Open Access

Matteo Ligorati, Jerome Geyer-Klingeberg, Timon E. Günther, Andreas W. Rathgeber

• The transition toward sustainable energy systems requires reliable and durable energy storage technologies, with lithium-ion batteries (LIBs) being central to electrification and the integration of renewables. However, the decline in capacity per cycle, referred to as the degradation rate (DegRate), decreases the lifetime and performance of LIBs, thereby limiting their ecological and economic benefits. To address this, we conducted the first meta-analysis of LIB degradation, drawing on 146 studies and 917 effect sizes. The analysis accounts for the heterogeneity in reported DegRates using explanatory variables grouped into battery differences, experiment-, measurement-, and publication-specific categories. Across studies, we found a median DegRate of 0.04%/cycle, with cut-off charge voltage and temperature emerging as the dominant influencing factors. Using meta-regression, we quantify the effects of these explanatory variables. Furthermore, we establish a forward-looking quantitativeThe transition toward sustainable energy systems requires reliable and durable energy storage technologies, with lithium-ion batteries (LIBs) being central to electrification and the integration of renewables. However, the decline in capacity per cycle, referred to as the degradation rate (DegRate), decreases the lifetime and performance of LIBs, thereby limiting their ecological and economic benefits. To address this, we conducted the first meta-analysis of LIB degradation, drawing on 146 studies and 917 effect sizes. The analysis accounts for the heterogeneity in reported DegRates using explanatory variables grouped into battery differences, experiment-, measurement-, and publication-specific categories. Across studies, we found a median DegRate of 0.04%/cycle, with cut-off charge voltage and temperature emerging as the dominant influencing factors. Using meta-regression, we quantify the effects of these explanatory variables. Furthermore, we establish a forward-looking quantitative benchmark: under extreme cold (0 °C) and very high charge cut-off voltages, model-implied mean DegRates for graphite-based systems reach 0.68–1.41%/cycle (14–29 cycles). For promising Si-based chemistries, the benchmark is more prospective with model-implied mean DegRates of 0.98–1.71%/cycle (11–20 cycles), with higher uncertainty in these sparsely covered regimes, as reflected in the confidence intervals. This study highlights critical gaps in the experimental matrix and, rather than relying on a single estimate, establishes quantitative benchmarks for LIB degradation that serve as a reference for future research, particularly for combinations of operating conditions that have not yet been experimentally explored.…