Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Choi, B.; Nishida, Akemi; Li, Y.; Muramatsu, Ken*; Takada, Tsuyoshi*
Proceedings of 26th International Conference on Nuclear Engineering (ICONE-26) (Internet), 9 Pages, 2018/07
After the 2011 Fukushima accident, nuclear power plants are required to take countermeasures against accidents beyond design basis conditions. In seismic probabilistic risk assessment (SPRA), uncertainty can be classified as either aleatory uncertainty, which cannot be reduced, or epistemic uncertainty, which can be reduced with additional knowledge and/or information. To improve the reliability of SPRA, efforts should be made to identify and reduce the epistemic uncertainty caused by the lack of knowledge. In this study, we focused on the difference in seismic response by modeling methods, which is related epistemic uncertainty. We conducted a seismic response analysis with two kinds of modeling methods; a three-dimensional finite-element model and a conventional sway-rocking stick model, by using simulated various input ground motions, which is related to aleatory uncertainty. And then we quantified the seismic floor response results of the various input ground motions of each modeling methods. For the uncertainty quantification related to different modeling methods, we further perform a statistical analysis of the floor response results of the nuclear reactor building. Finally, we discussed how to utilize the results from these calculations for the quantification of uncertainty in fragility analysis for SPRA.
Choi, B.; Nishida, Akemi; Itoi, Tatsuya*; Takada, Tsuyoshi*; Furuya, Osamu*; Muta, Hitoshi*; Muramatsu, Ken
Proceedings of 13th Probabilistic Safety Assessment and Management Conference (PSAM-13) (USB Flash Drive), 8 Pages, 2016/10
In this study, we address epistemic uncertainty in structure fragility estimation of nuclear power plants (NPPs). In order to identify and quantify dominant factors in fragility assessment, sensitivity analyses of seismic analysis results are conducted for a target NPP building using a three-dimensional finite element model and a conventional lumped mass model (embedded sway rocking model), and the uncertainty caused by the major factors is then evaluated. The results are used to classify epistemic uncertainty levels in a fragility estimation workflow for NPPs in several stages, and a graded knowledge tree technique, which can be used for future fragility estimations, is proposed.
Osakabe, Kazuya*; Masaki, Koichi; Miyamoto, Yuhei*; Katsumata, Genshichiro*; Katsuyama, Jinya
no journal, ,
Probabilistic fracture mechanics (PFM) analysis is a useful methodology for quantitative evaluation because failure probabilities can be calculated considering uncertainties of material properties. The recent studies about the classification of uncertainties of inputs for PFM analyses are introduced. In this paper, we describe overseas study on the handling of epistemic uncertainty and aleatory uncertainty in PFM analysis considering reliability.