Optimizing Performance of Low-Quality Graphite for High-Temperature Thermal Storage

Energy storage is a critical component of achieving the clean energy transition, due to the intermittency of common renewable generation technologies like wind and solar. Thermal energy storage (TES) is one cost-competitive option for large-scale, long-duration energy storage. TES stores energy in the form of heat, allowing use of a wider variety of cheaper storage materials than traditional electrochemical batteries.<br/><br/>The specific TES technology considered in this work is called thermal energy grid storage with thermophotovoltaics (TEGS-TPV). In the charge cycle, resistance heating is used to heat liquid tin to high temperatures (~2400°C), which is then flowed through graphite blocks to heat them. As the thermal storage medium, the graphite retains this heat for several (&gt;10) hours. During discharge, colder tin (~1900°C) is flowed through the graphite, and the heated tin at the outlet is flowed past TPV cells which convert the light emitted by the tin pipes at these high temperatures to electricity. Technoeconomic analysis indicates the cost per energy of TEGS-TPV is &lt;$20/kWh-e and the cost per power is &lt;0.40/W-e.<br/><br/>These promising technoeconomics are enabled by cheap (&lt;$0.50/kg) graphite as the storage material. However, this graphite is often low quality, formed by molding machining scraps or recycling used graphite. It is therefore porous, has poor mechanical strength, and may contain impurities. Understanding the high-temperature thermal properties of low-quality graphite is critical to evaluating its performance as a thermal storage material.<br/><br/>The first part of this study measures graphite’s high-temperature thermal properties including density, specific heat, and thermal conductivity, following ASTM standards. We use machines that measure properties up to 1500°C, and we extrapolate the measurements to the temperature region of interest with physics-based correlations. Using these properties, we can calculate the thermal diffusivity of the graphite, which directly correlates with thermal performance. Based on the measurements taken, we find low-quality graphite has low thermal conductivity (~10 W/mK) at room temperature, and worsening qualities with increasing temperature. In comparison, high-quality graphite has room temperature thermal conductivity around 100 W/mK.<br/><br/>The second part of the study investigates how best to operate the charge and discharge cycles given the less ideal properties of low-quality graphite. A transient 3D multiphysics simulation is set up coupling the liquid tin fluid flow to the conductive heat transfer within the graphite. Several flow configurations were tested, including varying the spacing and diameter of tin tubes within the graphite, considering flowing tin through several graphite blocks in series, or changing the flowrate as a function of time. An optimal design was determined to ensure near-constant power output through the storage duration.<br/><br/>Finally, the third part of the study develops a lumped capacitance model to accelerate computation of transient liquid metal and graphite temperatures. The developed model splits each block into 5 pieces and considers each as isothermal. This results in errors of less than 1% through the transient storage duration when compared to the 3D multiphysics model, with 3 orders of magnitude faster computation time. This allows rapid calculation of transient graphite and tin temperatures for varying charge/discharge profiles, for example based on daily generation and demand curves.<br/><br/>In conclusion, this study investigates the high-temperature properties of low-quality graphite to optimize its performance as a thermal storage material. By combining measurements with modeling, we determine ideal charge and discharge geometries, and develop a lumped capacitance model for accelerated transient calculations. Overall, our study helps to enable cheap long-duration storage to ease the transition to a clean energy future.