EN 10204: what 2.1, 2.2, 3.1 and 3.2 actually mean

What the standard covers

EN 10204, Metallic products — Types of inspection documents, does one job: it defines the types of document a manufacturer issues to state that delivered products comply with the order, and what testing sits behind each type. The current edition dates from 2004.

Note what it is not. It is not a material specification and it is not a quality standard. It says nothing about whether the steel is any good — that is the product specification’s job. EN 10204 only tells you what kind of evidence you are holding, and who stands behind it.

Specific and non-specific inspection

Everything else follows from this distinction, so it is worth getting straight.

Non-specific inspection is carried out by the manufacturer to their own procedures, on products made to the same specification by the same process. The material tested is not necessarily the material you received. It tells you what that production reliably produces.

Specific inspection is carried out before delivery, according to the product specification, on the products actually being supplied or on test units the supplied products form part of. It tells you about your material.

That single difference is what separates the 2.x documents from the 3.x documents, and it is why a 2.2 and a 3.1 can show similar-looking numbers while answering different questions.

The four document types

Type 2.1 — Declaration of compliance with the order

A statement from the manufacturer that the products comply with the requirements of the order. No test results. Based on non-specific inspection. Adequate where the material is not critical and nobody downstream needs numbers.

Type 2.2 — Test report

Compliance with the order, with test results, based on non-specific inspection. This is the one most often mistaken for a 3.1, because it has figures on it. The figures are real; they just do not necessarily come from your material.

Type 3.1 — Inspection certificate 3.1

Compliance with the order, with test results from specific inspection, validated by the manufacturer’s authorised inspection representative independent of the manufacturing department. That independence requirement is the substance of a 3.1. It is the trade default for anything structural, pressure-related or traceable, and it is what most customers mean when they say “certified”.

Type 3.2 — Inspection certificate 3.2

As 3.1, but validated by two parties: the manufacturer’s independent representative, and either the purchaser’s authorised representative or an inspector designated by the applicable official regulations. Required by certain specifications and contracts. More cost, more lead time, and no benefit at all where the specification does not call for it.

What changed in 2004

If you have older paperwork or an older specification in circulation, you will meet document types that no longer exist. The 1991 edition ran to more categories — including a 2.3, and 3.1 split into A, B and C variants alongside a 3.2 in two parts — distinguished largely by who did the validating.

The 2004 revision consolidated all of that into the four types above. Broadly, the old 3.1.B is the ancestor of today’s 3.1, while the variants validated by an independent or purchaser-side party map onto today’s 3.2. If a customer specification still asks for a “3.1.B certificate”, it is asking for a 3.1 in current terms — but confirm it rather than assume, because occasionally the old wording is doing deliberate work.

Cutting, and the stockholder’s document

EN 10204 describes what the manufacturer issues. Most material does not travel from mill to end user untouched: it goes through a stockholder, gets cut into several pieces, and each piece goes to a different customer.

The mill certificate describes the original heat or cast. Normal trade practice is for the stockholder to supply a copy of it together with their own certificate of conformity, tying the specific pieces delivered back to that heat number. The customer’s real question is never about the paperwork format; it is whether the chain from the bar in their hand back to the mill is unbroken and documented.

Where traceability breaks

The failures are consistent, and almost all of them are administrative rather than metallurgical.

• The offcut nobody re-identified. A bar is cut, the remnant goes back in the rack, and the identity travels with the piece that left. The remnant is now uncertifiable stock that looks identical to certified stock.
• Outprocessing. Material goes out for machining, testing or treatment and comes back with a new works order number and no reference to the heat it came from.
• Mixed heats on one line. An order filled from two casts, despatched on one note, with one certificate attached. Half the delivery now has the wrong paperwork.
• The certificate that lives in a filing cabinet. Or an inbox, or a shared drive named by supplier rather than by heat. It exists, and finding it takes twenty minutes that customer will not wait for.
• Re-keyed heat numbers. Every manual transcription of an alphanumeric cast number is an opportunity to create a piece of stock that can never be matched to its certificate again.

What good looks like in a system

Traceability is a data-modelling problem long before it is a paperwork problem. The systems that hold up under audit share the same handful of properties:

• The heat or cast is a record, not a text field. Certificates attach to it once, and everything cut from it inherits the link automatically.
• Every operation preserves parentage. Cutting, splitting and processing create new items that still know what they came from, so an offcut is never anonymous.
• Material leaving site and returning keeps its identity through the whole outprocessing round trip.
• The certificate pack assembles itself at despatch, covering exactly the items on the note — including the awkward case of two heats on one delivery.
• Customers can pull their own copies instead of ringing at four o’clock on a Friday.
• The numbers are read, not typed. Capturing heat numbers off the incoming document removes the single most common way traceability is lost.

This is the problem MetlSys was built around: an ERP for steel and stainless stockholders where EN 10204 traceability is part of the spine rather than a document-management feature bolted on beside it. If your trade has a problem shaped like this one, that is the conversation we like having.