(Invited) Lifetime Predictive Model of Capacity Loss, Resistance Increase, and Irreversible Thickness Growth
Sravan Pannala,Hamidreza Movahedi,Anna Stefanopoulou,Jason Siegel
DOI: https://doi.org/10.1149/ma2024-012514mtgabs
2024-08-09
ECS Meeting Abstracts
Abstract:We present a battery lifetime model that predicts Irreversible battery thickness change under constant pressure operation as the battery ages, along with capacity loss and resistance growth. These metrics are very important for predicting lithium-ion battery state of health (SOH) and remaining useful life due to their interdependence with battery packaging and safety. To design and operate battery storage systems safely, we must be able to predict and detect changes in all three SOH metrics over life accurately and simultaneously. Traditionally, these three SOH metrics have been modeled separately. However, we developed a single model that simultaneously predicts battery capacity, resistance, and expansion over its lifetime. In this work, battery degradation is modeled using the single particle model (SPM) framework, including mechanical damage in the electrodes, which results in loss of active material (LAM), and the side reactions for SEI growth and Li plating, which result in loss of lithium inventory (LLI). We introduce an equivalent stress and concentration dependence of stress to the mechanical damage model. Commonly used fatigue models assume symmetric cycles (zero-mean stress) [1]. To account for cycling conditions with different currents on charge and discharge, an equivalent stress that adjusts for non-zero mean stress is needed. The concentration dependency of strain and the resulting stress allows us to capture different degradation rates for cells cycled at different depths of discharge (DOD). Lithium plating at the separator interface is often described as a spatially non-uniform process using the pseudo-2D (P2D) model. This phenomenon can be captured in the SPM framework using a current-dependent dynamic term to account for spatial-nonuniformity of electrolyte dynamics at higher currents. Experimental data from cells cycling with periodic reference performance tests were used to tune the degradation model. In addition to the standard I, V, and T measurements, continuous measurements of the thickness change of the battery pouch cells were recorded. From this data, we extracted the capacity, electrode state of health (eSOH), resistance, and reversible and irreversible expansion. The test conditions were designed to excite different dominant degradation mechanisms (e.g. mechanical damage) by cycling the cells at different conditions (C-rate, temperature, preload pressure, and DOD). The tuning, which uses accelerated aging simulations [2] (to speed up the tuning process), results in a single set of parameters that predicts capacity loss, resistance growth, and irreversible thickness change as the battery ages, even for conditions that were not used for parameter tuning. The tuning is performed in two sequential steps. First, the parameters related to the electrochemical and mechanical degradation rates, which result in LLI and LAM, are tuned to match the extracted eSOH variables at each Reference Performance Test (RPT). The first tuning gives us the amount of plated Li, SEI growth, and mechanical damage in the particles. We then relate these quantities to irreversible expansion using a linear relationship in the case of SEI and mechanical damage and quadratic in the case of Li plating [3]. The same relationship is used for all cells, and this represents the 2nd tuning step. Our data set and calibration are unique in that the dominant degradation mode is the loss of lithium inventory due to the loss of active anode material. Most prior work on lifetime predictive models relied on SEI growth as the dominant degradation mode. Figure 2 shows the tuned model prediction of the three performance metrics over the cycle life of cells for various cycling conditions. The markers represent the measurement at each reference performance test, and the lines correspond to the model. References: [1] R. Burger and Y.-L. Lee, “Assessment of the mean-stress sensitivity factor method in stress-life fatigue predictions,” Journal of Testing and Evaluation , 2013 [2] S. Pannala et al. Methodology for Accelerated Inter-Cycle Simulations of Li-ion Battery Degradation with Intra-Cycle Resolved Degradation Mechanisms, American Controls Conferenc e, 2022 [3] S. Pannala et al. Consistently Tuned Battery Lifetime Predictive Model of Capacity Loss, Resistance Increase, and Irreversible Thickness Growth, Journal of the Electrochemical Society , 2023 Figure 1