Polyurethane Floor Coating Failure: Blistering, Delamination and Ambering

An industrial polyurethane floor coating rarely fails because of the resin. It fails at an interface, under a vapour pressure, or because the wrong grade was chosen for the light it receives.

Four patterns account for most of the failures we are called to: blistering, delamination, ambering and cracking, and each one names its own cause before anyone opens a tin.

Read the pattern before you read the datasheet

A floor failure has a shape, and the shape is evidence. Circular blisters mean pressure under the film, straight cracks mean movement above it, and a soft patch means chemistry reached it.

Age of the failure is the first thing to record and the last thing anyone writes down. A sheet that lifts in 6 weeks points at a missed overcoat window or at dust; the same sheet lifting after 3 years points at water arriving at the perimeter. Two identical photographs of PU blistering taken 2 years apart describe two different problems.

The instinct is to blame the product. In our project files, fewer than one in five failures we investigate turn out to be a defect in the material supplied.

That matters commercially, because the remedy differs. A material defect is a warranty claim; an interface or substrate problem is a repair, and the two are priced differently by everyone involved.

An industrial polyurethane floor coating is a layered system, and each layer can fail at its own interface. That is why a photograph of the surface is rarely enough to diagnose a floor failure, and why we ask for one of the opened edge.

So the first job is to decide which interface failed: resin to primer, primer to slab, or the film itself, which is where how the binder reacts stops mattering and the substrate starts. A knife and a moisture meter answer that in 20 minutes.

Figure 1: domes with a damp underside mean vapour, not a bad batch.

Where a plant has several rooms, the pattern usually repeats in the room with the same conditions. Finding it once is often enough to explain the rest of the building.

Why does PU blistering happen?

PU blistering appears as domes with a hollow centre, usually 5 to 40 mm across. They form because something is pushing from below and the film is strong enough to resist it until it lifts.

The pressure is almost always water vapour. A slab without a damp-proof membrane releases moisture continuously, and a resin film is a vapour barrier by definition.

An osmotic blister is the same mechanism with a different driver. Soluble salts in the slab draw water through the film by osmosis, and the blisters tend to be smaller, denser and concentrated in patches.

The test is simple: open one blister, and which medium reached it usually shows in the colour and the smell before any laboratory is involved. Damp concrete underneath means vapour; a dry, dusty surface means the primer never bonded; a liquid with a smell means a solvent was trapped.

The size of the domes matters too. Large, soft blisters suggest a slow vapour source across a whole slab, while small, hard ones suggest a local contaminant such as oil or a patch of release agent.

Figure 2: failure that starts at the edge is usually a perimeter water path.

Why does sheet lifting start at an edge?

Delamination usually begins at the perimeter of a room or at a joint, and spreads inward. It is the film separating from what is underneath, and it almost always starts where preparation was hardest.

Four causes cover most cases: dust left after grinding, a primer failure that never cured, a missed recoat time, and moisture arriving at the interface from a leaking drain or a failed seal.

A recoat window problem is the most common of the four and the easiest to prevent. Every product has a window within which the next coat will bond, and after it the surface has to be abraded again.

Where a floor fails in sheets within weeks of installation, the cause is almost always that window or the dust. Where it fails after 2 years at the edges, suspect water getting in at the perimeter.

Perimeter detail is worth checking on any industrial polyurethane floor coating that fails at the edge. A cove that has debonded, a sealant that has shrunk or a threshold that lets water under are all entry points for the same problem.

Primer failure: the interface nobody inspects

Primer failure is invisible once covered, which is why it causes the most expensive repairs. The bond between primer and slab is the weakest link in any industrial polyurethane floor coating, and nobody looks at it again after day one.

The usual triggers are a slab that was too damp for the primer chosen, laitance left in place, or a primer applied at the wrong coverage so that the surface stayed dusty.

Pull-off testing on a trial area is the only reliable check, and it takes an hour. We ask for it on every project where the slab history is unknown, because it is cheaper than a replacement floor.

Where a primer has failed under a sound topcoat, the repair is a removal rather than a patch. The film can look perfect and still be detached over square metres.

