Risk-informed Preventive Maintenance optimization

Abstract

The risk management group at the South Texas Project Electric Generating Station (STPEGS) has successfully developed a Preventive Maintenance (PM) optimization application based on a new mathematical model developed in collaboration with the University of Texas at Austin. This model uses historical maintenance data from the STPEGS work management database. Robust statistical analysis, coupled with an efficient algorithm generates an optimal PM schedule, based on a Non-Homogenous Poisson Process (NHPP) with a power law failure rate function. In addition, the risk associated with significant plant events triggered by a component failure is appropriately captured in the Corrective Maintenance (CM) cost estimates. The probabilities of such events are modeled via fault tree analysis, and consequences re expressed as monetary costs. The net cost of CM is then modified by a weighted sum of the probability of each event multiplied its monetary cost. The ratio of risk-adjusted CM cost to PM cost is used with the failure rate parameters to calculate the optimum PM frequency that minimizes combined CM and PM costs. The software can evaluate individual components or entire systems of components. Several low-risk ranked systems have been evaluated. In this paper we present the results of these evaluations