Deck Coating Substrate Compatibility: Plywood, Concrete, Metal, and Recoat

Acceptable panels

• Bond classification per DOC PS 1 / PS 2: Exterior (withstands repeated wetting) or Exposure 1 (same exterior-grade adhesive, not rated for sustained weather exposure). Per IBC 2021 § 2304.8, structural floor and roof sheathing must be Exposure 1 or Exterior; exposed exterior wall finish must be Exterior.
• APA grade stamp should be visible on each panel. Look for APA Rated Sheathing with a span rating (e.g., 24/16, 32/16, 40/20, 48/24). The first number is max roof support spacing in inches; the second is max floor joist spacing.
• APA Rated Sturd-I-Floor is a single-layer subfloor with tongue-and-groove (T&G) long edges that reduces seam telegraphing and deflection mismatch between panels. Helpful on deck assemblies receiving a bonded coating.
• Thickness per the system ESR. As a worked example, the Deck Flex W.M. Class A assembly under ICC-ES ESR-3672 calls out 5/8" minimum exterior-grade plywood with 2½ lb Deck Flex metal lath stapled at 22–24 #16 gauge staples per square foot. That is the substrate spec for that specific assembly. Other AC39-evaluated systems list different minimums. Read the ESR.
• Moisture content before coating. Industry practice is ≤15% moisture content measured with a pin or pinless moisture meter. AC39 itself specifies "structurally sound, clean, dry, free of contaminants" without a numeric cap. Verify against the specific system's ESR rather than treating 15% as a code number.

Prep before coating

• Fasteners set flush or slightly below the surface, no proud heads. Proud fasteners telegraph through every coat.
• Panel seams sealed per system ESR. Some systems call for a fabric-reinforced detail coat over seams, others for a butt-joint sealant plus fabric.
• Reinforcement fabric embedded at wall-to-deck transitions, penetrations, drains, cove risers, and door thresholds
• Panels dry to target before any bond coat hits. Tent and dehumidify if necessary, or reschedule.
• No power-wash within the cure window specified by the manufacturer; trapped moisture under a sealed film is how osmotic blisters form on plywood

Common failure modes

• Edge swell at panel perimeters where water has wicked into the core. Usually caused by a bad storage period, a long wet-in before coating, or coating over a wet panel.
• Seam telegraphing: coating cracks along panel butt joints as the assembly deflects. T&G panels or fabric-reinforced seam details reduce risk.
• Fastener pop: screw or nail heads back out under thermal cycling. Set fasteners well, use correct length, consider adhesive-attached blocking at high-stress points.
• Delamination from under-cured bond coat. Applying the next coat before the previous one has cured traps solvent or water under the film.

Cure, curing compounds, and laitance

• Industry benchmark: 28-day cure before coating. This is a practical baseline for strength gain and moisture release; it is not a hard ASTM floor. Early coating risks osmotic blistering as internal moisture vents through an already-sealed film.
• Curing compounds and form-release agents must be mechanically removed (shot blast, grind, scarify) before coating. They act as bond breakers under any acrylic, elastomeric, or polyurethane system.
• Laitance (the weak, cement-rich layer on freshly troweled concrete) comes off with the same mechanical prep. Do not coat over laitance.

Moisture testing

Concrete substrates need two measurements: surface moisture release (MVER) per ASTM F1869 and internal relative humidity per ASTM F2170 . They measure different things and both matter.

• ASTM F1869 (anhydrous calcium chloride test): 60–72 hour exposure of a sealed test dish on the slab surface, converted to lb of moisture per 1,000 ft² per 24 hr. A common ceiling in coating TDS is 3 lb / 1,000 ft² / 24 hr unless the system's ESR or TDS specifies otherwise. F1869 measures only surface moisture. It cannot see internal slab moisture that will release later.
• ASTM F2170 (in-situ relative humidity): holes drilled to 40% of slab depth (20% for slabs on grade dried from one side), sensors inserted and sealed, minimum 24-hour equilibration before reading. The numeric RH cap (typically 75% or 80%) is set by the coating or adhesive manufacturer, not by F2170. Cite F2170 for the method and coat per the manufacturer's RH limit.
• Run both tests on decks over occupied space and on any slab under 90 days old. F1869 alone is not enough for a walking deck application.

Alkalinity

• ASTM F710 surface pH test. Wet concrete runs pH 12–13; surface carbonation drops to pH 8–10 as the slab ages.
• pH >9 typically requires manufacturer consultation; alkaline surface can attack unprotected acrylic and cause saponification of oil-based primers
• Persistent high pH after cure can indicate moisture pushing fresh cement paste to the surface. Retest after further drying.

Surface profile (ICRI CSP)

ICRI 310.2R defines a 1–10 Concrete Surface Profile (CSP) scale, with standardized comparator chips. Different coating thicknesses demand different profiles.

• CSP 1–2: thin-film sealers and penetrating primers only (think rolled paint)
• CSP 3–4: thin coatings, thin-film waterproofing (acid-etch or light shot blast)
• CSP 5–7: heavier coatings, walking deck systems, traffic coatings (shot blast or scarify)
• CSP 8–10: high-build epoxy, polymer concrete overlays (heavy mechanical scarification)
• Carry an ICRI CSP comparator on the job; photograph the prepped slab next to the matching chip for the job file

Pull-off adhesion testing

• ASTM D7234 pull-off test for coatings on concrete (D4541's former Method A for concrete was replaced by D7234; legacy ESRs may still reference D4541 on concrete)
• Concrete tensile strength typically tops out at 400–500 psi. Many pull-off failures on concrete are substrate failures, not coating failures.
• Dolly glued to the cured coating, pulled perpendicular, instrument reports psi at failure plus failure mode (cohesive in concrete, adhesive at coating-concrete interface, cohesive in coating, etc.)

Crack and joint prep

• Static cracks: clean out, fill with manufacturer-specified crack filler or polyurethane sealant per C920 Class 25, detail-coat with reinforcement fabric
• Moving joints: maintain the joint through the coating with a backer rod and compatible elastomeric sealant; the coating should not bridge a moving joint rigidly
• Re-entrant corners and penetrations: fabric-reinforced fillet detail, not a hard 90° turn
• For sealant selection across construction joints, see the 2025 Sealant & Caulking Guide.

Common failure modes

• Osmotic blistering from residual slab moisture pushing through a sealed coating film
• Alkali attack on acrylic binders from high-pH surface, causing chalking, softening, or delamination
• Delamination from slick troweled surfaces (CSP 1 under a system that needed CSP 5–7)
• Pinhole porosity in thin coats over porous slabs. Air or moisture from the substrate breaks the film as it cures.
• Bond failure at the coating-concrete interface traced to residual curing compound or form-release agent

Always test first

• Pull-off adhesion per ASTM D7234 (concrete substrate under existing coating) or ASTM D4541 (metal substrate under existing coating) on at least 3 sample locations per deck
• Report the psi and the failure mode. Substrate cohesive failure (break inside the concrete or plywood) at 150+ psi is usually workable; adhesive failure at the existing coating-to-substrate interface at low psi is not
• Moisture test the substrate under the existing coating at the pull-off sample cuts. Coating-trapped moisture is a common and invisible failure driver.
• Inspect the visible film for alligatoring, chalking, delamination, blistering, soft spots, and intercoat cleavage between base and topcoat

Recoat window and intercoat contamination

• Every coating has a manufacturer-specified recoat window measured from last-coat cure. Within the window the next coat bonds chemically; past the window it has to bond mechanically (sand, abrade, or prime)
• Missed recoat windows produce intercoat delamination, the old cause of failure that does not show up until the first heat cycle or rain event
• Contamination between coats (dust, amine blush on some epoxies, laitance, efflorescence, pollen, silicone overspray) needs mechanical or solvent removal before the next coat
• Amine blush and efflorescence should be scrubbed and rinsed, then dried before top-coating

Strip vs recoat decision

Surface prep (SSPC)

• SSPC-SP 3 (Power Tool Cleaning): removes loose rust, loose mill scale, loose paint. Minimum for mild service, not acceptable under most AC39 traffic coatings.
• SSPC-SP 6 / NACE 3 (Commercial Blast): removes all visible rust, mill scale, and coatings; staining permitted on ≤33% of surface. Typical for walking deck traffic coatings.
• SSPC-SP 10 / NACE 2 (Near-White Blast): 95% contaminant-free. Required for high-performance vehicular traffic coatings and coastal-exposure jobs.
• Immediate primer application after blast. Flash rust on freshly blasted steel can form within hours, especially in humid environments.

Adhesion testing

• ASTM D4541 is the pull-off test standard for coatings on metal substrates (D7234 is for concrete)
• Typical acceptance thresholds from manufacturer TDS: 300–500+ psi for traffic coatings, failure mode reported alongside psi

Galvanic risk at mixed-metal flashings

• Where dissimilar metals meet (galvanized steel + aluminum, galvanized steel + copper), galvanic corrosion attacks the less-noble metal
• Isolate metals with a non-conductive gasket, elastomeric sealant, or compatible dielectric tape
• Coating over a galvanic cell does not stop the corrosion underneath. The cell has to be isolated first.

Why OSB fails under walking deck coatings

• OSB absorbs moisture vertically through the panel edges much faster than through the face. Any edge exposure (cut, seam, panel perimeter) wicks water into the core.
• Once wet, OSB swells at the edges and does not fully dry back to original thickness. The ridge is permanent.
• Swollen panel edges telegraph through every coating system as a raised seam that cracks and debonds under thermal cycling
• Long-term wet OSB supports rot and mold, often concealed under an intact-looking coating until structural failure

If OSB is found during a recoat inspection

• Remove and replace with exterior-grade or Exposure 1 plywood at the thickness the system ESR specifies
• Do not attempt to salvage OSB by sealing the edges, applying a barrier coat, or laminating plywood over the top. None of these restore the panel's core integrity.
• Check adjacent framing for rot or insect damage while the substrate is open

Which test for which substrate

Who reads the meter

• Surface F1869 kits are operable by any trained crew member; install per instructions, log start and finish times, and weigh the dish to gauge weight
• F2170 probes benefit from a single consistent operator across the tests on a job. Sensor handling and seating errors are common sources of misread RH.
• Pin-type plywood meters need fresh pins for each panel. Dull pins read high.
• Photograph meter readings in-situ with a tape measure or the panel stamp in-frame for the job log

What to record in the job file

• Test method and equipment make/model
• Reading, units, location (sketch or photo), and timestamp
• Ambient temperature, surface temperature, and relative humidity at time of reading
• Pass/fail against the system's TDS or ESR threshold, not against an internal shop rule

Plywood

• APA grade stamp verified on each panel: Exposure 1 or Exterior, span rating appropriate for the joist spacing
• Thickness meets system ESR minimum (example: 5/8" for Deck Flex W.M. under ESR-3672)
• Moisture content ≤15% (industry practice) or per specific ESR requirement. Pin or pinless meter, readings logged.
• Fasteners set flush, no proud heads, corrosion-resistant fasteners
• Panel seams prepped per system (butt-joint detail, fabric reinforcement, or both)
• Metal lath (if assembly requires) stapled at spec rate (e.g., 22–24 #16 ga staples/sqft for Deck Flex W.M.)
• Deck surface clean, dry, free of grease, dust, curing contamination
• No rain in forecast within manufacturer cure window

Concrete

• Slab age ≥28 days (or per ESR)
• Curing compound / form-release agent mechanically removed
• Laitance removed
• CSP profile per ESR. Verified with ICRI comparator chip, photographed.
• ASTM F1869 MVER tested, passing manufacturer threshold
• ASTM F2170 in-situ RH tested (24-hour minimum equilibration), passing manufacturer threshold
• ASTM F710 pH tested, passing manufacturer threshold
• Static cracks detailed with fabric-reinforced filler; moving joints sealed with compatible elastomeric sealant and backer rod
• Penetrations and drains fabric-reinforced

Recoat over existing coating

• Existing system chemistry identified (manufacturer, ESR if known)
• Pull-off adhesion (ASTM D7234 concrete / D4541 metal) passed on ≥3 samples, psi and failure mode logged
• Substrate moisture re-tested at sample cuts
• Visible film inspected for alligatoring, chalking, blisters, soft spots. Defects mapped and measured.
• Recoat vs strip-to-substrate decision documented and signed off
• Existing film abraded or primed per new system ESR requirements
• New system ESR Conditions of Use explicitly permits recoat over existing coating type (or a sample patch adhesion test is passing)
• Spot repairs complete and re-tested before full recoat

Does AC39 specify a numeric moisture content for plywood?

No. AC39 calls for plywood to be "structurally sound, clean, dry, and free of contaminants" without stating a numeric cap. The commonly cited ≤15% moisture content is industry practice, not a code minimum. Numeric thresholds come from the specific system's ESR or manufacturer TDS. Verify against the evaluated product being installed, not against a generic rule of thumb.

What is the concrete in-situ RH cap for deck coatings?

ASTM F2170 is the test method, not the acceptance threshold. The numeric RH limit (commonly 75% or 80%) is set by the coating or adhesive manufacturer on its TDS. It is not a code number and it is not in F2170. Cite F2170 for the method and coat per the manufacturer's RH limit for the specific system being installed.

Should pull-off adhesion on concrete be tested per ASTM D4541 or D7234?

ASTM D7234 is the current pull-off standard for coatings on concrete. D4541's former Method A for concrete was removed and replaced by D7234. ASTM D4541 remains the pull-off test for coatings on metal substrates. Legacy ESRs may still reference D4541 on concrete; cite whichever standard the specific ESR names, and use D7234 on new concrete work.

Can I coat over OSB on a walking deck?

No. OSB absorbs moisture vertically through panel edges, swells permanently, and telegraphs through any bonded coating as a cracked, raised seam. Most AC39-evaluated walking deck systems explicitly exclude OSB from their Conditions of Use. If OSB is found during a recoat inspection, remove and replace with exterior-grade or Exposure 1 plywood at the thickness the system ESR specifies.

Why does California SB 721 or SB 326 repair work care about substrate?

SB 721 (Health & Safety Code § 17973) and SB 326 (Civil Code § 5551) require inspection of exterior elevated elements on multifamily buildings. Repairs identified by the inspector typically have to use a code-compliant assembly, and for decks over occupied space that is almost always an ICC-ES AC39 evaluated walking deck system. An AC39 system is only code-accepted over the substrates named in its ESR Conditions of Use. That is why the substrate decision drives the repair scope, not the coating brand. For deadline and statute detail, see California Balcony Law (SB 721 / SB 326).

Do I need both ASTM F1869 and F2170 on every concrete job?

For walking deck coatings over occupied space, running both tests is good practice. F1869 measures surface moisture release; F2170 measures internal humidity at 40% slab depth. Surface dryness does not prove the slab will not vent moisture through a newly sealed film later. Slabs under 90 days old, slabs with unknown history, and any deck that will receive a Class A assembly are strong candidates for both tests. Single F1869 runs can be appropriate on well-aged slabs with documented curing history when the TDS allows.

What CSP profile does a cementitious-acrylic walking deck system need on concrete?

Depends on the system. Thin sealers and primers may work at CSP 1–2; thin coatings at CSP 3–4; most walking deck and traffic coatings at CSP 5–7; high-build epoxy and polymer overlays at CSP 8–10. The ESR or TDS names the required profile. Verify on site with an ICRI 310.2R comparator chip and photograph the prepped slab next to the matching chip for the job log.

Is CRC § R503 the code path for exterior balcony sheathing?

No. The California Residential Code § R503 covers wood floor sheathing in interior applications. Exterior balcony and deck sheathing on multifamily buildings falls under IBC / CBC Chapter 23 (wood), plus CBC § 1402 (exterior wall and weather-resistant barrier) and ICC-ES AC39 for the walking deck waterproofing system itself. Always confirm the code path with the authority having jurisdiction for your occupancy class and project type.