Exploiting Integrated Demand Response for Operating Reserve Provision Considering Rebound Effects

Electricity-driven thermostatically controlled loads (TCLs), e.g., air conditioners (ACs), have been widely utilized in demand response (DR) to provide operating reserve for power systems. However, the rebound effects may occur during the recovery process of DR, which can limit the operating reserve quality of ACs or even affect the reliable operation of power systems. With the community-level smart energy hubs (EH), the traditional electricity-driven TCLs can be expanded into multi-energy driven thermostatically controlled loads (MTCLs), e.g., household radiators. Under this circumstance, integrated demand response (IDR) can be exploited to coordinate the operation of MTCLs and provide more operating reserve resources while mitigating rebound effects. To this end, this paper proposes a two-stage IDR strategy to fully excavate the operating reserve provided by MTCLs. The first stage is to coordinate the energy consumption of ACs and household radiators to maximize the end-users’ thermal comfort and mitigate the rebound effects. To quantify the end-users’ thermal comfort, a modified predicted percentage of dissatisfied (PPD) index related to thermal environment parameters is introduced and simplified. Based on the energy consumption determined in the first stage, the energy conversion in EH is optimized in the second stage. Through the optimization in these two stages, a series of indices is established to evaluate the operating reserve in terms of aggregate capacity, duration, ramp rate, and smoothness. The case studies demonstrate that the proposed two-stage IDR strategy can provide high-aggregate-capacity and long-duration reserve resources in power systems while mitigating the rebound effects to maintain supply-demand balance and reliable operation of power systems. The analysis results of the test system show that the reserve capacity and duration obtained by the proposed model are 1.85 and 2.61 times those of the model without considering the multi-energy conversion, respectively.