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

Community and Q&A

Dense pack cellulose OVER closed cell foam in cathedral ceiling?

user-245916 | Posted in Building Code Questions on

I am considering the following roof assembly in climate zone 4a: Will i run into code or condensation issues?

Single plane roof 3/ 12 pitch

1) 6×10 timber rafters 4′ OC
2) 2×6 T&G decking perpendicular to and OVER (not pieced in between as a cosmetic ceiling) 6×10 rafters
3) 2×4 sleepers, 4′ OC, parallel to 6×10 rafters on top of T& G decking.
4) 2×10 boards perpendicular to and resting on top of 2×4 sleepers, 2′ OC.
5) 2″ closed cell applied from outside to top of 2×6 T&G. The 2×4 sleepers hold the 2×10’s off the T&G so the foam will be continuous underneath/prevent thermal bridging.
6) 5/8″ cdx plywood over 2×10’s
7) 30# roofing felt over plywood.
Cut holes through roof felt and apply dense pack cellulose to remainder of cavity on top of foam.
8) seal holes/patch felt
9) install metal roofing over roof felt directly or on 1×4 sleepers.

drawing of assembly attached

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. GBA Editor
    Martin Holladay | | #2

    You describe the 2x10s in your assembly as "boards." But boards are 1 inch thick or less. I would call the 2x10s "purlins."

    I don't think that this assembly complies with code requirements. In you want to combine a fluffy insulation like cellulose with foam in an unvented assembly, the foam has to go on the exterior side of the sandwich, to avoid the "cold sheathing" problem.

    Otherwise, you need to include a ventilation channel above the cellulose.

    To read more about insulated sloped ceilings, see How to Build an Insulated Cathedral Ceiling.

  2. charlie_sullivan | | #3

    I'm guessing that you are thinking this will work because the low permeability of the sprayfoam layer prevents moisture from getting into the cellulose from below and the roof prevent moisture from getting in from above. But that doesn't work. There will be some moisture in the cellulose, and in the winter, it will all migrate to the cold roof surface. And there are various ways more moisture can get in. When you confine moisture with no way for it to escape, it gets angry and will show no mercy in what it does to your decking.

  3. user-245916 | | #4

    I have read the article you mention along with just about every post or article on the subject. Whats to keep the second layer of sheathing that goes over the foam on the exterior side of the sandwich from being a "cold sheathing" problem? Code states that in a Permeable/impermeable unvented combo assembly the impermeable insulation must be above or directly below the roof decking. It doesn't say which roof decking. In my assembly the closed cell foam is above the 2x6 decking...doesn't that technically comply?

  4. user-245916 | | #5

    Yes charlie, that was my thinking. How does dense pack in walls ( without exterior rigid foam) not create the same issue? Wall applications like this seems to be a common and approved assembly.

  5. davidmeiland | | #6

    If you're building this new, I'm not sure why you wouldn't simply put rigid foam over the 2x6, then attach the 2x4s with very long screws, attach the plywood/felt/roofing to the 2x4s. Getting insulators up on the roof to install two different materials, without letting it rain during the process... that sounds tricky, and not something I would want to do on a project. You'll get a better and probably cheaper job this way.

  6. GBA Editor
    Martin Holladay | | #7

    Q. "How does dense pack in walls (without exterior rigid foam) not create the same issue?"

    A. Most walls can dry to the exterior when the sheathing gets damp. Most roofs can't. The code-required ventilation channel on the exterior side of the fluffy insulation is supposed to provide a drying mechanism, because most types of roofing don't allow moisture to escape to the exterior.

  7. iLikeDirt | | #8

    Gosh, this sounds complicated. Complicated means expensive and a lot of ways to design it wrong, install it wrong, etc.

  8. user-245916 | | #9

    David, I hear what your saying but the execution will not be tricky. The decking and felt can go on right behind the foam crew. Everything will be dried in in one day. The dense pack cellulose crew can come out any time to fill in the cavities because the roof will already be dried in. I want to stick with the sprayed foam vs rigid foam board for its sealing properties....I've never been a fan of butting panels and taping joints. The roof is a single plane shed roof 2000 SF... Just about as simple as they come. As far as price goes, I keep coming back to this assembly because it gets me high r value for low cost. 8" SIP's are coming in around 15k, This assembly will cost me 10K and have a significantly higher r value.

  9. GBA Editor
    Martin Holladay | | #10

    The way I see it, this type of roof assembly needs a ventilation channel above the top of the cellulose.

  10. user-245916 | | #11

    Thanks for the input Martin. Leaving an air gap would be more of a loose fill application in my assembly. I will probably stay with the dense pack and install 2" xps over the plywood decking.
    There seems to be an ongoing debate on this subject. I have found professionals and mfg's who say dense pack/unvented cathedral is the only way to go and others who wouldn't install it if their life depended on it. For example:

    "Dense pack non-ventilated cathedral ceilings are, as a building science study entitled Report On Roof And Wall Details: Upper Canada Post And Beam points out, "nothing more than well insulated exterior walls with insulating sheathing which is sloped.
    Cellulose has been used for years to dense pack cathedral ceilings with excellent results, outperforming the ventilated method. It is now being accepted as the preferred method of installation by building experts across the country."

  11. GBA Editor
    Martin Holladay | | #12

    Two inches of XPS above the topmost layer of roof sheathing will work in Climate Zones 1, 2, 3, or 4C. In the rest of Climate Zone 4, or anywhere colder, you'll need thicker rigid foam. Details are explained here: How to Build an Insulated Cathedral Ceiling.

    Applegate Insulation is correct if you are using a vapor-permeable roof underlayment and vapor-permeable roofing (for example, cedar shingles, concrete tiles, or slate). However, metal roofing is not vapor-permeable, so Appelgate Insulation's wall analogy doesn't hold in your case.

  12. Dana1 | | #13

    The notion that a roof is "nothing more than well insulated exterior walls with insulating sheathing which is sloped." is utter crap- an embarrassment to the authors who came up with that!

    Walls do not get direct & copious rain-wetting for hours or days on end, nor are they subject to being fully covered with snow & ice for weeks/months at a time.

    Walls are also not usually covered with extremely low permeance cladding layers, such as #30 felt + asphalt shingles (about 0.1 perms, on the verge of Class-II vapor retardency, only about 2x as permeable as 6-mil polyethylene)

    Roof claddings are necessarily nearly impervious to liquid water & water vapor on the exterior side, which is why moisture-suceptible roof decks need to be able to dry toward the interior, either in a ventilation channel or through the rest of the assembly.

    That said, in most US climates the roof deck can be protected from interior moisture drives with 1 - 2" of closed cell spray foam applied directly to the underside of the roof deck. That much foam runs about 0.4-1.5 perms (depending on density and thickness), and provides a non-susceptible condensing surface. With dense packed fiberglass that would still result in liquid water or frost formation on the interior side of the foam during cold weather, but if sufficiently air-tight to the interior there would not be enough moisture accumulation to cause damage, and long as there it isn't also extremely vapor open to the interior. While this would technically violate the letter of the code, it is not an extreme mold/rot hazard. If you used cellulose instead of fiberglass the cellulose would buffer that moisture in it's fiber without damage or loss of function.

    The solar reflectance also has an effect on the hazard level, since that affects the average temperature of the roof deck. A white/light colored roof is usually highly emissive in the deep infra-red, but also highly reflective of the solar spectrum, which leads to a higher hazard level than roofing with darker colors due to the induced lower temperature / slower drying of the roof deck. See Tables 3 & 4 of this document:

    Installing an interior-side smart vapor retarder it limits the hazard even further. With that additional feature it avoids being a direct code violation, but would simply be outside of the code. The IRC has plenty to say about what is necessary to be able to use only a Class-III vapor retarder on the interior, it is silent on Class-II vapor retarders. MemBrain or Intello Plus are both Class-II (not Class-III) vapor retarders when the proximate air is dry (under 35% for MemBrain) as it would be in winter, but when the humidity of the proximate air is high they become substantially more vapor open, into the Class-III range. Those characteristics allow the assembly to dry more quickly that it the moisture accumulates, provided the interior humidity is kept sufficiently bounded during the winter months.

  13. wjrobinson | | #14

    Dana, you are becoming a deep into the weeds theorizer extraordinaire of all things building enveloping for proper low energy consumption via proper care in super insulating. We need to send you to Bill Rose boot camp along with some time at Building Science put all your thoughts through some peer reviews.

    Keep up the great posts.... love them though I may have to go back to engineering school to follow along at some near future posting.

  14. user-6918827 | | #15


    I am curious what you ended up doing, and how the results turned out? I read that same article as you, and given everything else I have been reading about cathedral ceilings, including Martin's advice on my project, I saw this and thought that it might provide a less expensive and possibly better solution than what I am currently trying to achieve with venting.

    Have you had success with the dense pack approach?



  15. GBA Editor
    Martin Holladay | | #16

    I'm not sure whether Daniel is going to respond to a question on this 2-year-old thread.

    You mention "the dense-pack approach." I'll repeat what I said in this thread: There is nothing wrong with using dense-packed cellulose to insulate a roof assembly, as long as the assembly includes a ventilation channel above the top of the cellulose.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |