How to Work Out Where Condensation Will Form

Why this matters

Most condensation calls are diagnosed by looking for water. That works when the water is on a window and fails completely when it is on the back of the sheathing, because by the time it shows on the inside face the assembly has been wet for two winters. The temperature at every plane inside a wall is computable from numbers you already have, and comparing those temperatures to a dew point tells you where the water forms before you open anything. It also tells you when the goal you were about to sell the customer is unreachable, which is the more valuable half.

Before you drill, open, or enter anything

You will be putting a bit or a probe into a wall you cannot see inside. De-energize the circuits serving that wall at the panel, lock and tag them, and prove the conductors dead with a meter checked on a known live source before and after, per 29 CFR 1910.333(b)(2) and the live-dead-live sequence at NFPA 70E-2021, 120.5. Assume plumbing runs where you cannot see it and scan before you commit the bit.

In housing built before 1978, drilling or sanding a painted surface disturbs lead paint, and the hazard route is inhalation and ingestion of the dust, not skin contact. Work wet, contain the area, collect dust with a HEPA vacuum rather than a shop vacuum, and follow the EPA renovation, repair and painting requirements at 40 CFR Part 745 Subpart E, which apply to compensated work in pre-1978 target housing.

Loose fibrous insulation is a respiratory hazard, so wear a fitted particulate respirator rather than a nuisance mask for any attic or cavity work. If you find loose granular fill of unknown origin, stop, do not disturb it further, and treat it as presumed asbestos-containing until tested. For crawl space and attic entry hazards, the sibling article on dew point carries the confined-space and atmosphere-testing requirements.

Step 1: Identify the air masses and get a dew point for each

Not the rooms, the air masses. A house frequently has three or four: conditioned living space, an attic, a crawl space or basement, and sometimes a garage. Each carries its own water content.

Measure or compute dew point for each, from co-located dry bulb and relative humidity taken with a probe that has had time to settle. Skipping this step and using an outdoor forecast dew point is the single most common way this calculation goes wrong, because the air that reaches the plane in question is usually indoor air.

Step 2: List the layers and get honest R-values

Write the assembly out in order from inside to outside, each with its R-value, and include both air films. The films are small and they are not optional; the interior film alone is comparable to a sheet of gypsum.

Two things break this step:

A cavity is not its label. A batt compressed around wiring, stuffed rather than fitted, or with a gap at the top is delivering less than its rated R. If you can inspect it, inspect it. If you cannot, run the calculation twice, once at label value and once at a reduced value, and see whether the conclusion changes.

Framing is a separate path. A stud has roughly a fifth the R-value of the cavity beside it. That makes the interior surface over a stud colder than the interior surface over the cavity, which is why dust ghosting shows up as stripes on the wall exactly where the framing is. It also makes the sheathing behind a stud warmer than the sheathing behind the cavity, because more heat is arriving there. So the coldest sheathing is mid-bay and the coldest interior surface is on the framing. Run whichever path answers the question you were asked.

Step 3: Pick the outdoor temperature that matches the question

This decides the answer and it is the step people skip.

• "Will it drip tonight?" Use the design outdoor temperature or the actual forecast low. You are asking about an event.
• "Is this assembly going to rot?" Use the mean outdoor temperature for the coldest month. You are asking about accumulation over weeks, and a design-hour temperature will condemn every wall in a cold climate.

Wetting for a few hours during a cold snap is normal and assemblies dry from it. Wetting continuously for a month at a rate above the drying rate is a failure. Same math, different input, different verdict.

Step 4: Compute the temperature at each interface

The temperature drop across any layer is proportional to its share of the total R-value. So for any plane in the assembly:

Plane temperature = indoor temperature minus (total temperature difference x R inboard of the plane divided by total R)

Do it for every interface, not just the one you suspect. Cheap, and it occasionally shows you that the plane you were worried about is fine while a different one is not.

Step 5: Ask which air can actually reach that plane

A computed temperature only matters if humid air gets there. Two routes reach a plane inside an assembly: diffusion through the materials, which is slow and steady, and air leakage through holes, which is fast and carries far more water. The sibling article on how moisture moves through a building covers the difference in rate and the signature each leaves.

For this procedure, the practical form is: if there is an air path from the conditioned space to the plane, use the conditioned space's dew point. If the assembly is genuinely airtight to that plane, diffusion still delivers water, more slowly, and the same dew point applies with a longer time constant.

Step 6: Find the first plane below dew point

Compare each computed plane temperature against the dew point of the air that reaches it. The first plane going outward that sits below dew point is the condensing plane, and in a heating-climate wall it is almost always the inside face of the sheathing, because that is where the temperature has dropped through nearly all the insulation and there is a material change to condense on.

