Strategies for the uncertainty quantification of fuel cycle scenarios

Fuel cycle simulators' results are very dependent on the uncertainties of the input parameters. To assess this dependence 2 types of methods are used: local methods like sensitivity analysis and global methods. Global methods provide more information like the interaction between input parameters or the existence of non-linear effects. The main drawback of global methods is their excessive computational costs. Global methods are characterized by the Sobol indices that are integrals defined in a multi-dimensional space that grows with the number of parameters. In this presentation other approaches for the calculation of the Sobol indices are presented. They include different sampling techniques that speed up the convergence, or the build of a surrogated method that reduces the number of simulations. This document is composed of an introduction and the slides of the presentation.