Verification of the accuracy of Doppler broadened, self-shielded multigroup cross sections for fast power reactor applications

Abstract

Over the last 5 yr the IAEA nuclear data processing code verification project has attempted to improve the accuracy of the results produced by processing codes. The first round of verification only considered unshielded, cold, group-averaged cross sections. In the second round of verification the effects of Doppler broadening and self-shielding are considered. In this paper we present verification results for Doppler broadening and self-shielding. One of the important results presented in this paper is that the original SIGMA1 method of numerical Doppler broadening as proposed by Cullen and Weisbin (1976) has now been demonstrated to be inaccurate and not capable of producing results to within required accuracies. Fortunately, due to this study, the SIGMA1 method has been significantly improved and the new SIGMA1 is now capable of producing results to within required accuracies. Although this paper presents results based upon using only one code system, it is important to realize that the original SIGMA1 method is presently used in many cross-section processing code systems; the results of this paper indicate that unless these other code systems are updated to include the new SIGMA1 method, the results produced by these code systems could be very inaccurate. In this paper we review the objectives of the IAEA nuclear data processing code verification project as well as the requirements for the accuracy of calculation of Doppler coefficients and the present status of these calculations. Next we present the initial results of Doppler broadening and self-shielding calculations and point out the inconsistency of the results which led to the discovery of errors in the original SIGMA1 method of Doppler broadening. Analysis of the errors found and improvements in the SIGMA1 method are presented. Improved results are presented in order to demonstrate that the new SIGMA1 method can produce results within required accuracies. Guidelines are presented to limit the uncertainty introduced due to cross-section processing in order to balance available computer resources to accuracy requirements. Finally cross-section processing code users are invited to participate in the IAEA processing code verification project in order to verify the accuracy of their calculated results.