Western Merchant Plant Outlook
If the events of the past few years — price spikes in the Midwest, the California electricity crisis and the Enron bankruptcy — demonstrate anything, it is that electricity industry restructuring and competitive wholesale markets have not solved the riddle of the industry’s historical swings between undercapacity and overcapacity.
In the past, under cost-of-service regulation, the consequences of a utility’s poor timing for resource additions were generally absorbed in rates, although a utility could face adverse financial consequences from imprudent investments or inadequate service reliability.
In today’s market, merchant energy companies are worried that they will suffer a financial bust for building too much generating capacity. However, building too little capacity does not guarantee booms. Power shortages lead to price spikes, and price spikes produce volatile markets, which in turn lead to politicized power markets, regulatory uncertainty and turmoil in the financial markets. Thus, high prices are never sustainable because of the effects of politics and of market responses.
Despite the slowdown in power plant development, many analysts and industry watchers predict the US will have excess generating capacity for years to come.
However, a review of the historical ups and downs of power plant development, particularly in the western US, suggests that industry and market expectations rarely match the actual outcome. In fact, the current tightening in the financial markets may have provided the discipline power plant developers needed to sustain their financial health in the future.
This article explains the factors that contribute to the swings between under- and overcapacity in the power industry and discusses whether power markets in the western US are headed for a period of excess capacity.
Understanding the Variables
The cornerstone for electric resource planning is the assumption of economic rationality. That is, new power plants will be added only when costs are expected to be covered by revenues.
However, forecasting the need for power or future power prices is not easy. Even in the days when vertically-integrated utilities undertook sophisticated integrated resource planning processes and could count on recovering some costs through rates, reality seldom matched expectations. The Pacific Northwest provides a good example: over the past 25 years, this area has seen major shifts in expectations from under- to overcapacity every 5-7 years.
Certain characteristics of power markets make the industry prone to periods of under- and overcapacity. First, developing power projects requires long lead times. Project development activities for a 500 to 1,000 megawatt gas-fired, combined-cycle facility can easily require four to six years: one to two years to develop a viable site, about one year for permitting and another two to three years for construction. Controversial or difficult projects may require a decade or more. If major pipeline or transmission upgrades are required, the overall timing can again be eight to ten years.
Second, factors influencing either the demand- or supply-side of the load-resource equation can tip the capacity balance into an under- or overcapacity situation. These factors include the economy, policy decisions and the weather, among other things.
On the demand side, economic growth or recession will affect electricity demand. Yet, while long-term economic trends are difficult to forecast, it is even more difficult to forecast short-term variations in economic activity. This results in demand forecasts that either overstate or understate the actual demand for power. For example, in the early 1990s, electricity demand was projected to grow at about 2% per year. However, in the early 1990’s, California suffered its greatest economic downturn since the Great Depression. The economic slump depressed electricity demand growth and resulted in an overcapacity situation. On the other hand, in the late 1990’s, California’s economy expanded and, as a result, electricity demand soared and the state suffered from a shortage of generation capacity. Although demand growth was about 2% over the ten-year period — matching expectations for the long term — the state still experienced periods of over- and undercapacity.
In early 2002, the question for California is whether Silicon Valley’s high-tech firms will spark a new economic boom in northern California or relocate to states with more favorable business climates, leaving behind a new rust belt of yesterday’s technologies and dot-com dreams.
Short-term variations in electric demand can lead planners and policy-makers to act in ways that can exacerbate the over- or under-supply of generation. In the early 1990s, as a result of the slowing of electricity demand growth, policy-makers made several decisions affecting future resource additions. They terminated a resource solicitation. They encouraged the buyouts and shutdowns of “qualifying facility” projects. They provided Southern California Edison with economic incentives to close a troubled unit at the San Onofre nuclear generating station. Dams were removed from service for environmental reasons. The state’s power market was restructured with an exclusive orientation towards spot power (and without any capacity markets). Development of any significant new resources was delayed. Thus, once demand grew faster than expected in the late 1990s, the state was under-resourced.
As seen last year, policy decisions can have a substantial impact on electricity demand. One of the most important policy lessons to come out of California’s recent crisis is that it is possible to achieve significant energy savings in a very short time frame. The California Energy Commission estimates that the various education programs and utility demand-side management efforts reduced California’s peak demand by about 3,000 megawatts during the summer of 2001.The Pacific Northwest also achieved large savings in terms of its loads. The Bonneville Power Authority cut electricity demand by thousands of megawatts by buying back power from certain industrial customers such as aluminum smelters.
