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

Cathedral Ceiling Insulation Sandwich Plan

RickTeachey | Posted in Energy Efficiency and Durability on

Hello: been lurking on this site for years an it has been INCREDIBLY helpful– THANK YOU.

Based on all my reading and gathered knowledge, I believe the below plan will work for rehabbing the uninsulated, very leaky attic of my 125 yr old house. I am shooting to get R-40 but would be happy-ish with about R-36 (I’m in Dayton, OH).

I plan to insulate the entire roof using a cathedral style approach. All the roof slopes are 12:12. There are 3 gable ends and one hip roof end — two ridges meeting at a right angle at the top.

I plan to insulated the whole thing cathedral style because even though the attic is finished in the style similar to a cape cod, and I could theoretically go on top of the ceiling with batts or cellulose, I really want to have a nice attic to work in because have a lot of electrical and HVAC work ahead of me, and long term the HVAC will be in the “half attic” above the finished ceiling. Of course I will be removing the currently uninsulated drywall in the cathedral portion of the ceilings to do this, and will be insulating behind the kneewalls as well. But I’m willing to do that.

Here’s my “sandwich” design:

Shingles (existing)
Sheathing (existing)
Air Gap (0.5″) – open to vents top and bottom of rafter bays
Tyvek or similar vapor permeable air barrier and radiant barrier
5.5″ high density fiberglass batts (but compressed to 5.0″)
— BOTTOM OF 2X6 rafters —
3.5″ foil-faced polyiso (Firestone fire-rated, taped at seams)
Furring strips

Questions about this plan:
1. Is it a good plan? My main concern here is I am a little scared of getting condensation against the air/radiant barrier under the sheathing. Should I be concerned about this if I have good air continuity above the barrier, and if the polyiso foam is taped and all penetrations sealed well, and the air gap goes all the way down to the soffits?

2. Since I am compressing my bats against the top foil face, I am not benefitting from the radiant barrier. Would I be better off using 3.5″ batts with the reduced R value, but now with an air gap between the batts and foam so that the top polyiso radiant barrier can work properly?

3. If this is a good design, are there any suggestions on a good product for the vapor-permeable air/radiant barrier? Does such a product exist? Should I be considering using baffles instead? Are they vapor permeable?


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
    BILL WICHERS | | #1

    First, code minimum is 1" for a ventilation space in a roof assembly. Many (including myself) prefer a bit more than that, around 1.5". Your proposed 1/2" is too small. Aside from that, your assembly looks OK except for the following things you may want to consider:

    1- Tyvek isn't a radiant barrier. You'd need perforated metalized poly here if you want a radiant barrier. The perforated part is important to make it vapor permeable. There are vapor permeable radiant barrier products on the market. You're going to want something somewhat rigid here though, since you need something the batts can be compressed against that won't bow too much and block your ventilation channel. I would use thin plywood here, 1/4 or 3/8". You don't need a lot of strength.

    2- The extra insulation is more useful here than a radiant barrier, especially since you might already have a radiant barrier on the top of the insulation stackup (see #1, above). You don't need a radiant barrier in the middle of the assembly. The foil facing of the polyiso has the gap provided by the furring strips between the polyiso and the drywall, so you have an interior side radiant barrier already too.

    3- Yes, these products exist. A quick Google search to get an example comes up with this:
    I believe LP also makes a sheathing product with integral radiant barrier that could potentially be used for the "thin plywood" I mentioned in #1, above, which would get you your rigid seperator and radiant barrier in one piece. If this product isn't super pricey, and is vapor open (I've never used it), I'd use that instead of trying to make something up with seperate plywood and radiant barrier leayers since it would save a step and make the labor easier.

    Note that you shouldn't get too hung up on radiant barriers, the insulation itself is a bigger benefit to overall thermal performance of the assembly.


  2. Expert Member
    Akos | | #2

    I would skip any radiant barrier or baffles and just go with either HD or mineral wool 2x4 batts. These are pretty stiff and will stay in place if installed with a bit of care leaving the vent space open.

    Most of your R value comes from the 3.5" polyiso. An extra inch or two of fluffy will barely effect your overall assembly R value never mind actual heat loss.

    Keep it simple.

  3. RickTeachey | | #3

    Akos: I considered doing what you describe, but after reading about wind washing, I decided I need an air barrier between the to of the batts and the sheathing. Am I wrong?

    1. Expert Member
      Akos | | #5

      Take a look at this:

      Regular comfortbatt looses very little R value due to wind washing. FG batts loose a bit more but within the noise overall take into account assembly R value. Focus your extra effort on air sealing, much higher gains there.

      My own home has pretty close to your stackup in part of the roof (2x6 mw batts in 2x8 rafter+2.5 rigid poyiso on the interior). No baffles just the batts installed with a gap above.

      1. RickTeachey | | #8

        I am now leaning towards this plan.

        I did a sort of desktop energy audit last night for my house and concluded, based on recent average temperature and energy losses and assuming about 85% of energy loss is out of my attic (rather than through my new, pretty tight windows and doors and brick walls) that my attic R-value is currently about 8. It could be as high as maybe 12 if my energy loss out the attic is less than I think.

        I have concluded that the overwhelming portion of energy usage reduction will be present even if my new R-value or more like R-38 (15 batts, 21 foam, 2 air gap at sheetrock) rather than R-42 (19 batts, 21 foam, 2 air gap at sheetrock):

        1 - 8/38 = 78.9% reduction in conductive heat loss
        1 - 8/42 = 80.9% reduction in conductive heat loss

        That extra 2.0% would obviously be nice, but it's far from critical. I'm going to go ahead with the air gap.

        And even if I do have some wind washing, making the batts totally useless (obvioudly dubious, but this would leave a total of R-23), my conductive heat loss reduction is still:

        1 - 8/23 = 65.2%

        And even if my current R is more like 12 (also dubious):

        1 - 12/23 = 47.8%

        But even in that case, that 47.8% reduction is only the conductive heat loss and FAR FAR more of my current heat loss is via convection (focus on air sealing, as you said). So I'm probably still well above a 70% reduction for the entire project I'd say.

        Thanks for assisting me with thinking through this!

  4. RickTeachey | | #4

    Thanks, Zerphyr!

    I like the thin wood idea-- seems so easy. So the wood isn't going to significantly retard water vapor and make me want to jump out my attic window one day?

    1. Expert Member
      BILL WICHERS | | #7

      No, the plywood is pretty vapor open.

      Note that Akos' idea of using high density batts and not compressing them makes the rigidity of that barrier much less of an issue. I personally would still want a seperator between the air space and the insulation, but it is true that high density batts (especially mineral wool) are much less affected by air currents than are lower density insulating materials.


      1. RickTeachey | | #9

        Thanks for helping me think through it! I think I am going to go with the 2X4 batt insulation with no air barrier, and a total R-value of 38 rather than the 42 I wa shooting for. I think the effort compared with the relative return on investment just isn't there, especially since I am paying to have all of this done.

  5. Jon_R | | #6

    Always good to read code. Both local and the IRC - in this case 2018 section R806.

    I agree with using highly vapor permeable materials to create vents. Or nothing at all. Low perm plastic vents - let's see the data showing that they don't excessively inhibit moisture leaving the cavity.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |