ABSTRACT

A major purpose of a fault tree is to assist in the calculation of the probability of failure, P(F) say, for systems built by simpler subsystems or components. A fault tree is constructed deductively until a suitable level of detail is reached. Input to the fault tree calculations are probabilities P(Fi) for base events Fi, where Fi denotes failure of subsystem or component i. Often it is assumed, and so will we, that the base events are independent. If for any i, Fi represents the event that subsystem i is down at a particular point in time, then P(F) is the probability that the system is down at that particular time point. Moreover, in the stationary case, P(F) aswell as allP(Fi)donot depend on time.