The impact of variations in the weather on loads is fairly obvious. Weather conditions vary around their expected values, but seldom precisely reflect average conditions. Mild weather reduces either summer cooling or winter heating loads, just as severe weather conditions strain the electricity system. Variations in annual rainfall also affect both hydroelectric generation and agricultural pumping loads. Milder weather and higher hydroelectric generation depressed power markets in 1999, while low hydro conditions in the Pacific Northwest in 2000 and 2001 contributed to supply shortages.
Third, the balance between supply and demand can be as readily influenced by shortfalls in supply as increases in loads. Along with natural variations in the level of potential hydroelectric generation, the expected level of supplies can be higher or lower than anticipated. Plant operators can increase the output or availability at existing power plants over time, while environmental restrictions, relative economics, or forced outages can reduce the availability of supplies. A portion of generation throughout the western US in the past two years has been unavailable because of emissions limitations and higher than expected forced outages following periods of higher than anticipated operations.
Fourth, power plant development can be adversely influenced by regulatory and market uncertainty. Throughout the West, project development has been dampened by both the tsunami of price shocks from California rolling through the interconnected, regional market and the resulting desire of other western states to erect firewalls around California rather than form a regional electricity market.
In California’s case, uncertainty over who will buy the power from new plants has stalled new plant development. California’s power problems left few creditworthy entities to pay for power supplies. The Pacific Gas & Electric Company filed for bankruptcy, while Southern California Edison spent almost one year on the precipice of bankruptcy. Utilities throughout the western US were financially weakened by the need to buy high-priced power in wholesale markets to cover their net short positions for their retail customers. It is impossible to use project financing for the construction of a new power plant without either an offtake agreement with a creditworthy entity or a functional wholesale market with a variety of creditworthy buyers.
California’s Department of Water Resources, or DWR, is one of the few creditworthy power buyers in California today. At one time, it appeared that the DWR’s portfolio of power purchase contracts might provide the credit support for the construction of new projects. However, this program is mired today in regulatory uncertainty as some California public officials attempt to renegotiate these contracts. Moreover, the bond sale needed to finance the DWR’s power purchases is stalled by disputes among the governor’s office, the DWR, California’s new power authority, the California public utilities commission and the State treasurer’s office, among others.
Finally, volatile fuel markets can dramatically change the nature of the optimal supply mix. In recent years, the conventional assumption was that low-cost gas combined with very efficient turbine technology would allow new combined-cycle power plants to displace the operation of many existing power plants. However, volatile gas markets have revived developers’ interest in coal, nuclear and renewable technologies as resources to be built into diversified power portfolio.
Resource Diversity in the West
The electricity market in the western US is characterized by a diverse mix of coal, nuclear, hydroelectric and gas generation.
Particular subregions are dependent on specific resource types. The resource mix of the Pacific Northwest is dominated by hydroelectricity from the Columbia River system, which has limited long-term storage. The resource mix of the inland Southwest historically has been overly dependent upon baseload coal and nuclear generation. In 1989, baseload resources accounted for almost 70% of the Southwest’s capacity. This fraction has decreased to under 50% today. California has been, and continues to be, particularly dependent on oil- and gas-fired generation.
Political boundaries obscure the natural, interrelated nature of the Western power market. Electricity demand peaks in the winter in the Pacific Northwest but peaks in the summer in California and the inland Southwest. Thus, seasonal exchanges — either through the market or through long-term agreements —allow one subregion to provide power in an off-peak season to another during a peak season. For example, hydroelectricity from the Pacific Northwest or coal and nuclear power from the Southwest can displace gas-fired generation in California in the summer months, thus mitigating to some degree California’s resource imbalance. This exchange of energy is then reversed when needed. Swings in hydroelectric availability or power plant outages also can be buffered more readily across the requirements of the region as a result of the complementary nature of the region’s resource mix.
The region’s resource mix has historical roots. In the Pacific Northwest, the Depression-era Works Progress Administration dams on the Columbia River dominated the region for the majority of a century. California’s environmental requirements and regulatory climate led to its dependence on gas-fired generation. In the 1970s, high oil prices and expected load growth led to major coal and nuclear construction programs throughout the West, but particularly in the Rocky Mountain and Southwest desert states. Many of these coal and nuclear plants became operational in the mid-1980s just as oil and gas prices crashed.
Bust, Boom, Bust?
The coal and nuclear plant development programs, coupled with California’s “qualifying facility” and demand-side management programs and a regional recession, led to an overhang of regional surpluses into the mid-1990’s. Inland utilities that had added resources as a bet on power sales into the wholesale market faced financial ruin as a result of excessive reserve margins. One utility, Public Service of New Mexico, found itself with a reserve margin of over 80%.
