An epoxy floor almost always delaminates for a reason below the coating: moisture or vapour pressure in the concrete, insufficient surface profile, laitance left in place, contamination by oil or a curing compound, osmosis, or poorly controlled thickness and recoating. Diagnosis starts with where the failure occurs (in the concrete, at the interface or between coats); a lasting repair treats the cause before any new coating goes down.
A floor that peels in sheets, sounds hollow or breaks out in blisters is more than a cosmetic problem. Lifted edges catch forklift wheels, wash water gets under the film and the failure spreads. Reapplying the same coat in the same spot often gives the same result a few months later. This guide helps maintenance and facility managers understand why an industrial epoxy floor delaminates and how to repair it for good, in Quebec and Ontario. It complements our general article on coating failure modes.
How do you recognize epoxy floor delamination?
Delamination shows three signs: areas that sound hollow under a hammer or chain, edges that lift or sheets that come loose, and blisters, either dry or filled with liquid. The back of a loose sheet is the most useful clue for what follows.
- Hollow areas. The coating is still in place but no longer bonded; these areas grow with traffic.
- Loose sheets. The coating comes away in pieces, often starting at a joint, crack or edge.
- Blisters. Closed bumps that burst under wheels; what they contain points the diagnosis.
- Cracks in the film. They sometimes follow an underlying joint or crack in the concrete.
Before any repair, photograph the floor, mark the areas on a plan and keep a few loose sheets.
Where does the failure occur and what does it reveal?
Failure occurs at one of three levels, each pointing to a family of causes: within the concrete, at the concrete-coating interface, or between two coats. Looking at the back of a sheet often settles the question.
| Where the failure is | What the back of the sheet shows | Likely causes | Confirming test |
|---|---|---|---|
| Within the concrete | A layer of concrete or cement paste stuck to the coating | Laitance, weak surface concrete, frost or carbonation | Pull-off test to ASTM D7234 on a sound area |
| At the interface | Smooth, clean back, concrete imprint without particles | Insufficient profile, contamination, moisture | Moisture test, contamination check, profile review |
| Between coats | Lower coat intact on the concrete, upper coat loose | Recoat window missed, amine blush, contaminated surface between coats | Review of application records and data sheets |
| Liquid-filled blisters | Liquid under the film, often alkaline | Osmosis or vapour pressure | Liquid analysis, moisture testing |
This table shows tendencies; an on-site assessment with testing confirms the cause. Our coating inspection service records these observations in a report.
Can concrete moisture make epoxy come off?
Yes. Concrete moisture is one of the most common causes of delamination: vapour rising from a slab, especially a slab on grade without a vapour retarder or a young slab, builds pressure under an impermeable film and can lift or blister it. Ordinary epoxy applied over overly damp concrete bonds poorly from the start.
- Slab on grade without vapour retarder. Ground moisture rises continuously, even through an old slab.
- Young slab. The concrete has not finished drying.
- Washdown and leaks. Water entering through a joint or drain wets the concrete under the film.
Tests are done before application using the method in the specification, for example ASTM F2170 (relative humidity in the slab) or ASTM F1869 (vapour emission rate). Our article on concrete moisture before epoxy coating explains these methods and moisture-tolerant systems.
A special case is osmosis: liquid-filled blisters form when water moves through the concrete toward soluble salts or water-sensitive components trapped under the film. The concentration difference draws water in, which collects and lifts the coating. The liquid is often alkaline.
- Signs. Closed blisters that keep coming back, liquid under the film, sometimes odour or staining.
- Factors. Damp slab without vapour retarder, salts in the concrete or aggregates, water-sensitive system components.
- Repair. Opening blisters is not enough: moisture must be addressed, the coating removed in the affected area and a suitable system chosen, sometimes with a moisture barrier.
Why do preparation and laitance cause peeling?
Preparation and laitance cause peeling because epoxy bonds mechanically to open, sound concrete: on a smooth, power-troweled surface or one covered with friable laitance, the coating bonds to a weak layer that gives way under load. Washing or acid etching alone does not replace mechanical preparation.
- Insufficient profile. The profile is chosen for the system thickness and described by CSP profiles; see our article on concrete surface preparation by shot blasting or grinding.
- Laitance. A weak layer of cement paste brought to the surface during finishing; it is removed by shot blasting or grinding.
- Degraded concrete. Scaling, carbonation, frost or chemical attack; the weak layer is removed and repaired first.
- Residual dust. Preparation not vacuumed leaves a layer that blocks adhesion.
Prior concrete repairs are covered in our article on industrial concrete repair before coating.
Which contaminants prevent adhesion?
Contaminants that prevent adhesion include oils and greases, curing compounds and sealers applied during construction, densifiers and surface hardeners, waxes and film-forming cleaners, and some process chemical residues. They form a film between concrete and coating, often invisible.
- Oils and greases. Common in shops and machine areas; they soak into concrete and bleed back out after coating.
- Curing compounds and sealers. Applied to a new slab, they remain on the surface unless removed.
- Densifiers and hardeners. They close the concrete's porosity and reduce bond; compatibility is checked before choosing the system.
- Waxes and cleaners. On an old coating or maintained concrete, they leave a film.
- Process products. Sugars, fatty acids or chemicals in production areas.
An adhesion test on a small trial area, after preparation, confirms that contamination has been removed.
Can poorly controlled thickness or recoating cause delamination?
Yes. A coat too thick in a single pass creates shrinkage stress that pulls on the bond, a coat too thin wears through and lets water in, and a coat applied outside the recoat window bonds poorly to the one before it. Application conditions matter as much as the product.
- Thickness. Follows the data sheet, pass by pass; readings are recorded.
- Recoat window. Once it is missed, the previous coat must be sanded; our article on pot life and recoat windows explains these limits.
- Cold and ambient humidity. Below the dew point or in cold weather, amine blush can form on the surface and block adhesion of the next coat; see our article on cold weather application.
- Active joints coated over. Joint movement cracks the film, then water gets in and lifts the edges.
How is a delaminated epoxy floor repaired for good?
A delaminated floor is repaired for good by confirming the cause, removing the coating back to a well-bonded area, treating the concrete (moisture, contamination, laitance, repairs), then applying a compatible system with clean termination edges. The repair is local if the cause is local, and general if the cause affects the whole slab.
- Assessment. Hollow area survey, examination of sheet backs, moisture and adhesion testing.
- Outline. Saw-cut repair edges in sound coating to avoid feathered edges.
- Removal and preparation. Removal of unbonded coating, shot blasting or grinding, degreasing.
- Treat the cause. Moisture-tolerant system, removal of the contaminated layer, concrete repair.
- Application. A system compatible with the existing one, under data sheet conditions, with thickness readings.
- Verification. Adhesion tests on the repair and records handed to the client.
Flooring is part of our protective coatings service; we work in industrial settings only.
Frequently asked questions
Can loose areas simply be glued down or recoated?
No. A loose area that is not treated at the cause will come loose again. Unbonded coating must be removed, the concrete prepared and the cause addressed before a new system goes down.
How can you tell whether the whole floor will come off?
A hammer or chain survey and adhesion tests on areas that look sound show whether the failure is local or widespread. The origin (general moisture or local contamination) also guides the answer.
Is the manufacturer responsible for delamination?
It depends on the cause. Most delamination comes from the substrate or the application, not the product. A documented assessment with testing supports the discussion with the parties involved.
Can delamination appear years after installation?
Yes. Rising moisture, water entering through a joint or oil contamination can act slowly. That is why regular inspection catches hollow areas early.
Our crews and qualified subcontractors work from LaSalle, including in Mauricie and Centre-du-Québec, Saguenay-Lac-Saint-Jean and Ontario. Send photos of the loose areas, the back of a few sheets and the floor's history through the request a quote form: pricing on consultation, reply within 24 business hours.
Hicham M, P Eng, PMP
Engineer and project manager (PMP) at Induscoat. Over 16 years of experience in industrial coatings, composite repairs and wear protection on mining, energy and petrochemical sites in Canada and internationally.
