“Centralized vs Decentralized Demand Response: Evidence from a Field Experiment” 

Megan Bailey* (University of Calgary), David Brown (University of Alberta), Blake Shaffer (University of Calgary), and Frank Wolak (Stanford University)

We report on a large-scale experiment to evaluate the effectiveness of centrally-managed (utility-controlled) versus customer-controlled demand response to dynamic prices by residential customers. Both groups of households have wireless communication devices installed in their dwelling to control as many as three types of appliances using a mobile phone app: (1) baseboard thermostats, (2) hot water heaters, and (3) electric vehicle chargers. Centrally-managed households can use the mobile app to opt-out of a utility-controlled demand response event, whereas customer-controlled households use the mobile app to reduce remotely the electricity use of any of these appliances. We find that the centrally-controlled households reduce demand by 27% on average during 3-hour demand response events and customer-controlled households reduce their demand on average by 6% during 3-hour demand response events, even though both sets of households receive the same financial compensation for the same percentage demand reduction. ​The percent demand reduction of customer-controlled households is not statistically different from that of a third group of experimental households. Households in that third group receive the same compensation for demand reductions during demand response events, but have no ability to remotely control their appliances and only receive the demand response event signal on the mobile phone app.



*Denotes presenter