Reliability-based nuclear power plant maintenance resource allocation in order to reduce unfavorable operational publicity

Abstract

Investigations of unplanned shutdowns at the Seabrook and the Pilgrim Nuclear Power Stations were carried out to improve the understanding of the cause of those events, which had been reported in Boston newspapers. Failures of turbine-generator systems and nuclear steam supply systems reduced plant reliability, while failures of safety-related systems did not. Risk analysis with fault tree modeling was carried out to quantify risk of major components in turbine-generator systems, and in nuclear steam supply systems. The risk was evaluated in terms of the frequency of unplanned shutdowns in order to identify which components are important to plant reliability, As a result, feedwater pumps, major valves installed on main steam lines, the chemical and volume control system, and switchyards are identified as risk significant components, Other equipment was also ranked according to their respective risk significance, Uncertainty analysis for failure rates of components was carried out using Monte Carlo approach to compare the result of fault tree analyses with historical data. Taking uncertainty into consideration, the analytical data and the historical data showed good agreement in regard to reliability over different time spans of operation, Maintenance practices for major components at the Seabrook Nuclear Power Station was investigated in order to correlate this practice and the components I risk significance. Some components, such as Feedwater Pumps and Main Steam Isolation Valves, are maintained by the condition-based methods; however, because they have high risk significance in terms of plant shutdowns, preventive maintenance is preferable for these components.