Lithium-ion batteries are inherently dangerous. Short circuits, mechanical abuse, battery overcharging, and design and manufacturing flaws can all result in a fire or explosion. In 2017, the Federal Aviation Administration reported that there was one lithium-ion battery fire every 10 days on planes. But lithium-ion batteries have also become the leading energy storage devices in cell phones, medical devices, electric cars and renewable power grids. That’s why understanding how and why they fail is a matter of personal and national security.
Researchers at the University of Colorado Denver and the National Renewable Energy Lab (NREL) may have found a way to make lithium-ion batteries safer by detecting failures before the batteries catch fire. Using a new, real-time algorithm, researchers will be able to estimate the charge and health of individual electrodes within the batteries. The algorithm will track electrode-level failures as they happen and provide a chance to take corrective action (e.g. shutting down the battery) before the failure becomes catastrophic.
The study was published in the journal IEEE Transactions on Industrial Electronics.
“Our biggest challenge is that no one can see what’s going on inside of a battery,” said Satadru Dey, PhD, assistant professor of electrical engineering in the College of Engineering, Design and Computing at CU Denver. “Most research on real-time battery management algorithms explored batteries from cell-level point of view. But to know specifically what’s happening inside the battery, we need to zoom in further and focus on the individual electrodes.”
Lithium-ion batteries fail fast
That’s because when a lithium-ion battery fails, it fails fast. As a battery charges or discharges, lithium ions move between the battery’s two electrodes through a gel-like, highly flammable electrolyte. The moment of failure – whether from bad design or too much heat – typically kicks off a “thermal runaway”: a build-up of heat that grows in an uncontrolled positive feedback loop until it ends with fire or an explosion, sometimes within seconds of the initial failure.
The Battery Management System (BMS) often found in cars and other larger systems uses algorithms to tell batteries how to charge or discharge, monitor how much charge is left, and sometimes track general battery health. For now, it can do little to respond to internal battery failure issues.
“Our goal is to figure out how to rig the system to take some corrective action at the first moment of fault in the battery. To achieve this, the first step is to detect the fault at a very early stage,” said Dey.
Battery safety lags behind innovation
In the study, researchers measured the voltage across the electrodes, determining the differential between the two. They then created a scheme for a real-time estimation that infers the separate electrode information, giving the researchers a sense of the capacity and charge of each electrode.
“The more information we have about the individual electrodes, the better corrective action we can take,” said Dey. “If we know the negative electrode is dying faster than the positive electrode, or which electrode faulted, we can think of corrective action that might slow down such fault. On the other hand, the more we learn about battery electrode failure mechanisms, the more likely we can build better batteries.”
For now, the estimation can detect the fault within 100 seconds, but that will need to be much faster in the future, said Dey. Eventually, the goal is to develop an advanced BMS that can trigger a sort of internal corrective action at the first moment of fault to stop the battery from going into thermal runaway conditions. It’s important to Dey, because he believes battery safety has always lagged behind innovation.
“When people talk about batteries, it’s often about how to improve efficiency, energy and power outputs,” said Dey. “But safety should always be more important. No one wants to drive around an efficient car that could explode.”