Corrosion protection of a steel hall: how many microns, and who actually measures them

You ask the contractor how the hall is protected against corrosion and you get the answer: “it is galvanised” or “a primer and two coats”. Neither of those is a specification — they are descriptions of a process without a single figure that can be checked later. A real specification to ISO 12944 has five items: the corrosivity category of the environment, the durability class, the surface preparation, the coating system with its thickness, and the way all of it is accepted. Without those five, nothing stands in the contract that you could bring a claim about.

The first question: how aggressive is the environment

ISO 12944-2 divides atmospheres into six corrosivity categories, from C1 (heated offices) to CX (extreme coastal and industrial environments). The category is not an impression but a measurable quantity: it is defined by the loss of thickness of low-carbon steel in the first year of exposure. In C3 that is more than 25 up to 50 micrometres, in C4 more than 50 up to 80, and in C5 more than 80 up to 200 micrometres — per year, on bare steel.

Those figures apply to the first year and slow down over time, because the corrosion products partly protect the substrate. That means an estimate made by multiplying the first-year rate by the number of years is conservative, not optimistic — which is the good side, but also the reason why the durability of a coating is never calculated that way.

Two traps keep recurring. The first: a hall does not have one category. The outside in Slavonia is regularly C3, while the inside depends on the use — a dry goods warehouse can be C2, whereas a wash bay, a dairy or a building with permanent condensation and contamination goes into C4 or C5. The second: categories are not guessed from memory for unusual spaces. For such cases the standard requires a measurement or a comparison with an existing building in the same microclimate — and if nobody has done that, the quotation should at least state that the value is an engineering estimate and not a standard.

The second question: the durability class is not a guarantee period

ISO 12944-1 defines four durability classes: low up to 7 years, medium 7 to 15, high 15 to 25 and very high more than 25 years. This is where the greatest misunderstanding in the whole story arises.

The end of durability is likewise not a matter of impression. The standard ties it to a measurable state: when rusting to ISO 4628-3 reaches grade Ri 3 on about 10 % of the surface, the system has come to the end of its design life and it is time for a major refurbishment. Until then, smaller repairs are carried out at the places where the coating regularly fails first — the edges, the welds and the connections.

Surface preparation — and the clause almost nobody reads

For coating systems on steel the standard requires blast-cleaning to grade Sa 2½ to ISO 8501-1 and a roughness profile within the prescribed range to ISO 8503. That is the part contractors usually know about and comply with. The problem lies elsewhere — at the edges, the welds and the surface imperfections.

ISO 8501-3 prescribes three grades of preparation for them: P1 (no preparation), P2 (partial) and P3 (full — edges rounded, weld spatter removed, surface smooth). The difference is enormous, because on a sharp edge the coating thins out of its own accord, and that is where it starts to rust first.

That is why the P grade is a sentence that must stand in the quotation, just like the thickness of the coating. If it is missing, a later complaint about rust coming out of the edges has no basis — the contractor has delivered what was agreed.

How many microns — and how that is actually measured

ISO 12944-5 gives tables of coating systems: for every corrosivity category, durability class and type of binder it prescribes the minimum number of coats and the nominal dry film thickness (NDFT). The tables are separated by substrate — one for blast-cleaned steel, another for hot-dip galvanised steel. The figure is not guessed: the contractor chooses the system and copies the NDFT from the table, so that figure must stand in the quotation.

What almost never stands there, and what decides a dispute, is how that thickness is accepted. ISO 19840 sets the rule known in practice as the “80 % rule”: the mean measured thickness must be at least equal to the NDFT, and no individual measurement may be below 80 % of the NDFT. In addition, measurements on a blast-cleaned surface have to be corrected for the roughness profile — otherwise the instrument also measures the peaks of the profile and shows a thicker film than is actually there.

Galvanised and coated: what the standard really says about duplex

Hot-dip galvanising to EN ISO 1461 is not paint but a metallurgical layer: for steel 6 mm thick and above a mean thickness of 85 µm and a local minimum of 70 µm are required. Zinc is consumed slowly — in C3 it loses less than 2.1 µm in the first year — so in the milder categories galvanising on its own lasts for decades.

In C4 and C5 the picture is different. There the zinc goes at several micrometres a year, and the standard treats it as a base for a coating and not as a replacement for it. The combination of zinc and a coating (in practice “duplex”) has no separate durability multiplier in ISO 12944-5; what the standard really gives is more concrete and more useful: on a galvanised substrate, for the same category and the same durability class, 20 to 33 % less coating is needed than on blast-cleaned steel. For C5 and very high durability that means a system of 240 µm in two coats on zinc — while the same requirement on bare steel calls for a noticeably thicker film and more coats.

The figure “duplex lasts 1.5 to 2.5 times longer than the sum of the individual protections” that you will find in the brochures does not come from ISO 12944, but from the specialist literature on galvanising. It is not meaningless, but it cannot be invoked in a contract — whereas the percentage from the tables of the standard can.

Application and the only visual check the client has

ISO 12944-7 governs the conditions of application. The key requirement is that the temperature of the substrate be unambiguously above the dew point — the standard prescribes no figure for the difference here, but requires the dew point to be measured and the work to be done only once the condition is met. In practice that means that coating on the building site in late autumn or early in the morning cannot be improvised: without measuring the humidity and the dew point, the application does not comply with the standard, no matter how it looks.

A second thing from ISO 12944-8 is rarely asked for and costs nothing: every coat of the coating must be in a different shade. That is the only way for someone without an instrument to see whether a coat has been skipped — at a damaged spot or an edge you can see how many lie underneath. If all three coats are the same colour, after drying there is no visual proof that the third was applied at all.

The reference area: the benchmark agreed before the dispute, not after it

A reference area is a part of the structure protected to the agreed specification and accepted before the rest of the work is done. It serves as a benchmark: everything that follows is compared with it, and later it also serves for monitoring the ageing of the coating. EN 1090-2 in Annex F.7.3 provides for reference areas as a rule for categories C3 to C5 — that is, for almost every hall outdoors in Croatia. That is not an additional service that is charged extra, but a part of the work that has to be asked for.

That last item eats more coating than people expect. The structure arrives protected from the plant, but on the building site it is bolted, welded and scratched — and repairs in the field are carried out in conditions that are never like those in a paint shop. A quotation that says nothing about it is not cheaper; it has merely left that item for later.

Corrosion protection is the only item on a hall that is not seen when it is good and is seen too late when it is bad. The difference between a system for C3 and a system for C5 on the same structure is not large in euros compared with the price of the steel — but it is the difference between renewing the coating in twenty years and a partial renewal in seven. More about the execution classes and the documentation you are entitled to ask for is in the article on EN 1090-2, and about why all of it is written into the design and not into an e-mail, in the article on the project documentation.

We make steel structures and prefabricated halls from our own production in Đakovo, and we write the protection specification according to the corrosivity category from the design — not as a flat rate. If you have the dimensions and the use of the building, we can start from that: about steel structures and an enquiry with the dimensions.