A framework analyzing system status and human activities: Illustrated using 2011 Fukushima nuclear power plant accident scenarios

摘要

Systematically modeling human activities and their interactions with complex systems is important in decision making processes. In order to determine appropriate human activities bringing systems to some desired status, integrated analyses combining both components are necessary. However, this can be difficult due to the complexity of the problems and the lack of adequate data. In this work, a framework capable of analyzing system status and human activities as an integrated set is proposed based upon use of a time-dependent Bayesian Network. The 2011 Fukushima Daiichi nuclear power plant accident scenarios are utilized in this work in order to illustrate the framework. Scenarios presented in this work are developed based upon the authors' interviews with Tokyo Electric Power Company (TEPCO) staff who were directly engaged in Fukushima accident mitigation. In this work, we focus upon problems that TEPCO staff found challenging during the Fukushima accident. Two types of tasks, mitigation activities during ongoing accidents and preparation tasks before potential accidents, are considered. The capability of the framework for providing system status, based upon rapidly changing human activities, can enable decision-makers to select appropriate mitigation strategies for use during emergencies. The capability of the framework in comparing multiple preparation efforts can enable decision-makers to identify appropriate strategies in order to prepare for similar future situations. Beyond the examples illustrated here, the proposed framework can be applied to broader fields and support corresponding decisions as long as corresponding human activity and system models are available.