Acoustic Emission to Provide Early Warning Detection of Lithium-Ion Battery Thermal Runaway
16 affiliations
- 1Electrochemical Innovation Lab, Department of Chemical Engineering
- 2University College London
- 3Advanced Propulsion Lab, UCL East Marshgate, 7 Sidings Street, Stratford, London, UK, E20 2AE
- 4Faraday Institution, Quad One, Harwell Science and Innovation Campus, Didcot, UK, OX11 0RA
- 5School of Chemistry
- 6University of New South Wales
- 7AcouBatt Ltd, 83 Ellerton Road, London, UK, SW18 3NH
- 8University of Oxford, Holywell House, Oxford, UK, OX2 0ES
- 9Harwell Science and Innovation Campus
- 10Diamond Light Source Ltd.,
- 11Department of Engineering Science
- 12University of Oxford
- 13Stadium Business Centre, Unit 20,, Wembley, HA9 0AT
- 14Rutherford Appleton Laboratory, Harwell Campus, Didcot, Oxfordshire, UK, OX11 0QX
- 15Faculty of Science and Engineering, University of Manchester, Oxford Road, Manchester, UK
- 16University of Sydney
Abstract
Understanding lithium-ion battery (LIB) degradation pathways and failure mechanisms is essential given the rare but potentially severe associated hazards. Recent regulations on thermal runaway prevention are driving the need for advanced early detection diagnostic techniques. Acoustic methods, including Acoustic Emission (AE) and Ultrasonic Testing (UT), are promising operando monitoring tools owing to their cost-effectiveness, non-invasive nature, and real-time capability to detect physical and electrochemical changes within cells. This study demonstrates the first use of AE as a diagnostic tool for monitoring stateof-health (SoH), state of safety, and early failure onset. AE measurements were correlated with high-speed synchrotron X-ray radiography, linking acoustic signals to real-time physical changes such as gas formation, cell-volume expansion, and cracking. Experiments across a range of SoC and SoH, covering pristine, aged, and cells cycled below 0 ºC, highlight the robustness of the approach. Machine learning algorithms trained across multiple chemistries and formats enabled real-time signal interpretation. AE reliably identified degradation mechanisms and issued failure alerts across a spectrum of severity, 2 including in lithium-plated cells, with synchrotron imaging confirming the physical origin of key signals. Integrating AE into battery management systems could significantly enhance LIB safety across automotive, grid storage, and manufacturing applications without compromising cost.
Related
- Cell-Wise Fault Warning for Lithium-Ion Battery Packs Using Local Outlier Factor and Validity-Gated Extreme-Value Thresholds
- Acoustic emission-based monitoring for size-controlled robotic powder grinding.
- Detection of Thermally Induced Damage in Large-Format Lithium-Ion Batteries Using Electrochemical Impedance Spectroscopy
- First-principles investigation of penta-NiO₂ monolayer for selective detection of thermal-runaway gases from lithium-ion batteries.
- A Noise-Aware Calibrated Multi-Scale Voltage-Residual Method for Internal Short-Circuit Detection in Lithium-Ion Batteries
BibTeX
@misc{fordham2026acoustic,
title = {{Acoustic Emission to Provide Early Warning Detection of Lithium-Ion Battery Thermal Runaway}},
author = {Arthur Fordham and Inez Kesuma and Thomas Dore and Charlie Kirchner-Burles and Isabel Antony and Xiaoxia Guo and Julia Weaving and Mark Buckwell and Jian Peng and Pranay Gandharapu and Joshua Cruddos and Lara Rasha and Alexander Liptak and Stefan Michalik and Genoveva Burca and Wesley M Dose and Alexander J. E. Rettie and James B. Robinson and Thomas S. Miller and Paul Shearing and Rhodri Jervis},
howpublished = {ChemRxiv},
year = {2026},
month = sep,
doi = {10.26434/chemrxiv.15008458/v1},
url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15008458/v1},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - Acoustic Emission to Provide Early Warning Detection of Lithium-Ion Battery Thermal Runaway
AU - Arthur Fordham
AU - Inez Kesuma
AU - Thomas Dore
AU - Charlie Kirchner-Burles
AU - Isabel Antony
AU - Xiaoxia Guo
AU - Julia Weaving
AU - Mark Buckwell
AU - Jian Peng
AU - Pranay Gandharapu
AU - Joshua Cruddos
AU - Lara Rasha
AU - Alexander Liptak
AU - Stefan Michalik
AU - Genoveva Burca
AU - Wesley M Dose
AU - Alexander J. E. Rettie
AU - James B. Robinson
AU - Thomas S. Miller
AU - Paul Shearing
AU - Rhodri Jervis
PY - 2026
DA - 2026/09/07
PB - ChemRxiv
DO - 10.26434/chemrxiv.15008458/v1
UR - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15008458/v1
AB - Understanding lithium-ion battery (LIB) degradation pathways and failure mechanisms is essential given the rare but potentially severe associated hazards. Recent regulations on thermal runaway prevention are driving the need for advanced early detection diagnostic techniques. Acoustic methods, including Acoustic Emission (AE) and Ultrasonic Testing (UT), are promising operando monitoring tools owing to their cost-effectiveness, non-invasive nature, and real-time capability to detect physical and electrochemical changes within cells. This study demonstrates the first use of AE as a diagnostic tool for monitoring stateof-health (SoH), state of safety, and early failure onset. AE measurements were correlated with high-speed synchrotron X-ray radiography, linking acoustic signals to real-time physical changes such as gas formation, cell-volume expansion, and cracking. Experiments across a range of SoC and SoH, covering pristine, aged, and cells cycled below 0 ºC, highlight the robustness of the approach. Machine learning algorithms trained across multiple chemistries and formats enabled real-time signal interpretation. AE reliably identified degradation mechanisms and issued failure alerts across a spectrum of severity, 2 including in lithium-plated cells, with synchrotron imaging confirming the physical origin of key signals. Integrating AE into battery management systems could significantly enhance LIB safety across automotive, grid storage, and manufacturing applications without compromising cost.
ER -