A dynamic model of a sodium/salt PCM energy storage system Zebedee Kee, Joe Coventry and John Pye* Research School of Engineering, Australian National University, Canberra, Australia. *Tel: +61 2 6125 8778; email: john.pye@anu.edu.au. 1. BACKGROUND Thermal energy storage (TES) implemented in Concentrated Solar Power (CSP) addresses the issue of time mismatch between energy demand and supply. With TES, CSP plants can be operated flexibly to ensure power supply matches demand, maximising revenue. Direct two-tank TES with a molten eutectic mixture of NaNO 3 and KNO 3 is the currently dominant commercial choice in CSP. However, chemical stability issues limit operating temperatures which in turn limit the thermal conversion efficiency of the CSP plant’s power cycle. It is therefore desirable to explore alternative storage/transport media such as liquid sodium which are suitable for high temperature applications (Coventry et. al. 2015). In this work, we present a novel TES system involving a sodium heat pipe in direct contact with NaCl PCM. This combination is appealing due to high storage temperatures, low receiver losses and the potentially minimised cost of the storage subsystem. 2. INTRODUCTION NaCl has a low cost and high melting point of around 1073K, which is also similar to the saturation temperature of sodium heat transfer fluid (HTF) at slightly sub-atmospheric pressures. A dynamic system model of the HTF-PCM storage within a CSP system was implemented in OpenModelica evaluate key dynamic aspects of the system such as the temperature response of the PCM whilst heat is added/removed from the storage vessel, and movement of liquid sodium between the receiver and storage trays. Several design parameters which include PCM container dimensions, quantities of PCM and HTF material, and charging/discharging rates could then be optimised for maximum exergetic efficiency, and in future, minimum levelised cost of electricity (LCOE). Figure 1: Left: Conceptual design of the TES subsystem. Right: Equivalent model used in simulation. 3. DESIGN CONCEPT A simplified model of the HTF-PCM configuration includes a storage vessel containing saturated Na liquid-vapour at a temperature T Na in contact with the top surface of the NaCl contained in trays (Figure 1). Heat is delivered to the isochoric vessel via a sodium boiler receiver at T Na , which operates for 6 hours a day and is shut down for the remaining 18 hours. In the 18 hours, heat is discharged from the storage vessel into a Carnot power cycle at T Na . During the charging process, Na vapour condenses on top of a pool of Na liquid. Excess condensing sodium overflows from the sides of the tray walls to the bottom of the vessel, where it can be pumped back for re-boiling in the receiver. During discharging, liquid sodium is boiled off the top of the PCM surface and must be replenished by pumping from the bottom of the vessel. 4. MODEL ASSUMPTIONS The sodium HTF was modelled as a single component, two- phase mixture at temperature T Na . T Na is calculated at each time step using specific enthalpy and specific volume constraints and equations provided by Fink and Leibowitz (1995). The sodium receiver was modelled as an isothermal blackbody cavity receiver with fixed concentration ratio (CR) direct normal irradiance (DNI) for 6 hours each day. During the remaining 18 hours, the system the discharges at a rate   output which is set to 1/3 of   input or until the total energy stored by the vessel returned to zero. During charging, the heat transfer process between HTF and PCM was assumed to be conduction-dominated due to the low Prandtl number of liquid metals, and thus modelled via an extra thermal resistance term. During discharging, the effective thermal resistance of the liquid Na layer was assumed to be zero due to the large heat transfer coefficient associated with pool boiling. The temperature gradient within the liquid Na pool is assumed to be small enough such that the effect on the two-phase Na HTF model is negligible. Heat transfer within the PCM was modelled using a numerical scheme involving 1D finite-difference, enthalpy formulation with mushy node idealisation (Sharma et. al. 2009; Dutil et. al. 2011). This numerical scheme was described to have relatively simple implementation, with a single governing equation for both solid and liquid phases. All PCM trays were assumed to experience identical heat transfer, and as such, were represented using a single equivalent tray (Figure 1). ARGESIM Report 55 (ISBN 978-3-901608-91-9), p 23-24, DOI: 10.11128/arep.55.a55198 23 MATHMOD 2018 Extended Abstract Volume, 9th Vienna Conference on Mathematical Modelling, Vienna, Austria, February 21-23, 2018