Construction cycles have been observed in a variety of industries including automobile manufacturing, real estate construction and commodity production (Achi 1998, Ford 1999, Meadows 1970, Sterman 1999). The cycles have not been as prominent in the regulated electric industry because of the Investor-Owned Utilities (IOUs) clear obligation to build the power plants needed to keep pace with demand. Nevertheless, IOUs have felt the impacts of building too much or too little capacity (Ford 1997). IOUs responded to the challenge with improved forecasting methods and by shifting investments away from capital intensive, long lead time generating technologies. Large nuclear and coal plants were no longer the top choices. By the 1990s, the IOU's marginal resource was substantially smaller, cheaper, cleaner, and it could be initiated only a few years in advance of need (note 1). With the prospects for modular, short lead time construction, the traditional IOUs could look forward to an era when construction cycles would not be a serious problem.

But the electricity industry will no longer rely on traditional rules. When restructuring is completed, the IOUs will be under no obligation to serve any or all of the demand. We will look to the market to deliver the power plants that will be needed in the future. The paper addresses the challenge of designing a market that can do the job in a stable manner.

Before focusing on the western US, it's useful to learn from others with more experience in deregulation/restructuring. Appendix A summarizes the experiences in the United Kingdom (UK). It concludes that there is little evidence that the UK is progressing toward a market designed to stimulate steady investments in new generating capacity.

It's also useful to review previous modeling efforts. Appendix B provides a brief review starting with an exemplary application of computer simulation in the UK. Appendix B then turns to modeling activities by the California Energy Commission (CEC) and the Northwest Power Planning Council (NPPC). These organizations have used a variety of models to forecast market clearing energy prices. Unfortunately, their models rely on a "perfect foresight" assumption that precludes the possibility of simulating a construction cycle. Investors do not have perfect foresight; they can only proceed with their own best estimates of future prices. Consequently, NPPC planners warn that their price forecasts

are likely to result in a smoother pattern of future prices than the actual market is likely to achieve. Most commodities tend to go through cycles of over and under supply and electricity is not likely to be different. The real electricity market is likely to experience price cycles of several years duration, in addition to the hourly and seasonal cycles that the model captures quite well. (NPPC 1998, p. 25)

The NPPC warning arises from a commonly held view that the transition to competitive electricity markets will create an industry with some of the standard problems observed in other commodity industries. It's useful to consider the cycles in other industries because their experiences could shed light on analogous problems in competitive electricity markets.

Electricity Versus Other Commodities

The construction cycle is chronic problem for agricultural commodities (Meadows 1970) and metallic commodities such as aluminum smelting (Ford 1999, Kaufmann 1983). The aluminum industry is well known to utility planners because of the its huge electricity requirements. It's also well known for the tremendous price volatility shown in Figure 1. Figure 1 shows price swings of plus-or-minus 100% in just a few years. Near the end of the 1980s, for example, the price of ingots increased to over 100 cents/pound. Within the next two years, it had fallen to only 50 cents/pound. Another upswing in the late 1980s brought the price back to over 100 cents/pound, but it returned to around 60 cents/pound by the end of the decade. It's no exaggeration to call the aluminum industry a "roller coaster" industry (Kaufmann 1983; Oregonian 1995).

Figure 1. Price of aluminum ingots in cents/pound, 1987 $ (Ford 1999).

Utility planners are concerned that the restructuring of the western electricity markets could lead inadvertently to another "roller coaster" industry. If the analogy with aluminum smelting is extended further, there are concerns beyond the construction cycle. For example:

Krapels (1996) argues that electricity markets could adopt the "classic" condition of commodities "that capability to supply chronically exceeds typical demand." Newbery (1995, p. 49) makes an analogy between aluminum and the UK electric system. He describes aluminum as an industry "where competition leads to both volatility and periodic price spikes, during which producers recover most of their fixed costs." Backus (1996, p. 18) believes that competitors in commodity markets come to prefer highly volatile prices and plan their "tactics" to exploit the volatility for high rewards. He warns us to expect similar tactics in competitive electricity markets.

My discussions with utility planners have revealed a general concern that a construction cycle could emerge in the western US, but there was no consensus on whether the cycles would be more volatile than in other commodity industries. Electricity is fundamentally different from other commodities because it can not be stored in inventory as a buffer between producers and consumers. Production and consumption must occur simultaneously. This unique requirement leads some planners to reason that an electricity production cycle could be more volatile than the cycles in other commodity industries. This reasoning seems to make sense because a reduction in the size of target inventories can be a destabilizing factor for agricultural commodities (Meadows 1970).

But a contrary view also makes sense. The simultaneity of production and consumption requires the industry to carry an extra reserve of generating capacity. Some planners argue that this extra capacity could serve as a "buffering element" that might be missing from other commodity industries. They reason that an electricity production cycle could turn out to be less severe than cycles observed in other industries.

As we consider these views, we should remember that the requirement for simultaneous production and consumption is just one of several complexities that defy our ability to reason by analogy (Loehr 1998). Although a construction cycle could be a real possibility, we are unlikely to learn more through analogies with other industries. It makes better sense to apply computer simulation to examine the dynamics of an electricity construction cycle.

Organization of the Paper

Part 2 describes a model designed to simulate the potential for cyclical behavior in the electricity markets in the western USA. The model simulates the key feedback mechanisms that give rise to construction cycles.

Part 3 presents simulations which reveal that cyclical behavior could emerge under a wide range of assumptions on demand, fuel costs, construction costs and investors' behavior.

Part 4 illustrates the use of computer simulation to search for a more stable design. It reviews the arguments for and against capacity payments, and it argues that constant capacity payments deserve closer study. Computer simulations are presented to show the magnitude of a capacity payment that would dampen the cyclical tendencies. The paper concludes by showing "the price of stability" to consumers buying electricity at wholesale and at retail. The concluding analysis reveals that we can achieve more stable electricity markets, and the "price of stability" is surprisingly small.