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

Need help insulating an old non-vented cathedral ceiling

user-1085290 | Posted in General Questions on

I am looking into the proper way to insulate a room in my home that has a cathedral ceiling. I am in zone 3 (southern California). The house was built in 1957 and the ceiling assembly is as shown in the uploaded photo. (drawing). the room is 12’x12′ FYI.

From top(roof) to bottom… this is the assembly.

The roof is a metal 3 dimensional shingle (Metro shingle) so there is some venting “air flow” beneath the shingles. There is Titanium UDL50 synthetic underlayment, then 1×4 skip sheathing, 2×4 rafters, and then drywall.

My plan as of now is to add 2″ to the 2×4’s so i can insulated with Roxul R23 and then drywall the ceiling with 5/8 sheetrock making sure it is as air tight as possible.

I am just looking for some advice if this method should be acceptable or problematic.

It is an unvented assembly (other than leakage throughout wall via air gaps and the 1×4 skip sheathing) and i plan to completely seal every bit of leakage i can find.

Thanks in advance.

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 | | #1

    J. Loc,
    In your climate zone, roofs should be insulated (according to the 2012 IRC) to a minimum of R-38.

    Titanium UDL50 synthetic underlayment has a permeance of only 0.05 -- meaning that it is a vapor barrier. If you want to insulate with mineral wool insulation under this vapor barrier, you must have a ventilation channel between the top of the mineral wool insulation and the underside of the synthetic roofing underlayment.

    If you don't want to create this ventilation channel -- and doing so will require soffit vents and a ridge vent -- then you will need to insulate with spray polyurethane foam, not mineral wool.

  2. user-1085290 | | #2

    Thanks for the info Martin.

    I wish i could do soffit and ridge vents, but it is not possible. The Garage meets up to lower side of the room not allowing soffit vents. I was definitely hoping to NOT have to use poly foam. If i do the foam myself it is 3x the cost. If I have to hire someone...even more. Hmmm.

    Another question is closed or open cell foam? I's guessing closed cell?

    Thanks again for anyone who has some advice.

  3. Expert Member
    Dana Dorsett | | #3

    Closed cell foam is a waste of good foam when thermally bridged by rafters. At R7/inch and 5.5" you'd be at R38 meeting the letter of code, but it would underperform R38 fiber between deeper joists/rafters due to the much higher thermal bridging of ~R6.5 rafters instead of ~R13 rafters (twice the thermal bridging.)

    You can get there without spray foam, but to hit R38 or U0.030 (the code-prescribed alternative to R38 ) will necessarily be a deeper/thicker assembly. ( R23 rock wool in fattened-up 2x4s with 3" nailbase polyiso would get you to ~U0.030 if the rafters are 24" o.c. )

    Using a vapor permeable underlayment would offer a lot of moisture relief for the skip sheathing, and in zone 3B climate letting skip sheathing dry toward the exterior under the 3-D metal shingles would be fine. Is Titanium UDL50 the only underlayment approved by the shingle vendor, or is that something that you just picked for other reasons? There are vapor permeable synthetic underlayement options out there, eg:

  4. user-1085290 | | #4

    Dana, The roof and titanium underlayment are existing so there is no changing that (easily anyway). When you say "Closed cell foam is a waste of good foam when thermally bridged by rafters" are you recommending that I not use it?


  5. user-1085290 | | #5

    The more I try to look into finding an affordable solution, the more confusing it is due to so many opinions.

  6. GBA Editor
    Martin Holladay | | #6

    J. Loc,
    I don't know who designed the building, or how old it is, but it is a difficult assembly to insulate. If you can't remove the roofing, you're stuck with either closed-cell spray foam (the preferred option from a performance standpoint, but not an environmental standpoint) or open-cell spray foam.

  7. Expert Member
    Dana Dorsett | | #7

    The difference in "whole assembly-R" performance between 5.5" of open cell foam and 5.0" of closed cell foam at a 7% framing fraction (typical of CA roofs with 24" o.c. rafters in existing building surveys) is about R8-R9. You can't really do a full cavity fill of closed cell, since it's not trimmable, so don't count on more than 5", even if it's more than that in some places, one thin spot will take a toll on the average performance.

