The present work addresses electric infrastructure incidents: rescheduling is needed in order to optimally share the available power among the different trains. We propose a rescheduling process based on dynamic multiphysic railway simulations that compute physical quantities such as the train speed profiles, voltage along the catenary lines, temperatures … The optimization problem to solve is characterized by a large number of continuous and discrete variables, constraints on dynamic outputs (typically voltage limits), and a high computation cost. We use sensitivity analysis in order to analyse the behaviour of this complex system and provides us more information around the adjustment operations helping to reorganize train traffic in an optimal way. Our approach is based on statistics of the output of simulations of the dynamic railway system, with predefined variation ranges of the input parameters. Two parts of analysis are considered: variance decomposition based sensitivity analysis (generalized Sobol indexes) used for prioritization and fixing factors (quantitative analysis); regional sensitivity analysis (qualitative analysis) used for factor mapping. The proposed approach has been tested on a simple case, with a nominal traffic running on a single track line. The considered incident is the loss of a feeding power substation. The variables to be adjusted are: incrementing of the train departure times and speed reduction at given positions. The results show that the incrementing of train departure times is the most influential variable and we got Pareto-optimal front that meets multi-criteria such as: minimize travelling time, train delays, traction energy …