Battery Cell Grading and Sorting: OCV, Internal Resistance, and K-Value

Battery Cell Grading and Sorting: OCV, Internal Resistance, and K-Value

After formation and aging, every lithium-ion cell must be graded and sorted before it can be assembled into a module or pack. This process—measuring each cell's Open Circuit Voltage (OCV), internal resistance (IR), and self-discharge rate (K-value)—determines which cells ship to customers and which are destined for secondary applications or recycling. The grading process is the final quality gate in cell manufacturing, and its precision directly affects pack performance, safety, and warranty cost.

OCV: The First Pass/Fail Gate

After the formation cycle (typically 24–72 hours depending on chemistry), each cell is allowed to rest for a defined aging period (7–14 days for EV-grade cells, 3–7 days for consumer cells) before OCV measurement. The OCV reflects the cell's thermodynamic state of charge and, critically, reveals internal micro-shorts and lithium plating that occurred during formation.

For NMC cells (3.6–3.7 V nominal), the post-aging OCV at 50% SOC is typically 3.65–3.70 V. The acceptance band is tight:

  • EV Grade (A): OCV within ±5 mV of the batch median
  • Energy Storage Grade (B): OCV within ±15 mV of the batch median
  • Secondary Market (C): OCV within ±50 mV of the batch median
  • Reject: OCV deviation >50 mV or any downward voltage drift during the resting period that exceeds 0.5 mV/day (indicative of an internal short circuit)

A cell that reads 3.66 V at the start of aging and 3.63 V after 14 days—a 3 mV drop—exhibits a K-value of 0.21 mV/day. While within the ±5 mV total deviation band, this rate of voltage decay suggests elevated self-discharge and may warrant a B-grade assignment even if the absolute OCV is acceptable.

Internal Resistance: AC vs. DC Measurement

Internal resistance is the lumped parameter that determines the cell's voltage drop under load and its heat generation during charge and discharge. Two measurement methods are standard:

AC Internal Resistance (1 kHz Impedance):
A small AC current (typically 10 mA) at 1,000 Hz is applied, and the voltage response is measured. The result is primarily the ohmic resistance (RΩ) of the current collectors, tab welds, and electrolyte ionic conductivity. For a 21700 cylindrical NMC cell, typical AC IR is 10–25 mΩ. For a 100 Ah prismatic NMC cell, 0.3–0.6 mΩ.

AC IR is fast (measurement time <1 second per cell) and is the primary screening tool on high-throughput grading lines processing 10,000+ cells per day.

DC Internal Resistance (Pulse Method):
A DC current pulse (typically 1C for 10 seconds) is applied, and the voltage drop is measured. The DC IR includes both ohmic resistance and charge transfer resistance (Rct) at the electrode-electrolyte interface, making it a more functionally relevant measurement for predicting voltage sag under real loads.

The DC IR is always higher than the AC IR. For the same 21700 cell, DC IR might be 25–40 mΩ compared to 10–25 mΩ AC. The ratio DC IR / AC IR is typically 1.5–2.5 for healthy cells; ratios exceeding 3.0 indicate excessive charge transfer resistance, often due to inadequate electrolyte wetting or SEI formation issues.

The K-Value: Self-Discharge Screening

The K-value (also called the OCV drop rate) is the most time-consuming measurement in the grading process but is essential for identifying cells with internal micro-shorts or contaminated electrolyte. The test protocol:

  1. Charge the cell to a specified SOC (typically 30–50%, where the OCV–SOC curve is steepest and most sensitive to capacity change).
  2. Record OCV at t = 0.
  3. Hold the cell at a controlled temperature (typically 25 ± 0.5°C) for 7–14 days.
  4. Record OCV at t = end.
  5. Calculate K = (OCV_initial − OCV_final) / days.

For automotive-grade cells, a typical K-value acceptance threshold is <0.2 mV/day (equivalent to approximately 0.05% SOC loss per day at 50% SOC). Cells with K-values between 0.2 and 0.5 mV/day may be accepted for stationary energy storage applications where self-discharge tolerance is higher. Cells exceeding 0.5 mV/day are rejected and routed for root-cause analysis.

The relationship between K-value and capacity fade over the cell's life is not linear: a cell with a K-value of 0.3 mV/day at 14 days is approximately 3–5× more likely to experience accelerated capacity fade (≥20% loss in 1,000 cycles) compared to a cell with a K-value of 0.1 mV/day.

Sorting into Groups for Module Assembly

Cells destined for the same module or pack must be matched for OCV, capacity, and IR to prevent current imbalance. The sorting algorithm divides the production batch into "bins" or groups:

  • OCV Bin Width: 2–5 mV per bin
  • Capacity Bin Width: 0.5–1.0% of nominal capacity per bin
  • IR Bin Width: ±5% of batch median for AC IR

A typical sorting matrix for a production batch of 10,000 prismatic cells (100 Ah nominal) might yield:

Bin OCV (mV) Capacity (Ah) AC IR (mΩ) Yield (%) Application
A1 3,665–3,670 100.0–101.0 0.35–0.40 15% Premium EV pack
A2 3,660–3,675 99.0–102.0 0.30–0.45 50% Standard EV pack
B 3,650–3,685 98.0–103.0 0.25–0.50 25% Energy storage system
C Out of spec Out of spec Out of spec 10% Recycle/secondary use

Within a 96-cell series module for an EV pack, all 96 cells should come from the same OCV and IR bin. Mixing bins within a module guarantees unbalanced aging, which limits the entire pack's usable capacity to that of the weakest cell and accelerates degradation at the extremes.

Grading Line Throughput and Equipment

A modern automated grading line for 21700 cylindrical cells processes 200–300 cells per minute (12,000–18,000 cells per hour). The key stations are:

  1. OCV/IR Station (inline): 4-wire Kelvin contact probes measure OCV (to ±0.1 mV) and AC IR (to ±0.1 mΩ) in under 0.5 seconds per cell.
  2. Aging Racks: Temperature-controlled rooms (25 ± 1°C) with capacity for 10–14 days of production volume. For a 500,000 cell/day factory, the aging room holds approximately 6 million cells at any given time.
  3. Post-Aging OCV Station: Second measurement after aging; K-value calculated by the Manufacturing Execution System (MES) from the delta between pre-aging and post-aging OCV readings.
  4. Sorting Conveyor: The MES directs each cell to a bin conveyor based on the sorting algorithm. Pneumatic gates actuate in <100 ms to route cells without slowing the line.

The Cost of Inadequate Grading

A single cell with an undetected micro-short in a 7,200-cell EV battery pack can drain its parallel group over weeks, creating an imbalance that the BMS continuously compensates for—increasing heat generation and reducing range. If the micro-short progresses to a hard internal short, it can trigger thermal runaway in the module. The recall cost of replacing one battery pack is approximately $10,000–$15,000. The cost of thorough grading for that entire production batch is approximately $0.02 per cell. Do the math.

Grading and sorting is the last manufacturing step before the cell enters a module. Invest in precision measurement, maintain tight binning, and remember that every cell you let through with a borderline K-value is a future warranty claim waiting to happen.

Scroll to Top