Two-minute warning for battery fires

Researchers at the US National Institute of Standards and Technology (NIST) have developed an algorithm capable of detecting the sound of a lithium-ion battery cell overheating. This technology provides a critical early warning system against catastrophic battery fires.

Using a convolutional neural network, the NIST researchers built a software package that can accurately identify the characteristic noise of an overheating lithium-ion battery cell with 94% accuracy.

The research builds on prior studies of “venting acoustics” in lithium-ion batteries, which identified the sound of a rupturing safety valve with 92% accuracy. The venting noise is a distinct “click-hiss” sound, emitted when a battery’s safety valve ruptures, typically about two minutes before ignition. The solar industry has, similarly, explored sound analysis for early detection of inverter malfunctions.

Video Source: NIST

For the study “Development of a Robust Early-Stage Thermal Runaway Detection Model for Lithium-ion Batteries,” the NIST researchers collaborated with Xi’an University of Science and Technology, in China, to conduct 38 “thermal abuse tests” on single-cell lithium-ion batteries.

Those experiments simulated extreme conditions, such as overheating, to observe the progression of thermal runaway events. The study focused on 18,650 lithium-ion batteries, commonly found in laptops and electric vehicles. Each battery tested featured a graphite anode, a LiNiCoAlO₂ cathode, a nominal capacity of 3.2 Ah, and a voltage of 3.7 V.

The team meticulously recorded the sounds generated during each thermal runaway event, focusing on the distinctive “click-hiss” sound of the safety valve rupture. By studying the audio characteristics of the 38 recorded events, researchers identified unique acoustic patterns that reliably indicate an impending thermal runaway.

To ensure the algorithm could distinguish those sounds from background noise, the team applied advanced machine-learning techniques. They trained a convolutional neural network using a diverse dataset that included the battery sounds recorded and augmented samples. By tweaking the pitch and speed of the recordings, they created more than 1,000 unique audio samples to simulate a variety of real-world scenarios.

“We tried to confuse the algorithm using all kinds of different noises, from recordings of people walking, to closing doors, to opening Coke cans,” said Wai Cheong “Andy” Tam, one of the lead researchers. Despite these challenges, the algorithm achieved a 94% success rate in identifying the safety valve rupture sound, making it a promising candidate for integration into early warning systems.

In the image above, the acoustic signals from Experiment 30 show how the algorithm distinguishes different sounds from the critical event of a safety valve rupture. The top graph (3a) displays various noises: whispering, flipping a light switch, moving the camera, closing a door, using a hammer, and the ignition of a battery. The safety valve rupture is characterized by its high amplitude and prolonged oscillation, making it distinguishable from the sharp, singular peaks of a flipping switch and the more complex oscillations of hammering.

The researchers have submitted their device for a patent. They are also exploring refinements such as testing with different battery types and incorporating alternative microphones to enhance the system’s accuracy and versatility.

The need for advanced battery fire detection systems has become increasingly urgent, especially after high-profile incidents including a 2019 battery fire in Arizona. In that event, a shipping container filled with batteries exploded, injuring eight firefighters and a police officer. The explosion rendered the entire team unconscious, with one firefighter being thrown 73 feet by the force of the blast.

One safety innovation developed in response to such incidents is remote “off-gassing” detection, which allows gasses to be vented from battery storage containers before they can ignite. In the NIST experiments, the off-gassing process is visible in the video after the safety valve ruptures. This build up of gasses was the primary cause of the Arizona explosion when the container was opened.

From pv magazine USA.