Water in air is invisible right up until the moment it is not. The instant a surface gets cold enough, that same air lets go of some of its moisture and puts it on the surface. That temperature has a name. It is called the dew point, and it is the single most useful number in building science that most people never hear about.
This note is about what dew point actually is, why it explains most of the moisture failures we get called to investigate, and how to think about it without pulling out a psychrometric chart.
The idea in one paragraph
Warm air can hold more water than cold air. When warm humid air touches a cold surface, the surface cools the layer of air right next to it. If that layer cools past its dew point, water condenses out onto the surface. This is why a cold glass of water sweats on a July afternoon. The air near the glass cooled below its dew point, and the water vapor in that air had nowhere to go except onto the glass.
What sets the dew point
The dew point of an air mass depends on two things: how warm the air is, and how much water it is carrying. Warmer, wetter air has a higher dew point. Cooler, drier air has a lower dew point. On a mild spring morning in Oklahoma with 75-degree outdoor air at 90% relative humidity, the outdoor dew point is around 72 degrees. That means anything outdoors colder than 72 (metal handrails, shaded glass, the underside of a canopy) will bead water. Indoors, the same rule applies.
A concrete example
| Indoor condition | Dew point | Surfaces at risk |
|---|---|---|
| 72°F, 40% relative humidity | 46°F | Nothing normal in the building |
| 72°F, 55% relative humidity | 55°F | Cold water pipes, chilled water lines |
| 72°F, 65% relative humidity | 60°F | Uninsulated ductwork, cold exterior walls in winter |
| 72°F, 75% relative humidity | 64°F | Most exterior walls, some window frames |
Where this shows up in real buildings
Sweating pipes. Chilled water piping runs around 45 degrees. If it is not insulated and the space it runs through is humid, the pipe surface will drip. This looks like a plumbing leak but is actually a physics problem. The fix is insulation with a proper vapor barrier, not tightening fittings.
Ductwork ceiling stains. Cold supply ductwork running through a hot, humid attic or plenum will sweat on the outside. Water drips down onto the drywall, causing ceiling stains that look like a roof leak. Again, the fix is duct insulation and vapor barrier, not roof repair.
Cold exterior walls in winter. On a cold winter day, the interior surface of a poorly insulated wall can drop into the fifties. If the indoor humidity is high, moisture condenses inside the wall cavity, invisible until mold surfaces on the drywall six months later. This is why humidity control matters in winter, not just summer.
Window frames and glazing edges. Aluminum window frames without a thermal break get very cold. Condensation forms on the frame, runs down onto the sill, and rots the wood or drywall below. This is a common finding on 1970s and 1980s buildings.
The inside face of exterior sheathing in a mixed climate. This is the wall-assembly version of the problem, and it is why WUFI-style hygrothermal analysis exists. Building envelopes fail from the inside out when the wrong materials are stacked in the wrong order for the climate.
What actually fixes it
Every condensation problem has three levers, and any real fix uses at least two.
Warm the surface up. Insulation is the most direct fix. An uninsulated cold-water pipe sweats at 55 percent RH. An insulated one does not sweat until the room is above 90 percent, which means practically never in a conditioned building. Same principle for ductwork, walls, and window frames.
Add a vapor barrier where the vapor is trying to go. The vapor barrier goes on the warm side of the insulation, which is the side facing the humid air. Get this wrong and the insulation itself becomes a sponge. The rules differ by climate zone. In Oklahoma we are mixed-humid, which is one of the harder climates to get right.
Lower the indoor humidity. If you cannot warm every surface, you can dry the air. This is what the latent load note is about. Get the space below the surfaces' dew points and condensation stops. This is often the fastest fix in existing buildings where opening up walls to add insulation is not practical.
What to ask your engineer
- What is the design indoor humidity, and what dew point does that imply?
- Which exposed surfaces in the building are colder than that dew point at design conditions? Piping, ductwork, walls, windows?
- Where is the vapor barrier in each exterior assembly, and is it on the right side for our climate zone?
- Has any hygrothermal analysis been done on the exterior walls, or is the assembly a code-minimum guess?
- If we have had prior moisture or mold complaints, has anyone actually measured surface temperatures during the complaint season?
Most condensation investigations start with somebody insisting there must be a plumbing leak or a roof leak. In our experience, the leak is almost always the air itself.
The one-line version
Condensation is a temperature problem, not a water problem. Warm the surface, control the humidity, and place the vapor barrier on the right side of the wall.