Money, Power and Trade: What You Never Knew About the Western Energy Crisis

Deck: 
Fortnightly
Fortnightly Magazine - May 1 2001

Fortnightly



If you think you've heard it all about the power crisis, consider this fact:

In early 1999, the administrator of the Bonneville Power Administration (BPA) alerted the Northwest Power Planning Council (NPPC) about a possible energy shortage in the Pacific Northwest. And in March 2000, just two months on the eve of the crisis, the NPPC confirmed the analysis, noting that there was a 24 percent probability of not being able to serve all of the load sometime before 2003. But the NPPC also commented that the BPA had painted an "overly bleak picture," since imports and exports, in conjunction with hydropower reserves, would prove flexible enough to deal with the situation. If even noticed, the warnings were dismissed by both the region and California.

Now skip forward another year. As recently as Feb. 16, after the crisis was known to all, the Washington Energy Facility Site Evaluation Council turned down an application by the National Energy System to construct a 660-megawatt gas-fired turbine at the border with British Columbia.

"On balance," said the state council, "the significant environmental and social costs of the facility, if located at the site proposed, outweigh the resulting energy benefits it would provide only to most competitive bidders of the Western states power grid."

Simply put, it seems the siting council feared that the plant owner might export the output out of the state, to power-hungry customers in California who were willing to pay more.

These two anecdotes speak volumes on the power crisis that evolved in California and out West. The one shows the importance of free trade, the other its total collapse.

And in one final ironic twist, it turns out that the collapse of trade likely began at the very place that for centuries has served as one of the great trading hubs of North America, starting long before the Europeans came to the New World.

Celilo station on the Columbia River is the northern terminus of the DC (direct current) line that interconnects the electricity grids of Southern California and the Pacific Northwest. The Celilo station is aptly located at what was once Celilo Falls, where the Columbia River plunged over abrupt cliffs in a swirl of white water and spray, blocking navigation upstream. The falls are now covered by a subdued reservoir behind the Dalles Dam.

Celilo Falls was once one of the great trading hubs of the North American continent, where as many as 10,000 natives would gather to fish and trade. At the end of the Columbia Gorge, the falls marks the breach between the lush Pacific Coast with its blackish loam, wild rhododendrons, honeysuckle, roses, fern, moss, and evergreen firs, and the bleak interior with its sun-bleached soil, barren lava, sagebrush oak, and pine. At Celilo Falls, the principal medium of exchange was dried salmon, preserved to carry families through long cold winters. Here the boat people from the Lower Columbia, Puget Sound, and the Inside Passage would trade salmon for the Navajo's bright turquoise and silver. They would trade for Aztec corn and the buffalo hides of the Plains Indians. The Celilo Falls trading hub is only vaguely remembered, yet it may well symbolize the possible demise of the western power market.

The healthy benefits of trade that have been bestowed on the West's power consumers are now threatened as confusion, mismanagement, fear, greed, and parochial interests govern decision-making. Trade is essential to the western power system, because the transmission network was designed to take advantage of diversity in seasonal load and generation costs. For the first two decades after the North-South Intertie was built, the bulk of trade was arranged through long-term exchanges and the sale of "economy," or surplus hydroelectric power. Through this trade, utilities in the South would make power available to utilities in the North during the winter, and the flow would reverse in the summer. The interconnection also allows consumers and generators to take advantage of significant differences in variable cost. The West's transmission lines traverse vast empty spaces to connect population centers with far-flung generation. Unlike many of the power pools of the eastern and midwestern United States, the Western power system has distinct directional flows. %n1%n The benefits of exchange and trade are substantial (em the pieces fit together like a jigsaw puzzle, allowing individual utility seasonal peaks to be smoothed and cost reduced.

Now reconsider that warning from the BPA and how the NPCC reacted. %n2%n

Bonneville's analysis had demonstrated that the Pacific Northwest could be expected to run substantial deficits (em that is, to rely on imports (em from October 1999 to March 2000, unless water conditions were well above average. It seems as if everyone in the Western Systems Coordinating Council (WSCC) was planning on importing power during peaks, but no one planned for exports. The WSCC is not just interconnected, it is interdependent (em much more so than the relatively self-sufficient systems in other areas of the United States. When trade collapses, each sub-region of the WSCC must rely on its own resources; apparent abundance gives way to forced outages and capacity constraints.

