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

Insulation retrofit partial cathedral ceiling

Sheffield_Bodger | Posted in PassivHaus on

I’m remodeling a 125 year old folk Victorian house in northeast Ohio. Building zone 5. Rarely see colder then -5.
So far my wall system has 1.5″ of Eps foam and a rains screen followed by poplar wood siding. The wall cavity is insulated with r-13 faceless fiberglass batts. I’m using a vapor permeable paint on the interior. I’m comfortable with my wall system, it’s my roof that has been daunting to me. The upstairs had a low ceiling. I added collar ties to the rafters to raise the ceiling. I’m confident in the structural integrity of this. However, I now have 3′ of cathedral ceiling. It’s. 12 pitch with 5.5″ deep rafters. Which are at least 2″ wide. Because the majority of ceiling is attic, I planned on venting the roof, from eave to beyond the top of the blow in fiberglass. The plan was to use 1.5″ closed cell panels to act as baffles with 1.5″ of ventilation. Then to spray foam an additional 2″ of closed cell spray foam on bottom side of baffle. That brings me to roughly r-20. With the addition of 2″ Eps foam scabs, on bottom of rafters followed by 2″ of wood scabs parallel. After closed cell is cured, I want to fill the remainder with open cell, which would eliminate thermal bridging.
This all seems well, however, I have not reached optimal r value. Only about r-35.Is this a safe an effective system?!?
Any thoughts would be appreciated.
Also, the roof has new nailers spaced 1′ apart and tar paper covering, with a new metal roof.

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. walta100 | | #1

    First, I have to ask about your choice of poplar for your new siding from what I understand poplar generally does not survive long in outdoor applications.

    Sounds like the second floor is a half story. Half story retrofits are notorious for being almost impossible to air seal! The sloped part of the ceiling is very shallow and making room for insulation and roof ventilation means something is going to be compromised. The way I see it you simply do not have the room for a ventilation channel or the luxury of cheap fluffy insulation. As much as I hate spray foam you desperately need its one redeeming quality the fact the it packs the highest R value in the smallest space while blocking air flow. Ever packed full of CC foam you maybe forced to fer down the ceiling to get the R value you want.

    Did I say I hate spray foam, half story building and conditioned attics? Now I am going to prepose the trifecta of conditioning the attic of your half story home with spray foam.


  2. Sheffield_Bodger | | #2

    Thanks for the reply!
    Poplar does not have good rot resistant. However, it takes paint extremely well. With back priming, a 3/4" air gap and high quality paint, I don't have any concerns with that.
    The upstairs was a full upstairs. It was only 7' tall. Which is why I decided to raise the roof tie, which was sitting on wall plates, to 8' which is now connected to the rafters. So I only have 3' of cathedral ceiling the remaining ceiling is flat attic.
    No matter what, I need to add more depth to the rafters. Not sure what R value I should aim for if I have good ventilation and cover the living side of the rafters with insulation to reduce thermal bridging.

    1. Expert Member
      DCcontrarian | | #6

      Thermal bridging is the least of your worries.

      Moisture management is the big concern with ceilings under roofs. They need to be either ventilated or spray foamed. For your setup I would recommend ventilation. Then the problem becomes getting enough thickness to fit the vent channel and the insulation. You have 2x6 rafters, you're going to need about 14 inches of thickness -- only on the sloped section.

      If you were to extend the underside of the rafters by nine inches, does your plan still work?

      1. Sheffield_Bodger | | #7

        I'm planning to ventilate. If you look at my original post, I've outlined how I intend to do this. Do you think I need more r value?

      2. Sheffield_Bodger | | #8

        If I use 1.5"losed cell boards followed by 2" of closed cell foam that would come close to the depth of my existing rafter. I would then fasten 2"+ of open cell scabs parallel to the rafters and then 2"+ of wood perpendicular to rafters. Fill that with open cell foam. This assembly would be vented and also allow inward drying and give a thermal break. The only thermal bridging would be from the collar ties.
        Does this seem sound?

