The Bathroom Trim That Peeled in Sheets
Why this matters
Three weeks after a hall bathroom trim repaint, the customer called: the new white enamel on the door casing was coming off in long strips when her fingernail caught an edge. A peel like that in a bathroom gets blamed on steam first, and a crew that accepts that answer repaints the same failure. This case is told backwards from the shop's own record, because the record held the cause in what it did not say. The original job plus the fix took more than twice the hours of doing it right once, and the change that came out of it costs a cotton ball per room.
The record, as written
The lead pulled the original work order before the callback visit:
WORK ORDER - hall bath trim and door
Scope: clean, 2 coats semi-gloss latex enamel
Crew: 1 painter, 1 day
Prep: degreaser wash, rinse, dry
Notes: trim glossy, cleaned well
Tested: (blank)
Primer: none
Hours: 5.0 (clean 1.0, two coats 4.0)
Three things in it matter. The trim was described as glossy. The prep was cleaning only, with no scuff sanding. And the line for what the surface was tested for was empty, which in this shop's form means nobody tested it. A record that says what was done but not what was checked leaves the cause of any failure to be reconstructed from the wall.
What the wall showed
The house was built in 1984, so the lead presumption did not apply to the callback inspection. The lead went through the evidence in the order it was cheapest to read:
The peel itself. The new coat came off in sheets several inches long, intact and flexible, with almost nothing holding it. A coat that has bonded and then failed usually breaks into flakes; a coat that peels in a continuous rubbery sheet was never stuck.
The back of the sheet. Clean white enamel, smooth and slightly glossy, with no old paint and no dust on it. If the failure had been in the old layers, the back of the sheet would have carried pieces of them. If the failure had been dirt, the back would show it. A clean back means the new coat let go at the surface of the old one.
The surface underneath. Under the peeled strip, the old trim was hard, cream colored and glossy, noticeably glossier than the new semi-gloss had been.
The alcohol test, done three weeks late. A degreaser wipe on the exposed old surface, rinse, dry, then a cotton ball with denatured alcohol, about twenty firm strokes. Clean cotton, surface unchanged. Treat it as oil. The same test on a piece of the new enamel softened it and put color on the cotton: latex, as expected.
Ruling out the bathroom. Steam was the customer's theory and a fair one: a fresh latex coat in a steamy room before it cures can soften, blister or show surfactant leaching (sticky, glossy brown drips). But the outside face of the door casing, in the dry hallway, peeled exactly the same way at a fingernail, and the inside of the medicine cabinet trim, which hardly sees steam, did too. Moisture can make a weak bond show faster; it was not what made this one weak. There were also no blisters and no drips.
Ruling out cure time. Latex enamel takes longer to reach full hardness than it takes to dry, and a three-week-old coat can still be softer than it will be. But softness and peeling are different failures. A coat that is not fully hard dents and prints; it does not lift off in sheets from a clean, glossy substrate.
The reconstruction. The trim had been oil enamel from the original construction, never repainted. The painter washed it well, which removed the grease but did nothing about the gloss, then laid latex over it. Latex grips by tooth and by wetting a surface it can bond to; glossy cured oil gives it neither. The bond was weak from day one and let go the first time a fingernail pried an edge. 'How to Test Glossy Trim and Prime Latex Over Oil' covers the chemistry and the procedure in full.
Why the crew missed it
The painter was experienced and did not skip anything he thought was needed. The gap was that "glossy" was written down as an observation and not treated as a question. A glossy finish on 1980s trim is exactly the condition where oil is likely, and nothing in the shop's walkthrough at the time required a test to be done and recorded. Cleaning felt like prep because it is prep, and the trim looked ready.
Reading the peel before reading anything else
The shape of a paint failure narrows the cause before any test, and it is worth carrying as a habit:
| What comes off | What it usually means |
|---|---|
| Continuous rubbery sheets, clean back | New coat never bonded to a slick or contaminated surface |
| Brittle flakes with old layers on the back | An old layer in the stack failed; the new coat was the last straw |
| Flakes with wood fibers on the back | The wood surface itself let go, often weathered or wet wood |
| Blisters, some with water in them | Moisture behind the film or steam on an uncured coat |
Every peeled piece in this bathroom matched the first row, which is why the lead went looking for a slick substrate rather than a wet one.