    To do 5"+ safely and with good adhesion it would have to be done in three lifts of no more than 2" at a time, with a cooling/curing period between lifts to avoid the fire hazard.

    At 5" you're looking at about 0.2 perms which is getting on toward "dries never" performance. It's not a true moisture trap, but is way more vapor tight than ideal.

    At the same thickness, you did full cavity fill of open cell foam on the 2x4 rafters then installed a continuous layer of 2" of unfaced Type-II EPS, and long-nailing /long-screwing the ceiling gypsum to the rafters you'd have about the same thermal performance, and nearly three times the drying capacity as 5"+ of closed cell foam. The 3.5" of open cell foam can be safely installed in a single pass.

    With credit for air films and ceiling gypsum it'll end up with a U-factor about U0.040 give or take, which is higher than code, but still not terrible. The benefit of having an R8.4 thermal break over the framing is considerable from a temperature-striping point of view, even though the overall performance is slightly less than 5" of R6.5/inch closed cell foam in a 5.5" rafter bay.

    If the framing fraction is higher than 7% (which it might be) the performance balance tips fairly quickly toward the o.c. foam + 2" EPS solution.

  8. user-1085290 | | #8

    Martin, Thanks again for the input. The house is 60 years old, and yes.... it is definitely turning out to be a difficult assembly to insulate ( or at least figure out how to do it "correctly"). For such a small room, it is going to be expensive either way I go apparently. The roof was recently done and so redoing the roof to change out the titanium underlayment is not feasible.

    Thanks for the information Dana, although some of it is too scientific and over my head as far as my building experience goes.

  9. GBA Editor
    Martin Holladay | | #9

    J Loc,
    Unfortunately, the insulation decisions were delayed too long.

    The insulation decisions should have been made at the same time as the roofing work. That way, you might have chosen a vapor-permeable underlayment, which would have been preferable, or you might have chosen to install rigid foam insulation above the roof sheathing.

  10. user-1085290 | | #10

    Martin, I wish I had the option of making that decision. We did not have the ceiling open or decide to redo the drywall on the ceiling in this room until after the roof had been completed. It is what it is now and I will go with what you have suggested.

    If I may pick your knowledgable brain, I have another question. Another room in our house has an exposed rafter beam ceiling (4x6) and T&G decking( many homes of this era here in SoCal have these). The assembly from top to bottom is: The 3-D metal roofing tiles, titanium underlayment, 1x4 skip sheathing over the 1x6 T&G which is the finished ceiling inside. Am I going to have any issues with the titanium underlayment trapping moisture above my T&G ceiling now?


  11. GBA Editor
    Martin Holladay | | #11

    J Loc,
    The answer to your latest question depends on whether you intend to insulate the ceiling -- and if you intend to insulate it, how you plan to do so.

    In theory, the Titanium underlayment acts like a plastic bag or a tent. If you are camping in cold weather, and the interior of the tent is warm and humid, you can get condensation on the inside of the tent. In the case of the underlayment, I don't think you'll get much condensation -- and if you do, it will evaporate at almost the same rate that it forms, so you won't have a problem.

    If you intend to insulate this assembly, you have to carefully consider how this vapor barrier will perform.

  12. user-1085290 | | #12

    Martin, No... I do not plan to insulate this section of the roof, we would lose the exposed T&G and open beam design. The T&G has been caulked/sealed (grooves and along the 4x6 on the interior as well fyi).

    Thanks again for your help.

  13. user-1085290 | | #13

    I am still doing my research and had a couple more questions that hopefully guys can help me with.
    You mentioned closed cell not being trimmable, Can you elaborate or share some links to info regarding why you can't trim it? I have been searching but have come up empty handed. There is a lot of info on methods and tools on how to trim closed cell after over filling, but very little on that it should not be trimmed.

