An Aging Simulation Model for the Emergency Diesel Generators of Angra 2 by Supplementary Variables

Abstract

The extension of the useful lifetime of nuclear power plants is an issue of great importance and concern. From the reliability point of view, this problem requires the consideration of time-dependent failure rates and possible failure dependencies. This analysis has been typically performed through a Markovian approach. To illustrate this point, we have developed a computerized reliability analysis of the emergency diesel generators (EDGs) of a four “loop” PWR plant, considering the hypothesis of aging and perfect repair by using Supplementary Variables to cast the initially Nonmarkovian model into a Markovian one. In order to perform such analysis and to simulate aging effects, a nuclear plant has been taken for reference, which has been commercially operating for only six years. Failure rates were obtained from similar EDGs of another plant, already under aging, while repair data were taken from its technical specifications. Discontinuous repair rates were considered in order to improve maintenance strategies. Several ages were attributed to these equipments, allowing the calculus of the failure probability as well as their availability according to each regarded age. In this sense, the EDGs behavior as to aging can be obtained in detail and decisions concerning maintenance and useful lifetime extension can be made on a stronger basis. To get the desired results in terms of reliability figures and due to the discontinuous repair rates that had to be taken into account, a new numerical method that uses a part of the analytical solution, called Euler Iterative + Characteristic, has been developed in order to solve the differential equations systems, making the solving of the system faster and more efficient.