The Missing Link in Electricity Markets

What is the missing link in today's electricity markets? No, it is not centrally managed capacity markets, although they may be an important precursor to the emergence of what is missing. It is definitely not the imposition of price caps: soft, hard or otherwise. They work against the interests of an efficient market. Virtual bidding along with locational electricity pricing improve price formation and price discovery in wholesale markets, but they are necessary, not sufficient conditions for an efficient and effective market. Electricity markets, whether centrally organized with retail choice, or operated under a more traditional structure, need to intrinsically link reliability conditions and marginal supply costs to the prices consumers pay for energy to achieve the level of efficiency that can and should be realized. Fortunately, industry stakeholders have noticed what's missing, and initiatives to remedy shortcomings are being launched to fulfill a variety of interests.

Initiatives in Markets with Competitive Choice

Encouraged by the success of the National Grid's (NGrid) pioneering market-linked default service (for customers over two MW), New York utilities are extending real-time pricing (RTP) types of default service to all customers above 500 kW, linking hourly commodity prices directly to day-ahead New York Independent System Operator (NYISO) marginal prices (LBMPs). A study of NGrid's RTP customers showed that at least one-third of them responded affirmatively and substantially when hourly pries rose above typical levels. Competitive retailers report that many of these same customers are demanding an equivalent service, hourly prices tied to market LBMPs, but backed up by hedging options that can be applied to a portion of the load, or that can be exercised on short notice. This is the naturally occurring demand response and associated product diversity that competitive markets (and customers) have been waiting for.

Utilities in New Jersey, Maryland, and part of Pennsylvania have followed suit, but with a different wrinkle. Hourly default service prices for larger customers (800 kW and above in some jurisdictions) are linked to PJM's real-time market-clearing prices. The risk associated with unannounced, or unanticipated, high prices forces customers to decide for themselves how much exposure to undertake. Those that can and do respond enjoy greater savings compared to paying day-ahead market based prices, which on average are five percent or so higher. But, the real-time price discovery may cause many customers to hedge that would respond with day-ahead pricing. The effectiveness of this default design in achieving price response awaits a comprehensive study of how customers have adapted.

The Connecticut PUC is evaluating how to comply with a legislative mandate to employ a three-part, time-of-use (TOU) price structure for default commodity pricing for customers over 500 kW. One proposal under consideration would combine aspects of a TOU schedule and day-ahead LMPs, resulting in substantial and quantifiable benefits to both participants and all other electricity consumers.

Illinois recently enacted a law that requires utilities to make RTP available to residential customers on a voluntary basis, perhaps fueled by the results of the ongoing Commonwealth Edison (ComEd) and Community Energy Cooperative residential RTP pilot. Participant satisfaction in that ground breaking initiative has remained high, even when savings opportunities were modest due to low price volatility. Residential customers may be more willing to actively take control of when and how they use electricity than many previously thought.

Load Curtailments as Market Resources

The eastern ISOs have been at the forefront of promoting demand response in the competitive retail environment. They and FERC recognized the importance of demand response and that it was not going to bloom naturally, at least not as fast as needed. Concerns about the possibility of shortfalls of operating reserves caused market participants to allow end-use customers to offer load as a capacity resource available for dispatch on short notice. There are now over 2,000 MW of such resources fully integrated into the market operations of PJM, ISO-NE and NYISO. The recent heat spell in the northeast affirmed the need for these resources to avert forced outages, the consequences of which can have sizable economic effects. The political ramifications may be even more potent, and therefore important, given that they are measured in votes rather than volts.

ERCOT (Electric Reliability Council of Texas) has led the way in mining customer load management and control capabilities as an ancillary service. It has over 1,000 MW of demand response available to provide balancing services, and some of these customers are equipped to provide regulation. PJM recently opened up elements of its ancillary services markets to demand response resources, and New York is working toward the same end. ISO-NE is launching a pilot to demonstrate the value of demand response resources, even those available for small customers through device control, as dispatchable operating reserves.

System reliability is primary obligation of an ISO. Thus, its fostering of demand response to meet that obligation cost-effectively is compelling. However, it is not as clear that treating demand response as an energy resource improves net market efficiency. Scheduling load into the day-ahead market (and paying LMP) and dispatching loads into spot electricity markets (and paying them real-time LMP, or higher) can reduce price volatility, and more importantly be welfare improving, which means that it promotes the efficient use of all societal resources. ISO studies have demonstrated, however, that net welfare improvements are generally associated with conditions characterized by highly elevated market LMPs. As the ISOs have reported, a large portion of the scheduled and dispatched load reductions are associated with relatively low LMP, which can result in reduced, not increased, market efficiency as measured by social welfare. A recent ISO-NE sponsored study demonstrated the substantial advantages of achieving price response autonomously, as the result of customers paying retail prices that are directly and continuously linked to wholesale market LMPs, compared to those attributable to loads bidding as an energy resource in wholesale electricity markets. ISO-NE is working with its counterparts in utilities and state regulatory bodies to explore ways to foster price responsive behaviors through commodity rate plans that better link wholesale and retail prices both spatially and temporally.

Linkages are Equally Important in Vertically Integrated Utilities

Reports of the demise of vertically integrated utilities were apparently premature. They are and may define the market operations for the foreseeable future. However, they will likely be expected to adopt practices that have proven to improve the efficiency of supply in competitive markets. One such practice will be to foster demand response by offering customer pricing plans that contemporaneously link marginal supply cost and usage prices. Based on past experience, they are up to the task.

Many utilities began fostering demand response in earnest in the 1970s, beginning with interruptible rates. Over the next two decades, a wider variety of time-differentiated pricing plans were introduced. The list of accomplishments is long, but a few initiatives are particularly instructive.

Florida Power and Light is notable for its large and sustained commitment to harnessing the inherent diversity in home owner's need for air conditioning services to serve as dispatchable operating reserves. Xcel Energy's extensive load control program reflects its need to manage peak winter demand to assure system reliability. Gulf Power's residential critical peak pricing plan uses established day-types to link residential energy prices to system conditions. High prices replace the tariff schedule rate when load reductions are likely to improve system reliability. Like its French predecessor, which is the standard offer service for residential customers, this is a hybrid of time-of-use and real-time pricing (RTP). Several pilots have been launched to establish how customers respond to event-based pricing, and define the systems and enabling technology it takes to achieve that response.

RTP has been offered in over 50 jurisdictions since its introduction in California in the mid-1980s. The two-part, revenue neutral design has enjoyed sustained and demonstrable success in a few applications, most notably Duke, CSW (now AEP), and Georgia Power, and served a valuable niche in many other utilities. Utilities that assume the long-term obligation to serve the entire market load will likely have renewed interest in RTP programs because they offer additional resources to contribute to ensuring least cost supply, and they provide customers with choices in how they purchase electricity.

California, which is reconstituting its market around vertically integrated utilities, has been at the forefront in examining what role demand response can and should play in an efficiently run market. The results of its pricing pilot are being used to assess different ways to foster the development of persistent price responsive behaviors, including making time-differentiated rates the standard for buying electricity, and, developing an infrastructure that makes price response practical and beneficial to all retail customers.

Making the Right Linkages

Demand response is too important to the operation of efficient markets to be relegated to program status: it needs to be treated strategically and fully integrated into both physical and commercial operations. We are still learning how customers use and value electricity, so it's too soon to conclude which customers want and can benefit from dynamic pricing, or to stop being creative in designing pricing plans. It's just as important to recognize that pursuing demand response relentlessly, without goals and guidance, can make things worse, not better. Improved measures of the impact on the physical operation of markets and the subsequent level and distribution of benefits from demand response are needed to guide program design and to direct expenditures to foster demand response. Most importantly, they are needed to provide customers with an incentive to take control of electricity expenditures. The link may be missing, but fortunately interest in finding it is growing.