The modelling and simulation of nuclear systems allow scientists and engineers to probe aspects of the operation of these systems beyond what has been precisely measured. This has numerous applications, including interpolation between known operational scenarios and extrapolation into the unknown, providing information to underpin safety, optimise performance and design new facilities and experiments. As computational resources have increased in the past decades and with higher-fidelity modelling approaches, leveraging novel simulation methods has become critical to the nuclear industry. However, the potential value of all simulation is tightly bound to the body of evidence, which supports the accuracy of the methods in the application domains.
The NEA Expert Group on Multi-physics Experimental Data, Benchmarks and Validation (EGMPEBV) was created in 2014, under the aegis of the Nuclear Science Committee (NSC), to address the needs of member countries in certifying and utilising benchmarks for the validation of novel multi-physics codes. These benchmarks and the guidance on their use will help experts validate their codes, enabling their use in the wider nuclear community. Under the second task force created in the EGMPEBV, experts compared their approaches and their relative strengths and limitations, including a critical review of the current state-of-the-art in verifying, validating and performing uncertainty quantification with coupled multi-physics simulation.
This report provides a self-contained review of the current approaches for single and multiphysics validation, including both coupled and uncoupled uncertainty quantification and propagation. Through a detailed inter-comparison study, this report compiles the common elements that are required for best practices.