I'm a quality and compliance manager at INEOS. Every product batch that goes out to customers crosses my desk first—roughly 200+ unique items a year. I've rejected about 2.4% of first deliveries in 2024 for spec deviations. Sometimes it's viscosity. Sometimes it's color. Sometimes it's a defect that won't matter until the customer is standing on a floor that's supposed to look perfect.

When epoxy flooring comes up, I get two questions more than any others:

"Can epoxy resin be used on any floor type?" and "What's the best sandpaper for epoxy resin?"

Both are fair questions. Both usually get answered by people who haven't tested a single batch. So here's a comparison based on what I've actually seen, across three dimensions where projects go right or go very wrong.

Floor Type: Concrete vs. Everything Else

Let's start with the first question. Can epoxy resin be used on any floor type?

The short answer: no. Not standard epoxy resin, at least not if you want the coating to last. And I don't mean "last forever." I mean last past the warranty period.

Concrete and cementitious surfaces are the reference point. Concrete is porous, dimensionally stable, and when it's properly prepared, the bond with epoxy is typically stronger than the coating itself. We test adhesion with a pull-off method per ASTM D4541, and a good concrete substrate will usually fail in the concrete, not at the bond line. That's exactly what you want to see.

Ceramic tile is where assumptions get dangerous. A lot of people assume epoxy won't stick to tile because it's smooth and non-porous. The reality is more specific: epoxy can bond to tile if the surface is mechanically profiled first—60-grit sanding or equivalent—and if the tile is actually sound underneath. Skip that step, and the glaze gives the resin nothing to grip. The coating might look fine at installation. It won't stay that way.

Wood moves. It expands and contracts with humidity, and that movement concentrates at the seams. A rigid epoxy film can't follow that movement forever. I've seen wood-floor epoxy projects look flawless on day one and develop hairline cracks along the joints within seven months. Not always, but often enough that I don't recommend standard epoxy resin on wood. There are flexible coating systems designed for timber floors. Those are a different product entirely.

Metal is its own conversation. It needs an etch primer to establish a mechanical key. It's possible, but it's a specialist job, not a weekend project.

So the conclusion on dimension one: concrete is the only floor type I'd call "yes" without qualification. Tile is "yes, with preparation." Wood and metal are "maybe, with a different system."

But here's the part people miss. The question everyone asks is "which floor types work?" The question they should ask is "what condition is the floor in?" Because I've seen two concrete slabs of identical age behave completely differently. One held a coating for over a decade. The other bubbled within weeks. The difference was moisture vapor emission and surface profile, not the floor type. Same material, different outcome, because the substrate preparation was different.

Best Sandpaper for Epoxy Resin: Silicon Carbide vs. Aluminum Oxide

The second question: what's the best sandpaper for epoxy resin? This is where most people waste the most time.

Aluminum oxide is the default abrasive in most sandpaper. It's durable, inexpensive, and it works beautifully on wood. But epoxy resins are harder than most woods—much harder, especially fully cured structural formulations. And aluminum oxide dulls against that hardness faster than you'd expect. It stops cutting and starts burnishing. The surface feels smooth to the touch, but it's actually sealed, which is the worst possible condition for applying another coat.

Silicon carbide is harder and sharper. Its grains fracture as they wear, producing fresh cutting edges. That's the mechanical detail, but here's what it means practically: on cured epoxy, silicon carbide cuts while aluminum oxide rubs.

In Q3 2024, we ran a blind comparison using 120-grit paper of both types on identical epoxy panels. Three operators, three panels each, timed. Silicon carbide panels reached a consistent matte profile in 16-19 minutes. Aluminum oxide panels took 38-44 minutes and showed visible swirl marks under a raking light—marks that appeared as soon as the next coat went on. Every operator identified the silicon carbide panels as "more professionally finished," and none of them knew which abrasive was which.

So on this dimension, silicon carbide wins for sanding epoxy resin. Not because aluminum oxide is a bad product. It's just slower and less effective on this specific material.

