Rapid Sorting and Health Assessment of Retired EV Battery Packs

Introduction#

The rapid growth of electric vehicles has created a pressing challenge: what to do with the massive volume of retired battery packs that still retain 70–80% of their original capacity. These packs are prime candidates for echelon utilization in stationary energy storage, backup power, and low-speed vehicle applications. However, the wide variation in remaining useful life (RUL) and cell-to-cell consistency makes direct reuse risky and inefficient. Without accurate health assessment and sorting, second-life batteries can suffer from accelerated degradation, safety hazards, and poor economic returns.

This case study examines a real-world project where a fleet of retired lithium-ion battery packs from commercial electric buses was processed for echelon utilization. The core technical challenge was to evaluate remaining life and consistency quickly and accurately, using capacity fade models and internal resistance evolution laws. The solution combined rapid sorting, state-of-health (SOH) assessment, and regrouping into homogeneous modules. The results demonstrate how a systematic approach can turn retired packs into reliable, cost-effective energy storage assets.

Capacity Fade and Internal Resistance Evolution#

The first step was to characterize the aging behavior of the retired packs. Capacity fade was modeled using a semi-empirical approach that combines calendar aging and cycle aging. For each pack, historical data—including charge/discharge cycles, temperature exposure, and depth of discharge—was used to fit a degradation curve. The model predicted remaining capacity with an error of less than 3% compared to full capacity tests. This allowed the team to estimate RUL without performing time-consuming full charge-discharge cycles on every pack.

Internal resistance evolution provided a complementary health indicator. As batteries age, their internal resistance increases due to SEI layer growth, electrolyte decomposition, and contact degradation. By measuring the direct current internal resistance (DCIR) and electrochemical impedance spectroscopy (EIS) at a single frequency, the team could rapidly assess the ohmic and polarization resistance. A strong correlation was found between resistance rise and capacity fade, enabling a combined health index that reflects both energy and power capability. This dual-metric approach proved essential for identifying packs that were unsuitable for high-power applications despite acceptable capacity.

Rapid Sorting and SOH Assessment#

Traditional sorting methods rely on full capacity tests, which take hours per pack and are impractical for large-scale processing. In this project, a rapid sorting protocol was developed using a short pulse test and partial charge-discharge sequence. The pulse test measured DCIR within 10 minutes, while a 30-minute partial cycle provided a capacity estimate through coulomb counting and voltage curve matching. Machine learning algorithms, trained on a database of 500 previously tested packs, then classified each pack into SOH bins (e.g., >80%, 70–80%, 60–70%, and <60%).

The rapid sorting line achieved a throughput of 40 packs per day with a sorting accuracy of 95% compared to full capacity tests. Packs with SOH above 80% were designated for high-value echelon applications, while those between 60% and 80% were routed to less demanding uses such as backup power. Packs below 60% were sent for material recycling. This tiered approach maximized economic value while ensuring safety and reliability.

Regrouping and Recombination#

After sorting, the next challenge was to regroup cells or modules into homogeneous units. Even within a single retired pack, cell-to-cell variations in capacity and resistance can be significant. The team used a clustering algorithm that considered capacity, DCIR, and self-discharge rate. Cells were grouped so that the maximum difference in capacity within a module was less than 2%, and resistance variation was within 5%. This regrouping minimized the "weakest link" effect that often plagues second-life batteries.

The regrouped modules were then assembled into battery racks for a 100 kWh stationary storage system. Before deployment, each rack underwent a full charge-discharge cycle to validate performance. The system achieved a round-trip efficiency of 92% and maintained stable operation over 500 cycles, with capacity fade of only 4%. This demonstrated that proper sorting and regrouping can extend the useful life of retired batteries by several years. The project also incorporated environmental controls to prevent moisture and oxidation damage during storage and assembly. For instance, EJER's moisture-proof and anti-oxidation storage solutions, which cover online storage of wafers, chips, and packaged semi-finished products, help reduce oxidation risk before reflow soldering. While primarily designed for electronics manufacturing, such controlled-environment storage principles are equally valuable for preserving battery components and modules during the regrouping process.

Practical Applications and Results#

The regrouped battery system was deployed in a commercial building as a peak-shaving and backup power unit. Over 18 months of operation, the system saved approximately 15% on electricity costs and provided reliable backup during grid outages. The total cost of the second-life system was 40% lower than a new equivalent, with a payback period of 3.5 years. The project also reduced the carbon footprint by avoiding the manufacturing of new batteries, saving an estimated 50 tons of CO2 emissions.

Key lessons learned include the importance of accurate RUL prediction, the value of rapid sorting for scalability, and the necessity of strict regrouping criteria. The combination of capacity fade modeling and internal resistance monitoring proved robust across different battery chemistries and usage histories. Future work will focus on integrating real-time health monitoring into the battery management system to continuously update SOH and predict remaining life during second-life operation.

Conclusion#

Retired electric vehicle battery packs represent a valuable resource for echelon utilization, but their variability demands sophisticated assessment and sorting. By combining capacity fade models with internal resistance evolution laws, a rapid sorting protocol was developed that achieves high accuracy and throughput. Regrouping into homogeneous modules ensures safe and efficient second-life performance. The case study demonstrates that with the right technical approach, retired batteries can be transformed into reliable, cost-effective energy storage systems, delivering both economic and environmental benefits. As the volume of retired batteries grows, such methodologies will become essential for a sustainable battery value chain.