Why Do Epoxy Floors Peel in Bowling Green? Karst, Moisture and Bad Prep
September 9, 2026
Most coating failures on the Lost River sinkhole plain trace to water moving up through the slab or to a surface that was never ready, and both are testable before anyone opens a bucket.
The ground under Bowling Green moves water
Bowling Green sits on a broad, low-relief sinkhole plain with more than 1,500 documented sinkholes, above the Lost River cave system in the St. Louis and Ste. Genevieve limestones. The Lost River groundwater basin alone drains 55 square miles of Warren County, with shallow groundwater flowing along chert beds and surfacing at swallets, cave streams and springs. Rain off Lovers Lane or Scottsville Road does not run to a creek; it goes down and comes back up wherever the rock lets it.
The Kentucky Geological Survey puts Bowling Green, Russellville and Munfordville in the Western Pennyroyal, the karst belt holding the state's terrain most densely pocketed with sinkholes, and counts about 55 percent of Kentucky as underlain by rock that can develop karst. A slab here is a lid on a drainage system; the karst country slab guide explains why that shapes every coating decision.
How water reaches the bond line
Concrete is porous. When the ground beneath a slab is wetter than the air above it, water vapor moves up through the pores; the trade calls it vapor drive. Bare concrete breathes it off. A coated slab traps it, and pressure builds at the bond line until the coating lifts in blisters, then sheets.
Bowling Green makes vapor drive seasonal. The airport normals record 50.10 inches of precipitation over 126 days a year, with May the wettest month at 5.03 inches. In 2010, 258 millimeters of rain, about ten inches, fell in 48 hours, pushed the Barren River to its highest crest since 1962 and flooded sinkholes faster than the caves could drain them. A floor that tested dry in September can carry a very different load the week after a storm like that.
The two tests that catch it
Two ASTM test methods settle the question, and a written scope for any Bowling Green floor should name at least one.
- ASTM F1869, the anhydrous calcium chloride test, measures the moisture vapor emission rate at the top of a bare slab as pounds of water evaporating from 1,000 square feet in 24 hours. Published epoxy and polyaspartic data sheets refuse a coating at 3 pounds or more.
- ASTM F2170, the in-situ relative humidity test, sets probes in drilled holes to read the humidity inside the concrete, not just at its skin. One epoxy data sheet stops at 85 percent RH; a polyaspartic sheet at 80.
- Neither test is valid over an old coating, sealer, curing compound or patch; the F1869 scope says so. Test after grinding, on bare concrete.
- Both standards say results describe the slab only at the time and in the spots tested, so on karst ground test close to the install date and in several locations, including the low corner where water sits.
- A failed slab gets a moisture-mitigating primer specified for the measured reading, not a thicker coat of the same epoxy; the cost page explains why mitigation is its own line.
Failures that have nothing to do with water
The second family of failures is surface prep. Concrete cures with a weak, dusty skin of cement paste called laitance, and slabs poured under a curing compound or sealed later carry a film nothing bonds to. A manufacturer's epoxy data sheet requires laitance and curing compounds removed and a profile of CSP 2 to 3 under ICRI Guideline 310.2R. Acid etching, light sanding and pressure washing do not get there; one manufacturer states plainly they are no longer acceptable because they leave curing agents behind.
Diamond grinding is the garage answer; shot blasting takes over on large commercial slabs. Concrete grinding and surface prep to a named CSP is the difference between a coating keyed into the slab and one resting on a film. Oil-soaked bays in older garages off Nashville Road need degreasing first, because oil in the pores keeps rising through a fresh coating. Cool, damp air adds a third culprit from the slab guide: amine blush, a waxy film on curing epoxy that must come off before the next coat.
Salt, hot tires and the Kentucky freeze line
Bowling Green is not a northern city, but it salts. KYTC District 3, which covers Warren County, spreads roughly 6,000 tons of salt a year and lays anti-icing brine ahead of storms. That brine rides home on tires and dries on the garage floor, where bare concrete wicks the chlorides into its pores. With 81.4 nights at or below freezing but only 9.7 days that stay frozen, the surface crosses the freeze line roughly 72 times a year, and salted, saturated concrete flakes under that cycle. A coated slab sheds the brine to a squeegee.
The summer failure is hot-tire pickup: a tire that has run hot on I-65 parks on a thin coating that softens and pulls up with the tread, and the airport logs 47.4 days a year at 90 degrees F or hotter. Thin garage paints are the usual victims; a full-thickness garage floor coating with a topcoat rated for hot-tire staining is the fix.
What to ask before anyone opens a bucket
Ask for the moisture readings in writing with the test named, the number of spots and the date; the surface profile as a CSP number with the method; and the primer if the slab fails. On this ground a contractor who tests is not being cautious; one who skips it is guessing.
Those questions apply from Plano to Plum Springs and out to Smiths Grove and Franklin, where the same limestone runs. Use the quote form to reach epoxy flooring across Bowling Green and Warren County and ask for the tests as line items.
More from the blog
- Winter Epoxy Installs in a Heated Bowling Green Basement or Shop
- Recoat or Strip: What to Do With a Ten-Year-Old Peeling Garage Coating in Warren County
- Getting a Garage Floor Coated Before the Holidays in Bowling Green
Quick Answers
Why is my garage floor epoxy peeling after one year?
Usually moisture from below or a surface that was never profiled. In Bowling Green the slab sits over karst drainage, and vapor pushing up through the concrete lifts a coating from the bond line in blisters that turn into sheets. The other common cause is coating over laitance, an old sealer or a curing compound. A calcium chloride or in-slab RH test plus grinding to CSP 2 to 3 prevents both.
What moisture reading is too high for epoxy?
Published data sheets refuse a coating at 3 pounds per 1,000 square feet per 24 hours on the ASTM F1869 calcium chloride test, and at 80 to 85 percent relative humidity on ASTM F2170 in-slab probes, depending on the product. Above those readings the specification changes to a moisture-mitigating primer rated for the measured number before any color coat goes down, and the test should be repeated after mitigation.
Does the Corvette Museum sinkhole mean my slab could collapse?
Collapse is rare; movement and water are common. The February 2014 Skydome sinkhole that swallowed eight Corvettes was a dramatic exception. Most slabs on the sinkhole plain stay put but move water, and a coating cannot bridge a slab that is actively settling or cracking. Fix settlement and structural cracks first, then test moisture, then coat. The data page keeps the local figures in one place.