Quantitative Reliability Evaluation of Automatic Protective Equipment Used for Nuclear Reactors

Abstract

Reliability assessment of nuclear reactor protective system is presented via two reliability measures (or indices). These measures provide consistent estimate of the performance of the equipment and are: i) Time dependent probability of failure ii) Fractional dead time. Quantitative estimates of these reliability measures, when components have non-exponential failure distributions, are presented and compared with those obtained using exponential distributions. This comparison reveals some interesting features and leads to further insight into the behaviour of the protective system. The comparison shows that the exponential component failure distribution always provides optimistic estimates of the reliability measures which means that the system may not, actually, be as reliable as the exponential estimates would indicate. The difference in the reliability measures between the exponential and non-exponential cases is quite significant specially in the temporal evolution of the system failure probability. The difference in the fractional dead times (FDT) is dependent on testing interval. If the testing interval is long then the difference in the FDT between exponential and non-exponential cases is quite considerable (of the order of 50% for the example considered in this paper). On the other hand, if the testing interval is short, the difference is not very significant (of the order of 1%).