Uncertainty Analysis on the Fission Products Behavior Simulation Module of CINMEA Code based on Correlation Coefficients Analysis Derived from Linear Regression

다국어 초록 (Multilingual Abstract)

Uncertainty Analysis on the Fission Products Behavior Simulation Module of CINMEA Code based on Correlation Coefficients Analysis Derived from Linear Regression Yongjun Lee Department of Nuclear Engineering Graduate School of Hanyang University In South Korea, efforts are underway to develop a domestic comprehensive severe accident analysis code for regulatory use. The CINEMA code, which is under development and improvement, is a comprehensive severe accident analysis code designed to evaluate the overall phenomena and consequences occurring during severe accident scenarios, like existing codes such as MELCOR and MAAP. CINEMA consists of four sub-modules, each responsible for simulating in-vessel phenomena, ex-vessel phenomena, fission product behavior, and data exchange between modules. Among these, the SIRIUS module predicts the behavior of radioactive materials by simulating the release, transport, deposition, and removal of fission products during severe accidents. As the module responsible for fission product behavior, SIRIUS is currently under development but lacks information on the size and distribution of the inherent uncertainties. Therefore, this study establishes relevant uncertainty variables for SIRIUS and performs an uncertainty analysis to determine the potential uncertainty range of the consequences and to compare the importance of the output variables impacting FP behavior. Methodologically, this thesis focuses on the LBLOCA scenario, selecting uncertainty variables based on the operational certifications of the OPR1000. The sampling of uncertainty variables took the methodology of LHS for efficient and evenly distributed sampling. The minimum sample size was determined using the Wilks’ formula, resulting in a total of 153 sample sets to ensure the reliability of the analysis. The FOM chosen for this analysis includes airborne aerosols and release fraction, both of which are analyzed for the two fission product groups in SIRIUS. The framework required for the uncertainty analysis was automated using an in-house Python-based pre- and post-processing system. Through the code execution for uncertainty analysis, a horse tail plot was generated, and statistical analysis was conducted based on the primary data. As a result, CSF and SSF were identified as the most significant uncertainty variables, showing the highest correlation with airborne aerosol concentration and release fraction. In particular, CSF was found to promote particle growth and removal, thereby reducing the airborne aerosol, while SSF was observed to inhibit particle settling, thereby increasing the aerosol suspension time. In the early stages of the accident, both variables showed a strong correlation with the reduction or retention of airborne aerosols, respectively, with their correlation diminishing over time. In addition, Gap Release Temperature showed a strong positive correlation at the early phase of accident, but its correlation decreased rapidly to almost zero. This suggests the role for GAPT in delaying the initial release of radionuclides. Other variables exhibited relatively low correlations, indicating limited impact on the FP behavior. The uncertainty analysis results obtained in this study provide valuable data for identifying dominant uncertainty variables within the SIRIUS module and are expected to contribute to improving the reliability of the CINEMA code. Furthermore, the methodologies and findings of this study can serve as essential foundational data for the localization of regulatory technologies to improve the severe accident safety of domestic nuclear power plants.

목차 (Table of Contents)

• 1. Introduction 1
• 1.1 Motivation. 1
• 1.2 Pervious Uncertainty Analysis Review 3
• 1.3 Overview of CINEMA code 9
• 1.4 Objective 10
• 2. Preliminary Setup for CINEMA Code Uncertainty Analysis 12
• 2.1 Target System and T/H Calculation 12
• 2.2 SIRIUS FP Model and Overview of Uncertainty Analysis 20
• 2.3 Post/Pre-Processing for Statistical Analysis 27
• 3. Methodology for Uncertainty Analysis 30
• 3.1 Sampling Methodology 30
• 3.2 Uncertainty Variables 34
• 3.3 Production of T/H Information Files for SIRIUS Module 41
• 3.4 Post-processing Methodology and FOM Selection 46
• 4. Results and Discussion 50
• 4.1 Code Execution Result 50
• 4.1.1 Horse Tail Plot 51
• 4.1.2 Airborne Aerosol Distribution 55
• 4.1.3 Release Fraction to CTMT 65
• 4.1.4 Release Fraction to RCS 76
• 4.2 Statistical Analysis 85
• 4.2.1 Statistical Representative Values at specific time 85
• 4.2.2 Time-series Correlation Coefficient for Airborne Aerosol 89
• 4.2.3 Time-series Correlation Coefficient for Release Fraction 95
• 5. Summary and Conclusion 106
• Bibliography 109
• 국문요지 114
• 감사의 글 116