Geometric Optimization of Latent Heat Thermal Energy Storage for Carnot Batteries based on Multiscale Simulation

Kurzfassung

Discretized Multiscale Framework The efficient implementation of Carnot Batteries based on subcritical Rankine cycles requires the minimization of temperature differences in the dominant evaporation and condensation processes. Latent Heat Thermal Energy Storage (LHTES) systems are a preferred option because the absorption and release of heat take place under approximately isothermal conditions. Given the low thermal conductivity of cost-effective Phase Change Materials (PCMs), the development of effective heat transfer solutions is essential for the thermo-economic optimization of the LHTES system. To achieve sufficiently high-power densities, heat transfer structures are integrated into the storage volume. A multi-level simulation methodology is used in the design of these structures, as shown in Fig.1: Level 1 analyzes the 2D cross-section of the finned tube and PCM; Level 2 evaluates axial variations along a single full-length tube; and Level 3 aggregates these to model the full storage unit, providing the basis for PCM storage models that integrate into complete Carnot Battery system simulations on Level 4. The successful development of this multi-level methodology requires the identification of computationally efficient models with adequate accuracy at each level.

Translation from First to Second Level A critical bottleneck in this framework is the high computational cost of analyzing complex 2D extended surfaces at the first discretization level. To address this, the authors have developed a computationally efficient methodology that characterizes finned geometries using a novel, distance-based framework [1]. This approach translates complex 2D fin cross-sections, such as star, snowflake, and organic profiles, into 1D Cumulative Distribution Functions (CDFs) based on shortest-path heat transfer. This simplified model eliminates the need for extensive numerical calibration, enabling rapid integration of Level 1 effective properties into the higher levels for complete and computationally efficient LHTES system optimization.

This work will provide insight into the purely geometrical characterization of finned tubes immersed in PCM and the subsequent calculation of effective properties for use at the single-tube level.