Where a primer has failed, an osmotic blister often appears at the same time, because water is reaching the interface and reacting there rather than passing through. The two signs together are strong evidence of a slab problem.

How does moisture vapour cause a failure?

Moisture vapour moves through concrete continuously, driven by the difference in humidity between the slab and the air above it. A resin floor stops it, and the pressure builds underneath.

The reading that matters is not the surface dampness but the emission rate, measured with the correct method for the slab. A plastic sheet taped down overnight is a rough indicator and a poor specification basis.

Three remedies exist: a surface damp-proof membrane, a moisture-tolerant primer system, or a ventilated detail at the perimeter. Each has a cost, and the choice depends on the emission rate rather than on preference.

Where a plant cannot tolerate the downtime for a membrane, a moisture-cured primer is often the practical answer, provided the emission rate is within its declared limit.

An industrial polyurethane floor coating specified without a moisture reading is a bet on the slab. We have never seen that bet lost gracefully, and the repair is always more expensive than the test would have been.

A membrane adds cost and a day, and it is the only reliable answer above the emission limit. Below it, a moisture-tolerant system works, and the difference between the two options is worth stating in writing.

Ambering: cosmetic, predictable, and often avoidable

Ambering is a colour change rather than a structural failure. An aromatic binder reacts to ultraviolet light and turns from grey towards amber, usually within one to two summers.

The film stays sound in most cases, which is why we describe it as cosmetic. The exception is a thin clear sealer, where the colour change is dramatic and the owner reads it as a defect.

Where colour matters, an aliphatic topcoat is the answer, and it should be specified before the floor is ordered rather than after the first complaint.

Where it does not matter, the colour change is a maintenance note rather than a repair. Repainting an ambered floor that is structurally sound is a decision about appearance, not about service life.

Where a plant has already ambered a floor, the decision is not whether to fix it but whether the colour affects anything. In a racked warehouse it rarely does; in a customer-facing packing hall it usually does.

Cracking and what it is telling you

A crack in an industrial polyurethane floor coating is usually a message from the slab. Shrinkage cracks appear in the first months, movement cracks follow joints, and impact cracks radiate from a point.

The pattern distinguishes them. A crack that follows a construction joint is movement; one that wanders across a bay is shrinkage; a star pattern is impact from a dropped load.

Filling a crack without treating the cause is the most common wasted repair. Where the slab is still moving, the filler will fail within a season, and the second repair is more expensive than the first.

A crack that has been filled and reopened is telling you the movement is still there. At that point the answer is a joint, not another filler, and the industrial polyurethane floor coating should be re-detailed rather than patched again.

Where the failure is at a joint rather than in the field, the crack is telling you about the building. A joint that has opened and closed has to be detailed as a joint, and a filler chosen for appearance will fail at the same rate as the one it replaced.

How do you diagnose a failure in one visit?

A structured visit takes about 2 hours: photograph the pattern, open one defect, measure moisture, check the recoat history against the installation date, and read the cleaning records.

Table 1: The five checks and what each one rules out.

That sequence is deliberately ordered by cost. Opening a blister costs nothing and eliminates half the hypotheses before any instrument is used.

Record everything as you go. A file with photographs, readings and dates is what turns a floor failure into either a warranty claim or a repair, and it is the difference between a short conversation and a long one.

Where a floor fails in the first season, the installation record is the evidence. Mix ratios, batch numbers, slab readings and the time each coat went down are all recorded on a good job and lost on a cheap one, and the difference shows up only when something goes wrong.

Where the file is thin, the conversation defaults to blame. That is a commercial problem rather than a technical one, and it is the reason we photograph every stage of an installation even when nobody asks.

Repair options and what each one is worth

Diagnosis of a floor failure in one visit follows a fixed order: read the shape, open one defect, test the slab, then price the repair. The decision turns on how much of the film is detached and whether the cause has been removed.

Table 2: Repair choice by failure extent.

