Replies

1. 
   kevin_in_denver | Aug 11, 2013 02:37pm | #1

   If you're going to all this trouble, you better get it right. That's probably not enough insulation. Venting the roof plane in this assembly is necessary, but even that isn't the complete solution.

   The valley makes it hard to vent. You need some foam on top of the roof plane.

   100% Ice & water shield is worth the extra money.

   Here's the article that explains the best way: https://www.greenbuildingadvisor.com/blogs/dept/musings/how-build-insulated-cathedral-ceiling

2. 
   davidmeiland | Aug 11, 2013 02:55pm | #2

   I agree with Kevin that foam on top of the roof deck is a good idea, but be prepared to modify the fascia and rake boards for the extra thickness. That can mean removing and re-installing the gutters, adding trim, exterior painting, and other stuff. You might investigate whether installing closed cell spray foam instead of blown fiberglass with get you the R-value you need without the other added work, for a comparable cost. I would definitely not rely on a fiberglass-only solution here--not enough R value.

3. GBA Editor
   Martin Holladay | Aug 12, 2013 09:25am | #3

   Dave,
   First of all, I'm not a fan of the practice of installing Ice & Water Shield over the entire roof, because it makes drying to the exterior impossible, and because it is basically a band-aid for a bad roof design (especially if it is an attempt to address ice damming).

   I strongly suggest that you read my article on ice dams: Prevent Ice Dams With Air Sealing and Insulation.

   There are 4 suggested remedies for ice dams. Listed in order of importance, they are (1) air sealing, (2) insulation improvements, (3) venting, and (4) Ice & Water Shield. The first two remedies are by far the most important. The last two don't really address the problem, although they can sometimes be part of the solution. If someone decides to implement (3) and (4) without first implementing (1) and (2), they would be making a big mistake.

   It is a code violation to install an air-permeable insulation like fiberglass in a cathedral ceiling unless you have a vent channel above the top of the insulation layer. This vent channel will be counterproductive unless you also include an air barrier (for example, sealed plywood) above the insulation layer.

   Whatever type of insulation you install, you need to be sure that the insulation layer AT LEAST meets the minimum code requirement for your climate zone. In your zone (zone 6), you need at least R-49 to reduce heat flow through your roof assembly. That is the code minimum.

4. Expert Member
   Dana Dorsett | Aug 12, 2013 06:18pm | #4

   Unless there's vented roofing such as standing seam on purlins, drying toward the exterior is impossible anyway, whether it has the membrane or not. Composite shingles on #30 felt has a permeance in the 0.1 range- no drying to speak of happening there! Peel & stick membranes can be very effective at air-sealing plank roofs, and shouldn't be dismissed as a band-aid in every case, even if it kinda is a band-aid in THIS case (but it could be a necessary band aid.)

   But Optima in an unvented roof (even dense-packed) would be inviting roof-rot, but would be just fine with as little as 1" of closed cell foam applied to the underside of the roof deck (which would air-seal at least as well as the Ice & water shield), yet still have about 1-perm of drying capacity through the ~6" of Optima. Using that little foam would be a code violation, but it wouldn't cause an issue if you have a standard composite shingle type roof as opposed to light-metal or "cool roof" shingles. And by not making it super-vapor-tight on the interior you'd still be able to add the R25 (4" rigid polyiso) on the exterior to make it code legal without creating a moisture trap at the roof deck. For a sanity-check on the flash-foam + fiber approach, see: http://www.buildingscience.com/documents/reports/rr-1001-moisture-safe-unvented-wood-roof-systems

   In particular, note Table 3, the column just to the right of center headed "1" ccSPF + spray fiberglass", and the "Dk asphalt" row of the 6A Minneapolis rows. That simulation approximates your condition. But that was for an R49 total, whereas with 2x8 rafters you'd have a more favorable R-ratio between foam/fiber at ~R6 foam/R26 fiber, with fewer condensing hours at the foam/fiber interface, and a modestly warmer roof deck, so the roof deck is actually safer than with R43 fiber, R6 foam in the simulation.

   But it doesn't really fix the ice damming, not completely- an R30-ish level will reduce the ice-damming potential, but there is still meltwater happening along the rafter striping, and with a deep snowpack and optimally bad freeze/thaw weather events you can still end up with at least SOME ice damming.

   Stripped to the roof deck it's possible to then air-seal the plywood seams, apply the appropriate slip-surface for roofing iso, and add the code-prescribed R24 as prescribed by code. Then with 7-1/4 inches of dense packed Optima alone in the cavities you'd be in great shape from a total-R point of view, at a bit over R50, there would be an R25 thermal break over the ~R8.5 rafters for FAR less melt-striping, and with all but the most exceptional snowfall followed by freeze/thaw cycling the ice damming would be mostly a not-so-fond memory.

   http://publicecodes.cyberregs.com/icod/irc/2012/icod_irc_2012_8_sec006.htm

   If there's any way to get to code within budget, that's the way to go, even if it means scrapping the Optima idea for crummy low density R25 batts in the rafter bays in order to pay for the foam-over.