Conditioned attic roof detail
I am in the process of fine tuning the details of a one and a half story house I would like to build next year. It will be in Northern New York on the border of the 5A and 6A climate zones The upstairs has a knee wall and a couple access doors to the attic space. I think I have decided to just run the zip-system up the walls and around the top of the roof rafters to simplify the air sealing. The roof rafters are engineered 12″ I joists, so I will fasten 2X4’s for an 18″ roof overhang from the wall, but I plan on running the air barrier straight up the wall to the roof by cutting and taping the wall sheathing around the protruding rafters. Then I would fasten the last 18″ of roof decking down to the rafters after the transition from the wall to roof transition is taped. This overhang will be vented. I would like to use 3″ of used craft faced polyiso board on the roof deck held down with fern strips for most of the roof. On the bottom 3′ I would like to use plywood covered with ice and water shield, over top of the foam. I would put a rib style(not standing seam) metal roof on top of everything. The metal roof would not have seal strips at the bottom and I would use a vented cap so that it would vent and release heat in the summer time. Any water that entered the bottom would then drain down the ice and water layer. The 12″ roof rafters would then be filled in with cellulose resulting in a R-56 or so roof assembly.
I hope I have been fairly clear and haven’t confused to many people. What does anybody think of this system? Do you see any foreseeable problems that I may run into. This assembly seems like it would work to me but those have been famous last words. Maybe I am coming up with an even more complicated solution to an only complicated problem.
Thanks,
Dillon
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
First, 3" of polyiso is insufficient for dew point control at the roof deck in zones 5 or 6 at code-min total R values. For an R49 total-R you need at least R20 above the roof deck for zone 5, R25 for zone 6, and more is always better. And that's at a total-R of R49. For higher than code total-R you'll need a minimum of 40% of the R to be outside the roof deck for zone 5, or 50% for zone 6.
Moreover, in your climate with the polyiso on the cold side of the assembly you'll need to derate to about R5/inch to be predictive of it's average mid-winter performance (important for dew point control aspects.) A 3" /R15 polyiso layer is woefully inadequate for your R56 cavity fill. With an ~R55-ish cavity fill you'd need 11" (a derated R55) of exterior polyiso in zone 6 (making it an R100+ roof), or 7.5" (a derated R37.5) in zone 5.
Since you're on the edge between the zones and you're using 3" stock, the safe compromise would be 9", (three layers of 3".) That makes the polyiso layer about R45 to R55 cellulose, and a 45% fraction, splitting the difference between the prescriptives for zones 5 & 6.
See: http://publicecodes.cyberregs.com/icod/irc/2012/icod_irc_2012_8_sec006.htm
http://publicecodes.cyberregs.com/icod/irc/2012/icod_irc_2012_11_sec002.htm
Thanks for the response, 9" of deck insulation seems excessive, odd looking and labor intensive defeating the whole purpose of the idea. A traditional vented attic seems to make more since. Back to the drawing board.
Thanks
Dillon,
For more information on this issue, see How to Build an Insulated Cathedral Ceiling.
If you decide that you are in Climate Zone 5, all you need is R-20 above the roof sheathing -- in other words, 5 inches of EPS or 4 inches of XPS. If you use polyiso, I think you would be fine with 4 inches of polyiso.
Dillon, before you go back to the drawing board, here's an example from Building Science Corp that has a similar roof assembly as you propose. I think Dana's right but here's an example from a reputable company, where there's more insulation value in the rafters than above the roof deck.
http://www.buildingscience.com/documents/case-studies/cs-ma-westford-hfh/?searchterm=habitat
Martin: "If you decide that you are in Climate Zone 5, all you need is R-20 above the roof sheathing."
That's true from a letter of code point of view, but that prescription is based on an R49 total center-cavity R value, not a high-R roof.
From a dew point control it really needs to be 40% of the total, as recommended by BSC in their Deep Energy Retrofit Guide for Massachusetts (all of which is in zone 5.) The relevant figures for the proposed assembly is on pages 50 & 51 of the guide, and note the 40% number in the angled text bracketing the exterior rigid insulation on both figures::
http://www.buildingscience.com/documents/guides-and-manuals/gm-mass-save-der-builder-guide
This is consistent with the IRC precriptive of R20 for zone 5 when the total R is R50 (R49, is code min for MA, per IECC 2012) but R20 would fail at a total R of R76 (R20 exterior, R56 in the cavity.) At only 26% of the total R on the exteiror the structural roof would be at significant risk in a zone 5 climate, especially the COLD edge of zone 5 (as in this case.)
Mark Fredericks' example from BSC (http://www.buildingscience.com/documents/case-studies/cs-ma-westford-hfh/?searchterm=habitat ) shows R26 on the exterior, with R40fiber in the cavity, in a Westford MA (zone 5) location. That's an R26/R66= 39% fraction of the total-R on the exterior. I'd have some issues with using the labeled R value for that R13 x 2 polyiso however, given that during the 3 months of winter the outer layer will be performing much lower than it's performance when tested at 75F in an ASTM C518 test (the standard for labeling purposes.) In Westford MA the mean outdoor temp in January is in the low 20s F, a temperature at which most polyiso is performing at a bit less than R4/inch. The inner layer is going to be a bit warmer, so the average will be closer to R20 than R26 during the coldest months, which is marginal at best. Using 3" of EPS (labeled R12.6) instead of 2" polyiso (labeled R13) for the exterior layer would be safer from a dew point control point of view, since the EPS would GAIN performance at lower outdoor temps, offsetting the derated performance of the polyiso.
I am in zone 5, but only by about 10 miles. I figured erroring on the side of caution.
If you're within 10 miles of the zone demarcation, splitting the difference at 45% of total-R is prudent, but you'd probably still get away with only 40% using the derated R5/inch.
FWIW: Most fiber-faced roofing iso is 2 or 3lbs per cubic foot, higher density than the foil-faced stuff, and tests at R5.5/inch @ 75F rather than R6-6.5/inch (the typical labeled R for foil faced goods.) Derating it to R5/inch is really the right thing to do from a design point of view. It's not clear how the higher density affects the derating curves, but my WAG is that it won't be any worse than lower density stuff. More on derating curves for polyiso:
http://www.buildingscience.com/documents/information-sheets/info-502-temperature-dependent-r-value
https://www.greenbuildingadvisor.com/blogs/dept/musings/cold-climates-r-5-foam-beats-r-6