Standards

EN 10204 3.1 or 2.2: which certificate does your customer actually need?

The difference is specific versus non-specific inspection, not price. A guide to the four EN 10204 document types and the situations each one belongs in.

The four document types, in one table

EN 10204:2004 defines four inspection documents. Most arguments about them come from skipping straight to the numbers without the two columns that actually separate them.

EN 10204:2004 inspection documents
TypeDocumentInspectionTest resultsTraceable to your deliveryValidated by
2.1Declaration of compliance with the orderNon-specificNoneNoManufacturer
2.2Test reportNon-specificYesNoManufacturer
3.1Inspection certificateSpecificYesYesManufacturer's authorised inspection representative, independent of the manufacturing department
3.2Inspection certificateSpecificYesYesManufacturer's representative and an independent inspector, or the purchaser's representative

The line that matters is specific versus non-specific

Everything else follows from this. Non-specific inspection means the manufacturer tested material from its own production, chosen by its own procedures, of the same grade. Specific inspection means the manufacturer tested the material it is about to ship to you.

So a 2.2 test report is not a cheaper 3.1. It is a different claim. It says: here are real numbers from this grade, produced under this quality system. It does not say those numbers came from your heat, and it is not supposed to.

This is why the 2.2-versus-3.1 question cannot be answered by price. If the customer needs to tie a mechanical property to the piece in their hand, only a 3.1 does that. If they need evidence that the grade performs as specified, a 2.2 does it and a 3.1 is a more expensive way to reach the same place.

Who may issue what, and what a stockholder has to do

A point that trips up processors and merchants more often than the type definitions do. The inspection document originates with the manufacturer. When material passes through a stockholder or a processor, EN 10204 requires that the original inspection document, or a copy of it, is transferred with the material, and that the intermediary adds whatever documentation is needed to keep the delivered product traceable to that document.

In practice that means the burden that lands on a processor is not signing certificates. It is proving the chain: this bundle came from that heat, which is covered by that mill certificate. Cutting, bundling and re-labelling all break that chain unless something records it.

The certificate is rarely what fails an audit. The link between the certificate and the physical material is.

Six situations and the document each one calls for

  1. Structural steel into a CE-marked fabrication. 3.1, because the execution standard requires properties tied to the delivered material.
  2. Pressure equipment in the scope of PED. 3.1 at minimum, and 3.2 where the conformity assessment route or the notified body requires third-party witnessing.
  3. Commodity grade into a non-critical application. 2.2 is usually sufficient, and asking for 3.1 out of habit adds cost with no consequence attached to it.
  4. Customer's own specification stricter than the grade standard. 3.1, and check the test frequency clause, because the standard minimum may not satisfy the specification.
  5. Material resold from stock, several years old. Whatever the mill issued, plus your own documentation linking the bundle to that document. Reissuing it under your name is not a substitute.
  6. Tender that says "EN 10204 certificate" and nothing else. Ask. It is an incomplete requirement, and guessing high costs money while guessing low costs a rejected delivery.

What each one costs you

Set aside the price the mill charges, which is usually modest, and look at the operational cost, which is not.

Issuing 3.1

  • Traceable to heat and batch, which is what most audits ask for
  • Accepted wherever a 2.2 is accepted, so it never fails on type
  • Aligns with the batch-level data a Digital Product Passport will need
  • Removes the argument at incoming inspection

The cost of that choice

  • Requires an authorised representative independent of production to validate
  • Requires you to hold the heat-to-delivery link, reliably, for years
  • Slower to issue when the source data is on scanned paper
  • Over-specified on commodity material, where it buys nothing

Where it goes wrong in practice

Four failure modes account for most of the rejected deliveries we hear about. A 2.2 supplied where the contract said 3.1, discovered at the customer's incoming inspection. A 3.1 issued from a mill certificate covering a different heat, because the link was maintained by memory. Transposed digits in a mechanical property, copied by hand from a scan. And a certificate that is correct but cannot be produced within the window the customer allows, which in commercial terms is the same as not having it.

Every one of those is a data problem wearing a standards costume.

My take

The industry treats 2.2 as the budget version of 3.1, and that framing causes more trouble than any technical detail in the standard. They are different claims about different material. One is a statement about a grade, the other is a statement about your delivery. A discount does not convert one into the other.

My second opinion is less popular with quality departments. Defaulting to 3.1 on everything is not the safe choice people think it is. It is safe on paper and expensive in practice, because every 3.1 you issue is a promise that you can still reconstruct the heat-to-delivery link years later. Issue thousands of those on top of a folder of scans and you have manufactured an audit liability, politely, in triplicate.

Pick the type the application genuinely requires, then make sure you can defend the link behind it. A defensible 2.2 beats an indefensible 3.1 in every room where it actually matters.

The link is the part software can help with. Our page on EN 10204 certificate automation describes how the pipeline keeps heat and batch attached to the certificate it generates, and why OCR alone fails on mill test certificates explains why reading them is harder than it looks. If certificate volume is becoming the bottleneck, book a discovery call.

Related reading

Next step

Stop assembling certificates by hand

If the heat-to-delivery link lives in someone's head and a folder of scans, the two-week audit will show exactly where it breaks. Bring a sample of real certificates and you get a written go or no-go.

Free 45-minute call. Then a two-week data audit with a written go/no-go before you commit to anything. An engineer, not a sales rep, replies within one business day.