An application of genetic algorithms to surveillance test optimization of a PWR auxiliary feedwater system

Abstract

Nuclear power plant systems are comprised of both on-line and standby components. Standby components differ from on-line ones, as they might be unavailable due to unrevealed failures. The usual procedure employed to reveal failures before real demands is to submit the component to surveillance tests. Surveillance test policies might deal with two conflicting scenarios: the test frequency must be sufficiently high in order to reveal failures before demands, but, on the other hand, it must be low enough due to its influence on the component unavailability. Standard surveillance test policies for typical nuclear power plants usually consist of periodic tests for which the frequencies are often higher than necessary for obtaining the optimal availability. In this work, a new surveillance test policy optimization method, based on genetic algorithms, is applied to the Angra-I (Brazilian PWR) auxiliary feedwater system. The new probabilistic model has been developed in order to comprise the following features: (1) aging effects on standby components when they undergo surveillance tests; (2) revealing failures during the surveillance tests implies corrective maintenance, and, consequently, increasing outage times; (3) components are distinct (i.e., each has distinct test parameters, such as outage time, aging factors, etc); (4) tests are not necessarily periodic. The results, when compared to those obtained by standard test policies, show improved overall availability at the system level. © 2002 Wiley Periodicals, Inc.