Thermal runaway modeling of lithium-ion batteries at different scales: Recent advances and perspectives

Gao J., Qin Z., Zhao G., Liu Y., Zhang W., Yao H., Zheng Y., Lin Y., Huang Z., Li J.

High-energy-density lithium-ion batteries (LIBs) based on LiNixCoyMn1-x-yO2 cathodes necessitate a cost-effective and straightforward electrode modification technique to enhance both lithium storage capacity and thermal safety performance in industrial settings. This paper introduces an economically viable, convenient, and industrially feasible approach: the integration of minute quantities of TiO2 and TiN additives into the LiNi0.6Co0.2Mn0.2O2 (NCM622) cathode, aimed at enhancing the performance of NCM622-TT||Graphite pouch full batteries, particularly in terms of lithium storage and thermal safety. Remarkably, under testing conditions of 45 °C and 1C rate, the TiO2&TiN-modified pouch full battery demonstrated a capacity retention rate of 82.6 % after 1000 cycles, representing a substantial improvement of over 53 % compared to the pure NCM622|| graphite pouch full battery. Furthermore, needling tests revealed a notable reduction of 12.4 °C in the average temperature increase during thermal runaway in the modified LIB, indicating a significant enhancement in safety. Comprehensive characterization and mechanistic analysis suggest that the incorporation of TiO2 and TiN enhances interfacial compatibility and stability, mitigates side reactions during cycling, and indirectly improves ion transport kinetics. Unlike traditional approaches involving electrode material modifications and battery structural design, modulating the NCM electrode offers a promising avenue for advancing the research and development of nickel-rich NCM LIBs, with considerable practical implications for industrial applications.

Khan S.A., Hussain I., Thakur A.K., Yu S., Lau K.T., He S., Dong K., Chen J., Xiangrong L., Ahmad M., Zhao J.

Battery energy storage systems (BESS) are essential for integrating renewable energy sources and enhancing grid stability and reliability. However, fast charging/discharging of BESS pose significant challenges to the performance, thermal issues, and lifespan. This paper provides not only an overview of the recent advancements of battery thermal management systems (BTMS) for fast charging/discharging of BESS but also machine learning (ML) approach to optimizing its design and operation. Various thermal management strategies are highlighted in this review, such as liquid-based, phase-change material-based, refrigerant-based, and ML-based methods, offering improved thermal performance and better safety for fast charge/discharge applications. Overall, this paper provides a comprehensive and critical analysis of the current advancements and prospects of BESS thermal management and identifies the current research gaps and future directions for developing a more efficient and reliable BESS.

Liu M., Zeng Z., Wu Y., Zhong W., Lei S., Cheng S., Wen J., Xie J.

In the past few decades, rapidly advanced lithium‒ion batteries (LIBs) technologies have revolutionized our lives by powering portable electronic devices and transportation tools. But surged risks pertaining thermal runaway (TR) of LIBs have brought adverse concerns for their further applications, especially in grid‒scale energy storage. As the blood of LIBs, electrolytes serve as the “initiator and accelerator” of substance‒energy conversion reactions triggering TR. Therefore, executing the functionalized design for electrolytes to cut off these reactions have been recognized as a critical solution to mitigate TR. However, due to the lack of clarifying intricate relationship between mentioned reactions and TR, the targeted design of functional liquid electrolytes (LEs) is difficult in making effective progress. Herein, this review, for the first time, analyzes the affiliation‒mechanisms, while summarizing achieved progress in functional LEs to enhance LIB safety. Meanwhile, the review puts forward the design principles of functional LEs to aiming at each type of unfavorable substance‒energy conversion reactions.

Adam M.L., Moses O.A., Mailoa J.P., Yu Hsieh C., Yu X., Li H., Zhao H.

Investigating the role of electrodes' physiochemical properties on their output voltage can be beneficial in developing high-performance batteries. To this end, this study uses a two-step machine learning (ML) approach to predict new electrodes and analyze the effects of their physiochemical properties on the voltage. The first step utilizes an ML model to curate an informative feature space that elucidates the relationship between physiochemical properties and voltage output. The second step trains an active learning model on the informative feature space using Bayesian optimization to screen potential battery electrodes from a dataset of 3656 materials. This strategy successfully identified 41 electrode materials that exhibit good electronic conductivity and host highly electronegative anions. This work provides an efficient strategy to discover novel electrode materials while integrating domain knowledge of chemistry and material science with ML in materials research.

