Osmotic Blistering: When the Blisters Are Full of Liquid

Cut one open, and note when they arrived. Those two together settle it.

A bubble is trapped gas. It forms while the film is still wet, from air rising out of the slab as it warms or from air whipped into the mix, and it is there by the time the floor cures. Cut one and air escapes from a dry void.

An osmotic blister is trapped liquid. It forms after the floor has cured, sometimes weeks or months after, because water in the slab is being drawn through the film towards the salts and residues on the other side. Cut one and liquid runs out, often slightly soapy or alkaline to the touch and sometimes discoloured.

Timing is the second half of the test. Domes that were there on day one are bubbles. Domes that appeared on a floor that looked perfect at handover are almost always this.

Osmosis, which is why the pressure is so much larger than people expect from something as passive as a damp slab.

Water moves through a semi-permeable membrane towards the side with more dissolved material in it. A cured coating is exactly such a membrane, and the underside of it collects alkaline salts and unreacted residues from the concrete. Water in the slab is drawn towards that concentration, arrives under the film faster than it can pass out through it, and collects.

Once liquid collects it keeps drawing more, because the solution under the film gets no weaker. The pressure that develops is easily enough to lift a well bonded coating off a sound slab, which is why a floor can pass every adhesion test at handover and blister anyway.

This is the important difference from a bubble. A bubble is a defect that happened. An osmotic blister is a process that is still running.

It relieves the symptom for a while and does nothing to the cause, so plan on it coming back.

Draining a blister removes the liquid that is there now. It does not remove the moisture in the slab, the salts under the film or the concentration difference driving the whole thing, so the blister refills, usually in the same place and often larger.

Patching over a drained blister is worse, because you have added film thickness exactly where the pressure is highest and given it a fresh surface to work on.

The only repairs that hold address the slab. That means removing the coating over the affected area at minimum, testing the concrete properly, and then either installing a moisture vapour barrier sized to the reading or finding and fixing the water source. Which of those it is depends on the number, not on the appearance.

Because the taped down sheet is not a test, it is a hint, and it is the single most common reason a floor like this gets coated at all.

A sheet of plastic taped to a slab overnight tells you whether enough moisture arrives at the surface in that time to condense visibly. It does not measure how much moisture is in the slab, how deep it goes, or what it will do over the following year. A slab can leave a sheet dry and still be far too wet to coat.

The two tests that produce a number are in situ relative humidity, where a probe is sealed into a hole drilled to a specified depth, and the calcium chloride test, which weighs how much moisture arrives over a set period. Both are standardised, and both are cheap next to removing a floor.

If a floor blistered and the only evidence anybody has is a dry sheet of plastic, the slab has not actually been tested yet.

No, and it is worth being blunt because it is the intuitive move and it makes things worse.

Every additional layer is more barrier. The water is not being stopped from arriving, it is being stopped from leaving, so a thicker film raises the pressure that builds underneath rather than resisting it. Floors treated this way tend to fail later and more dramatically, over a wider area, because the coating that finally lets go is heavier and better bonded.

The products that do work are designed for the job: moisture vapour barriers formulated to tolerate high substrate moisture and to be applied directly to a wet slab, sized against a measured reading.

The distinction that matters is between a coating that resists moisture and a system that manages it. Adding more of the first does not produce the second.

Age is a poor guide, because most of these slabs are not drying out, they are being re-wet.

A slab on grade with a failed or missing vapour retarder underneath it is connected to the ground water table permanently. It does not dry, whatever its age, and a dry looking surface tells you only about the top few millimetres.

After that: drainage running towards the building rather than away, a leaking service under or beside the slab, planting or irrigation against an external wall, a hydrostatic head where the floor sits below the surrounding ground level, and rooms where the vapour drive reverses because the space is air conditioned and the ground is warm.

This is why the honest answer to a high reading is sometimes to investigate rather than to specify a barrier. A barrier manages vapour. It does not fix a leaking pipe.

In nearly every case, yes, and that is the uncomfortable part.

An in situ relative humidity or calcium chloride test before coating produces a number, and that number sits in a band that says clear, borderline, barrier required, or stop and find the source. Osmotic blistering is what the top of that scale looks like once a floor has been laid over it and given time.

What makes it common is that the test is easy to skip. It costs money on a job that has not been won yet, it takes time that the schedule has not allowed for, and a slab that looks and feels dry gives everybody a reason not to bother.

So the practical prevention is procedural rather than technical: test before quoting, put the reading in the quote, and price the barrier the reading calls for rather than hoping the slab is better than it looks.