GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted
Building Science

Building a Vaulted High-Performance and Foam-Free Roof Assembly

Josh Salinger shows us how to create more space and vent a roof, and do it all with a reduced environmental footprint.

Josh Salinger, owner of Birdsmouth Design-Build, is on-site in Portland, Ore., to discuss the vaulted high-performance and foam-free roof assembly on their latest ADU build.

Vaulted roofs are a popular option because they create usable living space where an attic would usually be. Typically, plastic forms such as spray-foam insulation or rigid foam are used to deter condensation and moisture issues in a roof assembly.

In this video, Josh shares his plastic-free solution to manage moisture and allow for full-depth insulation with proper venting while creating a high-performance vaulted roof assembly.

Read more about all the details of this roof assembly in Josh’s article, “A New Take on Insulating a Roof.”



  1. Jon_R | | #1

    What are your thoughts about the alternative method using "RafterCaps" (or similar) as described here?

    After blowing in insulation, what is the average vent size after accounting for some upward bow to the Tyvek?

    1. user-1116814560 | | #4

      Have to buy and install a lot of rafter caps (proprietary) assume this wont be a cost savings versus shimming out roof CDX with 1 x strips…also this is a very “fussy” install with the sticky peel and stick would be problematic on top of the roof versus just laying out the commercial tyvek and nailing , taping exposed seams with tyvek tape . Most peel and stick is hard to handle even when just laying it out flat in one long sheet ! And the commercial tyvek is going to be yopu send layers of WB as the CDX /ZIP will have to be covered with its own WRB.

      1. Expert Member
        MALCOLM TAYLOR | | #5

        'versus shimming out roof CDX with 1 x strips"

        The material needs to be 2"x to both meet code and be effective.

      2. Jon_R | | #6

        No need for peel and stick and if you are duplicating the function of a RafterCap, you can use 1x strips (attached to the upper side of the rafter, not the upper edge). This does leave less room for fluffy insulation. But might be worth it if you wanted a more that 1" vent (which is my guess for the average vent size with above rafter 2x4s after accounting for Tyvek bowing caused by dense pack insulation).

      3. Expert Member
        Joshua Salinger | | #8

        The issue with using plywood sheathing and strips above (likely needed 2x's for ventilation) is if you are using plywood in lieu of a highly permeable membrane, it in effect becomes the primary vapor control layer (the least permeable part of the assembly) as the plywood has a perm rating of around 10 perms. The smart membrane below will vacillate between (roughly) .1 and 10. When the plywood becomes the primary vapor control layer, then this part of the assembly becomes the point of concern since its on the cold side of the assembly. It is essentially open to the universe as the vented area above is no different than the area above the roofing-- the same as just stuffing a rafter bay with fluffy insulation.

        If one is using a highly vapor open membrane such as Tyvek, which has a perm rating of 53 perms, the lower membrane on the warm side of the assembly becomes the vapor control layer. The lower membrane controls the vapor diffusion into the assembly and is warm so there is no condensation, the upper membrane is wide open so the vapor that gets in can escape through it into the vented area above (this depends on the season, RH, interior temps, etc., but the principle stands). The key to this detail is that the smart membrane is not just controlling the vapor, but also is the air barrier.

    2. Expert Member
      Joshua Salinger | | #7

      One of the issues we were trying to solve with this roof system is how to vent roofs that aren't simple gables. The problem with the 'RafterCaps' or creating a vent chamber below the roof sheathing via baffles or a site made chase is what does one do if one has a valley, hip, dormer or other obstruction to either the bottom or the top of the rafter bay. By bringing the vent chamber over the top of the valleys, hips, etc. it completes the ventilation from the intake to the ridge. It also has the benefit of using this space for insulation instead of air.

      I have gotten the Q about the bowing many times and I don't have an empirical answer to it. I assume it is bowing, especially at 24" OC and a good dense pack install. That being said, I also know that cellulose settles over time and whatever bowing there is will likely be reduced, but as to how much is anyone's guess. It probably has a lot to do with the quality of the install. In a strange twist, a poorly installed cellulose job would likely reduce the bowing issue.

