Statistical considerations on safety analysis

Abstract

Alerting experience with a well-acknowledged safety analysis code initiated the authors to pay attention to safety issues of complex systems. Their first concern was the statistical characteristics of such a code. We point out a remarkable weakness of the so called 0.95/0.95 methodology: when repeating the search for the tolerance limit, we get a higher value with non-negligible probability. We propose the sign test as an alternative method. We point out the correct form of Wilks' formula when the number of parameters subjected to limitation is two or more.