Giving up today's customer to retail wheeling could help cut losses tomorrow.
Estimates of stranded investment for U.S. investor-owned electric utilities (IOUs) range from as little as $20 billion to as much as $500 billion (em more than double the shareholder equity in U.S. utilities. These potential losses can be traced to above-market book values for some utility-owned power plants and long-term power-purchase contracts, as well as deferred income taxes, regulatory assets, and public policy programs.
A utility can react to the potential loss of retail customers in two ways.1 It can let the customer go, and then resell some or all of the lost load on the wholesale market. Or it can offer its at-risk customers the lower, competitive wholesale price and thereby retain the sale. These two options, keep the sale (KEEP) or relinquish it (FORFEIT), often yield very different estimates of stranded investment. (In principle, regulators could cap utility recovery at the smaller of these two amounts.)
Estimates of stranded investment for the two options depend strongly on interactions between the utility's generating assets (both utility-owned units and long-term power-purchase contracts) and the wholesale power market, but in different ways. Several key factors in the wholesale market will affect the level of stranded investment: wholesale prices, utility marginal production costs, transmission capacity, the percentage of retail load at risk, and the difference between wholesale purchase and sale prices.
Under the KEEP option, a utility loses revenue in two ways:
s Lost Margin. For those customers that wheel, the utility loses the margin between its embedded cost of generation and the market price of power.
s Lost Allocations. The utility may incur losses associated with rate-design differences in cost allocation and subsidies, if any, among retail customer classes.
Under the FORFEIT option, a utility's revenue loss will reflect three factors:
s Avoided Costs. The probable revenue savings represented by the product of the retail load at risk and the difference between the utility's embedded cost of generation for those customers that wheel and its marginal cost of production.
s Potential Profit. The possible revenue gain from the sale of some or all of the lost load on the wholesale market, based on the difference between the utility's marginal cost and the price it can get on the wholesale market. (The fraction of lost sales that the utility can resell, and the profitability of doing so, depend on the utility's marginal cost and available capacity for each generating unit, wholesale prices, and transmission constraints and prices.)
s Lost Allocations. The same effects (tariff differences in cost allocation and subsidies) as with the second factor in the KEEP option, which create losses associated with rate design.
At first glance, one might consider the FORFEIT option unrealistic. If the utility can sell at wholesale and make money doing so, why not keep the wheeling customers? Several reasons make both options viable. First, the departing customer may have reasons for purchasing from another firm that go beyond the simple matter of price differential. Second, once a utility relinquishes a sale, and gains relief from its obligation to serve that customer's year-round, hour-by-hour loads, it can pick and choose when it will sell this "surplus" power on the spot market. Third, transmission losses and congestion might otherwise limit a utility's ability to sell on the spot market. Finally, corporate or regulatory policy may limit the utility's willingness or freedom to offer a discount to customers.
Base Case: Modest Growth, Constant Costs, Limited Wheeling
To illustrate the differences between the KEEP and FORFEIT options, we used data from the U.S. Energy Information Administration to create a hypothetical electric utility (Utility A) suitable for analysis. We did not create a typical U.S. investor-owned utility, however; rather, we created one with a substantial amount of stranded commitments.
In 1995, Utility A's retail sales totaled 22,700 gigawatt-hours; peak demand reached 4,610 megawatts (Mw). At an average retail rate of 8.25 cents per kilowatt-hour (¢/Kwh), Utility A's revenues totaled $1.75 billion, for a net income of $148 million and a return on equity of 11.5 percent. Utility A also owns generating units with a combined capacity rating of 4,210 Mw. Two long-term power-purchase contracts provide another 1,600 Mw of capacity, bringing Utility A's reserve margin up to 26 percent.
With inflation at 3 percent a year, modest load growth of 1.1 percent per year, no fuel price hikes, no new generating units coming on line, and no new investment in existing generating units, Utility A's assets will decline from $4 billion in 1995 to $3.3 billion by 2005. In addition, all production costs remain constant in real dollars, causing retail rates to decline from 8.25 to 6.7¢/Kwh over the same period.
Assume further that retail wheeling begins in 1996, with 10 percent of each customer class remaining eligible from 1996 through 2005. (In reality, however, the timing and amount of retail wheeling would depend on utility rates, wholesale prices, customer price elasticity, and the outcomes of the restructuring proceedings occurring around the country.) Assume also that retail sales and transmission rates are the same in the retail wheeling case as they are in the base case. Thus, for purposes of this analysis, all costs of retail wheeling fall on utility shareholders.
By choosing a FORFEIT option (eligible customers choose wheeling service to buy from competitors), Utility A incurs a loss in earnings attributable to the lost load, but its annual loss in earnings falls from year to year, from $25 million in 1995 to $19.2 million in 2000, and to -$0.1 million in 2005. This reduction in losses in later years reflects a decline over time in the fixed-cost component of electricity costs, as well as a rise in wholesale prices after 2000. On a net-present-value (NPV) basis, the earnings loss totals $100 million (em 8.2 percent of utility equity as of 1995. Nevertheless, Utility A's potential loss is small (compared to that estimated for many U.S. utilities) because we assume that only 10 percent of the utility's load lies at risk.