    @ 5.5" of closed cell, you mention that puts the assembly into a "never dries" condition. Does "never dries" mean if there was a leak? Or just general moisture traveling through the assembly? If the drywall on my ceiling is installed fairly tight, then with 5-5.5" of closed cell foam, then the UDL50 Titanium underlayment and vented metal roof on top, Shouldn't moisture in the assembly not be an issue unless I ever have a roof leak?

    If I go with 5-5.5" of CC foam, Is there any option/suggestions to help with the thermal bridging?

    Attached is an actual photo of the ceiling structure as it is now with the added 2" to get to 5.5".

    Thanks again for sharing your knowledge and experience.

  14. GBA Editor
    Martin Holladay | | #14

    J. Loc,
    Professional installers of closed-cell spray foam do their best to avoid trimming closed-cell foam, because it is dense and very hard to trim. Tools are available (electric rotary trimmers) that can shave cured closed-cell spray foam even with the studs, but the tools are expensive and the process is time-consuming. So in most cases, if a customer wants to fill 2x6 studs, the spray foam contractor aims for 5 inches of foam, not 5.5 inches of foam.

    Here is a link to a site that shows the equipment used to trim closed-cell spray foam:

    In your case, you have installed a vapor barrier (the synthetic roofing underlayment) on the exterior side of your sheathing. Dana is right -- if you install closed-cell spray foam under the sheathing, you'll have a "never dry" scenario, since you will have encapsulated the wooden elements of your roof assembly in two impermeable layers. Is that a problem? In my mind, it's far from ideal. But people do it. Of course, if your roof leaks, your roof sheathing may rot. But that's always the case with roof leaks -- so you may be willing to accept the situation.

  15. user-1085290 | | #15

    I have two more insulation projects in the same room that I am trying to resolve, Your advice would be greatly appreciated.
    The first is how to insulate the "knee wall" (see attached rough sketch) that separates the attic from this room with the cathedral ceiling. It is 2x4 construction, it has the original plaster over perf plasterboard and has had a newer layer of 1/2 applied at some point.

    I was thinking of spraying 2" of closed cell on the back of the drywall and filling the rest of the cavity with mineral wool. The framing is only 2x4's, But I was thinking on finishing it off with (2x6) R23 Roxul batts, and using insulation spikes (if needed) to hold them in place. because Roxul is so easy to trim, I would be able to notch out the Roxul around the studs keep a consistent layer of insulation.

    What would be your recommendation to properly insulate this section from the attic?

    My other project is insulating a bay window (see rough sketch) in the same room. It has no insulation above or below the window. I just discovered this while redoing the trim work outside the window. The glass is dual pane so I do have that working for me. The window faces south and gets sunshine all day long fyi. The roof over the window is basically the same as the room with the cathedral ceiling. ( from top to bottom- metro cottage metal shingle, ugl50 titanium underlayment, 1x6 T&G, and 2x4 rafter tails).

    And a reminder... I am in SoCal / Zone 3

  16. GBA Editor
    Martin Holladay | | #16

    J Loc,
    The kneewall is an exterior wall (with conditioned indoor air on one side and exterior air on the other). Detail it like any other exterior wall (although you don't need siding). The more R-value, the better. In your climate zone, the easiest approach is probably some type of fluffy insulation between the studs, and a continuous layer of thick rigid foam on the attic side of the studs. Pay attention to airtightness when you install the rigid foam (tape the seams). In most jurisdictions, building codes require a layer of drywall on the attic side of the rigid foam for fire protection.

    For more information, see “Two Ways to Insulate Attic Kneewalls.”

    Concerning your bay window: for the "attic" above the bay window, detail it like any attic (if there is enough room). If the depth isn't enough, closed-cell spray foam will provide more R-value per inch than any alternative.

    To insulate under the bay window, you'll need to install one or more layers of thick rigid foam on the exterior side of the projecting shelf. Trim the rigid foam at the perimeter with a weather-resistant trim board like cellular PVC, painted red cedar, or painted white cedar. The underside of the rigid foam can be finished with unvented soffit material. As usual, pay attention to airtightness when you perform this work.

  17. user-1085290 | | #17

    Ok, Got it.

    So I had it backwards on the kneewall with spraying the CC foam on the back of the drywall to air seal and then insulating. With treating it like an exterior wall like you mentioned, makes more sense.

    At the top of the kneewall where it meets the rafter in the attic ( see circled location in drawing marked "spray foam" ). could I fill this area (2x4 depth) with CC foam to air seal and insulate? Otherwise it would be tough to do any other way.

    Regarding the bay window... On top is pretty tight access. I think CC foam as you suggested is my best option. One concern I had was the foam distorting the 3/4 ply when it expands and cures. Should I be worried about this? Maybe make very light passes with the foam to fill it up slowly?

    I added an updated drawing that i came up with yesterday regarding the underside of the bay. I originally left out the 4x4 window supports that hold up the window to simplify things but added them now. I was thinking of adding 2" of foam on the underside of the bay and then building a 2x4 frame filled with roxul and a ply bottom to seal it all off.

    Looks like i got that backwards as well? Should I reverse that layout and keep the 2x4 frame with roxul directly under the 3/4 ply that is the bay window floor, then insulate under the 2x4 frame with the 2" of rigid foam?

    This is the foam I was considering using under the bay window. Would this also work for the kneewall?

  18. Expert Member
    Dana Dorsett | | #18

    J Loc: As I stated : "It's not a true moisture trap, but is way more vapor tight than ideal."

    If the wood is sufficiently dry when the foam goes on, it's only a problem if the roof leaks.

    "If I go with 5-5.5" of CC foam, Is there any option/suggestions to help with the thermal bridging?"

    There's nothing you can do to make the thermally bridging rafters less conductive other than to put insulation over the rafter edges. At a 7% framing fraction with 5" of closed cell foam you're looking at a U-factor of about U0.40. Code max is U0.30.

    Rather than 2" of wood, it's worth considering installing 4" of foil faced polyiso edge strips as a thermal break, quadrupling the R-value of the framing fraction, and use two lifts of open cell foam. Use 1x4 furring through-screwed to the rafters with pancake head timber screws for hanging the ceiling gypsum prior to the spraying the o.c. foam. That brings the center-cavity depth up to 3.5" + 4" + 0.75"= 8.25". Using R3.7/inch ocSPS would make the center-cavity R 30.5, but the framing fraction would be nearly identical. Adding another R1-ish for roof deck & ceiling gypsum brings it to R31.5, and another R1-ish for interior & exterior air films brings it up to R32.5 "whole-assembly R."

    A whole assembly R of R32.5 is a U-factor of U0.031, which is close enough to code max that only the most rigorous of inspectors would fail it.

    Use a "vapor barrier latex" on the ceiling gypsum, which will drop the vapor permeance to under 1 perm (a Class-II vapor retarder) but it wouldn't be much lower than 0.5 perms, which is plenty of drying capacity for your air-tight stackup.

    This would be both greener and cheaper than the 5-5.5" of closed cell foam, at the cost of at most 2.25" of headroom.

    For the 2x4 kneewalls, R15 fiberglass or rock wool would be fine. Rather than 2" edge strips with R23s, or wasting 2" of closed cell foam between studs, adding just 1.5" of continuous foil-faced polyiso on the attic side of the studs to provide the necessary air barrier to the "back" side of the batts would fully double 2x4 + fluff kneewall performance.

    There's a place for closed cell foam in home construction, but that place isn't between studs or rafters.

  19. user-1085290 | | #19

    Thanks for the added information Dana, It is all very helpful.
    I will be handling the kneewall and cathedral ceiling like you and Martin suggested.

    I just need to sort out insulating my bay window and this part of the house will be done.

  20. GBA Editor
    Martin Holladay | | #20

    J. Loc,
    Don't worry too much about the order of the layers under the bay window -- rigid foam above, or rigid foam below, won't make much of a difference. The main point is to pay close attention to airtightness.

    Yes, you can use spray foam above your attic kneewall.

  21. user-1085290 | | #21

    Thanks Martin!

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |