Model order reduction of a once-through steam generator via dynamic mode decomposition

Once-Through Steam Generators (OTSG), as the heat exchanger and the radioactive shield that connecting the primary system and the secondary system, is of vital importance for nuclear safety and economy. Whereas large-scale and coupling simulation models can provide high-fidelity estimations of the flow and heat exchange in OTSGs, there is an extra computational burden when applied to multi-query tasks such as optimization and uncertainty analysis. Model Order Reduction (MOR) methodology provides an alternative for the multi-query tasks. Few studies exist related to nuclear reactor components and to the OTSG using MOR; as such, this paper introduces DMD, which is data-driven and is suitable for any problem without restriction, to build the Reduced Order Model (ROM) aiming to offer accurate and fast estimation for the dynamic operating characteristics of the OTSG. The results show the established DMD model can accurately simulate the system state and improve greatly the computational efficiency.