ENERGY AND MINING SECTOR BOARD DISCUSSION PAPER PAPER NO.16 MARCH 2006 Centralized Purchasing Arrangements: International Practices and Lessons Learned on Variations to the Single Buyer Model Beatriz Arizu, Defne Gencer and Luiz Maurer THE WORLD BANK GROUP The Energy and Mining Sector Board AUTHORS DISCLAIMERS Beatriz Arizu (barizu@mercadosenergeticos.com) is an execu- The findings, interpretations, and conclusions expressed in tive with Mercados Energéticos, a consulting firm based in this paper are entirely those of the authors and should not be Argentina. In this capacity, Mrs. Arizu has been involved in attributed in any manner to the World Bank, to its affiliated the restructuring of power sectors in about 20 countries. organizations, or to members of its Board of Executive Before joining Mercados, she worked for the Energy Secretary Directors or the countries they represent. in Argentina. CONTACT INFORMATION Luiz Maurer (lmaurer@worldbank.org) is a Senior Energy Specialist at the World Bank. Before joining the Bank, he To order additional copies of this discussion paper please worked as an executive and as a consultant in the oil and gas contact the Energy Help Desk: +1.202.473.0652 and utility industries in Latin America and in the United States. energyhelpdesk@worldbank.org Defne Gencer (dgencer@worldbank.org) is a Power Sector This paper is available online www.worldbank.org/energy/ Reform Consultant in the Energy and Water Department at the World Bank. Before joining the Bank, she worked for the Energy Market Regulatory Authority of Turkey as a specialist on power sector regulation. The material in this work is copyrighted. No part of this work may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or inclusion in any information storage and retrieval system, without the prior written permission of the World Bank. The World Bank encourages dissemination of its work and will normally grant permission promptly. For permis- sion to photocopy or reprint, please send a request with complete information to the Copyright Clearance Center, Inc, 222 Rosewood Drive, Danvers, MA 01923, USA, fax 978-750-4470. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, World Bank, 1818 H Street N.W., Washington DC, 20433, fax 202-522-2422, e-mail: pubrights@worldbank.org ENERGY AND MINING SECTOR BOARD DISCUSSION PAPER PAPER NO.16 MARCH 2006 Centralized Purchasing Arrangements: International Practices and Lessons Learned on Variations to the Single Buyer Model Beatriz Arizu, Defne Gencer and Luiz Maurer The World Bank, Washington, DC THE WORLD BANK GROUP The Energy and Mining Sector Board Copyright © 2006 The International Bank for Reconstruction and Development/The World Bank. All rights reserved CONTENTS ABBREVIATIONS AND ACRONYMS ....................ii FOREWORD ..............................................................1 EXECUTIVE SUMMARY ............................................3 1 . INTRODUCTION ................................................4 2 . FORMS OF THE PURE SINGLE BUYER MODEL ..................................5 3 . ADVANTAGES AND DISADVANTAGES OF THE PURE SINGLE BUYER MODEL ................7 4 . EXPANDING THE CONCEPT OF THE SINGLE BUYER ................................16 5 . MITIGATING SOME PROBLEMS OF THE SINGLE BUYER WHILE PRESERVING ADVANTAGES ..................................................21 6 . CONCLUSIONS AND RECOMMENDATIONS ....................................36 ACKNOWLEDGEMENTS This paper could not have been written without the assistance of many friends and colleagues inside and outside the World Bank. These individuals helped us develop an understanding of the wholesale market design and implementation issues in developing and developed countries. They generously shared their time and insights to make sure that we understood the issues at hand and their implications for the countries we focused on. They include Ali Saleh Al-Barrak, Tariq Al-Betairi, Suman Babbar, Philippe Benoit, Ashley Brown, William Dunn Jr., Jonathan Halpern, Dilma Rousseff, Bernard Tenenbaum, Mauricio Tolmasquim, Fareed M. Zedan, and Fiona Woolf. Of course, none of them should be held responsible for any errors of fact or interpretation that remain. We would like to give special recognition to James Barker, Laszlo Lovei, and Tjaarda Storm van Leeuwen, our diligent peer reviewers. We would like to particularly commend Laszlo for pioneering the discussion of this important and controversial topic at the World Bank and inspiring us to further investigate its implications. ii Finally, we wish to acknowledge the financial assistance of the World Bank’s Energy and Water Department (EWDEN) in preparing this report. We owe a special debt of gratitude to the department's director, Jamal Saghir, who provided resources and encouraged us to pursue this topic. ABBREVIATIONS AND ACRONYMS AES AES Corporation, USA BLCP BLCP Power Limited CCEE Chamber of Electric Energy Commercialization, Brazil CDEEE Dominican Corporation of State Electricity Companies CFE Federal Electricity Commission, Mexico DWR California Department of Water Resources EDF Electricité de France S.A., France EGAT Electricity Generating Authority of Thailand ESKOM Electricity Supply Commission, South Africa EVN Electricity of VietNam IDA International Development Association IPP Independent Power Producer IPTL Independent Power Tanzania Ltd MW Megawatt MWh Megawatthour PPA Power Purchase Agreement PRG Partial Risk Guarantee PURPA Public Utility Regulatory Policies Act ROE Return on Equity SPP Small Power Producers TANESCO Tanganyika Electric Supply Company Limited TEAS Turkish Electricity Corporation TEIAS Turkish Electricity Transmission Company TETAS Turkish Electricity Trading and Contracting Company TSO Transmission System Operator WPPI Wisconsin Public Power Inc. US$ U.S. Dollars FOREWORD The World Bank has been approached by some developing countries asking for advice on how to expand their generation sector when a comprehensive restructuring of their power sectors is not technically feasible or politically desirable. Under those circumstances a question frequently asked is whether or not the Single Buyer is the only remaining alternative to attract private capital. And how should it be implemented? In this dialogue, the Bank’s client countries have expected some rather pragmatic advice on how the Single Buyer model should be implemented to maximize its benefits and avoid its drawbacks. In order to review worldwide experience with Single Buyers and other Centralized Purchasing Arrangements, the Bank selected several case studies, with a heavy 1 emphasis on developing countries. Different models have been scrutinized, in search for relevant lessons or directions to the developing world. This paper is an attempt to approach this concept with a fresh perspective and examine how it was implemented, what objectives it served, how it changed over time, and what arrangements have been introduced to overcome its drawbacks, while preserving its positive attributes. The paper provides some lessons and recommendations that may be useful both for countries which are contemplating to restructure their power sectors and for those at more advanced stages of power sector reform. Jamal Saghir Director, Energy and Water Chairman, Energy and Mining Sector Board 2 EXECUTIVE SUMMARY The expression “Single Buyer” has been dubbed by the alternative means of expanding their generation power industry to designate a commercial arrangement capacity. Reasons are many. Firstly, several of those whereby one entity, usually government owned, has the countries are financially constrained, plagued by inefficient, monopoly over wholesale electricity trading in a region. non-creditworthy state owned utilities. Secondly, smaller The Single Buyer, in its purest form, is also the sole seller countries remain unconvinced that a competitive of wholesale energy to all end consumers, either directly wholesale market is a workable proposition that can be or via a distribution company. In this capacity, a Single introduced in the short run.1 Thirdly, some large Buyer may be an electric utility or an entity created countries, not necessarily financially constrained, prefer specially to assume the energy trading role. a Single Buyer to maintain the government’s control over the development the power sector. In part, this The Single Buyer model was introduced as an initial concern is a somewhat disproportionate reaction to the step in power sector reform. It started in the United problems brought about by the California crisis.2 States as part of the PURPA approach, with the objective of increasing competitive pressure on integrated utilities The issue then becomes what those countries should and promoting cogeneration. Later on, a similar model do to balance the need to expand their power sectors was adopted in the developing world, this time with the with the necessity to avoid implementing institutional 3 objective of attracting new investments in generation arrangements that they may later regret. To help countries from the private sector, mainly in countries facing address these challenges, the Bank has carried out an serious energy shortages. By introducing the Single empirical analysis of the implementation of the Single Buyer model, many developing countries were able to Buyer model and other variations of centralized purchasing achieve remarkable results in terms of attracting much arrangements in the last 15 years, in around world. This needed new capital into distressed power sectors, paper is an attempt to shed new light into the controversial, particularly in Asia. Thousands of MW were contracted oftentimes polarized discussion of the Single Buyer model and commissioned. Undoubtedly, in those cases the by extracting some lessons learned and best practices model helped relieve power shortages and in this around the world. respect support economic growth. An interesting finding of the study is that there are Despite the initial optimism, the model was not flawless. several ways to implement what is broadly referred to as Experiences in many parts of the world have shown that a “Single Buyer” model. In some of the cases reviewed, the Single Buyer fell short of expectations. It created a the relevant entity assigned as Single Buyer was not series of unanticipated problems such as excess capacity, “single” and in other cases it was not a “buyer” at all. high tariffs and stranded costs. It also lacked transparency One observed common feature in most of the case and accountability, and in some cases exacerbated studies is the existence of one central entity aggregating corruption problems. Because of the rigidity of the the load, playing some role in the procuring of energy contractual arrangements put in place, the Single Buyer to serve that load, and allocating this energy among impeded the development of competition and the different consumers or distribution companies. Given evolution of power sector reform. those characteristics, we chose the expression “centralized purchasing arrangement”, to more properly capture the Nevertheless, the Single Buyer model continues to multitude of commercial arrangements that can be in be considered as a valid wholesale energy trading place.3 The pure Single Buyer model is one among arrangement and is being adopted in some countries, many forms of centralized purchasing arrangements. most of which do not foresee any other workable 1 Not only distressed and small power systems are considering the implementation of Single Buyers. The Province of Ontario, in Canada, has just implemented a similar arrangement, the rationale being the pressing need of promoting a significant expansion in the power system, since many of its existing plants will be decommissioned in the short term. 2 For a detailed discussion on the California crisis, please refer to The California Electricity Crisis. Sweeney, James L, Hoover Institution Press, Stanford, California, 2002 and to The California Power Crisis: Lessons for Developing Countries (200k pdf). World Bank Energy and Mining Discussion Paper No. 1, John Besant-Jones and Bernard Tenenbaum, Washington D.C. April 2001. 3 In the absence of a more accurate denomination, a centralized purchasing arrangement is intended to refer to a commercial process where an agent is responsible for some form of coordination among the multiple tasks involved in wholesale energy procurement By examining the broad range of variations in design 1. INTRODUCTION and implementation of energy trading models around the world, our evaluation showed that sometimes the The analysis builds upon previous studies carried out for pure Single Buyer is actually the only alternative, Tunisia, Saudi Arabia, and Brazil, as well as on the particularly in distressed power systems. Even under hands-on experience developed by the World Bank in dire circumstances, it is still possible to minimize many countries which faced similar challenges of mobilizing of its drawbacks with a careful design at the outset. It the necessary investment in their power sectors. This is essential to procure energy competitively and build paper also examines how the Single Buyer concept or flexible contractual and institutional arrangements that variations thereof, were initially implemented and have envisage a smooth transition towards a more competitive recently evolved in countries like Cambodia, Honduras, power sector, when conditions permit. Hungary, Italy, Jordan, Mexico, Northern Ireland, Panama, Thailand, Turkey, Ukraine, the United States The results of this study also indicate that it is neither and Vietnam. The countries were selected after a screening necessary nor desirable to grant a de jure monopoly and clustering process, with a view to ensuring that the status to a Single Buyer, even at the early stages of cases offered a good geographical coverage. power sector reform. Such a monopoly may backfire, by creating future rigidities that will defeat the main A “pure” Single Buyer may be defined as a structural 4 objectives of the reform process. Experiences have and commercial arrangement where one entity, usually shown that more often than not, Single Buyer models a government owned agency or company, has the have resulted in unaffordable prices, lack of transparency, monopoly over wholesale energy procurement and and virtually little or no competition in power generation. resale. We identify two major types of the “pure” Single Buyer: the integrated and the unbundled one, which are differentiated based on the degree of vertical integration or the extent of unbundling in the sector. Single Buyer models are centralized purchasing arrangements, built on the idea of one entity aggregating the load, playing a major role in the procurement of energy to serve that load, and in allocating this energy directly to large or small consumers either via “marketers/retailers” or via distribution companies. However, in many of the cases analyzed – which were still referred to as Single Buyers - the relevant entity was not single, nor it was involved in any buying or selling of energy. In this sense, they were not “pure” single buyers. In light of the multiple variations in the application of the concept, we tried to define a broader concept of “Centralized Purchasing Arrangements.” We use this term to embody commercial and organizational structures that have somewhat misleadingly been referred to as “Single Buyers”. The comments and criticisms to the “Single Buyer” model, discussed in this paper, refer to this arrangement in its purest form. Using the selected case studies, this paper examines 2. FORMS OF THE PURE SINGLE BUYER how the pure Single Buyer model, as well as other MODEL centralized purchasing arrangements, have worked around the world. For each case, we discuss the positive The “pure” Single Buyer may be organized in two different and negative aspects of the existing arrangements, as ways: vertically integrated or functionally unbundled. well as any proposed changes to deal with specific Each one of those organizational alternatives will be shortcomings. We then extract general lessons that may discussed in detail below. be useful both for countries which are contemplating to restructure their power sector and for those at more The Integrated Single Buyer advanced stages in the reform effort. The introduction of an Integrated Single Buyer model in the developing world occurred in the late 1980s and early 1990s, and was one of the first steps towards opening the power sectors to competition in developing countries. Some examples of this model include Tanzania, Nigeria, Cambodia, Mexico, and Thailand. According to this approach, a vertically integrated utility 5 plays the role of buyer in the wholesale market. The utility may own generation assets, but not sufficient to meet its entire load. Therefore, the utility acquires energy from Independent Power Producers (IPPs) via long-term contracts. By choosing this arrangement, countries have been able to maintain their power sector structures relatively intact; thereby averting strong backlash, which would have emerged in the case of more fundamental reform efforts. This arrangement spurred private sector generation investments in financially constrained power systems. Countries embarking upon this kind of arrangement were in most cases seeking rapid relief from acute supply shortages, or expected future shortages if no new investment came in place. It was also expected that this model would be able to create some form of competition “for” the market, as multiple bidders should in theory compete to build new generation and sell energy to the integrated utility. The electricity trading arrangements were usually characterized by long-term power purchase agreements (PPAs). Those were typically 15- year or more power contracts, in many cases denominated in US dollars to hedge the producer against currency exchange risks. Some countries took advantage of the possibility to create competition “for” the market through competitive procurement (e.g. Thailand, Mexico), while others accepted unsolicited offers and entered into negotiated deals. This model was extensively used in countries at early involvement in providing sovereign guarantees. Such stages of power sector reform. In Asia, for example, kinds of arrangements created contingent liabilities and Pakistan and Thailand had their state-owned integrated are not convenient in the long term. power company purchasing energy from IPPs to supply their own distribution divisions. The same model has The Unbundled Single Buyer been applied in Mexico, due to the political difficulties of reforming the sector and in unbundling CFE, the The main feature of this model is that the generation state owned utility. In those countries, some form of business is separated from the “wires” and from the competition was introduced “for the market,” on an retail businesses. This separation may be implemented incremental basis. Tanzania, Indonesia, Malaysia, the as functional and accounting unbundling (ring fencing), Philippines, Dominican Republic, Nigeria, Cambodia, or by the creation of independent companies (or India, and many other countries decided to embrace subsidiaries of a holding company). Purchasing functions this integrated model. may be carried out by the generation company or by an independent entity. Many of those countries had no alternative but to adopt an integrated Single Buyer at the outset. Only the A typical institutional arrangement has been to combine existing state owned companies with sovereign guarantees transmission assets and power system operation functions 6 were sufficiently creditworthy to attract lenders and with wholesale trading activities. The transmission service power project developers. Attempts to have distribution provider, which is also the system operator,4 plays the companies (or affiliated divisions) buying directly from role of the monopoly wholesale purchaser. This approach greenfield power plants would not have succeeded, was followed initially in Northern Ireland, in the early given the lack of creditworthiness of those utilities as stages of the power reform process. off-takers. A benefit of this approach is that the Unbundled Single When a utility is not creditworthy due to its poor financial Buyer does not own generation assets and should in performance, subsidized tariffs, low collections or high theory dispatch and provide network access to all technical and non-technical losses, PPAs tend to require available generation on a non-discriminatory basis.5 sovereign guarantees. Therefore, in such case the Therefore, this arrangement should increase the level success of attracting private investors is heavily dependent of comfort among new players interested in making on Government acting as a guarantor, which should new investments in generation. It also potentially paves only be considered as a transitory measure. At the the way for an embryonic competitive energy market, outset of the IPP programs, these guarantees helped to where all generation companies are on a level playing increase the comfort level of private investors and of the field as far as dispatch and access are concerned. investment bank community. The possibility of shifting risk away from developers was received with great Since the first Unbundled Single Buyer was introduced, enthusiasm. For some, it seemed that the Single Buyer it has been implemented in several jurisdictions, under would be able to solve the problems of the power sector different market structures. It has been adopted, for without the country being forced to confront the thorny example, in countries at the early stages of power sector issues of cost reflective tariffs, reduction of non-technical reform, which needed to attract private financing for new losses and non-payment by consumers. However, this generation or rehabilitation of existing power plants. It also created a burden on the Government since the critical has been proposed as a transitional arrangement in power problems – average tariffs that do not recover costs or sector reform as an attempt to gradually foster competition inefficient losses – were not addressed and resolved. and at the same time attract private investment. The integrated Single Buyer model was usually successful in enabling new generation capacity, though in many cases this apparent success was due to the Government 4 By system operator we mean the dispatch function, control centers and associated activities to coordinate generation and maintain a reliable and balanced transmission system and load supply. 