Laying a Resinous Floor

The slab decides, not the datasheet

Every resinous failure worth arguing about traces back to a decision made before the first pail was cracked. Delamination, blistering, osmotic bubbling, edge peel, telegraphed cracking — none of these are coating defects in origin. They are substrate conditions that the coating faithfully reported some weeks later, when the client was already walking on it.

So the sequence a resin applicator works to is not mix, prime, coat, topcoat. It is: find out what the slab is doing, decide whether it can be made to stop doing it, and only then choose a system that suits the answer. A slab that is still giving up water needs a different specification from one that is bone dry but polished to a closed, glassy skin. A slab contaminated to 4 mm depth with hydraulic oil needs neither — it needs removal.

Two properties govern nearly all of it. Surface profile determines whether the resin has anywhere mechanical to key into, because on a dense trowelled slab chemical adhesion alone is a coin toss. Moisture — both free water and vapour driving up from below — determines whether that key survives, because hydrostatic and osmotic pressure will lift a perfectly bonded film off a perfectly profiled slab if the drive is high enough and the coating is impermeable enough.

Read those two before pricing, before scheduling, before ordering. A resin job that has been sold on square metres of coating and then discovers a 95 percent RH internal reading is no longer the job that was sold. That conversation belongs at survey stage, not on day two of a shutdown.

Slabs that are still wet: reading the drive, not the surface

Concrete does not dry in the way that word suggests to a client. A ground-bearing slab loses water from its upper surface long before its lower half has equilibrated, and a slab on a well-detailed vapour barrier still holds a large reservoir of mix water that has nowhere to go but up. Touch-dry means nothing. What matters is the relative humidity deep in the slab, because that is what will drive vapour into the underside of a cured film.

In-situ probe testing to ASTM F2170, Standard Test Method for Determining Relative Humidity in Concrete Floor Slabs Using in situ Probes, is the method most resin specifications now call for, and it exists precisely because surface tests are easily fooled. Probes are drilled to a specified fraction of the slab depth — different for slabs drying from one side versus two — sealed, and left to equilibrate before reading. The equilibration period is not optional padding; a probe read too early reads the drilling, not the slab. Where surface-based methods are still specified, ASTM F1869, Standard Test Method for Measuring Moisture Vapor Emission Rate of Concrete Subfloor Using Anhydrous Calcium Chloride, may appear instead, and in UK and European practice BS 8203 and BS 8204 describe hygrometer-based assessment. Which regime applies is a contractual matter — the specification, the manufacturer's warranty conditions and the local standard between them decide the threshold, and those thresholds genuinely differ.

What the number changes is the system, not the effort. Below a manufacturer's stated limit, a standard primer is fine. Above it, the options narrow to a moisture-tolerant epoxy vapour-suppression primer applied at a film thickness the manufacturer states for that RH band, or a change of specification entirely. Applying a normal primer to a wet slab and hoping is how you buy a re-do at your own cost eighteen months later.

Watch for the localised sources that a probe grid misses. A slab cast against a wall with a failed damp-proof course, a floor duct that sweats, an area under a leaking chiller, a section of slab that had standing water during construction. Blistering that appears as a rash in one bay and nowhere else is almost always a local drive, not a batch fault. Map wet zones on the survey drawing and price them separately.

New slabs bring a second timing problem. A slab that lost surface water too fast during finishing has a weak, crazed upper skin that grinds away to reveal a sound layer beneath — good, if you allowed for the extra removal. Assessing evaporative conditions at the time of the pour, when you are coordinating with the concrete contractor on a fast-track programme, tells you how much of that skin to expect.

Slabs that are dry but closed: opening the surface

A power-floated slab that passes every moisture test can still reject a coating, because a burnished surface presents almost no mechanical key and often carries residual curing compound. Resin sitting on that surface bonds to a thin, weak layer that shears away under wheel loads. The remedy is mechanical, and there is no chemical shortcut that reliably substitutes for it.

