quality · procurement
Batch certificate vs catalogue value
Updated 2026-11-10
A catalogue value describes what a product is designed to do. A batch certificate describes what came out of the reactor on a specific day. Buyers routinely accept one in place of the other, and then discover the difference when a batch behaves differently from the sample.
The three things a certificate must name
- The batch. A number that matches the drum, the carton or the pallet label.
- The method. GB/T 528 for tensile, GB/T 13477.5 for skin time, GB/T 7124 for shear. A result without a method cannot be compared.
- The measured value. Not the specification range — the actual result, with the test date.
A certificate that lists only “conforms to specification” is a declaration, not an analysis.
What to compare against
Compare the measured value with:
- the catalogue typical value (is the batch in the expected region?),
- the specification limit (is it inside the agreed window?),
- the previous batch (is the process drifting?).
The third comparison is the one most buyers skip and the one that predicts problems. A tensile result that has fallen 8% over four batches is still in specification and still a warning.
What a COA cannot tell you
- How the product behaves on your substrate.
- Whether the batch was stored correctly after testing.
- Whether the sample tested is the sample shipped.
Those require a retained sample and a supply agreement that references it.
Reading a trend, not a number
A single certificate tells you one batch was inside its window. Four certificates told one after another tell you whether the process is stable.
batch tensile (MPa)
2401 3.52
2402 3.44
2403 3.31
2404 3.24
Every value is inside a 3.0–4.0 MPa specification, and the trend is a 8% fall in four batches. That is the pattern that precedes a real problem: a raw material change, a curing drift, or a line running faster than the reaction allows. No single certificate reveals it.
Ask for the last four certificates with the shipment and plot the two properties that matter for your joint. It takes ten minutes and it is the cheapest supplier monitoring available.
What to do when a batch is out
The sequence matters more than the reaction:
- Quarantine the pallet, keep the labels, and photograph the cartons.
- Test the retained sample at the factory and a sample from your pallet, same method.
- Compare with the certificate — if they disagree, the question is sampling or storage, not chemistry.
- Decide by joint criticality, not by batch size: a windscreen batch and a floor-seam batch have different tolerances.
A supplier who answers this sequence in a day is a supplier worth keeping. One who asks you to send the pallet back before discussing the result is telling you how the next dispute will go.
The sample was not the shipment
A certificate proves a batch was tested. It does not prove the pallet you received is that batch, or that it was stored correctly after testing.
Three checks close the gap:
- Batch number on the carton matches the certificate.
- Retained sample at the factory can be re-tested if a dispute arises.
- Your own retained sample from the delivery, stored cool, is the only physical evidence you control.
The last one costs one cartridge per delivery and settles arguments that no amount of email can.
What a usable certificate looks like
| Field | Example | What it tells you |
|---|---|---|
| Product and model | KA518 | Which specification applies |
| Batch number | 2408-17 | Links the paper to the drum |
| Production date | 2026-08-14 | Shelf life starts here, not at shipment |
| Test date | 2026-08-16 | Whether the result is from this batch |
| Property | Tensile strength | What was measured |
| Method | GB/T 528 | How it was measured |
| Result | 3.48 MPa | What came out, not what was specified |
| Specification | 3.0–4.0 MPa | Whether the result passes |
| Signature | Name and role | Who stands behind it |
A certificate missing the batch number or the method is a declaration. A certificate with all nine fields is a record you can compare across shipments, and the comparison across four shipments is where the useful signal is.
What this means for buyers
- Ask for the batch certificate with the shipment, not on request.
- Check the batch number matches the packaging.
- Keep the last four certificates and look at the trend, not the single result.
- Put the accepted test method into the purchase contract — it prevents later arguments.