Experimental investigation of parametric cell-to-cell variation and correlation based on 1100 commercial lithium-ion cells

In order to prevent battery modules from inhomogeneity during operation, the integrated battery cells should be matched. Therefore, the cell-to-cell variations of relevant cell parameters due to manufacturing tolerances need to be quantified. Regarding cell matching, three points are lacking in current literature: first, reported parameter analyses have either based their statistic analysis on a too small number of cells. Or, second, few different cell parameters have been considered - hence, it was not possible to discuss which parameter would be beneficial to be used for classification. Third, potentially sensible correlations between different cell parameters have not been determined and discussed adequately. This paper provides a unique combination of analysing multiple different cell parameters and discussing their correlation - both based on a large number of examined cells: we investigate the parametric cell-to-cell variation and correlation of 1100 commercial production fresh LiFePO4-graphite cells which originate from two batches. The cell parameters are experimentally determined by conducting DC check-up (CU)- and AC EIS-measurements under monitored temperature and relaxation conditions for all 1100 cells. The data is statistically analysed for 15 different parameters: different discharge capacities, different DC and AC impedances, the mass and mean temperature of the cells. The determined relative variation of capacity and impedance is small: 0.28% and 0.72% respectively, which corresponds to a variation ratio of 1:2.2. The variation of the cell impedance allows no conclusion about the variation of the cell capacity. From the results, we derive four major implications concerning recommended characterisation parameters for the development and modelling of battery modules as well as for the quality control during cell production. Our experimentally determined parametric variation values and drawn conclusions are valuable for model fittings and battery pack analyses which have up to now been based on assumptions about cell-to-cell variations. The data set for all 1100 cells and 15 parameters is provided as supplementary material [1].