What Conditions Make a LiFePO4 Capacity Acceptance Test Comparable?

Agree what the capacity number is intended to prove

An incoming acceptance test, a development characterization and an equipment runtime trial answer different questions. Acceptance asks whether a delivered item meets an agreed controlled requirement. Characterization explores behavior across conditions. A mission trial asks whether the equipment completes a task. Decide which decision the result will support before choosing a procedure or comparing supplier reports.

A capacity label needs its rating conditions. Victron's Lithium Smart technical table, for example, identifies capacity at a stated temperature and supplies different temperature-dependent figures. That is evidence that the conditions belong beside the number, not evidence that the same reductions apply to another manufacturer's pack. Ask the supplier for the governing model specification and distinguish minimum guaranteed values from typical or nominal reference values.

Freeze preparation and discharge conditions together

The starting charge procedure determines what is available at the beginning of the discharge. Specify the approved charger or charge profile, completion rule, any rest period and the way temperature is established. A charger indication by itself may not provide enough information to reproduce preparation on another bench. Do not introduce conditioning cycles unless they are part of the agreed procedure.

Then identify the discharge behavior. A constant-current reference test is different from a constant-power load or a repeating equipment profile. Keysight's discharge-capacity guidance describes controlled loading with voltage monitoring and a specified cutoff. A qualified test plan should retain those conditions and use the approved product limits; this article is not a procedure for discharging an unfamiliar pack to an improvised endpoint.

Use amp-hours and watt-hours for the questions they answer

Amp-hours quantify delivered electric charge. Watt-hours quantify delivered energy and therefore include the changing terminal voltage. Two results can be close in amp-hours yet differ in energy if their voltage profiles differ. Multiplying every measured amp-hour by a nominal voltage produces an estimate; integrating measured voltage and current over time provides the delivered-energy result at the selected boundary.

As an original arithmetic example, a hypothetical 10 A discharge lasting 9.6 hours delivers 96 Ah. If the time-integrated energy record is 1,200 Wh, that is the energy result for that test. The pair implies an average discharge voltage of 12.5 V. These invented planning values illustrate the relationship only; they are not a test of a Jetray 100 Ah model or a proposed pass threshold.

A regulated host may require energy until its own minimum input voltage, while a reference capacity specification may use a different approved endpoint. Retain both results under their own labels if both are required. Do not extend a test below the permitted limit to make its amp-hour total match a label.

Record why the discharge stopped

A test can end because the defined external cutoff is reached, the load reaches an operating limit, the pack interrupts output, or an environmental or equipment safeguard acts. Those endpoints are not interchangeable. Record pack voltage, current, temperature and available status near the event so a reviewer can distinguish the intended endpoint from an interrupted test.

Measurement location matters too. A voltage measured at the electronic load includes the effect of intervening wiring, while a pack-terminal measurement describes a different boundary. The test owner should document the sensing arrangement and consider calibration, timing, resolution and uncertainty. Reporting additional decimal places does not remove a systematic measurement error or make two different procedures equivalent.

Keep raw traces and sample identities with the calculated totals. A final spreadsheet cell without the underlying time record makes later comparison difficult, particularly when a firmware change or protection event alters the endpoint.

Set an adjudication rule before a borderline result appears

Agree how a result near the acceptance boundary will be handled. The rule should address measurement uncertainty, invalid runs, permitted repeat tests and responsibility for reviewing deviations. It should not permit repeated cycling until a desired answer appears. If supplier and buyer procedures differ, reconcile the method before treating the disagreement as a product defect.

Sampling also needs a separate commercial decision. A development sample result does not establish every delivered unit's conformity, and an incoming sampling plan does not prove long-term cycle life. Define which checks apply to qualification, routine release and dispute investigation. Preserve revision and lot traceability so a valid result can be associated with the correct supply configuration.

The linked Jetray products can anchor a model-specific discussion. Request the current rating conditions, approved preparation and acceptance limits rather than attaching a generic LiFePO4 test copied from another product family.

1. Identify whether the decision concerns conformity, characterization or mission performance.
2. Approve preparation, environment, load and endpoint as one procedure.
3. Record Ah, Wh and stopping reason with the actual raw data.
4. Apply the agreed uncertainty and retest rule before making a purchasing decision.

Capacity acceptance questions