In order to accurately describe the potential risk of catastrophic accidents in a power system, a new united uncertainty method to assess catastrophic accidents was proposed based on their inherent physical properties and the uncertainty theory. The catastrophic accidents were regarded as the uncertain events, and the concepts of credibility measure and global fuzzy security index were developed to describe the likelihood and severity of catastrophic accidents. The N-1 contingencies were taken as the initial accidents and the most possible catastrophic accident sequences were identified through the risk ranking of all N-k contingencies in each simulation layer. The WSCC nine buses system and a regional practical system were simulated. Compared with the traditional methods, the proposed method was more accordant to practical power system operating conditions and had better accuracy and validity.
