1) Prepare the Battery Bank

1.1 Assemble and baseline

• Assemble the battery at its installation site with the final inter-cell links, cables, and torque per manufacturer spec.
• Record open-circuit voltage (OCV) for every unit before any charging.
• Ensure interconnects and cable gauges match the intended operating configuration.

1.2 Charge protocol before testing

1. Equalize charge
   • Duration: 24 hours
   • Setpoint: 2.40 V/cell (≈ 14.4 V/unit for 12 V blocks with 6 cells)
2. Float charge
   • Duration: 3–7 days
   • Setpoint: 2.30 V/cell (≈ 13.8 V/unit for 12 V blocks)
3. Log voltages
   • Measure and record each unit’s voltage during both equalize and float. Look for outliers.

Tip: Stabilized, fully floated batteries yield more reliable capacity numbers and reduce scatter test-to-test.

2) Define Test Settings (Load, Time, and End Voltage)

2.1 Choose the test type

• Constant Current (A): Hold current constant; commonly used for telecom DC plants.
• Constant Power (W/cell or W/unit): Hold power constant; typical for UPS where bus voltage varies during discharge.

2.2 Set the End-of-Discharge Voltage (EOD)

Use the manufacturer’s recommended EOD per cell. Common values for VRLA are 1.75–1.80 V/cell. For 12 V blocks (6 cells), multiply by 6 (e.g., 1.75 V/cell → 10.5 V per 12 V unit).

2.3 Pick the rated duration

Choose the rated time from the datasheet (e.g., 1 h, 3 h, 10 h, 20 h). Your target load comes from the table for that duration and EOD.

3) Temperature Correction

Battery capacity is temperature-dependent. Ratings are usually given at 25 °C (77 °F). If you test at a different temperature, adjust the planned discharge rate using the manufacturer’s temperature correction factors.

3.1 Example

• Datasheet current for 1 h at 1.75 V/cell, 25 °C: ( I_{25} = 61.5 \text{ A} )
• Actual test temperature: 15.6 °C (60 °F)
• Correction factor (example): 0.93 at 60 °F

Adjusted one-hour test current: [ I_{\text{test}} = I_{25} \times 0.93 = 61.5 \times 0.93 = 57.2 \text{ A} ]

Rule of thumb: Colder → less capacity (reduce test load), hotter → more capacity (increase test load), but always use the table from the battery’s datasheet.

4) How to Calculate Capacity

4.1 Rated capacity percentage (time-based)

For a constant-current test at the rated hour-rate:

[ \text{Rated Capacity (%)} = 100 \times \frac{t_{\text{actual}}}{t_{\text{specified at }25^\circ\text{C}}} ]

Where:

• ( t_{\text{actual}} ) = discharge time you achieved to EOD at the corrected load
• ( t_{\text{specified at }25^\circ\text{C}} ) = datasheet duration at 25 °C

4.2 Alternate expression (Ah-based)

If you sum ampere-seconds or ampere-hours during the run (e.g., via a battery monitor):

[ \text{Rated Capacity (%)} = 100 \times \frac{\text{Ah}\text{actual}}{\text{Ah}\text{rated at 25^\circ\text{C}}} ]

Keep the discharge mode consistent with the rating you’re testing (constant current vs constant power). Don’t mix.

5) Test Setup & Instrumentation

5.1 Equipment checklist

• Load bank sized for the required current or power at the system voltage
• DC ammeter (for current) and digital voltmeter (for each unit and string total)
• Timer (or data logger)
• Battery cell monitoring system (optional but recommended)
• Torque tools per hardware spec

5.2 Wiring and verification

• Connect the load bank across the full string (e.g., 48 VDC systems with hundreds of amps may require 10 kW+ loads).
• Verify polarity, shunt orientation (if used), and that metering reads correctly at a small pilot load before the real run.

6) Run the Discharge

1. Confirm ambient temperature and apply the temperature-corrected load setpoint.
2. Start the timer and begin discharge.
3. Log string voltage, current/power, and individual unit voltages at regular intervals (e.g., every 5–10 minutes for short tests, 30–60 minutes for long tests).
4. Stop the test the moment the string reaches EOD under load (do not “coast” after EOD).
5. Record the final time and immediately remove the load.

Safety: VRLA can heat up near end of discharge. Watch for abnormal unit voltages, temperature rise, or loose connections.

7) Interpret and Report Results

7.1 Pass/fail guidance

• Many standards accept strings ≥ 80% of rated capacity as “serviceable,” but follow your organization’s threshold.
• If the string fails:
  • Identify weak units (lowest per-unit voltages under load, abnormal IR/conductance if measured).
  • Replace weak units and retest after the full charge protocol.

7.2 Minimum report contents

• Battery model, count, configuration, installation date
• Charge history before test (equalize/float durations and setpoints)
• Temperature and correction factor used
• Test mode (constant A or constant W), EOD, rated duration
• Measured time to EOD and calculated Rated Capacity (%)
• Per-unit voltage logs (at intervals)
• Observations, anomalies, corrective actions

8) Worked Mini-Example

• Battery: 12 V VRLA (6 cells), rated 100 Ah @ 20 h to 1.75 V/cell at 25 °C
• Test ambient: 20 °C; correction factor from datasheet table: 0.97
• Rated 20 h current at 25 °C: ( I_{25} = 5.0 \text{ A} )

Adjusted current: [ I_{\text{test}} = 5.0 \times 0.97 = 4.85 \text{ A} ]

Discharge result: Reached EOD in 17.8 h.
Capacity: [ \text{Rated Capacity (%)} = 100 \times \frac{17.8}{20.0} = 89% ]

Interpretation: At 89%, the string is below nominal but above a typical 80% replacement trigger—monitor trend or plan maintenance.

9) Practical Tips & Common Pitfalls

• Do not skip float stabilization (3–7 days). Under-charged strings under-perform.
• Hold the setpoint steady. For constant-current tests, keep current within ±1–2%.
• Measure individual units. Weak blocks often hide behind “OK” string voltage.
• Apply the correct EOD. A higher/lower EOD can inflate/deflate measured capacity.
• Use the right correction table. Use the manufacturer’s factors for your exact model.
• Document everything. Good logs make trend analysis and warranty claims easier.

Conclusion

A good capacity test is mostly good preparation: proper charge, correct temperature correction, the right EOD, and disciplined logging. With those in place, the math is simple—and the result is a trustworthy Rated Capacity (%) you can compare across years to plan maintenance and replacements.