Double Air Barrier
Designing my own passive house inspired home in 4A. The kitchen and living area will have scissor trusses. I recently found and attached a roof assembly that uses a double air barrier and scissor truss am an wondering if this is a problem. Specifically as you move towards the peak and there is to cellulose on the underside of the roof sheathing and the roof is airtight. Not sure if this is a concern or if I am overthinking this.
The second attachment is my modified version with materials that I have access to. Should I remove the roof Zip and use plywood with a highly permeable peel and stick to increase the ability to dry to the exterior or is this not a concern?
Finally, I am assuming that I should (or it would at least not cause a problem to) install and vapor diffusion port at the ridge so that any trapped moisture can get out into the 2×4 ventilation space with no resistance. Is this a correct assumption?
Thanks in advance!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Mick,
Unfortunately your assembly won't work or meet code with sheathing under the vent space. You can't use permeable insulation without either a vent space below, or continuous insulation above.
Code also doesn't allow the use of vapour-diffusion ports except in climate zones 1, 2 and 3
What does work is the same assembly without the lower sheathing. The moisture can move through the permeable membrane into the vent space. Josh Salinger has a good blog on those types of roofs, which perform very well: https://www.greenbuildingadvisor.com/article/building-a-vaulted-high-performance-and-foam-free-roof-assembly
Malcolm,
Thank you for your reply and alternative idea.
I thought the same thing until I saw this video. Is this bad advice or is there something different about the assembly Steve is suggesting that I am missing? https://www.youtube.com/watch?v=SooIofGUJdc
Assuming I built the assembly that Josh showed in the video how would I attach the roof sheathing? would nailing to the 2x4s with 2" nails be enough or would I need to nail through the plywood and 2x4 into the roof truss?
Mick,
There are two downsides.
One is code related. IRC R806.2 appears to preclude having permeable insulation in contact with the roof sheathing - unless you have an unvented assemble with continuous insulation above.
The other is practical. With Steve's roof any moisture has to diffuse through the roof sheathing before the vent space can remove it. Plywood is between 2 and 10 perms depending on how damp it is. Zip is only 3. A permeable membrane like Josh uses is between 30 and 70 perms, so you get considerably more drying.
Steve's roof also means you need a second basically redundant layer of sheathing.
I think the minimum nail length for attaching sheathing under your code is 2.5", so they would penetrate the top chord of the trusses slightly.
Malcolm,
Thank you for your response and knowledgeable advise.
Mick, I have designed one project using your proposed assembly, with diagonal 2X strapping over 5/8" Zip and standing-seam metal over that. At the interior we used 7/16" Zip sheathing as air control and vapor control, because the budget was extremely tight and the interior had to be semi-finished. That system cost less than using the variable permeance membrane and drywall I had first spec'd at the interior. The builder is the most diligent I know; he achieved 0.14 ACH50, with mostly double-hung windows, there is a whole-house ERV, etc. so I'm not worried about it, but I would not do it that way again.
I have built and/or designed roofs using the Sarking membrane system that Josh shows in the video that Malcolm posted, and that's my preferred way to design vaulted ceilings/roofs. I have had a lot of pushback from builders who have not built that system before, that installing the exterior membrane directly over the sheathing (edit: I meant framing) is too challenging, but I've done it myself and it's really not bad at all, just a few extra steps compared to the traditional approach.
Michael,
"I have designed one project using your proposed assembly"
I've advised Mick that his proposed assembly doesn't meet code because it is an un-vented assembly with permeable insulation against the sheathing, and no exterior insulation. Have I got that wrong?
Malcolm,
I am, for all intensive purposes clueless when it comes to building science but for what it is worth the way that I interpret the code this it not compliant. The reason I was considering this approach is because the code is not always right in my opinion (they do
not have to live with the consequences, I do) and I have seen plenty of other builders use a similar strategy without problems.
Mick,
I understand Steve's logic. There isn't much difference between the vapour having to move through sealed plywood baffles, and a layer of plywood sheathing - and both are a lot easier than if the baffles were made of something impermeable.
- But, the main purpose of venting is to keep the sheathing from accumulating moisture. It doesn't make sense to me to design an assembly that intentionally adds moisture to the sheathing before removing it. Doing that to baffles has no real consequences.
- Comparing the two I can't see any advantages, in terms of complexity, cost, or resilience, to Steve's assembly over Josh's.
- The starting point for every decision has to be: What am I legally required to do? There may be times when a code requirement may need to be challenged because it creates poor outcomes, but I can't get behind an argument that you should ignore code because you haven't seen a problem with some practice.
Malcolm, my argument, which the AHJ accepted, was that the Zip roof sheathing is a class 3 vapor retarder when damp, which is when you need drying to the exterior, so because we installed ventilation above the sheathing then it should be considered a vented roof assembly. The IRC only requires a class 1 or 2 vapor retarder on the interior in climate zones 6 and higher, which the Zip sheathing we used meets when dry. I would have only done that with the builder I worked with, with humidity control at the interior and with cellulose insulation. We installed data loggers and found that the moisture was getting high near the sheathing in late winter/early spring before the monitors died (I chose the wrong type) but based on other projects, the moisture load should have dissipated to both sides after that. I just did another project for the same clients, eight-ish years later, and they said the house is performing well with no issues. The owner is a former and still part-time renovation carpenter so he knows what to look for.
In my opinion, that assembly is probably safer than a common approach of using low-perm rigid foam or zero-perm plastic vent baffles, especially if they are airtight, and I haven't seen or heard about problems with those approaches, as long as there is vapor control and air control from the interior and humidity levels are within reason.
If I were to over-vent a roof again, I like Dr. John Straube's rule of thumb that in a heating-dominated climate, with certain caveats, the exterior side of the assembly should be roughly five times as permeable as the interior side. We actually had that, with roughly 0.5-perm dry Zip on the interior and roughly 2.5-perm "damp" Zip on the exterior. But I still wouldn't do it that way again, as I am not confident that even Zip sheathing will hold up over 50+ years of moisture cycling.
Mick,
It sounds like Michael is saying that he would be comfortable with your proposed assembly if it used plywood as the sheathing, and that a good case can be made that it meets code.
Malcolm, thanks for the Cliff's Notes version--I know I can be wordy! I would indeed be ok with plywood, but prefer Josh's assembly.
One other thing to note is that if there is metal roofing going directly over the roof battens without an additional layer of sheathing, the roofing will condense water that will fall on the sheathing, so make sure it's waterproof. Or install the metal on sheathing and it won't be a problem.
Michael,
I didn't put that very well. Not so much trying to edit your post as saying I should row back my assertion it definitely doesn't meet code.
Michael
"that installing the exterior membrane directly over the sheathing is too challenging"
I am assuming you meant to say "trusses" and not "sheathing?"
I am trying to get a blower door number like .14 and that is why I am looking at this assembly because I can have two continuous air barriers. One continuous inside and one continuous outside. I am curious as to which membrane you spec when building this assembly. Tyvek and be careful about taping the seams or do you use something like mento 3000 connect that has an adhesive strip on the ends of each sheet?
Yes, sorry, trusses, or rafters which I still use regularly.
It's been a long time since I completed or fully specified a Sarking membrane roof. Back then, 10+ years ago, I used Mento Plus or Mento 1000, but in the meantime products like Mento 3000 have reached the market. In theory, Tyvek or Typar should work, and I spec those for walls when I can to save money, but for a roof I prefer something bulletproof. With Mento Plus, you can put your full body weight on it between 24" o.c. trusses with no problems. I'm not sure if I'd trust Tyvek for that. Tyvek is supposed to be airtight but I prefer to stay within one system when it comes to membranes and tapes, and I prefer European tapes in many cases.
Michael,
I was looking at using the mento 3000 connect. Thank you again for your input!