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 on addition

Howard Gentler | Posted in Energy Efficiency and Durability on

I’ve been procrastinating on how to insulate, and need to decide soon.  I’d like to run this by GBA folks for input.  This is a 16′ square addition with a gable roof (metal with a 6/12 pitch in zone 6) with no obstructions, dormers, etc. An idea now is as follows:
3″ of closed cell foam against the roof deck, then fiberglass batt below that.  I could do a 6″ batt, but since that combination would create an R value of only about R-37, I thought of rigid foam below the batts/rafters.  Using 2″ polyiso would give me about another R-12/13 (for a total of R-50 or so), and taped well would provide a good air barrier.  I think this layer would need to be somewhat permeable to allow drying to the inside, so if polyiso with a permeable facing would not work I could use unfaced EPS with an R of about 8.4.

Is this legit?  If the rigid foam in that location is a no-no, I could go with R-30 (9.5″ thick) fiberglass batts for R-48 with the air barrier being taped sheetrock.

I’m interested in avoiding exterior rigid foam or a number of reasons.

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. flatrooftrouble | | #1

    I'm curious... how do you plan on securing the polyiso. i'm visualizing spray foam 3" then 6" batt would stick out passed the framing a few inches. then somehow a layer of polyiso. Like maybe long screws with some kind of spacer to keep them from compressing the batt?

  2. Expert Member
    Dana Dorsett | | #2

    With a full sheet of polyiso on the interior side it creates a moisture trap for the fiber insulation layer, and the average temperature at the spray foam/fiber boundary is cold enough to condense (or frost) in winter. In zone 6 at least half the center-cavity R to be closed cell foam.

    At 4" of HFO blown closed cell runs about R27-R28. A 5.5" deep rock wool batt is good for R23. That's 9.5 and a total R of R50+, and more than half the R is the spray foam. But a 2x10 is only 9.25" deep. Compressing the batt to 5.25" to make it fit drops the center cavity R to R49-R50, and a slightly higher foam ratio. This assembly could use standard latex ceiling paint as the vapor retarder.

    If you want to use R25 fiberglass, 4" of HFO blown foam a of a full layer of foam on the interior installing Bonfiglioni edge strips sufficient to deepen the total cavity depth to 11-11.25" would also get you there. A strip made of 1" polyiso and 1x furring would add 1.75" of depth, for a total of 11". With 4" of foam you'd have 7-7.25" of space for the R25 fiberglass, which performs at R24 @ 7.25", R23 @ 7.0".

    This assembly too only needs standard latex ceiling paint as the vapor retarder.

    If using 4" of (climate damaging) HFC blown foam the foam layer is only R24-ish, so using high density fiberglass R21s (compresssed to 5.25" is ~R20) would give a bit more dew point margin than R23 rock wool (~R22 @ 5.25"), though compressed R23s would technically still make it. With the Bonfiglioni strip approach and ~R24 spray foam go with only 3/4" foam for the edge strips, adding only 1.5" of total depth, not 1.75".

  3. Howard Gentler | | #3

    Flatroof: securing the polyiso is easy. I've done it with adequate length cap nails temporarily. We were reinstalling t & g pine boards over the polyiso and used long enough panel type screws in the tongue, holding both. Furring strips would be appropriate if using sheetrock, but one could probably use long sheetrock screws (without furring strips) to support both sheetrock (though I bet sheetrock installers prefer furring strips) and polyiso.

    Dana: It sounds like you feel the polyiso below the batt is a no-no, but other options you suggest seem workable. I know there are options for adjusting the rafter depth, either by adding strips as you suggest, or with slight compression, which does not seem to result in much of an R hit. Thanks for your response.

    1. Expert Member
      Dana Dorsett | | #5

      You're welcome, and you're right. Putting vapor barrier on the interior side of an UNVENTED roof assembly is asking for trouble. It's fine if it's all dry when assembled, and never leaks air or water into the space between the spray foam and the interior foam, and most of the roof would probably stay dry for a long time. But true perfection on air sealing or water leakage even in the near term (let alone over the lifecycle of the building) is a lot to ask. Leaving the ceiling relatively vapor open and putting the first condensing plane (the spray foam/fiber boundary) where it's average winter temp will be no cooler than ~40F is a lot safer.

      Compressing most mid and high density batts by 1/4" takes about an R1 hit from it's labeled R. For low density batts it can be even less than a R0.5 hit. Very low density R19s (everybody's favorite punching bag when criticizing fiberglass & batts :-) ) are manufactured at 6.25" thickness and test at the labeled R19 at that thickness. But they perform at R18 when compressed to 5.5" in a 2x6 bay. That's about an R0.3 hit for every 1/4" of compression from it's rated tested thickness. Compressing it to 3.5" it performance at R13, fully 2.75" thinner than it's tested value. That's about an R0.55 hit for every quarter inch, but really the last quarter inch from 3.75" to 3.5" is closer to an R1 hit, since the density is already considerably higher- it's not really linear over large compression ranges.

      The foam in a Bonfiglioni strip provides a significant thermal break on the framing fraction, which improves the overall performance more than merely the center-cavity R. Just as putting high R/inch between low R/inch framing is disproportionately robbed of it of it's potential by the thermal bridging, putting high R/inch insulation over the framing elements as a thermal break disproportionally improves the whole-assembly performance. At 9.25" a typical 2x10 has an R value of about R11. Adding an inch of polyiso and a 3/4" thick 1x furring boosts that to about R18, whereas adding 1.75" of solid wood to achieve the same depth would only bring it to R13. That's nearly a 30% reduction in the thermal bridging at the same depth.

  4. Expert Member
    Zephyr7 | | #4

    I just did a reroof myself similar to, but slightly larger than, your project. Assuming you are using a spray foam contractor to apply closed cell spray foam and not one of the little DIY kits, you may find that it’s cheaper to just have the contractor apply a thicker layer of spray foam to get up to the R value you want. I went for R38, or a bit over (all I could fit in the roof assembly) and the extra cost over a thin layer of foam and the labor/materials to install batts was worth it.

    If you do go with a layer of spray foam and then batts inside, I’d look at mineral wool batts instead of fiberglass. You can get a bit more R per inch that way, and they’re easier to handle.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |