Colombia - The power sector and the World Bank, 1979-1987 (Vol. 1 of 3) : Overview

Report No. 8893 Colombia The Power Sector and the World Bank, 1970-1987 (in Three Volumes) Volume 1: Overview June 28,1990 Operations Evaluation Department FOR OFFICIAL USE ONLY Document of the World Bank This document has a restricted distribution and may be sed by recipients only in the performance of their official duties. Its con ts may not otherwise be disclosed without World Bank authorization. ABREVIATIONS AND ACRONYMS AR - Accounts Receivable COL$ - Colombian Peso CONPES - National Economic and Social Policy Council CORELCA - Corporacton Electrica do la Costa Atlantica CPP - Country Program Paper CPS - Central Projects Staff, World Bank CVC - Corporacion Autonoma Regional del Valle del Canca DNP - Departamento Nacional de Planeacion DS - Debt Service Coverage Ratio ECOPETROL - Empresa Colombiana de Petroleos EEEB - Empresa de Energia Electrica de Bogota EPH - Empresas Publicas de Medellin EHCALI - Empresas Municipales do Cali FEN - Financiera Electrica Nacioal GDP - Gross Domestic Product gWh - Gigawatt-hour (106 kWh) IC - Long-run average incremental cost ICEL - Instituto Colombiano de Energia Electrica IERR - Internal Economic Rate of Return IFRR - Internal Financial Rate of Return ISA - Interconexion Electrica S. A. JNT - Junta Nacional do Tarifas kW - Kilowatt kWh - Kilowatt-hour LAC - Latin American and Caribbean Region LRMC - Long-run marginal cost HHE - Hinisterio de Mines y Energia OED - Opirations Evaluation Department, World Bank OMS - Operational Manual Statement, World Bank OPS - Operations Policy Staff, World Bank ROA - Return on Assets 201 - Return on Investment SAR - Staff Appraisal Report SF - Self-Financial Ratio TWh- Terawatt-hour US$ - United States dollar WC - Working Capital *4 THE WORLD bANK FOR OFFICIAL USB ONLY Washington. D.C. 2043i U.S.A. OIke of DruW-eowi Opeatms tviumm June 28, 1990 NMORANDM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECTs COLOMBIA - THE POWER SECTOR AND THE WORLD BANK, 1970-1987 Attached for information is a copy of a report entitled 'Colombia - The Power Sector and The World Bank, 1970-1987* prepared by the Operations Evaluation Department. Attachment This document has a restricted distribution and may be used by recipints only in the Performance of their oficial duties. Its contents may not otherwise be disclosed withou,World Bank authorastion., CURRENCY EXCHANGE RATES Year Mid-Year (COL$ per US$) 1970 18.4 1971 19.9 1972 21.9 1973 23.6 1974 26.1 1976 34.7 1978 38.1 1980 47.3 1981 54.5 1982 64.1 1983 78.9 1984 100.8 1985 142.3 1986 194.3 1987 242.6 FOR OFFICIAL USP ONLY COLOMBIA THE POWER SECTOR AND THE WORLD BANK, 1970-1987 VOLUME Is OVERVIEW TABLE OF CONTENTS Page No. P re f a c e . . . . . . .. . . . . . . . . . * . . . . . . . . . . . . . . . . . . . . . . . . . # Summary and Recommendations ........................................ iii I. INTRODUCTION........ **................ 1 Development of the Sector................................. 1 History of Bank Relations with the Sector................. 5 II. SECTORAL OBJECTIVES AND INSTITUTIONAL ISSUES.............. Introduction.............................................. 8 Sector Objectives ......................................... 8 Sector Regulation ........................................ 9 Restructuring the Sector.. ............................... 9 Structure and Role of ISA............. ................... 10 The Role of the Bank ...................................... 10 Conclusions and Recommendations........................... 11 III. ELECTRICITY DEMAND AND SUPPLY ISSUS...................... 14 Introduction .............................................. 14 Sales Structure ......................... . . .... ..... 14 Demand Forecasts........................................... 15 Power Rationing and the Cost to the Economy............... 15 Economic Cost of the Surplus Generating Capacity.......... 16 System Losses ............................................. 17 Conclusions and Reccamendations........................... 18 IV. TARIFF ISSUES .......................................... .. 19 Introduction .............................................. 19 Achievements .............................................. 20 Costs of Service and Tariffs at Retail Levels............. 21 Distortions in Retail Tariff Structures................... 22 Subsidies and Income Distribution......................... 22 Outstanding Issues...................................... 24 Costs of Service and Bulk Tariffs......................... 25 Tariff Estimates at Appraisal of Bank Operations.......... 26 Bank Performance on Tariff Issues......................... 28 Recommendations ........................................ .. 29 This document has a restricted distribution and may be used by recipients only in the performance of their oftcial duties. Its contents may not otherwise be disclosed without World Bank authorization. TABLE OF CONTENTS (Continued) Page No. V. FINANCES....................... ............... . 30 Introduction............................................. 30 Self-Financing Ratio.............. . .... ................ 31 Return on Assets............. . ... .... . .................. 33 Debt Service Coverage Ratio............................... 33 Debt and Equity ........................................... 34 Accounts Receivable............. ...................... 34 The Bank's Performance .................................... 35 Absence of a Sectoral Approach...................... 35 Bank Lending Terms........................................ 35 Handling Exchange Risk.................................... 37 Financial Performance Standards............. 38 The Bank's Use of Financial Indicators.................... 39 Recommendations ............................ ....... ...... 41 VI. SECTOR INVESTMENT......................................... 42 Investment Planning and Decision Process.................. 42 Multi-Objective Planning .................................. 43 Implementation Time ....................................... 44 Project Costs............................................. 44 Balancing the Investment Program.......................... 44 Recommendations................ .......................... 45 VII. CONCLUSIONS .................................... ... .. . 46 Policy Issues for Colombia ................................ 46 Policy Issues for the Bank...... ......................... 47 ATTACHMENT. Comments from Interconexion Ele trica, S.A................... 49 MAP IBRD No. 17690Rl COLOMBIA THE POWER SECTOR AND THE WORLD BANK, 1970-1987 PREFACE 1 * This review of the Bank's lending to the Colombian power sector during the period 1970-1987 is the first OED study that exclusively addresses the Bank's operations and related issues in the power sector of a country. The first suggestions to carry out such a study came during discussic.s between Colombian officials and the Director General of the Operations Evaluation Department (OED) at the Bank's 1986 Annual Meeting. The proposal reflected Government's and the Bank's growing concern about the clear deterioration in the sector's performance during the 1980s. 2. The Approach Paper to the present study, which the joint Audit Committee (JAC) of the Bank's Board of Executive Directors approved in January 1988, was based on an extensive review of documents available in Washington and on preliminary discussions with Colombian officials in December 1987. During four weeks in February and March 198§, the study team comprising three OED staff and a Colombian consultant stayed in Colombia, collecting information from the various entities involved in the sector and interviewing persons who played a major role in power development and operations. Members of the team visited facilities, completed or under construction, that the Bank helped or is helping finance. On returning to Washington, the team discussed the issues raised in the present report with Bank officials from both Operations and Policy Departments, who, in the review period were involved in power sector lending to Colombia. 3. An early draft of this report was submitted to the Bank's Operations Departments and the Policy, Research and External Affairs complex for comments in the fall of 1989. After taking into account these comments, a revised draft was submitted, in the winter of 1989-90, to the Colombian authorities and power utilities concerned for their comments. Such comments were received from the lead institution of the power sector, Interconexion Electrica S.A. (ISA), and are reproduced as an Attachment to this volume (Vol. I). These comments have been taken into account also in preparing this report. 4. The report is in three volumes. Volume I--the Overview--first presents the Summary and Recommendations of the review. It then sets out in seven brief Chapters the highlights or OED's findings with a minimum of background discussion. Volume II--the Technical Report--sets forth the justification of the findings under the chapters that discusst (i) the sector's setup and history over the review period, (ii) a few general issues, some of them far wider than the subject of the present study, but addressed here because they cropped up during the investigations for this review, (iii) institutional aspects, (iv) issues related to demand and supply, (v) the perennial tariff problem, (vi) the partly related pervasive financial - ii - issues, (vii) selected questions concerning sector investment, and (viii) the links between the power sector and macro-economy. Volume III--Annexes--then presents supporting evidence for the detailed discussions in Volume II. 5. It is important to note that the study's first main focus is on the sector, which means that, in general, issues concerning individual utilities are dealt with only insofar as they are relevant for the sector as a whole. The second main focus is on the Bank's relations with Government and the sector in the context of the latter's development over the period under review. 6. The review pursued issues which appeared important and for which OED considered that constructive conclusions and recommendations could be arrived at. However, some issues were not addressed though originally intended to be studied. The most important among these is doubtlessly the environment, an area in which the period under review witnessed substantial progress, to a good part associated with successful efforts in the power sector. 7. The study team would like to thank the many Bank people it interviewed who helped by searching their memories--some happy, some less so--of things past and by candidly discussing their views of the issues set forth here. It is, of course, particularly grateful for the many contributions it received in Colombia from Government, especially Departamento Nacional de Planeacion (DNI) and from the sector entities, in particular Empresa de Energia Electrica de Bogota (EEEB), Empresas Publicas de Medellin (EPM), Instituto Colombiano de Energia Electrica (ICEL), Corporacion Autonoma del Valle del Cauca (CVC), Empresas Municipales de Cali (EMCALI), Interconexion S.A. (ISA), and Corporacion Electrica de la Costa Atlantica (CORELCA). EEEB, EPM, and ISA even went so far as to %irganize seminar-type gatherings during which the team was able to hear from many officials, in charge during the review period (1970-1987), their views on the issues. Many thanks also go to the numerous persons, formerly and/or presently active in the sector, who were willing to be interviewed individually, as well as to the Bank's Resident Representative for his views and the logistical help he provided the team. COLOMBIA THE POWER SECTOR AND THE WORLD BANK. 1970-1987 SUMMARY AND RECOMMENDATIONS 1. The Sector 1. During the period 1970-87, in Colombia, seven companies generated, transmitted, and distributed virtually all the electric power for public consumption: - the municipal companies in the main cities: Empresa de Energia Electrica de Bogota (EEEB), Empresas Municipales de Medellin (EPM), and Empresas Municipales de Cali (EMCALI); - the Government owned Corporacion Autonoma del Valle de Cauca (CVC) in the Cauca area, Corporacion Electrica de la Costa Atlantica (CORELCA) in the northern part of the country, and Instituto Colombiano de Energia (ICEL) in the rest of Colombia; - Interconexion S.A. (ISA), a company owned by the other utilities of the sector. 2. The Ministry of Mines and Energy (MME), the National Planning Department (DNP), in particular its National Tariff Commission (JNT), and, since its creation in 1982, the National Electricity Fund (PEN) were the main Government entities responsible for the sector and its regulation. 2. Bank and IDB Lending to the Sector 3. In the period under review, the Bank made 15 loans for over US$1.9 billion to the sector. The Inter-American Development Bank (IDB) contributed another US$1.9 billion. The two institutions together provided about 80% of the long term foreign borrowing the sector incurred over the period. 3. The Macro Setting During the Review Period. 4. In forming judgments about the failures and successes of the Bank's lending to the Colombian power sector in the review period, it is essential that the overall setting be understood, both within Colombia and in the global economy, in which Bank financed operations were prepared, appraised and implemented. Unlike many other developing countries, Colombia was not seriously affected by the "first oil shock" of 1973-74 when oil prices quadrupled since it was then a small marginal net oil exporter. By 1976, however, it had become a net oil importer, a status it retained until 1986 when it re-emerged as an oil exporter, just as oil prices collapsed. T-e Colombian economy remained fairly buoyant from the mid to late 1970s due to the boom in coffee prices; however, the effects of the "second oil shock" in 1979-80 weakened the economy which also experience increased inflation-- rising to an annual rate of 29% in 1979. This weakening was compounded by the world recession of 1981-83 which witnessed a collapse in coffee prices; iv a decline in the country's manufactured exports; higher priced petroleum imports; severe increases in real interest rates and a concomitant escalation in debt servicing costs; a rapid decline in international inflation levels which was not matched in Colombia, thereby resulting in an increasingly overvalued peso. The major devaluation of the peso was undertaken in 1985 which, along with increased trade liberalization and a more restrictive public sector expenditure, laid the basis for the economy's adjustment. 5. In this rapidly changing macroeconomic enviroiment, there is no doubt that power sector planning and operations have become much more complex. Real increases in costs of construction, equipmttit and fuels in the mid-1970s reversed a long-term decline in the real costs of electric power that had lasted until the early 1970s. Uncertainty and risk increasingly became the hallmarks of any scenario about the future -- be it power demand, exchange rates, or fuel prices. Under these conditions the search for least- cost solutions to power development itself became increasingly fragile. 4. Sector Development 6. Between 1971 and 1986, the number of subscribers in the sector tripled to some 3.8 million. So did sales, which reached about 20 TWh (billion kWh), more than doubling the pe capita consumption. Capacity installed more than quadrupled, to some 7.L *W (million kW). 7. During the review period characterized by fast demographic growth, the segment of the total population with access to electricity increased by more than 15 percentage points, to a level in excess of 60%, overall; the rural. population with such access increased from 15% of the total rural population to 45%. 8. The period also witnessed major changes in sales patterns, as the institutionally and financially weak utilities, ICEL and CORELGA, increased their share of the market, from 30% to 421. at the expense of the stronger companies, in particular. EEEB and EPK. whose share declined from 51% to 42%. At the same time, residential consumption, on average heavily underpriced, grew from 41% to 48% of sector salez both developments exacerbated the sector's financial problems. The high level of residential electricity sales in Colombia (about one-half of total sales) is anomalous, given the country's development level -- for example, such sales in Argentina and Brazil only account for one-third and one-fifth, respectively, of total sales. 9. Actual developments in the sector. in tne review period, fell short of what had been forecast In particular. Lower-than-expected growth in demand in the 1980s due to the unforeseeable recession, and rigidity in the generation expansion program led to excess capacity late in the decade. Most of the issues that were identified in OED evaluation studies of twenty years ago are still unresolved, in particular: the complex and cumbersome instititional structure of the sector; poor financial performance in individual utilities and the sector at large; very low and distorted electricity tariffs; a sectoral investment program that overly stresses generating plants, at the expense of distribution, and suffers from large V time and cost overruns; and a project selection procedure that has, in some cases, led to results that are economically and financially less than optimal. 5. Institutional Development 10. In the 1980s, JNT developed into a reasonably strong and accepted tariff regulatory body, which, increasingly led the way in the necessary tariff reform started ir the mid-1980s, after the Government and the sector reluctantly accepted the concept of tying the price of electricity to its economic cost, albeit with large discounts for residential users. 11. ISA, which at the beginning of the review period had been operating for less than three years, became--albeit temporarily--the largest generating utility, as its output represented 31 percent of the country's total. ISA achieved this beyond meeting its primary responsibility, which is to assure the power exchanges between the main systems in the country. ISA became also the central planner of generation and transmission plant for the interconnected system which increased in coverage in the 1970s until 1984 when all major areas of the country weze interconnected. Most recently, it assumed a coordinating function in distribution planning. 12. ISA's shareholders, i.e. the other major utilities if the sector, prevented the company from achieving the originally envisaged monopoly of building and operating all major additions to the interconnected system, an outcome that, in Colombia's fiercely regionalistic environment, was predictable. They also tended to introduce into ISA's planning their preferences at an unduly early stage, at the expense of national considerations. This did not necessarily lead to uneconomic solutions, though in the past 20 years there may have been such instances. However, the interventions seriously affected the transparency of the investment planning and decision-making process. 6. The Bank's Role and the Sector's Performance 13. In the 1970s, Government, the Bank, and, to a lesser extent, the sector shared interconnection and the development of ISA as two main objectives for the sector. In the early part of the 1980s, when the lack of agreement on the future of ISA became undeniable and the economic environment seriously deteriorated, the focus of the dialogue between Government, the sector, and the Bank became less precise. Concurrently, the Bank relegated the sector approach, which emphasizes sector development over that of individual utilities, to a second priority, which represented somewhat of a shift of emphasis to that taken in the second half of the 1970s. Accordingly, it shifted, for a time, its operations toward individual utilities, with varying emphasis on sectoral issues. Nevertheless, the operation with FEN in 1984 and, in particular, the Sector Adjustment Loan in 1987, led the Bank some way toward an overall sector approach. 14. Despite the variations in its approach, the Bank, in the course of the period 1971-86, supported major sector initiatives, such as: vi - the creation of stronger regional utilities, with the major existing companies as a nucleus; - the preparation of a development master pla; * systematic generation and transmission planning, based on the least present cost method; * balancing investment between generation and transmission, on the one hand, and distribution, on the other; - reduction of systems losses; - a unified prssentation of financial data and progress toward compatible accounting in the sector utilities; * marginal cost pricing; and - generating a reasonable contribution to investment from revenues. 15. Success in some of these endeavors was achieved, but, in general, it was limited: - Although decided at the highest level of Government, the creation of new regional aggregations of utilities did not proceed. - Preparation of the sector development master plan was drastically curtailed. - Generation and transmission planning, as carried out by ISA, vastly improved, albeit with the flaws related to its shareholders' untimely interventions. In addition, the appropriateness of the search for least-cost optimal power expansion paths has arisen, in hindsight, due to the increasing uncertainty and risk associated with critical variables, such as demand and fuel prices. - Despite all efforts, the investment program remained as unbalanced as ever, at the expense of distribution. - The Colombians ultimately accepted the principle of pricing according to long-run incremental costs, but implementation has been slow, leaving the sector far from the goals set, as tariff levels were much too low and tariff distortions worsened, implying large subsidies to residential consumers (especially those with high income), partly at the expense of industry and commerce. From the perspective of the Bank's role, there was a tendency since the mid-1980s to downplay sompwhat the role of tariffs in the sector's financial problems. This conveyed an incorrect impression that internal resource mobilization problems were less severe than they were (and always have been) and that adjustments could proceed at a very slow pace. vii - Overall systems losses increased from about 17% to 25% of net power generated, whereas in the EEEB and CORELCA systems the increase was far larger, ranging from around 13-14% at the beginning of the period to about 25-26% at the end. - Producing consolidated financial data for the sector is still .xceedingly difficult and the results are fraught with 4nconsistencies. - Internal cash generation fell far short of what was hoped for and of what would have eased the heavy financial burden the sector imposes on Government, as the sector's revenues never contributed more than 10% of its investment after 1977. This is to be viewed against the background of reasonably satisfactory self-financing ratios (of 40% and above) being achieved by EPM and EEEB during part of the 1970s. - The shift in 1976 of Bank financing terms being determined on a country and no longer on a project basis, allied with a hardening of these terms, had severe implications for the financing of hydro projects, such as those in Colombia. The effects of these changes were not compensated for by better internal resource generation. 16. Despite the above setbacks, the Bank, all in all, helped the sector achieve substantial progress, especially in the late 1970s and early 1980s. Some of the disappointing results have to be seen as associated with the dramatic deterioration of performance that came with the financial crisis in the mid-1980s, when Government did not allow the sector to adjust its tariffs for the effects of the massive devaluation of the Col$. This lack of success in the power sector was particularly disappointing for the Bank, as it coincided with a successful dialogue between Government and the Bank on changes required in Colombia's economy. 7. The Main Issues and the Ways to Tackle Them 17. In OED's view, the Colombian power sector needs, specifically, a set of Government sector objectives in line with similar objectives for the wider energy sector, which, in turn, should be properly embedded in Government's macroeconomic policies. To translate Government sector policy into ccncrete measures, to coordinate implementation of such measures, to monitor progress achieved, to assess success, and to determine possibly required remedial action, calls for a strong regulatory body. 18. Both the set of specific sector objectives and the strong regulatory framework are presently missing. These issues, which OED considers to be at the root of most others, urgently need addressing, as, of course, do more short-term problems, in particular, the poor financial performance. The years following the review period, i.e. 1988-89, have witnessed some promising steps toward strengthening regulation, in particular the approval in Congress of legislation creating the Energy Board, which is bound to play a central role in sector regulation. viii 19. It is crucial to take into account that, whether the specific sector objectives imply development toward overall centralization (highly unlikely in the Colombian context), or toward a set of central and regional utilities with substantial autonomy, or even toward various degrees of privatization, the strong regulatory body will always be needed. A further crucial point is that ISA's setup, as conceived in 1968 on the basis of a compromise and as subsequently adjusted, while providing for a long time a reasonable framework, has run its course. Indeed, the often untimely interventions of the shareholders in the investment planning and decision process, and these utilities' tendency to deny ISA's bulk tariff its role as the most important single electricity price signal in the country seem to call for a thorough rethinking of the functions ISA has to fulfill and of a proper framework in which these functions should be discharged. The new concept, which e.g. could involve the reassignment of some functions, will have to acknowledge ISA's major achievements and to capitalize on the proven strengths of the organization and of its staff. 20. Once the above basis is set right, the framework will exist that will allow Government and the sector to tackle decisively and successfully the other unresolved issues, identified above. 21. The Bank--most likely together with IDB--beyond helping Government and the sector address the short-term problems, should be able to play a central role in the establishment of the new basis for the sector's development and operations. However, OED concludes that the Bank should continue to support the sector in the short term only if Government and the sectot prove willing to tackle decisively the broader, more long-term issues. Developmer.cs after the close of the review period suggest that there is such increased willingness. 22. As it did at least once in the recent past, the Bank should induce the partners in the sector to discuss candidly the sector issues, in particular, the politically sensitive sector structure, the role of Government in the sector, as well as the array of possible solutions. This time, however, it may have to help Government and the entities involved in the sector arrive at the positions from which decisions would be made. It should then be willing to accompany them on the arduous and long path of implementation. Experience accumulated during the review period suggests that, should a consensus it can endorse remain elusive, the Bank should refrain at least for a time from further lending to the sector. When lending, it should: - consistently take the sector approach; - avoid the pervasive, at times utterly unrealistic optimism, that has charactrized several of the operations reviewed here; and - insist on action by Government and the sector each time before it submits an operation for Board consideration. ix Adopting a positive attitude toward regionalism and helping mobilize, to the extent possible, the strengths of the phenomenon to the advantage of sector development would further help the Bank in its dealings with the Colombian power sector. 8. Sector Specific Recommendations 23. The general approach to the sector issues outlined above, suggests a series of more specific recommendations set forth below. Institutional Framework 24. In this area, the review concludes that once the specific sector objectives are defined, Government will have: - to spell out the strategy according to which the sector decides to meet or not to meet demand, i.e. which demand should be met to what extent (and what are the cost implications at the macro and micro level of so doing); - to set the ground rules for using national resources in the power sector, within the broader framework of an energy sector policy, and the national economy; - to clarify the basic rules of commercial operation of the utilities, in particular what elements of costs revtnues would have to cover; - to define, in general terms, the type of operations that Government would subsidize, e.g. elements of rural electrification and such re-financing measures in order to put the sector back on a financially sound footing; and - to spell out the principles of setting electricity bulk and retail tariffs, and their relation to long run incremental costs, as well as the measure of cross-subsidies between consumer categories that will be tolerated. Power Sector Regulation 25. In the area of power sector regulation, the Bank should assist Government in setting up the Energy Board, under whose purview power sector regulation would take place. The responsible entity should, in general, translate Government objectives and policies into sector rules, and exercise, inter alia, the following functions, presently only loosely coordinated: - define for the individual utilities areas of autonomy as large as possible in order to encourage the deployment of initiative, entrepreneurship, and regional dynamism, but limited such as to avoid endangering the pursuit of clearly identified national objectives; x - strengthen sector planning and operational coordination, i.e., in the present setup, support ISA in carrying out its main tasks; - establish and oversee the decision making mechanism connecting technical and economic sector planning, financing, and project implementation; - set, in accordance with the general Government objectives for the sector, the operational and financial targets for the individual utilities, as well as associated incentives and penalties; - oversee creation and operation of joint ventures to build and own new plants, following rules defined in accordance with the above first point; - monitor operations, identify shortcomings, and agree with the utilities on remedial action; - set the tariffs in agreement with the general objectives and rules set forth by Government as suggested above and to over-see their implementation; - link up with the proper Government institutions, in particular the Ministry of Finance, to make sure that Government contributions established on the basis of the sector's investment and operational planning in accordance with the general Government objectives, come forth at the proper time, in the proper amounts; - define the rules of interchange between utilities and, in particular, of sharing surpluses and deficits within the sector. Re-alignment and Concentration of Utilities 26. Re-alignment and concentration of utilities, as envisaged in several instances during the review period, could undoubtedly represent a substantial step towards increased efficiency in the sector and the Bank should induce the sector to define as soon as possible its feasibility. ISA's Structure and Role 27. The Bank, as ISA's main sponsor, should assist Government and the sector in re-examining ISA's structure and role and in defining adjustments in the setup and the modus operandi that may well imply the re-assignment of some of ISA's functions, as outlined earlier, but should preserve the practical approach to planning, shield planning from political influence before the start of the decision making process proper, and to define realistically and unambiguously the role of the central generating and transmission utility. xi TarLfU 28. In terms of tariff policy--beyond the suggestions above--the Bank should: - continue to put its full weight behind Government to help implement and further improve the present tariff rules and to support JNT, or whatever organization takes over its tariff regulatory role, in future; - insist on substantial action towards economic pricing of electricity each time before it submits a new operation for Board approval; - assess systematically the degree to which tariff levels set according to long-run incremental cost principles cover the full financial costs of the sector and individual utilities; - emphasize, within its own organization and in Colombia, the role of the bulk tariff as a key factor in the pricing of electricity in Colombia, and, as such, should, in connection with future lending, consistently insist on action towards economic pricing of bulk electricity; Finanjes 29. It seems evident that utilities and sector finances need to be turned around. As suggested above, this will have to be done on the basis of a new broad consensus on the institutional setup, on the respective roles of Government (at central, regional, and local level) and utilities, and on the objectives to be pursued. Once basic principles are agreed, their application will include a financial component, that in many steps, will hopefully allow the utilities to re-establish their cash generation capacity, strengthen their capitalization, bring accounts receivable to acceptable levels, and thus allow the companies to operate in a transparent way within an economically and financially reasonably efficient sector. 30. To create one of the main bases for the above development the Government and the sector should achieve within a few years reasonable uniformity in the accounting systems of the utilities, such as to allow both Government and the Bank: - to prepare meaningful consolidations of the individual utilities' accounts permitting a clear evaluation of sector performance, and - to monitor the performance of the individual utilities and to set their financial goals using indicators that allow adequate comparisons between the various companies. xii 31. In OED's view, there is little basis for establishing financial transfer mechanisms between the utility companies servicing the five major markets. The political consensus does not exist for their creation and on economic grounds it is likely to reward inefficient management in some markets. The absence of such mechanisms implies that tariff levels in each market be set so that costs are covered in that market. To the degree that this cannot be fully achieved in some markets, either on equity grounds or because of market structure (for example, in poorer rural markets), then the onus would be on Government to transfer directly to the affected local utility resources to cover such deficits. In this way, transparency is achieved regarding who pays how much, to whom, and for what. Other avenues could also be explored. For example, utilities continue to be exempted from income taxes and, as such, fiscal policy measures oriented toward creating a mechanism for cost compensation have never been assessed. This is an avenue of analysis that could yield benefits as an alternative to excessive reliance on inter- or intra-sector subsidies. 32. In the financial area, again the Bank should insist on demonstration by Government and the sector of willingness to perform, before Board presentation of new operations, and on compliance with agreed targets during implementation. In this connection, OED further recommends that, as part of a broader effort in improving power sector borrowers' financial monitoring, performance under financial covenants be subject to audit as part of the external auditors' report on the annual financial statements of Bank borrowers. 33. Finally, OED considers that the Bank should review, in the light of what has occurred in the Colombian power sector, the indicators it uses to monitor and covenant financial performance of such a sector and determine possible improvements, in particular, in measuring revenue performance. Investment Program 34. The sector urgently needs a more balanced investment program with a proper share for subtransmission and distribution. It will have to address the shortcomings of planning and decision making in generation and transmission through (i) institutional measures; (ii) improved comprehensiveness of alternatives considered; (iii) better uniformity of project preparation at the selection stage; (iv) more realistic time and cost scheduling; and (v) a bolder approach to sensitivity and risk analysis, implying large variations in the main project parameters. Improvements in time scheduling seem the most promising avenue to better costing of projects; the time estimates should, in particular, take into account that one or the other of the risks perceived are nearly bound to materialize. Linkages With Macro-economy 35. In times of higher uncertainty, programs that provide higher flexibility may be preferable even if, on the basis of a straight least cost analysis they may not represent the economic optimum. Specifically, it may be valuable for the sector to consider a program of large plants meeting the xiii lower limit of projected demand development and a complementary program of smaller plants with shorter lead times covering the difference to the most probable demand. Such solutions might well result from comprehensive multi- objective planning, a tool whose systematic application might provide substantial benefits to the sector. 36. OED's findings confirm that proper financial management of the power sector is an essential requirement in any effort to achieve macroeconomic targets such as balance-of-payments stabilization, reduction of the public sector fiscal deficit, and accelerating non-inflationary economic growth. They further suggest, that, in connection with power sector planning, Government and the sector should use tools that allow to capture better than was done in the past, on the one hand, the critical linkages with the macroeconomy and, on the other, the impacts the alternative investment sequences and their associated financial and pricing strategies can have on the macroeconomy. Implications for Bank Lending 37. Arising out of the lessons learned from Bank lending to the Colombian power sector, the present review suggests several recommendations which lead beyond that sector, into the Bank's involvement in other sectors and other countries. They can be summarized as follows: - When sponsoring measures to improve the performance of one sector, the Bank should assure their consistency with the broader, especially macro-economic policies agreed with Government. - The discount rate used in the traditional determination of the least-cost power development can strongly affect the choice of the type of plants and the most desirable sequence of projects to be proposed for implementation. There is a need for the Bank to clarify and justify its position on the degree to which this rate really reflects the opportunity cost of capital, so that the rate used in project analysis truly reflects economic conditions in the sector and the country. - The pervasive optimism--often not justified by experience-- reflected in many aspects of the Bank's presentation of programs and projects to its Board of Directors identified in the course of the present study, seems to be a general issue that it should address in a wider context. This may also apply to the pressure to lend, which, whether actual or only perceived, played a major, most likely detrimental, role in several projects reviewed here, especially those prepared in the 1980s. - In the context of the Colombian power sector operations, Central Projects Staff (CPS), and later Operations Policy Staff (OPS) (in this case the Energy Department), have been singularly unsuccessful in their role as quality controllers. This raises the question of whether the reorganized Bank, in which the quality control function is now carried out in much more immediate contact with the staff xiV processing the operations, and not by an independent group, has, in the light of the Colombian experience, a better chance of success. The review also identifies a temporary lack of broad vision and strategy the Bank would have shared with Government and sector as one of the factors that limited the success of the operations reviewed here. This is likely to hold for some activities beyond the sector and the country discussed here. The review further concludes that in Colombia, in part because of the above lack of vision and strategy, from the end of the 1970s, the Bank pursued, with limited resources, too many objectives at the same time, as valid as these objectives may have been. This shortcoming may also be more widespread than identified in this review, and may justify an analysis in a broader context. The review further illustrates the importance of Bank power sector lending strategy and operations being supported by thorough power sector work. This was important especially for a sector as complex as that of power in Colombia. However, at important junctures in the review period, the absence of such sector analysis appeared to have affected negatively the Bank's power lending strategy. There is a need in most countries for some form of strong independent regulatory body for the power sector, to act as a buffer between Government, the utilities and consumers. This is an essential element of the reforms needed in many countries. In countries like Colombia with strong regional sensitivities, the establishment of a central utility, such as ISA, for generation and transmission ownership is unlikely to be fully effective or commercially viable so long as it lacks support from the regional utilities and is concurrently owned by such utilities. Significant underpricing of electricity (US2.5/kWh for residential users who represented about one-half of total sales) was a major factor in the poor performance of the Colombian power sector, as it is in many countries. This highlights again the need to strive for the economic pricing of electricity and the minimizing of subsidies in all Bank power operations. Power investment, in several countries, has come to represent a substantial portion of total public sector investment and external debt. Greater attention must be given to the linkages between the levels of resource allocation to the sector and the proper financial management of the sector, on the one hand, and the achievement of macroeconomic objectives, on the other. The Bank should review its guidelines regarding the financial indicators it uses to monitor, and bind by covenant, financial performance of its power sector borrowers. xv The increased uncertainty and risk associated with critical variables such as power demand and fuel prices, necessitates that more focus be given ex ante to the flexibility and affordability of power investment programs. The latter, in turn means that financing plans should be complete and firm, at least for the first few years, and that given the importance of internal cash generation, as an alternative to excessive debt and budgetary impact, the necessary tariff levels be in place at the time the investment is decided with appropriate regulatory provisions to ensure that they will be maintained in the face of inflation and other adverse circumstances. The lower than expected growth in electricity demand in the 1980s due to the unforeseeable recession, and the rigidity in generation expansion programs led to excess capacity in Colombia and in many other countries. Clearly, methodologies for estimating power demand have not been perfected yet and research in this area should continue, while taking into account the difficulty of predicting periods of economic recession. Likewise, studies should be undertaken to produce scenarios which would confer greater flexibility to expansion plans. The high cost of surplus generating capacity also suggests the need for more cautious system planning. One possible approach involves targeting the long-lead time high investment cost projects (such as hydro and coal) to satisfy the assured baseload portion of the load curve, while leaving the short-lead time lower investment cost projects (such as combined cycles and gas turbines) to cover the remaining supply as the market pattern of demand emerges. 