Carrots, Sticks and Other Smart Grid Tricks: The Capacity Value of Retail Demand Response

Various flavors of retail demand response or "demand side management" have been employed for decades by electric utilities as a way to reduce or defer system upgrade costs, and the number of utilities experimenting with retail demand response has grown along with technologies that enable finer-grained metering and pricing of electricity consumption. The existing literature suggests that consumers do respond to time-varying retail prices; this response is enhanced by information technology; and retail demand response can be an effective tool for lowering the costs of meeting installed capacity or other system investment requirements. A recent set of experiments involving electric utilities in Vermont, however, suggests that the capacity value of retail demand response may not be as high as previous studies have suggested. A randomized control trial involving thousands of electric ratepayers in Vermont over two summers used three different price instruments (time-invariant rates, peak-time rebates and peak-level pricing), different modes of communicating prices to customers, and information technology to examine the potential for peak-time load reductions. Over the course of the two summers, we observed results consistent with previous studies and the behavioral economics literature: Consumers did respond to prices, reducing usage at a level consistent with prior experiments; this demand response was higher for customers exposed to peak-level prices as opposed to rebates; and information technology tended to increase the magnitude of response. A finer-grained examination of the persistence of customer response, however, revealed substantial variation in the magnitude of load reductions across time. These variations appear uncorrelated to weather or other exogenous variables, and were not isolated to any one price or information treatment. Because the persistence of response on a day to day basis was so low, the capacity value of the retail demand response programs is unclear. Analysis of attrition behavior reveals additional lessons for the design of retail demand response programs related to socioeconomic status of households and how utilities communicate with customers involved in time-varying pricing programs.
Seth Blumsack is Associate Professor of Energy Policy and Economics and Associate Department Head in the John and Willie Leone Family Department of Energy and Mineral Engineering at The Pennsylvania State University and Chair of the Energy Business and Finance program. At Penn State he also serves as a faculty member in the Operations Research program; Co-Director of the Penn State Energy and Environmental Economics and Policy Initiative; and holds the John T. Ryan Faculty Fellowship in the College of Earth and Mineral Sciences. He also holds a position on the External Faculty of the Santa Fe Institute and is an Adjunct Research Professor with the Carnegie Mellon Electricity Industry Center and the Centre for Energy and Mineral Economics at the Curtin University of Technology in Australia. He earned a B.A. in Mathematics and Economics from Reed College in 1998, an M.S. in Economics from Carnegie Mellon in 2003, and a Ph.D. in Engineering and Public Policy from Carnegie Mellon in 2006.