Understanding battery degradation is essential for accurate health diagnostics and long-term performance prediction. Degradation mode analysis (DMA), a non-destructive technique, leverages the voltage response, which reflects the cell’s thermodynamic equilibrium, to dissect aging mechanisms . Since DMA relies on accurate thermodynamic data, open circuit voltage (OCV) measurements provide the best representation of the cell’s intrinsic state. However, obtaining true OCV requires extended rest periods or low charge rates, conditions achievable in controlled laboratory environments but impractical in real-world applications. To approximate OCV under practical constraints, pseudo-OCV measurements taken at slightly higher C-rates are often used, though these introduce a resistance component that can distort the interpretation of degradation mechanisms .
This study addresses this challenge by quantifying the influence of resistance on DMA. Using cycle-aging data from two commercially significant lithium-ion cells, LG M50T 21700 (an energy cell) and Molicel P45B (a power cell), with NMC811 cathodes and C/SiOx composite anodes (notably higher silicon in the Molicel), we explore how resistance affects voltage-based indicators of degradation in energy versus power cells. Cells are cycled across the full state-of-charge (SoC) window, with reference performance tests (including C/10 pseudo-OCV and resistance measurements) at defined charge throughput intervals. By performing DMA on the C/10 pseudo-OCV both with and without resistance correction, we isolate resistance’s impact on observed degradation modes.
An additional complexity arises from hysteresis due to the SiOx component in the anode, causing disparities between charge and discharge profiles that evolve differently with aging . Ideally, a cell without hysteresis would exhibit identical charge and discharge potentials under OCV, yielding consistent DMA results. However, hysteresis introduces deviations, particularly in cells with prolonged cycling. This study quantifies the evolution of hysteresis with aging and assesses its impact by separately analyzing charge- and discharge-based DMAs, a focus not explicitly addressed in prior research.
To identify diagnostic metrics minimally impacted by resistance and hysteresis, this analysis also includes differential voltage and capacity diagnostics. By rigorously quantifying these effects, the study contributes to a more precise understanding of degradation pathways, enhancing the predictive power of DMA and extending its applicability for real-world battery health monitoring.