E N E R G Y & M I N I N G S E C T O R B O A R D D I S C U S S I O N P A P E R S E R I E S P A P E R N O . 1 A P R I L 2 0 0 1 The California Power Crisis: Lessons for Developing Countries John Besant-Jones and Bernard Tenenbaum THE WORLD BANK GROUP The Energy and Mining Sector Board AUTHORS ACKNOWLEDGMENTS This paper was prepared by John E. Besant-Jones and Bernard W. Tenenbaum, Energy and Water Department, Private Sector Development & Infrastructure, under the sponsorship of the World Bank's Energy and Mining Sector Board. It benefited from the comments of several staff members of the World Bank Group, including Jamal Saghir, Alastair Mckechnie, Laszlo Lovei, Alan Townsend, Barbara Kafka, Robert Bacon, Ranjit Lamech, Mangesh Hoskote, Charles Feinstein and Kseniya Lvovsky. Comments were also received from Sabine Schnittger (Frontier Economics), Harvey Salgo (La Capra Associates), Michael Rosenzweig (NERA), Jim Barker (BDR, Inc.) and Daniel Yergin (Cambridge Energy Research Asso-ciates). DISCLAIMER The views published are those of the authors and should not be attributed to the World Bank or any other affiliated organizations; nor do any of the conclusions represent official policy of the World Bank or of its Executive Directors or the countries they represent. The authors are solely responsible for any errors of fact or interpretation that may remain. CONTACT INFORMATION To order additional copies please call the Energy Help Desk. 202-473-0652 energyhelpdesk@worldbank.org This Paper is available online www.worldbank.org/energy/ 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 permission 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. E N E R G Y & M I N I N G S E C T O R B O A R D D I S C U S S I O N P A P E R S E R I E S P A P E R N O . 1 A P R I L 2 0 0 1 The California Power Crisis: Lessons for Developing Countries John Besant-Jones and Bernard Tenenbaum The World Bank, Washington, DC THE WORLD BANK GROUP The Energy and Mining Sector Board Copyright © 2001 The International Bank for Reconstruction and Development/The World Bank. All rights reserved CONTENTS Contents ...............................................................................ii 3. Allow vesting contracts as a form of insurance for distributors purchasing from a new spot market. .........15 Acknowledgements..............................................................iii 4. Save full retail competition for last. .................................15 Foreword...............................................................................v 5. Starting points matter. ....................................................16 Acronyms .............................................................................vi 6. The economic regulatory system must be open, Units of Measurement..........................................................vi independent, credible and not prone to bankrupting Introduction..........................................................................1 reasonably efficient firms................................................17 Overview of the California Reform and its Lessons .............3 6.1 Distribution ...................................................................17 Why the Reform? ....................................................................3 6.2 Market Regulation and Monitoring..................................19 The Nature of the Reform ........................................................3 6.3 Division of Authority between National Key Features...........................................................................3 and State Regulators......................................................21 i How the California Reform Differs from 6.4 A Caveat: Regulating State Enterprises Is Other Power Sector Reforms.............................................4 Different from Regulating Private Companies ...................21 The Reform Process.................................................................4 7. Economic and environmental regulators should talk to each other................................................22 The Crisis...............................................................................4 Part II From Reform to Crisis in California ........................23 The Lessons............................................................................5 1. Background ..................................................................23 Overall Design of the Power Market.........................................5 2. The Indicators of the California Power Crisis....................23 Requirements for Competition to Work in the Wholesale Power Market..................................................5 3. Main Parameters of the California Power Market..............25 Introducing Competition to the Wholesale Power Market ...........6 4. Formation of the New Power Market under the 1996 Reform ................................................................26 Introducing Competition to the Retail Power Market...................7 4.1 New Market Structure ....................................................26 Regulating Power Markets........................................................8 4.2 New Market Operating Arrangements.............................28 Part I Lessons from California or What the Power Minister Needs to Know.....................................9 4.3 New Market Regulatory Framework.................................30 1. Start with limited forms of competition that 5. Main Factors that Led to the Crisis ..................................31 can evolve to full wholesale competition............................9 5.1 Market Design Flaws......................................................32 1.1 Cost-Based Spot Markets with Obligations for 5.2 Exogenous Factors.........................................................35 Capacity and Ancillary Services ........................................9 5.3 Exodus of Funds by Utilities ............................................37 1.2 Multiple Buyers, Multiple Sellers in Bilateral Markets.........10 6. Could the crisis have been avoided?...............................38 1.3 Single-Buyer Model .......................................................11 Selected Bibliography.........................................................41 2. Move to a full bid-based spot market only once the necessary conditions are in place......................13 ii FOREWORD The U.S. electric power industry, the last major energy industry should be scrapped, while others argue that the deregulation in that country subject to traditional utility regulation, is being is still a worthwhile endeavor to make the electric power opened up to widespread competition. Some states allow their industry more efficient and customer-oriented, and that prob- retail electricity customers to choose their electricity supplier. lems such as California's can be solved by adjusting market Competitive trading of wholesale electricity and the emergence rules. A third group argues that California's power crisis is a iii of independent grid operators have spread to many regions of failure of market design rather than a failure of deregulation. the United States. The number of independent power producers and marketers competing in the U.S. retail and wholesale power Deregulation of power markets would be rejected on false markets has increased substantially over the past few years. grounds if the causes of the California crisis were largely specific to the design of the California reform. In view of this However, these new markets have not emerged without uncertainty, the World Bank has a duty to its clients and itself problems. California introduced competition to its retail and to gain an understanding of what has happened in California, wholesale power markets in 1998, but has experienced a and to draw lessons from the California experience that are major crisis during 2000 and into 2001. This crisis has applicable to other countries. The purpose of this paper is to provoked a major debate about the risks, as well as the fulfill this duty. In so doing, the paper also assesses whether rewards, of deregulating power markets to allow competition. the crisis could have been avoided through better market In fact, the California power crisis is giving deregulation a bad design and management. Overall, the paper concludes that name, both in the United States and beyond to other countries much of the crisis was avoidable. Nevertheless, the paper also that are reforming their power sectors. identifies many invaluable lessons for other countries that are considering or implementing power sector reform, and I here- This characterization is somewhat misplaced, however, since with commend it to all who are involved in this endeavor. the California reform is more precisely characterized as part deregulation and part re-regulation. Nevertheless, some observers argue that the California experiment with deregulation JAMAL SAGHIR Director, Energy and Water Chairman of Energy and Mining Sector Board ACRONYMS CPUC California Public Utilities Commission CTC competitive transition charge iv FERC Federal Energy Regulatory Commission IOU investor-owned utility IPP independent power producer LADWP Los Angeles Department of Water and Power NOx nitrogen oxides PG&E Pacific Gas & Electric PPA power-purchase agreement PURPA Public Utilities Regulatory and Policy Act QF qualifying facility RECLAIM Regional Clean Air Incentives Market RMR Reliability Must Run RTC retail emissions credit SCAQMD South Coast Air Quality Management District of California SC scheduling coordinator SCE Southern California Edison SDG&E San Diego Gas and Electric SMUD Sacramento Municipal Utility Department TOU time-of-use UDC utility distribution company UNITS OF MEASUREMENT GWh gigawatt-hour MW megawatt MWh megawatt-hour TCF thousand cubic feet INTRODUCTION "California's new electricity market ended up being designed in a highly politicized process.... [W]hat emerged was the most complicated electricity market ever created.." --Professor Paul Joskow , The New York Times, January 13, 2001 "This is a dreadful mess for a state that is held up around the world as a model of innovation...." --The Economist, January 20, 2001, p.57 "Its problems are largely manmade." 1 --Newsweek Magazine, April 3, 2001, p. 23 The California power crisis is so sudden and serious that it is prompting policymakers in many countries as well as other U.S. states to look for lessons that can be applied to the reform of their own power sectors. Concerned policymakers around the world are asking: If things can go so badly wrong with a reform that did not involve wholesale privatization of the electricity supply industry in such a rich and sophisticated economy, what are the implications for much less well- endowed countries embarking on the full menu of reform including privatization? When a power sector reform like California's fails, political authorities are inevitably under strong pressure to "do some- thing" to solve the crisis. In a recent special session of the California legislature called by the governor, legislators intro- duced more than 75 bills intended to solve one or more aspects of the crisis. Unfortunately, quick-fix "solutions" often lead to outcomes that can be inconsistent with the original reform objectives and can produce outcomes that are even worse than the conditions that triggered the reform. For exam- ple, at the time of this writing (March 2001), the California and federal governments have proposed or undertaken actions including the following: · Price caps. The Federal Energy Regulatory Commission (FERC) imposed price caps that may deter the investment needed to overcome the current supply shortage. · Forced sales. The U.S. Secretary of Energy issued several orders that required generators and natural gas suppliers to continue selling to non-creditworthy California buyers. · Government energy trader. A new state law authorizes the requires restructuring and privatization. But like all human state government to spend up to $10 billion on the state's endeavors, power sector reform can be done well or done credit to purchase wholesale electricity that can be resold to poorly. The principal lesson of California is that good inten- the three large privately owned utilities. tions are not enough. Any reform must pay close attention to · "Nationalization" of the grid. The State of California may starting points, the particular problems that need to be solved, become the new owner of the portion of the high-voltage and the appropriateness of the path selected for solving these transmission grid currently owned by the three large private- problems. ly owned utilities. · "Balkanization" of wholesale electricity trade. The state's The paper is organized into three parts. It begins with an Assembly has passed a bill that would make it difficult to overview of the key features of the 1998 California power sec- export electricity produced from new generating plants tor reform: how it differs from reforms elsewhere, the events located in California to buyers outside the state. and actions that have put it in a crisis mode, and the main lessons that can be learned from the crisis. Many elements of the California reform package are peculiar to a complicated and unusual market design that was the out- Following the overview, the main text is divided into two parts. come of a political compromise reached by various stakehold- Part I discusses in depth the lessons learned, which concern er groups. Many of these features will have no immediate, or mainly the establishment and regulation of a mandatory, 2 even near-term, relevance for most developing countries. wholesale power market based on spot pricing. Since this is Since this paper has been written mainly for power sector offi- not a near-term option for many developing countries, the cials in developing countries, it focuses selectively on the sub- paper also describes other, more-limited forms of competition stantive lessons of the California crisis that pertain to the that may suit their situations. Although privatization was not an design of power sector re-form in these countries. element of California's reform, the state's experience does indirectly provide important lessons for the privatization and In developing countries, the California power crisis may be regulation of distribution enterprises and new market entities in creating the impression that power reform is too risky. The developing countries. power crisis in California does not justify this conclusion. For many developing countries, the status quo in the power Part II details the specific reforms initiated in California, sector is the riskiest alternative of all. The status quo often cre- reviews the factors that led to the crisis, and examines whether ates a drag on economic growth through inadequate and the crisis could have been avoided through better market poor-quality power supply. In addition, limited government design and management. The paper draws on numerous funds are frequently diverted to the power sector that would sources such as published articles, reports and websites, as otherwise be available for schools, clinics and roads. well as the working experience of World Bank staff in numer- Therefore, most countries simply have no alternative to a sub- ous countries. stantial and basic reform of the sector that almost always OVERVIEW OF THE CALIFORNIA REFORM The Nature of the Reform AND ITS LESSONS Key Features Why the Reform? · The three privately owned utilities were "encouraged" to sell off their generating plants but without any vesting contracts · California's economy in the early 1990s. to buy back the output of plants. - Major statewide recession. High unemployment. Loss of · In return, the utilities were allowed to recover their "stranded industry and jobs to other states. The state's governor costs" (i.e., anticipated above-market costs) associated with believed that continued high electricity prices (about 50 the two high-cost nuclear power plants and the state-mandated percent higher than the U.S. national average in 1996) purchases of power from certain IPPs through a "competitive would drive many industries out of the state. transition charge" on consumers' electricity bills. · The state government mandated a 10-percent reduction in · Pre-reform electricity sector retail rates. Retail rates were frozen for four years or until stranded costs were recovered. Actual consumer bills went - Three-fourths of the state's consumption was supplied by down little because the reduction in rates was largely offset three large vertically, privately owned utilities: Pacific Gas by the competitive transition charge. & Electric (PG&E), Southern California Edison (SCE) and · Retail (residential, commercial and industrial) customers were San Diego Gas and Electric (SDG&E). The rest of the given the right to choose alternative electricity suppliers. 