GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Audio Play Icon Headphones Icon Plus Icon Minus Icon Check Icon Print Icon Picture icon Single Arrow Icon Double Arrow Icon Hamburger Icon TV Icon Close Icon Sorted Hamburger/Search Icon

Community and Q&A

Conflicting insulation advice from spray foam installers

Shawn Sanders | Posted in General Questions on

I’m trying to determine the best solution for insulating our conditioned attic/unvented roof/cathedral ceiling and our walls. I have read so many articles/blogs/reports on GBA and from Building Science that my head is spinning. I know more about Joe Lstiburek then I care to, and I realize that fights can break out in blogs over who knows more about vapor drive. I’m probably staring my own theory war by posting this, but here goes. This will long, but I want to cover the details of the build upfront.

Climate Zone 4 Mixed, but very close to the line of Zone 5 – Owen County. We have very large seasonal swings in temperature and high humidity during parts of the year. It can reach 0 degrees F and on a few occasions even less during the winter. We have highs in the mid to upper 90s in the summer, and occasionally hitting 100 or more can happen

Current Build Status:
Framing done and dried in (windows/doors not yet sealed inside with canned spray foam)

Roof Stack-up (conditioned attic portion):
2×8 Trusses
3/4″ OSB sheathing
Feltex synthetic underlayment
8:12 pitch standing seam metal roof – 0 penetrations through the roof i.e. no chimneys/vents/air stacks/etc

Roof Stack-up (cathedral ceiling portion):
2×12 rafters
same as above

Wall stack-up:
2×6 wood stud wall
1/2″ OSB
1″ XPS Foam
Tyvek Commercial D Wrap
DC-14 DrainMat by Green Guard (provides and R value of 1 according to manufacturer)

Top Half
1/2″ Rough-sawn Plywood with 3/4″ Kiln dried Poplar spaced 1.5″ to 2″. (looks like reverse board and batton but uses less wood)
Bottom Half
Jumbo tex and wire mesh over DC-14, then Coronado manufactured stone

8″ poured concrete walls on 3 sides (walk out)
14″ Floor Joist on top of basement walls
1″ xps foam from footer on the same plane as the wall 1″ XPS
An additional 1.5″ xps foam from top of concrete wall 4′ down (4×8 sheets turned sideways) to better insulate near the frost line

Walk-out Wall of basement
same as exterior 2×6 walls above

Now to my question, how to insulate from here on out? I feel I have addressed thermal bridging in the exterior walls but I have no exterior foam on the roof. I wanted 4.5″ of XPS, but the contractor could not be convinced to do that much rigid foam on an 8:12 pitch – he also felt the thermal bridging in the rafters would result in very little energy loss anyway. I held my ground on the walls, at least. He also suggested a cathedralized attic to make use of the otherwise wasted space in the portion of the home that would not have cathedral ceilings.

In order to do that, we planned to do spray foam directly under the roof deck. 1 of the 3 installers in my area refuses to use open cell foam, 1 will only use open cell, and the other says he recommends open, but will do closed cell if I sign a waiver for water damage due to a roof leak!!

Here is what I wanted to do:
4-5″ ccSPF under the roof deck
BIBS fiberglass (or installing batts myself) in the remainder of the rafters
1″ of foil faced poly foam attached to the bottom side of the rafters
T&G ceder for the ceiling in the cathedral section and dry wall in the bonus attic room section
NOTE: I realize I will not be able to insulate the 8″ trusses in the cathedralized attic as much until I add scabs on to match the 12″ rafters of the cathedral ceiling section.

Walls: BIBS fiberglass, Roxul, or Open Cell foam

Floor Joists: 3″ ccSPF followed by batts/roxul

So, how would you insulate this build? I’ve done so much researching and reading that my head is clogged with conflicting ideas. I think I would have preferred to NOT do spray foam if I had it to do over, but that is what the builder suggested before I began researching so much. With an unvented roof, it’s really about my only option at this point. Not that spray foam can’t be done right, it’s just that it seems to be a bit riskier and we still don’t know all the effects it MAY have down the road.


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. Stephen E | | #1


  2. Stephen E | | #2

    Should have used scissor trusses. Would have given you cathedral ceilings, cheaper cellouse insulation and cost less long term. Even I joist at 11 7/8 would have been better. That aside might have a solution. Use roxul between 2x8 and 2x12 joist. Mattering on the stage of build put foam panels on top of roof and strap with boards. Connect metal roof to straps.

  3. Shawn Sanders | | #3

    Thanks for the tips, but I am already past that stage. See the attached PDF for a look at the build. Hindsight is 20 20 so I'm looking for answers that look forward based on my current situation.

  4. Joe Suhrada | | #4

    As to the closed cell option and the guy wanting you to sign a waiver for roof leaks: what are you concerned about as to your roof leaking, could I ask? Is this a snap lock roof or a mechanical seamed roof? You said the pitch was 8/12 which is rather healthy... Are you unsure of the job your roofer did or something only because my guess is that this roof will never, ever leak. Especially if it is mechanically seamed. Not that I am advocating closed cell but the are medium density foams that might fit the bill.

  5. User avatar GBA Editor
    Martin Holladay | | #5

    Before I give you advice on the best way to insulate, I thought I'd point out that your roofer chose the wrong type of roofing underlayment.

    Like most synthetic underlayments, Feltex synthetic underlayment can only be used over vented attics, not over unvented roof assemblies. See this link: ICC-ES Evaluation report of Feltex synthetic roofing underlayment. That report advises architects and builders (on page 3, in section 5.5): "Installation is limited to roofs with ventilated attic spaces."

    Does this limitation matter? Technically, yes -- failure to follow the manufacturer's installation instructions is a code violation. But practically, probably not.

    If the underlayment ever fails, however, Feltex doesn't have to back up its product, because it wasn't installed according to the manufacturer's instructions.

  6. User avatar GBA Editor
    Martin Holladay | | #6

    Concerning your insulated roof assemblies: Your case is similar to many that show up on our Q&A pages. After the framing is complete, and the roofing has been installed, and the house is dried in, we get the question: "How should I insulate my cathedral ceilings?"

    I hate to say it, but at this point it's too late to insulate your roofs correctly, because you didn't install the needed rigid foam above your roof sheathing. It seems like you know that. Even though you know it, it's still too bad. Rigid foam above the roof sheathing is always the best way too go.

    It sounds like you are leaning toward a flash-and-batt job (or, technically speaking, a flash-and-blow job) using open-cell spray foam. Before you choose that approach, you should read this article: Open-Cell Spray Foam and Damp Roof Sheathing.

    If this were my house, I wouldn't use open-cell spray foam unless there was a way to create a ventilation channel from soffit to ridge under the roofing underlayment, using stiff baffles, before the spray foam is installed. There are just too many reports of damp roof sheathing when open-cell spray foam is installed directly against the roof sheathing.

    Closed-cell spray foam is safer from a moisture perspective, but worse from an environmental perspective (because of the issue of blowing agents with a high global warming perspective). The good news is that there is a new type of spray foam on the market -- a closed-cell spray foam that uses an environmentally friendly blowing agent. It's Lapolla Foam-Lok 2000 4G. You might want to see if any of your spray foam contractors offer it.

  7. Shawn Sanders | | #7


    I appreciate the time you took to answer my question, and in fact you are correct. I already knew it was best to use foam on the outside, but I was talked out of this due to difficulty in installation that the contractor had faced before. I wish now I had subbed out to another I just have to make due with what we have. I knew that the Feltex was not vapor permeable, but fortunately, this standing seam roof is very unlikely to ever leak with no penetrations. I think I'll be OK here.

    You are right, I was leaning toward flash foam and blown in, but the issue came about from Open vs Closed and arguments between foam contractors. I have thoroughly read the Open-Cell Spray Foam and Damp Roof Sheathing article and that's what led me to believe ccSPF was the answer. However, I got such heavy resistance I thought there might be some new studies done that I could show to the Demelac and Icynene dealers. They both have ccSPF available, but didn't want to use it in our area. Is this just ignorance on their part?? The dealer who is recommending ccSPF is very, very high, even when considering the added board ft costs of ccSPF - out of budget I'm afraid. Unfortunately, I don't think the venting is an option as they have already installed the non vented ridge cap. With no escape for the incoming are from the soffit, I'm sure moist warm air would accumulate at the peak. So now I'm stuck trying to convince them ccSPF is the way to go. Lastly for the cathedral ceiling, if I used ccSPF, followed by blow in or Roxul, are there any concerns with using the foil faced poly, like Thermax for example, attached to the bottom of the rafters/trusses to stop thermal bridging?

    As to the walls, do you think open cell in the walls would be OK? I like the idea of the air sealing properties it has, but I have a pretty good air barrier stack up on the exterior. With caulk and spray can I think Roxul or BIBS would also be a good insulation option with out the risks. There are open cell foams from Icynene that are fairly green options from what I have read. What do you think?

