Abstract
Source term estimation is one of the key steps for nuclear emergency decision-making, and it is usually of the most significant uncertainty and difficulty than the other steps. As revealed by Fukushima nuclear accident, we are so difficult to know what had exactly happened to the reactors due to the widespread unavailability of instrumental indications and power supply. As a matter of course, this dilemma leads to the hard decision-making about timely actions of emergency response based on the unconfident understanding of the reactors’ status and the radioactive releases to the environment. However, we’d like to insist that we unfortunately are not aware of the effective utilization of some existing valuable information sources such as Probability Risk Assessment (PRA) during the nuclear emergency response decision making. Hence, in this paper, we propose a new approach to estimate the emergency source term based on PRA knowledge and insights, including logic models and quantitative information (reliability data, release category and so on). The approach intends to intelligently diagnose the status of reactors and to give out the potential source terms in a fast and real-time manner, so that emergency response decision making can be significantly improved.
