Incorporating Demand Resources into Optimal Dispatch

Abstract

This paper presents a methodology for integrating demand response into optimal dispatch algorithms for electric power systems. Current methods to dispatch demand do not typically account for the impact of load shifting to later time periods. A critical component to properly dispatch demand resources is the inclusion of the rebound effect. Since the time scales for many demand response implementations are on the order of hours, once a demand resource has been dispatched, it is likely unavailable for re-dispatch during the same day. It is also likely that dispatched demand will increase in subsequent time periods. Incorporating the limited number of daily dispatches and the rebound effect into the optimal dispatch of demand resources is therefore necessary. This paper first provides a framework for incorporating demand resources into optimal dispatch and then presents a numeric example that compares optimal demand dispatch programs with and without the rebound effect. This comparison demonstrates the inefficiencies associated with a large-scale demand response program that does not take the rebound effect into account.