Customer Response to Day-Ahead Market Hourly Pricing: Choices and Performance

In response to the problems of extreme price volatility and market power observed in some restructured electricity markets, policymakers and analysts are considering the relative roles of pricing and other regulatory and market interventions to improve their performance (Clarke, 2003; Flippen, 2003). Most agree that limited demand response (DR) at the retail level hampers the development of efficient wholesale markets. Relying on conceptual studies and anecdotal evidence, some have pointed to time-varying pricing, particularly real-time-pricing (RTP), as a mechanism to enable demand response (DR) and improve the linkage between wholesale and retail markets (Borenstein, 2002; Flippen, 2003; Horowitz and Woo, 2005; Turvey, 2003). Unfortunately, there is little publicly available information to help policymakers assess how well RTP actually works to elicit DR or to characterize its actual impacts on wholesale markets. Furthermore, in restructured electric markets, the new choices available to retail customers create a complex set of incentives. A few studies have examined industrial customer experience with RTP and found modest response (Boisvert et al., 2004; Herriges et al., 1993; Schwarz et al., 2002). California regulatory agencies and utilities recently sponsored a statewide pricing pilot for residential and small commercial customers and found load reductions ranging from 5- 15% in response to high price signals from a critical peak price tariff (Charles River Associates, 2005). However, all these studies examined voluntary RTP programs implemented in jurisdictions without retail choice. This research sheds light on how well retail pricing strategies actually promote demand response in restructured electric markets with retail competition. It examines the experience of 149 large customers of Niagara Mohawk Power Corporation (NMPC), an upstate New York utility, that have been exposed to hourly prices indexed to day-ahead, wholesale spot market prices as the default service under retail competition since 1998. Their hourly load and price data over five summers (2000- 2004) are supplemented by two phases of detailed customer survey and interview results to estimate demand models and to provide quantitative and qualitative context to model results.1 Detailed information on data sources, survey administration and response and demand modeling methodology is available in Goldman et al. (2005). Findings from this study are discussed in terms of customer choices in adapting to RTP as the default utility service in a competitive retail market environment, and customer performance, the actions customers undertook in response to hourly prices, their degree of price response and the aggregate impact on loads during high-price events.