Seismic PSA Research Articles

A hybrid approach of partially applying BDD for seismic PSA quantification

The binary decision diagram (BDD) method provides a means to calculate the exact probability of a fault tree, but cannot solve large fault trees for nuclear power plant PSAs. However, the BDD method can be used as a supplementary means to increase the accuracy of PSA quantification, especially for a seismic PSA which includes many non-rare events.Since it is difficult to solve a large PSA model using BDD method, several approaches have been developed that partially apply the BDD technique to small-sized branches of a large PSA model. This paper proposes an approach of partially applying BDD technique to large-sized branches. The results from the pilot application to a seismic PSA model shows that the partial BDD approach of this paper provides a relatively accurate way to quantify each sequence as well as the overall core damage in a seismic PSA model.

Overview of the seismic probabilistic safety assessment applied to a nuclear installation located in a low seismicity zone

Deterministic and probabilistic nuclear safety analysis methodologies have been developed and updated based on operational experience, investigation of past incidents or accidents, and analysis of postulated initiating events in order to maintain the protection of workers, the public and the environment. The evaluation of accident sequences and the total radiological risk resulting from off-site releases are general objectives addressed by these methodologies. There are hazards that continually challenge the safety of a nuclear facility or its nearby area. In particular, seismic events represent a major contributor to the risk of a nuclear facility. Different levels of ground motion induced by earthquakes may be experienced by the structures, systems and components (SSCs) of the installation. In this context, a seismic hazard analysis, seismic demand analysis and seismic fragility analysis must be carried out in order to characterize the local seismic hazard and what are the seismic demands on SSCs, allowing an adequate seismic classification of SSCs, even in installations located in sites with low seismicity. In this article, a general description of the Seismic Probabilistic Safety Assessment (Seismic PSA) methodology is presented, with emphasis on their support studies, aiming at applying the methodology described in this article to an experimental nuclear installation containing a PWR reactor designed for naval propulsion to be installed in a low seismicity zone in Brazil.

A new methodology for modeling explicit seismic common cause failures for seismic multi-unit probabilistic safety assessment

In a seismic PSA, dependency among seismic failures of components has not been explicitly modeled in the fault tree or event tree. This dependency is separately identified and assigned with numbers that range from zero to unity that reflect the level of the mutual correlation among seismic failures.Because of complexity and difficulty in calculating combination probabilities of correlated seismic failures in complex seismic event tree and fault tree, there has been a great need of development to explicitly model seismic correlation in terms of seismic common cause failures (CCFs). If seismic correlations are converted into seismic CCFs, it is possible to calculate an accurate value of a top event probability or frequency of a complex seismic fault tree by using the same procedure as for internal, fire, and flooding PSA.This study first proposes a methodology to explicitly model seismic dependency by converting correlated seismic failures into seismic CCFs. As a result, this methodology will allow systems analysts to quantify seismic risk as what they have done with the CCF method in internal, fire, and flooding PSA.

Development of efficient complete-sampling-based seismic PSA method for nuclear power plant

In this study, we propose an enhanced complete-sampling-based seismic PSA quantification technique that can precisely treat the partial correlations among components of seismic fragility information. Specifically, we have developed an improved seismic PSA quantification code by combining advantages of two representative methodologies: EPRI separation of variables approach-seismic fragility input based quantification method and the JAERI seismic fragility input based SECOM2-DQFM technique [31]. The most important feature of the proposed method is that it maps the EPRI seismic fragility input into the JAERI seismic fragility input space and then utilizes the improved DQFM algorithm as a seismic PSA quantification method. The improvement in the existing DQFM algorithm lies in extracting samples only from the median seismic intensity level (Am) and re-using such samples by scaling their values, instead of extracting samples of the response R and the capacity C regarding all seismic intensity levels. From the results of applying the proposed method to simple examples and an actual nuclear power plant, it is confirmed that the method proposed in this study derives results for system seismic fragility and seismic risk that are close to the exact solutions using fewer samples than the existing DQFM does.

