Results of Root Cause Analyses of Spring Operated Pressure Relief Valve Failures

A subset of 44 failed spring operated pressure relief valves (SORV) from a set of over 5600 Process Equipment Reliability Database (PERD) proof tests was subjected to root cause of failure analysis (RCA). This paper reports on the findings of those 44 analyses and includes specific examples of root causes (RC) identified. We considered SORV failures detected in both new and used valves over a period of almost seven years. We categorized the failures using the PERD taxonomy and, additionally, divided the failures into systematic (where reasonable changes to processes or procedures might prevent future failures) and random failures based on each RC. Further, we identified a number of subcategories of RC because they suggest possible actions to significantly reduce failure rates and improve safety performance. We considered how remedial attention to some root causes has markedly improved proof test results at the facility that provided the proof test data. We provide recommendations for conducting such RCA as well as for initiating a RCA program and discuss how time to valve failure (important information for computing valve failure rates) can be estimated based on the RCA findings. Non-Newtonian fluids are outside the scope of this paper.