Implications of Probabilistic Risk Assessment for Fusion Decision Making

Publisher Website
Google Scholar
Abstract
The potential value of probabilistic risk assessment (PRA) tools to fusion safety and economic issues is discussed. The main results and implications of a systematic examination of these general issues via PRA tools are reported. It is concluded that PRA methodology, tools, and thinking are useful to fusion research in the process of further improving fusion concepts and ideas. The MARS and STARFIRE designs are examined for possible answers to questions posed by using PRA tools. Several general magnetic-fusion design insights result from the study, including the following: 1. possible fault interactions must be minimized by decoupling fault conditions 2. the reliability of the vacuum boundary appears vital to maximizing facility availability and minimizing safety risk 3. economic analyses appear to be incomplete without consideration of potential availability loss from forced outages. A modification to PRA formalism called the “fault interaction matrix” is introduced. The fault interaction matrix contains information concerning what initial fault condition could lead to another fault condition, with what frequency. Thus, the fault interaction matrix represents a way to present and measure the degree to which a designer has decoupled possible fault conditions in his design. Such decoupling is crucial to enhancing fusion safety and facility availability.