Emergency management for a nuclear power plant using fuzzy cognitive maps

Abstract

This paper explores the application of fuzzy cognitive maps (FCM) to emergency operating procedures (EOPS), to represent the decision-making process during abnormal situations in a nuclear power plant (NPP). The decision-making process in a NPP is a complex process, due to the many elements involved in its operation, and the permanent attention demanded by its maintenance. At the present time, the decision making process in a NPP is analyzed and developed by reactor operators, based on a set of instructions as well as flow charts to mitigate the consequences of a broad range of transients, accidents and multiple equipment failures, whose main characteristic is to be linear representations of events within a scenario. One of the main objectives of this paper is to present a method based in FCM that could be applied in the development of EOPS, and show some simulations, specifically the loss of coolant accident (LOCA) scenario in a boiling water reactor (BWR) with the Mark II containment design was studied. The FCM-based method represents with high fidelity the expert reasoning (the human expert is very important) and the interpretation of the results aids instantly to the reactor operators in the surveillance of the reactor proper functionality due that they have the responsibility of the decision taking in emergency situations. The simulations results show that the FCM predict properly the phenomenon in the reactor vessel and primary containment.