Development of probabilistic risk assessment methodology using artificial intelligence technology, 3; Automatic fault tree creation tools for failure mode level fault tree

Futagami, Satoshi; Kondo, Yuki; Yamano, Hidemasa; Kurisaka, Kenichi

Proceedings of Probabilistic Safety Assessment and Management & Asian Symposium on Risk Assessment and Management (PSAM17 & ASRAM2024) (Internet), 9 Pages, 2024/10

Integration of level 3 probabilistic risk assessment for nuclear power plants with transportation simulation considering earthquake hazards

Shimada, Kazumasa; Sakurahara, Tatsuya*; Farshadmanesh, P.*; Reihani, S.*; Mohagehgh, Z.*

Annals of Nuclear Energy, 197, p.110243_1 - 110243_12, 2024/03

https://doi.org/10.1016/j.anucene.2023.110243

This research improves the realism of Level 3 probabilistic risk assessment (PRA) for nuclear power plants (NPP) to avoid subjective expert judgment when setting evacuation behavior for residents. Therefore, the evacuation speed output by the traffic simulation code MATSim was input to the level 3 PRA code MACCS. Furthermore, to set the priority of the places where road closure is to be considered, a method to evaluate the road closure risk due to the earthquake using the natural disaster risk assessment code HAZUS was developed. Then, the relationship between the evacuation routes and the radiation dose was evaluated for the case study of the Sequoyah NPP adopted in the SOARCA study conducted by the US NRC. As a result, the present study found an evacuation route with low closure risk but causing high radiation dose of residents when it is closed. This showed effectiveness of the proposed Level 3 PRA methodology for supporting decision-makers to enhance evacuation routes.

Development of probabilistic risk assessment methodology using artificial intelligence technology, 1; Automatic fault tree creation

Futagami, Satoshi; Yamano, Hidemasa; Kurisaka, Kenichi; Ujita, Hiroshi*

Proceedings of PSAM 2023 Topical Conference AI & Risk Analysis for Probabilistic Safety/Security Assessment & Management, 8 Pages, 2023/10

To create an innovation for efficient and effective social implementation of nuclear power plant PRA, automatic construction tool for fault tree architecture and automatic failure judgment tool to construct reliability database are developed by using AI and digitization technology. This paper describes overall development plan of PRA methodology using the AI technology and the progress of automatic FT creation tools development.

Development of probabilistic analysis code for evaluating seismic fragility of aged pipes with wall-thinning

Yamaguchi, Yoshihito; Nishida, Akemi; Li, Y.

Proceedings of ASME 2022 Pressure Vessels and Piping Conference (PVP 2022) (Internet), 7 Pages, 2022/07

https://doi.org/10.1115/PVP2022-84009

The wall-thinning is one of the most important age-related degradation phenomena in nuclear piping systems. Furthermore, in recent years, several nuclear power plants in Japan have experienced severe earthquakes. Therefore, failure probability analysis and fragility evaluation of piping systems, taking both wall-thinning and seismic response stresses into consideration, have become increasingly important in seismic probabilistic risk assessment. In Japan Atomic Energy Agency, in order to evaluate the failure probability of aged piping system with wall-thinning, a probabilistic analysis code PASCAL-EC was developed. In this study, to evaluate the seismic fragility of a wall-thinned pipe, a model of seismic response stress considering the wall-thinning effect, a failure evaluation method for wall-thinned pipes, and functions related to uncertainties treatment for important influence parameters have been introduced to PASCAL-EC. In this paper, the improved PASCAL-EC is outlined and preliminary results of the seismic fragility evaluation performed using this code are provided.

Development of guideline on seismic fragility evaluation for aged piping

Yamaguchi, Yoshihito; Katsuyama, Jinya; Masaki, Koichi*; Li, Y.

Proceedings of ASME 2021 Pressure Vessels and Piping Conference (PVP 2021) (Internet), 9 Pages, 2021/07

The seismic probabilistic risk assessment is an important methodology to evaluate the seismic safety of nuclear power plants. In this assessment, the core damage frequency is evaluated from the seismic hazard, seismic fragilities, and accident sequence. Regarding the seismic fragility evaluation, the probabilistic fracture mechanics can be applied as a useful evaluation technique for aged piping systems with crack or wall thinning due to the age-related degradation mechanisms. In this study, to advance seismic probabilistic risk assessment methodology of nuclear power plants that have been in operation for a long time, a guideline on the seismic fragility evaluation of the typical aged piping systems of nuclear power plants has been developed considering the age-related degradation mechanisms. This paper provides an outline of the guideline and several examples of seismic fragility evaluation based on the guideline and utilizing the probabilistic fracture mechanics analysis code.

User's manual and analysis methodology of probabilistic fracture mechanics analysis code PASCAL-SP Ver. 2 for piping (Contract research)

Yamaguchi, Yoshihito; Mano, Akihiro; Katsuyama, Jinya; Masaki, Koichi*; Miyamoto, Yuhei*; Li, Y.

JAEA-Data/Code 2020-021, 176 Pages, 2021/02

JAEA-Data-Code-2020-021.pdf:5.26MB

In Japan Atomic Energy Agency, as a part of researches on the structural integrity assessment and seismic safety assessment of aged components in nuclear power plants, a probabilistic fracture mechanics (PFM) analysis code PASCAL-SP (PFM Analysis of Structural Components in Aging LWR - Stress Corrosion Cracking at Welded Joints of Piping) has been developed to evaluate failure probability of piping. The initial version was released in 2010, and after that, the evaluation targets have been expanded and analysis functions have been improved based on the state-of-the art technology. Now, it is released as Ver. 2.0. In the latest version, primary water stress corrosion cracking in the environment of Pressurized Water Reactor, nickel based alloy stress corrosion cracking in the environment of Boiling Water Reactor, and thermal embrittlement can be taken into account as target age-related degradation. Also, many analysis functions have been improved such as incorporations of the latest stress intensity factor solutions and uncertainty evaluation model of weld residual stress. Moreover, seismic fragility evaluation function has been developed by introducing evaluation methods including crack growth analysis model considering excessive cyclic loading due to large earthquake. Furthermore, confidence level evaluation function has been incorporated by considering the epistemic and aleatory uncertainties related to influence parameters in the probabilistic evaluation. This report provides the user's manual and analysis methodology of PASCAL-SP Ver. 2.0.

Improved Bayesian update method on flaw distributions reflecting non-destructive inspection result

Katsuyama, Jinya; Miyamoto, Yuhei*; Lu, K.; Mano, Akihiro; Li, Y.

Proceedings of ASME 2020 Pressure Vessels and Piping Conference (PVP 2020) (Internet), 8 Pages, 2020/08

https://doi.org/10.1115/PVP2020-21430

We have developed a probabilistic fracture mechanics (PFM) analysis code PASCAL4 for evaluating failure frequency of reactor pressure vessels (RPVs). It is known that flaw distributions have an important role in failure frequency calculation in PFM analysis. Previously, we proposed likelihood function to obtain more realistic flaw distributions applicable for both case when flaws are detected and when there is no flaw indication as the inspection results based on Bayesian update methodology. Here, it can be applied to independently obtain posterior distributions of flaw depth and density. In this study, we improve the likelihood function to enable them to update flaw depth and density simultaneously. Based on the improved likelihood function, an example is presented in which flaw distributions are estimated by reflecting NDI results through Bayesian update and PFM analysis. The results indicate that the improved likelihood functions are useful for estimating flaw distributions.

Improvement of probabilistic fracture mechanics analysis code PASCAL-SP with regard to PWSCC

Mano, Akihiro; Yamaguchi, Yoshihito; Katsuyama, Jinya; Li, Y.

