Subfloor and Crawlspace Drying Technique

Why this matters

A subfloor drying job run from one dehumidifier at one end of a 40-foot crawlspace looks identical, on the equipment log, to one run correctly: same dehu, same air movers, same daily numbers written down. The difference shows up three weeks later when the far end, the section the airflow never actually reached, is still damp enough to feed a musty smell back up through the floor register. A subfloor and the crawlspace under it are really two connected drying problems worked as one assembly: air has to move across the joist bays from below and reach the subfloor from above, in a space too tight to walk and too easy to under-ventilate at one end of. Get the zoning and the per-bay tracking wrong and the job reads dry on the log while a section of the house is still open.

Before anyone goes under the house

Bulk standing water, category posture, and the extract-versus-encapsulate call belong to the crawlspace decision matrices, not this technique; confirm that work is already done before treating a crawlspace as a drying zone (see References). What this technique adds is the entry discipline for a space where a tech will be working repeatedly, not just walking through once. Test the atmosphere (oxygen, flammability, carbon monoxide and hydrogen sulfide where sewage is a possibility) before the first entry and again at intervals through an extended work session, not once at the start of the job. Work with an attendant stationed at the access point on any crawlspace that requires more than a quick look, with an agreed communication method that does not depend on voice carrying through the floor system, a hard line or a rope-tug signal if radios do not reach. De-energize any circuit that runs exposed through the space unless it is confirmed dead with a meter, and treat standing dampness under a live circuit as a stop condition, not a detail to work around.

If the atmosphere monitor alarms at any point, evacuate immediately and ventilate the space; do not re-enter on a fixed wait time and an assumption that it cleared. Re-test on the same monitor and confirm a safe reading before anyone goes back in, every time, including the second and third entries of the same job.

Compartmentalizing a long or open crawlspace into workable zones

A single dehumidifier and a couple of air movers placed at one access hatch will dry the joist bays nearest that hatch and do almost nothing 30 feet away; airflow in a low, obstructed crawlspace does not travel the way it does in an open room. Hang poly baffles between joist bays or across the crawlspace width to break a long run into zones sized to what one dehu and its air movers can actually cover, then treat each zone as its own chamber with its own equipment and its own readings, the same logic that governs any staged drying assembly, applied below grade. A zone boundary at a natural break, a mid-span support wall, a plumbing chase, or the practical limit of a low, tight section, both improves airflow reach and makes it obvious on the log which equipment serves which readings.

Working the joist bays from below and the subfloor from above as one assembly

From below, point air movers or an injection hose along the joist bays rather than across them, so moving air sweeps the full length of each bay instead of just the entry point; a mover aimed straight at the soil or perpendicular to the joists moves a lot of air without covering much wet surface. From above, use existing floor registers, an access panel, or a pulled toe-kick as entry points for a probe or a short injection run rather than cutting into sound subfloor; the goal is to work both faces of the same assembly without opening material that does not need to be opened. Where the subfloor is also being dried from a room above (a mat system or room-level dehumidification), coordinate so the two chambers are not fighting each other, the crawlspace pulling moisture up into the room while the room chamber is trying to hold a lower vapor pressure than the crawlspace can support.

Sizing dehumidification for a below-grade chamber

Crawlspace air runs colder than room air, and a refrigerant dehumidifier's capacity drops sharply as intake temperature falls, which is the reason a desiccant unit is the default choice below grade rather than the refrigerant unit that might be sized correctly for the room above (see References for the full selection logic). Undersizing dehumidification for a compartmentalized zone shows up the same way it does above grade: air movers running, gauges showing flow, and the zone's grain depression barely moving because there is not enough dehumidification capacity to hold the vapor pressure gradient that actually pulls water out of the joists and subfloor.

Reading joist bay by joist bay

Read every zone's joist bays individually with a deep probe, logged against an unaffected reference joist read elsewhere in the same crawlspace on day one, using the same 4-percentage-point convergence default and 2-point-per-72-hour progress check used on any structural drying zone. While the probe is in hand at each bay, run the other hand along the subfloor underside and the panel seams at every visit; edge swell or delamination is a tactile finding a probe reading alone will not catch, and it belongs on the log the same day it is felt, not just at final walk-through. That finding routes to the OSB dry-versus-cut-versus-replace decision tree; this technique's job is to report it accurately and promptly, not to make the replace call itself.

