Why Does Epoxy Peel Off a Garage Floor?

We get called to a lot of peeling garages, and the homeowner almost always opens with some version of “the epoxy was bad.” It usually was not. Modern coating chemistry is very good. What fails is the connection between the coating and the concrete, and that connection is made during prep — hours before anyone opens a can.

First, diagnose it yourself in thirty seconds

Find a loose piece, pull it up and turn it over.

• Grey concrete stuck to the back. Good news, in a way. The coating held onto the concrete and the concrete itself came apart — meaning the surface layer was weak. This is the signature of coating over laitance, or over a slab that was never properly profiled.
• Clean, smooth back, like the underside of a sticker. The coating never bonded at all. Something was between it and the concrete: a sealer, a curing compound, dust, moisture, or a surface too closed to key into.
• Chalky white powder on the back. Moisture. Water has been travelling up through the slab, carrying minerals with it, and depositing them right at the bond line.

That single check narrows the cause enormously. Here is the full list.

Cause 1: acid etching instead of grinding

This is the big one, and it is the reason most DIY kits fail. Acid etching is what the instructions in the box tell you to do because it is the only prep a homeowner can do without renting equipment. It cleans the surface and opens it a little. It does not remove the dense, closed laitance layer that formed when the slab was troweled.

The result is a coating firmly attached to a layer that is itself barely attached. Everything holds for a while. Then heat, a hot tire, or ordinary seasonal movement shears it, and the floor comes up with grey concrete on the back.

Diamond grinding removes that layer outright. It is loud, dusty and slow, and it is the single most important hour of the whole installation.

Cause 2: moisture coming up through the slab

Concrete is porous and a slab sits on ground that is sometimes wet. In a garage with no vapor barrier under it — routine in North Texas homes built before the 1990s — moisture moves upward continuously. Bare concrete lets it evaporate off the top and nobody notices. Seal the top and it has nowhere to go, so it collects at the bond line and pushes.

The tell is bubbles or domes rather than flat peeling, often appearing in the same spots each spring, and that chalky deposit on the back of a chip. It is also why we take a moisture reading on every slab before deciding what goes on it. If the numbers are high, the floor gets a moisture-mitigating primer. Skipping that step to save an hour is how a floor blisters a year later.

Cause 3: coating over an existing sealer or curing compound

Builders frequently spray a curing compound on a new slab to control how fast it dries. It is invisible. It also sits on the surface for years and nothing will bond through it. The same applies to a penetrating sealer a previous owner applied for dust control.

There is a crude field test: splash a cup of water on the bare slab. If it beads and sits, something is sealing it. If it darkens and soaks in within a minute, the concrete is open. Grinding removes all of it either way, which is the point.

Cause 4: applied in the wrong window

Epoxy in particular has a real temperature floor. Below roughly 55°F it stops cross-linking properly and you get a floor that never reaches full hardness and stays slightly soft forever. Too hot is its own problem — the pot life collapses, the material kicks in the bucket, and what goes down is already partly reacted.

North Texas gives us both extremes. This is a large part of why our standard system is polyaspartic: its usable temperature range is far wider, and it is much harder to install on a bad day.

Cause 5: film too thin, plus a hot tire

A single-coat roll-on system might be 8 mils thick. A ground-and-broadcast floor is 20 to 24. Thin film has less mechanical strength and less thermal mass, and a tire that comes home hot in August can soften it, stick to it, and lift a plate-sized disc when you back out.

If your garage has round or oval bare patches roughly the size of a tire contact patch, this is what happened. It is the most common single failure we are called out for, and it is essentially always a thin coating on marginal prep — the two problems compound.

What fixing it actually involves

There is no patch. Whatever caused the failure is still under the floor, and new material bonded to failing material fails with it. The repair is:

1. Full removal. Grind the old system off down to sound concrete across the whole slab, not just the loose areas.
2. Diagnose the cause. Moisture test, check for sealer, look at what the old bond line tells us.
3. Address it. Moisture primer if the readings call for it. Crack repair. Full re-profile.
4. Install properly. Base coat, full flake broadcast, UV-stable topcoat.

The good news is that the concrete underneath is nearly always fine, and a ground slab that has had a failed coating removed usually has a better profile than it ever did new. You are not starting from a worse position than a fresh garage — you are just paying for the removal.

How to not be here again

Ask the next contractor one question: how are you opening the concrete? If the answer involves acid, a pressure washer, or the word “clean,” you are buying the same floor you already have. If it involves a diamond grinder and a profile, you are buying a different one.

What we check before quoting a failed floor

When we come out to a peeling garage, the quote depends almost entirely on four findings, and we would rather you know what we are looking at:

1. How much is still stuck. Counter-intuitively, a floor that is 90% loose is cheaper to fix than one that is 90% bonded. Loose material comes up fast; well-bonded material has to be ground off.
2. What is on the back of the chips. Grey concrete, clean plastic, or white powder — each one points at a different cause and a different remedy.
3. Slab moisture. A meter reading, taken in several places. If it is high, the new floor needs a moisture-mitigating primer, and that has to be in the quote rather than discovered later.
4. Whether the cracks have moved. A crack that was filled and has re-opened tells us the slab is active, which changes how we treat the joints.

The one case where the coating really was the problem

It is rare, but it happens: a water-borne epoxy applied at the right thickness over properly ground concrete can still fail if it was installed outside its temperature window, or if the two components were not mixed in the correct ratio and mixed thoroughly. Under-mixed epoxy never fully cross-links, and you end up with a floor that stays faintly soft forever and picks up every tire mark.

You can usually identify it because the failure is uniform rather than patchy — the whole floor is slightly tacky or rubbery rather than lifting in specific places. If that is what you have, the cause is the mix or the conditions, not the prep.

How long a proper repair takes

For a standard two-car garage, removal and reinstallation is usually a single long day, occasionally two if the old coating fights us or the moisture readings call for a primer that needs to cure before the base goes on. You will get a straight answer on which one it is after we have seen the floor, not after we have started grinding.