Why Epoxy Floors Peel and Bubble in Columbia, MD -- and What Prevents It

September 9, 2026

Most failed coatings in Columbia's 1970s and 1980s garages trace to the slab, the prep or the weather on install day, and every one of those can be tested before the first coat.

Failure one: moisture pushing up through Piedmont clay

Columbia sits on the Piedmont, where the ground is weathered crystalline rock rather than sand. The Glenelg soil series, whose type location is in Howard County, is clay loam over saprolite from micaceous schist; the Baile series, also typed in Howard County, is poorly drained silty clay loam in upland depressions and footslopes. Both hold water against a slab, and BWI records 45.00 inches of precipitation a year on 121.3 days, with no dry month. The builder-grade slabs under the older villages were, in the site's words, troweled fast and never sealed, and a slab of that era cannot be assumed to have a vapor retarder.

That water arrives as vapor. It condenses under a coating; the first signs are blisters, cloudy domes and a topcoat that lifts with a skin of concrete. Two standards measure the risk. ASTM F1869 seals anhydrous calcium chloride under a dome for 60 to 72 hours and reports pounds of moisture per 1,000 square feet per 24 hours; readings under about 3 are comfortable, 3 to 5 marginal and above 8 high. ASTM F2170 drills probes to 40 percent of slab depth, three for the first 1,000 square feet and one per additional 1,000, and reads relative humidity after 24 hours with the space at service conditions for 48 hours first. Under about 75 percent is comfortable; a representative epoxy data sheet caps it at 85 percent. Fail either test and the answer is a moisture-mitigating epoxy membrane, not a thicker topcoat.

Failure two: no profile under the coating

Epoxy bonds mechanically, so it needs open pores and a measurable roughness. ICRI Technical Guideline 310.2R-2013, Selecting and Specifying Concrete Surface Preparation for Sealers, Coatings, Polymer Overlays, and Concrete Repair, describes that roughness as a concrete surface profile from CSP 1 to CSP 10 and requires removal of laitance, dirt, oil, films, paint and old coatings. Acid etching only reaches CSP 1 and most coating manufacturers no longer accept it. Diamond grinding produces CSP 1 to 3, shot blasting CSP 3 to 7, and the example data sheet asks for CSP 3 to 4 under its epoxy.

Laitance, a curing compound from the original pour, or a clear sealer someone rolled on in the 1990s will all look like clean concrete and reject the coating anyway. Concrete grinding and surface prep removes them and opens the profile in the same pass, which is why the site's guide tells owners to insist the prep be written as grinding.

Failure three: coating on a dew point morning

The example data sheet requires the slab to be at least 5 degrees F above the dew point and the air no wetter than 85 percent relative humidity. On a normal July 15 at BWI the 5 a.m. air is 69.9 degrees with a 65.3 dew point, a spread under 5 degrees, and a slab that cooled overnight sits below the air. Coating then traps condensation; the result is a hazy blush, poor gloss and a topcoat that peels in sheets a season later.

Cure timing compounds it. At 68 degrees the recoat window is 8 to 24 hours; at 86 degrees it is 6 to 12. Prime on a hot afternoon, return two days later, and the topcoat has nothing to grab.

Failure four: brine, freeze-thaw and scaling

Maryland's highway agency sprays salt brine, a solution of 22 percent salt and 78 percent water, on highways before frozen precipitation, and Howard County plows more than 1,000 miles of county roads. That brine rides home on every tire and puddles on the garage floor. The National Ready Mixed Concrete Association lists freezing and thawing, excessive calcium or sodium chloride deicers, low air content and poor curing as the causes of scaling, the flaking and peeling of a finished surface. BWI gives the slab 93.2 nights a year at or below 32 degrees F and 19.3 inches of normal snow.

Bare concrete absorbs the brine and scales; a coated floor sheds it. Howard County's own winter guidance notes rock salt stops working below 15 degrees F, which is exactly when people pile more on. A garage floor coating with a sealed topcoat turns that puddle into something a squeegee handles.

Failure five: hot tires and thin paint

BWI averages 33.9 days a year at or above 90 degrees F. A car parked after a Route 29 commute puts tires far hotter than that onto the floor, and a thin one-part garage paint softens, sticks to the rubber and comes up in tread-shaped patches. Hot-tire pickup is a thickness and chemistry problem: thin films soften under a hot tire, while a properly primed epoxy body coat with a urethane or polyaspartic wear layer takes the same tires without marking. The weather figures behind this live on the Columbia epoxy by the numbers page, and garage and basement floor coatings across Columbia's ten villages start from the same five checks.

The pre-coat checklist that prevents all five

None of these failures needs a village application to fix, since the covenants and coatings guide confirms interior floors sit outside committee review. What they need is a written scope that answers these before the first coat.

• Moisture test named (F1869 or F2170), reading recorded, primer chosen from that reading.
• Prep written as diamond grinding to a stated CSP, with old sealers and curing compounds removed.
• Dew point, slab temperature and humidity logged on install morning, with the 5 degree rule honored.
• Recoat window stated for the actual forecast temperature, not a generic day count.
• System build stated in mils, with a topcoat suited to brine and hot tires.

More from the blog

• How to Read an ASTM F1869 Calcium Chloride Report Before Signing a Columbia Epoxy Quote
• Coating a Columbia Basement Floor: Efflorescence, Vapor Drive, and When a Coating Is the Wrong Answer
• Getting a Columbia Garage Floor Ready Before the First Brine of the Season