Research On Uncertainty Analysis Method Based On Perturbation Theory

To develop nuclear data uncertainty analysis, statistical methods and mechanism methods are both applicable. In this paper, the mechanism method perturbation theory is applied to carry out sensitivity and uncertainty analysis of core effective multiplication factor about the few-group macroscopic cross section parameters.However, the traditional analysis focuses on sensitivities of the effective multiplication factor of microscopic cross section different reaction channel, in this paper we focus on the process of uncertainty propagation, namely nuclear data uncertainties transfer from the microscopic cross section to macroscopic cross section,and then influence the core effective multiplication factor. To get nuclear data uncertainty information, firstly we need to obtain sensitivity information of the reactor integral parameter about the few-group macroscopic cross section parameters, and multiply nuclear cross-section uncertainty covariance matrix to get uncertainty of core effective multiplication factor. The innovation point of this paper is that the sensitivity of the assemblies and the covariance matrix of the reactor core from the covariance matrix of the assemblies are got.The sensitivity calculation formulas of thermal neutron reactor effective multiplication factor about the few-group macroscopic cross section are obtained based on the perturbation theory. And the program PETMUS coupled the program CITATION to calculate sensitivity and uncertainty is developed. We research on the AP1000 reactor core, and other reactor cores are also suitable for the theory. In this process, we need to carry out critical calculation of the reactor core. Therefore diffusion equation rather than transport equation is applied. This program PETMUS couple the core diffusion program CITATION to calculate the sensitivity of keff to the few-group macroscopic cross section. The method of direct perturbation is applied to test and verify the sensitivity of the few-group macroscopic cross section ∑a, v∑f,∑s and D. It shows that the research method is correct. Combined with SCALE6.1 calculated the covariance matrix of component itself and between the different components, the uncertainty of the core effective multiplication factor is got, and then analyze the uncertainty results.In order to verify the uncertainty calculation of reactor core effective multiplication factor,another partner’s statistical method in our research group is used to compare the results,through statistical sampling methods we get uncertainty of effective multiplication factor,the result calculated by the two methods is very similar.It is proved that the research result is reasonable and feasible. At last, the statistical methods is applied to get the uncertainty of the normalization power distribution of the reactor, the flux distribution and the adjoint flux distribution, and then analyze the uncertainty results.