Repair cost is driven by area of detachment, not by area of visible damage, and a moisture vapour problem is priced by the bay rather than by the blister. A blister is the visible part of a larger unbonded patch, so the repair boundary has to be found by tapping outward from the defect until the note changes. Patching to the visible edge is how a second repair gets scheduled inside a season.

The trap is patching a primer failure. A patch bonded to a detached film has nothing to hold on to, and it will lift with the surrounding area at the first thermal cycle.

Table 3: What each failure pattern points to, in order of likelihood.

Preventing the four failures we see most

Almost every floor failure we investigate could have been prevented by four decisions at specification stage, and none of them adds much to the price. All four are written into the build-up that was specified when the specification is written properly.

Measure moisture properly, test adhesion on a trial area, write the recoat window into the method statement, and specify the binder for the light the room receives.

Every repair should be photographed before and after, with the date and the area written on the same sheet. That record is what a warranty conversation is conducted on, and it is also the baseline figure for the next failure in the same room.

That is the whole list. It is short because the mechanisms are simple, and it is ignored because each item costs a day or a test that a cheaper bid can leave out.

The four checks are also the basis of a fair comparison between bids. A supplier who has measured moisture and tested adhesion is quoting an industrial polyurethane floor coating for this building; one who has not is quoting a floor for a building like it.

It is worth asking a bidder what would make them walk away from a project. A supplier who names a moisture level, a slab condition or a temperature has a limit; one who never has a limit is selling an industrial polyurethane floor coating without a warranty they expect to honour.

Where a client asks for a specification review, we run those four checks first. In most cases the answer is not a different resin but a different sequence.

That is also the most useful outcome for the client. A sequence change costs a day; a resin change costs the contract, and in most of the cases we see the resin was never the problem. Two failures in the same room within a year point at the specification of the industrial polyurethane floor coating rather than at the batch.

Two failures in one room rarely share a cause, but usually share a detail. Before the third repair, list them by location and look for the construction detail they have in common; the pattern is usually a perimeter, a drain or a joint, not a batch. If the floor is under warranty, photograph and sample before anyone patches.

A repair removes the evidence that would settle the claim, and is worth more than the patch. Inside the warranty period the sequence differs from a repair: record, notify, then agree the method before anyone lifts a section. The repair specification should name the material family, not the colour, because a patch in a different chemistry will fail at its own edge.

For a plant with several failing rooms, that changes the order of work. Fix the sequence and the drainage first, then repair the floors, and the repairs last longer than they would if the causes were still present. On a site with three failing rooms, that sequence is usually the difference between one shutdown and three, which is the number most plants remember.

Frequently asked questions

Why is my polyurethane floor blistering?

Almost always moisture vapour pushing from the slab, or a solvent trapped under the film. An osmotic blister has the same look and a different cause, so open one and inspect the underside.

What causes delamination at the edges of a room?

Water entering at the perimeter, or preparation that was weakest where access was hardest. Sheet failure within weeks is usually a missed recoat window or dust.

Does ambering mean the floor has failed?

No. It is a colour change from ultraviolet exposure in an aromatic binder, and the film usually stays sound. Where colour matters, an aliphatic topcoat is the fix.

Can a blistering floor be repaired without replacement?

Under about 5 per cent of the area, with the cause removed, yes. Where the primer has detached over large areas, patching will fail and removal is the honest option.

How do I know whether the slab is the problem or the primer is?

An emission test rather than a surface reading. Opening a blister and finding damp concrete underneath is the fastest confirmation that the slab is driving the failure.

What is a recoat window and why does it matter?

The period after which a coat will no longer bond chemically and must be abraded. Missing it is the commonest cause of delamination in floors that fail within a year.

Is a failed floor covered by warranty, and what has to be recorded before anyone repairs it?

Only if the cause is a material defect. Substrate moisture, missed windows and cleaning chemistry are normally excluded, which is why the diagnosis matters before the claim is made.

Why owners call Kaida Paint after a failure

Kaida Paint investigates PU blistering and sheet lifting with the same five checks described here, and we report the outcome whether or not it favours us. Where the cause is the slab we say so and quote the membrane, rather than selling a thicker industrial polyurethane floor coating that would fail in the same place for the same reason.