Heat exchanger sizing and shared-exchanger operation of a transcritical CO2 Carnot battery

This study investigates the physical sizing and shared-exchanger operation of transcritical CO 2 Carnot Batteries, focusing on the transition from thermodynamic optimization toward practical implementation. Building on a previously established thermodynamic framework (Shakirov et al. 2026), this work addresses the physical design of heat exchangers (HEX) using a discretized UA-LMTD methodology for Printed Circuit Heat Exchangers (PCHEs), identified as highly suitable for capturing the steep, non-linear thermophysical property variations of CO 2 near the pseudocritical region. The proposed sizing methodology is validated against experimental benchmark data from the literature, achieving agreement within 2.9%, well within the 11.97% experimental measurement uncertainty. Applying this framework to the optimal baseline cycle reveals a critical operational constraint: the required heat transfer areas for the charging and discharging cycles differ substantially. Consequently, utilizing identically sized heat exchangers for both operational modes without strategic mass flow regulation leads to severe thermal imbalances. To overcome this limitation and enable shared-exchanger operation with a single set of heat exchangers, linear and dynamic bypass mixing strategies are evaluated alongside terminal approach tolerance ( τ ) calibration. These strategies utilize controlled bypass streams to adjust the temperature and thermal capacity of the working fluid, while calibrating τ per unit matches the discharge areas of the three high-temperature exchangers to within 0.14% of their installed charge-cycle values. Furthermore, holding the cold thermal energy storage at −3.0 °C enables a single 6330 m 2 latent unit to serve both modes. With a single set of heat exchangers, the shared cycle achieves a round-trip efficiency of 54.52%, surpassing the 53.92% baseline with dedicated exchangers at equal input power: the mixing strategy delivers a high turbine inlet temperature (116.91 °C) while yielding a lower pump inlet temperature that reduces parasitic compression work. The system operates at 100% nominal storage mass flow over an identical 7.0-hour discharge duration. For the studied cycle configuration, dynamic mixing converges identically to the linear mixing optimum, establishing linear mixing as the preferred architecture due to its lower piping and control complexity.