3 state was served by large and small municipal utilities. · A non-profit, independent system operator (Cal ISO) was - High electricity prices were caused by expensive nuclear created to operate the transmission facilities owned by the power and green power. Specifically, massive cost over- private utilities (about 75 percent of the state's high-voltage runs on two major nuclear power plants and stateman- grid). The Cal ISO also operated a bid-based real-time dated purchases of power from independent power pro- energy market as well as several other markets to acquire ducers (IPPs) using renewable and other technologies at grid support services (i.e., ancillary services). prices significantly higher than the costs of traditional · A separate Power Exchange (Cal PX) was created to operate technologies. a bid-based, centralized market for forward (day-ahead and - There was surplus generating capacity just prior to the day-of) power sales. The two largest private utilities were reform (April 1998). required to buy and sell all of their electricity through the - Approximately 20 percent of California's electricity supply Cal PX. was imported from neighboring states. · Both the Cal ISO and Cal PX were governed by large - The three privately owned utilities were regulated by the boards, each of which was made up of more than 30 California Public Utilities Commission (CPUC) under a stakeholder and non-stakeholder members. traditional U.S.-style cost-of-service regulatory system with · The retail electricity rates of individual privately owned utili- some targeted incentive mechanisms. The CPUC ties continued to be regulated by the CPUC. Even though described the existing regulatory system as "fragmented, the Cal PX and Cal ISO were under the regulatory jurisdic- outdated, arcane and unjustifiably complex." tion of FERC (the national electricity regulator), the CPUC and the state government had substantial de facto influence · Expectation. over their actions. The two regulatory entities, the CPUC - The new market system would lower prices by encouraging and FERC, sometimes issued conflicting orders. competition among existing and new wholesale and retail · The coverage of the reform was incomplete. Municipal utili- suppliers and by reducing regulation. ties were given the option of not participating in these new arrangements. In general, they chose not to participate. How the California Reform Differs from states had less impact because, unlike California, only 5 to Other Power Sector Reforms 10 percent of their overall supplies are purchased on the · Initially, the major private distribution companies were not spot market. allowed to buy outside of the spot markets. (No vesting or · Mandated rolling blackouts throughout the state since forward contracting was allowed.) Hence, they were totally December 2000 seriously disrupted the state economy (the exposed to the price volatility of the Cal PX spot markets. sixth largest in the world). Even more widespread blackouts · Distribution companies and others who serve retail cus- are expected in the upcoming summer. tomers were not required to own or have under contract · Some evidence indicates that the growing shortage of gen- sufficient generation capacity to meet their peak demands. erating capacity, combined with certain features of the com- · No provision was made for passthrough of wholesale pur- plex wholesale market design, may have allowed some gen- chase power costs to retail rates until full recovery of strand- erators to exercise market power. ed costs or March 2002 (whichever came first). · Limited or no pass-through of wholesale costs to retail cus- · The complicated design involved multiple, sequential whole- tomers has forced the two largest private companies to sale markets operated by two new separate entities (the Cal incur around $12 billion in unfunded liabilities since April PX and the Cal ISO). In other U.S. regions, the ISO and PX 2000. They are on the verge of bankruptcy. are combined in a single entity. · The Cal PX ceased to operate its two markets on January 31, 2001. 4 The Reform Process · The reform operated by "political consensus." The final ver- The Lessons sion of the reform package reflected a compromise among competing stakeholders. It was incorporated in a bill that Overall Design of the Power Market was passed unanimously by the California legislature. · A poorly designed power market will not operate properly, · Criticisms of the final design by outside power sector reform and inadequate attempts or delays in correcting market dis- experts were generally ignored by state and national politi- tortions will spill over into a serious financial crisis. cal and regulatory authorities. · The California power reform crisis offers many valuable les- sons on "what not to do" for reformers of power sectors, The Crisis particularly for the establishment and regulation of a mandatory, wholesale power market based on spot pricing. · The highly contentious siting and permitting process for new · The California experience indirectly provides important les- generating plants blocked the installation of any major new sons for the privatization and regulation of distribution generating plants for more than 10 years. California's enterprises and new market entities in developing countries, installed generating capacity declined by about 1,200 MW even though privatization was not an element of California's between 1997 and 2000. reform. · Wholesale markets operated by the Cal PX and Cal ISO · The California experience also provides a lesson about cri- worked reasonably well for the first two years (1996­98) sis management: there is no way out that is quick, painless while the initial surplus of generating capacity disappeared. or cheap. "Quick-fix" solutions to basic design flaws usually Less than 2 percent of residential customers exercised their fail and may aggravate the problems. Any real solutions will option to pick new electricity suppliers because new suppli- impose heavy costs on stakeholders such as suppliers, con- ers could not offer substantial reductions in consumers' sumers, shareholders, and legislators. electricity bills under the rate freeze and competitive transi- tion charge during the reform transition period. Requirements for Competition to Work in the · A shift in market fundamentals occurred: large increases in Wholesale Power Market electricity demand in California and neighboring states, · Spot markets for wholesale power require careful design of reduced availability of hydropower in California and the market rules and price regulation to allow participants to Pacific Northwest, and big increases in the prices of gas manage their trading risks efficiently. and pollution permits to emit nitrogen oxides (NOx). · Competition requires adequate capacity to meet demand · Wholesale spot prices skyrocketed starting in the spring of without experiencing supply constraints (generation, trans- 2000. California utilities paid around $11 billion more for mission, fuel, etc.). The market must provide signals and electricity in the summer of 2000 than in the summer of incentives for investment in new generating capacity when 1999. Similar wholesale price increases in neighboring needed. These can be provided by various means, such as imposing a capacity obligation on distribution companies · Vesting contracts should be allowed as a form of insurance purchasing power in the market, setting up a parallel for distributors purchasing from a new spot market. A vest- capacity market to the energy spot market, or developing a ing contract that fixes the sale price for trade between exist- forward energy trading market whose prices signal expecta- ing or new generators and distributors for five or more years tions about future supply/demand balances. should be established before the market goes into opera- · Competition requires that investors in new supply capacity tion. They also provide at least initial protection against face no major barriers to entry to the wholesale power mar- market power. ket. These barriers include uncertainty and expense in facing · The spot market can evolve from a cost-based to a price- delays to the permitting process, regulatory uncertainty based system as the power market becomes more competitive. about after-the-fact price reviews, and regulatory constraints · Alternative trading arrangements to spot markets, such as on managing trading risks efficiently by means such as bilateral trading among multiple buyers and multiple sellers, hedging instruments. should be considered for small power systems and as transi- · The design of a competitive power market is too complex tional arrangements until the benefits of a spot market are and delicate to be dominated by heavy political compromis- considered to outweigh the risks. es that are intended to shield stakeholders from the conse- · Bilateral trading becomes unsustainable as the only trading quences of the reform. Market design should be firmly guid- method when the complexity of balancing system supply ed by sound economic principles. with demand in real time becomes unmanageable as the number of buyers and sellers increase. Commercial transac- 5 · New competitive trading arrangements in a wholesale tions cannot be divorced from physical realities of power power market should be introduced carefully to provide system operation. scope for dealing with design flaws as well as settling-in · A temporary single-buyer arrangement can be considered-- problems. but with strong reservations--in situations where bilateral trading or spot markets need substantial time for develop- Introducing Competition to the ment of power purchasers and sellers. Wholesale Power Market · Most developing countries should start with limited forms of Introducing Competition to the Retail Power Market competition that can evolve to full wholesale competition · Retail tariffs should be aligned with the costs of wholesale through spot markets once the sector can manage full com- power. Regulators should avoid rate freezes that expose dis- petition without uncontrollable market power. The creation tributors to the possibility of an unsustainable squeeze on of bid-based spot markets should generally not be their top their cash flow occurring when rising wholesale power costs priority. approach or even exceed fixed retail rates. · A mandated, deregulated, wholesale bid-based spot market · Regulators should encourage and even require suppliers to should be pursued only if certain conditions are likely to be allow large users to adjust their demand for power in real satisfied. Some of these prerequisites are also required for time, through smart metering and other means, since com- other, more limited forms of competition. But the conse- petition works properly only when both suppliers and users quences of not satisfying these conditions are most dramatic interact in the market (i.e., prices must be seen by both the and harmful in a mandated and deregulated spot market. demand and supply sides of the market). · Price-based spot markets are generally too risky for small- · Interruptible supply tariffs work only when consumers do not to-medium­sized power systems because of these systems expect to be called more than occasionally to reduce their will lack sufficient bidders to maintain effective competition. demand on the power system. Power outages are enor- · Cost-based spot markets, such as those developed in Latin mously costly for consumers who have already adjusted to America, offer a simpler and less risky alternative that can using grid power. Hence blackouts are symptomatic of yield competitive benefits for medium-sized power systems, enormous macroeconomic losses. This shows in turn the complemented by imposing a capacity obligation on distri- potential gains from reforming systems in such a way that bution companies. such a situation is avoided. · Likewise, it is simpler and less risky to impose obligations on · Small retail power users should have the option of avoiding generators and distributors to provide ancillary services (i.e., exposure to the high price volatility that can occur in spot grid support services) as a condition for being connected to markets for power. Power suppliers or other entities should the grid, rather than trying to synchronize one or more sep- be given regulatory scope to absorb this volatility through arate markets for ancillary services with an untested spot risk management techniques. energy market. · One or more commercially viable entities must have a legal requiring that the system operator's governing board be obligation to provide adequate supplies for consumers who composed of non-market participants (i.e., non-stakehold- prefer to deal with a default supplier rather than shop ers). Governance boards composed of stakeholders should around in the market for a supplier. not be too large or dominated by one or more classes of · In countries where the power supply industry is under state market participants. ownership and is due to be privatized and opened up to · Price caps should be used only as a last resort, since they competition, stranded costs for past investments by utilities introduce distortions with unintended consequences and do need not be recovered through surcharges on consumers' not correct the causes of the problem that they address. bills. This is because these costs will generally be absorbed · The system operator should monitor markets carefully and by the state through the proceeds received from the sale of continuously for signs of trouble--such as unusual price these assets. movements that may indicate abuse of market power--and · Full retail competition should be saved for last. In countries give the system operator the authority to penalize those who that have not achieved substantial household electrification, violate market rules. it will generally be more productive to focus on encourag- · An independent and expert market surveillance group ing competition to serve those who do not currently have should be created outside of the system operator. It should access to electricity, than on retail competition for those issue periodic public reports assessing the state of the market who already have access. and mobilize quickly when a problem arises. The members of 6 the group must be perceived as independent and objective. Regulating Power Markets · Regulation of fuel and power markets should be coordinated, · The economic regulatory system must be open, independ- especially the linkage between electricity and natural gas ent, credible and not prone to bankrupting reasonably effi- markets when most new generating plant burns natural gas. cient firms. · In large countries it is important to divide regulatory respon- · Regulatory "certainty" for power purchases by distributors is sibilities rationally between national and state regulators to of no value if, as in California, it can lead to bankruptcy of avoid unnecessary conflicts. It is not enough to simply say, efficient firms. The regulatory system must be designed to as in India, that electricity is a "concurrent subject" with reg- allow the cost of power purchases that are beyond the con- ulation shared by national and state regulatory authorities. trol of a distributor (e.g., mandated purchases in the spot The nature of the "sharing" has to be defined precisely to market, assigned purchases under a vesting contract or pur- avoid costly and distracting conflicts. chases under a previously reviewed bulk supply tariff) to be · The economic regulator for the power sector and the envi- automatically passed through in retail tariffs. ronmental regulator need to work together. Each is in a · If there is a spot market, the regulator should encourage position to undermine the work of the other. The ultimate hedging by allowing distribution entities to recover hedging success of both regulators requires a change in their mind- costs if hedging opportunities are available (rather than for- sets. The power regulator has to accept that compliance bid it until it is too late, as in California). with strict environmental standards is an integral element of · The governance of the system operator should be kept inde- power sector reform. The environment regulator must rec- pendent of the market participants. Independence can be ognize the need to work constructively with developers of achieved directly by prohibiting market participants from new generating plants to help achieve compliance with having an ownership interest in the system operator and agreed-upon environmental standards. PART I Those countries that have adopted this approach in Latin LESSONS FROM CALIFORNIA OR America have generally experienced significant increases in WHAT THE POWER MINISTER NEEDS TO KNOW private investment combined with clear improvements in oper- ating efficiency. 