Updating operating rules against voltage collapse using new probabilistic techniques

Abstract

A risk-based statistical approach has been carried out to assess voltage stability limits in the South Eastern part of France's subsystem and to construct new operating rules to avoid voltage collapse. The possibility of taking a large number of power system states as well as the uncertainties of the models into account enables us to assess accurately and robustly the stability limits of the system. This has lead Electricite de France and National Grid Company to develop the ASSESS tool to performs this kind of study. After this statistical study, a probabilistic refinement introducing probabilities of power system states has been carried out to validate the rules. Bayesian Belief Networks proved to be helpful for the evaluation of a power system state probability, because of their ability to model the uncertainties and dependencies of variables.