A Simulator-Based Study of Human Errors in Nuclear Power Plant Control Room Tasks

Abstract

The purpose of this study was to empirically establish error rates for control selection and operation during proceduralized tasks in nuclear power plant control rooms. Data were collected in high fidelity simulators from licensed operators and students in the final stages of training prior to licensing by a Performance Measurement System (PMS) which recorded control manipulations and plant parameters. The PMS data were compared to a list of Procedural Task Elements (PTE) derived from the procedures used in responding to the casualties. Any action called for by the PTE list and not taken was considered an error of omission. A subject matter expert reviewed the PMS record of actions for errors in selection or operation of controls. The observed error rates (OERs) for commission errors were in close agreement with the human error probabilities (HEPs) given in NUREG/CR-1278. The omission OER was an order of magnitude higher than that given in NUREG/CR-1278. However, this number may be an artifact of the way simulator training is conducted, as the operators did not have the time to recover omitted actions that would be available in a real plant.