About eight months after we ran a narrow strip of turf along the side-yard fence line, the corner nearest the gate started curling up — maybe an inch and a half of lift, just enough to catch a shoe on the way through. I actually noticed it with my foot before I noticed it with my eyes, a soft give underfoot right where I usually step through the gate, and it took a second pass a couple days later, crouched down and running a hand along the seam, to see that the backing had actually pulled free from the paver edge underneath rather than just relaxed a little. The rest of that same edge, six feet in either direction, was still flat and bonded tight when I pressed on it. That it was only the one corner, not the whole run, turned out to be the actual clue.
Why Only One Spot Lifts, Not the Whole Edge
An edge that lifts along its entire length usually means the adhesive failed everywhere at once — bad batch, wrong product, or it never cured properly to begin with. A lift confined to one corner or a short stretch is almost always local: that spot gets walked on, brushed against a gate swinging open, or sits somewhere water collects and re-wets the bond line every time it rains. In our case it was the gate — every time it swung open, the edge of the gate frame grazed that corner, and eight months of that eventually worked the glue line loose.
There was a second factor stacked on top of the gate contact, and I didn’t clock it until the corner was peeled back: that paver edge sits about a half-inch lower than the run on either side, a small grade dip I never shimmed during install. Every time it rained, a thin skim of water sat right there for an hour or two before draining off, re-wetting that corner’s adhesive on a cycle the rest of the edge never dealt with. Neither the gate contact nor the low spot alone would likely have lifted it in eight months — it was the two together, mechanical stress on a bond line already softened by standing water, that did it.
The Fix: Clean, Dry, Re-Glue, Weight It Down
- Pulled the lifted corner back fully, past where it was still bonded.
- Scraped off the old adhesive residue from both the turf backing and the paver edge underneath with a putty knife.
- Let the spot dry completely for a full day — the old glue had actually still been slightly damp underneath, which is likely part of why it let go.
- Applied a fresh bead of polyurethane turf adhesive (not general construction adhesive — it doesn’t flex the same way and cracks under repeated movement).
I was tempted to skip the full-day dry-out and just towel the spot off, since it wasn’t visibly wet by the time I got to it. Glad I didn’t. Adhesive over a paver edge that’s damp below the surface doesn’t cure evenly — it can skin over fine on top while staying soft underneath, which is a reasonable description of what had already failed once at that exact spot. The extra day cost me nothing but patience, and I wanted the bond starting from bone dry rather than betting the new glue would somehow do better than the old glue did in the same wet conditions.
- Pressed the edge down firmly and weighted it with a couple of paving stones for 48 hours instead of the usual 24, since the gate contact meant this spot needed a stronger cure before it took any more abuse.
It’s held for four months since, through the gate swinging the same way it always did.
Keeping It From Happening Again
Anywhere an edge sits near something that moves or brushes against it — a gate, a mower turning point, a hose reel you drag past — give that spot extra adhesive coverage and extra cure time before it goes back into regular use, not just the standard treatment you’d use on a straight run of edge nobody touches. And check that specific kind of spot again after the first really hot stretch of weather, since heat softens adhesive temporarily and makes it more vulnerable to exactly this kind of repeated contact.
I also went back and packed a thin wedge of paver base under the low side of that corner to bring the grade level, so the same puddle can’t form there again. If I’d caught that dip during the original install instead of eight months later, the gate contact alone probably wouldn’t have been enough to lift the edge — worth checking any new edge for low spots with a level before you glue it down, not after.
Telling This Apart From a Seam Coming Apart
It’s easy to mix this up with a seam failing between two pieces of turf, but they feel different once you know what to check. A lifting edge like ours stays a single piece of turf pulling away from whatever it’s glued to underneath — the backing itself is intact, it’s just the bond that’s gone. A failing seam shows a gap opening between two separate pieces, usually with the seam tape visible and crumbling in the gap, and it tends to run along the whole seam rather than staying at one corner. Press down on a lifted spot: if you feel one continuous sheet of backing flex under your hand, it’s an edge-bond problem like this one. If you can see backing on both sides of a widening line, that’s a seam, and it calls for seam tape and seam glue rather than the edge adhesive I used here.
As for how long a re-glue like this should hold: four months in, through a full stretch of hot-weather afternoons and the gate swinging the same as always, it hasn’t budged. I’d expect it to hold about as long as the original bond did somewhere in the neighborhood of years rather than months, now that the standing water that undermined it in the first place has nowhere left to sit.