Enhancing incremental capacity and differential voltage diagnostics for lithium-ion battery cells connected in parallel

Voltage profile diagnostic techniques such as Incremental Capacity Analysis (ICA) and Differential Voltage Analysis (DVA) are powerful tools that can extract degradation modes directly from cell voltage data using measured half-cell voltage profiles and simple models. As multiple individual cells are connected in parallel, cell-to-cell variations and cell interconnection resistances will alter the current distribution and measured voltage, blending and smoothing features in the voltage profile, impacting degradation diagnostics. This work experimentally investigates the application of voltage profile diagnostics on 16 NMC811/Graphite 21700 cells connected in parallel. Experimental results show that whilst the influence of cell-to-cell variation is small for unaged cells, the influence of cell interconnection resistance creates inhomogeneous current distribution between parallel cells which can significantly reduce the accuracy of ICA and DVA, even at low C-rates of C/25, due to reduced definition of degradation features caused by a distribution of phase transitions between cells. This effect is most significant in ‘U-shape’ parallel strings where the positive and negative connections are located close to the same cell compared to ‘Z-shape’ where the connections are at opposite ends. A numerical model is presented and validated to quantify inhomogeneous current distribution at low C-rates. A simple voltage fitting process is introduced and validated to compensate for these effects by conducting a resistance based voltage shift prior to fitting half-cell voltage profiles to obtain a corrected low C-rate voltage profile, improving the accuracy of degradation diagnostics in parallel cells. • Impact of parallel connections on voltage diagnostics experimentally investigated • Blending of voltage signatures seen at C/25 due to inhomogeneous current • Busbar resistance and external connections main cause of inhomogeneity • Numerical model proposed and validated to study inhomogeneous distribution • Novel method for correcting parallel voltage blending proposed and validated