Development of a Seismic Probabilistic Safety Assessment Quantification Tool

Shanghai Nuclear Engineering Research and Design Institute (SNERDI) has been studying seismic risk analysis for nuclear power plant for a long time, and completed seismic margin analysis for several plants. After Fukushima accident, seismic risk has drawn an increasing attention worldwide, and the regulatory body in China has also required the utilities to conduct a detailed analysis for seismic risk. So, we turned our focus on a more intensive study of seismic probabilistic safety assessment (PSA/PRA) for nuclear power plant in recent years. Since quantification of seismic risk is a key part in Seismic PSA, lots of efforts have been devoted to its research by SNERDI. The quantification tool is the main product of this research, and will be discussed in detail in this paper. First, a brief introduction to Seismic PSA quantification methodology is presented in this paper, including fragility analysis on system or plant level, convolution of seismic hazard curves and fragility curves, and uncertainty analysis as well. To derive more accurate quantification results, the binary decision diagram (BDD) algorithm was introduced into the quantification process, which effectively reduces the deficiency of the conventional method on coping with large probability events and negated logic. Finally, this paper introduced the development of the seismic PSA quantification tool based on the algorithms discussed in this paper. Tests and application have been made for this software based on a specific nuclear power plant seismic PSA model.

Analysis of common cause failure effect on system reliability in seismic PSA

In seismic PSA of nuclear power plant, the equipment failure probability is affected by two parts – response and fragility. In this paper causations of CCF in the above two parts is analyzed, the sensitivity of CCF factor is evaluated under different peak ground acceleration levels. When the peak ground acceleration is much lower than the component seismic resistance capacity, difference between the results of CCF factor β = 0.05 and β = 1 is about one magnitude, while the results of β = 0.05 and β = 1 are very similar when the peak ground acceleration is close to or even higher than the component seismic resistance capacity. Then a model for system reliability analysis based on Monte Carlo (MC) simulation is proposed: the effect of structure's response can be analyzed in the system reliability model, and the CCF methods used in internal event PSA are also suitable for CCF in response and fragility of the equipment in seismic PSA. The reliability of accumulator system in AP1000 is analyzed as an example under different seismic hazard curves, the system failure probability difference between uncertainties of structure's response βS = 0.1 g and βS = 0.4 g is about one magnitude, and the results of CCF factor of β = 0.05 and β = 1 are very similar since the component failure probability is close to 1 when peak ground acceleration is high.

DEVELOPMENT OF AN INTEGRATED RISK ASSESSMENT FRAMEWORK FOR INTERNAL/EXTERNAL EVENTS AND ALL POWER MODES

From the PSA point of view, the Fukushima accident of Japan in 2011 reveals some issues to be re-considered and/or improved in the PSA such as the limited scope of the PSA, site risk, etc. KAERI (Korea Atomic Energy Research Institute) has performed researches on the development of an integrated risk assessment framework related to some issues arisen after the Fukushima accident. This framework can cover the internal PSA model and external PSA models (fire, flooding, and seismic PSA models) in the full power and the low power-shutdown modes. This framework also integrates level 1, 2 and 3 PSA to quantify the risk of nuclear facilities more efficiently and consistently. We expect that this framework will be helpful to resolve the issue regarding the limited scope of PSA and to reduce some inconsistencies that might exist between (1) the internal and external PSA, and (2) full power mode PSA and low power-shutdown PSA models. In addition, KAERI is starting researches related to the extreme external events, the risk assessment of spent fuel pool, and the site risk. These emerging issues will be incorporated into the integrated risk assessment framework. In this paper the integrated risk assessment framework and the research activities on the emerging issues are outlined.

A shaking table test for an evaluation of seismic behavior of 480V MCC

Abstract In this study, a seismic behavior of the electrical cabinet system in Nuclear Power Plants (NPPs) was evaluated by the shaking table test. A 480 V Motor Control Center (MCC) was selected for the shaking table test, and a real MCC cabinet for the Korea Nuclear Power Plant site was rented from a manufacturing company. For the shaking table tests, three kinds of seismic input motions were used, which were a US NRC Reg. guide 1.60 design spectrum, a UHS spectrum and Floor Response Spectrum of PAB 165′ at the Ulchin NPP site. The UHS input motion was particularly selected for an evaluation of high frequency seismic effects. The PGA levels for the shaking table test were scheduled from 0.2 g to 3.0 g in order to assess the real seismic capacity of the electrical cabinet system. For the evaluation of the structural seismic amplification effects, three directional accelerations were measured at three points outside of the cabinet system, and regarding incabinet response amplification, accelerations were measured at two points where electrical equipment such as relay were mounted. The seismic amplification effect is determined on the outside and the inside of a cabinet as input seismic motion, and compared to the results which were calculated by analytical method based on NUREG/CR-5203. Finally, with the seismic fragility evaluation for the 480 V MCC cabinet, the previous seismic result from a seismic PSA report in Korea is found to be slightly conservative.

