Electrical, thermal and life-time models are a cornerstone in today’s lithium ion battery (LIB) research and engineering. Despite significant effort has been undertaken to design and perform adequate test matrixes to cover all relevant battery operating points, the production induces variances of battery parameters and their influences on battery models have been neglected as a point of interest in academia and industry.
For this work more than 4.000 commercial 18650 cells from Sony/Murata were tested with real-time electrochemical impedance spectroscopy (RT-EIS) and for each cell an electrical equivalent circuit model was parameterized. After the initial EIS characterization 10 cells, which represent the production spreads, were chooses for cycling aging tests to evaluated the influence of production inhomogeneous on battery lifetime. The aging results were mapped back on the whole cell population.
The results show a significant variation in the circuit parameters as well as lifetime, which are led back to production inhomogeneities from different production batches and production lines. The variations of each equivalent circuit parameters were statically analyzed regarding standard deviation and mean value. Furthermore, for each cell an electrical digital twin was generated and application related performance parameter such as heat generation, voltage drop, capacity and lifetime were calculated. An deviation of more than 20 % in the battery model parameters were observed.
The results show the production induced variation in the cell performance is a significant driver for discrepancies between laboratory models and the real-life behavior of lithium-ion cells and systems. Bases on the developed measurement and assessment method relevant cells can be chosen from large production batches to consider production induced inhomogeneities for LIB characterization and modeling. The proposed RT-EIS method is fast enough to be integrated in pilot and large scale LIB production for end-of-line (EOL) testing as well as incoming good inspection for battery integrators or as initial characterization in research facilities.