“A Reappraisal of probabilistic risk analysis”
By: John Whittaker
The Article is an evaluation of probabilistic risk analysis carried out for the risk associating to the public safety. Fault-tree analysis, probability calculus and concept of acceptable risk are assessed by the help of recent examples to examine in this article with an intention to highlight the problems associated with these technological processes. Fault-tree deals only with the failures that the designers can conceive and not with all the potential failures in the system. It is an appropriate tool for the safety analysis of public safety. It is shown by the examples that all the faults and failures in a system are not foreseen and it can be results of unexpected interaction of different component, which is not considered in the fault-tree mechanism. All the probability calculations are not correct all the time as to use the probability calculus with the confidence, probability estimates must be derived from repeated trials. Sometimes the data such as for chemical reactors are very small to imaginary and hence cannot be trusted for public safety. The probability risk analysis makes people believe that new systems are better than the old ones despite the lack of tangible evidence. It is assumed that the events in the failure sequence are independent of each other but the research has shown they are interconnected. Acceptable risk is a rational concept; as long as a technological hazard does not present a risk that is quantitatively greater than those to which one is normally exposed, and then the risk should be acceptable. But it is not the way people make everyday decisions as it is not an effective way to provide assurance to public safety. Examples from the past and recent accidents are used to expose these flaws in the probabilistic risk assessment conducted for public safety. It is recommended that the theory of probability risk analysis should be abandoned and researches...