Cabinet Refinishing Job Standard
Purpose
A kitchen is the only surface in residential painting the customer touches every day, judges at arm's length under task lighting, and closes against itself before the coating has finished curing. Two failures produce almost every cabinet callback: the finish that peels because grease or furniture polish was still on the door when the primer went down, and the door that sticks because it was re-hung before the product could resist blocking. This standard puts a measured cleanliness check and a measured adhesion check in front of the first topcoat, and a written cure schedule in front of the handover.
Scope
On-site and shop refinishing of previously finished wood, MDF and factory-coated cabinet doors, drawer fronts and boxes in residential kitchens, baths and built-ins, from survey to handover.
Not covered: raw new cabinetry; conversion varnish and other shop-booth chemistries, which the cabinet decision matrices own; thermofoil and laminate delamination, which is a replacement conversation rather than a coating one; and box replacement or door remanufacture.
Roles and responsibilities
| Role | Owns | Hands off |
|---|---|---|
| Lead painter | The cleanliness check, the adhesion test and its verdict, the release to topcoat | Does not release to topcoat until the tape test result is written on the ticket |
| Crew painter | Numbering, degreasing, sanding, dust removal, racking | Hands the lead a numbered test door, not a rack, for the adhesion test |
| Office | Books the cure window with the customer and confirms kitchen-down days in writing | Sends the written cure schedule to the customer before demolition of the kitchen's use begins |
Procedure
1. Survey, number and strip the hardware before anything is removed from the wall. Photograph every elevation, note damage, delamination, swelling at the sink base and any door that already rubs, then number each door and drawer front on a hidden edge with a matching bag for its hinges and pulls. Acceptance: a photographed count matching the numbered piece count exactly, one labelled bag per piece, and a written note of every pre-existing defect. Wrong, and the stop rule: a swollen or delaminated door does not come back from a coating, so it is quoted as a replacement now rather than discovered at re-hang. Hazard: overhead doors come off toward you, so a second hand supports the door before the last screw backs out.
2. Degrease every surface, then prove it with a water-break test. Wash with a degreaser suited to cooking soil, rinse with clean water, and check the rinse: a clean surface holds a continuous unbroken film, a contaminated one pulls the water into beads. Acceptance: a continuous water film holding 30 seconds without beading, checked on the door above the range and the one beside the sink, the two dirtiest positions in any kitchen. Wrong, and the stop rule: beading means residue is still there, so wash again rather than sanding over it, because sanding drives grease and silicone polish into the substrate where no primer reaches. Hazard: degreasers are commonly alkaline, so gloves and eye protection per the safety data sheet, no misting into the air you breathe, and if it reaches an eye, flush at an eyewash for 15 minutes and get medical attention rather than judging it by feel.
3. Deglaze to a uniform dull surface and remove the dust you just made. Sand every coated face to a consistent flat sheen, break the gloss in profiles with an abrasive pad, then vacuum with a filtered extractor and follow with a tack cloth or damp wipe. Acceptance: no glossy patches under raking light, and a white cloth wiped across the face comes away clean. Wrong, and the stop rule: shiny spots in profiles are where the finish lifts first, so re-sand rather than trusting the primer to bond. Hazard: sanding is an inhalation route whatever the coating, so use a filtered extractor and a respirator under a written program per 29 CFR 1910.134. If the housing is pre-1978 and the coating has not been tested, this SOP stops here and the RRP workflow governs, because disturbed lead paint is a dust hazard controlled under 40 CFR 745 Subpart E for occupants and 29 CFR 1926.62 for the worker.
4. Prime one test door and prove adhesion with a tape test before any production coating is applied. Apply the bonding primer to a single numbered door at the specified film, let it reach the recoat time on the data sheet, then run a cross-hatch tape test per ASTM D3359 in the edition your specification references, using the lattice method that standard assigns to thin films. Acceptance: classification 4B or better, meaning under 5 percent of the lattice squares removed. Wrong, and the stop rule: 3B or worse means the surface is still contaminated or under-deglazed, so the whole set goes back to step 2 and nothing is topcoated on a promise; log the classification and door number both times. Hazard: the razor knife cutting the lattice is the sharpest thing on the job, so the door is clamped rather than held and the cut runs away from the hand.
5. Spray the topcoat to the specified film, checking wet film on the first piece of each rack. Set tip and pressure from the sprayer ticket, spray a full rack, and take a wet film reading on the first piece before the rest goes up. Acceptance: wet film inside the range that produces the data sheet's dry film at the product's stated solids, and no runs, sags or dry spray at profile edges. Wrong, and the stop rule: heavy film on a cabinet does not just sag, it stays soft for weeks and blocks at the first close, so a heavy rack is laid flat and held to the top of the cure range rather than re-hung on schedule. Hazard: atomized coating and solvent are an inhalation route, so ventilate and use a respirator under the same 29 CFR 1910.134 program, and on a solvent-borne product bond and ground per NFPA 33 in the edition your authority having jurisdiction has adopted, with ignition sources including the kitchen's own pilot flames shut down first.
