|
Description
We analyze a number of unstudied aspects of retail electricity competition. We first explore the implications of load profiling of consumers whose traditional meters do not permit the measurement of their real-time consumption. The combination of retail competition and load profiling does not yield the second-best prices given the non-price-responsiveness of retail consumers. We then examine the incentives that electricity retailers have to install each of two types of advanced metering equipment. Finally, we consider the implications of physical limitations on the ability of system operators to cut off individual consumers relying instead on "zonal" rationing of large groups of individual consumers
1. Introduction
* Competition in the supply of electricity to retail consumers is being introduced in many countries. In the United States, more than a dozen states have introduced retail electricity competition, and the European Union has issued directives requiring that retail electricity markets be opened to competitive supply in all member countries. Yet there has been little study of the properties of retail electricity competition in the context of some of the unusual attributes of electricity supply, demand, pricing, and metering. The article analyzes a number of hitherto unstudied aspects of retail competition in electricity markets. Its starting point is that final consumers may not react to the real-time prices that emerge in wholesale electricity markets for (at least) three reasons: First, they do not have incentives to properly adjust their consumption to real-time prices (RTPs) if only their total consumption over a given period is recorded, i.e., they are on a traditional meter. Second, even if their consumption is recorded on a real-time basis, transaction costs associated with monitoring the evolution of hourly prices and constantly optimizing the use of equipment may be large for small consumers, although they may be reduced by Internet-based technologies in the future. Third, consumers, even if they want to, may not be able to adjust their consumption freely. They may be constrained by the physical attributes of distribution networks as they are currently configured; in particular, rationing usually occurs at the level of zones rather than individual consumers (what we refer to as the joint interruptibility problem).
In order to analyze competition among electricity retailers or load-serving entities (LSEs) for the final (retail) consumers, it is convenient to refer to the taxonomy of Table 1.
Price-sensitive consumers, who are studied in standard theory, are endowed with real-time (RT) meters and either autonomously or through communication with the LSE adjust their demand efficiently to the evolution of the wholesale spot market price.
Price-insensitive or partially price-sensitive consumers with real-time meters are endowed with RT meters but are unaware or only partially aware of RT prices and therefore do not adjust their consumption perfectly as real-time prices vary from minute to minute and hour to hour. At the extreme, they are fully (RT) price-insensitive. Such consumers are not irrational; rather, they trade off the transaction costs outlined above and the savings in their electricity bill. Although these consumers do not react to real-time prices, their actual real-time consumption can be measured and assigned to their LSEs for wholesale market settlement purposes.
Consumers on traditional meters are metered only once a month or every few months (in some countries meters are read even less frequently), and pay a per-kWh electricity charge that is independent of the actual timing of their overall consumption. The case of consumers on traditional meters can be decomposed into two subcases, depending on the way a consumer's LSE is charged for its energy purchases.
In the case of a monopoly local distribution company, this company pays the real-time wholesale market price of the consumer's consumption: even though the LSE is then unable to measure the realized profile of any given consumer with a traditional meter in its distribution area, it observes and pays for the costs it incurs in the wholesale markets associated with the realized total consumption profile of all such customers in the area in the aggregate.
Under retail competition, by contrast, an LSE other than the local monopoly distribution grid owner and serving such a consumer pays a unit electricity charge based on the "load profile" of the consumer. That is, it pays the average wholesale price for the load profile that is representative of the consumer's class regardless of the actual time pattern of the individual customer's consumption and the relationship between this actual physical consumption and the contemporaneous RT wholesale prices.
Note that Table 1 does not consider cases in which the customer is endowed with an RT meter, yet the retailer is subject to load profiling; presumably, the presence of an RT meter at the customer level also allows the LSE to be charged for and pay the actual wholesale market costs associated with the customer's RT profile. (1)
The article focuses on three possible failures of price signals to adequately reflect the scarcity conveyed by real-time wholesale market price signals. The first failure arises at the consumer level when only her aggregate consumption is measured because she is not on an RT meter. Because the consumer then does not pay more when consuming mainly at peak when wholesale prices are high than when spreading consumption more equally across peak and off-peak hours, the consumer consumes too much at peak and too little off peak. By contrast with the nonresponsiveness created by the absence of real-time consumption measurement, nonresponsiveness generated by the consumer's transaction costs is shown not to result in a market failure. The second failure occurs at the retailer's level, when the latter's individual consumers' real-time intake of power purchased in the wholesale market again is not measured by the system, so that the retailer is charged for power purchased in the wholesale market on the basis of some estimated consumption load profile rather than the LSE's consumers' actual load profile. The third failure may arise from the joint interruptibility problem.
Section 2 analyzes the behavior of load serving entities in a world in which consumers are homogeneous (possibly up to a scale parameter) and on traditional meters. The first subsection characterizes the second-best optimum and shows that it can be implemented in the absence of retail competition provided that the monopoly retailer is permitted to use two-part tariffs and, in particular, is not constrained to use linear prices. By contrast, the second subsection shows that (with or without the incumbent distributor) retail competition, which implies load profiling of consumers with traditional meters, leads to a retail price equal to the average wholesale power cost and differing from the socially optimal retail price. In a sense, retail competition plus load profiling transforms a second-best situation (with consumers free riding on the RT structure) into a third-best one (in which both consumers and LSEs free ride on the RT structure).
Section 3 first takes on the case in which consumers are on real-time meters but, in order to economize on transaction costs, do not react or only partially react to the real-time prices. We find that with homogeneous and rational consumers, retail competition leads to the second-best optimum; this is no longer true when a uniform pricing constraint is imposed (in the second subsection). The third subsection compares and contrasts our results to those in Borenstein and Holland (2005), hereinafter BH.
Section 4 extends the analysis to situations in which consumers differ in other aspects than just scale (i.e., they have different load profiles) and investigates the possibility of adverse selection and competitive screening.
Section 5 shows that given the price inefficiencies associated with load profiling, LSEs face the fight incentives when offering their customers enhanced metering equipment, and so subsidies for such equipments are not warranted. (2)
Last, Section 6 analyzes the implications of limitations in the controllability of the distribution circuits. These limitations imply that price-sensitive consumers may be rationed along with everyone else, and that LSEs cannot generally demand any specific level of rationing that they desire to reflect their consumers' valuations. At best one can then elicit only the aggregate willingness to pay for reliability in any given joint interruptibility zone. The section discusses... |
More articles from RAND Journal of Economics
Screening when some agents are nonstrategic: does a monopoly need to e..., December 22, 2006
Spin-outs: knowledge diffusion through employee mobility., December 22, 2006
Submarkets and the evolution of market structure., December 22, 2006
Coordination versus differentiation in a standards war: 56K modems.(Co..., December 22, 2006
Dynamic monopoly pricing and herding., December 22, 2006
Looking for additional articles?
Click here
to search our database of over 3 million articles.
|
%20%20&position=1)