  3. walta100 | | #3

    “Not sure what R value I should aim for”
    Seems like R38 would likely be code min value and above R60 generally has no ROI.

    The big problem with half stories is that the conditioned air leaks out faster than you can make it. Your R value is irrelevant when the conditioned air simply blows thru and around the insulation. The point where the upstairs floor ends and a vented attic begins is the air sealing challenge.

    You might find something interesting in these articles.


    1. Sheffield_Bodger | | #4

      You seem to be very educated in these matters. The one thing we are not on the same page with, is the overall framing design. This is not a cape cod half story. It's a balloon frame house with exterior walls extending 7' above the second floor. I raised said 7' ceiling to 8' by raising the rafter or roof tie up from the wall plate to 16" higher up the rafter. This changed loading which was appropriately dealt with.
      The issue being is how and where to achieve an appropriate air barrier and what insulation method should I use for the cathedral section of this ceiling.
      Of a 18' wide building there is, because of raising the rafter tie, that leaves about 3' of cathedral ceiling on each eave side. 12 pitch ceiling for 3', then 14' of flat ceiling and then another 3' of 12 pitch. I don't know how to get and appropriate air barrier (and where that should be) and the proper type and sequence of insulation.
      I'm now worried about the thermal bridging of the rafter tie.
      I have a lot of confidence with wood movement and structural integrity. However, I'm struggling with this..

      1. walta100 | | #29

        “The one thing we are not on the same page with, is the overall framing design. This is not a cape cod half story.”
        I am liking your house more already.

        I still dislike spray foam so try to avoid it if at all possible.

        Did you consider the 5 cathedral ceilings from the article?

        Are you willing to increase the depth of the sloped part of the ceiling to get a reasonable R value without spray foam?

        Seems the only place left for the air barrier is the surface of the drywall or just behind it. After the drywall is taped and before the trim gets installed you should consider having a blower door directed air sealing work done.

        The way I see it thermal bridging with wood framing over 5.5 inches depth is not a big issue if your R value goal is reasonable.


  4. Sheffield_Bodger | | #5

    Another way to explain my ignorance.
    If I vent this, which I would like to, does open cell make more sense then closed? So I still need a vapor barrier?
    Any suggestions are appreciated.
    I'll draw up things if it could help.

  5. Expert Member
    DCcontrarian | | #9

    OK, read the original post more closely this time. Here's what I'm thinking: a 2x6 nominal batt of rock wool gets you R-23, 2.5 inches of polyiso gets you another R15, that's R38 total.

    Put a spacer between the roof sheathing and one layer of polyiso, then the rockwool, then a second layer of polyiso. However thick the spacer plus the first layer is, rip a piece of 2x to that width and screw it to the underside of the rafters. So if you did a 1.5" spacer and 1" of polyiso, you'd use a 2.5" rip. Attach the rip with 3.5" star drive deck screws.

    However thick the first layer of polyiso is, make the second layer enough to get to 2.5" total. The second layer is continuous across the rafters.

    Over the second layer run pieces of 1x3, 16" on center, perpendicular to the rafters. Screw them to the rafters through the second layer with long drywall screws, 3.5" if you have 2" of foam in the second layer. Tape the seams of the second layer and that's your vapor barrier and air barrier.

    I don't have strong feelings about the thickness of the first layer vs. the second layer of foam. The spacer could be either 1" or 1.5". The total thickness of the spacer plus the first layer shouldn't be more than 2.5" or it's going to be difficult to attach the rip of 2x. I could see a case for anything from 1/2" to 1-1/2" on the first layer. I guess if you did 1/2" for the first layer that makes the second layer 2", which makes more of the insulation continuous where it does more good. Also gives you a thinner rip which is easier to work with. But I think it may come down to what sizes of polyiso are locally available, I've had difficulty recently getting all sizes.

    With 1/2" as the first layer you may not even need spacers, at least not in the section behind the rock wool. The rock wool will be a friction fit between the rafters and can hold the foam in place.

  6. Sheffield_Bodger | | #10

    Thanks so much!
    You don't have any concerns sandwiching foam on both sides of the rock wool?
    Do you have any particular concerns with the method I outlined?
    Really appreciate your advice.

    1. Expert Member
      DCcontrarian | | #11

      Spray foam has some practical issues. For small areas like this it's expensive and hard to get someone to do. Closed cell foam doesn't generally form an even layer so it's hard to ensure you have the thickness you want.

      What I laid out is something that can be done by a person of ordinary ability with ordinary tools and materials that are readily available.

  7. Sheffield_Bodger | | #12

    Got ya. Very sound advice. However, I happen to have a friend who has been doing spray foam for years, who has no problem doing a small job.
    I certainly don't think spray foam is a cure all. I believe it's value is in jobs similar to mine.
    Last time I used spray foam was on a low pitch roof.
    The roof was over a small bump out and another over a bathroom. Both roofs where abutting an exterior sheared wall. I put two layers of staggered 1.5" of foam boards over the rafters, then I screwed plywood through the foam into the rafters. Followed by 2" of closed cell on bottom side of foam. The used batts to finish the assembly. Thoughts on that?

    1. Expert Member
      DCcontrarian | | #18

      Even if you use spray foam instead of rock wool, I would still:
      1. have a vent channel against the sheathing
      2. fur the rafters with 2x rips to get more thickness
      3. run continuous insulation across the rafters
      4. run 1x3 strapping over the continuous insulation to hold the drywall

  8. Expert Member
    Akos | | #13

    Take a look at this:

    In your case since you want more R value, I would use rigid insulation for the vapor/air barrier on the low section and continue it for the sloped roof section. You can strap out the rigid with 1x4 and install your interior drywall over that.

    1. Expert Member
      MALCOLM TAYLOR | | #15


      I like it, but I don't think it's been accepted as a prescriptive method yet, so would need an engineer to sign off.

      1. Expert Member
        Akos | | #19

        You can always use batts and build it with 1.5" gap and make up the R value delta with more rigid insulation. That would be perspective compliant for not much overall cost increase or labor.

        Air sealing at the various intersections would still be a PIA, this is why these types of roofs generally get spray foam in the end.

  9. Sheffield_Bodger | | #14

    My concerns about having a hard air barrier on the low side, is sealing around all the walls. It's easy on the slope. However, continuing that through all the walls on the flat section, could be troublesome.
    My proposed method leaves the rafter tie as a substantial thermal bridge. Which I have variable concerns with. I definitely agree that your method should work. It does make bathroom fan and chimney penetrations to be less the desired.
    There is definitely not an absolute best scenario here.
    I however am certain someone will suggest that unicorn.

    1. Expert Member
      Akos | | #17

      Are you knee walls structural? If not I would remove them and do the insulation and air sealing work. This would make the whole side a single plane which is pretty easy to detail.

      Your only ugly intersection now is at the ceiling/floor joist. That can be done by notching the rigid and sealing it with canned foam or with spray foam.

      Chimney is always complicated, unless in use, the best thing is to remove it and patch the roof while doing the work.

      Once done, the knee walls can be reinstalled over the rigid. You can also leave it out or build shorty ones as it will make the space feel much bigger.

      1. Sheffield_Bodger | | #20

        There aren't any knee walls. Just 7' exterior walls. I raised the rafter tie up to 8' to increase ceiling height. I agree the rafter tie which is also the ceiling joint is a troublesome connection.
        About 10" of it will be encapsulated with foam.

        1. Expert Member
          Akos | | #21

          Ahh, that is much simpler. In that case you are overthinking it.

