(Invited) Impact of Preload Compression and Aging on High-Temperature Li-Ion Battery Pouch Cell Venting

Abstract

Listen

Translate article

With the increase in electrification, addressing safety concerns from emergency responders and the reverse logistics teams who handle Li-ion battery (LIB) packs at the end of life is increasingly urgent. Battery failures and thermal runaway (TR) present several risks, including fire, fire reignitions, explosions, and toxic gas. Moreover, in dense areas with electric vehicles (EV), e.g. parking lots and charging stations, thermal runaway (TR) risks can quickly spread to adjacent vehicles. A combination of early detection and controlled fast discharge can mitigate the hazards by limiting TR propagation and removing otherwise stranded energy from the battery pack to prevent fire reignitions. However, the effectiveness of the discharge depends on several factors including the risk of first venting, which releases flammable electrolytes and can be avoided by detecting and limiting cell expansion from gas generation at high temperatures. Understanding the impact of model mismatch due to parameter identification errors and unmodeled behavior will be important for managing venting risks and informing thresholds for control and detection, especially at high temperatures where gas generation is highly nonlinear [1]. This work uses a pouch cell venting model to represent the two-staged expansion behavior observed during an external short circuit (ESC) in a fixed-displacement fixture, enforcing similar boundary conditions to those found in an EV pack. As illustrated in Figure 1, the model considers active material thermal expansion and gas generation due to electrolyte vaporization and solid-electrolyte interphase (SEI) decomposition. Venting occurs when the internal pressure overcomes the critical venting pressure.In this work, we will first identify parameters in the venting model that relate normal operation to LIB mechanical behavior during failure and quantify the sensitivity of vent timing to the parameters identified. The preload force is one example where the model predicts that increased preload leads to earlier vent timing, which was demonstrated by Jia et al. [2]. Additionally, previous work showed that aging at different temperatures affects the onset temperature of SEI decomposition [4]. Aging at cold temperatures leads to reduced SEI thermal stability due to plated/mossy Li, while aging at high temperatures leads to increased stability due to increased thickness [4]. The SEI decomposition reaction rate can be modeled as a self-limiting, Arrhenius reaction rate such that dxSEI/dt=-ASEIxSEIexp(-ESEIkbT), where xSEI is the mole fraction of Li in SEI to negative electrode active material [3]. The impact of cell degradation on vent timing can be modeled through the parameters related to thermal stability and SEI quantity, which include the frequency factor (ASEI) and initial xSEI (xSEI,0).The simulations show that the largest sensitivities of vent timing are to parameter errors in ASEI and xSEI,0, changing the vent timing during an ESC by up to 30 seconds. The SEI decomposition parameters are theoretically related to quantities captured during aging, including electrode-specific state-of-health (eSOH), electrochemical impedance spectroscopy (EIS) resistances, and ir/reversible expansion. Together, the aging data informs an appropriate range for the parameter sensitivity study. Representative results are shown for xSEI,0 in Figure 1, where the simulations illustrate how vent timing can be affected by the initial amount of SEI. Increased SEI quantity corresponds to more gas generation and faster venting, for example, in aged cells whose resistances can double over life, assuming a proportional relationship.Ultimately, studying the parameter sensitivity of cell venting to real-world operating factors increases the feasibility of health-aware failure mitigation responses. By leveraging models, we can better understand how to set detection thresholds and manage venting risks during fast discharge as cells age to minimize the safety risks associated with end-of-life EV pack failures.REFERENCES[1] Tran, Vivian, Jason B. Siegel, and Anna G. Stefanopoulou. “Extending a Multiphysics Li-ion Battery Model from Normal Operation to Short Circuit and Venting.” Journal of The Electrochemical Society. (Submitted)[2] Jia, Zhuangzhuang, et al. "The preload force effect on the thermal runaway and venting behaviors of large-format prismatic LiFePO4 batteries." Applied Energy 327 (2022): 120100.[3] Hatchard, T. D., et al. "Thermal model of cylindrical and prismatic lithium-ion cells." Journal of The Electrochemical Society 148.7 (2001): A755.[4] Börner, M., et al. "Correlation of aging and thermal stability of commercial 18650-type lithium-ion batteries." Journal of Power Sources 342 (2017): 382-392.Figure 1: Parameter sweep of xSEI,0, where Δσcell is the change in fixture compression stress due to cell expansion with contributions from gas generation (Δσgas) and thermal expansion (Δσthermal). Venting occurs when the cell’s internal pressure overcomes the critical venting pressure. Figure 1