Meanwhile, California represents less than half of the electricity load in the western states, but serves at the same time as the lynchpin of regional trade. California sits at the geographical hub connecting the winter-peaking system in the North with the summer-peaking system in the South. Moreover, the significance of California's role in western power markets has grown over the last decade, as the state has become more and more dependent on out-of-state imports to keep its system in balance. Thus, when the California power market tottered on the edge of collapse, the entire western market suffered.

With western markets joined at the hip, the siting council in Washington state saw it as prudent to block turbine construction. Put another way, the council concluded that local residents should not be expected to bear the environmental costs of a new merchant plant that would simply sell power for the highest price, presumably to residents of California.

Yet both that decision and its underlying reasoning are chilling. In an interconnected market, reduced supply means higher prices for everyone, not just for Californians. Attempts by each state to husband local supplies are myopic and counterproductive.

Only a year ago California was cited as a leader in energy market restructuring. Now politicians from Japan to Italy are rethinking their position on market liberalization, because obviously something has gone terribly wrong. They and others see the rolling blackouts and wrangling as evidence of a complex and imperfect market, yet nothing could be further from the truth.

The problem, in a manner of speaking, is very simple. On one hand, the richest state in the richest country in the world does not want to pay its bills. But the feeling is mutual, since it appears California's neighbors will be only too happy not to sell.

Virtually no one was prepared for the rapid demand growth experienced by western states in 2000. Yet the total power generation, as measured by the Edison Electric Institute (EEI), actually increased 7.6 percent in one year. %n3%n

The pace was more than double the year before and radically different from the 1 to 2 percent growth rates that had been both the historic norm and the planned growth for the next 10 years. Several reasons explain the rapid growth.

First, the summer of 2000 was hotter than the previous year, particularly in Nevada, Arizona, New Mexico, and Utah; the early winter in December was much colder than normal. Yet it takes more than abnormal weather to explain the sky-high growth rate. It turns out that electricity growth rates have been higher than usual even in off-peak periods, driven by an overheated high-tech economy. The rapid creation of jobs has accelerated population growth at the same time that the breakneck installation of servers, routing stations, and assorted communication facilities to support the Internet has kindled a rapid increase in power demand. This reflected a booming economy for the whole region: Rough estimates of regional GDP growth from 1999 to 2000 are around 8 percent for all the western states, with much of the growth centered in California. %n4%n

The general tightening of the West's power supplies had also been obscured by much better than average hydroelectric generation in recent years. The combined Canadian/U.S. WSCC region is unique in North America for its dependence on hydropower. The resource supplies about 40 percent of the region's generation capacity %n5%n and in most years more than one-quarter of the energy. Most of the capacity is located in British Columbia and the Pacific Northwest, where hydro supplies over three-quarters of the energy. Capacity figures by themselves can be misleading, because the actual energy generated from the system is limited by annual water flow. The energy extracted varies substantially, depending on the level of snow pack and reservoir fill that accumulates during winter months. In the banner year of 1997, total U.S. and Canadian WSCC hydro generation was 299 Terawatt hours (TWh) (em 99 TWh higher than generation in 1992. %n6%n This difference is nearly half of California's electricity energy demand and the equivalent output of more than a dozen nuclear power stations. As Figure 2 demonstrates, WSCC hydro generation is dominated by Columbia River water flow. Generation from other hydro systems, such as the Colorado and Sierra facilities, comes in relatively small amounts, often correlated to conditions along the Columbia.

Columbia water flows are now projected to be 54 percent of normal for the coming year. That is uncharted territory for the western power market (1977 was the year of the last major drought).

