Can Acoustic Sensors Detect a Cell Vent Early?

Yes — in laboratory and battery-storage testing, acoustic and ultrasonic sensing detects a lithium-ion cell venting before temperature, voltage or gas sensors register anything, because the vent is a mechanical pressure event before it is a thermal one. Reported lead times run to hundreds of seconds. On a vehicle deck the physics still hold; the open question is coverage and ambient noise, not whether the signal exists.

Two acoustic modalities, one early signal

"Acoustic" covers two distinct methods. Passive acoustic emission listens for the sound the cell makes: as internal gas generation raises pressure, the safety vent opens with a sharp release — a click-and-hiss the sensor can pick up. Active ultrasonic works the other way, transmitting an ultrasonic pulse through the cell and reading how gas generation, swelling and density change alter its attenuation and time-of-flight. Both register a failing cell before the exothermic temperature spike, and both precede an off-gas cloud reaching a remote gas sensor.

How much earlier — the lead-time evidence

The peer-reviewed lead times are large. One ultrasonic study reported early warning of venting 282–954 seconds ahead and of thermal runaway 377–851 seconds ahead across heating powers of 600–1200 W, with ultrasonic flagging the event earlier than a force sensor — previously regarded as the most responsive channel. Passive acoustic classifiers trained on venting signatures have reported 98.6% and 99.98% classification accuracy in recent work. Where the failure is a slow overcharge, warnings on the order of two hours before failure have been reported. The consistent finding across the 2025 literature is that acoustic and ultrasonic channels lead temperature, voltage and gas.

Why the vent is a pressure event first

The sequence explains the lead time. Before a cell goes into thermal runaway, electrolyte breakdown generates gas and internal pressure climbs; the vent then releases that pressure mechanically. That pressure release and the structural changes preceding it are acoustic and ultrasonic events that happen ahead of the runaway's heat, and well ahead of vented gas migrating across a large space to a fixed detector. Temperature and off-gas detection are downstream of transport and thermal lag; the acoustic signal is at the source. This is public-domain battery physics, and it is why the modality keeps appearing in early-warning research.

What it takes to work on a vehicle deck

Almost all of this evidence comes from battery-energy-storage, EV-pack and aerospace test rigs — controlled environments with the sensor close to the cell. A car deck is the opposite: large, ventilated, and full of ambient machinery, hull and sea noise, with hundreds of vehicles between any fixed sensor and a failing cell. The engineering problems are coverage (hearing a vent metres away, not millimetres), discrimination (separating a vent from a slammed door or a dropped lashing), and environmental robustness. None of these is a physics objection; they are the reasons acoustic sensing is positioned as one input in a multi-modal stack rather than a standalone deck detector.

Where it fits in a detection stack

The credible marine path is fusion, not replacement. An acoustic or ultrasonic vent event is most valuable as corroboration — tightening the confidence and the timing of a thermal or off-gas anomaly the other channels are already forming. NASA and US Department of Energy work has pushed ultrasonic monitoring toward battery-management integration; a marine deployment would follow the same logic but demand staged-test validation on representative decks before any lead-time claim is trusted. Used that way, the modality's strength — being first and being robust to the transport lag that slows gas and heat — is additive to a detection layer already built on infrared thermal and electrochemical sensing.

What this means for engineers and class

For classification-society and shipowner engineers, acoustic and ultrasonic detection is a maturing modality with strong laboratory lead-time evidence and limited marine validation — track it, and ask for representative staged-test data rather than citing a BESS figure. For detection designers, it is a corroborating channel that can shorten time-to-alarm and add robustness, provided the noise-discrimination and coverage problems are solved for the specific space. The signal exists earlier than the alternatives; converting that into a dependable vehicle-deck alarm is an engineering and validation problem, not a scientific one.

Sources

• Han et al., "Advances in Early Warning of Thermal Runaway in Lithium-Ion Battery Energy Storage Systems," Advanced Sensor Research (Wiley, 2025) — review of acoustic/ultrasonic and conventional early-warning methods.
• "Advanced ultrasonic detection of lithium-ion battery thermal runaway under various heating powers," Journal of Power Sources (2025) — ultrasonic lead times vs force/temperature sensors. [VERIFY: venting 282–954 s / TR 377–851 s at 600–1200 W — figures taken from the abstract; confirm against the full paper.]
• Venting-acoustic-signal recognition studies, Journal of Energy Storage (2025) — Vision Transformer and SE-Net-GRU classifiers. [VERIFY: 99.98% and 98.56% accuracy — from abstracts; confirm against the full papers.]
• NASA NTRS — "Local Ultrasonic Resonance Spectroscopy of Lithium Metal Batteries for Aerospace Applications" (ultrasonic NDE for internal battery monitoring).
• US DOE / OSTI — ultrasonic characterization of Li-ion thermal-runaway conditions for real-time BMS integration.
• [VERIFY: no published, peer-reviewed validation of acoustic/ultrasonic detection on an operational ro-ro / PCTC vehicle deck was found — marine application is characterised here as emerging, not demonstrated.]