A Fast and Automatic Methodology for Selecting Lithium-ion Cells for Second-Life Batteries

The increasing presence of electric vehicles (EVs) and hybrids (VEHs) characterizes the current urban mobility scenario. The key component of these vehicles - the battery - operates for about a decade in this application. After this period, they are usually removed from the EVs, as they no longer supply the energy demand required for this application. When this happens, the batteries still have about 80% of their rated capacity available, allowing them to be reused in several other applications (second-life). This work proposes the development of a method for the evaluation and selection of cells for different second-life applications based on tests and analyses carried out on batteries previously used in electric vehicles. The main contributions of this work are: (a) the study of fast, efficient, and well-established measurements in the literature, such as internal resistance (DC) and Electrochemical Impedance Spectroscopy (EIS) as parameters to estimate the degradation state of second-life LiBs; (b) design and analysis of an extensive database containing information on EIS, RI and capacity for 395 Lithium-Iron-Phosphate (LFP) cells; (c) study of the possibility of implementing Machine Learning techniques to correlate the ohmic measurement data with the real capacity of the cells. The experimental results demonstrated that EIS measurements present a strong correlation with the residual capacity of second-life cells, and can be applied as a rapid way of identifying the degradation state of these cells.