1 COLOMBIA THE POWER SECTOR AND THE WORLD BANK, 1970-1987 I. INTRODUCTION Development of the Sector, 1970-87 1. During 1970-87, the World Bank made 15 loans to Colombia for electric power for over US$1.9 billion (see Box 1.1). The Inter-American Development Bank (IDB), the other main source of foreign finance for the sector, contributed another $1.9 billion in 22 loans (see Box 1.2). Each contributed about 40Z of the long-term gross foreign borrowings of the sector--some US$5 billion. This was a period of rapid growth in installed power capacity and in coverage of the fast-growing Colombian population. Box 1.3 shows the main organizations that were active in the Colombian power sector and Table 1.1 gives a broad picture of how power operations developed: Table 1.1: POWER SUBSCRIBERS, GENERATION CAPACITY, AND SALES, 1971-89 Subscribers Generating Retail Sales Utility (thousands) Capacity in MN La in TWh 1971 1986 1989 1971 1986 1989 1971 1986 1989 EEEB 303 805 941 637 878 1158 1.8 4.8 5.8 EPM 183 414 491 579 1168 1369 1.6 3.9 4.5 CVC 231 485 547 262 860 821 1.1 3.0 3.4 ICEL 389 1,413 1,698 513 832 851 1.1 4.6 6.2 CORELCA 212 617 712 178 896 1,020 0.9 3.6 4.2 ISA - 0 1,885 2,652 - Total lb 1,348 3,831 4,530 2,169 6,519 8,370 6.7 20.4 24.6 Average annual growth rates, 1971-86 (%) 7.2 10.5 7.7 1986-89 (2) 5.8 8.7 6.4 /a Effective capacity. b Incorporates other smaller markets not included in those shown above. -2- BOX 1.1: WORLD BANK LOANS TO THE COLOMBIAN PWER SECTOR, 1970-87 Lean Date of Loan No. Borrower Pmject Approval Amount (US$ iTTTons) 0681 ISA Chivor Hydro 05/70 52.3 0874 EPH Guatape I' Hy1ro 01/73 56.0 1582 ISA San Carlos I Hydro 05/78 126.0 1583 Gov't 500-KV Interconnection 05/78 50.0 1628 EEEB Nesitas Hydro 11/78 84.0 1725 ISA San Carlos II Hydro 06/78 72.0 1807 EEEB Bogota Distribution I 03/80 87.0 1868 EPH Guadalupe IV Hydro 06/80 125.0 1953 EPH Pla as Hldro 03/81 85.0 1999 CORELCA Village lectrification 05/81 36.0 2008 EEEB Guavio Hydro 05/81 359.0 2401 FEN Power Development Finance 03/84 170.0 2449 EPM Rio Grande Multipurpose 06/84 164.5 2634 EEEB Bogota Power Distribution II 11/85 171.0 2889 Gov't Power Sector Adjustment 12/87 300.0 BOX 1.2: INTER-ANERICAN DEVELOPMENT BANK LOANS TO THE COLOMBIAN POWER SECTOR, 1970-87 Loan Loan Amount Approval NO. Borrower Project Nil. US$ Date OC-211 ICEL Subtr. Distrib 9.8 02/71 50-290 ICEL Subtr. Distrib 16.0 02/71 OC-214 ISA Chivor Hvdro 53.0 05/71 OC-249 EEEB Transa. Distrib 21.2 06/73 OC-258 CHEC Esmeralda/Insula Hydro 4.9 10/73 50-374 CHEC Esmeralda/Insula Hydro 2.4 10/73 OC-263 ISA Chivor II Hydro 58.6 12/73 IR-19 ISA San Carlos Hydro 96.6 12/77 IR-39 ISA Jaguas Hydro 111.7 11/78 IR-40 ISA Jaguas Hydro 30.0 11/78 OC-371 CVC Salvajina Hydro 51.5 12/79 50-608 ICEL Rural Electrification 50.0 08/80 IR-70 EPM Playas Hydro 106.4 12/80 IR-77 EEEB Guavio Hydro 172.8 06/80 OC-438 GOVERNMENT Small Hydro (ICEL) 25.0 08/83 IR-126 EEES Guavio Hydr Suppi 411.5 12/83 IR-147 ISA Jaguas Hydro SuppI 40.2 01/85 IR-160 ISA Jaguas Hydro Suppi II 52.0 05/85 IR-195 ISA Transmis ion 115.8 12/85 IR-237 FEN Distribution Program 80.0 12/86 OC * Ordinary Capital IR * Inter-Regional Capital SO * Special Operations -3- BOX 1.3: PRODUCING AND REGULATING POWER IN COLOMBIA Throughout the period under review, seven companies generated, transmitted, and distributed virtually all the power for public conscmption in Colombia: Three municipal ly-owned power companies--Empresas de Energia Electrica de Bogota (EEEB): Empresas Publicas de Medellin (EPM), and Empresas Municipales de Cali (ENCALI). The last of these is a distribution utility, but the first two also own and operate generating capacity and associated transmission facilities. A government-owned entity, Corporacion Autonoa Regional del Valle del Cauca (CVC), responsible for generating and transmitting bulk electricity to regional distributing agencies, in particular to ENCALI, the largest of these. instituto Colombiano de Energia Electrica (ICEL) and Cororacion Electrica de la Costa Atlantica (CORELCA). These government-owned entities provide-- through thirteen and seven subsidiaries, respectively--power supply outside the service area of the four utilities introduced above. CORELCA's area covers the departments along the Atlantic coast; ICEL's covers the rest of the country. Interconexion Electrica S.A. (ISA), a utility owned by EEEB, EPH, CVC, and ICEL (and since 1976, also by CORELCA) that is responsible for planning, constructing, owning, and operating major generation and transmission facilities in the system interconnecting its shareholders' grids. Four other organizations played a major role in the sector: The Ministry of Mines and Energy (NE), which exercises a policy and (limited) regulatory function in the sector. The National Planning Department (DNP) which, inter alia, reviews the sec- tor's investment programs and submits them for final approval to the National Council for Economic and Social Policy (CONPES). The Nati-*nal Tariff Commission (JNT) which operates within DNP and sets tariffs but, until 1986, lacked full regulatory responsibility in that area. The National Electricity Fund (FEN), which the government set up in 1982 to mobilize finance for power investment and to oversee the sector's finances. 2. Between 1971 and 1989, the number of subscribers in the sector increased about 3 112 times to some 4.5 million. Sales rose slightly more rapidly to about 24.6 teravatt hours (TWh), and installed capacity just about quadrupled, to some 8.4 gigawatts (GW). The share of the population with access to electricity increased by more than 15 percentage points, to more than 60%; in rural areas, the share increased from 152 to 452. 3. Actual developments in the sector fell short of what had been fore.sist. In particular, lower-than-expected growth in demand in the 19809 and rigidity in the generation expansion program led to excess capacity late in the decade. Most of the issues that were identified in OED evaluation studies of twenty -4- years ago1 are still unresolved, in particular: the complex and cumbersome institutional structure of the sector; poor financial performance it individual utilities and the sector at large; generally low and distorted electricity tariffs; a sectoral investment program that overly stresses generating plants, at the expense of distribution and suffers from large time and cost overruns; and a project selection procedure that has, in some cases, led to results that are economically and financially less than optimal. 4. The large power sector investment program, combined with the utilities' inadequate internal resource generation and large external borrowing undoubtedly had a negative impact on Colombia's economic growth and macrofinancial situation in the 1980s. (See Box 1.4.) OED's findings confirm that proper financial management of the power sector is an essential part of any effort to achieve macroeconomic targets such as balance of payments stabilization, reduction of the public sector fiscal deficit, and accelerating noninflationary economic growth. 1Two 1972 World Bank evaluaticn studies dealt with the Colombian power sector. The first assessed Bank lending to ten power companies, three of them in Colombia. Part of the second, a comprehensive review of all Bank opera- tions in Colombia, assessed the Bank's role in the sector during the period 1949-70. See Operations Evaluation Reports Electric Power (Report Z-17, March 1972) and Bank Operations in Colombia: An Evaluation Report (Report Z- 18, May 1972). -5- BOX 1.4: LINKAGES BETWEEN THE POWER SECTOR AND THE COLOMBIAN ECONOMY Although the share of the power sector in Colombia's GOP tripled between 1975 and 1986, it was still only 2% in the latter year. The sector's share in public investment rose from 20-25% in the mid-1970s to over one-third in the 1980s. This increase may have contributed to the simultaneous sharp decline of public investment assigned to the social sectors. There is little backward linkage from the power sector to the rest of the Colombian econovy, and up to 60 % of the sector's large investment expenditures were for imported goods and services. This high import intensity was the main reason why the power sector's foreign debt increased not only in absolute terms but also as a share of Colombia's total public external debt--from about 20% in the 1970s to one third in the mid-1980s. In that period total public external debt itself increased about fivefold. The power sector's foreign interest payments tripled between 1980 and 1987, by which time they represented 37% of total public sector foreign interest payments. Such a large allocation of external resources to the power sector is likely to have crowded out other sectors, possibly with better growth prospects, from access to external credit. Since the late 1970s the power sector has depended on government for funds, in amounts usually well over $150 million a year, and up to $300 million a year in the early 1980s. The power sector was one of the main contributors to the rise in public sector fiscal deficits. Poor internal resource mobilization meant that with the massive investments starting in the late 1970s, the sector t,rned to government to cover the substantial cash deficits that arose. In 1981-84 the sector's cash deficit was about 20% of the public sector's fiscal deficit, and in 1985 it reached one-third (about US$400 million), or about 1.2% of GP. History of Bank Relations with the Sector 5. During the review period the relations between government, the power sector, and the Bank went through two broad phases and the start of a third. First, in the 1970s, the emphasis of the dialogue was on interconnection, on the structure and role of ISA, and on sector finances as reflected in ISA's financial performance. Second, in the first half of the 1980s, the need to react to emergencies over- shadowed long-range considerations. Accordingly, the dialogue was less sharply focused; it shifted to distribution and village electrification and to issues associated with individual utilities with the sector issues being addressed more indirectly. By the middle of the decade, it returned to wider sectoral issues, especially financial performance. Third, at the very end of the period under review--1987, the preparation of the Power Sector Adjustment Loan set the stage for what appears to be a new phase of dialogue, focusing on broad sectoral issues that, though always discussed, were somewhat obscured by the exigencies of the early 1980s. 6. The Period 1971-78. Following the sector reorganization in the late 1960s, when ISA and CORELCA were created, and ELECTRAGUAS was reshaped and became ICEL, by the early 1970s, the sector was settling into its new setup, with ISA implement- ing the interconnection of the main regional systems (those of EEEB, EPM, CVC, and ICEL's main grids) and the 500 MW Chivor hydroplant. As electricity tariffs lagged - 6 - behind inflation and therefore deteriorated by about one third in real terms, ISA's shareholders did not fulfill their statutory financial obligations towards ISA which were an integral part of loan conditions agreed with the Bank. 7. In 1974 the persistent default led the Bank to suspend disbursements from the two power loans that were current. Although it resumed disbursements after about seven months, another three years of preparatory work and protracted negotia- tions were needed before the Bank was able to submit a new operation to its Board- -for ISA's San Carlos I hydro project and for the 500-kV line linking CORELCA's system with the other main utilities. The long preparation time was related to the hard stance the Bank (and the Inter-American Development Bank) took on tariff adjustments and on adequate modifications of ISA's bylaws. 8. Although, in 1976, in the Acuerdo de Sochagota, government and the sector agreed in principle on CORELCA becoming an ISA shareholder and on a tariff policy that was largely in line with the views of the main financiers, agreement on the specifics took much time and led to controversy between the Colombians and the Bank, as well as within the Bank. However, the extremely thorough preparation of the San Carlos I operation did permit the Bank to proceed quickly with two further generation projects (Mesitas with ZEEB and San Carlos II with ISA). 9. Th- Acuerdo de Cali. The year 1979, which separated the first from the second phase of Bank relations with the sector, was the year of the second major oil price increase, and in the sector, that of the Acuerdo de Cali. This agreement brought into the open a development that had been going on for quite some time, but which had never been clearly acknowledged. It led away from the concept underlying the creation of ISA, namely, that this utility would become the absolutely dominant entity in the sector and which, beyond planning and operating the national grid, would construct and own all major additions of generation and transmission facilities in the interconnected system. 10. At the time of the Acuerdo de Cali, the sector's main preoccupation was meeting the demand in the short term (i.e. the early 1980s) and the medium term (i.e. mid-to-late 1980s). The delays incurred in the construction of the main hydroplants and the protracted preparation and arrangements for financing of the San Carlos project induced Government and the sector to carry out two emergency programs largely involving thermal generation plants. However, these proceeded at a much slower pace than planned. 11. These delays in program implementation and a two-year outage of one of the large existing plants led to the sector's inability to supply temporarily up to 18% of demand in 1980/81 (but only around 3-5% of energy needs). Further but less stringent rationing occurred in 1982 and 1983. 12. The recognition by the late 1970s of impending power shortfalls, combined with the view in the sector that ISA was over-committed with the construction of Chivor II and San Carlos, whereas the experienced regional and municipal power companies had available implementation capability, resulted in the Acuerdo de Cali. - 7 - 13. In this agreement, the sector entities decided to break formally ISA's Bank backed monopoly on construction and ownership of new major plant in the inter- connected system. Indeed, the agreement essentially distributed the next generation investment plan among ISA and its shareholders. The Bank, eventually succeeded in limiting the erosion of ISA's role by obtaining, from Government and sector, assurances that ISA would ultimately own about one third of the capacity installed in the entire system. 14. The 1980s. The Acuerdo de Cali was followed in qAick succession by the start of the world recession, the emergence of the international debt crisis, and, in the power sector, a sharp drop in the growth rate of electricity demand that was detected only about 1983 largely because rationing in the 1980/81 had masked true demand levels. During this time, the Bank concentrated on projects that EEEB and EPM were implementing in accordance with the Acuerdo de Cali and on distribution and rural electrification- -areas that it had long identified as inadequately covered. 15. Bank lending increased to an unprecedented level of $692 million in five operations over about two years, culminating in 1981 in the operation for EEEB's 1,000-M Guavio hydroplant, supported by a $359 million loan. But construction problems at Mesitas and Guavio, leading to cost overruns; delays in most major projects, entailing further massive cost increases; and difficulties in providing local funds, as tariffs, though increasing, never caught up with requirements, caused seemingly endless financial difficulties in these operations. 16. The 1984 Bank loan to support the development of FEN, the organization set up to mobilize resources for power investment, tried to address some of these problems, as did the 1987 Sector Adjustment Loan, which had to deal with the further deterioration of the sector's financial performance after the large devaluation of the Colombian peso in 1985. 17. Despite variations in its approach, the Bank, through its lending over the entire review period, supported major sector initiatives such ass the creation of stronger regional utilities, with the major existing companies as a nucleus; the preparation of a development master plan; systematic generation and transmission planning, based on least-cost analysis; balancing investment between generation and transmission, on the one hand, and distribution, on the other; reduction of system losses; a unified presentation of financial data and progress toward compatible financial accounting in the utilities; marginal cost pricing; and a reasonable contribution to investment from sector revenues. 18. Success in some of these endeavors was achieved, but in general it was limited. For example, although decided on at the highest level of government, the creation of new regional aggregations of utilities did not proceed. Preparation of a sectoral development master plan was drastically curtailed. Generation and transmission planning, as carried out by ISA, improved vastly, albeit with the flaws stemming from the tendency of ISA's shareholders to introduce their project preferences at an unduly early stage, at the expense of nationally-oriented considerations. Despite all efforts, the investment program at the end of the review period remained as unbalanced as ever, at the expense of distribution. Colombians accepted, in 1986, the princip'e of pricing according to average long run incremental costs, but its implementation remained slow, leaving the sector far from the goals set. Substantial tariff distortions and subsidies persisted, implying very large subsidies to residential consumers (particularly to high income, high consumption residential subscribers), partly at the expense of industry and commerce. Overall system losses increased by about 50%, while those in EEEB and CORELCA roughly doubled. Producing consolidated financial data for the sector is still exceedingly difficult, and the results are fraught with incon- sistencies. Internal cash generation fell far short of targets, never contributing more than 10% to the sector's investment (after a momentary peak of around 15 percent in 1976), and failing to ease the heavy burden the sector imposed on government finances. 19. Despite the Bank's efforts and similar actions that were taken by the IDB, at the end of the period under review, the power utilities' performance was, with some exceptions, unsatisfactory and the sector remained a heavy burden on government. However, there are indications that the measures associated with the 1987 Power Sector Adjustment Loan are likely to bring about a better dialogue among government, the Bank, and the sector. II. SECTORAL OBJECTIVES AND INSTITUTIONAL ISSUES Introduction 20. Four broad issues played a prominent role in the review period and will continue to do so in the foreseeable futures - sector objectives; - sector regulation; - restructuring of the sector; and - ISA's structure and role. 21. A prerequisite for successfully addressing such broad sector issues is an equally broad vision of what the sector should accomplish in the medium and long term. In the 1970s the Government, the sector, IDB, and the Bank, largely shared such a vision, which focused, both the policy dialogue and lending activities, on implementing interconnection and developing ISA as a central planning institution and, as the main future generation and transmission utility. In the late 1970s, however, these parties failed to adjust their vision, when, at the time of the Acuerdo de Cali, parts of the original goals proved untenable in particular, ISA as the sole owner of large additions to the generating system. Therefore, in the 1980s, sector development, which already was affected by a hostile economic environment, also had a weaker sense of direction. Sector Objectives 22. Formulating a comprehensive set of precise sectoral objectives did not seem urgent in the 1970s. The substantial progress then being achieved seemed sustainable on the basis of measures addressing specific management, planning, and operational issues (e.g. generation investment and shareholders' contributions to -9- ISA). This reflected the above original broad vision of the sector's development. Therefore, in the 1970s, Government did not set sectoral objectives beyond the very broad ones calling fort meeting demand within justifiable limits; using resources optimally; and achieving financial self-sufficiency. 23. From the beginning of the 1980s, however, the need for specific sectoral objectives to be set became more pressing. This arose mainly because of two factors. First, the changes in the broad vision of sectoral development due to the Acuerdo de Cali, necessitated that a new consensus be forged between the Govern- ment, sector, Bank and IDB on what the new vision of sectoral development would be. Second, as the economic environment deteriorated in the early 1980s, success became elusive of fulfilling the above broad sector objectives. However, as no consensus on the new vision of sectoral development emerged between the parties, little progress was achieved towards further definition of specific sector objectives. Sector Regulation 24. In the period reviewed, the regulation of the sector was affected by the creation of the *DE and the strengthening of the JNT. The MME, through its Energy Division, oversees the utilities government owns directly (ICEL and CORELCA)z it also formulates government policies in the power sector and translates them into regulations. However, its resources have remained limited and so has its influence on sector regulation. 25. In the mid-1980s, after playing only a marginal role in sector regulation, JNT emerged as the entity that sets electricity tariffs, a development to which the Bank contributed substantially through its long-standing encouragement of improvements in tariff policy and implementation. 26. In 1987 Government endorsed a proposal made in a 1986 study, a *Basis for Formulation of a Colombian Energy Policy. sponsored by the United Nations Development Programme (UNDP) and the Bank, and introduced draft legislation into Congress for the creation of the Energy Board as the body that would regulate the entire energy sector and, hence, play a major role in regulating the power sector (see Box 2.1). The law was passed in 1989 and should provide an adequate framework for urgently required improvements in sector regulation. Restructuring the Sector 27. In the 1970s and 1980s, several initiatives supported by the Bank and involving restructuring appeared but did not progress beyond the planning stage, mostly for political reasons and because of the weak regulatory setup. The most important such initiatives were the merger of CVC and EMCALI, and the redistribu- tion of markets. Such redistribution was to combine the main municipal utilities with the ICEL subsidiaries surrounding them; the most immediate such clusters implied the consolidation of EEEB with the Cundinamarca and Meta utilities, and that of EPH with the Antioquia utility. - 10 - Structure and Role of ISA 28. ISA, a young institution at the start of the review period, had been created to build and operate all major additions to the interconnected power system. The Bank supported this concept of ISA's role. ISA has become a truly national entity since CORELCA became a shareholder. In the mid-1980s it was also the largest generating utility, with a 31Z share of the country's total. It is further the sector's central planner of investments in generation and transmission plant, and toward the end of the review period, it also assumed a coordinating function in distribution planning. But the individual interests of ISA's shareholders--which are the other major utilities in the sector--were often at odds with those of ISA itself, and prevented ISA from achieving the monopoly originally envisaged for it in building and operating major new plant for the interconnected system. This outcome was predictable, given the fierce regionalism in Colombia. 29. ISA's shareholders also tended to introduce their preferences at an unduly early stage in ISA's planning, at the expense of nationally oriented considera- tions. This may not necessarily have led to uneconomic solutions, but it impeded a proper assessment of the cost to the economy of the shareholders' preferences and a transparent choice of the most desirable solutions. These actions also prevented the bulk tariff that ISA charges for its supply and services from reflecting economic cost, and thus from allowing ISA to achieve satisfactory financial performance. Further, ISA's present operating rules appear to provide for the dispatch of energy essentially on the basis of financial rather than economic considerations. The Role of the Bank 30. The Bank was instrumental in most institutional improvements that occurred during the period under review: it insisted on CORELCA becoming an ISA sharehold- er, which occurred in 1976; It strongly supported ISA in its growing planning role; and, after the Acuerdo de Cali, it successfully impeded further erosion of ISA's role as a major constructor and owner of the sector's generating facilities. For a long time, however, the Bank failed to acknowledge that some of ISA's shareholders did not identify with ISA's planned role as agreed in 1968; there- fore, it operated on a somewhat unrealistic basis that, in the long run, proved untenable. 31. This shortcoming is associated with the Bank's obvious difficulties in dealing with Colombia's regionalism, which it seems to have considered, much of the time, an utterly negative force. Rather than squarely confront regionalism, the Bank chose to deny it the high profile it actually had. With hindsight, this failing impeded a fruitful, in-depth dialogue about the broad sectoral issues. In 1987, the Bank again assumed its original position that ISA alone should construct and own major additions to plant. Experience suggests this stance to be less than promising, in spite of progress achieved in this direction, because the underlying issue--the identification of the sectoral entities with the overall policy-- remains unresolved. - 11 - Conclusions and Recommendations 32. One of the prerequisites for addressing the institutional issues is full acknowledgment of some historical and political facts--in particular, Colombia's regionalism. A positive approach to this phenomenon, and the willingness to use its strengths, would be especially helpful. 33. Given the difficult circumstances at present, reaching a new consensus among government, the sector, the IDB, and the Bank on the broad direction of sectoral development--and the translation of this vision into precise and consistent objec- tives at the level of government, the energy sector, and the utilities--now has high priority. Acknowledging this urgency should facilitate actions to make the mechanisms of the power sector more transparent, a prerequisite for improved efficiency. Once the broad principles are defined, government will need to: - spell out the strategy according to which the sector decides to meet (or not to meet) demand--that is, which demand should be met and to what extent; - set the ground rules for using national resources in the power sector, within the context of an energy sector policy that is itself framed with reference to macroeconomic policies; - clarify the utilities' basic rules of conuercial operation, in particular the costs that revenues would have to cover; - define, in general terms, the type of operations that government would subsidize (operations that are not financially viable in the short run but serve specific socioeconomic goals--examples include elements of rural electrification and transitory refinancing measures to put the sector back on a financially sound footing); and - spell out the principles for setting bulk and retail electricity tariffs and their relation to average long-run incremental costs, as well as the measure of cross-subsidies among consumer categories that will be tolerated. 34. Whatever the outcome of the ongoing discussion about the Energy Board (see Box 2.1), the power sector sorely needs an organization to unify--or at least, firmly coordinate--the functions now exercised in the area of power sector regulation by the Electric Power Division of the MME, the DNP, the JNT within DNP, the PEN, and, to some extent, the ISA. * 12 - 80 2.1: THE ENERGY BOARD Created in V189 as part of the ambitious program supported by the 1987 Power Sector Adjustment Loan, the Energy Board is made up of the Minister of Energy and Hines, the Director of the National Planning Department, and represen- tatives of the energy agencies. On the basis of analysis by its staff, the Board sets policies, regulates the use of energy resources, and approves investment plans for the various subsectors. A committee reporting to the Ministry of Energy and Mines will monitor on a continuing basis all the power sector utilities. 35. Such a regulatory body, which would be under the purview of the Energy Board, should in general translate government objectives and policies into specific sectoral rules and should exercise, inter alia, the following functions, presently only loosely coordinateds - define for the individual utilities areas of autonomy that are as large as possible--in order to encourage initiative, entrepreneurship, and regional dynamism--but are limited in order to keep national objectives clearly in view; - strengthen sectoral planning and operational coordination (that is, in the present setup, support ISA in carrying out its main tasks); - establish and oversee the decisionmaking mechanism connecting technical and economic sectoral planning, financing, and project implementation; - set, in accordance with the general government objectives for the sector, operational and financial targets for the individual utilities, as well as associated incentives and penalties;2 - oversee creation and operation of joint venturbs to build and own new plants, following rules defined in accordance vi, the first point above; - monitor operations, identify shortcomings, and agree with the utilities on remedial action; 2These can be given operational effectiveness through Performance Contracts as practiced in some countries. These Contracts are entered into between the Regulatory Agency and the utility whereby the goals to be achieve- d, and the means by which they would be achieved, are agreed to ex-ante and then assessed ex-post by the Regulatory Agency. Appropriate penalties and incentives being provided for poor and good performance, respectively. - 13 - - set the tariffs in agreement with the general objectives and rules set forth by government and oversee their implementation; - link up with the proper government institutions, in particular the Ministry of Finance, to ensure that government contributions, established on the basis of the sector's investment and operational planning in accordance with the general government objectives, come forth at the proper time and in the proper amounts; - define the rules of interchange among utilities and, in particular, of sharing surpluses and deficits within the sector. 36. The parties will also have to address the issues related to the structure of the sector, among them the creation of more efficient units and the redistribu- tion of markets. The realignment and concentration of utilities would undoubtedly increase efficiency in the sector. The chances of its implementation would improve if the balance between autonomy and intervention through the regulatory process is ensured. This implies, first, that the new groups with the large municipal and regional utilities at their core should be free to organize themselves properly and, second, that government support is guaranteed to help maintain the financial viability of the clusters integrating some weak utilitieu. A candid analysis and discussion among all parties of the lack of success of the efforts to integrate (1) EEEB and ICEL's subsidiaries surrounding it; (2) EPM and ICEL's Antioquia subsidiary; and (3) CVC and EMCALI could be a valuable first step toward a meaningful dialogue. 37. The compromise on ISA's structure and role, as reached in 1968 and subsequently amended, served its purpose reasonably well for quite a while. However, in recent years the drawbacks of the current arrangements call for a thorough re-examination of ISA's functions and of the framework in which it must discharge them. The rethinking will of course need to acknowledge ISA's major achievements and to capitalize on the proven strengths of the organization and its staff. Such a rethinking may well imply the re-assignment of some of ISA's functions in the attempt to accommodate some of the following partially con- tradictory points: - There is an advantage in keeping central planning and execution (both in investment and operations) close together, as is presently the case in ISA, because this promotes a less theoretical approach than when planning is done in isolation. - There is a clear need to shield planaing at the early stages from direct political influence, as is exercised now within ISA's Board in other words, the political input into the decision process should come at a different level and at a later stage than it does at present. - The time may have come for government, mainly through proper regulation, to assume a more direct role in the power sector, in general, and in ISA, in particular--a role that would enhance the achievement of macroeconomic objectives and be more commensurate with the financial obligations government has to assume on behalf of the sector and ISA. - 14 - - It would be desirable to reduce the prominence of the plant ownership issue by redefining the role of the central generating and transmitting utility in the light of recent experience (inter alia, with project execution by ISA conducted jointly with one or several of the other main utilities). III. ELECTRICITY DEMAND AND SUPPLY ISSUES Introduction 38. On the supply side, the power sector registered important achievements during the review period, all of them supported by finance and advice from the Bank. First, electricity coverage rose from 45? of the population in 1970 teo 60% in 1986, when about 3.5 million households had access to electricity. Second, the interconnection between the southern and central systems, completed in 1981, and ISA's emergence as the largest utility made the system more flexible; energy transfers over the interconnections accounted for about 40? of the overall needs. Third, in 1984 the sector completed the main interconnection process, when it linked the Atlantic region (CORELCA system) with the rest of the country. This interconnection not only allows, in principle, economies in reserve capacity but has also created a large potential for substituting thermal energy from the (entirely thermal) CORELCA system for hydro energy from the other grids.3 Sales Structure 39. The share of residential consumers in total system sales increased from 41 in 1971 to 48Z in 1987. This significant shift reflects rapid growth, of 7.6 percent annually, in the number of (heavily subsidized) residential connections, coupled with a slow increase, of 5.6 percent annually, in industrial value added, while the ratio between industrial output and electricity consumption stayed roughly constant. This rise in the share of residential sales to virtually one- half of total sales has, in part, also been brought about by the very low pricing of electricity to households, which has encouraged uneconomic demand. 40. A second important change in sales structure was the decreasing share of the two major municipal utilities, EEEB and EPM, in total system sales, from 51? in 1971 to 42? in 1987. In contrast, the share of the regional utilities serving both urban centers and rural areas, ICEL and CORELCA, rose from 30? in 1971 to 41Z in 1987, in large part in response to government policy to increase electricity 3Up to 1987, the scale of energy transfer actually remained limited since the line operates at 230 kV and not the design 500 k, thereby restricting its carrying capacity. This has arisen because of a lack of compensating capacity in CORELCA if the line failed. In addition, large fuel subsidies in the CORELCA system, make increased bulk purchases from the power pool financially unattractive for the regional utility. - 15 - coverage. Because the latter companies have always been financially weak and less efficieut than the other main utilities, the structural change in demand has tended to exacerbate the sector's financial problems. Demand ForecaLs 41. Throughout the 1971-86 period the dominant influence on the growth of overall power demand was the growth and structure of gross domestic product (GDP). With the 1981-83 recession and the associated collapse of GDP growth, power sales grew by much less than had been forecast: growth between 1981 and 1986 was about 6Z a year, compared with the 102 annual rate that had been projected in the late 1970s for the 1980s. By 1986, overall sales were about 30 percent lower than had been projected. (Another reason, though much less important, for the lower sales was the escalation of energy losses.) 42. In the 1970s forecasts of power demand or sales were within the accepted range of 10? of actual outcomes, even though they were based on simple "statist- ical* models. This was, in large part, due to the more stable macroeconomic environment. But in the 1980s, the extreme macroeconomic uncertainties made power demand forecasting much more difficult. In particular, the severity and depth of the 1981-83 recession were not predictable during 1978/79, when the major decisions were being made about new capacity additions for the 1980s to cover demand projections that could have been viewed as somewhat high but not unreasonable. Power Rationing and the Cost to the Economy 43. Utility managers and political authorities invariably perceive power shortages as expensive for both the utilities and the economy. But the real economic cost of shortages depends on the nature, duration, and size of the outages as well as the markets affected. Although difficult, the valuation of shortages is important, since very high estimates, such as those discussed below, provide encouragement for increasing supply-side investments to improve system reliability beyond what is justified and, above all, what government is willing to ask consumers to pay for. 44. Advance notice was given in all three cases of supply shortfalls during the period reviewed (1977, 1980, and 1981), so that the costs of the outages were far lower than in the case of sudden breakdowns. Surprisingly, the President's Report for the Bank-supported Guavio operation (1981) estimated that the cost to the economy from an estimated 1,200 Gvh shortfall in that year would be about US$3 billion (at 1981 prices). OED has not been able to determine the basis for this estimate, except that the cost of rationing was not taken into account by ISA as part of its objective function during planning. From information that was - 16 - available at the time, OED recalculated an ex ante cost of such an outage at not more than US$500 million (1985 prices).4 Assessed from an ex post perspective, the outage costs in 1981 are a mere US$130 million (1985 prices), or only 42 of the figure in the President's Report. 45. There are two reasons for the still-large difference between OED's ex post and ex ante estimates. First, the onset of recession in 1981 reduced the expected level of demand in that year below what was anticipated when the Guavio project was submitted to the Bank's Executive Board, thereby lowering the actual power shortage. Second, as is normal policy, rationing, when it became inevitable, was planned to affect residential users most and industrial users least, thus minimiz- ing the cost to the economy. Economic Cost o! the Surplus Generating Capacit! 46. The shortages of 1981 coincided with the beginning of the recession and signif.cant weakening of power demand growth. However, by 1985, the Colombian power system was entering a period of excess generating capacity, which is expected to last until 1993/94. This excess capacity is due to the additions that were decided in the late 1970s on the basis of demand forecasts that largely extrapo- lated past trends. With much slower demand growth in the 1980s than foreseen, the power system was unable to respond to this development because of the rigidity of its investment plan associated with the dominance of large and long gestation hydro projects and the absence of smaller thermal projects (such as gas-fired combined cycle plants) with shorter lead times. 47. OED estimated the level and costs of the excess generating capacity by analyzing the energy and peak demand balances of the system between 1985 and 1993. The excess energy available was about 24Z of actual requirements in 1989 and is expected to decrease slowly to about 10Z by 1993. In contrast, effective installed generating capacity in 1989 exceeded peak demand by about 50%, with this expected to decline to around 39Z by 1993. Since the excess capacity is now unavoidable, its cost is really a 'regret." This 'regret* of excess capacity is estimated to range from $410 to $450 million (1985 prices), depending on whether one uses the information available in 1978, for an ex ante estimate, or the information available today, for an ex post estimate. 48. Of particular relevance is that the ex post estimates of the cost of excess capacity are about 3.5 times higher than the ex post estimate, of $130 million in 1985 prices, that OED made of the cost of the 1981 outage. This example il- lustrates that the highly visible and, hence, politically costly, effects of 4This estimate is based on the assumption of the rationing being appor- tioned among consumers in the same ratio as their share of sales. As such, it represents an upper bound on ex ante costs, since general rationing policy is to minimize outages to industrial users. - 17 - rationing have been less important, in economic terms, than the invisible and seemingly costless excess capacity. This suggests important lessons for system planning, which often can focus too strongly on outage avoidance and neglect surplus avoidance. System Losses 49. Power losses in the present context are measured as the difference between billed sales and net generation in percent (Z) of the latter; thus they exclude the generation plants' own use of power. The two components of power losses are, first, technical losses in the transmission and distribution (T&D) systems, which represent energy that is consumed as losses because of the physical characteristics of the T&D system; and, second, the nontechnical losses. The latter represent the sum of consumption and billing losses--the former accounting for the fact that not all consumption is accurately recorded and the latter that not all recorded consumption is accurately billed. 50. Technical losses are measurable and have technical solutions such as improved T&D system design (involving, for example, changes in voltage levels, capacitor additions, and reconductoring of lines). Nontechnical losses are more a function of the socioeconomic environment and of utilities' management and supervisory capability, since they relate to such shortcomings as incorrect meter calibration, meter tampering, illegal connections, legal but unmetered connections, and error and fraud in meter reading and bill processing. 51. Efforts to reduce the two types of losses thus require different types of capability. 'Often, a company may be better able to reduce technical losses than nontechnical losses. Further, though there are broadly applicable norms for technical losses, the level of nontechnical losses that is tolerable varies among companies. 52. Throughout most of the 1970s, total power losses remained reasonably stable in Colombia's regional markets (except in CORELCA's, where they increased). From 1978 on, aowever, they grew rapidly, except in the CVC and EPM systems.5 The deterioration reflects mainly an escalation in nontechnical losses. A substantial though not measurable part of the increase in these losses can be traced to the growth of unmetered but legal connections. For example, by 1986 the number of such connections in EEEB's system had reached 133,000, as one in five residences in the system fell into this category. Given the other problems faced by EEEB, in the form of cost overruns and delays in the Mesitas and Guavio projects, the Bank's and the utility's lack of early focus on this component of losses is understand- able. An attempt to address the problem was incorporated in 1985 into the Bogota 51n EPM they had always been high (18-20Z) but stable, and they remained so. In EEEB they surged from 16Z in 1976 to 26% in 1987. In ICEL and CORELCA they increased to levels comparable to that in EEEB. Indeed, in the EEEB system technical and nontechnical losses were 12Z and 4Z, respectively, in 1976 and 11% and 13% in 1986. - 18 - Distribution II loan, through financing for the installation of meters additional to those naeded to satisfy the requirements of new customers. The results of these measures should become apparent in the medium term. 53. A second factor, especially in the EEEB market, that has aggravated nontechnical losses has been the massively distorted tariff structure. In the review period EEEB charged commercial and industrial users close to 1002 and 50Z. respectively, above their long-run incremental supply costs, creating incentives for large-scale theft through sophisticated meter tampering and other fraud. Efforts by the Bank to have the EEEB tariff distortions corrected have not been successful and call for stronger measures in the future. Some recent measures toward this end may show promising results. 54. During the 1980s, contradictions surfaced in the Bank's approach to the issue of losses in EPM and 2BB. With EPM, in connection with the Playas (1981), Rio Grande II (1984), and Sector Adjustmant loans (1987), the Bank agreed that 18- 20% losses were realistic in EPM's system. Because no ambitious targets were set for loss reduction, it is fair to assume that 18-20% was considered a tolerable level in the given environment and company, which the Bank had always considered well-managed and reasonably efficient. With EEEB, in contrast, the Bank agreed in connection with the Bogota Distribution II operation (1985) and the Sector Adjustment Loan (1987), that the company would reduce its losses by one-half, to 13%, by 1992. Not only the sheer size and pace of the planned improvements tax credibility, but also the fact that at the time of these assumptions EEEB was already struggling with major financial, project (Guavio), and managerial problems, well known to the Bank. These difficulties would have made the company an unlikely candidate for performing so much better than 1PM. Conclusions and Recommendations 55. The Colombian power sector, from the supply perspective, has registered some important achievements during the review period which the Bank has been identified with. However, there are lessons to be drawn from those areas where either success has been limited or failure very evident. 56. This review has found that econometric demand models have demonstrated their validity and should be further used as a complement to the more traditional methodologies (e.g. ext;apolations), while recognizing that incorrect GDP projec- tions lead to equally large deviations between power demand forecasts and outcomes. Nevertheless, with the benefit of hindsight, the Bank should encourage the sector authorities: - to subject proposed investment sequences to greater variations in the range of projected power demand than in the past; and - to incorporate investment strategies implying greater flexibility, for example, by including some smaller projects with shorter gestation periods, so that the investment program can be adjusted more easily to unforeseen demand changes. The introduction of multi-objective planning could provide a further tool to analyze such alternative strategies. * 19 - 57. In designing futue power loss reduction programs, the Bank needs to assess critically the degree to which overall losses can be reduced significantly, as. long ass (i) utility profitability is not an objective at the forefront of the sector's strategy; (ii) government subsidies to the companies remain pervasivel and (iii) utility managers have little incentive and face no penalties in connection with success or failure to achieve loss reduction objectives. 58. In terms of its procedures, the Bank needs to enforce more carefully than in the past ten years that its project presentations, for example, in Staff Appraisal Reports (SARs) include at least three years of histotical data for key parameters, such as sales, income, tariffs. etc. to ensure that the starting point of projections properly reflect reality. IV. TARIF? ISSUES Introduction 59. Seriously inadequate tariff levels and increasingly severe distortions in tariff structures, both in individual markets and in the sector as a whole, were a persistent source of tension in the relationship between the Bank and the Colom- bian power sector in the period reviewed. This was not a new development; in its 1972 review of the Bank's power sector lending, OED concluded that the Bank had been well justified in applying pressure for aoequate tariff levels among its bor- rowers. That review also pointed out, however, that the Bank had not given proper weight to the relation between the cost of supply and the tariff. It especially highlighted the substantial cross-subsidies from industrial to residential consumers then prevailing in the EEEB system. It further found that no account had been taken of the effects of tariffs on income distribution. 60. In the eighteen years covered by the present review, three periods of tariff development are identifiable (see Table 4.1).6 But looking at the 1971-87 period as a whole, the average retail tariff levels in the CORELCA, CVC, ICEL, and EMCALI markets (representing about 602 of sales) showed virtually no change in real terms. Although in the EEEB and EPM mar'- -t tariffs increased modestly in the review period (by 3.42 and 1.62 annually zespectively), the overall tariff performance was poor; the more so in view of t. large investment program the sector undertook in the period. 6The first, 1971-75, saw the average national unit revenue decline by some 82 annually in real terms, with the greatest deterioration in the CVC, EMCALI, and EEEB markets. The second period, 1975-83, witnessed a systematic improvement in all regional markets, with the average national unit revenue increasing by about 92 annually. The EEEB market experienced the most rapid advance, at some 132 annually. The final period, 1983-87, saw a uniform retreat, with a decrease of some 4Z annually in real terms. - 20 - Table 4:1 ANNUAL PERCENTAGE CRANGES IN NATIONAL AVERAGE Unit Revenue in Real Terms, 1971-87 Residential Commercial Industrial Total 1971-75 - 9.2 - 6.5 - 6.4 - 8.1 1975-83 6.6 10.8 10.9 9.2 1983-87 - 7.8 - 2.2 - 2.6 - 4.2 1971-87 - 1.3 2.9 2.9 1.2 Achievements 61. Up to the early 1980s, virtual anarchy prevailed in Colombian power tariffs, with almost as many tariffs as utility companies (more than 30) and, above all, the lack of a common basis for setting tariffs in the different markets. This situation was perhaps tolerable as long as the systems remained independent; with interconnection, however, better coordination was needed. 62. Increasing dissatisfaction with the state of affairs led in 1984 to the first major reform through Decree 2545, which focused, correctly, on residential tariffs. The Decree defined, for both the variable and the fixed-rate components, a uniform structure for the sector -- a significant advance. At the same time, the fixed-rate element of the tariff was used to achieve some limited level of intra- residential consumer cross-subsidies from high to low income households. These intra-sector cross-subsidies were, however, much smaller than the inter-sector cross-subsidies from industrial and commercial users to residential consumers as a whole. Decree 2545 met stron resistance, especially from EEE, which had the most distorted tariff structure. 63. Another important advance occurred in September 1986 when the government, through JNT and with the Bank's support, assumed full control of retail tariff policy by issuing JNT Resolution 086. This removed the utilities' discretion to request and apply authorized rates and required them to apply the tariffs set by JNT. For the first time the retail tariff rates also explicitly incorporated criteria based on efficiency and an economic rationale, in that they were linked to the average long-run incremental costs (IC) of supplying electricity to different consumers. Socioeconomic considerations were reflected in 20-80Z discounts from IC for various types of residential consumers and up to 20% margins above IC for the other types of consumers. 7To overcome this difficulty, in February 1986 JNT passed Resolution 020, which set out specific provisions for EEEB to conform with the national policy determined in Decree 2545. - 21 - 64. The sector also achieved a significant reduction in the large disparities that existed in the 1970s in the electricity subsidies between rural and urban households. Traditionally, average residential tariffs for rural utilities have been significantly higher than for the large municipal companies, albeit in a measure unrelated to the differential costs of service. In particular, cross- subsidies from industrial/commercial to residential users were much higher in municipal markets. Since 1983, however, these differences with their associated negative rural/urban income distribution effects have markedly diminished. Costs of Service and Tariffs at Retail Levels 65. A comparison of the estimated average long-run incremental cost (IC) of supplying electricity to consumers overall and the revenues of the sector shows that the annual shortfall, already significant in the 1970s, had grown by 1986 to some US$470 million (1986 prices). This represented in 1986 about 9z of the sector's net fixed assets employed. Indeed, the cumulative cost of pricing electricity well below its average IC for the seven-year (1980-86) period was roughly US$2.3 billion (1986 prices). 66. The long history of very low retail tariffs relative to IC raises the question of how the sector survived financially. The reasons appear to be threefold. First, ICEL, CORELCA, and to a lesser extent CVC traditionally received significant capital transfers from the central government. Between 1978 and 1985, central government capital transfers to the sector were some $1.1 billion (1986 prices), or about 30Z of the total central government capital transfers in that period. Second, until the mid-1970s, EEB and EPH owned predominantly low-cost hydro supply systems, which resulted in their average costs being lower than the IC of the system as whole. Thus for a while they could sustain low tariffs, until costly expansion programs began in the late 1970s. Third, tariffs were kept lower than IC through recourse to massive borrowing. 67. The immediate financial consequences of not adjusting tariffs to IC are modest. Indeed, during the long construction periods of hydro projects, financial difficulties do not appear while external financing is available. Problems develop, however, as soon as debt servicing begins. Since 1977/78, the Colombian authorities have tried to adjust retail tariffs to allow them to catch up with IC. Some progress was achieved between 1978-83; however, much of this was undone by 1986 when the sector's average unit revenue deteriorated to 213 of the average IC in the wake of the 1985/86 peso devaluations, which were not reflected in the power tariffs. Due to the slow pace of tariff adjustment, which was agreed to in connecticn with the Bank's 1987 Sector Adjustment Loan, the catch-up period now extends well beyond 1992. Existing agreements suggest that even then, retail tariffs will still not have reached even close to IC. The sector will then, however, be about to enter a new cycle of expansion in investments and will therefore face larger claims on financial resources which will compound the difficulties. - 22 - Distortions in Retail Tariff Structures 68. Combined with a low level of unit revenues relative to the overall system average IC, the review period saw an increasing distortion in tariff structures, especially in the EEEB market. Such distortion reflects itself in the tariffs charged to specific users bearing little relation to the structure of the supply costs to these end-users. For example, in 1978, EEB's residential, commercial, and industrial tariffs were about 332, 66Z, and 502, respectively, of their estimated supply costs, but by 1986 these had changed to 33Z, 1902, and 1502 of their supply costs. In other companies the picture, though less dramatic, showed a similar pattern. This has resulted in three discernible effects. First, as noted earlier, excessively high tariffs to commercial and industrial users have provided incentives for large-scale theft and meter tampering, aggravating non- technical losses. Second, excessively low tariffs to residential users (which already represent the largest consumer block) have over-stimulated electricity demand, especially for end-uses such as cooking, where electricity is not the most economic choice. Third, as discussed further below, given the structure of electricity demand, the overall level of unit revenues cannot be raised without reducing the degree of distortion in tariff structures. Subsidies and.Income Distribution 69. During the review period, subsidies for residential electricity consumers grew increasingly costly. Between 1978 and 1987, for example, their estimated annual economic cost (in 1986 prices) rose from some $230 million to $400 million. The subsidies tended to be poorly targeted, benefiting relatively well-off consumers. 70. Residential electricity subsidies are achieved through two mechanisms. The first involves intra-sectoral transfers within the residential sector itself, as high consumption (essentially by higher-income consumers) is priced above the average price paid by residential users, and the revenue used to cross-subsidize low consumption (largely by low-income households) priced considerably below this average price. In RPM, the relation between the marginal tariffs for the highest and lowest residential tariff blocks increased from just over 2:1 in 1968 to more than 11:1 in 1986. 71. The second subsidization mechanism is intersectoral transfers from industrial and commercial consumers to residential consumers. Such a policy, if used in moderation, can help to redistribute income without creating distortions that become untenable. During the review period, however, such caution was thrown aside. 72. The structure shown in Figure 4.1, for the Bogota market, is illustrative of similar structures found in other regional markets. The figure shows the most negative aspects of the effects of the subsidies to residences according to separate income and consumption classifications. Based on January 1989 tariffs high-income households (strata 5 and 6) enjoyed absolute subsidies per user (US$200 annually) that were about 50Z higher than those of poor households (up to stratum 2). The small proportion of high-consumption users enjoyed subsidies 2-3 times higher than those of low-consumption users, representing a large fraction of users. - 23 - Figure 4.1 S ~BSIDY PER CUSTOMER BY CONSUMPTION BLOCK BOGOTA suas0y-u8s Pen veAn PencETGs MCUSTOU M BC2~ 15 MfTHLY CONauMPflON NANGE resaoun - % of euonus - ls SUBSIDY PER CUSTMR SY SOCAL STRATUM suessY-uJss en Y<An sencTAGE OP CusIOMn (% 20 30 0 1 2 3 4 8 6 sanosa= c RCW« mgnu o a% et msmam9s W nRaao "Colombia-Energy Pricing Study, 1989, World Bank Report - 24 - 73. For a long time there have been inconsistencies in Colombian pricing policy on sources of household energy. For example, though kerosene is an important cooking fuel for low-income urban households living at lower altitudes, it enjoyed no subsidy comparable to that provided for electricity, LPG, and natural gas--all fuels that are, almost by definition, less accessible to the poor because of their high distribution costs. These pricing issues did not seem to find a place in the Bank's energy pricing dialogue with government. Outstanding Issues 74. Resolution 086, which was mentioned earlier, formed the basis for the treatmene of retail tariffs in the 1987 Power Sector Adjustment Loan. But though the Resolution marked a dramatic conceptual advance, as currently structured it seems unlikely to improve retail tariff structures and levels significantly or to allow appropriate generation of revenues during the next several years. 75. There are four main reasons why: - First, for residential consumption below 400 kWh/month the permissible discounts with respect to IC are so large that the average prices paid by households in most markets would remain between one third to one half of the reference IC. Since residential consumers account for the largest block of sales, it is virtually impossible for the average revenue per kWh to reach a level close to the overall IC even if the maximum levels of the resolution are applied. - Second, the proposed pace for adjusting tariffs to target levels appears too slow, given the scale of the financial crisis facing the power sector. - Third, the implications for fuel-substitution in the cooking fuel market are very negative if the tariffs for the second consumption block (200-400 kWhmonth) are set at the minimum level (50% of IC) enunciated in the Resolution. In that case, price signals would continue to encourage consumers to use electricity instead of, liquefied petroleum gas (at low and high altitudes) and kerosene (at low altitudes) for cooking, whereas from the standpoint of economic supply cost, the reverse consumption preference is desired. - Fourth, Resolution 086 does not provide guidance for setting tariff levels when an individual utility's financial costs well exceed the overall system IC. This is a very real problem currently facing EZEB. 76. The severe distortions in the retail tariff structure, which result from pricing electricity for residential uses much below--and that for commercial and industrial consumption much above--their respective costs of service, have become intimately linked to the problem of further increasing average unit revenues. Moreover, it needs to be recognised that though the estimated average IC for low- voltage (LV) residential users in the review period has been low by international standards, at around 6.5 US centslkWh in current terms, it has been much higher than the average residential tariff of about 2.5 US cents/kWh. - 25 - 77. Yet another problem that has remained in the review period, despite interconnection, has been the cost differences between the different power markets in Colombia. These arise due to differences in: network costs and load densities; supply costs due to differences in generation mix and dependence on bulk imports; and management efficiency. These differences will persist into the future. Similarly, tariff differences will persist between markets in the measure that costs are reflected in tariffs and that a financial transfer mechanism cannot be implemented within the sector. 78. Tariff 'unificationw has become a desirable goal provided it is interpreted in terms of uniformly applying principles relating tariffs to costs across the sector, and not equality of tariffs across regions. Yet, tariff differences cannot become or remain too large since this exacerbates regional political difficulties. This suggests that *acceptable, tariff differences between regions is one area of policy that now needs agreement. 79. There appears in OED's view, to be little basis for establishing financial transfer mchanisms across the sector between the utility companies serving the five major markets. The political concensus does not exist for the creation of such mechanisms and on economic grounds it is likely to reward inefficient management in some markets. The absence of such mechanisms implies that tariff levels in each market be set so that costs are covered in that market. To the degree that this cannot be fully achieved, either on equity grounds or because of market structure, then the onus would be on Government to transfer directly to the utility financial resources to cover such deficits. In this way, transparency is achieved regarding who pays how much, to whom, and for what. Other avenues could also be explored. For example, utilities continue to be exempted from income taxes and, as such, fiscal policy measures oriented toward creating a mechanism for cost compensation have never been assessed. This is an avenue of analysis that could yield benefits as an alternative to excessive reliance on inter- or intra-sector subsidies. Cost of Service and Bulk Tariffs 80. The bulk tariff between ISA and its shareholders may be regarded as the most important power price in Colombia because it directly affects nearly 40Z of the electricity used. Had the bulk tariff been closer to the average IC at the interconnection level, ISA's financial performance clearly would have been significantly better during the 1980s. For example, during the review period, the ISA bulk tariff varied between 50 and 60% of the average IC at the interconnection level. 81. Conflicts about the bulk tariff relate to ISA's structure, for two reasons: its shareholders presumably want a return on their investment and, if they are large electricity importers, they also want very low bulk tariffs. This applies particularly to ICEL which also possesses markets in which the share of residential sales is high. Indeed, at times it has seemed that the shareholders consider their position strengthened if the position of their company, ISA, is kept weak. 82. Traditionally, decisions about generation investment by companies have been driven more by the utilities' perceived desire to minimize imports of energy from other companies (in part because they consider their own generation more reliable * 26 - and less costly than imports) than by the market price of bulk energy traisfers. Rather than serving as a fundamental price signal, conveying economic information regarding desirable and undesirable generation investments and consumption patterns, during the review period, ISA's bulk tariff had essentially an accounting role, which was used to allow the company to generate funds only to cover its operating costs and debt service. Because the bulk tariff was so low, between 1977 and 1986 ISA's self-financing ratio after debt service was negligible (as described in Chapter V of this Volume). Tariff Estimates at Appraisal of Bank Operations 83. OED examined the tariff levels estimated ex ante by the Bank during its economic and financial appraisal of operations processed during the review period, and assessed how appropriate these estimates were. The key question asked was: how did the position adopted in SAR9 regarding future tariff levels compare with what was perceivable at the time that tariffs should have been? From a sectoral perspective, there were three distinct points at which such estimates could have been madet 1980181--the Guavio appraisal; 1983--the FEN appraisal; and 1986/87- -the Sector Adjustment Loan appraisal. See Table 4.2. Table 4.2: ESTIMATED INTERNAL RATES OF RETURN (IRR) Ex-Ante IRR in OED Ex-Ante Bank Report IRR 1. Guavio SAR (1981) Consolidat ecor Not presented 12.8% Investment 1980-88 2. EM 2AR (1984) consolidated etor Not presented 6.5% Investment 1983-87 (to achieve 10.3% tariffs needed to double 3. FEN SARA(194 EEBI t Not presented 4.4% Program 1983-87 (accounting for ISA's share in Guavio) 8.9% (excluding entire Guavio investment) 4. ) Lo Consolidated Sector 13.0% 9.3% Investment 1987-90 (after correcting for sector's unit cash Operating and malatenance costs) 84. In the case of the Guavio appraisal, OED estimated the ex ante internal rate of return (IRR) at 12.81 on the basis of the consolidated sector investment program for 1980-88 and the projected tariffs for the sector that were given in - 27 - the SAR. Even though the IRR was not presented in the SAR, what this illustrated was that the tariff levels proposed at the time seem to have been appropriate to ensure the economic and financial viability of the 1980-88 program. 