PROMETHEE and Fuzzy PROMETHEE Multicriteria Methods for Ranking Equipment Failure Modes

The objective of this work is to develop and implement a computer program of a methodology for the identification and ranking of failure modes of equipment in operation on electric power substations. It is proposed to rank the problems (failure modes) assuming a multicriteria context in opposition to the empirical methods today adopted. The methodology to support multicriteria decision PROMETHEE is compared with the fuzzy-PROMETHEE method, which is associated with the theory of fuzzy sets. In the fuzzy- PROMETHEE the input data are treated as fuzzy numbers, with the purpose of considering the uncertainty contained in the data. Using the fuzzy-PROMETHEE ranking one gets a more realistic failure mode ranking, considering the lack of data. The severity of the effects associated with the occurrence of each failure mode was used as a criterion for evaluating the methodology developed. It is known that functional failures affect businesses in different ways and may compromise the reliability of the system, operating costs, or even the safety or the environment. Therefore, different degrees of severity in terms of economical, operational, environmental, and safety impacts were attributed. A fuzzy inference system to obtain the overall severity of each crash mode, where the entries are the specific severities above, was built. With the overall severity of each failure mode it is possible to evaluate the risks associated with each one of them. Having a list with the prioritization of failure modes, a methodology for prioritization of actions most appropriate for the reduction or elimination of the consequences of each mode of failure can be applied. The major contribution of this work is to make available a refined model, considering multiple criteria analysis and the interests of different decision makers, for a maintenance plan to be carried out. This plan should aim to increasing operational reliability of equipment and reducing the maintenance overall costs.