Cell Cooling Coefficient, thermal gradients and their implications on degradation

Summary:

Active cooling of Li-ion battery packs is a demanding aspect of current electric vehicles. The correct surface to maintain safe and optimal temperatures to reduce degradation is not always an obvious choice. In this work we combine experimental studies on degradation and thermal capability of different cooled surfaces, with discretised 3-D electro-thermal models to highlight the difference between external and internal cell thermal gradients and the subsequent effects on critical performance and degradation.
The CCC is a thermal metric for a battery cell that defines the amount of heat that can be rejected through specific surfaces for a given thermal difference between the cooling surface and cell maximum. A 5 Ah Kokam pouch cell was found to have a CCCtab of 0.332 W/K under tab cooling, and a CCCsurf of 0.987 W/K under single sided surface cooling. These CCC values indicate that surface cooling is 3 times better to maintain safe operating temperatures.
Our previous experimental work compared the rate of degradation of the 5 Ah Kokam cells when under tab and dual-sided surface cooling. The tabs and surfaces of the cells were held at 20°C using Peltier elements. This study found that at high rates (2C charge, 6C discharge) surface cooling resulted in a significant reduction in available capacity, and a rate of degradation 3 times higher than tab cooling.
To investigate the cause of this accelerated degradation under the two cooling schemes, a 3-D electro-thermally coupled ECN model was developed. The model was validated and the accelerated degradation was successfully reproduced. The validated model enables access to experimentally unattainable information, such as internal temperatures and states of charge. The model predicts that surface cooling leads to lower average temperature within the cell than tab cooling, but a much higher internal temperature gradient.
The CCC correctly predicted that the maximum surface temperature for tab cooling would be hotter than surface cooling. However, the model predictions also alert to the fact that internal thermal gradients within the electrode stack are higher in surface cooling, and that it is the electrode stack thermal gradient that is the most critical one for degradation, not the surface-measured thermal gradient. These results highlight the fact that accounting for intra-cell inhomogeneities (I, T, SOC, etc.) is critical to correctly modelling and predicting degradation.
Cylindrical cells are another format of Li-ion cells that has significant industry relevance. For this reason, the CCC for base cooled cylindrical cells has also been defined. An LG M50 cell has been found to have a CCCbase of 0.139 W/K, approximately 60% lower than the tab cooled pouch cell. To determine if this will translate to significant internal gradients and degradation rates, a similar 3-D electro-thermal model has been developed for this cylindrical cell. This model has been further validated for the CCC experiment itself, and was shown to be able to capture the behavior under a wide range of operating regimes, including the high frequency pulsing and extreme boundary conditions of CCC testing.
For a 2C fast discharge under base cooling, with an external can gradient of approximately 25°C, the model predicts an internal jellyroll gradient of only ≈2°C. As predicted by the poor CCC value, the external gradient is extremely high in a high-rate discharge. However, the model indicates that despite this, the internal gradient is comparatively low, thus may result in a moderate level of degradation. This work highlights the importance of modelling cell level internal gradients when attempting to determine degradation.