Li-ion Battery Safety Awareness via Vibration Measurements for Sustainability

Transitioning towards renewable energy has been identified by the United Nations Development Program (UNDP) and the Intergovernmental Panel on Climate Change (IPCC) as an important approach to battle climate change. One of the most efficient ways to store renewable electricity is using rechargeable batteries. With the transition to electrical vehicles for transportation and integration of large-scale batteries into power grids, the global battery market, valued at approximately $112 billion in 2021, is projected to grow more than four-fold by 2030, with Li-ion batteries leading the segment. [1] Due to their efficient charging, high discharge power, large energy density, and long cycle life, Li-ion batteries are widely used in many applications. For example, in 2023, California’s CAISO grid had 5,000 megawatts (MW) of lithium-ion battery storage. [2] Ensuring the safety of large-scale Li-ion batteries in use is crucial for sustainability.

Traditionally, voltage, current, temperature, and other electrical characteristics are monitored to assess battery performance. However, relying solely on these signals may not provide sufficient early warning due to the rapid onset of thermal runaway in Li-ion batteries. In addition, sometimes electrical signals fail to distinguish damaged batteries that are more prone to thermal runaway events. Changes in the battery state are reflected in coupled alterations in the battery's electrical, thermal, and mechanical properties. The mechanical properties of a packed battery are influenced by electrical and thermal processes and can also indicate the battery's electrical and thermal status.

This project adopts vibrational signals to monitor the health of Li-ion batteries in a nondestructive fashion, as vibration detection is widely used for health monitoring of other power equipment, such as wind turbines and generators, due to its high sensitivity, easy acquirability and clear physics interpretation. In this study, vibration signals were collected using sensors placed on the surface of the batter cells, which do not interfere with the battery's electrical or thermal operation. Vibrational signals were recorded from multiple Li-ion punch rechargeable cells during normal charging and discharging cycles, as well as during artificially induced overcharging, overdischarging and shorted states, within the frequency range of 0-23 kHz. The data was processed using Gaussian smoothing, as well as other dimension-reduction techniques, such as principal component analysis (PCA) and t-distributed stochastic neighbor embedding (t-SNE), and the deviation of battery state from a normal health condition is determined through a specifically-defined metric in a quantitative way.

It was found that the vibration spectrums of the batteries varied under different charging and discharging states. A noticeable difference was also present between healthy and unhealthy batteries. This presentation discusses the signal processing techniques used and the differences between the vibration spectrums of various battery states, focusing on frequency and amplitude.

This experimental study strongly suggests that vibration detection can be effectively implemented for battery health monitoring, providing early warning of abnormal states.

References:

1. Statista Research Department. (2023, October 19). Size of the global battery market from 2018 to 2021, with a forecast through 2030, by technology (in million U.S. dollars)[Review of Size of the global battery market from 2018 to 2021, with a forecast through 2030, by technology (in million U.S. dollars)]. Statista. https://www.statista.com/statistics/1339880/global-battery-market-size-by-technology/
2. California Independent System Operator. (2023, July 7). Special Report on Battery Storage [Review of Special Report on Battery Storage ]. https://www.caiso.com/documents/2022-special-report-on-battery-storage-jul-7-2023.pdf