Because droughts are infrequent, their occurrence plays havoc with the economics of alternative power supplies. During good water years many of the region's oil and gas generators will lie idle or generate at prices not much higher than marginal cost. That means that capital costs will have to be recovered primarily during a period of drought. That, in turn, implies a spot market with many years of low-priced power, punctuated by episodes of extremely high prices. %n7%n

Following a string of high water years, supplies in 2000 were modest, with natural water flow lower than normal. In order to refill the reservoirs, generation had to be curtailed. During the critical months of June, July, August, and September, WSCC hydroelectric generation was an average of over 6,000 MW per hour less than the previous year (em roughly the output of seven to 10 nuclear reactors. %n8%n The drop in hydro supply occurred just when California and the Southwest were experiencing peak demand. Since California typically imports electric energy from the Pacific Northwest and the desert Southwest, the combination of lower hydro supplies and peak southwest cooling demand meant that existing oil and gas generators inside the state had to be driven at exceptionally high rates. In California, overall power output increased 13 percent, as the state's oil and gas generators were forced to displace imports that had been available in previous years. %n9%n For the months of June through September of 2000, natural gas generation in western U.S. states increased 62 percent from 1999. %n10%n The consequence of the unprecedented increase in thermal generation output was a shortage of natural gas delivery infrastructure, a draw down of gas inventories, and a rapid depletion of emissions credits, which drove marginal production costs and prices to historic highs.

October brought cooler weather and loads dropped. However, due to the intensity of use in summer months, California's oil and gas generators were in poor repair and many were down for maintenance, or were short on emissions credits. Likewise, nuclear and coal units scheduled maintenance or fuel replenishment during the normally low-demand shoulder season.

California's utilities and ratepayers might have slipped through the net, were it not for Murphy's Law and mercurial Mother Nature. Winter in the Pacific Northwest, which is usually awash in rain, sleet, and snow, saw unusually dry weather, coupled with a severe early-December cold spell. The mere threat of the Arctic Front and its impact on gas inventories was enough to panic the market. In December the region experienced record price spikes, as illustrated in Figure 3. And, since that period, bilateral prices in the Pacific Northwest (Mid-Columbia, Wash.) have been the highest in the region, reflecting an increasingly serious water shortage, compared to the Southwest (Palo Verde, Ariz.).

The consequences of the December price spikes were immediate and serious. California's two largest utilities were forced to buy large volumes of power at prices that they could not pass on to consumers. Southern California Edison (SCE) and Pacific Gas & Electric (PG&E) already were weakened from the summer spikes, and the new round of price increases rang alarm bells on every trading floor in the region. As a consequence suppliers began to question whether they would ever be paid. %n11%n On Dec. 13, the "dirty 13" bluntly refused to sell to the California Independent System Operator (ISO), which provoked a stage 2 emergency and threatened blackouts and chaos over the Christmas holidays. A day later on Dec. 14, the secretary of energy, under the auspices of the Federal Power Act, ordered power suppliers to continue sending energy to California. The secretary's action was followed on Dec. 15 by an order issued from the Federal Energy Regulatory Commission (FERC) that aimed to temper California markets by introducing so-called "soft price caps." %n12%n

The combination of cold weather and federal action took its toll. Hydroelectric generation in the Pacific Northwest increased in order to meet local load and supplement California supplies. The consequence was a rapid drafting of reservoirs, increasing the risk that the expected low runoff in May, June, and July would not be adequate for them to refill the pools. The power crisis was exacerbated in December by a record-setting cold spell in the South, Midwest, and Atlantic states, which severely strained natural gas inventories and delivery systems across North America, and particularly in California.

As the WSCC entered the shoulder season in February and March, the sense of crisis and urgency abated for a short period. Snow levels in the Sierra Nevada range improved to near normal, but precipitation in the mountains feeding the Columbia River drainage remained near record lows. Snow levels in the Cascade Mountains of Oregon and Washington are the second-lowest on record. Thus, there is increasing concern about the ability of the region's generation resources to meet summer peaks. For this reason, spot prices remain high and forward prices for the summer have risen significantly. The problem is exacerbated by the fact that California is unprepared to cope with high load levels, its interruptible load program in turmoil since late 2000. Virtually none of the long-term contracts entered into by the California Department of Water Resources will begin delivering power until late in 2001 or 2002. The cost for California to procure power on the spot market from June through September will likely total no less than $2 billion per month. If the summer is hot or if qualifying co-generation and small power facilities (QFs) are allowed to charge spot prices, this figure could approach $4 billion. Even with the 3 cents per kilowatt-hour rate hike of March 26, funds appear inadequate to meet such expenses. The state controller has forecast that the cost of power purchases by the state will exceed $26 billion in the next 18 months and will likely lead to a deficit of up to $7 billion in the state budget next year (from the current surplus). %n13%n

