PT Notes
Uncertainty and Subjectivity in Using Risk Matrices
PT Notes is a series of topical technical notes on process safety provided periodically by Primatech for your benefit. Please feel free to provide feedback.
Risk matrices are used in process safety to provide a simple means of rating and ranking the risks of events such as hazard scenarios using severity and likelihood values. In particular, they have become a key aspect of performing process hazard analysis (PHA) such as Hazard and Operability (HAZOP) studies. Risk matrices appear to provide a straightforward means of estimating risks but this appearance is deceptive. The design and use of risk matrices poses a number of challenges. These challenges include addressing uncertainty and subjectivity in estimating risks.
Risk estimates are subject to various types of uncertainty. However, risk matrices do not address uncertainties explicitly, other than by providing ranges for risk estimation. Also, risk estimates are subjective in nature and rely upon the opinions of risk matrix users, which introduces additional uncertainty. Lack of consideration of uncertainties in risk estimates can lead to the acceptance of risks that exceed tolerable levels and a higher likelihood of catastrophic accidents. Thus, process safety practitioners must be aware of how uncertainty and subjectivity should be addressed when using risk matrices.
Practitioners tend to think of severity and likelihood values in linear terms because their levels frequently are defined using cardinal numbers. However, the levels commonly represent severities and likelihoods in logarithmic space and differ by orders of magnitude. The risk matrix cell with the highest severity and lowest likelihood covers the largest part of risk space. Thus, scenarios that are assigned to high severity and low likelihood levels (i.e. the events of greatest interest in process safety) unfortunately have the most uncertain risks.
Furthermore, risk ranking fundamentally is a subjective process open to the unpredictability of analyst opinions and subject to the effects of human and psychological factors such as heuristics and cognitive biases. For example, the availability heuristic can result in biased estimates of event consequences and likelihoods. Arguably, subjective judgment is the largest source of uncertainty in using risk matrices.
Uncertainties are a key aspect of risk characterization and must be addressed in risk-informed decision making that uses risk matrices. Moreover, the ways in which human and psychological factors affect the use of risk matrices must be recognized and addressed by practitioners.
Guidelines are provided for addressing uncertainties in risk estimates and managing the subjective judgment involved in making them are provided in the article:
Addressing subjectivity and uncertainty in using risk matrices, Loss Prevention Bulletin, Issue 252, pages 17 - 20, December, 2016.
You may contact Primatech for further information by clicking here.
To comment on this PT Note, click here.
Copyright © 2017, Primatech Inc. All rights reserved.
enteiconOf Interest
