Systematic failures are the result of oversights, lapses, mistakes, and decisions of not only people, but of organizations. The UK HSE showed that most accidents involving control and safety system failures were due to incorrect and incomplete specifications, or design errors. This also includes the requirements for the application program (software implementation). If the requirements are not clear and complete, personnel developing the application program will make assumptions and judgments. Unfortunately some may not align with those of the owner’s design team. This may result in systematic failures.
Some companies implement tasks during the analysis and design stages with names such as “IPL Select” or “LOPA Reconciliation”. The result of such studies is often a “refinement” of the control and/or safety system. Examples have ranged from identifying additional final elements to avoid the hazard, eliminating the use of shared instrumentation between protection layers, addressing response time issues, and assessing control system protection layers for full independence of a function against the initiating event and other protection layers. The benefit of such studies is that it’s easier and less expensive to make necessary changes to systems while the design is still on paper. It’s very expensive, and in some cases not even possible, to make design changes after systems have been installed.
In the end, it all boils down to people. It is imperative that all personnel be competent. New people entering the industry need an opportunity to learn. Yet they need training, mentoring and reviews of their work in order to prevent systematic failures from creeping in and causing accidents.
To read more examples of systematic failures throughout the lifecycle, and to learn how to reduce them, read the full paper “Methodologies in Reducing Systematic Failures of Wired IPLs” by Rick Hanner of aeSolutions and Tab Vestal of Eastman.
Process Safety & Risk Management Industrial Safety Instrumented Systems