SIGA Majcoat vs VaproShield SlopeShield
I’m designing my LAST house– my retirement home, after having built several homes over my lifetime. My last was in Vermont in 2000, using ICF, in-floor radiant heat (walkout basement with concrete slab, upper floor with 1.5-inch poured gypcrete), standing seam roofing, closed-cell spray foam under roof deck.
My new home will be in a completely different climate, in Prescott Valley, AZ, having 4200 hdd and (I believe) 500 cdd. Elevation is 5000 ft. It will be a 2450 SF ranch with 4:12 pitch roof, using trusses. Floor to be slab-on-grade, well insulated. Walls to be “C-SIPS” or “SCIPS”, meaning two concrete wythes with EPS foam between and wire trusses joining the two wythes through the foam; wire mesh is embedded in each wythe in addition to rebar reinforcement (looking at GCT but there are others out there). Roof will be fully hip.
The building width is 32 feet, or 35 feet including the overhangs, in the direction the main trusses will run. So the single-side roof slope will be about 18-1/2 feet in length (along the slope).
Attached is my intended detail for the roof and eaves. The trusses stop 1/2-inch inboard of the exterior face of the inner structural mortar layer, which I plan to make 4-inches thick for thermal mass. The trusses are sheathed with 1/2-inch OSB or plywood (code requires solid blocking at joints, or T&G, for this sheathing, for thermal barrier). Cover the plywood with 100% coverage of Ice & Water Shield– this is my vapor barrier and interior air barrier, so it should run down between the truss tails and the wall EPS layer (where I show a piece of plywood to continue the truss sheathing). However it also serves as a “last-ditch” water barrier, in case the upper roof were to leak; this suggests it should run out over the wall EPS and outer concrete wythe, which is what I show in the drawing. In either case it will be sealed with caulking against the wall EPS, together with the foam above the truss sheathing.
Above the truss sheathing and IWS will be 3 layers of poly-iso rigid foam boards, each 2-inches thick, with joints staggered vertically and horizontally. I imagine these joints should be taped?
The rigid foam layers are then held down by 2×2 “super-rafters” that extend to form the eave overhangs, providing venting above the foam and below the upper roof deck. At hips and valleys these 2x2s will be interrupted (segmented) to allow air flow up to the ridge. They will be screwed to the rafters using 10-inch-long #10 screws as are used for SIPS. Yes, I am concerned about accurately hitting each truss top chord through the 6-inch thickness of foam. A drill guide to maintain accurate 90-degrees might help, but I’ve also considered piggy-backing a flat 2×4 on each truss top chord (what a waste of good lumber!!); either before or after sheathing.
Finally the super-rafters get sheathed by plywood, probably 3/4-inch T&G, to provide a sound deck for the finish roofing, which will be Decra Villa Tile stone-coated steel panels. These are designed for “direct to deck” installation, so no battens above the plywood.
My question involves the rain barrier between the roofing metal and the plywood. I’m sure this needs to be vapor permeable, and I realize 30# roofing felt (tar paper) would probably work fine. But I’m going to be doing a lot of this myself, and don’t expect it to move along very quickly, so a self-adhered rain-proof layer would be great. Hence my question about SIGA Majcoat vs VaproShield’s SlopeShield. I believe the latter is somewhat more expensive, but it states 51 perms vs 34 perms for the Majcoat.
I’m of course also open to general critique and comments about my design. I haven’t explained in detail how I’ve arrived at many of the specifics. I know one problem is the HardieSoffit only has 5 sq inches of free vent area per foot, which equates to a continuous opening about 0.4-inch (10mm) wide. I would think I want about 3 times that much, just to approximate the venting channel created by the super-rafters. I haven’t worked out the ridge vent detail, but I know there are several good solutions up there.
Finally I should add that I plan to install feet, or anchors, for future PV panels at the time of installing the Decra roofing, since it is not feasible to add them later (requires removing panels all the way up to the ridge!).
If you’ve read this far, thank you for your time and consideration!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Peter,
1. Why do you think that your roofing underlayment needs to be vapor-permeable? After all, you have a ventilated air space under the top layer of roof sheathing. Moreover, you aren't going to get much (or any) outward drying through your metal roofing. So in this case, there is nothing wrong with choosing a roofing underlayment that is vapor-impermeable.
2. I'm a little worried about the structural support for the three layers of rigid foam that are cantilevered over the roof sheathing at the eaves. Those three layers of rigid foam are kind of floating, supported only by the vertical EPS in your wall panels. What happens when a 200-pound roofer stands on that cantilevered foam?
-- Martin Holladay
Hi Martin,
Thank you for your fast reply!
1. My reasoning was only so that any water getting below the metal tiles could dry downward, because as you say there won't be much drying upward through the metal. Somewhere I thought I have read that metal tile roofing lasts longer if the underlayment is permeable so moisture cannot get trapped under the roofing. I believe this would only apply to "direct-to-deck" metal panels, not the batten mounted style. I may be wrong about this, so I will ask Decra directly whether they have evidence and/or a recommendation one way or the other.
2. Yes I should have shown a 2x8 (ripped to width) closing off the bottom of the insulation boards. See attached marked-up drawing. I think this is also a good idea to help minimize air flow up between the foam boards. I guess each layer of foam should be sealed to the one beneath it, all around the perimeter? That plus taping all seams and staggering them should help minimize air flow through the foam assembly. BTW, the 2x I've added in red would rest on the outer concrete wythe. But how to attach it and keep it from tilting is not obvious-- I need to work on that. This 2x also supports the super-rafters of course-- that is possibly more critical than supporting the foam layers; that may be what you were actually referring to. Without it the 2x2 would dent the foam when stepped on, I suspect. Also I plan to use 1.25 PCF EPS in the wall, so if it is reasonably accurately cut at the top miter face, it can support quite a bit of weight in case someone stepped directly over it via the foam layer(s) above (during installation of the foam and before the upper sheathing gets installed). Another option would be to run the outer concrete wythe up accurately to be flush with the top of the truss top chord, then let the truss sheathing run across the EPS and rest on that outer wythe (with waterproofing layer to protect sheathing from concrete moisture).
3. Since posting my original question I've read more on GBA about PERSIST, roof design with rigid insulation, etc. Most surprising to me was the temperature dependence of poly-iso's R-value-- I hadn't realized the dependence is so large. I'm going to re-think the type of foam, possibly do EPS for upper layer(s), poly-iso for lower layer(s). Also in Yavapai County, AZ, the current code for energy is the 2006 IECC, while the building code is the 2012 IRC. I have both of those in hard copy, which I have studied. Our zone is "4 Dry" and the required ceiling R-value is 38, compared to the newer codes that specify 49. My heat loss spreadsheet shows an annual savings of just $24 when I change the roof from 38 to 49. The cost/SF would be between $0.50 and $1.00, depending on which foam I use, so between $1500 and $3000 total. Even if the cooling cost changed the same amount as the heating cost, it's hard to justify a 30-to-60-year payback. So I am not inclined to exceed the local code requirement for the roof. Oh, and actually the 2006 IECC allows the "ceiling" R-value to drop to 30 if the insulation extends completely over the top of the wall at the eaves (as with an "energy heel" truss). Of course the code assumes wood rafters or trusses or ceiling joists and therefore assumes some amount of reduction from the nominal R-value of the installed insulation. With foam on top of sheathing, I only have one steel screw per 4-sq-ft to reduce my actual R-value below the nominal value. [Which actually reduces the total R-value by almost 9% if I've done the math right! I think I will investigate SST screws!]