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

Unvented cathedral with metal roof: Is open-cell foam with vapor retarder OK?

slonnett | Posted in General Questions on

My question is regarding a small addition that we are building in Zone 5A in southeastern Michigan on a vacation home.  We connected a screened porch to the main house with a small Foyer addition (roughly 10 x 20) that has an insulated concrete slab-on-grade floor (electric NuHeat mat planned under porcelain tile flooring), 2×6 exterior wood stud walls, and 2×12 sloped rafters @ 16″ o.c. with a 3:12 pitch to form an unvented cathedral ceiling.  The cathedral ceiling finish will be stained tongue-and-groove wood planking.  There is an Mid-Century style single entry door with a large sidelight and transom on the east wall (short side of the single-slope cathedral) and one fixed window (24″ x 48″) on the west wall (high side of the single-slope cathedral).

We re-roofed the entire house as part of the addition, and this new Foyer structure has standing seam metal roof over one continuous layer of ice & water shield over plywood sheathing per Michigan 2015 IRC building code.

Initially, I had planned for the ceiling to be insulated with closed-cell foam  or a hybrid of CCF/OCF to R-49 (even though code only requires R-38).  Throughout the many discussions with my contractor and possibly the fact this is not our full-time residence and I’m not onsite to oversee everything, something got “lost in translation” and the spray foam company the contractor hired INSTEAD installed 6″ of open-cell Icynene-Lapolla foam in the unvented cathedral ceiling.  The foam application was applied to a 6″ depth and directly to the underside of the roof sheathing boards.  The cavity is not filled, only about half of the rafter cavity is filled.  It appears to be installed correctly… it’s just not what I thought I wanted.  But maybe this will still work out – I’m struggling to make sense of all the conflicting data and publications about the hazards/risks/benefits of OCF or CCF for OUR particular situation.

I accepted responsibility for not making my design criteria more clear to my contractor before they foamed the job, and he accepted responsibility for dropping the ball on this and not asking for my review of their proposal before beginning — and for trusting the foam company’s recommendations, but that’s not what I’m grappling with now.  He’s a good guy, and we’re continuing the project with him; I have to figure out what I want to do now though.

The question is whether or not I can still salvage this assembly by installing a layer of 1/2″ gypsum board then Certainteed MemBrain (to act as a vapor retarder)?  (2015 IRC 806.5.4)  I can still install the stained tongue-and-groove wood plank ceiling over the gypsum board, so that’s not an issue.

I spoke with the manager of the spray foam company, and while I wasn’t impressed with his technical knowledge and I disagreed with his salesman’s recommendations to our contractor, we arrived at two options to move forward:

A) Leave the 6″ OCF in place in the unvented cathedral ceiling and spray MORE open-cell foam to fill the rafter cavity – then install one layer of 1/2″ gypsum board on the underside of the rafters and a vapor retarder (either via coating or membrane product). They will not charge for the additional foam, and this assembly meets code (2015 IRC 806.5.4).
B) Rip out the OCF and install 5.5″ of closed-cell foam in the cathedral ceiling. This comes with an additional price tag of about the same fee as the first installation, so essentially I’ll be paying twice for the job!

My concern is whether or not the unvented assembly (open-cell foam + 1/2″ gyp. + vapor retarder) will be OK over time and actually improve the roof’s ability to dry to the warm side and dissipate any vapor that may form or diffuse inside the rafter cavity?  Or am I gambling by leaving the open-cell foam in there at all?  We are already behind schedule for completion, so I have to make a decision quickly.

Some have suggested that, due to the presence of the continuous ice & water shield above the roof sheathing boards, we would actually be CREATING a vapor problem if we used closed-cell foam on the cathedral ceiling because any vapor that forms (either by condensation or night radiation) is truly trapped and has nowhere to go.  There is absolutely no desire to tear off the roof and add rigid foam board on the exterior because it’s a brand new roof, nor can I add venting to the rafter bays at this stage.

Please advise.  Thank you –

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. Expert Member
    Dana Dorsett | | #1

    Option "A" is not very risky, especially if the metal roof doesn't have a "cool roof" finish, and is not facing north.

    It is even less risky if the place isn't occupied by bathing, breathing, cooking mammals all winter, but still ventilated at some minimum level. Square feet of conditioned space x 0.01 cfm should be enough ventilation to keep the indoor dew point averages tracking reasonably with outdoor levels when unoccupied, with no active moisture sources. That can be verified with indoor humidity monitors and adjusted if needed.

    With air-impermeable open cell foam and a smart vapor retarder like MemBrain (or even vapor barrier latex paint) there is no air-transported moisture, only vapor diffusion, and the diffusion rates through MemBrain are extremely low during the winter when the indoor air is dry.

    Option "B" doesn't create any problems, but the moisture content of the roof deck should be verified to be well under 20% before installing the closed cell foam.

    1. slonnett | | #2

      Thank you, Dana!

      The roof slope faces the west, and it does not have a 'cool roof' finish. My concerns are eased by your comments. We do keep the house at 55 degrees in the winter and while it's more occupied in the summer, we do still use the home year round. And because it's the Foyer -- and away from all living, breathing, bathing, cooking mammals :) -- the indoor humidity levels can be easily managed.
      I'm still curious about whether "Option B" would create any new problems by doubly-sealing the rafter cavity to vapor/moisture?

      1. Expert Member
        Dana Dorsett | | #3

        Option B will work just fine if the wood is sufficiently dry before the closed cell foam goes on. Once the closed cell foam is installed moisture can still move in and out of the roof deck, but only at extremely low rates, and the moisture content of the roof deck won't change much with the seasons/years.

        Option A is higher performance than Option B, due to both the longer path through the rafter (11.25"/ R13.5 instead of 5.5" / R6.6) and the higher center cavity R value ( R40+ compared to R33-R39).

        The dew point of the air in the foyer will be about the same as in adjacent spaces, whether occupied or not, but if wintertime use is intermittent and you're not actively humidifying the indoor air, with even modest ventilation rates the average humidity is going to be low (uncomfortably low for some people.) But with even the modest solar heating of a west facing roof pitch, air-impermeable insulation and an interior side smart vapor retarder there isn't much risk even if the indoor humidity rises to 40% RH @ +70F during the winter periods that it's occupied.

  2. Peter Yost | | #4

    Knowing is better than not: and knowing is measuring:


  3. slonnett | | #5

    Thank you to both Dana and Peter for weighing in on my questions and sharing your advice and expertise. It gave me great peace of mind knowing that our decisions were backed by this great shared community of knowledge and science - our project is wrapping up soon, and the outlook is good that this roof assembly will perform well for many years to come. Thank you again!

  4. Jon_R | | #6

    > installing a layer of 1/2″ gypsum board then Certainteed MemBrain

    For any future readers with similar mistakes, note that R806.5 requires "a Class II vapor retarder coating or covering in direct contact with the underside of the insulation". So be careful that there is a complete, trimmed flat cavity fill, then MemBrain, then drywall (to provide the required direct contact).

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |