WALL CONDENSATION CHECK · GUIDE
Will my wall condense? One wall, worked month by month
The wall: 1/2 in drywall with latex paint, an R-21 fiberglass batt in a 2x6 cavity, 7/16 in OSB sheathing and housewrap, with vented siding. The climate: an invented cold-climate year with a 20 F January, roughly a northern US city, and a house held at 70 F and 35% relative humidity in midwinter. Every number below is the calculator's output on that example, which you can load on the tool page with one click.
The layer values are typical rounded figures, not one product's data sheet, and the climate is invented. Change them to yours.
1. January: where is the dew point?
Indoor air at 70 F and 35% carries vapour at 876 Pa. Its dew point is 41 F: any surface colder than that, with that air touching it, gets wet. Temperature falls through the wall in proportion to R-value. The batt holds almost all of the wall's R-22, so nearly the whole drop from 70 F to 20 F happens inside the cavity, and the inner face of the OSB sits at 22 F (-5.6 C). Saturation pressure there, over ice, is only 380 Pa.
Without condensation, vapour pressure would fall in a straight line in proportion to vapour resistance. The paint (5 perm) and the batt (20 perm) let vapour through easily; the OSB (1.5 perm) does not. So the straight line reaches the OSB face still far above 380 Pa. That is physically impossible: the vapour pressure at the OSB face is clamped at saturation, and the difference between vapour arriving through the paint and batt and vapour leaving through the OSB condenses there. In January that is 8.2 g/m² per day.
2. The winter adds up
Carried month to month, the first month that condenses after a dry month is November. By the end of March the OSB face holds 744 g/m², about three quarters of a litre on every square metre of wall. If all of it went into the OSB (taking 7/16 in OSB as about 7 kg/m² dry, an assumption), that is roughly eleven points of moisture content on top of whatever the panel started at. In April the flow reverses: the wall dries both ways, inward through the vapour-open paint as well as outward, and the face is dry again by the end of June. It is wet for 7 months of the year.
Glaser is known to be pessimistic here because it ignores the water that the wood and the batt absorb and the sun that warms the sheathing. It is also blind to air leaks, which in a real wall can deliver far more water to the same OSB face than diffusion does. Read the 744 as "this wall is vapour-sensitive", not as a measured amount.
3. Three changes
| Wall | Peak at the OSB | Months wet | Indoor RH that first wets it |
|---|---|---|---|
| As above (paint, batt, OSB) | 744 g/m² | 7 | 17% |
| Plus 1 in XPS outside the OSB | 316 g/m² | 5 | 25% |
| Plus 2 in XPS outside the OSB | 61 g/m² | 2 | 31% |
| Polyethylene instead of paint, no foam | none | 0 | 92% |
Exterior foam works by warming the sheathing, not by blocking vapour: with 2 in of XPS the OSB face sits above freezing in January, its saturation pressure roughly doubles, and most of the condensation goes away. Polyethylene on the warm side stops vapour before it reaches the cold face, which in this heating-dominated climate removes the winter problem entirely. It also stops the wall drying inward, which matters in a cooled house in a humid summer, and this model does not have the wet-cladding, sun-driven summer case that shows it.
The last column is the most useful single number. It is the highest indoor humidity, held constant all year, at which nothing condenses anywhere. If the house will run above it, the wall needs a change or a closer look.
Common questions
Why does the OSB get wet and not the drywall?
Because it is the first layer on the cold side of the insulation that slows vapour down. Almost all of the temperature drop happens inside the batt, so the inner face of the OSB is close to the outdoor temperature, while the paint and the batt let indoor vapour reach it easily.
Does exterior foam always help?
In this example it does, because it warms the sheathing: with 2 in of XPS the OSB face stays above freezing in January. Foam that is too thin for the climate helps less, and a foam that also blocks vapour slows drying to the outside, so the amount matters.
Is 744 grams per square metre a real measurement?
No. It is what the steady-state Glaser model gives for typical layer values and an invented climate. The model has no moisture storage, sun or air leakage, so read it as a sign the wall is vapour-sensitive, not as the water a real wall would hold.
Try your own wall
Open the wall condensation check
The research model behind the tool, with the same physics in Python and a climate-by-retarder table: house-dewpoint.