Rapid State-of-Health (SoH) Determination and Second-Life Grading of Aged Automotive Battery Modules Via Electrochemical Impedance Spectroscopy (EIS)

Abstract

The number of electric vehicle (EV) batteries reaching end-of-service is set to increase from thousands to tens of thousands per annum by 2025. These end-of-service batteries typically retain significant capacity and power delivery capability, and their re-use in so-called ‘second-life’ applications has been proposed as a means to extend the battery value chain, offset a portion of their up-front cost and minimise waste by deferring recycling. Accurate grading of used battery assets is essential in order to assign them an appropriate valuation in the second-life market. These grading activities account for the majority of the cost incurred in repurposing batteries at present, due largely to the time required for testing, which typically employs DC techniques taking up to several hours to perform a capacity measurement. Rapid grading methodologies are required to reduce processing time and achieve the levels of throughput required to handle increasing quantities of end-of-service batteries. Electrochemical Impedance Spectroscopy (EIS) is a promising technique for determination of SoH, offering results in minutes instead of hours. However, EIS measurements are affected by extraneous factors such as temperature, state-of-charge and the experimental setup, and do not provide a direct measurement of capacity. The deployment of EIS for battery grading therefore requires a careful approach in order to eliminate noise factors and ensure robustness, and necessitates modelling in order to interpret output in terms of capacity. Here, we present a methodology for rapid grading of aged automotive Li-ion batteries using EIS, and demonstrate its application to battery modules from the 1st generation Nissan Leaf. An empirical model has been developed, which is capable of inferring the residual capacity of an aged module from measurements of its EIS spectrum, open-circuit voltage and temperature made within 3 minutes using a Solartron impedance analyser. The model was derived from data gathered using a design-of-experiments approach, employing numerous used battery modules covering a range of states-of-health. EIS measurements were recorded under different conditions in order to understand and correct for the influence of key noise factors. We demonstrate that the SoH of a vehicle battery can be determined in minutes using EIS, instead of hours using traditional DC techniques, with no loss of accuracy, repeatability or reproducibility. The process is now being implemented for grading of battery modules at Nissan’s Sunderland plant in the UK, and significantly reduces the time required for complete characterisation of a used battery pack. Reduction of characterisation time in this way affords a corresponding reduction in repurposing costs, thus improving the viability of battery re-use as part of the circular economy. This work was undertaken within the UK Energy Storage Laboratory project funded by the UK Government Department for Business, Energy and Industrial Strategy (BEIS) and in conjunction with Nissan Motor Manufacturing UK and Element Energy.