Development of a Dynamic Simulation Approach to Mission Risk and Reliability Analysis

This paper describes a NASA-funded project to develop reliability analysis software capable of modeling complex, highly dynamic systems over the duration of a mission, taking into account variation in input parameters and the evolution of the system. It is designed as an extension to GoldSim, a simulation program which is widely used for Performance Assessment in the nuclear arena, most notably at Yucca Mountain. To illustrate the GoldSim approach to reliability modeling, two NASA examples that have previously been evaluated using classical PRA models were developed using the simulation approach. Issues surrounding the translation of the classical PRA models to a simulation-based approach are discussed, and areas where the simulation approach provided additional insights into the system’s behavior are highlighted.