By retaining the loads of its retail-wheeling customers, Utility A would lose only $82 million (em 18 percent less than it would lose under the FORFEIT scenario. In this case, then, the utility should choose the KEEP option if possible, retaining the customer at risk and charging them market-based rates.
Variable Case: Price & Grid Fluctuate; Wheeling Expands
How do stranded-investment estimates change as conditions vary from those discussed above?
Lower Prices Favor FORFEIT.
As wholesale prices increase, stranded-cost estimates fall for the FORFEIT option (see Figure 1 on page 23). Indeed, if wholesale prices are sufficiently above the utility's cost of generation, then stranded investment can be negative (i.e., the utility can earn more money selling on the wholesale market at spot prices than sell- ing to retail customers at embedded-cost prices). This negative relationship between prices and stranded investment depends on transmission capacity, however. Where transmission capacity is low, stranded investment is nearly invariant with wholesale prices.
Under the KEEP option, the utility's losses also depend on wholesale price. As the wholesale price increases, the utility's loss decreases because the difference between its production cost and the wholesale price declines. Because the utility's production cost depends partly on its purchase or sale of electricity to meet native loads, transmission constraints affect stranded-investment estimates. These effects are more pronounced under the FORFEIT option.
Constraints Affect FORFEIT More. The relationship between transmission capacity and the level of stranded investment under the FORFEIT option depends on the relationship between wholesale prices and utility marginal cost (see Figure 2). When wholesale prices are close to marginal cost, transmission capacity has little effect on stranded-investment estimates because there is little profit to be earned on purchases and sales. Where wholesale prices fall below marginal cost (in which case the utility would purchase power), increasing transmission capacity increases the amount of stranded investment. The utility will buy as much power as it can up to the transmission limit, but under the retail-wheeling scenario, the utility's marginal cost is lower than in the base case, so it has less need to buy on the spot market. Thus, increases in transmission capacity benefit the base case more than the wheeling case, leading to increases in stranded-investment estimates. The reverse is true at higher wholesale price levels. Here, as the transmission capacity is increased, the utility sells more power. In the retail-wheeling case, the utility's marginal cost is again lower so it can sell more power on the wholesale market. Thus, at high wholesale prices, the utility earns more money on the spot market with retail wheeling than without.
Under the KEEP option, the utility's losses depend only slightly on transmission constraints. As explained above, the utility's production costs may depend on its wholesale purchases or sales to meet native loads. The KEEP and FORFEIT estimates are closest to each other when transmission capacity is large.
More Choice Favors KEEP. Losses under the FORFEIT option increase more than linearly with the number of retail-wheeling customers (see Figure 3). With low wholesale prices as retail wheeling increases, the utility buys less and less (because its marginal cost decreases with increased wheeling), and eventually begins to sell on the spot market. At high levels of retail wheeling, the utility is unable to sell as much as its marginal cost would suggest because of transmission constraints.
Losses under the KEEP option increase linearly with the number of retail-wheeling customers. Thus, per wheeling customer, KEEP losses are invariant with the amount of wheeling because neither prewheeling production costs nor wholesale prices depend on the amount of retail wheeling.
Input Costs a Wash. Other factors affect stranded-investment estimates but have no effect on the utility's interactions with the wholesale market. Because these factors affect the utility's income statement and balance sheet directly, they have the same effects under both options. These "accounting" factors include the fixed costs of generation (utility-owned as well as power-purchase contracts), regulatory assets, administrative and general costs, and the costs of social programs.
A Final Analysis
The amount of stranded investment a utility will face depends on whether it keeps or loses sales at risk of retail wheeling (see Table). Factors that affect wholesale markets are crucial under the FORFEIT option: variable production costs, wholesale-power prices, and transmission constraints, for example. Fixed generating costs (initial cost, depreciation, taxes, and fixed O&M), public policy programs, and income-tax rates are examples of accounting factors.
Under the KEEP option, utility revenues reflect the supply price (purchased-power price) at wholesale. Under the FORFEIT option, resale revenues depend more on the market-based sales price, which, in our analyses, is equal to or less than the supply price. Also, the FORFEIT option remains vulnerable to the fact that the utility may find itself unable (because of generation and transmission constraints) to resell all of its free capacity and energy on the wholesale market. On the other hand, the FORFEIT option allows the utility to choose when to buy and sell on the wholesale market, based on the time-varying relationship between its marginal costs and wholesale prices.
Clearly, when the wholesale price is higher than the utility's marginal cost, the utility is better off keeping the sale. How much better off will depend on transmission capacity, the percentage of customers that wheel, and the difference between supply and sales prices in the wholesale market. The KEEP option is generally preferable where the wholesale price increases relative to the utility's marginal cost, the percentage of customers that wheel increases, or the difference widens between supply and sales prices in the wholesale market. t
Stan Hadley, Eric Hirst, and Lester Baxter are all researchers in the Energy Division at Oak Ridge National Laboratory in Tennessee. Their research focuses on electric industry restructuring issues, such as stranded commitments, ancillary services, and public policy functions.
1 For additional details see E. Hirst, S. Hadley, and L. Baxter, Methods to Estimate Stranded Commitments for a Restructuring U.S. Electricity Industry, ORNL/CON-424, Oak Ridge National Laboratory, Oak Ridge, TN (January 1996).
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.