Modelling a molten salt thermal energy system – A validation study.

Highlights • Easily define a mixture of molten salt and non-condensable gas for rigorous dynamic simulation. • Enables system-wide thermal and hydrodynamic analysis for energy storage processes. • New experimental data on operating a thermal energy storage facility using molten salt. • The heat exchanger performance is influenced by trapped non-condensable gas. • Anomalous sudden changes in the hydrodynamic losses uncovered. Abstract Thermal energy storage (TES) plays a crucial role improving the efficiency of solar power utilization. Molten salt (MS) has gained a strong position as a thermal fluid in applications where solar power is stored and used overnight to provide dispatchable energy production. Novel process and operating concepts are being developed for TES systems that require reliable engineering tools. System-wide dynamic simulation provides a virtual test bench and analysis tool for assisting in process and control design and operational issues. Proper characterization of the thermal fluids in simulation tools is critical for successful simulation studies. In this paper, we report the experimental and modelling work related to counter-current heat exchange and free drainage test runs in CIEMAT's multi-purpose MS test loop at Plataforma Solar de Almería in Spain. We present a general method to define MS and non-condensable gas within a homogeneous pressure-flow solver. We present modelling of an indirect MS TES system connected to a thermal oil loop through TEMA type heat exchangers, model calibration with half of the experimental data, and finally, validation simulations against rest of the data. All these experimental data are previously unpublished. The model predicts the system behaviour with good agreement regarding temperatures, pressures, flow rates and liquid levels. The simulations suggest that the heat exchangers' shell sides suffer from trapped non-condensable gas which significantly affects heat transfer, heat loss to ambient air and hydrodynamic losses. Our results contribute to thermal-hydraulic, system-wide modelling and simulation of MS processes. Furthermore, the results have practical implications for MS TES facilities with respect to system design, analysis and operation. [ABSTRACT FROM AUTHOR]