ENRON International has begun building a $150-million, 80-megawatt independent power project in Piti, Guam. Enron signed a 20-year energy conversion agreement to develop the baseload, slow-speed...
Metering, Loads & Profiles: Let the Cherry-Picking Begin
oft-stated goal that electric industry restructuring will benefit all customers.
Then why is load profiling used? Reliance on load profiling stems from two interrelated and widely held notions about electric industry restructuring. First, most believe that competition and its promised benefits ought to be available to all customer classes. Second, many say restructuring should reduce costs for all customers. It is this broad idea (em competition for all (em that finds small customers participating in retail choice programs with their usage patterns and associated costs estimated by load profiles rather than explicitly (and relatively expensively) metered. Because most utilities have some type of load research capability 'on the shelf,' adapting the utility's existing load research effort to accommodate the new retail choice regime becomes a low-cost way to allow small customer participation. Yet even the data processing systems needed to support load profiling are likely to be costly. %n4%n
It is certainly true, however, that load profiling could prove feasible as a way to allow universal retail choice given the time allotted for implementation. To apply the ever-popular telephone analogy, think of load profiling as measuring the amount of traffic flowing through a switch and dividing by the number of customers served by that switch to determine the traffic patterns and costs imposed by customers on their carrier. A less extreme analogy might have a sample of telephone customers metered and the results of the sampling process used to estimate the usage patterns of the applicable class. However, even this latter regime would represent an unacceptable basis to determine usage and costs imposed by customers on telephone carriers, because the variation in usage patterns across billing telephone numbers is thought to be large and, happily, precise usage metering is already available. Why then is such a regime to be allowed to determine cost causality in the restructured electric industry? Is it that smaller electric service customers, when grouped into rate classes, are believed to have homogeneous use patterns?
The current practice in today's electric industry makes winners and losers out of different customers in a single rate class by charging unit prices derived from class characteristics that do not precisely represent the costs imposed by those customers. Customers with beneficial cost characteristics who are "underappreciated" by prices in their current rate class may respond by opting out of one class into another more beneficial rate class that may reward the transferee with a lower bill. For the current monopoly service provider, such transfers are internal. But, with retail choice, a customer can choose a new supplier that offers either lower prices or the option of a new load profiling class, without changing usage patterns.
When customers migrate from their current class, this self-selection cherry-picking could degrade the load shape of the remaining customers in the class. If the load shape degrades, and load profiles do not adjust for the loss of these customers, another supplier (including the supplier of last resort) ultimately will find itself disadvantaged. If load profiles adjust quickly, then the remaining customers can be expected to pay more for