Modeling and validation of the DYNASTY natural circulation loop with the RELAP5 code

he DYNASTY (DYnamics of NAtural circulation for molten SalT internallY heated) facility, developed at the Energy Laboratories of Politecnico di Milano, offers a unique platform for investigating natural circulation dynamics under conditions of internal heat generation. Primarily focused on advancing the understanding of these critical processes, DYNASTY provides high-quality experimental data essential for validating numerical models. Beyond its primary scope, DYNASTY also delivers key insights for the study of passive safety systems based on natural circulation in light-water small modular reactors and Generation-IV reactors. This paper presents the development and validation of a novel RELAP5 model of the DYNASTY natural circulation loop. The RELAP5 code was selected for its established reliability in modeling thermal-hydraulic systems, providing efficient and accurate predictions. Model validation was conducted using experimental data from recent campaigns, which tested the facility's behavior under various conditions, including different cooling fan speeds and heating configurations. In addition to validating the model's predictive capabilities, a sensitivity analysis was conducted to assess its robustness under different numerical methods and nodalization schemes. The RELAP5 model demonstrates accurate reproduction of the DYNASTY experiments, with a maximum root mean square error of 7.25 °C, recorded in a case lasting more than 23000 seconds. This model provides a solid foundation for integrating additional physical and engineering aspects beyond thermal-hydraulics, expanding its application in the development of reliable passive safety technologies. Finally, in the context of advancing nuclear safety, this paper contributes valuable insights into the validation of passive safety systems models based on natural circulation, addressing critical challenges and paving the way for enhanced safety and efficiency in future deployments.