Knee point prediction for lithium-ion batteries using differential voltage analysis and degree of inhomogeneity

• Published in: Journal of power sources 2024-11, Vol.621, p.235210, Article 235210
• Main Authors: Fuhrmann, Marion, Torcheux, Laurent, Kobayashi, Yo
• Format: Article
• Language: English
• Subjects: Degree of inhomogeneity/heterogeneity; Differential voltage analysis; Knee point prediction; Large-format cells; Lithium-ion battery
• ISSN: 0378-7753
• DOI: 10.1016/j.jpowsour.2024.235210

Lithium-ion (Li-ion) battery is increasingly recognized as a leading energy storage solution for stationary applications, promising durability and efficient energy management. Yet, a crucial challenge lies in predicting the inflection point, commonly referred to as the “Knee Point,” in the capacity trend, as it is crucial for estimating the real operational life of the system. In response to this critical issue, we are introducing a new methodology designed to anticipate the appearance of the Knee Point when operating a battery. Our approach involves detailed analysis of dV/dQ (DVA) to monitor and predict the crucial moment when the battery is likely to reach its Knee Point. Additionally, we validate this approach by conducting post-Knee Point cell disassembly studies, allowing us to observe and confirm the appearance of lithium (Li) plating. It has been applied to large-format commercial Li(NiCoMn)O2 (NMC)/graphite cells, for which the degree of inhomogeneity quantification is particularly crucial. This method therefore makes it possible to predict the risk of Li plating and the upcoming appearance of a Knee Point using the distribution of the anode capacity. [Display omitted] •Capacity fade of large-format Li(NiCoMn)O2/graphite cells is investigated.•The occurrence of a Knee Point (inflection point) due to Li plating is predicted.•Differential voltage and post-mortem analysis are used to quantify Knee Point risk.•Dispersion of graphite capacity (inhomogeneity) is estimated from the 2nd stage peak.•2nd stage peak broadening is a valuable indicator for Knee Point risk quantification.