Application of perturbation analysis to sensitivity computations of generating units and system reliability

Power generating units do not contribute equally to system reliability. The contribution or reliability importance of units can be determined by the sensitivity of system indices to input uncertainties. Units with high sensitivities are the primary candidates for system upgrades or improvements. This paper extends the infinitesimal perturbation analysis (IPA) to complex Monte Carlo based generating system reliability programs and computes system and unit sensitivity measures. Generating units are represented by 3- and 6-state models and various important system operating rules and policies are recognized. Results from a case study are presented and discussed.