In this paper one presents a data envelopment analysis and fuzzy sets based hybrid model application, in order to establish a priority ranking for the failure modes evaluated in a failure mode and effect analysis. The main objective is to present the applicability of this hybrid model to analyze complex systems, as the chemical and volumetric control system of a pressurized water reactor power plant, where a great bunch of failure modes must typically be evaluated. The presented modeling makes it possible the consideration of linguistic variables, i.e., it permits that one considers most effectively expert opinion in a risk analysis context when scarce data is available, as may occur in the plant design phase. The results obtained demonstrated the powerfulness and the viability of use of the proposed hybrid modeling to establish a priority ranking among many failure modes concerning a complex system.
CVCS; FMEA; DEA; expert opinion; risk priority number