          Stuff the rafters with batts or spray foam. Either will work for a small section like that. If you are using batts, make sure the top is open to the attic space to allow for a path for drying through diffusion and use a variable perm membrane for the vapor retarder.

          Adding any rigid in the mix is more work and will barely budge the house energy use or effect the comfort of the place.

          The important detail for comfort and energy use here is air barrier continuity.

          In your case, the simplest would be the vapor retarder. This should be one continuous layer from the bottom of the wall, over the sloped section, flat ceiling and back down. Seams taped and edges sealed with acoustic caulk.

          1. Sheffield_Bodger | | #22

            I was avoiding vapor barriers on the wall because of the exterior insulation.
            I'm most certainly guilty of overthinking this.
            Thanks for all the advice.

  10. Sheffield_Bodger | | #16

    I value engineers. I am married to so much of builders math. I would like the common thoughts on condensation and air sealing to be on the mind of all builders detailing these systems. If all involved knows the imperativness of these details and the science behind, we would all be better off.
    Even the worst and dumbest builders acknowledges the need for footers and basic insulation. We just need to raise the bar and expand the education.
    After all. Insulating houses is a very new thing as far as building dwellings go.

    Thanks for all the advice.

  11. Expert Member
    Michael Maines | | #23

    I would keep it simple: 1" vent space, 4 1/2" of insulation of your choice, 2" of continuous polyiso on the interior. For small, tricky areas like yours, anything over R-30 is probably good enough, though of course more would be better.

    I echo Walta's concern about using poplar outdoors. There aren't many LESS rot-resistant woods. I once built a storm/screen door from poplar, primed and painted all surfaces, and it lasted less than ten years.

    I am also not as confident in your raised rafter ties (not collar ties, which are by definition near the ridge). I have seen many roofs with kinks of 1-3", in once case almost 6", from similarly placed rafter ties. People under-estimate the load they put on rafters. They also significantly under-estimate the number of fasteners needed--see table R802.5.2:

  12. Sheffield_Bodger | | #24

    Polar siding has been used for many years. There are lots of buildings here sided with poplar. I do understand it doesn't have rot resistance. It's straight stable when dry and takes on paint very well. With the rain screen and back priming and 2 coats of paint, which was applied when the wood was 12-14% mc, I feel pretty confident. I have an out building done like this without a rain screen and also no insulation that shows no indication of rot. According to Dr. Gene Wengert, insects are more of a concern then rot.

    The rafter ties where only raised up 16" and are new 2x10. Nailed to the sides of the existing rafters. Underneath rafter ties is another new rafter leg which is attached to the old rafters. From there I have plywood gussets stapled connecting all members together. Hurricane clips have been added as well.
    In the main attic, I installed collar ties, which were glued and stapled to connect to rafters. Also some plywood gussets to connect tops of rafters together.
    The last things mentioned were done to help with wind load. I understand that I moved the load point up from the rafter foot to 16" higher. Which is why I did the extra steps.

    Are we talking about the same species. Yellow or Tulip poplar Liriodendron Tulipifera Not a Populus species.
    I was quite comfortable with my siding choice, up until yesterday... As with most wood that is outside, if it has the ability to dry easily and has an appropriate paint job, I think ( now pray) that it will be fine

    1. Expert Member
      Michael Maines | | #26

      Yes, tulip poplar, not populus. I have to be extra cautious with my material recommendations because I design homes for others, but if it's a locally common material and your own home, with a ventilated rain screen and a well-done paint job, it might be just fine.

      1. Sheffield_Bodger | | #27

        Thanks for all your valuable input.
        Whats your company name?

      2. Sheffield_Bodger | | #28

        If you're interested in some professional consultation, I'd be interested in paying for further advice. I appreciate all that you and the others have said.

  13. Sheffield_Bodger | | #25

    The Collar ties were installed in plane with the rafters at a 45⁰ angle and screwed into bottom of rafters. The plywood gussets were glued and stapled to both sides of that connection.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |