Epoxy Floor Peeling from a Weak Substrate: Causes and Fixes

An epoxy floor peeling in patches and exposing the concrete beneath is one of the most frustrating failures for a project owner: a floor coated not long ago is already damaged, and recoating exactly the same spot only peels again after a short while. In many cases, the fault does not lie in the topcoat but in the slab supporting it — concrete too weak to hold the coating firmly.

This article discusses one cause only: peeling caused by a weak substrate. That is the case where the concrete surface has low strength, a weak and crumbly cement surface layer, a lot of dust, or poor internal cohesion, so that the coating separates together with a thin layer of the substrate. This article does not discuss peeling caused by subfloor moisture — a different cause with its own mechanism and testing methods — nor does it repeat the general diagnosis of floor failures or the detailed patching process, since both have their own articles (see the links at the end).

Illustration of peeling: an old coating cracked and peeling away, exposing the surface beneath. The photograph is not a Kadier epoxy floor. Photo: Eva Bronzini / Pexels.

1. How does peeling from a weak substrate differ from other types of peeling?

The most important distinguishing point is the separation surface. When a coating peels because of poor adhesion to a substrate that is still sound, the separation surface is usually relatively clean: the coating comes away while the substrate beneath remains hard and intact. When peeling is caused by a weak substrate, the coating usually pulls a thin layer of substrate material with it — cement dust, small aggregate grains, or a thin patch of concrete surface — stuck to the underside of the peeled piece. In other words, the coating does not separate from the substrate; the substrate itself separates just beneath the coating.

This difference decides the remedial direction. If the substrate is still sound, the main work is to clean and re-prepare the surface and then recoat. If the substrate is weak, recoating the same weak surface means the new coat still adheres to a substrate that will separate again. In that case, the most important work is not the new coating but treating the substrate until it is strong enough before coating.

A second distinguishing point is the extent. Peeling from a localised contamination spot is usually confined to that spot. Peeling from a weak substrate usually spreads across a zone: a whole area of equally poor-quality substrate peels in the same pattern, even though the coating above was applied the same way throughout. So when peeling spreads across a zone rather than appearing at just a few isolated points, substrate quality should be suspected first, rather than only re-examining the coating.

2. Signs of a weak substrate, before and after peeling

Before application, a weak substrate usually shows itself through signs that can be seen and felt. The surface gives off dust easily when swept or rubbed by hand: after sweeping, a fine layer of dust remains, or light rubbing already dislodges cement grains. The surface scratches easily: a hard object drawn lightly across it leaves a deep groove and crumbled powder, rather than just a faint scratch on a hard surface. There may be honeycombed, porous areas, or a thin surface mortar layer that lifts slightly when tapped. These signs do not replace measurement, but they are good reason to stop and inspect more carefully rather than priming immediately.

Once the coating has peeled, look at the underside of the peeled piece and the exposed substrate. If the underside carries a lot of dust and substrate grains, and the exposed substrate is also crumbly and easily scratched into powder, the cause very likely lies in the substrate. If peeling is widespread and the peeled pieces show the same type of separation surface, that is further evidence of a systemic substrate cause, not an isolated application error at one spot.

Signs of a weak substrate need to be distinguished from signs of subfloor moisture. Subfloor moisture usually comes with blistering in bubbles, dampness or ingress marks, and peeling that may recur with the wet season. A weak substrate can peel even when the slab is dry, and its separation surface carries substrate material as described above. When both sets of signs appear together, both causes need to be addressed, because strengthening the substrate without resolving moisture still leaves the coating at risk of failing again.

3. Why does a weak substrate destroy adhesion?

An epoxy coating adheres to the substrate through its bond with a sound, clean concrete surface. That bond is only as strong as the weakest link in the chain: coating — primer — substrate surface — concrete body. When the substrate surface is a weak, crumbly, dusty cement layer, the weakest link lies within the substrate itself. Loads in use — forklifts passing, heavy objects dragged across, temperature changes making the coating expand and contract — will tear apart precisely that weak substrate layer, and the coating peels with it.

Several common forms of weak substrate occur. The first is a weak cement surface layer, sometimes called laitance: a thin layer rich in cement and water that rises to the surface during compaction and trowelling, and when dry becomes brittle and crumbly, adhering poorly to the concrete body beneath. Coating epoxy over this layer is little different from sticking something onto a layer of powder: when the powder separates, the coating goes with it. The second is a honeycombed, porous concrete surface caused by inadequate compaction or by a mix and curing regime that was not good enough, leaving the surface insufficiently dense to anchor the primer. The third is a deteriorated substrate: an old surface worn down, weathered, or previously coated with an old coating that has itself become weak, so that the new coating adheres to an intermediate layer that is no longer strong enough.

One often-overlooked cause is a substrate that had not yet reached a stable state when coated. Concrete that is still young and has not yet developed its full strength, or a patching mortar repair that has not yet hardened enough, still has a weak surface that keeps changing after coating. The coating hardens faster than the substrate beneath it, and that mismatch concentrates stress in exactly the weakest surface layer.

4. Testing the substrate, at the level of principle

Testing for a weak substrate does not need to start with complex equipment; the sensible sequence runs from observation, to simple trials, and only then to quantitative measurement when a major decision is at stake.

The first step is to survey the whole floor, not just the peeled area. Record the dusty, honeycombed, cracked, and crumbly-surface zones, and compare them with the extent of peeling: if the peeled zones coincide with the poor-substrate zones, the weak-substrate hypothesis becomes stronger. Tap around the edges of peeled areas to establish how far peeling extends, since peeled edges commonly extend further than the eye suggests.

The next step is simple surface trials: scratching and rubbing to assess crumbliness and dust levels, and checking whether the surface layer separates in thin patches. These trials give only a qualitative conclusion — whether the substrate shows signs of weakness — not a number for acceptance. For a number, a pull-off adhesion test is needed, in which a test dolly is bonded to the surface and pulled off to see the failure load and the failure location: failure within the substrate shows the substrate is weaker than the bond, while failure at the interface shows an adhesion problem. How to read the results of this test, and the meaning of strength indicators, are covered in the two in-depth articles on adhesion strength and on the pull-off method, linked at the end.

An important principle: test results are valid only for the exact area tested. A floor can have sound zones interspersed with weak zones, so testing one point and extrapolating to the whole floor easily leads to a wrong decision. When peeling spreads across a zone, testing should be done zone by zone, and remedial decisions should also be made zone by zone.

5. Remedial approaches, at the level of principle

Remediation of peeling caused by a weak substrate follows a single principle: the substrate must be strong enough before recoating. Any plan to overcoat a weak substrate without strengthening it only postpones the next peel.

For a localised peeled area on a substrate that is generally still sound, the approach is to remove thoroughly the peeled portion together with the weak substrate layer clinging to it, until a sound surface is exposed, and only then re-prepare the surface and recoat with the correct coat system from primer to topcoat. Surface preparation — grinding, dust removal, crack repair — has its own detailed article, as does the process for patching and recoating individual areas; this article does not repeat the detail of either, but states only the principle: do not apply a new coat over any remaining weak substrate, because what remains will be the starting point of the next peel.

For widespread peeling caused by a systemic weak substrate, spot repairs are usually not enough. The approach then is to treat the substrate at zone scale: remove the weak surface layer across the whole zone, strengthen or rebuild the surface layer with a suitable material, and only then recoat the whole zone. The choice of strengthening material and coating system must rest on an actual survey and on the technical documentation of the specific product, not on habit. Among the Kadier systems, the roller-applied system and the self-levelling system serve different needs in terms of thickness and function; which system suits a strengthened substrate must be checked against the technical documentation of that system and the actual service conditions of the floor.

One mistake to avoid is applying an extra thick primer coat in the hope that the primer will “hold” the weak substrate together. A primer’s role is to penetrate and anchor into the surface, but it does not turn a crumbly substrate layer into a sound one. If the weak surface layer is still there, the primer only adheres to that weak layer, and the whole coating system still separates with it under load. Conversely, grinding away the weak layer is a step that no coating, however thick or good, can replace.

Finally, the decision between spot repair and zone rework should rest on three factors: the peeled area relative to the floor area, the degree of substrate weakness found by testing, and whether the substrate is likely to keep deteriorating. When peeling has already spread widely and the substrate is weak over a wide area, reworking the zone from the start is often more durable and less costly in the long run than repeated patching.

6. Prevention from the start: three decisive points

Preventing peeling from a weak substrate is far cheaper than remediation, and only requires holding three points. First, do not coat a substrate that has not been inspected: take time to observe and test the surface for crumbliness and dust before deciding to coat, especially with an old substrate or one that has just been repaired. Second, remove the weak surface layer before coating: grinding down to a sound surface is a mandatory step, not an optional one, when the substrate shows signs of weakness. Third, apply the correct coat system from primer to topcoat in accordance with the product’s technical documentation, and only bring the floor into service once the coating system has stabilised as the manufacturer directs — bringing it into service too early loads an unfinished system and an unstable substrate at the same time.

In addition, during service, pay attention to zones of substrate that showed signs of weakness even before peeling: those zones belong on the periodic inspection list, so that peeling is caught early while it is still small enough for spot repair.

7. Note on specifications

This article presents principles and methods for recognising, testing, and treating peeling caused by a weak substrate at a general framework level; it gives no adhesion strength, compressive strength, or other product-specific technical value for any Kadier product. The reason: Kadier had not published an official product technical data sheet (TDS) for its coating systems at the time this article was written. Any strength value, consumption rate, or specific application condition for a Kadier system applies only after checking against the official TDS of that very system, or against actual test results for the project. The standards TCVN 9014:2011 and TCVN 12574-2018 mentioned in the article serve only as general reference bases; they do not replace the product TDS and do not replace an actual survey of each project. This article also gives no price, makes no warranty commitment, and cites no specific project or certification, since those subjects require verified real figures and records.

Uncategorized