Options for the treatment of uncertainty in seismic probabilistic safety assessment of nuclear power plants

One of the most complex cases for assessing the nuclear power plant safety is the evaluation of the response of the plant to an earthquake and calculation of the core damage frequency related with this. Plant level fragilities are convolved with the seismic hazard curves to obtain a set of doublets for the plant damage state. The standard methodology of the description of randomness and epistemic uncertainty of the fragility is based on the use of lognormal distribution. In the practice, because of large number and variety of types of components, variety of failure modes, further simplification is needed in spite of simplicity of the mathematic description of the fragility and its uncertainty. Sophisticated modeling and screening methods have to be applied for plant fragility development requiring enormous experience. Several practical assumptions utilized in the seismic PSA showing certain analogy with interval type description of uncertainties. In the paper an attempt is made for outlining some new options for nuclear power plant seismic fragility development based on the interval and p-box concept. The possibility for derivation of conditional probability of failure for cumulative absolute velocity is also highlighted.

CURRENT STATUS AND IMPORTANT ISSUES ON SEISMIC HAZARD EVALUATION METHODOLOGY IN JAPAN

The outlines of seismic PSA implementation standards and seismic hazard evaluation procedure were shown. An overview of the cause investigation of seismic motion amplification on the Niigata-ken Chuetsu-oki (NCO) earthquake was also shown. Then, the contents for improving the seismic hazard evaluation methodology based on the lessons learned from the NCO earthquake were described. (1) It is very important to recognize the effectiveness of a fault model on the detail seismic hazard evaluation for the near seismic source through the cause investigation of the NCO earthquake. (2) In order to perform and proceed with a seismic hazard evaluation, the Japan Nuclear Energy Safety Organization has proposed the framework of the open deliberation rule regarding the treatment of uncertainty which was made so as to be able to utilize a logic tree. (3) The b-value evaluation on the "Stress concentrating zone," which is a high seismic activity around the NCO hypocenter area, should be modified based on the Gutenberg-Richter equation.

Seismic PSA method for multiple nuclear power plants in a site

The maximum number of nuclear power plants in a site is eight and about 50% of power plants are built in sites with three or more plants in the world. Such nuclear sites have potential risks of simultaneous multiple plant damages especially at external events. Seismic probabilistic safety assessment method (Level-1 PSA) for multi-unit sites with up to 9 units has been developed. The models include Fault-tree linked Monte Carlo computation, taking into consideration multivariate correlations of components and systems from partial to complete, inside and across units. The models were programmed as a computer program CORAL reef. Sample analysis and sensitivity studies were performed to verify the models and algorithms and to understand some of risk insights and risk metrics, such as site core damage frequency (CDF per site-year) for multiple reactor plants. This study will contribute to realistic state of art seismic PSA, taking consideration of multiple reactor power plants, and to enhancement of seismic safety.

Development of Evaluation Method to Determine Fragility of Power Plant Equipment Based on Bayesian Analysis

In the seismic design of nuclear power plant, it is recently considered to apply the probabilistic method in addition to the deterministic method. The deterministic method secures the seismic safety of a nuclear power plant by veryfying that the calculated stress on the equipment is lower than allowable stress in the design criteria. The seismic probability safety assessment (Seismic PSA) method secures seismic safety by veryfying that the probability of reactor core failure is lower than a certain value. In seismic PSA, it is important to estimate the fragility of the equipment for precise assessment. Therefore, in this paper we dealt with a method using the Bayesian approach to generate a fragility curve of power plant from a multi-specimen test, and investigated the influence of the sample size driving the fragility evaluation test

Development of Synchronized Control Method for Shaking Table with Booster Device (Verification of the Capabilities based on Both Real Facility and Numerical Simulator)

In the seismic design of nuclear power plant, it is recently considered to use probability method in addition to certainty method. The former method is called Seismic Probability Safety Assessment (Seismic PSA). In case of seismic PSA for some components of a nuclear power plant using a shaking table, it is necessary for some limited conditions with high level of accelerations such as actual conditions. However, it might be difficult to achieve the test conditions that a current shaking table based on hydraulic power system is intended for the test facility. Therefore, we have been planning out a test method in which both a current and another shaking table called a booster device are applied. This paper describes the verification test of a synchronized control between a current shaking table and a booster device.

Development of the DQFM method to consider the effect of correlation of component failures in seismic PSA of nuclear power plant : Watanabe, Y. et al. Reliability Engineering & System Safety, 2003, 79, (3), 265–279

Learning a classifier from an imbalanced dataset is an important problem in data mining and machine learning. Since there is more information from the majority classes than the minorities in an imbalanced dataset, the classifier would become over-fitted to the former and under-fitted to the latter classes. Previous attempts to address the problem have been focusing on increasing the learning sensitivity to the minorities and/or rebalancing sample sizes among classes before learning. However, how to efficiently identify their optimal mix in rebalancing is still an unresolved problem. Due to non-linear relationships between attributes and class labels, merely to rebalance sample sizes rarely comes up with optimal results. Moreover, brute-force search for the perfect combination is known to be NP-hard and hence a smarter heuristic is required. In this paper, we propose a notion of swarm fusion to address the problem – using stochastic swarm heuristics to cooperatively optimize the mixtures. Comparing with conventional rebalancing methods, e.g., linear search, our novel fusion approach is able to find a close to optimal mix with improved accuracy and reliability. Most importantly, it has found to be with higher computational speed than other coupled swarm optimization techniques and iteration methods. In our experiments, we first compared our proposed solution with traditional methods on thirty publicly available imbalanced datasets. Using neural network as base learner, our proposed method is found to outperform other traditional methods by up to 69% in terms of the credibility of the learned classifiers. Secondly, we wrapped our proposed swarm fusion method with decision tree. Notably, it defeated six state-of-the-art methods on ten imbalanced datasets in all evolution metrics that we considered.

235 発電機器のベイズ解析に基づいた損傷確率評価手法の開発

In the seismic design of nuclear power plant, it is recently considered to use probability method in additon to certainty method. The latter method is to secure seismic safety of nuclear power plant by corroborating that calculated stress of equipment is lower than allowable stress of design criteria. And the former method, which is called Seismic Probability Safety Assessment (Seismic PSA) , is to secure seismic safety by corroborating that probability of reactor core failure is lower than a certain value. In Seismic PSA, it is important to estimate fragility of equipment for precise assessment; therefore, in this paper we dealt with a method using Bayesian Approach for generating a fragility curve and a design fragility curve from a multispecimen test.

358. 地震PSA用のヒューマンエラーのモデル化の試み

358. 地震PSA用のヒューマンエラーのモデル化の試み

Study of rationalized safety design based on the seismic PSA for an LMFBR

Seismic PSA was carried out for a typical liquid metal cooled fast breeder reactor (LMFBR) in order to study the rationalized seismic design, maintaining and/or improving safety during seismic event. The seismic sequence quantification identifies the dominant structures, systems and components (SSCs) to the seismic core damage frequency (CDF). The sensitivity analyses by reducing or increasing the seismic capacity for SSCs are used to examine the optimized seismic design in view of safety and economical aspects. The LMFBR-specific risk-significant SSCs are reactor coolant boundary, decay heat removal coolant path and reactor control rod, which are different from those of light water reactors (LWRs). The electrical power supply system has a minor contribution to the seismic CDF. The sensitivity study shows that passive safety features of LMFBRs are important to maintain and/or enhance seismic capacity. The passive safety includes the decay heat removal capability via natural circulation and safety measures without depending on the support systems such as alternating current (AC) electrical power, for example. On the course of seismic sequence quantification, a methodology to evaluate the probability of seismic-induced multiple failure has been developed and applied to the decay heat removal function. The results suggest the multiplicity of the triply redundant system is to be considered for the significant components such as the decay heat removal path when one considers the difference in the seismic response.

A proposal for safety factor estimation in seismic PSA of structure by stochastic FEM technique

Abstract The study of safety factor estimation in seismic probabilistic safety assessment has been a focus of researchers' interest for some time. Most of the papers dealt with the safety factor estimation by means of experimental knowledge or numerical analysis such as the Monte Carlo technique. This paper presents a new method to gain parts of structural response factors analytically by utilizing the statistical theory of extremes and the stochastic FEM technique. First, the stochastic FEM is adopted to obtain the statistical response values considering the uncertainty of structural material, through the spectral analysis due to an earthquake as a stationary random process. As second step, combining the results by statistical theory of extremes for a structure having deterministic material properties with the previous results by spectral analysis considering material uncertainty, with the aid of the total representation theorem against uncertainty modeling, statistics related to maximum values of response could be calculated under the consideration of structural material uncertainty. Those results are shown in good agreement with Monte Carlo simulation. Finally, using the obtained results, it is shown how to get the structural safety factors related to structural material uncertainty analytically.

原子力発電所に対する地震ＰＳＡ研究の動向

原子力発電所に対する地震ＰＳＡ研究の動向