The Glue-Down Plank That Lifted Over a New Slab

Why this matters

A new slab looks dry long before it is. When glue-down vinyl plank goes over one that is still giving off moisture, the failure arrives in weeks, it covers the whole room, and the replacement is a full tear-out: the planks are rarely reusable once the adhesive lets go. This case is worth walking through because the job file looked complete. There was a moisture entry, a number, and a checkmark. The record was the problem.

The call

A homeowner called four weeks after the shop installed glue-down luxury vinyl plank in a new 14 by 40 foot family-room addition, about 560 square feet on a 4-inch slab on grade. The planks at the far end had lifted at the ends, several seams had opened into ridges, and in the middle of the room the floor felt "spongy" underfoot. Pressing on a plank pushed a soft bubble along under it.

That last cue is called pillowing: the adhesive has softened and the plank floats on it. Lifting at the ends and pillowing across the field together point at the adhesive bond failing everywhere, not at one bad plank or one bad seam.

Reading the job file backwards

The lead tech pulled the job file before going out, and read it from the failure back toward the slab.

The install record. Installed in early August, one day, adhesive lot numbers logged, rolled with a 100-pound roller as the adhesive maker requires. Nothing here explains a field-wide failure; a bad lot is possible but rare, and the lot numbers let the shop check it later.

The conditions record. "HVAC on." No temperature or humidity reading. A call to the homeowner filled the gap: the addition's ductwork was tied in the week before flooring, and before that the room had been open to the outdoors through a framed doorway all summer. So the slab had spent its first ten weeks in humid summer air with no conditioning, and the HVAC had run for about a week when the floor went in.

The moisture record. One line: "Slab checked with concrete meter, reads dry, OK to install." No method, no locations, no number against a limit. The installer had swept a pinless concrete moisture meter across the room and seen it read in its dry band.

The pour date. Not in the file. The general contractor's schedule put it in late May, roughly ten weeks before install.

That is the gap, and reading it backwards makes it visible: the only moisture evidence was a surface instrument reading taken on a slab ten weeks old that had never been conditioned. A handheld concrete meter reads the top fraction of an inch through electrical impedance. It is good for surveying a slab to find the wettest area to test. It is not an acceptance test, and no flooring or adhesive maker's limit is written in its units. The surface of a slab dries first; the moisture that re-emulsifies an adhesive comes up from deeper in the slab once the surface is sealed by flooring.

Making the slab testify

The tech did not argue the theory on site. He pulled planks in three places, confirmed the adhesive was soft and stringy with no bond to the slab, photographed it, and arranged a proper test.

Safety first on the pull. Lifting planks and scraping adhesive is low-hazard on a new slab, with no old adhesive involved. Drilling the probe holes is not: drilling concrete releases respirable crystalline silica, so the holes were drilled with a vacuum-shrouded drill and HEPA vacuum under the shop's written exposure control plan per 29 CFR 1926.1153.

Conditions before testing. ASTM F2170 requires the slab and the air space above it to be at service conditions for at least 48 hours before readings, in substance; check the current edition. The planks were pulled from the test areas, the HVAC ran on its normal setting, and the tech logged the room at the start and end of that period.

Depth. A slab on grade over a vapor retarder dries from the top only, so F2170 places the probe at 40 percent of the slab thickness. On 4 inches that is 1.6 inches. A slab drying from both sides (elevated, no retarder below) goes to 20 percent.

Count. F2170 calls for three tests in the first 1,000 square feet. The addition is 560 square feet, so three probes. One went where the meter survey had read wettest, which put it near the old foundation wall where the new slab met it.

RH RETEST AFTER FAILURE - ADDITION SLAB
Method   ASTM F2170 in-situ probes
Slab     4 in on grade, poly retarder below
Depth    40 percent = 1.6 in
Room     71 F / 44 percent RH, HVAC 48+ h
-----------------------------------------
Probe  Location                RH     Temp
P1     north wall, 3 ft in     93 %   68 F
P2     room center             91 %   69 F
P3     old foundation wall     95 %   68 F
-----------------------------------------
Adhesive maker's limit         85 % max
Margin over limit              6 to 10 pts
Original file entry            "meter reads dry"

The 85 percent limit above is illustrative; the real number is on the adhesive's data sheet and differs between adhesives, including between two adhesives approved for the same plank. Every probe was over: P2 by 6 points, P1 by 8, P3 by 10. The wettest reading sat where the survey said it would, which is the meter doing the one job it is for.

