Fast Screening and Sorting of Commercially Available Second-Life Batteries from Former Mobility Applications for Construction of Small Energy Storage Systems

Lithium-ion batteries are the most used cell technology in electric mobility applications, because of their high capacity, energy density, and cycle stability. Many cells retire with a remaining State of Health of 80%, at the same time, forecasts show an increasing availability of up to 200 GWh/y of retired batteries in 2030. Often, these cells are directly recycled while their performance is still sufficient for less demanding applications, without giving a battery cell a second life (SL). This paper evaluates pre-sorted commercially available cells for SL applications reclaimed from former mobility applications. These cells are presorted based on a single ac resistance measurement with a frequency of 1 kHz. Since a single measurement provides little information for later use and complete charging and discharging cycles are time-consuming, the cells are classified for later use by a fast-screening method. This paper makes use of hybrid pulse power characterization (HPPC) for feature extraction. Clustering is performed with the k-means algorithm. The benchmark is provided by 143 cells; an additional 254 cells were screened with the use of a fast HPPC micro cycle. Finally, the paper provides researchers with an open-source dataset of HPPC cycles of 397 cells with different states of health.