Imperfect Maintenance Scheduling for High Pressure Feedwater Heater System in Nuclear Power Plant

Abstract

As a complex and safety-critical system, the reliability of a nuclear power plant (NPP) has significant importance and effective maintenance program is one of the key considerations for its safe operation. High Pressure Feedwater Heater (AHP) system as an important subsystem in NPP needs to be endowed with an effective maintenance strategy. In normal operation, the AHP system works on heating of feedwater flowing through two high pressure (HP) heater banks A and B. Typically, component inner degradation and sensor control malfunction would affect feedwater outlet flow and temperature. This would further cause power coastdown and increase the operation and maintenance cost. This paper focuses on the optimization of the imperfect maintenance scheduling for an AHP system served in an NPP based on the Markov process and fault tree analysis. In particular, the component condition is modeled as a multi-state Markov process so to determine the component’s reliability curve under maintenance intervention. The system reliability function is then obtained by fault tree analysis that characterizes all the undesired outcomes with respect to feedwater outlet flow and temperature. To achieve a cost-effective maintenance satisfying system reliability requirement, a problem-specific genetic algorithm is developed and ultimately determines optimal periodic maintenance intervals for critical components in the AHP system. This study provides an understanding of the maintenance effects on the system reliability, and can also work as the basis to improve the nuclear maintenance management plan.