Simulation-based dynamic probabilistic risk assessment of an internal flooding-initiated accident in nuclear power plant using THALES2 and RAPID

Kubo, Kotaro; Zheng, X.; Tanaka, Yoichi; Tamaki, Hitoshi; Sugiyama, Tomoyuki; Jang, S.*; Takata, Takashi*; Yamaguchi, Akira*

Proceedings of the Institution of Mechanical Engineers, Part O; Journal of Risk and Reliability, 237(5), p.947 - 957, 2023/10

https://doi.org/10.1177/1748006X221091604

Probabilistic risk assessment (PRA) is a method used to assess the risks associated with large and complex systems. However, the timing at which nuclear power plant structures, systems, and components are damaged is difficult to estimate if the risk of an external event is evaluated using conventional PRA based on event trees and fault trees. A methodology coupling thermal-hydraulic analysis with external event simulations using Risk Assessment with Plant Interactive Dynamics (RAPID) is therefore proposed to overcome this limitation. A flood propagation model based on Bernoulli's theorem was applied to represent internal flooding in the turbine building of the pressurized water reactor. Uncertainties were also taken into account, including the flow rate of the floodwater source and the failure criteria for the mitigation systems. The simulated recovery actions included the operator isolating the floodwater source and using a drainage pump; these actions were modeled using several simplifications. Overall, the results indicate that combining isolation and drainage can reduce the conditional core damage probability upon the occurrence of flooding by approximately 90%.

Dynamic probabilistic risk assessment of seismic-induced flooding in pressurized water reactor by seismic, flooding, and thermal-hydraulics simulations

Kubo, Kotaro; Jang, S.*; Takata, Takashi*; Yamaguchi, Akira*

Journal of Nuclear Science and Technology, 60(4), p.359 - 373, 2023/04

https://doi.org/10.1080/00223131.2022.2100837

Probabilistic risk assessment (PRA) is an essential approach to improving the safety of nuclear power plants. However, this method includes certain difficulties, such as modeling of combinations of multiple hazards. Seismic-induced flooding scenario includes several core damage sequences, i.e., core damage caused by earthquake, flooding, and combination of earthquake and flooding. The flooding fragility is time-dependent as the flooding water propagates from the water source such as a tank to compartments. Therefore, dynamic PRA should be used to perform a realistic risk analysis and quantification. This study analyzed the risk of seismic-induced flooding events by coupling seismic, flooding, and thermal-hydraulics simulations, considering the dependency between multiple hazards explicitly. For requirements of safety improvement, especially in light of the Fukushima Daiichi Nuclear Power Plant accident, sensitivity analysis was performed on the seismic capacity of systems, and the effectiveness of alternative steam generator injection by a portable pump was estimated. We demonstrate the use of this simulation-based dynamic PRA methodology to evaluate the risk induced by a combination of hazards.

Uncertainty analysis of dynamic PRA using nested Monte Carlo simulations and multi-fidelity models

Zheng, X.; Tamaki, Hitoshi; Takahara, Shogo; Sugiyama, Tomoyuki; Maruyama, Yu

Proceedings of Probabilistic Safety Assessment and Management (PSAM16) (Internet), 10 Pages, 2022/09

https://www.iapsam.org/PSAM16/

Quasi-Monte Carlo sampling method for simulation-based dynamic probabilistic risk assessment of nuclear power plants

Kubo, Kotaro; Jang, S.*; Takata, Takashi*; Yamaguchi, Akira*

Journal of Nuclear Science and Technology, 59(3), p.357 - 367, 2022/03

https://doi.org/10.1080/00223131.2021.1971119

Dynamic probabilistic risk assessment (PRA), which handles epistemic and aleatory uncertainties by coupling the thermal-hydraulics simulation and probabilistic sampling, enables a more realistic and detailed analysis than conventional PRA. However, enormous calculation costs are incurred by these improvements. One solution is to select an appropriate sampling method. In this paper, we applied the Monte Carlo, Latin hypercube, grid-point, and quasi-Monte Carlo sampling methods to the dynamic PRA of a station blackout sequence in a boiling water reactor and compared each method. The result indicated that quasi-Monte Carlo sampling method handles the uncertainties most effectively in the assumed scenario.

Evaluation of risk dilution effects in dynamic probabilistic risk assessment of nuclear power plants

Kubo, Kotaro; Tanaka, Yoichi

Proceedings of 31st European Safety and Reliability Conference (ESREL 2021) (Internet), p.810 - 817, 2021/09

https://doi.org/10.3850/978-981-18-2016-8_477-cd

Probabilistic risk assessment (PRA) is a method of effectively evaluating risks in nuclear power plants and is used in various agencies. Dynamic PRA is attracting considerable attention, as it enables realistic assessment by reducing the assumptions and engineering judgments related to time-dependent failure probability and/or human action reliability. However, it is difficult to remove all assumptions and engineering judgments. Therefore, their effects on assessment results should be understood. This study focuses on the "risk dilution effect," which arises from assumptions about uncertainty. Results showed that this effect causes a difference of about 10% to 20% in the relative change of the conditional core damage probability in the station blackout scenario. This effect should be fully considered when using dynamic PRA in critical decision-making, such as that on regulations.

A Comparative study of sampling techniques for dynamic probabilistic risk assessment of nuclear power plants

Kubo, Kotaro; Zheng, X.; Tanaka, Yoichi; Tamaki, Hitoshi; Sugiyama, Tomoyuki; Jang, S.*; Takata, Takashi*; Yamaguchi, Akira*

Proceedings of Joint International Conference on Supercomputing in Nuclear Applications + Monte Carlo 2020 (SNA + MC 2020), p.308 - 315, 2020/10

Dynamic probabilistic risk assessment (PRA) is a method for improving the realism and completeness of conventional PRA. However, enormous calculation costs are incurred by these improvements. One solution is to select an appropriate sampling method. In this paper, we applied the Monte Carlo, Latin hypercube, grid-point, and quasi-Monte Carlo sampling methods to the dynamic PRA of a simplified accident sequence and compared the results for each method. Quasi-Monte Carlo sampling was found to be the most effective method in this case.