Sometimes we do things with one goal in mind and find that we get the opposite, unintended result. Such experiences have been all too frequent in my consulting practice in a coastal environment. In the process of trying to do a better job of controlling indoor residential humidity, we actually can make the situation worse. In fact, many times things get so bad, occupants are forced to move out and expensive remediation is required.

ASHRAE recommends we keep our indoor humidity below 60 percent and our thermostat at or above 75˚F for indoor summer comfort. New high-end residential equipment is designed to make 75˚ more comfortable by offering a dehumidification option. We are much more comfortable with 55 percent relative humidity (rh) than 65 percent, given the same sensible temperature of say 75˚. Weathermen call this the “real feel” temperature. Plus, lower humidity prevents mold from spawning. So great — we can now control temperature and humidity, and most of the time the theory works.

New variable-speed equipment can operate in dehumidification mode by lowering the fan speed. This in turn slows airflow over the coil, allowing more moisture to be removed. Typically the system goes into dehumidification mode at the beginning and end of a cycle calling for cooling based on thermostat set points. For example, a rise in room temperature causes a call for cooling. If the rh is above the set point, the cooling process starts and ends with operation in dehumidification mode. The fan ramps up, slowly holding at a few hundred feet-per-minute velocity, condensing moisture out of the air very effectively. After a set period of time, the fan accelerates to full speed to deliver designed cfm cool air to the space to satisfy the latent load. Of course moisture continues to be removed during this process.

So far, so good. Once we have the sensible and latent loads satisfied, the unit cycles off. But, what if we have not reached the rh setting? This is when things get interesting. The unit does not cycle off. Rather it cycles back into dehumidification mode. Now we are back to slow air flow, coil temperature drops off a cliff, and so does discharge air temperature. A typical LAT/RAT delta-T can double from 15 degrees to 30 degrees. The unit continues to over-ride sensible settings trying to reach the rh setting until we get 2-3 degrees below the set point. This can take longer since we have less air mixing given the slower flow. Only when either the room is sub-cooled to the maximum allowed, or rh drops below the set point, does the unit cycle off. Meanwhile, we have sub-cooled air flowing in ductwork and supply registers, which are typically metal. This cold air discharges against building materials at the supplies, causing condensation when the unit shuts off and warmer air comes into contact with these surfaces.

The Envelope

Now we need to take a look at how the system interacts with its environment. Here we’re mostly concerned with ductwork and its location. If our duct is in conditioned space, there usually is no problem. But, typically ducts are in unconditioned attic or crawl spaces. So we potentially have R-6 or R-8 nominal insulation intended to separate air with a delta -T of 30-60 degrees difference in an uninsulated attic space. (e.g. supply air at 55˚ and attic temperature of 115˚). With the sub-cooled dehumidification mode air, we can see our delta T driven to over 70˚. As a result, our insulation begins to sweat.

This can also happen in a crawl space without a large delta-T. If the typical foil-jacket vapor barrier is torn or insulation seams have separated, condensation is frequently found starting on the duct as water vapor follows vector forces moving inside the vapor barrier opening. As this water accumulates inside the duct insulation in a crawl space, it collects on the bottom of the insulation blanket and cannot escape. As a result, we get more and more cold water accumulating inside the insulation, resulting in more thermal runaway. The bottom of the duct sweats as the insulation loses its R-value. As the water continues to accumulate in the insulation we have catastrophic loss of R-value bringing the exterior of the insulation below dew point conditions and sub-cooling the adjacent structure resulting in building materials dropping below dew point.

Continued water accumulation spreads in a building structure and subfloor, eventually reaching the finished floor and causing cupping, swelling, and eventual buckling. As this moisture moves from wet to dry, thermal and moisture runaway accelerates, and water ponds on the vapor barrier as well. We see mold growing in the crawl space on building materials but so far we have not had intrusion of spores into the living space. Wrong.

Negative Indoor Conditions

Suppose the supply ductwork leaks and we lose 5 percent or more of our supply air. Typically, return static pressures are lower and there are fewer connections so let’s assume we are returning 100 percent of system air. In our exemplar home we have over 10 tons of capacity.

For a nominal 400 cfm-per-ton, we are losing at least 200 cfm. Add bath and kitchen exhaust, as well as dryer and range hood, and the house really begins to suck. And, as the house goes negative, air in the attic space is being heated, expanded, and becomes more positively pressurized. Warm humid air begins to move into the house from the attic via all available penetrations. Examples include ceiling penetrations for fixtures and bath fans, wall switches, and penetrations for registers including floor registers. Water vapor moves from wet to dry so we also have moisture moving from the attic toward the living space. Gypsum wallboard is transparent to moisture, so where there is no vapor barrier more moisture enters the living space. We are now losing control of humidity.

Moisture and Mold Runaway

Now the situation gets ugly. Warm, moist air infiltrates penetrations. It also condenses against the insulation paper facing a vapor retarder, where it comes into contact with the cooler room temperature sheetrock. The problem is especially acute near supply registers, where temperatures drop sharply below dew point conditions. The insulation paper facing gets saturated, in turn wetting the wallboard or subfloor. The warm, moist air entering penetrations contacts indoor surfaces below the dew point and condenses. As conditions worsen, the area becomes wetter and indoor rh rises. The equipment is running in dehumidification mode as long as controls allow. Instead of lowering rh, as intended, the dehumidification mode sub-cooling makes matters worse by lowering temperatures even further below dew point conditions. Now, we are experiencing a runaway nightmare. Wet gets wetter and rh rises beyond the point of no return. Obvious evidence of water damage worsens and, of course, with it comes the inevitable mold.


Mold spores require a food source and sufficient moisture to grow. Building materials, particularly cellulose based products, provide excellent nutrients for molds to grow indoors. All that is needed now is sufficient moisture for mold spores to grow and colonize the wet building materials. Mold will begin to actively grow within 24- to 48-hours when sufficient moisture is present.

The kinds of molds that grow depend on the amount of water present and the duration of wet conditions. Some molds require relatively low moisture levels, while others require persistently moist conditions.

Condensation in crawlspaces and attics may cause mold colonization on building materials. Indoor air quality is subsequently significantly degraded because attic and crawlspace air moves into the occupied areas of a building carrying with it mold spores. Air sampling conducted when evaluating some of these cases has shown significant amplification of mold spores indoors, particularly in water-damaged buildings. Overall, levels of mold spores are detected in excess of 15 times higher indoors when compared to ambient mold spore levels outdoors. Mold spore types usually associated with water-damaged buildings, such as Penicillium/Aspergillus, were as much as 35 times higher indoors when compared to outdoors, in areas where significant mold colonization existed on the back side of drywall.

Other mold types identified in the air include Stachybotrys, which indicates that building materials have been subjected to long-term chronically wet conditions due to repeated condensation.

Surface sampling of visible mold colonization may show Aspergillus growing inside supply ductwork. Penicillum may be found colonizing joists in the crawlspace. Stachybotrys is commonly found colonizing the paper on the HVAC filter on the return side of an air handler. This would indicate high rh in the return air which condensed on the cold fan coil cabinet.

Super-cooled supply air blowing on adjacent wall board causes condensation on the back of drywall, which may lead to the colonization of Stachybotrys on the attic side of drywall.

To control wet conditions and resulting mold colonization on building materials, rh should be controlled and supply air systems should be designed so that air is not discharged on adjacent building materials causing condensation. RH should be controlled in crawlspaces by using adequately insulated ducts, the crawlspaces should be enclosed and dehumidified, and duct leakage should be controlled to prevent negative pressure conditions in the occupied spaces of the building.

We recommend attics also be sealed and insulated at the roof deck in addition to insulation on the ceiling/wall board. Best practice is to also install an attic outdoor air dehumidifier which discharges via supply ductwork, reducing or eliminating negative indoor conditions.

Publication date: 8/6/2012