Since the 1970s oil and gas markets have transformed themselves into highly efficient and sophisticated commodity markets. Spot transactions lie at the heart of these markets. They determine the daily flow of energy from suppliers to consumers. While the system is highly efficient, it can be risky. That is because the demand for energy remains highly insensitive to price in the short-run (em nearly straight up and down on a chart. In the short term, consumers will pay almost anything for heat, light, and motion. Prices are kept reasonable, because supplies are usually abundant. In fact, most of the time the supply of energy seems to be infinite. It is infinite, in effect, until demand approaches production capacity. But once capacity constraints are reached the supply curve switches from a flat plane to a vertical pillar. At the transformation, energy prices explode. Abundant surplus turns to critical shortage quickly and without much warning. Let's call this kink in the energy supply curve the "devil's elbow." It acts as the principal cause of volatile energy prices.

When faced with critical demand and uncertain supply, it is wise to buy insurance. Such prudence seems to have escaped the attention of California's regulators. California's deregulation plan was based on the assumption that private entrepreneurs would offer "contracts for differences" (CFD) to moderate swings in spot prices. This market never developed in part because the three largest participants, the investor-owned utilities, were discouraged from such activity. Spot purchases were deemed per se reasonable, but the treatment of profits or losses from a CFD or New York Mercantile Exchange (NYMEX) futures contract (even after recovery of stranded costs) was purposely left vague. After it had recovered its stranded costs, SDG&E petitioned the PUC for performance-based ratemaking (PBR) to provide an incentive for the utility to do better than prices posted at the California Power Exchange (PX). The PUC turned down the request and San Diego stuck to the spot market. %n14%n

Recognizing the chasm in market structure, the PX introduced block forward contracts in 1999, which allowed participants to purchase peak power months or quarters in advance. These contracts were no different than a bilateral forward contract, except that the PX acted as intermediary and pricing was transparent. Since the power was procured through the state-sanctioned exchange, the purchases were considered per se reasonable. Nonetheless, the PUC put strict limits on the volume of forward power the utilities could buy, limiting participation to the utilities' "net short position," defined according to average quarterly volume. This artificial limit left the utilities vulnerable to daily and hourly peaks (em the very period of times in which protection from price spikes was critical. As late as July 2000, the PUC continued to enforce strict limits and turned down SCE's request for open-ended participation in PX forward markets. %n15%n

The reluctance of California officials to encourage forward markets stemmed from their earlier experience with gas markets.

Back during the 1980s and early 1990s, long-term gas prices usually stood much higher than spot prices, because the industry had vastly overbuilt production and transportation facilities. Also, as the volume of spot gas trading grew, the market matured, liquidity blossomed, and spot transactions proved reliable. Spot gas prices would spike, but only for short periods in response to pipeline outages or unusual weather. No one questions that California's reliance on spot gas markets for over a decade was beneficial to its industry and utilities. However, that success offered no similar guarantee for the power industry, where markets remained immature and where spare capacity depended on an unpredictable hydro supply. In addition, the gas market transition from bundled services for retail customers to unbundled supply occurred over a 10-year period, starting with the largest and most sophisticated customers.

In thermal electricity systems, the power prices tend to go to extremes (em prices spike to extraordinary levels when load reaches system capacity, or else fall down more or less to the cost of fuel when demand abates. Hydro systems can work in a very different way, however.

With hydropower, usually there is more than enough capacity to meet peak demand. However, the system can be "energy" constrained. The energy constraint arises because natural stream flow plus the water stored in reservoirs may not prove adequate to meet the load over an annual cycle. In a market setting, that means that the alternative cost of generating power today depends on the value foregone of generating some time in the future. There are, of course, all kinds of other constraints (em fisheries management, shipping, irrigation, recreation, etc. Nonetheless, hydro systems ought not to produce pricing extremes except in the most extreme high demand/ low supply situations. In theory, if water (rather than capacity) is short, both off-peak and on-peak prices will rise, until the current price is more or less equal to the net present value of expected prices in the future. In this case, however, expectations for load and replacement energy are critical, because precipitation for future water periods is unknown.

Since market disturbances began in May 2000, the western power market has been acting like an energy-constrained hydro system. Prices often peaked in response to capacity constraints, but did not return to fuel costs plus a small margin. Instead off-peak prices have remained high, acting as if the system had an energy constraint. In fact, it often has. The fundamental increase in demand described earlier and drop in hydro generation together have meant that oil and gas peaking units have had to increase utilization far beyond expectations. These units, however, depend on the availability of fuel, storage capacity, and the pipeline delivery infrastructure. These constraints frequently limit the total number of hours, or proportion of time, a unit may generate. Most generators contract in advance for a given volume of gas, and arrange storage based on expected utilization. Sometimes they may simply not be able to exceed their planned level of utilization. At other times, however, over-generation means that they will have to rely on spot market purchases of fuel, pipeline deliveries, storage, and emissions credits. These constraints may combine to create an economic limit on the energy that can be produced. The limit can be exceeded only at high prices. Thus, off-peak prices do not decline (em the cost of generating stays high because it reflects the opportunity foregone to sell into peak market conditions.

Environmental regulations often impose outright restrictions on generation by limiting annual emissions. Other local ordinances may constrain activity to particular times of the day or season. For example, the city of Pasadena limited the use of combustion turbines under its jurisdiction to 300 hours per year in 2000. (The city recently has increased the limit to 1,300 hours per year. It has 226 MW of oil and gas fired capacity, but these generating units are old, heavily polluting, and not suitable for prolonged use.) %n16%n

The problem of obtaining emission credits also contributes to the rising marginal cost. During the summer, emission credits in the South Basin environmental district increased more than forty-fold, from a monthly average of $.85 in January to $36.93 per pound of NOx emissions in September. %n17%n

The rise in demand and contraction in supply that has gripped the western power market presented an unexpected and serious problem. Yet it need not have turned into a tragedy. It was the failure by politicians and regulators to heed or understand market signals that compounded the problem and created a crisis. Because a variety of federal and state decision-makers did not understand the nature of trade in the WSCC, their decisions have greatly inhibited the normal flow of electricity.

False Signals. When prices first rose in May 2000, the California ISO maintained a real-time price cap of $750 per megawatt-hour, high enough to have little or no practical impact on the market. However, as higher wholesale prices worked their way through SDG&E's retail rates, politicians applied pressure on the ISO board to lower the cap to $250 per megawatt-hour. The board compromised. On July 1 it lowered the cap to $500, followed on Aug. 7 by a drop to $250. The lower caps did have serious impacts, although not those intended by its architects.

Because the ISO maintained price caps, the PX did not have to. Those purchasing power in the exchange would never bid more than the ISO cap, instead they would shift demand to the real-time market. In a normal market that would never have posed a problem (em just one or two days of dislocation. But problems in the western power market compounded as shortages continued through most of June and into July and August. More and more of the load was served through the real-time market, which remained highly prone to panic buying. On July 28 the phenomenon peaked, when as much as 28 percent of load had to be met by real-time purchases. %n18%n

The ISO compounded the migration to the real-time market through its own actions. Initially, the ISO procured reserve capacity at a price up to $750 per megawatt-hour, and could pay up to $750 per megawatt-hour if the energy was called upon. As load migrated to ISO markets, however, unscheduled demand became highly volatile. That only increased the need for reserves, taking energy out of day-ahead markets where the extra supply would have dampened prices. Moreover, the ISO possessed authority to make "out-of-market" purchases from out-of-state suppliers at whatever prices the market would bear. It did not take in-state suppliers long to discover that they could sell higher-priced power to marketers in Oregon and Arizona, which could then be resold to the ISO. (Some have called this practice "megawatt laundering.")

Normally power flows south from the Pacific Northwest to meet summer air conditioning loads. In the summer of 2000, day-ahead schedules indicated huge congestion differentials for moving power north instead of south. This effect was "paper" congestion, however, because the day-ahead contracts were resold to the ISO in real time and the power never actually left the state.

Misallocation. The ISO attempted to clamp down on such market aberrations by lowering the cap paid for reserves. Reserve margins declined as a consequence, signaling successive stage 1 and stage 2 emergencies. These dwindling margins led, in turn, to other unintended impacts.

For example, hydropower capacity remains an excellent reserve facility, because it can generate almost instantly without the ramping-up time of thermal units. Yet hydro generators could not offer reserve capacity to the ISO, because in the event of an emergency, they would be forced to generate even if reservoir levels dictated that the water be held back. As a consequence, bids into the ISO reserve market tended to come from thermal capacity. When that thermal capacity went unused, it took available energy out of the market.

In an effort to make up thermal shortfall and maintain reliability, the Northern California hydroelectric system became significantly overused. During November and December 2000, the ISO complained that some of California's reservoirs were drawn down to the intake valves, providing a vivid picture of sputtering sand and mud. These facilities, however, did more than just provide energy. Their location helps balance load and generation centers in the grid. The lack of balance increased congestion on "Path 15," the major transmission line running between Northern and Southern California, once again reducing available energy.

Regional Disruptions. The chaos in California disrupted normal trade flows throughout the West, causing at least one major transmission facility to become underused, thus undercutting regional trade.

When the new California power market opened in April 1998, its two principal load zones were Northern California and Southern California. Path 15, which connected the two zones, was seldom congested. And what little congestion occurred was usually restricted to a few off-peak periods, in a south-to-north direction. (The northerly flow allowed hydro reservoirs in the north to be refilled during off-peak periods by thermal units in the south). However, Path 15 was not the only way to move power from Southern to Northern California. Oregon is interconnected to California through the high-volume DC line that runs from the Celilo Station to Los Angeles, and by AC lines (alternating current) that run from the Oregon border to Northern California. These lines can be used to bypass Line Path 15 and wheel power to Northern California. As the regional power crisis has unfolded, this bypass has seen less and less use, not for technical reasons, but to avoid the risk of default or underpayment. Power can only move on this path by changing title through multiple owners, which of course compounds financial risk. As the credit risk of selling to California's utilities increased, trade in the WSCC dried up, exacerbating the crisis and pushing Northern California into rolling blackouts.

Missed Opportunities. The response of regulators and politicians to the unwinding crisis has been astonishing. The gas and power price increases in May could have been taken as a signal of looming infrastructure problems. Instead, the nearly universal reaction was to accuse suppliers of gaming, withholding supplies, conspiring, and jacking up prices. Obviously new generating units could not be built in a few months, but California could have implemented conservation measures, industry buy-back programs, lifeline rates, or any number of policies aimed at bringing the market back into better balance.

Even if suppliers were responsible for the price spikes, they would have far less leverage following the implementation of conservation programs. Instead, industrial interruptible programs were over-used in 2000 and early January 2001, and suspended altogether as the tariffed limit on disruption time was exceeded and customers reacted to the penalty charges. The PUC had changes finally underway as of early April, but given the time needed for customers to set up response efforts, the turmoil will damage the effectiveness of the interruptible programs for 2001.

Meanwhile, the price spikes experienced in November and December were unexpected. They provoked a political response that led to a financial crisis.

In late fall and early winter it became evident both to SCE and PG&E that they would be unable to continue purchasing high-priced power if the costs could not be passed along to the retail market. Summer peak purchases already had exhausted most liquid funds; borrowing capacity was rapidly reaching its limit at the two companies. In November and December, SCE and PG&E applied for rate increases in excess of 30 percent, the minimum they believed would be necessary to reassure lenders and allow them to continue purchasing power. The PUC turned down the requests and instead granted a temporary rate increase of approximately 7 to 15 percent, depending on customer class. The rate increase was so modest that it unraveled the tenuous finances of SCE and PG&E.

Contract Price Creep. As regulators moved slowly in allowing SCE and PG&E to recover wholesale power costs, other events conspired to ratchet up prices for purchased power, even when secured by contract.

