A cascading failure model based on AC optimal power flow: Case study

Abstract

Simulating the grids cascading failure process is an essential means of preventing cascading failures. In traditional cascading failure models, DC power flow models are applied widely, but reactive power characteristic cannot be reflected. This study improves and applies an AC-based Cascading Failure model (called ACCF model), which captures bus load shedding and branch failures, all via AC power flow and optimal power flow analyses. Taking the IEEE 30- and 118-bus power systems as case studies, the ACCF model is proved feasible. With case studies, this study reveals that during the cascading failure, the broken branches are not necessarily close to the initial faulty elements, and some of the affected nodes/branches are “far” away from the initial faulty nodes. And as the initial branch failure probability increases, the system real power loss probability function gradually changes from approximate power distribution to a normal distribution. Meanwhile, the study also discovers that as the initial branch failure probability further increases, the system real power loss changes from a normal distribution to a distribution that appearing to be symmetric with the loss function under a low initial branch failure probability. The findings could facilitate grids safety and stable operation, as well as grids disaster prevention and relief.