How to Test a LiFePO4 Battery's Real Capacity: Discharge Test, Cutoffs and Records

The only dependable way to verify a LiFePO4 battery’s delivered capacity is to measure a controlled discharge against a clearly documented test boundary. A voltage screenshot, a BMS percentage, an internal-resistance reading or one short inverter run cannot prove the battery’s actual amp-hours or watt-hours.

The practical answer is:

• charge the battery with the exact approved profile and record the end condition;
• let the battery rest for the time required by the test plan;
• discharge it with a known, stable load while logging current, voltage, time and temperature;
• stop at the cutoff defined by the exact battery, cell, BMS or test specification;
• calculate amp-hours and watt-hours from measured data; and
• repeat the test when the result will be used for a purchase, warranty, pack-matching or system-sizing decision.

This guide answers one customer question: how can an owner verify the real capacity of a LiFePO4 battery or pack without turning an uncontrolled inverter run into a misleading “Ah test”? It is an evidence and measurement workflow, not a universal discharge-voltage recipe. The exact charge limits, cutoff, temperature range and current limits always come from the battery and BMS documentation.

Quick Answer: What Makes a LiFePO4 Capacity Test Valid?

A useful capacity test has five properties:

1. the starting state is defined, normally by a documented full-charge procedure;

2. the discharge current is measured rather than guessed from an inverter’s AC output;

3. the end condition is defined before the test begins;

4. the measurement records include time, current, voltage and temperature; and

5. the result states the test conditions instead of presenting one number as an absolute property.

If any of these are missing, the result can still be a field observation, but it should not be labelled as the battery’s confirmed rated capacity.

Capacity, Energy and Usable Capacity Are Different

Capacity is usually reported in amp-hours (Ah). Energy is reported in watt-hours (Wh) or kilowatt-hours (kWh). They are related but not interchangeable:

Ah describes charge flow. Wh describes charge flow multiplied by the battery voltage during that flow. A simple constant-current illustration is:

Ah = current in amps × discharge time in hours

Wh is better calculated from the logged samples:

Wh ≈ sum of voltage × current × time interval

If a 100Ah example battery supplies a stable 20A for 4.6 hours, the simplified result is 92Ah. If its logged average terminal voltage during that test were 13.0V, a rough energy illustration would be about 1,196Wh. These are arithmetic examples only, not AmpBird specifications and not a prediction for any particular cell, kit or finished battery.

Usable capacity is narrower than the nameplate number. It depends on:

• the manufacturer’s charge and discharge boundaries;
• the BMS charge or discharge permissions;
• current and voltage drop;
• temperature;
• the cutoff selected by the inverter or load;
• the battery’s condition and age;
• the measurement instrument and sampling interval; and
• whether the test stops because of a planned cutoff or an unexpected protection event.

The official Victron Lithium NG documentation gives a useful example of why the boundary belongs in the result: its nominal-capacity data is tied to a stated temperature and discharge-rate condition, and its technical data defines an end-of-discharge value for the exact battery model. That is a product-documentation example, not a universal AmpBird limit. See the Victron Lithium NG operation guidance and technical data.

Why a BMS Percentage or Voltage Screenshot Is Not a Capacity Test

A BMS may show pack voltage, cell voltage, current, temperature, alarms and an estimated SOC. A dedicated monitor may integrate current through a shunt. These are useful operating signals, but neither display automatically proves the battery’s delivered capacity.

The result can be misleading when:

• the configured battery capacity is wrong;
• a load or charger bypasses the current sensor;
• the BMS disconnects before the planned test cutoff;
• the voltage falls under load and triggers an inverter early-shutdown;
• the current sensor has an offset or poor resolution;
• the battery begins at an unknown SOC;
• the load varies with voltage or temperature; or
• the test is stopped by a user before the defined end condition.

The AmpBird guide to LiFePO4 cell voltage and data evidence is useful for keeping cell evidence separate from a finished-pack capacity claim. A cell’s OCV, resistance or listing capacity is not the same as the delivered capacity of a protected series/parallel battery.

Decide What You Are Testing Before Connecting a Load

The test plan changes depending on the object:

Do not use a pack result to grade individual cells unless the pack is disassembled and each cell is tested with an approved procedure. Do not use a cell test to promise the capacity of a finished pack. Wiring resistance, BMS behavior, temperature and the weakest series cell can change the system result.

Safety Gate: When Not to Run the Test

A capacity test is a controlled discharge of stored energy. It can create high current, heat and a large amount of energy at the load. Stop the plan and obtain qualified help when:

• the battery is swollen, leaking, cracked, wet, mechanically damaged or unusually hot;
• the terminals, busbars, cables, fuse or enclosure show heat or damage;
• the BMS, charger, inverter or battery model is unknown;
• the battery has already entered an unexplained protection state;
• the test load is not DC-rated for the voltage and current;
• the load can become unattended or overheat;
• there is no reliable way to isolate charging sources; or
• the proposed cutoff would require defeating a BMS or bypassing protection.

Never bypass a BMS, short a protection device, defeat a temperature sensor or continue after an abnormal smell, noise, heat rise, swelling, smoke or unstable voltage. A capacity test is not a repair procedure.

Step 1: Collect the Evidence and Exact Manuals

Before charging, create a test record with:

• battery, cell or module model and serial or batch information;
• chemistry and nominal voltage class;
• nameplate Ah and, if provided, nominal Wh;
• series and parallel arrangement;
• BMS model and firmware or configuration revision when available;
• charger and load model;
• current-measurement instrument and accuracy or resolution;
• temperature sensors and measurement location;
• the intended charge endpoint;
• the intended discharge current;
• the intended cutoff;
• the ambient and battery temperature range; and
• the reason for the test: receiving inspection, troubleshooting, matching, warranty evidence or system planning.

The AmpBird supplier-verification guide is useful when a seller’s datasheet, test report and listing disagree. For a finished battery, the manufacturer’s battery and BMS documentation takes priority over a generic cell-data assumption.

If a listing, datasheet and BMS configuration show different capacity, current or cutoff fields, record the conflict. Do not silently choose the value that produces the most attractive test result.

Step 2: Charge to a Documented Starting Condition

The starting state must be repeatable. Use the exact charger profile and battery instructions. Record:

• charger model and selected chemistry profile;
• charge voltage and current as observed;
• start and end time;
• pack voltage and cell voltages if available;
• battery and ambient temperature;
• the point at which the charger changed stage or stopped;
• the BMS charge permission and any alarm; and
• the condition used to declare the test “full.”

“The charger showed green” is not a sufficient record. A full-charge condition may depend on voltage, taper current, absorption time, BMS permission or a manufacturer-specific sequence. The Victron Lithium SuperPack NG configuration guidance is one official example that calls for the correct LFP charging configuration and a complete first full-charge sequence; do not copy its settings to another battery.

Do not charge a battery below the permitted temperature or outside its current and voltage limits. The battery manual, BMS and charger must agree before the test continues.

Step 3: Rest Before Discharge

After charging, disconnect the charger and let the battery rest for the time specified by the test plan. Record:

• the rest duration;
• open-circuit or resting pack voltage;
• individual cell voltages where the BMS exposes them;
• battery temperature;
• any current still flowing; and
• whether a heater, display, communication device or auxiliary load remains connected.

The rest period is not a universal number. A short rest may leave surface-charge effects; a long rest may expose a standby load or self-discharge issue. The point is to make the starting condition visible and repeatable.

If a monitor is used, confirm that all current paths are inside the intended measurement boundary. A shunt can document current, but it cannot make a bypassed load part of the test.

Step 4: Choose a Controlled Discharge Current

Select the current from the exact test specification and the battery’s approved operating limits. C-rate is a useful way to describe the current:

C-rate = discharge current ÷ rated amp-hour capacity

For a 100Ah example, a 0.2C test corresponds to 20A. For a 314Ah example, a 0.2C test corresponds to 62.8A. These are arithmetic examples only. The actual test current must follow the battery or cell documentation, the BMS limit, the load capability, the cables, the fuse and the safe test environment.

