Power Distribution Systems Failure Response Against Extreme Events: a Stochastic Approach

Abstract

Power Distribution Systems are a critical infrastructure, and there is a growing interest related to their capacity to deliver and maintain power supply to the final customer. In this sense, the failure response of a typical Distribution System (DS) must be correctly assessed, especially when considering Low Frequency, High Impact (LFHI) events, such as extreme weather scenarios. Previous studies focus on topological and qualitative approaches, thus not considering the impact of extreme weather on the DS’s reliability metrics related to the frequency and duration of the failure. In this work, a fragility curve model was used to sample time until failure and time to repair inside a daily time window of a real Brazilian DS while taking weather variables into account. The Monte Carlo method was used to verify both power flow parameters and weather variables’ influence on the frequency and duration of failures. A gradual node removal approach was also modelled to investigate the impact of energy not supplied. Results show that the weather variables values can drastically change the failure response of the DS in both the time window and node removal approaches. Furthermore, when considering redundancy (alternative energy paths), in the most extreme weather scenario, an increase in system redundancy did not improve failure response, contrary to what is expected. In this sense, an extreme weather failure response analysis is recommended in any investment study related to a DS - such as switch or protective device placement.