Impact of Relaxation Time on the Accelerated Cycle Ageing Tests of Lithium-Ion Batteries

Abstract

As lithium-ion batteries are becoming the energy storage technology of choice for transport electrification, the accurate estimation and prediction of their service life remains a critical aspect for their successful adoption and integration. The established methods of lifetime estimation employ battery degradation models along with accelerated ageing test data. Although accelerated methods have been widely used in the literature over the last decade, there are still some discrepancies regarding the testing methodology and parameters. Specifically, in cycle ageing experiments, the relaxation time between cycles is chosen arbitrarily and is typically kept short to enable higher cycle throughput. Although this parameter is considered to have minimal influence on the test results1, its impact on the ageing rate and cell cycle life has not been thoroughly investigated. Furthermore, previous studies demonstrated significant capacity recovery after prolonged rest periods (≥2 days)2, however, their contribution is usually considered only at the end of the cycling ageing test3.In this work, the impact of cycling relaxation and rest periods in cycling ageing tests is systematically investigated. Commercially available cylindrical lithium-ion cells were subjected to long-term cycling at 10ºC with different discharge relaxation times. During cycling, three rest periods, i.e., 1min, 10min and 60min were investigated for the discharge relaxation steps, while the charge relaxation time was kept constant for all the cells at 1min. The cells were cycled in the manufacturer’s full voltage window with 0.3C/1C charge/discharge current respectively, and three cells were used in each test case to ensure the reproducibility of the results. At regular intervals, reference performance tests (RPT) were performed at 25°C for cell characterisation. After 6 months of ageing, a rest period of 2 weeks was incorporated into the ageing test, in order to investigate the influence of reversible losses in the course of accelerated cycling. During the rest period the cells were fully discharged and remained stored in open-circuit condition at 10°C. An RPT was conducted prior to the start as well as at the end of this period, before the cycling ageing regime was restarted.The relative capacity fade and resistance increase for the different relaxation times after approximately 8 months of ageing are presented in Figure 1. As illustrated, there is a clear increasing trend between cell degradation and relaxation time, i.e., cells that are rested for 60min after discharge are exhibiting the fastest rate of capacity fade and resistance increase. This result is counter-intuitive, since the longer the relaxation period the less cycles are performed between successive RPTs. Furthermore, the rest period is spent at low SOC, minimising the calendar ageing contributions, which are already expected to be comparatively small, due to the sub-ambient test temperature (10ºC). This behaviour was investigated further through differential voltage analysis (DVA), which revealed that the degradation modes across the three test cases exhibited minor differences, with higher losses of active material observed for the cells cycled with 60min relaxation and higher degrees of lithium inhomogeneity observed for the cells cycled with short relaxation (1min, 10min). The intermediate rest period led to significant capacity recovery for all cells, which reached up to >10% of the apparent capacity fade measured prior to the cycling break in the short relaxation test cases. The resistance values however were not affected, remaining almost constant before and after the rest period. The reversible losses were attributed to the anode overhang, due to the very low storage SOC, and to the increase in homogeneity of lithium distribution due to the increase in the peak intensity of the DVA curves. The capacity recovered through these mechanisms was effectively utilised once the ageing cycles were restarted and the ageing rates of each test case remained consistent, indicating that the cycling break did not affect the cell ageing trajectory.The results of this work demonstrate the strong influence of relaxation time on the cycle life of lithium-ion cells, which are cycled in otherwise similar conditions. This issue, which has not been reported before, needs to be considered in the design and interpretation of accelerated ageing tests. Similarly, reversible capacity losses need to be accounted and quantified in order to avoid the underestimation of cell cycle life. M. Reichert, D. Andre, A. Rösmann, P. Janssen, H. G. Bremes, D. U. Sauer, S. Passerini, and M. Winter, Journal of Power Sources, 239 45-53 (2013). B. Epding, B. Rumberg, H. Jahnke, I. Stradtmann, and A. Kwade, Journal of Energy Storage, 22 249-256 (2019). M. Lewerenz, P. Dechent, and D. U. Sauer, Journal of Energy Storage, 21 680-690 (2019). Figure 1