(54z) Pitfalls of Using the Wrong Risk Matrix in PHA and LOPA

A Process Hazard Analysis (PHA) is an important tool for identifying and managing the hazards associated with a chemical process. Hazards may include safety issues, environmental impacts, business impacts, or reputational issues. It is now common to use a risk ranking scheme to discriminate among the potential hazard scenarios so that resources to reduce risk are allocated on a rational basis. Additionally, the PHA methodology has been extended to Layer or Protection Analysis (LOPA), which allows quantification of risk gaps based on a consistent methodology and industry data. The risk ranking process is driven by a company’s individual Risk Matrix, which governs the decision-making process and ultimately determines the extent of mitigation measures necessary to reduce risk. If the risk matrix used during a PHA or LOPA session has not been carefully reviewed, a number of issues may arise over the course of the study. On several occasions, the authors of this study had to request that company management revise their risk matrix to avoid unintended conclusions and decisions during the study. This paper aims to identify some of these traps and unforeseen consequences of using a risk matrix in PHA and LOPA.

Typically, hazard scenarios of the most interest in a PHA are those which have a high severity and a low likelihood of occurrence. Such scenarios are often subjected to the LOPA process to ensure that sufficient safeguards are in place. When looking at high severity, low likelihood scenarios, it is easy to unintentionally demand an overly conservative risk criterion. When this criterion is applied in a PHA or LOPA study, the results can lead to dramatic measures for risk mitigation, as a result of unacceptable risk levels or requirements for added safeguards which may not be warranted. Difficulties that the authors have faced in using corporate risk matrices have included (a) overly conservative risk ranking of high severity, low likelihood scenarios, (b) improper decision granularity and (c) risk matrices unsuited for the specific operation under review. These and other related issues are addressed with specific examples from PHA and LOPA projects conducted by the authors.