85. OED reached a different conclusion, however, in the case of the FEN appraisal, which the Bank based on the 1983-87 sectoral investment program of some $6.4 billion (1983 prices). At the time of the FEN appraisal in 1983, conditions in the power sector differed radically from those of three years earlier. Energy losses in virtually all markets (except EPH) had increased significantly. The growth rate of energy sales had sharply declined. The Mesitas and Guavio projects had already experienced significant cost overruns and delays. The Latin American debt crisis had erupted, real interest rates were high and nzw foreign borrowing was difficult. In the period 1981-83 internation- al inflation collapsed, while that in Colombia continued at about 252 annually, resulting ir an increasingly overvalued peso. Somehow these radically different perspectives did not filter through the FEN SAR. OED estimates that, using the tariff levels set out in the FEN SAR for the overall sector, the ex ante IRR for the 1983-87 p:ogram was only 6.5Z.8 To yield a 10.3% return, the average unit revenue would have had to be nearly double what was presented in the SAR. 86. Somehow these negative ex-ante signals were not illuminated in the FEN SAR, with the result that appropriate tariff adiustments were not called for. However, the Bank, to its credit, financed only .he ongoing investments and not any of the new ones contained in the 1983-87 s ctoral plan. Subsequently (in 1985) it refused to participate in financing t' Urra project, which was one of the new investments contained in the 1983-87 1ectoral plan. It also secured a covenant, covering the whole sector, that limited new investments only to projects that were economically justified and whose financing would not affect the existing financial obligations of the companies. All in all, however, the FEN appraisal appears, in the judgment of OED, to have marked a shift both in the Bank's perception of the sector's problems and in the measures it considered that the authorities needed to take to bring an escalating financial problem under better control. 87. For the Power Sector Adjustment Loan of 1987, the President's Report estimated the IRR at 132 for the 1987-90 overall sector investment program. It assumed that the sector's average unit revenue would rise by 3Z annually between 1987 and 1992 in real dollar terms; that the investment program would be cut back by almost 602 in real terms with respect to the five-year investment tranche 8This low rate of return on the sectoral investment program was evident also for individual power companies' 1983-87 investments. For example, in the case of EEEB, OED evaluates the ex ante IRR for the 1983-87 investment program at only 4.4Z (with the tariffs proposed in the FEN SAR), even if ISA's share in the Guavio investment is accounted for. Indeed, even after excluding the entire Guavio investment, the ex ante IRR for such an eviscerated EKEB investment program remains a modest 8.9?. This again shows either that the 6.9 USo/kWh tariff (1983 prices) proposed for EEEB was too low or that a drastically reduced investment program, even without Guavio, was still too large for the company's projected electricity demand. - 28 - reviewed at the time of the FEN appraisals and that the sector's unit cash operating and maintenance costs would fall by some 6.52 annually in real dollar terms. 88. OED found that the evidence between 1971 and 1986 did not provide a basis for assuming that the sector's unit cash operating and maintenance (O&M) costs would decline in the 1987-92 period at the rate assumed in the President's Report. Indeed, the sharp decline (in dollar terms) that took place in these costs in 1984-86 was unique; it was partly due to the accelerated peso devaluations in 1985186 and partly to the dramatic drop in fuel usage after interconnection of the CORELCA system in 1984. 89. Using the same 3Z rate of real increase for unit revenue for the sector as in the President's Report, but a still-optimistic 1.5? annual decline in unit cash O&M costs, OED re-calculated the ex ante IRR at 9.31. To increase the return to about 13? would have required an average unit sector revenue of 5.5 US cents/kWh at 1987 prices over the entire 1987-92 period--or about 302 above the level that was proposed in the President's Report. This indicates that, again, the level of retail tariffs for the overall sector assumed at appraisal was inadequate. Bank Performance on Tariff Issues 90. During almost the entire review period, the Bank viewed tariff policy largely from a financial perspective. In the mid-1970s, in connection with the preparation of the San Carlos I operation, it raised the issue of pricing according to economic cost, but acceptance of the principle was slow. 91. After OED's 1972 report had pointed to the pervasive subsidies to electricity consumers, the Bank agreed with the sector, in the context of the San Carlos I operation, on the preparation of ntudies that should have illuminated the costs of these subsidies, and the questions of who subsidized whom, in what markets, and by how much. These efforts obviously fell short of providing the necessary insights, another casualty of the absence of an overall sector view by the Colombian authorities and the Bank. Thus, the Bank did not in the 1980s take into account, in its definition of the "appropriatew financial level of tariffs, the repercussions on public sector finances of major financial deficits in the power sector. 92. However, the Bank can consider itself instrumental in the major changes in tariff policies of the mid-1980s, which led to the conceptual advance of linking retail tariffs to economic costs, albeit with high discounts. However, from both the economic and financial points of view, retail tariff structures and levels are still highly inadequate, in particular in EEMB. 93. The Bank appears to have shared the Colombian view of the ISA bulk tariff merely as an accounting tool, and not as an economic signal and a prime lever for retail tariffs across the country. This view is responsible for ISA's consistent failure to generate enough internal resources. As explained above, however, financial strength was not a goal of ISA's shareholders, which raises - 29 - the issue of the compatibility between ISA's financial well-being, its shareholders objectives and the future role of the Bank in supporting such arrangements. 94. The Sector Adjustment Loan of 1987 supported the important advance in tariff policy constituted by Resolution 086, discussed earlier in this chapter. However, it called for a level of tariffs about 30% lover than the ex ante requirement that OED estimates to have been necessary. This emerged four years after the inappropriate tariff adjustments called for at the time of the PEN appraisal. Such playing down of the role of tariff levels in the sector's problems has given the incorrect impression that problems of internal resource mobilization were less severe than they actually were (and had long been) and that adjustment in the sector could proceed at the leisurely pace envisaged. 95. Although the Bank maintained an intense dialogue with the Colombian authorities on the issue of tariffs throughout the period, government took substantive action to raise tariffs only after the informal suspension of disbursements on Bank loans to power companies in the mid-1970s. OED understands that from 1983 onward--as the technical, financial, and tariff performance indices of power companies and the sector deteriorated again, the Bank's senior management again considered suspending disbursements and delaying new loans to the sector. The then-prevailing opinion favored persuasion over coercion, although the Bank did indicate (in OED's view justifiably) that it would not then finance the Urra project. 96. The Bank's willingness to take a tougher position on the sector's tariff and financial issues in the mid-1970s, and its reluctance to do so in the mid- 1980s, has to be seen in the context of macroeconomic developments. In the mid- 1970s the macroeconomic situation was generally sound and growth prospects strong. By 1984185, however, Colombia was in the throws of macroeconomic adjustment, involving significant peso devaluations, trade liberalization, and restrictions on public sector investments. In such a setting, macroeconomic concerns prevailed over those of a sectoral character. The fight against inflation took priority over raising electricity tariffs (although the two bore little linkage), and the enforcement of sector financial covenants was relaxed. However, this raises the broader problem of ensuring consistency in the Bank's approach to macroeconomic adjustment measures and those in individual sectors of the economy which transcend the scope of this study (see Chapter VII of this volume). Recommendations 97. The past seventeen years of Bank support for the Colombian power sector provide lessons for the institution's potential further involvement in the sector. These lessons derive from the above discussion of the shortcomings of Colombian tariff policy and of the Bank's performance in assessing the appropriateness of tariff levels: - The Bank should make a broad sectoral approach to tariffs the center of its involvement, to improve tariff levels, tariff structure, and resource mobilization. * 30 - - It should continue to put its full weight behind goverament to help the tariff regulatory agency, at present JNT, to make further tariff adjust- ments, to better reflect the economic costs of supply and effectively recognize that the bulk tariff is a key factor in the pricing of energy. - It should insist on substantial action toward economic pricing of electricity before it submits any new operation to the Board for approval. - It needs to assess systematically the degree to which tariff levels set according to long-run average IC principles cover the full financial costs of the sector, and individual utilities. - It should emphasize, within its own organization and in Colombia, the role of ISA's bulk tariff as a key factor in the pricing of electricity in Colombia and accordingly, in connection with future lending, should consistently insist on action toward the economic pricing of bulk electricity. V. FINANCES Introduction 98. During the period under review, the sector's financing and financial performance were never fully satisfactory, and from the early-1980s were poor. This chapter discusses the financial performance of the sector as a whole and of the main individual utilities that had direct relations with the Bank, (EEES, EPM, and ISA), with regard to self-financing, return on assets, debt service coverage, development of debt and equity, and accounts receivable. It also analyzes the role that the Bank played in support of improvements in the sector and in utility finances. 99. OED encountered difficulties in establishing a clear picture of the sector's financial development in the review period. It was able to assemble consolidated financial statements for the entire sector only for the years 1976, 1980, and 1983-86. These data are to varying degrees incomplete and tentative, but nonetheless telling, and they form the basis for the flow of funds shown in Table 5.1.9 These results suggest that during the 1976-86 period, from the 9OED also prepared a flow of funds statement based on a consolidation of the audited financial statements of four companies in the sector--EEEB, SPM, ISA, and ICEL. In 1986 these four entities represented about two-thirds of the sector's long-term assets. The companies not included, due to resource constraints, were CORELCA Empresa (which excludes CORELCA's seven sub- sidiaries), Betania, CVC, and ENCALI. The findings of this partial consolida- tion are discussed in Chapter VI of the Technical Report (Volume II). - 31 - sector standpoint, its revenues did not fully cover debt-service costs in 1980 and then after the mid 1980s and the contribution of its revenues to investments was marginal, becoming negative after 1985. All in all, a very disappointing performance in light of the level of investments over the period. Table 5.1: POWER SECTOR: SOURCES AND APPLICATIONS OF FUNDS (percent of total resources mobilized) sucs1976 1980 1983 1984 1985 1986 Net Income Before Depreciation & Debt Service 31 30 27 32 33 39 Borrowing- 54 48 55 61 61 56 Other Sources 15 22 18 7 6 5 (largely Gov't) Total Wm W W W WWo i Applications Debt Service 28 31 23 27 36 41 Investment 59 71 76 76 71 56 Adjustments in Working Capital 1 -2 1 -3 h7 3 Total AW W W W WWT N 1 Nemo Items: Sector Self-Financina Ratio (Net Internal 5% -1M 51 78 -48 -4 Cash Generation after Debt Service to Investment) Index of Total Applica- 100 250 428 475 477 339 tions in Constant 1976 Col$ (1976*100) Self-Financing Ratio 100. Table 5.2 shows the average self-financing ratio (net internal cash generation after debt service in terms of total annual investment) for the three main Bank borrowers in the sector. Table 5.2: SELF-FINANCING RATIOS IN BEEB, EPM, AND ISA (percent) 1972 1974 1976 1978 1980 1982 1984 1986 EEEB 40 27 43 54 21 9 5 6 EPM - - 40 24 89 42 24 9 ISA - - -12 9 6 9 -3 -6 - 32 - 101. EEB's cash operating costs (calculated in terms of kWh sold and constant prices) more than doubled during the period 1971-86. This increase partly reflects the rising cost of energy purchases and partly the company's failure to improve its efficiency, as reflected--especially in the 1980s--by the increase in energy losses. 102. Since EEB's investments remained at a modest level during the 1970s, and because from 1976 onward its tariffs increased substantially in real terms, the utility's contribution to investment generally remeined higher than 30Z and peaked at more than 50% in the 1976-78 period. Later, after the large investments in the Mesitas hydroplant had begun, BEB's contribution to investment deteriorated to 16% in 1981. From 1982 on, the even larger expenditures on the Guavio facilities reduced the self-financing ratio even further, although tariffs continued to rise in real terms until 1985. At this stage, the failure to allow the utilities to recoup the effects of the massive devaluation of the peso pushed EEEB's self-financing ratio down to an utterly unsatisfactory 62, where it has essentially remained since then. 103. During the early 19709, EPM's unit cash operating costs were similar to EEES's, but in later years they remained roughly constant, rather than increasing. EPM held down its costs by supplying relatively low-cost energy from its own plants, and was even able to export surplus energy to the interconnected grid. It also kept its losses at about a constant, albkit relatively high, level (18-20% of energy sent out). 104. In the early 1970s, when SPM was not undertaking major construction work, its self-financing ratio was high (702 in 1971), but the construction of Guatape II in the mid-1970s, combined with declining tariffs in real terms, led to a deterioration (282 in 1975). In 1980-81, after the tariff improvements in connection with the preparation of the San Carlos I project, the self-financing ratio peaked at more than 80%, while EPH was not yet overly burdened by the new investments in the Guadalupe IV and the Playas hydro projects. These projects, together with the start of the Rio Grande II plant in 1984 and the effects of the devaluation mentioned above, reduced EPM's contribution to investment to a poor 9Z by 1986. 105. Until ISA commissioned its first plant, Chivor, in 1977, the company's revenues were limited to those from transfer of electricity through the interconnection. They were complemented by contributions from the utility's shareholders to ISA's operations and investment. As discussed below in this Chapter, the rates that ISA was allowed to charge for its supply and services were intended solely to cover operating costs and debt service, which they barely did from the late 1970s to 1982. After that, they fell substantially short of the goal. Therefore, ISA's contribution to investment was negligible (as planned) until 1982 and negative from then on. 106. Of the three utilities whose finances OED did not analyze in detail in the context of the present review, CYC achieved results comparable to those of EM. ICEL and CORELCA have more decentralized operations through weakly managed - 33 - subsidiaries largely serving rural areas. By nature, these subsidiaries are less able thAn the large municipal utilities to contribute a high percentage to investment under the tariff policies that have prevailed in Colombia during the past two decades.10 Return on Assets (ROA) 107. The return on net fixed revalued assets in operation (ROA) was the overall resource mobilization indicator convenanted by the Bank in all its operations in the review period. EMR_ never achieved the covenant level of the ROA during 1978-84, even though no large assets were commissioned and thereby added to the asset base in this period. The company's corresponding shortfall in cash amounted to some US$90 million (1986 prices), or about one-half the cost overrun of the Mesitas project, which had to receive supplementary financing through the FEN loan from the Bank. By 1985186, REEB's return on assets had risen to around 15?, but this increase permitted only a r.rginal 5-6? contribu- tion to its investment program from its own resources. The reasons were the company's high debt service costs and the still considerable size of its investment program. 108. EM's ROA remained below covenant levels through most of the review period. Nevertheless, the company's cash flow provided reasonable debt service coverage and self-financing (except in 1986). 109. ISA was the only company in which the ROA was generally close to the level set by the Bank in its covenants. Indeed, the covenant figure of 9? ROA merely reflected ISA's bylaws, which required the company's revenues only to cover its debt service and operating costs. This is perhaps a good illustration of the Bank dealing pragmatically with a difficult situation, by nurturing a new institution it was instrumental in establishing but whose setup reflected a reluctant compromise on the part of the regional and municipal power entities. Debt Service Coverage Ratio (DS) 110. From 1981 onward, with the economic downturn, the dbt service coverage ratio deteriorated significantly for EEES and for ISA. In RPM again the decline was less dramatic in RPM than in the other two companies. In ISA's case, the indicator fell to around 1.0 in 1983 and 1984, after which exclusion of interest payments on ISA bonds held by its shareholders was required to avoid the ratio decreasing below 1.0. 10For example, in the case of CORELCA Empress (which excludes its sev!n subsidiaries), the self-financing ratio varied between -25% and -40? for the period 1980 to 1986, even after including government subsidies on natural gas. - 34 - Debt and Equity (DIE) 111. Low self-financing implies large borrowings and equity contributions. Since the latter were limited in Colombia, the ratio of debt to equity (DIE) could be expected to increase. This indicator, bowever, is not covenanted in Bank power operations, since the DS ratio is the preferred debt limitation covenant in the case of state-owned power companies enjoying a strategic monopoly position in a closed market in which government exercises full price control. On the basis of the utilities' audited accounts (which reflect the partial revaluation of fixed assets corresponding to the revaluation only of outstanding foreign liabilities at the prevailing exchange rate), the debt-equity ratios of EE3B and ISA increased significantly in the review period; in 3PM the increase was quite modest. However, the debt-equity ratios of 3318 and ISA as estimated from a pro forma full revaluation of assets (presented in Bank SARs), were about 1.5 in 1986. Accounts Receivable (AR) 112. Table 5.3 sets forth for the Bank's main borrowers in the sector, the partly unsatisfactory development of accounts receivable. It shows EEEB's deteriorating commercial performance, and EPM's temporary weakness in the early 1980s, followed by a strong recuperation to a satisfactory level by the mid- 1980s despite the difficult macroeconomic circumstances. ISA's poor record essentially reflects the shareholders' policy of Improving their cash flow by not paying ISA for its supply and services. Table 5.3s ACCOUNTS RECEIVABLE, IN MONTHS OF BILLINGS EEB EPM ISA 1971 2.0 1.3 - 1980 4.8 2.6 7.4 1986 4.8 1.5 11.3 The Bank's Performance 113. The main financial benefit to Colombia from the Bank's involvement in the Colombian power sector has been the mobilization of resources in amounts which could not have been achieved otherwise. Indeed, in the review period, the Bank contributed about US$1.9 billion in 15 loans to the sector. The Bank's performance regarding the financial aspects of this lending, five areas warrant discussions the lack of a sector-wide financial view; the handling of changing maturity, grace period, and repayment term-; the lack of focus on the foreign exchange risk; the Bank's weakness in setting and enforcing satisfactory levels of financial performance for the sector and some of its borrowers; and inconsistency in the use of financial indicators. - 35 - Absence of a Sectoral Approach 114. During the first half of the review period one of the major constraints both the Bank and the Colombian government faced was the difficulty of forming a sector-wide perspective, especially of financial performance and development. The absence of such a sectoral view was a product of the fragmentation of the sector itself, largely an outgrowth of fierce Colombian regionalism. It prevented government from properly assuming its policy-setting and regulatory role, and also limited the Bank's ability to plan its operations within an overall financial framework. 