• The issue of how long a battery pack will remain in useful operation, to guarantee a warranty for an electric vehicle battery, is a challenge faced by most car manufacturers, as they haven’t been around long enough for them to know their full lifetime
• Working as part of an APC6 funded project, WMG at the University of Warwick developed a battery model capable of forecasting the battery health for realistic usage cases
• Researchers at WMG experimentally tested the batteries for several usage cases and the model predicted the evolution of battery health with 98% accuracy

Extensive research, carried out by researchers from WMG, University of Warwick included characterisation, performance, safety and degradation testing of an EV battery, with degradation being the key focus to understand the impact on battery pack warranty.

To predict the degradation of the battery capacity researchers had to build a complex battery model, incorporating the key physics that causes batteries to age. The model was capable to predict scenarios under which a battery health will gradually fade and fulfil the warranty requirements and scenarios where the battery health will suddenly decrease after a certain usage duration.

Knowing when a battery’s health, can suddenly decrease, known as the ‘knee-point’ effect is a hotly studied problem among lithium-ion battery researchers. Researchers found that, for the particular battery investigated, by avoiding deep discharges and reducing the number of fast-charging a week, enabled the battery to perform and last the expected lifetime of the battery pack.

The aged cells were subsequently disassembled and examined for failure evidences. Fully discharging the battery at different rates demonstrated a thin film to deposit on the electrode and deform the electrode causing the active material in the batteries to delaminate. These effects can bring about a sudden reduction in the battery health.

Dr Dhammika Widanalage, from WMG, University of Warwick comments:

“We worked with the project partners to understand the needs of the battery ageing model and the usage scenarios. Using the data from our laboratories, we were then able to calibrate our model and also predict the voltage response and capacity fade of the battery under new usage cases with a very high accuracy of around 98%.”