Distributed minimum spanning tree approach for critical load restoration using microgrid formation in resilient distribution systems

Abstract

Power grid is a vital infrastructure for society that can be severely damaged when facing high-impact and low-probability (HILP) events. Following these events, the power supply to critical infrastructures, known as critical loads (CLs), is often interrupted. This interruption can result in substantial financial, social, and security damage. Therefore, it is crucial to quickly restore de-energized CLs based on their priority and importance to society. This paper presents a distributed heuristic algorithm using a Distributed Minimum Spanning Tree (DMST) approach to restore CLs. The proposed algorithm offers several advantages over centralized optimization solutions, including robustness to single-point failures, scalability to large-scale networks, and reduced computational burden. In this algorithm, each controllable distributed generator (CDG), which has a restoration ancillary service contract with the distribution system, individually forms an islanded microgrid to restore CLs. The restoration process aims to maximize the number of restored CLs, enhance post-restoration reliability, and adjust restoration times. The effectiveness of the proposed approach is demonstrated through several case studies on the IEEE 123-bus test system. The results show that the algorithm successfully restores the maximum number of CLs.