimated and actual costs for the main items, civil works and electro-mechanical equipment. The table permits the fol--wing conclusions: - the original contract amounts are in line with the estimated costs, which confirms the advantage of using bid prices for the appraisal estimates; - contract additions (para.3.04) exceeded by far the allowances for physical contingencies; - escalation was reasonably in line with the allowances for foreign exchange price contingencies; however, as a result of the unexpectedly high rate of inflation from 1978, the local cost escalation exceeded by far the allowance made at appraisal. - 45 - Table 3.04 Cost Analysis of Civil Works and Equipment Contracts (in millions) Appraisal Actual Difference Foreign Local Foreign Local Foreign Local Civil orks $ I Cost/original contract amount 55.5 30.4 59.6 35.1 7.4 15.5 Physical contingencies/ contract additions 4.4 3.1 21.6 10.2 390.9 229.0 Price contingencies/ escalation 14.4 9.2 16.6 78.8 15.3 756.0 Total 74.3 42.7 97.8 124.1 31.6 190.6 Mechanical and Electrical Equipment Cost/original contract amount 45.1 2.7 41.9 3.5 -7.1 30.0 Physical contingencies/ contract additions 6.9 0.3 11.2 0.3 62.3 62.3 Price contingencies/ escalation 12.2 1.6 10.7 5.1 -12.3 218.7 Total 64.2 4.6 63.8 8.9 -0.6 93.0 3.22 The main additions to the project, which the physical contingencies were supposed to cover,occurred in the civil works contracts and amounted to US$ 31.4 million (36%) of the basic contract price and were due to: - changes in the powerhouse superstructure design to allow for higher earthquake loads; - unanticipated adverse geological conditions in a few places; - use of materials (cement, steel, etc.) and equipment imported by the contractor because they were not available locally due to the economic crisis; -46- - major economic dislocations forcing the contractor to incur significant expenditures to keep the work going; - work outside original scope in connection with the resettlement scheme, and the modification of the intake for the water works; and - a 12% increase of the contractual quantities due to a series of mostly minor changes except for the higher cost in civil structures due to higher earthquake loadings. 3.23 Unforeseen items added USS 4.2 million (47%) to the basic contract price for mechanical and electrical services because of: - supply of non-contractual parts for the generating plant; - installation of overhead lines between powerhouse and switchyard; - provision of a telephone system; - supply and erection of all mechanical and electrical installation for the modified water works intake; - final painting of all mechanical equipment; and - VRA's purchase of contractor's site equipment at contract end. 3.24 The addition to the switchyard equipment amounted to about $0.3 million (20%) of the basic contract price mainly because the contractor had to supply switching equipment for a 5 MVA transformer bay, and his installation supervision work substantially exceeded the quantities included in the corresponding provisional sum of the contract. Finally, t]e amount of engineering work was 16% greater than forecast due to increased scope of work and extension of the contract period. Financing 3.25 Table 3.05 below shows the financing packages, as defined at appraisal compared with the actual funding. The conditions for the various loans and credits were as follows: - Arab agencies (ABEDA, Kuwait, Saudi Fund): US$ 73 million for 20 years including a grace period of 5 years at an interest rate of 2 to 4%; - CIDA: US$ 39.0 million for 50 years including a grace period of 10 years, at no interest; - EDF: US$ 10.0 million for 40 years including a grace period of 10 years at an interest rate of 1%; - EIB: US$ 11.0 million for 15 years including a grace period of 4 1/2 years at an interest rate of 8.5%. -47- Table 3.05 Financing of Foreign Exchange Component ---Million US$--- Lender 'r+^&s Estimated Actual (1) Saudi Fund, Kuwait Civil Works Kpong Fund and ABEDA 73.0 90.8 (2) Joint financing EDF, Generating equipnent, powerhouse EIB, and IBRD crane, transmission Kpong to Tema and Tema to Accra, reactive power compensation, engineering, review Board 54.6 55.8 (3) CIDA Engineering Kpong, gates and hoists 39.0 29.7 (4) IBERD Preliminary works Rpong, resettlement, Sefwi-Wiawso, Bihiani, transmission 5.4 5.4 (5) OPEC-Fund Gate erection a/ - 3.7 (6) VRA - 11.5 Total 172.0 196.9 a/ Originally included in (2) 3.26 Actual financing of the foreign exr-bange component of the project cost can be characterized as follows: - in connection with the increase in the cost of civil works the Saudi Fund, the Kuwait Fund, and ABEDA increased their commitment by US$ 17.8 million; - the contribution of EDF, EIB, and IBRD were in the order of magnitude foreseen; - only about three quarters of CIDA's credit was used as the corresponding package did not cost as much as originally envisaged; - the OPEC Fund filled a financing gap left in the package jointly financed by EDF, EIB, and IBRD, and; - VRA had to finance the remaining shortfall of US$ 11.5 million in foreign exchange. - 48 - Disbursements 3.27 Initially, disbursements of the IBRD loan fell behind schedule because of the delays incurred at the award stage of the Kpong contracts. Later, works at Kpong proceeded as quickly as planned, disbursements still lagged about one year and towards the end of construction the transmission facilities were significantly behind schedule. Annex 3 compares the disbursement profiles. Operation 3.28 Since July 1981, when the first units of Kpong started operating, they have run to VRA's full satisfaction with only minor interruptions caused by minor equipment flaws. An inspection in early 1984 showed that the equipment was in very good condition. The only shortcoming observed was a superficial roughness of practically all the piston rods of the turbine governors. VRA consulted the supplier to have this defect analyzed and corrected. The efficiency of the equipment is also satisfactory. Though the output of a unit is nominally 40 MV, the actual power generated easily reaches 42 MV. Since mid-1983, when it became clear that there would be large deficits in the runoff of the Volta river, VRA drastically reduced production to save water. The water level of the Volta Lake at Akosombo is currently below the safe turbine operating limit. Thus, for several months Kpong has generated mostly with only one turbine running. This is likely to continue at least until mid-1984 and will end only if the rainy seasons refill the lake to safe operating levels,which has not yet occurred. After protracted finishing work, the addition to VRA's transmission system (the Accra-Tema line, the connection of the Sefwi-Wiawso-Bibiani area and the power factor correction equipment) has operated satisfactorily. However, by early 1984 most of these facilities had operated for less than a year. Consequently, operational experience with them is still limited. Performance of the Consultants 3.29 The performance of both consulting firms involved in the project, for the Kpong development and for the transmission system additions was quite good, particularly taking into account the difficult economic environment in which they had to plan and help the contractors carry-out their work. After successful completion of the feasibility study, project design, and preparation of the complicated set of tender documents under severe time con- straints, the Kpong consultant implemented the bidding, the tender evaluation and the award procedure without any major problem. The consultant's professionalism and their familiarity with the working ways of the various agencies involved was a decisive factor in getting the project started. Together with the main contractors, in particular the one responsible for the civil works, they conceived and helped carry out imaginative measures to overcome the obstacles created by the near breakdown of the economy. 3-30 The Engineering Review Board suggested, and VRA endorsed, among other matters, the changes in the Kpong consultant's design discussed in paragraphs 3.04. None of these modifications should be considered design corrections. However, VRA - 49 - feels that the necessity of reinforcing the generator brackets, which only appeared at the erection stage and delayed operation of the first unit by three months, could have been detected earlier. 3.31 An engineering firm had great difficulties with the contractors for the additions to VRA's 161 kY network. Therefore, in spite of its efforts, the works experienced long delays. The firm was, however, successful in urging the contractors to remedy the technical shortcomings that had at an earlier stage seemed to endanger the quality of the works. Performance of the Contractors 3-32 The performance of the main contractors for the Kpong plant was generally satisfactory. However, that of the civil works contractor deserves a special mention. Indeed, by taking over or helping complete works (e.g. preparatory works, resettlement, adjustments at the water works intake) that the local contractor could not master under the adverse conditions prevailing in Ghana at the time, the main contractor contributed a great deal to the timely completion of the plant. The high quality of all the works of the Kpong complex reflects the competence of the contributing companies. In contrast, both contractors involved in the reactive power compensation and the transmission lines connecting Accra and Tema and the grid with the Sefwi- Wiawso-Bibiani area, did not perform as well. Their organisation was weak and they lacked the management and motivation necessary to carry out under the admittedly adverse prevailing conditions the technically relatively simple jobs for which they were responsible. Performance of the Borrower 3.33 VRA performed well. It has benefitted from the fact that it had a seasoned management team under a strong leadership which was not notably affected by a change in the Managing Director. The reinforcement of the management by an expatriate during project implementation was also a success. 4. OPERATING PERFORMANCE 4.01 During the entire project period VRA's sales remained below those estimated in 1976. Whereas they were forecast to increase from 4,180 GWh in 1976 to 5,860 GWh in 1982, they actually grew from 4,081 GWh in 1976 to 5,181 G'Wh in 1981 but fell off to 4,799 GWh in 1982 due to the worsening water shortage in the Akosombo reservoir. The shortfall in VRA's sales is mainly due to ECG's own sales remaining about constant from 1978 onwards because of the economic disruption in the country. Furthermore, VALCO did not require the full energy for its five potlines until 1980, instead of 1977 as foreseen at the time of project appraisal. This was mainly due to the breakdowns of VRA's system discussed in paragraphe 4.02 below. Annex 5 shows a comparison of actual load, generation, and sales figures with their 1976 estimates. It also shows the systems losses, which were consistently lower than forecast. 4.02 In 1977, the generating units at Akosombo developed mechanical and electrical problems related to cracks in the speed rings of all six turbines and to stator faults on unit 4. VRA's plans called for repairing the speed ring of unit 4 while replacing the coils of the stator and then for the - 50 - successive reconditioning of the other units. However, two times in a row the system collapsed. In part due to a strike joined by most of VRA's engineers, VRA was not in a position to restore the system before VALCO's smelting pots cooled requiring the removal of the solidified flux and alumina with pneumatic hammers. This took about ten weeks. In April 1978, the system collapsed again, this time because someone manually operated the C02 fire protection. Due to lack of coordination between VRA and VALCO and to VALCO's personnel working to rule, VALCO's potlines froze again, causing again a damage to VALCO in excess of US$ 40 million in repairs and lost production. 4.03 The analysis of the system collapses pointed to several shortcomings in the maintenance of protection and control equipment as well as in the level of training of the operating personnel. This prompted YRA to appoint a Power Equipment Review Board integrated by three specialists of Ontario Hydro (Canada) and the electromechanical specialist of the Kpong Engineering Review Board to inquire about the accidents and to propose corrective measures. Upon VRA's request, the Bank agreed to finance the cost of this new board under the loan. In January 1979, the experts submitted their report recommending: - modifications in the protection system including the installa- tion of automatic load sheding relays; - a complete renovation of VRA's telecommunication systems; and - a special training of control room personnel. These measures, which VRA carried out at a cost of some USS 5 million, were successful. 4.04 The construction schedule for Kpong foresaw the commissioning of the first unit at the end of 1980. However, in accordance with the demand projections, the Akosombo plant could only meet the demand until 1977, on the basis of the average year runoff of the Volta river. Therefore, the operations plan foresaw from 1978 to 1980 an overdraft on the Akosombo reservoir to cover the expected energy deficits with respect to the availability in an average year. As the demand did not develop as expected, and as the accidents referred to in paragraphe 4.02 above further reduced consumption, the drawdown was not necessary to the extent foreseen. Thus, the depletion of the water reserve for operational reasons was limited. Nevertheless, by the end of 1978 the water level in the lake fell below safe operating limits due to two consecutive dry years (1977-1978). In 1979 the runoff was substantially above average and brought the level up well above the design operating limits. However, since 1980, the river discharge has been so low that, even with Kpong fully available from late 1981 and consumption well below the forecast, the water volume was depleted to such an extent that towards the end of 1982 VRA was forced to reduce output drastically. Thus, VRA was not able to supply VALCO. As the drought continued in 1983, VALCO had to stop aluminum production altogether. 4.05 At the time of appraisal, following the appraisal mission recommendation, VRA and ECG temporarily planned to supply electroboilers with interruptible surplus hydroelectric energy. As, during the entire project period, no surplus energy was available and since there is little chance that this would happen in the future, VRA and ECG discussed with the present and potential boiler owners an early changeover to oil firing. This disruption - 51 - considerably slowed down industrial production for which the boilers are needed. 5.0. FINANCIAL PERFORMANCE General 5.01 The following discussion of the development of VRA's finances during the implementation of Kpong project covers only VRA's power activities. VRA carries out the non-power operations as Government's agent. It also owns all the shares of the Volta Lake Transport Company, but it keeps separate accounts for the non-power activities. 5.02 The years 1977-1982, during which VRA carried out the Kpong project, were characterized by a rapid deterioration of Ghana's economy: severe inflation and depreciation of the Cedi. Therefore, the Cedi-figures in VRA's financial statements (Annexes 9, 10, and 11) bear little resemblance to those estimated in 1976, as the 11% annual inflation rate estimated at appraisal was considerably below the actual average of 32%. Furthermore, the official exchange rate greatly overvalued the Cedi, thus significantly distorting US$ equivalent figures. Hence, the following paragraphs concentrate on the expected and actual trend of VRA's main financial indicators, although the unreliability of the values corresponding to the actual financial figures also reduces the validity of conclusions drawn from the indicators calculated. Financial Plan 5.03 With the limitations indicated above, the analysis of the flow of funds (Annex 11) shows that VRA financed about 37% of its 1977-1982 capital investment program with internally generated funds. This result corresponds very closely to the appraisal estimate of 32%. The cost of the actual program amounted to US$ 270 million, against an estimated US$ 380 million; the difference between these two figures stems from the fact that, contrary to what VRA had planned at appraisal, it did not start construction of a new plant (Bui) during execution of Kpong. Financial Indicators 5.04 Table 5.01 below shows variations in the main financial indicators. - 52 - Table 5.01 Actual and Estimated Financial Indicators 1977 1979 1981 1982 Average tariff (mills/kWh) actual 7.8 18.2 35.3 37.0 appraisal 7.6 8.2 9.3 9.5 Sales in GWh actual 4,799 4,524 5,181 4,799 appraisal 5,060 5,510 5,680 5,860 Rate of return actual 0.9 2.0 5.8 1.6 on equity (%) adjusted actual 5.7 6.5 7.3 7.3 appraisal 6.9 7.9 8.8 8.9 Operating ratio actual 0.72 0.58 0.49 0.60 appraisal 0.45 0.42 0.40 0.37 Current ratio actual 3.9 2.2 2.1 2.2 appraisal 4.2 4.9 4.4 3.2 Debt/equity ratio actual 16:84 32:68 30270 2B:72 appraisal 28:72 35:65 38:62 39:61 Receivables (% billings) actual 28 33 44 48 appraisal 20 20 20 20 5.05 The above figures show that the rate of return was consistently lower than estimated at appraisal. The covenant (Section 5.05 of Loan Agreement) called for a minimum return on equity of 7% in 1977, 8% in 1978- 1980, and 9% thereafter, adjusted to take into account the results of the revaluation of assets which VRA was to carry out in 1977. At appraisal net fixed assets at end of 1976 were estimated at about 0 300 million; the above covenant states that if the value of revalued assets were different from said amount the minimum rate would be adjusted in inverse proportion to the relation of assumed to actual revalued net assets. By October 1977 ( four months after the date foreseen in the Loan Agreement) VRA had determined the value of its assets at December 31, 1976 at about 0 370 millioi. Therefore, the actual lower limits for the rate of return were 5.7% for 1977, 6.5% for 1978-80, and 7.3 % thereafter, levels which VRA never achieved. 5.06 The main reasons for unsatisfactory earnings, which prevented VRA from paying a dividend, except for 1977, can be summarized as follows: - Sales were substantially lower than expected and not associated with correspondingly lower costs, this was related to the downturn of the economy as a whole, to the 1977-78 system breakdowns temporarily interrupting supply to VALCO, and to the supply curtailments associated with the drought from 1982 onward; and - 53 - - While Government granted VRA a series of substantial tariff increases, these fell short of YRA's needs. Essentially for the same reasons, the operating ratio was consistently higher than expected. Only in 1981, as a consequence of massive tariff increases, the operating ratio and rate of return came within 25 % of the forecast values. 5.07 The debt/equity ratio stayed lower than forecast mainly because the asset revaluation reserve, which is the dominant part of equity, grew at approximately the rate of devaluation of the Cedi and VRA did not incur as much additional foreign debt as expected, as it did not start the post-Kpong program as originally envisaged. 5.06 Accounts receivables were consistently higher than estimated in 1976, mostly because of ECG's overdues, in turn caused by ECG's inability to obtain adequate tariff increases. Conclusions 5.09 With all the shortcomings highlighted above, it is fair to state that under the conditions prevailing in Ghana during project execution, it was no small feat for VRA to have fared financialy as it has. This, of course, was not only VRA's doing; it was also related to VRA's foreign exchange earnings. Furthermore, VRA took maximum advantage of its special position as the country's most important parastatal company. 6. INSTITUTIONAL PERFORMANCE Management 6.01 All in all, VRA remained well managed throughout the project period. Since 1966, a competent, forceful, and highly respected engineer was the authority's Chief Executive. He retired in 1980 and the former Deputy Chief Executive (Engineering) replaced him. The Bank concurred with the latter appointment. 6.02 In 1977, as agreed during loan negotiations, VRA appointed a Chief Financial Officer, a Canadian, whose services CIDA helped finance. This specialist focused on loan and contractual aspects, costing, budgeting, and tariffs. 6.03 In 1977-78, YRA went through a difficult time when political turmoil penetrated the organization and led to industrial action which was instrumental in the accidents that led to the freezing of VALCO's potlines in 1978. In the early 80's VRA also started to suffer from the economic disruption and general demoralization in the country, as several senior officers retired or went on prolonged leave without replacement. One contributing factor to this exodus is VRA's salary scale which is closely related to that of the civil service. Thus, in recent years VRA's middle management, in both the financial and engineering sectors, suffered serious depletion, as many specialists took positions in Ghana's private sector or went to other West African countries. - 54 - Personnel 6.04 Throughout the implementation of the Kpong project, VRA's personnel remained about constant at some 2,500 people of which 1,250 were carrying out the power activities proper, 850 servicing the Akosombo township, and 400 working in connection with VRA's agency activities (shipping, resettlement, Volta Lake research, etc.). 6.05 In the late 70's a certain unrest came into the middle and lower echelon personnel leading to work-to-rule measures. The situation had improved by 1980. However, demoralization still persists today. Training 6.06 VRA's training has traditionally been very good, as the authority year after year had about 100 people attending courses at local institutions and 10 to 20 overseas. Nevertheless, the three system breakdowns in 1977-78 showed that control room personnel needed more specialized training, which VRA provided. As a continuation of this action, VRA is installing and manning a training center at Akuse (near Kpong) with the financial assistance of CIDA (about US$ 3 million). Merger of VRA and ECG 6.07 Since its inception, VRA has operated rather independently under the control but with little direct intervention of Government. Being a major earner of foreign currency it had and still has great advantages over other parastatals, in particular ECG. It was, and continues to be, well managed in spite of the major economic difficulties it is currently experiencing. Therefore, it is natural that a merger of YRA with ECG, which for years has suffered from sub-standard management, has been under consideration off and on for the past 20 years. In 1976, the Bank concluded that the merger would not bring major benefits to Ghana. However, with the dramatic deterioration of ECG's management in the past 6 years, the issue has emerged again. It is to be analyzed in 1984 as part of a new proposed project. Autonomy 6.08 Until project conclusion, VRA's autonomy, which in the Bank's view has been a major factor for the authority's efficiency, remained essentially intact except for the temporary application of the complicated major contract approval procedure which in 1977-78 hampered project implementation. Covenants 6.09 Annex 6 sets forth the princial covenants in the Loan and Guarantee Agreement. The Annex shows that VRA complied with all covenants except for the rate of return one. - 55 - 7. ECONOMIC JUSTIFICATION Actual and Prospective Consumption 7.01 In VRA's system, energy consumption is the dominant supply factor and not power demand. Therefore, the following considerations exclusively concern energy. Nevertheless, VRA has made sure that power does not become a problem. 7.02 Annex 7 shows actual and projected consumption and generation. The projections assume that in 1985 the water accumulation at Akosombo will have normalized to a point permitting VALCO to start again operations with two potlines and a third in 1986. However, the fourth potline would restart production in 1990 and the fifth only when a further plant will have been commissioned. The projections further assume that the ECG demand will return to the 1981 level until 1987 and grow thereafter at about 3.5% per year. This leads to required generation by 1986 which is less than 60% of that estimated at project appraisal. Generation Attributable to Kpong 7.03 Annex 7 also shows the generation attributable to Kpong which essentially consists of three parts. The first of these is the actual generation at Kpong which, had the plant not existed, could have been generated at Akosombo but only at the expense of a further drawdown of the lake level or earlier restrictions. Second, in the years 1985 and 1966, when the refilling of the lake is assumed to take place, the production assignable to Kpong and Akosombo is assumed in proportion to the installed capacities, again because without Kpong the return to normal operation would take longer. Finally, once normal operation is reached again, only the incremental generation above that possible at Akosombo is assigned to Kpong. As a review of the hydrology of the Volta river is likely to show that the average production at Akosombo and Kpong has to be assumed lower than projected, the present re-evaluation uses three assumptions which assign to Akosombo and Kpong 100%, 90%, and 75% of the originally planned productivity. 7.04 The energy attributable to Kpong could improve if Ghana could export to Ivory Coast more energy than presently anticipated. However, an improve- ment of the Volta river runoff usually parallels a similar improvement in the availability of water in the other West African rivers. Therefore, a considerable increase of the energy transfer to the Ivory Coast would not be likely. Least Cost Solution 7.05 The procedure applied to determine whether the Kpong investment was economically sound, is the same used in the appraisal report, which is to compare Kpong with an equivalent thermal plant (the comparison slightly favors Kpong because the thermal plant could have been somewhat smaller; further refinements were not necassary for such ex post facto review). Comparison of Kpong within a long range investment program was not done because it did not exist then or now. Based on a price for crude of US$ 20 per barrel, which is on the low side for 1981, the equalizing discount rates for the three assumptions used for the average potential hydrogeneration (para. 7.03) are - 56 - between 8% and 10%. With fuel costs at US$ 25 per barrel, these rates increase by some three percentage points. Therefore, the decision to build Kpong was sound but not clearly the least cost solution. Return on Investment 7.06 The return on investment was estimated as the discount rate that equates the present values of the benefits and costs associated with the 1976-1982 VRA investment program, plus ECG's (distribution). Benefits were measured by the forecast revenues from the sales of electricity at the average retail level, usir-g the tariff in effect on July 1, 1984 (see Annex 7). The return for the program is about 16% which compares marginally with the opportunity cost of capital for Ghana, estimated to be between 15% and 20% Project Achievements 7.07 Although the return on investment is unsatisfactory, the implementation of the Kpong project has provided Ghana with facilities that will ease the starting phase of an economic recovery by providing industries and foreign exchange earning activities with electric energy at a still rather advantageous cost. Furthermore, Kpong is an important generating element in the developing regional electricity network presently interconnecting Ghana, Togo, Bnin and Ivory Coast. Bank involvement was also instrumental in helping to limit the negative effect of the economic disruption on VRA's management and thus on the authority's future. 8. BANK PERFORMANCE Assessment of Borrower's Capability 8.01 At appraisal, the Bank assessed VRA's capability in terms of management, project implementation and operation at a high level. The execution of the project under unexpectedly difficult conditions essentially confirmed this judgement. The circumstances leading to the system breakdown showed that there were some shortcomings in VRA which had not been detected. However, the decisiveness with which VRA's management corrected the deficiencies under adverse circumstances again confirms the correctness of the Bank's original judgement. The Bank's Role 8.02 The Bank played a crucial role in coordinating the interventions of the various lending agencies in the context of the project. Not only was this role decisive at the project preparation stage but also at the procurement stage, where Bank staff assisted VRA and the consultants in the fine tuning of the various contract packages. It also helped the financing agencies by providing them with the results of the Enk supervision missions. The Bank also supported Government and VRA in their negotiations with VALCO and thus helped Ghana obtain a fair retribution for the resources used. This support, requested by the Government, consisted of the Bank advising the Ghanaians on the most app:opriate and most experienced consultants needed for the negotiations. - 57 - Supervision 8.03 Annex 8 sets forth the schedule of the Bank's supervision missions. The average interval between such missions was about 10 months and, over the 6 years of project implementation, two engineers, --ur financial analysts - two of them in the same team - and a consultant participated in the supervision. Relations between Bank staff and Government and VRA officials were at all times very good. 9. CONCLUSIONS 9.01 The project did achieve its main objective which was to provide Ghana with an additional power plant and associated facilities utilizing one of the country's -main resources, water. It also helped preserve the efficiency of one of the country's main institutions, VRA. Unfortunately, the unforeseen economic depression has not allowed Ghana to reap the full benefits of the investment in the short term. Nevertheless, Kpong and the related transmission installations will be a substantial asset in the country's expected recovery. GHANA: VRA - KPONG HYDROELECTRIC PROJECT LOAN 1380-OH Main International Contracts and Their Finanoing Contract Scope to Work Contractual Dates Contractor Foreign Financing Tender Tander Contract Call Receipt Award K3 - Civil Works Construction of principal civil 08/76 02/77 08/77 Joint Venture Kuwait Fund, Saudi works, including provision of Impregilo-Recohi Fund, BADEA camp facilities and equipment, (Italy) transportation to other con- tractors and construction of resettlement villages K4 - Turbines, Design, supply and installation 09/76 03/77 08/77 Joint Venture IRD, EDP and BIB i Generators of turbines, generators and Thoshiba-Boving and Power- powerhouse crane (Japon/UK) house Crane K5 - Transformers Design and supply of generator 10/77 01/78 08/78 Oy Stromberg IBRD, EDF and BIB transformers (Finland) K6 - Spillway Supply of spillway radial gates 08/77 10/77 02/78 Dominion Bridge CIDA Gates and hoists, stop logs, including Company Limited embedded parts (Canada) K7 - Powerhouse Supply of powerhouse operating 10/77 12/77 03/78 Canron Limited CIDA Gates gates and hoists, bulkhead gates, (Canada) H trashracks draft tube gates and 1 gantry cranes Contract Status Contractual Dates Contractor Poreign Financing Tender Tender Contract Call Receipt Award K8A - Mechanical Design, supply and installation 03/78 09/78 04/79 Sadelmi IBRD, EDF and EIB and Elec- all mechanical and electrical (Italy) trical auxiliary services, installation Services of generator transformers and switchyard equipment K8B - Gates Installation of spillway and 03/78 09/78 04/79 Sadelmi OPEC Fund Erection powerhouse gates (Italy) K9 - Transmission Design, supply and installation 05/78 10/78 05/79 Sadelmi OPEC Fund Lines of 161-kV transmission lines (Italy) K1O - Switchyard Design and supply of switchyard 04/78 08 02/79 Merlin 0rin IBRD, EDF and BIB Equipment equipment for Kpong switchyard (France) and additional equipment for Akoeombo switchyard N1 - Transmission Design, supply and installation 06/78 10/78 05/79 Energoinvest IBRD, EDF and BIB Line of 161-double circuit line (Yugoslavia) Acora-Tema N2 - Power Factor Design, sipply and installation 06/78 09/78 06/79 Bharat Heavy IBRD, DF and EIB Correction of 25 NVAR sinchronous condensor Equipment Equipment at Prestea and 25 NVAR static (India) condenseor at Kumasi N N3 - Transmission Design, supply and installation 06/78 10/78 05/79 Energoinvest IBRD 0 Line of 161-single circuit line (Yugoslavia) Dunkwa-Akiwinso (Snfwi, Wiawo, Bibiani area) March 1984 Annex 2 -60 - KPONG HYDROELECTRIC PROJECT COMPLETION REPORT Estimated and Actual Project Costs in million units Appraisal estimate 11 Actual Foreign Local Total Foreign Local Total a) Kpong Power Plant US$ 0 US$ US$ I US$ Preliminary works 1.7 4.5 5.6 3.0 4-9 6.4 Civil engineering works 75-3 42.7 112.4 97.8 124.1 133.0 Electrical and mechanical works 61.3 4.6 65.3 63.8 8.9 66.3 Resettlement 1.7 11.3 11.5 1.6 9-5 5.4 Engineering and management 15.9 3.6 19.0 14.8 - 14.8 Owner's cost - 4-8 4.2 - 17.2 5.0 Review Board 0.2 - 0.2 0.1 - 0.1 Transmission facilities 5.9 1.0 6.8 6.6 4.5 7.9 Subtotal 162.0 72.5 225.0 187.7 169.1 238.9 b) Reactive Power Compensation 3.8 0.2 4.0 - - - c) Sefwi-Viawso-Bibiani transmission system 3.0 1.1 4.0 - - - d) Additional transmission Tema-Accra 3.2 0.4 3.5 - - - Subtotal 10.0 1.7 11.5 9.1 9.4 11.7 Grand Total 172.0 74.2 236-5 196.8 178.6 250.6 1/ Contingencies included in individual items. Annex 3 GHANA KPONG HYDROELECTRIC PROJECT COMPLETION REPORT Schedule of Disbursements (in million US$) IERD FY Cumulative Disbursements at end and semester of semester Appraisal estimate Actual 2nd 1977 3.6 1st 1978 7.3 2nd 1978 9-9 3-6 1st 1979 12.6 4-8 2nd 1979 18.0 8.0 1st 1980 23.4 13.8 2nd 1980 28.4 18.6 1st 1981 33.5 27-4 2nd 1981 36.2 29.8 1st 1982 39.0 35.2 2nd 1982 37.6 1st 1983 39.0 1/ 1/ Last payment 8.12.82 - 62 - Annex 4 GEANA: VRA - KPONG HIDROELECTRIC PROJECT LOAN 1380-GH Disbursements by Categories Actual disbursements against the various categories are as follows: Categg Description Original Allocation Actual Disbursement 1-A & B Equipment Parts A, C, E 30,400,000.00 32,831,202.07 & Engineering services Parts C, E 2-A & B Equipment & Engineering 2,700,000.00 3,279,536.80 services Part D 3 Review Engineering Board 200,000.00 38,969.84 Part D 4 Construction, Equipment & 2,200,000.00 2,850,291.29 Vehicles for preliminary works and resettlement Unallocated 3,500,000.00 N/A TOTAL 39,000,000.00 39,000,000.00 March 1984 -63 - .6-Annex 5 GMANA: KPONG HYDROELECTRIC PROJECT LOAN 1380-GH VRA Operational Characteristics 1976 1977 1978 1979 1980 1981 1982 Peak load in XW - 1976 estimate 570 684 730 768 786 803 831 - actual 554 640 632 660 703 Gross Generation in GVh - 1976 estimate 4320 5230 5530 5706 5790 5900 6090 - actual 4174 4394 3721 4631 5276 5349 4891 Sales in GYh - 1976 estimate 4180 5060 5340 5510 5580 5680 5860 - actual 4091 4303 3652 4524 5130 5181 4799 Sales to VALCO in GMh - 1976 estimate 2700 3330 3333 3330 3330 3330 3330 - actual 2645 2784 2086 2908 3319 3303 3008 Sales to other clients in GWh - 1976 estimate 1480 1730 2010 2180 2250 2350 2530 - actual 1446 1519 1566 1616 1811 Losses in % of generation - 1976 estimate 3.2 3.2 3.4 3.4 3.6 3.7 3.8 - actual 2.0 2.1 1.9 2.3 2.8 3.1 1.9 March 1984 -64 - Annex 6 GHANA: VRA - KPONG HIDROELECTRIC PROJECT LOAN 1380-GH Compliance with Major Covenants Loan Agreement Section Description Status 3-02 (a) Employment of consultants - done (see para. 2.02) 4.01 (b) Appointment of Director of Finance by July 1, 1977 - done (see para. 6.02) 4.04 Periodical inspection of works after completion - organised 4.05 Power for boilers to be - done insofar applicable interruptible (see para. 4.05) 5.04 Revaluation of assets - done (see para. 5.05) 5.05 Rate of return - not met (see para. 5.05 ) 5.07 Debt limitation - met VRA had to seek Bank concurrence for further indebtedness 5-09 Non-power operations - met March 1984 GHANAs LOAN 1380-GH VRA: KPONG HYDROELECTRIC PROJECT Economic Re-evaluation 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1995 a) VRA's sales and generation in GWh Total sales with Kpong 5,181 4,799 2,384 1,060 3,022 3,732 3,839 3,952 4.070 4.190 4,980 5.530 Total generation with Kpang 5,350 4.970 2,470 1,100 3,130 3,860 3,980 4,090 4,220 4,340 5,160 5,740 (100%) 5,250 4,970 2,090 950 2,630 3,260 3,980 4,090 4,220 4,340 5o160 5,400 Total generation without Kpong (1902) 1/ 5.250 4,970 2,090 950 2,630 3,260 3,980 4,090 4,220 4,340 4,860 4,860 (and ray water usage) ( ( 752) 2 5,250 4,210 2,090 950 2,630 3,260 3,980 4,030 4,030 4,030 4,030 4,030 ( (1OOZ) 150 760 320 150 500 600 3/ - - - - - 340 | Generation Attributable to Kpong ( ( 902) 150 760 360 150 500 - 600 / - - - - 300 870 ( ( 752) 150 760 380 150 500 - 600 - - 60 190 210 730 730 ( (1002) 145 735 365 145 485 580 - - - - - - 335 Sales attributable to Kpong ( ( 902) 145 735 365 145 485 580 - - - - 290 840 ( ( 752) 145 735 365 145 485 580 - 60 185 205 705 705 ( (100)z 2.1 10.6 5.3 2.1 6.8 8.1 - - - - - 4.8 b) Benefits of Kpong in million US ( ( 90%) 2.1 10.6 5.3 2.1 6.8 8.1 - - - - 4.1 11.8 ( ( 75%) 2.1 10.6 5.3 2.1 6.8 8.1 - 0.8 2.6 2.9 9.9 9.9 If average generating capability 902 of original estimate i.e. 4,860 Mh for Akosombo and 870 GWh for Kpong V average generating capability 75% of original estimate i.e. 4.030 GWh for Akosombo and 730 GWh for Kpong V Contribution of Kpong to faster return to normal reservoir operation -66- Annex 7 Page 2 of 3 Basic Assmptions for Preceding Table - Total sales assume that it will take until 1992 to reach again the sales level of 1981 as VALCO would work with only 3 potlines until then and only in 1991 add again the fourth line, the fifth still remaining in reserve. - After the droughts of the 70s and the 80s the hydrology is likely to have to be revised in particular to take into account retention in the under- ground. As the potential generations at Akosombo are likely to require a reduction against the figures used in the appraisal report, the table uses 75% and 90% (i.e. 4,030 GWh and 4,860 GVh) for Akosombo, and 730 GWh and 870 GYh for Kpong. - During the years 1981 to 1984 when the drawdown of Volta lake took place generation was attributed to Akosombo and Kpong in proportion to their capacity. This is justified as, without Kpong, the corresponding energy would not have been available. Indeed, Akosombo could have produced it using more water which later would have been lacking. Least Cost Solution As in the project appraisal least cost solution results from a comparison of Kpong with a thermal alternative. The main assumptions on the thermal plants compare as follows. The monetary values for the re-evaluation correspond to the 1981 cost level as Kpong was commissioned in that year. 1976 appraisal estimate 1981 figures Installed capacity MW 160 160 144 122 (1oo%) (90%) (75%) Productibility GWh 970 970 870 730 Capital Cost USS/kW 400 500 Fuel Cost US/barrel 12 20 Transmission investment million US$ 1.6 2.0 Equalizing discount rates - with fuel at US$ 12/barrel 11% - with fuel at US$ 20/barrel 7.5% 8.5% 9% - with fuel at US$ 25/barrel 10% 11% 12% -67- Annex 7 Page 3 of 3 VOLTA RIVER AUTHORITY - VRA KPONG HYDROELECTRIC PROJECT, LOAN 1380-GH Cost and Benefts Streams - Return on Investment Year Investments Generation & Operation and Total Transmission Distribution Total Maintenance Costs I. COSTS (in million Cedis) 1 1976 61 - 6.1 6.1 2 1977 7.8 12.2 20.0 20.0 3 1978 87.0 25.1 112.1 112.1 4 1979 216.6 9.3 225.9 5 1980 218.6 10.2 228.8 228.8 6 1981 225.1 19.9 245.0 245.0 7 1982 118.0 18.7 135.7 136.7 8-25 - - 2.5 2.5 Total Sales Increase in Total Sales through Project Benefits (GWh) Sales over 1981 (GWh) (GWh) (Cedis X 106 a/ II. BENEFITS (in million Cedis) 1 198- 4779 (382) - 2 1983 2384 (2797) - 3 1984 1060 (2159) - 4 1985 3022 (2159 - 5 1986 z732 (1449) - 6 1987 3839 (1342) - 7 1988 4107 (1074) - 8 1989 4394 (787) - 9 1990 4702 (479) - 10 1991 5031 (150) - 11 1992 5383 202 202 400 12 1993 5761 580 580 1148 13 1994 6165 984 970 1920 14-25 1995 6595 1219 970 1920 Rate of Return = 19.6% a/ At ave-age retail tariff of Cedis 1.92/kWh, effective on July 1, 1984 -68 - Annex 8 GHANA KPONG HYDROELECTRIC PROJECT COMPLETION REPORT Schedule of Supervision Missions Date of Duration Mission Staff No. of months Imission in days since last mission June 1977 10 Engineer No.1 15 Fin.Anal.No.1 1/ July 1978 7 Engineer No.1 13 February 1979 7 Engineer No.1 7 November 1979 7 Engineer No.1 10 Fin.Anal.No.2 Consultant 2/ October 1980 7 Engineer No.1 11 Engineer No.2 Fin-Anal.No.3 July 1981 7 Engineer Bo.2 9 Fin.Anal-No.3 Fin.Anal.No.4 November 1981 7 Fin-Anal.No.3 4 1/ members of appraisal team 2E/ resettlement specialist GHANA KPORG HTVROELECTRIC PROJECT COPLETION REPORT Cosparative Intas. 8tat.ens,LI[6-198 tin *tLl0r.na of EInt Cedis) Sales (vh 1 000) Iw 1.0 1.0 1.n 1.2 1.1 1.4 1.0 1.5 1.1 1.6 1.1 1.7 1.0 1.8 VALCO 2.6 2.7 2.8 3.3 2.1 3.3 2.9 3.3 3.3 3.3 3.3 3.3 3.0 3.3 1ines 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 CIn 0.2 0.2 0.2 0.2 0.2 0.3 0.3 0.3 0.4 0.3 0.5 0.3 0.5 0.3 Almolbo 0 0.0 0 00 0 0 0 0.0 0.0 0 0 0.0 0.0 0.1 Pris- (Mill/kVh> EW 9.761 9.360 12.878 11.940 16.821 12.080 22.277 12.400 as.068 12.640 "6.07 12.620 79.217 12.560 VALCO 5.175 5.170 4.973 5.360 8.210 5.450 12,827 5.620 13.191 5.870 15.886 7.370 13.787 7.470 Nine. 10.630 10.340 13.827 13.960 20.492 13.950 34.825 13.930 65.827 13.930 89.456 13.930 92.006 13.930 Cza 9.948 9.340 11.223 9.280 24.506 8.850 40.992 8.650 39.594 8.850 39.40 8.850 58.665 8.850 Akosomh 09989 224 8 7 990 D22-39 e90 2.2 W*.9i*1 1 .7 Oe rtag RevenuAs. ed 9.6 9.4 13.3 14.0 17.9 16.5 22.8 18.8 47.4 19.9 85.6 21.0 79.2 22.9 VIL00 13.7 14.0 13.8 17.8 17.1 18.1 37.3 18.7 43.8 19.5 52.5 24.5 41.4 24.9 sina 3.0 2.8 3.6 3.9 5.1 4.1 9.0 4.2 17.9 4.4 24.5 4.6 23.7 4.9 C&B 1.5 1.6 2.0 2.3 5.3 2.8 12,3 2.8 17.4 2.8 18.6 2.8 30.6 2.8 Akoscabo 0. _. 9. 0.4 0.7 _0 1.1 0.4 b. 0.4 1.6 0. 1.2 20. Total maln revenuss 28T 28.1 3ti 38.3 4wT 4 9 8 W, 44d I2.T4" iU¶ 51" I7T 55.8 Other 0.8 4.6 1.1 1.0 2.6 6.7 5.5 11.0 8.6 16.9 10.9 28.6 12.7 52.0 Total 28.9 32.6 34.5 39.3 48.7 48.6 88.0 55.9 136.7 64.0 193.6 81.9 188.8 107.8 ¥81.rtion 1.8 1.4 2.9 1.4 4.3 1.4 5.2 1.4 9.0 1.4 12.2 1.4 16.2 1.4 Transafislon 1.7 1.5 2.8 1.5 3.7 1.5 5.1 1.5 8.2 1.5 10.6 1.5 11.0 1.5 Adatnistration 2.7 2.0 3.8 2.0 5.1 2.0 8.0 2.0 9.9 2.0 13.0 2.0 15.7 2.0 Akosombo Tovnöhup 0.9 0.7 1.6 0.7 2.2 0.7 2.9 0.7 4.7 0.7 6.9 0.7 6.0 0.7 Health a Satfsy 1.1 0.8 1.5 0.8 2.0 0.8 2.5 0.8 4.4 0.8 6.1 0.8 5.9 0.8 Prias conttngenaies • 1.1 - 1.9 - 3.4 - 4.9 - 6.0 - 7.8 - 9.1 [pong - - - - - - - - - - 0.5 - 0.5 Deproiation _95 16:3 10:7 27.8 12:0 41:0 13 45.8 18.4 23._ Total fl =f4 W__. fl. -9l- K-O: IM- 2U»2 11. OporatIng %acnse 16.0 18.3 9.5 21.5 15.2 28.1 36,7 32.7 59.5 38.1 99.0 48.9 75.3 68.0 son-operattag aopsaes - - 3.5 - 5.0 - 11.0 - 10.9 - 10.3 - 8.4 - internt chrgsed to operations 4.0 4.2 .L 4.0 5j8 8.3 .4 11.4 3.1 7.2 2.8 401 11.4 fst incowe 12.0 14.! 2.3 17.6 4.4 24.4 17.3 29.3 37.1 35.0 81.5 46.1 26.s 56.5 Dats of return on equity 12.3 8.9 0.9 6.9 0.8 7.9 2.0 7.9 3.3 7.9 5.8 8.8 1.6 8.9 Averaga 1977-1982 2.4 8.1 Iata of return on not rasd asdsta 10.0 8.0 3.1 6.9 2.4 8.2 4.0 8.8 5.2 9.4 6.0 8.4 3.5 8.9 Average 1977-1982 4.0 8.4 Covenanted te. of return on qats j - - 5.7 7.0 6.5 8.0 6.5 8.0 6.5 8.0 7.3 9.0 7.3 9.0 Operating ratin 45 44 72 45 69 42 58 42 56 40 49 40 60 37 11 Forsat fagrs are the Loan Agreement figuren; the asoult figurco are the forcesso figure edjusted bl the fater 300/370. The fattor te *ospose4 of the 1 300 stilton not firmed aset valut sattnated at apprataat and the 0 37n atual reulte of the revaluation aerats as of December 31, 1976. March 12, 1984 OHANA KP )NG HYDROELECTRIC PROJECT COMPLETION REPORT Comparative Balance Sheets, 1976-1982 (in millions o current Cedia) 1976 1977 1978 1979 180 91 1982 Actual Foreasst Actual Forecast Actual "orecast Actual Forecast ActaUldPrecast Ac 96_eca ActualFPorecas Asseta Praat in operation 200.6 373.1 571.4 421.2 987.3 469.5 1,339*3 527.8 1,789.3 591.8 2,825.2 940.6 3,013.3 1,046.1 Lees Depreciation 43.1 76.1 124.3 94.0 186.3 115.0 285.3 139.6 5. 168.4 740.2 205.4 79a.1 251.8 Net plant 15. 9. 47.1 327.2 801.0 354.5 1,054.0 38.T~843. ,8. 3-5.2 2,215.2 794.3 Work in progress 15.1 7.0 18.2 45.3 146.5 117.7 341.2 209.4 555.9 306*3 68.9 142.6 112.9 327.6 Current assetsi Cash and banks 8.1 5.1 16.0 11.3 43.5 13.8 82.1 18.4 71.5 18.9 72.t 24.9 73.9 16.4 Accounta receivable 11.2 7.3 9.6 7.9 20.3 9.7 28.6 11.2 63.7 12.5 85.6 16.4 83.9 21.6 Inventories 2.1 1.- 2.0 1.9 2.2 2.0 5.0 2.0 6.4 2.1 ...± 2.1 11.2 2.2 Total -14 14,0 2q.6 j.O W 0 T -.5 TI 115 .7 31.6 141.6 33.8 -14 4 3.4 190 4.2 Total assets 194.0 318.0 492.9 393.5 1,013.5 497.7 1,510.9 629.2 1,924.3 763.5 2,318.3 921.2 2,497.1 1,162.0 ................... .......... ..... Rangoon a....0 SU*m....a .Now ws.... moxam 420.080 mumm.. ltabilitie Equity 59.1 59.1 59.1 59.1 59.1 59.1 59*1 59.1 59*1 59.1 59.1 59.1 59.1 59.1 Retained earnings 29.7 27.9 34.4 43.5 60.8 5.8 89.7 93.1 144.7 126.1 227.5 t70.2 252.9 224.7 Revaluation reserve 17.6 142.7 ~jI.- 175.4 602.0 211.4 849.0 20.4 11,0998 293.1 1.8. 33* 1,419.1 420.- Total 10. -4 366 28. T 3 99. 402.6 163.6 47 T 1,6. 568.9 1.731.1 704.3 Long term debt 80.5 84.0 79.2 110.5 270.0 155.7 459.9 220.1 597.3 277.2 672.3 342.3 689.7 445.2 Current liabilities: Accounts payable 5.8 3.0 5.7 3.6 18.2 3.8 46.4 3.9 56.2 4.0 52.7 4.1 42.6 4.2 Finan-.ial charges 3 1.3 1.4 1. 3.5 2.0 6.9 2.6 .2 4.0 2 _ 33.7 8.3 Total 7.1 4.1 : H 21.7 5.8 53*3 =i 64 8.0 78.1 10.0 .-3 -12.5 Total liabilities 194.0 318.0 492.9 393.5 1,013.5 497.7 1,510.9 629.2 1,924.3 763.5 2,318.3 921.2 2,497.1 1,162.0 ........ .............. ....... ........ .......** 0.0.0 .....U a. aw.4 Ummw. .0. ... a.. . Debt/Debt + Equity 43 27 16 28 27 32 32 35 32 37 30 38 28 39 Debt/Equity 0.7 0.4 0.2 0.4 0.4 0.5 0.5 0.5 0.5 0.6 0.4 0.6 0.4 0.6 Current ratio 3.0 3.3 3.9 4.2 3.0 4.4 2.2 4.9 2.2 4.2 2.1 4.4 2.2 3.2 Receivabler/revenue () 39 20 28 20 42 20 33 20 47 20 44 20 48 20 Receivables (days) 141 72 102 72 152 72 119 72 170 72 161 72 174 72 Annual revaluation (a) 81 80 20.3 11 34.8 11 34.0 11 35.8 It 34.5 11 35 11 March 12, 1984 OHANA KPON HYDROELECTRIC PROJECT COMPLETION REPORT Comparative Flow of Funds Statements 1976-1982 (in millions of current Cedia) 1977 1n J1i 190 198 1 1977-19a2 Actual Forecast Actual Forecast AcuI ore ActualForecasf ActualIMorecas ActuallForet114t Acua orecast Actual Forecast Internal Sources Operating Income 20.2 18.3 9.5 21.5 15.2 28.1 36.7 32.7 59.5 381 99.0 48.9 75.3 68.0 295.2 2Y7.3 Depreciation 4.6 6. 12.3 .5 16.3 10.7 27.6 12.0 4. . 45.8 18.4 58.8 23.8 201.8 8. Total 24.8 25.2 21.8 31.0 31.5 38.7 64.3 44.7 100.5 51.5 144.8 67.3 134.1 91.8 497.0 325.0 Operational requirementas Working capital 7.2 0.5 0.1 0.1 (3.7) 1.3 (20.5) 0.8 26.4 0.7 7.5 1.7 4.6 2.7 14.4 7.3 Debt service 8.6 8.9 5.6 9.7 10.7 11.6 19.9 14.3 23.7 19.2 20.5 26.3 75.2 42.2 158.6 123.3 Dividends 2.0 2.0 1.0 2.0 - 2.0 - 2.0 - 2.0 - 2.0 - 2.0 1.0 12.0 Total 17.8 11.4 _.7 11.8 _.0 1.9 (0.6) 17.1 50.1 21.4 28.0 29.9 79.8 47.0 174.0 1420 Net available from operations 7.0 13.8 12.1 19.2 24.5 23.9 64.9 27.5 50.4 30.2 116.8 37.3 54.3 44.8 323.0 182.9 Construction requirements 6.1 10.6 7.8 44.4 87.0 71.6 216.6 93.0 218.6 92.7 225.1 103.2 118.0 170.8 873.1 575*7 Balance to finance (0.9) (3.*) (4.3) 25.2 62.5 47.7 151.7 65.5 168.2 62.5 108.3 65.8 63.7 126.0 550.1 392.7 Financed byt Borrowings 0.2 - 1.4 31.3 62.8 50.3 190.3 70.0 157.7 63.0 106.2 71.8 S6.2 117.5 584.o 403.9 Other 1.1 - 2.2 - 27.2 - - - - - 2.7 - (0 7) -_ 31 - Total 1.3 - 3.6 31.3 90.0 50.3 190.e 70.0 157.7 63.0 108.9 71.8 65*5 117.5 616.0 403.9 Surplue (deficit) of funds 2.2 3.1 7.9 6.1 27.5 2.6 38.6 4.5 (10.6) 0.5 0.6 6.0 1.8 (8.5) 65.8 11.2 Accumulated surplue (deficit) of funds 8.1 3.1 16.0 9.3 43.5 11.8 82.1 16.4 71.5 16.9 72.1 23.0 73*9 14.4 73.9 14.4 Net available from operations/ construction requirements (U) 115 130 155 43 28 33 30 30 23 33 52 36 46 26 37 32 Debt service covereas 2.9 2.8 2.5 3.2 2.9 3.3 3.2 3.t 4.2 2.7 7.1 2.6 1.8 2.2 3.1 2.6 March 12, 1984 - 72 - GRANA ELECTRICITY CORPORATION OF GHANA PROJECT COMPLETION REPORT THIRD POWER PROJECT - LOAN 1381-GH, CREDIT 689-GH 1. INTRODUCTION 1.01 In Ghana public electricity supply is the responsibility of Volta River Authority (VRA) and Electricity Corporation of Ghana (ECG), both statutory corporations owned by the Government of Ghana. Created in 1961 to build and operate the first hydropower plant on the Volta river at Akosombo, VRA supplies power to BCG, to Volta Aluminium Company (VALCO), which operates a smelter at Tema, to several gold, diamond, manganese, and bauxite mining industries, and to the CommunautZ Electrique du Benin (CEB), which supplies the national power entities of Togo and Benin. ECG is responsible for the distribution of power to all other consumers and for the generation of electricity for public supply in areas that cannot be economically connected to VRA's high voltage system. The government of Ghana established ECG in 1967 to reorganize, under Loan 310-GH to VRA, the Electricity Division of the Ministry of Works and Housing as a government-owned public utility with authority to conduct its business according to commercial principles. Today, the inistry of Fuel and Power (MFP) supervises ECG, which is directed by a board of seven members including ECG's and VRA's Managing Directors. 1.02 The Bank Group has assisted both entities of the power sector. VRA received Loan 310-GH for US$ 47 million in 1963, Loan 618-GH for US$ 6.1 million in 1969 and Loan 1380-GH for US$ 39.0 million. Loan 1380-GH was approved in 1977 at the same time as the Loan/Credit package for the ECG's Third Power Project. Loans 310-GH and 618-GH helped finance the Akosombo hydroelectric plant on the Volta river and associated transmission facilities, and Loan 1380-GK the Kpong hydro-plant downstream from Akosombo and further transmission installations. ECG has been the beneficiary of two previous IDA operations: Credit 118-GH for US$ 10 million in 1968 and Credit 256-GH for US$ 7.1 million in 1971. All the projectb have been satisfactorily completed without undue delay. OED's audit of the first four oper-.tions in the power sector comments on the desirability of (i) merging VRA and ECG (ii) revaluaing the sector's assets, (iii) achieve a higher rate of return, and (iv) solve the overstaffing problems. 1.03 In 1976, ECG had an installed generating capacity of only 81 MV in 28 diesel plants. The largest such plants were those at Tema (33 XV) and Accra (15 MW). The dependability of some of this equipment was low, but it did not affect the quality of service as the units served only to back up the normally reliable supply from VRA's Akosombo hydro-plant. Only isolated systems had to rely entirely on ECG diesel plants. As these systems were small and as in the interconnected system transmission was and remains VRA's responsibility, ECG only owned subtransmission networks and distribution facilities at a voltage less than 33 kV. Currently, ECG supplies about 900 GVh to some 105,000 residential, 35,000 commercial, and 250 industrial premises. - 73 - 1.04 Ghana has never had a long term program for rural electrification. However, Government has supported this activity and each year, within its budget constraints, ordered ECG to implement specific facilities judged to have high priority. Usually the economic viability of the ventures was not analyzed. Under Credit 256-GH, Government undertook not only to finance investment in rural electrification but also to reimburse ECG for operating losses in rural scheiues. 1.05 In 1975, Government asked the Bank to continue supporting ECG's efforts to meet the growing demand and to improve service by helping finance a further subtransmission and distribution project. This report describes the preparation 7id implementation of the third power project on the basis of the consultant's completion report, the IBRD supervision reports and files, and the findings of recent missions to Ghana by one of the Bank's financial analysts and a consultant. 2. PROJECT PREPARATION Origin and Preparation of the Project 2.01 ECG's own staff prepared the project. The Bank was instrumental in its formulation through three identification and pre-appraisal missions carried out in 1974 and 1975. The Bank carried out the appraisal in November- December 1975, but loan approval and signature took place only in March 1977, essentially because the VRA project (Loan 1380-GH), with which the ECG operation was to be coordinated, took longer to process than the ECG project. Furthermore, only after long discussions was the Bank willing to include the Kumasi-Kumawu scheme, which is essentially rural, into the project. 2.02 In 1976, ECG had hired an engineering consultant to assist in pro- curement, to carry out the final design of the electrical facilities and to adpervise their installation. As the consultant completed most of the evaluation of bids before the project was submitted to the Bank's Board of Directors, the cost estimates in the appraisal report were based on bid prices. Project Role 2.03 Ghana's power sector objectives for the period 1976-1981 were to: a) increase the firm generating capacity to enable the sector to meet the demand growth beyond 1978; b) carry out the feasibility studies for the generating and transmission facilities required for the years 1983 onwards; c) extend hydropower supply to areas where it is moee economical than that based on diesel generation; and d) extend and re-inforce subtransmission and distribution systems to improve the quality of service and meet the load growth. 2.04 Whereas VRA pursued objectives (a), (b) and the transmission part of - 74 - (a), ECG was responsible for reaching objective (d) and the subtransmission and discribution component of objective (c) above. 2.05 VRA and ECG prepared separate development plans: VRA for large scale power generation (which until now and in the foreseeable future is hydro) and high voltage transmission; ECG for subtransmission and distribution. In 1975, when the Bank appraised ECG's Third Power Project, VRA had identified the 160 NW Kpong hydroelectric plant on the Volta River, 30 km downstream from the then already operating Akosombo plant (912 NV), as the most economical next facility for meeting load growth. VRA's planning also foresaw the associated extensiorn and reinforcement of its transmission system. 2.06 ECG, which has incrrRsingly concentrated on distribution, prepared only a medium term plan, which, in 1976, was essentially the physical part of the Third Power Project. ECG and VRA have effectively coordinated their planning as: - VRA and ECG offices are located in the same building; - the Chief Executive of each entity is a member of the Board of Directors of the other; and - managers of one utility at times transfer to the other. 2.07 At the time the project was appraised, it was without any doubt the appropriate step for meeting the demand of ECG customers and hence provide the economy with the electric energy it needed, and at one of the lowest costs in the region. A further project objective was ECG's institutional strengthening, particularly in the field of management, staffing, training, systems operation, and maintenance. Project Description 2.08 The project comprised the following components: a) Construction of about 240 km of 33 and 22 kV subtransmission lines and 11 new 33/11 kV substations with a capacity of 150 MVA, and the expansion of 6 existing stations by about 60 MVA, the facilities supplying mainly: - the Aboso glass factory (Tarkwa - Aboso scheme) - the Teschie housing program (Teschie - Nungua scheme) - the Weija water works (Weija scheme) - the agglomeration of Tema (Tema scheme) - the Sefvi-Wiawso-Bibiani area - the Kumawn area (Kumasi-Kumawu scheme); 2/ b) Improvement of distribution (415 /24 0 V) in particular in Accra/Tema-; c) Supply of miscellaneous equipment and material for rehabilitating the existing subtransmission and distribution systems; d) Vehicles to replace and expand existing transportation facilities which, due to the lack of foreign exchange, could not be kept in good ct.ndition; and 2/ See ECG's comments, Appendix I, p. 28. - 75 - e) Consultant and engineering services for - the execution of the project and - a power tariff study. 2.09 The USS 9 million IDA credit was intended principally to help Government finance the foreign exchange component of the Kumawu rural scheme (USS 2.8 million). The remainder of the credit and the entire US$ 9 million IBRD loan were to finance the rest of the items on a pro rata basis. Accordingly, Government relent US 6.1 million of the IDA credit to ECG on the same terms as the Bank loan, i.e. for 20 years including a five years grace period and an interest rate of 8.5%. Furthermore, Government passed on to ECG the US$ 2.9 million for the Kumawu scheme as a grant. 3. PROJECT IMPLEMENTATION, OPERATION AND COST Conditions of Effectiveness 3.01 The loan and credit documents contained only the usual requirements for execution and ratification of Loan Agreement, Project Agreement, and Subsidiary Loan Agreement. No events delayed the procedures. Thus, loan and credit became effective on June 10, 1977, that is 2 1/2 months after the date of signature and two weeks before the date set in the Loan and Development Credit agreements. Changes in Project 3.02 ECG did not introduce any changes of substance into the project. The most important modification concerned the addition of two substation extensions and of connection cables between substations in Kumasi. However, due mainly to the difficult economic situation which developed in 1978 and which impeded the orderly execution of the project, ECG sometimes had to use equipment and material earmarked for maintenance to carry out system extensions and vice versa. 3/ Implementation Schedule 3.03 Except for part of the Tema distribution extension, the physical project items suffered delays of between 1 1/2 and 3 years. Thus, final disbursement of the loan was postponed from December 1980 to February 1982. The delays were mainly due to the deterioracion of the economy, which at times, in particular in 1979, led to severe disruptions of most activities in the country. In the context of project execution, this caused: - scarcity of materials such as fuel, tires, cement, steel due to shortages of foreign exchange; - long delays in the processing of payments involving foreign exchange; - late payment of local currency expenditures because ECG was 3/ See ECGTs coments, Appendix I, p. 23. - 76 - short of cash, due to insufficient tariffs; and a rather pervasive staff pessimism and low morale, which also affected many contractors and, together with the reasons mentioned (para. 3.03 above), made many of them reluctant to keep up the work pace. 3.04 A more project specific reason for the delays relates to the planning and execution of the civil works, which ECG separated from supply and erection of equipment and entrusted to local consultants and contractors. Mis- understandings between ECG and its main consultant, concerning the latter's role in assisting the local firms, unduly delayed the start of installations. Only after the consultant, at ECG's requ5it, had instructed the civil works consultants did the corresponding engineering actually start. But the limited technical ability of both local consultants and contractors allowed only slow progress, often associated with substandard work that had to be corrected or even redone. This experience led ECG to return to its earlier practice of making the installation contractors also responsible for the associated civil works. Under the circumstances prevailing in Ghana, this certainly had the best chances to lead to timely completion and to the required quality of the facilities. 3.05 ECG completed the Tarkwa-Aboso scheme in February 1982, 38 months late, though it had started work 5 months before schedule. The main delay occurred at the Tarkwa connection to the VRA system, where VRA and ECG took an unduly long time to define the details of the interphase. Work on the Teschie-Nungua scheme started about one month early but ended about 14 months late, in November 1980. The Weija scheme, which had started about one month late, was completed in October 1980 with a 16-months delay mainly due to the reasons already mentioned (para. 3.03). 3.06 The improvement and extension of the Tema system including part (b) of the project (para. 2.08), essentially consisted of the construction or extension of 3 substations, of the connection of the Tema steel plant and of Tema Textiles, and of the installation of a 33 WV overhead line between two substations. Whereas the line was completed three months early, the other items suffered long delays, the last of them (switchgear at Station H) was completed in June 1981, i.e. 28 months late. 3.07 ECG had to coordinate its work on the Sefwi-Wiawso-Bibiani scheme with that of VRA, which, under Loan 1380-GH, was extending its 161 kV transmission system to supply the scheme. The loan and credit documents included as a condition of disbursement of funds for this item that VRA issue a letter of intent concerning the construction of the facilities feeding the scheme. VRA sent that letter in early 1978 but incurred a three-year delay in the execution; therefore, when BCG completed its scheme in May 1982, i.e. 26 months late, VRA was still not in a position to supply it. This finally occured in 1983. The slippage in ECG's own work was also due to the previously stated general reasons. 3.08 The Kumasi-Kumawu scheme, including the reinforcement of the 33 kV system in Kumasi, has not yet been entirely completed. A temporary connection between two substations in Kumasi is limited to I1kV capacity until ECG can carry out a permanent rearrangement of its system. Only then will it be possible to install the 33 kY cable planned to replace the temporary 11 kY - 77 - connection. Work on the Kumasi-Kumawo line, though started about 9 months early, was only completed in October 1980, i.e. 17 months late. 3.09 The supply of equipment and materials for maintenance and normal development under Part (c) of the project proceeded essentially as foreseen. The same applies to the vehicles procured under Part (d) of the project. 3.10 Under Part (e), Consulting and Engineering Services of the project, ECG contracted a consulting engineering firm to assist in engineering and procurement, and to supervise equipment installation. As mentioned earlier, ECG also let contracts for the engineering of civil works to local consultants, but this did not lead to satisfactory results. It would be preferable in future to follow the consultant's suggestion, which ECG now endorses, that only c -onsultant be made responsible for a project like ECG's. This would not. vent using local consultants as subcontractors of the main consultant or wucing the installation contractors responsible for associated civil works. 3.11 As completion of the project as a %.ole was delayed by some three years, the consultant had to extend their services in Ghana to May 1982, when its personnel left the country as the project facilities were completed except for the connection of the Sefwi-Wiawso-Bibiano scheme (para. 3.07) and the cable installations in Kumasi (para. 3.08). From its headquarters, the consul- tant continued to administer the contracts. At this late stage this mainly involved the processing of contractors' cia:ms, of which a fer r_nor ones are still pending 4/. 3.12 In 1978, also under Part (e) of the project, ECG appointed an economic consultant to carry out a sector study aimed at defining costs of supply and proposing a tariff policy for the sector. EdF completed the assignment in 1980, rhen it submitted a report suggesting a tariff schedule based on marginal cost pricing and designed to provide the revenues ECG needed to reach and maintain a sound financial position. However, the rapidly deteriorating economic situation and the effects of the droughts on VRA's operations prevented implementation. Procurement 3.13 ECG, assisted by its consultant procured the project items in accordance with the Bank's Guidelines for Procurement, except for US$ 400,000 (USS 500,000 forecast) of materials and of subtransmission and distribution equipment which, for compatibility and standardization reasons, the original supplier furnished. Before contract award the consultant made sure that the prices of the latter equipment were reasonably in line with those obtained under international competitive bidding. 3.14 By the end of 1976, or three months before signing of loan and credit, the engineering consultant had already carried out the tender procedure and ana- lyzed the bids for all the supply and erection contracts. ECG issued letters of intent in early 1977 before the validity of the bids expired. Although the ten- der documents included the usual preference clause for locally manufactured equipment, no Ghsnaian supplier presented a winning bid. By mid-1977, when credit and loan became effective, only the spare parts contract had not yet been signed. This happened in the second half of 1977. Annex 1 lists the main contracts. 4/ See ECG's comments, Appendix I, p. 28. - 78 - 3.15 In parallel with the above procedure, RCG, assisted by local consultants, carried out the tendering for the civil works which, by mid-1977, led to the signature of all contracts except one, which was finalized later in the year. 3.16 To contract the consultants for the tariff study ECG invited pre- selected consultants of 5 nationalities to present proposals based on terms of reference approved by the Bank. ECG selected a firm which best responded to these terms and offered a reasonable price. The Bank concurred with the procedure and ECG's award recommendation. Costs 3.17 Table 3.01 below compares estimated and actual costs of HCG's Third Power Project. The actual costs are those of SWEB's completion report, as modified to include works that ECG carried out with it.q own forces (e.g. clearing and some erection of distribution items). It seems that these additions, especially those in local currency, are substantially under- stated. Furthermore, the US$ equivalent of the local currency component has been calculated on the basis of average official exchange rates in the various years that expenditures took place. As the exchange rate was grossly distorted, the US$ equivalent figures are at best indicative of orders of magnitude. Table 3.01 Estimated and Actual Project Costs (in million of USS) Appraisal Estimate Actual Foreign Local Total Foreign Local Total Item Tarkwa-Aboso Scheme 0.8 0.4 1.2 1.2 0.8 2.0 Teshic-Nmgua Scheme 1.1 0.5 1.6 0.8 0.7 1.5 Veija Scheme 0.9 0.5 1.4 1.0 0.6 1.6 Tema Scheme 1.9 0.4 2.3 2.8 1.1 3.9 Sefri-Viawso Scheme 2.8 1.6 4.4 3.6 4.0 7.6 Kunasi-Kuman Scheme 2.8 1.7 4.5 3.1 2.2 5.3 Supply of spares etc. 2.6 - 2.6 3.5 - 3.5 Vehicles 1.5 - 1.5 1.3 - 1.3 Consultancy 0.3 0.2 0.5 1.1 0.5 1.6 Others - - - 0.4 0.8 1.2 14.7 5.3 20.0 18.8 10.7 29.5 Contingencies 3.3 3.4 6.7 - - - Total 18.0 8.7 26.7 18.8 10.7 29.5 Prom the above figures, it can be concluded that: - the actual foreign cost was slightly (about 4%) above the appraisal estimate; - 79 - - total cost in US$ equivalent exceeded the estimated amount by only 10%; and - the local costs in Cedis, in spite of inflation, and even if one assumes that another 2 or 3 million Cedis should be added to make them comparable with the estimated figures (as ECG accounts do not seem to include all assets), could not have been more than 40% above the amount envisaged in 1976. 3.18 As there were no extraordinary project modifications, it is possible to compare the appraisal estimates for physical and price contingencies with the amounts actually paid for variations and escalation. Table 3.02 below sets forth this comparison. Table 3.02 Contingency Allowances and their Use (in millions of USS) Foreign Local Total Physical contingencies: appraisal estimate 0.6 0.2 0.8 Actual costs of additions or changes 1.3 1.0 2.3 Price contingencies: appraisal estimate 2.7 3.2 5.9 Actual 2.9 6.0 8.9 Total contingencies (appraisal) 3.3 3.4 6.7 Total incremental cost (actual) 4.2 7.0 11.2 Table 3.01 and 3.02 show that the appraisal estimate of costs was very accurate for the foreign exchange component, due mainly to having been based on bid prices. The difference between estimated and actual Cedi costs is apparently due to the devaluation of that currency during the project period. Allowances for physical contingencies were clearly too low. Those for price contingencies were quite adequate for the foreign exchange component but too low for the local currency portion. This large difference is not surprising as neither Government nor the Bank foresaw Ghana's economic deterioration. - 80 - Disbursements 3.19 Due to procedural difficulties in Ghana, disbursements of both the loan and the credit started only four months after effectiveness. However, these picked up quickly as funds were needed to reimburse the Bank of Ghana for the US$ 1.5 million it had advanced for consultants fees and down payments for equipment which the Bank had agreed to finance retroactively. By the end of 1980, total drawdowns had nearly reached the level foreseen at appraisal. Later, however, disbursements fell again behind schedule because of the delays incurred in installations and construction. Annex 2 shows planned and actual disbursements, and Annex 3 shows the use of credit and loan by categories. Operations 3.20 As far as ECG'e engineers are already able to judge, the equipment installed under the project operates satisfactorily. However, as indicated below, many facilities have been tested but are not yet operating under normal conditions. Indeed, the various schemes' present utilization may be as low as 20% of their capacity. 3.21 The Tarkwa-Aboso scheme, though complete since Narch 1982, has not yet provided significant service because the installation of the Aboso glass factory was delayed due to the depressed state of the economy. Later, however, the Aboso Glass Factory was supplied with power during the period VRA curtailed its supplies 5/ to ECG as a consequence of the severe drought that had depleted Volta Lake (Akosombo plant). 3.22 The Teschie scheme was able to supply some new industries but the housing development was executed at a very slow pace due to the financial difficulties facing the State Housing Corporation and to the country-wide lack of building materials. 3.23 The Veija scheme is fulfilling its purpose of improving the reliability of operation of the water works. Furthermore, it could soon provide service to the Weija irrigation project, as the pumping station was recently commissioned. 3.24 In the Tema scheme, EGG was to supply energy to two new electroboilers at the Tema textile factory. However, for most of 1983 the factory was closed for lack of raw material. Recently, it received material but due to the prevailing power curtailment, ECG could not supply the energy required and had to suggest to the company to rehabilitite and operate its oil-fired boilers. 3.25 In 1983, VRA was able to hook up ECG's Sefti-Wiawso-Bibiani scheme. When VRA can provide electricity, ECG may discontinue operation of its diesel plants in Sefwi, Wiawso, and Bibiani. The same applies to the Awaso bauxite mine and the State Gold Minirg Corporation which has built its own 33 kV line to connect its facilities with the VRA/ECG system. ECG expects that a sawmill and a plywood factory in the area will start operating in 1984. It would seem that this region, which has a substantial actual and potential export-oriented production, is developing in spite of the difficult economic situation. 3.26 The Kumasi-Kumawu scheme has permitted ECG to discontinue operating its diesel station at Mampong and seems, within the present constraints, to 5/ See ZCG's co-mments, Appendix I, p. 28. - 81 - have brought a considerable improvement in the economic activity and the living standard in the communities along the main line. 3.27 The vehicles acquired under the project have proved to be of good quality. ECG engineers state that many are still operational. 3.28 Naintenance of ECG's system, including the new facilities, is not satisfactory. The utility is often not in a position to carry out preventive maintenance because it lacks spares and tools. The items purchased under the project have been used and Government, in 1982, could not provide enough foreign exchange to permit ECG to import the materials required. Though in..1983 Government was able to grant ECG a larger import license, ECG could not utilize the whole license, because the banks could not establish the necessary Letters of Credit. 6/ Performance of consultants 3.29 The performance of the consultants employed to assist ECG in design, procurement and implementation of the electrical part of the project was satisfactory. However, there are diverging views on the responsibility of the local civil work consultants for the delays experienced in this part of construction, which had the unfortunate effect of impeding the completion of several of the main items before the economic crisis set in and further hampered work. 3.30 The contract between ECG and its consultant stated that ECG would be responsible for all civil works at substation sites, for accesses, and for services associated with these items. ECG, of course, also assumed the responsibility for the standard of design and the quality of these items as well as for their integration and cocrdination with the electrical works. ECG obviously assumed that the main consultant would provide the information and instruction to the civil works consultants although the contract did not explicitly foresee such a service. The consultant in turn, on the basis of the contract provision outlined above felt that, beyond holding a watching brief and attempting to step up progress by written comment to ECG and unofficial contact with the local consultants, it had no obligation to provide further services in this matter. Ultimately, but very late, ECG asked its consultant to provide the local consultants with the details required and to give them the initial instructions, which the foimer did in March 1978. 3.31 It seems that there was, if not a lack of communication, certainly a lack of understanding among the various parties involved. It is also likely that ECG's coordinating capacity was already affected by the shortage of staff. The local consultants also appeared to have been overtaxed by their job and affected by ECG's lack of strong guidance. Performance of Contractors 3.32 G experienced no problems with the equipnent suppliers, as these delivered their equipment and materials in good condition and essentially on time. The erection contractors also performed satisfactorily under the very difficult circumstances they faced. To begin with, they were confronted with delays which started at the design stage of the civil works and compounded later by those experienced during construction. Thus, erection, except that of the 33 kV overhead lines, started late and the contractors suffered the full impact of the economic situation in the country. As a result of the . various deviations from schedules, the consultants, in order to prevent 6/ See EC's conens, Appendix I, n. 28. - 82 - further slippage and interruption, directed the installation contractors to install equipment whenever it was possible. Therefore, it was sometimes expedient for the electrical contractor to complete minor civil works himself. Taking into account that payments often came late, it is no small feat that all equipment was eventually satisfactorily installed. 3.33 The performance of the lozal contractors was rather poor, but it is likely that under a stronger guidance they might have done better. However, one should take into account that the performance of these firms, which are financially weak and fully paid in Cedis, often after long delays, was affected by the adverse economic conditions in the country. 3.34 In any event, ECG's decision to make the electrical contractors responsible for the associated civil works seems sound. ECG has already accumulated some experience with this method as it is applying it successfully in the electrification of the Volta region, financed by KfW. Performance of the Borrower 3-35 It is evident that, since 1978, ECG was not in a position to provide the strong leadership that project implementation would have required, as the utility was burdened by tasks which it easily could have delegated (e.g. the coordination of civil and erection work). The reason for this weakness lies with the institutional shortcomings which, in part, were also responsible for ECG's poor operational performance. 4. OPERATING PERFORMANCE 4.01 ECG's expected operating performance waa not achieved. Between 1977 and 1981: - instead of growing by the projected 9% per year, sales grew at less than 5%. In 1982 they fell to 928 GWh, the lowest value since 1975; - instead of improving, the reliability of ECG's system diminished substantially; and - instead of improving, the financial situation of ECG deteriorated. 4.02 The main reason for ECG's operational deterioration was the poor state of Ghana's economy. In the period 1977-1981, despite residential sales growth of about 8% per year, total energy sales stayed about constant until 1Sr81 as industrial and commercial consumption fell by about 15%. In 1982 and apparently also in 1983, a reduction of consumption in all categories took place as in addition to the economic problems the drought forced VRA to curtail energy production and thus the supply to ECG. ECG, in turn, had to "iscontinue periodically service to its customers. The situation may improve at the earliest in late 1984 if the year's ixflows to the Akosombo reservoir are sufficiently abundant to bring the reservoir level, which in January 1984 was about 4 a below the lower rule curve, above the curve again. Annex 4 compares planned and actual development of the main operational characteristics. - 83 - 4.03 The deterioration of system reliability also contributed to the shortfall of sales. Already in 1979, operation of ECG's distribution system, particularly in the Accra-Tema, Kumasi, and Takoradi areas, was reaching a critical level with frequent outages and consequent disruption of industrial production, as well as of commercial and residential activities. Since then, the situation has further deteriorated due to the combined effects of the following main factors: - lack of foreign exchange allowed ECG to import only small quantities of spare parts and materials beyond those included in the project; and - deficient quality and quantity of maintenance work due to poor management and lack of adequately trained and motivated staff. 17 4.04 ECG statistics indicate that since 1977 the energy unaccounted for varied between 9% and 15%, which compares rather poorly with the 10% appraisal target. Thesz figures, however, should be used with caution, as the relialility of the statistics is doubtful. 4.05 Operations have further been hampered by recurrent strikes and shortages of fuel and lubricants, which at times forced ECG to stop operation of its generating plants in isolated systems. 4.06 During negotiations, ECG committed itself to making surplus energy generated by VRA available on an interruptible basis for steam generation in industries. Moreover, ECG was to replace the low tariff conceded to various clients for relatively large firm power contracts with the standard industrial tariff, whenever such contracts expired. In late 1977, the utility took steps to meet these commitments. Unfortunately, the initial measures were delayed as they had to be clarified and adjusted. By the time ECG was ready for implementation, VRA's curtailments of energy generation forced ECG to inform the industries using boilers that it could not hold up basic supply, let alone provide interruptible excess energy. 5. FINANCIAL PERFORMANCE 5.01 The years 1977-1982, during which ECG carried out its third power project under Bank/IDA financing, were characterized by a rapid deterioration of Ghana's economy with increasing inflation, devaluation of the currency and a sharp overvaluation of the Cedi. Therefore, the Cedi figures in ECG's financial statements (see Annexes 8, 9, and 10) show little resemblance with those estimated in 1976, when the project was appraised. Furthermore, the distortion of the exchange rate takes much meaning away from US$ equivalents of Cedi amounts. Hence, the following analysis concentrates on the estimated and actual development of the main financial indicators bearing in mind that the uncertainty of the values corresponding to the actual financial figures also reduces the validity of the conclusions drawn from the indicators derived from such figures. 5.02 Table 5.01 below, which should be considered under the limitations indicated above, compares the actual development of sales and of the main financial indicators with that expected at appraisal time. - 84 - Table 5.01 Actual and Estimated Financial Indicators 1977 1979 1981 1982 Sales in GWh - actual 967 972 1005 928 - 1976 estimate 1096 1403 1550 1698 Revenue per kWh sold (0) - actual 0.041 0.082 0.245 0.269 - 1976 estimate 0.038 0.042 0.055 0.063 Rate of return on revalued rate base (%) - actual 1.1 3.1 16.7 10.6 - 1976 estimate 10 8 8 8 - covenanted (adjusted 7.8 6.2 6.2 6.2 after actual revaluation) Operating ratio (%) - actual 96 94 82 86 - 1976 estimate 87 91 92 94 Current ratio (M) - actual 2.2 1.6 1.4 1.0 - 1976 estimate 3.8 4.3 4.5 4.4 Debt/equity ratio - actual 36/64 71/29 40/60 32/68 - 1976 estimate 34/66 28/72 27/73 26/74 Receivables as a % - actual 65 70 59 62 of power revenues - 1976 estimate 24 26 25 25 Internal contribution to - actual 61 2 114 98 investment () - 1976 estimate 23 137 45 10 These figures, together with those of the Annexes 8, 9, and 10, illustrate the worsening trend of ECG's finances, which, in the framework of general economic deterioration, was due to the following more specific main reasons: - Government granted tariff adjustments (e.g. 230% betweei march and July 81 alone) but often late and not in imounts sufficient to compensate fully for the increases in costs. It also happened that VRA could increase its rates far earlier and in larger amounts than ECG. This lag was due to Government's ability to increase VRA's tariffs without referring to Parliament. Deferrals of Government's approval of ECG's tariff increases were apparently due to its reluctance to provide them to such an inefficiently run utility. - Although ECG revalued its assets annually in an environment of galloping inflation, the valuation of the foreign debt at the official exchange rates -- which produced an overvalued Cedi -- resulted in substantial distortions (e.g. an artificially lower debt/equity ratio). - 85 - 5.03 The rate covenant (Section 5.05 of Loan Agreement) essentially called for a return of at least 8% on net fixed assets in operation, adjusted to take into account the results of the revaluation of assets ECG was to carry out in 1977. At appraisal, net fixed assets at the end of 1976 were estimated at about 0 69 million. If, following the revaluation, the value of ECG's net fixed assets exceeded this amount, the covenant required a proportionate reduction in the minimum rate of return. By December 1977, i.e. about six months after the date foreseen in the Loan Agreement, ECG determined the value of its assets as of December 31, 1976, at about 0 96 million. Therefore, the lower limit for the rate of return became 6.2%. ECG's actual rate of return was far below this until 1981. In 1980 ECG was heavily in deficit and recovered in 1981 and 1982 sufficiently to produce a rate of return on net fixed assets substantially above the required 6.2%. However, its relatively high cash generation in 1981/1982 was needed to avoid a financial breakdown of the corporation, as during the period 1977-79 revenues had barely met operating costs including depreciation, and in 1980 not even cash operating costs. Thus, for long periods ECG was not in a position to pay VRA for energy purchased and to meet its debt service obligations. In this desperate situation, the higher 1981-82 revenues allowed to improve somewhat ECG's current ratio frc,i a low of 0.9 in 1980 to 1.0 in 1982, which is still far below the 4.4 estimated at appraisal time. 5.04 The current illiquid position is due to ECG's inability to collect its bills promptly. Throughout the project period, receivables were always in excess of 7 months' revenues, which compares poorly with the maximm of three months' revenues called for in Section 5.09 of Loan Agreement. 5.05 The debt/equity ratio behaved erratically.. In 1978, debt increased as a consequence of project requirements and a first devaluation of the Cedi. The ratio also increased as retained earnings were depleted because of insufficient revenues. By 1982 it was back at about 30/70, i.e. as forecast, as a consequence of improved earnings, increased asset revaluation reserve and approximately constant debt, the latter, of course, largely a consequence of a low level of investment and valuation of foreign debt at the official exchange rate. 5.06 The appraisal team had forecast that net internal cash generation over the project period would finance about 50% of ECG's construction program. Actually, it covered 57% of investment but on a reduced construction program. 6. INSTITUTIONAL PERFORANCE Management 6.01 At the time of appraisal, ECG's management seemed to be performing moderately well, though numerous improvements were necessary- During project implementation, the quality of management seriously deteriorated and ECG did not or could not implement many of the improvements agreed during negotiations. The predominant reason for this was ECG' s inadequate pay, which is even lower than that of VRA, and impedes the recruitment of suitably experienced staff. Daring the past eight years the situation has worsened as Government remains unwilling to make ECG's pay scales competitive with those of the private sector, allegedly because such a step would invite the latter to raise further its own salary scales. Low salaries together with a pervasive demoralization and lack of motivation, with even high level employees having constantly to worry about finding the essentials for subsistence (food, gasoline, etc), has reduced the quality of ECG's management to crisis level. Most top positions are filled by people who are not officially appointed but are acting incumbents. This may reflect the fact that many of the managers are considered inadequate, and often rightly so, for the task they have been assigned. This state of affairs further discourages those who are capable of filling senior positions but whose capability is not recognized. 6.02 ECG' s nren-compliance with the specific covenants referring to management highlights the general trend of events. ECG should have appointed a Chief Engineer in July, 1977. As the utility did not comply until late 1977, the Bank agreed to the appointment of the Acting Chief Engineer on the condition that ECG hire an expatriate to support the Managing Director. The utility was not able to do this, mainly because of a growing lack of foreign exchange. In the meantime, upon the General Manager's retirement, the Chief Engineer has become Acting General Manager and a new Acting Chief Engineer was appointed. During negotiations ECG had also agreed to appoint, by July 1977, two highly qualified accountants who would lead the teams responsible for financial accounts, budgeting, and stores. The Bank agreed to postpone the deadline first to January 1, 1978, then to July 1, 1978. In 1979, ECG still had not found adequate personnel; even worse, both the Chief Accountant and his Deputy left ECG with no experienced successors in sight for either position. The lack of qualified staff also forced ECG's management to subordinate the Commercial Engineer, i.e. the head of the commercial division, to the Chief Engineer. Given the importance of the commercial function, this measure is only understandable as a judgement on the calibre of the man in charge. Personnel 6.o In the course of project implementation, the staff first increased from 5,200 in 1976 to 6,200 in 1979 and then diminished to 5,140 in 1982. Sales per employee were about 160 MWH in 1975 and thereafter varied between 175 and 215 MWh. Though these figures indicate that a slight improvement took place, they form no reasonable basis for a positive judgement as the statistical data is uncertain and the decrease in staff is dae less to a conscious effort to reduce personnel than to ECG' s incapacity to fill vacancies, in particular in posts requiring reasonably trained people. During negotiations, ECG had agreed that (i) with the assistance of experts it would draw up by January 1, 1978, a staffing plan, (ii) thereafter implement the plan as agreed with the Bank, and (iii) until such agreement had taken place, not to increase staff. To determine the staffing plan, ECG used the services of the - 87 - Management Development and Productivity Institute, sponsored by Government, UNDP, and ILO, which studied ECG's manpower and training needs and submitted its suggestions in the second quarter of 1979, i.e. some 16 months late. However, no further action resulted until 1980, when the Bank hired a consultant to study ECG's problems. Slightly later, Government appointed a technical committee to inquire into these problems. The conclusions of the Bank's consultant and the committee coincided; however the consultant recommended a more massive technical assistance than the committtee. The consultant's final report is one of the basis used for the planned next IDA operation with ECG. Training 6.04 At the time of appraisal ECG needed to assess its training requirements and to improve and expand its existing programs. During negotiations, it agreed to survey its needs by July 1, 1977, and to designate a senior staff member responsible for training. Ultimately, Government asked the technical committee which analyzed ECG's staffing problems (para. 6.03) to also submit proposals concerning ECG's training. Implementation of these proposals is planned in the context of the next IDA operation as, at this stage, ECG is not in a position to finance such a program. During the implementation of the Third Project, ECG's personnel manager succeeded with the extremely limited manpower at his disposal in training up to 100 junior staff and up to 22 senior staff per year locally and up to 5 persons per year overseas, but these efforts remained far short of the goal of formalizing manpower training. Covenants 6.05 Annex 5 sets forth the principal covenants in the loan and credit documents and the degree to which they were followed. This Annex shows that ECG did not fulfill practically any of the financial covenants and that there was considerable delay in fulfilling the organizational covenants, if at all. 7. PROJECT JUSTIFICATION The Power Market 7.01 Annex 4 compares actual and appraisal forecast figures for generation, sales, and peak load. It illustrates the effect of the economic crisis, which the appraisal did not anticipate. Indeed, instead of growing as expected in 1976 at about 9% annually, sales were slightly above those of 1976. The differences were mainly due to ECG's capability to supply (para. 4.01). The 1982 figures already reflected the curtailment of VRA's supply related to the excessive drawdown of the Volta Lake. Thus in 1981 (the last year of the appraisal estimate), ECG's actual sales amounted to only 58% of the estimated values. As it is not possible to assume that in the short or medium term the above shortfall will be compensated, it is evident that the economic indicators of the program after its implementation are substantially below the levels expected at appraisal. 7.02 ECG's current sales forecast assumes that, after a drought related dramatic drop from about 800 GWh in 1983 to 520 GWh expe,-ed in 1984, both the run-off of the Volta river and the economic situation will permit a moderate recovery back to the 1977-80 level. ECG hopes to reach this level by 1987 and with a modest 5% per year increase thereafter until at least 1990. However, in - 88 - view of the uncertainty that the current drought has brought about, the economic reevaluation of the project assumes that in future VRA will not be able to provide significant amounts of interruptible energy. Project and Demand 7.03 At appraisal, the project items were deemed necessary to meet the expected increase in demand. Although the latter did not materialize as forecast, the project items are needed to meet the same type of loads as foreseen in the appraisal as the load may build up in the second half of the 80's, or 7 to 10 years later than envisaged in 1976. Least Cost Solution 7.04 The executed subprojects essentially consist of subtransmission and distribution equipment combined in a way that they represent the least cost way to meet the demand. Had the load forecast been accurate the works could have been deferred for a few years or executed at a slower pace. Nevertheless, they would still constitute the least cost way to meet the demand. Return on Investment 7.05 The return on investment was estimated as the discount rate that equates the present values of the benefits and costs associated with the 1977-1982 ECG investment program. Benefits were measured by the forecast revenues from the sales of electricity at the average retail level, using the tariff in effect on July 1, 1984 (see Annex 6). The return for the program is about 36% which compares quite favorably with the opportunity cost of capital for Ghana, estimated to be between 15% and 20%. Project Achievements 7.06 This report deals in various places with specific project objectives. They can be summarized as follows: - The project has provided ECG and Ghana with several facilities which, if properly maintained, will ease the starting phase of an economic recovery by serving industries, and in particular foreign exchange earning activities. - The provision of material and equipment for operation and maintenance has prevented ECG from a total operational collapse. - The EdF study has provided a basis for a reasonable tariff structure by introducing the concept of marginal cost pricing. Now that the principle is set, the specific quantitative proposals will have to be adjusted to take into account developments which the original study did not anticipate. - Bank involvement, though it could not prevent the deterioration of ECG's operation and management, was instrumental in limiting the negative developments and in preparing for a recovery that is the main purpose of the planned next operation. - 89 - 8. BANK PERFORMANCE Assessment of Borrower' s Capability 8.01 It is evident that the Bank's evaluation of ECG's capability to carry out the project was rather optimistic. One of the main reasons for this optimism was the appointment in early 1975 of the former Director of Engineering of TRA to the position of Managing Director of ECG. The new officer took project preparation, which had been lacking, energetically in hand and thus allowed the Bank to appraise the project in 1976, as planned in 1974. 8.02 Unfortunately, ECG was never able to hire the personnel required to strengthen its management and administration; on the contrary, it soon started to lose key people with nobody in sight to replace them. The Bank had from the start seen that one of the main issues was ECG's salary scale which was (and remains) intimately linked to that of public service, and was far too low to permit the successful recruitment of capable Ghanaian personnel. As the Bank did not see any possibility to sever this link, it aimed at the solution of individual problems by agreeing on covenants which foresaw the filling of key positions, the preparation of a staffing plan, and the definition and implementation of a training program, all measures that ultimately could not succeed without a change in salary policy. From hindsight, the question is whether a limited but perhaps more realistic objective, such as putting ECG in the top category of Government-owned companies, would not have been worth pursuing. It would certainly not have prevented, but it may have helped limit the deterioration of ECG' s management. 8.03 An early agreement on a technical assistance package providing for expatriate managers in some key positions, or on a kind of sponsorship (twinning) between a utility in a developing country and the Borrower, might have helped to limit the derioration in ECG. 8.04 Also with hindsight, but on a more specific level, tha Bank should have forcefully intervened in the discussion about the responsibility for the design and supervision of civil works and tried to prevent ECG from taking responsibilities which it would not be able to discharge successfully. Supervision 8.05 Annex 7 gives a schedule of the Bank's supervision missions. The average interval between such missions was 10 months and over the five-year project implementation 3 engineers and 3 financial analysts were successively responsible for supervision, which is not unusual. While the frequency might have been intensified, it is unlikely that more intensive supervision would have provided the basis for any material improvement in project-related performance as the problems encountered derived mainly from fundamental problems with Government's management of its institutions and the economy. Working Relationship 8.06 Relations between Bank staff and Government and ECG officials were at all times good, even when ECG's handling of the special contracts with some industries - which in part was based on ECG' s misunderstanding - caused the affected companies to express some anger (para. 4.06). - 90 - 9. CONCLUSIONS 9.01 The project achieved its main objective which was the installation of the facilities; it did not succeed in helping improve management and operations. In fact, both EOG's conduct of business and the reliability of its electricity supply greatly deteriorated. The project was nevertheless partly successful in providing facilities important for the future development of the country though Ghana will reap the full benefit of these installations some seven years later than expected at the time the Bank appraised the project. The reasons for the above shortcomings were mostly beyond ECG's and the Bank's control, as they were largely determined by political and economic developments in the country which were, in part, related to changes in the international environment. 9.02 It was unrealistic from the Bank's point of view not to try to obtain assurances from Government that it would allow ECG to improve its employment conditions outside the framework of civil service and the complex of the other parastatal companies and then to have agreed on measures that ECG could not be expected to implement without a revision of its salary scales. At the time of appraisal it was certainly not possible to predict the depth of the economic crisis that was to come. Nevertheless, the economy was already struggling and the prospects were not good, in particular one could foresee that the availability of foreign exchange for ECG's operations would at best be precarious. Therefore, it would seem that the Bank could have been more decisive in helping ECG prepare contingency plans. The question is whether, as soon as it became evident that the expected growth in consumption would not materialize in the short term, it should have redirected the utilization of equipment and materials to maintain existing facilities rather than carry out the extensions in full. As the actual developments showed, this wnuld have been possible to some extent. Indeed, ECG used some of the maintenance material and of the spare equipment for extensions and vice versa employed equipment earmarked for project items for emergency repairs. 9.03 The main lesson the Bank should draw from this project concerns its flexibility in responding to unexpected developments. The Bank should not readily agree to modify its projects but, nevertheless, it might better serve the interests of the country and of the specific project goals by supporting, if not inducing, project changes when these might contribute to improved project efficiency or at least to cut financial losses. It seems that in this project some such changes might have been possible. Annex 1 -91 - Page 1 of 2 GHANA ELECTRICITY CORPORATION OF GHANA THIRD POWER DISTRIBUTION PROJECT Loan 1381-G/Credit 689-GR List of Foreign Suppliers and Contractors Description of Plant and Country ot Equipment Origin Supply and delivery of 33 kV England cables and accessories Supply and delivery of 11 k England cables and accessories Supply and delivery of MV England cables, accessories and feeder pillars Supply and delivery of England pilot cables and accessories Installation of cables and Ghana jointing work ionstruction of 33 kV England overhead lines Supply and delivery of England insulators Supply and delivery of India insulators Supply and delivery of England overhead line fittings Supply of overhead line Finland conductors Supply, delivery and erec- England tion of 11 kV switchgear -92- Annex I Page 2 of 2 Description of Plant and Country of Equipment Origin Supply and delivery of Yugoslavia 33 kV switchgear Supply, delivery and Yugoslavia erection of connections to 33 kV/11 kV transformers Modification of existing Yugoslavia switchgear Modification to existing West Germany switchgear Supply, delivery and Yugoslavia erection of 33 kV switchgear in Kumasi Supply and delivery of 11 kV fuse switch and isolator units Supply and delivery and West Germany erection of cooling fans station E Tema Modification to existing England switchgear Supply, delivery and Scotland erection of 33/11/6.6 kV transformers Supply and delivery of England cars, trucks and lorries Supply and delivery of England fork lift trucks Supply and delivery of Canada chain link fencing and ancillary equipment Annex 2 GRANA ELECTRICITY CORPORATION OF GHANA THIRD POWER POWER DISTRIBUTION PROJECT Loan 1381-GICredit 689-GE Disbursements Cumulative disbursements by end of semester of Bank FY in million US$ Appraisal Actual Estimate Credit Loan 1st semester 1977 6.1 2nd semester 1977 10.6 1st semester 1978 13.1 6.3 2nd semester 1978 15.1 9.0 2.8 1st semester 1979 16.6 - 3.6 2nd semester 1979 17.1 5.3 1st semester 1980 18.0 6.4 2nd semester 1980 7.2 1st semester 1981 7.9 2nd semester 1981 8.4 1st semester 1982 8.5 2nd semester 1982 8.9 2nd semester 1982* 9.0 *) drawdown completed July 1982 - 94 - Annex 3 page 1 of 2 pages GRANA: LOAN 1381 AlD CREDIT 689 ECG: THIRD POWER PROJECT CORPLETION REPORT Allocation of Proceeds of Loan Credit 689-GH The expectd and actual amounts disbursed under Schedule 1 of the Development Credit Agreement compare as follows: BS1 MATE US$ ACTUAL US$ All Parts except A-5 & D Category 1 Equipment, materiels, 4,100,000-- 8,817,463.37 supplies, spare parts and vehicles Category 2 Erection of lines and 300,000.-- - substations, and related works Category 3 Consultants' Services 100,000.-- 156,721.62 Part A-5 Category Equipment, materiels, 2,900,000.-- 25,815-01 supplies, spare parts and vehicles. Erection of lines substations, and related works Part D Category 5 Consultants Services 100,000.-- Category 6 Unallocated 1 ,500,000.-- - 9,000,000.-- 9,000,000 - 95- Annez 3 page 2 of 2 pages Loan 1381-GH Disbursements under the above Loan Account were completed on July 7, 1982. The Bank by its cable dated August 10, 1982, informed the Borrower that the undisbursed balance of US$ 1,018.45 is cancelled effective August 10, 1982. Category/Description Original Amount Amount Disbursed usS usS I - Equipment, materials, 6,700,000.-- 8,033,724.75 supplies, spare parts, and vehicles II - Erection of lines and 500,000.-- - substations and related works III - Consultants's Services 150,000.-- 965,256.80 IV - Unallocatd 1,650,00.-- n.a. TOTAL 9,000,000.-- 8,998,981.55 Amount cancelled 8/10/82 1,018.45 Total disbursed 8,998,981.55 8,998,9981.55 -96- Annex 4 GHANA ELECTRICITY CORPORATION OF GHANA THIRD POWER DISTRIBUTION PROJECT Loan 1381-Gi/Credit 689-GH Operational Characteristics 1976 1977 1978 1979 1980 1981 1982 (1) Energy purchased and generated - actual GWh 1029 1101 1118 1086 1116 1151 1001 (2) Energy sold - 1976 estimate GWh 1045 1218 1419 1576 1633 1740 - actual GWh 918 967 1019 972 952 1005 928 (3) = (1) - (2) losses - actual GWh 119 118 99 114 164 146 93 - in Z of (1) 12 11 9 11 15 13 9 (4) non coincidental peak load - 1976 estimate MW 183 201 237 272 291 308 - actual MW 181 191 192 197 201 209 180 (5) number of employees 5160 5860 6100 6190 5830 5350 5140 (6) daily rated workers % of (5) 46 n.a. n.a. n.a. 33 36 36 (7) sales per employee MWh 178 165 167 157 163 189 177 1/ These figures do not appear in the projections. -97- Annex 5 Page 1 of 2 GRANA ELECTRICITY CORPORATION OF GHANA THIRD POWER DISTRIBUTION PROJECT Loan 1381-CH/Credit 689-GM Major Covenants - Status of Compliance Loan Agreement Section Description Status 3.02 Employ engineering consultants - done 4.01 (b) By July 1/77 - appoint Chief Engineer - done (para.6.02) - employ two qualified - not fulfilled, furthermore accountants position of Chief Accountant and Deputy Chief Accountant vacant (para. 6.02) 4.01 (c) By January 1/78 - prepare manpower - prepared by local consultant requirement plan 16 months late (ysra. 6.03) 4.01 (d) By July 1/77 - survey feasibility of - done by Government appointed upgrading operating efficiency committee (para. 6.04) through training, and - appoint senior training officer - done in November 1979 (personnel manager, para. 6.04) 4.04 From January 1/78 ensure customers - attempt to comply led to with electro-boilers supplied only misunderstandings; after 1979 no by surplus hydro power on surplus hydropower interruptible basis 5.02 Audited accounts to Bank not later - not fulfilled (para. 5.05) than 6 months after December 31 (closing date) 5.04 By June 30/77 revalue fixed assets - not fulfilled as of December 31/76; thereafter, revalue fixed assets each year 5.05 (a) Take action to ensure 6.2% return - not fulfilled, several (the 8% modified by revaluation) on insufficient tariff increases net fixed assets in service plus %para. 5.10) 5% in lieu of working capital allowance -98- Annex 5 Page 2 of 2 Loan Agreement Section Description Status 5.05 (b) By July 31 each year basis financial - not fulfilled projections - review adequacy of tariffs and - provide Bank with results of review 5.06 Except as Bank agrees, Borrover not - not met; Bank has approved to incur debt unless current net incurrence of new debt revenues cover maximum future debt service 1.5 times 5.07 Receivables less than or equal to - not met; equivalent to about three months of billing seven months in average 5.08 (a) Apply uniform rates to industrial - done customers, except existing contracts 5.08 (b) Not extend existing contracts - done 5.10 By December 31/77 - Employ financial - not done experts to assist in revaluation of inventories and review inventory management sys tem; - Inform Bank of results and proposed action 5.11 Not undertake construction (other - done than Project and Rural Electrifica- tion) with annual expenditure over $2 million without Bank concurrence Guarantee Section No. Description Status 2.02 Government to provide funds as - done required for project expenditures 3.03 Government to allow ECG to increase - doDe but in insufficient amounts tariffs when VRA iucreases rate to ECG Credit 3.01 By June 30/78 carry out tariff study - done with 20 months delay with assistance of consultants -99 - Annex 6 GHANA ELECTRICITY CORPORATION OF GRAIL THIRD POVER DISTRIBUTION PRDJECT Cost and Benefit Streams - Return on Investment Operation and Year Investments Naintenance Total Costs I. Costs (in million Cedis) 1 1977 12.2 12.2 2 1978 25.1 25.1 3 1979 9.3 9.3 4 1980 10.2 10.2 5 1981 19.9 19.9 6 1982 18.7 18.7 7-25 - 1.00 1.0 Total Sales Increase in Total Sales tbrough Project Benefits a/ (GWh) Sales over 1981 (GWh) (GWh) (Cedis I 10U) II. Benefits 1 1982 928 (76) 2 1983 786 (218) - - 3 1984- 484 (520) - - 4 1985 80 (196) - - 5 1986 860 (144) - 6 1987 925 (79) 7 1988 990 (14) - 8 1989 1059 55 55 255.4 9 1990 1133 129 129 255.4 10 1991 1212 208 208 411.8 11 1992 1297 293 293 580.1 12 1993 1388 384 384 760-3 13 1994 1485 481 481 952.4 14 1995 1598 585 585 1158.3 15 1996 1700 696 696 1378.1 16 1997 1819 815 815 1613.7 17 1998 1946 942 942 1865.2 18-25 1999 2083 942 942 1865.0 Rate of Return = 36% a/ At average retail tariff of Cedis 1.98/kWh, effective on July 1, 1984. 100 Annex 7 GHANA ELECTRICITY CORPORATION OF GHANA THIRD POWER DISTRIBUTION PROJECT Loan 1381-GH/Credit 689-GH Schedule of Supervision Missions Date of Duration Mission staff No of months since Mission in days last mission 6/77 5 Engineer No. 1 - 12/77 7 Engineer No. 1 6 Fin. An. No. 1 2/79 8 Engineer No. 1 14 10/79 8 Engineer No. 2 8 Fin. An. No. 2 10/80 7 Engineer No. 3 11 Fin. An. No. 3 7/81 10 Engineer No. 3 9 Fin. An. No. 3 EOECTRICITY CORPORATION 07 OHAIA THIRD POVER DISTRIBUTION PROJECT Loan 1281-0y/CRBDIT 689-011 Inoome Statement (million U) 1977 1970 179 1980 1981 19a Act. Foreoait Act. Foreoast a, o t o recast Act. forecast Act, Forecas Sales in GVh 967 1096 1019 1277 972 1403 952 1465 1005 1550 928 1698 Revenue per kWh (g) 0.041 0.038 0.053 0.039 0.082 0.042 0.095 0.048 0.245 0.055 0.269 0.063 Revenuess Energy Revenue 39.7 41.6 54.0 50.1 80.0 58.6 90.7 70.7 245*8 85.2 250.0 107.5 Other Revenue 4.1 - 1.5 - 1.5 - 1.5 - 2.6 - 1.5 - Total 43.8 41.6 55.5 50.1 81.6 58.6 92.2 .70.7 248.4 85.2 251.5 107.5 Operating Costs: Purchases from VRA 13.8 14.2 18.6 16.8 24.1 19.2 40.6 20.4 85.4 21.5 62.2 23.4 Other Costa 20.8 16.1 26.1 21.8 37.1 27.1 53.1 35.8 93.1 46.7 119.4 66.1 Depreoiation 7.5 5.8 10.7 6.5 15.3 7.2 20.6 8.6 26.0 10.2 35.2 11.3 Total 42.1 36.1 55.4 45.1 76.5 53*5 122.3 64.8 204.9 78*4 216.8 100.8 Operating Income 1.7 5.5 0.1 5.0 5.1 5.1 (30.1) 5.9 43.9 6.9 34.7 6.7 Interest charged to operation. 2.2 1.5 4.6 1.3 8.4 1.1 9.5 1.0 7.6 2.2 9.5 2.0 Non-operating expenses (0.9) (0.5) 8,1 (0.5) 8.3 (0.5) 7.4 (0.5) 7.4 (0.5) 7.4 (0.s) Net Inoose 0.4 4.5 (12.8) 4.2 (11.7) 4.4 (47.0) 5.3 28.9 5.0 17.8 5.1 Operating ratio (M) 96 87 100 90 94 91 133 92 82 92 66 94 Rate of Return on rate base (%) 1.1 10 0 8 3.1 8 (14.5) 8 16.7 8 10.6 8 March 1984 GHANA BLECTRIOITY CORPORATION OF GHANA THIRD POWER DISTRIBUTION PROJECT Loan 1381-OH/Oredit 689-01 lalance Sheet (Million 9) 1977T~reas 1978 1979. 1980 1981 18 Acti recast Actual Forecast AotuaT Feoast AotuaT Forecast Aotua"orecast Aotus -orocast Assets Fixed Asetes Plant in operation 173.3 123.1 238,3 137.1 333.6 152.7 456.7 192.4 620.6 214.1 837.8 236.2 Leae aooumul. depreciation 66.6 _2.3 10044. A 4. 78.9 24.2 '96.2 32.6 7117. .4 j1 12 Net plant in operation 100T 70.7 137.9 72.5 183.7 73.8 232.5 96.2 293.0 97.1 360.4 97.0 Work in Progress 16.5 15.0 55.0 21.7 54.8 252 61.7 12.6 75.3 27.3 94.0 44.8 Investments 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 015 0.5 0.5 0.5 Current assetes Cash 6.4 9.1 1.8 9.6 1.9 13.0 0.3 14.2 5.0 14.5 7.0 10.3 Receivables 25.7 10.4 31.5 13.4 56.1 15.4 80.2 18.4 144.8 21.4 155.0 27.4 Inventories 90 8.0 12,0 80 3 8. 0 8.0 29.8 8.0 45.0 8.0 Total current asset. -7. T . 73.3 T t IbT.T . T 207.0 T5.7 Total asset. 164.8 113.7 238.7 125.7 312.3 135,8 398.2 149.9 548.4 168.8 661.9 187.9 Liabilities Equity: Government equity 17.3 17.3 17.3 17.3 17.3 17.3 17.3 17.3 17.3 17.3 17.3 17.3 Asset revaluation reserve 45.9 33.3 16*9 41.2 55.6 49.1 120.3 57.2 202.4 67.8 275.3 78.5 Retained earnings 12.5 3.0 0 j52 JU.& 1. 6 (§8.9) . (3. 0) 2 12a 27.1 Total equity 75.7 63.6 33.9 73.6 61.0 84.0 78.7 95.4 189.7 109.0 280.4 122.8 Long term debt 43.0 32.3 134.4 33.7 151.6 32.8 147.6 34.8 128*7 39.4 131.0 44.0 Customer contributions 27.3 10.5 3901 10.5 54.1 10.6 76.4 10.6 106.1 10.7 145.8 10.7 Current liabilities 18.7 .1 31.3 7.9 4 8.5 115.9 _2. 123.9 a I- C jA _- Total liabilities 164.8 113.7 238.7 125.7 312.4 135.8 398.2 149.9 548.4 160.8 661.9 187.9 Current ratio 2.2 3.8 1.4 3.9 1.6 4.3 0.9 4.5 14 4.5 2.0 4.4 Debt/Equity ratio 36/64 34/66 80/20 31/69 71/29 28/72 65/35 27/73 40/60 27/73 32/68 26/74 Receivables as % of revenues 65 24 50 26 70 26 88 25 59 25 62 25 March 1984 GHANA, RLECTRICITY CORPORATION OF GHANA THIRD POWER DISTRIBUTION PROJECT Loan 1381-GH/Credit 689-OH Sourcen and Applicatione of Funds (Million 1) 1977 1978 1979 1980 1981 V82 Actual Forecast Actual Forecast Aotual Forecast Actual Forecast Actual Forecast Actual Forecast Sources Internal cash generation: Operating Income 1.7 5.5 0.1 5.0 5.1 5.1 (30.1) 5.9 43.9 6.8 34.7 6.7 Depreciation 7.5 5.8 10.7 6.5 15.3 7.2 20.6 8.6 26.0 10.2 35.2 11.3 Consumer contributions 3.7 2 2.2 0.1 0.1 2.9 0.1 - 0.1 2 0.1 Gross internal generation 12.9 13.6 13.1 11.6 22.1 12.4 (6.6) 14.6 73.2 17.1 72.4 18.1 Lees: - debt service 2.2 6.5 4.6 4.6 8.4 4.7 9.5 4.9 7.6 5.4 9.5 7.0 - dividends - 2.0 - 2.0 - 2.0 - 2.0 - 2.0 - 2.0 - increases in non cash working capital 3.2 1.7 (3.7 2.6 13.6 .0 (18.1)W 4 .. j ! w Net cash generation 7.5 3-4 12.2 2.4 0.2 3.7 2.0 4.4 22.6 6.5 18.3 1.7 Borrowings 7.3 11.0 9.5 3.6 8.0 1.3 3.9 4.2 2.0 7.0 2.4 8.0 Other sources (O0) 0.5 (1.0) 0.5 1.1 0.5 2.7 0.5 - 0.5 - 0.5 Total sources 14.7 14.9 20.7 6.5 9.3 5.5 8.6 9.1 24.6 14.0 20.7 10.2 Applicationst Construction 12.2 15.0 25.1 6.1 9.3 2.7 10.2 8.8 19.9 14.4 18.7 16.4 Increases in cash balance 2.5 (2.1j k.4) O U[ - 2.8 16 .3 4.7 (0.i) 2.0 C.2 Total applications 14.7 14.9 20.7 6.5 9.3 5.5 8.6 9.1 24.6 14.0 20.7 10.2 Internal contribution to investment () 61 23 49 39 2 137 20 50 114 45 98 10 March 1984