Development of Perturbation Theory Based Model for Sensitivity and Uncertainty Analysis

The multigroup cross sections used in the reactor physics calculations have associated uncertainties. The cross section uncertainties, present in the evaluated nuclear data libraries in the form of covariance matrices, are considered as the most significant uncertainty source in the reactor physics analysis. The quantification of these uncertainties is necessary to have confidence in simulation models and ascertain safety margins in operating and future nuclear facilities. A perturbation theory based model has been developed for sensitivity & uncertainty (S&U) analysis capability to the indigenous lattice physics analysis code VISWAM. The sensitivity profiles of k∞ for different reactions such as capture (n,γ), fission (n, f) etc. of significant isotopes such as 235U, 238U etc. are estimated using adjoint flux weighting method based on perturbation theory. The uncertainty quantification in k∞ for these reactions is estimated using the sandwich rule. In the sandwich rule, the relative covariance matrices for a reaction are weighted with the corresponding sensitivity profile. The multigroup relative covariance matrix library required for the uncertainty calculation is generated using NJOY 2016 code system. The accuracy of S&U model developed in VISWAM code has been verified by analyzing Exercise I-1 of the OECD/NEA Benchmark for Uncertainty Analysis in Modeling (UAM) for Design, Operation, and Safety Analysis of LWRs. The Exercise I-1 of the benchmark is dedicated to evaluate uncertainties for a lattice cell due to different microscopic cross sections. The results obtained using VISWAM are compared with other published results for the cell physics benchmark in the present paper.