New Light in Crawlspaces by Joseph Lstiburek

Insight
By Joseph Lstiburek, Ph.D., P.Eng., ASHRAE Fellow

Crawlspaces stink, they rot, and are just plain icky. Photograph 1 shows the modern crawlspace, which is a forest of water droplets on the underside of fiberglass batt insulation. The exposed wood floor joists are rotting. The house over this crawlspace is not shabby (Photograph 2). Did I mention that this is a vented crawlspace (Photograph 3)? Oh, by the way, there is a continuous plastic sheet ground cover and excellent drainage. Everything in this crawlspace was done “right.” It has code specified ventilation, a continuous impermeable ground cover, excellent drainage and still we have a mess. What caused this mess? The floor insulation. No way. Yup.

Photograph 1: Icky Crawlspace—Note the condensation on the underside of the fiberglass insulation and the rot at the exposed portions of the crawlspace floor joists.

Photograph 2: Nice House Over Icky Crawlspace—This house with the crawlspace problem is in Washington, DC

Photograph 3: Ventilation Opening—Crawlspace has plenty of cross ventilation and good drainage.


Think of the good old days—the Civil War, WWI, the Great Depression, WWII—crawlspaces were uninsulated. They were ventilated and they didn’t have ground covers—and they didn’t have problems. Why? The floor framing was pretty much always warmer than the ground (Figure 1). Everywhere, even in air-conditioned buildings. This was a pretty big deal as the temperature of the floor framing was above the dew point temperature of the exterior air that was used for “ventilation.” And the old floor finishes tended to breathe—they were relatively vapor open. No one had heard of vinyl flooring yet.

Figure 1: Old Crawlspaces—Old uninsulated and ventilated crawlspaces had warm floor assembly surfaces due to heat flow downward from occupied space above.

The ground in crawlspaces is cold,1 much colder than the outside air during the summer months. In an irony not appreciated except by building science geeks ventilation air in the summer months in most parts of North America brings moisture into crawlspaces and deposits this moisture on surfaces that are below the ventilation air dew point. In the good old days this was the ground—or the ground cover which of course is at the same temperature as the ground. It was not typically the floor framing. And who cared if the ground or the ground cover was wet as long as the wood framing was not.

We don’t get much ventilation air change in crawlspaces—the typical ventilation air change rate in a crawlspace is approximately 1 air change per hour (ach).2 In determining crawlspace surface temperatures we can pretty much ignore the ventilation air change.3 We can’t ignore the ventilation air in the moisture balance but we can in the energy balance. Crawlspace surface temperatures are dominated by radiation and conduction, not by convection (Figure 2). And, as pointed out, old crawlspace floor framing was not only warmer than the ground but also warmer than the ventilation air dew point.

Figure 2: Old Crawlspace Temperatures—Surface temperatures are dominated by radiation and conduction, not by convection. Old crawlspace floor framing was not only warmer than the ground but also warmer than the ventilation air dew point.

That all changes when we install insulation in crawlspace floor framing. The most common insulation installed in this location are fiberglass batts. When fiberglass batts are installed between floor joists the exposed bottom edges of the floor joists become much colder (Figure 3). The surface temperature of the underside of the fiberglass batt insulation is also much colder than the floor sheathing—within one or two degrees of the ground temperature. The energy picture within the crawlspace is radiation dominated—the floor assembly surfaces are in essence radiation coupled to the ground. The exposed surface of the fiberglass batt insulation is below the dewpoint temperature of the air in the crawlspace as is the exposed portion of the wood floor joists and condensation forms on both the surface of the insulation and the surface of the exposed wood (go to Photograph 1 again).

Figure 3: Insulated Crawlspace Temperatures—The surface temperature of the underside of the fiberglass batt insulation is colder than the floor sheathing—within one or two degrees of the ground temperature. The exposed surface of the fiberglass batt insulation is below the dewpoint temperature of the air in the crawlspace as is the exposed portion of the wood floor joists and condensation forms on both the surface of the insulation and the surface of the exposed wood.


Now let’s look at the wood floor joist moisture content. Wood is hygroscopic—it “sees” relative humidity not vapor pressure. So we are going to need a psychrometric chart and a sorption curve for wood. The temperature of the wood floor joist is 75 degrees F at the top (i.e. interior temperature). We have to figure out the crawlspace conditions. The air in the crawlspace comes from the outside—who knew? Let’s pretend that this particular crawlspace is in Washington, DC. The average dew point of the exterior air during the summer months in Washington, DC is 65 degrees. Let’s bring this air into the crawlspace—so therefore the dew point of the air in the crawlspace is 65 degrees. Recall that the top of the floor joist is 75 degrees. The floor joist sees the dew point of the air in the crawlspace (we can ignore the vapor permeance characteristics of the fiberglass batt insulation since it is so vapor open—just pretend that we have air rather than insulation here—but not just any air—air with a huge temperature drop— “insulating air”), but because the floor joist is 75 degrees at this location, the relative humidity at this location is 70 percent yielding a wood equilibrium moisture content of 13 percent. The floor joist is “dry” at the top and “wet” at the bottom. Why wet at the bottom? The surface of the wood is cold, below the dew point of the air in the crawlspace and therefore condensation forms on the wood. At fiber saturation the moisture content of wood is 28 percent (go to Figure 3 again).

The wood floor joist moisture content increases in the downward direction as the wood becomes progressively colder (Figure 4). Another way of saying this—the warmer the wood, the drier the wood. Duh. If we were to wrap the floor joists completely with insulation we would warm up the wood thereby lowering the relative humidity the wood “sees” thereby lowering the moisture content. This is a pretty neat strategy that I will refer to later. So how low do we need to lower the moisture content of the wood? Below 19 percent to never see rot. Below 16 percent to never see mold. We could do this a couple of ways. One way is with spray foam insulation (Figure 5).

Figure 4: Moisture Dynamics—The wood moisture content increases as we move down the floor joist. At the same time the vapor drive is upward. The entire shaded area “sees” the same vapor pressure (dewpoint) due to the vapor openness of the fiberglass insulation. The entire vapor resistance is at the floor sheathing and flooring. If only the floor could breathe…it used to in the old days when we had wood floors and not vinyl…