Choose a current that is:

• stable enough to measure;
• low enough for the test equipment and wiring;
• representative of the question being answered;
• compatible with the battery’s continuous discharge limit; and
• documented so another person can repeat it.

Do not confuse an inverter’s AC output rating with battery-side DC test current. A 1,000W AC load may draw substantially more than 1,000W from a low-voltage battery after conversion losses, and its current may change as the battery voltage falls.

Step 5: Define the Cutoff Before the Test

The cutoff is one of the most important parts of the result. It may be:

• an exact cell or pack voltage from the applicable test specification;
• a BMS-controlled end condition;
• a load-specific shutdown voltage; or
• a manufacturer-defined remaining-capacity or protection condition.

There is no universal LiFePO4 cutoff that can be copied into every cell, pack or inverter. A series count changes pack voltage, a load changes voltage drop, temperature changes behavior and the BMS may disconnect before the external meter reaches a planned voltage.

Write down:

• the cutoff voltage and where it is measured;
• whether the voltage is under load or after the load is removed;
• the allowed current at the cutoff;
• the BMS low-voltage or discharge permission behavior;
• the stop condition for cell imbalance or temperature; and
• the recovery procedure after the test.

If the BMS opens its discharge path, mark the result as “BMS-protection termination,” not as a clean capacity result to the planned voltage. A protection event can be useful evidence, but it changes the meaning of the test.

Step 6: Use a Measured DC Load When Possible

A controlled DC load is usually easier to interpret than an inverter powering household equipment. The load should be rated for the battery voltage, current, heat and duration. It should provide:

• stable or logged current;
• a defined stop condition;
• a safe heat-dissipation path;
• an independent voltage measurement;
• a current measurement that is inside the test boundary; and
• data logging or a written sampling method.

An inverter and a known AC load can be used for a system-energy test, but it answers a different question. It includes inverter conversion loss, standby consumption, low-voltage shutdown, load cycling and AC-side power-factor or measurement effects. Label it as a system test and do not compare its Wh result directly with a cell or battery nameplate Ah number.

The AmpBird 48V wiring and fuse guide can help map the protected current path. It does not provide a universal capacity-test cutoff or authorize a particular load.

Step 7: Log More Than the Final Ah Number

At each sample, record:

Do not fill missing data by interpolation and then present the result as measured capacity. If a logger stops or a current lead is moved, mark the test invalid or report the affected interval separately.

Step 8: Calculate Ah and Wh from the Log

For a stable current, the basic Ah calculation is:

Ah = current × time

For changing current, add the current contribution from each interval:

Ah ≈ Σ(current at interval × interval duration)

For energy, use the measured voltage and current together:

Wh ≈ Σ(voltage at interval × current at interval × interval duration)

Use consistent units. If current is in amps and time is in hours, the result is Ah. If voltage is in volts, the result is Wh. If the logger uses seconds, convert seconds to hours before multiplying.

Example only:

• measured current: 20A;
• elapsed discharge time: 4.6 hours;
• measured Ah: 20 × 4.6 = 92Ah;
• approximate average load voltage: 13.0V;
• rough energy: 92 × 13.0 ≈ 1,196Wh.

The energy calculation should use the logged sample-by-sample values when available. Average-voltage multiplication is a screening estimate, not a replacement for the raw log.

Step 9: Interpret the Result Without Overclaiming

Compare the result with the exact reference condition:

A result below the advertised number is a reason to check the test conditions before opening a dispute:

1. Was the battery genuinely full?

2. Was the current stable and measured at the battery?

3. Was the cutoff the same as the reference?

4. Was the temperature inside the stated range?

5. Did a BMS, inverter or fuse end the test early?

6. Did a cell reach an abnormal voltage before the pack?

7. Was the current path free of bypass loads?

8. Was the instrument calibrated or at least cross-checked?

If the conditions cannot be matched, report the number as “capacity delivered under this field test,” not as a universal capacity verdict.

