Running a supply counter means you see the good work and the bad work, and the bad work usually arrives in a Ziploc bag. Somebody sets a curled chip of coating on the counter, tells you the floor went down four months ago, and asks what was wrong with the material. Nine times out of ten the material was fine. The chip is telling a different story, and if you know how to read it, it tells you exactly which step got skipped.
Chris would honestly rather teach you the boring parts than sell you a second set of material for the same slab. So here are the failures that walk back through the door, what each one looks like in the hand, and the small step up front that would have kept it from happening.
Moisture drive, and the chip that comes back with clean concrete stuck to it
This is the one that costs the most money and gets blamed on the resin the most often. The floor looks perfect for weeks. Then a bubble shows up, then a patch of them, then a section lifts and the customer can peel it up in a sheet like old wallpaper.
Flip the chip over. If the back of it is coated in a thin, chalky, pale layer of concrete, moisture won. Water vapor moving up through the slab carried dissolved salts with it, those salts crystallized right at the bond line, and the crystal growth pushed the coating off the slab. The coating did not release from the concrete. It took a paper thin skin of concrete with it, which is why the failure often reads as a bond problem when the bond was actually stronger than the top of the slab was.
Slabs on grade with no vapor retarder under them are the usual suspects, and so is anything poured recently. Concrete is wet for a long time. Freshly troweled slabs in the valley can read high for months, and a warehouse floor that was dry all summer can drive moisture in February when the ground under it is saturated and the building is heated. Interior slabs below grade are their own category of trouble.
The five minute step is a moisture test, and there are two worth running. A plastic sheet taped down overnight tells you something fast and free. A calcium chloride test tells you a number you can write on a job sheet and hand to a customer, and it is the number that decides whether you need a moisture vapor barrier under your build or whether you can go straight to primer. We stock the test kits for exactly this reason. A kit costs a fraction of what a tear out costs, and running one is the difference between a professional decision and a hopeful one.
Here is the part installers hate hearing: if the number comes back high, the answer is a vapor barrier, not a thicker coat of the same product. A heavy build of standard epoxy does not stop vapor. It just gives the vapor a bigger, more expensive thing to push off the slab.
Not enough profile, and the coating that peels in clean sheets
Different chip, different story. This one comes off in wide flexible sheets and the back is glossy and smooth. No concrete on it at all. You can see the trowel marks from the original slab printed on the underside like a fingerprint.
That is a mechanical bond failure and it means the coating never had anything to grab. Resin does not glue itself to concrete chemically the way people imagine. It flows into the open pore structure and the roughness of the surface and locks in there once it cures. If the surface is closed, hard troweled, sealed, or polished, there is nothing to lock into. The floor is held on by surface tension and optimism.
Acid etching is where most of these start. Etching gets sold as prep, and on a soft, porous, unsealed slab it can open the surface up some. On a hard troweled slab it barely touches it, and it leaves behind residue and softened paste that you now have to neutralize and remove. Grinding is the answer, and it is not a close call. A CSP 2 to CSP 3 profile is the working target for most thin film builds, and heavier builds want more tooth than that.
The check takes less time than reading this paragraph. Pour a small puddle of water on the ground slab. If it wets in and darkens the concrete within a few seconds, you have an open surface. If it beads up and sits there looking at you, keep grinding or figure out what is still on that floor. Run the same check in four or five spots, because slabs are not uniform, and the corner the crew rushed is exactly where the failure will start. If your machine is not aggressive enough for the slab you are on, that is a tooling problem and not a technique problem. Come talk to us about tooling and prep before you fight a floor with the wrong metal bond.
Mix ratio, and the parts of the bucket nobody scraped
Soft spots. Tacky patches that never hardened while the rest of the floor came out glass hard. Areas that print under a rolling toolbox six months later. Sometimes a whole floor that stayed rubbery and gummy no matter how long anybody waited.
Two part resin is a chemical reaction with a stoichiometry, not a paint you tint to taste. The A side and the B side have to meet in the ratio the manufacturer published, because the molecules have to find their partners. Extra hardener does not cure it faster or harder. It leaves unreacted amine in the film, which stays soft and eventually turns yellow and greasy. Short hardener leaves unreacted resin, which stays soft in a different way. Both fail. Both look like bad product.
The other half of this is incomplete mixing, and it is more common than bad measuring. Somebody mixes a kit for thirty seconds with a paddle spinning too fast, whips air into it, and pours. The unmixed material is not distributed evenly through the pail, it is concentrated on the sides and in the bottom corner and under the lip where the paddle never reached. Those are the soft spots on the floor, and they land wherever that part of the pail got poured.
The habit that fixes it: mix by the clock, not by eye, at the speed the data sheet calls for. Scrape the sides and the bottom with a flat stick partway through and keep going. Then do the thing that separates the pros from the people who will call us in April, which is the second pail. Pour the mixed material into a clean pail and mix again briefly. Whatever clung to the first pail stays in the first pail instead of going on the floor. It costs you one pail and about forty seconds.
Also, respect the kit. Splitting a kit to do a small patch is where ratio errors live. If you need small quantities, use a scale, weigh both sides, and write the weights down. Guessing at half a kit by looking at the fill line on a pail is not measuring.
The recoat window closed while everybody was at lunch
This chip delaminates between coats, not from the slab. The primer or the base is still stuck tight to the concrete. The topcoat came off the base cleanly and the underside is smooth and shiny. Two layers that never became one.
Every product has a recoat window with two edges. Too early and you trap solvent or disturb a film that has not gained enough strength, and it wrinkles or lifts. Too late and the previous coat has cured out so completely that there are no reactive sites left for the next coat to chemically link into. Past that window, the next coat is sitting on a hard slick plastic surface and is bonded to it about as well as tape.
The window is temperature dependent and the data sheet numbers are written for the temperature on that data sheet. At 90 degrees in a Tracy garage in July, a window listed as six to twenty four hours might be closed at nine hours. In a January warehouse at 52 degrees, that same window stretches out and the early edge moves way later than you expect. Read the sheet, then adjust it for the slab temperature you actually have, not the number on the wall thermostat.
If you miss the window, the fix is not a prayer and a heavier coat. Abrade it. Scuff sand or lightly screen the surface to give the next layer mechanical tooth, wipe the dust, and proceed. Ten minutes of abrasion recovers a job. Ignoring a blown window buys you a callback in the fall. When you are stacking a system, it helps to be running components designed to go together, which is most of why we build out full epoxy systems rather than selling loose parts and wishing you luck.
Dew point, amine blush, and fisheyes
Two related failures live here and both come from the air, not the bucket.
Amine blush is the waxy, greasy, sometimes blotchy film that shows up on a curing epoxy surface when humidity is high. Amine hardener reacts with carbon dioxide and moisture in the air and forms a carbamate on the surface. It feels slick. It can look like a hazy bloom or a mottled oil slick. If you coat over it, the next layer is bonded to a wax film, which is to say it is not bonded. Blush is also water soluble, which is the good news: scrub it off with warm water and a stiff pad, rinse, let it dry, and you have a clean surface again. The mistake is scuffing over it dry, which just spreads it around.
Fisheyes and craters are the other half. If the slab is colder than the dew point, moisture is condensing on the surface faster than you can see it, and you are rolling resin onto a microscopically wet floor. The resin pulls away from those spots and leaves round craters. Cold slabs in an unheated shop on a foggy delta morning are exactly the setup for it, and the slab lags the air temperature by hours, so a warm morning does not mean a warm floor.
The five minute step is a thermometer and a hygrometer, and the rule installers should have memorized: the slab needs to be at least five degrees above the dew point and heading up, not down. An infrared gun reads the slab in a second. Reading the slab in four corners takes another minute. If the numbers are wrong, the honest move is to wait or heat the space, and that conversation with the customer is far shorter than the conversation about why the floor has to come off.
Contamination, which is usually silicone and usually not yours
The last common failure is the one nobody can see. The floor goes down beautifully and cures with round bare craters or dull patches in odd spots, or a section that never bonded at all in a place with no obvious explanation.
