%0 Conference Paper
%F Oral
%T Nuclear data uncertainties and adjustments using deterministic and Monte-Carlo methods along with PWR measurements
%+ École Polytechnique de Montréal (EPM)
%+ Laboratoire de Neutronique (IRSN/PSN-RES/SNC/LN)
%+ Paul Scherrer Institute (PSI)
%A Salino, Vivian
%A Rochman, Dimitri
%A Dumonteil, Eric
%A Malvagi, Fausto
%A Hébert, Alain
%< avec comité de lecture
%B M&C 2023 - The International Conference on Mathematics and Computational Methods Applied to Nuclear Science and Engineering
%C Niagara Falls, Ontario, Canada
%8 2023-08-13
%D 2023
%K Total Monte-Carlo
%K BFMC adjustment
%Z Physics [physics]/Physics [physics]/Computational Physics [physics.comp-ph]Conference papers
%X Nuclear power plants safety relies partly on predictions from numerical simulations. Their uncertainties must therefore be evaluated. In PWR reactor physics, the simulations' uncertainties are nowadays estimated on the basis of the differences between existing measurements and the corresponding simulations. This approach leads to weaknesses in the estimation of uncertainties. For example, a novelty can induce more deviations than those seen in the past. Moreover, in some (common) cases, these uncertainties are based on differences between measurements and adjusted simulations, i.e. carried out after the measurements and by forcing the simulations as close as possible to the measurements. The residual deviation thus leads to consider compressed uncertainties. This method of evaluating uncertainties can therefore be improved. To further increase robustness, the approach proposed in a recently published Ph.D. thesis is to return to the fundamentals : the sources of errors in the simulations. The two main types discussed in this paper are the following. On the one hand, physical and numerical approximations introduce deterministic biases, which can be evaluated as the difference with Monte-Carlo. On the other hand, the nuclear data uncertainties, that arise from the limited human understanding of nuclear physics, are propagated in this article with a Total Monte-Carlo approach.
%G English
%2 https://irsn.hal.science/irsn-04095487v1/document
%2 https://irsn.hal.science/irsn-04095487v1/file/Salino_ND.pdf
%L irsn-04095487
%U https://irsn.hal.science/irsn-04095487