6. Verify hardness before anything is re-hung, and verify it on the slowest piece. After the data sheet's dry-to-handle time, press a thumb hard onto a hidden area of the piece sprayed last and coated heaviest, hold ten seconds, and look for a print. Acceptance: no visible thumbprint and no tack, on the last and heaviest piece rather than a convenient one. Wrong, and the stop rule: a print that remains means the film is still soft and will block at the first close, so the schedule moves and the doors stay on the rack, with the office told the same day so the customer hears it from them and not from a stuck door. Hazard: racked doors are a tipping hazard in a room a family still walks through, so racks sit out of traffic and are weighted or tied.
7. Re-hang, reset the hardware, and fit bumpers on every closing face. Hang each numbered door on its own bagged hinges, adjust for even reveals, reinstall pulls, and fit a felt or rubber bumper at each corner where a door or drawer front meets the frame. Acceptance: even reveals across each bank, every door closing without metal contacting coating, and a bumper on every closing face. Wrong, and the stop rule: a door that has to be forced is rubbing, not adjusted, and rubbing removes a fresh finish in days, so it comes back off rather than being closed harder. Hazard: pinch points at hinges, so the door is supported by a second hand or a shim stack while a screw starts.
8. Hand over with the cure schedule in writing, not as a verbal caution. Walk the kitchen with the customer, then hand over a written note with the product name, the date sprayed, the date the finish reaches the data sheet's full cure, and what to avoid until then: no shelf liner against a painted interior, no cleaning products on the faces, doors closed gently. Acceptance: written schedule handed over and acknowledged, same dates on the job file. Wrong, and the stop rule: a customer who needs the kitchen back before the cure date does not get a shortened cure, they get a resequenced job with the least-used bank last. Hazard: none, a conversation held in a finished kitchen.
The record this produces
The cabinet file carries the numbered piece list with a photograph per elevation, the pre-existing defect notes, the water-break result with the two positions tested, the ASTM D3359 classification and door number for every test including failures, the wet film readings per rack, the hardness check date and the piece it was taken on, and the signed cure schedule.
The lead reads the tape test before releasing to topcoat, which is its whole purpose. The office reads the cure dates when a customer calls about a sticking door, because a door that stuck partway through the stated cure is a schedule conversation and one that sticks well past it is a product or adjustment problem instead.
One filled-in run: 24 doors and 6 drawer fronts, occupied kitchen
Thirty pieces numbered and bagged, photographed count matches at thirty, so step 1 clears. One base door beside the dishwasher shows swelling at the bottom rail and is quoted as a replacement rather than coated. Step 2: after the first wash the range-side door beads within about ten seconds, so it is washed again and then holds a continuous film for the full 30 seconds, as does the sink-side door. Step 3: no gloss under raking light, wipe cloth clean.
Step 4 fails. Door number 11 is primed, given the data sheet recoat time, and tape tested. The lattice flakes along the cut edges and at intersections with roughly a tenth of the squares removed, which classifies as 3B, below the 4B gate, so nothing is topcoated. The set goes back to step 2, where the second wash reveals a slick residue on the doors nearest the range hood consistent with a silicone furniture polish rather than cooking soil. All pieces are re-washed and re-deglazed, door 11 is re-primed and re-tested, and the retest shows no squares removed, which classifies as 5B and clears the gate.
Step 5: wet film on the first piece of each of five racks lands inside the specified range. Step 6 is taken on rack five, the last sprayed, at the data sheet's dry-to-handle time, and ten seconds of thumb pressure leaves no print. Step 7: reveals even across all four banks, bumpers at every closing face. Step 8: written schedule handed over.
Self-check against the criteria as written: step 2 acceptance is a continuous water film held 30 seconds; first attempt beaded, FAIL; second held the full 30 seconds, PASS. Step 4 acceptance is 4B or better; 3B on door 11, FAIL, stop rule taken back to step 2; 5B on retest of the same door 11, PASS. The count is 24 doors plus 6 drawer fronts, 30 numbered, one replaced rather than coated, so 29 coated.
The exception this run hit is the one that matters most: a failed adhesion test is not a reason to press on with better technique, it is a reason to go back two steps. The one it did not hit is a customer who cannot give up the kitchen for the cure window, handled by sequencing the least-used bank last rather than by shortening a dry time.
References
- ASTM D3359, Standard Test Methods for Rating Adhesion by Tape Test, in the edition your specification references, for the lattice method and the 0B to 5B classification.
- Coating manufacturer technical data sheet, for film build, recoat window, dry-to-handle and full cure times. Label directions govern where they differ from anything here.
- 29 CFR 1910.134, for the written respiratory program covering both sanding dust and spray aerosol.
- 40 CFR 745 Subpart E and 29 CFR 1926.62, for lead-safe work practices when a pre-1978 coating is disturbed.
- See related: the cabinet refinishing HowTo for technique, the cabinet product decision matrices for chemistry selection, and the airless sprayer setup SOP for tip and pressure.