Sensitivity analysis of a passive decay heat removal system under a post-loss of coolant accident condition

Liu, Q.; Homma, Toshimitsu

Journal of Nuclear Science and Technology, 49(9), p.897 - 909, 2012/09

https://doi.org/10.1080/00223131.2012.712481

Passive safety features are now of interest to the design of future generation reactors. Though passive safety systems are considered to be more reliable, the large uncertainty associated with the system response can not be ignored. It is necessary to identify the uncertain inputs that have the important impact on the uncertainty of the system performance. In this study, two global sensitivity measures, the first-order sensitivity index and the total-order sensitivity index, are applied to a passive decay heat removal system of a gas-cooled fast reactor for identifying the key uncertainty inputs. It is found that the uncertainty in the system pressure contributes the most to the uncertainty in the system outputs. In addition, the cooler wall temperature, the Nusselt number and the friction factor in the mixed convectional flow regime also have small impact on the uncertainty of the system outputs.

A Simple method for estimating the structure temperatures and the cesium revaporization inside the reactor pressure vessel, 2; Feasibility study for the Fukushima Daiichi Nuclear Power Plant

Liu, Q.; Ishikawa, Jun; Maruyama, Yu; Watanabe, Norio

Journal of Nuclear Science and Technology, 49(5), p.486 - 495, 2012/05

https://doi.org/10.1080/00223131.2012.677130

This work examines the feasibility of the method proposed in the preceding part of a two-part paper. The base case study shows that water injection via the core spray line is more effective to cool the uncovered core and to reduce the release of CsOH. Sensitivity study is conducted by introducing the dimensionless decay heat, which combines the effects of the ratio of the flooded core, the ratio of the slumped core and the leak ratio of the injected water on the steam generation rate associated with the forced convection cooling in the reactor pressure vessel (RPV). The results show that the temperatures of the uncovered core and the other structures increase with the dimensionless decay heat. So does the release rate of CsOH. The relationships of the measurable RPV wall temperature with the temperatures of the uncovered core and the other structures as well as the release characteristics of CsOH are also examined in this work.

GSALab computer code for global sensitivity analysis

Liu, Q.; Homma, Toshimitsu; Nishimaki, Yuichiro*; Hayashi, Hiroko*; Terakado, Masato*; Tamura, Satoshi*

JAEA-Data/Code 2010-001, 57 Pages, 2010/03

JAEA-Data-Code-2010-001.pdf:16.81MB

For a risk assessment model of an engineering system, the uncertainties in the model inputs propagate through the model and lead to the uncertainty in the model output. In order to evaluate the model output uncertainty and the contribution of each model input to the output uncertainty, the computer code GSALab, which is based on Monte Carlo simulations, has been developed. It is composed of three parts, namely, random samples generation, uncertainty analysis and sensitivity analysis. In the part of sensitivity analysis, several global sensitivity indicators, including the popularly used variance-based indicators, are implemented. In addition, the GUI (Graphical User Interface) of GSALab has been developed for the user's convenience. In addition to risk assessment models, it is also possible to use GSALab for uncertainty and sensitivity analysis of a wide class of mathematical models.

A New computational method of a moment-independent uncertainty importance measure

Liu, Q.; Homma, Toshimitsu

Reliability Engineering & System Safety, 94(7), p.1205 - 1211, 2009/07

https://doi.org/10.1016/j.ress.2008.10.005

The uncertainty of input parameters are transferred through the model to the output and leads to the issue of output uncertainty. The study of how the uncertainty in the model output can be apportioned to the uncertainty in the inputs is the job of sensitivity analysis. In this paper, the authors analyzed a newly proposed uncertainty importance measure, Delta. The core calculation method of Delta is based on the measurement of the area between the conditional PDF (Probability Density Functions) and the unconditional PDF of the model output. The authors found that the area between these two PDFs is equivalent to two times of the algebraic sum of the vertical deviations between their corresponding CDFs at the intersection points of these two PDFs. Therefore, the author proposed a new method to calculate Delta. The applicability of this new method is proved by applying it to two test models.

User's manual of SECOM2-DQFM; A Computer code for seismic system reliability analysis

Liu, Q.; Muramatsu, Ken; Uchiyama, Tomoaki*

JAEA-Data/Code 2008-004, 70 Pages, 2008/03

JAEA-Data-Code-2008-004.pdf:4.54MB

SECOM2-DQFM is developed for seismic reliability analysis of complex engineering systems, such as nuclear power plants. Suppose that the seismic hazard curve of the location site of a plant, the fault tree model and the event tree model of this plant are known. If the capacities and responses of components are available, the conditional occurrence probability (and frequency) of the top event of the fault tree model can be estimated by using SECOM2-DQFM. In addition, the importance of each basic event as well as the occurrence probability (and frequency) of each accident sequence can also be calculated. One feature of SECOM2-DQFM is that the method of Direct Quantification of Fault Tree using Monte Carlo simulation (DQFM) is adopted to evaluate the concurrent failure probability of multiple components. This report is summarized as the user manual of SECOM2-DQFM.

