Performability of electric-power systems modeled by non-homogeneous Markov chains

This paper treats a particular type of Markov system: time nonhomogeneous Markov system in discrete time. In order to adapt measures to this kind of system, performability measures are formulated, such as new indicators more dedicated to electrical power systems like 'instantaneous mean load curtailed' and 'mean energy not supplied on a time interval'. This method considers hazard-rate time variation in order to: 1) obtain more-accurate and more-instructive indicators; and 2) access new performability indicators that cannot be obtained by classical methods. An example from an Electricite De France electrical substation illustrates this.