Concrete Surface Profile, as defined by the International Concrete Repair Institute in Guideline No. 310.2R, Selecting and Specifying Concrete Surface Preparation for Sealers, Coatings, Polymer Overlays, and Concrete Repair, is the language to use here. The CSP chips run from a near-smooth acid-etched texture through progressively more aggressive profiles to heavy scarification. Coating manufacturers state a required CSP band against system thickness — thin roller-applied films need a shallower profile, self-smoothing screeds and broadcast systems need a deeper one. Specifying by CSP number rather than by machine type is what makes the requirement enforceable, since two operators with the same grinder can produce very different surfaces.

Method follows the target. Diamond grinding opens a closed surface and removes curing compound with reasonable dust control on a vacuum-shrouded head. Shot blasting produces a more uniform, deeper profile quickly across open floor, and is the workhorse for large industrial areas — but it rides over undulations rather than correcting them, and it will not reach within a hand's width of walls and columns. Scabbling and scarifying go deeper still and bruise the substrate, which then needs a further pass. Whatever the primary method, budget properly for edge work; the hand-ground perimeter strip is where most programmes lose a shift they did not plan for.

Pull-off adhesion testing to ASTM D7234, Standard Test Method for Pull-Off Adhesion Strength of Coatings on Concrete Using Portable Adhesion Testers, applied to a trial patch, settles arguments before they become claims. A cohesive failure in the concrete is the result you want. A clean break at the interface says the preparation was insufficient, and finding that out on a two-square-metre sample is a good day.

Dust is not a nuisance, it is a bond breaker. Vacuum, then vacuum the vacuum's path, then check by pressing a clean gloved hand to the floor. Compressed air moves dust sideways into the wet edge you are about to work into.

Slabs that are contaminated: what grinding will not fix

Oil, grease, hydraulic fluid, sugars, food acids and silicone all defeat resin adhesion, and unlike laitance they are not confined to the surface. Contaminants migrate down the pore structure and back up again under heat, which is why a workshop floor can look clean after grinding and then show fisheyes and craters within minutes of the primer going down.

Test rather than assume. Water dropped on the prepared surface should darken it and soak in within a short period; beading indicates residual contamination or an unbroken sealer. Do this in a grid across the area, because contamination is patchy and follows the machine positions, drip trays and traffic paths of the previous occupant.

Degreasing precedes preparation, not the other way round. Grinding contaminated concrete drives oil deeper and smears it across clean areas. Alkaline degreasers with hot water and mechanical agitation, followed by thorough rinsing and drying, will handle surface loading. Deep-set oil in a slab that has been soaking for decades may need the top layer physically removed — and where it cannot be removed, the honest answer is a system chosen for its tolerance, or a different floor finish entirely.

Legacy coatings deserve their own decision. A sound existing epoxy can sometimes be abraded and overcoated, but only after establishing what it is, whether it is bonded, and whether the new system is compatible with it. Old chlorinated rubber, bitumen or a failing polyurethane comes off. Sample-test in a corner; if the existing film lifts in sheets when scraped, the whole floor is coming up regardless of what the tender allowed.

Sawn joints and cracks need resolving at this stage too. Movement joints must be carried through the resin, not coated over — a resin film bridging a moving joint will crack in a straight line and take a strip of coating with it. Static cracks are opened out, filled with a resin mortar and allowed to cure before the main application.

Slabs that are out of tolerance: flatness and falls

Surface regularity is the third condition that gets decided before resin is opened, and it is the one clients notice most. A thin roller-applied coating follows every undulation in the substrate; it does not correct anything. A self-smoothing resin screed evens out small deviations but will not build falls or fix a slab that dishes across a bay.

Where a floor must drain, the falls have to exist in the substrate. Resin cannot be laid to a fall by trowelling in the way a sand-cement screed can, except in the heavier trowel-applied polyurethane screeds, and even those have limits on how much fall they can hold before slumping. If a wash-down area needs a fall to gullies and the slab is flat, that is a screeding operation before the resin trade arrives.

Flatness tolerance classes are set by the relevant national standard and by the specification. In the UK, Concrete Society Technical Report 34, Concrete Industrial Ground Floors, is the usual reference for industrial floor flatness, while ASTM E1155 covers the F-number method used in North American practice. Do not assume the incoming slab meets the class the resin specification wants; survey it, record the deviations, and issue them before you take possession of the area. A resin contractor who coats an out-of-tolerance slab has, in practical terms, adopted the tolerance problem.

Where levelling is needed, an underlayment before the resin adds a further cure period and a further moisture consideration to the programme, since the levelling layer itself has to dry down to the resin manufacturer's limit.

Coverage on a real profile, not a flat plane

The moment preparation is signed off, the quantity question changes shape. Manufacturers quote coverage against a theoretical smooth surface, and the surface you have just created is deliberately not smooth. A CSP band opened for a broadcast system has substantially more real area than its plan area, and the primer coat fills that texture before it builds any film at all. Under-ordering here is one of the most common commercial errors on a resin package.

Porosity compounds it. A ground slab that has been opened up drinks the first coat, particularly in areas that were weaker to begin with, and a single primer pass over a thirsty substrate can leave dry patches that show through the finish as dull, absorbent shadows. Plan for the possibility of a second primer application on porous zones and price it as a contingency rather than discovering it mid-shift.

Estimating the base and build coats against the actual area, the stated wet film thickness and a realistic wastage allowance is what keeps the last bay from being coated out of a different batch. Batch variation in pigmented resins is real, and running short on a continuous pour is worse than over-ordering by a pail.

Working the pot life: conditions inside the film

With the substrate resolved, the remaining risks move into the material itself and the air around it. Epoxy and polyurethane systems are exothermic; once mixed, they generate heat, and heat shortens working time non-linearly. A batch that gives twenty-five minutes at one temperature may give half that on a warm afternoon, and a full pail left standing after mixing will run away far faster than the same material spread out on the floor.

Mix to the stated ratio by weight or volume as the manufacturer specifies, at the stated speed, for the stated time, scraping sides and base. Under-mixed resin cures soft in streaks. Over-fast mixing entrains air that surfaces as pinholes in a self-smoothing layer. Decant into a second clean container and remix briefly if the system calls for it — unmixed material clinging to the original pail wall is a recurring source of soft spots.

Substrate temperature must sit above the dew point by the margin the manufacturer states, and that margin exists to stop condensation forming on the surface immediately ahead of the wet edge. A slab is thermally massive and lags air temperature by hours; measuring air temperature and assuming the slab matches it is how amine blush and adhesion failures happen on a cold morning in an unheated warehouse. Take slab temperature directly and calculate dew point from measured air temperature and relative humidity.

Wet film thickness is checked as you go, with a comb, at a frequency that lets you correct before an entire bay is wrong. Applied too thin, an epoxy will not develop its stated chemical resistance and may cure with a patchy sheen. Applied too thick, a fast system can exotherm, discolour, wrinkle or crack. The gauge takes a moment; the remedy for a mis-gauged bay does not.

Ventilation matters for cure and for the people doing the work. Solvent-free systems still demand air movement in enclosed spaces, and confined-space rules apply in pits, tanks and basements regardless of what the resin label says about VOC content.

Cure, handover and the first six weeks

Resin floors are handed over long before they are fully cured, and that gap causes more disputes than any application defect. Light foot traffic, full foot traffic, wheeled traffic and full chemical resistance each arrive at different times, and every one of those times stretches in cold conditions. A client who parks a loaded forklift on a floor at the light-traffic stage will print the tyres into it permanently, and the argument afterwards is unwinnable without a written handover schedule.

Protect the floor during the cure window as though it were a finished product, because it is. Sheet it against dust and drips from following trades, keep doors closed against wind-driven grit, and hold the temperature — a floor left to cure in an unheated building through a cold snap may never reach its stated hardness.

Inspect and record before demobilising. Photograph the prepared substrate, log moisture readings with their dates and probe locations, keep batch numbers against areas, and file the adhesion test results. When a blister appears in nine months, that file is the difference between a defect investigation and an argument.

Set the maintenance expectation explicitly. Resin floors are chemically resistant, not indestructible; the wrong cleaning chemistry will dull a gloss finish and aggressive scrubbing pads will burnish it unevenly. A short written regime handed to the facilities team, naming what to use and what to avoid, keeps the floor looking like the sample the client approved.