The target of this study is improving the electrical power systems performance estimation because of its importance in complex modern networks investment that have a lot of advanced equipments. Having a complete representation of the electrical system makes diagnosis thing much accurate and reliable which is used in electrical networks operation and maintenance. We focus in this research on implying different conditions and factors which affect operation and we use fuzzy petri nets for that purpose because of its characteristics in handling with incomplete and uncertainty data . we make our simulation on a real diagram of a refinery to compute its truth degree and discuss the results , we match that adding changing factors at operation contribute in improving networks estimation accuracy and support decisionmaking process .