115. The preparation of the first consolidated sectoral financial statements in connection with the Guavio appraisal in 1980 was a landmark, both technically and politically, in that it required a joint effort by the utilities toward a common goal. Later, consolidated statements were prepared in 1983 in connection with the appraisal of the FEN operation, which was carried out expecting that FEN would become the institution that would develop a reliable sector-wide financial view. By 1986-87, at the time of the Power Sector Adjustment Loan, another sectoral financial consolidation was done which included the historical years 1984-85, and projections for 1986-92. The apparent lack of consolidated sectoral financial statements other than for part of the post-1980 period was somewhat surprising to OED, given the level of sophistication of the country and its power sector, and the nearly forty years of Bank involvement in the sector. Bank Lending Terms 116. At the beginning of the review period Bank loans were quite advantageous to IBRD power sector borrowers in terms of maturity, grace periods, and provisions on debt service payments (in the form of equal semi-annual sums for interest and principal).11 In 1976, however, in the wake of the first oil price crisis and in an effort to recycle its resources more rapidly, the Bank hardened its loan terms, which became defined on a country and no longer on a project basis. This meant that, in the case of Colombia, even for such long-lead time projects as hydro-plants, the maximum loan maturity was 17 years, with a grace period of 4 years, while simultaneously repayment terms moved to a Olevelized, principal repayment schedule. Under these conditions, for hydro projects like those in Colombia, principal repayments jumped to a level some three times higher than that of depreciation charges, while loan amortization, on average, started 15 months before the plant began operation and 23 months before the Bank made the last disbursement on the loan. 117. Although these changed conditions applied to all IBRD borrowers, the effect on hydro projects was especially marked. Some comparisons on the basis 11The 1970 and 1972 Chivor I and Guatape II loans were, respectively, for 30 and 25 years and with grace periods of 7 and 5 years--reasonably close to the construction times required for these hydro projects. The mismatch between principal repayments and depreciation charges (the latter based on a 40-year economic life) was modest and did not call for special measures. - 36 - of the average loan life are instructive. For example, in the Mesitas loan, the average loan life would have been 8.5 years under constant annual payments of interest and principal but 7.7 years under levelized principal repayments; the actual loan life turned out to be around 6.7 years, as shown in Table 5.4. The impact of the shortening of grace and repayment periods was even more dramatic. For example, whereas loan lives for the Chivor I and Guatape II loans were 16 years and more at appraisal and, despite delays, resulted in actual lives longer than 14 years, four loans made between 1978 and 1981 had loan lives of only 7.7 to 8.0 years at appraisal and actual lives of only 3.2 to 6.7 years. 118. The maximum loan balance available to the borrower is another way of illustrating the effects of the changes in loan terms (see Table 5.4). The Chivor I and Guatape II loans were expected to reach a 1002 balance for a short time, and nearly did so. While the loan balances of the post-1976 operations referred to above were expected to reach 882 to 962, they actually ranged only from 31% to 801.12 Fluctuating exchange rates may further worsen the outcome for all the loans discussed here. Table 5.4: AVERAGE LOAN LIVES AND MAXIMUM LOAN BALANCE Loan No. 681 874 1628 1668 1953 Borrower ISA EPH EEEB EPN EPH Project Chivor I Guatape II Nesitas Guadalupe IV Pla Year of Loan 1970 1972 1978 1980 191 Loan Amount 52.3 56.0 84.0 125.0 85.0 Maturity (Years) 30 25 17 17 17 Grace Period (Years) 7 5 4 4 4 Loan Life (Years) Projected 19.0 16.0 7.7 8.0 7.8 Actual 18.8 14.3 6.7 5.4 3.2 Maximum Loan Balance in Terms of Loan Amount Projected 100% 100% 96% 95% 88% Actual 99% 94% 80% 58 31% 119. In its appraisal and other reports the Bank gave little weight to the financial consequences of the 1976 changes in its IBRD lending terms, although these were significant for power investment programs heavily weighted with long- lead hydroelectric projects, such as those in Colombia. In fact, the President's Report for the 1987 Colombia Power Sector Adjustment Loan seems to have been the first official Bank document to highlight the changes in Bank loan maturities and grace periods as contributing to the sector's financial problems. That 12The Playas operation was an extreme case. In this instance, the shortening of the grace and repayment periods, long delays, and the fact that a project change reduced costs and led to disbursements amounting to only about half of the loan amount reduced loan life to about 3.2 years and the maximum loan balance to a mere 312 of -.he loan amount. - 37 - report, however, played down the significance of the impact, which took the form of substantially increased debt service requirements. 120. The much shorter loan lives since 1976 associated with Bank loans for hydro power development in countries like Colombia, necessitates that the financing of such projects rely much more heavily on more permanent sources of funds, including internally generated resources and, where feasible, equity rather than on the massive debt financing that has characterized the Colombian power sector over most of the review period. However, the Bank, in its dialogue with the government and sector authorities, seems to have foregone this quite potent argument for adjusting revenues. With the benefit of hindsight one can also note the potential for marshalling equity funding from the private sector, as an alternative to scarce public funds. Of course, it must be stressed that, even though the IBRD's lending terms tightened in 1976, the institution remained one of two sources of long-term finance on a reasonably significant scale. Handling Exchange Risk 121. In the late 1970s and early 1980s, the real value of the US dollar expressed in Colombian pesos increased by about one third. The increase was not reflected in the actual exchange rate, however, and the exchange risk borne by the utilities built up considerably during the period. This risk materialized in the mid-1980s when a major devaluation of the Colombian peso against the US dollar occurred. This devaluation was one of the measures government used to shift the emphasis of production away from nontradables (such as power), which consume large amounts of foreign resources aud from which the benefits accrue only slowly, to tradables. The devaluation substantially raised the power companies' costs, especially debt servicing. For example, in 1985 the exchange losses consumed 45Z and 43? of ISA's and EEEB's operating margins, respectively. Government did not take the next step, however, which would have further supported the shift in the production structure and provided a basis for a better financial performance of the sector: to allow the power companies to pass on the additional cost to the consumer. 122. Although the exchange risk had consistently been underestimated in the past and the risk of exchange losses had been growing since the mid-1970s, Bank appraisal reports of the early and mid-1980s did not discuss the issue. This may have been because the Bank assumed--if so, against experience--that materialization of the risk would be offset by higher tariffs. The issue was (and still is) complicated by inconsistencies in the presentation of exchange losses on the books of various Bank borrowers as well as in the external auditors' discussion of this item. The Bank did not address this problem, either. 123. Another source of exchange risk manifested itself in 1986, when the US dollar depreciated relative to the yen, deutsche mark, and Swiss franc. The exchange loss and risk situation has been aggravated by cross-currency fluctuations on Bank loans disbursed under the currency-pool system. Loans from the Bank's currency pooling system had yielded benefits for borrowers while the - 38 - dollar was appreciating against these currencies (1980/81 to 1985), but this benefit became a cost for currency-pooled loans compared with nonpooled loans when the dollar depreciation started in 1986. Financial Performance Standards 124. For operations in the Colombian power sector, since the late 1970s the Bank has progressively lowered its standards of financial performance --whether those required for Board presentation of new operations, those that were set forth in covenants, or those it tolerated without suspending disburse- ments or refusing to proceed with further lending. One of the reasons for the Bank's increased tolerance of lower standards of financial performance in the power sector during the mid-1980s was the difficult macroeconomic situation the Colombian authorities were seeking to manage (with Bank support). In this setting the larger macro concerns were focussed on, and deterioration in the power sector was tolerated until macroeconomic adjustment was achieved. But the Bank has also persistently made optimistic assumptions about the future, even though experience and the situation at the time of appraisal did not warrant them. 125. The easing of the conditions of Board presentation is evidenced by the fact that conditions, which in earlier operations were typically attached to Board presentation and tended to delay the latter, became conditions of effectiveness (often delaying the latter) or conditions to be fulfilled during project implementation. Accordingly, the average time that elapsed between loan approval and loan effectiveness, which had been 6 months for loans approved before 1978, increased to 16 months for later loans. 126. The lowering of performance standards specified in covenants and the tolerance of inadequate performance was unevenly distributed among companies. It was particularly pronounced in the case of EEEB, and less so in the case of EPM. For ISA, the performance standards remained low during the entire review period, essentially in accordance with ISA's statutes. This uneven treatment of individual borrowers is a further sign of the lack of a sectoral strategy. Instead of pulling the sector together, the Bank's attitude tended to lead the individual entities even farther apart. 127. In several instances the Bank appeared lax in its treatment of deterior- ating financial performance. For example, EEEB's return on net fixed revalued assets in service in the period 1978-84 fell significantly below the level specified in covenants, and the company conspicuously failed to achieve the covenant level of its self-financing ratio during 1981_86.13 A second instance, which began with EEEB and then was extended to all borrowers in the sector, involved relaxing the definition of the self-financing ratio. This was achieved by introducing into the numerator noncash items, such as increases in employees' pension fund reserves--a practice that has become quite common in Bank projects. 13This level had been set at 35% between 1982-85 and at 551 thereafter; however, the company never achieved more than 10% in the years 1982-86. - 39 - In the case of ZEB, this practice tended to reduce the shortfall from the self- financing ratio level in the covenants, though it did not eliminate it. From a cash management standpoint such inclusion may be correct, but it appears inappropriate to include increases in employee pension fund reserves as a resource to cover debt service or Investments in fixed assets, because creditors have no claims on this resource.14 128. The Bank also seems to have been excessively flexible in failing to bind EPM's self-financing ratio by covenant, even at a time when the company was building two hydro projects. The financing plan for EPM's 1980-86 investments envisaged that net internal cash generation would contribute as much as 601 of the financing required, and self-financing was thus a vital resource. Instead of making it subject to covenant, the Bank imposed a condition of loan effectiveness that called for balancing the financing plan for the 1980-82 period--a poor substitute, under the circumstances. 129. By contrast, the covenant level of the ROA indicator for ISA in the San Carlos I loan is a good example of the Bank dealing pragmatically with a difficult situation in helping a new institution through its early years of existence. In this instance, it was known at the outset that the covenant level of 9% for the ROA was too low and that ISA was very likely to have difficulty covering its debt service with that level of resource mobilization. However, what was positive in this case was that this was candidly stated in the SAR. The Bank's Use of Financial Indicators 130. Financial indicators are the tools by which the financial performance of entities can be judged. The findings of this review suggest that it would be prudent for the Bank to re-examine its guidelines for financial covenants in power operations, especially with regard to three areas: - the case for including works-in-construction in the asset base of the ROA indicator; - the problem of fixed asset revaluation when done on a pro forma basis; and - the desirability of including in covenants the self-financing ratio as a complement to the ROA indicator, especially where major investments with long gestation periods are contemplated. 14As long as such noncash items did not exceed 5-7Z of gross cash generation, as estimated in the Mesitas SAR, their inclusion was unlikely to lead to misunderstandings. However, in the case of REEB, in the early 1980s, such noncash items turned out to be 20-30% of gross cash generation. There- fore, their use had become an issue and should have been addressed. - 40 - 131. Traditionally, the Bank, in its power sector lending operations has given prominence to the rate of return on assets (ROA), specifically net fixed revalued assets in operation. This indicator is essentially designed to provide, when compared with the average (or better the marginal) cost of the funds used to finance the assets, a measure for the efficiency of a company in using its operating assets. However, in practice, in the context of power sector lending, the Bank uses the ROA to determine the company's gross resource mobilization capacity to generate enough funds to meet costs--operating costs, debt service, and, above all, to contribute to its investment program. Clearly, when the objective is to monitor the efficiency of fixed asset use in operation, works- in-construction should not be included in the fixed assets base. However, since the Bank uses the ROA as an indicator of the resource mobilization capacity, it would seem to have merit to include plant-in-construction in the fixed asset base, as this might help support the generation of higher revenues at an earlier stage as needed in this type of situation. 132. A further implication of the present reliance on a return on assets in operation became evident in the Colombian power sector in the 1980s. As long implementation periods have led to debt service becoming due long before (in the case of Bank loans on average 15 months before) the assets financed were commissioned, total debt service quickly absorbed most of the internal cash generation. Therefore, even notionally reasonable returns on revalued assets in service (for example the case of ISA) did not generate any contribution to investment. The above suggests that the Bank re-examine the type of rate of return on assets covenant it should use in its power lending operations, especially in light of the much shortened loan life and of the lumpy, long-lead time character of hydro investments. The case for including works-in-construc- tion in the asset base becomes quite strong in the case of fast growing small and medium sized power companies, where assets in construction often represent a very large part of total fixed assets. 133. Another problem the Bank has encountered in Colombia when using the concept of return on net fixed revalued assets in operation concerns the revaluation process itself. Colombian law and accounting practice only allow partial revaluation of outstanding foreign liabilities according to the prevailing peso exchange rate. This has meant that the full revaluation of fixed assets in service was done purely on a pro-forma basis for the determination of the asset base used in the key resource mobilization indicator of the Bank. One of the problems that arose in regard to MEB was the inaccuracy of the full revaluation process which became apparent only some years after the event. Traditionally, the pro-forma full revaluation of fixed assets has been performed on a quarterly basis, starting from a 1976 fixed asset base. However, in fulfillment of a condition of the Mesitas loan, BEEB commissioned a study to establish the value of these assets. This study, which was completed in mid- 1984, three years later than originally envisaged, showed that, as of the end of 1982, EEEB's revalued total assets were some 40% higher than the previous routine revaluations had suggested. This meant that, since 1976, EEEB's key rate of return had been increasingly overstated by amounts unknown at that point in time. Of course, this resulted in seriously compromising the effectiveness of the Bank's most important financial indicator. * 41 - 134. Finally, it was only in the case of EDEB that the Bank simultaneously made both the self-financing ratio and the ROA subject to covenant. Of course, this did not prevent poor financial performance. But in OED's view there is some merit in the Bank including both of these indicators in covenants in cases of large, long-gestation investment programs in the power sector. This especially arises when full asset revaluation is not part of the law and practice in the country in question, and the Bank has to resort to a pro-forma revaluation. Recommendations 135. During most of the period under review, Government contributions to the sector were far beyond those economically and socioeconomically justified. This was mostly because low tariffs kept internal cash generation inadequate, but in part because of unrealistic financing plans and inadequate operational performance. 136. It seems evident that the sector's finances must be turned around. In OED's view, this will need to be done on the basis of a new broad consensus about the sector's institutional structure, the respective roles of government (at central, regional, and local level) and the utilities, and the objectives to be pursued. Once basic principles are agreed, it is hoped that their application will include a financial component that will allow the utilities to recover financially and to operate in a transparent way, within an economically and financially reasonably efficient sector. 137. A financially sound sector starts with sound financing plans and with firmly committed resources from all sourcess (1) revenues and their collection must improve (through higher tariffs and lower accounts-receivable); (2) costs should be limited (through more efficient operation, in particular reduced losses); and (3) capitalization must be re-established at the proper level (through higher equity financing) to allow the sector again to raise loans in the capital market and thus make it less dependent on IDB and Bank support. To this effect, there is a need to establish, first in broad terms and then in more detail: - what the consumer will ultimately have to pay for, what government will subsidize, and through which channels government will provide the external subsidy; - which cross-subsidies among tariff categories will be implemented; - what utility revenues will have to cover: cash operating costs, debt service, proper adjustment to working capital, a reasonable contribution to investment, and an equitable remuneration for equity; - why accounts receivable in individual entities are high and how to remedy this shortcoming; and - what amounts of equity are needed, and how soon. * 42 - 138. Starting from the above--and taking into account the regulatory framework, which also should be substantially strengthened--government, the sector, IDB, and the Bank should chart a course to reach the agreed financial targets, setting precise, coherent, and achievable intermediary goals for the sector and all the utilities. 139. An important early objective should be a reasonably uniform presentation of the utilities' accounts, so as to permit a reliable consolidation that would provide a clear view of sectoral finances. From the Bank's standpoint, as part of a broader effort to improve power sector borrowers' financial monitoring, performance under financial covenants should become subject to audit as part of the external auditors' report on the annual financial statements of Bank borrowers. Finally, the Bank should also review its guidelines regarding the financial indicators it uses to monitor, and bind by covenant, financial performance in the power sector. VI. SECTOR INVESTMENT 140. Some of the main issues related to the sector's investment program today were already prominent in OED's 1972 reviews - the investment decision process and its relation to planning; - time and cost overruns; and - the imbalance between investment in generation and transmission, on the one hand, and in distribution, on the other. Investment Planning and Decision Process 141. As noted in Chapter II, throughout the past two decades, the shareholders of ISA exercised a strong influence on ISA's planning. They introduced essentially political input into the decision aking process at an early stage, thus hampering the definition of a development plan, which either was clearly identifiable as the least cost solution or whose deviation from this solution was known. In the early 1970s, this drawback was compounded by the limitation of the availability of reasonably prepared projects that led essentially to the implementation of the projects that happened to be at hand (Chivor I and II, and Guatape II). From the mid-1970s on, other factors played an increasing roles the desire of the individual utilities to have as much as possible of the generation facilities needed to meet the demand in their own supply area under their own control; the continued high rate of growth of demand projected; the delays incurred in implementation, which increased the threat of rationing; and the prospects of supply shortages in the early 1980s. 142. These factors, together with a persisting shortage of sufficiently prepared projects, again led to the implementation of the projects which happened to be at hand. These evidently included the developments which ISA's sharehold- ers preferred and which were not necessarily those that would have come out of an exhaustive least-cost analysis based on national economic parameters. This - 43 - situation persisted until the mid-1980s, when the downturn in demand growth provided a respite. Since the end of the review period, some steps have been made which may address part of this problem.15 It is unlikely, on the other hand, that a stricter analysis would have led to a substantially different order of implementation of the hydroplants. However, OED concludes, on the basis of admittedly rough calculations, but without using hindsight unfairly, that in the late 1970s the case for an almost exclusively hydro development was already debatable, even under strict application of the least-cost principle. This became especially so in the early 1980s, when the sector decided to construct the Guavio hydroplant. Indeed, the risks associated with the construction of large underground facilities in the Eastern Cordillera were known at the time from the implementation of the Chivor plant and from the driving of the Chingaza tunnel, both essentially in the same area as Guavio. Moreover, such lumpy additions to the system always imply a high rigidity of the investment program. 