Closed-cell foam versus thermal bridging
I have read several of Martin’s articles and reader responses regarding the benefits of using closed cell spray foam on the underside of unvented roof decks. Because it is vapor impermeable, you can spray the underside of the roof without fear of vapor migration to the cold sheathing…so long as it is the proper R-value….. which would then cause rot over time. That’s all good. In a cathedral ceiling or gable end wall, what happens to the rafters or studs which cannot be encased in the foam due to the need for drywall? Since they themselves are a thermal bridge aren’t they exposed to the possibility of vapor transmission and condensation…and therefore rot over time??
Anyone’s clarification on this specific item would be most helpful!
Thanks.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Jon,
Q. "In a cathedral ceiling or gable end wall, what happens to the rafters or studs which cannot be encased in the foam due to the need for drywall? Since they themselves are a thermal bridge aren't they exposed to the possibility of vapor transmission and condensation, and therefore rot over time?"
A. Softwood lumber has an R-value of R-1.25 per inch, so the exposed surfaces of the lumber don't get cold enough for condensation of significant moisture accumulation. So you can safely use the "flash-and-batt" approach in wood-framed cavities without worrying about rafter rot or stud rot, as long as the layer of closed-cell spray foam meets recommended minimum R-values. (Of course, flash-and-batt recommendations assume that the rafter bays or stud bays are fully filled with fluffy insulation, not just air. The fluffy insulation also helps keep the framing members warm, limiting moisture accumulation.)
There is a thermal penalty to the flash-and-batt approach, of course, since flash-and-batt doesn't address thermal bridging through the rafters or studs. If there is any way to install a continuous layer of rigid foam on the exterior side of the wall sheathing or roof sheathing, you'll end up with an assembly that performs better than a flash-and-batt assembly.
The R value of the closed cell foam is irrelevant in terms of protecting the roof deck from rot- it's vapor permeance is, and that vapor permeance drops with thickness. At 1 perm or lower it's a Class-II vapor retarder, which only lets moisture out in or out slowly. The R-value of the closed cell foam is only relevant to the total R in a flash & batt, and determines the amount of moisture than can accumulate in the fiber insulation layer, not the roof deck.
Closed cell foam isn't impermeable, but it's only slightly vapor permeable. Dry softwoods have roughly the same vapor permeability as closed cell foam- about 1 perm @ 1" (order of magnitude- varies by foam vendor and wood species & density.) So a rafter that isn't encapsulated doesn't admit appreciably more or less moisture to the cold side of the assembly than closed cell foam would, which is why heavy timber log structures don't automatically rot from interior moisture drives over a season.
Thanks for both answers.
Dear Jon: That's a good question -- and Martin provided a good answer to the condensation question. Wood is a surprisingly good conductor of heat / cold, however, so it would be smart to avoid the 'thermal bridges' created by dozens of exposed rafters. But you need a solid surface to mount the drywall panels. We solved this problem on a recent project by screwing short pieces of 2"x4" to the rafters -- that extend past the surface of the rafters by 2" at every location where drywall needs to be screwed down. This allows the spray foam installer to easily cover the rafters and roof sheathing, with either a full-depth fill, flash-and-batt, or a couple of inches of closed cell followed by a deeper layer of open cell.
I hope this is helpful,
Mark