Resilience metrics for improved process-risk decision making: Survey, analysis and application

The use of metrics in risk assessment and management of chemical process systems is a well-known area of research. There are numerous sources, which provide a comprehensive list of such metrics categorized into leading or lagging and technical or social. However, the significance of these metrics to overall risk and resilience assessment of the system is not well established. Process Resilience Analysis Framework (PRAF) is an advanced method of risk assessment with three phases - avoidance; survival; and recovery, which include twenty-four resilience metrics covering both technical and social factors. In this paper, we are reporting on the statistical analysis of the resilience metrics survey conducted within the process industry. The survey respondents present a wide variety of experience and employment sectors. This paper aims to answer the following three research questions related to resilience metrics- what are the most important metrics for each of the 3 phases?; are there any differences in viewpoints of various groups of survey respondents?; and what is the relative level of importance for each of the metrics? Answers to these research questions are critical in the quantification of overall process resilience and also provide the essential information for senior management to make informed risk decisions. Therefore, a PRAF survey based on Likert type questionnaire was conducted, which produced categorical responses. Methods and techniques such as ordinal alpha, Kruskal-Wallis test, and polychoric correlations, relevant to analyze categorical responses have been used in the programming language R.