Stochastic reliability analysis

It is well known that the so-called independent failure analysis can produce meaningless results for redundancy structures. Apart from exhausting possibilities of modelling e.g., shared components explicitly, to date the preferred solution to overcome this drawback is to apply parametric common cause failure (CCF) models. Recently an alternative class of approaches has been proposed. This type of stochastic approach allows to elicit definitions, concepts and models of conventional CCF analysis. This paper reviews the generalized binomial distribution model for the reliability analysis of redundancy structures. Furthermore, procedures for data evaluation in order to determine the parameters of this approach are proposed