Markovian and semi-markov models for availability evaluation of NPP subsystems and equipment

The behavior of many subsystems and equipment in industrial facilities, including the nuclear power plants (NPPs) need to be modeled by means of time variant models over their lifetime. As regards the availability of subsystems and equipment in an NPP, it has to be evaluated over an operational cycle. The random character of relevant parameters related to the system under consideration and its loading conditions makes necessary to use stochastic models. In this paper we adapt some methods to evaluate the availability of systems based upon Markov and semi-Markov type models. Time intervals for maintenance and repair of subsystems (in NPPs) are taken into account as subintervals of an operational cycle. Specific distributional assumptions are selected for the time lengths for the most typical states. Thus, we extend our earlier research concerns reported in our contributions to SMiRT 15 and other conferences [6], [7], [8]. (authors)