1. Start with limited forms of competition that The designers of the reform should consider imposing two can evolve to full wholesale competition. types of obligations: Competition is intended to produce operational and invest- · A capacity obligation on distribution enterprises and other ment efficiencies. Alternative forms of competition exist that load-serving entities to avoid complete reliance on a new are less complex than the mandated, centralized competition short-term market to induce investments in new generation model adopted in California. These alternatives can be imple- capacity. mented separately or in combination. None of these alterna- This requirement--currently in effect in Eastern United tives precludes moving to a deregulated, bid-based spot mar- States, Texas and several Latin American countries--means ket in the future. that anyone who sells electricity to retail customers must also have enough generation capacity (either owned or 1.1 Cost-Based Spot Markets with Obligations for under contract) to meet customer demands. An alternative, Capacity and Ancillary Services used in Chile and Argentina, is to require that the pool or sys- 7 If participation in a competitive wholesale market is mandat- tem operator acquire capacity from generators on behalf of ed, then a less risky alternative is to begin with cost-based bid- those who buy from the pool using administratively determined ding (as in four Latin American countries and in New England capacity payments that are in addition to the pool price. These until recently, and as proposed for Ghana) rather than price- two approaches would work in either a cost-based or a bid- based bidding (as in California, Colombia, El Salvador and based spot market. Only California appears to have introduced the United Kingdom). a mandatory spot market without any accompanying capacity obligation or capacity payment mechanism. A cost-based spot market based on generators' actual or esti- mated variable production costs is easier to establish and pro- · Initial obligations on generators and distributors to provide vides more protection against market power than a bid-based ancillary services (i.e., grid support services) as a condition spot market. It represents a relatively natural extension from for being connected to the grid. the traditional merit-order dispatch systems used in many pre- As practiced in Latin America, England and Wales, this is reform, vertically integrated power systems. While the cost- generally easier than trying to synchronize one or more sep- based bid market determines day-to-day dispatch patterns, arate markets for ancillary services with an untested spot there is usually a parallel "free" market in which generators, energy market. Once the basic energy market is functioning distributors and others can enter into hedging contracts to lock well, it may be less costly to acquire ancillary services in future prices and revenues. After several years of opera- through market mechanisms. tional experience, the cost-based spot market can evolve into a bid-based spot market. The three principal advantages of a 1.2 Multiple Buyers, Multiple Sellers cost-based market or pool are that it in Bilateral Markets Pools that operate a mandatory spot market, whether bid- or 1. Ensures efficient dispatch (if generators tell the truth about cost-based, are one form of a multi-buyer, multi-seller market. their production costs), There is, however, an alternative form of a multi-buyer, multi- 2. Makes it difficult for generators to exercise market power, seller market that does not require creating a pool. This alter- and native allows distributors, large industrials or both to buy 3. Is easier to implement because it builds on what the system directly from generators and other suppliers through one-on- or grid operator was doing prior to the reform. one bilaterally negotiated transactions. The bilateral transac- tions could be for short-, intermediate- or long-term supplies. North Americans and Europeans often consider the Latin It has been suggested that this type of market would be easier American cost-based approach to spot markets an inferior to implement in developing countries because it would be vol- form of competition. But the reality is that it has worked even untary and does not require the complicated protocols or soft- if it does not fit a textbook definition of perfect competition. ware of a mandatory spot market. Such voluntary markets involving one-on-one bilateral trans- workable once the number of buyers and sellers rises above a actions have existed for many years in the United States and threshold level because it becomes increasingly difficult to continental Europe. One big difference, however, is that the match a group of bilaterally negotiated power-sales agree- buyers and sellers were usually vertically integrated utilities ments of varying durations. These agreements produce hard- with sufficient generating capacity to meet all of their energy to-predict physical demands on the grid, requiring a grid needs. In general, these traditional vertically integrated utilities operator to balance the overall supply and demand of elec- participated in these markets to "fine-tune" their supply needs tricity on a moment-to-moment basis. (i.e., to lower their supply costs in certain hours rather than to meet their basic supply needs). Allowing industrial customers to participate in such markets raises other concerns. If industrial customers have been subsi- The question then is whether this type of market is feasible in a dizing residential and other customers (which is the case in different type of industry structure. Specifically, is it a viable many countries), the industrial customers will no longer be a option in an unbundled power sector (separate enterprises for source of cross-subsidies if they can buy from other suppliers. generation, transmission, system operation, distribution and This, in turn, may lead to the need for a big immediate retail supply) in which buyers would have little or no supply of increase in retail tariffs for non-industrial customers, rather their own and therefore would have to rely on the market for than a series of phased-in increases over a longer period of most or all of their supply needs? Moreover, would it work in a time that could be managed by phasing the exodus of non- 8 developing country where there is simply not enough generat- industrial customers from the market. ing capacity to meet the demands of all connected customers? It is not enough to simply provide distribution companies with Because developing countries lack experience with this type of the opportunity to participate in such a market. Distribution market, there are no clear-cut answers. One major concern, companies must also be given incentives to be efficient and however, is the issue of "balancing." Even if a distribution intelligent buyers. In particular, the regulatory system must company is able to contract for all of its expected needs, its include explicit incentives that allow distribution companies to moment-to-moment demand will rarely be exactly equal to the earn higher profits if they find more economical supply sources. amount for which it has contracted. Therefore, there has to be some balancing mechanism. (This balancing problem does 1.3 Single-Buyer Model not occur when the trading is among vertically integrated The single-buyer model requires that all generation supplies enterprises because buyers will have their own generation sup- be procured by an entity specifically mandated to fulfill this plies as well as the technical capacity to self-balance.) If the function, and that this entity in turn be the only seller of bulk balancing mechanism is an organized market with more than power to distributors and large users of power. a few generators and distributors, the cost and complexity of This is the "toe in the water" approach to introducing compe- setting up this residual balancing market may be almost the tition. In principle, it is the most limited form of competition same as a full, mandatory spot market. because it allows competition only for one-time competitive procurements for relatively well-defined products--the supply One possible, less costly alternative to an organized balancing of base, intermediate or peaking power for a specified period market would for distributors to acquire most of their needs of time. In practice, however, it is often poorly implemented through one or more supply contracts with generators, and because the single-buyer entity is usually an existing state- then hire a generation company or the system operator to be owned power enterprise that is not a skilled buyer and that responsible for meeting the moment-to-moment fluctuations in may be forced into signing high-priced and poorly designed its demand. Under this approach, the balancing would be per- power-purchase agreements (PPAs) through political or com- formed by the hired agent rather than by a balancing market. mercial pressure exerted by its government owners. Furthermore, it carries a substantial risk that the political and Regardless of whether the underlying industry structure is bun- commercial interests that benefit from this approach will block dled or unbundled, it appears that voluntary bilateral markets further reform by ensuring that it remains in force. are feasible only if there is (1) little congestion on the grid (i.e., ample transmission capacity), (2) a small number of Although single buyers tend to be state-owned enterprises, buyers and sellers and (3) an independent operator who has state-owned entities usually have limited experience in pur- complete knowledge and effective operating control of the chasing power, and this lack of experience may put the future entire interconnected grid. This type of market may not be budget revenues of their governments at considerable risk. This could be the case in California, where an existing state Other options besides the single-buyer model exist for purchas- agency has become the de facto buyer for about 50 percent ing wholesale power in small power systems. One approach of the short- and long-term supply needs of the three privately worth considering is a "joint action agency." This is a common owned utilities. This happened because generators in model used by groups of small power systems in several parts California and neighboring states were no longer willing to of the United States, including California (e.g., the Northern sell to these three companies, which account for about 75 California Power Agency). A joint action agency is essentially a percent of California's retail sales, because the three compa- buying cooperative made up of small distribution systems that nies could no longer pay for their power purchases. But simply pool their demands and hire purchasing expertise. It is different replacing these companies with a state-controlled single buyer from the pure single-buyer model in two important respects. will clearly not be a solution if the three utilities are not First, the buying cooperative is an entity created and governed allowed to charge tariffs to their retail customers that are high by the buyers rather than a separate, government entity that is enough to allow them to pay for the power that they will now not accountable to its customers (the current norm in many purchase from the state agency. Almost exactly the same situa- developing countries). Second, it is voluntary. If a small distri- tion exists in the Indian state of Orissa, which, like California, bution system believes that it can do better job by purchasing was the first state in its country to undertake significant reform. on its own, it always has the option of "going off on its own" as The four privately-owned distribution companies in Orissa are long as it satisfies its previous purchase commitments. unable to pay for the power purchased by the state-owned single buyer because their retail tariffs have been set too low. 2. Move to a full bid-based spot market only 9 once the necessary conditions are in place. Because the single-buyer model in developing countries often postpones an essential element of reform (i.e., raising retail A full bid-based spot market provides helpful price signals prices to cover costs), it frequently forces governments to offer needed by consumers and potential investors when the neces- backup payment guarantees they usually can't afford because sary conditions are in place. It is not, however, the highest ultimate consumers are "insulated" (at least temporarily) from reform priority in a power sector that is starting from a base of bulk power costs. However, what consumers do not initially pervasive under-pricing, significant cross-subsidies, over- pay for in electricity rates, they (and those who do not have staffing, high technical and commercial losses and widespread access to electricity) will eventually pay for in higher taxes or in political interfer-ence. The danger of trying to create such a the reduction of other government services (e.g., hospitals, spot market too soon in the reform process is that the effort roads and schools) that are "crowded out" because of the required to make it work properly will divert attention and subsidies or guarantees that now go to the electricity sector. In resources from trying to solve more fundamental problems. It California it has been reported that a state government sur- is a potentially time-consuming distraction when more basic plus of several billion dollars will soon be exhausted because problems need to be addressed. of the need to cover power purchases by the state buying agent. Standard & Poor's, a U.S. credit rating agency, put the A mandated, deregulated, and bid-based spot market should state on a credit watch "with negative implications" when the be pursued only if certain conditions are likely to be satisfied. state began to purchase power. Some of these prerequisites are also required for other, more- limited forms of competition. But the consequences of not sat- Countries with small power systems may be tempted to consid- isfying these conditions will not be as dramatic or as harmful er adopting a single-buyer model because unbundling gener- as they would be in a mandated and deregulated spot market. ation and distribution into a number of small entities, com- The conditions include the following: bined with sophisticated market mechanisms, may not be a realistic option for such systems. In the more than 100 coun- · Market power is not pervasive. There are sufficient non- tries with installed capacity of less than 1,000 MW, the poten- affiliated suppliers in each segment of the system load curve, tial number of operators and distributors in the bulk supply no serious bottlenecks exist in the transmission system, and market may simply be too small to support workable, ongoing control of fuel supply is not under the control a major competition unless the country has strong interconnections to generator. This condition is unlikely to be fulfilled in a country neighboring countries. Moreover, trying to introduce sophisti- with a small power system and few interconnections with cated trading arrangements could divert attention from other power systems in neighboring countries. higher priorities, such increasing supply, reducing losses and · Distributors have the money to pay for their power purchas- providing electric power to those who are currently unserved. es and distribution costs (i.e., retail tariffs are cost reflective and are not artificially suppressed for political reasons). based spot market system involving multiple sellers and buy- Competitive power markets will fail unless distribution enti- ers requires significant expenditure on real-time metering, ties and other buyers are commercially solvent. California bidding protocols, settlement and market-making software started with commercially viable distribution entities but then and communication and data transmission equipment. pushed them towards bankruptcy by forcing them to buy in Much of these costs are independent of the size of the a spot market in which prices skyrocketed and the regulato- power market. California is a rich state, so it was able to ry system (which was the result of a political compromise) finance a veritable army of consultants working under prevented the two largest distributors from passing these extremely tight deadlines to install the necessary hardware, high bulk-power costs through to their retail customers. develop the protocols and write the corresponding software. · Buyers and sellers in a deregulated market have the means In contrast, most developing countries will not have these and incentive to hedge price volatility in forward spot mar- resources. And even if they did, these limited resources kets, through intermediate and long-term contracts, etc., and would produce bigger and more immediate benefits if used are not forced to rely completely on mandatory, short-term in extending service to unserved households, putting in retail bulk power markets. Apart from vesting contracts (see meters where such meters don't exist and making transmis- below), volatility in spot electricity prices can be hedged with sion and distribution investments to improve the basic quali- a variety of other financial instruments such as futures con- ty of current service. tracts, options and derivatives. The market for such instru- · There is a "workout" of high-priced power purchase agree- 10 ments are not easy to create, can be manipulated if there is ments with IPPs or an explicit stranded-cost mechanism in not enough volume and, more importantly, may divert atten- place before the market becomes operational. A wholesale tion from more critical "first order" tasks such as raising tar- market will generally not work unless this happens. iffs so that distribution entities can recover their total cost. · There are few bottlenecks on the transmission system that Policymakers sometimes fail to appreciate that it is more diffi- would block transactions and create segmented markets. If cult to create a bid-based spot market in electricity than in there are bottlenecks, a workable and efficient system exists other energy commodities because of the basic physical reali- for pricing congestion. For example, transactions in a day- ties of electricity production and consumption: ahead or hourly energy market should not be arranged in isolation from whatever congestion exists on the grid. · Electricity is very expensive to store. · The market and system operator are genuinely independent · It is subject to rapid changes in demand. in ownership and decision-making from market participants · There are pervasive externalities on the grid. For example, (generators, distributors, retail and wholesale suppliers and physical failure at one location can cause the collapse of final customers). The governance system in California the entire grid supply. resembled a mini-legislature and was susceptible to deadlocks. · Its demand and supply must be balanced on a moment-to- · New generation and transmission capacity can be built moment basis. without excessive delays in permitting and siting (i.e., supply · The demand for electricity (on a real-time basis) can be very can respond to market prices). In California, the susceptibil- unresponsive to price increases. ity of the siting and permitting process to legal challenges by nearby residents was a major barrier to entry for new 3. Allow vesting contracts as a form of insurance for generators. In developing countries, similar delays could be distributors purchasing from a new spot market. caused by weak environmental agencies that are adminis- tering cumbersome administrative processes. Before the market goes into operation, the government or its · Retail tariffs are designed so that at least large- and medi- privatization agency should establish a vesting contract that um-sized customers can "see" spot market prices on an fixes the sale price for trade between existing or new genera- hourly basis and can cut their consumption in response to tors and distributors for five or more years. (The same tech- high prices (i.e., demand can respond to high prices). nique, which is sometimes described as "allocated PPAs," can Consumers cannot "respond" to a price that they cannot also be used when new distribution entities are created even in see. California distribution companies are now pursuing a the absence of an accompanying spot market.) Vesting con- crash effort to install real-time meters and tariffs for their tracts provide "insurance" in case the market design is flawed, large customers before summer 2001. and they provide revenue and cost certainty to generators and · Sufficient time, money and human resources are available distributors in the early years of reform. In most countries that to develop the new market system. A fully developed, bid- have created short-term markets, vesting and other hedging instruments may cover as much as 80 to 90 percent of total Full retail competition (i.e., allowing every retail customer the power trade. This was not the case in California, however. The right to pick their electricity supplier over an existing distribu- largest distributors were required to sell generating plants and tion network) is expensive and complicated to implement. In were not allowed to repurchase the output of these plants England and Wales, it has been estimated that the initial hard- using vesting contracts. Instead, they were required to pur- ware (metering, data transfer and telecommunications systems) chase almost all of their supply needs in the newly created and software has cost more than US$1 billion so far. spot market. This is the functional equivalent of requiring that everyone buy their airplane tickets for a particular flight in a It appears that other countries (Australia and Norway) and mandatory auction that takes place 30 minutes before the other U.S. states (Pennsylvania) have had more success with scheduled departure. full retail competition than California. But it also important to remember that these countries, like California, are starting However, vesting contracts are not risk-free for distribution with full household electrification. companies. If the contract prices are high because of corrup- tion or a non-competitive or poorly negotiated procurement In countries that have not achieved substantial household process, future distribution companies and their customers electrification, it will generally be more productive to focus on may not be able to pay the high prices. In such cases, a vest- encouraging competition to serve those who do not currently ing contract will simply perpetuate a bad outcome and lead to have access to electricity, rather than on retail competition for "stranded costs" when and if competition is introduced. those who already have access. For example, in poor, rural 11 Starting power sector reform with a legacy of high-priced PPAs areas, the competition may be for the right to receive a gov- is like starting a race with a 20-kilogram weight on each leg. ernment subsidy (whether it is for capital, operating costs or both) in return for an obligation to provide a specified level of Vesting contracts can also be used with the creation of sepa- grid or off-grid service (Argentina and Chile). In other coun- rate distribution entities through privatization or divestiture, tries, privately or cooperatively owned mini-grids with an even if these actions are not accompanied by the creation of a accompanying generating unit (i.e., a mini-privatization) in spot electricity market. Such contracts reduce uncertainty for rural areas can be encouraged if regulatory licensing require- potential investors in both distribution and generation. They ments are kept to a minimum and the mini-grid providers are also allow the regulator to focus in the early post-privatization allowed to offer electrical service with lower quality-of-service years on distribution costs and performance (e.g., wires' costs, standards than the main grid distribution companies. If the technical and non-technical losses, billing and collections) that mini-grid operator wants the option of being connected to the are under the more direct control of distribution entities. main grid for enhanced reliability, then the key regulatory issue is the terms and conditions of the backup service that is pro- Vesting contracts are a transition mechanism. When the con- vided to it by the main grid distribution company or a separate tracts expire or when the distribution companies make addi- generation company. The general rule is that the regulator tional power purchases, the regulator will need to establish a should not impose regulatory requirements above and beyond system to ensure that the distribution entity purchases econom- the willingness and ability of people to pay. ically to protect its captive retail customers. And the regulatory system must provide incentives for distribution companies to Policymakers should also consider adopting a simpler version enter into a portfolio of purchase contracts to continue to of retail competition--by tying the energy prices paid by resi- hedge price risks. dential customers to a measure of market prices paid by industrial customers who have access to competing suppliers. 4. Save full retail competition for last. This "piggybacked" form of retail competition should be easier and less costly to implement than full retail competition. A Retail competition did not succeed in California for several variant of this approach has been adopted in Chile. reasons relating to the specific design features (e.g., a 10-per- cent mandated rate reduction combined with a rate freeze, the 5. Starting points matter. recovery of stranded costs through a competitive transition charge) of the California retail competition program. But even The starting conditions in power sectors vary enormously if California had been successful in introducing retail competi- among reforming countries. The "starting points" are particu- tion, this does not imply that most developing countries should larly important in four areas: make retail competition an early action in their reform programs. 1. Prices. lated bulk power market in a small country with weak inter- Are retail power prices above or below costs? In connections to neighboring countries. The better strategy is to California, the pre-reform prices were high, but in many privatize what already exists, provide subsidies for rural electri- developing countries the prices are too low to recover fication and strengthen interconnections to neighboring coun- costs. It is virtually impossible to undertake any serious tries (Central America) before contemplating a deregulated, power sector reform (including the creation of ongoing bulk bulk power market. power markets) unless a government is politically committed to closing the revenue-cost gap as its first priority. Basically, it makes little sense to start a power sector reform without first deciding which problems need to be solved. If a 2. Capacity. country moves too quickly to a complex bulk power market Is generation capacity adequate to meet the demand in that is inappropriate to its current problems, it runs the risk of the power market? In California, the reform started with a losing what may be a "once-in-a-generation" chance to make cushion of excess capacity, while many developing coun- fundamental reforms in its power sector. Power sector reform is tries have a shortage of capacity. Is there potentially a highly political process. Policymakers need to be alert to the enough within-country generation capacity (assuming weak fact that the necessary political support will quickly disappear interconnections to other countries) to make it worth think- unless the reforms produce "early wins" that are readily dis- ing about a national bulk power market? In Africa, among cernible to the general public. 12 the 34 sub-Saharan countries that each have less than 1,000 MW of installed capacity, spot markets and other 6. The economic regulatory system must be open, forms of ongoing bulk power competition, while interesting independent, credible and not prone to to read about, are largely irrelevant to their immediate bankrupting reasonably efficient firms. problems (unreliable service, high losses and insufficient generating capacity). Independent regulatory commissions are necessary but not sufficient for sustainable power sector reform. It matters little to 3. Coverage. investors that a regulatory commission is "independent" if the Is there full electrification? California has full electrification commission issues tariff decisions that make it difficult or coverage. In many developing countries in Asia, Africa and impossible for a reasonably efficient distribution company to Latin America, large segments of the population lack recover its total costs (purchase power plus wires costs). access to electricity. For example, of the 34 countries of sub-Sahara Africa, more than 90 percent of the countries 6.1 Distribution have less than 20 percent household electrification. Multi-Year Tariffs Most developing countries that have successfully privatized dis- 4. Institutions. tribution have given potential investors reasonable certainty Will investors and consumers trust regulatory and govern- about the initial revenue stream for 5 to 8 years through a ment institutions to honor commitments and treat them multi-year tariff formula that is fixed in the law or a concession fairly? In California, the state and national regulators have agreement (akin to a contract between the government and existed for more than 60 years and have established a the investors). Because this tariff-setting system is usually an good track record of honoring their commitments. In many integral and legally binding element of the overall privatization developing countries, the regulator is a new institution, its package, the regulator may have very little to do with setting responsibilities vis-ŕ-vis the government may not be clear, tariffs in the initial post-privatization period. This has been the and previous governments may have a history of reneging norm in Bolivia, Chile, El Salvador, Georgia, Guatemala, on agreements. In effect, there is often an "institution gap" Moldova and Peru. as well as a "supply gap." Multi-year tariffs are the regulatory equivalent of going on The reform transition strategy should reflect starting conditions "autopilot." Although they reduce risk for investors (and have and country characteristics. For example, in a country starting been adopted in almost every country that has successfully pri- with suppressed prices (i.e., prices that are less than costs) and vatized distribution), they may be difficult to implement if there a shortage of supply, there is a greater political risk to intro- is considerable uncertainty about the initial levels of cost, con- ducing deregulated bulk power competition than in another sumption and losses and the level of investment needed in country that starts with cost-reflective prices and a surplus of meters, lines, transformers and substations to raise service to supply. Similarly, it makes little sense to try to create a deregu- acceptable standards. A multi-year tariff that turns out to be too generous to new private companies also runs the risk of a Ireland, Scotland and the Netherlands are useful models. The political backlash that could lead to after-the-fact windfall choice of benchmarks is critical. Several Latin American coun- profit taxes or even re-nationalization. Consequently, it may tries have adopted an index based on six-month estimates of make sense in some countries to combine a multi-year tariff nodal prices as the purchased power benchmark. However, with a profit- and loss-sharing mechanism outside of a pre- they have found that the distribution companies will simply rely specified dead band. Such a sharing mechanism increases the on their legal right to buy all of their power needs at these political sustainability of the reform. prices and not attempt to engage in any hedging transactions. Any multi-year tariff should also be combined with perform- Blaming the Regulator ance standards so consumers can experience some improve- A politician can do few things more unpopular than raising ments in service to balance the pain of tariffs that are initially electricity tariffs. The Governor of California was quoted as likely to be higher. However, the performance benchmarks saying that he could have solved the crisis in "20 minutes" if must be developed with considerable care. In particular, any he had been willing to raise retail tariffs. Although political benchmarks must (1) take account of starting points (e.g., authorities in developing countries are often initially nervous in technical and non-technical losses on the system), (2) recognize allowing the creation of "independent" regulatory commis- that not all customers may want or can afford the same levels sions, they frequently discover the political convenience of of quality, (3) be able to be objectively measured and (4) be attributing necessary but unpopular tariff increases to the inde- bounded with respect to their financial impact on the enterprise. pendence of their regulatory commissions. The principal bene- 13 fit is the ability to say that the tariff increases are beyond one's Purchased Power control. For example, when the California Public Utilities Like regulatory "independence," regulatory "certainty" is of no Commission announced average retail tariff increases of 40 value if, as in California, it can lead to bankruptcy of efficient percent, on top of an earlier 10 percent increase, the Governor firms. For distribution companies, it is especially important that was quoted as saying: "I can't order or direct an independent the regulatory system must be designed to allow for the auto- body. I've not given any advice to them on the subject of a rate matic pass-through to retail tariffs of purchase power costs increase." (Washington Post, March 27, 2001, p. A2) that are beyond the control of the distributor (e.g., mandated purchases in the spot market, assigned purchases under a 6.2 Market Regulation and Monitoring vesting contract or purchases under a previously reviewed bulk Governance of System Operators supply tariff). Where the distributor has some discretion in its The governance of the system operator should be kept inde- purchases (e.g., post-privatization purchases for incremental pendent of the market participants. Independence can be demand growth), the regulatory system should create incen- achieved directly by prohibiting market participants from hav- tives for the distribution company to minimize its purchase ing an ownership interest in the system operator and requiring power costs. It appears that such incentives did not exist in that the system operator's governing board be composed of California. The privately owned utilities were generally reluc- non-market participants (i.e., non-stakeholders). But it may not tant to pursue potentially cost-reducing, long-term purchases always be possible or desirable to create a non-stakeholder in 1999 for fear that the purchases would be found "impru- board in some developing countries. Therefore, the alternative dent" in a later, after-the-fact regulatory review. is to create a stakeholder board where no entity or class can dominate board decisions. The failure of the California stake- Incentives to Hedge holder board suggests four lessons: If there is a spot market, the regulator should accommodate hedging by allowing distribution entities to recover hedging 1. The board cannot be too large or it will be ineffective as a costs if hedging opportunities are available (rather than forbid decision making body. (The California system operator it until it is too late, as in California). There needs to be a bal- board had 25 voting members before the Federal electrici- ance in the regulatory system. The regulator should not write a ty regulator dissolved it.) blank check by accepting all hedging costs, nor should the 2. The voting rules must ensure that one or two classes can- regulator discourage distributors from hedging because they not control the board's decisions. fear disallowance of profits under after-the-fact "prudency" 3. The regulator must be able to step in and make a decision reviews. The better approach would be to establish before-the- if the board is deadlocked. fact price benchmarks for wholesale power purchases to 4. Consumer representatives or advocates should be viewed encourage efficient buying. The indexed purchasing power as market participants. benchmarks created by the electricity regulators in Northern Price Caps nections with the industry. The surveillance group must have a Once a market has been created, price caps should be used broad mandate. It should be charged with assessing not only only as a last resort if serious structural or market design flaws the performance of the market, but also the actions of the sys- emerge. When prices go up, the natural instinct of most politi- tem operator and the regulator. (For example, in California cal authorities is to impose price caps. But price caps distort the market surveillance group has concluded that the "soft markets, and they treat symptoms rather than causes. If the price cap" imposed by the FERC would probably worsen the underlying problem is a shortage of generation capacity, a existing supply shortage.) Finally, the market surveillance price cap will not help with the two needed solutions: increas- group should work with the system operator but must have the ing supply and restraining demand. As the former FERC chair- clear right to issue reports without the prior approval of the man observed: "We cannot `price cap' California out of a system operator. supply shortage." Self-Regulation With any price cap, there is always a danger that it will be set Where organized spot or balancing markets are created, too low. For example, it appears that the price caps imposed industry "self-regulation" of the accompanying grid and com- in California were at times below the (historically high) vari- mercial codes should be encouraged. In California, these able production costs of some old generating units, and so technical advisory groups were able to make some technical prevented these units from operating profitably when the sys- improvements in grid and market operation. The regulator 14 tem needed their output. If price caps are put into place, they need not formally approve every decision or arbitrate every should be applied comprehensively across all markets in dispute, but the regulator must have the legal right to inter- which a generator might sell. If they are imposed piecemeal, vene on its own initiative or in the event of a formal complaint generators will simply sell in other markets where the price is by a market participant. not capped at all or capped at a higher level (as happened in California), thus defeating the purpose of the caps. Price caps Regulation of Fuel and Power Markets must be a temporary, last-resort measure. If they are kept in Regulators must coordinate the regulation of fuel and power place for too long, they will reduce the pressure to deal with markets--especially the linkage between electricity and natural the underlying problems and will ultimately prevent the market gas markets when most new generating plant burns natural from developing as originally planned (as happened with the gas. For example, if a generator is owned by or affiliated with wholesale electricity market in the Ukraine). a company that provides natural gas transportation to com- peting generators, this corporate relationship could be used to Monitoring by System Operators put its competitors at a competitive disadvantage. Regulators should require the system operator to monitor mar- kets carefully and continuously for signs of trouble--such as 6.3 Division of Authority between National unusual price movements that indicate abuse of market and State Regulators power--and give the system operator the authority to penalize In large countries (e.g., Argentina, India, Brazil, Canada, those who violate market rules. The system operator has China, Russia and the United States), it is important to divide detailed knowledge of daily operations and therefore is in a regulatory responsibilities rationally between the national and unique position to serve as the regulator's "eyes and ears." In state regulators to avoid unnecessary conflicts. It is not enough California, several (but not all) of the recommendations made to simply say, as in India, that electricity is a "concurrent sub- by the Cal ISO's monitoring unit, as well as an external moni- ject" with regulation shared by national and state regulatory toring unit (see below), were adopted by regulators. authorities. The nature of the "sharing" has to be defined pre- cisely to avoid costly and distracting conflicts. The areas of Monitoring by Outsiders regulation actions that are likely to cause friction include An independent and expert market-surveillance group should be created. It should issue periodic public reports assessing · Transmission siting and certification the state of the market and mobilize quickly when a problem · Transmission tariffs arises. The members of the group must be perceived as inde- · Bulk power tariffs pendent and objective. A small- or medium-sized country · Grid codes might have to hire experts from outside the country because · Commercial and governance rules for regional trading enti- most knowledgeable people within the country will be per- ties and grid operators. ceived, at least initially, as being biased because of past con- In California and the rest of the United States, the division of environmental regulators has been indifference to compliance by regulatory authority has not always been clear or appropriate. state-owned power entities because of government reluctance to Also, political authorities need to recognize that the division of face the costs of enforcing compliance. As power sectors become regulatory authority will probably have to change as the indus- increasingly privatized, however, governments and their envi- try structure changes. In particular, a division of regulatory ronmental regulators are re-discovering the local and global authority that may have been workable under a vertically inte- importance of compliance with environmental standards, and grated industry structure may break down as the industry are willing to put more effort into enforcing these standards. moves to an unbundled, vertically de-integrated structure. The California experience shows that reform of the way that the 6.4 A Caveat: Regulating State Enterprises Is power sector is regulated economically should be coordinated Different from Regulating Private Companies with environmental regulation of the sector. Environmental reg- Although California provides many useful lessons in "how not ulation contributed substantially to the high bulk supply prices to regulate," there is a hidden assumption behind these les- because it acted as a significant barrier to increasing the sup- sons. It is that the enterprise that is being regulated will ply of electricity in California. The problem was not so much respond to the incentives created by the regulatory regime. This the standards themselves (which continue to be strict), but how may not be true in many developing countries that have recent- they were implemented. Specifically, it took almost twice as ly created new, separate electricity regulatory bodies that are long to get state and local siting and permitting approvals for a regulating government-owned enterprises. These regulatory new generating plant in California as it did in any other U.S. 15 entities often borrow regulatory techniques that were developed state. The legal and political system allowed inhabitants near to exploit the profit-maximizing objectives of private companies, the sites of the proposed facilities and environmental groups to and try to apply these techniques to public enterprises. block or substantially delay the siting and permitting process for most new generating plants. As a consequence, supply stagnat- However, the inescapable reality is that most public enterpris- ed, while demand steadily increased. es, despite lengthy and expensive programs to "commercialize and corporatize" them, still usually act like public enterprises. The specifics of power sector environmental regulation--deter- In particular, because they do not pay much attention to prof- mining which pollutants should be controlled and at what lev- its and commercial performance, many of the attempts to cre- els, and deciding whether market or non-market control ate regulatory incentives are lost on them. As a consequence, mechanisms should be used--are beyond the scope of this regulators who find themselves regulating public enterprises paper. However, it is clear that decisions about the substance often spend considerable time writing impressive orders filled and process of environmental regulation cannot be undertak- with directives that, in the words of one new Indian electricity en in isolation from power sector reform decisions. Most elec- regulator, read like "pretty poetry" but which are "rarely read tricity regulators would prefer to oppose unduly restrictive envi- and almost always ignored." ronmental standards that raise costs at precisely the moment when electricity prices may need to go up for other reasons. While it is relatively easy to produce a list of regulatory lessons Similarly, most environmental regulators tend to take the nar- that can be learned from the California experience, many of row view that their mandate is only to ensure compliance with the lessons will be inapplicable to a developing country unless environmental standards. In particular, they do not feel any the state-owned power enterprise can be made to act like a real responsibility for the overall success of power sector commercial enterprise (which seems to be rare) or until the reform or, more immediately, whether a particular plant does state enterprise is privatized. or does not get built. 7. Economic and environmental regulators The reality is that these regulators need to work together. Each should talk to each other. one is in a position to undermine the work of the other. The ultimate success of both regulators requires a change in their In many developing countries, environmental standards that mindsets. The power regulator has to accept that compliance apply to the activities of state-owned power entities sector with strict environmental standards is an integral element of have been either non-existent or loosely enforced. Where power sector reform. The environment regulator must recog- standards exist, state-owned enterprises, operating with tight nize the need to work constructively with developers of new budgets and lax maintenance standards, have often acted as generating plants to help achieve compliance with agreed if compliance were a low priority. Similarly, the attitude of most upon environmental standards. PART II increased under a record-breaking heat wave; and (2) in the FROM REFORM TO CRISIS IN CALIFORNIA winter months of 2000­2001, when power supply fell sharply under seasonally low hydropower output and heavy withdrawals 1. Background from service of old thermal power units for maintenance. The reform of the California power market is often character- The serious nature of California's power crisis is shown by ized as a process of deregulation. In fact, the reform involved numerous indicators for the state's economy that has been the limited deregulation by introducing price-based competition in engine of high-technological growth in the United States. an elaborately structured wholesale power market, and it Resolution of the crisis is proving difficult and is imposing changed the way that the power market is regulated. It did not heavy costs on the stakeholders--suppliers, consumers, share- involve divestiture of state-owned assets. Hence the reform is holders, legislators, etc. The consensus is that there is no way more precisely characterized as part deregulation and part re- out of the crisis that will be quick, painless or cheap. regulation. The reform also involved some restructuring of market functions by: · Wholesale electricity prices during 2000 were more than three times the 1999 level. Huge spikes in wholesale power · Obliging the incumbent utilities to sell some of their power prices occurred during the summer months of 2000. The generating capacity to independent suppliers, market was declared dysfunctional by all who studied it 16 · Unbundling their distribution arms from their generation then. and transmission arms, · Retail electricity prices in the San Diego area in 2000 were · Placing responsibility for grid operation with an independent up to three times higher than in 1999; one household system operator, and reported, for example, an increase in monthly electricity bill · Establishing separate markets for energy and ancillary services from $129 to $353 for the mid-December to mid-January . period. Most U.S. states have started or plan to start programs to · The first sustained series for decades of brownouts and deregulate their power markets. California was one of the first blackouts occurred during the months of November 2000 to start because of its desire to lower its retail electricity prices. to February 2001, when system demand was seasonally Competition in the power market was introduced through low, forcing temporary closures of businesses and social divestiture of generating capacity by incumbent utilities, devel- institutions. opment of new power plants by IPPs, and extension of compe- · Industrial and commercial users of electricity have been tition gradually to retail supply. California's progress in adopt- paying massive penalties rather than cutting their power ing policies that give consumers the right to choose their elec- usage under interruptible supply contracts. Electricity is so tricity supplier--the key and ultimate indicator of competition vital for Silicon Valley that even a one-day power outage, in the market--ranks about average for the 24 U.S. states that such as the one that occurred in June 2000, reportedly cost have already implemented reforms, according to the Retail as much as $100 million in lost output. Energy Deregulation Index (a scorecard developed by the · The two main power utilities are facing bankruptcy, claiming Center for the Advancement of Energy Markets). that they have accumulated some $12 billion in uncompen- sated costs because of the high prices that they have been Part II of this paper proceeds in the following sections. First, it paying for wholesale electricity from power generators. Each summarizes the indicators and consequences of the California was losing around $400,000 per hour on electricity trading power crisis. It then outlines the main parameters of the during January 2001. They currently lack the credit to pur- California power market, describes the formation of the new chase wholesale power, and their debt rating has been power market under the 1996 reform, and reviews the factors slashed to junk-bond status. that led to the crisis. It concludes by assessing whether the cri- sis could have been avoided. The crisis has had the following immediate consequences: 2. The Indicators of the California Power Crisis · A Stage 3 alert to power consumers, which had seldom been declared up to the end of 2000, was declared for an The California power crisis of 2000-2001 has had two distinct unbroken series of 32 days during January and February phases: (1) during the summer months, when demand rose 2001. A Stage 3 alert is the severest indication of an sharply because the power load from air-conditioners impending power system brownout or blackout, when the system capacity reserve margin falls below 1.5 percent of so they have traditionally not competed with each other for peak demand. business, except for new industrial customers. · The state government has declared several dozen statewide · California currently has about 53,000 MW of installed gen- emergencies to urge consumers to conserve electricity, but erating capacity with the following distribution of ownership: this has not been much help. Public agencies comprising the LADWP and SMUD 23% · The financial crisis caused by the default on payments by the main utilities has threatened to spread to the banking Renewable energy producers and co-generators, supplying under long-term contracts based on Public community. Utilities Regulatory and Policy Act (PURPA) legislation 22% · The Federal Secretary for Energy invoked emergency powers Investor-owned utilities (IOUs) 15% on December 13, 2000, to order power generators to con- IPPs, most of which is held by five major power firms tinue selling into the California power market. (AES, Reliant, Duke, Southern and Dynergy) based · Natural gas suppliers threatened stoppage of deliveries of outside the state 40% natural gas to the main power utilities this winter, because they are concerned about the utilities' ability to honor pay- In addition, California's imports of power provide about ment commitments. 5,000 MW towards meeting system load. · The state government has enacted measures that place it firmly in the center of the California power market (e.g., · California's installed generating capacity by type of genera- becoming the principal buyer of energy for the two largest tor is as follows: 17 utilities), thus effectively flying against the world-wide trend towards deregulation and privatization of electricity trade. Hydropower 24% · The main organized wholesale energy market--the Coal-fired steam generators 6% California Power Exchange--has ceased to function effec- Oil and/or gas-fired steam generators 37% tively and faces extinction, because of the utilities' loss of Nuclear 8% credit on the exchange and a move to long-term contracts Combustion turbines and combined cycle plant 8% for bulk power in response to the crisis. Geothermal, wind, solar, municipal waste, etc. 17% · Serious power shortages in California are expected to con- tinue for the next two years, especially during the summer · The sources of the 275,800 GWh of wholesale supply months. of electricity in 1999 by type of energy resource were as · Serious impacts on California's economy are a concern, follows: including threats by businesses to move away, and the repercussions on the rest of the country. Hydropower 15% · Other states are reconsidering plans to deregulate their Coal 13% electricity markets. Nevada, for example, has postponed Oil and/or gas 31% power deregulation plans, in part to stop generators from Nuclear 15% selling electricity to higher-margin markets in California. Geothermal, wind, solar, municipal waste, etc. 8% Regulators in Arkansas are recommending a two-year delay Energy imports 18% to their plans. This distribution did not change much throughout the 1990s. 3. Main Parameters of the California Power Market · Peak load on California's interconnected power system in 2000 was about 51,400 MW including the loads on the The main parameters for the California power market in 2000 public agency systems. The breakdown of this load by serv- are summarized herewith. ice area was as follows: · Retail supply of electricity in California is dominated by three investor-owned utilities (IOUs)--Pacific Gas & Electric PG&E 41% (PG&E), Southern California Edison (SCE) and San Diego SCE 38% Gas and Electric (SDG&E)--and two municipal vertically SDG&E 6% integrated monopolies--Los Angeles Department of Water LADWP 10% and Power (LADWP) and the Sacramento Municipal Utility SMUD 5% Department(SMUD). Their service areas are discrete zones, · Retail electricity consumption by sector in 2000 was as fol- 4.1 New Market Structure lows: The reform established a new market structure (shown in Figure 1) that promotes competition. Separate markets were Residential 30% created for energy, transmission and ancillary services that are Commercial 36% procured every hour at market-priced rates through pool- Industrial 21% based transactions. Bilateral transactions are also allowed for Agricultural 7% some participants in the market. The structure was designed to Other categories 6% avoid imposing administratively determined commitments, such as capacity obligations, on market participants. · Retail electricity prices--expressed in terms of average tariff This new market structure was established by the following yield of U.S.cents/kWh--by consumer category for means: California during 2000 are given below. They show that California's electricity tariffs are about one-third higher than · A Power Exchange (Cal PX) was created by January 1998. Cal the U.S. average. PX is set up as a non-profit public benefit corporation under California legal statutes. It acts as a market place in which California U.S. Average generators and suppliers compete to meet demand for electric 18 Residential 10.6 8.3 energy. It functions as an auctioneer and as such does not Commercial 9.9 7.3 engage in energy trading on its own account. To ensure the Industrial 6.2 4.5 viability of Cal PX, the AB 1890 statute requires the IOUs to Other 3.7 6.1 sell energy produced from their own power stations (mainly All Sectors 9.0 6.7 hydro and nuclear) and purchase energy on behalf of cus- tomers who had not changed to another supplier (nearly all customers) from the PX during the four-year transition period 4. Formation of the New Power Market under the to 2002. Their retail arms--called Utility Distribution 1996 Reform Companies (UDCs)--and electricity marketers purchase energy from Cal PX and resell electricity to their customers. Before the reform, the IOUs were vertically integrated and · Independent System Operator (Cal ISO) was established to were able to recover their costs of generating and supplying operate the statewide transmission system impartially for electricity through the bundled rates that they charged their buyers and sellers of bulk electricity. Any supplier that meets customers, as long as the sector regulator--the California the regulated reliability standards has access to the system. Public Utility Commission (CPUC)--approved these costs as Cal ISO operates as an independent, non-profit agency. It being "reasonable" and prudently incurred. does not own any generation, transmission or distribution systems, and relies entirely on services supplied from its The reform of the California power market was implemented markets to meet the demands on the statewide power system. according to CPUC's restructuring order issued in December · The IOUs continue to own the transmission facilities and 1995, which led to the enactment of Assembly Bill 1890 (AB receive a fee for the use of these facilities. The Federal 1890) by the California legislature in September 1996. The Energy Regulatory Commission (FERC) regulates these objective of the reform was to reduce the costs of electricity transmission use fees and the Cal ISO system operation because California's electricity prices were much higher than fees, as well as many of the operating, commercial and the national average under traditional regulation. At the same technical protocols of Cal ISO and Cal PX. time, however, the concern was that competition would push · Other than the three Californian IOUs, participation in Cal wholesale prices so low as to render unviable the investments PX is voluntary for all buyers and sellers of bulk power such in new power capacity needed to meet growth in demand, as municipalities, IPPs and out-of-state producers. They can while exposing consumers to high price volatility. AB 1890 trade electricity using a variety of means (e.g., bilateral con- was thus designed to deal with these conflicting objectives. tracts). Figure 1. Electric Supply in California 19 · The non-PX participants must submit schedules with the Cal 4.2 New Market Operating Arrangements ISO through entities known as scheduling coordinators (SC). The reform established separate markets for electric energy, The SCs are the only point of contact between these partici- ancillary services, and congested transmission capacity that pants and Cal ISO, and they number around forty. They are operated in parallel by Cal ISO and Cal PX according to coordinate scheduling activities continuously, and each SC market operating procedures approved by FERC. They were submits a "balanced" schedule to the Cal ISO in which the launched in April 1998 (except for the Block-Forward market, quantity of energy supplied equals the quantity demanded. which was launched in July 1999). They are operated as auc- Cal PX also submits a day-ahead schedule to Cal ISO. tions carried out sequentially throughout the day, with bids for · PG&E and SCE were required to sell at least 50 percent of demand and supply. The final price is the highest supply bid their generation plants to IPPs or to place them in separate that is accepted to clear the market. new companies, in order to mitigate their market power by reducing their scope for anti-competitive "self-dealing." · The energy market is structured primarily as a day-ahead SDG&E was required to divest all its generation assets (but auction by Cal PX, with bidders allowed to submit different its parent company was allowed to merge with the local gas quantities and prices for each hour. This auction is accom- supplier). The capacity sold amounted to about 7,500 MW panied by hour-ahead auctions for energy to allow for by PG&E, 10,600 MW by SCE, and 2,200 MW by SDG&E, divergences in demand or supply from the day-ahead bids. totaling 20,300 MW. Hence ownership of about 40 percent Such divergences may occur from unexpected changes in of the total installed capacity in California was transferred weather conditions or generating plant availability. to IPPs. · The day-ahead and hour-ahead markets are independent · A California Energy Market Oversight Board was estab- and are closed separately. Upon closing, the winners are lished comprising members appointed by the state governor financially and operationally obligated to provide the servic- and legislature, in addition to large stakeholder governing es that are selected by Cal ISO. boards for the Cal ISO and Cal PX. · Since scheduled transactions seldom match the actual load on the power system, Cal ISO calculates, in real time, the amount of energy needed to balance total system demand. on Cal PX sales, as well as through the Cal ISO ancillary It conducts a real-time auction for providing supplemental services market. energy or for backing off demand to achieve this balance. · Open and flexible scheduling opportunities are characteris- Bidders submit prices up to 45 minutes prior to the start of tic of the market framework. For example, a generator may each operating hour. They indicate the prices at which they bid into multiple Cal ISO markets and have multiple deliv- are willing to change their generation or purchases in real ery points. It can have a bilateral transaction with another time. Cal ISO uses these bids to balance total generation market participant, sell a portion of its output to Cal PX, sell and load in real time. Prices are established in this market another portion to the Cal ISO ancillary services markets, every five minutes. and export a part of its output out of state. · Upon certification by Cal ISO, SCs can participate in any or · A generator faces a complex set of decisions concerning all any of the day-ahead, hour-ahead, and real-time mar- whether to sell capacity into an earlier or later auction, as kets. SCs are not required to schedule all of their expected well as between selling it for energy or ancillary services. load and generation in the day-ahead market. They may Each decision to sell potentially forecloses opportunities to elect to bid for less than their expected load in the day- sell into other markets. For example, a sequence of deci- ahead market, and then cover their remaining load in the sions facing a generator could be whether to bid (1) into hour-ahead energy market. Deviations from their day-ahead the day-ahead energy market at 7 a.m., for which the or hour-ahead schedules are allowed by Cal ISO, and settled results are declared about one-and-a-half hours later; (2) 20 on the basis of real-time energy imbalance market prices. into the ancillary services market at 11 a.m., for which the · Every day, Cal ISO collects energy schedules from the SCs results are declared by 1 p.m.; or (3) into auctions through- and assesses the viability of each schedule. Individual out the day in the hour-ahead and real-time energy mar- schedules accepted by Cal ISO are aggregated into a mas- kets. Market participants have the opportunity to place bids ter schedule that is checked to ensure that it can be accom- up to five hours before power flows in the Cal PX day- modated by California's bulk power grid in a reliable and ahead market, two hours ahead in the Cal ISO hour-ahead safe manner. If Cal ISO identifies power system problems market ancillary services market, and 45 minutes ahead in such as congestion in parts of the grid, it provides the mar- the Cal ISO real-time imbalance energy market. kets with an opportunity to adjust schedules in order to alle- · Apart from these centralized markets, there are separate viate the problems. bilateral transactions involving parties such as Californian · Cal PX operates a Block-Forward market that allows partici- generators who are not obligated to trade through the Cal pants to enter into electricity supply contracts for physical PX, out-of-state generators and Californian buyers other delivery up to six months into the future. These contracts than the three UDCs. provide a hedge against spot-market price volatility. · Cal ISO purchases ancillary services (for black starts, fre- Figure 2. Overview of Market Operations quency control, spinning, non-spinning and replacement Bids Bids reserve generating capacity available at short notice) in an unbundled manner from generators through long-term con- tracts and competitive bidding. Schedules Schedules · Cal ISO ensures reliable operation of the transmission grid of Other of Other PX Auction PX Auction Schedule Schedule by holding an auction for allocating congested transmission Coordinators Coordinators capacity among the various system users after Cal PX has established preliminary hourly day-ahead prices for energy. PX Schedules PX Schedules To facilitate this allocation, Cal ISO accepts "adjustment bids" for both the day-ahead and hour-ahead markets. Day Ahead Hour Day-Ahead & Real Time These bids reflect the prices at which SCs are willing to pro- Hour-Ahead Market Ahead Trade Market cure more energy or curtail loads from their preferred sched- Market Deliveries ules. If market participants do not submit sufficient adjust- California ­ ISO Ancillary Services Market ment bids, Cal ISO levies a congestion management charge on the schedule that utilizes congested transmission lines. · Generators receive no capacity payments or payments for Bids from PX Participants and Other Market Participants start-up costs in the energy market. Hence they must recover their fixed costs through direct payments received for energy The market operating arrangements are depicted in Figure 2. 4.3 New Market Regulatory Framework facilities during the 1980s. The high prices (averaging The reform changed the way the power market is regulated as around 17 U.S. cents per kWh) paid to the QFs under the follows: terms of these contracts would make these plants uncom- petitive under anticipated market conditions (i.e., condi- · Commitment of the contractually agreed capacity with Cal tions that prevailed before 2000) without the CTC. The ISO for a specified term (generally one to two years) of prices in many of these contracts were tied to CPUC pre- power plants sold by the IOUs as "Reliability Must Run" dictions of world oil prices, but these predictions proved (RMR) to maintain system stability and to overcome local to be inaccurate. congestion on the transmission system. "RMR" designation for a generating unit means that the · Imposition of a 10-percent rate reduction for all residential owner must commit to maintaining the unit and to respond- and small users from January 1, 1998, to last for four years. ing on a best-efforts basis to a directive from Cal ISO to This reduction was funded by the issuance in December operate the unit. The owners of RMR units are required to 1997 of $6 billion worth of 10-year rate-reduction bonds bid all of their contracted capacity into Cal PX. Hence they by a special purpose trust authorized by the state. do not participate fully in the Cal PX market. Ironically in view of the events during 2000, Cal ISO designated RMRs · Regulation of the distribution component of retail soon after the new market started because of concerns tariffs for the UDCs about ultra-low clearing prices in its imbalance market. In will be based on performance-based rate-making. 21 this situation, the relatively high-cost thermal power genera- tors in southern California would not win business in the · Initiation of retail competition. market and therefore have little incentive to participate in it. Suppliers have competed actively for the business of large Cal ISO was concerned about the availability and disper- commercial and industrial users. Retail competition has not sion of sufficient reserve capacity so that the transmission progressed beyond 2 percent of the market in the market system could absorb the loss of major transmission lines for residential users (except for a niche market for "green between northern and southern California. power") because of the freeze on retail rates and the inclu- sion of the CTC in customers' electricity bills. · Introduction of a competitive transition charge (CTC) on customers' electricity bills for the recovery of the IOUs' · The California Public Utilities Commission (CPUC) continues stranded costs arising from the introduction of competition. to regulate the UDCs' distribution activities. These costs refer to the relatively high operating costs and debt-service obligations (usually referred to as stranded In addition, fossil-fueled power generation is subject to strict costs) for some of the IOUs' generating plants built before and a rather unique environmental regulation that pre-dates the 1990s. The CTC is computed for each user's bill as the the 1996 power market reforms. In particular, a Regional difference between the regulated rate and the cost of sup- Clean Air Incentives Market (RECLAIM) for Nox. Retail ply. The regulated rate is frozen for all retail users until the Emissions Credits, or RTCs had been established with the total IOU that serves them has recovered its stranded costs under allowed emissions in a district to be lowered over time so as the CTC. California's utilities had recovered more than $11 to reduce urban smog. Regulated firms are allocated a fixed billion under the CTC by the summer of 2000, and SDG&E number of RTCs for NOx emissions for each year, and they had fully recovered its costs so that its rates were unfrozen. are required to redeem these RTCs according to the amount The transition cost-recovery period lasts up to December of their NOx emissions. Regulated firms can buy RTCs from 31, 2003, after which retail sales are no longer frozen by other firms to overcome a shortage for meeting their require- statute. ments, and sell RTCs in excess of their needs. These trades set - The CTC is also used to help recover the high costs of up a market in RTCs, both for the current year and for future power procured by the IOUs under PURPA-mandated years ("vintages") in order to prevent a "NOx spike" of higher- contracts with certain renewable generation and co-gen- than-anticipated emissions. Firms are not allowed to combine eration facilities (termed qualifying facilities, or QFs). RTCs of different vintages. These QFs provide up to 30 percent of the electricity pro- duced in California. This high proportion reflects the state's aggressive pursuit of electricity from these types of 5. Main Factors that Led to the Crisis · Lack of risk-mitigation options for distributors. The UDCs were not allowed full access to forward markets, The California crisis centered around the three UDCs and and so were not able to develop a risk-minimizing power port- their suppliers through the Cal PX. Other power entities, such folio. During 2000 they acquired only about 6 percent of their as the municipal utilities that chose not to participate in the energy from forward markets, in contrast to 34 percent from Cal PX, have not been so affected by crisis. This difference their own generating plants and 60 percent from other suppli- indicates that design flaws in the Cal PX market are a major ers on the Cal PX market. They were not even allowed to sell source of factors that led to the crisis. their power plants with long-term vesting contract protection Nevertheless, a number of factors exogenous to the market against price volatility. Instead, they have had to rely on design worsened the problems created by the design flaws. In volatile spot markets. Hence, they were forced to "sell long and particular, the crisis arose out of an unpredicted combination buy short," which is disastrous for a trader in any commodity. of events. Undoubtedly the most important was the shortage of power supply relative to demand. In the summer crisis, · Demand inelasticity. demand increased to around 51,400 MW--30 percent above Lack of demand elasticity by UDCs in the energy markets the winter level. In the winter the supply capacity was reduced arises from their inability to curtail their demand to avoid by more than 20 percent as thermal plants were taken out of paying high prices, because of their obligation to serve the service for deep maintenance, and an unusually dry end to the demands of their captive customers. Just as a relatively 22 year 2000 in the Pacific Northwest left reservoir levels low and small amount of tightening of the supply/demand balance thus limited the amount of hydropower that California could in the absence of any demand elasticity produced the sum- import. The other factors have exacerbated this problem. mer price spikes in the Cal PX market, so a relatively small amount of loosening of the supply/demand balance in the 5.1 Market Design Flaws presence of some demand elasticity would have significantly Structural and operational flaws in the California power mitigated the pressures that produced price spikes. market became evident within a year after the ISO and PX went operational in 1998: · Price caps. Facing virtually no supplies in the real-time balancing ener- · A mismatch between the regulated retail market and the gy market to meet system imbalances, the Cal ISO was deregulated wholesale market. authorized by FERC to impose during 2000 progressively While wholesale electricity prices and natural gas prices are lower "soft" price caps on bids in the real-time balancing deregulated, retail electricity prices are fixed for the UDCs energy market, starting at $750/MWh during the summer until they have recovered their stranded costs through the and dropping to $250/MWh by the end of the year. CTC or by December 31, 2003, whichever is sooner. Hence Payments made by the UDCs above the price cap would be increases in wholesale power costs cannot be passed through subject to scrutiny and cost-justification by Cal ISO in retro- to retail users, thus exposing the electricity distributors to spect. These levels would amply cover the costs of power huge potential losses under their obligation to serve their generation under normal trading conditions, but $250/MWh customers. This flaw does not become serious unless whole- was insufficient to cover even the variable operating costs of sale prices rise above the retail rates, which they were not the older power plants during the periods of very high gas expected to do at the time that the reform was being intro- prices and high costs of NOx emission permits. The situation duced. This flaw may be only transitory, but it has con- appeared to provoke generators into raising their bids for tributed to the onset of the crisis during the transition period. supply during off-peak periods to recover their losses under the price caps during peak periods. These caps also · Lack of economic incentives for adequate capacity to appeared to limit prices to below the opportunity costs of maintain supply reliability standards. other units providing replacement reserve, hydro units con- The UDCs were not obliged to contract capacity, nor were strained by lack of water, and thermal units constrained by generators recompensed specifically for providing capacity. emissions limits, as well as exporters to neighboring markets Long-term forward contracting of energy by the UDCs was which were also experiencing high prices. also not allowed. Finally, the lack of forward energy markets for some years ahead suppressed the price signals that · Market arbitrage by generators. would have helped the distributors and investors in generat- Since the markets for energy, transmission congestion rights ing capacity to assess the need for new capacity. and ancillary services are cleared sequentially, rather than together, Cal ISO faces heavy demands on coordination to created by transmission constraints. This potential takes the prevent arbitrage by market participants that leads to ineffi- form of artificial scarcity of power created by power genera- cient dispatch of generating plants and higher prices than tors to drive up prices and earn huge profits. The potential predicted under models of these competitive markets. This for abuse of market power by generators increases signifi- sequencing gives incentives to generators to collect high cantly during periods when supply falls short of demand. premiums for real-time energy and ancillary services by Some experts contend that the generators' exploitation of withholding supply (or by putting in such high bids as to be market power caused a significant portion of the huge price sure that they won't be accepted) from the day-ahead ener- spikes for a few hours during 2000 in the California whole- gy market, and then bidding more supply into the other sale electricity market. Others go further by alleging persist- markets. Such profit-maximizing incentives for generators ent and serious abuse of market power by generators. bidding into these multiple markets may account for some Likewise, some observers allege that common ownership of of the observed price spikes under supply shortages during one of the main gas suppliers and critical gas pipeline 2000. For example, a generator would set a bid in the capacity in southern California created the conditions for energy market for a segment of capacity to cover at least market power in this market. After auditing plant outages in the foregone expected earnings in the ancillary services California, however, FERC staff stated that they did not find market for that segment, and this bid could be a very high evidence of certain practices that indicate abuse of market hourly rate to cover these foregone earnings if the generator power by the audited companies. It is generally acknowl- expects the segment to be dispatched for only one or two edged that it is difficult to distinguish from available data 23 hours in the energy market. Likewise, prices in some mar- the exercise of inappropriate market power from the kets for ancillary services could be driven up by considera- exploitation of legitimate scarcity rents when a market is in tions of foregone earnings in markets for other ancillary short supply. services. Some observers also allege that the repeated rounds of bidding under the market structure provide gener- · Market governance. ators with scope to "game" the system by adjusting their Poor governance structures contributed to the problem. The bidding strategies to their advantage merely by observing large size and politicization of the boards of Cal ISO and each others bidding behavior without collusion in the Cal PX, through quotas of stakeholders each representing accepted legal sense. their own interests, hampered attempts to focus on getting the market to work. The governance arrangements for Cal · Market arbitrage by UDCs. PX give to some parties the voting power to block changes Since the Cal PX capped prices in the day-ahead energy to market rules, which was done out of concern about put- market at a much higher level ($2,500/MWh) than the Cal ting market power in the hands of the UDCs. This led to the ISO's cap in the real-time balancing market, the UDCs prohibition of trading on forward markets by the UDCs. have kept down their demand purchases in the day-ahead Likewise, it is alleged that generators have too much power market by under-scheduling their during hours when price in Cal ISO, which they have used to block proposals to spikes would otherwise be likely to occur. They have done force them to schedule their entire output in the day-ahead this to keep the price in this market below the cap in the market. In late 2000, FERC ordered the replacement of Cal real-time balancing market, thus effectively capping the rate ISO's stakeholder board by a non-stakeholder board. they pay at the lower level of the latter. Purchases on the real-time balancing spot market have constituted a higher · Retail competition. proportion of total traded energy in Cal PX (20­30 percent Less than 2 percent of California's retail electricity users of the total energy procured) than in other U.S. states and have migrated from the incumbent UDCs to alternative other countries that have forward contracts in their power Energy Service Providers (ESP). Most ESPs have exited the markets, since a balancing market usually handles less than California market after their failure to attract customers. The 5 percent of total trade. This feature appears to have con- failure to develop retail competition in California results tributed significantly to the large volatility in prices in Cal PX. from a policy of charging retail users a default price equal to the wholesale power price, rather than the retail market · Market power. price, and by allowing the UDCs the right to provide default The potential for market power is likely to exist in a deregu- service. Default service refers to electricity supply provided lated price-bid market such as the California wholesale to those customers that are not receiving service from a market, especially in the presence of local market segments competing supplier. It is a regulatory device used to smooth the transition to a competitive retail market or as a long- · Power stations and transmission facilities are old. Nearly 60 term alternative to it. The amount by which the default serv- percent of California's power plants are at least 30 years ice price exceeds the wholesale price dictates the level of old, and now need more maintenance and thus longer out- customer savings and supplier earnings, which are funda- age periods than modern power plants. The withdrawal of mental drivers of retail competition. Generally, the higher about 10,000 MW of this plant for maintenance, as usual the default price relative to the wholesale price, the more during the low-demand winter period, helped create the intense the competition and switching to new suppliers. end-2000 supply shortages. The presence of these flaws raises the issue of how the An Unexpected Increase in Demand process for reforming the California market was managed. A · The growth of Internet-based power consumption based on consensual process was adopted, so that interested parties Silicon Valley industries spearheaded a 25-percent increase influenced the design in ways that possibly caused these flaws. in statewide demand during the 1990s, but this statistic This process resulted from the difficulty in changing market hides the real problem. From 1988 to 1998, electricity structures when (1) the ownership of the means of supply is demand grew at an average rate of only 1.3 percent per diversified among private interests that possess property rights year. In 1999 and 2000, however, electricity demand on by virtue of their ownership and (2) other parties, such as con- the Cal ISO system surged unexpectedly. In June 2000, sumer and environmental advocacy groups, have the legal energy demand was 12.5 percent higher than in June 24 right to mount strong legal challenges in defense of their inter- 1999, and peak demand was 6.2 percent higher. ests, as in California. · Demand for electricity in the summer of 2000 was pushed up by air conditioning loads under the highest temperatures 5.2 Exogenous Factors for May to July recorded for 106 years. Constraints on Expanding Supply · Retail demand was not sensitive to increases in the costs of · No new power generation capacity has been commissioned wholesale power since the tariff rates for most consumers in in California since 1992 because (1) uncertainty about the California were frozen until the utilities collected all their new power market deterred investors until the new market stranded costs under a regulated surcharge on customers' structure and regulations were finalized in 1996, and (2) electricity bills. In addition, lack of demand elasticity by retail subsequently excessive delays occurred in obtaining siting electricity buyers arises because they only discover the prices permits for new power stations in the face of local opposi- that they are paying after the transaction, and then only in tion when investors submitted applications. terms of an average monthly price rather than hour-by-hour · Investors have been deterred from entering the California prices. Relatively few users have time-of-use (TOU) meters. power market by the expense and uncertainty of the extenu- · Failure to meet demand reliably for electricity--especially ated permitting process for new power stations and trans- through blackouts and brownouts--is enormously costly for mission lines, exacerbated by the ability of people dwelling power users who have already adjusted to using grid power. in the vicinity of the proposed facilities to initiate numerous Californian users of electricity showed their willingness to legal challenges. The propensity of California's consumer pay huge penalties under interruptible supply contracts and environmental groups to use ballot measures to oppose rather than reduce power consumption when called upon to new power plants has added to the delays and uncertainty do so by their suppliers. for investors in these plants. However, in the last two years the state has licensed nine new power plants (totaling A Steep Increase in the Cost of Wholesale Power During 2000 10,600 MW), and five (totaling 2,900 MW) are under con- · The market clearing price in the day-ahead Cal PX energy struction. These plants will contribute significantly to easing market oscillated between $25 and $50/MWh during the supply shortage, but only in about two years' time. 1998, 1999 and the first half of 2000, and then rocketed · A drop in imports of power from neighboring states to over $150/MWh in June, July and August 2000 during occurred because of low hydropower production caused by an extreme heat wave. The steep increase in price occurred a drought and a growth in demand for electricity in these when supply started to fall below demand, even though markets. Environmental safeguards to protect fish popula- prices did not move discernibly beforehand as the margin tions in the Pacific Northwest region further limited the water diminished between supply and demand. Electricity markets available for generating electricity. These imports formerly do not have the price stabilizing mechanism of buffer stocks provided an important source (20 percent) of California's because electricity cannot be stored economically. power needs, especially during the peak demand period in summer months. · The average price of natural gas across the country also 1980s, the CPUC authorized the creation of holding compa- shot up during 2000 due to growth in demand, because nies, in which the utilities were relegated to the status of sub- gas is the fuel of choice for the huge amount of power-gen- sidiaries. The parent companies were permitted to pursue erating capacity recently commissioned or under construc- other, unregulated businesses as long as those activities did tion. The shortage also reflects a slowdown in gas explo- not compromise the utilities' ability to serve customers or the ration during the second half of the 1990s, when oil and capital needs of the utilities. thus gas prices were low. This price increase occurred when much more gas was used in 2000 than in 1999 for gener- Independent audits of SCE and PG&E released by the CPUC ating power in California because of higher demand for recently showed that the UDCs transferred billions of dollars to power and lower supply from other power-generating sources. their parent companies during the first years of deregulation. · The price of natural gas in California reached extraordinari- The parent of SCE received $4.8 billion and the parent of ly high levels during a spell of cold weather in December PG&E received $4.6 billion between 1997 and 2000 from 2000 (gas is used for space heating as well as power gen- their Californian utilities. These funds were derived from the eration). In December gas sold daily on spot markets at sale of their Californian generating plants, the surpluses major terminals averaged around $11 per thousand cubic earned through the sale of power in Cal PX from their remain- feet (TCF), compared to around $2.5/TCF in the preceding ing generating plant, and the recovery of stranded costs under years. This increase in gas price added about $75/MWh to the CTC. The parents used this cash to finance most of their the operating cost of a typical old power plant in Southern dividends and for the acquisition or construction of power 25 California that was supplied with gas bought on the spot generating capacity in other states and abroad. The parent market. Daily prices reached at times more than $60/TCF companies of these UDCs have instituted so-called ring-fenc- at the southern border of California during the first week of ing provisions designed to prevent bankruptcy courts or any- December 2000, partly due to bottlenecks in the California one else from using the parents' unregulated assets to cover gas pipeline system. However, a large proportion of gas the debt of the UDCs. These steps have aroused considerable purchases by gas traders and suppliers was hedged, and controversy in California. hence they were less exposed to gas price volatility than UDCs were to electricity price volatility. 6. Could the crisis have been avoided? · The design of NOx emission regulations--restrictive levels of annual emission permits complemented by a market for In assessing the impact of the design of the California power emission credits--has caused owners of older generating market on the current crisis, the issue is whether design flaws plants in California to pay a high price for these credits. have made a serious situation unmanageable. The fact that this Given power supply shortages, these plants were under arrangement worked without major trouble for the first two years pressure to utilize their capacity above the level that would indicates how easy it was to fall into a false sense of security allow them to meet the NOx emission standards. In the while market fundamentals were heading for a crisis. In the case South Coast Air Quality Management District of California of California, these fundamentals were strongly rising demand, (SCAQMD), the allowed NOx level was reduced on July 1, no new capacity, decline in hydropower output, and surging 2000, which reduced the supply of NOx RTCs just when natural gas prices. Once the crisis hit the market, the opportuni- demand for them increased. Consequently the cost of a vin- ty for making adjustments smoothly had been lost and the tage 2000 RTC increased from around $3/lb. NOx impact was magnified by the flaws in the market design. between 1997 and mid-2000 to around $45/lb. NOx by end-2000. This increase in price for NOx emission credits Other states experienced spikes in wholesale electricity prices pushed up the variable operating costs of a typical Southern similar to those in California, but only for a few days at a California power plant by around $30/MWh. time. Only California experienced a persistent series of such spikes throughout the summer of 2000. Retail prices in some 5.3 Exodus of Funds by Utilities other states have also risen by similar proportions to the tre- The holding structure adopted by the three IOUs has enabled bling of rates in the San Diego area. Likewise, natural gas these companies to keep substantial funds out of reach of the prices have risen on average by similar amounts across the creditors of the UDCs as the latters' debt mounted through United States, although they have risen much more at times in 2000. If they had been available, these funds would have parts of southern California due to pipeline congestion. But been sufficient to defer the current financial crisis, and thus to the other states have not experienced the brownouts and provide some time for implementing corrective measures to financial crisis that afflict California. prevent the development of the financial crisis. From the mid- Two avoidable design flaws stand out: and CPUC generally granted these requests, sometimes months later. In July 2000, PG&E asked CPUC for emergency 1. UDC's unhedged exposure to spot prices, especially when authority to buy power outside Cal PX, which CPUC approved tight supply conditions were foreseeable. The regulators in August in the face of the full-blown crisis. eventually tried to help the UDCs diversify this risk, as described immediately below, but their efforts appeared to The UDCs sometimes hesitated to use their freedom fully to be a case of "too little, too late." enter into such contracts because of concern about CPUC's 2. Retail prices capped at levels that depended on low prices ability to cut their profits later in a "prudency review" if it in the wholesale power market for sustainability. Despite deemed the contract terms unacceptable. This might occur if intense political and consumer opposition, the CPUC has spot prices dropped below the level of prices under long-term recently approved an emergency rate increase of 9 to 15 contracts before the contracts expired. So both options open percent to relieve some of this pressure. to the UDCs were risky, and generally the spot market was chosen by them. The utilities could have tested the proposed structure in the market before taking irreversible steps, for example by offering The market based NOx credit trading system, whose perceived their generating plants for sale with vesting contracts on terms advantage is reduction in the cost of achieving compliance for that were affordable under the capped retail prices. A lack of the industry, in fact appeared to contribute to the increase in 26 takers from IPPs for such contracts would have indicated that marginal supply costs of electricity when supply was con- the proposed structure was unsustainable. strained in the Summer of 2000. For example, "NOx spikes" can occur on days when electricity demand is greatest (due to The higher-than-expected prices that the IPPs paid for the air-conditioning load, for example), because electricity spot IOU's generating plants possibly indicated that they expected prices can then rise sufficiently to encourage plant operators spot prices to be much higher than the levels at which the to pay high prices for NOx RTCs so as to run power plants at UDCs could survive within the capped retail rates. Other maximum output. This indicates the possibility of interaction explanations for these high observed prices are the potential between environmental and energy costs when both are deter- value of generation capacity on the plant's site, and the mined by market clearing prices. expectation of obtaining major gains in operating efficiency. Inadequate transition arrangements also appear to have con- After experiencing extreme (up to that time) price spikes during tributed to the crisis. The Californian "big-bang" approach to the summer of 1998 shortly after Cal PX opened, SCE sought deregulation is open to the risks of unexpected market condi- CPUC's permission to buy 2,000 MW--about 10 percent of tions, as well as the unexpected ability of participants to the peak summer demand of its customer base--outside the "game" the market. Market rules and highly sophisticated soft- Cal PX. This move was opposed by consumer groups, electric- ware, hardware and telecommunications systems were devel- ity sellers and other stakeholders. CPUC rejected SCE's oped in only 12 months, completely independently of any request on the grounds that such purchases would weaken market participants. A structured transition strategy is needed Cal PX and put the smaller electricity sellers at a competitive that is based on planning for steps that might be taken if cru- disadvantage on the Cal PX. cial assumptions, such as continuation of surplus power supply capacity and low natural gas prices, proved to be wrong. In Cal PX tried to help the UDCs protect themselves from price particular, the IOUs mistakenly anticipated earning huge mar- fluctuations by offering forward contracts for up to 18 months gins during four competition-free years in which to recover in April 1999. CPUC gave the UDCs permission to enter into their stranded costs. Cal ISO was forced to make ad-hoc such contracts, with limits on how much electricity they could adjustments such as introducing price caps to deal with these buy that way, and so Cal PX opened its Block-Forward market unexpected events; these adjustments provided quick fixes but in July 1999. As prices kept rising, the UDCs asked for more, led to further problems. California's inclination to rely on power imports, rather than Overall, three conclusions may be drawn from the California expand its own supply capacity, exposed it to developments power crisis: beyond its control. Neighboring states object to being energy farms for California, whereby the latter avoids the environ- 1. The flaws in the design of the California market con- mental consequences of building new generation capacity tributed substantially to the financial crisis of California's while benefiting from the output. They are also unhappy about main utilities. the increases in prices in their power markets that they attrib- 2. Efforts to deal with the crisis in the presence of these flaws ute to events in the California market. could not have succeeded. 3. A properly designed power market could have coped with One indicator of whether California could have avoided its the factors leading to the crisis. Because the reforms crisis by better market design is the existence of workable already undertaken in the California power market prevent deregulation of a power market elsewhere under similar mar- a return to the pre-reform structure, the state's only option ket conditions in the United States such as in Pennsylvania, is to correct these flaws and move forward to a better- Texas and Illinois. Another indicator of California's specific designed market. vulnerability is the experience of its neighboring states under similar supply constraints and growing demand. Wholesale power prices during the summer months of 2000 also rocket- ed in these states, partly due to the rise in California's whole- 27 sale power prices, but their utilities did not hit the severe finan- cial crisis that has floored the state's main utilities. In Pennsylvania, where the state restructured the electricity market with far less political influence on the design, the state PUC set a high cap on wholesale prices to secure an upper limit, and did not require utilities to sell their generation plants. Buyers and sellers are allowed to choose whether to exchange in the power pool or through direct contracts with financial hedging through "contracts-for-differences." A capacity market exists in parallel with the energy market. They have not experienced the shortages faced by Californian power users for these reasons and also because the Pennsylvania power system benefits from extensive intercon- nections with other regional power markets; also, coal is wide- ly used for power generation, which hedges against increases in natural gas prices. Independent power producers are devel- oping nearly 40,000 MW of new generation capacity in the state. Retail competition is promoted by a high default cost (considered to be too high by some commentators) and by mandatory reallocation of retail customers from the incumbent suppliers, so that around 10 percent of customers have switched supplier. SELECTED BIBLIOGRAPHY Alaywan, Ziad. "Evolution of the California Independent Kucewicz, William P. "Power Politics: Drawing Lessons from System Operator Markets." Electricity Journal 13, No. 6 (July California's Electricity Crisis." Unpublished paper. October 19, 2000). 2000. Available at www.geoinvestor.com/archives. Alonso-Zaldivar, Ricardo and Nicholas Riccardi. "Audit Finds Lynch, Michael W. "Don't Blame Deregulation for the Golden No Orchestration of Hikes." Los Angles Times, February 3, State's Electricity Snafus." Reason Magazine (January 4, 2001. Available at www.latimes.com. 2001). Available at www.reason.com. Berry, Carolyn and William Meroney. California Power Crisis: Moore, Adrian T. and Lynne Kiesling. Powering Up California: Implications for Power Sector Reform in Emerging Economies. Policy Alternatives for The California Energy Crisis. Reason Presentation at Brown Bag Lunch Seminar. Energy Markets and Public Policy Institute. January 2001. Available at Reform Thematic Group. The World Bank. January 11, 2001. www.rppi.org. Available from carolyn.berry@nera.com or william.meroney@ferc.fed.us. Murphy, Kim. "Energy Secretary Rejects Calls for Power Price Caps." Los Angeles Times. February 3, 2001. Available at Borenstein, Severin. The Trouble With Electricity Markets (and www.latimes.com. Some Solutions). University of California Energy Institute. Working Paper PWP-079. Program on Workable Energy Ruff, Larry E. "Where California Went Wrong." Unpublished Regulation. January 2001. Available at www.ucei.org. paper. Available at http://www.ksg.harvard.edu/hepg/. 28 Center for the Advancement of Energy Markets. Retail Energy Samuelson, Robert J. "The American Energy Fantasy." Deregulation (RED) Index 2001, Washington, D.C., February Newsweek Magazine. January 29, 2001. P. 47. 2001. Available at www.caem.org. Simon, Richard. "State Gets Little Sympathy at Hearing." LA CERA. Beyond California's Power Crisis: Impact, Solutions, and Times. February 3, 2001. Available at www.latimes.com. Lessons. Special Report for North American Power and Western Research Services. March 2001. Available from Smith, Rebecca. "Crossed Wires: Major Kinks Emerge in Gov. www.cera.com. Davis's Plan to Power California." The Wall Street Journal. March 8, 2001. p. 1. Hogan, William W. Coordination for Competition: Electricity Market Design Principles. AEI-Brookings Joint Center Stoft, Steve. Soft Price Caps and Underscheduling Penalties: Conference: "The California Electricity Market Meltdown: The How Would the FERC Plan Affect California Electricity Markets? End of Deregulation?" March 1, 2001. Available from University of California Energy Institute. Working Paper PWP- www.aei.brookings.org. 079. Program on Workable Energy Regulation. November 2000. Hildebrandt, Eric. Market Analysis Report. Presentation at the Market Surveillance Committee Meeting. California ISO. Wolak, Frank A. What Went Wrong with California's Re- December 1, 2000. Available from www.caiso.org. Structured Electricity Market (And How To Fix It). Presentation at the AEI-Brookings Joint Center for Regulatory Studies Joskow, Paul and Edward Kahn. "A Quantitative Analysis of Conference: "The California Electricity Market Meltdown: The Pricing Behavior In California's Wholesale Electricity Market End of Deregulation?" March 1, 2001. Available from During Summer 2000." Unpublished paper prepared for www.aei.brookings.org. Southern California Edison Company. November 22, 2000. Available at www.stoft.com/x/cal/20001121-Joskow-Kahn.pdf U.S. Federal Energy Regulatory Commission. Order Directing Remedies For California Wholesale Electric Markets. Docket Kahn, Michael and Loretta Lynch. California's Electricity No. EL00-95-000, et. al. December 15, 2000. Available at Options and Challenges: Report to Governor Gray Davis. www.ferc.fed.us/electric/bulkpower.htm. California Electricity Oversight Board and the California Public Utilities Commission. August 2, 2000. Available at http://www.cpuc.ca.gov/published/report. THE WORLD BANK GROUP The Energy and Mining Sector Board The World Bank 1818 H Street N.W. Washington, D.C. 20433 USA