Cell Capacity Grading Report: How to Read the Data

Two suppliers can quote the same 280 Ah LFP prismatic cell and deliver products that behave nothing alike. The datasheet is identical because it describes the design. The grading report is different because it describes the batch. If you read one document before signing a lithium cell purchase order, read that one.

Key takeaways

The datasheet describes the design; the grading report describes the batch. If you read one document before signing a cell purchase order, read the grading report.
Four columns matter: measured capacity in Ah at a stated rate and temperature, AC internal resistance in milliohms at 1 kHz, self-discharge (K value) in mV per day, and open-circuit voltage at dispatch.
Spread matters more than the average. A lot with a 3 mΩ average and a 4 mΩ spread makes a worse pack than one with a 5 mΩ average and a 1 mΩ spread.
For matched prismatic LFP, a capacity spread within about 1% and a resistance spread within about 10% of the mean is a reasonable expectation.
Ask for the raw per-cell CSV rather than a summary, check the lot number, cabinet number and date, and write the bin, the maximum spread and the K value into the purchase order.

What grading is, and why it exists

Cells come off the production line with real variation — a percent or two of capacity, a few milliohms of internal resistance, slightly different self-discharge. Grading (also called binning) sorts them into groups tight enough to be used together. In a series string the weakest cell sets the pack capacity and the highest-resistance cell sets the heat, so the spread matters more than the average.

The sequence is: formation, where the cell is charged for the first time at low current and the SEI layer is built; ageing, typically one to two weeks at room temperature or a few days at elevated temperature; grading, where the cell is cycled at a defined rate and its actual capacity measured; and finally binning by capacity, resistance and open-circuit voltage.

The columns that matter

Capacity (Ah)

Measured discharge capacity at a stated rate and temperature — usually 0.5C at 25 °C. A 280 Ah cell delivering 281 Ah is normal; the same cell delivering 272 Ah is a lower bin and should be priced that way. Check that the rate is stated. Capacity measured at 0.2C is always higher and is the easiest number to inflate.

AC internal resistance (mΩ)

Measured at 1 kHz, this predicts how much the cell will heat and how far its voltage will sag under load. Spread matters more than the absolute value: cells going into one pack should sit within a few milliohms of each other. A lot with a 3 mΩ average and a 4 mΩ spread will make a worse pack than a lot with a 5 mΩ average and a 1 mΩ spread.

Self-discharge / K value (mV per day)

Resting voltage drop per day after charging. It is the test that catches the defects capacity misses — microscopic metal contamination and separator damage show up here first. Cells with an abnormally high K value are the ones that go into thermal trouble months later.

Open-circuit voltage at dispatch

Cells shipped for assembly usually sit around 30 to 50 percent state of charge for transport compliance. A batch with widely scattered voltages has either been sitting mixed stock or has not been re-matched since grading.

Lot number, cabinet number and date

Without these three the report is decoration. A lot number lets you trace a failure back; a cabinet number tells you the data came from a real grading cabinet; a date tells you the cells were graded recently rather than two years ago in someone else’s file.

Reading the distribution, not the average

Ask for the raw file — a CSV with one row per cell — rather than a summary sheet. Then look at four things: the minimum, the maximum, the spread between them, and whether any cell sits outside the stated bin. A healthy lot clusters. A suspicious lot shows suspiciously round numbers, identical values repeated in blocks, or a distribution with no natural scatter at all, which usually means the file was typed rather than measured.

For prismatic storage cells, a capacity spread within about one percent and a resistance spread within about ten percent of the mean is a reasonable expectation for matched lots. Pouch cells for traction packs are typically binned tighter. Whatever the chemistry, the rule is the same: the tighter the bin, the more you pay and the better the pack behaves.

How fake grading reports are made

The hero cell — a report for one good sample presented as the batch average
A summary without per-cell rows, so the spread cannot be checked
A file with no cabinet number, no date, or a date that predates the cell model
Capacity quoted at a lower rate than the application actually uses
Grade A claimed with no grading data at all, on the word of the sales person
Refurbished cells re-marked with a fresh lot code, often visible as laser marking over an older code

What to write into the purchase order

Grading data is only useful if it is contractual. Put these in writing before you pay:

Cell brand, model, and the exact bin: minimum capacity in Ah at a stated rate and temperature
Maximum AC internal resistance and the maximum spread within the delivered lot
Maximum self-discharge (K value) and the minimum resting voltage at dispatch
The right to receive the raw per-cell grading file for the shipped lot before balance payment
Sampling method and acceptance number — for example AQL 1.0 on a random draw from the delivered pallets
Retention samples: a sealed box of cells from the same lot kept for later comparison
Remedy defined in advance: replacement, credit, or return at the supplier’s freight cost

Acceptance belongs in the same document: the test method, the pass threshold and who pays if a batch falls short. Sample approval and incoming inspection gives a practical checklist to attach to the order.

Where our cells sit

Pouch cells are stocked with grading data attached to the lot. The Farasis P58 is rated 60.5 Ah at 275 Wh/kg, the Farasis P70 69 Ah at 281 Wh/kg, and the Farasis S45 semi-solid-state cell 46.7 Ah with a 2,500-cycle rating. LFP prismatic cells — 280 Ah, 314 Ah and similar — are sourced to order from tier-one makers and graded on arrival rather than taken on trust. Terms and grading standards are on our lithium cells wholesale page; if you want the cells turned into packs, the custom pack OEM process starts from the same data.

Finished packs built on graded cells include a 72V 30Ah LiFePO4 pack and a 74.6V 148.5Ah NMC module.

Frequently asked questions

What is a capacity grading report?

A per-cell record produced after formation and ageing, giving measured capacity, AC internal resistance, self-discharge and open-circuit voltage for each cell in a lot. It is the evidence that the cells were actually measured and sorted, rather than sold on the strength of a datasheet.

What does Grade A mean for lithium cells?

An industry convention, not a standard — there is no authority issuing grades. In practice Grade A should mean the cell meets the maker’s full specification and carries a traceable lot code. Always ask for the grading data behind the claim.

How much capacity spread is acceptable in one pack?

Cells assembled into the same series string should be matched to within roughly one percent of capacity and a few milliohms of internal resistance. Wider spreads do not fail immediately — they cost you usable capacity and cycle life, because the pack is limited by its weakest cell.

Why is internal resistance reported at 1 kHz?

AC internal resistance measured at 1 kHz is fast, repeatable and comparable between suppliers, which makes it the industry default for grading. It is not the same as DC resistance under load, but for matching cells within a lot it is the right measurement.

Can I verify a grading report myself?

Partly. Check internal consistency first: does the average match the claimed bin, is the scatter natural, are the cabinet and lot numbers present. Then verify physically — measure internal resistance on a random sample yourself or have a third party do it, and keep retention samples for later comparison.

Do I need graded cells for a small order?

Yes, especially for a small order. Matching matters more as the series string gets longer, and a small pack is often a 13S or 20S string where one weak cell is a large fraction of the total. Buying ungraded cells is cheaper per unit and more expensive per pack.

Ask for the grading file

Send the cell model, quantity and application and we will return the grading data for the lot we can actually ship, together with the export file. Start on WhatsApp, or see lithium cells and LiFePO4 battery packs.