(78l) Looking across Industries to Improve Human Reliability Data for QRAs

While meticulous work is often done to determine failure probabilities of SIS for QRAs, the work done to determine failure probabilities of humans performing specific tasks is often glossed over. Yet, human reliability data is just as important to the quality and accuracy of the QRA as non-human failure probabilities. The problem lies in the lack of availability of statistically reliable human failure probabilities for various tasks within the process industries. This paper will present and analyze the strengths and weaknesses of a variety of sources of such data. The sources of human reliability data that will be explored come from the nuclear industry, the aviation industry, the medical field, and of course the process industries. More specifically, the most emphasis will be placed on discussion of the NRC's HERA program and the possible applicability and worth of the program to the process industries. This presentation is based on work done for CCPS' Human Reliability Project.