Capacity Test Versus a Battery Health Check

A capacity test answers how much charge or energy was delivered under one boundary. It does not answer every health question.

Use other evidence for:

• cell matching and resistance consistency;
• enclosure or terminal damage;
• BMS temperature-sensor placement;
• current sharing in parallel modules;
• charger compatibility;
• insulation, fuse and cable condition;
• self-discharge during storage;
• communication faults; and
• abnormal cell spread under load.

The AmpBird cell-matching guide and shipment-inspection guide cover evidence that a pack-level discharge test cannot replace. A battery can deliver a plausible net Ah result while one cell, terminal, BMS sensor or branch connection needs attention.

Parallel Banks: Test the Bank and the Branches Separately

A bank-level test may tell you how much energy the complete parallel group delivered. It does not prove that each module shared current evenly.

Before testing a parallel bank:

• identify every module and its BMS state;
• verify that the external load is connected to the intended bank boundary;
• check branch fuses, cable length and terminal condition;
• decide whether branch current will be measured;
• record per-module voltage and temperature where available; and
• stop if one module heats, disconnects or diverges abnormally.

The AmpBird parallel-battery guide separates net-bank capacity from branch-current sharing. Do not add, remove or mix modules during a comparison test without recording the change.

Common Capacity-Test Mistakes

Mistake 1: Calling a BMS SOC screenshot a capacity test

SOC is an estimate based on the monitor’s configuration and current boundary. It is not the same as a controlled discharge result.

Mistake 2: Using voltage alone to calculate Ah

Voltage can be useful for plausibility and cutoff checks, but it does not measure charge delivered over time.

Mistake 3: Draining to an arbitrary voltage

A cutoff copied from another battery can over-discharge one pack, stop another pack early or produce a result that cannot be compared with its datasheet.

Mistake 4: Testing through a changing AC load

Household equipment cycles, inverter standby draw and low-voltage cutoffs make the result difficult to reproduce.

Mistake 5: Ignoring battery temperature

Capacity, voltage sag and BMS behavior can change with temperature. Record it rather than assuming the test is at a standard condition.

Mistake 6: Treating a BMS cutoff as proof of empty capacity

A BMS may disconnect because of one low cell, current, temperature or communication condition. Record the reason before interpreting the Ah total.

Mistake 7: Comparing a cell test with a finished-pack test

The series/parallel architecture, BMS, wiring and cutoff make the boundaries different.

Mistake 8: Reporting only the headline number

“92Ah” is incomplete without the starting condition, current, cutoff, temperature, measurement point and stop reason.

Mistake 9: Repeating full-depth tests unnecessarily

A full discharge can stress the battery and adds little value when the question can be answered with a safer controlled check. Follow the manufacturer’s service and warranty guidance.

Mistake 10: Continuing after an abnormal event

Heat, swelling, unstable voltage, smell, smoke, unexpected BMS action or damaged hardware is a stop condition, not a reason to extend the test.

A Copyable Capacity-Test Record

Use this record for receiving inspection, troubleshooting or a controlled comparison:

Save the raw file or photographs with the record. A table copied into a message is less useful than a record that preserves the time series and the test conditions.

Product and Configuration Path

Capacity evidence should be collected after the battery object and measurement boundary are defined:

1. identify the exact cells, modules or finished battery;

2. verify the datasheet, batch and test condition;

3. define the charge, rest, load and cutoff conditions;

4. choose a safe measurement method;

5. record the result with the exception and stop reason; and

6. ask for a configuration or evidence review when the test does not match the intended system.

The LiFePO4 cell collection, DIY battery kits collection, battery components collection and home battery systems collection are product-path references, not a promise that every variant includes a tester, load bank, shunt, BMS or capacity certificate. For a project-specific review, send the exact model, series/parallel arrangement, intended load, charger/inverter, target current, cutoff source, test log and installation environment through AmpBird contact support.

Final Decision Rule

Use a controlled discharge test when the decision depends on delivered capacity: receiving inspection, cell matching, warranty evidence, troubleshooting or system planning. Use a monitor, voltage reading or resistance measurement as supporting evidence, not as a substitute for the defined Ah/Wh test.

Before accepting the result, ask:

1. What exactly was tested?

2. What charge condition started the test?

3. What current actually flowed?

4. At what temperature?

5. Where was voltage measured?

6. What exact cutoff ended the test?

7. Did the BMS or another protection device stop it?

8. Is the result Ah, Wh or an AC-side system number?

9. Can the test be repeated under the same conditions?

10. Does the test answer the customer’s actual decision?

If those answers are missing, the next step is better evidence and a safer test plan, not a more confident capacity claim.

Frequently Asked Questions

How do you test a LiFePO4 battery’s capacity?

Charge it with the approved profile, rest it under a documented condition, discharge it with a measured load and record current, voltage, time and temperature until the exact approved cutoff. Calculate Ah and Wh from the log and report the test conditions with the result.

Can a battery monitor prove LiFePO4 capacity?

Not by itself. A monitor can show current and estimate SOC within its measurement boundary. Confirmed capacity requires a controlled charge and discharge test with a defined cutoff.

What load should I use for a LiFePO4 capacity test?

Use a controlled DC load when possible, rated for the battery voltage, current, heat and duration. An inverter and AC load can measure delivered system energy, but the result includes conversion losses, standby power and inverter shutdown behavior.

What cutoff voltage should I use?

Use the cutoff in the exact cell, battery, BMS or test specification. There is no universal LiFePO4 cutoff that is safe and comparable for every series count, current, temperature and battery design.

Is a 0.2C test always required?

No. The current must match the reference condition you are trying to reproduce and stay within the exact battery and test-equipment limits. A 0.2C example is not a universal AmpBird test setting.

Can I test a battery with an inverter?

You can measure an inverter system’s delivered energy if the AC load and DC input are logged, but it is not the same as a controlled battery Ah test. Inverter efficiency, standby draw, low-voltage cutoff and load cycling must be reported.

Why did the BMS shut down before the expected Ah?

Possible causes include one low series cell, current or temperature protection, a configured discharge floor, wiring voltage drop, an inaccurate SOC setting or a battery fault. Record the BMS reason and stop condition before repeating the test.

Does a lower capacity result prove the cells are defective?

No. First compare charge condition, current, cutoff, temperature, measurement boundary and stop reason with the reference test. A pack-level result cannot identify one defective cell without additional evidence.

Can I test each battery in a parallel bank separately?

Only with a safe, documented isolation procedure that the battery and BMS architecture allows. A bank-level test does not prove branch sharing. Do not disconnect or bypass a live module to create a test path.

Should I discharge a new battery to zero to test it?

No. Use the manufacturer’s approved test boundary and stop conditions. A BMS cutoff is not a license to defeat protection or repeatedly perform unnecessary full-depth discharges.

How should I report a capacity-test result?

Report the object, model/batch, charge condition, rest, current, cutoff, temperature, measurement instruments, elapsed time, Ah, Wh and stop reason. Attach the raw log or clear time-series record.

What should I send AmpBird for a capacity review?

Send the exact battery or cell model, batch/serial, series/parallel arrangement, BMS information, charger and load, measured current and voltage, temperature, cutoff source, raw log, stop reason and the decision you need to make.

Technical References

• Victron Lithium NG operation — example of capacity data tied to a defined temperature and discharge-rate condition, plus battery-operation boundaries.
• Victron Lithium NG technical data — example of nominal capacity, discharge-rate note and end-of-discharge data for one exact battery model.
• Victron Lithium SuperPack NG configuration and settings — example of following the exact LFP charging configuration and completing a full-charge sequence before operation.
• Victron SmartShunt operation — current integration and SOC-monitoring concepts; a monitor remains different from a controlled capacity test.

These references illustrate why a capacity result must name its current, temperature, cutoff and measurement boundary. They do not approve a particular AmpBird cell, battery, BMS, charger, inverter, load bank or capacity claim.