Abstract:

This paper presents a case study on approaches of addressing the error-forcing context (EFC) quantitatively for a misdiagnosis in a decision-making process guided by an emergency operating procedure (EOP). Two approaches are presented for the EFC-specific assessment of the human error probability (HEP). In the first approach this HEP is determined by expert judgement supported by a scale of both description of reference contexts with respect to the cognitive impact of information available for the operators and conditional HEPs for a subset of these contexts. Based on a description of the way the task is cognitively understood and processed by the human, the second approach derives the HEP from query to a database of human failures and successes observed from operating experience. The approaches are illustrated in a simple case study of an excerpt of a human reliability analysis (HRA) carried out for a loss of service water scenario postulated for the concept of a medium-sized gas-cooled reactor.

The objective is to estimate the probability of a specific misdiagnosis of the status of the heat sink of the emergency decay heat removal (EDHR) system. The analysis is supposed to account for the possibility of a partial failure of this heat sink as an EFC. Furthermore options are discussed to integrate the result of context-specific quantification in the process of assessment of the total HEP (i.e., over all contexts). The results show that a systematic compilation of cognitive demand contexts supports the assessment of a context-specific error probability. The discussion on total HEP modelling identified that the results rely on the adequacy of assumptions, concerning the degree of coverage of the HEP assigned to nominal scenario and the type of HEP distribution, which may deserve further investigation.