5 One possible way to prevent discrimination is to oblige the integrated utility to procure energy from third party producers and pay the equivalent to the buyer’s avoided cost, as determined by the relevant State regulatory commission. This was the case in the US as a result of the PURPA legislation. 3. ADVANTAGES AND DISADVANTAGES OF plants during the 1990s, Southeast Asia alone was able SINGLE BUYER MODEL to attract more capital than all of Latin American efforts both in privatization and new projects.7 This section discusses the advantages and disadvantages of the “pure” Single Buyer model.6 It relies upon empirical As a result of the success, during the period 1991-97, evidence and lessons extracted from the case studies that contracts for 137 greenfield private power projects for form the basis of this document. 100 GW or more and worth US$65 billion were concluded. About half of those generation projects Advantages of the Single Buyer Model consisted of build-own-operate and about half of build- operate-transfer schemes. Asia accounted for 103 The main observed advantages of the Single Buyer contracts worth US$54 billion, and Latin America for 28 model are shown in Box 1. Each one of those projects worth US$6.6 billion.8 advantages will be discussed below. A lot of this success was driven by the existence of the Single Buyer model, offering long term PPAs with Box 1: Advantages of the Single Buyer Model sovereign guarantees. 1. Ability to attract capital by offering long term PPAs Being a workable alternative for distressed power 7 sectors 2. Being a workable alternative for distressed power sectors to initiate reform In distressed power systems, distribution companies or the integrated power utility are not creditworthy buyers 3. Ability to centralize purchases, developing expertise due to, in many cases, absence of cost-recovery level and benefiting from economies of scale tariffs, high losses due to lack of meters, fraud and 4. Ease of implementation, in terms of not requiring theft, as well as low collections rates. In this kind of fundamental changes in sector structure, deal- environment and until those inefficiencies are resolved, making and acceptability by the prevailing culture. it is practically impossible to envision a scenario with multiple, credible buyers and sellers, particularly in small systems. The “natural” solution has been to assign only one buyer backstopped by government (sovereign) Ability to attract capital by offering long term guarantees. Even if a country recognizes that fixing PPAs some of the basic commercial and technical problems is essential to the sustainability of the power sector, it is The Single Buyer model had the merit of attracting perceived that results are unlikely to be achieved in a significant capital to the energy sector, especially in short time frame.9 Moreover, there may be complex power generation. In addition, it enabled countries to issues of affordability10 and political resistance that have access to modern technologies that would have cannot be addressed overnight. Under those otherwise been difficult for them to adopt. In that sense, circumstances, a Single Buyer model has been viewed the model was successful. In terms of new, greenfield as the only supply-side alternative able to buttress 6 Since this section refers to the “pure” Single Buyer, the advantages and disadvantages herein described may not fully apply to “enhanced” versions of the Single Buyer. Those include, inter alia, models with competitive procurement, the phase out of the monopoly status, and more flexible contractual arrangements. 7 This comparison does not take into account the residual government role in the power sector. Reforms in Latin America reduced government involvement, and privatization reduced fiscal debt, therefore relieving funds for other priority areas such as health, education and poverty alleviation. IPPs in Latin America had no sovereign guarantees, even in smaller countries such as Nicaragua. On the contrary, increasing generation capacity in Asia entailed contingent liabilities and significant costs to the government. 8 In the 1990-99 period, AES, an Independent Power Producer (IPP) based in the US, was the most important sponsor of independent power projects around the world, with 35 projects and a total investment of $12.7 billion. AES, Enron and Electricité de France together accounted for more than half of the 156 projects and the total investment of $68.2 billion during this period. 9 Real world experiences have shown that improving efficiency and reducing non technical losses can be achieved faster than many times believed. What several times has delayed initiating changes was the fear of the incumbent utilities and other beneficiaries of the status quo. 10 Issues on affordability involve, inter alia, a broad discussion on the issue of targeting and efficient subsidy delivery mechanisms. capacity expansion and mitigate chronic power shortages.11 Box 2: Aggregation of Demand and Economies of What the holders of these views failed to recognize was Scale – Municipal Utilities that they were witnessing a quick fix for a supply shortage. Unless the root causes of the inefficiencies discussed The typical Single Buyer entity is a mandated, top- above are resolved, the PPAs paid in full, and network down, government-owned enterprise—but it does not activities received sufficient funds for investment and have to be designed this way. An alternative form of maintenance, a system would most likely continue to single buyer would be a voluntary, bottom-up buying experience a power crisis. Worldwide experience has agency, formed by a group of distribution utilities (for demonstrated that resolving shortages can never be as example municipal distribution utilities) to acquire simple as signing many PPAs and building many power power supplies on their behalf. The joint action stations at any cost. agency would perform this function by buying power through contracts or by purchasing ownership Ability to centralize purchases, develop expertise interests in existing or new generating units that may and benefit from economies of scale also be owned by one or more vertically owned utilities, or on spot markets where they exist. Such aggregating buyer entities have existed for many In very small, undeveloped or fragmented electricity years in the United States. For example in Wisconsin systems it may be difficult to build and replicate energy 37 small municipal distribution utilities, serving a 8 procurement expertise among all distribution companies. population of about 250,000 people, created a The required expertise includes, inter alia, the preparation jointly owned company called Wisconsin Public Power of power purchase agreements, tasks related to Inc. (WPPI) to cover their aggregated power needs. procurement (competitive or single source), negotiation The goal of WPPI is to be “a smart, efficient and with suppliers, and contract management. This knowledge aggressive aggregator of power for the benefit of our was virtually non-existent (and unnecessary) in a fully member communities.” This means developing and integrated state owned company, where all the generation maintaining a diverse power supply portfolio from was produced “in house,” but became essential when which members may offer their customers electricity new IPPs enter the market, particularly for efficient and that will be more attractive than other options. In general there is little or no regulatory oversight of the fair cost entry. The Single Buyer offered a possibility of actions of the joint action agencies. The rationale for developing this expertise in a centralized fashion.12 this “hands-off” regulatory approach is that contrary to the typical “top-down” approach of a single buyer There are other ways to accelerate the power the joint action agency is owned and governed by its procurement learning process and achieve economies beneficiaries. of scale. For example, expertise may be outsourced, as was done in Northern Ireland. Furthermore, Source: Bakovic, Tenenbaum and Woolf. (2003) economies of scale13 in purchasing may be achieved by the aggregation of multiple distribution companies represented by one single buying agent, achieving larger volumes and having more bargaining power over suppliers, particularly when the multiple off-takers are credible buyers. An interesting example is represented by the aggregation of small municipal utilities in the U.S. (Box 2). 11 Most countries facing power shortages have underestimated the importance of working on the demand side to alleviate the crisis. This discussion goes beyond the scope of this document. For a thorough analysis on the subject, please refer to Maurer, Rosenblatt and Pereira – “Implementing Power Rationing in a Sensible Way.” ESMAP 305/2005. The World Bank (August 2005). Available at http://www.worldbank.org/esmap 12 The original assumption was that a centralized transmission company or Ministry staff could easily obtain this expertise if procurement were to be organized in a centralized fashion. Some countries did jumpstart in the learning curve. The World Bank has provided technical assistance and expertise in the areas of procurement and contracting. 13 It is also considered the issue of “lumpiness” of investments. Benefits of scale in generation can be achieved by the aggregation of smaller buyers, therefore justifying the building of more economic large-scale generation. Ease of implementation and acceptability for the Disadvantages of the Single Buyer Model15 prevailing culture Despite many of the undeniable advantages, the Single A Single Buyer model is viewed as an option that Buyer model fell short of expectations in many parts of should be relatively easy to implement. It does not the world. In reality, it turned out to be not as easy to require radical changes in the institutional and commercial implement in an efficient and sustainable manner as it frameworks. Put simply, the incumbent utility relinquishes was originally perceived. For many, the Single Buyer its right to build new capacity and signs a power purchase model brought more problems than those it was able to agreement with an independent developer. The solve and the model was eventually strongly criticized as organizational structure and institutional arrangements a “dangerous path toward competitive electricity remain basically the same. It is much simpler than, for markets.”16 Empirical evidence suggested that, in many example, unbundling the power sector, creating an (albeit not all) cases, a government owned purchasing independent system operator and a wholesale market monopoly invited corruption, weakened payment administrator. Building and operating the power plant discipline, and imposed high contingent liabilities upon becomes a developer’s responsibility. The Single Buyer the government.17 The most important disadvantages of does not have to get involved in construction, financing, the Single Buyer model are summarized in Box 3, system integration, technology evaluation and other followed by a discussion of each issue. similar tasks, particularly when the power plants are 9 built on a turn-key basis.14 Box 3: Disadvantages of the Single Buyer Model State-owned, vertically integrated utilities prefer the 1. Demonstrated failure to be “good’ buyers Single Buyer to other alternatives that represent more radical restructuring of the power sector (that may have 2. Lack of transparency as a pathway to corruption an impact on the utilities themselves). The Single Buyer preserves the utility’s control over the energy traded and 3. Rigid contractual and institutional arrangements usually does not represent any immediate threat to the organization structure or employment levels. 4. Creation of contingent liabilities From a transaction perspective, keeping the utility and 5. Inability to eliminate risks for developers as the regulatory framework virtually intact has also its own expected merits. Government, developers and investment banks have to focus their attention primarily in the design of a PPA and related documents. Despite not being an easy Demonstrated failure to be “good’ buyers task, it is still much simpler and faster than dealing with the multi-faceted and politically thorny issues of a far- Over the years, most Single Buyers, either integrated or flung power sector reform. Therefore, the Single Buyer, unbundled, have failed to develop a good track record PPA and sovereign guarantees trilogy has come to be in terms of sound procurement practices. They generally seen as the ideal solution from a fast, straightforward lacked commercial acumen and contractual expertise. deal-making mentality; yet oftentimes at heavy eventual Furthermore, they tended to “gold-plate” the system by cost to the consumers, power system and society at large. planning more capacity and signing more PPAs than necessary. In some cases, overoptimistic demand forecasts resulted in excessive contracting and 14 An exception may be the case of a small system with low access, which still has to grow significantly. In such case, many technical characteristics are imposed by the Single Buyer as part of central planning of the sector. 15 This section benefited from excerpts and cases at “Private Gain – Public Risk? The International Experience with Power Purchase Agreements of Private Power Projects,” by Peter Bosshard from the International Rivers Network, published on November 20, 2002. 16 Lovei, Laszlo. (2000). 17 This viewpoint had a great impact among World Bank client countries. It presented a harsh, nonetheless candid evaluation on the dismal performance of Single Buyer models around the world. recessions in the mid 1990s. As a consequence, stranded Box 4: Single Buyer procurement and the California investment costs built up, contracts had to be renegotiated, Crisis and in some cases significant financial defaults resulted from the “euphoria” in contracting capacity. In an attempt to reduce the high energy price volatility and to mitigate a serious financial crisis faced by distribution utilities in 2001, the Governor Even in jurisdictions with stronger governance and of California ordered that the Department of Water more developed markets, Single Buyers did not have Resources (DWR), a small state owned utility, be in the incentives to make sound commercial decisions, charge of the procurement of energy on behalf of particularly in times of crisis and when acting under the state's three largest investor owned utilities to political pressure. A similar situation was seen in meet a substantial part of the market requirements. California during the power crisis, where a state Those large utilities were close to bankruptcy due to government branch played the transitory role of a their inability to fully pass through power purchasing quasi Single Buyer, as described in Box 4. costs to their retail customers, generating a financial gap of several US$ billions. DWR had to buy energy Lack of transparency as a pathway to corruption from generators and re-sell to distribution companies. It was similar in some aspects to a Single Buyer type Decisions made by a Single Buyer have been criticized of arrangement. DWR did not have the required 10 skills to carry on such a complex task, particularly in for not being sufficiently transparent. Lack of transparency a situation where it was suspected that generators opens up an avenue for corruption – or at least creates were exercising market power that was aggravating a strong perception thereof. an existing shortage in available generation capacity. DWR, under intense political pressure, ended up The Single Buyer model per se does not necessarily signing medium and long-term contracts that turned lead to corruption. Some conditions should be in place out to be unsustainable.18 Billions of US$ were for corruption to manifest and hurt the customer or committed in contracts that soon became stranded. the taxpayer. Most often than not, those conditions are DWR had the merit of introducing energy contracts of systemic nature, and are related to poor governance as financial hedges in a situation which was already mechanisms. out of control. However, for some experts, it should have exercised its monopsony power and carried out Despite not being the culprit for systemic corruption, the a more competitive process to force prices down. After high price contracts were signed, there were attempts Single Buyer centralizes many key decisions in the capacity to fix the problem by renegotiating the agreements, expansion process in the power sector. Those decisions but resulting prices were still very high and may include, inter alia, system planning, which determines confidence in the sanctity of contracts was the new projects or the generation capacity to be built,19 undermined. Recovering those stranded costs from how much additional generation capacity should be customers would have entailed significant tariff procured, the technical and financial requisites of the increases that were unlikely to be implemented. potential bidders/sellers, draft procurement documents Instead California taxpayers will have to bear the and PPAs, selection of the sellers or the PPA award consequences of such decisions for many years to criteria, and how contracts should be negotiated. By come. centralizing those decisions, the Single Buyer may put in place a conduit for undue political interference which Source: Sweeney (2002) and World Bank analysis. may lead to corruption. construction of new plants. This was typical in Asian Therefore, the excessive centralization of key decisions countries, where available capacity could not be absorbed in environments with poor governance provides by demand growth due partly to overly optimistic load incentives for corruption to manifest itself. Some of growth forecasts and partly to unexpected economic those most relevant decisions will be discussed in turn. 18 It must be noted that the contract prices were not considered so high at the time they were signed, compared to those that prevailed in the market during the shortage, with prices reaching historical peaks. 19 Some Single Buyers plan specific generation projects, whereas others only determine how much new generation capacity is needed. Centralized Planning of the System It is generally accepted that power procurement per se is the most sensitive area to assure transparency, government Planning is a genuine and useful technical component accountability, and to avoid corruption in a Single Buyer to support capacity expansion. However, it can be type of arrangement. Lack of transparency has in many manipulated to serve the interests of politicians and cases triggered a lot of genuine political opposition.22 technocrats.20 The more centralized the process without consultation with or inputs from stakeholders in the Many examples suggest a clear link between the lack of power sector, the more it is subject to undue interference. due process in planning and procurement and the Very often, load forecasts are extremely optimistic, development of unwarranted, often uneconomic private leading to excessive contracting and over capacity. power projects, as shown in Box 5. Over optimism does not affect attracting investors as, under most Power Purchase Agreements, the load growth risk is shifted to the off-taker. The Single Buyer, Box 5: High Costs Due to Non-Competitive Power often responsible for the demand forecast and planning Projects and Oversupply and backed by Government guarantees, has limited In the case of Enron’s Dabhol thermal power plant in incentives to provide a realistic forecast. Maharashtra, a company executive admitted that Enron spent “approximately US$20 million” on the Non competitive procurement “education and project development process alone, 11 not including any project costs”. In September 1995, In many cases, energy is procured by a Single Buyer Maharashtra’s government said in court that the from unsolicited projects. Procurement processes not Dabhol PPA had been procured on a non-competitive always follow a codified competitive tender or a widely basis “by fraud and misrepresentation.”23 publicized auction procedure. Selected projects are not necessarily part of the least cost expansion plan but In the Philippines, there were concerns that many result from sometimes obscure bilateral negotiations onerous PPAs could impact generation costs. between developers and the host countries from which Independent investigations suggested that the government was pushing the speedy approval of projects are solicited at the time.21 some of the most expensive power deals and justified signing more contracts despite warnings from within This approach has resulted sometimes in contracts for the government and from the World Bank that an power projects which were not needed, extremely impending oversupply could push prices up.24 attractive terms and high prices offered to the investors, and a shift of all major risks to the buyer and host In the Dominican Republic, CDEEE management governments. Favorable terms of investment, hidden alleges that PPAs with several IPPs, as well as costs and an oversupply of power have in several subsequent negotiations held under the “Acordo de countries caused power prices to soar and led to Madrid” represent energy overprice of 25%. Two of political turmoil over power tariffs, subsidies and the the most expensive PPAs represent a loss to the utility renegotiation of contracts. of about US$ 7 million per month.25 Source: Author’s analysis 20 For a discussion on how discount rate, one single methodological element in a planning process, may radically impact the least cost expansion options, refer to Becker and Maurer. Plan 2010 ignores variations in the cost of capital. Gazeta Mercantil. Brazil. (1990) 21 This is not always the case. In Thailand and Mexico, which will be described later in this paper, a transparent competitive process was put in place, showing very satisfactory results. 22 “The most perverse corruption is in the procurement of goods and services”, the Uganda Debt Network commented in May 2001. “A high proportion of grand corruption cases result from procurement” the Ministry of Finance, Planning and Economic Development confirmed in its latest Poverty Reduction Strategy Paper Progress Report. 23 The World Bank in several instances expressed its views contrary to the development of Dabhol. 24 Rimban and Samonte-Pesayco (2002). 25 Delgado, Esteban. Diario Libre. Republica Dominicana. August 26, 2005. In 1995, the state power utility TANESCO signed a PPA Confidentiality of the PPAs with investors for a 100 MW diesel-fired power plant. At the same time, a natural gas fired project was being The issue of confidentiality of Power Purchase Agreements evaluated, to be partially financed by the World Bank. compounds the absence of competitive procurement. The two projects together resulted in excess capacity. Developers adamantly claim that the PPA contains Furthermore, there was a suspicion of malpractice in the commercially sensitive information and therefore should contract signed with the investors, which triggered a not be disclosed.26 The argument for disclosure, so many lengthy arbitration procedure. Box 6 provides details times rebuffed by signing parties, is that the PPA is not a about this judicial and contentious renegotiation process. contract between two private parties, but commits a government, when providing guarantees, and consumers who are ultimately the ones who will pay the tariffs, over Box 6: The Tanzanian IPTL Project a long period (typically 15 to 30 years). After extended legal battles, the project, There is a growing trend towards making those PPAs as Independent Power Tanzania Ltd. (IPTL), started transparent and open to the public as possible: supplying power to the national grid in January 2002. The World Bank’s International Centre for the • In the case of the Bujagali project, the World Bank 12 Settlement of Investment Disputes ruled that Inspection Panel argued that “full disclosure of the PPA Tanzania was entitled to reduce the payments under is vital if the intent is to place the public in a position the PPA signed by a previous administration. The to analyze, understand, and participate in informed PPA, as originally signed, agreed paying actual costs discussion about viability of the project and its impact incurred by IPTL, which turned out to be a huge burden to the country’s economy. On August 31, on the economy and well-being of Ugandans;” 2001, Tanzania’s Finance Minister wrote in a Letter • On November 12, 2002, the Uganda High Court of Intent to the International Monetary Fund: ruled that the PPA of the Bujagali hydropower project TANESCO defaulted on the PPA and international be released to the public; arbitration was called. • In 2002, the Supreme Court of Belize ordered the public release of the PPA for the Chalillo hydropower “TANESCO's financial problems will be further project on the Macal River; aggravated by new financial obligations following the completion of the arbitration case concerning the costs of the Independent Power Tanzania Limited (IPTL) power plant. (…) Under normal circumstances, there will only be limited use of IPTL's capacity for some years to come and, presently, TANESCO lacks the resources to make these payments. (…) The government will allow TANESCO to increase the average electricity tariff to accommodate partly the IPTL's financial requirements and compensate for inflation since the last increase in 1998. (…) The above measures notwithstanding, TANESCO will require TSh 26 billion in support from the budget during 2001/02.” “In the final analysis it appears that IPPs have often inflated supply prices for utilities,” the World Bank concluded in a Public Policy for the Private Sector Note in December 1998. Source: World Bank Analysis 26 In some cases, where the pass-through of power purchaser costs does not depend on the contractual prices, developers are even reluctant to submit their contracts to regulatory scrutiny. Rigid contractual and institutional arrangements purchases.”29 Ensuring reliability and maintaining system reserves entails fixed costs and these costs will have to More often than not, long-term power purchase be paid for in one way or another. The pure Single agreements have proved to be very rigid and inflexible, Buyer model poses the risk of aggravating this problem, leading to stranded costs and creating obstacles to the as a result of potential lack of discipline and perverse migration towards a competitive power sector.27 incentives in preparing load forecasts and contracting for new capacity. The efficient design of a PPA requires that energy price component should correspond to the plant variable From an institutional standpoint, long-term contracts generation costs, and fixed costs (e.g. investment) with high energy take-or-pay obligations are a hindrance should be part of the capacity payment. However, to to the development of a least-cost dispatch mechanism reduce market or dispatch risks, developers prefer to or efficient use of available energy resources, let alone charge most of their expected margins to the capacity of a competitive market.30 The rationale is as follows: If payment itself.28 new, cheaper generation is built, it will be not able to compete with the existing, high volume take-or-pay In part, investors fear that they are in the hands of one energy contracts, with artificially set, very low marginal 31 Single Buyer, whose behavior in dispatching the system costs. As a result, efficient plants (with lower variable and in planning future expansions is not perceived as costs) will not be dispatched while less efficient plants 13 absolutely neutral or professional. Therefore, eliminating will be instructed to run. As a result, “cheaper” plants the dispatch risk (or transferring it to fixed payment) are not in a position to challenge more expensive seems to be the right thing to do, transforming the PPA existing ones if there are energy take or pay provisions into a de facto” take or pay” contractual arrangement. or clauses that oblige dispatch as base load generation. Mitigating the dispatch risk becomes necessary in a The Single Buyer model in tandem with high “take of Single Buyer model where a monopoly buyer makes pay” clauses responds poorly when electricity demand decisions for new generation entry, generation and falls short of projections (e.g. if there is a macroeconomic dispatch scheduling. More often than not, the Single crisis or capacity turns out to be much higher than Buyer overestimates demand growth, resulting in excess necessary). Ideally, electricity prices would fall, stimulating capacity and aggravating the above problem. demand, and revenue losses would be allocated to private financiers, best equipped to manage market risks. Under the Single Buyer model, however, wholesale electricity prices rise because take-or-pay quotas must be spread over a shrinking volume of electricity 27 This is not a judgment as far as the caliber of the financial advice provided to host governments is concerned. It simply reflects that energy in developing countries is to a large extent a “seller’s market.” Energy is in short supply and reluctant investors are in a good bargaining position to shift risks to the off-taker, particularly in those countries facing chronic power shortage problems. 28 An interesting example is a PPA originally signed between the Cuiabá Power Plant (formerly developed and operated by Enron) and Eletronorte, a state owned company in Brazil, which played the role of a quasi-Single Buyer operating on a regional basis. Despite being the result of a competitive process, bidding documents were not clear on the mix between fixed and variable payments. Having more expertise and commercial acumen than its counterpart, Enron’s proposal included a minor part of the payment as variable, as low as US$ 2-3/MWh, and was awarded the contract. The developer became virtually hedged against dispatch risks, a critical feature for a plant to be commissioned under a very uncertain regulatory scenario and also behind a transmission constraint. A similar situation arose in Panama, where price offered for energy was zero and capacity very high. This is a safer strategy for investors in the early stages of market reform, where the rules of the game are still being defined. 29 Laszlo Lovei, a World Bank lead expert and one of the first professionals to call the power sector’s attention to the long terms problems introduced by the Single Buyer. 30 Energy take or pay contracts establish an energy volume that must be paid independently on whether it is generated or not , as long as the Generator complies with a level of availability agreed in the contract. As the take or pay energy must be paid, it is considered for dispatch as zero variable cost and dispatched, even if actual generation variable costs are higher than other generation that is displaced by the take or pay generation. Some PPAs have included not only take or pay clauses but also must run provisions (e.g. mandatory baseload dispatch of available contracted generation). 31 The marginal cost of a generation plant with a take or pay contract could be considered zero until the volume of the contracted take or pay is reached. While some power pools allow generators to bid a very low marginal cost (“inflexible generation,” others force generators to bid the actual cost, not to distort competition and energy efficiency. This has put pressure in renegotiating and reducing the take or pay clauses. Contingent Liabilities Transferring risks and contingent liabilities to a third party (off-taker) distorts an economic comparison The introduction of a Single Buyer model often leads among multiple supply options. Box 7 illustrates the to the build up of significant liabilities to the host issue at hand. government and to their taxpayers. On the one hand, if the PPAs signed with a Single Buyer require sovereign guarantees, the government will backstop the payments Box 7: Comparing Contract Economics With for capacity and energy, in case the utility is not able Different Risk Profiles to do so. Given the financial distress in of many distribution companies, (including the Single Buyer’s A frequent problem faced by governments is how to lack of creditworthiness), those guarantees happen choose among a wide range of supply options, each more often than it would be desirable. one with a specific contract risk profile. For example, which one is a more advantageous option: A contract In the case of the Unbundled Single Buyer model, requiring a return on equity (ROE) of 18%, where the developer bears a significant portion of the market merging the wholesale purchasing function with and dispatch risk, or a capacity contract with an transmission wires creates a potential liability for the implicit ROE of 14%? Or perhaps a “back to the old transmission company far greater than its asset base. days”, cost-plus regulated contract with ROE of 9%. 14 It may jeopardize the firm’s capacity to finance the wires Which one benefits (or hurts) the poor most? expansion, therefore increasing the risk of transmission congestion and, in countries with low electrification This question becomes relevant when the relative rates, limiting access. attractiveness of PPAs, whose risks have been fully loaded to the energy purchaser or to the If a Single Buyer contracts excess capacity, it represents government. Needless to say, those ROE’s (or tariffs a burden on the utility and/or on the government, resulting from a competitive procurement process) particularly in light of the rigid contractual arrangements are not comparable at face value, since the contracts described in the previous section. A few examples entail very different risks for the off-taker, customers and taxpayers. As a caveat, great care should be illustrate this point. taken when analyzing the attractiveness of deals or PPA’s whose risks have been taken away from Indonesia’s national power utility had a capacity of developers and disproportionately transferred to the 14,000 MW in 1998. After the Asian financial crisis, customers off-taker, customers, the government, and this was sufficient to fulfill the country’s demand. Yet ultimately, the tax payers. since 1994, the Suharto government concluded Power Purchase Agreements for private power projects which A transaction which may look financially attractive to would add another 11,000 MW to the system. During the customer (e.g. reasonable tariffs or embedded the economic crisis, low cost state-owned power plants ROEs) may be distorted by a biased distribution of had to be stopped in order to absorb the much more risks and contingent liabilities. This is particularly expensive private power. concerning in countries with very low market penetration rates, where shifting risks from the customer to the taxpayer may eventually result in a In the case of Enron’s Dabhol power plant in very regressive taxation, with significant impact on Maharashtra,32 an official investigation committee the poor. found that the power demand projections that justified the project were “based on extremely overoptimistic An evaluation of contractual options has to take those assumptions”. India’s Central Electricity Authority warned factors into account, at least on a qualitative basis. that Maharashtra’s power utility would need to stop plants generating power at 0.50-0.80 Rupees/kWh in Source: Authors’ analysis order to buy power from Dabhol at 3.47 Rupees/kWh, because of take or pay provisions in the PPA. 32 Eventually costing 7.8 Rupees per kWh. Inability to eliminate risks for developers as Box 8: Issues Regarding the Indonesia Single Buyer expected IPP Program Governments believed that the Single Buyer model was During the 1990s, escalating electricity demand and a “risk-free” proposition for customers and for developers. limited public resources throughout East Asia left Long term, hard currency denominated PPAs, which many countries little choice but to invite foreign hedged investors against market volatility, in tandem investment in new power plants. While the programs with sovereign guarantees, seemed to be an ideal brought positive results by reducing (or in some situation for private investors. cases eliminating) power shortages, and while it is difficult to quantify the cost of power shortages had the IPP programs not been implemented, they also The model was touted as a perfect solution for attracting created serious problems that were exacerbated by generation that would compete for the market. Its the regional financial crisis: advocates alleged that investors would be willing to accept lower rates of return for the capital invested, (a) In some countries, costly power shortages were as illustrated in Box 7. As a corollary, tariffs would be reduced, but in many cases more capacity was lower. It seemed to be a win-win situation. added than was needed, and some corruptly benefited from lucrative transactions. However, reality proved to be quite different. In some 15 cases, an environment conducive to corruption and (b) Governments often protected the IPPs against nepotism inflated prices. Investors were not always market risks (and sometimes fuel supply risks) through convinced about the alleged macro-economic stability; long-term take-or-pay contracts, and created huge nor could they possibly count on the long-term sanctity contingent liabilities for the countries, particularly since the currency risks were often covered by of contracts resulting from sweetheart deals. Therefore, indexing power purchase prices to hard currencies. high rates of return were in many cases embedded in the contract prices.33 (c) While subsidies for capital costs were reduced, the IPPs sometimes increased the Government- On the government side, overoptimistic expectations owned utilities’ supply costs, often through of modest rates of return (which in some cases did not uncompetitive bidding practices. materialize) ignored the fact that in a typical Single Buyer arrangement most risks - such as demand (d) Although IPP programs were intended to break up volatility, market prices, exchange rate devaluation, sector monopolies, the terms of their PPAs often created and stranded costs - were fully allocated to the off-taker. rigidities that caused the systems to operate inefficiently In certain countries, such as Thailand, Malaysia, the and complicated the sector’s further liberalization. Philippines, and Indonesia those risks ended up These issues were particularly pronounced in representing billions of US dollars. Indonesia, where 26 primarily US$-denominated PPAs were signed between the IPP sponsors and PLN In some cases, prompted by macroeconomic crisis, for about US$18 billion, which added roughly governments defaulted on their PPA obligations, and 11,000 MW of capacity. Funds for these IPPs were investors had to post significant losses in their books, secured from international sources, but even for agreements with apparently “rock-solid” predominantly from unsolicited proposals. Although sovereign guarantees. With the benefit of hindsight, it the Government did not issue explicit guarantees, was naïve to assume that countries hit by major crises “letters of support” were given to the IPPs through would honor their contractual obligations, particularly in which the Ministry of Finance (MOF) or the Ministry cases where the deals were conspicuously lopsided and of Mines and Energy (MEMR) required PLN to subject to heavy public criticism. In sum, the risk profile of perform its obligations. As the result of the crisis and the Single Buyer model was grossly misinterpreted, from devaluation of the Rupiah, the Government—faced with huge debt—postponed some IPP projects and the point of view of developers, customers and taxpayers. directed PLN to reimburse only part of its obligations 33 The expected rate of return for the Dhabol project, for example, was about 24% p.a. Box 8: Issues Regarding the Indonesia Single Buyer 4. EXPANDING THE CONCEPT OF SINGLE IPP Program, con’t BUYER to the operating IPPs. Investor response was mixed, Variations of Single Buyer Models depending on whether each took a short or long-term view of its involvement; in the most extreme cases, the Recent studies carried out by the World Bank have shown34 Government was sued and in one, an international that there is not a single way to implement what has arbitration panel ruled in favor of the IPP. imprecisely still referred as “the” Single Buyer model. The concept has been over stretched in the last few years. The Government has settled or renegotiated all the disputes with the IPPs, excluding the case mentioned For example, the entity referred to as the “Single Buyer” above, and has done a commendable job in in Austria encompasses more than one wholesale buyer. renegotiating prices, keeping its commitment and In July 1997 the Austrian Federal Economic Minister protecting the contracts. However, at least two aspects of the agreements have created serious problems. proposed a draft for electricity reform with several They have produced a huge contingent liability for the “Single Buyers”, corresponding to the existing 15 Government and the PPAs have captured, for the next provincial utilities, the 5 municipal utilities and the 30 years, a sizeable share of the power market under national electricity wholesale company. 16 a set tariff and “dispatchability” level which heavily affects the future market. In other cases, what is referred to as a Single Buyer is not even a “buyer,” since it does not necessarily own Source: World Bank (2004) the energy traded. Its responsibilities could be more appropriately described as a “broker” or load aggregator, rather than a single, monopolistic agent, since there may Several large-scale PPA programs, based on Single be multiple buyers and sellers in the market. Buyers, ended up failing. As an example, Box 8 illustrates some of those aspects observed in the In Italy, the Acquirente Unico, is also referred as a development of the IPP program in Indonesia. “Single Buyer,”35 despite the fact that: (i) it is involved in the procurement of energy only for small customers who did not switch suppliers (a condition called, default supplier, in an environment of full retail competition); and (ii) discussions were going on as to the merit of having the Single Buyer acquiring energy, in light of the likely creditworthiness of distribution companies. Single Buyers may also be introduced under a great variety of industry and commercial arrangements. Given this diversity, what are the common features among those models referred as “Single Buyers” The common thread is the fact that all those models, in one way or another, entail some form of centralized purchasing arrangements. That is, there is a mechanism and in some cases an organization responsible for some form of coordination in the procurement and contracting of power. It refers to an arrangement where the coordination of all, or a significant part of, the wholesale power procurement is carried out by an organization through a codified process. 34 See references at the end of this report, documents related to the Single Buyer between 2003 and 2004. 35 NERA. David Salant. The Acquirente Unico in Italy. The power procurement activity, in a broad sense, may This distinction leads to two different possible roles encompass a series of functions, including: played by a central agent – active and passive. • Designing standard conditions for bilateral contracts • Active: The central agent takes ownership of energy or PPAs;36 and payment risk. In this case, the agent is the bulk • Drafting bilateral contracts or PPA for a specific purchaser and reseller through back-to-back contracts procurement; of bulk supply. It receives payments from distribution • Administering the procurement process (procurement companies, exporters, and eligible consumers, and documents, auction, contract awarding, etc ;) makes payments to bulk sellers (generators, importers). • Negotiating contracts; • Passive: The central agent (or process) does not • Assigning or signing contracts with distribution directly buy energy to resell, but instead acts as a companies; load aggregator, and intermediates purchases or • Administering long and medium term power conducts the administration of power procurement. It purchases.37 may also assist distribution companies or other purchasers in contract invoicing and contract disputes. Taking “Title” For the Energy – A Fundamental Aspect The difference between active and passive agents is a fundamental aspect as far as the performance of the 17 A “pure” Single Buyer, by definition, takes the ownership procurement function is concerned. Many of the flaws title for the energy traded. It signs PPAs with and criticisms of the pure Single Buyer happen by virtue independent producers and re-sells this energy to of its role as an active buyer. As the name suggests, in distribution companies. this case, a central agent has more control and ownership of the process of procurement, which lends However, there may be variations of this purchasing itself to be more centralized and secretive. Also, Single arrangement where no central agent takes the title for Buyers take price and market risk, which may result in the energy traded. In such centralized purchasing significant financial imbalances between generators and arrangements, the central agent’s role in this case is to the Single Buyer – to be passed through to customers aggregate demand from multiple distribution and /or to taxpayers. companies, coordinate the energy procurement process, and allocate energy contracts volumes among the off- When distribution companies are the buyers, under the takers (usually distribution companies). orchestration of a centralized purchasing process, some of those problems may be mitigated. For instance, This demand aggregation role has been observed in procurement tends to be more transparent, as there are some States in the US, New Jersey being one of the more players involved in the decision-making.38 pioneers. The idea is also gaining momentum in some Moreover, distribution companies have a better developing markets, such as in Brazil. As of 2004, it is knowledge about their markets and are not willing to mandatory that all the energy procured by Brazilian over contract, if they will fully bear the financial costs of distribution companies be acquired via an auction-like their good or bad decisions. process, coordinated by the Ministry of Mines and Energy and approved by the regulatory authority. Those observations are of directional nature, and by no means should be construed as prescriptive. The real flaws or successes of any energy commercial arrangement depend on the design details and on careful implementation. 36 This is a separate function that, as defined in this paper, does not involve in competitive Electricity Markets the market operator. 37 The expression “long and medium term contracts” refers to all bilateral contracts agreed in advance. Trading (purchases and sales) in a Day Ahead Market or Balancing Market are considered in this document as short term. 38 As already pointed out, the success and behavior of a procurement agency depends not only on market design issues, but also on structural ones. Therefore, a pure Single Buyer, even taking the title for the energy, may still circumvent some of the corruption problems if there is a strong will by the government to work on a transparent, efficient and competitive energy purchasing process. Possible Functions Carried Out by a Central 4) Power system operator41 Agent • Performing transmission and generation maintenance outage coordination In the Single Buyer models the central agents are • Scheduling and dispatching generation. accountable for more than buying and selling of energy, • Administering and purchasing ancillary services if they do indeed perform those functions as part of • Balancing the system and managing congestion centralized purchasing arrangements, and may carry out • Administering the allocation of international many other responsibilities, which are usually bundled in interconnection capacity vertically integrated companies. 5) Market operator42 Empirical evidence has revealed that the trading function • Administering spot market or a balancing may be associated with one or more activities. The level mechanism of bundling on these activities depends on the market • Administering commercial metering and contract and industry structure in place. Some of those most settlement important associated functions include the following: There is no ideal blueprint on how those functions 1) Transmission planning39 should be grouped and which ones should be carried 18 • Planning transmission expansion and upgrades, out by a central agent under a centralized purchasing including international interconnections arrangement. At the early stages of power sector reform (e.g. a Single Buyer that is an integrated utility) the 2) Provider of transmission services central agent is assigned many functions in addition to • Providing transmission services on existing grid, the core functions of buying and selling. Those including international interconnections additional functions may include capacity and • Expanding and/or upgrading the transmission transmission system planning, dispatch and system system operation, maintenance coordination of transmission • Administering third party access and generation assets, energy trading, and many others. • Submitting and implementing regulated Table 1 illustrates a typical “unbundling” of those transmission charges functions as power sector evolves. The cells marked with • Managing congestion and losses (optional)40 a cross show the functions typically assigned to various central agents, demonstrating how the number of 3) Planning generation expansion and cross border functions goes down as the sector is restructured and trading moves towards competitive market arrangements. As the • Forecasting load requirements of each table shows, at the early stages of power sector reform, distribution company and/or of the system as a the central agent is responsible for carrying out all or whole most of the functions described above. As the sector • Evaluating technical and least cost expansion evolves, there is a tendency to unbundle the power alternatives sector. When wholesale competition is possible, with multiple buyers and sellers, the role of the central agent ends up being significantly reduced, focusing on the purchasing of electricity and management of the resulting contracts. At later stages of power sector 39 Transmission expansion per se (and/or ownership of transmission assets) is a transmission company typical responsibility, not a function of the Single Buyer or of any entity in centralized purchasing arrangements. Even if both roles (transmission and wholesale purchasers) are combined into one company, it is important to establish a functional separation between the buying/selling of energy and running the transmission business. Approval of, or regulation on decisions regarding new transmission investment, should be assigned to the Regulatory Authority. 40 If the transmission service provider is merged with system operator, the result is a Transmission System Operator (TSO), sometimes also referred to as a “Transco” type of organization. – Specific regulations will establish who bears the cost of transmission losses and congestion and how those costs (or recognized costs) should be recovered. 41 Two functions have been differentiated, system operator and market operator, depending on electricity trading arrangements. However, both functions may be combined in one: a system and market operator, as has been the case in the US. 42 Depending on the market design, those responsibilities may belong to the system operator. There is also some ambiguity as far as the correct definition of market and system operation is concerned, but such market design considerations goes beyond the scope of this paper. Table 1. Evolution of Functions of Central Agent VARIATION OF CENTRAL INTEGRATED PURCHASING SINGLE RESTRUCTURED WHOLESALE RETAIL PROVIDER OF AGENT BUYER SINGLE BUYER COMPETITION COMPETITION LAST RESORT Status Active Active Active Active Passive Role Integrated Transmission Market Wholesale Procurement utility Company Operator Trader Agency FUNCTIONS Integrated Utility x Transmission Planning x x Transmission Services x x System Operation x x Generation Planning x x 19 Market Administration x x x Purchaser of Electricity x x x x Administrator x x x x x Procurement of Contracts/PPAs development, the central agent becomes a passive of the table show the country and company/entity that buyer and no longer acquires energy in the market. originally operated as the central agent prior to the Instead, it is solely responsible for administering and reform, respectively. In the second row, if the cell with reassigning contracts among distribution companies. the entity name is blank, it means that prior to reform, That is the case, for example, in countries with full retail there was no entity with a centralized purchasing role, as competition where the utility plays the role of provider of in the case of Brazil and Italy. The cells with check last resort. marks show the functions assigned to the original central agent. The blue-shaded cells with check marks Worldwide experience shows that there are also multiple correspond to the functions that remain with the central ways of organizing those functions, ranging from a agent in the new model, while the un-shaded cells with complete merging of all responsibilities under a single checkmarks mean that a particular function is now organization, to a situation where the central agent is assigned to another entity, instead of the original central solely a broker. Table 2 illustrates this great variety. agent. The cells with crosses identify the functions assigned to a newly-created central agent that did not The responsibilities assigned to central agents change exist prior to reform. as power sector reform evolves. Proposed or ongoing reforms in several countries’ power sectors envision a Tanzania is an illustrative example. Currently, TANESCO, gradual “unbundling” of central agent functions, resulting the vertically integrated utility, plays the role of an active in less responsibility and more specialization. Table 2 Single Buyer. The new sector design envisages the provides a few examples from different countries, showing establishment of an unbundled administrator of long and how the allocation of functions is in a state of flux, as medium term purchases, whereby contracts are signed power sector reform progresses and the centralized directly between the IPP and Distribution Companies. purchasing roles are consolidated. It uses symbols and color-coding to illustrate the change. The first two rows Table 2. Proposed Centralized Purchasing Arrangements HONDURAS INDONESIA CAMBODIA TANZANIA THAILAND SICHUAN VIETNAM UKRAINE JORDAN NIGERIA BRAZIL ITALY Entity TANESCO Dispatch Energo- NEPCO market Center (utility) (utility) (utility) (utility) (utility) (utility) (utility) (utility) NEPA EGAT ENEE SPEC EDC EVN PLN FUNCTIONS Transmission Planning ✔ ✔ ✔ ✔ ✔ ✔ Provider of Transmission ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ Services Planning Generation ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ Expansion and Cross Border Trading 20 System Operator ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ Market Operator ✔ ✔ ✔ ✔ x Administrator of Long ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ x x ✔ and Medium Term Purchases Purchaser of Electricity ✔ ✔ ✔ ✔ ✔ ✔ ✔ x ✔ Table 3. Perceived benefits of the Single Buyer model that can be preserved Perceived Benefits of the How a Centralized Purchasing Single Buyer Model Arrangement Can Still Contribute Ability to attract PPAs can still be offered, providing the necessary comfort for investors to capital by offering build power plants. However, some centralized purchasing arrangements do long term PPAs not require central ownership for the energy, neither the Single Buyer needs to be granted a de jure monopsony/monopoly status Being a workable Similar results can be achieved, as long as credit, market, and regulatory alternative for distressed risks are minimized, at least at the critical initial phases of the power reform power sectors to initiate reform Ability to centralize A centralized purchasing arrangement can equally include mechanisms to purchases, developing rapidly develop procurement capabilities, achieve economies of scale, and expertise and benefiting start disseminating this expertise among the potential future players, as from economies of scale conditions permit Ease of implementation, Ease of implementation is potentially deceiving, as new problems may result in terms of not requiring in the future, such as stranded contracts. Less opposition from the fundamental changes in incumbent utility is not necessarily a plus. As a minimum, the buyer has to sector structure, and the develop competitive, transparent procurement mechanism. consequent acceptability for the prevailing culture 5. MITIGATING SOME PROBLEMS OF THE Box 9: Critical measures for mitigating problems SINGLE BUYER WHILE PRESERVING associated with the Single Buyer ADVANTAGES 1. Organizational structure and key procurement This section presents some practical considerations that processes should be as transparent as possible could improve the Single Buyer model and make it more efficient. Some modified institutional and commercial 2. The purchasing agent should not be granted a arrangements, can address some problems associated monopoly status with the “pure” Single Buyer model and at the same time preserve many of its perceived benefits. 