Journal of Nuclear Engineering and Radiation Science, 5(3), p.031505_1 - 031505_8, 2019/07

https://doi.org/10.1115/1.4042115

Probabilistic fracture mechanics (PFM) analysis is expected as a rational method for the structural integrity assessment because it can consider the uncertainties of various influence factors and can evaluate the quantitative value such as failure probability of a cracked component as the solution. In the Japan Atomic Energy Agency, a PFM analysis code PASCAL-SP has been developed for the structural integrity assessment of piping welds in nuclear power plants. In the latest few decades, a number of cracks due to primary water stress corrosion cracking (PWSCC) have been detected in the nickel-based alloy welds in the primary piping of pressurized water reactors (PWRs). Thus the structural integrity assessment taking account of PWSCC has become important. In this paper, we improved PASCAL-SP for the assessment considering PWSCC by introducing the several analytical functions such as the evaluation models of crack initiation time, crack growth rate and probability of crack detection. By using improved PASCAL-SP, the failure probabilities of pipes with a circumferential crack or an axial crack due to PWSCC were evaluated as numerical examples. We also evaluated the influence of a leak detection and a non-destructive examination on the failure probabilities. On the basis of the numerical results, we concluded that the improved PASCAL-SP is useful for evaluating the failure probability of pipe taking PWSCC into account.

Development of a stochastic biokinetic method and its application to internal dose estimation for insoluble cesium-bearing particles

Manabe, Kentaro; Matsumoto, Masaki*

Journal of Nuclear Science and Technology, 56(1), p.78 - 86, 2019/01

https://doi.org/10.1080/00223131.2018.1523756

If an insoluble cesium-bearing particle is incorporated into the human body, the radioactivity will move as a single particle. In this case, it is impossible to estimate the number of disintegrations by considering the average behavior of countless nuclei. Then, a method was developed to simulate the behavior of the particle stochastically; and a biokinetic model was constructed to consider the characteristics of insoluble particles. Combination of the method and the model enables to estimate the number of disintegrations, and consequently the internal doses considering the stochastic behavior of the single cesium particle. We evaluated a probability density function of committed equivalent and effective doses and its 99th percentile value and arithmetic mean by repeating the above described procedure, and compared them to the reference values based on the existing models. As a result, the 99th percentile value of committed effective doses was 70 times the reference value when the number of incorporated particles was one, and consequently the dose level was quite low. When the exposure level is 1 mSv in committed effective dose, the uncertainty originating in the insolubility of cesium particles was negligibly small.

Verification methodology and results of probabilistic fracture mechanics code PASCAL

Masaki, Koichi; Miyamoto, Yuhei*; Osakabe, Kazuya*; Uno, Shumpei*; Katsuyama, Jinya; Li, Y.

Proceedings of 2017 ASME Pressure Vessels and Piping Conference (PVP 2017) (CD-ROM), 7 Pages, 2017/07

A probabilistic fracture mechanics (PFM) analysis code PASCAL has been developed by Japan Atomic Energy Agency (JAEA). PASCAL can evaluate failure frequencies of Japanese reactor pressure vessels (RPVs) during pressurized thermal shock (PTS) events based on domestic structural integrity assessment models and data of influence factors. In order to improve the engineering applicability of PFM to Japanese RPVs, we have performed verification of the PASCAL. In general, PFM code consists of many functions such as fracture mechanics evaluation functions, probabilistic evaluation functions including random variables sampling modules and probabilistic evaluation models, and so on. The verification of PFM code is basically difficult because it is impossible to confirm such functions through the comparison with experiments. When a PFM code is applied for evaluating failure frequencies of RPVs, verification methodology of the code should be clarified and it is important that verification results including the region and process of the verification of the code are indicated. In this paper, our activities of verification for PASCAL are presented. We firstly represent the overview and methodology of verification of PFM code, and then, some verification examples are provided. Through the verification activities, the applicability of PASCAL in structural integrity assessments for Japanese RPVs was confirmed with great confidence.