143. Already at the time of the decision on Guavio, an alternative development could have foreseen the large hydroplants meeting only the demand that would be exceeded in any reasonably foreseeable scenario. Once the development of demand would have become more definite, the difference between demand and the capacity available with the large hydroplant could have been covered by installing facilities with a shorter lead time, inter alia thermal plants using non- exportable coal or gas. Multi-Objective Planning 144. A development like the one sketched above might not have been part of the strict least-cost solution and, therefore, would have been in contradiction to the Bank's policy at the time. But taking into account the reduced engineering and, hence, financial risks and the lesser difficulties to mobilize financing, it could well have proven to be more desirable than the exclusive hydro development. This fact, together with the conclusions from the financial analysis suggests that, for the future, incorporating multi-objective analysis in its approach to defining investment in generation and transmission could reinforce the sector's planning capability. Which tool to choose for such planning is less important than the principle. Indeed, its essence is dealing at the same level with various development aspects (e.g. economics, finance, and environment) and the associated risks, with the focus of the analysis resting on the trade-offs between the aspects considered. The Energy and Industry Department of the Bank is currently supporting research to look at effective ways (including multi-objective planning) of introducing risk and uncertainty into power system planning and thereby relieve the dependence on least cost planning based on deterministic planning models. 151n particular, the shareholder pressures referred to above may be reduced as a result of new arrangements established between ISA and its shareholders on the standardization of all studies of hydro and thermal projects. - 44 - Implementation Time 145. In its 1972 study, OED found that most power sector projects, and especially hydroelectric plants, suffered delays up to three years. Such delays were also typical in the 1970s and 1980s. As can be expected, the delays were shorter for transmission and thermal generation (the projects with shorter construction times) and longer for hydro plants. 146. The 15 hydro projects reviewed for this study were in general more complex, with longer planned lead times than earlier projects. Their average delay between the decision to implement and their commissioning was about two years (one year before the start of construction and another during construc- tion). Preconstruction delays were mainly caused by major design changes becoming necessary after the decision to implement; institutional issues (who builds? who owns? who partcicipates?); and protracted procurement. The construction delays, in turn, were mostly related to geological difficulties, especially in tunneling; resettlement issues; lack of local currency funds; and a deliberate slowdown after the 1982-83 sharp reduction in demand growth. Project Costs 147. In the 1972 review OED found that substantial cost overruns were pervasive but were tending to decline. The decline continued in the 1970s and 1980s. 148. For the Bank projects completed between 1971 and 1987, plus Guavio, to be completed in 1993, in five out of nine cases the actual cost (in constant prices of the year of the estimate) was within a 15% range of the estimated base cost plus physical contingencies. In two cases (Guadalupe IV and Playas) the actual base cost was lower because of major project changes after project appraisal. In the two others (Mesitas and Guavio), the actual base cost was 50% and 902 higher, respectively, than the estimate because of the difficulties encountered (inter alia, those related to the geology of the sites). 149. There has also been an improvement in the Bank's estimates of price contingencies. Much of the vast difference between estimated and actual costs in current terms is related to the cumulative effect of delays and inflation. This further suggests that, unless time scheduling improves, substantial cost overruns are likely to persist. Balancing the Investment Program 150. The 1972 OED review pointed out the generally inadequate investment in subtransmission and distribution in the 1950s and 1960s. In the period reviewed here, the Bank persistently stressed that this imbalance in the investment program continued, and from 1980 on it directly addressed the issue in EEEB and CORELCA through specific distribution and village electrification operations. - 45 - The financial crisis in which the sector found itself later in the 1980s, however, forced it to concentrate on completing projects already in progress, which were mostly in generation, leaving few resources for distribution investment. 151. Thus the imbalance in investment has not diminished. Indeed, the few reliable figures available indicate that, in general, investment in subtransmis- sion and distribution remained substantially below 25% of total investment. The imbalance of course did not help the reduction of losses, especially technical losses. Nevertheless, the new role of ISA as a general coordinator in the field may help the sector to inch toward standardization and optimization of distribution investment. Recommendations 152. The sector urgently needs a more balanced overall investment program with a proper share for subtransmission and distribution. It will need to address the shortcomings of planning and decision making in generation and transmission through: - institutional measures (see Chapter II); - considering a broader range of alternatives when assessing possible development plans; - requiring more uniform preparation of projects at the selection stage; - more realistic time and cost scheduling; taking into account that some of the risks implied in their implementation are likely to materialize; and - sensitivity and risk analysis that addresses all the major risks involved and reflects the wide range of probable outcomes. 153. The sector should also emphasize improvements in time scheduling because such advances seem the most promising avenue to better costing of projects; the time estimates should, in particular, take into account that one or another of the perceived risks are almost certain to materialize. 154. In times of high uncertainty, programs that provide flexibility may be preferable even if, on the basis of a straight least-cost analysis, they do not represent the economic optimum. Specifically, it may be valuable for the sector to consider a program of large plants that will meet the projected baseline demand and a complementary program of smaller plants with shorter lead times that will cover the most likely difference in demand above the baseline. 155. Systematic adoption of comprehensive multi-objective planning might substantially benefit the sector. The findings further suggest that, in connection with power sector planning, government and the sector should use tools that allow analysts to capture, better than was done in the past, the - 46 - critical linkages with the macroeconomy, on the one hand, and the macroeconomic impact that alternative investment sequences and their associated financial and pricing strategies can have, on the other. 156. OED also suggests that the Bank support ISA in fully assuming its new role in coordinating distribution. This will help the sector to strike a rational balance between generation and transmission investment, on the one hand, and distribution investment, on the other, and will also contribute to reduce losses. VII. CONCLUSIONS 157. OED concludes that in the short term, the Bank should continue to support the Colombian power sector only if government and the sector prove willing to tackle decisively the broad, long-term issues. Developments since the close of the review period suggest that their willingness has increased. Policy Issues for Colombia 158. Colombia lacks basic goals for its power sector--specifically, a set -f government sectoral objectives in line with similar objectives for the wider energy sector, which themselves should be properly embedded in macroeconomic policies. To translate government sectoral policy into concrete measures, to coordinate the implementation of such measures, to monitor progress achieved, to assess success, and to determine possibly required remedial action--all require a strong regulatory body. This is so regardless of whether the broad sectoral objectives decided upon imply centralization (highly unlikely in the Colombian context) or a set of central and regional utilities with substantial autonomy--or even various degrees of privatisation. 159. Both the set of clear objectives and the strong regulatory framework are at present missing. These issues, which OED considers to be at the root of most others, urgently need addressing. In the long term, even financial success will have to take second place to the definition of clear government objectives and strong regulation. The years following the review period, 1988-89, have witnessed some promising steps toward strengthening such regulation through the Energy Board, for which Congress approved legislation in 1989. 160. In countries, such as Colombia, with strong regional sensitivities, the establishment of a central utility, such as ISA, for generation and transmission ownership is unlikely to be fully effective or commercially viable so long as it lacks support from the regional utilities and is concurrently owned by such utilities. 161. Beyond helping government and the sector to address the short-term problems, the Bank--most likely together with IDB--should be able to play a central role in establishing a new basis for the sector's development and operations. The central government, the regional and local authorities affected, the utilities, IDB, and the Bank need to agree on - 47 - - a long-term view of desirable sectoral development; - specific sectoral objectives: - a timetable for dealing with the issues; and - the forum for discussing the problems, reaching conclusions, and bringing the proposals to the stage where final decisions can be made and Laplementation started. Policy Issues for the Bank 162. Several of the conclusions of this review raise questions for the Bank that lead beyond the Colombian power sector. 163. First, in the first half of the 1980s in Colombia, Bank-sponsored measures to improve power sector performance--in particular, in the financial field--came into conflict with macroeconomic policies, also formulated using subsLantial analytical contributions from the Bank. Ultimately, the broad policies prevailed and the power sector's performance fell short of the objectives set. How to ensure consistency in the Bank's approach to the economy as a whole and to individual sectors is an issue that transcends the present study. 164. Second, the discount rate (which serves as a measure of the opportunity cost of capital) traditionally used to determine the least-cost power development can strongly affect the types of plants and the choice of sequence in which proposed projects are implemented. In the Bank's lending for power in Colombia, the discount rate used in the estimation of returns on projects has only been increased slightly over the past 15 years, even though in the early 1980s capital, especially in developing countries, became quite scarce. A clarifica- tion and justification of the Bank's position on this issue, which arises in many other sectors and countries, seems urgently needed. 165. The pervasive optimism--often not justified by experience, but reflected in many aspects of the Bank's presentation of programs and projects to its Board of Directors--identified in the course of the present study may also need to be addressed. This optimism may also be correlated with the pressure to lend, which, whether actual or only perceived, played a major and most likely detrimental role in several projects reviewed here, especially those prepared in the 1980s. 166. In the Colombian power sector operations reviewed, the Bank's Central Projects Staff (CPS), and later its Operations Policy Staff (OPS)--in this case the Energy Department--were singularly unsuccessful in their role as quality controllers. This raises the question of whether the re-organized Bank, in which the quality control function is carried out in much more immediate contact with the staff processing the operations, has, in the light of the Colombian experience, a better chance of success. 167. A lack of a shared broad vision and strategy, on the part of the Bank, the goverment, and the sector, limited the success of the operations reviewed - 48 - 167. A lack of a shared broad vision and strategy, on the part of the Bank, the government, and the sector, limited the success of the operations reviewed in this study. This deficiency is likely to be relevant also for some activities beyond the sector and country discussed here. In Colombia, in part because of this lack of shared vision and strategy, the Bank ased its limited resources to pursue too many objectives at the same time, valid as these objectives may have been. 168. Significant underpricing of electricity (USc2.5/kwh for residential users who represented about one-half of total sales) was a major factor in the poor performance of the Colombian power sector, as it is in many countries. This highlights again the need to strive for the economic pricing of electricity and the minimizing of subsidies in all Bank power operations. 169. Power investment, in several countries. has come to represent a substantial portion of total public sector investment and external debt. Greater attention must be given to the linkages between the levels of resource allocation to the sector and the proper financial management of the sector, on the one hand, and the achievement of macroeconomic objections, on the other hand. 170. The increased uncertainty and risk associated with critical variables, such as power demand and fuel prices, necessitates that more focus be given to the flexibility and affordability of power investment programs. 171. Finally, the Bank should review its guidelines regarding the financial indicators it uses to monitor, and bind by covenant, financial performance of its power sector borrowers. - 49 - ATTACHMENT Page 1 of 3 COLOMBIA POWER SECTOR REVIEW 1970-1987 SOME COMMENTS ON THE DOCUMENT -COLOMBIA, THE POWER SECTOR AND THE WORLD BANK (1970-1987)' PREPARED BY THE OPERATIONS EVALUATION DEPARTMENT - ORD (DRAFT OF FEBRUARY 20, 1990) 1. The document refers to the existence of political and regional pressures by ISA shareholders during the studies for the various generation and transmission expansion plans executed by ISA. It is our view that, beginning with the study carried out during 1988, these pressures may have been significantly reduced or even eliminated as a result of the execution of the studies on the standardization of all hydroelectric or thermal electric projects, with final feasibility studies now carried out by ISA and shareholders; this standardization is being implemented by ISA with the agreement of the shareholders. 2. Table 1.1, Vol. I, indicating the number of subscribers, generation capacity and sales in the sector, contains some significant errors, particularly as regards generating capacity in 1971 and 1986. It should be noted that during 1986 only 200 MW of El Paraiso-La Guaca came into service; the other 400 MW came on-stream during 1987. The table below shows the figures that according to our statistics are the correct ones. The figures for 1989 have been added. The result is to alter the growth rates shown in the document. Table 1.1: Power Subscribers, Generation Capacity, and Sales, 1971-89 Subscribers Generating Retail Sales Utility (thousands) Capacity in MW a in TWh 1971 1986 1989 1971 1986 1989 1971 1986 1989 EEEB 303 805 941 637 878 1158 1.8 4.8 5.8 EPM 184 414 491 579 1168 1369 1.6 3.9 4.5 CVC 231 485 547 262 860 821 1.1 3.0 3.4 ICEL 389 1,413 1,698 513 832 851 1.1 4.6 6.2 CORELCA 212 617 712 178 896 1,020 0.9 3.6 4.2 ISA - - - 0 1.885 2652- Total /b 1,348 3,831 4,530 2,169 6,519 8,370 6.7 20.4 24.6 Average annual growth rates, 1971-86 (Z) 7.2 10.5 7.7 1986-89 (Z) 5.8 8.7 6.4 /a Effective capacity. lb Incorporates other smaller markets not included in those shown above. - 50 - ATTACHMENT Page 2 of 3 3. As regards the analysis of the period 1971-1978, the suspension of disbursements in 1974 and the further three years required for the Bank to mount a new operation, viz. the ran Carlos I and 500 kV line loans, were the direct origins of the rationing that occurred in the National Interconnected System during 1980 and 1981. Construction on these two projects should have started in the period 1974-1977. In addition, given the certainty that this rationing would be required, the sector decided to construct emergency thermal plants, which were not part of the least cost plan, in an attempt to reduce the deficit. Unquestionably this entire situation, in conjunction with the tariff problem from 1971 to 1975, caused the financial problems that have burdened the sector since that time. This fact is only very lightly touched on in the OED document. 4 The secure and economical carrying capacity of the 500 kV line linking CORELCA to the Central System is much less than the 960 MW referred to in the OED report. From an electricity standpoint, reliability problems, voltage regulation and temporary and dynamic stability in the secure operation of the CORELCA system, have created a situation in which the maximum transfer capacity of this line under current conditions is 290 MW. This statement is demonstrat- ed in the attached document entitled *Reliability Analysis to Determine the Interchange of Maximum Economic Benefit Among Interconnected Areas., 5. As regards the ex-post evaluation made in Annex 4D.2 of the energy and power supply in the interconnected system it must be stated that, at the end of the summer season in 1987 and 1988 in particular, had the actual generating capacity not existed, it would have been necessary to resort to rationing owing to particularly adverse hydrological conditions. The con- clusion, therefore, is that all the projects in the 197711978 generation program which came into service during these years (1987 and 1988) could not have been postponed to a later date. Hence the latest dates for the following projects in the 1977/1978 generation program are: Guadalupe IV November 1985 Tasajero February 1985 Termoguajira II November 1987 Betania November 1987 Therefore, the dates of possible entry into service given in Annex 41.7 are not feasible, especially those for Guadalupe IV and Betania; as a result the conclusions derived from this analysis as regards the costs of excess capacity are not valid. 6. It is repeatedly asserted throughout the document that the projections for electric power demand for the 1980s were very high, resulting in a larger generation program thanwas in fact needed. While this is true, it creates the impression that this only occurred in Colombia, whereas in practice the same thing happened in many industrialized and developing countries. This leads us to the conclusion that the methodologies for estimating the demand for electricity have not yet been perfected and that research in this area should - 51 - ATTACHMENT Page 3 of 3 continue, while also taking into account the difficulty of predicting periods of economic recession such as occurred at the beginning of the 1980s. Likewise, studies should be made to produce scenarios conferring greater flexibility on expansion plans. 7. In general terms, the OND document contains a very comprehensive analysis and diagnosis and presents in a coherent and explicit form, probably for the first time, all the problems facing the sector, while recommending possible solutions. In brief, it is a very well-documented and well-written report which it must be hoped will be taken into account by the agencies and individuals who have to take decisions regarding the future of the electricity sector in Colombia. ISA Interconexion Electrica, S.A. Planning Office April 1990 IBRD 17690R1 cio AM E Rt l Å mtao p i a 2 Termogu0pro lvostec'O3 - A Santoo Mortoro- r2 b Oorronque ch.. edep 8' .U,a/ A \af A MAt ENEZ, C a UrrU -u u 4 8aaramong ''- G4udolpe Å4,,> ,'ra.cobemne Caotf,s'6 8rbosc pReogr AI Joguas M 11 a esý0 Q.-bdo0 An&n arpo Cho,, Carm*b pemm,r To,c Guavmo Arffen "Q Iou o4 0, -bg. OTA 4.. ByBna?.oren:IIo An¶chicoyo Cal 1,ne COLOMBIA e,ania ELECTRIC POWER SYSTEMS FuIuQE , ' i-t"",*, 5YSTEYS Florenc.o .SO$ kv E P M 0 1 JO kV = CVC Hydror,er plants C(:REiC Thermal power plants ICEL Pao Substahons • Mocoo Rwvers Department,Intendenc and Comisarto boundaries - C U A D 0 R''e° nd'a' E",ud'.es JUlY 1988