This content is not included in your SAE MOBILUS subscription, or you are not logged in.
Development and Validation of a Reduced Order Model Incorporating a Semi-Empirical Degradation Model for Pouch Type LiFePO4/Graphite Cells
ISSN: 2167-4191, e-ISSN: 2167-4205
Published March 28, 2017 by SAE International in United States
Citation: ZHAO, X., Bi, Y., and Choe, S., "Development and Validation of a Reduced Order Model Incorporating a Semi-Empirical Degradation Model for Pouch Type LiFePO4/Graphite Cells," SAE Int. J. Alt. Power. 6(2):279-289, 2017, https://doi.org/10.4271/2017-01-1218.
We propose a reduced order model (ROM) for LFP/graphite cells derived from the electrochemical thermal principles that considers degradation effects and validated against experimental data obtained from a large format pouch type LFP/graphite cell whose nominal capacity is 20Ah. The characteristics of the two-phase transition and path dependence were taken into account in the ROM using a shrinking-core model with a moving interface that presents lithium rich and deficient phase. Different currents (0.1/1/3/4C) were applied to fresh cells at different ambient temperatures (25/35/45°C). Comparison between simulated results of the ROM and the collected experimental data shows a good match. The path dependence was also analyzed experimentally. For degradation model, side reaction is treated as the predominant cause of degradation of cells, which are affected by the operating conditions, such as temperature and SOC cycling range. At every 30 or 40 cycles, the capacity and impedance of five cycled cells were measured under different temperatures and SOC cycling ranges. The final capacity fades were 5.81%, 10.95%, 12.17%, 14.17%, and 23.54%, respectively. Degradation was accelerated by elevated temperature and high SOC cycling range. A semi-empirical degradation model was incorporated into the developed ROM. Three key parameters, the volume fraction of active anode material, the resistance of SEI and deposit layer, and the effective diffusion coefficient in the electrolyte, were extracted from experiments using the nonlinear least square method. The simulation results and experimental data are in a pretty good match.