By 2000, the regional surpluses were absorbed by load growth. California’s story of supply shortages has been told many times in the past year. What is less well known is that the reserve margin for the Arizona-New Mexico-Southern Nevada region was just as bad as California’s, if not worse. The North American Electric Reliability Council reported that this region had negative reserve margins in 2000 and 2001, with frequent rolling blackouts averted only by a combination of the absence of severe regional heat waves, plants remaining on line and imports from other regions. Even so, rolling blackouts occurred in the Las Vegas area on July 2, 2001.
These regional shortfalls rippled throughout the whole western US. California was rudely surprised when it assumed it could rely on imported surplus power during its peak periods, primarily from the Bonneville Power Authority’s dams. In 2000 and 2001, the Pacific Northwest experienced one of its lowest hydro years on record. As a result, electricity imports into California dropped an average of 2,000 megawatts for the period from May through August in 2000 and almost 3,500 megawatts in August.
The shortfall in hydroelectric generation required significantly greater gas-fired generation, which in turn led to a congested gas transportation system, higher gas prices, greater air emissions and challenging operating schedules for the state’s aging fleet of existing gas-fired power plants. This should not have been a surprise. A well-circulated 1999 California Energy Commission study pointed out California’s vulnerability to supply shortages during statewide or regional heat storms.
In 2000 and 2001, price spikes in the West caused a surge in potential project development. In an average year, the western region requires about 3,000 megawatts of net resource additions, not counting capacity that is needed to make up for the low existing reserve margin nor that needed to replace retirements. Thus, with a four to six year development cycle, at least 12,000 to 18,000 megawatts should be in permitting or construction at any given time. Along with the 41,000 megawatts of approved plants or those undergoing review prior to 2000, over 60,000 megawatts of additional potential projects were announced in 2000 to 2001. If all of these plants were to come on line by 2010 instead of 2005, the WSCC-wide reserve margin would balloon to 60%. For reference, region-wide only about 11,200 megawatts (net) came on line in the WSCC from 1997 through 2000.
Analysts always knew there would be some attrition as project developers ran the gauntlet from press release to operating project. The higher the stack of press releases in the power plant gold rush era, the larger the number of canceled projects or deferral notices when forward price curves reflected the impacts of those potential projects. Thus, a boom of potential projects appears to have gone bust. Power Markets Week recently reported that developers have announced almost 85,000 megawatts of potential projects nationwide as either having been put on hold until markets look brighter or scrapped altogether. In California, nearly 5,000 megawatts of projects have been cancelled or postponed. In the West, nearly 60,000 megawatts of projects have been delayed, although not all for economic reasons. Calpine alone recently announced that it was placing 34 plants, totaling 15,100 megawatts, on hold. In December, Mirant said that it would defer or cancel any new plants beyond those that are under construction. Nonetheless, projects totaling over 30,000 megawatts have recently become operational or are already under construction, which will address near-term supply shortage concerns in the West. Furthermore, better hydro conditions, higher retail prices and continued conservation will reduce the need for thermal power plant output in 2002 relative to 2000 and 2001.
Even if only a fraction of the announced plants materialize, a strain will be put on the gas and electricity delivery infrastructure. A recent study by the Western Governors Association estimated that if most of the new demand is met through gas-fired generation, then $2.1 billion of pipeline expansions and upgrades would be needed. If the demand is met by using coal-fired generation located at the minemouth, an investment of at least $8 billion in high-voltage electric transmission infrastructure would be needed.
Historically, the timing of electricity generation capacity additions has never been particularly optimal. Long project development lead times combined with uncertain supply, fuel price and demand forecasts often give rise to over- and undercapacity periods. The consequences are often felt most strongly at the local level.
Regional markets for electricity are needed to facilitate the flow of power between areas that have either too much or not enough generating capacity. Regional markets also will provide some benefits in terms of load and resource diversity. Capacity markets, such as an “available capacity” or “installed capacity” market, may smooth out the power plant development cycle while some form of capacity payment should dampen the volatility of energy markets. Demand-side management can fill the gap when regions face unexpected surges in electricity demand and insufficient capacity. Developers will be able to mitigate their own risks by building diverse, national portfolios to act as a hedge against regional fluctuations in capacity markets. The challenge for the industry is to develop a political consensus on these issues that will lead to regulatory certainty and a favorable long-term investment climate.
The financial community can also play a role in smoothing out the project development process by seeking out objective, independent assessments of market risks and mitigation strategies as part of the due diligence process.
Industry restructuring and the expansion of competitive markets has led to a race among merchant developers to translate competing development plans into operating projects, particularly in regions which are seen as having too little capacity or a suboptimal supply mix. Perhaps fortunately, the plans of merchant energy developers have been influenced by Wall Street’s expectations as to emerging supply and demand trends. Each developer individually may have an incentive to build as much as it can. The financial community has a broader perspective and acts as a useful brake.