2025 Spring Meeting and 21st Global Congress on Process Safety

(118b) A New Look at Safety Metrics in a World of Evolving Operational Practices

Authors

Julia Bukowski - Presenter, Villanova University
Dr. William Goble, Villanova University
On-going changes to the chemical industries’ operating practices (including testing and maintenance) which aim to proof test safety instrumented function (SIF) components on various schedules rather than test the entire SIF loop at once, and to further increase proof test intervals for some SIF components to as much as 10 years, will continue to affect achievable safety. The safety metrics currently used may no longer provide all the information needed to ensure the preservation of current safety levels and aid in further optimizing safety. This paper considers the limitations of the existing safety metrics average probability of failure on demand (PFDavg) and probability of failure per hour (PFH), presents another metric to provide additional safety information, analyzes more comprehensive models of SIF safety that account for many parameters relevant to changing industrial practices including rate of SIF demand (d), and demonstrates the importance to safety calculations of correctly modeling repair/recovery strategies when a SIF fails. Significantly, it identifies the ratio “rate of SIF restoration to fully functional with probability one (P(OK)=1)” to “rate of SIF demand” as an important additional variable impacting PFDavg. This ratio also defines a much wider range of applicability for the PFDavg metric than currently recognized in safety standards. This means, going forward, realistic evaluation of a safety instrumented function’s (SIF) safety requires that the SIF demand rate (d) be known at a facility, not merely assumed.