Study on Thermal Runaway Propagation Characteristics of Lithium Iron Phosphate Battery Pack under Different SOCs

Overview

The study asks how state of charge (SOC) changes the way thermal runaway propagates through an LFP battery pack. Three 42 Ah prismatic LFP cells (about 917 g each) were placed in a row with a 1500 W copper heating plate against cell 1, and temperatures, heating rates, mass loss and venting behaviour were recorded at 100%, 50% and 0% SOC.

At 100% SOC the first safety valve ruptured after 1791 s of heating with the plate at 265 °C, and runaway followed about 6 s later. The runaway interval was 7 s from cell 1 to cell 2 and 37 s from cell 2 to cell 3. Compared with 50% SOC, the cell 1→2 interval at 100% SOC was 71.4% shorter and the cell 2→3 interval 87.2% shorter. At 100% SOC the runaway neighbour supplied 71.6% of the heat absorbed by cell 2 and 89.9% of that absorbed by cell 3 (68.6% and 77.9% at 50% SOC), so most of the heat pushing downstream cells into runaway came from the runaway neighbour rather than the external heater.

The 100% SOC pack ejected sparks when the safety valve ruptured; the 50% and 0% SOC packs only produced smoke, and the 0% SOC pack's overall temperature rise rate stayed below 1 °C/s. Under the same SOC, runaway behaviour of cells 1 and 2 was more intense than cell 3, with maximum temperature, temperature rise rate and mass change rate ordered cell 1 > cell 2 > cell 3.

Limits: one cell type, only three cells and a one-sided heating plate; thermal barriers, cooling and larger cell arrays in real modules were not examined. The conclusion is that pack fire and explosion risk rises with SOC, and runaway is more intense closer to the heat source.

Key points

1. Test cells were 42 Ah prismatic LFP cells (about 917 g each), arranged three in a row with a 1500 W copper heating plate against cell 1.
2. At 100% SOC the first safety valve ruptured after 1791 s of heating, with the plate at 265 °C, and runaway followed about 6 s later.
3. At 100% SOC the runaway interval was 7 s from cell 1 to cell 2 and 37 s from cell 2 to cell 3.
4. Compared with 50% SOC, the cell 1→2 interval at 100% SOC was 71.4% shorter and the cell 2→3 interval 87.2% shorter.
5. At 100% SOC, the runaway neighbour supplied 71.6% of the heat absorbed by cell 2 and 89.9% of that absorbed by cell 3 (68.6% and 77.9% at 50% SOC).
6. The 100% SOC pack ejected sparks when the safety valve ruptured; the 50% and 0% SOC packs only produced smoke, and the 0% SOC pack's overall temperature rise rate stayed below 1 °C/s.

Using it in design

• In inter-cell thermal barrier design for battery modules, base the design on the thermal contribution of the runaway neighbour (70-90%); neighbour-to-neighbour heat, not an external source, drives propagation.
• Consider limiting SOC (for example during storage or maintenance) as an operating control for propagation risk; at 0% SOC the temperature rise rate stayed below 1 °C/s.
• Take spark ejection at high SOC into account in fire and explosion protection design (vent direction, distance to combustible material).

Who should read it

Engineers assessing safety and thermal runaway propagation of LFP battery packs, UPS batteries and BESS.

Document contents

1. 1. Introduction
2. 2. Experiment (2.1 Battery, 2.2 Experimental Device and Battery Modules Description, 2.3 Experiment Setup)
3. 3. Results and Discussion (temperature and safety venting comparisons, 3.2 Quality Analysis of TR and Its Propagation, 3.3 LFP Battery TR Behavior Analysis)
4. 4. Conclusions

Related standards

• UL 9540A
• NFPA 855

Read the document

Publication details

Authors

Minghao Zhu; Jiajie Yao; Feiyu Qian; Weiyi Luo; Yin Chen; Luyao Zhao; Mingyi Chen

Publisher

Electronics (MDPI AG)

Year

2022

Volume

12

Issue

1

Pages

200

Document language

English

Source type

Scientific

Access

Open licence — downloadable

DOI

10.3390/electronics12010200

Licence

CC BY 4.0 — share with attribution

Licence evidence

www.mdpi.com

Citation