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

Increasing the R-Value of an existing cathedral ceiling

andy.kuss | Posted in Energy Efficiency and Durability on

When attempting to increase the R-value of a poorly performing cathedral ceiling, what is the most cost-effective interior application? There are exposed rustic wooden rafters every 4 feet or so that protrude from the ceiling 6″.

My thought was to apply 2″ rigid foam to the interior with 1X3 furring strips in between existing ceiling and the rigid board taking extra care to make the ceiling airtight prior to installation and then taking great care to make sure the rigid board is airtight with spray foam around perimeter and seams. Lastly, would be applying drywall to the rigid board.

The cheapest (and possibly the best thermal performing) method seems like it would be filling the large cavities between the exposed rafters with unfaced high-density fiberglass batts and then running the rigid foam board and drywall over the rafters. You would obviously lose the look and feel of the exposed rafters, but if the homeowner is willing to part with that look in hopes to achieve comfort and energy savings without paying out the wazoo for spray foam then wouldn’t this be a good option?

My main concern is moisture build up within the ceiling assembly. Would I want to use EPS rather than polyiso. Being in Ohio at that, where it is hot and humid in the summer and cold and dry during the winter it seems like let the ceiling breathe somehow would be important. The roof was replaced within the last couple years.

Thanks for any insights!

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

    We need to know more information.

    You mention that you are in Ohio (Climate Zone 5), so we know that building codes (at least the 2012 IRC) require a minimum of R-49 ceiling insulation in your location. But you didn't tell us whether your existing ceiling has any insulation, or, if so, how much.

  2. andy.kuss | | #2

    This ceiling has insulation, sorry:) Based off IR scan and exterior assessment of the framing rafters, as well as the year of construction (late 70s), I am assuming a maximum R value of R-26 in the closed cathedral ceiling. It is inaccessible, so short of drilling test holes it hard to say exactly what is in there. Another X factor is that I don't know if it is vented. If it is, it is safe to say that isn't done properly with vent channels. I think we could dense pack the cavities with cellulose even though there appears to FB batts in there already, but that doesn't seem ideal and doesn't address thermal bridging. thanks!

  3. Expert Member
    Dana Dorsett | | #3

    It's worth drilling those holes, if you're planning full-width sheet foam, since the foam may reduce the drying capacity toward the interior. There could be venting or not, and there could be poly vapor barriers- you simply don't know. It would be best to fully gut the ceiling so you can see what's really going on there, but that may not be in the cards.

    Assuming that's not happening, rather than a full-width foam solution which may inhibit drying toward the interior, snugly install continuous R23 rock wool (provided it makes the full depth of the exposed rafter beams) snugged up to original gypsum, and a layer of Certainteed Membrain as an interior side air-barrier/vapor retarder for the fiber, with 1x furring on the rafter beams on which to install the new ceiling gypsum layer.

    The variable vapor permeance of MemBrain means it won't create a moisture trap- the roof deck will still be able to dry as quickly toward the interior as it could previously, but moisture diffusion from the interior toward the roof deck in winter will still be less than it would be through 2" of EPS. With something like 50% of the total-R on the exterior of the original ceilng gypsum that gypsum layer (which becomes the first condensing surface) not going to become wet even if you didn't have the vapor retarder. But if unvented the roof deck could still be taking on excessive moisture burdens without the vapor retarder. It's cheap stuff, way cheaper (and more effective) than 2" of EPS that you may more may not be able to air seal adequately.

    If R23 rock wool batts don't have sufficient loft to fully fill the depth, go with a higher loft but unfaced product like a low-density R22 which will still hit about R20-R21 when compressed to about 6", from it's manufactured 6.5-6.75" loft at which it's R22 was measured. R23 rock wool an R21HD fiberglass batts really are only about 5.5", so if it's really 6", use something else.

    The R49 code minimum prescriptive is based on the whole-assembly U-factors at typical framing fractions with 16-24" o.c. joists or rafters, not 48" o.c. beams of unspecified width. If you knew what was on the other side of the ceiling gypsum and the width of the beams it's possible that it would make code on U-factor basis, even if it wasn't quite R49 at center cavity.

