Study Of Key Issues Of The Best Estimate Plus Uncertainty Methodology For Reactor Thermal Hydraulic Analysis

Compared with traditional conservative methods,the best estimate plus uncertainty(BEPU)method can provide more real analysis results and safety margins,and improve the economy and operation flexibility of nuclear power plants on the premise of ensuring safety.Code Scaling,Applicability and Uncertainty(CSAU)method is a representative guidance framework for the BEPU method application practice.This thesis proposes solutions and methods for the problems of code applicability evaluation,code scaling capability and the impact of model scale factors on uncertainty analysis in the CSAU method,and conducted application research based on reactor test facilities.In the evaluation of the thermal hydraulic program code based on the separate effect test facilities,according to the principle of expectation maximization algorithm,the algorithm program is written to solve the interval range solution problem of the physical model coefficient of the thermal hydraulic code.Combined with the separate effects test facility Marviken,the 95% confidence intervals for the discharge coefficient and thermal non-equilibrium coefficient of the critical flow model are [0.6877,1.2847] and[0.1022,0.1739],respectively.It can be used as input parameters for uncertainty analysis.In the evaluation of thermal hydraulic code simulation capability based on the integrate effect test facilities,the overall simulation capability of the code was quantitatively evaluated by combining the fast Fourier transform method.The numerical modeling and transient simulation of the typical accident conditions of the ACME facility and the PUMA-E facility were carried out,and the overall performance index was quantified to be less than 0.3.This verifies the accuracy of the code simulation facility target accident conditions,and provides technical support for the application of system code to the safety assessment of target Generation III reactor.In the thermal hydraulic code capability evaluation based on the comparison of different codes,the numerical model conversion process of different codes was constructed to realize the reliability evaluation of code simulation for the extended operation of the accident.Combined with the test data,the reliability of the newly constructed facility TRACE models was verified.the reliability of the newly constructed models.Then,the simulation of the extended conditions of the accident was carried out.The reliability of code simulation results and the safety of the reactor under the accident condition were confirmed by the comparative analysis of the simulation results of different codes.An analysis method was expanded to solve the scaling capability evaluation of the code by using the "ideal scale extrapolation" of the facility model.A numerical model of the ideal extrapolation of the ACME facility to the scale of the prototype power station was established.The results show that the simulation of important phenomena of the two scale models is well consistent,and the overall quantitative accuracy evaluation index is less than0.3,which proves the scale-scaling ability of the code for the third-generation PWR.That is,the system code verified by the small-scale test bench is suitable for the simulation analysis of large-scale prototype reactors.Aiming at the influence of scale factors on uncertainty analysis of bench model,a comparative study of uncertainty across scales,the same working conditions,and the sampling combination of the same input parameters was designed and implemented.The results show that the average deviation brought by the numerical model scale factors to the target parameters of interest is 6.39%,and the uncertainty envelope range of the output results changes according to the designed proportional scale,and the most significant parameter affecting the minimum core collapsed level is core decay heat.In summary,aiming at the code model evaluation problem in the CSAU method,this thesis develops an evaluation program that uses the expectation maximization algorithm to quantify the parameter distribution of the physical model,and combines the fast Fourier transform method to realize the quantitative evaluation of the code simulation accuracy,and constructs model conversion frameworks to realize the code simulation reliability analysis of the accident extended conditions.For the problem of code scale capability evaluation and the influence of model scale factors on uncertainty analysis,new solutions were proposed.The research results have reference guiding significance for establishing the independent intellectual property rights BEPU method framework.