Wang G., Gao W., He X., Peng R., Zhang Y., Dai X., Ping P., Kong D.

The recently emerged cell-to-chassis (CTC) technology tremendously raises the energy density of the battery pack by directly integrating lithium-ion batteries into the chassis fame, while it also brings more safety concerns involving thermal runaway propagation (TRP). However, the TRP behavior and the relevant prevention measures for the CTC battery system have not been comprehensively studied yet. In this work, a three-dimensional model was developed within the framework of OpenFOAM to investigate the TRP behavior in CTC battery packs. The heat generation was described by the electrochemistry reaction kinetics fitted in stages and the heat balance was used to address the propagation of TR. This model captures the evolution process of TRP in CTC system and five typical heat transfer modes between the TR and normal cells are identified according to simulation results. The subsequent numerical study indicates that enhancing heat dissipation can reduce the demand on critical thickness of the thermal insulation layers to inhibit TRP. However, local failure of thermal insulation can initially lead to small-scale TR and further contribute to a larger heat accumulation, putting forward higher standards for critical parameters to cease TRP. As the thickness of insulting layers increases to 2 mm, TRP can be completely ceased even when local failure of insulation occurs. This work enhances the understanding of TRP mechanism and its prevention, which can serve as new guidelines for the safety protection of rapidly developing non-module battery pack technologies.

Karmakar A., Zhou H., Vishnugopi B.S., Mukherjee P.P.

The thermal safety of lithium‐ion (Li‐ion) batteries continues to remain a critical concern for widespread vehicle electrification. Under abuse scenarios, thermal runaway (TR) of individual energy‐dense Li‐ion cells can be dominated by various exothermic mechanisms due to interelectrode crosstalk, resulting in an enormous heat generation response that can potentially lead to thermal runaway propagation (TRP) in a battery module. Herein, a hierarchical TRP analytics approach is developed, which includes cell‐level thermokinetic and electrode crosstalk interactions derived from accelerating rate calorimetry characteristics of a representative high‐energy 18650 cylindrical Li‐ion cell with Ni‐rich cathodes and Si–C anodes. The hierarchical TRP model, coupled with multimodal heat dissipation, demonstrated for an exemplar energy‐dense Li‐ion battery module configuration, determines TRP criticality at module level for a wide range of conditions, including ambient temperature, intercell spacing, trigger cell location, external heating power, and heat dissipation coefficients. Potential propagation pathways have been identified, and their underlying attributes in terms of propagation speed, heat release from exothermic reactions, critical thermal energy input, and heat dissipation to surroundings have been quantified. This hierarchical approach, including thermal transfer and chemical interelectrode crosstalk during TR, can provide high‐resolution TRP analytics for energy‐dense Li‐ion battery modules and is scalable to packs.

Chen L., Bao X., Lopes A.M., Xu C., Wu X., Kong H., Ge S., Huang J.

The estimation of the state of health (SOH) of lithium-ion batteries (LIBs) is of great significance to ensure the safety and reliability of the battery management system. Equivalent circuit model (ECM) and data-driven based methods are commonly used to estimate the SOH. Each method has pros and cons, but combining them is challenging. In this paper, a new approach integrating ECM and data-driven methods is proposed for SOH estimation. Firstly, the internal resistance of a first-order ECM of the LIB is identified using particle swarm optimization (PSO). Secondly, a fractional-order three-learning strategy PSO is adopted to optimize a back-propagation neural network (BPNN). Finally, the internal resistance of the ECM, voltage, current and time of the LIB are used as input to the optimized BPNN to predict the SOH. Different battery datasets from NASA and CALCE are used to verify the effectiveness of the proposed technique. The results show that the maximum root mean square error (RMSE) of the new method does not exceed 1.35%, and the error of the best SOH prediction is just 0.39%. Moreover, the highest and lowest prediction interval coverage probability (PICP) are 100% and 85.71%, respectively. Compared with other approaches, the proposed method reveals faster convergence speed, superior accuracy, and better generalization ability.

Song Y., Wang L., Sheng L., Zhang M., Liang H., Ren D., Cui H., Zhang H., Xu H., He X.

High-energy lithium-ion batteries (LIBs) are growing in developing and adoption, but are associated with a rapid capacity fading as well as a high risk of thermal runaway. Apart from the decay of electrode materials, electrolyte and interphases, the imperceptible interaction between electrodes, i.e., crosstalk, is emerging as a critical contributor to the failure of high-energy battery. In this perspective, we pioneer in summarizing and commenting on the imperceptible phenomena. Firstly, the origins of crosstalk are accountable to the release of transition metal ions and singlet oxygen species from the cathode, as well as to the oxidation or reduction of the electrolyte at electrodes. Subsequently, chemical and electrochemical approaches, in-situ spectroscopies and numerical simulations that are accessible for crosstalk detections are outlined. Furthermore, the crosstalk suppression strategies are classified into three aspects including the modification of electrode materials, design of favorable electrolytes, and fabrication of functional separators. Moreover, the essential role of crosstalk in life predictions and safety alerts through machine-learning and artificial intelligence is highlighted. Focusing on electrode crosstalk, this work is expected to be instrumental in enriching the failure theories of high-energy batteries while improving their survivability and predictability.

Park S., Mallarapu A., Lim J., Santhanagopalan S., Han Y., Choi B.

Owing to the complexity of coupling mechanical and electrical solvers for finite element analysis (FEA) of mechanical abuse of a battery, the simulation should be performed by incorporating a local short circuit model activated based on the status of battery elements. However, the approach should also consider various other conditions of the battery simulated, such as the initial state of charge (SOC), for precise simulations of the internal short. This study proposes an approach developed with a mechanical model and single-particle battery model, along with a reaction kinetics model for thermal abuse using LS-DYNA. Several key parameters were selected, such as the activation criteria, kinetic abuse and element integration points during the internal short circuit induced by the indenter. The impact of these key parameters in the FEA on thermal runaway response for various initial SOCs from 25to 100 % were analyzed. The FEA results obtained were compared against experimental results with various parameters such as activation criterion of internal short, kinetic abuse based on accelerating rate calorimetry (ARC) data. Finally, the root causes of the discrepancy between the FEA and experimental results are discussed regarding integration point what solver can make difference during calculation.

Peng R., Ping P., Wang G., He X., Kong D., Gao W.

Large-scale Energy Storage Systems (ESS) based on lithium-ion batteries (LIBs) are expanding rapidly across various regions worldwide. The accumulation of vented gases during LIBs thermal runaway in the confined space of ESS container can potentially lead to gas explosions, ignited by various electrical faults. However, a systematic simulation and assessment of the battery vented gases explosion under deflagration venting design still lack. In this work, a three-dimensional combustion model was developed within the frame of open source computational fluid dynamics code OpenFOAM based on a full-scale container, and the LIBs vented gases in realistic proportion were selected as the combustion gas. Coupled boundary conditions were introduced to enable the response of explosion vent doors and top deflagration vent panels on pressure. The internal and external overpressure, flame temperature, and wind velocity fields were employed to assess the gas explosion hazards to ESS container structure and surroundings. The results demonstrate that altering the vent door pressure, without the top vent panel, still leads to serious explosion accidents. There will be unacceptable overpressure for the container structure, as well as serious visible flames and high-speed airflow invading the external environment. Lower vented gas concentrations can reduce explosion hazards, and introducing the vent panel design aids to promote such reduction. The overpressure within the container is significantly decreased by guiding the top external secondary combustion through the vent panel, and the influence range for the environments is also significantly reduced on X-axis. The ignition location can affect the propagation of gas combustion within the container, and the proposed complete vent panel design minimizes the impact on the container structure and surroundings. The findings of this work can offer new references for the safety design of the LIBs ESS.

Wang G., Ping P., Peng R., Lv H., Zhao H., Gao W., Kong D.

Thermal runaway (TR) and the resulting fire propagation are still critical issues puzzling the application of lithium-ion batteries in energy storage system (ESS). A fire propagation model including accurate TR propagating process assists in understanding the battery failure mechanism and determining the safety-optimal design of ESS, while its development is hindered by the complexity of simulating large-scale spatial system and interactions between TR and fire. In this work, a coupled semi reduced-order model (SROM) toward real-scale ESS is developed to capture battery TR and fire propagation behavior. Wherein, meshless methods are implemented for battery cluster by constructing thermal resistance network to simulate heat generation and transfer, which simultaneously couples a mass flowing network to address gas generation and subsequent jet. Full-order CFD model is adopted to simulate burning behavior in external fluid with higher precision. This model can accurately capture cross-scale parameters, including temperature evolution at cell-level and heat release rates (HRR) at system-level, as confirmed by experiments. Simulation results elucidate the failure propagation mode and mechanism from cell-to-cell to module-to-module levels. The significant impact of triggering position on fire behavior is also revealed that TR originating from the cluster center causes rapider fire growth and larger peak HRR during fire propagation. The SROM covers entire phenomena chain from cell-level to system-level, which can serve as new guidelines for designing and running safer ESS.