What the customer was told
The conversation happened at the door casing, with the peeled strip in hand:
- What happened, in one sentence: the old trim was oil-based enamel, the new paint was water-based, and it was put on without the primer that makes the two stick. That was the shop's miss.
- What the fix is: strip the new coat, dull the old one, prime with a primer that bonds to it, then repaint. Two visits because the primer has to cure before the topcoat goes on.
- What it costs her: nothing. The shop owns a missed test.
- What she would see afterward: the same white semi-gloss, and a fingernail test at an edge that she was invited to repeat after a week.
The customer had theorized steam. The lead showed her the hallway side of the casing peeling the same way, which settled the question faster than any explanation of chemistry.
The repair
- Strip the failed coat. Everything that would lift was peeled and scraped off with a putty knife. Most came off in sheets. Where the new latex held more tightly, a tape check (a field version of the ASTM D3359 tape method, in substance) decided it: a small X cut through the coating, firm tape pressed over it and pulled sharply. Anything that lifted came off; anything clean on the tape was left and treated as part of the substrate. Time: 3 hours.
- Clean, then degloss. Degreaser wash, rinse, dry. Then scuff sanding with a 180 to 220 grit sponge until the whole surface was uniformly dull under a raking light, with no shine left in the casing profiles. A liquid deglosser was used in the tightest profile coves where the sponge could not reach, with the bathroom fan running and the window open, no flame or pilot nearby, and used rags spread to dry outdoors before going into a closed metal can, because solvent-soaked rags can heat up and ignite on their own. Vacuum and a damp wipe to remove dust. Time: 2 hours.
- Bonding primer. One coat of a bonding primer chosen per 'Bonding Primer Vs Stain Blocking Primer Vs No Primer Decision', applied at the film the data sheet calls for and left for the full time its data sheet gives before topcoating. Time: 1.5 hours.
- Check the primer. After that cure time, the X-and-tape check on a primed spot inside the closet-facing side of the casing came away clean. That is the go signal; a primer lifting on the tape would have meant stopping and going back to step 2.
- Topcoat. Two coats of the same semi-gloss latex enamel, brushed per 'How to Brush Enamel on Baseboard and Casing'. Time: 4 hours. The customer was asked to run the bath fan during and after showers and to avoid scrubbing the trim until the enamel had cured for the period on its data sheet.
The hours, checked
The original job was 5 hours. The repair was 3 plus 2 plus 1.5 plus 4, which is 10.5 hours. Together that is 15.5 hours of labor on one bathroom's trim.
Done right the first time, the job would have been the repair without the stripping: 2 hours of cleaning and scuffing, 1.5 hours of primer and 4 hours of topcoat, 7.5 hours. So the shop spent 15.5 hours where 7.5 were needed, a little over twice as much, all of it unbillable. The test that would have flagged it takes a few minutes, and the 2.5 extra hours of prep and primer it would have triggered could have been bid and paid for.
What changed in the walkthrough
The shop added a line to its bid walkthrough, filled in for every room with trim or doors being repainted:
- Alcohol test result for each type of trim in the room (doors, casing, baseboard, window trim), recorded as latex or oil.
- Tape check result on any trim that reads latex and will not be primed.
- Primer called for, yes or no, and which one.
A blank in any of those three fields stops the work order from going to the schedule. The point is the blank: "glossy" in a notes field is an observation anyone can write without acting on, while an empty test field is something the office can see before the crew arrives.
What would have changed the call
- A pre-1978 house. The peel inspection is minor, but the scraping and scuff sanding in steps 1 and 2 are paint disturbance, and in target housing done for pay they become RRP work under a certified renovator once they pass the interior minor-repair limit of 6 square feet per room, with containment and cleaning verification inside the bathroom. See 'Lead Paint RRP Protocol Reference'.
- The alcohol test reads latex and the tape check fails. Then a previous painter already put latex over oil, and the job is the same strip and prime, found before the bid instead of after the callback.
- Blisters and brown drips instead of sheets. That points to steam on an uncured coat or to moisture behind the trim, and the investigation goes to ventilation and leaks, not adhesion.
References
- ASTM D3359, Standard Test Methods for Rating Adhesion by Tape Test (current edition), used here as a field version in substance
- Primer and topcoat manufacturers' product data sheets
- See related: How to Test Glossy Trim and Prime Latex Over Oil
- See related: Bonding Primer Vs Stain Blocking Primer Vs No Primer Decision
- See related: Identifying the Existing Sheen and Coating Before a Repaint