A C5 site eats coatings. ISO 12944-2 puts carbon steel loss in category C5 at 80 to 200 µm per year, and ISO 9223 puts first year zinc loss there at 4.2 to 8.4 µm. Run those numbers against a plant that has to stand for 25 years and the answer is obvious: no single layer gets you there.
So we stack two. ZAM steel as the substrate, then 240 µm of paint on top. That combination is a duplex system, and this article is the actual specification we quote for C5 projects: what each layer does, what the numbers are, and the three places where a 240 µm specification quietly stops being 240 µm on site.
This is an option, not our standard build. ZAM is what we run as standard on steel structures. The 240 µm stack is what we add when the site classifies C5 and the owner wants the paint film as well.
What C5 actually asks for
ISO 12944-1 splits durability into four ranges: low up to 7 years, medium 7 to 15, high 15 to 25, very high over 25. Those are design lives for the coating, not warranty terms.
For category C5 the paint systems in ISO 12944-5 land in a 240 to 320 µm band, usually across three layers. The lab qualification in ISO 12944-6 scales with the durability range you are targeting:
| C5 durability range | Water condensation, ISO 6270-2 | Neutral salt spray, ISO 9227 | Cyclic ageing |
|---|---|---|---|
| Low | 240 h | 480 h | not required |
| Medium | 480 h | 720 h | not required |
| High | 720 h | 1440 h | 1680 h |
| Very high | see cyclic | see cyclic | 2688 h |
Read that table honestly and 240 µm of paint on bare carbon steel sits at the bottom of the C5 band. That is exactly why we do not put it on bare carbon steel.
Why ZAM alone is not the whole answer at C5
ZAM is a hot dip Zn-Al-Mg coating, nominally 91% zinc, 6% aluminium, 3% magnesium. Its strongest property is what happens at damage. Nippon Steel describes the mechanism plainly: zinc, aluminium and magnesium leach out of the coating layer and deposit a fine zinc based protective film containing magnesium across the exposed cut edge. A shear line heals itself. A galvanized cut edge does not do that as well.
The accelerated numbers are large. Nippon Steel’s catalogue puts Zn-Al-Mg at 10 to 20 times conventional hot dip galvanized in salt spray. Their own 8 year outdoor exposure gives about 4 times. Both figures are correct and they are 5x apart, which tells you everything about converting salt spray hours into field years. Do not do it. We ran our own 24 month marine and arctic field exposure for the same reason.
Here is the limitation. ZAM is a coil product. The alloy goes on at the mill, on a continuous line, before anyone cuts anything. You cannot re-dip a fabricated frame in Zn-Al-Mg the way you can drop a welded assembly into a galvanizing kettle. Sheared and punched edges are fine, because that is what the self healing behaviour is for. Welds are the harder case: the heat affected zone burns the coating off a band far wider than the bead itself.
On a C5 site, that band is where the structure starts.
The stack, layer by layer
| Layer | Type | Coats | DFT per coat | Total |
|---|---|---|---|---|
| Substrate | Hot dip Zn-Al-Mg coated steel, 91% Zn / 6% Al / 3% Mg | mill applied | per coil spec | per coil spec |
| Primer | High build tie coat for zinc coated substrates | 3 | 60 µm | 180 µm |
| Topcoat | Acrylic polyurethane | 2 | 30 µm | 60 µm |
| Paint total | 5 | 240 µm |
Paint goes on after fabrication. Not before. That is the point of the whole exercise, because the paint is what covers the weld zone the coil coating never reached.
Why a tie coat and not a zinc rich primer
Zinc rich primer over a zinc alloy coating adds nothing. The substrate is already the zinc reservoir, and a thicker reservoir is not the problem you have. The problem you have at that interface is adhesion, because a zinc surface is smooth and chemically active in a way blasted steel is not.
A tie coat is formulated for exactly that. It wets the zinc surface and gives the barrier layers something to bond to.
One hard rule that comes with it: no alkyd anywhere in the stack. Alkyd binders saponify against zinc and the film releases. Hempel’s technical guideline for hot dip galvanised substrates says it in four words. Alkyds should never be used. That applies to the topcoat as much as the primer.
Why 180 µm of primer in three coats instead of one thick pass
Three reasons, all of them practical.
Sag control. 180 µm of wet film in one pass on a vertical column runs. You get a fat bottom edge and a starved top edge, and the DFT survey finds it.
Solvent release. Thick single coats trap solvent under a skinned surface. That comes back later as blistering, usually after the first hot week.
Pinhole probability. Every coat has holidays. Three independent coats have to line their holidays up in the same spot to make a path to the substrate. They almost never do.
