Structural Steel Fire Protection: How Thickness Is Set

Steel does not burn. That is exactly why people underestimate it. It softens, quietly and predictably, and a beam that has lost half its strength looks identical to one that has not.

Every fire protection decision on a steel frame comes down to one number: the temperature at which that particular member stops carrying its particular load. Get that number and the thickness of coating follows from it. Skip it and you are buying paint by the drum instead of by the calculation.

Buyers ask us for intumescent coating by the litre and by the price. The useful conversation starts a step earlier, with the schedule of members and the fire period the design calls for. Simfy Exim sources these coatings; the specification belongs to the project’s fire engineer, and the thickness on the drawing is not something a supplier gets to adjust.

Why does steel need protecting if it cannot catch fire?

The MDPI review of intumescent coating qualification states that the critical temperature ranges between 500 and 600 °C and is the one at which the steel element loses its load-bearing capacity at a specific load. That last phrase does the work. A column carrying sixty per cent of its capacity has a different critical temperature from an identical column carrying thirty per cent. The same review notes that higher critical temperatures require lower fire protection thickness, which is why a structural engineer who has done the utilisation calculation properly can often cut the coating bill without touching the fire rating. The saving comes out of the analysis, not out of the product.

What is a section factor and why does it change the price?

This is why two members with the same fire rating on the same drawing can carry very different thicknesses. A light purlin has a great deal of surface for very little metal, so heat reaches the core almost immediately. A heavy column has the opposite ratio and behaves like a thermal flywheel. Exposure matters too: a beam with a concrete slab on top is exposed on three sides rather than four, which lowers the ratio and the thickness with it. When a coating schedule arrives with a single thickness for every member on the job, that is a sign nobody has run the section factors — and it means either wasted material on the heavy sections or an unprotected slender one.

What does EN 13381-8 actually assess?

Note the word ranges. The output of the standard is not a single pass or fail but an assessment across a field of conditions, which is then published as a set of thickness tables. A specifier reads into those tables with three inputs — the section factor of the member, the critical temperature the engineer has calculated, and the required period — and reads out a dry film thickness. That is the entire mechanism, and it explains why an intumescent product cannot be compared on price per litre. Two coatings at the same price may demand very different thicknesses for the same member, and the one that looks dearer per litre can be cheaper per tonne of steel protected.

How is the test fire defined?

The standard curve is a severe, fast-rising fire and it is what almost every published thickness table is based on. The slow heating curve matters for a narrower set of cases — smouldering scenarios, or buildings where the fire load burns for a long time at lower intensity. Reactive coatings behave differently under slow heating because the chemistry that produces the char has time to run before the steel is hot, and a product that performs well under the standard curve is not automatically the right answer under a slow one. If a project specification calls for slow-heating data, ask for it explicitly at enquiry stage; it is a separate assessment and not every product carries it.

How thick does the coating actually go?

That range is enormous in practical terms. At the bottom of it, a coating goes on in a single works coat and behaves like heavy paint. At the top, it is several passes with drying time between each, and the labour and programme implications outweigh the material cost. A thickness of 3000 μm on a large steel package is a genuinely different construction operation from one at 600 μm, and it belongs in the programme discussion, not only the purchase order. This is also where site application starts to compete with off-site: applying two millimetres of coating in a controlled shop is a different proposition from doing it on a scaffold in the rain.

How much strength does steel actually lose?

Dlubal, in a worked interpolation from that code, uses those two anchors to derive a factor of 0.498 at 591 °C — which is the practical shape of the problem. The curve is steep in exactly the band where critical temperatures fall. A hundred degrees is not a margin; it is most of the remaining capacity. This is why a design that pins a member’s critical temperature at 620 °C rather than 550 °C is making a real structural claim, not a paperwork convenience, and why the fire engineer rather than the buyer sets it.

What does the fire period mean in practice?

Buyers occasionally read a 120-minute rating as two hours of safety. It is better read as a comparative measure: a member protected to 120 reaches its critical temperature twice as far into the standard test as one protected to 60. What that converts to in a real fire depends on the fuel load, the ventilation and the compartmentation around it, which is the fire strategy’s job rather than the coating’s. The value of the classification is that it is reproducible. Two products carrying the same period under EN 13381-8 have been asked the same question and given the same answer, which is precisely what a purchasing decision needs.

Intumescent coating or board: which suits the job?

The decision is rarely about fire performance, because both routes can reach the same classification. It is about what the steel is for. Exposed structure in an atrium is a design feature, and boxing it defeats the point. Steel above a suspended ceiling is invisible either way, so the question becomes which trade can work faster in that space. Coating is also more sensitive to what happens after it is applied: it needs a compatible primer beneath it and, in many cases, a topcoat above it, and damage during follow-on trades has to be made good and recorded. Board is more robust to knocks and less forgiving of complex geometry.

What does a buyer need to send before we can quote?

We ask for them in that form because they are exactly the inputs the thickness tables need. Where a buyer only has section sizes and a period, we can quote a range and say honestly that it is a range — but we will not pick a critical temperature on the engineer’s behalf, because that number is a structural judgement about how hard each member is working. Nor will we quote a single average thickness across a mixed frame to make the number look tidy. If you want to see how we handle this kind of scope discipline on the sealing side of the same package, our note on specifying firestop in a tender follows the same logic, and passive fire protection sets out the wider package.

Where should the coating be applied?

Thickness is the reason to prefer the shop. Reactive coatings are measured wet and checked dry, and hitting a target within tolerance across a whole frame is far easier under a roof with a calibrated gauge and a controlled temperature than it is at height. The counter-argument is handling: coated steel gets chained, lifted and bolted, and every one of those operations can break the film. A workable compromise is to apply the bulk in the shop, leave connection zones bare, and complete them on site under a written touch-up procedure with recorded readings. What matters is that somebody owns the record. A frame with no thickness readings has no evidence behind its fire rating, whatever was bought.

Does the origin of the coating matter?

We source country-agnostically for that reason — India is one important origin among the network, not the default, and we do not operate local offices or hold local stock unless the quotation says so. What we fix first is the technical envelope: which periods, which section factor range, which critical temperatures, and whether slow-heating data is needed. Then we source against that envelope so the commercial comparison is genuinely like for like. If you are also weighing the suppliers rather than the products, our note on verifying a supplier covers what an audit can and cannot show, and FOB, CIF and DDP explains why two quotations on different terms are not two prices for the same thing.