By Pablo Figueroa, commercial flooring consultant
When an epoxy floor peels, my first question is never which product was used. My first question is where it let go. A floor that peels in clean sheets between coats is telling a different story than a floor that lifts off the concrete, blisters in one wet area, or cracks and lifts along a joint line. The location and the shape of the failure point to the cause, and the cause decides whether a repair will hold.
Here is the short answer. In my experience across hundreds of epoxy and resinous jobs, peeling usually traces to four causes, in this order: a missed recoat window between coats, an existing coating or sealer that was never bonded well in the first place, moisture working against the bond, and movement at joints and cracks telegraphing through the system. I check them in that order because the first two are the most common, the easiest to confirm from the peeled piece itself, and the ones a simple recoat will repeat if they are missed.
First, read the peeled piece
Before you call anyone, pick up a loose piece if you can do it safely and look at both sides. Is the underside smooth and the same color as the coat below it? That suggests separation between coats. Is concrete stuck to the underside, or is the underside clean while bare concrete shows below? That tells you about the bond to the slab. Is the area damp, discolored, or limited to low spots and areas near drains, coolers, or exterior walls? That points toward moisture. Is the peeling following a straight line or a crack pattern? That points toward joint and crack movement. Write down where the peeling is worst, where it started, and how old the floor is. That record is more useful than a general statement that the floor is failing.
Cause 1: The recoat window was missed
Epoxy coats bond to each other in two ways. Within the recoat window, the next coat can bond chemically to the coat below it. After that window passes, the surface needs mechanical preparation, usually sanding and a tack wipe, so the next coat has something to grip.
When that step is skipped, the system can look fine at handover and then separate cleanly between coats months later. In completed job records I review, delamination between coats after a missed recoat window, without sanding and tack wiping, is a recurring pattern, including on a job where the slab also carried an old cure and seal compound that complicated the bond.
How to check it: peeling that comes up in sheets, with a smooth underside and a smooth, intact coat still on the floor below, fits this cause. Ask for the install records. What were the coat dates and times, and what preparation was done between coats? If no one can answer that, treat it as a warning sign for any repair plan that simply adds another coat on top.
What fixes it: the loose material has to come off, the remaining sound coating has to be properly prepared, and the repair system has to be bonded to a prepared surface, not to a smooth, closed surface that already rejected one coat.
Cause 2: What was already on the slab was not bonded
A new epoxy system is only as sound as what it is bonded to. If it goes over an existing coating that was already losing its bond, the new system inherits that failure. The same is true over cure and seal compounds, old sealers, and other surface treatments that were never removed or tested. The new floor may be well mixed and well applied and still peel, because the weak layer was underneath it from day one.
How to check it: sound the floor. Walk it and tap suspect areas. A bonded coating sounds solid. A debonded area sounds hollow. Check edges, drain rims, door thresholds, equipment feet, and the edges of previous repairs first, because those details take the most water, impact, and movement. Small test areas, where a contractor opens a limited section to inspect the bond line and the slab beneath, confirm more than any surface inspection alone.
What fixes it: unbonded existing coatings need to come off to a sound surface before a new system goes down. Recoating over a hollow sounding area is not a repair. It is a delay, and the peeling usually returns in a larger area.
Cause 3: Moisture working against the bond
Moisture reaches a coating bond in more than one way. Vapor can move up through the concrete and push against a sealed surface. Liquid water can be present during installation from sources inside an operating building, such as cooler and freezer melt water, leaking pipes, and cleaning water that never fully dried. Water can also get behind a system afterward at gaps, edges, and wall lines, and roof leaks can attack a bond from above.
In job records, moisture related complications show up alongside the need for relative humidity testing and a moisture vapor barrier before coating, and alongside slow overnight cure and bubbling or soft sections when conditions were wet or cold. Cold slabs make this worse, because a slab that is well below the system minimum temperature will slow cure and weaken the result even when the room air feels warm enough.
How to check it: look for a pattern. Blisters, bubbles, discoloration, damp concrete under a lifted section, and failures concentrated in low spots, near drains, near coolers and freezers, or below a known roof leak all point this way. The proper next step is moisture testing of the slab before any recoat is priced, not after a second failure.
What fixes it: fix the water source first, test the slab, and use a system that matches the measured conditions, including a moisture vapor barrier where the test results and the system requirements call for one. A standard recoat over an active moisture problem will fail the same way.
Cause 4: Joints and cracks moved underneath it
Concrete joints and cracks are movement points. If a rigid coating is spread straight over untreated joints and cracks, that movement telegraphs through the coating. The coating cracks along the line below it, water and traffic work into the crack, and peeling starts along that line and spreads outward.
How to check it: map the peeling against the floor. Peeling and cracking that follow straight saw cut lines, control joints, or a visible crack pattern fit this cause. Also check whether joints were treated as their own scope item before coating, with the joint detail written into the proposal, or whether they were simply coated over.
What fixes it: joints and cracks need their own treatment in the scope, before the coating system, and the treatment has to match the movement and traffic the joint will see, especially under forklifts and hard wheeled traffic. Coating over an untreated moving joint and hoping it holds is not a plan.
A simple order of checks you can use
- Look at the underside of a peeled piece and note whether separation is between coats or off the concrete.
- Sound the surrounding floor for hollow areas, especially at edges, drains, thresholds, and old repairs.
- Map the damage against low spots, drains, coolers, exterior walls, and any known leaks, and arrange slab moisture testing if the pattern fits.
- Map the damage against joint and crack lines and ask how those joints were treated before coating.
- Ask for install records covering coat timing, intercoat preparation, surface preparation, and moisture testing.
That sequence costs little, disturbs little, and usually narrows a peeling floor to one primary cause before anyone prices a fix.
When a spot repair can work, and when it cannot
A spot repair can make sense when the peeling is small, isolated, the surrounding floor sounds solid, moisture testing is within the repair system limits, and the cause was local, for example a single untreated crack or a single damaged detail. A spot repair is a poor bet when large areas sound hollow, when peeling separates cleanly between coats across a wide area, when moisture is active, or when the same area has already been patched and failed again. In those cases the problem is the bond system, not one spot, and repeated spot repairs usually cost more over time than addressing the cause once.
Limitations
This article describes patterns seen across many completed epoxy jobs, typically in the range of about 330 to 1,860 square feet with a median around 850 square feet, and larger systems exist well beyond that range. Your floor has its own history, products, and conditions. Moisture readings, slab temperature, and the actual bond condition of your existing floor outrank any general pattern here. Treat this order of checks as triage, not as a verdict on your specific floor.
Frequently asked questions
Our epoxy floor is only a couple of years old and already peeling. Was it a bad product?
Occasionally a product contributes, but look first at the four causes above. Process and moisture problems commonly peel in clean sheets at one interface or in one wet pattern area, which is a clue worth checking before blaming the material alone.
Can we just recoat over the peeling areas?
Only after the cause is fixed and the loose material is removed to a sound, prepared surface. Recoating over a missed recoat window problem, an unbonded old coating, or active moisture repeats the failure.
Why is it peeling at drains, edges, and joints first?
Because that is where water, movement, and hand finished details concentrate. The open field of a floor is the easiest part to install well. Drains, cove, joints, and thresholds are where conditions test the bond first.
How long should a repair take to assess?
A first assessment is a site visit with sounding, photos, a damage map, and a moisture testing plan where the pattern calls for it. Pricing a durable repair before that assessment is guessing, and guessing is how a floor gets recoated over its own cause.
Questions about epoxy floor peeling? For a free consultation, call Pablo Figueroa at (209) 607-2549 or email pablo@dtindustrialflooring.com.