Use of uncertainty importance measures to complement risk importance measures in PSA

Liu, Q.; Homma, Toshimitsu

Proceedings of 9th International Probabilistic Safety Assessment and Management Conference (PSAM-9) (CD-ROM), 7 Pages, 2008/00

Fussel-Vesely (FV) and Risk Achievement Worth (RAW) are two commonly used measures in importance ranking of basic events in PSA. Both measures are based on point-estimates of the risk. However, realistic failure characteristics of components are associated with some kinds of uncertainty. The uncertainties of component failures are propagated through the model and bring about the uncertainty of the model risk. Therefore, it is necessary to take uncertainty into consideration when the contribution of a basic event to risk is estimated. By using two fault tree models as examples, the authors calculated the FV and the RAW as well as the uncertainty importance of each basic event. The results show that the uncertainty importance ranking of each basic event does not always agree with the ranking with regard to FV (or RAW). It is argued that uncertainty importance measures provides complementary perspectives of the roles of a basic event in determining the risk.

The Indication of a moment independent measure and its new calculation method

Liu, Q.; Homma, Toshimitsu

no journal, , 

In risk assessment problems, the variations of input parameters are transferred through the model to the output, and bring about the issue of output uncertainty. The study of how the uncertainty in the model output can be apportioned to the variations of input parameters is the job of sensitivity analysis. Borgonovo argued that an ideal measure should be global, model-free and moment-independent. He proposed a new uncertainty importance measure, which is called Delta. The calculation method of Delta is based on measuring the surrounded area between two PDFs (Probability Density Functions) of the model output. By analyzing this calculation method, the authors found that the area surrounded by two PDFs is equivalent to two times of the algebraic sum of the vertical deviations between their corresponding CDFs at the intersection points of these two PDFs. Therefore, the author proposed a new method to calculate Delta. An analysis shows that this measure will improve the calculation accuracy of Delta.

Effect of correlations of component failures and cross-connections of EDGs on seismically induced core damage of a multi-unit site

Muramatsu, Ken; Liu, Q.; Uchiyama, Tomoaki*

no journal, , 

A preliminary seismic PSA study was conducted to examine core damage frequency (CDF) and core damage sequences with consideration of the effect of correlation of component failures for a two-unit site. The effectiveness of the cross connection of emergency diesel generators (EDGs) between adjacent units was also examined. The results showed that calculated CDF depends on the assumptions on component correlations and when the rules for assigning correlation coefficients defined in NUREG-1150 program was adopted, the CDF of a single unit, the CDF of this two-unit site and the frequency of simultaneous core damage of both units increased. When cross-connection of EDGs between two units was available, the CDF of a single unit, the CDF of this two-unit site as well as the frequency of simultaneous core damage of both units decreased. In addition, the CDF of this two-unit site was smaller than the CDF of a single unit site.

Uncertainty and sensitivity analysis of risk assessment models, 3; A Comparison of several uncertainty importance measures for ranking influential input parameters of risk assessment models

Liu, Q.; Homma, Toshimitsu; Hato, Shinji

no journal, , 

When the risk information obtained from the PSA study of nuclear power plants is to be used for safety-related decision making, it is necessary to reduce the uncertainty in the risk. Therefore, it is necessary to identify the influential components that contribute to this uncertainty. In this work, we applied several uncertainty importance measures (including the measures proposed by the authors) to a risk assessment model and ranked the importance of each input parameter. It is found that although the ranking orders of the input parameters with regards to these measures are a bit different, the use of several or more importance measures can increase confidence in the ranking of key input parameters.

Development of GSALab computer code for global sensitivity analysis

Liu, Q.; Homma, Toshimitsu

no journal, , 

In order to evaluate the model output uncertainty and the contribution of each model input to it, the computer code GSALab, which is based on Monte Carlo simulations, has been developed. It is composed of three parts, namely, random samples generation, uncertainty analysis and sensitivity analysis. The random samples are generated based on the probability distribution of model input variables in the part of random sample generation. The statistics as well as the distribution of the model output are computed in the part of uncertainty analysis. In the part of sensitivity analysis, several global sensitivity indicators, including the popularly used variance-based indicators, are implemented. In addition, the GUI (Graphical User Interface)of GSALab has been developed for the user's convenience. In addition to risk assessment models, it is also possible to use GSALab for uncertainty and sensitivity analysis of a wide class of mathematical models.