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

Closed and open cell foams on unvented cathedral ceiling

NormanWB | Posted in Energy Efficiency and Durability on

I am planning a cathedral ceiling over a great room that will have skylights, so based on what I have read on this site, it seems an unvented rafter assembly with closed cell foam is my best option. However, Martin’s recent “Musings” in Fine Homebuilding noted that closed cell foam is hard to trim, so most installers stop filling the bay a bit short, wasting R value.

I want to get the maximum insulation in the rafter bay, so I am wondering if it would be cost effective to use enough closed cell foam to seal the sheathing from moisture (2″ flash w.o batt), followed by open celled foam that can be trimmed a day or so later.

Also, would it be worthwhile to put foam board on the underside of the rafters after the trimming to address thermal bridging?

Greenville, SC

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
    ARMANDO COBO | | #1

    I work in CZ3, in TX. I recommend the 2" sprayed closed cell foam against the roof decking and 8" of open cell under that, for a total R38 min. assembly, but you'll have some thermal bridging.
    If you were to install metal or tile roof, I would install 1.5" rigid foam on top of the deck. The 1.5" is equals the 2x4 nailer edge you need for the roof perimeter, then the battens. If asphalt shingles, you could install a nailer base plywood (3/8 min.) on top of the rigid foam.

  2. user-2310254 | | #2

    Is this new construction?

    Martin's article on flash and batt ( indicates your can combine closed cell and open cell or another air permeable insulation. In Zone 3, you only need an inch (R-5) or somewhat less of closed cell from a moisture-safety standpoint. The balance of air permeable insulation needs to be applied directly to the CC, and it is also important to avoid air leakage in the drywall installation (no concealed lights, for example).

    If you have a simple roof line, you might consider installing rigid foam on top of the sheathing. You can use reclaimed sheets that will be much less expensive than new material--although you might have to give up the skylights.

  3. NormanWB | | #3

    THought I had responded, but don't see the post, so I must not have, so sorry for the delay.

    It is new construction. The main roof line is a hip ranch with this great room gable running perpendicular to it with a nice wooded river view. Roofing material will be asphalt shingles until Tesla's price point gets much lower. :) Lots of shade on the lot from mature hardwoods, so sun mostly around noon and in the winter.

    Wife and I do not want to get rid of the skylights and want to avoid the extra expense of another layer of sheathing on the exterior. Hence, the thought of using the foam board on the interior to mitigate thermal bridging. Would use permeable foam board (no foil facing) to allow drying to interior.

  4. Expert Member
    Dana Dorsett | | #4

    Closed cell foam spray polyurethane foam is one of the least-green insulation options going, due to the high-impact blowing agents (with a few exceptions) and the high polymer per R. As Steve Knapp points out, and inch is enough to be protective of the roof, and sufficient for dew point control on up to another R35 or more of fiber insulation or open cell foam. Open cell foam delivers more than twice the R per lb of polymer, and uses (very low impact) water instead of HFCs as the blowing gent.

    An inch of closed cell + 8-9" of open cell foam or cellulose / rock wool / fiberglass / open cell foam would give you the code-min R38 (for zone 3) with a 9-10" path (R10+) through the thermally bridging rafters. The vapor permeance of the insulation stackup would be about 1 perm, a minimal Class-II vapor retarder, sufficient for protecting the roof deck, but still plenty of drying capacity for seasonal drying toward the interior.

    A 6-7" all closed cell foam would have a significantly more thermally conductive R7-R8.5 thermal bridge, and would run about 0.2 perms, with only very slow drying capacity. Foam board on the interior would lower the vapor permeance even further?

    How deep are the rafters? If not quite deep enough to be able to get there with a flash-closed cell + fiber or open cell solution building 1.5 wide edge strips of rigid polyiso + 1x furring (or 3/4" plywood ripped to 1.5" width) glued & through-screwed to the rafters will add the necessary depth while providing a significant thermal break on the framing without limiting the drying capacity. A 2x8 rafter is 7.25", you need at least 9" for a 1" flash ccSPF + 8" of fluff or low-density foam. An edge strip of 1.5" polyiso + 3/4" wood would fatten that up to 9.5", while more than doubling the R of the framing fraction. Use only purpose-made foam board construction adhesive to glue the foam to the furring and to the rafter edge.

  5. EthanT | | #5

    Dana, thank you! I think we can no longer continue to let foam application suggestions on "Green Building Advisor" pass us by without pointing out the horrible environmental impact of foam.

  6. Expert Member
    Dana Dorsett | | #6

    But some closed cell foam IS green (in color, anyway! :-) )

  7. NormanWB | | #7

    Still waiting on the engineering to come back on the rafters, but I like the idea of adding the polyiso to the rafters and, if needed, 3/4 plywood., to get more insulation in the rafter bay. A lot simpler than rigid board.

    Note on Dana's "green" closed cell. This image perfectly illustrates the point being made by Martin in his "Musings", as it shows the lack of insulation in the last inch or so when using closed cell.

  8. Jon_R | | #8

    > code-min R38 (for zone 3) with a 9-10" path (R10+) through the thermally bridging rafters.

    I'm curious how the thermal bridging of the rafters allows compliance with the IRC in the case of unvented enclosed rafter assemblies using flash and batt at higher R values. For example., it may require R25 of air-impermeable insulation directly below the sheathing, not "R25, but feel free to replace sections of that with solid wood of less R value".

  9. Expert Member
    Dana Dorsett | | #9

    For unvented insulated roof assemblies code requires ( R5, in zone 3) the air-impermeable insulation directly under the roof deck, with the air-permeable insulation directly below the air impermeable layer. The thermal bridging does not affect those prescriptives.

    "5.1.3. Where both air-impermeable and air-permeable insulation are provided, the air-impermeable insulation shall be applied in direct contact with the underside of the structural roof sheathing in accordance with Item 5.1.1 and shall be in accordance with the R-values in Table R806.5 for condensation control. The air-permeable insulation shall be installed directly under the air-impermeable insulation."

    At higher R-values it's prudent to keep the same ratio of impermeable to permeable insulation, but that's far more important in zones 4 & up than in zones 1-3. R25 is the minimum for zone, at an R49 total, or a bit more than 50% of the total. If one were shooting for R80 it would be prudent to install R40 of closed cell foam. It's the average temperature at the foam/fiber boundary that determines whether & how much condensation/adsorption moisture will collect there over a winter, and that average temperature is largely a function of the R-value ratio.

  10. Expert Member
    Dana Dorsett | | #10

    Code allows all sorts of stuff than can be problematic when taken to the nth degree. Code compliance on it's own isn't guaranteed to be problem free- some amount of judgement (or calculation/modeling) is prudent.

  11. Jon_R | | #11

    So code would allow steel joists from the roof sheathing to the wallboard, perhaps causing significant condensation*?

    IMO, the moisture performance of partitions using continuous insulation is clearer and better than thermally bridged discontinuous insulation (ie, at the same R value ratio, flash-and-batt or cut-and-cobble are riskier).

    * In Winter, at the steel/foam/fiber corner and at the steel/fiber boundary. In summer, at the steel/roof-sheathing boundary area.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |