Identification of Critical Outages Leading to Cascading Failures in Electrical Power Systems

The extended black-outs that have occurred in the near past, have stimulated a renewed interest in the assessment of power system vulnerability. Accurate post-mortem analyses of the causes involved in a major black-out have shown that, in most cases, the outage of a line or a generator is the initiating event of cascading faults that may rapidly lead to a catastrophic failure. The role of hidden faults has been emphasized, but a systematic screening of the protection system in search for hidden faults remains a formidable and almost impossible task for today power engineers. The aim of the present paper is that of carrying out a more effective screening of single and multiple contingencies looking for those situations which are likely to jeopardize system security and to expose possible hidden faults. The problem of finding the worst branch or generator outages is recast in the form of a linear programming problem with mixed (real and binary) variables which is solved by efficient branch-and-bound techniques implemented in the CPLEX optimization package. Tests carried out with reference to the Italian EHV network and to the European UCTE system, confirm the effectiveness of the proposed contingency screening approach.