Investigating trade-offs in immersion and cold-plate cooling

Minimising degradation of cells is a key goal across all scales of research, from the atomic and electrochemical up to pack and techno-economic scales. A reduction in cell degradation comes with a choice of benefits to the end-user. A pack can be operated more agressively, yielding an improvement in performance, or fewer cells can be used within the pack, reducing cost and improving sustainability credentials. Less degradation would mean these benefits can be achieved without any reduction in pack life, thus giving improvements in performance and cost.

One method to improve the lifespan of existing cells is to improve their thermal management. It has been suggested that choosing appropriate thermal management strategies can improve the lifespan of a typical EV pack by several years [1]. Various studies have compared the relative performances of immersion and cold-plate cooling, eg. [2], which show immersion cooling strategies to yield more homogeneous cell temperatures, which should in turn lead to lower degradation. Nevertheless, links between cooling strategy and pack lifetime have not received the same attention..

This secondment investigates whether immersion cooling would lead to improvements in pack costs and energy densities. An immersion cooling system is heavier and more complex than cold-plate cooling, and would therefore be expected to increase cost and decrease energy densities. Nevertheless, reduced degradation would mean fewer cells are needed to offer the same pack lifespan. Hence, the fundamental focus of this secondment is whether the additional cost and mass of an immersion cooling system would be offset by needing fewer cells in the pack.

The project uses a distributed electro-thermal cell model, coupled to empirical ageing simulations. Through careful construction of cell boundary conditions, both immersion and cold-plate cooling can be studied. We compare the two cooling strategies by seeing how many cycles a cell is able to complete, for a simple power-based drive cycle. It is observed that the immersion-cooled cell lasts up to 40 % longer than a cell under cold-plate cooling, suggesting that the extra weight and cost of immersion cooling may well be offset by a substantial increase in cell robustness.

Further work would be needed to scale to pack-level, by comparing how few immersion-cooled cells are needed to provide the same lifespan as a cold-plate cooled pack. Nevertheless, the initial results from this project suggest that this may well be a fruitful area of research.

The project has been a valuable opportunity to share expertise across institutions. We have held rich and varied discussions about models, experiments, and parameterisation, and identified several areas for future collaboration opportunities over the coming months and years.

LinkedIn Abstract:

Mark Blyth is undertaking a 9 week staff exchange from the University of Bristol to CEA (Grenoble). His usual research focus covers the spatial couplings between electrical and thermal behaviours in cells, and he has been using these skills at CEA to investigate the effects of cell cooling strategies on their lifespan.