The words Grade A travel further than the document behind them. LiFePO4 cell selection for home batteries is usually settled by a label, and the label is the least informative item in the file. It reports the outcome of one sorting step, taken against one specification, at one moment in production, and it says nothing about the spread inside the delivery that will be assembled into packs.
What the label is assumed to prove
Grade A is not a standard, not a certification and not a test method. It is a sorting word, and two suppliers can use it while describing two different criteria. If the label is the only document in the file, the acceptance criteria for the delivery do not exist, and there is nothing to reject a shipment against later.
Where the assumption breaks
Cell selection is one of several influences on how a home battery behaves, and the one that is hardest to correct later. A pack is built from cells that share a balancer, a busbar, a compression arrangement and one set of current limits. Two consequences follow for buyers.
The first is that a uniform nominal capacity hides a distribution, which is why decides usable capacity long before the datasheet does. Loose capacity matching is tolerable in a small series string and becomes a balancing problem in a long one, as make clear.
The second is that incoming inspection cannot recover a wide distribution. If the internal-resistance spread within a delivery is wider than the balancing current can absorb, the pack will pass its end-of-line test and start reporting imbalance in the first year, which is exactly the failure mode that makes a warranty claim expensive: the evidence of why it happened was in the lot data, months before, and nobody kept it.
The evidence worth requesting
Reading the distribution matters as much as receiving it. The figures worth having are the spread rather than the average, the sample size behind it, and whether the numbers describe the full lot or a sample drawn from it. Compare distributions only when the temperature, the measuring current and the cut-off condition are the same, because a resistance reading taken at a different current is a different measurement and cannot be ranked against another supplier's number. A supplier who sends the spread and the condition it was measured under has answered the question; a supplier who sends one nominal value has restated the datasheet.
The standards that sit under this list are worth naming separately, because they answer different questions. IEC 62619 covers safety requirements and tests for industrial lithium cells and batteries, IEC 63056 covers cells and batteries used in electrical energy storage systems, UL 1973 covers stationary battery systems, UL 1642 covers lithium cells at cell level, and UN 38.3 governs transport testing. None of them certifies a pack because a cell passed, and a cell certificate does not cover the assembly. Designations must be confirmed against the edition in force for the destination market, and clause-level references are what a purchase file should quote; they were current as of 2026-09.
Three questions that separate a file from a claim
Ask which specification and which test condition the Grade A statement was made against. Ask for the capacity and internal-resistance distribution of the actual lot being shipped, stating the temperature and the current at which the resistance was measured, because a resistance figure without its measurement condition cannot be compared with another supplier's figure. Ask what triggers a change notification when the cell supplier, the cell model or the assembly process changes, and in what timeframe. An answer that begins with "all our cells are Grade A" has not answered any of the three: it restates the label.
Two comparison mistakes decide cell questions more often than the cells do. Comparing capacity at one C-rate against capacity quoted at another changes the ranking of two suppliers, so capacity has to be compared on the same basis: the same discharge rate and the same cut-off voltage. Comparing a per-cell figure against a per-pack requirement does the same thing in the other direction, because the pack figure already contains the matching and the losses.
What cannot be changed later
The cell model and its format, the series and parallel arrangement of the pack, the compression and busbar design, and the current limits written into the protection logic are all fixed at the design stage. A later cell change is not a substitution; it can invalidate certification evidence and the warranty terms that were quoted with it, which is why the change-notification question belongs in the purchase file rather than in a conversation after the first shipment.
What travels with the shipment matters for the same reason. The lot identifiers, the matching data and the end-of-line results for each serial number are the only evidence that survives to a warranty conversation, and an installer or distributor has no way to inspect the cells that are already inside a sealed pack. The handover pack for each unit should therefore carry the lot identifier and the end-of-line result as standard, so that the installer, the distributor and the owner hold the same record at the moment a claim is opened, rather than reconstructing it months later. The batch records behind the packs Ruibit supplies are kept retrievable by serial number, which is what turns a claim about a failed unit into a document review instead of a discussion about memory.
Cell selection is a data question, not a badge question. Ask for the distribution, the condition it was measured under, and the records that connect the lot to the unit that was sold, in the form set out in . The collects the rest of this cluster.
FAQs
1. Does a Grade A label mean the cells are better?
Not on its own. Grade A reports the result of one sorting step, against one specification, at one moment in production. Two suppliers can both use the term while describing different criteria, and the label says nothing about the spread inside the delivery.
2. What should I ask for instead of a certificate?
The lot identifier with its date code, the capacity and internal-resistance distribution of the delivered lot, the condition those figures were measured under, and the end-of-line results per serial number. A certificate without a lot reference cannot be traced to a production window.
3. Why does one lot behave differently from another?
Because matching is a distribution, not a single value. If the resistance spread within a delivery is wider than the balancing current can absorb, the pack passes its end-of-line test and reports imbalance months later. The cause was visible in the lot data before assembly.
4. Which comparison mistakes distort cell decisions?
Comparing capacity at different discharge rates, and comparing a per-cell figure against a per-pack requirement. Both change the ranking of two suppliers without changing either product. Compare on the same basis and state the measurement condition every time.
5. What makes a cell-related warranty claim hard to win?
Missing records. The cells sit inside a sealed pack, so nobody can inspect them later. If the lot identifier and the end-of-line results do not travel with the unit, a claim depends on memory instead of documents.