Koenig B.C., Zhao P., Deng S.

With the increasing demand for applications in vehicles and energy storage systems, lithium-ion batteries have attracted extensive research interest. Thermal runaway in these batteries, which can lead to rapid temperature rise, flammable gas release, and even fires, directly imposes safety concerns. Previous studies have utilized differential scanning calorimetry (DSC) to infer thermal kinetic mechanisms for individual battery components. However, the commonly used Kissinger analysis involves oversimplified assumptions which lead to erroneous quantification of the kinetic parameters and inaccurate physical interpretation of the results. Here, we propose an extension of the recently developed Chemical Reaction Neural Network (CRNN) framework that is not limited by the simplifications from traditional optimization methods and can learn complex, multi-step thermal kinetics from DSC data. After a proof of concept, we leverage this novel approach to improve recent thermal decomposition models for nickel–cobalt–manganese oxide (NCM) cathodes. Its generality and enhanced learning capability enable improved models with better agreement to the data that account for the actual physical coupling between multi-step reaction pathways. The successful development of the CRNN approach to learn such thermal kinetic models from simple DSC data demonstrates its potential to advance thermal runaway modelling for lithium-ion batteries and other complex kinetic systems.

Zhang P., Lu J., Yang K., Chen H., Huang Y.

A coupled simulation model of the 18650 lithium-ion batteries (LIB) thermal runaway (TR) is presented in this study, which includes TR decomposition reaction, gas generation and combustion processes, solid particles ejection and particles heat transfer process. The model considers a combination of solid heat conduction, gas convection, flame and particles radiation, which is validated to accurately capture the temperature evolution and two typical jet processes during TR. Model validation conducts with experimental measurements for the temperature of the battery surface. The simulation results show that the “ignition” time of the flammable gas mixture is delayed and the rate of flame temperature increase is slowed down when the effect of solid particles is considered. The radiation heat transfer rate and convection heat transfer rate on the adjacent battery surfaces are 80.3W and 12.76W, and are approximately 4.29 times and 1.76 times larger than the results calculated by the model without solid particles, respectively. The difference between these two can be further magnified in confined space. The model developed in this study combines the effect of the solid particles’ radiation into the TR simulation innovatively. It can provide a more accurate calculation method for the prediction of TR propagation.

Li Q., Li R., Yang Z.

A computational fluid dynamics (CFD) solver for a GPU/CPU heterogeneous architecture parallel computing platform is developed to simulate incompressible flows on billion-level grid points. To solve the Poisson equation, the conjugate gradient method is used as a basic solver, and a Chebyshev method in combination with a Jacobi sub-preconditioner is used as a preconditioner. The developed CFD solver shows good performance on parallel efficiency, which exceeds 90% in the weak-scalability test when the number of grid points allocated to each GPU card is greater than 2083. In the acceleration test, it is found that running a simulation with 10403 grid points on 125 GPU cards accelerates by 203.6x over the same number of CPU cores. The developed solver is then tested in the context of a two-dimensional lid-driven cavity flow and three-dimensional Taylor-Green vortex flow. The results are consistent with previous results in the literature.

Kim M., Jeon J., Hong J.

To diagnose and elucidate thermal runaway accompanying gas evolution of a lithium-ion battery, it is essential to understand the thermal side reactions that lead to thermal runaway inside a lithium-ion battery. It is very useful to make a reliable model that represents these reactions to analyze thermal runaway processes in order to secure battery safety and overcome high costs of large-scale experiments. This study proposes the reaction mechanism and the reaction model through the design of experiments with the combination of battery components such as a cathode, an anode, an electrolyte, and a separator. To develop the reaction mechanism, the peak temperature and calorific value of each reaction are obtained by using a differential scanning calorimeter. The change of mass and produced gas from each reaction are identified by using an online thermogravimetry-mass spectrometer. Based on these measurements, the reaction model is developed by estimating kinetic parameters obtained from the Kissinger analysis. The reaction model exhibits root-mean-square-error of 1.91 mW, 21.79 mW, and 4.53 mW in the electrolyte, the cathode and the anode, respectively, as compared to differential scanning calorimeter results, confirming its high fidelity. The proposed model illustrates the variation of volume fractions of each phase inside a lithium-ion battery to simulate electrochemical performance degradation during thermal runaway stage. The change in internal pressure is also evaluated by using the change in mass and volume of each phase. Based on the mechanism and model derived from this study, it is possible to pinpoint the electrochemical performance degradation and heat generation characteristics during thermal runaway.