Why acrylic polyurethane on top, only 60 µm of it
The topcoat is not the barrier. The 180 µm underneath is the barrier. The topcoat is the UV layer and the colour layer, and it is meant to be the part that goes first.
Acrylic polyurethane holds gloss and colour under sun in a way an epoxy will not, because epoxy chalks. Sixty micrometres across two coats gives even colour and full hiding over the primer without adding film you do not need.
It is also the maintenance layer. When a C5 site comes back for coating work in year twelve, you are recoating 60 µm of topcoat, not rebuilding 240 µm.
Surface preparation is where this system is won or lost
Everything above is straightforward. This part is where jobs fail.
ASTM D6386 sorts zinc coated surfaces into three conditions and they are not prepared the same way. Newly coated, under 48 hours, needs profiling because the surface is smooth and clean. Partially weathered, from two days to about a year, is the hardest of the three: zinc reaction products have started to build and organic residues have to come off first. Fully weathered, past a year, is the easiest, and the zinc compounds that are already there should be left alone.
Sweep blasting parameters for zinc coated surfaces are not steel blasting parameters. The figures below are the published ones, not ours. Abrasive size and hardness and the angle come from ASTM D6386. Pressure, standoff and the atmospheric limits come from Hempel’s technical guideline for hot dip galvanised steel as a paint substrate:
- Non-metallic grit. Aluminium silicate or corundum, 0.2 to 0.5 mm, Mohs hardness 5 or less. (ASTM D6386 gives the same range as 200 to 500 micrometres.)
- Nozzle pressure 2.5 to 4 bar. (Hempel)
- Angle of incidence 30 to 60 degrees, not perpendicular. (ASTM D6386)
- Standoff around 50 cm. (Hempel)
- Relative humidity below 85%, steel surface at least 3 °C above dew point. (Hempel)
- Target is a uniform matt grey with a dense profile and no unswept patches. (Hempel)
Blast zinc the way you blast steel, square on and at full pressure, and you take off the coating you are trying to protect.
Two things to get right at the purchase order, not at the paint shop
Passivation and oiling. Coil is often supplied with a chromate or chromium free passivation, or lightly oiled, to survive storage. Both of them are release agents as far as your primer is concerned. If the steel is going to be painted, order it untreated, or budget the time to strip the treatment before blasting.
Or order the treatment that is designed for painting. Nippon Steel’s ZP phosphate treatment is specified for post paintability, and their catalogue says it gives paint adhesion good enough to omit pre-paint surface preparation. That changes the whole prep story and the cost with it. It is a coil ordering decision, and by the time the steel is on the shop floor it is too late to make it.
Wet storage stain is the other one. White rust on stored coil has to come off completely before any paint goes on. It is a corrosion product with no cohesive strength, and paint applied over it fails at that layer.
Three checks on site
1. Film thickness, and the gauge zeroing trap
Acceptance follows ISO 19840. The arithmetic mean of all readings has to be at or above the nominal dry film thickness. Every individual reading has to be at or above 80% of NDFT. Readings that fall between 80% and 100% have to be fewer than 20% of the total. Minimum reading counts scale with area: 5 up to 1 m², 10 above 1 to 3 m², 15 above 3 to 10 m², 20 above 10 to 30 m², 30 above 30 to 100 m².
Now the part that catches people. A magnetic induction gauge measures everything non-magnetic sitting on top of the steel. On a ZAM substrate that includes the zinc alloy layer. Zero the gauge on bare steel and every reading comes back high by roughly the coating thickness, and a thin paint job passes.
Zero on an uncoated ZAM offcut from the same coil. Keep the offcut with the QC kit. Write it into the inspection and test plan.
2. Adhesion, on a coupon
Pull off to ISO 4624 or cross cut to ISO 2409. Run it on a sacrificial coupon prepared and painted in the same batch, not on the delivered structure. A pull off test on a member you are about to ship leaves a hole in the coating you just paid for.
3. Stripe coats
This is the one that turns a 240 µm specification into a thin edge. Paint pulls back from sharp edges as it cures, because surface tension does not care about your specification. Edges, bolt holes, weld seams and cut ends all end up thin unless someone brushes a stripe coat on before each full coat.
Prepare edges to ISO 8501-3 grade P2. Then stripe them. If the inspection report has full DFT on the flats and nobody looked at the edges, the report is measuring the wrong thing.
How much does the duplex actually buy you
The American Galvanizers Association puts duplex service life at 1.5 to 2.3 times the sum of the two systems used separately. It is a real effect and the mechanism is easy to follow: the paint slows the zinc consumption, and the zinc stops undercutting wherever the paint gets damaged. Neither one is doing its job alone.
Two caveats we will state rather than let you find out later. That figure comes from hot dip galvanized work, not from Zn-Al-Mg, so treat it as a direction and not a number for a warranty schedule. And it assumes both systems are allowed to weather naturally with no maintenance, which is not how anyone runs a plant.
What this system does not do
It does not turn ZAM into stainless. On a C5 site the fasteners usually decide the outcome, and they are not painted. SS304 A2-70 to ISO 3506-1 is the standard specification. Where salt is constant, SS316.
It does not fix a bad detail. ISO 12944-3 exists because water sitting in a closed section will beat any coating system. Drainage, no traps, access for later maintenance.
It does not remove galvanic risk. An aluminium rail bolted to painted steel through a stainless fastener is still a bimetallic couple. The paint is the insulator right up until it is scratched.
And it is a specification, not a guaranteed service life. The durability range you get depends on the actual products, the actual preparation and the actual site.
On evidence, including ours
We have not run the ISO 12944-6 laboratory qualification on this exact stack. That test is run by the paint supplier on the specified system, and 1440 hours of salt spray plus 720 hours of condensation takes roughly ten weeks. If a tender needs it, ask at the enquiry stage and we will quote it as a line item.
What we will not do is describe a coating system as passing a test it has not been through.
Our reports are listed with the issuing body, report number and year on the certifications and test reports page. The salt spray report XMIN2309001187PL01 (ISO 9227, CASS, 2023) is one of them and can be checked against SGS directly. Note what that means and what it does not. A report covers the part it was run on. It does not cover a product line, and it does not cover this coating system.
Ask any supplier for the report number. A supplier who cannot give you one is telling you something.
Specification wording you can paste into a tender
Steel members: hot dip Zn-Al-Mg coated steel, coating alloy nominally 91% Zn, 6% Al, 3% Mg. Coating mass to be stated on the mill certificate. Coil supplied untreated, or with a phosphate treatment suitable for post painting. No chromate passivation. No oiling.
Sequence: fabrication complete before painting. Edges prepared to ISO 8501-3 grade P2.
Surface preparation: to ASTM D6386 for the applicable surface condition. Sweep blast with non-metallic abrasive 0.2 to 0.5 mm, Mohs hardness 5 or less, nozzle pressure 2.5 to 4 bar, angle of incidence 30 to 60 degrees. Apply only at relative humidity below 85% and with the steel at least 3 °C above dew point.
Primer: high build tie coat formulated for zinc coated substrates. 3 coats. NDFT 180 µm. No zinc rich primer. No alkyd binder in any layer of the system.
Topcoat: acrylic polyurethane. 2 coats. NDFT 60 µm. Colour to RAL as scheduled.
Total paint NDFT 240 µm. Stripe coat all edges, bolt holes, cut ends and weld seams before each full coat.
Acceptance: dry film thickness to ISO 19840, gauge zeroed on an uncoated offcut of the same coil. Adhesion to ISO 4624 or ISO 2409 on a coupon prepared in the same batch.
FAQ
Why not just order thicker ZAM instead of painting it?
Coating mass on a continuous hot dip line has a practical ceiling, and more of it does nothing for the weld heat affected zone. That zone is the reason the paint is there.
Why not batch hot dip galvanize after fabrication instead?
That is a legitimate alternative, not a worse one. Batch galvanizing after fabrication covers welds and gives you thicker zinc in one operation. What you give up is the Zn-Al-Mg behaviour at later mechanical damage, and thin sections can distort in the kettle. Different spec, different trade. We will quote either.
Does the paint cancel out ZAM’s self healing?
No. Under a scratch the alloy layer still works exactly as it does on bare ZAM. The paint just means you get to find out far less often.
Can we get a colour match?
Yes. Colour lives in the acrylic polyurethane topcoat, so RAL matching is a topcoat decision and does not change anything underneath it.
Is 240 µm enough for offshore?
Offshore is CX, not C5, and ISO 12944-9 covers it separately. Different conversation. Tell us it is offshore at the enquiry stage.
Related resources
- How to specify corrosion protection for solar mounting systems, C1 to CX
- ISO 9223 corrosion classes for solar mounting, C1 to C5
- What is ZAM steel coating
- Zn-Al-Mg in extreme environments: 24 month marine and arctic field exposure
- Choosing roofing screws for solar: coating class vs corrosion environment
Send us the site coordinates and the design life and we will come back with the corrosion specification and a structural calculation. If it classifies C5, this stack will be on the quote as an option with the cost broken out, so you can decide whether you want it.