Using Reference Electrodes to Validate Parameterization of Individual Electrode Soh in the Single Particle Model with Electrolyte

Sravan Pannala,Everardo Olide,Jason Siegel
DOI: https://doi.org/10.1149/ma2023-01251671mtgabs
2023-08-28
ECS Meeting Abstracts
Abstract:Most model parametrization techniques for reduced order physics-based Lithium-ion battery models like the single particle model rely on minimizing the error in the measured and predicted cell terminal voltage. The battery terminal voltage is the difference between the voltages of the positive and negative electrode solid phase potentials. If the individual contribution of each electrode overpotential to the terminal voltage, is not measured, there could be multiple combinations of parameters that provide similar RMSE for voltage prediction, yielding non-unique solutions and local minima. Misattributing the parameters for kinetics and diffusion in each electrode could result in increased plating or over-conservative designs at higher operating currents. Adding a reference electrode to the battery allows us to measure the individual electrode potentials experimentally. Using this extra measurement, we can parametrize the kinetic and diffusion reactions in positive and negative electrodes with better confidence [1], and validate the electrode-specific state of health parameterization, specifically the increased resistances due to SEI growth and shift in the stoichiometric window as the cell ages [2,3]. In our experiments, we used a pouch cell manufactured in the U of M Battery Lab with Targray NMC 622 Single Crystal cathode powder, Superior SLC 1520-T anode powder, and a custom fabricated Lithium Iron Phosphate (LFP) reference electrode deposited on a porous electrode layer. Further details of the reference electrode are given in [4]. The model was parametrized using discharge data at different C-rates and a modified Hybrid Pulse Power Characterization test. Finally, the RMSE for both the modeled negative and positive electrode potentials are calculated for the validation data set which uses the US06 Drive cycle. References: [1] Zhengyu Chu et al., Journal of Energy Storage, Volume 27 (2020). [2] Suhak Lee et al 2020., J. Electrochem. Soc., 167, 090531 (2020). [3] Hannah M. Dahn et al., 2012 J. Electrochem. Soc. 159, A1405 (2012). [4] Taylor R. Garrick et al., J. Electrochem. Soc. 168, 010530 (2021).
What problem does this paper attempt to address?