(43ak) Guidelines for Developing Site-Specific Human Error Rates and Human IPLs for LOPA Using a Safety Climate Approach

Layer of Protection Analysis (LOPA) is a simple order of magnitude quantitative risk analysis technique commonly applied in the process industries.  This powerful tool is especially useful for refining specific hazard scenarios during a process hazard analysis or a risk analysis.  In LOPA, a hazard scenario is decomposed into a series of events, which may include the frequency or probability of equipment failures, personnel actions, and physical processes.  By one estimate, 80 percent of incidents involve human error; thus, most risk analyses consider the likelihood and effect of human factors on potential scenarios.  For LOPA, human error includes events initiated by the human in the work system and events where the human element failed to act as an independent layer of protection against an adverse event.  

Although the LOPA method is widely used, there are few guidance sources for refining the human error component.  The currently available standard LOPA estimates for the occurrence of human error assume optimal human factors conditions and that work systems are static environments unchanging over time. Work systems are, in fact, dynamic environments, which evolve over time through planned changes, as well as unplanned adaptations.  Thus, no work system is completely free of error at any given time.  Current LOPA estimates, therefore, fail to account for the dynamic nature of work systems where change from original conditions is not only possible but even necessary from time to time.  This paper will examine the deficiencies in the LOPA human error probability estimates and propose a strategy that utilizes a safety climate approach to understanding error causation and incorporating it as a fundamental building block to human error estimation.