Extraction and classification of transportation incidents potentially caused by natural events emerged from the Fukushima NPP accident

In view of the state of the Fukushima NPP accident, CSS of IAEA has decided on an action plan to review the safety regulations for nuclear facilities. The transport regulations and guides are also included in the plan and Japan, as the country directly concerned, has to be engaged in the task by TRANSSC through a gap-analysis at a specialists’ meeting. In arguments of that, it was pointed out that impacts given by natural events should be re-examined and boundaries of test condition between routine, normal and accident should be clarified. These issues were discussed in the technical meeting (TM-44891) held on July 2013. This study provides a framework for identifying and classifying transport incidents potentially caused by natural events such as extreme earthquake and tsunami, in order to separate significant ones from the others using risk concept. The hazard matrix is proposed for finding potential incidents efficiently and comprehensively. Each incident is roughly rated by two risk indices representing an occurrence frequency and severity of package influence. The risk matrix is then used to classify them by certain risk criteria and clearly show a group of incidents that may have significant risk. In fact, these are a kind of risk-finding and screening process in probabilistic risk assessment. As a case example showing the applicability of our techniques, practical efforts were made among local experts to find incidents in domestic transport of spent nuclear fuels. As a result, a total of over 300 incidents were identified in terms of 22 types of natural events, four types of accident loads (impact, fire, immersion and burial) and four transport modes (land, sea, cargo and temporal storage). After classification, only 9 incidents were left on a cell group of possibly exceeding accident conditions. Note that the risk is rated at the maximum imaginable if uncertain. Further site survey will be required to know true frequency and severity of such incidents in quantitative manner.