A Study on Behavior Characteristics-based Load Flexibility Assessment in Residential Area

Abstract

Caused by the widespread of renewable energy resources, oversupply and oversecurity of reserve capacity in power system are serious concerns. Hence, system operators need load flexibility (LF) to mitigate the concerns. However, there is an inefficiency in aggregated LF operations due to individual customers’ load uncertainties based on their human behaviors. Therefore, in this paper, a novel LF potential estimation model is proposed based on resident behavior analysis. The proposed scheme comprises a resident behavior profile generation model (RBPGM) through deep learning and probabilistic clustering approaches and a virtual LF operation model (VLFOM) considering human comfort. Simulations were conducted in a residential environment in San Diego, California. The RBPGM superiority was confirmed through the internal clustering evaluation methods. Accordingly, VLFOM result confirmed LF potential depending on resident behaviors. A segmentation uncertainty assessment method was presented and interpreted from multiple perspectives. It was verified that customer targeting, one of LF operation strategies, to select appropriate residential customers for participation in LF aggregates load resources efficiently. Finally, the results presented that the proposed model is effective for the oversupply mitigation and CO2 reduction, while also suggesting meaningful implications of future LF market settlement design.