      The only way to really find out would be to peel off the roof after insulation. Unfortunately, I have yet to find a client who is willing to have us rip off the roof of their new home to check on the insulation install...

  2. scottohara | | #2

    Thank you Josh. I really like that detail. Have you used it on a lower sloped roof, say 2 in 12, that is common on a more modern home design? I build in the SF Bay Area where the modern design is popular but I've stopped doing it for spec homes because of the amount of foam needed. This seems like a good alternative if the air would move through the vent on a low slope.

    1. Expert Member
      Joshua Salinger | | #9

      My hunch is that the lesser sloped roofs wouldn't perform as well. I can't say if it wouldn't perform at all since there are so many variables with not just roof pitch but other factors such as roof color, size of vent cavity, etc. One needs 2 holes and a driving force. The driving force of stack effect would be reduced if the slope is minimized. If I were to do it on a low sloped roof I would make the vent chamber as large as possible and go for a dark colored roof. Peter Yost at his 'Wing Nut' testing blog has some great information on this here:

      1. scottohara | | #10

        That was my thought as well. Just about all of the houses I build are in WUI so adding vent area may be a challenge. Thanks for the link, I'll check out Peter's information.

      2. Jon_R | | #11

        +1 on lots of variables with no method of accounting for their effects.

        Note that wind is the larger driving force and this doesn't change much with slope. Also that with the high performance Majrex and careful air sealing, venting requirements go down significantly.

        > SF Bay Area ... amount of foam needed

        A huge effect. Given the monthly average outside air temperature of the three coldest months, the IRC allows zero inches, even with no foam. So my guess is that a 1" vent in a 2:12 roof would perform very well.

  3. user-1116814560 | | #3

    Elegant, not expensive, easy to maintain quality control-brilliant.

  4. sidekickarchitects | | #12

    Thanks so much for sharing this fantastic assembly, Josh. A few questions if you have a moment:

    1) I'm not quite clear on how/where the Tyvek Commercial Wrap WRB terminates at the eaves/rake. I see that the WRB at the walls is lapped with the self-adhesive and connected to the variable-perm air barrier BELOW the rafters, but doesn't that leave an opening for water to infiltrate between the self-adhesive at the top plate and the WRB above the rafters?

    2) Is it safe to assume this assembly does not work with skylights since the vent chambers above/below skylight would not have a continuous path to the ridge, or can one leave the gap between parallel and perpendicular strapping to allow for cross-bay ventilation?

    3) Any concern about loss of shear-resistance in the roof diaphragm due to the 1 1/2" strapping-thickness gap between sheathing and structural rafters? I know our structural engineers get nervous about Zip-R sheathing walls of any significant rigid thickness for this reason.

    4) Do you see any reason this assembly would not work for Los Angeles' marine climate and/or with 2x10 rafters (other than the reduced insulation thickness in the rafter bays)?

    1. Expert Member
      Joshua Salinger | | #13


      1) I'm not sure if I follow the question... Maybe you are referring to the space between the top of the top plates and the bottom of the roof sheathing on the exterior? If one had an open soffit, one could simply place WRB in between the rafters at the blocking if one was concerned here. With the overhangs I don't see how water could possible get into this assembly and if it did somehow, it is all vapor open and would dry out.

      2) It works fine with skylights for the same reason it works with hips, valleys and dormers. In fact, this was one of the issues we were trying to solve with this assembly. Leave gaps at the adjacent strapping bay for ventilation. If you go to the article in FHB this is shown in the illustration.

      3) We worked with our structural engineer as we designed the assembly. We are in a seismic zone and we were able to get it to calc without any issues.

      4) It should work well in all climates, especially one like LA. You are also able to get away with 'vapor diffusion ports' in that CZ which would save you some effort and costs. If your structural engineer let's you get away with 2x10's and it meets code minimum you should be fine, although I can only recommend going with a beyond code level insulation, if possible.

      1. sidekickarchitects | | #14


  5. acarbs12 | | #15

    Have you seen they seem similar to what you are doing...

Log in or create an account to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |