Thermal runaway dynamics of a cylindrical 18650 Li-ion (NMC) cell: model calibration, ROM development, and emission characterization

Thermal runaway dynamics of a cylindrical 18650 Li-ion (NMC) cell: model calibration, ROM development, and emission characterization

WoS Article

Lithium-ion batteries (LIBs) are widely used in modern energy systems; however, their safety under extreme operating conditions remains a critical concern. This work presents an experimental and numerical study of the thermal runaway of a single 18650 LiNiMnCoO2 (NMC) cell with a nominal capacity of 2.6 Ah. The cell was subjected to controlled external heating until ignition, and the temperature evolution was recorded to capture the full runaway sequence. The onset temperature was approximately 190 degrees C, followed by a self-accelerating overheating phase exceeding 480 degrees C within a few seconds. A physics-based model was developed incorporating four key exothermic reactions (SEI decomposition, electrolyte-anode and electrolyte-cathode reactions, and electrolyte decomposition). The model was calibrated using experimental data and accurately reproduced the measured temperature profiles with a deviation of +/- 8 degrees C. Subsequently, it was reduced to a second-order reduced order model (ROM) using the proper orthogonal decomposition (POD) method. The ROM predicted the thermal response with an accuracy of +/- 3 degrees C while reducing computation time from 5 min to < 2 s, enabling its application in Battery Management Systems (BMS) and digital twins. Simultaneously, particles and gases released during the runaway were captured on microfilters and analyzed using scanning electron microscopy (SEM)/energy-dispersive spectroscopy (EDS). The detected elements (O, F, Ni, Mn, Cu) confirmed LiPF6 electrolyte decomposition and partial oxidation of electrodes, demonstrating the strong coupling between thermal and chemical degradation processes. The combination of experimental testing, CFD modeling, and model reduction provides a comprehensive framework for studying and predicting thermal runaway in NMC 18650 cells. The acquired data and validated models can serve as a reference basis for safety analyses and for the implementation of predictive diagnostic methods in modern battery systems.

Lithium-ion batteries (LIBs) are widely used in modern energy systems; however, their safety under extreme operating conditions remains a critical concern. This work presents an experimental and numerical study of the thermal runaway of a single 18650 LiNiMnCoO2 (NMC) cell with a nominal capacity of 2.6 Ah. The cell was subjected to controlled external heating until ignition, and the temperature evolution was recorded to capture the full runaway sequence. The onset temperature was approximately 190 degrees C, followed by a self-accelerating overheating phase exceeding 480 degrees C within a few seconds. A physics-based model was developed incorporating four key exothermic reactions (SEI decomposition, electrolyte-anode and electrolyte-cathode reactions, and electrolyte decomposition). The model was calibrated using experimental data and accurately reproduced the measured temperature profiles with a deviation of +/- 8 degrees C. Subsequently, it was reduced to a second-order reduced order model (ROM) using the proper orthogonal decomposition (POD) method. The ROM predicted the thermal response with an accuracy of +/- 3 degrees C while reducing computation time from 5 min to < 2 s, enabling its application in Battery Management Systems (BMS) and digital twins. Simultaneously, particles and gases released during the runaway were captured on microfilters and analyzed using scanning electron microscopy (SEM)/energy-dispersive spectroscopy (EDS). The detected elements (O, F, Ni, Mn, Cu) confirmed LiPF6 electrolyte decomposition and partial oxidation of electrodes, demonstrating the strong coupling between thermal and chemical degradation processes. The combination of experimental testing, CFD modeling, and model reduction provides a comprehensive framework for studying and predicting thermal runaway in NMC 18650 cells. The acquired data and validated models can serve as a reference basis for safety analyses and for the implementation of predictive diagnostic methods in modern battery systems.

Thermal runaway, 18650 Li-ion cell, NMC chemistry, CFD modeling, Reduced-order model (ROM), Particle emissions, Safety

Thermal runaway, 18650 Li-ion cell, NMC chemistry, CFD modeling, Reduced-order model (ROM), Particle emissions, Safety

VYROUBAL, P.; SEDLAŘÍK, M.; KAZDA, T.; ČUDEK, P.; PREISLER, L.; MAČÁK, M.

2026

23.01.2026

Springer Nature

Monatshefte für Chemie

1

Republic of Austria

9

https://link.springer.com/article/10.1007/s00706-025-03424-3