Hybrid simulation-optimization approach for power distribution systems resilience assessment

Electrical power is central to our modern society, and minimal delivery disruption underlies other critical systems ranging from healthcare to defense. Consequently, power distribution systems’ resilience is a strategic field, with resilience assessment playing a vital role in understanding and planning resilience enhancement strategies for this complex system. This paper presents a simulation-optimization approach to assess power distribution systems’ resilience. The method combines a Monte Carlo simulation to obtain failure and repair times for weather conditions and a metaheuristic to derive a service restoration plan for each component’s state transition during resilience estimation. We demonstrate the approach’s applicability on two example distribution systems under mild, adverse, and extreme weather scenarios. We evaluate the effect of service restoration during resilience estimation alone and in the presence of distributed generation in a grid-connected mode for two distribution systems. In addition to showing emergent resilience due to service restoration and distributed generation integration, which differ for each system, a resampling experiment is conducted to investigate the uncertainty around usual resilience metrics. Results show the benefits of our method in comprehensively assessing power distribution systems’ resilience and highlight the need to carefully examine the uncertainty associated with resilience indices during assessment or enhancement studies.