Deterministic-Realistic Hybrid Methodology (DRHM) For LOCA Licensing Analysis

The 1988 amendment of the 10 CFR 50.46 rule allowed the use of realistic physical models to analyze loss-of-coolant accident (LOCA). Best-estimate LOCA methods are now extensively employed within the nuclear industry. Both the realistic or best-estimated LOCA methods and conservative LOCA methods are the current licensing LOCA safety analysis methods. A realistic analysis requires the execution of various realistic LOCA transient simulations where the effect of both model and input uncertainties are ranged and propagated throughout the transients, but a procedure is developed to assess the code and determine its biases and uncertainties. A new methodology called Deterministic-Realistic Hybrid Methodology (DRHM) for large break loss of coolant accident licensing analysis with conservative Appendix K models plus statistical plant-related parameters analysis is presented in this paper. By using the DRHM, while satisfying related safety regulation and analysis requirements, a greater thermal margin can be generated as compare to the conservative LOCA methods traditionally.A key step in a best-estimate analysis is the assessment of uncertainties associated with physical models, data uncertainties, and plant initial and boundary condition variabilities. For the assessment of physical model uncertainties need to be associated with lots of separated-effected test and integral-effected test data, a relatively realistic technique have been adopted in the DHRM. In the DRHM methodology, a conservative model for LOCA analysis was adopted with statistical treatment of plant status uncertainty. The conservative LOCA model satisfy the requirements of conservative evaluation models (EM) in related regulation (10CFR50, Appendix K) by modifying related models or correlations of the best estimate (BE) program RELAP5/MOD3, in an effort to form a licensing LOCA analysis code. So the procedure that developed to assess the thermal/hydraulic code and determine its biases and uncertainties was replaced by integral conservative assessment.For the assessment of uncertainty associated with data uncertainties, initial and boundary condition variabilities, of which all the plant-related parameters that are potential contributors to the overall uncertainty were determined by phenomena identification and ranking table (PIRT) and plant's technical specifications. The plant-related parameter uncertainty can be evaluated from data or known operation ranges plus measurement uncertainty. The nonparametric statistics was adopted in the DHRM, which derives from direct Monte Carlo methods. However, instead of attempting to obtain information with regard to underneath probability distribution function (PDF) of the measure, the PDF is ignored (distribution-free) and nonparametric statistics is used to determine a bounding value of the population with a given confidence level. So that the bounding value of the 95th percentile tolerance interval with a 95% confidence level can be determined directly for each of the 10 CFR50.46 acceptance criteria, that is, peak clad temperature (PCT<2200F), local maximum oxidation (LMO<17%), and core-wide oxidation (CWO<1%).Application of LOCA analysis to the LOFT L2-5 double-ended cold leg guillotine large break loss-of-coolant accident integral test is performed in the paper. With the modified RELAP5/MOD3, the results of the experimental simulation in the thermal-hydraulic responses for LOFT/L2-5 are assessed to be in good accordance with experimental data. Furthermore, the large break LOCA transient description which divided into a number of time periods in which specific phenomena are occurring was performed. Based on the non-parametric statistical approach and following a direct Monte Carlo technique, all uncertainty attributes are randomly sampled from their respective distributions to obtain a population of 124 transients. For each of the transients, all uncertainty attributes are simultaneously sampled. Results are then ranked in terms of PCT, LMO and CWO, the top-ranked value for each of the three outcomes ensures the 95-percent probability coverage at 95-percent confidence level for PCT, LMO and CWO. The upper bounding case which all the code inputs were in the worst condition was calculated and compared with PCT95/95. In addition, parametric statistics methods were also performed in the paper. By passing the Chi-square goodness-of-fitting test for normality, theoretical normal distribution curve accounting for the sample mean and the standard deviation obtained from the 124 runs was generated, then the bounding value of the 95th percentile one-side tolerance interval with a 95% confidence level can be deduced. Furthermore, a sensitivity analysis to the effects of various initial operation condition parameters on PCT is presented.