Price Perturbations for Privacy Preserving Demand Response With Distribution Network Awareness

Demand response (DR), where electricity consumption is shifted in response to incentive signals, can ease the transition to renewable generation. However, when many devices simultaneously respond to these signals there is the potential for violating local network constraints. Many of the proposed solutions require consumers to hand over control of their devices to third-parties. Here, we propose a method of using price distortions to coordinate distributed resources, which is both robust to local constraints and privacy preserving. We formulate the price distortion setting problem as a mixed-integer linear programming problem. Conditions are derived under which the method can guarantee constraint violations are eliminated within one time step. The method was tested using case studies involving both electric vehicles and smart heating/cooling systems. We show that the proposed method, under a scenario with maximum DR participation, can achieve 98% of the theoretical lower bound on the number of constraint violations. Furthermore, our method out-performs the benchmark where some devices opt-out of DR.

I. Introduction

Demand response (DR) broadly refers to altering power demand to achieve economic or grid-level goals. This is distinct from historic operation of the power grid, which has focused on altering power supply to match demand. Renewable generation presents increasing challenges for transmission system operators (TSOs), as they need to cope with greater uncertainty on the supply-side. The proliferation of devices such as electric vehicles (EVs) and smart thermostats in distribution networks has resulted in a large amount of latent flexibility being present at the low voltage level [1]. This presents an opportunity to improve the operation of the power system, without incurring additional infrastructure cost [2].