Cell-to-cell variation beyond parameter analysis — Identification and correlation of processes in Lithium-Ion Batteries using a combined distribution of relaxation times analysis

Cell-to-cell variations inside a battery pack can result in inhomogeneities, leading to an accelerated, uneven aging. Until today these variations are solely quantified by parameter-based studies, which only offer a limited interpretability. To tackle this inconclusiveness, cell-to-cell variations and their effects on the electrode processes of lithium-ion batteries are investigated in this work. For this purpose 92 cells are characterized by impedance, pulse, and pseudo open-circuit voltage measurements, and the data obtained is compiled as publicly available data set. The data is examined by a holistic distribution of relaxation times analysis, comprising complementary distribution of relaxation times approaches. This enables a reliable identification of process variations between cells. First, the confidence intervals of both the raw data and the distribution of relaxation times are examined. Subsequently, the individual processes, their characteristic time constants, and their polarization are determined. The analysis reveals that a normal distribution only applies in one of the examined cases and can even be excluded in four cases. Finally, a correlation analysis is conducted, allowing the categorization of the identified processes into cell winding, electrochemical interface processes, and solid-state diffusion. The potential physicochemical reasons behind are discussed and the work thus contributes to a deeper understanding of cell-to-cell variations.