Real‐time risk assessment and decision support

Abstract

Managers and operators of major hazard facilities make complex decisions as a part of their daily work activity. These decisions are made against the background potential for a major accident. Such decisions may be required to account for daily changes in a large number of factors including plant condition and performance, operational status, knowledge and experience of personnel, interactions with other activities, and the effectiveness of processes. The information involved in the decision comes from multiple sources and may be difficult to assess. Technical risk assessments provide a useful picture of major accident risk but some widely accepted approaches suffer from some significant problems which limit their value as tools for operational decision making. The article describes investigations into an approach that addresses how these difficulties may be addressed in day‐to‐day assessments. It describes a method and tool in which risks can be monitored in real‐time and so enable safer decision making. The method is applicable to the assessment of a wide range of major accident hazard scenarios. The article will describe how the tool addresses problems in some alternative approaches. A significant feature of the approach is its ability to identify the most probable causes of risk. The speed of the assessment points to its potential use in real‐time detection and control systems. The method employs a Bayesian net to perform the risk assessment. Bayesian nets have been used to aid decision making in many different situations and industries, but have received relatively little attention as risk assessment and decision tools in major hazard industries. The article will include a description of the benefits offered by this technology as well as a view of its limitations. © 2014 American Institute of Chemical Engineers Process Saf Prog 34: 183–190, 2015