Security Constrained Optimal Power System Dispatch Considering Stochastic Power Facility Failures Under Extreme Precipitation

Abstract

Power systems are experiencing more extreme precipitation events and the resulting waterlogging disasters than ever before due to climate changes, which may further induce power facility failures or even cascading contingencies, depending on the operation states of a power system. Therefore, it is essential to establish a waterlogging induced stochastic power facility outage model along with the appropriate power system scheduling strategy against extreme precipitation events. In this paper, the rainstorm waterlogging model is constructed to calculate the water depth across power systems according to geographical information and possible rainfall scenarios. The latter and the resulting water depth distribution is uncertain and so are the power facility failures. Based on typical failure scenarios extracted from the initial scenario library, an optimal unit commitment scheduling strategy considering load-shedding and system operation costs is proposed to minimize the adverse effect of extreme precipitation events, complying with the constraints on generator start-up and shut-down, and post-contingency frequency stability. The effectiveness of the proposed method is verified by case studies based on a modified IEEE 24-bus test system. Simulation results demonstrate that the method can optimize the operation cost as well as enhance the power system resiliency under extreme precipitation events.

Keywords: Extreme precipitation, Waterlogging disasters, Power facility failure model, Security constrained unit commitment