Good idea for R-38 under a 2×4 shake roof?
I want to expand a top-floor bedroom to include an uninsulated 16’x9’ corner of my 1920s craftsman (zone 5). Its southwest exposure, 7/12 roof is cedar shingles on skip sheathing, across 2×4 rafters which bird-mouth onto the exterior wall exposing 2” of the backside of the frieze boards; exposed rafters outside, no vents.
If I fur out the rafters with 2x6s, then filled between them (top down) with a 1” gap, 2″ double foil sided polyiso board R-12.9, 5½” fiberglass batting R-21, then nailed to underside of 2×6’s with 1” XPS R-10, then ½” sheetrock. Then add vents to the frieze boards. Can you think of issues created by this plan i.e. condensation, ice dams…?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
James,
There are a couple of issues here. The main one is that you are proposing a foam sandwich, with rigid foam on both sides of your fiberglass batts. That's not ideal.
In order to retain the possibility that any trapped moisture can dry (albeit slowly) to the exterior, it would be better to choose a different material for the ventilation baffle. Instead of using 2-inch-thick foil-faced polyiso as your ventilation baffle, you might want to use thin plywood or 1-inch-thick EPS. That will allow some drying to the exterior. Needless to say, you should pay attention to airtightness when installing your baffles.
By the way, 1-inch-thick XPS does not provide R-10; it provides R-5.
If you want to use 2-inch-thick polyiso, I suggest that you install it on the interior side of your fiberglass batts instead of the proposed XPS.
Oh ya, typo’d the R5.
Thanks for the response. I’ve been finding it hard to plan a retrofit of manageable combos of the available materials that covers all the aspects (R38 min., vapor/air/radiance barriers, bridging, venting, +++) w/o being 2’ thick, re-roofing or $$$$ . I’m open to ideas.
I like the plywood idea but isn't it a vapor barrier also? What about a wrap like "Tyvek"(weather barrier: allows air & vapor to pass but sheds water & inhibits wind) replacing the plywood on top? So then (top down), air gap, tyvek, fiberglass batting, foiled ISO, air gap, XPS, drywall?
James,
Plywood is not a vapor barrier. It is a "smart" vapor retarder with variable permeance. As the plywood begins to get damp, its vapor permeance increases.
I think that you will find that it is much easier to create an airtight seal with plywood than with strips of Tyvek. The steps required to install many thin strips of Tyvek in an airtight manner in each rafter bay are daunting. You can try it if you want, though.
Thanks Martin. Agreed, Tyvec would be a major install head-ache. Only issue with the plywood is it makes the materials stack pretty tight and I might lose some of the R value of the Batt's due to compression. Otherwise, this makes an R38, vapor, air, and thermal envelope with a radiant barrier about 9" thick. Aside from money and time, do you see any other issues this scenario might be hiding? Also, would it matter which grade plywood and which sealant/adhesive I use?
James,
Either 1/2" plywood scraps or 1/4" lauan would work. So would stiff cardboard or thin polystyrene. You can also use a commercial product like AccuVent if you want.
If these details are new to you, you might want to read this article: How to Build an Insulated Cathedral Ceiling.
Scraps are good, I like scraps. I've read that article many, many times, weighing my options. If it wasn't for the shingles or if they were overlain with 3-tab I could eliminate the ventilation space create a "hot roof" in theory, right. But that doesn't seem like a good idea to me; 3-tab could leak and the shingles need to breath I think. That's why I don't see spray foam on the underside of skip sheathed shingles right?
In zone 5 with only 2" of polyiso on the exterior you'll have wintertime frost in the R21 fiberglass layer. It won't damage the fiberglass, but there will be a (temporary) degradation of performance during the coldest weather. If instead you made that 2-3" of EPS, the fiberglass would dry sufficiently quickly into the vented air gap. Martin's suggestion about flipping the stackup with the foil faced goods on the interior side works too.
As a general rule, unless you need the lower vapor retardency of XPS in your stackup (rare, but it happens), it's both greener & cheaper to go with EPS due to the much lower impact blowing agent used. It's the same polymer, and after 50 years or so after the blowing agent has leaked out of the XPS, the R/inch is going to be about the same. In Europe where XPS is blown with CO2 as opposed to an HFC soup (the biggest component of which is HFC134a, at about 1400x the global warming potential of CO2), 1.5lbs density XPS is about R4.2 @ 1", just like 1.5lbs density EPS.
James,
As long as your roof has skip sheathing and cedar shingles, you have no need for a vent channel. This type of roof dries rapidly to the exterior.
What your rafter bays lack, however, is a top-side air barrier.
If you knew that your roof would always have skip sheathing and cedar shingles, you could forget about the ventilation channels. But you never know when someone will decide to install plywood and asphalt shingles.
If you want to create an unvented roof, you can. I would recommend that you first install a layer of cardboard in each rafter bay, pushed up against the skip sheathing. Then a spray foam contractor could install spray foam against the cardboard.
Thanks Dana, I didn't know about the XPS mfg process. My thought of ISO on top was to combined radiant and water barrier for when the roof did eventually leak. As I understand it now, even with a plywood or EPS top crust and ISO bottom crust the sandwich would still release vapor out the top. In that case, a stack could be air, EPS, Batt, ISO//foil, air, XPS, Sheetrock and get good drying outward.
Thanks Martin,
Honestly I thought the shingles had to have the 1" air space and read where some advice a 2" space. Since I lost my crystal ball I can't tell what will happen to the roofing. I figure there's about 15 years life left in the roof and by that time...; with this plan I'm wanting to keep as many future options open as possible.