A Network-Constrained Demand Response Game for Procuring Energy Balancing Services

TL;DR

This paper introduces a privacy-preserving, network-constrained demand response game that incentivizes energy consumers to provide balancing services while ensuring secure distribution network operation and demonstrating scalability through test system simulations.

Contribution

It proposes a decentralized market clearing algorithm for a Generalized Nash Game with physical network constraints, ensuring efficiency and privacy in energy balancing services.

Findings

Low market efficiency loss demonstrated

Physical network constraints significantly impact results

Algorithm scalable to large test systems

Abstract

Securely and efficiently procuring energy balancing services in distribution networks remains challenging, especially within a privacy-preserving environment. This paper proposes a network-constrained demand response game, i.e., a Generalized Nash Game (GNG), to incentivize energy consumers to offer balancing services. Specifically, we adopt a supply function-based bidding method for our demand response problem, where a requisite load adjustment must be met. To ensure the secure operation of distribution networks, we incorporate physical network constraints, including line capacity and bus voltage limits, into the game formulation. In addition, we analytically evaluate the efficiency loss of this game. Previous approaches to steer energy consumers toward the Generalized Nash Equilibrium (GNE) of the game often necessitated sharing some private information, which might not be practically feasible or desired. To overcome this limitation, we propose a decentralized market clearing algorithm with analytical convergence guarantees, which only requires the participants to share limited, non-sensitive information with others. Numerical analyses illustrate that the proposed market mechanism exhibits a low market efficiency loss. Moreover, these analyses highlight the critical role of integrating physical network constraints. Finally, we demonstrate the scalability of our proposed algorithm by conducting simulations on the IEEE 33-bus and 69-bus test systems.