Li M., Hu H., Lu L., Zhang H.

Lou Z., Huang J., Su Z., Zhang D., Wei X., Yao H.

Cheng Y., Gao D., Zhao F., Yang Q.

Mama M., Solai E., Capurso T., Danlos A., Khelladi S.

Peng R., Kong D., Ping P., Gao W., Wang G., Gong S., Yang C., Gao X., He X.

Guan J., Chen M.

Park S., Lee H., Choi S., Lim J., Kim S., Song J., Ali M., Kwon T., Doh C., Lee Y.M.

Ibrahim O., Abdul Aziz M.J., Ayop R., Dahiru A.T., Low W.Y., Sulaiman M.H., Amosa T.I.

Considering the rapidly evolving microgrid technology and the increasing complexity associated with integrating renewable energy sources, innovative approaches to energy management are crucial for ensuring sustainability and efficiency. This paper presents a novel Fuzzy Logic-Based Particle Swarm Optimization (FLB-PSO) technique to enhance the performance of hybrid energy management systems. The proposed FLB-PSO algorithm effectively addresses the challenge of balancing exploration and exploitation in optimization problems, thereby enhancing convergence speed and solution accuracy with robustness across diverse and complex scenarios. By leveraging the adaptability of fuzzy logic to adjust PSO parameters dynamically, the method optimizes the allocation and utilization of diverse energy resources within a grid-connected microgrid. Under fixed grid tariffs, the investigation demonstrates that FLB-PSO achieves grid power purchase and battery degradation costs of $1935.07 and $49.93, respectively, compared to $2159.67 and $61.43 for the traditional PSO. This results in an optimal cost of $1985.00 for FLB-PSO, leading to a cost saving of $236.09 compared to the $2221.10 of PSO. Furthermore, under dynamic grid tariffs, FLB-PSO incurs grid power purchase and battery degradation costs of $2359.20 and $64.66, respectively, in contrast to $2606.47 and $54.61 for PSO. The optimal cost for FLB-PSO is $2423.86, representing a cost reduction of $237.23 compared to the $2661.08 of PSO. The FLB-PSO algorithm proficiently manages energy sources while addressing complexities associated with battery storage degradation. Overall, the FLB-PSO algorithm outperforms traditional PSO in terms of robustness to system dynamics, convergence rate, operational cost reduction, and improved energy efficiency.

Huang Y., Zhao Y., Bai W., Cao Y., Xu W., Shen X., Wang Z.

Lai M., Lv J., Ge X.

It is essential to ensure thermal safety in lithium-ion (Li-ion) batteries to facilitate their increased use in electric cars, thereby enhancing the safety of individual batteries. When high-power Li-ion batteries are subjected to abusive conditions, they frequently experience thermal runaway (TR) due to a sequence of exothermic reactions resulting from electrode contacts. These events generate a significant amount of heat and have the potential to cause thermal runaway propagation (TRP) throughout the battery module. This paper presents an in-depth TRP analysis approach, combining thermo-kinetic and electrode interaction assessments at the cell level. The method is implemented on the TR properties of a cylindrical Li-ion cell equipped with nickel-rich and silicon-carbon electrodes. The developed TRP approach, incorporating various heat dissipation techniques, is applied to a prototype high-energy Li-ion chamber to evaluate TRP susceptibility under different conditions, including ambient temperature, cell spacing, initiating cell location, external heating power, and heat dissipation rates. Additionally, this research provides insights into the various TRP pathways, focusing on aspects such as the propagation rate, the exothermic processes' thermal, required thermal energy input, and external heat release. The comprehensive model, considering both heat transfer and chemical interactions between electrodes under TR, can address accurate TR- propagation examinations for batteries. Moreover, it can be expanded to larger battery systems, ultimately improving overall battery safety.

Kong D., Wang G., Wen J.X.