3. The purchasing agent should not take ownership for the energy traded, and instead act like a Benefits to be preserved broker, when market conditions permit This section addresses some of the key perceived 4. Roles and responsibilities of the purchasing agent should be unbundled to a reasonable extent in benefits of the Single Buyer that one wants to maintain order to enable some checks and balances by preserving some form of centralized purchasing arrangement. This alternative form may (or may not) 5. More flexible contractual arrangements should include a separate entity in charge of procuring energy be utilized 21 and/or taking the title. Table 3 summarizes the benefits of the Single Buyer and how a more flexible central 6. There should be a game plan to allow power purchasing arrangement can still preserve most of sector reform to evolve those benefits. 7. Sovereign guarantees should be granted on a Mitigating Some Problems of the Single Buyer selective basis Model 8. There need to be clear, codified rules to assure transparency, accountability and good performance The problems associated with the Single Buyer model can be mitigated with measures designed to incorporate greater flexibility, transparency and a longer term vision into the power sector reform process. Some of the design A Sound System Planning Process factors critical for success are described as follows. A good planning process is a very important step in Organizational structure and key procurement determining the need for additional generation capacity processes should be as transparent as possible within reasonable load projection ranges. It encompasses the selection of potential sites, sequencing of plants, This paper underscores the importance of transparency and main technologies to support expansion. It is meant as an antidote to cronyism and corruption. Procurement to be an analytical process, but as already pointed out of new generation encompasses several phases and planning can be interfered with by politicians and starts with power system planning, when load is forecasted others. Therefore, efficient and adequate planning may and preliminary least cost target projects are determined be a necessary, but not sufficient, condition for a to achieve generation adequacy. Once a portfolio of transparent and sustainable power sector. preferred projects is selected, the next step is to choose the most efficient among those – ideally via a competitive, In countries at the early stages of power sector reform, technology-neutral procurement type of process. planning activities are usually under the responsibility of the government owned utility or are sometimes under Transparency, the presence of codified rules such as a Grid the purview of the Energy Ministry. However, many of Code, due process, regulatory independence, and many the poorest developing countries do not yet have other attributes are essential to assure that the projects and sufficient capacity in power sector planning. They lack sellers are selected on their own merits. A transparent the skills, software models and processes, and often mechanism should contain all the following elements. must rely on consultants’ or on donors’ advice on an ad-hoc basis. Unfortunately, the advice may be biased Box 10: Dominican Republic – Lack of an towards some particular technology, project, or source Integrated Planning Approach of funds. As an attempt to alleviate chronic power shortages If there is a structured and elaborate planning capability, that had been plaguing the Dominican Republic for it is less likely that unsolicited, single sourced projects many years, the state owned company, acting as a will emerge. Or if they do, it becomes more conspicuous. quasi Single Buyer, decided to sign several PPAs with Countries that have no tradition of carrying out least new private developers. More than US$1 billion was cost planning on a continuous basis, may not be capable invested, including a natural gas terminal and of assessing whether or not a particular unsolicited natural gas fired power plants. Today, the country project is indeed the best alternative for its power system, has significant excess capacity, but is still facing daily taking into account other potential generation alternatives, black-outs. Despite all the capacity made available, impact on transmission investments, and interconnection generators are not able to collect money from to neighboring countries. It is easy for a country to fall distribution companies, which, in turn, complain about theft, rampant fraud and non-payment. It is a victim to suboptimal options under those circumstances. so called “financial black out.”43 More than 500,000 customers are not metered. To aggravate matters, A sound planning process should also take into account there is a culture of non-payment. Both factors have 22 the demand side of the market. Demand should be an resulted in a wasteful use of expensive resources. integral part of the planning process, including, inter alia, considerations on price elasticity, energy efficiency, The approach followed to alleviating shortages was rational use of energy, affordability, and willingness to biased toward supply side solutions, which were pay. Those aspects are often ignored or addressed ineffective to deal with a much more comprehensive superficially in the overall plans. Demand forecasts do and intermingled problem of supply and demand not properly take into account the demand price elasticity imbalance. and therefore underestimate the importance of proper pricing on the energy conservation efforts. Box 10 Source: Author’s analysis. illustrates the specific case of Dominican Republic. In that country, a distorted tariff policy has led to explosive demand growth. The power system has been undermined by unsustainable subsidies, not necessarily targeted to planning and evaluation technical work. However, an the poor. Lack of revenues or collection in the sector integrated organizational structure does not offer the has resulted in chronic financial “blackouts,” despite necessary checks and balances to assure a transparent existence of excess capacity. process unless a consultation or public hearing process is included. Developing a sound system plan is a must. In tandem with this effort, donors and multilateral institutions should In more developed power systems, the central purchasing help poor countries prepare systematic and codified agent has a more limited role in system planning. It planning processes, spelling out who and when should serves to aggregate the demand forecast prepared by carry on those plans, how often they should be revised, distribution companies and to procure the energy what the kinds of public hearings or industry consultation volumes on a consolidated basis. However, the agent are required to assure maximum transparency, who has does not “second guess” demand forecasts and does the necessary approval authority to alter the plans. not necessarily decide on the best generation projects. Generation investors will compete to provide the At the early stages of power sector reform, developing cheapest energy and procurement will take place on a the utility’s system planning internal capability and competitive basis. This subject of competitive enhancing in-house planning skills is a crucial step and procurement will be described in the next section. should be considered a sector priority. A central purchasing agent could in principle carry out most of the system 43 See Maurer, Pereira and Rosenblatt. (2005) Competitive Procurement for Power Box 11: World Bank Views on the Need for Competitive Procurement This report considers a competitive procurement for power (capacity and energy) as the single most important The World Bank, in a Public Policy for the Private Sector step to mitigate many of the problems associated with Note of December 1998, acknowledged that “bidding the pure Single Buyers. seems to have reduced PPA prices by 25 percent on average”, and recommended “discouraging the In principle, the procurement of all energy by a central widespread practice of noncompetitive procurement agent (in particular by a monopolistic Single Buyer) of goods”. “[P]rovided that sufficient interest can be should be carried out through some form of competitive attracted from bidders, governments, and utilities mechanism. This statement holds true both for new, can obtain better terms for the host country under greenfield plants, as well as for existing generation to competitive bidding for proposals from IPPs than be contracted or re-contracted. Exceptional situations under noncompetitive negotiated deals.”45 apart, such as natural disasters and the like,44 there Competitive bidding also reduces the risk of political should always be room for competitive procurement. resistance against projects. “The transparency of the This proposition becomes especially important when the process by which a project concession is awarded also entity at the core of the central purchasing arrangement is a strong determinant of the level of political risk takes title of the energy, serves captive consumers, faced by the sponsors. In general, competitively bid 23 and/or acts as a supplier of last resort. concessions face low levels of political risk and command better financing terms in the marketplace.” Many professionals from the World Bank have been strong advocates of competitive procurement, as In a World Bank study published in 2003,46 Arizu, illustrated in Box 11. Maurer and Tenenbaum, pointed out that a competitive bid is the best methodology to assure Several countries have also realized the importance of transparency and fairness in the energy procurement competitive procurement and have made some progress process. Additionally, it brings an elegant solution to the complex issue of pass through of the power in this direction. The first area of progress was to establish purchase costs, since prices resulting from the the principles or conditions under which a central agent competitive bid should be the basis for pass-through. should be allowed to procure competitively or directly Lastly, a competitive bidding process, if carried out (on an exceptional basis). on a level playing field basis, mitigates many of the problems resulting from self-dealing transactions. In Cambodia, the power sector model was designed For all those reasons, the authors strongly based on acknowledgement of the fact that unsolicited recommend the use of mandated competitive projects would exist for some time. Under the current procurement for the great majority of the energy market design, generation procurement guidelines acquired by distribution companies. The regulatory establish how unsolicited projects should be administered. agency should provide the basic guidelines for the According to this design, unsolicited projects are auction, and monitor the process to ensure fairness included as alternatives in the system plan and are and transparency. selected if they correspond to the least cost option. Source: Authors’ analysis. This allows for transparency and least cost decisions in new generation entry. 44 A shortage crisis should not be an excuse to waive competitive procurement either. Except in a few situations, where load shedding is caused by unexpected unavailability of existing power plants, rationing can be anticipated and prepared for. It is possible to carry out a competitive process on an expedited basis. That was the case in Brazil during the 2001-2002 rationing crises when the government acting as a central purchaser for thermal back-up reserves, decided to establish 3-year contracts for about 3,000 MW of new thermal generation to help mitigate the crisis. Questions were raised about the high prices paid for this capacity, which simply reflected the tight supply demand balance at that time. The procurement process was transparent and competitive despite being carried out under a very short timeframe. 45 Robert Bacon and John Besant-Jones (2001). 46 Arizu, Maurer and Tenenbaum (2004). Some countries have adopted competitive procurement of private capital. In one of the schemes, the government on a regular basis. Thailand is an example of bids the rights to use a certain hydropower development competitive procurement processes which have been site. The winner is the developer who offers the highest used for more than 10 years. Details are described in price for the concession (similar to a royalty, to be paid Box 12. over the life of the concession). According to this scheme, the winner has the freedom to trade this energy Competition is also possible for the development of the way the investor pleases, either via long-term hydropower. Brazil, for example, has developed two contracts or on a merchant basis. More recently, Brazil alternative schemes that have successfully attracted has introduced a variant of the scheme, where the plant private capital and enabled competition. Reductions in site and an associated PPA are put for bid jointly. The cost and time of construction have been remarkable, as winner of a particular concession is the one who a result of competition process and of the participation promises to deliver energy at the lowest price. Recent Box 12: Thailand’s Experience in Energy Procurement In Thailand, state-owned Electricity Generating Authority of Thailand (EGAT) is responsible for generation and 24 transmission, while power is distributed by two state-owned entities. A 1992 Amendment of the EGAT Act ended EGAT’s monopoly over power generation, allowing private sector investment in power generation projects in the form of Small Power Producers (SPP) and IPP , both of which sell power to EGAT under single buyer modality. SPPs are small power projects that are either cogenerators or renewable energy, which can sell power directly to consumers and in addition sell to EGAT up to 90 MW for each project. EGAT has since 1992 used a competitive procurement process to buy energy from SPPs or from IPPs. Since its inception, 110 proposals have been received and 71 SPPs have been accepted and obliged to supply power. The bidding uses an objective and transparent scoring and evaluation process modeled after the PURPA system in the US. In August 1994, EGAT commenced solicitation of large-scale IPPs, for 1,400MW for 1996-2000 (Phase 1) and a further 2,800MW for 2001-2002 (Phase 2). In July 1996, the Phase 1 program was increased to 1,750MW and Phase 2 to 4,200MW and PPA (20 and 25 year) was signed with 7 IPPs with the combined capacity of 5,950MW. Economic slowdown and associated slowdown in power demand led EGAT’s power development plan to be adjusted downward. The last of the 7 IPPs, the BLCP project, with total financing requirements of US$1.4 billion reached financial closure in 2003 and was funded by US$400 million equity and US$1 billion debt without Government guarantee for EGAT contractual payment obligations under 25-year PPA. Private investors and commercial and multilateral/bilateral lenders/guarantors view EGAT credit favorably. EGAT is rated high-BBB, the same with the sovereign government, because of: proven strong support from the Ministry of Finance (which has previously guaranteed all foreign debt and 90% of domestic debt of EGAT; the existing regulations/ EGAT Charter obliging the government to cover cash flow deficiencies in the event EGAT cannot obtain from other sources); strong market position; automatic tariff adjustment mechanism which protects EGAT from increases in uncontrollable costs (fuel price, foreign exchange fluctuation, power purchases and inflation) and sound operating history. The legal system, foreign exchange markets and local financing markets are relatively well developed in Thailand, allowing limited-recourse project finance for large-scale projects. The Government has scrapped plans for the introduction of a competitive power pool due to opposition from utilities and unions and uncertainties about the complexity of pool operations; and decided to adopt the so- called Enhanced Single Buyer Model for the electricity supply industry which provides for: (i) EGAT continuing to be a sole buyer, dispatcher and bulk transporter of electricity; and (ii) EGAT and the state owned distribution companies introducing accounting separation for each of their business units and to be corporatized as a whole and listed on the stock exchange through IPO. Source: Adapted from Tomoko Matsukawa. World Bank (2005) Unpublished. auctions carried out in 2004, where most of the supply Box 13: Alternatives to introduce competitive was hydropower based, proved to be quite transparent procurement in hydro-generation and competitive. (a) Option 1: Administrative Allocation. The key When the power system in Brazil was being restructured feature of this option is that distribution and retail in 1997, several options were considered for competitive companies would be allocated particular sites by procurement to build hydro plants. These options, the Indicative Planner and given responsibility for illustrated in Box 13, may serve as workable alternatives undertaking competitive tenders for the concession to other countries. At that time, Option 1 was suggested contracts. The distribution and retail companies who as a transitional measure to ensure that new capacity tendered the concession would undertake to purchase continued to be built for the period before the new the energy from the concession holders. The bidder power sector model was implemented. Both Options 2 offering the lowest tariff per kWh of firm energy and 3 are competitive forms of tendering concessions.47 would be awarded the concession. A pre-determined standard concession fee would be payable to the Competitive procurement requires rules and skills, not Granting Authority. always readily available in some of the World Bank’s client countries. For example, designing an energy (b) Option 2: Simple tender to all eligible parties. auction is a very specialized task, and a lot of Under this option, the Granting Authority would 25 customization needs to be done in terms of definition undertake the tendering of concessions directly. All of objectives, specification of the product to be offered, interested parties would be eligible to bid for the contract preparation, auction design and process concession and would be responsible for identifying monitoring. Professional and independent views and a market for the energy. The bid would consist of the expert advice may be acquired. concession fee to be paid to the Granting Authority in each hear of the concession, which would be The World Bank has also supported some of its client equal to or greater than the standard fees defined countries to develop processes and organizations to by the Granting Authority. The winning bidder would be the one offering the highest present value of carry on competitive procurement for power. A recent concession fees. experience in Vietnam is described in Box 13. (c) Option 3: Purchase offers followed by concession Confidentiality of the PPAs tender. This option is based on the idea that ahead of the tender for the generation concession, distribution There is a long tradition in the power sector that energy and retail companies (and possibly eligible consumers) contracts are commercially sensitive and therefore should would make binding offers to buy energy from the not be publicly disclosed. This may be valid in the non- project. Generators would then bid for the concession franchised business,48 but not for a regulated company with known purchase commitments, on the basis of where the full costs and consequences of contracting are the highest present value of concession fees. The to be borne by the customers (and in some cases winning bidder would be free, but not obliged, to taxpayers). The long standing argument for contract take up the long term contracts offered by the confidentiality has been recently challenged in courts. distribution and retail companies. This would allow self-producers or IPPs with other potential markets to This paper shares the voiced concerns about contract bid for the concessions. confidentiality, but claims that there should be a better trade-off between commercial sensitive information and Source: RE-SEB Project, Coopers & Lybrand transparency. Contracts resulting from a competitive Consortium, Eletrobrás-MME. Brazil. (1998) procurement process that are paid by consumers (regulated tariffs) should, almost by default, be in the 47 Brazil ended up moving to Option 2 and subsequently migrated to a variation of Option 3. 