Why the readings settle it

Three things had to line up for this to be moisture rather than adhesive or workmanship.

• The failure was field-wide, not seam-by-seam or confined to one trowel pass. A workmanship error (wrong notch, adhesive flashed too long, no rolling) usually shows in a pattern that follows the installer's work.
• The adhesive was soft, not dry and unbonded. An adhesive that never grabbed shows a clean, dry transfer pattern on the plank back. One that bonded and then failed in moisture is soft, stringy and often smells sour. That is re-emulsification, and alkalinity carried up with the moisture speeds it.
• The slab was over the limit at every location, measured by the method the adhesive maker writes its limit in.

The lot-number check with the adhesive maker came back clean, which closed the last alternative.

If the probes had read under the limit

The tech went out ready for the readings to come back clean, because the case turns completely if they do. With every probe at or under the limit, moisture is no longer the explanation and the investigation moves to the install itself, in this order:

• Transfer on the plank backs. Pull planks from the worst area and the best area. Good transfer covers the back with ridges collapsed flat. Ridges still standing, or a dry, lightly marked back, say the plank never fully met the adhesive: the notch was worn or wrong, or the plank went down after the adhesive's working time. See related: Trowel Notches, Open Time and Adhesive Transfer.
• Temperature at install. A slab in the low 60s F slows a pressure-sensitive adhesive's tack and lengthens its flash time; an installer working to the clock rather than to a touch test sets planks into adhesive that is not ready. The install record had no slab temperature, which is a second gap in the same file.
• Rolling. No roller, or rolling once in one direction, leaves an unseated field. That usually shows as hollow spots under a tap rather than pillowing.
• Sunlight and heat. A wall of west glass can drive a plank's surface well past the room's air temperature and pull ends up in a band near the windows only. That is a heat and expansion problem, not a bond problem, and its pattern is a stripe, not a field.

None of these produce soft, re-emulsified adhesive across a whole room, which is part of why the tech expected the slab to be the answer before the probes went in. Expecting it is not the same as proving it, and the probes are what the adhesive maker and any claim adjuster will ask for.

The fix

The floor had to come up. Glue-down plank that has pillowed rarely re-bonds, and the residue left on the plank backs makes reuse impractical. The repair had a fork, and it was the homeowner's to choose with the facts in hand.

Path 1, wait and retest. Leave the slab bare with the HVAC running and retest per 'How to Run an In-Situ RH Test on a Slab' until every probe is at or under the adhesive's limit. Nothing about the waiting time is predictable on a new slab. The old "one month per inch of thickness" rule describes a drying environment most slabs never get and is not an acceptance criterion; the test is. The cost to the customer is weeks of bare concrete with no firm end date.

Path 2, mitigate. Prepare the slab to the profile the mitigation maker requires, apply a moisture mitigation system rated by its maker for the measured RH and accepted by the flooring maker, then install on the system's primer per its cure schedule. No RH retest is needed over a system rated for the reading, but the readings go in the file because the rating has to cover them. Shot blasting or grinding to profile is silica-generating work, done with a shrouded grinder and HEPA vacuum or a blaster with dust collection under the same 1926.1153 plan, never dry swept.

The homeowner chose path 2. The adhesive residue came off with the slab profiling, the mitigation went down, and new plank went in on the same adhesive.

Confirmation. The tech checked the room at 30 and 90 days: no lifted ends, no pillowing, seams tight, with the room logged at 70 and 72 F and 45 and 48 percent RH.

What the shop changed

The fix was technical. The cause was a record that let a survey stand in for a test.

• The job form now has three required moisture fields: method (F2170 or F1869), each location with its reading, and the limit copied from the adhesive's data sheet. "Meter reads dry" will not save.
• Pour date is a required question on any new slab, along with how long the space has been conditioned. A slab younger than a season, or one that has never seen HVAC, triggers a scheduled F2170 test before the install date is confirmed.
• The meter stayed in the kit, with its job named on the form: find the wettest spot so a probe goes there.

References

• ASTM F2170, Standard Test Method for Determining Relative Humidity in Concrete Floor Slabs Using in situ Probes, current edition.
• 29 CFR 1926.1153, Respirable Crystalline Silica (construction).
• Flooring, adhesive and moisture mitigation manufacturers' data sheets and installation instructions.
• See related: Slab Moisture Testing for Flooring; How to Run an In-Situ RH Test on a Slab; How to Glue Down Luxury Vinyl Tile.
• See related: Respirable Crystalline Silica and the Construction Standard.