A global sensitivity analysis framework for guiding the design improvement of new concepts of small modular reactors

Small Modular Reactors (SMRs) are new concepts of Nuclear Power Plants (NPPs) whose safety assessment typically cannot rely on design-specific operational experience and field data. Assessing their safety requires a careful combination of engineering expertise, and of physical (laboratory) and virtual (simulation) experiments. Modeling the system response accurately, in consideration of the engineering design and physical aspects of SMRs requires computationally burdensome computer models. Although interpretable, it is difficult to consis-tently identify the most important inputs and provide safety margins, especially for high-dimensional correlated inputs and outputs. To address these issues, we propose a Global Sensitivity Analysis (GSA) framework that combines recently introduced global sensitivity indices based on Optimal Transport (OT) with a graphical visualization tool based on CUmulative SUm of NOrmalized Reordered Output (CUSUNORO). The OT-based indices allow us to identify the most important inputs, whereas the graphical representation of the latter en-ables us to identify the safety critical ranges of these inputs. The proposed framework is applied to a demon-strative case study concerning Small Modular Dual Fluid Reactors (SMDFRs), a new SMR concept. The numerical experiments show how the approach can help interpreting the simulation results and suggesting design im-provements to keep operational settings away from the Critical failure Region (CR).