DIRECT-EVALUATION ALGORITHMS FOR FAULT-TREE PROBABILITIES

Traditional fault-tree methodologies have estimated the probability of system failure based upon a prior determination of minimal cut sets. Recent direct-evaluation methods offer an alternative which is exact rather than approximate and which often is faster than cut-set based methods. Furthermore, these direct-evaluation methods are easily implemented and used in a microcomputer environment. We examine and compare some of the most promising approaches that have emerged during the past few years. A large number of examples are included to illustrate not only the manner in which the various direct-evaluation methods differ from each other but also the potential for such methods to replace cut-set methods for probability calculations involving systems of moderate size.