Performance degradation of lithium-ion cells is inevitable during their application. Hence, state of health (SoH) estimation is crucial for battery management, especially in electric vehicles (EV), where the demands on the cell performance are more stringent. Conventional SoH estimation is often based on the cell capacity or impedance. Nevertheless, online capacity and impedance measurements are rarely achievable owing to the diverse and dynamic operation conditions. An alternative indicator should replace capacity and impedance to make SoH estimation more manageable.

In our work, an aging test matrix, including calendar and cycle life aging, is carried out on a commercial lithium-ion pouch cell. The stress factors are temperature (25 °C, 40 °C, 55 °C) and discharge rate (1C, 3C, 5C). Performance tests are conducted every two weeks to measure the actual capacity of the tested cells in a climate chamber at 25 °C. The performance test consists of two continuous constant current (CC) and then constant voltage (CV) charge/discharge cycles with 1C (5 A). The cut-off voltages are 2.7 V and 4.2 V with a cut-off current of C/20 (0.25 A). During the performance tests, a temperature sensor is attached to the surface center of each cell to monitor the surface temperature evolution.

During the analysis of the surface temperature evolution from the performance tests, a strong cooling effect is detected in the 1C CC charge process after about 0.4 h charging. The dT/dt curve of a new cell is shown in Fig. 1, along with its temperature evolution curve, to identify the position of the maximum cooling effect, i.e. the dT/dt minimum. Then, t₀ is defined as the period from the charge beginning to the dT/dt minimum during 1C CC charging. In Fig. 2, a linear correlation between t₀ and the cell actual capacity is noted with a coefficient of determination (R²) of 0.93. This correlation is analyzed for the aforementioned cells aged under both calendar and cycle life conditions. Furthermore, various stress factors have little influence on the linear correlation behavior of the cells. Accordingly, t₀ could be employed as a new indicator for SoH estimation of lithium-ion cells.