48 Example of an IPP selling to free customers, at freely negotiated terms and conditions. public domain and there should be no room for further Box 14: Vietnam’s experience with competitive negotiation. If all energy is to be procured competitively, power procurement as a corollary, all contracts should be publicly disclosed. Electricity demand in Vietnam increased at about This should be the case even in a situation where the 14% per annum over the period of 1995-2000. government assumes that only a few bidders will The Government decided to promote private sector participate in the process. There should still be participation in order to address the need for contestability and the principles of competition apply. increased power generation. The World Bank assisted the Government to prepare and conduct Countries are moving in the direction of more bidding for the Phu My 2-2 (715MW gas-fired) transparency. In several cases where an Electricity power project under the BOT scheme, which sells Regulatory Commission exists, a new PPA or contract power to a state-owned integrated utility EVN under proposed by a central purchasing agent requires prior a 20-year PPA and is supplied gas from a newly- approval by the Regulator, and such approval is subject developed offshore gas filed via a state-owned PV to compliance with procurement rules or guidelines. under a 20-year GSA. The government issued the One interesting case is Cambodia, where the Regulator Request for Proposals in late 1997, and offered an International Development Association (IDA) Partial must assess and inform the Government and the public Risk Guarantee (PRG) backstopping government in general about the expected impact of a new PPA in 26 guarantees (covering Vietnamese contractual the final consumers’ tariffs, thereby creating a climate undertaking and foreign exchange availability/ for an open discussion on issues such as affordability convertibility/ transferability) up to US$75 million in and the risks that may be faced by the investor. support of commercial debt financing as an option to the bidders. Six strong international consortia The Single Buyer should not be granted a submitted bids and the EDFI-led consortium was monopoly status awarded the project in 1999. Key project documents including the BOT Contract were signed and the Single Buyers should not be given a de jure monopoly Investment License was issued in 2001. Financing status, even if competition is not workable at the outset. documents were signed in 2002 and the plant was It is possible to preserve the benefits of scale and commissioned in early 2005. expertise, observed in pure Single Buyers, even in the The project became the first infrastructure project in absence of a monopoly status. The market in which the country for which the sponsor was selected Single Buyers (or central purchasing agents) operate through competitive bidding and the largest foreign should be perceived as contestable, even if competition investment outside the oil and gas sectors. Tariffs is not workable at the outset. proposed by bidders reflected the Government undertaking, prospective availability of IDA PRG and The existence of a monopoly wholesale purchaser (as enhanced competition, and resulted in much lower present in both in the Integrated Single Buyer and the tariff level compared with what the Government had Unbundled Single Buyer) is one of the factors that have negotiated for other private power projects. The lead to the establishment of long-term (and sometimes closure of this deal heightened the interest of the inflexible) PPAs. Investors want to make sure that, once Government in promoting private projects in key the generation capacity has been built, the purchasing infrastructure sectors such as power generation. entity will not abuse its monopoly position. As a The Government hopes that the IPPs (including ones consequence, the PPA design has tended to allocate to be set up by the state-owned entities) will account most (or all) market risks, such as load growth, to the for over a half of total incremental power production purchaser or the government (through government and recently announced its intention to start preparing guarantees or implementation agreements). This can the next batch of BOT power projects for competitive lead to inefficiencies and eventually stranded costs as bidding to meet energy future demand needs. discussed before.49 Source: Adapted from Tomoko Matsukawa. World Contract design is about risk allocation. If there is only Bank. (2005) Unpublished. one wholesale buyer in the power sector, it becomes very difficult to create more flexible trading 49 This issue has been evaluated in different papers. References can be found at the end of this paper. arrangements or avoid allocating certain risks to the • In Brazil, the central agent is only responsible for the purchaser. Arrangements should be put in place that do coordination of energy procurement for the regulated not create a monopoly position for the Single Buyer. For market (captive consumers). Buyers and sellers in the example, even if a Single Buyer has a de facto monopoly non-franchised market (eligible customers) may trade position at the outset, nothing in the legislation should energy at unregulated, bilaterally negotiated contracts prevent the entrance of new players, sales by IPPs to and are not subject to the procedures of mandated non-franchised customers (e.g. large consumers competitive procurement. Brazil’s recent experience connected directly to transmission), and imports (or with energy procurement is described in Box 15. exports) carried out by third parties. Whenever practical for larger systems, a central procurement function The central purchasing agent should not take should be combined with a spot market, whereby ownership for the energy traded, and instead act electricity trading arrangements allow greater diversity like a broker, when market conditions permit and create price benchmarks, which may serve as a reference to future contracting.50 The role of the central agent as the owner of the energy traded should be reduced or completely eliminated over A clear power sector reform vision should be established time. Our preference is for a model in which contracts at the outset, to move the power away from a de facto are signed directly between buyers and sellers and the monopolistic buyer towards a scenario where other buyers central agent acts as a neutral coordinator of the 27 and sellers are invited and encouraged to participate. procurement process – or in other words, an “honest broker.” This has the advantage of mitigating the There are several possible ways to create this negative effects of corruption and kickbacks, oftentimes contestability in the market, including the following: present in large scale, obscure commercial transactions. • In Panama, the Transmission Company, acting as a Direct contracts require distribution companies to be central agent, was responsible at the outset for creditworthy, which may not be the case at the outset for administering the procurement of up to 85% of the many of the Bank’s client countries. However, it should Distribution Companies’ needs. Anything above this be part of the vision in establishing a centralized threshold could be contracted directly by the purchasing arrangement. Distribution Company from third parties or alternatively through the administration of the central Some practical alternatives to implement this kind of agent. This obligation of the Distribution Company “hands-off” scheme include: lasted for a transition period of five years after which the role of the Transmission Company as centralized • In Ukraine, there was a mandatory pool based central purchasing arrangement ended; agent. This situation avoids many of the problems • In Indonesia, the Single Buyer is designed to be associated with having an active buyer. The phased out on a regional basis (as opposed to disadvantage of this approach is that it discourages volumetric reductions) as its tariffs become cost the establishment of long-term contracts, which also reflective and the energy is subsequently traded in a serve as important financial hedges. In the absence of competitive electricity market; those hedges, distribution companies do not have the • In Mexico, IPPs are free to export energy to the U.S. instruments to manage price volatility and may Self-generation is also encouraged. Despite its experience serious financial difficulties, such as the appearance as a Single Buyer, CFE, the state ones observed during the California crisis. integrated utility, in practice does not operate as such; Furthermore, it increases credit risks for generation companies, since they will not have a contract with a specific entity which may be taken to court. Hedging contracts of various durations, either physical, financial or a combination of the two, should be an integral part of any sustainable market design. 50 With the understanding that spot prices are more volatile, as they reflect short-term marginal costs, while long term contracting is meant to reflect long run marginal costs. Box 15: Recent Brazilian Experience in Energy Procurement (Auctions) Until 2003, Distribution Companies in Brazil were allowed to freely negotiate contract terms and conditions with their bulk energy suppliers for the procurement of energy. Self-dealing was permitted up to a certain level of their energy requirements (20%) but derogations had to be established down the road to facilitate the privatization of state owned generators. Distribution Companies had a mandatory requirement to acquire at least 85% of their energy needs in forward contracts with durations of two years or more. The balance could be acquired in short- term contracts or in the spot market. There were also regulatory limits (administrative caps) set on the total costs of energy that Distribution Companies were allowed to pass through to the regulated tariffs of their captive customers, with some incentives for efficient procurement. Those caps were not free from controversy. After the 2001-2002 energy crisis, when 20% of energy consumption had to be rationed almost on a country- wide basis, the existing procurement system went through a major review. It was acknowledged that it was not conducive to long term contracting, which in turn did not enable generation expansion (oftentimes cited as one of the culprits of the rationing crisis). Furthermore, in a desperate attempt to bring more capacity during the crisis, the former administration was more flexible in terms of allowing self-dealing and increasing the pass-through of generation costs caps to regulated tariffs (VN, or Normative Value in Brazil). Greenfield natural gas fired thermal plants were indeed built but some of those contracts are now stranded creating a burden on customers 28 in some particular concession areas. In early 2003, a new administration proposed a pure single buyer model as a possible alternative to address those challenges. Additionally, this model would help capture the “economic rent”51 of depreciated hydro plants, and distribute it evenly to a customer base of about 40 million users. After an analysis of international experiences with single buyer models, the idea was overruled. Instead, the Ministry of Mines and Energy came up with an innovative scheme whereby the Market Administrator (CCEE) would work as “broker.” According to this scheme, distribution companies should prepare their demand forecasts which would be centrally aggregated. All the energy to be acquired by distribution companies had to take place in a competitive manner via an auction mechanism carried out by CCEE in a centralized purchasing arrangement.52 The arrangement was called a “pool.” Distribution companies were not allowed to enter into new bilateral contracts outside this centralized arrangement except for the existing contracts or PPAs, which were honored.53 In these auctions each distribution company would buy a “quota” of the energy from the pool. As a result each would sign forward contracts with all the winning generators. An elaborate procurement mechanism based on a simultaneous declining price auction was put in place. The first auction was carried out in December 2004, with the participation of 18 Generation Companies and 35 Distribution Companies. Roughly, 600 bilateral contracts were signed, each Distribution Company with each Generation Company, in a criss-crossed kind of contractual web. 17,000 MW of firm generation capacity were traded, involving about US$ 28 billion in 8 year contracts for delivery in 2005, 2006 and 2007.54 Given its sheer size, it was called a “mega-auction.” Contrary to a typical Single Buyer, no government owned special centralized purchasing entity took title for the energy traded, but simply played the role of an honest broker, setting the rules of the game and instructing the market administrator to play the role of the auctioneer. The price paid by distribution companies in the auction is the basis for pass-through of power costs to regulated tariffs of captive customers. Therefore, the mechanism created a more straightforward and transparent way to deal with the thorny issue of setting a normative value for pass thorough purposes, something that Brazil had struggled for more than 5 years without finding a satisfactory solution.55 The auction itself was a success, and proved that this centralized procurement model worked. For some investors, it lacked the desired transparency and to some extent it was a change in the rules of the game, since it no longer allowed privatized generation companies to freely negotiate their energy with distribution companies. The auction was also criticized on the ground of the resulting low prices and loss of market value for large generators, both public and private. It is reasonable to assume that prices reflected supply and demand conditions at that time (Brazil had a capacity glut of 8,500 MW post 2001 rationing) but some investors contend that large state owned generators bid artificially low prices. Box 15: Recent Brazilian Experience in Energy Procurement (Auctions), con’t. The real test for the new power sector model will take place for the expansion of new capacity. Initial auction have simply re-contracted generation which was already existent in the market (the so-called old energy). Therefore, they did not entail the building of any additional MW. The success of the new model and the auction process should be measured by the real commitment to build new plants to be made by private investors, both domestic and foreign. A first test carried out in December in 2005 has been successful in terms of attracting investors to build the required new capacity. However, lack of interest from large international players is still a point of concern. Furthermore, it turned out that difficulties in getting environmental licenses in place prior to the auction were underestimated. The jury is still out, and Brazil will have to go through a learning curve as the competitive procurement process is unfolded. Source: Authors’ analysis • As described in Box 13, no central agency in Brazil is most developing countries do not have retail in charge of taking ownership for the energy. The competition in place, some of the principles of market operator, based on clear guidelines set forth coordinated procurement process adopted in New 29 by the government and by the regulator, coordinates Jersey may still be applicable in some developing the auctions and allocates the energy among countries. Box 15 provides more details about the distribution companies based on their own load model adopted in New Jersey. forecasts. Therefore, it ends up being multiple buyer and seller contractual arrangements where the central Roles and responsibilities of the central agent agent does not take the title for the energy. should be unbundled to a reasonable extent in • A more elaborate competitive procurement model is order to enable some checks and balances observed in the New Jersey in the US. In New Jersey, there is vibrant wholesale competition and also full A “pure” Single Buyer is a monolithic organization. retail competition. However, not all customers have Most procurement functions are carried out internally, chosen alternative suppliers. The regulator allows sometimes with minimal regulatory oversight and public them to stay with the incumbent distribution company. scrutiny. Some unbundling is advisable, in terms of who This company, in turn, has the obligation to serve plans the system, determines energy requirements, sets those customers as a supplier of last resort. The rules for the procurement and monitors results. The regulator put in place an interesting scheme to practical degree of unbundling depends on the stage of encourage distribution companies to buy energy the development of the power sector and the level of effectively. It is based on a central agent arrangement. maturity of the organizations. The model has been considered very successful, has served as a benchmark and has been adopted in At a minimum, the government or the regulator should other adjoining states in the US. A similar model is establish the rules and procedures for carrying out the now being considered for Italy. Despite the fact that functions described in the previous paragraph. An 51 Studies developed in 1998 proved that there were indeed depreciated power plants in perfect operating conditions – therefore with economic rent. However, there were also plants heavily stranded, whose “accounting cost” was as high as US$ 270/MWh. On the average, the accounting cost, on a full costing basis, for the entire portfolio of generation plants was about US$ 36/MWh, or very close to the long run marginal cost of expansion. Therefore, on the average, there was not too much economic rent to be captured. 52 This concept of pooling has nothing to do with how the power system was (or is) operated on a real time basis. Brazil has always operated its power system as a security constrained, least cost centralized dispatch, run by an independent system operator (ONS). Forward energy contracts have always been financial arrangements (financial hedges), and the new commercial pooling concept did not alter the way the system is physically dispatched. 53 Or for transaction outside of the regulated environment, where generation companies were allowed to sell their energy to large customers or to marketers – but not to regulated distribution companies. 54 17,000 MW contracts of firm generation capacity correspond approximately to 25,000 MW of installed capacity 55 A discussion about this issue can be found in Arizu, Maurer and Tenenbaum. (2003) interesting example is in Jordan where the Grid Code56 Initially, it may be necessary or desirable to group not only establishes the central agent procedures for several functions together under a central agent. system planning, dispatch and generation expansion, However, as a minimum, some functional separation but also assigns to the central agent57 the right to audit should be planned at the outset, to build more the physical dispatch conducted by the system operator. transparency and to pave the road for future unbundling. Currently, both activities are under the same ownership This is the case in European Union accession countries structure. However, functional separation can create a that are working to comply with EU directives. For fledgling culture of relative independence with internal instance, when the Turkish Electricity Corporation (TEAS) checks and balances. was restructured in 2001 as part of the power sector Box 16: Mandated Competitive Procurement to Serve Retail Customers In the United States, the New Jersey regulator has mandated competitive procurement by all distribution companies. New Jersey has full retail competition, whereby every customer has the legal right to buy from an alternate supplier. But even though all customers have the “right to choose,” very few retail customers have actually exercised this right. This then requires that some entity be the “default supplier” or “supplier of last resort” for the energy needs of these customers. In New Jersey the four traditional distribution companies were 30 required to assume this role for the first two years following the introduction of retail competition. However, starting in the third year, the reform law in New Jersey required that the default service (known in New Jersey as the “basic generation service”) be provided through a competitive auction. There have been several such auctions. One carried out in February 2003 attracted 20 bidders who competed for the right to supply about 18,000 MW of capacity and associated energy with an estimated value of US$5.2 billion. The auction was conducted over the Internet using a type of auction known as the “clock auction.” This auction involves multiple bidding rounds in which prices tick down until the amount of energy supplied equals the amount of energy sought. Bidders bid for the right to serve tranches that represent 2% of the residential, commercial and industrial loads of each of the four distribution companies for periods of either 10 or 34 months. In the future, it is anticipated that the right to serve one third of the retail load will be subject to competition every year. The winning bidders will assume a one year supply obligation. The distribution companies do not actually conduct the procurement. The Regulatory Commission hires its own consulting firm to review the design and implementation of the auction. The actual auction process was conducted by NERA, an economic consulting firm, whose services were paid for by the distribution companies. The prices to serve residential and small businesses were close to US$ 50/MWh. However, these are not final prices to customer. Each of the distribution companies will add on a “distribution margin” to the winning bids for generation service. (More complete information on the New Jersey auction can be found at www.bgs- auction.com.) Contrary to a pure Single Buyer model, in New Jersey the distribution companies do not take title to the power. They are acting as agents for those retail customers taking the default service as opposed to the more traditional role of distribution companies as buyers and resellers of power. The 15 winning bidders bear demand risk—the risk that demand will be greater or smaller than anticipated. Also, distribution companies continue to bear the risk of non-payment. Self-dealing is allowed. Actually, at least three of the winning bidders were affiliates of the buyer. But there have been no allegations of unfair favoritism towards these bidders. Presumably, this reflects the fact that the actual auction was conducted by a third party and the winners were selected solely on the basis of their willingness to supply at particular prices rather than the more subjective combination of price and non-price criteria, such as auctions conducted in other places in the US (e.g. Florida). 56 A set of norms and procedures on how, inter alia, the power system should be planned, expanded and operated. 57 An Unbundled Single Buyer Box 16: Mandated Competitive Procurement to Serve Retail Customers, con’t. This kind of competitive auction tied to serving default customers, could also be used in situations where there is no mandatory retail competition. Even in the absence of retail competition, a Regulatory Commission could require that a distribution company conduct an auction to serve some or all of the energy needs of their retail captive customers. Alternatively, the Regulatory Commission could itself set the rules and organize the auction in a centralized fashion. The New Jersey central purchasing arrangement has been considered very successful in the U.S., among many other attempts to have distribution companies procuring energy in their capacity as providers of last resort. It has recently been adopted by the state of Maryland, and other states are about to adopt the system. As already mentioned, its application goes beyond places where there is full retail competition, and the concept can be used by some developing countries interested in some sort of competitive procurement. Source: World Bank analysis. reform process it was separated into three functional More flexible contractual arrangements should be 31 companies. Under the current arrangement, the Turkish utilized Electricity Transmission Company (TEIAS) serves also as the system and market operator. The trading functions, As discussed before, one of the drawbacks of the Single on the other hand, are carried out by the Turkish Electricity Buyer model is that it relies upon very rigid, capacity Trading and Contracting Company (TETAS). Prior to the based, long term PPAs. Once signed, a serious buyer is reform both functions used to be under the purview of committed with 15-20 year energy contracts, bearing most TEAS which operated as an Unbundled Single Buyer. risks for price, market, and exchange rate, among others. Sector regulation may include rules and codes published The rationale for this rigidity is well known. Developers, and amended through consultation processes with those particularly in un-chartered and uncertain territories, participating or trading in the industry (participants). want to hedge as much political and market risk as Non-discrimination is ensured by the implementation possible. Furthermore, in the absence of a market, any of these rules through an entity (the system operator, the developer faces a monopsony-like situation. In a market operator, or the system and market operator if predominantly seller’s market, and in countries oftentimes they are combined) that is independent from the generation plagued by chronic shortages, off-takers are seldom in and trading activity (i.e. is not involved in the business a good position to impose risk allocation schemes 58 of selling and purchasing electricity). Other mechanisms fundamentally different from industry norms. However, can increase predictability and transparency, by monitoring some flexibility on risk allocation is workable, as long as and enforcing compliance with the rules and codes. governments perceive it as desirable. Independent auditing and market surveillance functions of both the system and market operators are also Moving away from a rigid contractual environment to a important, to assure compliance, performance and new one where competition is fostered entails major efficiency. changes. The first, possibly the most important step, is to make sure that there is competition “for the market,” that is, new energy is procured competitively. The second is how to make sure that all the generation plants in compete “in the market” for dispatch on a day 58 The issue of concern is that the system operator or the market operator must provide transparent and non discriminatory services, and therefore should not be involved in the commercial interests of those participating in the production, sale or resale of electricity. The role of the market or system operator must be carefully designed in order to prevent their competition with generators in purchasing and reselling ancillary services, including balancing energy. to day and hourly basis, producing energy at the least The techniques to be used to integrate a PPA into the possible cost.59 This feature is referred as “integrating” power market will vary depending on the nature of the the PPA into the power market. Reconciling the need to contract, the specifics of the power sector or market dispatch at minimum cost with honoring contractual design, and other important political considerations. obligations is not always an easy task. In general, Some possible techniques to give more “flexibility” to traditional IPP programs in power sectors without pre- the PPAs, include:60 established codes and rules, most of the rights and obligations of the IPP are (or are supposed to be) • Forced market integration spelled out in its PPA, which cannot (and should not) • Forced contract renegotiation be overruled by subsequent market or dispatch rules • Market Rules modification unless the investor voluntary agrees to do so. • Voluntary renegotiation • Virtual Generation/Managed PPA A fundamental principle is contract sanctity. A contract • Contract buy-out breach entails a significant reputational risk for the country, discouraging new investors to further participate A detailed discussion on each one of those approaches in the sector.. On the other hand, entering into long goes beyond the scope of this paper. Table 5 presents a term, rigid contracts implies a burden for the customer summary of the main features and key considerations 32 and may impede a smooth transition to a more surrounding each one of those approaches. competitive power sector. There is not a single, perfect solution. Unless there is demonstrable evidence of In summary, there is not a single solution to deal with the corruption and contractual malpractices, the underlying problem of integrating existing IPPs into new markets. assumption is that power contracts are to be honored. There is a delicate balance between creating a climate of contract sanctity, conducive to new investments, and This underscores the importance of having good contracts at the same time permitting a better integration of energy at the outset. Designing or negotiating contracts is not an contracts into the market. Table 6 evaluates those easy task. Any amendments have to be mutually agreed approaches against a common set of criteria. among parties. In some cases, such as in Dominican Republic or Indonesia, parties have mutually agreed to Some important recommendations deriving from this reduce the price of the contract. As a quid pro quo, the analysis follow: buyer extended the terms of the power agreement. Contract renegotiations, abiding by the principle of • The forced market integration and forced sanctity, also took place in Thailand, following a significant renegotiation approaches have been attempted but devaluation of the local currency. More often than not, are very unlikely to achieve a satisfactory outcome contract renegotiation is a tedious, lengthy process. It is from the point of view of IPPs or consumers; worth trying even if chances of success may not be high, • The adaptation of market rules, on a transitional or unless the contract sets forth severe penalties in case of exemption basis could, under some circumstances, renegotiation. In the case of termination, the government prove successful provided that the market can still has to be prepared to pay the necessary compensation work efficiently and competitively with respect to other for terminating the contract in a manner consistent with generating plants; its relevant clauses. For such alternative, it is recommended • The virtual/managed PPA approach could be a useful that PPAs include a buy-out clause. technique provided that there is an incentive to maximize above market costs; Governments, or centralised purchasing agents acting • Voluntary renegotiation stands a much better chance on their behalf, should make serious efforts to design of succeeding than forced negotiation if there is contracts that are “bankable” and at the same time may innovative thinking to find mutually beneficial solutions be smoothly integrated into the rest of the power sector and its ongoing reform. 59 Equivalent to a least cost, security constrained central dispatch. 60 This discussion on the design of flexible contracts and on the integration of IPPs is based on “Integrating Independent Power Producers into Emerging Markets”, by Woolf, Fiona and Halpern, Jonathan. World Bank Working Paper, November 2001. • Contract buyout has the advantage of crystallizing above-market costs at an early stage, but inexperience with the newer market and the likelihood of having to make mid course adjustments to market rules may result in contract bidders demanding a new market risk premium which will magnify the above-market costs. Table 5. Options for Integrating PPAs into Wholesale Power Market Forced Market Integration • IPPs forced to become market participants, regardless of contract firms. • Alternative tried in Guatemala without success. • Difficult choices for IPPs – wither to accept or to seek damages. • Dangerous option for Governments – may be construed as expropriation, leading to endless litigation and arbitrage. Forced Contract • Contacts are changed to become compatible with technical and trading Renegotiation rules. • Forces IPPs to take market risks that were not present before. • Requires immense efforts, but may produce small results • IPPs may be resistant to look for consensus. 33 • Undermines credibility among investors. • May be a costly option for the Government, involving recovery of stranded costs: e.g. Dominican Republic (Madrid Agreement) Market Rule Modification • Market rules imply additional obligations (and costs) for the IPPs. • Government may provide exemptions to smaller plants. • However, those exemptions may compromise efficiency. • Possible solution is to include capacity payments in the market. Voluntary Renegotiation • To be decided on a case by case basis. • More efficient and may lead to more practical results. • Some very successful cases reported, such as in Pakistan. • Some additional sources of revenues may be given to IPPs, for example, the option to sell ancillary services in the market. • A contract facilitation process may help, if perceived as independent – e.g. Ontario, Thailand. Virtual Generation- • Entails designing contract in such a way that IPP is protected if changes Managed PPA occur in market, or transferring the contract to a successor entity, public or private. They become the virtual owners of the generation. • Off-taker, not the IPP, will bear market risk – in IPP perspective, contract is “ behind the market” • Successful example in Brazil – IPP had a contract with a generation company which assumed market risks. • This option has the advantage of preserving the contractual rights and minimizing changes in the PPA. Contract Buyout • Either the contract holder or third parties may buy the contract when sector restructuring occurs. • There has to be consent from IPP lenders, bondholders and possibly equity investors. Tried in Ontario, without success. • However, appears to be a promising alternative, particularly if buyout conditions are established at the outset. Source: Adapted from Woolf and Halpern (2001) 61 Fraser. (2005) Table 6. Evolution of Options for Integrating PPAs into Markets RENEGOTIATION -MANAGED PPA MODIFICATION NEGOTIATION INTEGRATION MARKET RULE GENERATION VOLUNTARY CONTRACT CONTRACT BUYOUT FORCED FORCED VIRTUAL MARKET EVALUTATION CRITERIA May adversely Yes Yes No No, if process No Yes, unless affect credibility is transparent IPP of Government and fair compensation is fair May deter future Yes Yes No No, if process No Yes, unless investors is transparent IPP and fair compensation is fair May adversely Yes Yes Yes, unless No, IPP No Yes, unless affect economic exemptions unlikely to IPP is fully position of IPPs are granted agree to any compensated reflected in PPA deterioration 34 May reduce impact Yes Yes Unlikely Yes Yes, if an No, unless on consumers of (depends on (depends on effort is buying price above-market cost outcome) outcome) made to offered by reduce above third party is market costs discounted to reflect market uncertainties May adversely No No Yes Maybe No No affect efficiency or (depends on (but it may be (depends on competition in the outcome) transitional) outcome) electricity market Source: Adapted from Woolf and Halpern (2001) There should be a game plan to allow power a de facto sole buyer, signing contracts directly with the sector reform to evolve sellers. Despite not being desirable, it is probably the only workable solution in distressed power systems, A centralized purchasing arrangement (either via single where other buyers such as distribution companies are or multiple buyer models) has to be implemented with a not minimally creditworthy or are not institutionally clear view on how the power sector should evolve. The established entities. first important step is to have flexibility built into the energy contracts, as discussed in the previous section. As the sector evolves and distribution companies are Another important step is to define the role and intensity of created and/or become more creditworthy, the model the central purchasing agent, if any, and how this role may move towards a non-monopolistic centralized should be adjusted as power sector reform moves forward. purchasing arrangement where distribution companies are allowed to be buyers from IPPs. The Single Buyer In principle, it is possible to envision three basic stages may co-exist with other buyers.62 Over time, some of the of power sector reform, each one resulting in a different original responsibilities of the pure Single Buyer may be role for the central purchasing function. phased out. A central purchasing function will still be desirable, to achieve economies of scale and conduct Initially, in a vertically integrated industry, it is competitive procurement on behalf of the diverse buyers. reasonable to assume that the central agent (e.g. a In more competitive power markets, the central agent Single Buyer) will play a more active role, possibly being may play the role of an energy aggregator and 62 As we alluded before, it will not be a “single” buyer technically speaking. procurement coordinator, with no ownership for the places where electricity is considered a public good that energy traded in the market. This has been the case in consumers should receive even if they do not pay. New Jersey and Brazil. In the former, procurement takes The issue of eliminating guarantee requirements is place to serve captive loads acting as the provider of largely in the hands of the government. If investors last resort, while in the latter procurement takes place perceive a dysfunctional power sector, they will require on behalf of local distribution companies. sovereign guarantees as a must. To reduce contingent liabilities ensured by those guarantees, governments Sovereign guarantees should be granted on a have to work hard to treat energy as a commodity, selective basis preserve property rights, and put in place the proper tariff and other regulatory mechanisms to create a healthy The need for guarantees depends on the creditworthiness and self-sustainable power sector. Many of those things of the distribution companies, who are in most cases are cannot be achieved overnight, but a clear direction of the off-takers. If distribution companies in the sector are change and demonstrable commitment is necessary. financially viable and present a good credit, sovereign guarantees should not be seen as needed as part of the There needs to be clear, codified rules to assure centralized purchasing arrangement. That is the case in transparency, accountability and good South Africa, where the government made it very clear performance that given its good financial standing, ESKOM, acting 35 as a Single Buyer, would not receive any form of The government, and the independent regulator (in guarantees for the PPAs to be signed with third parties. countries where such entity is functional), should do The vision is that ESKOM, as a state company, should their best to put in place rules and procedures, in the act on an arm’s length basis, and no contingent entire supply chain, to guarantee that energy is acquired liabilities should be imposed upon the government. in the most efficient and transparent way. When other country experiences are evaluated, some Procurement rules and guidelines to be followed by the kind of government support seems to be a requisite in central purchasing agent should be drafted by the sectors where financial viability and sustainability are Regulator (or the Ministry if it is vested with the of concern. More often than not, companies are not regulatory authority), and should also monitor and creditworthy because tariffs are not cost reflective assure compliance. and/or there is a culture of non-payment for electricity and/or inefficiencies or fraud in the metering, billing “Among the standard rules on procurement which and collection cycle. should be covered by the national jurisprudence are the following: open, public competition must be the rule Investors will seek sovereign guarantees based on their and actual practice for all procurement decisions above perception of how committed governments are in their a relatively low value threshold; any exceptions should endeavors to create a well-functioning and sustainable be possible only in truly exceptional circumstances (e.g., power sectors. A good indicator of the seriousness of natural disasters).”64 government in pushing the reform agenda and improving efficiencies is to what extent resources are being allocated The Regulator’s role is to strengthen the transparency of to the installation of meters, reduction of technical losses, the centralized purchasing arrangements. Several combat on fraud and theft,63 enforcement on the need possible functions assigned to the Regulator have been for prompt payment of electricity bills, and pass-through identified, as shown in Box 17. of power purchase costs. Investors are very skeptical in 63 There are two problems associated with this item. One is fraud, by which staff inside the utility offers “services” to consumers not to pay or not be billed. The other is when consumers themselves tamper meters or make their own illegal connections. 64 Wiehen. World Commission on Dams. (2002) 6. CONCLUSIONS AND RECOMMENDATIONS Box 17: Regulator’s Role in Centralized Purchasing Arrangements Worldwide experience shows that creating a Single • Licensing the centralized agent: When the central Buyer with a monopoly status can be problematic. purchasing agent is subject to licensing, its Such rather rigid energy procurement arrangement obligations can be set in the license conditions should be avoided whenever possible. regarding different functions, in particular economic procurement and competitive tenders for power In developing or transition countries, this type of procurement. The Regulator then has the role of arrangement has been successful in facilitating IPP monitoring and enforcing these conditions. investments and privatization, but it has been disappointing in many other aspects. It may lead to corruption, • Approval of generation expansion planning. This discriminatory behavior, excess capacity, high prices, is applicable when a central purchasing agent is increase in public sector liabilities, and even prevent the responsible for the planning function. future development of a competitive power market. • Issuing guidelines or procedures for competitive procurement by the central purchasing agent. If it is the only alternative, then the Single Buyer model needs to be approached with caution. 36 • Approval of standard PPAs and standard tender Governments and those who advise them should be documents. aware of its drawbacks, do not view it as a permanent solution and plan a transition to phase • Approval of new PPAs, subject to compliance out or reduce the wholesale purchasing monopoly as with procurement procedures or guidelines. soon as possible. • Periodic audits of the central purchasing agent’s If initially, given the small scale or difficult situation procurement processes, or of the system operation faced by a power sector, a multiple buyer/sellers dispatch function, when it is assigned to the central scheme is not workable, some form of centralized agent or when it is assigned to the system operator. purchasing arrangement may be required. Experience • Introduction of a tariff level and structure that has shown that very small, low income countries have aims at improving efficiency and cost recovery. indeed relied upon the Single Buyer model when they could not find a workable alternative to attract investment • Approval of bulk supply tariffs and pass through for the generation capacity, badly needed to reduce of the power procurement costs when power is power shortages and to improve their distressed electricity procured competitively, or making sure that the sectors. In case of lack of creditworthy buyers in the sector, central agent proves that the purchase was countries have resorted to wholesale power purchases economic and at least possible cost. carried out by a central government owned entity, with contracts backed up by sovereign guarantees, in order to • Formalize distribution companies’ tariff methodology attract much-needed new generation investments. There to address concerns on pass through and recovery may be legitimate reasons to start with a Single Buyer of distribution company purchase costs. type of arrangement. Countries in distressed situations and facing chronic The broad range of centralized purchasing power shortages should examine rational Demand arrangements may offer the same advantages Side Management (DSM) alternatives. associated with the single buyer, but allow the much needed flexibility. There is no “one size fits all” Bridging the supply demand gap requires more than approach to a centralized purchasing arrangement. expansion of the generation capacity. In many instances, It is crucial that each country takes the specific only supply driven alternatives do not suffice or are too characteristics of its electricity system into consideration expensive (not affordable) to be implemented. Unless and incorporates its own objectives in power sector energy is used rationally, customers may not have reform in order to design a sustainable solution. financial resources to cope with the expensive capacity expansion alternatives promoted by the Single Buyer or The institutional and commercial mechanisms related by other similar kind of arrangement. Experience has to a centralized purchasing arrangement have to be shown that promoting more efficient use of energy offers carefully tailored to the specifics of each country. This many opportunities that are by far cheaper than increasing will enable the country realize its objectives of building traditional or renewable supply sources. In many countries a sustainable, efficient electricity sector and at the same there is a significant potential to be harnessed. This is time allow the government to play a necessary role in not to say that DSM will preempt capacity expansion, the initial period. There are many ways to set up a but it should necessarily be considered part of the centralized purchasing function. Its prime role as a 37 equation.65 In many cases, it may be a better substitute buyer (or agent) involves the decision on how much to to supply alternatives. buy. It may have to be an active buyer, in the absence of creditworthy institutions, to support generation Centralized purchasing arrangements do not need expansion. It can also be responsible for system operation to have a monopolistic entity responsible for the and market administration in the initial phases of power energy buying and selling in a country. sector reform. A central procurement entity can be assigned a role of market aggregator, coordinating A Single Buyer model is a well known form of centralized power procurement, or acting as a buyer of last resort. procurement, which is given a quasi-legal monopoly An important point is that the central agent does not status over the trading of wholesale electricity in the need to take title over the energy to play these roles, market. However, there is not a single way to implement unless there are no other ways to have reliable buyers. it. While designing or deciding a centralized purchasing approach for a power sector, it is important to look at A central purchasing agent should ideally be what other countries have done. There is a wide range unbundled from generation and any other activity of possibilities to capture economies of scale, develop competing with its trading of energy. expertise, and deal with the issue of non creditworthy buyers. Among the range of possibilities, there are To make a sector attractive for generation investors, instances where the buyer is not “single” and may be the central procurement agent should not own generation subject to competition with other buyers. In other cases, assets, particularly if it is responsible for system operation it is not even a buyer, since it does not take title for the and market administration. IPPs should be assured that energy traded directly between generators, distribution they are not going to be unduly discriminated. The level companies and large customers. of unbundling needs to be defined and implemented with a macro vision and blueprint for the power sector reform as a whole. 65 Maurer, Pereira and Rosenblatt (2005). Whenever possible, monopoly rights should not be In order to avoid many of the problems associated granted to a central agent as this may be very with the integrated or unbundled Single Buyer difficult to eliminate in the future. At the outset, model, one must make sure that all energy is plan the phasing out of the central agent’s role as a procured competitively, in a transparent process, key energy player, as the power sector evolves. while maintaining the sanctity of existing contracts. Direct one to one negotiation with unsolicited Very small, poor countries with distressed power projects should not be acceptable. There are many sectors may have to initially rely on a model with a success stories in energy auctions or competitive Single Buyer resembling a vertically integrated power tenders for new generation. sector. However those countries should avoid granting permanent monopoly rights to this buyer, in order to A transparent and competitive procurement process is allow the sector to evolve into a more flexible model perhaps the most important aspect to be taken into when possible. One can avoid making the central account, at any phase of power sector reform, no purchasing agent a pure monopoly (a Single Buyer) matter how small or distressed a system is. Many by giving customers or distribution utilities the prerogative problems related to corruption (or perception thereof), to buy a growing portion of their energy needs in the oversupply, nepotism, contract disputes and arbitration, market as the sector grows or the market matures, or by and high energy prices derive from unsolicited offers, 38 implementing the right of a distribution company to side deals, and the lack of a sound least-cost system procure directly from generators if it can obtain lower expansion planning process, codified in the regulatory prices than the central agent. framework. Competitive processes for energy procurement, such as energy auctions for existing and new energy, As part of the transition plan to phase out the role are becoming increasing popular and promising. and dominance of the central agent, it would be useful to create initial contractual arrangements that Centralized purchasing arrangements also make give investors a reasonable degree of certainty, but sense as part of the second generation reform at the same time are flexible enough by including process. provisions for the healthy evolution of the power sector. Centralized purchasing arrangements are not only relevant in developing country, small power sectors, To avoid a straightjacket, the contractual arrangements where procurement skills, lumpiness of investment, and between the central agent and the buyers and sellers, lack of credit worthiness of distribution companies are have also to be flexible, include buy out and transferability real causes of concern. Its value has also been conditions, and provisions regarding any temporary role acknowledged in second generation power sector of the central agent that will disappear later. It is reforms in developed competitive electricity markets, understandable that developers and their financiers whereby the purchaser acts like a broker or coordinator would rather have long term, risk-free PPAs for the entire of a transparent competitive procurement process. duration of the asset pay-back period. However, these arrangements can include special provisions that add flexibility and avoid the self-perpetuation of the central agent, and, in particular, avoid creating barriers or stranded costs that prevent a country from moving towards a more competitive or multi buyer market in the future. The interests of the investors and those of the consumers need to be balanced. There is no magic solution, but experience has shown that it is possible to find a compromise. Some of the alternatives include - shorter term contracts, particularly for existing generation (as opposed to green field), buy-out clauses, and a gradual decline in distribution companies’ obligation to acquire energy from the central agent. Like the Single Buyer model, other centralized purchasing arrangements can indeed help expand generation capacity and improve reliability of supply. However, purchasing arrangement should not be blindly amended just for the sake of expanding capacity. A power sector has to be sustainable, and the ultimate goals should be, in countries with low electrification, to increase access, and in all countries to provide an affordable and reliable electricity service to consumers. Expanding generation capacity via a Single Buyer or other centralized purchasing arrangements should not be perceived as an end in itself. Experience has underscored the importance of building an ultimately sustainable energy sector, where tariffs are adequate, energy is used efficiently and smartly targeted subsidies are implemented whenever needed. Examples abound 39 where countries were indeed able to attract sufficient generation capacity but at high costs, excess supply needs, and therefore consumers were not able to pay for the energy produced. In such instances, little or no improvement in access or service quality was noticed, despite the changes in the institutional procurement mechanisms and availability of power. Examples also exist of countries with low electrification where a Single Buyer model was implemented but energy access has not increased. A Single Buyer model or any other centralized purchasing arrangement should not view capacity expansion an end unto itself, but as part of a broader program to help the poor and make energy more affordable and reliable. In that sense, the track record of Single Buyers in serving the poor is relatively modest. Little correlation was established between capacity expanded via Single Buyers and a significant increase in energy access. REFERENCES • Lovei, Laszlo. 2000. The Single Buyer Model. Viewpoint. No. 255. Washington, D.C.: World Bank. • Arizu, Beatriz, William Dunn, and Bernard Tenenbaum. 2002. Transmission System Operators. • Maurer Luiz, Mario Pereira, and Jose Rosenblatt. Lessons from the Frontline. Energy and Mining Sector 2005. Implementing Power Rationing in a Sensible Board Discussion Paper Series, Paper No. 4. Way. ESMAP Series, No. 305-2005. Washington, Washington, D.C.: World Bank. D.C.: World Bank. • Arizu, Beatriz, Luiz Maurer, and Bernard Tenenbaum. • Maurer, Luiz and Tjaarda Storm van Leeuwen. 2003. 2004. Pass Through of Power Purchase Costs, Energy Issues to Consider: Flexibility of the Single Buyer and Mining Sector Board Discussion Paper Series, Model. Riyadh. (Draft for Discussion) Paper No.10. Washington, D.C.: World Bank. • Petrov, Konstantin, Paul Giesbertz, Tjaarda Storm van • Bakovic, Tonci, Bernard Tenenbaum, and Fiona Leeuwen, and Rene Mendonça. 2003. Market and Woolf. 2003. Regulation by Contract. Energy and Industry Design in the Power Industry: Application for Mining Sector Board Discussion Paper Series. the Maghrebian Countries. Washington, D.C.: World Washington, D.C.: World Bank. Bank. (Unpublished Draft) 40 • Barker, James, Luiz Maurer, and Tjaarda Storm van • Petrov, Konstantin, Paul Giesbertz, and Annelies Leeuwen. 2003. The Evolution of the Single Buyer Huygen. 2002. Electricity Generation Activity reform in Model and Introduction of Competition. Washington, Tunisia, KEMA Consulting. D.C.: World Bank. (Unpublished) • Parikh, Kirit. 2002. Thinking Through the Enron Issue. • Bacon, Robert and John Besant-Jones. 2002. Global Economic and Political Weekly, XXXVI (17): Pages Electric Power Reform, Privatization and Liberalization 1463-14–72. Volume XXXVI, Number 17. of the Electric Power Industry in Developing Countries, Energy and Mining Sector Board Discussion Paper • Rimban, Luz and Sheila Samonte-Pesayco. 2002. Trail No.2. Washington, D.C.: World Bank. of Power Mess Leads to Ramos. The Philippine Center for Investigation. • Becker, João and Luiz Maurer. 1990. Plan 2010 Ignores Variations in the Cost of Capital, Gazeta • Salant, David. 2003. On the Role of the Single Buyer: Mercantil, Brazil. Acquirente Unico. NERA. • Bosshard, Peter. 2002. Private Gain - Public Risk? The • Sweeney, James L. 2002. The California Electricity International Experience with Power Purchase Crisis. Hoover Institution Press Publication No. 513, Agreements of Private Power Projects, International Stanford, CA. Rivers Network. • Wiehen, Michael. 2002. Transparency and Corruption • Fraser, Julia. 2005. Lessons from the Independent on Building Large Dams, World Commission on Power Experience in Pakistan. Energy and Mining Dams, Thematic Review V.4. Transparency Sector Board Discussion Paper No. 14. Washington, International, Germany. D.C.: World Bank. • Woolf, Fiona and Jonathan Halpern. 2001. • Henisz, Witold and Bennet Zelner. 2001. The Political Integrating Independent Power Producers into Economy of Private Electricity Provision in Southeast Emerging Wholesale Power Market, Washington, Asia, Reginald H. Jones Center Working Paper, the D.C.: World Bank. Wharton School University of Pennsylvania and McDonough School of Business Georgetown University, WP 2001-02. THE WORLD BANK GROUP The Energy and Mining Sector Board The World Bank 1818 H Street N.W. Washington, D.C. 20433 USA