New roof with rigid foam under 3/4 inch plywood sheathing?
I am building a shop with an apartment in my back yard in Dallas, TX and need some advice on the roofing. I want a completely sealed envelope, and the plans call for 2×8 rafters. I want to avoid baffle vents to maximize the R-30 space in the 2×8 rafters and install Owens Corning R30C paperfaced fiberglass insulation in the rafter cavities.
My plan is to put 3/4 FOAMULAR 250 rigid XPS directly on the rafters and then put the 3/4 sheathing over it. My thought is that this would save some money over the alternative, which would be to put the 5/8 osb deck down first, followed by the rigid foam, followed by another layer of deck.
I feel like I’m missing something because putting the rigid foam directly on the rafters just feels wrong. Please enlighten me.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
You are missing the effect on structural strength of the roof. But maybe there are alternatives to make up for it.
Avoid XPS (it's bad for the environment).
It could be designed so the first layer of sheathing applied directly to the rafters is 3/8" thick OSB or PLY. The 3/8" would provide the structural diaphragm needed for the building. Saves on the material cost.
Hi Hellecaster.
I have yet to see a project with a roof assembly designed like this. Roof sheathing, as you know, is a significant structural component. If you are really interested in going down this road, you should consult an engineer (buying the second layer of roof sheathing will likely cost you less, though).
Agreed
The more tried & true method would be (as you described) to put the structural roof deck directly on the rafters, and install the foam board above, with a 5/8" nailer deck above the foam through-screwed to the structural roof deck.
Stuffing an R30C into a 2x8 rafter bay delivers about R27:
https://s3.amazonaws.com/greenbuildingadvisor.s3.tauntoncloud.com/app/uploads/2018/08/08062722/Compressing%20fiberglass_2-700x310.jpg
With as little as 2" of exterior 2lb roofing polyiso (R11- R11.5) you'd meet the IRC code minimum R38, but beat the performance of R38 between rafters, due to the R11-ish thermal break over the rafters.
If only going for R30 (local code), installing 3/4" of foil faced polyiso (R5) would be sufficient to beat R30 and is protective of the unvented roof deck. Some roofing materials can be long-nailed in place through 3/4" foam.
Better yet, R30 rock wool performs at R30 in a 2x8 rafter bay, and with 3/4" polyiso would pretty much meet IRC 2018 minimums on a U-factor basis, despite being not quite R38 at center cavity.
Alternatively installing 1" rafter edge strips on the interior side to get to the full 8.25" R30 loft of an R30C and painting the ceiling gypsum with half-perm "vapor barrier latex" primer or a sheet of 2-mil nylon sheet behind the ceiling gypsum installed with longer fasterners would also not be too risky in a zone 3 climate. See: Table 3:
https://www.buildingscience.com/sites/default/files/migrate/pdf/BA-1001_Moisture_Safe_Unvented_Roofs.pdf
(R30 rock wool + v.b.latex or 2-mil nylon would meet R30 at low risk without the edge strips.)
My framer has suggested I substitute the 2x8 for 2x10 rafters and then put in 1" site built baffles and ridge vents and just install OC R30C. The price difference between the 2x8 and 2x10 was marginal. Does this sound right? Can I still add foam to the roof deck? Would I still use Vapor Barrier latex primer on the gypsum?
[Can I still add foam to the roof deck?]
- Nope. The open air channel between the baffles and the roof deck would negate the purpose of the insulation above it.
You can add foam on the interior side.
What you are proposing is a homemade version of Zip-R. Huber doesn't allow it to be used on roofs I suspect because of the point-loads from the rafters would crush the foam - and the same logic applies to doing it with separate layers.
From a practical perspective, I sure wouldn't want to be the framer trying to install sheets of foam before sheathing. it would be both dangerous and time consuming.
.
I agree. That is the part that sounded 'off' to me.