CALIFORNIA POWER CRISIS Lessons for Developing Countries 23871 April 2001 Sector and, Assistanc e Ser. e ail v-S 'l T.' ~~~~~~~~~~ i- Energy Energy Sector and Management Mining Assistance Sector Programme Board BSMAP,^ U JOINT UNDP / WORLD BANK ENERGY SECTOR MANAGEMENT ASSISTANCE PROGRAMME (ESMAP) PURPOSE The Joint UNDP/World Bank Energy Sector Management Assistance Programme (ESMAP) is a special global technical assistance program run as part of the World Bank's Energy, Mining and Telecommunications Department. ESMAP provides advice to governments on sustainable en- ergy development. Established with the support of UNDP and bilateral official donors in 1983, it focuses on the role of energy in the development process with the objective of contributing to poverty alleviation, improving living conditions and preserving the environment in developing countries and transition economies. ESMAP centers its interventions on three priority areas: sector reform and restructuring; access to modern energy for the poorest; and promotion of sus- tainable energy practices. GOVERNANCE AND OPERATIONS ESMAP is governed by a Consultative Group (ESMAP CG) composed of representatives of the UNDP and World Bank, other donors, and development experts from regions benefiting from ESMAP's assistance. The ESMAP CG is chaired by a World Bank Vice President, and advised by a Technical Advisory Group (TAG) of four independent energy experts that reviews the Pro- gramme's strategic agenda, its work plan, and its achievements. ESMAP relies on a cadre of en- gineers, energy planners, and economists from the World Bank to conduct its activities under the guidance of the Manager of ESMAP, responsible for administering the Programme. FUNDING ESMAP is a cooperative effort supported over the years by the World Bank, the UNDP and other United Nations agencies, the European Union, the Organization of American States (OAS), the Latin American Energy Organization (OLADE), and public and private donors from countries including Australia, Belgium, Canada, Denmark, Germany, Finland, France, Iceland, Ireland, It- aly, Japan, the Netherlands, New Zealand, Norway, Portugal, Sweden, Switzerland, the United Kingdom, and the United States of America. FURTHER INFORMATION An up-to-date listing of completed ESMAP projects is appended to this report. For further in- formation, a copy of the ESMAP Annual Report, or copies of project reports, contact: ESMAP c/o Energy and Water The World Bank 1818 H Street, NW Washington, DC 20433 U.S.A. CALIFORNIA POWER CRISIS Lessons for Developing Countries A P R I L 2 0 0 1 Energy Energy Sector and Management Mining Assistance Sector Programme Board |3ES1AA4TD W nU EF, 601 v 14 11 ~~~~~~~~~The World Bank Copyright ( 2001 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing April 2001 ESMAP Reports are published to communicate the results of the ESMAP's work to the development community with the least possible delay. The type- script of the paper therefore has not been prepared in accordance with the proce- dures appropriate to formal documents. Some sources cited in this paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, or its affiliated organizations, or to members of its Board of Ex- ecutive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no re- sponsibility whatsoever for any consequence of their use. The Boundaries, col- ors, denominations, other information shown on any map in this volume do not imply on the part of the World Bank Group any judgement on the legal status of any territory or the endorsement or acceptance of such boundaries. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the ESMAP Manager at the address shown in the copyright notice above. ESMAP encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Contents 4. Save full retail competition for last . ............ 18 5. Starting points matter ............................... 18 6. The economic regulatory system must be Contents. iii open, independent, credible and not prone Acknowledgements ...........................................v to bankrupting reasonably efficient firms. .20 Foreword ............................. 6.1 Distribution ........................... 20 Acronyms .................................. 6.2 Market Regulation and Monitoring ........... 21 UnitsofMeasurement............... 6.3 Division of Authority between National and State Regulators ................................ 23 Introduction..........................................1 6.4 A Caveat: Regulating State Enterprises Is Overview of the California Reform and its Different from Regulating Private Lessons ...............3 Companies ...... 23 Why the Reform? ...........3 7. Economic and environmental regulators The Nature of the Reform ..................,,.3 should talk to each other . ................ 24 Key Features ........................3 Part II From Reform to Crisis in How the California Reform Differs from California .27 Other Power Sector Reforms ....................4 1. Background ......... 27 The Reform Process .............4 2. The Indicators of the California Power The Crisis ......4 Crisis .. 27 The Lessons ........................................ .5 3. Main Parameters of the California Power The Lessons ..Market ...,,,,,,,.,..,,,29 Overall~ ~ ~ ~ ~~~~~~~~~~~~~~~~Mre .........................................................29 Overall Design of the Power Market .................4. Formation of the New Power Market under Requirements for Competition to Work in the the 1996 Reform .30 Wholesale Power Market ............................6 4.1 New Market Structure ......... 30 Introducing Competition to the Wholesale Power Market.6 4.2 New Market Operating Arrangements..32 Power M arket ..............................................6 Introducing Competition to the Retail Power 4.3 New Market Regulatory Framework ..34 Market ....................................,,,,,,,,,,,,..7 5. Main Factors that Led to the Crisis . 36 Regulating Power Markets . ...........8 5.1 Market Design Flaws .36 Part I Lessons from California or What the 5.2 Exogenous Factors . .39 Power Minister Needs to Know .............. 11 5.3 Exodus of Funds by Utilities ..................... 41 1. Start with limited forms of competition that can evolve to full wholesale competition.. 11 6. Could the crisis have been avoided? ........... 42 1.1 Cost-Based Spot Markets with Selected Bibliography . .45 Obligations for Capacity and Ancillary Services ............. 11 1.2 Multiple Buyers, Multiple Sellers in Bilateral Markets ................ 12 1.3 Single-Buyer Model ................ 13 2. Move to a full bid-based spot market only once the necessary conditions are in place. 15 3. Allow vesting contracts as a form of insurance for distributors purchasing from a new spot market ............. 17 Al Acknowledgements 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 Associates). 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. v IA Foreword Deregulation of power markets would be rejected on false grounds if the causes of the The U.S. electric power industry, the last California crisis were largely specific to the major energy industry in that country subject design of the California reform. In view of to traditional utility regulation, is being this uncertainty, the World Bank has a duty opened up to widespread competition. Some to its clients and itself to gain an states allow their retail electricity customers understanding of what has happened in to choose their electricity supplier. California, and to draw lessons from the Competitive trading of wholesale electricity California experience that are applicable to and the emergence of independent grid other countries. The purpose of this paper is operators have spread to many regions of the to fulfill this duty. In so doing, the paper United States. The number of independent also assesses whether the crisis could have power producers and marketers competing been avoided through better market design in the U.S. retail and wholesale power and management. Overall, the paper markets has increased substantially over the concludes that much of the crisis was past few years. avoidable. Nevertheless, the paper also identifies many invaluable lessons for other However, these new markets have not countries that are considering or emerged without problems. California implementing power sector reform, and I introduced competition to its retail and herewith commend it to all who are involved wholesale power markets in 1998, but has in this endeavor. experienced a major crisis during 2000 and JAMAL SAGHIR into 2001. This crisis has provoked a major debate about the risks, as well as the Director,CEnergy and Water rewards, of deregulating power markets to Chairman of Energy and Mining Sector Board allow competition. In fact, the California power crisis is giving deregulation a bad name, both in the United States and beyond to other countries that are reforming their power sectors. This characterization is somewhat misplaced, however, since the California reform is more precisely characterized as part deregulation and part re-regulation. Nevertheless, some observers argue that the California experiment with deregulation should be scrapped, while others argue that the deregulation is still a worthwhile endeavor to make the electric power industry more efficient and customer- oriented, and that problems such as California's can be solved by adjusting market rules. A third group argues that California's power crisis is a failure of market design rather than a failure of deregulation. vii Acronyms CPUC California Public Utilities Commission CTC competitive transition charge 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 Incen- tives Market RMR Reliability Must Run RTC retail emissions credit SCAQMD South Coast Air Quality Management District of Cali- fornia 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 viii Introduction governments have proposed or undertaken actions including the following: "Califormia's new electricity market ended up Price caps. The Federal Energy being designed in a highly politicized Regulatory Commission (FERC) process.... [W]hat emerged was the most imposed price caps that may deter the complicated electricity market ever created.." investment needed to overcome the -Professor Paul Joskow, The New York current supply shortage. Times, January 13, 2001 "This is a dreadful mess for a state that is held Forced sales. The U.S. Secretary of up around the world as a model of Energy issued several orders that innovation...." required generators and natural gas -The Economist, January 20, 2001, p.57 suppliers to continue selling to non- "Its problems are largely manmade." creditworthy California buyers. -Newsweek Magazine, April 3, 2001, p. 23 Government energy trader. A new state law authorizes the state government to The Califoria power crisis is so sudden and spend up to $10 billion on the state's . .l . . . ~~~~~credit to purchase wholesale electricity serious that it is prompting policymakers in many countries as well as other U.S. states that can be resold to the three large to look for lessons that can be applied to the privately owned utilities. reform of their own power sectors. * "Nationalization" of the grid. The State Concerned policymakers around the world of California may become the new are asking: If things can go so badly wrong owner of the portion of the high-voltage with a reform that did not involve wholesale transmission grid currently owned by the privatization of the electricity supply three large privately owned utilities. industry in such a rich and sophisticated "Balkanization" of wholesale electricity economy, what are the implications for trade. The state's Assembly has passed a much less well-endowed countries bill that would make it difficult to export embarking on the full menu of reform electricity produced from new including privatization? generating plants located in California to When a power sector reform like buyers outside the state. California's fails, political authorities are Many elements of the Califoria reform inevitably under strong pressure to "do package are peculiar to a complicated and something" to solve the crisis. In a recent unusual market design that was the outcome special session of the Califormia legislature of a political compromise reached by called by the goveror, legislators various stakeholder groups. Many of these introduced more than 75 bills intended to features will have no immediate, or even solve one or more aspects of the crisis, near-term, relevance for most developing Unfortunately, quick-fix "solutions" often countries. Since this paper has been written lead to outcomes that can be inconsistent mainly for power sector officials in with the original reform objectives and can produce outcomes that are even worse than deelopngconties,oit focue selively . . . ~~~~~~~~on the substantive lessons of the California the conditions that triggered the reform. For crisis that pertain to the design of power example, at the time of this writing (March 2001), the California and federal 1 In developing countries, the California of distribution enterprises and new market power crisis may be creating the impression entities in developing countries. that power reform is too risky. The power crisis in California does not justify this Part II details the specific reforms initiated conclusion. For many developing countries, in California, reviews the factors that led to the status quo in the power sector is the the crisis, and examines whether the crisis riskiest alternative of all. The status quo could have been avoided through better often creates a drag on economic growth market design and management. The paper through inadequate and poor-quality power draws on numerous sources such as supply. In addition, limited government published articles, reports and websites, as funds are frequently diverted to the power well as the working experience of World sector that would otherwise be available for Bank staff in numerous countries. schools, clinics and roads. Therefore, most countries simply have no alternative to a substantial and basic reform of the sector that almost always requires restructuring and privatization. But like all human endeavors, power sector reform can be done well or done poorly. The principal lesson of California is that good intentions are not enough. Any reform must pay close attention to starting points, the particular problems that need to be solved, and the appropriateness of the path selected for solving these problems. The paper is organized into three parts. It begins with an overview of the key features of the 1998 California power sector reform: how it differs from reforms elsewhere, the events and actions that have put it in a crisis mode, and the main lessons that can be learned from the crisis. Following the overview, the main text is divided into two parts. Part I discusses in depth the lessons learned, which concern mainly the establishment and regulation of a mandatory, wholesale power market based on spot pricing. Since this is not a near-term option for many developing countries, the paper also describes other, more-limited forms of competition that may suit their situations. Although privatization was not an element of California's reform, the state's experience does indirectly provide important lessons for the privatization and regulation 2 Overview of the California Public Utilities Commission (CPUC) Reform and its Lessons under a traditional U.S.-style cost-of- service regulatory system with some targeted incentive mechanisms. The Why the Reform? CPUC described the existing regulatory system as "fragmented, * California's economy in the early 1990s. outdated, arcane and unjustifiably Major statewide recession. High complex." unemployment. Loss of industry and * Expectation. The new market system jobs to other states. The state's governor would lower prices by encouraging believed that continued high electricity competition among existing and new prices (about 50 percent higher than the wholesale and retail suppliers and by U.S. national average in 1996) would reducing regulation. drive many industries out of the state. * Pre-reform electricity sector The Nature of the Reform - Three-fourths of the state's consumption was supplied by three Key Features large vertically, privately owned The three privately owned utilities were utilities: Pacific Gas & Electric "encouraged" to sell off their generating (PG&E), Southem Califomia Edison plants but without any vesting contracts (SCE) and San Diego Gas and to buy back the output of plants. Electric (SDG&E). The rest of the state was served by large and small * In return, the utilities were allowed to municipal utilities. recover their "stranded costs" (i.e., - High electricit pricesweranticipated above-market costs) by expensive nuclear power aud associated with the two high-cost byrexpensiven power.Specific , mane nuclear power plants and the state- gren pwer Spciicaly,masiv mandated purchases of power from cost overruns on two major nuclear cain Ip roha a "ometitive power plants and statemandated ..tain .haroge o "conmpers purchases of power from transition charge" on consumers' purchases Of power from elcrct bils independent power producers (IPPs) using renewable and other * The state government mandated a 10- technologies at prices significantly percent reduction in retail rates. Retail higher than the costs of traditional rates were frozen for four years or until technologies, stranded costs were recovered. Actual consumer bills went down little because capThere wasj surpustp o ghereforang the reduction in rates was largely offset (April 1998). by the competitive transition charge. ( Approxim998). 20 percent of m* Retail (residential, commercial and - Approximately 20 percent of industrial) customers were given the California's electricity supply was right to choose alternative electricity imported from neighboring states. suppliers. - The three privately owned utilities * A non-profit, independent system were regulated by the California operator (Cal ISO) was created to 3 operate the transmission facilities owned price volatility of the Cal PX spot by the private utilities (about 75 percent markets. of the state's high-voltage grid). The Cal Distribution companies and others who ISO also operated a bid-based real-time , ,, , , ~~~~~serve retail customers were not required energy market as well as several other to own or have under contract sufficient markets to acquire grid support services generation capacity to meet their peak (i.e., ancillary services). demands. * A separate Power Exchange (Cal PX) No provision was made for passthrough was created to operate a bid-based, of wholesale purchase power costs to centralized market for forward (day- retail rates until full recovery of stranded ahead and day-of) power sales. The two costs or March 2002 (whichever came largest private utilities were required to first) buy and sell all of their electricity through the Cal PX. * The complicated design involved multiple, sequential wholesale markets * Both the Cal ISO and Cal PX were oprtdb.tonwspraeette govemed~~~~~~~ bylrebad,.aho hc operated by two new separate entities governed by large boards, each Of which (the Cal PX and the Cal ISO). In other was made up of more than 30 U.S. regions, the ISO and PX are stakeholder and non-stakeholder c edin aie entity. members. ~~~~~~~combined in a single entity. members. * The retail electricity rates of individual privately owned utilities continued to be The Reform Process regulated by the CPUC. Even though the Cal PX and Cal ISO were under the * The reform operated by "political regulatory jurisdiction of FERC (the consensus." The final version of the national electricity regulator), the CPUC reform package reflected a compromise and the state government had substantial among competing stakeholders. It was de facto influence over their actions. The incorporated in a bill that was passed two regulatory entities, the CPUC and unanimously by the California FERC, sometimes issued conflicting legislature. orders. Criticisms of the final design by outside * The coverage of the reform was power sector reform experts were incomplete. Municipal utilities were generally ignored by state and national given the option of not participating in political and regulatory authorities. these new arrangements. In general, they chose not to participate. The Crisis How the California Reform Differs from . The highly contentious siting and Other Power Sector Reforms permitting process for new generating plants blocked the installation of any * Initially, the major private distribution major new generating plants for more companies were not allowed to buy than 10 years. California's installed outside of the spot markets. (No vesting generating capacity increased by just s ~~~~~~~521 MW between 1996 and 2000. or forward contracting was allowed.) Hence, they were totally exposed to the 4 Wholesale markets operated by the Cal two largest private companies to incur PX and Cal ISO worked reasonably well around $12 billion in unfunded liabilities for the first two years (1996-98) while since April 2000. They are on the verge the initial surplus of generating capacity of bankruptcy. disappeared. Less than 2 percent of * The Cal PX ceased to operate its two residential customers exercised their markets on January 31, 2001. option to pick new electricity suppliers because new suppliers could not offer substantial reductions in consumers' The Lessons electricity bills under the rate freeze and competitive transition charge during the Overall Design of the Power Market reform transition period. * A shift in market fundamentals occurred: * A poorly designed power market will not large increases in electricity demand in operate properly, and inadequate California and neighboring states, attempts or delays in correcting market reduced availability of hydropower in distortions will spill over into a serious California and the Pacific Northwest, financial crisis. and big increases in the prices of gas and * The California power reform crisis pollution permits to emit nitrogen oxides offers many valuable lessons on "what (NOx). not to do" for reformers of power * Wholesale spot prices skyrocketed sectors, particularly for the starting in the spring of 2000. California establishment and regulation of a utilities paid around $11 billion more for mandatory, wholesale power market electricity in the summer of 2000 than in based on spot pricing. the summer of 1999. Similar wholesale * The California experience indirectly price increases in neighboring states had provides important lessons for the less impact because, unlike California, privatization and regulation of only 5 to 10 percent of their overall distribution enterprises and new market supplies are purchased on the spot entities in developing countries, even market. though privatization was not an element * Mandated rolling blackouts throughout of California's reform. the state since December 2000 seriously * The California experience also provides disrupted the state economy (the sixth a lesson about crisis management: there largest in the world). Even more is no way out that is quick, painless or widespread blackouts are expected in the cheap. "Quick-fix" solutions to basic upcoming summer. design flaws usually fail and may * Some evidence indicates that the aggravate the problems. Any real growing shortage of generating capacity, solutions will impose heavy costs on combined with certain features of the stakeholders such as suppliers, complex wholesale market design, may consumers, shareholders, and legislators. have allowed some generators to exercise market power. * Limited or no pass-through of wholesale costs to retail customers has forced the 5 Requirements for Competition to Work in Introducing Competition to the the Wholesale Power Market Wholesale Power Market * Spot markets for wholesale power * Most developing countries should start require careful design of market rules with limited forms of competition that and price regulation to allow participants can evolve to full wholesale competition to manage their trading risks efficiently. through spot markets once the sector can * Competition requires adequate capacity manage full competition without to meet demand without experiencing uncontrollable market power. The supply constraints (generation, creation of bid-based spot markets transmission, fuel, etc.). The market should generally not be their top priority. must provide signals and incentives for * A mandated, deregulated, wholesale bid- investment in new generating capacity based spot market should be pursued when needed. These can be provided by only if certain conditions are likely to be various means, such as imposing a satisfied. Some of these prerequisites are capacity obligation on distribution also required for other, more limited companies purchasing power in the forms of competition. But the market, setting up a parallel capacity consequences of not satisfying these market to the energy spot market, or conditions are most dramatic and developing a forward energy trading harmful in a mandated and deregulated market whose prices signal expectations spot market. about future supply/demand balances. Price-based spot markets are generally * Competition requires that investors in too risky for small-to-medium-sized new supply capacity face no major power systems because of these systems barriers to entry to the wholesale power will lack sufficient bidders to maintain market. These barriers include effective competition. uncertainty and expense in facing delays Cost-based spot markets, such as those to the permitting process, regulatory developed in Latin America, offer a uncertainty about after-the-fact price reviews, and regulatory constraints on sim ld les i tive that .' . . . ~~~~~can yield competitive benefits for managing trading risks efficiently by medium-sized power systems, means such as hedging instruments. complemented by imposing a capacity * The design of a competitive power obligation on distribution companies. market is too complex and delicate to be Likewise, it is simpler and less risky to dominated by heavy political impose obligations on generators and compromises that are intended to shield distributors to provide ancillary services stakeholders from the consequences of (i.e., grid support services) as a the reform. Market design should be condition for being connected to the firnly guided by sound economic grid, rather than trying to synchronize principles. one or more separate markets for * New competitive trading arrangements ancillary services with an untested spot in a wholesale power market should be energy market. introduced carefully to provide scope for Vesting contracts should be allowed as a dealing with design flaws as well as form of insurance for distributors settling-in problems. 6 purchasing from a new spot market. A costs approach or even exceed fixed vesting contract that fixes the sale price retail rates. for trade between existing or new * Regulators should encourage and even generators and distributors for five or require suppliers to allow large users to more years should be established before adjust their demand for power in real the market goes into operation. They time, through smart metering and other also provide at least initial protection means, since competition works properly against market power. only when both suppliers and users * The spot market can evolve from a cost- interact in the market (i.e., prices must based to a price-based system as the be seen by both the demand and supply power market becomes more sides of the market). competitive. Interruptible supply tariffs work only * Alternative trading arrangements to spot when consumers do not expect to be markets, such as bilateral trading among called more than occasionally to reduce multiple buyers and multiple sellers, their demand on the power system. should be considered for small power Power outages are enormously costly for systems and as transitional arrangements consumers who have already adjusted to until the benefits of a spot market are using grid power. Hence blackouts are considered to outweigh the risks. symptomatic of enormous * Bilateral trading becomes unsustainable macroeconomic losses. This shows in as the only trading method when the turn the potential gains from reforming complexity of balancing system supply systems in such a way that such a with demand in real time becomes situation is avoided. unmanageable as the number of buyers * Small retail power users should have the and sellers increase. Commercial option of avoiding exposure to the high transactions cannot be divorced from price volatility that can occur in spot physical realities of power system markets for power. Power suppliers or operation. other entities should be given regulatory * A temporary single-buyer arrangement scope to absorb this volatility through can be considered-but with strong risk management techniques. reservations-in situations where * One or more commercially viable bilateral trading or spot markets need entities must have a legal obligation to substantial time for development of provide adequate supplies for consumers power purchasers and sellers. who prefer to deal with a default supplier rather than shop around in the market for a supplier. Introducing Competition to the Retail Power Market * In countries where the power supply industry is under state ownership and is * Retail tariffs should be aligned with the due to be privatized and opened up to costs of wholesale power. Regulators competition, stranded costs for past should avoid rate freezes that expose investments by utilities need not be distributors to the possibility of an consumers' bills. This is because these unsustainable squeeze on their cash flow costs w i lly be isobedaus the occurring when rising wholesale power costs will generally be absorbed by the 7 state through the proceeds received from governing board be composed of non- the sale of these assets. market participants (i.e., non- Full retail competition should be saved stakeholders). Goverance boards for last. In countries that have not composed of stakeholders should not be achieved substantial household too large or dominated by one or more electrification, it will generally be more classes of market participants. productive to focus on encouraging * Price caps should be used only as a last competition to serve those who do not resort, since they introduce distortions currently have access to electricity, than with unintended consequences and do on retail competition for those who not correct the causes of the problem already have access. that they address. * The system operator should monitor Regulating Power Markets markets carefully and continuously for signs of trouble-such as unusual price movements that may indicate abuse of * The economic regulatory system must be market power-and give the system open, independent, credible and not opetor-th e to e those prone to bankrupting reasonably wopvato e maurkt rules. efficient firins. ~~~~who violate market rules. efficient firms. * Regulatory "certainty" for power * An independent and expert market purchases by distributor is ofnosurveillance group should be created p urcase by distribo is fno vlue outside of the system operator. It should iasrin Cfornia,iit cans lea issue periodic public reports assessing bankruptcy of efficient firms. The the state of the market and mobilize regulatory system must be designed to the whe arket arises.iThe allow the cost of power purchases that quickly when a problem arises. The are beyond the control of a distributor members of the group must be perceived (e.g., mandated purchases in the spot as independent and objective. market, assigned purchases under a * Regulation of fuel and power markets vesting contract or purchases under a should be coordinated, especially the previously reviewed bulk supply tariff) linkage between electricity and natural to be automatically passed through in gas markets when most new generating retail tariffs. plant burns natural gas. * If there is a spot market, the regulator * In large countries it is important to should encourage hedging by allowing divide regulatory responsibilities distribution entities to recover hedging rationally between national and state costs if hedging opportunities are regulators to avoid unnecessary available (rather than forbid it until it is conflicts. It is not enough to simply say, too late, as in California). as in India, that electricity is a ''concurrent subject" with regulation * The governance of the system operator shared by national and state regulatory should be kept independent of the authorities. The nature of the "sharing" market participants. Independence can has to be defined precisely to avoid be achieved directly by prohibiting costly andedistracting conflicts. market participants from having an ownership interest in the system operator * The economic regulator for the power and requiring that the system operator's sector and the environmental regulator 8 need to work together. Each is in a position to undermine the work of the other. The ultimate success of both regulators requires a change in their mindsets. The power regulator has to accept that compliance with strict environmental standards is an integral element of power sector reform. The environment regulator must recognize the need to work constructively with developers of new generating plants to help achieve compliance with agreed- upon environmental standards. 9 10 Part I dispatch patterns, there is usually a parallel Lessons from California "free" market in which generators, distributors and others can enter into or hedging contracts to lock in future prices What the Power Minister and revenues. After several years of Needs to Know operational experience, the cost-based spot market can evolve into a bid-based spot market. The three principal advantages of a 1. Start with limited forms of cost-based market or pool are that it competition that can evolve to full 1. Ensures efficient dispatch (if wholesale competition. generators tell the truth about their production costs), Competition is intended to produce 2. Makes it difficult for generators to operational and investment efficiencies. exes market por and Alternative forms of competition exist that exercise market power, and are less complex than the mandated, 3. Is easier to implement because it centralized competition model adopted in builds on what the system or grid California. These alternatives can be operator was doing prior to the implemented separately or in combination. reform. None of these alternatives precludes moving North Americans and Europeans often to a deregulated, bid-based spot market in consider the Latin American cost-based the future. approach to spot markets an inferior form of competition. But the reality is that it has 1.1 Cost-Based Spot Markets with worked even if it does not fit a textbook Obligations for Capacity and Ancillary definition of perfect competition. Those Services countries that have adopted this approach in Latin America have generally experienced If participation in a competitive wholesale significant increases in private investment market is mandated, then a less risky combined with clear improvements in alternative is to begin with cost-based operating efficiency. bidding (as in four Latin American countries The designers of the reform should consider and in New England until recently, and as proposed for Ghana) rather than price-based imposig two types of oblgations: bidding (as in California, Colombia, El * A capacity obligation on distribution Salvador and the United Kingdom). enterprises and other load-serving entities to avoid complete reliance A cost-based sp)ot market based onona ewsrtem mrkto generators' actual or estimated variable on a new short-term market to production costs is easier to establish and induce investments in new provides more protection against market generation capacity. This power than a bid-based spot market. It requirement-currently in effect in *. ~~~Eastern United States, Texas and represents a relatively natural extension sev Latin Aerican from the traditional merit-order dispatch several Lathn Amerwcan systems used in many pre-reform, vertically sels-eeri to raicomers integrated power systems. While the cost- stlls have eo genera based bid market determines day-to-day capacity he owne ner capacity (either owned or under contract) to meet customer demands. been suggested that this type of market An alternative, used in Chile and would be easier to implement in developing Argentina, is to require that the pool countries because it would be voluntary and or system operator acquire capacity does not require the complicated protocols from generators on behalf of those or software of a mandatory spot market. who buy from the pool using administratively determined capacity Such voluntary markets involving one-on- payments that are in addition to the one bilateral transactions have existed for pool price. These two approaches many years in the United States and would work in either a cost-based or continental Europe. One big difference, a bid-based spot market. Only however, is that the buyers and sellers were California appears to have usually vertically integrated utilities with introduced a mandatory spot market sufficient generating capacity to meet all of without any accompanying capacity their energy needs. In general, these obligation or capacity payment traditional vertically integrated utilities mechanism. participated in these markets to "fine-tune" Initial obligations on generators and their supply needs (i.e., to lower their supply costs in certain hours rather than to meet distributors to provide ancillary their basic supply needs). services (i.e., grid support services) as a conditionfor being connected to The question then is whether this type of the grid. As practiced in Latin market is feasible in a different type of America, England and Wales, this is industry structure. Specifically, is it a viable generally easier than trying to option in an unbundled power sector synchronize one or more separate (separate enterprises for generation, markets for ancillary services with transmission, system operation, distribution an untested spot energy market. and retail supply) in which buyers would Once the basic energy market is have little or no supply of their own and functioning well, it may be less therefore would have to rely on the market costly to acquire ancillary services for most or all of their supply needs? through market mechanisms. Moreover, would it work in a developing country where there is simply not enough 1.2 Multiple Buyers, Multiple Sellers in generating capacity to meet the demands of Bilateral Markets all connected customers? Pools that operate a mandatory spot market, Because developing countries lack whether bid- or cost-based, are one form of experience with this type of market, there a multi-buyer, multi-seller market. There is, are no clear-cut answers. One major however, an alternative form of a multi- concern, however, is the issue of buyer, multi-seller market that does not "balancing." Even if a distribution company require creating a pool. This alternative is able to contract for all of its expected allows distributors, large industrials or both needs, its moment-to-moment demand will to buy directly from generators and other rarely be exactly equal to the amount for suppliers through one-on-one bilaterally which it has contracted. Therefore, there has negotiated transactions. The bilateral to be some balancing mechanism. (This transactions could be for short-, balancing problem does not occur when the intermediate- or long-term supplies. It has trading is among vertically integrated 12 enterprises because buyers will have their case in many countries), the industrial own generation supplies as well as the customers will no longer be a source of technical capacity to self-balance.) If the cross-subsidies if they can buy from other balancing mechanism is an organized suppliers. This, in turn, may lead to the need market with more than a few generators and for a big immediate increase in retail tariffs distributors, the cost and complexity of for non-industrial customers, rather than a setting up this residual balancing market series of phased-in increases over a longer may be almost the same as a full, mandatory period of time that could be managed by spot market. phasing the exodus of non-industrial customers from the market. One possible, less costly alternative to an organized balancing market would for It is not enough to simply provide distributors to acquire most of their needs distribution companies with the opportunity through one or more supply contracts with to participate in such a market. Distribution generators, and then hire a generation companies must also be given incentives to company or the system operator to be be efficient and intelligent buyers. In responsible for meeting the moment-to- particular, the regulatory system must moment fluctuations in its demand. Under include explicit incentives that allow this approach, the balancing would be distribution companies to earn higher profits performed by the hired agent rather than by if they find more economical supply a balancing market. sources. Regardless of whether the underlying industry structure is bundled or unbundled, 1.3 Single-Buyer Model it appears that voluntary bilateral markets are feasible only if there is (1) little The single-buyer model requires that all congestion on the grid (i.e., ample generation supplies be procured by an entity transmission capacity), (2) a small number specifically mandated to fulfill this function, of buyers and sellers and (3) an independent and that this entity in turn be the only seller operator who has complete knowledge and of bulk power to distributors and large users effective operating control of the entire of power. interconnected grid. This type of market This is the "toe in the water" approach to may not be workable once the number of introducing competition. In principle, it is buyers and sellers rises above a threshold the most limited forg of competition level because it becomes increasingly because it allows competition only for one- difficult to match a group of bilaterally time competitive procurements for relatively negotiated power-sales agreements of well-defined products-the supply of base, varying durations. These agreements well-defined p e pply or a produce hard-to-predict physical demands intermediate or peaking power for a on the grid, requiring a grid operator to specified period of time. In practice, balance the overall supply and demand of however, it iS often poorly implemented electricity on a moment-to-moment basis. because the single-buyer entity iS usually an existing state-owned power enterprise that is Allowing industrial customers to participate not a skilled buyer and that may be forced in such markets raises other concerns. If into signing high-priced and poorly designed industrial customers have been subsidizing power-purchase agreements (PPAs) through residential and other customers (which is the political or commercial pressure exerted by 13 its government owners. Furthermore, it temporarily) from bulk power costs. carries a substantial risk that the political However, what consumers do not initially and commercial interests that benefit from pay for in electricity rates, they (and those this approach will block further reform by who do not have access to electricity) will ensuring that it remains in force. eventually pay for in higher taxes or in the reduction of other government services (e.g., Although single buyers tend to be state- hospitals, roads and schools) that are owned enterprises, state-owned entities "crowded out" because of the subsidies or usually have limited experience in guarantees that now go to the electricity purchasing power, and* this lack of sector. In California it has been reported that experience may put the future budget a state government surplus of several billion revenues of their governments at dollars will soon be exhausted because of considerable risk. This could be the case in the need to cover power purchases by the California, where an existing state agency state buying agent. Standard & Poor's, a has become the de facto buyer for about 50 U.S. credit rating agency, put the state on a percent of the short- and long-term supply credit watch "with negative implications" needs of the three privately owned utilities. when the state began to purchase power. This happened because generators in California and neighboring states were no Countries with small power systems may be longer willing to sell to these three tempted to consider adopting a single-buyer companies, which account for about 75 model because unbundling generation and percent of California's retail sales, because distribution into a number of small entities, the three companies could no longer pay for combined with sophisticated market their power purchases. But simply replacing mechanisms, may not be a realistic option these companies with a state-controlled for such systems. In the more than 100 single buyer will clearly not be a solution if countries with installed capacity of less than the three utilities are not allowed to charge 1,000 MW, the potential number of tariffs to their retail customers that are high operators and distributors in the bulk supply enough to allow them to pay for the power market may simply be too small to support that they will now purchase from the state workable, ongoing competition unless the agency. Almost exactly the same situation country has strong interconnections to exists in the Indian state of Orissa, which, neighboring countries. Moreover, trying to like California, was the first state in its introduce sophisticated trading arrangements country to undertake significant reform. The could divert attention from other higher four privately-owned distribution companies priorities, such increasing supply, reducing in Orissa are unable to pay for the power losses and providing electric power to those purchased by the state-owned single buyer who are currently unserved. because their retail tariffs have been set too low. Other options besides the single-buyer model exist for purchasing wholesale power Because the single-buyer model in in small power systems. One approach developing countries often postpones an worth considering is a "joint action agency." essential element of reform (i.e., raising This is a common model used by groups of retail prices to cover costs), it frequently small power systems in several parts of the forces governments to offer backup payment United States, including California (e.g., the guarantees they usually can't afford because Northern California Power Agency). A joint ultimate consumers are "insulated" (at least action agency is essentially a buying 14 cooperative made up of small distribution mandated and deregulated spot market. The systems that pool their demands and hire conditions include the following: purchasing expertise. It is different from the Market power is not pervasive. There pure single-buyer model in two important are sufficient non-affiliated suppliers respects. First, the buying cooperative is an in each segment of the system load entity created and governed by the buyers curve, no serious bottlenecks exist in rather than a separate, government entity the transmission system, and control that is not accountable to its customers (the of fuel supply is not under the control current norm in many developing countries). a major generator. This condition is Second, it is voluntary. If a small unlikely to be fulfilled in a country distribution system believes that it can do with a small power system and few better job by purchasing on its own, it always has the option of "going off on its interconnecons ri er s own" as long as it satisfies its previous purchase commitments. * Distributors have the money to pay for their power purchases and distribution costs (i.e., retail tariffs are cost 2. Move to a full bid-based spot reflective and are not artificially market only once the necessary suppressed for political reasons). conditions are in place. Competitive power markets will fail unless distribution entities and other A full bid-based spot market provides buyers are commercially solvent. helpful price signals needed by consumers California started with commercially and potential investors when the necessary viable distribution entities but then conditions are in place. It is not, however, pushed them towards bankruptcy by the highest reform priority in a power sector forcing them to buy in a spot market that is starting from a base of pervasive in which prices skyrocketed and the under-pricing, significant cross-subsidies, regulatory system (which was the overstaffing, high technical and commercial result of a political compromise) losses and widespread political interference. prevented the two largest distributors The danger of trying to create such a spot from passing these high bulk-power market too soon in the reform process is that costs through to their retail customers. the effort required to make it work properly will divert attention and resources from * Buyers and sellers in a deregulated trying to solve more fundamental problems. market have the means and incentive It is a potentially time-consuming distraction to hedge price volatility in forward when more basic problems need to be spot markets, through intermediate addressed. and long-term contracts, etc., and are not forced to rely completely on A mandated, deregulated, and bid-based mandatory, short-term bulk power spot market should be pursued only if markets. Apart from vesting contracts certain conditions are likely to be satisfied. (see below), volatility in spot Some of these prerequisites are also required electricity prices can be hedged with a for other, more-limited forms of variety of other financial instruments competition. But the consequences of not such as futures contracts, options and satisfying these conditions will not be as derivatives. The market for such dramatic or as harmful as they would be in a instruments are not easy to create, can 15 be manipulated if there is not enough high prices). Consumers cannot volume and, more importantly, may "respond" to a price that they cannot divert attention from more critical see. California distribution companies "first order" tasks such as raising are now pursuing a crash effort to tariffs so that distribution entities can install real-time meters and tariffs for recover their total cost. their large customers before summer There are few bottlenecks on the 2001. transmission system that would block * Sufficient time, money and human transactions and create segmented resources are available to develop the markets. If there are bottlenecks, a new market system. A fully workable and efficient system exists developed, bid-based spot market for pricing congestion. For example, system involving multiple sellers and transactions in a day-ahead or hourly buyers requires significant expenditure energy market should not be arranged on real-time metering, bidding in isolation from whatever congestion protocols, settlement and market- exists on the grid. making software and communication The market and system operator are and data transmission equipment. genuinely independent in ownership Much of these costs are independent and decision-making from market of the size of the power market. participants (generators, distributors, California is a rich state, so it was able retail and wholesale suppliers and to finance a veritable army of final customers). The goversance consultants working under extremely system in Califomia resembled a mini- tight deadlines to install the necessary legislature and was susceptible to hardware, develop the protocols and deadlocks. write the corresponding software. In contrast, most developing countries New generation and transmission will not have these resources. And capacity can be built without even if they did, these limited excessive delays in permitting and resources would produce bigger and siting (i.e., supply can respond to more immediate benefits if used in market prices). In California, the extending service to unserved susceptibility of the siting and households, putting in retail meters permitting process to legal challenges where such meters don't exist and by nearby residents was a major making transmission and distribution barrier to entry for new generators. In investments to improve the basic developing countries, similar delays quality of current service. could be caused by weak environmental agencies that are * There is a "workout" of high-priced administering cumbersome power purchase agreements with IPPs administrative processes. or an explicit stranded-cost mechanism in place before the market Retail tariffs are designed so that at becomes operational. A wholesale least large- and medium-sized market will generally not work unless customers can "see" spot market this happens. prices on an hourly basis and can cut their consumption in response to high Policymakers sometimes fail to appreciate prices (i.e., demand can respond to that it is more difficult to create a bid-based 16 spot market in electricity than in other contracts. Instead, they were required to energy commodities because of the basic purchase almost all of their supply needs in physical realities of electricity production the newly created spot market. This is the and consumption: functional equivalent of requiring that * Electricity is very expensive to store. everyone buy their airplane tickets for a particular flight in a mandatory auction that * It is subject to rapid changes in takes place 30 minutes before the scheduled demand. departure. * There are pervasive externalities on the grid. For example, physical failure However, vesting contracts are not risk-free at one location can cause the collapse for distribution companies. If the contract ofathe entire grid supplys prices are high because of corruption or a non-competitive or poorly negotiated * Its demand and supply must be procurement process, future distribution balanced on a moment-to-moment companies and their customers may not be basis. able to pay the high prices. In such cases, a * The demand for electricity (on a real- vesting contract will simply perpetuate a bad time basis) can be very unresponsive outcome and lead to "stranded costs" when to price increases. and if competition is introduced. Starting power sector reform with a legacy of high- priced PPAs is like starting a race with a 20- 3. Allow vesting contracts as a form kilogram weight on each leg. of insurance for distributors purchasing from a new spot market. Vesting contracts can also be used with the creation of separate distribution entities Before the market goes into operation, the through privatization or divestiture, even if government or its privatization agency these actions are not accompanied by the should establish a vesting contract that fixes creation of a spot electricity market. Such the sale price for trade between existing or contracts reduce uncertainty for potential new generators and distributors for five or investors in both distribution and generation. more years. (The same technique, which is They also allow the regulator to focus in the sometimes described as "allocated PPAs," early post-privatization years on distribution can also be used when new distribution costs and performance (e.g., wires' costs, entities are created even in the absence of an technical and non-technical losses, billing accompanying spot market.) Vesting and collections) that are under the more contracts provide "insurance" in case the direct control of distribution entities. market design is flawed, and they provide revenue and cost certainty to generators and Vesting contracts are a transition distributors in the early years of reform. In mechanism. When the contracts expire or most countries that have created short-term when the distribution companies make markets, vesting and other hedging additional power purchases, the regulator instruments may cover as much as 80 to 90 will need to establish a system to ensure that percent of total power trade. This was not the distribution entity purchases the case in California, however. The largest economically to protect its captive retail distributors were required to sell generating customers. And the regulatory system must plants and were not allowed to repurchase provide incentives for distribution the output of these plants using vesting companies to enter into a portfolio of 17 purchase contracts to continue to hedge (whether it is for capital, operating costs or price risks. both) in return for an obligation to provide a specified level of grid or off-grid service (Argentina and Chile). In other countries, 4. Save full retail competition for privately or cooperatively owned mini-grids last. with an accompanying generating unit (i.e., Retail competitio did not succeedin a mini-privatization) in rural areas can be Retail competition did not succeed in ecuaemf rgltr iesn Califormia for several reasons relating to the equrem ar kept ltora m imm ndith specific design features (e.g., a 10-percent irequ-rements prers arep a e m dmum and the mandated rate reduction combined with a m gelectrical service with lower quality-of- rate freeze, the recovery of stranded costs service s a th the mainygrid through a competitive transition charge) of stributincanis If the minigrid the California retail competition program. operato n the tio o ing But even if California had been successful in onected to the grid f enh introducing retail competition, this does notma grd for enhanced imply that most developing countries should reliability, then the key regulatory issue is make retail competition an early action in the terms and conditions of the backup makei retailn crogrmpetit an early actionin service that is provided to it by the main grid distribution company or a separate Full retail competition (i.e., allowing every generation company. The general rule is that retail customer the right to pick their the regulator should not impose regulatory electricity supplier over an existing requirements above and beyond the distribution network) is expensive and willingness and ability of people to pay. complicated to implement. In England and Policymakers should also consider adopting Wales, it has been estimated that the initial a simpler version of retail competition-by hardware (metering, data transfer and tying the energy prices paid by residential telecommunications systems) and software customers to a measure of market prices has cost more than US$1 billion so far. paid by industrial customers who have access to competing suppliers. This It appears that other countries (Australia and "piggybacked" form of retail competition Norway) and other U.S. states Norwa) and other U.S. tates should be easier and less costly to (Pennsylvania) have had more success with sholdmbe easieran le costlto full retail competition than California. But it implement than full retail competition. A also important to remember that these variant ofthis approach has been adopted in countries, like California, are starting with Chile. full household electrification. In countries that have not achieved 5. Starting points matter. substantial household electrification, it will The starting conditions in power sectors generally be more productive to focus on vary enormously among reforming encouraging competition to serve those who countries. The "starting points" are do not currently have access to electricity, particularly important in four areas: rather than on retail competition for those who already have access. For example, in 1. Prices. Are retail power prices above poor, rural areas, the competition may be for or below costs? In California, the pre- the right to receive a government subsidy reform prices were high, but in many developing countries the prices are too 18 low to recover costs. It is virtually track record of honoring their impossible to undertake any serious commitments. In many developing power sector reform (including the countries, the regulator is a new creation of ongoing bulk power institution, its responsibilities vis-a-vis markets) unless a government is the government may not be clear, and politically committed to closing the previous governments may have a revenue-cost gap as its first priority. history of reneging on agreements. In effect, there is often an "institution 2. Capacity. Is generation capacity gap aswl,sa"spl ,a. adequate to meet the demand in the power market? In Califomnia, the The reform transition strategy should reflect reform started with a cushion of starting conditions and country excess capacity, while many characteristics. For example, in a country developing countries have a shortage starting with suppressed prices (i.e., prices of capacity. Is there potentially that are less than costs) and a shortage of enough within-countryygenerationo. enough withicumint gnerat supply, there is a greater political risk to capacity (assumig weak ntroducing deregulated bulk power interconnections to other countries) to mtion t ingathe cunr that make ~ ~ itwrhtikn. bu competition than in another country that make ilt worth thinkg about a starts with cost-reflective prices and a nAtional bulkg power3 mub-Sarke n surplus of supply. Similarly, it makes little Afuntrica, a gthateah 3e lesubsharn sense to try to create a deregulated bulk countries that each have less than power market in a small country with weak 1,000t MW ofhinstalled c it sotgoin interconnections to neighboring countries. markets andpother formsetiof, ooin The better strategy is to privatize what bulkretipowrer compti, whlagel already exists, provide subsidies for rural interesting to read about, are largely eetiiainadsrnte irrelevant to their immediate problems intrconnetion to ntres (unrliale ervie, ighlosss ad iterconnections to neighboring countries (nsunreliabl senervingc,ahightlosse a(Central America) before contemplating a insufficient generating capacity). eeuae,bl oe akt * ~~deregulated, bulk power market. 3. Coverage. Is there full electrification? California has full electrification Basically, it makes little sense to start a coverage. In many developing power sector reform without first deciding countries in Asia, Africa and Latin which problems need to be solved. If a America, large segments of the country moves too quickly to a complex population lack access to electricity. bulk power market that is inappropriate to For example, of the 34 countries of its current problems, it runs the risk of sub-Sahara Africa, more than 90 losing what may be a "once-in-a-generation" percent of the countries have less than chance to make fundamental reforms in its 20 percent household electrification. power sector. Power sector reform is a highly political process. Policymakers need 4. oInstitutions. Willt ivestors and to be alert to the fact that the necessary consumers trust reulaostory ond political support will quickly disappear governments instieattio toem honorly Iunless the reforms produce "early wins" that commitments and treat them fairly? In California, the state and national are readily discermible to the general public. regulators have existed for more than 60 years and have established a good 19 6. The economic regulatory system lines, transformers and substations to raise must be open, independent, credible service to acceptable standards. A multi-year and not prone to bankrupting tariff that turns out to be too generous to reasonably efficient firms. new private companies also runs the risk of a political backlash that could lead to after- Independent regulatory commissions are the-fact windfall profit taxes or even re- necessary but not sufficient for sustainable nationalization. Consequently, it may make power sector reform. It matters little to sense in some countries to combine a multi- investors that a regulatory commission is year tariff with a profit- and loss-sharing "independent" if the commission issues mechanism outside of a pre-specified dead tariff decisions that make it difficult or band. Such a sharing mechanism increases impossible for a reasonably efficient the political sustainability of the reform. distribution company to recover its total Any multi-year tariff should also be costs (purchase power plus wires costs). combined with performance standards so consumers can experience some 6.1 Distribution improvements in service to balance the pain of tariffs that are initially likely to be higher. However, the performance benchmarks must Multi-Year Tariffs be developed with considerable care. In particular, any benchmarks must (1) take Most developing countries that have account of starting points (e.g., technical and successfully privatized distribution have non-technical losses on the system), (2) given potential investors reasonable recognize that not all customers may want or certainty about the initial revenue stream for can afford the same levels of quality, (3) be 5 to 8 years through a multi-year tariff able to be objectively measured and (4) be formula that is fixed in the law or a bounded with respect to their financial concession agreement (akin to a contract impact on the enterprise. between the government and the investors). Because this tariff-setting system is usually an integral and legally binding element of Purchased Power the overall privatization package, the Like regulatory "independence," regulatory regulator may have very little to do with "certainty" is of no value if, as in Califoria, setting tariffs in the initial post-privatization it can lead to bankruptcy of efficient firms. period. This has been the norm in Bolivia, For distribution companies, it is especially Chile, El Salvador, Georgia, Guatemala, important that the regulatory system must be Moldova and Peru. designed to allow for the automatic pass- Multi-year tariffs are the regulatory through to retail tariffs of purchase power equivalent of going on "autopilot." Although costs that are beyond the control of the they reduce risk for investors (and have been distributor (e.g., mandated purchases in the adopted in almost every country that has spot market, assigned purchases under a successfully privatized distribution), they vesting contract or purchases under a may be difficult to implement if there is previously reviewed bulk supply tariff). considerable uncertainty about the initial Where the distributor has some discretion in levels of cost, consumption and losses and its purchases (e.g., post-privatization the level of investment needed in meters, purchases for incremental demand growth), 20 the regulatory system should create k Governor of California was quoted as saying incentives for the distribution company to that he could have solved the crisis in "20 minimize its purchase power costs. It minutes" if he had been willing to raise appears that such incentives did not exist in retail tariffs. Although political authorities in California. The privately owned utilities developing countries are often initially were generally reluctant to pursue nervous in allowing the creation of potentially cost-reducing, long-term "independent" regulatory commissions, they purchases in 1999 for fear that the purchases frequently discover the political convenience would be found "imprudent" in a later, after- of attributing necessary but unpopular tariff the-fact regulatory review. increases to the independence of their regulatory commissions. The principal Incentives to Hedge benefit is the ability to say that the tariff increases are beyond one's control. For If there is a spot market, the regulator should example, when the California Public accommodate hedging by allowing Utilities Commission announced average distribution entities to recover hedging costs retail tariff increases of 40 percent, on top of if hedging opportunities are available (rather an earlier 10 percent increase, the Governor than forbid it until it is too late, as in was quoted as saying: "I can't order or direct California). There needs to be a balance in an independent body. I've not given any the regulatory system. The regulator should advice to them on the subject of a rate not write a blank check by accepting all increase." (Washington Post, March 27, hedging costs, nor should the regulator 2001, p. A2) discourage distributors from hedging because they fear disallowance of profits 6.2 Market Regulation and Monitoring under after-the-fact "prudency" reviews. The better approach would be to establish before-the-fact price benchmarks for Governance of System Operators wholesale power purchases to encourage efficient buying. The indexed purchasing The governance of the system operator power benchmarks created by the electricity should be kept independent of the market regulators in Northern Ireland, Scotland and participants. Independence can be achieved the Netherlands are useful models. The directly by prohibiting market participants choice of benchmarks is critical. Several from having an ownership interest in the Latin American countries have adopted an system operator and requiring that the index based on six-month estimates of nodal system operator's governing board be prices as the purchased power benchmark. composed of non-market participants (i.e., However, they have found that the non-stakeholders). But it may not always be distribution companies will simply rely on possible or desirable to create a non- their legal right to buy all of their power stakeholder board in some developing needs at these prices and not attempt to countries. Therefore, the alternative is to engage in any hedging transactions. create a stakeholder board where no entity or class can dominate board decisions. The Blaming the Regulator failure of the California stakeholder board suggests four lessons: A politician can do few things more 1. The board cannot be too large or it unpopular than raising electricity tariffs. The will be ineffective as a decision 21 making body. (The California system higher level (as happened in California), operator board had 25 voting members thus defeating the purpose of the caps. Price before the Federal electricity regulator caps must be a temporary, last-resort dissolved it.) measure. If they are kept in place for too 2. The voting rules must ensure that one long, they will reduce the pressure to deal or two classes cannot control the with the underlying problems and will board's decisions. ultimately prevent the market from developing as originally planned (as 3. The regulator must be able to step in happened with the wholesale electricity and make a decision if the board is market in the Ukraine). deadlocked. 4. Consumer representatives or Monitoring by System Operators advocates should be viewed as market participants. Regulators should require the system operator to monitor markets carefully and Price Caps continuously for signs of trouble-such as unusual price movements that indicate abuse Once a market has been created, price caps of market power-and give the system should be used only as a last resort if serious operator the authority to penalize those who structural or market design flaws emerge. violate market rules. The system operator When prices go up, the natural instinct of has detailed knowledge of daily operations most political authorities is to impose price and therefore is in a unique position to serve caps. But price caps distort markets, and as the regulator's "eyes and ears." In they treat symptoms rather than causes. If California, several (but not all) of the the underlying problem is a shortage of recommendations made by the Cal ISO's generation capacity, a price cap will not help monitoring unit, as well as an external with the two needed solutions: increasing monitoring unit (see below), were adopted supply and restraining demand. As the by regulators. former FERC chairman observed: "We cannot 'price cap' California out of a supply Monitoring by Outsiders shortage." An independent and expert market- With any price cap, there is always a danger surveillance group should be created. It that it will be set too low. For example, it should issue periodic public reports appears that the price caps imposed in assessing the state of the market and California were at times below the mobilize quickly when a problem arises. (historically high) variable production costs The members of the group must be of some old generating units, and so perceived as independent and objective. A prevented these units from operating small- or medium-sized country might have profitably when the system needed their to hire experts from outside the country output. If price caps are put into place, they because most knowledgeable people within should be applied comprehensively across the country will be perceived, at least all markets in which a generator might sell. initially, as being biased because of past If they are imposed piecemeal, generators connections with the industry. The will simply sell in other markets where the surveillance group must have a broad price is not capped at all or capped at a mandate. It should be charged with 22 assessing not only the performance of the United States), it is important to divide market, but also the actions of the system regulatory responsibilities rationally operator and the regulator. (For example, in between the national and state regulators to California the market surveillance group has avoid unnecessary conflicts. It is not enough concluded that the "soft price cap" imposed to simply say, as in India, that electricity is a by the FERC would probably worsen the "concurrent subject" with regulation shared existing supply shortage.) Finally, the by national and state regulatory authorities. market surveillance group should work with The nature of the "sharing" has to be defined the system operator but must have the clear precisely to avoid costly and distracting right to issue reports without the prior conflicts. The areas of regulation actions approval of the system operator. that are likely to cause friction include * Transmission siting and certification Self-Regulation * Transmission tariffs Where organized spot or balancing markets * Bulk power tariffs are created, industry "self-regulation" of the * Grid codes accompanying grid and commercial codes should be encouraged. In California, these * Commercial and governance rules technical advisory groups were able to make for regional trading entities and grid some technical improvements in grid and operators. market operation. The regulator need not In California and the rest of the United formally approve every decision or arbitrate States, the division of regulatory authority every dispute, but the regulator must have has not always been clear or appropriate. the legal right to intervene on its own initiative or in the event of a formal Also, political authorities need to recognize complaint by a market participant. that the division of regulatory authority will probably have to change as the industry Regulation of Fuel and Power Markets structure changes. In particular, a division of regulatory authority that may have been Regulators must coordinate the regulation of workable under a vertically integrated fuel and power markets-especially the industry structure may break down as the linkage between electricity and natural gas industry moves to an unbundled, vertically markets when most new generating plant de-integrated structure. burns natural gas. For example, if a generator is owned by or affiliated with a company that provides natural gas 6.4 A Caveat: Regulating State transportation to competing generators, this Enterprises Is Different from Regulating corporate relationship could be used to put Private Companies its competitors at a competitive disadvantage. Although California provides many useful lessons in "how not to regulate," there is a hidden assumption behind these lessons. It is 6.3 Division of Authoritybetween that the enterprise that is being regulated will respond to the incentives created by the In large countries (e.g., Argentina, India, regulatory regime. This may not be true in Brazil, Canada, China, Russia and the many developing countries that have 23 recently created new, separate electricity often acted as if compliance were a low regulatory bodies that are regulating priority. Similarly, the attitude of most government-owned enterprises. These environmental regulators has been regulatory entities often borrow regulatory indifference to compliance by state-owned techniques that were developed to exploit power entities because of government the profit-maximizing objectives of private reluctance to face the costs of enforcing companies, and try to apply these techniques compliance. As power sectors become to public enterprises. increasingly privatized, however, governments and their environmental However, the inescapable reality is that most regulators are re-discovering the local and public enterprises, despite lengthy and global importance of compliance with expensive programs to "commercialize and environmental standards, and are willing to corporatize" them, still usually act like put more effort into enforcing these public enterprises. In particular, because standards. they do not pay much attention to profits and commercial performance, many of the The California experience shows that reform attempts to create regulatory incentives are of the way that the power sector is regulated lost on them. As a consequence, regulators economically should be coordinated with who find themselves regulating public environmental regulation of the sector. enterprises often spend considerable time Environmental regulation contributed writing impressive orders filled with substantially to the high bulk supply prices directives that, in the words of one new because it acted as a significant barrier to Indian electricity regulator, read like "pretty increasing the supply of electricity in poetry" but which are "rarely read and California. The problem was not so much almost always ignored." the standards themselves (which continue to be strict), but how they were implemented. While it is relatively easy to produce a list of Specifically, it took almost twice as long to regulatory lessons that can be learned from get state and local siting and permitting the California experience, many of the approvals for a new generating plant in lessons will be inapplicable to a developing California as it did in any other U.S. state. country unless the state-owned power The legal and political system allowed enterprise can be made to act like a inhabitants near the sites of the proposed commercial enterprise (which seems to be facilities and environmental groups to block rare) or until the state enterprise is or substantially delay the siting and privatized. permitting process for most new generating plants. As a consequence, supply stagnated, 7. Economic and environmental while demand steadily increased. regulators should talk to each other. The specifics of power sector environmental regulation-determining which pollutants In many developing countries, should be controlled and at what levels, and environmental standards that apply to the deciding whether market or non-market activities of state-owned power entities cnt mehanimshle ue are sector have been either non-existent or bond thensce this p e Hoeve loosely enforced. Where standards exist, iear tha deisios pabou tH ersb ta state-owned enterprises, operating with tight an cess of enironment regulatio budgets and lax maintenance standards, have annote udtenvineiso lation cannot be undertaken 4 isolation from 24 power sector reform decisions. Most electricity regulators would prefer to oppose unduly restrictive environmental standards that raise costs at precisely the moment when electricity prices may need to go up for other reasons. Similarly, most environmental regulators tend to take the narrow view that their mandate is only to ensure compliance with environmental standards. In particular, they do not feel any real responsibility for the overall success of power sector reform or, more immediately, whether a particular plant does or does not get built. The reality is that these regulators need to work together. Each one is in a position to undermine the work of the other. The ultimate success of both regulators requires a change in their mindsets. The power regulator has to accept that compliance with strict environmental standards is an integral element of power sector reform. The environment regulator must recognize the need to work constructively with developers of new generating plants to help achieve compliance with agreed upon environmental standards. 25 26 Part 11 progress in adopting policies that give consumers the right to choose their electricity supplier-the key and ultimate From Reform to Crisis in indicator of competition in the California market-ranks about average for the 24 U.S. states that have already implemented reforms, according to the Retail Energy 1. Background Deregulation Index (a scorecard developed by the Center for the Advancement of The reform of the California power market Energy Markets). is often characterized as a process of Part II of this paper proceeds in the deregulation. In fact, the reform involved following sections. First, it summarizes the limited deregulation by introducing price- indicators and consequences of the based competition in an elaborately California power crisis. It then outlines the structured wholesale power market, and it main parameters of the Califomia power changed the way that the power market is mmprmtr fteClfrl oe reanguled. Ith did nt involve divestarket of market, describes the formation of the new stglate-nd. ats. Hnce .the refor power market under the 1996 reform, and reviews the factors that led to the crisis. It more precisely characterized as part concludes by assessing whether the crisis deregulation and part re-regulation. The reform also involved some restructuring of market functions by * Obliging the incumbent utilities to sell 2. The Indicators of the California some of their power generating Power Crisis capacity to independent suppliers, The California power crisis of 2000-2001 * Unbundling their distribution arms has had two distinct phases: (1) during the from their generation and transmission summer months, when demand rose sharply arms, because the power load from air- * Placing responsibility for grid conditioners increased under a record- operation with an independent system breaking heat wave; and (2) in the winter operator, and months of 2000-2001, when power supply . Establishin separatemarketsfell sharply under seasonally low * Estabshmig separate markets for hydropower output and heavy withdrawals energyand acilla sevices, from service of old thermal power units for Most U.S. states have started or plan to start maintenance. programs to deregulate their power markets. The serious nature of California's power California was one of the first to start crisis is shown by numerous indicators for because of its desire to lower its retail the state's economy that has been the engine electricity prices. Competition in the power of high-technological growth in the United market was introduced through divestiture States. Resolution of the crisis is proving of generating capacity by incumbent difficult and is imposing heavy costs on the utilities, development of new power plants stakeholders-suppliers, consumers, by IPPs, and extension of competition shareholders, legislators, etc. The consensus gradually to retail supply. Califormia's 27 is that there is no way out of the crisis that The crisis has had the following immediate will be quick, painless or cheap. consequences: * Wholesale electricity prices during * A Stage 3 alert to power consumers, 2000 were more than three times the which had seldom been declared up to 1999 level. Huge spikes in wholesale the end of 2000, was declared for an power prices occurred during the unbroken series of 32 days during summer months of 2000. The market January and February 2001. A Stage 3 was declared dysfunctional by all who alert is the severest indication of an studied it then. impending power system brownout or . Retail electricity prices in the San blackout, when the system capacity Dieg area in 2000 were up to three reserve margin falls below 1.5 percent Diego area in 2000I were up to threeg times higher than in 1999; one of peak demand. household reported, for example, an * The state government has declared increase in monthly electricity bill several dozen statewide emergencies from $129 to $353 for the mid- to urge consumers to conserve December to mid-January period. electricity, but this has not been much * The first sustained series for decades help. of brownouts and blackouts occurred * The financial crisis caused by the during the months of November 2000 default on payments by the main to February 2001, when system utilities has threatened to spread to the demand was seasonally low, forcing banking community. temporary closures of businesses and * The Federal Secretary for Energy social institutions. invoked emergency powers on * Industrial and commercial users of December 13, 2000, to order power electricity have been paying massive generators to continue selling into the penalties rather than cutting their California power market. power usage under interruptible Natural gas suppliers threatened supply contracts. Electricity is so vital stoppage of deliveries of natural gas to for Silicon Valley that even a one-day the main power utilities this winter, power outage, such as the one that because they are concerned about the occurred in June 2000, reportedly cost utilities' ability to honor payment as much as $100 million in lost output. commitments. * The two main power utilities are * The state government has enacted facing bankruptcy, claiming that they measures that place it firmly in the have accumulated some $12 billion in center of the California power market uncompensated costs because of the (e.g., becoming the principal buyer of high prices that they have been paying energy for the two largest utilities), for wholesale electricity from power thus effectively flying against the generators. Each was losing around world-wide trendtowards deregulation $400,000 per hour on electricity and privatization of electricity trade. trading during January 2001. They currently lack the credit to purchase * The main organized wholesale energy wholesale power, and their debt rating market-the California Power has been slashed to junk-bond status. Exchange-has ceased to function 28 effectively and faces extinction, other for business, except for new because of the utilities' loss of credit industrial customers. on the exchange and a move to long- * California currently has about 53,000 term contracts for bulk power in MW of installed generating capacity response to the crisis. with the following distribution of * Serious power shortages in California ownership: are expected to continue for the next Public agencies comprising 23% two years, especially during the the LADWP and SMUD summer months. Renewable energy 22% * Serious impacts on California's producers and co- generators, supplying under economy are a concern, including long-term contracts based threats by businesses to move away, on Public Utilities and the repercussions on the rest of the Regulatory and Policy Act country. (PURPA) legislation Investor-owned utilities 15% * Other states are reconsidering plans to (lOUs) deregulate their electricity markets. IPPs, most of which is held 40% Nevada, for example, has postponed by five major power firms power deregulation plans, in part to (AES, Reliant, Duke, stopowenerators from sellingelectricity Southern and Dynergy) stop generators from selling electricity based outside the state to higher-margin markets in California. Regulators in Arkansas are In addition, California's imports of power recommending a two-year delay to provide about 5,000 MW towards meeting their plans. system load. * California's installed generating capacity by type of generator is as 3. Main Parameters of the California follows: Power Market Hydropower 24% The main parameters for the California Coal-fired steam generators 6% power market in 000Oil and/or gas-fired steam 37% power market i 2000 are summarzed generators herewith. Nuclear 8% * Retail supply of electricity in Combustion turbines and 8% Califoria is dominated by three combined cycle plant investor-owned utilities (I-Us)- Geothermal, wind, solar, 17% Pacific Gas & Electric (PG&E), municipal waste, etc. Southern California Edison (SCE) and * The sources of the 275,800 GWh of San Diego Gas and Electric wholesale supply of electricity in 1999 (SDG&E)-and two municipal by type of energy resource were as vertically integrated monopolies-Los follows: Angeles Department of Water and Hydropower 15% Power (LADWP) and the Sacramento Coal 13% Municipal Utility Oil and/or gas 31% Department(SMUD). Their service Nuclear 15% areas are discrete zones, so they have Geothermal, wind, solar, 8% traditionally not competed with each municipal waste, etc. Energy imports 18% 29 This distribution did not change much regulator-the California Public Utility throughout the 1990s. Commission (CPUC)-approved these costs * Peak load on California's as being "reasonable" and prudently interconnected power system in 2000 incurred. was about 51,400 MW including the The reform of the Califormia power market loads on the public agency systems. was implemented according to CPUC's The breakdown of this load by service restructuring order issued in December area was as follows: 1995, which led to the enactment of PG&E 41% Assembly Bill 1890 (AB 1890) by the SCE 38% California legislature in September 1996. SDG&E 6% The objective of the reform was to reduce LADWP 10% the costs of electricity because California's SMUD 5% electricity prices were much higher than the national average under traditional * Retail electricity consumption by regulation. At the same time, however, the sector in 2000 was as follows: concern was that competition would push Residential 30% wholesale prices so low as to render Commercial 36% unviable the investments in new power Industrial 21% capacity needed to meet growth in demand, Agrinustural 7__ while exposing consumers to high price Agricultural 7% volatility. AB 1890 was thus designed to Other categories 6% deal with these conflicting objectives. * Retail electricity prices-expressed in terms of average tariff yield of 4.1 New Market Structure U.S.cents/kWh-by consumer category for California during 2000 The reform established a new market are given below. They show that structure (shown in Figure 1) that promotes California's electricity tariffs are competition. Separate markets were created about one-third higher than the U.S. for energy, transmission and ancillary average. services that are procured every hour at California U.S. Aver market-priced rates through pool-based Residential 10.6 8.3 transactions. Bilateral transactions are also Commercial 9.9 7.3 allowed for some participants in the market. Industrial 6.2 4.5 The structure was designed to avoid Other 3.7 6.1 imposing administratively determined All Sectors 9.0 6.7 commitments, such as capacity obligations, on market participants. 4. Formation of the New Power This new market structure was established Market under the 1996 Reform by the following means: Before the reform, the IOUs were vertically * A Power Exchange (Cal PX) was integrated and were able to recover their created by January 1998. Cal PX is set costs of generating and supplying electricity up as a non-profit public benefit through the bundled rates that they charged corporation under California legal their customers, as long as the sector statutes. It acts as a market place in 30 Figure 1. Electric Supply in California Powei Plaiits flhIOWnled Utlities (IOUJ) UDC Bti11dledUltility. Source: Berry and Hoskote (2001) - Adapted from California ISO Ctistomers which generators and suppliers generation, transmission or compete to meet demand for electric distribution systems, and relies energy. It functions as an auctioneer entirely on services supplied from its and as such does not engage in energy markets to meet the demands on the trading on its own account. To ensure statewide power system. the viability of Cal PX, the AB 1890 The IOUs continue to own the statute requires the IOUs to sell transmission facilities and receive a energy produced from their own fee for the use of these facilities. The power stations (mainly hydro and Federal Energy Regulatory nuclear) and purchase energy on C ( behalf of customers who had not trnmission usERfees lathe CaesO changed to another supplier (nearly all system operation fees, as well as many customers) from the PX during the o h prtn,cmeca n four-year transition period to 2002. Their retail arms-called Utility technical protocols of Cal ISO and Cal Distribution Companies (UDCs)-and PX. electricity marketers purchase energy * Other than the three Californian IOUs, from Cal PX and resell electricity to participation in Cal PX is voluntary their customers. for all buyers and sellers of bulk An Independent System Operator (Cal power such as municipalities, IPPs and out-of-state producers. They can ISO) was established to operate trade electricity using a variety of statewide transmission system impartially for buyers and sellers of means (e.g., bilateral contracts). bulk electricity. Any supplier that * The non-PX participants must submit meets the regulated reliability schedules with the Cal ISO through standards has access to the system. Cal entities known as scheduling ISO operates as an independent, non- coordinators (SC). The SCs are the profit agency. It does not own any only point of contact between these 31 participants and Cal ISO, and they sequentially throughout the day, with bids number around forty. They coordinate for demand and supply. The final price is the scheduling activities continuously, and highest supply bid that is accepted to clear each SC submits a "balanced" the market. schedule to the Cal ISO in which the quantity of energy supplied equals the * The energy market is structured quantity demanded. Cal PX also primarily as a day-ahead auction by submits a day-ahead schedule to Cal Cal PX, with bidders allowed to ISO. submit different quantities and prices for each hour. This auction is * PG&E and SCE were required to sell accompanied by hour-ahead auctions at least 50 percent of their generation forpener to aow r-diegcein plants to IPPs or to place them in fo nryt lo frdvrecsi seplrants tow IPPscormtopaies th in odemand or supply from the day-ahead separate new manies iorer to bids. Such divergences may occur mitigate their market power b- from unexpected changes in weather reducimgtthe scopea for. anti- conditions or generating plant competitive "self-dealing." SDG&E availability. was required to divest all its generation assets (but its parent * The day-ahead and hour-ahead company was allowed to merge with markets are independent and are the local gas supplier). The capacity closed separately. Upon closing, the sold amounted to about 7,500 MW by winners are financially and PG&E, 10,600 MW by SCE, and operationally obligated to provide the 2,200 MW by SDG&E, totaling services that are selected by Cal ISO. 20,300 MW. Hence ownership of * Since scheduled transactions seldom about 40 percent of the total installed match the actual load on the power capacity in California was transferred system, Cal ISO calculates, in real to IPPs. time, the amount of energy needed to * A California Energy Market Oversight balance total system demand. It Board was established comprising conducts a real-time auction for members appointed by the state providing supplemental energy or for governor and legislature, in addition to backing off demand to achieve this large stakeholder governing boards for balance. Bidders submit prices up to the Cal ISO and Cal PX. 45 minutes prior to the start of each operating hour. They indicate the prices at which they are willing to 4.2 New Market Operating Arrangements change their generation or purchases in real time. Cal ISO uses these bids to The reform established separate markets for balance total generation and load in electric energy, ancillary services, and real time. Prices are established in this congested transmission capacity that are market every five minutes. operated in parallel by Cal ISO and Cal PX according to market operating procedures * Upon certification by Cal ISO, SCs approved by FERC. They were launched in can participate in any or all any of the April 1998 (except for the Block-Forward day-ahead, hour-ahead, and real-time market, which was launched in July 1999). markets. SCs are not required to They are operated as auctions carried out schedule all of their expected load and 32 generation in the day-ahead market. ahead prices for energy. To facilitate They may elect to bid for less than this allocation, Cal ISO accepts their expected load in the day-ahead "adjustment bids" for both the day- market, and then cover their remaining ahead and hour-ahead markets. These load in the hour-ahead energy market. bids reflect the prices at which SCs are Deviations from their day-ahead or willing to procure more energy or hour-ahead schedules are allowed by curtail loads from their preferred Cal ISO, and settled on the basis of schedules. If market participants do real-time energy imbalance market not submit sufficient adjustment bids, prices. Cal ISO levies a congestion Every day, Cal ISO collects energy management charge on the schedule schedules from the SCs and assesses that utilizes congested transmission the viability of each schedule. lines. Individual schedules accepted by Cal * Generators receive no capacity ISO are aggregated into a master payments or payments for start-up schedule that is checked to ensure that costs in the energy market. Hence they it can be accommodated by must recover their fixed costs through California's bulk power grid in a direct payments received for energy reliable and safe manner. If Cal ISO on Cal PX sales, as well as through the identifies power system problems such Cal ISO ancillary services market. as congestion in parts of the grid, it * Open and flexible scheduling provides the markets with an opportunities are characteristic of the opportunity to adjust schedules in market framework. For example, a order to alleviate the problems. generator may bid into multiple Cal * Cal PX operates a Block-Forward ISO markets and have multiple market that allows participants to enter delivery points. It can have a bilateral into electricity supply contracts for transaction with another market physical delivery up to six months into participant, sell a portion of its output the future. These contracts provide a to Cal PX, sell another portion to the hedge against spot-market price Cal ISO ancillary services markets, volatility. and export a part of its output out of * Cal ISO purchases ancillary services state. (for black starts, frequency control, * A generator faces a complex set of spinning, non-spinning and decisions concerning whether to sell replacement reserve generating capacity into an earlier or later capacity available at short notice) in auction, as well as between selling it an unbundled manner from generators for energy or ancillary services. Each through long-term contracts and decision to sell potentially forecloses competitive bidding. opportunities to sell into other * Cal ISO ensures reliable operation of markets. For example, a sequence of . ... ~~~~~~~~~decisions facing a generator could be the transmission grid by holding an g g auction for allocating congested whether to bid (1) into the day-ahead transmission capacity among the energy market at 7 a.m., for which the various system users after Cal PX has results are declared about one-and-a- established preliminary hourly day- half hours later; (2) into the ancillary 33 services market at 11 a.m., for which * Commitment of the contractually the results are declared by 1 p.m.; or agreed capacity with Cal ISO for a (3) into auctions throughout the day in specified term (generally one to two the hour-ahead and real-time energy years) of power plants sold by the markets. Market participants have the IOUs as "Reliability Must Run" opportunity to place bids up to five (RMR) to maintain system stability and hours before power flows in the Cal to overcome local congestion on the PX day-ahead market, two hours transmission system. "RMR" ahead in the Cal ISO hour-ahead designation for a generating unit market ancillary services market, and means that the owner must commit to 45 minutes ahead in the Cal ISO real- maintaining the unit and to responding time imbalance energy market. on a best-efforts basis to a directive Apart from these centralized markets, from Cal ISO to operate the unit. The there are separate bilateral transactions owners of RMR units are required to involving parties such as Californian bid all of their contracted capacity into generators who are not obligated to Cal PX. Hence they do not participate trade through the Cal PX, out-of-state fully in the Cal PX market. Ironically 'Califomian buyers in view of the events during 2000, Cal genertanto and ISO designated RMRs soon after the other than the three UDCs. new market started because of concerns about ultra-low clearing Figure 2. Overview of Market Operations prices in its imbalance market. In this situation, the relatively high-cost BXds Bids thermal power generators in southern California would not win business in Scheduleschedchedulesof the market and therefore have little PX A uction Other Schedu le PX A uction Other Sche dulepatcaeCl Coord Coord incentive to participate in it. Cal ISO PX Schedules PX Schedules was concerned about the availability and dispersion of sufficient reserve D.y.Abfld & Day Hour Ilo.r-Ahead Real capacity so that the transmission Ahead Ahead I nT'. Time Market Market Market system could absorb the loss of major E:c.iror.i.l50 tl transmission lines between northern Ancillary Services Market and southern California. r ________________________________________________ * Introduction of a competitive Source: Berry and Hoskote (2001) -Adapted from California PX transition charge (CTC) on customers' electricity bills for the recovery of the IOUs' stranded costs arising from the introduction of The market operating arrangements are competition. These costs refer to the depicted in Figure 2. relatively high operating costs and debt-service obligations (usually 4.3 New Market Regulatory Framework referred to as stranded costs) for some of the IOUs' generating plants built The reform changed the way the power before the 1990s. The CTC is market is regulated as follows: computed for each user's bill as the difference between the regulated rate 34 and the cost of supply. The regulated reduction bonds by a special purpose rate is frozen for all retail users until trust authorized by the state. the IOU that serves them has Regulation of the distribution recovered its stranded costs under the component of retail tarifs for the CTC. Califoria's utilities had UDCs will be based on performance- recovered more than $11 billion under based rate-making. the CTC by the summer of 2000, and SDG&E had fully recovered its costs * Initiation of retail competition. so that its rates were unfrozen. The Suppliers have competed actively for transition cost-recovery period lasts up the business of large commercial and to December 31, 2003, after which industrial users. Retail competition has retail sales are no longer frozen by not progressed beyond 2 percent of the statute. market in the market for residential users (except for a niche market for - The CTC is also used to help "green power") because of the freeze recover the high costs of power on retail rates and the inclusion of the procured by the IOUs under CTC in customers' electricity bills. PURPA-mandated contracts with certain renewable generation and * The California Public Utilities co-generation facilities (termed Commission (CPUC) continues to