Step 7: Pick a lever, then check that the lever can reach

Three levers, and only three:

• Raise the plane's temperature by adding insulation outboard of it, which shifts the temperature profile without changing the assembly's inboard R.
• Lower the dew point of the air reaching the plane, by source control or dehumidification.
• Block the air path so much less water arrives, which does not change any temperature but changes the rate.

Then do the arithmetic on the lever you picked. Frequently one of them turns out to require a value nobody will accept, and finding that out on paper is the point of the exercise.

The worked pass

A cold-climate house. Indoor conditions 70 F and 35 percent relative humidity, which is a dew point of about 41 F. Read that off an instrument or a chart; the familiar rule of thumb overstates dew point by several degrees at this humidity and would put you at 47 F, which changes the answer.

The assembly, inside to out:

R inboard of the sheathing's inner face: 0.68 + 0.45 + 13.0 = 14.13. Total assembly R: 14.92.

The event question. Design outdoor temperature 10 F, so a 60 F difference. Temperature drop to the sheathing's inner face is 60 x (14.13 / 14.92) = 60 x 0.947 = 56.8 F. That plane sits at 70 minus 56.8, about 13.2 F. It is 28 degrees below the indoor dew point, and below freezing, so what forms there is frost rather than liquid.

The accumulation question. Coldest month mean outdoor temperature 25 F, a 45 F difference. Drop is 45 x 0.947 = 42.6 F, so the plane sits at about 27.4 F. Still 14 degrees below the 41 F indoor dew point, sustained for the month. That is the finding: any indoor air reaching that plane deposits water there all winter.

Lever one, lower the dew point. To keep the plane dry at that monthly mean, indoor dew point must sit at or below 27.4 F. At 70 F indoor that is about 20 percent relative humidity, which is drier than most households run in winter and drier than comfort guidance recommends. The lever exists but it cannot reach on its own.

Lever two, raise the plane. Adding continuous insulation outboard of the sheathing raises that plane without changing the R inboard of it. Solve for it: the allowed drop is 70 minus 41, so 29 F out of the 45 F available, a fraction of 0.644. Total R must be at least 14.13 divided by 0.644, about 21.9. Subtract the existing 14.92 and the continuous insulation needs to be about R-7.

Check that against R-7.5, a value that shows up in the residential code's continuous insulation table for exactly this purpose. Total R becomes 22.42, the fraction inboard becomes 14.13 / 22.42 = 0.630, the drop becomes 45 x 0.630 = 28.4 F, and the plane sits at about 41.6 F. Just above the 41 F dew point.

That the derivation lands within half a point of the code's tabulated value for a 2x4 wall in a cold zone is a useful check on the method rather than a coincidence: the table was built from this calculation.

Lever three, block the air. Air sealing the interior plane does not move a single temperature in that table. It changes how much water arrives at a plane that is still below dew point. In a real repair it usually accompanies lever two rather than replacing it, and it is what keeps the assembly's wetting rate under its drying rate during the cold snap that the monthly mean averaged away.

The conclusion the arithmetic forces. In this climate you do not keep the sheathing above the dew point on the design day. Nothing affordable does. You keep it above dew point at the monthly mean, keep the wetting rate low with air sealing, and leave the assembly a drying direction. Anyone who sells "no condensation ever" in a cold climate is selling something the physics does not offer.

Checking the prediction against the building

Contact temperature readings at the planes you can reach. Interior surface over a bay and over a stud should differ, and the direction confirms your framing math. Use contact rather than infrared on anything damp or reflective.

A smoke source or a blower door for the air path in step 5. If you assumed an airtight interior plane and smoke says otherwise, the whole time constant of your prediction changes.

Moisture content in the sheathing at the end of winter, taken through a small inspection port from the interior, with the port made in a location you can close cleanly. Wood at or above the high 20s in percent moisture content has been sustained wet and supports decay; wood in the low teens has been drying. That single reading, taken in late winter rather than in September, is what confirms or kills the prediction.

Repeat the prediction with the reduced cavity R-value if the port shows compressed or voided insulation. If the conclusion held both ways, you can act on it. If it flipped, you now know the repair has to include the cavity.

References

• ASHRAE Handbook, Fundamentals, for material R-values, surface air film resistances, and psychrometric relationships
• International Residential Code, Section R702.7, for vapor retarder classes and the continuous insulation values that permit each by climate zone
• 40 CFR Part 745 Subpart E, EPA Renovation, Repair and Painting Rule, for lead-safe work practices in pre-1978 target housing
• 29 CFR 1910.333(b)(2) and NFPA 70E-2021, 120.5, for de-energizing and proving dead before drilling into a wall
• See related: What Dew Point Predicts; How Moisture Moves Through a Building; Moisture Control in Residential Construction