Concrete Moisture

This is the page that saves floors. If you test for moisture and act on the result, you have removed the single most common cause of coating failure from your project. If you skip it, you will not find out for months, and by then the fix is grinding the floor off and starting again.

What moisture does to a coating

Concrete sits on the ground, and the ground holds water. Unless something stops it, that water travels up through the slab as vapor and leaves at the top surface, where it evaporates into the garage and nobody notices.

A cured coating closes that surface. The vapor keeps arriving and now has nowhere to go, so it collects at the bond line between the coating and the concrete — the one place in the whole system whose strength is entirely adhesion. Pressure builds under the film. The film lifts, first as small blisters, then as sheets that peel away with a thin gray layer of your slab still stuck to their underside.

Nothing about the coating causes this, and no ordinary floor coating survives it. It is the concrete’s water, and it has to be dealt with in the concrete.

Where the water comes from

• No vapor barrier under the slab. A sheet of polyethylene under the concrete is standard practice now and was routinely left out of older construction, particularly in a detached garage.
• Ground water and drainage. A downspout discharging at the foundation, ground sloping towards the garage, a high water table after heavy rain. These are the cheap problems, because they are fixable from outside.
• A slab that has never dried out. Poured on wet ground, poured in winter, or simply young.
• Water put in from above. A leaking door seal, snowmelt off a car all winter, or washing the floor down the week before you coat.

That last one is worth reading twice. Pressure-washing a slab and coating it two days later has failed floors that never had a moisture problem in the first place. Concrete takes up water quickly and gives it back slowly.

The plastic sheet test: a rough screen

ASTM D4263. It costs nothing, takes a roll of tape and an evening, and it is a screen rather than a measurement.

1. Tape a square of clear polyethylene sheet, roughly 18 by 18 inches, flat against the slab.
2. Seal all four edges completely. Any gap and the test tells you nothing at all.
3. Leave it at least 16 hours. Longer is better.
4. Lift it and look. Condensation on the underside, or concrete visibly darker than the slab around it, means water is moving.

Do several at once, in different parts of the floor: against the exterior walls, in the middle, at the door, and anywhere the slab looks stained, darker or chalky.

Be clear about what a pass means, which is very little. A dry sheet tells you the slab is not obviously wet. It cannot tell you the rate, because it does not measure one. A failed sheet is a definite answer and you should stop; a clean sheet is only permission to keep going.

The two tests that produce a number

Calcium chloride (ASTM F1869)

A small dish of dry calcium chloride is weighed, sealed to the slab under a dome for 60 to 72 hours, and weighed again. The weight it gained is the water that arrived, and the result is reported as pounds of moisture per 1,000 square feet per 24 hours. Kits are inexpensive and sold for exactly this job.

It measures what is leaving the surface, which makes it sensitive to what the air in the garage is doing. The slab has to be at service temperature and the space conditioned before and during the test, or the number you get describes the weather rather than the concrete.

Relative humidity probes (ASTM F2170)

Sleeves are drilled into the slab — to 40 percent of its depth where it can only dry from the top — and a sensor in each sleeve is left to equilibrate before it is read. The result is a relative humidity, in percent, from inside the concrete.

This is the test we would rather you ran. It reads the body of the slab rather than its surface, which is where the water actually is, and it is far harder to fool with a warm dry day. It costs more, and it puts small holes in your floor — which you patch during preparation anyway.

What the number has to be

Here is where we stop short, on purpose.

The primer data sheet does give a limit, and it is the one that governs: residual moisture in the slab must not exceed 4%, confirmed by the ASTM D 4263 plastic sheet test — tape a sheet of plastic down overnight, and any water on the underside means the slab is not ready. What none of the three sheets gives is a maximum vapor emission rate in pounds (ASTM F1869) or a maximum internal relative humidity (ASTM F2170), and we are not going to borrow either threshold from another manufacturer’s product and print it as though it applied to this one. The figures that circulate online — a number in pounds, a number in percent — belong to specific products tested under specific conditions. Repeating one here would be a guess wearing the clothes of a specification.

So run the test, write the number down, and send it to us with the age of the slab, whether it has a vapor barrier under it, and what the garage is built on. We will get you an answer for the system you are actually buying. Where we do not have it, we ask Duromar — which is a phone call, and better than a guess.

If the slab is wet

A moisture problem gets solved before coating, not during it and not with a thicker coat. Roughly in the order that these work:

• Fix what is putting the water there. Downspouts, grading and a blocked drain are the cheapest repairs on this list and they are often the entire problem. Fix them, then wait and test again.
• Give it time and air. A slab wet from an event — a flood, a wash-down, a wet build — dries out. A slab wet from the ground beneath it does not, no matter how long you wait.
• A moisture-mitigating primer. These exist, they work up to a stated limit, and they are applied to the bare slab under the system. ANVIL does not currently list one, so if your result comes back high this is a conversation with us rather than a product page.
• Do not coat it. If a slab is pushing water and cannot be fixed, a full-broadcast floor coating is the wrong product for it. We will tell you that, and we would rather tell you now.

New concrete

New slabs have to dry, and drying takes far longer than curing. The 28-day figure people repeat is about strength — it is when concrete is expected to have reached its design compressive strength. It says nothing about how much water is left inside it.

A slab that can only dry from one side, in an unheated garage, in a wet season, can still be too wet to coat months after it is strong enough to park a truck on. Do not count days. Test.

New concrete usually also carries a curing compound, which is a film and has to come off. That is covered in existing coatings and dealt with in floor preparation.