Abstract
The standby combustion turbine units in a nuclear generating station are periodically tested, and they are repaired if the test reveals a failure. The unique test procedure in this application is not accommodated by existing reliability models. In this paper, a stochastic model is developed, and the optimal time interval between two consecutive tests is computed. Comparisons with the current test practice suggest that improvements can be achieved in both the reliability and operating economy of such standby systems.

This publication has 4 references indexed in Scilit: