Influencing the human and organisational factors in process safety risk assessments

Abstract

The PSRA (process safety risk assessment) is the most important element of an effective process safety management program. A PSRA is an organized and systematic effort but there are also highly complex tasks related to it. The effort is to create a complete identification of causes and consequences and subsequently set of effectiveness of barriers to manage the risk to acceptance level. The critical task is that the information to be compiled about the chemistry needs to be comprehensive enough for an accurate assessment of the reactivity hazards, fire and explosion characteristics and toxic releases. A successful PSRA also requires an unambiguously Risk Assessment Metric or risk graph, the availability of reliable clean (failure) data and must be organized properly. The organizational part of the PSRA's have to deal with demanding circumstances like time pressure, culture, workload, costs, motivation, procedures of classification and competent resources to complete the risk assessment in time. These elements together are typical ingredients introducing the human factor in PSRA's and the presence of so-called cognitive biases. These biases can have a negative influence on the validity of the risk assessment and can lead to incorrect and hence ineffective recommendations. Cognitive biases are defined by Kahneman as the tendency of systematic deviations when assessing risk instead of objective and rational judgment. The occurrence of these cognitive biases can be clarified by the understanding the dual-system of thinking (system 1 and system 2), Kahneman. The objective of this paper is to identify what kind of human factors and organisational factors are present in a PSRA and what can be done to prevent them? To identify these biases, we have evaluated multiple PSRA's from our company. The results of this evaluation show that human and organisational factor can be avoided, and more important incorrect conclusions and implementation of ineffective recommendations can be prevented.