Battery management systems (BMS) for advanced portable electronics (e.g. cellphones, computers), medical devices (e.g. cardiac pacemakers), and transportation (e.g. electric vehicles) are essential to ensure reliable and safe operation of Li-ion battery (LIB) packs. Specific to EV applications, the U.S. Advanced Battery Consortium (USABC) has established several targets for deployed battery stacks, including: 1) thermal stability from -30 to +52 ^oC; 2) 15 year life; and 3) 300 mile driving range per charge.¹ To achieve the long driving range, it is important to increase the energy density of operating cells, but this often comes with the risk of reduced safety and potentially shorter life. The safety of high energy density Li-ion batteries is of major concern, due to its propensity to undergo catastrophic cell failure and release of energy through thermal runaway. Therefore, the BMS plays a pivotal role in improving performance, protecting battery packs from over-charge/discharge, thermal management, state of charge (SOC) estimation, drift correction, etc. by leveraging mathematical models to conduct state-predictions and assessment.² Onboard BMS modules must conduct fast, accurate, and continuous parameter estimation to assess the state-of-health/charge/power of the battery.

Physics-based models developed by Newman’s group in the 1990’s leverage PDEs to describe the transport of charge carriers (e.g. e^-, ions, holes) and reactants across different interfaces.^3,4 Newman’s pseudo-2D (P2D) models are able to capture the complex dynamics of LIBs under a variety of environments and operating conditions (e.g. high current, varying temperature). At low to moderate currents, where the current distribution and concentration gradients are negligible, the P2D model can be simplified to a single particle model (SPM). Both P2D and SPM models require many parameters (P2D ≃40 and SPM ≃22), which can be obtained with experiments and/or data fitting to describe electrodes, separator, and electrolyte properties. The extraction of the needed modeling parameters from experiments is traditionally difficult. Additionally, the computational load can be significant, requiring fine meshes to predict the concentration and voltage gradients.⁵ Therefore, it would be advantageous to develop other modeling strategies that have a considerably reduced computational load. Lower dimension or nondimensional models typically evaluate LIBs in terms of equivalent circuits (e.g. resistors, capacitors, inductors); however, the components of the equivalent circuit models cannot be directly associated to a specific phenomenon in a LIB – making them mostly empirical in nature and unable to fully describe the system over a wide range of operating conditions.

This study investigates a simplified numerical electrochemical-thermal battery model that requires only four unknown parameters (diffusion time constant, charge exchange current, ohmic overpotential, entropic heat coefficient), which require fitting and/or experimental validation. This presentation will detail the experimental determination of the critical modeling parameters and report an extremely high accuracy in predicting the voltage and temperature profiles of a commercial 50Ah Li ion battery.

References

1. Lu, J., Li, L., Park, J.B., Sun, Y.K.,Wu, F., Amine, K. Aprotic and Aqueous Li–O2 Batteries. Chem. Rev. 114, 5611–5640 (2014).
2. Zhang, R. Xia, B., Li, B., Cao, L., Lai, Y. Zheng, W., Wang, H., Wang, W. State of the Art of Lithium-Ion Battery SOC Estimation for Electrical Vehicles. Energies 11, 1820 (2018).
3. Fuller, T. F. & Newman, J. Relaxation Phenomena in Lithium-ion-Insertion Cells. 141, (1994).
4. Doyle, M. & Newman, J. Comparison of Modeling Predictions with Experimental Data from Plastic Lithium Ion Cells. J. Electrochem. Soc. 143, 1890–1903 (1996).
5. Ekström, H., Fridholm, B. & Lindbergh, G. Comparison of lumped diffusion models for voltage prediction of a lithium-ion battery cell during dynamic loads. J. Power Sources 402, 296–300 (2018).