qualifying facilities, or QFs). regulate the UDCs' distribution These QFs provide up to 30 activities. percent of the electricity I a produced in California. This high In addition, fossil-fueled power generation is proportion reflects the state's subject to strict and a rather unique aggressive pursuit of electricity environmental regulation that pre-dates the from these types of facilities 1996 power market reforms. In particular, a during the 1980s. The high prices Regional Clean Air Incentives Market (averaging around 17 U.S. cents (RECLAIM) for Nox. Retail Emissions per kWh) paid to the QFs under Credits, or RTCs had been established with the terms of these contracts the total allowed emissions in a district to be would make these plants lowered over time so as to reduce urban uncompetitive under anticipated smog. Regulated firms are allocated a fixed market conditions (i.e., number of RTCs for NOx emissions for conditions that prevailed before each year, and they are required to redeem 2000) without the CTC. The these RTCs according to the amount of their prices in many of these contracts NOx emissions. Regulated firms can buy were tied to CPUC predictions of RTCs from other firms to overcome a world oil prices, but these shortage for meeting their requirements, and predictions proved to be sell RTCs in excess of their needs. These inaccurate. trades set up a market in RTCs, both for the current year and for future years Imposition of a JO-percent rate ("vintages") in order to prevent a "NOx reduction for all residential and small spike" of higher-than-anticipated emissions. users from January 1, 1998, to lastfor Firms are not allowed to combine RTCs of four years. This reduction was funded different vintages. by the issuance in December 1997 of $6 billion worth of 10-year rate- 35 5. Main Factors that Led to the whichever is sooner. Hence increases Crisis in wholesale power costs cannot be passed through to retail users, thus The California crisis centered around the exposing the electricity distributors to three UDCs and their suppliers through the huge potential losses under their Cal PX. Other power entities, such as the obligation to serve their customers. municipal utilities that chose not to This flaw does not become serious participate in the Cal PX, have not been so unless wholesale prices rise above the affected by crisis. This difference indicates retail rates, which they were not that design flaws in the Cal PX market are a expected to do at the time that the major source of factors that led to the crisis. reform was being introduced. This flaw may be only transitory, but it has Nevertheless, a number of factors exogenous contributed to the onset of the crisis to the market design worsened the problems during the transition period. created by the design flaws. In particular, Lack of economic incentives for the crisis arose out of an unpredicted adequate capacity to maintain supply combination of events. Undoubtedly the reliability standards. The UDCs were most important was the shortage of power not obliged to contract capacity, nor supply relative to demand. In the summer were generators recompensed crisis, demand increased to around 51,400 specifically for providing capacity. MW-30 percent above the winter level. In Long-term forward contracting of the winter the supply capacity was reduced energy by the UDCs was also not by more than 20 percent as thermal plants allowed. Finally, the lack of forward were taken out of service for deep energy markets for some years ahead maintenance, and an unusually dry end to suppressed the price signals that the year 2000 in the Pacific Northwest left would have helped the distributors and reservoir levels low and thus limited the investors in generating capacity to amount of hydropower that California could assess the need for new capacity. import. The other factors have exacerbated this problem. *Lack of risk-mitigation options for distributors. The UDCs were not allowed full access to forward 5.1 Market Design Flaws markets, and so were not able to develop a risk-minimizing power Structural and operational flaws in the portfolio. During 2000 they acquired California power market became evident only about 6 percent of their energy within a year after the ISO and PX went from forward markets, in contrast to operational in 1998: 34 percent from their own generating * A mismatch between the regulated plants and 60 percent from other retail market and the deregulated suppliers on the Cal PX market. They wholesale market. While wholesale were not even allowed to sell their electricity prices and natural gas prices power plants with long-term vesting are deregulated, retail electricity prices contract protection against price are fixed for the UDCs until they have volatility. Instead, they have had to recovered their stranded costs through rely on volatile spot markets. Hence, the CTC or by December 31, 2003, they were forced to "sell long and buy 36 short," which is disastrous for a trader limit prices to below the opportunity in any commodity. costs of other units providing Demand inelasticity. Lack of demand replacement reserve, hydro units elasticity by UDCs in the energy constrained by lack of water, and markets arises from their inability to thermal units constrained by emissions curtail their demand to avoid paying limits, as well as exporters to high prices, because of their obligation neighboring markets which were also to serve the demands of their captive experiencing high prices. customers. Just as a relatively small * Market arbitrage by generators. Since amount of tightening of the the markets for energy, transmission supply/demand balance in the absence congestion rights and ancillary of any demand elasticity produced the services are cleared sequentially, summer price spikes in the Cal PX rather than together, Cal ISO faces market, so a relatively small amount heavy demands on coordination to of loosening of the supply/demand prevent arbitrage by market balance in the presence of some participants that leads to inefficient demand elasticity would have dispatch of generating plants and significantly mitigated the pressures higher prices than predicted under that produced price spikes. models of these competitive markets. Price caps. Facing virtually no This sequencing gives incentives to supplies in the real-time balancing generators to collect high premiums energy market to meet system for real-time energy and ancillary imbalances, the Cal ISO was services by withholding supply (or by authorized by FERC to impose during putting in such high bids as to be sure 2000 progressively lower "soft" price that they won't be accepted) from the caps on bids in the real-time balancing day-ahead energy market, and then energy market, starting at $750/MWh bidding more supply into the other during the summer and dropping to markets. Such profit-maximizing $250/MWh by the end of the year. incentives for generators bidding into Payments made by the UDCs above these multiple markets may account the price cap would be subject to for some of the observed price spikes scrutiny and costjustification by Cal under supply shortages during 2000. ISO in retrospect. These levels would For example, a generator would set a amply cover the costs of power bid in the energy market for a segment generation under normal trading of capacity to cover at least the conditions, but $250/MWh was foregone expected eamings in the insufficient to cover even the variable ancillary services market for that operating costs of the older power segment, and this bid could be a very plants during the periods of very high high hourly rate to cover these gas prices and high costs of NOx foregone earnings if the generator emission permits. The situation expects the segment to be dispatched appeared to provoke generators into for only one or two hours in the raising their bids for supply during energy market. Likewise, prices in off-peak periods to recover their losses some markets for ancillary services under the price caps during peak could be driven up by considerations periods. These caps also appeared to of foregone earnings in markets for 37 other ancillary services. Some power by generators increases observers also allege that the repeated significantly during periods when rounds of bidding under the market supply falls short of demand. Some structure provide generators with experts contend that the generators' scope to "game" the system by exploitation of market power caused a adjusting their bidding strategies to significant portion of the huge price their advantage merely by observing spikes for a few hours during 2000 in each others bidding behavior without the California wholesale electricity collusion in the accepted legal sense. market. Others go further by alleging Market arbitrage by UDCs. Since the persistent and serious abuse of market Cal PX capped prices in the day-ahead power by generators. Likewise, some energy market at a much higher level observers allege that common ($2,500/MWh) than the Cal ISO's cap ownership of one of the main gas in the real-time balancing market, the suppliers and critical gas pipeline UDCs have kept down their demand capacity in southern California created purchases in the day-ahead market by the conditions for market power in this under-scheduling their during hours market. After auditing plant outages in when price spikes would otherwise be California, however, FERC staff stated likely to occur. They have done this to that they did not find evidence of keep the price in this market below the certain practices that indicate abuse of cap in the real-time balancing market, market power by the audited thus effectively capping the rate they companies. It is generally pay at the lower level of the latter. acknowledged that it is difficult to Purchases on the real-time balancing distinguish from available data the spot market have constituted a higher exercise of inappropriate market power proportion of total traded energy in Cal from the exploitation of legitimate PX (20-30 percent of the total energy scarcity rents when a market is in short procured) than in other U.S. states and supply. other countries that have forward Market governance. Poor governance contracts in their power markets, since structures contributed to the problem. a balancing market usually handles The large size and politicization of the less than 5 percent of total trade. This boards of Cal ISO and Cal PX, feature appears to have contributed through quotas of stakeholders each significantly to the large volatility in representing their own interests, prices in Cal PX. hampered attempts to focus on getting Market power. The potential for the market to work. The governance market power is likely to exist in a arrangements for Cal PX give to some deregulated price-bid market such as parties the voting power to block the Califoaia wholesale market, changes to market rules, which was especially in the presence of local donett of concet about putting market segments created by market power in the hands of the transmission constraints. This potential UDCs. This led to the prohibition of takes the form of artificial scarcity of trading on forward markets by the power created by power generators to UDCs. Likewise, it is alleged that drive up prices and earn huge profits. generators have too much power in The potential for abuse of market Cal ISO, which they have used to 38 block proposals to force them to ownership of the means of supply is schedule their entire output in the day- diversified among private interests that ahead market. In late 2000, FERC possess property rights by virtue of their ordered the replacement of Cal ISO's ownership and (2) other parties, such as stakeholder board by a non- consumer and environmental advocacy stakeholder board. groups, have the legal right to mount strong Retail competition. Less than 2 legal challenges in defense of their interests, percent of California's retail as in California. electricity users have migrated from the incumbent UDCs to alternative 5.2 Exogenous Factors Energy Service Providers (ESP). Most ESPs have exited the California market after their failure to attract Constraints on Expanding Supply customers. The failure to develop retail competition in California results * No new power generation capacity has from a policy of charging retail users a been commissioned in California since default price equal to the wholesale 1992 because (1) uncertainty about the power price, rather than the retail new power market deterred investors market price, and by allowing the until the new market structure and UDCs the right to provide default regulations were finalized in 1996, service. Default service refers to and (2) subsequently excessive delays electricity supply provided to those occurred in obtaining siting permits customers that are not receiving for new power stations in the face of service from a competing supplier. It local opposition when investors is a regulatory device used to smooth submitted applications. the transition to a competitive retail * Investors have been deterred from market or as a long-tern alternative to entering the California power market it. The amount by which the default by the expense and uncertainty of the service price exceeds the wholesale extenuated permitting process for new price dictates the level of customer power stations and transmission lines, savings and supplier earnings, which exacerbated by the ability of people are fundamental drivers of retail dwelling in the vicinity of the competition. Generally, the higher the proposed facilities to initiate default price relative to the wholesale numerous legal challenges. The price, the more intense the propensity of California's consumer competition and switching to new and environmental groups to use ballot suppliers. measures to oppose new power plants has added to the delays and The presence of these flaws raises the issue ain to inestors ite of how the process for reforming the plants. for intestotwo yes Caifm i make wa aagd plants. However, in the last two years califorial markcet was managed. Ao that the state has licensed nine new power consensual process was adopted, so that plants (totaling 10,600 MW), and five interested parties influenced the design in (totaling 2,900 MW) are under ways that possibly caused these flaws. This cotrcin. Thes e p nt wi process resulted from the difficulty in contrutignicT lynto e int changing market structures when (1) the 39 supply shortage, but only in about two * Demand for electricity in the summer years' time. of 2000 was pushed up by air A drop in imports of power from conditioning loads under the highest neighboring states occurred because of temperatures for May to July recorded low hydropower production caused by for 106 years. a drought and a growth in demand for * Retail demand was not sensitive to electricity in these markets. increases in the costs of wholesale Environmental safeguards to protect power since the tariff rates for most fish populations in the Pacific consumers in California were frozen Northwest region further limited the until the utilities collected all their water available for generating stranded costs under a regulated electricity. These imports formerly surcharge on customers' electricity provided an important source (20 bills. In addition, lack of demand percent) of California's power needs, elasticity by retail electricity buyers especially during the peak demand arises because they only discover the period in summer months. prices that they are paying after the * Power stations and transmission transaction, and then only in termns of ^ .. . - .......... ~~an average monthly price rather than facilities are old. Nearly 60 percent of hour-by-houprcs relatiel few California's power plants are at least hour-by-hour prices. Relatively few 30 years old, and now need more users have time-of-use (TOU) meters. maintenance and thus longer outage * Failure to meet demand reliably for periods than modern power plants. electricity-especially through The withdrawal of about 10,000 MW blackouts and brownouts-is of this plant for maintenance, as usual enormously costly for power users during the low-demand winter period, who have already adjusted to using helped create the end-2000 supply grid power. Californian users of shortages. electricity showed their willingness to pay huge penalties under interruptible An Unexpected Increase in Demand supply contracts rather than reduce power consumption when called upon to do so by their suppliers. * The growth of Internet-based power consumption based on Silicon Valley industries spearheaded a 25-percent A Steep Increase in the Cost of Wholesale increase in statewide demand during PowerDuring2000 the 1990s, but this statistic hides the real problem. From 1988 to 1998, * The market clearing price in the day- electricity demand grew at an average ahead Cal PX energy market oscillated rate of only 1.3 percent per year. In between $25 and $50/MWh during 1999 and 2000, however, electricity 1998, 1999 and the first half of 2000, demand on the Cal ISO system surged and then rocketed to over $150/MWh unexpectedly. In June 2000, energy in June, July and August 2000 during demand was 12.5 percent higher than an extreme heat wave. The steep in June 1999, and peak demand was increase in price occurred when 6.2 percent higher. supply started to fall