Modeling and Simulation Needs and Capabilities for Artificial Intelligence Based Plant Reload Optimization Platform

The Risk-Informed Systems Analysis (RISA) Pathway Plant Reload Optimization Project-under the United States (U.S.) Department of Energy's (DOE's) Light Water Reactor Sustainability (LWRS) Program-aims to develop and demonstrate an automatized technology-inclusive platform that can generate optimized fuel load configurations for the reactor core of a nuclear power plant. The project targets to optimize reactor core thermal limits through the implementation of state-of-the-art computational and modeling techniques. During the development of the platform, the constraints were identified in computational tools. The main issue that was identified is that the tools need to be immediately applied to the optimization platform without significant development or update. The tools used in the platform should have the highest technical maturity so that they can be deployed to nuclear industry with ease. Hence, this study focused on reviewing the applicable computational tools in the field of the reactor core design and fuel performance analysis to give a snapshot on tool selection for the optimization platform. The benchmark study was therefore performed using well-designed case studies for the core design and fuel performance tools planned to be used in the plant fuel reload optimization framework. Three core design tools-VERA-CS, SIMULATE-3, and PARCS-are reviewed and compared with different hot zero power tests, physical reactor zero power physics tests, and a hot full power case. Two fuel performance tools-BISON and TRANSURANUS-are reviewed and compared with two instrumented fuel assembly test cases to analyze fuel performance under the long-term operation and loss of coolant accident event. Other parameters are benchmarked, including computational performance issues while coupling with Risk Analysis and Virtual Environment (RAVEN) and accident tolerance fuel (ATF) applicability.