The stop rule for a bay that will not converge

If a zone's bay has not dropped at least 2 points over a 72-hour window despite confirmed airflow reaching that bay, first confirm the baffle boundary is still sealed and the air mover is still aimed along the bay rather than having drifted or been bumped, since a below-grade setup gets disturbed more easily than a room chamber and nobody walks through to notice. If the setup is confirmed correct and the bay still will not move, that bay is escalated on its own for a source recheck (wicking from an adjacent wet zone, a second unidentified source, or standing moisture trapped under a low spot in the vapor barrier) rather than left running on the assumption that a crawlspace simply dries slower than a room.

Worked example: a 40-foot run under a kitchen and hallway

A slab-adjacent crawlspace took on water from a failed supply line, wetting the joists and subfloor above a 40-foot run under the kitchen and an adjoining hallway. Bulk water has already been pumped and the vapor barrier removed per the standing-water decision tree; this technique picks up from a wet but drained crawlspace. The hallway section is markedly lower and tighter than the kitchen section, so the run is split into two 20-foot zones with a poly baffle at the midpoint, each with its own dehumidifier and air movers run along the joist bays.

Day 0 deep-probe readings: Zone 1 (kitchen) averages 32 percent on the joists, Zone 2 (hallway) averages 29 percent. An unaffected reference joist elsewhere in the crawlspace reads 12 percent, setting the convergence target at 16 percent or lower. During the initial crawl-through, one subfloor panel seam in Zone 2, near the kitchen doorway, shows a raised ridge under a bare hand, edge swell, logged and flagged to the OSB decision tree the same day.

Day 3: Zone 1 reads 22 percent, a 10-point drop. Zone 2 reads 24 percent, a 5-point drop, slower even after compartmentalization because the lower clearance limits how far the air movers can throw usable airflow down that run.

Day 6: Zone 1 reads 16 percent, at target; its equipment is redirected to reinforce Zone 2's far end. Zone 2 reads 19 percent, a 5-point drop, holding pace.

Day 9: Zone 2 reads 15 percent, inside target. Both zones are converged on moisture. The flagged seam in Zone 2 is re-checked by hand at this visit: the ridge is still raised and does not press flat, confirming permanent edge swell independent of the moisture reading. That panel is flagged for cut-and-replace even though the zone it sits in passed on MC, because a board that reads dry and still carries a permanent ridge will telegraph through finish flooring regardless of what the meter says.

Closing out: pulling equipment, restoring the barrier, verifying from both sides

Pull air movers, hoses, and the compartmentalization baffles only after every zone in the run has been re-read and confirmed at target, and re-check every bay flagged for a tactile finding one more time before equipment leaves, since a ridge or a soft spot can be easy to miss on a single pass in low light. Hand off the ground vapor barrier to whatever plan the extract-versus-encapsulate decision already set for this job, temporary poly for now or a sealed encapsulation system, rather than treating barrier replacement as part of this technique. Final verification is a deep-probe reading at every bay logged during the job plus a from-above confirmation at any register or panel used as an access point, not a single reading taken at the access hatch and assumed to represent the whole run.

References

• See related: Crawlspace Standing Water: Extract vs Encapsulate vs Vent Decision Matrix, for the bulk-water and vapor-barrier decisions this technique assumes are already made.
• See related: Crawlspace Standing Water Pump vs Extract vs Dry Decision Tree, for Category posture and confined-space hazard scope before entry.
• See related: OSB Subfloor Swell Dry vs Cut vs Replace Decision Tree, for the replace-versus-dry-in-place call a tactile edge-swell finding routes to.
• See related: Desiccant vs Refrigerant Dehumidifier Selection, for the full below-grade capacity comparison summarized above.
• ANSI/IICRC S500, Standard and Reference Guide for Professional Water Damage Restoration, current edition, sections on structural cavities, crawlspaces, and confined-space safety.
• OSHA 29 CFR 1910.146, Permit-Required Confined Spaces, for atmosphere testing and attendant requirements.