below demand, even though prices did not move 40 discernibly beforehand as the margin UDCs were to electricity price diminished between supply and volatility. demand. Electricity markets do not The design of NOx emission have the price stabilizing mechanism regulations-restrictive levels of of buffer stocks because electricity annual emission permits cannot be stored economically. complemented by a market for The average price of natural gas emission credits-has caused owners across the country also shot up during of older generating plants in California 2000 due to growth in demand, to pay a high price for these credits. because gas is the fuel of choice for Given power supply shortages, these the huge amount of power-generating plants were under pressure to utilize capacity recently commissioned or their capacity above the level that under construction. The shortage also would allow them to meet the NOx reflects a slowdown in gas exploration emission standards. In the South Coast during the second half of the 1990s, Air Quality Management District of when oil and thus gas prices were low. California (SCAQMD), the allowed This price increase occurred when NOx level was reduced on July 1, much more gas was used in 2000 than 2000, which reduced the supply of in 1999 for generating power in NOx RTCs just when demand for California because of higher demand them increased. Consequently the cost for power and lower supply from other of a vintage 2000 RTC increased from power-generating sources. around $3/lb. NOx between 1997 and The price of natural gas in California mid-2000 to around $45/lb. NOx by reached extraordinarily high levels end-2000. This increase in price for during a spell of cold weather in NOx emission credits pushed up the December 2000 (gas is used for space variable operating costs of a typical heating as well as power generation). Southern California power plant by In December gas sold daily on spot around $30/MWh. markets at major terminals averaged around $11 per thousand cubic feet 5.3 Exodus of Funds by Utilities (TCF), compared to around $2.5/TCF in the preceding years. This increase The holding structure adopted by the three in gas price added about $75/MWh to IOUs has enabled these companies to keep the operating cost of a typical old substantial funds out of reach of the power plant in Southern California creditors of the UDCs as the latters' debt that was supplied with gas bought on mounted through 2000. If they had been the spot market. Daily prices reached available, these funds would have been at times more than $60/TCF at the sufficient to defer the current financial southern border of California during crisis, and thus to provide some time for the first week of December 2000, implementing corrective measures to partly due to bottlenecks in the prevent the development of the financial California gas pipeline system. crisis. From the mid-1980s, the CPUC However, a large proportion of gas authorized the creation of holding purchases by gas traders and suppliers companies, in which the utilities were was hedged, and hence they were less relegated to the status of subsidiaries. The exposed to gas price volatility than parent companies were permitted to pursue 41 other, unregulated businesses as long as opportunity for making adjustments those activities did not compromise the smoothly had been lost and the impact was utilities' ability to serve customers or the magnified by the flaws in the market design. capital needs of the utilities. Other states experienced spikes in wholesale Independent audits of SCE and PG&E electricity prices similar to those in released by the CPUC recently showed that California, but only for a few days at a time. the UDCs transferred billions of dollars to Only California experienced a persistent their parent companies during the first years series of such spikes throughout the summer of deregulation. The parent of SCE received of 2000. Retail prices in some other states $4.8 billion and the parent of PG&E have also risen by similar proportions to the received $4.6 billion between 1997 and trebling of rates in the San Diego area. 2000 from their Californian utilities. These Likewise, natural gas prices have risen on funds were derived from the sale of their average by similar amounts across the Californian generating plants, the surpluses United States, although they have risen earned through the sale of power in Cal PX much more at times in parts of southern from their remaining generating plant, and California due to pipeline congestion. But the recovery of stranded costs under the the other states have not experienced the CTC. The parents used this cash to finance brownouts and financial crisis that afflict most of their dividends and for the California. acquisition or construction of power generating capacity in other states and Two avoidable design flaws stand out: abroad. The parent companies of these 1. UDC's unhedged exposure to spot UDCs have instituted so-called ring-fencing prices, especially when tight supply provisions designed to prevent bankruptcy pric,spe whenetight spl conditions were foreseeable. The courts or anyone else from using the regulators eventually tried to help the parents' unregulated assets to cover the debt UDCs diversify this risk, as described of the UDCs. These steps have aroused immediately below, but their efforts considerable controversy in California. appeared to be a case of "too little, too late." 6. Could the crisis have been 2. Retail prices capped at levels that avoided? depended on low prices in the wholesale power market for In assessing the impact of the design of the sustainability. Despite intense political California power market on the current and consumer opposition, the CPUC crisis, the issue is whether design flaws have has recently approved an emergency made a serious situation unmanageable. The rate increase of 9 to 15 percent to fact that this arrangement worked without relieve some of this pressure. major trouble for the first two years indicates how easy it was to fall into a false The utilities could have tested the proposed sense of security while market fundamentals structure in the market before taking were heading for a crisis. In the case of irreversible steps, for example by offering California, these fundamentals were strongly their generating plants for sale with vesting rising demand, no new capacity, decline in contracts on terms that were affordable hydropower output, and surging natural gas under the capped retail prices. A lack of prices. Once the crisis hit the market, the takers from IPPs for such contracts would 42 have indicated that the proposed structure if it deemed the contract terms unacceptable. was unsustainable. This might occur if spot prices dropped below the level of prices under long-term The higher-than-expected prices that the contracts before the contracts expired. So IPPs paid for the IOU's generating plants both options open to the UDCs were risky, possibly indicated that they expected spot and generally the spot market was chosen by prices to be much higher than the levels at them. which the UDCs could survive within the capped retail rates. Other explanations for The market based NOx credit trading these high observed prices are the potential system, whose perceived advantage is value of generation capacity on the plant's reduction in the cost of achieving site, and the expectation of obtaining major compliance for the industry, in fact appeared gains in operating efficiency. to contribute to the increase in marginal After experiencing extreme (up to that time) supply costs of electricity when supply was price spikes during the summer of 1998 constrained in the Summer of 2000. For shortly after Cal PX opened, SCE sought example, "NOx spikes" can occur on days CPUC's permission to buy 2,000 MW- when electricity demand is greatest (due to about 10 percent of the peak summer air-conditioning load, for example), because demand of its customer base-outside the electricity spot prices can then rise Cal PX. This move was opposed by sufficiently to encourage plant operators to consumer groups, electricity sellers and pay high prices for NOx RTCs so as to run other stakeholders. CPUC rejected SCE's power plants at maximum output. This request on the grounds that such purchases indicates the possibility of interaction would weaken Cal PX and put the smaller between enviroenental and energy costs electricity sellers at a competitive when both are deter.ined by market disadvantage on the Cal PX. clearing prices. Cal PX tried to help the UDCs protect Inadequate transition arrangements also themselves from price fluctuations by adeate trit o thents The offering forward contracts for up to 18 appear to have contributed to the crisis. The months in April 1999. CPUC gave the Californian "big-bang" approach to UDCs permission to enter into such deregulation is open to the risks of Ucts, perniwith limits on how much unexpected market conditions, as well as the contracts, wt liiso hw muh unexpected ability of participants to "game"~ electricity they could buy that way, and so thexpet. arket of and highly Cal PX opened its Block-Forward market in the market. Market rules and highly July 1999. As prices kept rising, the UDCs sophisticated software, hardware and askd fr mre,andCPUC generally telecommunications systems were asked for more, andsometimeneralth developed in only 12 months, completely granted these requests, sometimes months independently of any market participants. A later. In July 2000, PG&E asked CPUC for structured transition strategy is needed that emergency authority to buy power outside is based on planning for steps that might be Cal PX, which CPUC approved in August in taken if crucial assumptions, such as the face of the full-blown crisis, continuation of surplus power supply The UDCs sometimes hesitated to use their capacity and low natural gas prices, proved freedom fully to enter into such contracts to be wrong. In particular, the IOUs because of concern about CPUC's ability to mistakenly anticipated earning huge margins cut their profits later in a "prudency review" during four competition-free years in which to recover their stranded costs. Cal ISO was 43 forced to make ad-hoc adjustments such as capacity market exists in parallel with the introducing price caps to deal with these energy market. They have not experienced unexpected events; these adjustments the shortages faced by Californian power provided quick fixes but led to further users for these reasons and also because the problems. Pennsylvania power system benefits from extensive interconnections with other California's inclination to rely on power regional power markets; also, coal is widely imports, rather than expand its own supply used for power generation, which hedges capacity, exposed it to developments beyond against increases in natural gas prices. its control. Neighboring states object to Independent power producers are being energy farms for California, whereby developing nearly 40,000 MW of new the latter avoids the environmental generation capacity in the state. Retail consequences of building new generation competition is promoted by a high default capacity while benefiting from the output. cost (considered to be too high by some They are also unhappy about the increases in commentators) and by mandatory prices in their power markets that they reallocation of retail customers from the attribute to events in the California market. incumbent suppliers, so that around 10 One indicator of whether California could percent of customers have switched supplier. have avoided its crisis by better market Overall, three conclusions may be drawn design is the existence of workable from the California power crisis: deregulation of a power market elsewhere 1. The flaws in the design of the under similar market conditions in the California market contributed United States such as in Pennsylvania, substantially to the financial crisis of Texas and Illinois. Another indicator of California's main utilities. California's specific vulnerability is the experience of its neighboring states under 2. Efforts to deal with the crisis in the similar supply constraints and growing presence ofthese flaws could not have demand. Wholesale power prices during the summer months of 2000 also rocketed in 3. A properly designed power market these states, partly due to the rise in could have coped with the factors California's wholesale power prices, but leading to the crisis. Because the their utilities did not hit the severe financial reforms already undertaken in the crisis that has floored the state's main California power market prevent a utilities. return to the pre-reform structure, the state's only option is to correct these In Pennsylvania, where the state restructured flaws and move forward to a better- the electricity market with far less political designed market. 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 44 Selected Bibliography Alaywan, Ziad. "Evolution of the California Independent System Operator Markets." Electricity Journal 13, No. 6 (July 2000). Alonso-Zaldivar, Ricardo and Nicholas Riccardi. "Audit Finds No Orchestration of Hikes." Los Angles Times, February 3, 2001. Available at www.latimes.com. Berry, Carolyn and Mangesh Hoskote, California Currents: A Review of A Restructuring Experiment in Progress and Lessons for Emerging Economies,Unpublished draft paper, Africa Infrastructure Department, Energy Division, TheWorld Bank, January 2001 Berry, Carolyn and William Meroney. California Power Crisis: Implications for Power Sector Reform in Emerging Economies. Presentation at Brown Bag Lunch Seminar. Energy Markets and Reform Thematic Group. The World Bank. January 11, 2001. Available from carolyn.berry@nera.com or william.meroney@ferc.fed.us. Borenstein, Severin. The Trouble With Electricity Markets (and Some Solutions). University of California Energy Institute. Working Paper PWP-079. Program on Workable Energy Regulation. January 2001. Available at www.ucei.org. Center for the Advancement of Energy Markets. Retail Energy Deregulation (RED) Index 2001, Washington, D.C., February 2001. Available at www.caem.org. CERA. Beyond California 's Power Crisis: Impact, Solutions, and Lessons. Special Report for North American Power and Western Research Services. March 2001. Available from www.cera.com. Hogan, William W. Coordination for Competition: Electricity Market Design Principles. AEI- Brookings Joint Center Conference: "The California Electricity Market Meltdown: The End of Deregulation?" March 1, 2001. Available from www.aei.brookings.org. Hildebrandt, Eric. Market Analysis Report. Presentation at the Market Surveillance Committee Meeting. California ISO. December 1, 2000. Available from www.caiso.org. Joskow, Paul and Edward Kahn. "A Quantitative Analysis of Pricing Behavior In California's Wholesale Electricity Market During Summer 2000." Unpublished paper prepared for Southern California Edison Company. November 22, 2000. Available at www.stoft.com/x/cal/20001121 -Joskow-Kahn.pdf Kahn, Michael and Loretta Lynch. California 's Electricity Options and Challenges: Report to Governor Gray Davis. California Electricity Oversight Board and the California Public Utilities Commission. August 2, 2000. Available at http://www.cpuc.ca.gov/published/report. Kucewicz, William P. "Power Politics: Drawing Lessons from California's Electricity Crisis." Unpublished paper. October 19, 2000. Available at www.geoinvestor.com/archives. Lynch, Michael W. "Don't Blame Deregulation for the Golden State's Electricity Snafus." Reason Magazine (January 4, 2001). Available at www.reason.com. Moore, Adrian T. and Lynne Kiesling. Powering Up California: Policy Alternatives for The California Energy Crisis. Reason Public Policy Institute. January 2001. Available at www.rppi.org. 45 Murphy, Kim. "Energy Secretary Rejects Calls for Power Price Caps." Los Angeles Times. February 3, 2001. Available at www.latimes.com. Ruff, Larry E. "Where California Went Wrong." Unpublished paper. Available at http://www.ksg.harvard.edu/hepg/. Samuelson, Robert J. "The American Energy Fantasy." Newsweek Magazine. January 29, 2001. P. 47. Simon, Richard. "State Gets Little Sympathy at Hearing." LA Times. February 3, 2001. Available at www.latimes.com. Smith, Rebecca. "Crossed Wires: Major Kinks Emerge in Gov. Davis's Plan to Power California." The Wall Street Journal. March 8, 2001. p. 1. Stoft, Steve. Soft Price Caps and Underscheduling Penalties: How Would the FERC Plan Affect California Electricity Markets? University of California Energy Institute. Working Paper PWP-079. Program on Workable Energy Regulation. November 2000. Wolak, Frank A. What Went Wrong with California 's Re-Structured Electricity Market (And How To Fix It). Presentation at the AEI-Brookings Joint Center for Regulatory Studies Conference: "The California Electricity Market Meltdown: The End of Deregulation?" March 1, 2001. Available from www.aei.brookings.org. U.S. Federal Energy Regulatory Commission. Order Directing Remedies For California Wholesale Electric Markets. Docket No. ELOO-95-000, et. al. December 15, 2000. Available at www.ferc.fed.us/electric/bulkpower.htm. 46 47 48 1EICAA AD The World Bank 1818 H Street, NW Washington, DC 20433 USA Tel 1.202.458.2321 Fax 1.202.522.3018 Internet: www.worldbank.org/esmap Email: esmap@worldbank.org The World Bank