    If you feel you need the thermal break at the rafter beams, use 1.5" polyiso strips on the bottom side of the beams, and fill in the space between with split R15 batts snugged up to each other (and to the polyiso edges), keeping the vapor retarder at the beam-edge level. Make the strips wide enough to overhang the beam edges by 3/4" or so on each side to get the full benefit of the R9-ish thermal break. With 1x furring 16-14" o.c. through-screwed through the foam strips to the rafter beam edges you should be able to install the split batts in there restrained by the furring without too much fuss. Tack the polyiso in place with cap nails, then add the furring as you add the split batts. With 23" wide x 47" long batts (typical for rock wool) you probably won't even have to do much batt trimming, running the split batts parallel to the rafters. (Depends on your exact dimensions and how much overhang you give the polyiso.)

  4. GBA Editor
    Martin Holladay | | #4

    Dana Dorsett's recommendations are based on assumptions about vapor diffusion and air leakage that are (apparently) based on WUFI modeling. Not all of Dana's recommendations are code-approved, and the performance of some of his recommendations is speculative rather than verified by field experience.

    The code allows insulated cathedral ceilings to be vented or unvented. If you have a vented ceiling, it has to really be vented -- meaning that you need soffit vents, clear air channels all the way to the ridge, and a ridge vent. You don't know whether you have that -- and until you determine what you have, it's impossible to give you good advice.

    Once you know more about the venting situation, it will be easier to give you advice. In the meantime, you make want to read this article: How to Build an Insulated Cathedral Ceiling.

  5. peaceonearth | | #5


    I have a similar situation (son's house actually) and have asked questions and gotten much advice here. We know more about the details than you do, but I think it is good advice to gut the rafter bays if you can. We will soon be doing that from below.

    It's hard to imagine no venting since everyone was doing it back then, but that is just an opinion. Many things were done horribly wrong even by professional builders. You should be able to see soffit vents and either a ridge or gable end vents.

    The wood members below the ceiling might be decorative, as ours are. They are toenailed into the pine v-groove ceiling boards. Some of them are chases for electricity for surface lighting.

    Ours is vented, but once we gut the rafter bays we will be assuring a 2" vent channel by using foam board (I've found 1" unfaced EPS for that area). We'll be putting back the 12" fiberglass batts (we have 13 or so inches of rafter depth) to the bottom of the rafters, then do full sheet foam beneath the rafters. We plan on using foil faced 2" polyiso secured with furring strips. The v-groove boards go back on attached to the furring strips, providing a reflective scenario for 2-3 extra R.

    Keep asking questions as you get more detail on your stack up. These guys know what they are doing.

  6. peaceonearth | | #6

    Correction: you couldn't have gable end vents in a cathedral ceiling. Guess I was thinking of my own house. You should be able to tell if you have a vented system, but march harder to determine, I think Martin would agree, if it is working as intended.

  7. Expert Member
    Dana Dorsett | | #7

    To be sure, I was recommending pulling it all down to know what's really up there, " It would be best to fully gut the ceiling so you can see what's really going on there, but that may not be in the cards."

    The rest is basically a primer on how to not make it worse than it already is, if in fact it's not all coming down. The original concept of 2" EPS + batts carries more risk. When you don't really know if it's vented or not, or whether & what pre-existing vapor retarders are in there, it's imprudent to decrease the drying potential toward the interior, but it IS prudent to use a more reliable air barrier such as 8-12' wide sheet membranes), and retarding the moisture uptake rates, which is what a more responsive & fairly vapor-open smart vapor retarder like MemBrain provides. Intello Plus would probably work too, but it's much more vapor-tight than MemBrain at humidity levels that can support mold growth, and would be more vapor tight than a perfectly sealed 2" layer of EPS. Air sealing perfection over the long term isn't very likely with sheet EPS due to both seasonal & long term dimensional changes, but is pretty reliable with flexible plastic sheet goods.

    A refining amendment: With 4' spans between the rafter beams it's probably much better to use 2x furring than 1x, and to screw them to the rafters with pancake head timber screws. With 1x furring spanning 4' the ceiling won't be very flat. With any furring with that few fasteners bearing the weight there's a good chance of it splitting or pulling over time if using fasteners of lesser load bearing surface. If it were 16" o.c. rafters ring-shank nails would suffice, but that's 3x the number of fastener points than you get with 48" spans. With 2x furring you're also less likely to make air leaks in the vapor retarder with the gypsum fasteners, which would have to be pretty stubby to work with 1x furring.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |