Physical Model-Based State of Health Understanding of Severely Aged Lithium-Ion Batteries Under Real-World Automotive Operational Life
Claudio Rabissi,Gabriele Sordi,Andrea Casalegno
DOI: https://doi.org/10.1149/ma2023-022416mtgabs
2023-12-22
ECS Meeting Abstracts
Abstract:Lithium ion-battery (LIB) technology, featuring upstanding energy and power density, satisfying lifetime, high round-trip efficiency and very fast dynamics in a reasonably economic package, rapidly became the undisputed ruler of portable power and it is now the main driver of the electrification of transportation sector. However, despite the fully commercial development, understanding and predicting degradation of such devices is still a great challenge for the scientific and technical community, especially when dealing with real-life-operation induced aging. The reusability of such devices in a circular economy perspective is a hot topic to the sector to improve sustainability, but requires understanding “how” batteries are faded rather than only “how much” they are, to enable a physically consistent and second-life-related estimation of residual lifetime. In the present activity, a detailed analysis of high-power LFP (lithium iron phosphate) cells samples operated on IVECO hybrid buses are performed to assess their possible reusability, in the frame of a joint research cooperation. A large batch of cells with different ages (from 9 years old to brand new) and different positions inside the modules are investigated. State-of-art electrochemical diagnostics are performed embedded in a multi-measurement optimized protocol (full discharge, pulse test, electrochemical impedance spectroscopy) developed after a sensitivity optimization for a model-based improved parameter identification [1], as visible in Figure 1. Residual performances are analyzed and compared with limited on-board data collected from vehicles BMS and analyzed by means of an appositely-improved physical modelling platform. Both capacity and power capabilities have sensibly decreased over the cells lifetime (see Figure 2 for the full discharges at 0.1C, 25°C) with a clear pattern attributable to cells position beneath the modules, prompting the importance of a homogeneous thermal management during operation. Power-based state of health of the samples is higher than a capacity-based SoH, foreseeing a possible reuse of the samples in a power-intensive second-life. Residual performance interestingly feature a high consistency with data recently published in [2], relative to similar system despite operated in a completely different geographic area; this strengthens the generalizability of the results. Aged cells diagnostics were interpreted with a previously developed Newman P2D physical model provided with heath exchange, to identify the degradation mechanisms through the EIS-based parameter identification procedure [1] involving both thermodynamic and kinetic aspects. Interpretation of thermodynamic aging modes has been conducted on all the tested cell samples, based on 0.1C thermodynamic analysis and the differential voltage. Incremental capacity curves of aged cells have been reproduced with a PSO-based (particle-swarm-optimization) identification of three ageing parameters: loss of lithium inventory (LLI) , mainly corresponding to solid electrolyte interphase (SEI) growth, and loss of active materials at negative (LAMn) and positive (LAMp) electrode, indicating loss of actives sites for the electrochemical reactions. Despite the severely low residual capacity, the model reproduces satisfyingly the features of the aged cells. Interestingly, a linear trend could be identified as common beneath all the samples involved in the analysis, indicating common aging mechanism appearing with different magnitude. As visible in Figure 3, the trend points out a two-steps aging path: (I) a first pathway mainly leading to LLI close to 20%, followed by (II) a second pathway with both LLI and LAMn occurring at the same time. In the literature, similar observations have been already performed by some authors in previous works, with a lower extent of degradation [3,4]. They agree in this interpretation: the first stage is usually associated to the growth of the SEI layer, while the second stage is usually associated to the onset of lithium plating (due to a dense SEI which inhibits the lithium intercalation into graphite). Confirmation of such interpretation is obtained by means of model based interpretation on non-equilibrium measurements, permitting estimation of physical parameters value and enabling identification of aging mechanisms, together with ex-situ measurements based on electrochemical and morpho chemical analyses. Additionally, several consistencies between on-line measurements, such as EIS, and equilibrium capacity loss of cells have been identified, theoretically discussed and tested under further accelerated aging tests, enabling possible strategies of SoH implementation based on fast and informative electrochemical measurements. References [1]Rabissi et al., doi.org/10.1016/j.est.2022.106435 [2] K. Ramirez-Meyers et al., doi.org/10.1016/j.est.2022.106472 [3] E. Sarasketa-Zabala et al., doi.org/10.1021/jp510071d [4] X.G. Yang et al., https://doi.org/10.1016/j.jpowsour.2017.05.110 Figure 1