What Dew Point Predicts

Why this matters

Relative humidity describes air at the spot and the second you measured it. Move that air to another room, wait until three in the morning, or push it into a wall cavity, and the relative humidity is a different number. Dew point is not. It travels with the air unchanged, which makes it the only humidity reading you can carry from where your instrument is to where the problem is. Techs who work in dew point can stand in a crawl space at noon and tell a customer which surfaces will be wet at dawn. Techs who work in relative humidity are guessing, and the guess is usually optimistic.

A sibling article covers the mechanism by which a cold surface makes water. This one is about using the number as a forecast.

The property that makes dew point portable

Dew point is the temperature at which the air you are holding would become saturated. It is set by how much water is actually in the air, and nothing else.

Heat that air and the dew point does not move. Cool it and the dew point does not move, right up until the air reaches the dew point and starts dropping water out. Only two things change it: adding water vapor, or removing it.

Watch what that does to relative humidity, which is what most people are actually reading. Take crawl space air at 78 F and 85 percent relative humidity. That is a dew point of about 73 F. Push the same air, unchanged, into a 90 F attic and the relative humidity reads about 58 percent. Nothing was added, nothing was removed, and the number on the meter fell by 27 points. A tech who reads 58 percent in the attic concludes the attic is doing fine. The dew point says otherwise, and the dew point is the one that predicts what happens when that air meets something cold.

That is the whole reason to convert. Relative humidity is a ratio to a moving reference. Dew point is a property of the air.

The gate

Every condensation question in a building reduces to one comparison, and it takes two readings.

Is the surface temperature, at the time in question, above or below the dew point of the air that can reach it?

Below, and you get liquid water on that surface. Above, and you do not, no matter how high the relative humidity reading was. The two cases below are the same air mass in the same building on the same afternoon, and the gate sends them opposite directions.

Case one: a surface fifteen degrees under

Same crawl space, dew point measured at 73 F. Running through it is an insulated supply trunk with a section of bare metal where the wrap was torn during a previous job. Surface temperature on the bare metal, taken with a contact probe rather than an infrared gun because the surface is damp and an infrared reading on a wet surface reports the evaporating film, comes back at 58 F.

Fifteen degrees under the dew point. That surface is wet whenever the system runs and this air is present, and it will stay wet. Water drips onto the joist below it, the joist stays damp, and in a year somebody is diagnosing a roof leak on the floor above.

Two levers close the gap, and they are not interchangeable.

Raise the surface. Repair the insulation and seal the vapor barrier at the seams. An adequately wrapped, vapor-sealed duct sits within a couple of degrees of the surrounding air, so the surface goes from 58 F to near 78 F, comfortably above 73 F. This fixes the duct. It fixes nothing else in the crawl space.

Lower the dew point. Seal and condition the crawl space so its air carries less water. This protects every cold surface down there at once, including the ones you have not found: refrigerant lines, cold water piping, the underside of a tile floor over an unheated space.

The choice is not about which is cheaper in hours. It is about how many cold surfaces exist. One damaged duct is a duct repair. A crawl space with a dozen cold surfaces and a dirt floor is a dew point problem, and repairing ducts one at a time in it is a subscription.

Case two: a surface a degree and a half over

Same air, dew point 73 F. The rim joist at the band, measured with a contact probe, reads 74.5 F.

By the gate, dry. And it is dry, right now, at two in the afternoon.

That is a margin of 1.5 F, and a margin that small is not a pass. Three things routinely move a surface by more than 1.5 F: overnight cooling of the ground and the framing that touches it, a change in the air handler's operation, and outdoor conditions that push the crawl space dew point up by a couple of degrees after a rain.

The field convention worth adopting is a 5 F margin: treat a surface as protected only when it sits at least 5 F above the dew point of the air that reaches it, and treat anything inside that band as a surface that will be wet on some nights. It is a practice convention rather than a code number, and it exists because both terms in the comparison drift.

So case two gets written up as a finding, not a pass: surface within 2 F of dew point, expected to condense under overnight conditions, recheck at the coldest part of the cycle or address the dew point.

Six things a dew point predicts before you can measure them

Where condensation will appear tonight, which is any surface that will fall below the dew point when the load drops and surfaces cool.

The lowest supply air temperature you can run before the ducts and diffusers in unconditioned space start sweating. If the space around the duct has a 73 F dew point, supply air cold enough to pull the duct exterior under 73 F will make water regardless of how well the system is performing.

Whether an evaporative process will do anything. Evaporation is driven by the difference between the water's surface vapor pressure and the air's. Air near its dew point cannot take much more water, which is why an evaporative cooler stops working in humid weather and why a floor dries in an afternoon in one season and takes three days in another.

What a dehumidifier can actually achieve. A refrigerant dehumidifier drops the space toward the dew point corresponding to its coil temperature. It cannot dry a space below that, and its capacity falls off as it approaches, so a machine that pulls a room from a 70 F dew point to 60 F may take an order of magnitude longer to get the last few degrees.

Where flue gas will condense, which is the same comparison with a different fluid. The sibling article on condensing appliances runs that case.

Which surfaces frost overnight even when the air stays above freezing. A surface with a clear view of the night sky radiates to it and can sit several degrees below the surrounding air. That is why a windshield frosts at 38 F ambient, and it is why a north-facing sheathing surface or an uninsulated attic-side duct is colder than the attic air around it.

Turning the prediction into a target number

The gate gives you a diagnosis. Reversed, it gives you a specification, and that is what you put on a proposal.

Take the crawl space. The coldest surface you cannot practically insulate is a cold water line running at about 60 F. Apply the 5 F margin and the required dew point is 55 F. Current dew point is 73 F.

The target is a dew point reduction of 18 F, and that is the number the equipment selection is made against. It is not "keep the humidity down," it is not a relative humidity setpoint that will read differently at every temperature the space passes through, and it is not a guess about a percentage. It is a measurable target that either holds or does not, checked with the same instrument on the follow-up visit.

Note what the number does not depend on: the temperature of the crawl space. If somebody heats the space, the relative humidity reading will fall and the customer will believe the problem is solved. The dew point will not have moved, the water line will still be at 60 F, and it will still sweat.

How to get a dew point you can trust

Measure it, or compute it from two co-located readings. Most field hygrometers display dew point directly. If yours does not, you need dry bulb and relative humidity taken at the same point at the same time by a probe that has equilibrated, which in a duct or a crawl space means minutes, not seconds.

Do not use the rule of thumb below about 50 percent relative humidity. The familiar approximation of subtracting one degree Celsius of dew point for every 5 percent of relative humidity below saturation is decent in humid conditions and increasingly wrong as air gets drier. At 70 F and 35 percent it overstates the dew point by several degrees, which is exactly the range where a wall cavity calculation lives. Read it off a chart or an instrument.

Identify the air mass, not just the room. A crawl space, an attic, and the living space are frequently three different air masses with three different dew points, and the useful question is which one actually reaches the surface you care about. Where a plane is fed by air that leaked from somewhere else, use the dew point of the air that leaks, not the air in the room you are standing in.

Take the surface temperature with contact, on anything damp or shiny. An infrared reading is fast and it is wrong on wet surfaces because evaporation cools the film, and wrong on bright metal because of reflectivity.

Before you go under the house to take any of this

A crawl space can be a permit-required confined space when access is restricted and the atmosphere cannot be assumed safe. The general industry standard is 29 CFR 1910.146 and the construction standard is 29 CFR 1926 Subpart AA, and a field-service shop can fall under either depending on the job. Test the atmosphere before you enter for oxygen, combustible gas, and carbon monoxide, ventilate it, and do not enter alone.

Rodent droppings are an inhalation hazard, not a housekeeping problem. Do not sweep or dry-vacuum them; wet the material down, wear a respirator with particulate cartridges rather than a dust mask, and ventilate before and during. Standing water plus any energized equipment under the house means the circuits serving that space get opened, locked, and proven dead per 29 CFR 1910.333(b)(2), with the live-dead-live sequence at NFPA 70E-2021, 120.5, before you go in.

References

• ASHRAE Handbook, Fundamentals, psychrometrics chapter, for dew point, humidity ratio, and their relationships
• 29 CFR 1910.146 (general industry) and 29 CFR 1926 Subpart AA (construction), OSHA confined space standards
• 29 CFR 1910.333(b)(2) and NFPA 70E-2021, 120.5, for de-energizing and proving dead before work in a wet space
• See related: Dew Point and Condensation, the Reasoning Behind the Fault; How to Work Out Where Condensation Will Form; What Condensing Actually Means in an Appliance