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

Community and Q&A

Vented vs. unvented roofs: Need primer

JohnPaulJames | Posted in Green Building Techniques on

I’m having a building designed for me [I gave a sketch of the floor plan I’d like to the draftsman].

Part of the roof will be engineered trusses.  Part will be 12″ I beams to allow a cathedral ceiling over the great room, and windows along the south facing wall.

Now trying to learn the best way to insulate the roof and do I need to the roof to be vented or un-vented.  Didn’t even know to ask the question [assumed all roofs were vented] until I joined GBA yesterday.

So, school me on roofs.



GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. Jon_R | | #1

    Joe Lstiburek on vented: "there is no better", "works everywhere", "best value proposition".

    Also "the drywall layer must be airtight" (which is different than saying the partition must be airtight at some layer).

    Martin Holladay: "You can still vent roofs that don’t have ridges".

  2. Expert Member
    ARMANDO COBO | | #2

    IMO, the best cathedral ceilings are non-vented, with taped and seal rigid foam over the taped sheathing, and dense packed cellulose in between the TJIs.
    The engineered truss roofs, depends on whether you need a non-vented attic for an HVAC system and/or ducts, or storage. If you don't have those requirements, I typically design a vented attic assembly.
    There is plenty of articles here in the GBA and the web to point you to the best practices for any king of roof and/or attic assemblies.

  3. GBA Editor
    Martin Holladay | | #3

    Start by reading this article: "How to Build an Insulated Cathedral Ceiling."

    If a roof has valleys, hips, or dormers, the vented approach is not possible. That type of roof needs to be an unvented assembly.

  4. JohnPaulJames | | #4


    The roof does not have valleys, hips or dormers. Just a simple, single, peak down the middle. 9:12 pitch.


  5. JohnPaulJames | | #5

    OP here, I've been reading and reading.

    Building on the border of zone 4 & 5 so want to use zone 5 requirements [although where I'm building there is no code or inspections imposed]. Zone 5 has R-49 as the minimum goal. I'd like to strive for R-60.

    Builders here are very traditional so trying to balance best practices with what I can get someone to actually build in the area.

    Will this system work?:

    1 inch of closed cell sprayed on the under side of the sheathing followed by 10 inches of Polyiso covered by dry-wall? The closed cell should act as an air barrier. The Polyiso should always be above the temps where its R values starts diminishing and any moisture should vent to the inside of the house.

    WRB on top of the sheathing then covered by a metal roof.

    Will I need venting? If so, can it be above the WRB and below the metal roof in the form of 1x4's or something similar? If necessary, there should be an inch available, in the rafters, if channels are put against the the underside of the sheathing and before the spray foam is applied and then a traditional soffit to roof ridge vent system could be used.

    What can I do to mitigate thermo bridging?

    The roof pitch is 9:12.

    Is there a better alternative that I'm not seeing? I've probably read 20 articles from this site in the last day.

    Still in the planning & design stages but fear I'll spec something out that I'll not find anyone willing to build [especially in this market where there is plenty of construction happening so labor is scarce].



  6. Expert Member
    RICHARD EVANS | | #6


    Martin's article on cathedral ceilings offers 5-6 options. Your option is a kind of hybrid of two of them. My fear with your idea would be that warm, moist air would eventually find its way around the 'cut and cobbled' polyiso and then condense on the underside of the cold, inadequately thick layer of spray foam.

    From a buildability standpoint, I would hate to be the builder that would have to cut that much polyiso and then have to seal the edges perfectly. As the rafters expand and contract with changing temperatures, the seals are likely to break down anyway.

    I think its much easier to build a vented attic in the colder climates. Also, an R-60 flat ceiling loses less heat than an R-60 cathedral ceiling because the cathedral ceiling has more surface area.

  7. JohnPaulJames | | #7

    Thanks Rick,

    Martin's article on cathedral ceilings was one of the first of the many articles I've read.

    I should have added that the primary HVAC system will be a ground loop heat pump. There will also be a EPA approved wood burning stove.

    To your points: momma say Cathedral ... so flat is out, vented attic is out.

    I accept your points and so my system is flawed.

    What will work, using the most 'traditional' building methods?



    1. Jon_R | | #11

      Consider a raised heel scissor truss to get a vaulted ceiling and some attic space.

  8. GBA Editor
    Martin Holladay | | #8

    An R-60 cathedral ceiling needs to be carefully planned. You'll need about 17 inches of cellulose if you use cellulose.

    If you want to combine exterior rigid foam with fluffy insulation between the rafters, you would need 41% of the insulation (in Zone 5) in the form of rigid foam -- so R-25 of exterior rigid foam (call it 5 inches of polyiso) along with about 10 inches of fluffy insulation between the rafters.

    More information on this last approach here: "Combining Exterior Rigid Foam With Fluffy Insulation."

    1. JohnPaulJames | | #9

      Thanks Martin,

      Yes, I read the 41% ratio in several articles last night and came to the same R-25, 5" calculation.

      The problem is, I don't believe I'll find a builder willing to put exterior foam sheet over the sheathing as it isn't a common practice here.

      Will spraying 1 " of closed cell in the engineered I Beam rafters [which has an 11 7/8" depth] against the underside of the sheathing then 10 inches of polyiso under that, then drywall work?



  9. GBA Editor
    Martin Holladay | | #10

    If you are thinking of using the cut-and-cobble approach, the answer is no, I don't think that's a good idea. More information here: "Cut-and-Cobble Insulation."

  10. brendanalbano | | #12

    This is fairly expensive solution, but hits R-60 in a 12" cavity and is installable with conventional construction practices:

    - 5" HFO blown closed-cell sprayfoam (e.g. Icynene Proseal HFO, Demilec Heatlok HFO High Lift, Lapolla FOAM-LOK 2000 4G): R-34
    - 7" Dense pack cellulose or dense pack fibergalss (JM Spider): R-26
    - TOTAL: R-60

    Jon R's suggestion of a vented assembly with a raised heel scissor truss may be more economical.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |