Development of Probabilistic Safety Assessment with respect to the first demonstration nuclear power plant of high temperature gas cooled reactor in China

Due to the unique concept of HTR-PM (High Temperature Gas Cooled Reactor-Pebble Bed Module) design, Chinese nuclear authority has anticipated that HTR-PM will bring challenge to the present regulation. The pilot use of PSA (Probabilistic Safety Assessment) during HTR-PM design and safety review is deemed to be the necessary and efficient tool to tackle the problem, and is actively encouraged as indicated in the authority's specific policy statement on HTR-PM project. The paper summarizes the policy statement to set up the base of PSA development and application activities. The up-to-date status of HTR-PM PSA development and the risk-informed application activities are introduced in this paper as the follow-up response to the policy statement. For open discussion, the paper hereafter puts forward several technical issues which have been encountered during HTR-PM PSA development. Since HTR-PM PSA development experience has the general conclusion that many of the PSA elements can be and have been implemented successfully by the traditional PSA techniques, only the issues which extra innovative efforts may be needed are highlighted in this paper. They are safety goal and risk metrics, PSA modeling framework for the non-water reactors, passive system reliability evaluation, initiating events frequencies and component reliability data estimation techniques for the new reactors and so on. The paper presents the way in which the encountered technical issues were or will be solved, although the proposed way may not be the ultimate best solution. The paper intends to express the standpoint that although the PSA of new reactor has the inherent weakness due to the insufficient information and larger data uncertainty, the problem of component reliability data is much less severe than people have conceived. The unique design conception and functional features of the reactors can influence the results more significantly than the component reliability data. What we are benefited from PSA is indeed the systematic way which PSA follows. This is more important especially for the new reactors. (C) 2011 Elsevier B.V. All rights reserved.