Can Spray Foam Rot Your Roof?
In the Q&A forum, Nick from Louisiana asked why these icicles appeared after he spray foamed the underside of his roof.
Ice dams are a familiar problem in New England and other parts of the country where winters are long and cold. Snow on under-insulated and under-ventilated roofs melts, pools and refreezes to form a dam. Water backs up under the shingles and much to the horror of homeowners often finds its way inside the building.
Spray foam polyurethane insulation is supposed to be a hedge against that problem. By forming an effective seal around rafters, and offering respectable R-values, foam should be blocking the migration of cold air into the roof where it can condense into water.
But a post by a New Orleans resident in our Q&A section showed just the opposite seemed to be happening.
Nick had hired a contractor to install open-cell foam on the bottom of his roof deck, converting the attic from a vented to an unvented space. When temperatures dropped below freezing, Nick noticed icicles forming at the soffit near an old, and by then unused, vent for a bathroom fan.
The roof itself was virtually new, and neither the insulation contractor nor the roofer could explain what was going on.
GREEN PRODUCT GUIDE
"This leads me to believe that it has to do with the cold weather on the asphalt roof and somehow the warm air in the attic is going through the insulation in that area to cause the problem," Nick writes. "The question is what can I do to stop the condensation?"
Open-cell foam is not a vapor barrier
Closed-cell foam, with higher densities and a higher R-value, can be an effective vapor retarder.
But not open-cell foam, It has a perm rating as high as 35 per inch, according to a technical bulletin from Fomo Products Inc.
And this, Robert Riversong points out, could be the source of the problem:
"The open-cell foam allows water vapor to diffuse through to the cold-at-night roof sheathingMaterial, usually plywood or oriented strand board (OSB), but sometimes wooden boards, installed on the exterior of wall studs, rafters, or roof trusses; siding or roofing installed on the sheathing—sometimes over strapping to create a rainscreen. where it condenses and runs down, possibly saturating the foam which does hold water," Riversong writes. "Open cell foam needs to be sealed with a vapor retardant paint. If the installer didn't inform you of that, he wasn't doing his job."
Riversong thinks the foam has the potential to absorb significant quantities of water, allowing it to pass through to a cold, condensing surface--the bottom of the roof deck. He argues the foam is a hygrophobic (water repelling) material that morphs into a hydrophillic (water loving) matrix as it cures.
When subjected to a standard ASTMAmerican Society for Testing and Materials. Not-for-profit international standards organization that provides a forum for the development and publication of voluntary technical standards for materials, products, systems, and services. Originally the American Society for Testing and Materials. test, he adds, an Icynene Inc. foam sample sucked up 34% of its volume in water after 96 hours of submersion. Installed at a density of 0.5 lb. per cubic foot, it has the potential to take in 2 1/2 gal. of water per cubic foot.
"I've heard of Icynene problems under leaking roofs, in which the foam soaked up water like a sponge," Riversong says.
Whoa, writes GBA advisor Michael Chandler. "Open cell foam is hardly hydrophillic. Leave a chunk of it floating in a bucket of water for a couple of days and see how much water it takes on. Hardly any, I've done it."
While Chandler thinks the ASTM test here is inappropriate, Riversong is undeterred.
"There are many examples of open-cell foam in walls and roofs that has become so wet from vapor diffusionMovement of water vapor through a material; water vapor can diffuse through even solid materials if the permeability is high enough. alone that it could be wrung out like a sponge," he writes.
"An anecdotal report on a blog told of Icynene in a basement ceiling under a bath that developed a leak. When the foam was pulled, the subfloor was black with mold and there were mushrooms growing in the foam."
The whole roof is not affected
If vapor permeance was the problem, wouldn't we expect to see evidence of water over the entire roof deck, icicles forming all along the eaves rather than in an isolated spot?
Rich Bev suggestions Nick look for a crack in the condensation line from air conditioning equipment installed in the attic, which may be contributing to a water and humidity problem.
"It may be just a direct leak from the A/C to a poorly installed spot in the insulation," he says.
A not-so-hot job of installing the insulation also could be playing its part.
Chandler suggests that some installers plug up rafter bays at exterior walls with plastic film so the foam doesn't leak out the soffits as its applied. But that keeps the foam from sealing wiring and plumbing penetrations in the top plateIn wood-frame construction, the framing member that forms the top of a wall. In advanced framing, a single top plate is often used in place of the more typical double top plate., he adds, and allows humid air to collect in soffits where it can condense.
Taking short cuts as the foam is sprayed can result in voids in the foam between the rafters and the roof deck--another channel for warm air. He suggests a look inside the soffit to make sure the foam has completely sealed the top plate in the wall, and to check that foam has completely filled the rafter bays.
"So much of this stuff is about caring enough about the building science to get the details right the first time," Chandler ways. "Spray foam is 'hot' and there are a lot of get-rich-quick types going into green building with little care for quality or integrity."
GBA senior editor Martin Holladay is thinking along the same lines: "Is there any chance that the spray foam has shrunk away from the rafters, leaving cracks that permit warm humid air to contact the cold roof sheathing?" he asks. "Perhaps the moisture transport mechanism isn't diffusion--perhaps it's air movement."
Our expert weighs in
We invited Peter Yost, the Director of Residential Services for BuildingGreen, to provide his expert opinion on what might be causing Nick's icicle.
Peter Yost's advice:
To identify the source of the moisture, the first step is to determine if the dampness problem is local or general. If the problem is vapor diffusion, the moisture will extend to all (or most) of the roof sheathing. If the moisture is due to a roof leak or an air leak, the dampness will be localized.
A clue: The exhaust fan looks awful suspicious
Nine times out of ten, moisture problems are caused by bulk water or air leaks. I bet this is one of those nine times. If the moisture occurs in just this one spot — a spot that has a soffit termination for a bath exhaust fan and a condensate line — it’s hard to imagine that this is just coincidence. Even though this exhaust vent was abandoned for a through-the-roof route, are there enough obstacles in this section of the attic at the eave to make the spray foam application in this section less than complete?
Pete's bet: An incomplete foam job
My first bet would be that the spray-foam installer had a hard time getting a proper airtight application of the foam in this area of the attic. My second bet would be he or she saved this tough section for last; contrary to popular belief, spray foam installations are quite quality-dependent (and even more so with open-cell foams because their expansion is more than three times greater than closed-cell).
The pictures of the opened soffit are too close up to really tell us much — other than stuff is wet and probably moldy. Does the opened soffit suggest that spray foam installation in this section of the attic at the eave was deficient?
- OSWALDO HERNANDEZ
Tue, 01/19/2010 - 20:08
Wed, 01/20/2010 - 07:33
Wed, 01/20/2010 - 09:51
Wed, 01/20/2010 - 11:26
Wed, 01/20/2010 - 11:37
Thu, 01/21/2010 - 14:51
Thu, 01/21/2010 - 21:04
Sat, 01/23/2010 - 08:33
Sun, 01/24/2010 - 18:13
Sun, 01/24/2010 - 20:39
Tue, 01/26/2010 - 07:06
Tue, 01/26/2010 - 07:32
Tue, 01/26/2010 - 08:43
Tue, 01/26/2010 - 09:07
Tue, 01/26/2010 - 09:08
Tue, 01/26/2010 - 09:34
Tue, 01/26/2010 - 09:57
Tue, 01/26/2010 - 10:12
Tue, 01/26/2010 - 10:34
Tue, 01/26/2010 - 10:57
Tue, 01/26/2010 - 12:18
Tue, 01/26/2010 - 12:21
Tue, 01/26/2010 - 12:29
Tue, 01/26/2010 - 12:56
Tue, 01/26/2010 - 13:24
Tue, 01/26/2010 - 13:26
Tue, 01/26/2010 - 13:27
Tue, 01/26/2010 - 13:30
Tue, 01/26/2010 - 13:34
Tue, 01/26/2010 - 13:39
Tue, 01/26/2010 - 13:44
Tue, 01/26/2010 - 13:48
Tue, 01/26/2010 - 13:50
Tue, 01/26/2010 - 14:15
Tue, 01/26/2010 - 14:46
Tue, 01/26/2010 - 14:50
Tue, 01/26/2010 - 14:59
Tue, 01/26/2010 - 15:05
Tue, 01/26/2010 - 15:15
Tue, 01/26/2010 - 20:12
Tue, 01/26/2010 - 20:31
Tue, 01/26/2010 - 21:39
Tue, 01/26/2010 - 23:24
Wed, 01/27/2010 - 08:08
Wed, 01/27/2010 - 08:26
Wed, 01/27/2010 - 16:19
Thu, 01/28/2010 - 08:35
Fri, 01/29/2010 - 00:35
Fri, 01/29/2010 - 01:17
Sat, 01/30/2010 - 21:06