Uncertain about Uncertainty

Abstract: It is well-known that considerable uncertainty characterizes Process Safety Management with respect both to frequency of undesired conditions and events and to the consequences thereof. In the extreme (and in all sorts of human endeavors), the uncertainty that surrounds our sense of “likelihood” may give rise to the Precautionary Principle or may even cloak a Black Swan. This short paper describes three general approaches to defining probability and focuses on the ambiguity that characterizes qualitative or subjective probability. Particularly with Risk Based Process Safety, PSM project teams very commonly use subjective probability, because it is simple and often because more definitive data are scarce — but what does it mean when a team of 7 members, or maybe just 2, decides that the chance a condition or event may occur is “rare,” or “likely,” or “a toss-up?” In fact, this problem is so common that we hardly consider it in the multitude of decisions that we make every day. This paper explores ambiguities inherent in such terms and presents results from group tests that demonstrate that, absent some disciplined definition, individual players may, and often do, assign substantially different values in making those judgments. These results show that it is very important to use clear, agreed and consistent definitions for the subjective probabilities that will apply in a project — whether a PHA, a MOC, a PSSR, or any other project element where uncertainty lurks. Teams must think carefully about how to define uncertainties as they characterize facilities and conditions that may enable or cause incidents.