Silicone is the classic offender. It comes from tire dressing, from lubricant sprays, from RV and boat detailing products, from the aerosol somebody used on a door hinge, from car wax. It migrates, it is invisible, and it does not wash off with a mop. Grinding it does not always solve it either, because the grinder can smear it and drive it deeper. If you know a garage has been a detailing bay, degrease before you grind and again after, and expect to fight it.
Curing compounds and previous sealers are the other family. A slab that was sprayed with a membrane curing compound after placement has a film on it designed to stop things from penetrating, which is doing its job perfectly and ruining yours. Those come off with grinding, but only if you actually get through them, which is why the water drop test above matters more than a visual read of a ground floor. Ground concrete and ground curing compound can look nearly identical.
Then there are the sneaky ones. Oil that wicked deep into a slab under an old truck bay and comes back up when the coating warms during cure. Concrete dust left in a corner after grinding, which acts as a bond breaker. A drywall crew that walked joint compound onto a clean floor overnight. Rain blown in an open roll up door onto a primed slab. Vacuum, tack, and inspect at floor level with a light raking across the surface before you open the first pail. You will find things standing up that you missed standing over.
Why the boring steps are the whole job
Read those six failures back and notice what they have in common. Not one of them is a material defect. Every single one traces back to something measurable that somebody chose not to measure, and every single fix costs minutes and a piece of equipment you can hold in one hand.
That is the ugly math of this trade. Prep and testing are maybe fifteen percent of the labor on a typical job and close to a hundred percent of the failures. And the failures are expensive in a way that stacks: you eat the material, you eat the labor twice, you eat the grinding to get the failed coating off, and you eat whatever the customer says about you afterward. Chris has been on the wrong end of that math and built the shop partly so that other installers do not have to learn it the same way.
If you want the whole sequence hands on instead of read off a screen, that is what the training academy is for. Bring a real problem slab you are quoting and we will work through what it needs.
Questions we get at the counter about failed floors
Q. Can I just recoat over a floor that is bubbling instead of tearing it out? No, and this is the most expensive shortcut in the trade. If the bubbles are from moisture drive, the vapor pressure that lifted the first coating will lift the second one too, and now you have twice the material to grind off. Find out why it failed first. If it is moisture, the slab needs a vapor barrier system, and the failed coating has to come off for that barrier to bond to concrete.
Q. Is acid etching ever acceptable prep? On a soft, porous, unsealed, never troweled slab it can produce some profile, and some manufacturers still list it as an option for thin film products. In practice we tell installers to grind, because you cannot tell by eye whether an etch opened a given slab, and the residue it leaves is its own liability. A grinder gives you a profile you can verify and repeat. Ask us about a grinder rental if you are not ready to buy a machine yet.
Q. My floor cured tacky in a few spots but hard everywhere else. Is the product bad? Almost certainly not. Isolated soft areas point at mixing, since unmixed material concentrates on the pail walls and bottom and lands wherever that portion got poured. A whole floor that stayed soft points at ratio or at pouring below the product minimum temperature. Bring us the batch numbers and the conditions you poured in and we will work backward with you.
Q. How long should I really wait after grinding before I coat? Long enough for the slab to be genuinely dry and genuinely clean, which depends on whether you ground wet or dry and on the conditions. Wet grinding puts water into an open profile and that water has to leave before resin goes on. The real answer is that the slab tells you, not the clock: check the moisture, check the slab temperature against dew point, and pull a light across the surface looking for dust.
Q. Does a more expensive resin prevent these failures? Better resin gives you better working properties, better clarity, better chemical resistance, and more consistency pail to pail, and that is worth paying for. It does not test your slab, profile your concrete, or scrape your pail. We field test everything before it goes on the shelf specifically so you can stop worrying about the product and put your attention on the six things above, which are the ones that actually decide whether the floor lasts.
Bring us the chip. Seriously. If a floor came back on you, put a piece of it in a bag with the batch numbers and the conditions you poured in, and come see us at the counter or call (209) 208-4656 and talk to the owner directly. Chris will read the back of that chip with you and tell you straight what happened, even when the answer is one nobody wants to hear. You can also send us the details ahead of time so we have the data sheets pulled before you walk in.