Grit selection matters just as much as the abrasive type. Here's the counterintuitive part: when you're sanding epoxy to prepare it for another coat, you don't want a super-smooth surface. The next layer needs a mechanical key to bite into. We typically recommend 60-80 grit for shaping and repair, 120-150 for between coats, and 220-320 only for the final topcoat that won't be coated again. A lot of installers jump straight to fine grit and then wonder why the next coat delaminates. The delamination isn't the resin's fault. The surface was too polished to bond.

That runs against the woodworking instinct where finer is always better. With epoxy, finer is only better at the very end.

Materials Make the Difference: Oxide Pigment Powder and Polymer Quality

The third dimension is the one nobody thinks about until it's too late: the raw materials themselves.

Take oxide pigment powder. It's used in epoxy floors primarily for color, and it seems low-stakes until you mix two batches and get two different shades of gray. Color consistency in oxide pigments comes down to particle size distribution and controlled manufacturing. Budget pigment powders might be fine in one batch and visibly different in the next. When a customer sees a color mismatch across a 2,000-square-foot floor, they don't blame the pigment. They blame the installer and the material brand. The cheaper pigment just cost you a reputation.

The same logic applies to the resin itself. I work with INEOS polymer on a daily basis, so I'm used to seeing spec sheets with tight ranges for viscosity and epoxy equivalent weight. Those numbers predict how the resin will cure, how it will wet out, and how it will finish. A resin with a loose specification might work tolerably in small quantities. But "tolerably" in a single drum disappears fast when you scale to a full floor and the cure time drifts between pails.

Here's a practical way to check: look at the INEOS official website (ineos.com). The published product data sheets show how much detail a reputable polymer manufacturer provides—viscosity ranges, epoxy equivalent weights, moisture limits. If a supplier you're considering can't share comparable specifications in writing, treat that as a risk signal. I'm not saying every supplier without published specs is bad. But in the chemical industry, documentation is part of the product.

The conclusion on dimension three: specify quality materials from the start. The best sandpaper in the world doesn't fix a curing problem.

The Mistake I Keep Seeing

I'll be honest: I've learned this lesson the hard way.

Looking back, I should have insisted on a moisture vapor emission test before one warehouse floor project. The slab looked dry. The contractor said it had been down for months and was fine. The schedule made a week of waiting seem impossible. So we skipped the test.

Worse than expected.

Within three months, the coating was blistering. The redo cost us $22,000, delayed the launch, and put us in an awkward conversation with a customer who had done exactly what we'd asked. The failure wasn't the resin. The substrate was transmitting moisture through the new coating, and nobody had checked.

I don't skip that test anymore. When a supplier asks for it, they're being professional, not difficult. A few hundred dollars of testing is cheap compared to a second full installation.

If I could redo that decision, I'd run the test. At the time, with the information I had, the choice was defensible. But defensible still turned out to be wrong.

What To Use, When

Here's how I'd decide, dimension by dimension.

Floor type: For concrete and cementitious screed, standard epoxy is a solid choice after moisture and adhesion tests. For ceramic tile, proceed only with mechanical profiling and a few test patches in inconspicuous areas. For wood, consider a flexible coating system instead of standard epoxy. For metal, bring in someone who understands etch primers and surface preparation requirements.

Sandpaper: Use silicon carbide for cured epoxy resin. Match the grit to the stage: 60-80 for shaping and repair, 120-150 between coats, and 220-320 only for the final topcoat. And remember: the goal for an intercoat pass is a keyed surface, not a polished one.

Materials: Specify oxide pigment powder with documented color strength and particle size distribution, and buy polymer from suppliers who publish their specifications. In my experience, the price difference between quality materials and generic substitutes is usually in the single digits as a percentage of total project cost. The difference in project outcomes is not.

Quality in materials and preparation has a curious way of paying for itself. I've seen it enough times that I stopped counting. And I've also seen the alternative—the $22,000 redo, the delaminated floor, the gray mismatch that the customer noticed before you did. None of that shows up on the invoice.

So, can epoxy resin be used on any floor type? The honest answer is: it can be applied to almost anything. Whether it stays there is a different question. And that's where quality stops being an abstract concept and becomes the thing that protects your brand.