  8. User avatar GBA Editor
    Martin Holladay | | #8

    In many ways, you have painted yourself into a corner. I don't like the idea of a foam sandwich -- one with closed-cell spray foam above the mineral wool insulation, and Thermax polyiso under the mineral wool insulation. But I know that people do it, and that this type of roof assembly usually works out OK. That approach does limit thermal bridging through the rafters.

    If you paid attention to airtightness when you installed your wall sheathing and the exterior XPS -- for example, by taping the seams of the OSB and taping the seams of the XPS -- then there really is no reason to install any spray polyurethane foam insulation between your studs. I think that dense-packed cellulose would be the best option there.

    If the OSB and XPS were installed without tape, then you may want to install open-cell spray foam between your studs.

  9. User avatar
    Dana Dorsett | | #9

    With 1" of closed cell on the under side of the roof deck is sufficiently low permeance to protect the roof deck from wintertime moisture, and 2" would be more than is actually necessary. At 4" or more it's ridiculously vapor tight, and there is no drying path for the roof deck. And it's environmentally unfriendly, using more than 2x the polymer of open cell foam, and most of it is blown with HFC245fa, with a very high global warming potential. (Icynene has some water blown 2lb foams that run ~R5/inch, as well as higher R/inch stuff blown with HFC245fa. Lapolla has a much lower impact HFO1234 blown 2lb foam that's almost as benign as the water blown stuff, runs about R6.5/inch.)

    With the 1" ccSPF on the roof deck, if you set your mesh on the 2x8 rafter element you'd have R26 of fiberglass, and ~R6 of ccSPF for R32. If you then sprayed 5" of half pound OPEN cell foam on the interior side of the mesh you'd meet IRC code min R49 with an R17 thermal break over the rafters, and an assembly that can still dry toward the interior.

    Or, skip the fiberglass install 1" polyurethane on the roof deck, and 12" of half-pound foam (in a minimum of 2 lifts, with a cooling off period between applications, but 3 lifts of 4" would be safer.)

    You DON'T want to put vapor-barrier facers like aluminum clad Thermax on the interior side of an unvented roof, EVER. It's a moisture trap. You can get away with that for decades, or it can fail in under 5 years- it's a crap shoot. On the catheddralized ceiling section, hit the roof deck with an inch of closed cell, cut 1.5" wide strips of the 1" Thermax and tack them centered on the rafters Over your mesh before blowing the BIBS, and fill in between the rafter edge strips with compressed split batts.

    On the walls, air-sealing the sheathing to the framing inside each stud bay and going with the BIBs would be higher performance than open cell foam, and a tighter-better fit than you're likely to achieve with rock wool batts. If you used Thermax rather than XPS on the exterior it would be easier to air-seal that layer, using foil tapes, and would be greener due to the much lower impact blowing agent, pentane, instead of the HFC134a used for blowing XPS. As the HFC blowing agent leaks out over a handful of decades that R5 fades to a performance of about R4.2, not that it makes a huge difference- even at it's degraded performance it's still sufficient exterior R for dew point control on a BIBS fiberglass 2x6 wall in a zone 4A climate. Don't forget to caulk between top plates and between the bottom plate & subfloor before closing it up with wallboard.

    The basement might be cheaper to use 2" + 2" EPS, or 1.5" + 1.5" at least. With symmetrical 2" EPS you're at R16.8 compared to R12.5 with your XPS approach. With symmetrical 1.5" you'd be at R12.6, and it would have the thermal mass advantage of the concrete. With 1" XPS on the exterior and 1.5" on the interior you's meet code min, but you'd lose the thermal mass benefit of the wall since most of the R is on the interior. To deal with the thickness transition at the sheathing above and below the foundation sill you can make a strip of EPDM- seam-tape Z-flashing lapped correctly with the housewrap &/or drainboard to divert the water to the exterior side of the EPS. Unlike XPS, EPS will not lose R-value over decades. EPS also blown with pentane, not a climate-damaging HFC, most of which is already gone before it leaves the plant. In some cases the pentane is recaptured at the factory to be burned for process heat. Take the interior EPS all the way down to the footing if the slab hasn't already been poured. At 1.5lbs density EPS typically costs less than 10 cents per R per square foot, compared to 12-13 cents per R-foot for XPS, so it's really about the same or even slightly lower material cost to do the entire interior & exterior with 1.5" of EPS as the proposed 1" full coverage +1.5" half-coverage with XPS.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |