Peru - Matucana Power Project

Report No. 868 CONFIDENTIAL CIRCULATING COPY TO BE RETURNED TO REPORTS DESK INTERNATIONAL BANK FOR RECONSTRUCTION AND DEVELOPMENT PROJECT PERFORMANCE AUDIT REPORT ON PERU MATUCANA POWER PROJECT (LOAN 511-PE) September 23, 1975 Operations Evaluation Department  PROJECT PERFORMANCE AUDIT REPORT PERU MATUCANA POWER PROJECT (LOAN 511-PE) PREFACE This report presents an audit of achievements under the Peru Matucana Power Project, for which Loan 511-PE of September 7, 1967, in the amount of US$17.5 million was fully disbursed in November 1973. This audit is based mainly on correspondence and reports in Bank files (Loan and Guarantee Agreements, Appraisal Report, President's Report, Progress Reports, Supervision Reports, and correspondence between the Bank and the Borrower), as well as discussions with staff members of the Bank and of Empresas Electricas Asociadas. A Project Completion Report, pre- pared by the Latin America and Caribbean Regional Office in April 1975, was also useful in the preparation of this report. In February 1975, a three-day visit was made to the Republic of Peru in connection with this report. The valuable assistance of Empresas Electricas Asociadas is gratefully acknowledged.  PROJECT PERFORMANCE AUDIT REPORT PERU MATUCANA POWER PROJECT (LOAN 511-PE) PROJECT DATA Loan Amount US$17.5 million Amount Disbursed US$17.5 million Date of Loan Agreement September 11, 1967 Date of Effectiveness December 28, 1967 Original Closing Date November 1, 1972 Final Disbursement Date November 1, 1973 Exchange Rates-(soles) Through September 1967 US$1 = S/.26.8 September 1967 to Present US$1 = S/.38.7  PROJECT PERFORMANCE AUDIT REPORT PERU MATUCANA POWER PROJECT (LOAN 511-PE) TABLE OF CONTENTS Page No. Summary i - iii I. Introduction 1 II. The Bank and the Loan 1 III. Project Implementation 4 IV. Project Costs 5 V. EEA's Financial Evolution and the Financial Plan 7 VI. Economic Justification and Operating Efficiency 11 VII. The Role of the Bank 15 VIII. Conclusions 17 List of Annexes 1. Project Composition at Appraisal 2. Original and Final List of Goods 3. EEA's Selected Operating Statistics 4. EEA's Income Statements 5. EEA's Expected and Actual Financial Plan, 1968-71 M1p: Power Plants in Central Peru  PROJECT FER.ORMANCE AUDIT REPORT PERU MATUCANA POWER PROJECT (LOAN 511-PE) SUMMARY Loan 511-PE for US$17.5 million to Empresas Electricas Asociadas (EEA) was signed on September 11, 1967, and fully disbursed on November 1, 1973, one year behind schedule. The purpose of this loan was to finance the foreign exchange cost and.interest during construction of a US$50.5 million project. The main component of the project was the Matucana hydro- electric power plant, which together with its transmission lines accounted for 75% of the project cost. The other components of the project were equipment for expansion of the distribution and communication systems, con- sultants' services, and interest during construction. The hydroelectric plant of Matucana was to be a run of the river plant, with a maximum capacity of 120 Mw. The purpose of this plant was to meet forecast increases in power demand in EEA's system during the 1971-73 period. In 1973, the first stage (342 Mw) of the Government hydroelectric project of Mantaro was expected to be in operation. This plant was to provide fo*r an interconnected systemn in the Peruvian Central Region where EEA's system is located (see map). After 1973, EEA was to meet further increases in power demand by purchasing power produced by the Government hydroelectric project of Mantaro. The project started slowly, and construction of Matucana did not begin until about one and a half years later than expected. Several rea- sons explain this delay. One reason was the 45% devaluation of the sol in September 1967, which produced a difficult financial situation for EEA un- til March 1968, when the Government granted a tariff increase. Another was the reluctance of the Swiss group that traditionally had been financ- ing EEA's expansion plan to continue its support. However, the Bank inter- ceded and the Swiss group finally agreed to resume its support of EEA in September 1968. Finally, a change in Government in October 1968 caused a further delay. Construction of Matucana finally started in February 1969. Despite these delays, Matucana was operational only about ten months later than expected at appraisal. Important reasons for the speed in completion of the works were the favorable geological conditions en- countered in the excavations and the good organization of the contractors, who were all Peruvian. The final cost of Matucana in local currency was about 67% higher than expected (excluding contingencies). This substantial cost overrun is - ii - mostly accounted for by price increases. The 10% contingency included in the project was not adequate insofar as the devaluation of the sol, about one week before consideration of the loan by the Board, had al- ready produced an estimated increase in Matucana's cost of about 13%. The financial performance of EEA during the 1968-71 period was uneven. EEA was able to obtain a rate of return on net fixed as- sets slightly higher than forecast (9.2% on the average against 8.8%). Operating income (before interest and taxes) during the 1968-71 period was 33% higher than forecast, but despite this satisfactory performance internal cash generation was little more than one quarter the expected amount. This shortfall is mainly explained by reduced depreciation and connection fees, and increased dividend payments and debt service. Con- sequently, EEA had to fill the gap through borrowing or stock sales. But, because of the political situation in Peru, EEA could not rely on long-term borrowing or stock sales and had to get the necessary funds through extensive medium- and short-term borrowing. This development shows the weakness of the indenture test on which the Bank relied to prevent the incurrence of excessive borrowing instead of the traditional debt service coverage ratio. Whether different financial covenants would have led to different financial performance by EEA is uncertain, however, since most factors causing the rapid growth of short-term debt were not of EEA's making. The rate of return for EEA on the incremental investment in Matucana versus a thermal alternative was 12% in the appraisal of this loan. The audit rate of return is higher than 12% because of the sub- stantial increase in oil prices. Matucana was also probably the best alternative in the context of the Central Region system, in light of the uncertainties at appraisal. At that time, completion of Mantaro was most likely to be delayed beyond 1973. A possible lower reliability of power supply from Mantaro also contributed to diminishing the relative advantage of the thermal alterna- tive. The completion of Mantaro almost on schedule and the slowdown in the growth of demand for electricity would have resulted in a low audit rate of return on Matucana in the broader context of the whole Central Region. But these factors were more than offset by the substan- tial increase in oil prices, and the audit rate of return on Matucana in this broader context is about 11%. The Bank played an important role in the successful completion of the project. It gave valuable advice to EEA in the selection of alternatives to fill the gap in energy generation during the 1971-73 period., The decision to allow only national competitive bidding for civil works - iii - was another good initiative of the Bank, which was later confirmed by the excellent performance of the Peruvian contractors. However, another Bank decision in the area of procurement -- the use of the Bank loan for imported goods and services -- seems to have had a negative effect on the project. It increased the pressure to obtain other sources of finance, adding to the financial difficulties of EEA at a time when market condi- tions for obtaining additional borrowing were not favorable. In contrast with its valuable participation in the construction of Matucana, the Bank's contribution in the broader context of the whole power sector was less important. There was a lack of major institutional objectives in relation to the organization and strengthening of the sector. The main Bank contribution in this broader context was the financ- ing of the feasibility studies for the hydroelectric project of Salto Sheque, which will contribute to the analysis of alternatives for meeting the power demand in the Central Region after 1978.  PROJECT PERFORMANCE AUDIT REPORT PERU MATUCANA POWER PROJECT (LOAN 511-PE) 1. INTRODUCTION 1.01 Loan 511-PE was the fourth Bank Group loan to the Peruvian electric power sector. All loans regarding this sector have been made to Empresas Electricas Asociadas (EEA), a privately-owned electric utility company serving the greater Lima-Callao area.1!/ EEA and its associate, Energia Hidroelectrica Andina (EHA), had a generating capacity of 492 Mw in 1965 and were the largest of the privately-owned electric power companies. EHA was established in 1946 with the purpose of financing, constructing and operating the power plants that were going to be used by EEA. EHA was created because a closed mortgage secured the debts of EEA, preventing it from obtaining loans in the substantial amount required to finance the expansion of its facili- ties. EHA had constructed two plants: Moyopampa (63 Mw) and Huampani (31 Mw), and all the power generated by them was bought by EEA (see map). 1.02 The Bank's association with EEA started in 1960 with Loan 260-PE for US$24 million. This loan helped finance the first stage of the Huinco (120 Mw) hydroelectric plant and the Marcapomacocha diversion. The second stage of this project was subsequently partially financed by the second Bank loan (365-PE) in 1963 for US$15 million. The third Bank loan (464-PE) in 1966 was for US$10 million and was intended to help finance the expansion of EEA's transmission and distribution system. The overall performance of EEA under these Bank loans was satisfactory. II. THE BANK AND THE LOAN 2.01 The EEA first approached the Bank in connection with the Matucana Power Project in 1966,and the loan for US$17.5 million was signed in September 1967. The purpose of this project was to meet forecast increases in power demand on EEA's system during the 1971-73 period. The system generating capacity would be fully utilized by 1970 and therefore a plant of about 100 Mw- 120 Mw to meet the capacity shortfall would be required for service in 1971 to meet forecast demand until 1973. EEA would meet further increases in power demand by purchasing power produced by the Government Hydroelectric Project of Mantaro, whose first unit (114 Mw) was expected to be in operation 1/ The Government is now the main shareholder of EEA, as a result of changes in legislation in 1972-73. - 2 - by early 1972 and the second and third units (114 Mw each) at six- monthly intervals thereafter to complete the first stage of the project (342 Mw) by early 1973.1/ 2.02 Two main issues were raised by the Bank before mounting an appraisal mission. The first was about the selection of the alternatives to meet the expected demand in the 1971-73 period. At the Bank's sug- gestion, several other alternatives were studied in addition to that originally proposed by EEA. The analysis showed that Matucana was still EEA's least cost alternative for discount rates below 12%. 2.03 The other issue raised by the Bank was the effect on EEA's financial plan, during the construction period of Matucana, of the purchase of the generating plants of EHA (Moyopampa and Huampani),which were an integral part of its own system. All the power produced by those plants had been bought by EEA, and as EHA was about to start the construction of a new hydro plant (Pativilca) in other concession area, the Bank sug- gested during the negotiations of the third power loan to EEA (Loan 446-PE in 1966) that EEA should try to purchase EHA's Moyopampa and Huampani generating plants. EEA therefore included provisions for the purchase of these plants in its financial plan during the construction period of Matucana. As it was not clear from EEA's financial plan what was the net effect of the purchase of EHA's plants, the Bank wanted assurance that the construction of Matucana would have priority over the purchase of EHA's plants and that Bank funds would not be channelled to ERA. The financial plan was revised by EEA and a plan for the acquisition of EHA's plants in three stages was presented to the Bank. 2.04 The Bank found these revisions satisfactory and an appraisal mission visited the country in March-April 1967. The mission found the Matucana project suitable for a Bank loan of about US$17.5 million (including US$1.5 million interest during construction). The submission of this report to the Loan Committee was delayed until late June 1967 because of the need to wait for a decision on the Bank's lending policy to Peru.2/ The loan was finally approved by the Loan Committee 1/ As demand developed, the Government planned to install three additional 114 Mw units to bring up the project to its full capacity of 684 Mw. This plant was expected to meet market requirements for the Central Region of Peru until about 1978. 2/ Because of the unsatisfactory financial situation of Peru in mid-1967, the Bank decided not to lend for any project requiring financial sup- port from the budget until sufficient public savings were assured to finance a substantial proportion of the public investment program. But the Bank decided to continue processing the loan, because Matucana did not need budgetary assistance. EEA was expected to raise the funds to cover the local expenditures on the project from its own earnings and from sales of its securities in the private market. -3- in July 1967, but with the proviso that it would not be used for the financing of any local currency costs, so that if Peruvian suppliers were successful in the bidding, EEA would have to use resources other than the Bank loan to finance purchases from them. This represented a change from the original version of the appraisal report and also from the procedure followed in previous Bank loans to EEA. 2.05 Negotiations were held in August 1967 . Agreement was reached on a new financial plan that incorporated the changes in use of the Bank funds just referred to and some changes that EEA wanted to introduce in its contract with the civil works contractors. These changes led to a reduction of about US$2.6 million in the foreign exchange cost of the project, but the original loan amount remained the same because some US$2.6 million of foreign exchange costs for transmission and distribu- tion works previously planned for financing under future loans were brought into the project. 2.06 Since the changes in the financial plan implied that some of the reserves built into the original plan were now being used, it was necessary to make an issue of debentures for Sw F 25 million (US$5.8 million) a condition for the effectiveness of the loan. 2.07 - - The loan for US$17.5 million was approved by the Board on September 7, 1967. The total cost of the project, including interest during construction, was estimated at US$50.5 million, of which US$16.0 million would be the cost of imported services and goods and US$1.5 million interest during construction. 2.08 The main component of the project was the Matucana hydro- electric power plant which, together with its transmission lines, accounted for 75% of the project cost (Annex 1). The Matucana plant would be essentially a run of the river plant located on the Rimac River, about 80 km from Lima. The plant would have a maximum capacity of 120 Mw and a firm energy of 440 GWh/year. The other project components were equipment for the expansion of the distribution and communication systems, consultants' services, and interest during construction. 2.09 In addition to the standard loan covenants, special covenants were included to ensure that construction of Matucana would have priority in the use of funds over the purchase of assets from EHA and that the Bank's approval be sought for any changes in the financial plan and for financing of the distribution system expansion in 1969-70. The usual rate covenants were not included because the Bank felt that the provisions of the Electric Industry Law -4- 1/ regulating tariffs were adequate. Instead, an additional covenant was included by which the Government agreed to inform the Bank of any proposed changes in the law and to give the Bank an opportunity to comment. It was also stipulated that any change in the law that may adversely affect the company would be an event of default of the loan. III. PROJECT IMPLEMENTATION 3.01 The project had a slow start,and construction of the Matucana plant did not begin until FebrUary 1969, about one and a half years later than expected. But despite this substantial delay, Matucana was in operation in October 1971, only 10 months later than scheduled. How- ever, the closing date of the loan was postponed by about one year because of the late inclusion in 1971 of the feasibility studies for the hydroelectric project of Salto Sheque. The loan was finally closed on November 1, 1973. 3.02 The first delay in initiation of the works arose because of the 45% devaluation of the sol on September 1, 1967, only a few days before consideration of this loan by the Executive Directors. This devaluation seriously affected the contribution to be expected from net internal cash generation to the financing of the project because the service on the large outstanding external debt increased proportionally. At the same time, the Government delayed consideration of a tariff increase to compensate for the effects of the devaluation. Another important contributing factor, also in late 1967, was an advance payment made to EHA for the purchase of its plants. This ,advance payment amounted to about 30% (US$1.8 million) of the proceeds from the sale of the debentures, and was in excess of the US$1.3 million ceiling agreed upon during negotiations. The advance was made to help finance the substantial cost overruns in construction of Pativilca (which - was finally completed in December 1967). 3.03 The combination of these factors led to a very tight financial situation until March 1968, when the Government finally granted a 21% tariff increase. This increase helped improve EEA's financial situation, 1/ The Electric Industry Law of July 1955, provided for periodic tariff adjustments aimed at producing a profit of 11.5% on the common share capital and free reserves of large electric companies. According to this law, tariffs were subject to review every three years, but they could be revised at shorter intervals at the request of the company or the State Electricity Authority. The law also provided for revaluation of installations on the basis of the current replacement value as determined by technical ap- praisals. - 5 - and the revaluation of the assets that was underway would have permitted a new debenture issue which the company expected to place in Switzerland. But another difficulty arose because the Swiss group that traditionally had been financing EEA's expansion plan was reluctant to continue to do so. The change in attitude of the Swiss group seems to have been the consequence of their disagreement with the top management of EEA and of the difficult financial situation in Peru at the time. 3.04 As a consequence of the lack of support from the Swiss group, the construction of Matucana was further delayed. By August 1968, it was already delayed by one 'year and the evolution of demand for power indicated that a shortage would occur by 1971. Since the construc- tion of Matucana did not appear financially viable, the alternative of a thermal plant at about half the cost was being seriously considered by EEA. But since this would not have been the best alternative, the Bank played an important role in helping EtA in its efforts to obtain the financial support of the Swiss group. These efforts were finally successful and by September 1968 the Swiss group indicated its willingness to continue to support EEA's financial plan. After a further delay due to the change in Government in October 1968, the work on Matucana started in February 1969. However, this delay was not a complete loss since EEA carried out additional site and design studies which resulted in improvements in project design. The principal design change affected the intake, surge chamber and penstock. The major improvement was in the penstock,which was going to be constructed as a lined pressure tunnel instead of a surface penstock. These changes were reviewed by the Bank, and were considered improvements over the original plan. 3.05 Construction proceeded more rapidly than expected at appraisal; the first stage was inaugurated in October 1971 and the second stage in February 1972. An important reason for the speed of construction was the favorable geological conditions encountered in the excavations and the good organization of the contractors, who were all Peruvian. IV. PROJECT COSTS 4.01 Comparison of actual versus expected project costs is only- possible in the case of the Matucana plant (75% of the estimated project cost). Lack of detailed information and changes in the scope of the work does not permit a meaningful comparison of expected and actual costs of the other project components. 4.02 The actual cost of Matucana was S/.1,474 million (US$38.1 million) while the estimated cost (including contingencies) was S/.971 million (US$36 million), as follows: -6- Appraisal Estimates L Actual Cost ---million) --- --- (million)--- Preliminary Works 48.8 1.8 64.9 1.7 Civil Works 656.0 24.3 779.6 20.1 Mechanical Equipment 169.6 6.3 325.1 8.4 Other Costs 96.7 3.6 305.3 7.9 Total 971.1 36.0 1,474.9 38.1 /a Including contingencies. lb Exchange rate: 1 US$ = S/.26.8 /c Exchange rate: 1 US$ = S/.38.7 4.03 The cost overrun in terms of local currency (excluding contingencies) was then about 67% and it is mainly explained by price increases before the start of construction. The devaluation of the sol in 1967 directly increased Matucana's costs, in soles, about 13%. Internal prices rose by 34% between 1966, the year in which the costs were based, and 1969, the year in which construction actually started. This gave rise to another 24% increase in the project cost. The combined effect of devaluation of the sol and higher domestic prices produced then about a 37% increase in the total costs (in soles) before construction actually started in early 1969. Additional price increases in the construction period (1969-71) led to a total cost increase at- tributable to higher prices of about 51%. 4.04 The total contingencies included in the project were only 10%, and they were clearly inadequate insofar as the devaluation of the sol,about one week before consideration of the loan by the Board,had already produced an increase of about 13% in the cost of Matucana. 4.05 The list of goods for the project (Annex 2) suffered considerable changes. There was a reduction of about US$2 million in items related to Matucana (Items 1-3) and a corresponding increase in the expenditure on communication and distribution equipment and on consulting services. The main reason for the reduction in the ex- penditure related to Matucana was that local industry was able to supply some of the items that were assumed would be imported. The increase in distribution expenditures is explained by a substantial increase, particularly after 1970, in the provision of electric service - 7 - to low income suburban areas. This development is illustrated by the evolution of the number of residential customers, which increased by 42% in 1970-73, as against 21% in 1967-70 (Annex 3). In the case of the communication equipment (Item 6 of Annex 2), there was a con- siderable change in the scope of the works, which accounts for the increase in expenditure on this item. 4.06 Finally, the increase in the cost of consulting services is explained by the changes in design in Matucana, consulting services for the design of the communications center and the feasibility study for the Salto Sheque hydro project. This feasibility study was added in 1971 as a means of contributing toward the study of alternatives for meeting the demand of the Central Region of Peru after 1978, when Mantaro would be fully loaded. V. EEA'S FINANCIAL EVOLUTION AND THE FINANCIAL PLAN 5.01 EEA was able to achieve a net operating income (before taxes and interest) over the 1968-71 period which was sufficient to obtain a rate of return on net fixed assets slightly higher than forecast (9.2% on the average as against 8.8%) (Annex 4). But despite this satisfactory performance, only 16% of the requirements of funds during the 1968-71 period were financed from sources other than borrowing, while the appraisal assumed that internal funds would finance 62% of the total requirements. 5.02 The satisfactory rate of return on net fixed assets was obtained despite devaluation of the sol. The Government granted a 21% tariff increase in 1968, and at the same time net fixed assets were revalued by about 40%. The rate of return on the revalued net fixed assets was 9% in 1968 as against 8.9% estimated at appraisal, and continues to be slightly higher than the appraisal estimates over the whole 1968-71 period (see below). A major factor explaining this satisfactory development was EEA's tight control on operating exygnses which were only 4-7% higher than forecast over the whole period- The role of this successful containment of operating expenses became even more important because of the lower than expected increase in the demand for electricity: 28% by 1971, compared with 33% forecast at appraisal, over the 1968 level: 1/ Another contributing factor was a sharp reduction in depreciation allowances in 1968 and following years. -8- Rate of Return on Net Fixed Assets Operating Operating Operating Appraisal Revenue Gwh Sold Expenses Income Estimates Actual (actual as a percentage of estimated) --------(%)------- 1968 115 98 107 138 8.9 9.0 1969 114 95 106 139 8.8 9.3 1970 111 94 105 129 9.2 9.9 1971 110 94 104 128 8.5 8.8 5.03 Net operating income (before interest and taxes) was S/.448 million higher than forecast in the 1968-71 period, but despite this satisfactory performance, there was a shortfall in internal cash generation (Annex 5). The reason for this uneven performance can be found in the comparison of the expected and actual financial plan. But a direct comparison is not possible because of the devaluation of the sol and its effect on the borrowing and debt service of EEA, which was almost entirely in foreign currency. In order to make the comparison possible the borrowing and debt service in the financial plan at appraisal have been adjusted to reflect the effect of'the devaluation of the sol (Annex 5). On that basis, the expected and actual relative participation of each source of funds in the financial plan for the 1968-71 period were: Expected /- Actual ------------------------(M --------- Net Internal Cash Generation 38 9 Sales of New Stock 10 7 Borrowing Long-term Borrowing 51 53 Short- and Medium-term Borrowing 1 _1 Total Funds Required 100 100 La These estimates have been derived from the appraisal forecasts, which have been adjusted to reflect the devaluation of the sol. Net internal cash generation was little more than one quarter the expected amount, falling short by S/.716 million (Annex 5). A breakdown of the contribution of the different items to the S/.716 million actual shortfall follows: -9- Effects on Net Internal Cash Generation Positive Neaative -----("/. million)----- Net Operating Income +448 Depreciation and Expansion Fund - 310 Connection Fees - 242 Taxes and Director's Bonus - 13 Debt Service - 421 Dividends - 178 Total +448 -1,164 Net Effect -716 5.04 This shortfall in net internal cash generation is mainly explained by reduced depreciation and connection fees and increased dividend payments and the higher debt service. Consequently, EEA had to fill the gap through borrowing or sale of stock. The shortfall in depreciation is due mainly to a reduction in the depreciation rate which was introduced by the Government in 1968 at the time of the revision of tariffs. The actual rate applied was 3% while at appraisal it was assumed to be 5%. This new rate is reasonable because EEA's system is almost all hydroelectric, but it diminished the funds obtained through internal cash generation. 5.05 The increase in dividend paymnts was due to EEA's decision to distribute its commercial profits to shareholders in cash rather than by issuing stock dividends, as it had previously done. In this way, EEA attempted to improve the quotation of its shares. No sales of. shares had taken place since 1968, and rumors about the likely nationalization of the company had produced a significant reduction in the quoted price of the shares. The quotations improved somewhat in 1970, after the distribution in cash of the dividends, but changes in the law did not permit new issues of shares for private subscription. According to the new electricity law in 1972, all capital increases haVe to-be subscribed entirely by the Government. 5.06 The increase in debt service was caused by the need to rely on substantial short and medium term borrowing and the unexpected increases in worldwide interest rates. Changes in the political situation in 1968 made it very difficult for EEA to raise additional funds from long-term borrowing,and this explains the increase in debt service over tiat forecast (adjusted for the devaluation of the sol) by about 25% (S/.421 million) during the 1968-71 period (Annex 5). Most of it is explained by increases in the medium and short term debt service which was S/.444 million higher than expected. This development - 10 - shows the weakness of the provisions of the indenture test as a means of preventing the incurrence of excessive debt. In this loan, the Bank relied on the open-ended indenture created in 1957 with Schroder Trust Company (New York) as a trustee instead of the traditional debt service coverage ratio. But the actual indenture tests were above the minimum required over the whole period, while at the same time EEA was able to accumulate substantial short- and medium-term debt, as can be seen in the following table: Evolution of Key Financial Indicators 1968 1969 1970 1971 Appraisal Appraisal Appraisal Approisat Estimate Actual Estimate Actual Estimate Actual Estimatc Acttml Indenture Tests 1. Income before interest and taxes as percentage of consolidated interest (minimum 150%) 220 230 226 228 243 217 259 200 2. Net fixed assets as percentage of consolidated debt (minimum 150%) 213 184 195 180 210 173 201 16G Interest and Long-Term Debt Tests Times interest covered by income before interest 2.0 2.1 2.0 2.1 2.1 1.9 2.3 1.6 Times debt-service covered by income before interest plus depreciation 2.4 1.8 2.4 1.2 2.4 1.6 2.3 1.0 Debt/equity ratio 47/53 54/46 50/50 55/45 49/51 58/42 48/52 62/38 The main reason for this failure of the test is that as it is based on the "consolidated" debt, it does not reflect the effect of Bank overdrafts, debenture notes and other forms of short-term borrowing. The traditional debt-service covenants would have been more effective in preventing the substantial increase in borrowing. The total debt, not only the consolidated one, is used in the estimation of the debt-, service ratio and would then reflect not only the effect of interest on the debt but also changes in the term structure of the debt. - 11 - 5.07 Whether different financial covenants would have led to different financial performance by EEA is uncertain, however, since most of the factors causing the rapid growth of short-term debt were not of EEA's making. VI. ECONOMIC JUSTIFICATION AND OPERATING EFFICIENCY 6.01 Matucana was justified in the appraisal report as the least cost alternative for meeting the expected demand on EEA's system during 1971-73. In that context, Matucana was clearly more advantageous than any of the thermal alternatives for rates of discount lower than 12%. Therefore, the rate of return on the incremental investment in Matucana versus a thermal alternative was also 12%. 6.02 But while Matucana was clearly the best alternative for EEA, it is very difficult to arrive at a definitive answer on whether Matucana was also the best alternative in the broader context of the Peruvian Central Region as a whole, in light of the information available at the time of appraisal of Loan 511-PE. A crucial factor in the analysis of alternatives within this broader context was the Government Hydroelectric Project at Mantaro, since an interconnected system serving all Central Peru was to be built at the same time. The first stage of this project (342 Mw) was expected to be in operation in 1973 and to meet the increases in demand in the Central Region, which were estimated at about 100 Mw/year, until 1975 when it would be fully loaded and the second stage of the same project (342 Mw) would start operation. Mantaro was then expected to meet the increases in demand in the Central Region until about 1978. 6.03 The first stage of Mantaro was already being constructed at the time of the appraisal of Loan 511-PE and it was to be completed regardless of the alternative chosen to meet the demand for power during the 1971-73 period. The investment of the first stage of Mantaro could then be regarded as a "sunk cost" from an economic point of view with regard to the decision to fill the 1971-73 gap in the demand for power. Therefore, the relevant cost for the energy supply from Mantaro to the EEA system was not the full cost, but only the variable running costs, that is, only those costs that were directly associated with the operation of the plant. Therefore, a thermal plant that would operate only during the 1971-73 period and would then be kept in reserve, with its power production being taken over by Mantaro, might have been a better alternative than Matucana. The main implications of this alternative are that the first stage of Mantaro would have been fully loaded one year earlier, and therefore the second stage of this project would have had to be moved ahead about one year. In the same way, after 1978 about 100 Mw of generating capacity/year would have had to be moved ahead one year. A rough estimate of the rate of return on the investment - 12 - in Matucana versus this alternative indicates a return of about 9%.kI This result suggests that from a purely economic point of view Matucana would not have been the least cost solution in the broader context of the Central Region as a whole. 6.04 However, there are at least two other important factors that do not permit a definitive answer to this problem. The first is the likelihood that the completion of Mantaro would be delayed beyond 1973. Mantaro is located in a relatively inaccessible part of the high Andes, and although construction appeared technically feasible, some difficulties were already being encountered at the time of the appraisal of Loan 511-PE. It was then regarded as highly probable that Mantaro's construction schedule would not be met. Any delay in comple- tion of Mantaro would require the thermal alternative to operate beyond 1973 and therefore, its relative advantage over Matucana would diminish. The rate of return of Matucana is highly sensitive to these delays and a two-year delay in completion of Mantaro would raise the return to about 11%. 6.05 The second factor that diminished the relative advantage of the thermal alternative was a possible lower reliability of power supply from Mantaro. The river on which Mantaro is located was known to be susceptible to serious bank slides and the transmission lines would go through a very difficult and mountainous area. It is difficult to incorporate these elements into the rate of return on the investment in Matucana, but they would definitely have a positive effect on it. 6.06 The net effect of the two factors of uncertainty presented is to raise the estimated 9% rate of return on the additional investment necessary for Matucana. It is, however, difficult to assess the magni- tude of that increase given the information available at the time of appraisal of Loan 511-PE, but it seems that the decision to build Matucana was probably correct in the broader context of the Central Region, given the uncertainties at the time of appraisal. 6.07 The audit rate of return on Matucana for EEA is estimated to be higher than the 12% rate of return expected at appraisal because of the substantial increase in oil prices. 6.08 Matucana was in full operation one year later than expected, but this delay did not affect the profitability of the investment 1/ Assuming that the cost of capital is about 10% and using fuel prices expected at the time of the appraisal. - 13 - because of the lower than expected growth in the demand for electricity.!! Despite this lagging demand growth, the plant was necessary in EEA's system in 1973. It has been operating normally, and the energy generated has been in general more than the appraisal expectation of about 540 Gwh/year. 2/ 6.09 The lower than expected growth in the demand for electricity and the completion of the first stage of Mantaro in late 1973, almost as scheduled, seem to imply that in the context of the Central Region system the investment in a thermal plant that would have discontinued operations after the completionof Mantaro would probably have been a better alternative than Matucana. However, the substantial increase in oil prices 3/ has more than compensated the negative effect of those factors and this explains the audit rate of return of about 11% for the extra investment in Matucana (compared with the thermal plant) in the broader context of the Central Region power system. 6.10 The lower than expected growth in demand for electricity raises the issue of whether the forecast was too optimistic or whether there was an actual slow down in demand growth. The latter seems to be the answer in this case because the forecast increase in demand was in line with actual growth in the preceding years (Annex 3). For example, the system peak demand (SPD) grew by 11% in 1966 and the forecast overall growth in SPD during the 1967-71 period was 46%, which was very close to the 44% actual increase in SPD in the preced- ing four-year period (1963-67). Therefore, the forecast of SPD seems adequate in light of the information available at appraisal. 6.11 The reason for the lower than expected growth in SPD is then an actual slow down in its growth after 1967. This slow down can be closely associated with the reduction in the rate of growth of industrial 1/ In the appraisal of Loan 511-PE, it was assumed that system peak demand (SPD) and sales of electricity would grow at about 10% per annum, but the actual rate of growth during the 1967-71 period was about 8%. 2/ The energy generated by Matucana has been: 1972 - 475 Gwh; 1973 - 604 Gwh; and 1974 - 678 Gwh. 3/ Oil prices in 1973, the only year in which the thermal alternative plant would have had to operate if Matucana had not been built, were about four times higher than in 1966, the year of project appraisal. - 14 - and commercial demand (about 80% of total demand) in the period 1967-71 as compared with 1963-67, as can be observed in the following table: Total Increase in Demand 1963-67 1967-71 M-------------------( -------- Industrial 72 53 Commercial 27 17 Residential 39 33 A major factor explaining this slowdown seems to be the difficult economic situation that followed the devaluation of the sol in late 1967. 6.12 Another contributing factor to the lower than expected growth in SPD has been an improvement in the system load factor (SLF). The SLP had increased from an average of 55.9% in 1963-67 to 58.3% in 1967-71 (Annex 3), producing better utilization of the existing generating capacity. One important reason for this improvement might have been the increasing importance of industrial consumers in EEA's total system demand. 6.13 A clear indication of the increasing importance of industrial consumers in EEA's system is the evolution of the participation of industrial consumers in EEA's total energy sales. This participation grew from an average of 38% in 1962-66 to 44% in 1967-71 (Annex 3). 6.14 This increasing importance of industrial consumers also played an important role in explaining the reduction in the system losses (Annex 3), which decreased from an average of 11.4% in 1963-67 to 9.2% in 1967-71 -- a development closely associated with the growing importance of industrial customers which can be supplied at a higher tension (10, 30 and even 60 Kv) than residential consumers. Finally, some technological changes introduced in the design of the transmission system had also contributed to the reduction of energy losses. 6.15 System losses is not the only area in which important improve- ments are noticeable. Other examples of the eff-iciency of EEA's opera- tions are the lack of a major increase in personnel despite the substan- tial increase in energy sales and the close relation between the revenue from energy sales and the energy billed in each year. In the case of personnel, the average number of customers served by one employee rose by about 50% -- from 127 in 1963-66 to 184 in 1967-73 -- (Annex 3). - 15 - 6.16 The efficiency in the revenue collecting system is impressive. In 1966, the revenue from energy sales was already 95.6% of the total energy billed in the year. After 1966, there has been a steady increase in this indicator and by 1973, revenue from energy sales was 99.2% of the energy billed (Annex 3). 6.17 The investment in Matucana was then worthwhile for EEA, earning a rate of return higher than the 12% expected at appraisal. Although EEA already had a high level of efficiency in its operations in 1966, there were important improvements thereafter. VII. THE ROLE OF THE BANK 7.01 The Bank's role in this project has to be analyzed in two dimensions: one in the context of EEA and the construction of Matucana and the other in the broader context of the whole power sector. This distinction is necessary because the Bank played a more important role in relation to EEA and the successful completion of Matucana, than in the broader context of the power sector. 7.02 The Bank gave valuable advice to EEA in the selection of alterna- tives to fill the gap in energy generation in 1971-73. Other alternatives in addition to the one originally suggested by EEA were considered by the Bank but Matucana was finally selected as it was clearly the best alterna- tive available to EEA. The suggestion to purchase the EHA plants which were an integral part of EEA's system was another good Bank initiative. This measure, which was not finally carried through because of the cash shortage, would have led to the integration of EEA's system_and at the end it would also have had a positive effect on the net cash flow of EEA. 7.03 Another good Bank decision during project preparation was to allow for only national competitive bidding in relation to civil works. This decision was taken on the grounds of the considerable experience of local contractors in tunnelling, which was the main item of civil works; the fast implementation of the project proved this to be correct. The works were completed almost as scheduled despite the substantial delay in getting underway, and the performance of the contractors was regarded as very good by the Bank's supervision missions. 7.04 But another Bank decision in the area of procurement -- the exclusive use of the Bank loan for imported goods and services -- seems to have had a negative effect on the project. The effect of the Bank's decision was to increase the pressure to obtain other sources of finance for the project. It turned out that about US$2 million less than expected of the loan was spent on items related to Matucana, contributing therefore to the financial difficulties of EEA at a time when the market conditions for obtaining additional borrowing were not favorable. - 16 - 7.05 Another Bank decision in the preparation of the project that later on did not prove to be adequate was the reliance on the indenture test as a means of preventing the incurrence of excessive debt. As it turned out, EEA incurred substantial medium- and short-term borrowing which was not included in the indenture test. The debt-service ratio would have reflected to a higher degree than the indenture test the deteriorating financial situation and would have permitted more active Bank participation in the financial aspects of EEA. 7.06 During implementation, the Bank played a crucial role in helping EEA obtain the financial assistance of the Swiss group at a moment in which the construction of Matucana was seriously jeopardized and other less efficient alternatives for meeting the expected increase in demand were studied. 7.07 In contrast with this valuable participation in the construc- tion of Matucana, the contribution of the Bank in the broader context of the power sector as a whole was only minor. The Bank has to be credited, however, with a contribution toward power sector planning when it accepted inclusion of the financing of the feasibility study for the hydroelectric project of Salto Sheque in 1971. This study con- tributed to the analysis of the different alternatives for meeting the power demand in the Central Region after 1978 when Mantaro would be fully loaded. However, tlck of consideration of the power sector as a whole was a general weakness of the Bank's approach which was also reflected in the lack of major institutional objectives in the Bank's work with the power sector. All previous Bank loans as well as this one were made to EEA,which was the most efficient and well organized entity in a power sector characterized by lack of coordi- tion and planning. Power development had been haphazard as new entities were created over the years to meet the needs for electricity in various parts of the country, and government-owned autonomous corpo- rations for power development had been poorly organized. The power' sector as such was therefore almost nonexistent and one may wonder whether the Bank in its association with EEA since the early 1960s could have also contributed to the organization and strengthening of the sector. 7.08 This lack of major institutional objectives by the Bank appeared later when the new government that followed the revolution in 1968 was seriously interested in overcoming the lack of coordination in the power sector. In 1969, the Government requested UN assistance to formulate legislation to create a new power entity to be called Electro Peru, which would coordinate all future power development. This was a good opportunity for the Bank to take a more assertive role, but unfortunately it did not make any specific offer to the Government to finance power development. Thus, an opportunity to -17- contribute to strengthening of the power sector was lost. 7.09 Electro Peru was finally organized by the Government in 1972 and at the same time the General Directorate of Electricity was strengthened and, with the assistance of consultants, the Government undertook feasibility studies of power development on a regional basis with the ultimate goal of creating a national power grid. Creation of Electro Peru implied a radical change in organization of the power sector and set the basis for a state-owned power system. But there are still important problems in the administrative and planning orga- nization of Electro Peru that have prevented adequate coordination of the sector and which perhaps could have been avoided if the Bank had taken earlier a more assertive role in development of the whole power sector. In the last few years, the Bank has tried to play a more active role, but so far its efforts to make a loan to Electro Peru and therefore help in building up this institution have not been successful. VIII. CONCLUSIONS 8.01 Construction of Matucana was successfully completed despite EEA's financial problems. and the impoirtn.t institutional changes that took place in Peru during the period of this project. The efficient organization of the local contractors and the experience and efficiency of EEA were major factors explaining the successful completion of the project. 8.02 The substantial increase in oil prices is the main reason for the higher than expected rate of return to Matucana in the context of both EEA and the Central Region power system. If oil prices had not increased, the audit rate of return for EEA would still have been satisfactory, but that for the Central Region system would have been below 10%. It is very unlikely, however, that the factors which would have produced such a low rate of return,(the lower than expected growth in the demand for electricity and the completion of Mantaro on schedule) could have been foreseen at appraisal. n 4 a a «,, .e J Me 6 ANNEX 1 PROJECT PERFORMANCE AUDIT REPORT PERU MATUCANA POWER PROJECT (LOAN 511-PE) Project Composition at Appraisal Local Total Currency Foreign Exchange Cost S/. S/. US$ S/. ----(milli1On)---- I. Matucana Power Plant 1. Preliminary Works 44.1 0.3 0.01 44.4 2. Civil Works Intake 63.2 5.4 0.2 68.6 Tunnel and Forebay 315.4 27.0 1.0 342.4 Valve Chamber 29.8 5.4 0.2 35.2 Penstock 63.0 35.0 1.3 98.0 3. Powerhouse Civil Works 47.6 5.4 0.2 53.0 Electrical and Mechanical Equipment 42.6 127.0 4.7 169.6 4. Construction Equipment 5.0 13.5 0.5 18.5 II. Transmission and Distribution Lines, Communication Equipment, etc. 1. 220 Kv Transmission Line 48.5 48.5 1.8 97.0 2. 220 Kv Substations 29.7 54.0 2.0 83.7 3. Communication System 10.8 14.8 0.55 25.6 4. Meters 5.4 16.2 0.6 21.6 5. Distribution Cables 8.1 9.4 0.35 17.5 III. Consulting Services 39.7 29.7 1.1 69.4 IV. Interest and Other Charges 56.5 40.5 1.5 97.0 V. Contingencies 82.4 40.5 1.5 122.9 TOTAL 891.8 472.6 17.50 1,364.4 US$ equivalent (33.0) (17.5) (50.5) ANNEX 2 PROJECT PERFORMANCE AUDIT REPORT PERU MATUCANA POWER PROJECT (LOAN 511-PE) Original and Final List of Goods (US$) Original Final 1. Electric and Mechanical Equipment 6,600,000 5,782,824.44 2. Construction Materials 1,000,000 191,928.26 3. Construction Equipment 500,000 151,594.83 4. 220 kv Transmission System 1,800,000 2,159,369.55 5. 220 kv Substations 2,000,000 1,551,968.87 6. Communication Equipment 550,000 1,760,252.75 7. Meters 600,000 1,302,488.48 8. Distribution Cables 350,000 1,057,316.35 9. Consulting Services 1,100,000 2,242,256.44 10. Interest and Other Charges during Construction 1,500,000 1,300,000.00 11. Contingencies 1,500,000 Total 17,500,000 17,499,999.97 PROJECT PERFORMANCE AUDIT REPORT PERU MATUCANA POWER PROJECT (LOAN 511-PE) EEA's Selected Operating Statistics 1963 1964 1965 1966 . 1967 1968 1969 1970 1971 1972 1973 Generating Capacity Installed (Mw) 229.7 294.3 423.5 485.1 485.1 485.1 485.1 485.1 545.1 594.1 584.1 Substation Capacity (Mva) 408.0 443.8 498.8 513.4 532.5 549.0 566.8 597.4 657.4 Distribution Transformers (Mva) 389.8 874.2 941.4 954.2 936.9 944.0 1,046.2 Primary Feeders (km) 2,576.0 3,101.0 3,345.0 Secondary Lines Built (km) 3,728.0 3,402.0 4,503.0 Number of Customers (thousands) 240.1 257.4 272.5 293.4 320.8 345.3 357.9 381.7 454.7 487.4 516.1 Residential (thousands) 191.7 206.7 218.8 236.9 262.6 284.4 296.5 318.1 389.3 421.9 452.6 Industrial (thousands) 3.1 3.4 3.5 3.7 4.3 4.4 4.4 4.4 4.5 4.6 4.6 Commercial (thousands) 44.6 46.6 49.5 51.9 53.0 55.5 55.9 58.0 59.8 59.7 58.1 Others (thousands) .7 .7 .7 .9 .9 1.0 1.1 1.2 1.1 1.2 .8 Average Monthly Consumption/ Customer (kwh) Residential 145 148 149 148 147 145 147 149 132 130 131 Commercial 183 405 421 445 464 457 474 488 502 538 531 Industrial 9,143 9,279 10,290 11,405 11,545 12,887 14,160 15,239 16,981 17,195 18,493 Overall Average 329 337 347 360 370 378 389 396 367 366 374 Energy Sales by Customer Category (Gwh) Residential 332.5 368.3 390.2 421.7 462.9 494.3 521.9 567.5 615.7 656.1 709.4 Industrial 343.1 376.7 430.8 505.6 591.7 681.3 739.8 807.4 906.0 957.7 1,019.5 Commercial 204.1 226.1 248.8 276.6 259.4 280.0 291.8 287.7 303.3 323.9 370.4 Public Lighting and Other 47.9 51.9 57.3 65.1 110.9 109.2 117.2 149.8 176.4 203.1 215.6 Total 927.6 1,023.0 1,127.1 1,269.0 1,424.9 1,564.8 1,670.7 1,812.4 2,001.4 2,140.8 2,314.9 Gross Generation (Gwh) 1,078.4 1,173.1 1,273.4 1,423.1 1,586.2 1,733.3 1,842.3 1,994.3 2,204.1 2,345.7 2,555.5 Losses as % of Gross Generation 12.2 11.4 11.4 10.8 10.2 9.7 9.3 9.1 9.2 8.7 9.4 Annual Load Factor (%) 55.5 55.9 55.8 55.8 56.7 58.3 59.5 58.2 59.2 60.0 59.8 System Peak Load Actual 221.5 239.5 260.0 291.0 319.0 338.0 353.0 391.0 425.0 444.0 488.0 Forecast 320.0 352.0 387.0 425.0 467.0 513.0 564.0 Revenue from Energy Sales as % of Energy Billed 98.8 95.6 96.1 95.7 88.1 96.4 100.0 96.9 99.2 Employees 2,055 2,171 2,138 2,176 2,190 2,123 2,122 2,116 2,248 2,310 2,388 Customers/Employees 119 120 130 138 148 163 169 180 202 211 216 Energy Sales/Employees (Gwh) .460 .478 .527 .583 .650 .737 .787 .856 .890 .926 .969 PROJECT PERFORMANCE AUDIT REPORT PERU MATUCANA POWER PROJECT (LOAN 511-PE) EEA's Income Statements (SI. million) 1968 1969 1970 1971 Forecast Actual Forecast Actual Forecast Actual Forecast Actual Energy Sales (millions kwh) 1,592.0 1,564.8 1,750.0 1,670.7 1,925.0 1,812.4 2,117.0 2,001.5 Average Revenue per kwh (centavos) 72.0 84.4 72.0 86.1 72.0 85.1 72.0 83.5 Operating Revenue 1,146.0 1,320.4 1,260.0 1,439.2 1,386.0 1,541.4 1,524.0 1,671.6 Operating Expenses Operations 501.0 554.5 582.0 650.8 655.0 723.2 766.0 816.1 Energy Purchased from Hidrandina 70.0 122.3 60.0 122.8 50.0 120.3 30.0 121.3 Depreciation 256.0 288.0 305.0 333.0 Contribution to Expansion Fund 15.0 10.0 15.0 10.0 15.0 10.0 15.0 10.0 Total Operating Expenses 842.0 899.6 945.0 1,002.4 1,025.0 1,075.9 1,144.0 1,186.0 Net Operating Income 304.0 420.4 315.0 436.8 361.0 465.5 380.0 485.6 Other Expenses Interest Paid (net) 122.0 106.0 143.0 163.0 Financing Expenses 10.0 14.2 20.0 15.4 15.0 18.7 10.0 15.6 Director's Fees and Taxes 9.7 12.5 11.7 14.6 12.7 18.8 12.7 40.5 Net Income 162.3 209.8 173.3 219.7 190.3 231.5 194.3 214.3 ANNEX 5 PROJECT PERFORMANCE AUDIT REPORT PERU MATUCANA POWER IROJECT (LOAN 511-PE) EEA's Expected and Actual Financial Plan, 1968-71.1/ (S/. million) Expected Actual Difference (Actual-Expected) SOURCES OF FUNDS Internal Cash Generation Net Operating Income before Interest 1,360 1,808 448 Depreciation and Expansion Fund 1,242 932 -310 Connection Fees 668 426 -242 3,270 3,166 -104 Less Taxes and Director Bonus 57 70 + 13 Interest Long Term Debt 841 775 Medium and Short Term Debt 70 179 911 954 + 43 Amortization of Debt Long Term Debt 639 692 Medium and Short Term Debt 168 493 807 1,185 +378 Total Debt Service 1,718 2,139 +421 Dividends 553 731 +178 Net Internal Cash Generation - 942 226 -716 Capital 250 175 - 75 Borrowifg Long Term Loans 1,242 1,274 Medium and Short Term Loans 78 654 Other Loans - 148 Less Discounts and Fees - 58 - 65 1,262 2,011 748 TOTAL SOURCES OF FUNDS 2411 USES OF FUNDS Construction Program (excluding interest) bateo Reservoir 50 97 *Matucana Plant 1,117 1,217 Distribution System Expansion 1,1C6 1,275 Total Construction 2,273 2,589 +316 Acquisition of Hidrandina Plants 211 - -211 Changes in Working Capital 181 - 178 TOTAL USES OF FUNDS 2454 11 1/ The expected figures had been adjusted-for the devaluation of the sol.  IBRD 11765 AUGUST 1975 PATIVLCA 40 MW £ PARAMONGA 24MW PAUCARrAMBO /05 MW CERRO DE PASCO Lago de Junin HUA CH6 ý fu42urq SAYANOJUN MÅL PÅASO S4MW - CTARMA 0MARCAPOMACOCHA LA OROYA LA OROYA lOMi PACHACHACA 13MW AO' HUINCO 240MW MATUCANA 120 M W à CALLAHUANCA 6T7M MATUCANA AVYOPARMPA 63 Mw A NTAHUASASHUA PUCUSU.. PE RU POWER PLANTS IN CENTRAL PERU EXISTINGOHUANCAVELICA Thermal plants Hydro plants Headrace tunnels ------ AAUH KILOMETERS o o