Two directions Vehicle-to-grid and vehicle-to-load capabilities are creating new thermal management challenges as batteries cycle more frequently between charging and discharging states. While most residential bidirectional systems operate at relatively modest 7–11 kW levels, the cumulative effect of repeated thermal cycling can accelerate battery degradation if not properly managed. The key challenge lies in maintaining tight temperature control during these bidirectional power flows, particularly in extreme ambient conditions. Effective thermal management can extend the operational envelope for vehicleto-everything applications, allowing battery systems to participate in grid services even during cold weather when the economic value may be highest. Some implementations now incorporate predictive preconditioning algorithms that warm batteries to optimal temperatures in anticipation of scheduled discharge events. Thermal system durability also becomes more critical in bidirectional applications. Components such as pumps and valves must withstand significantly increased duty cycles compared with conventional singledirection charging scenarios. This has driven development of more robust cooling system components with enhanced wear resistance and maintenance-free operation targets. Cool intelligence Artificial intelligence (AI) and machine learning are transforming battery thermal management from reactive to predictive systems. Modern implementations use AI in several key ways: forecasting thermal loads based on driving patterns and charging history, optimising coolant flow in real time to minimise energy consumption and detecting early signs of thermal anomalies. The most advanced systems combine reduced-order physical models with machine learning trained on extensive operational data. This hybrid approach maintains the interpretability of physicsbased models while exploiting AI’s pattern recognition capabilities for subtle thermal behaviours. Some implementations have demonstrated 15–20% reductions in cooling energy use. Emerging ‘smart virtual sensing’ techniques show particular promise, using AI to infer internal battery temperatures from easily measurable external parameters. This could eventually reduce or eliminate the need for extensive internal sensor arrays while maintaining or improving thermal monitoring accuracy. Looking ahead The coming half-decade will likely see several significant advances in battery thermal management reach commercial maturity. Immersion cooling appears poised for mainstream adoption, with its ability to enable faster charging, improve safety and increase energy density through tighter cell packing. Early implementations suggest potential for 40% faster charging rates and up to 22% longer battery life. The competition between immersion and flexible ribbon cooling will be worth watching. Solid-state battery commercialisation will drive corresponding innovations in thermal system design. While solidstate cells typically operate at higher temperatures than conventional lithium-ion cells, maintaining inter- and intra-cell temperature limits is still key, and warm-up is important. Their thermal management often demands tighter tolerances, favouring compact cooling systems capable of maintaining temperatures within narrow windows. Some designs may reduce cooling capacity requirements by 30% while maintaining safety margins. Sustainable materials will play an increasingly prominent role, with biobased PCMs and recyclable coolant formulations helping meet stringent environmental regulations. The European Union’s evolving sustainability mandates in particular are driving this trend, pushing developers to consider full lifecycle impacts in thermal system design. Integration will remain a central theme, both at the component level through structural cooling solutions and at the system level through tighter coupling with vehicle HVAC systems. These developments collectively point toward a future where thermal management becomes less of a standalone system and more of an intrinsic, optimised property of the pack. Acknowledgements The author would like to thank the following for their help with this feature: Michael Ruscigno, co-founder at Baknor; Mark Payne, senior manager for research, and Robert Timmis, expert in thermal management modelling at Castrol (BP); Sergio Grunder, scientist at DuPont; Boris Meisterjahn, senior manager, sales, battery systems, and Gero Mimberg, manager, thermal systems at Kautex Textron; Florian Wiedrich, head of sales, and Franz Poehn, team lead r&d, testing, simulation at Miba; Mike Tomlin, principal engineer, attributes, Scott Porteous, principal engineer, attributes, Temoc Rodriguez, GTE, electrification, and Nicholas Higginson, principal engineer, design & development at Ricardo; Gaetan Bouzard, Simcenter industry lead, battery & heavy equipment at Siemens Digital Industries Software; Francois Bordes, power thermal cluster director at Valeo. 71 E-Mobility Engineering | September/October 2025 Valeo’s modular thermal management approach encompasses refrigerant plate, coolant plate and immersion cooling systems, all exploiting heat pumps (Image courtesy of Valeo)
RkJQdWJsaXNoZXIy MjI2Mzk4