To understand this situation, recall that California's utilities have three sources of supply: (1) their own resources not sold off in the restructuring, (2) long-term purchase contracts previously approved by the PUC, mostly with QFs, and (3) outright purchases from independent power producers (IPPs). Recall also that the deregulation scheme required the utilities to purchase their loads from the California Power Exchange during a four-year transition period.

In order to ensure a robust and liquid exchange, the utilities bid their own resources and contract purchases into the Exchange as supply. The PX price, used for valuing utility supplies, simply washed out. For utilities, important cost components were: (1) their own cost of generation, (2) the prices they paid for contract purchases, and (3) incremental load served by net purchases from the PX or the ISO. Typically, the generation cost did not increase for utility-owned units, but contract purchase prices did, because most were tied to spot natural gas prices. Thus, the actual costs of about two-thirds of SCE's and PGE's supplies were significantly above the prices that could be passed on to the consumers, even after the modest PUC rate increase.

In December, when the PUC limited retail rate increases, it did nothing about wholesale prices in QF contracts, even when it became clear that the formula would result in valuations three to four times above retail prices. %n19%n The nuances of the regulatory failing may have been lost to the general public, but it did not take long for Wall Street's analysts and IPP suppliers to conclude that the utilities were headed for insolvency, thus propelling the state headlong into a financial crisis.

FERC's Futile Response. The FERC did no better in managing unfolding events, focusing on market structure rather than on market fundamentals. As a consequence, both the cause and cure of California's market collapse was misread.

Beginning early in the summer of 2000, it became obvious that forward contracting should have been made an integral part of the original market design. In its Dec. 15 order, however, the FERC jumped to a simplistic conclusion. It saw only a single market structure flaw (em the utilities' mandated dependence on purchases from the PX. The commissioners concluded that by terminating the mandatory buy-sell requirement, they could somehow terminate high spot prices. Nothing, of course, could be further from the truth. Prices at the PX simply mirrored spot prices in the far larger bilateral markets, which in turn reflected market fundamentals.

Along with ending mandatory purchases and sales in the PX, the FERC ordered the exchange to implement soft price caps, with no corresponding requirement for alternative suppliers in the bilateral market. The FERC order, in combination with the inability of SCE and PG&E to pay for power purchased from the PX, caused the exchange to terminate its operations at the end of January 2001, and declare bankruptcy on March 9.

Crises create confusion and confusion leads to poor decisions. Both federal and state regulators are deeply confused about the differences between contracts and markets. Pushing utilities (or state agencies) into long-term purchase contracts does not eliminate the need for reliable and transparent spot markets; rather, it makes them all the more necessary.

Likewise, the FERC effectively terminated the California Power Exchange, the region's largest hourly spot market, presumably because it disliked the messenger as much as the message. Yet these quick fixes are likely to be regretted, because they bind the state to inflexible, high-cost supplies (em the very thing it sought to eliminate when it restructured in 1998. Mature, healthy markets depend on a robust mix of spot and forward trading; in contrast, many of the actions taken by federal and state regulators actually reduce liquidity and reliability and will contribute to higher costs and prices.

Electricity demand, more than any other commodity, is subject to the vagaries of weather. If utilities plan to meet extreme demand peaks, either through their own resources or through contract purchases, they will frequently have substantial surpluses. Such surpluses can often be economically disposed through spot sales. In parallel fashion, if a utility is short it may plan to supplement supply through short-term purchases. Which strategy will prove the most cost effective depends on the skill of the utilities' planners and traders. In all cases, however, an interconnected region is better off to encourage diversity of strategies and trade. Otherwise, the consequence will be significant under-utilized capacity and higher costs.

The WSCC was unprepared for the simultaneous onslaught of reduced hydro supplies and an unexpected spurt of demand growth in 1999 and 2000. The institutional framework for solving the shortfall was an immature and heavily regulated wholesale market. Despite the imposed constraints and market immaturity, the WSCC achieved what all markets are supposed to do. It reallocated resources and balanced the system, however imperfectly.

Meanwhile, however, the market has failed in a more important arena. It is causing politicians, regulators, and the public to lose confidence in liberalization and deregulation in the United States and worldwide. Any backtracking here would portend of an especially disappointing outcome. California's energy problems are unique. They should not provide an excuse to return to inefficient and distorting regulation. F


It was the price jump afterward that mattered.

Critics of the California market structure say it is suicidal to allow flexible wholesale prices when retail prices are frozen. This critique is too simplistic, however. In actuality, however, it was not the freeze that wreaked havoc as when the PUC lifted the rate freeze in the fall of 1999 for San Diego Gas & Electric Co., which had succeeded in recovering its stranded costs.

THE SCHEME. The press has widely reported that California froze electricity rates charged by the state's three largest utilities for a certain transition period. At the same time, it forced them to buy and sell at wholesale only through the day-ahead and real-time spot markets run by the California Power Exchange, where prices remained free to rise and fall. That scheme allowed California's utilities to recover stranded costs by pocketing the difference (if any) between what they paid for wholesale power (em estimated at the start to run about $25 per megawatt-hour (em and what they could pass on to customers through the frozen rates (em pegged at about $55 per megawatt-hour.

THE TRANSITION. When the rate freeze was lifted in San Diego, spinning wholesale prices out of control in the summer of 2000, the extra cost was passed to San Diego consumers who watched, horrified, as retail rates more than doubled in the early summer months. Southern California Edison (SCE) and PG&E had not yet claimed stranded cost recovery, so rates for their customers remained frozen, with the utilities absorbing the loss. And with competing suppliers unable to beat the capped retail rates, the direct access market responded in a predictable fashion; customers flocked back to the price-capped utilities. The return of retail customers accelerated the utilities' cash drain.

THE REACTION. In the end, consumer reaction in California was so deep-rooted that politicians and regulators would not accept rate increases as a solution, thus setting up SCE and PG&E for insolvency. The real problem was the design of the California market, which mandated the sale of utility generating assets, while blocking development of retail market competition and risk management. (em S.A.V.V., F.H.P.

  1. Most other U.S. systems have limited interchange capacity, even in key market areas like the New York City metro area. The direction of flows in the WSCC, of course, varies with the season and other factors.
  2. For the record, the council praised BPA's work but thought it saw shortcomings: "Bonneville's White Book analysis is very valuable in terms of signaling the need for concern. However, there are several assumptions that could result in the White Book analysis painting an overly bleak picture. First, the analysis does not account for imports beyond existing firm contracts. In reality, the Northwest is part of a Western system with a good deal of seasonal diversity in loads that makes a certain degree of reliance on imports a cost-effective choice. Second, the White Book analysis does not reflect how the hydroelectric system can frequently (but not always) be operated to manage through relatively short-term supply problems. Finally, the White Book chooses not to speculate about possible new resource development." See Northwest Power Planning Council, White Paper, March 2000, p. 1.
  3. Edison Electric Institute, Weekly Power Output.
  4. The West Census region overlaps with the WSCC. Growth estimates are based on U.S. GDP growth during period adjusted upward to reflect greater population and job creation growth in the West.
  5. NERC Electricity Supply Demand, 2000, based on year 2000.
  6. Energy Information Administration and Statistics Canada.
  7. The alternative would be a mixed market with spot supplies and long-term contracts that include capacity payments. This would produce levelized prices over multiple years.
  8. It is instructive to note that the loss of even one or two nuclear plants in Texas is sufficient to cause gas prices to increase.
  9. Edison Electric Institute, Weekly Power Output.
  10. Energy Information Administration.
  11. At least as of March 2001, their fears have been confirmed. Many have received no or only small partial payments for December supplies.
  12. The soft price caps set a maximum purchase price of $150 per megawatt-hour unless higher prices could be justified on a cost basis.
  13. Dow Jones, Energy Wire, March 28.
  14. The proposed settlement with SDG&E sat idle for months before it was turned down.
  15. Cal.PUC Resolution E-3683, July 6, 2000.
  16. Los Angeles Times and personal communication with city officials.
  17. Monthly averages from the South Coast Air Quality District, as analyzed by Robert McCullough.
  18. WWW.CAISO.COM
  19. As of early April, the quagmire still was not resolved. Most of the QFs had received only a tiny portion of their expected revenue. This shortfall has been especially difficult for cogenerators, who face extremely high natural gas prices. Many have had to shut down.

Articles found on this page are available to Internet subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.