Quantitative Cell Classification
Article Reprints
Electric vehicle (EV) batteries are a critical element in the global transition to green energy. They must fulfill a range of quality criteria depending on the application type, such as high power density, long cycle life, and low self-discharge. However, it is also essential that these batteries consistently achieve these criteria.
A single underperforming cell can drag down the performance of a full EV battery pack, and a single defective cell may pose a safety risk. Therefore, testing of cells before they leave the gigafactory is critical. Similarly, these cells must undergo testing by automotive manufacturers before they are integrated into modules and packs.
However, testing methods may introduce errors and uncertainties, such as false negatives, where an unacceptable cell erroneously passes inspection, or false positives, where an acceptable cell is erroneously rejected. Therefore, an acceptance threshold must be chosen that balances between both errors. Avoiding false negatives is generally more critical, but the financial and environmental cost of reducing production yield through false positives in testing is not negligible either. Reducing the prevalence of false positives in classification by a small percentage has a substantial real-world impact on the efficiency of gigafactories.
This study shows how metrological electrochemical impedance spectroscopy (EIS) calibration and uncertainty analysis can improve cell classification, and hence production efficiency.