Wood Moisture Content and the Match to the Subfloor

Why this matters

A hardwood floor that cups the first summer or opens gaps the first winter was usually condemned the day it went down, by a moisture difference nobody wrote down. The flooring was dry, the subfloor was wet, or both were fine on average and one corner was soaking. The match between the two is a number you can read in fifteen minutes, and it is the number a claims inspector will ask for first.

What moisture content means and where it settles

Moisture content (MC) is the weight of water in the wood as a percentage of the wood's oven-dry weight. Wood is always trading water with the air around it and drifts toward an equilibrium moisture content (EMC) set by the air's temperature and relative humidity.

The USDA Wood Handbook's EMC table is the reference most of the trade uses. At about 70 F, air at 30 percent RH holds wood near 6 percent MC, 40 percent RH near 7.7 percent, 50 percent RH near 9.2 percent, and 60 percent RH near 11 percent. That is why "in-service conditions" matters more than any number on a carton: the floor will end up wherever the house runs, and the job is to install it close to that point.

Two consequences follow from that table:

• A house is a range, not a point. A home in a heating climate that runs 30 percent in January and 55 percent in August swings its floor through roughly 4 points of MC over the year (about 6.2 to 10). The goal is to install near the middle of the range the house actually runs, with the HVAC on.
• The subfloor is wood too. Plywood and OSB carry their own MC and move toward the same EMC. A subfloor that is still drying out from a wet framing season will hand its water to the flooring above it.

The match rule

The rule as commonly stated in NWFA guidance: for strip flooring under 3 inches wide, the flooring and the wood subfloor should read within 4 percentage points of each other; for plank 3 inches and wider, within 2 points. NWFA guidance also commonly states that a wood subfloor reading over 12 percent is not ready to receive flooring, whatever the flooring reads. Check the current edition of the NWFA installation guidelines, and where the flooring maker's instructions are tighter, the maker wins. Some engineered products publish their own window, and some solid-plank makers ask for an even closer match than 2 points.

Why width tightens it. Wood moves across its width as MC changes, and a wider board has more width to move. A 5 inch plank responds to a given MC change with more than twice the movement per board of a 2-1/4 inch strip, and that movement concentrates at fewer joints. A difference a strip floor shrugs off as hairline gaps shows up in plank as cupping.

Which way it fails. When the subfloor is wetter than the flooring, the bottom of each board takes on water and swells while the top stays drier, and the edges rise above the centers: cupping. When the flooring is the wetter of the two and dries in place, it shrinks and opens gaps. Neither shows on install day. Both show the first time the season turns.

Reading a pin meter honestly

Pin meters read electrical resistance between two pins, and resistance depends on species as well as water. Most are calibrated to one reference species, and a reading on oak, maple or a tropical species has to be corrected, either with the meter's species setting or with the maker's correction chart. The correction can move a reading by a point or more, which is the whole width of the plank tolerance. Pin meters also read differently with temperature; meters that do not compensate automatically come with a temperature chart.

Pinless meters read a field to a set depth through the surface, and they are set by specific gravity (density) rather than species name. A pinless meter on the wrong density setting reads confidently wrong, and the depth it reads can include whatever is under a thin board.

Practical rules that hold for both:

• Correct before you compare. Log corrected values, and log the setting you used, so a reading taken by someone else can be checked.
• Read the subfloor and the flooring with the right setting for each. Plywood and OSB are not the flooring species. Use the meter's setting for the subfloor material or the maker's guidance for panels, and say which.
• One reading proves nothing. A single board can be an outlier from the edge of a bundle, a single subfloor spot can sit over a supply line. Wood MC varies board to board within a carton by a point or more. Only a spread of readings tells you whether you are looking at a material or a spot.

How many readings, where, and how to lay them out is a procedure: see related: How to Take and Log Wood Moisture Readings Before Hardwood. This card is about what the numbers mean once you have them.

Worked example: the 600 square foot log

The job: about 600 sq ft of 2-1/4 inch solid red oak strip over 3/4 inch plywood in a living room, dining room and hall of a 1990s house. The HVAC has run for the week the flooring has been on site, and the thermostat shows 70 F and 38 percent RH. From the EMC table, that air holds wood a little under 7.7 percent. The crew takes 10 flooring readings from boards pulled across several cartons and 12 subfloor readings spread across the three rooms, pin meter, corrected for species and temperature.

JOB MOISTURE LOG  - 600 SQ FT, 3 ROOMS
Flooring: 2-1/4 red oak strip, pin, oak setting
Subfloor: 3/4 plywood, pin, panel setting
Air: 70 F / 38 percent RH, HVAC running 7 days

FLOORING (corrected)    SUBFLOOR (corrected)
 F1  7.2                 S1  10.2  living
 F2  7.6                 S2  10.6  living
 F3  7.0                 S3  10.4  living
 F4  7.8                 S4  10.8  living
 F5  7.4                 S5  10.5  dining
 F6  7.1                 S6  10.3  dining
 F7  7.5                 S7  10.7  dining
 F8  7.3                 S8  10.6  dining
 F9  7.6                 S9  10.4  hall
 F10 7.5                 S10 10.5  hall
                         S11 10.9  hall
                         S12 14.1  dining, at
                                   kitchen wall
Avg F1-F10: 7.4          Avg S1-S11: 10.5
                         Avg S1-S12: 10.8

Step one: the flooring. The ten readings run 7.0 to 7.8, average 7.4 (sum 74.0 over 10). That is a tight spread close to the house's EMC, so the flooring has acclimated.

Step two: the subfloor, and the outlier. Eleven readings run 10.2 to 10.9 and average 10.5 (sum 115.9 over 11). The twelfth, S12, reads 14.1 at the dining room wall shared with the kitchen. That is 3.2 points above the next highest reading and over the 12 percent ceiling on its own.

Step three: the match. Against the body of the subfloor, 10.5 minus 7.4 leaves 3.1 points. Under the 4 point strip rule, this floor passes.

Step four: what averaging would have hidden. Fold S12 into the average and the subfloor reads 10.8 (130.0 over 12), and 10.8 minus 7.4 is 3.4, still a pass. A crew that only logs the average installs over a wet spot and never knows. S12 is not a rounding problem, it is a location: the crew pulls the kick plate on the kitchen side and finds a weeping dishwasher supply. That gets fixed by whoever the shop sends for plumbing, the area dries with the HVAC running, and it is re-read before any flooring covers it.

Step five: the same log under plank. Suppose the customer had chosen 5 inch red oak plank instead. The same 3.1 point difference fails the 2 point plank rule by 1.1 points. The fix is not to wet the flooring up to the subfloor: the house runs near 7.7 percent EMC, so the subfloor is what is out of step with the building. Leave the HVAC running, re-read the subfloor over the following days, and install when the difference closes to 2 points or less on the corrected readings, or when the maker's own window is met if it is tighter.

What changes the call

• A dry flooring delivery in a humid climate. If the flooring reads 6 percent in a house that runs 10 percent EMC all summer, the subfloor may match the house fine and the flooring is the outlier. Let it acclimate toward the house, verified with the meter, rather than installing tight and watching it swell.
• Engineered flooring. The cross-ply core moves much less than solid, and many engineered makers publish their own subfloor limit rather than a match window. Use theirs.
• New construction. Framing that stood in weather can read in the high teens. No amount of acclimating the flooring fixes that; the building has to dry with heat and air running.
• Concrete. A slab is not tested with a wood meter at all. See related: Slab Moisture Testing for Flooring.

How a wrong log shows up later

The failure that comes back is rarely a bad meter. It is a log with three readings, all from the same carton, taken at the truck, with no species setting written down. When the floor cups, nobody can show what the subfloor read, and the claim inspector's first question has no answer. A log with corrected values, locations, the setting used, and the air conditions on the day is what lets a shop say the floor was installed in tolerance and point to the water source that arrived later.

References

• NWFA Wood Flooring Installation Guidelines, current edition (moisture match and subfloor limits)
• USDA Forest Products Laboratory, Wood Handbook (equilibrium moisture content tables)
• Moisture meter maker's species and temperature correction charts
• See related: How to Take and Log Wood Moisture Readings Before Hardwood
• See related: Slab Moisture Testing for Flooring