Experimental and Reduced-Order Modeling Research of Thermal Runaway Propagation in 100 Ah Lithium Iron Phosphate Battery Module

Overview

The study asks how thermal runaway propagation (TRP) in energy storage batteries can be predicted far more efficiently than with conventional finite-element models. A four-cell module of 100 Ah, 3.2 V nominal, 130 × 36 × 211.8 mm prismatic LFP cells at 100% SOC (with 5 mm mica sheets and a heating plate at one end) was heated at one end to trigger runaway, and propagation was measured with thermocouples, voltage and mass loss. The heat-transfer problem was then converted into a state-space equation using reduced-order theory for linear time-invariant systems, giving a reduced-order model (ROM) based on the Arnoldi method for Krylov subspaces.

Experimental results: cell 1 vented at 558 s of heating and went into runaway at 762 s. The LFP cells did not burn or explode, but every cell went into runaway in turn, each losing about 20% of its mass. Venting phases (vent opening to runaway) lasted 204 s, 399 s, 156 s and 309 s for cells 1-4. Maximum cell surface temperatures were 452.4-539.2 °C, and runaway reached the back face 77-118 s after the front face, an internal propagation speed of about 0.12 mm/s.

The ROM ran in 8 min 54 s versus 328 min 49 s for FEM (about 40× faster), with critical-feature errors within 10%. In the model, raising the runaway trigger temperature from 190 to 200 °C, cutting total heat release from 100% to 90%, raising convective h from 15 to 35 W·m^-2·K^-1, or adding an interface thermal resistance of hR = 100 W·m^-2·K^-1 each suppressed propagation. Of a cell's runaway energy, more than 70% heats the cell itself, over 20% leaves with vent gases, about 2% passes to the next cell, 1% to the previous cell and about 1% is lost by convection and radiation.

Limits: one cell type, only 100% SOC, a single four-cell module and one heating configuration; the model's parameter adjustment analysis is not validated for real pack designs. Because runaway progressed without flame, vent-gas flammability and explosion are outside this study.

Key points

1. Cells: 100 Ah LFP, 3.2 V nominal, 130 × 36 × 211.8 mm, all tested at 100% SOC; the module had 5 mm mica sheets and a heating plate at one end.
2. Cell 1 vented at 558 s of heating and went into runaway at 762 s; the LFP cells did not burn or explode, but every cell went into runaway in turn, each losing about 20% of its mass.
3. Venting phases (vent opening to runaway) lasted 204 s, 399 s, 156 s and 309 s for cells 1-4.
4. Maximum cell surface temperatures were 452.4-539.2 °C; runaway reached the back face 77-118 s after the front face, an internal propagation speed of about 0.12 mm/s.
5. The ROM ran in 8 min 54 s versus 328 min 49 s for FEM (about 40× faster), with critical-feature errors within 10%.
6. Raising the runaway trigger temperature from 190 to 200 °C, cutting total heat release from 100% to 90%, raising convective h from 15 to 35 W·m^-2·K^-1, or adding an interface thermal resistance of hR = 100 W·m^-2·K^-1 each suppressed propagation in the model.
7. Of a cell's runaway energy, more than 70% heats the cell itself, over 20% leaves with vent gases, about 2% passes to the next cell, 1% to the previous cell and about 1% is lost by convection and radiation.

Using it in design

• In LFP module design, evaluate inter-cell thermal resistance (interface insulation) together with measures that cut heat release to stop propagation; in the model an interface resistance of hR = 100 W·m^-2·K^-1 alone suppressed it.
• Run the propagation parameter sweep with the reduced-order model and verify only the selected design with detailed FEM or testing.
• Use the finding that even flameless LFP runaway takes the whole module into runaway in turn, with about 20% mass loss, as input to vent-gas management and thermal insulation design.

Who should read it

Engineers and researchers working on battery module and BESS safety and thermal runaway propagation modelling.

Document contents

1. 1. Introduction
2. 2. Methodology (2.1 Battery Sample, 2.2 Geometry, 2.3 Reduced-Order TRP Model: reduced-order thermal model, TR model, boundary condition, modeling procedures; 2.4 Experimental Design: EV-ARC test, module TRP test)
3. 3. Results and Discussion (3.1 Experimental Result, 3.2 Simulation Result: model validation, model prediction of characteristic location, model parameters' adjustment analysis, simulation efficiency compared with FEM; 3.3 Energy Flow Analysis: energy flow composition, energy flow distribution characteristic)
4. 4. Conclusions

Related standards

• UL 9540A
• NFPA 855
• IEC 62619

Read the document

Publication details

Authors

Han Li; Chengshan Xu; Yan Wang; Xilong Zhang; Yongliang Zhang; Mengqi Zhang; Peiben Wang; Huifa Shi; Languang Lu; Xuning Feng

Publisher

Batteries (MDPI AG)

Year

2025

Volume

11

Issue

3

Pages

109

Document language

English

Source type

Scientific

Access

Open licence — downloadable

DOI

10.3390/batteries11030109

Licence

CC BY 4.0 — share with attribution

Licence evidence

www.mdpi.com

Citation