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

Is it advisable to use nail base insulation panels over a roof insulated with closed cell spray foam?

Jodi Gunderson | Posted in Energy Efficiency and Durability on

We are renovating an old railroad station, turning it into a residence and are gearing up to put new roofing on the building. Built in 1888, the building has massive overhangs on all sides and the rafters are mostly 2x6s with 4x6s spaced about 8′ apart supported by large brackets. We had planned to use closed-cell spray foam insulation in order to get the maximum r-value in the relatively shallow rafter bays. Also, we want to maximize the available head room in the second floor rooms. The underside of the roof will also be the ceiling with t&g beadboard installed directly to the rafters. I have had an engineer help out with some of the structural aspects – we have sistered many of the rafters and added some additional rafter ties, etc.

The roofer has concerns with the spray foam. He has seen a lot of dry rot from spray foam insulation with no ventilation under the roof decking. I’ve read here that closed cell spray foam is what to use in an unvented cathedral ceiling application. Not wanting to reduce the headroom upstairs, I decided to look into nail base insulation panels to boost the r-value and reduce thermal bridging. The nail base insulation we’d like to use will have a 1 inch air space below the OSB sheathing so that should help extend the life of the roofing material (most likely asphalt). One of my concerns is with the old roof deck being sandwiched between the cc spray foam and the polyico nail base. I would love to hear some thoughts and suggestions from you folks regarding this particular situation. Our code officer tells me that the min. r-value for roofs here is 38. I’d like to go above the minimum!

The building is located right next to a lake in upstate New York (fingerlakes region). I’m am probably leaving out some more critical information here but I’ve got to stop somewhere. Thanks for any help – and for this website where people can continue to learn about building science and best practices.


  1. GBA Editor
    Martin Holladay | | #1

    An R-38 assembly is OK. An R-49 assembly is even better.

    If I were you, and I were planning to install nailbase above the roof sheathing, I would use open-cell spray foam between the rafters. The open-cell spray foam will allow the roof sheathing to dry inward.

    I wouldn't worry about the open-cell foam allowing outward vapor diffusion -- after all, the nailbase will keep the roof sheathing nice and warm during the winter, so condensation or moisture accumulation in the sheathing is unlikely or impossible.

  2. D Dorsett | | #2

    In a US climate zone 6 climate (most of the Finger Lakes region is zone 6) at least 50% of the total R value needs to be on the exterior of the roof deck if using a more vapor-open insulation on the interior, for residential construction, per the IRC. The presumptive interior moisture levels have a dew point of about 40F. If the humidity is higher than that in winter there needs to be more exterior R.

    With 2x6 framing you get about R20-R22 on between the rafters with open cell foam, so you need at LEAST R20 (R30 would be better) for your exterior nailbase.

    If it's less than that you can still use open cell foam, but you'll need a class-II vapor retarder between the open cell foam & the interior, such as vapor-barrier latex, or a membrane type variable permeance vapro retarder such as Certainteed MemBrain, or Intello Plus.

  3. Jodi Gunderson | | #3

    Many thanks for the information and suggestions. My next question was going to be about the amount of insulation needed above the roof deck.
    Would the vapor retarder be beneficial even with the R20 - R30 nailbase - or only necessary if there isn't the 50% or more R-value above the roof deck? There are quite a few wall partitions on the second floor so I'd need to detail any sort of membrane well.
    Looks like we're going to have pretty tall fascia boards!
    Back to the subject of insulation: about 1/3 of the exterior walls (2x6s again) have blown in cellulose in the stud bays. This was used in part because it was done a few years ago and because much of the interior wallcovering is chestnut beadboard that we wanted to leave undisturbed. We had planned to spray foam in the remaining exposed stud bays as well as the roof. I like the cellulose for its environmentaly friendliness. Now I'm wondering if I should just stick with cellulose for the walls and if possible, in the rafters bays too. Dense packed by professionals preferably (we have an older blower we've used thus far). I doubt that the cellulose I've installed is at it's ideal density. Any suggestions there? Should we still consider the cc spray foam for the walls in order to get the R-value as high as we can there? The siding and exterior trim is finished so it's too late to add insulation on the exterior.
    Man, the things I'd have done differently if I knew what I know now.
    Thanks again for the help!

  4. GBA Editor
    Martin Holladay | | #4

    For more information on the amount of insulation needed above the roof deck, see:
    How to Build an Insulated Cathedral Ceiling.

    If you are in Climate Zone 6, and you are following this method, the code requires you to install at least R-25 above the roof sheathing.

    For still more details on this technique, see How to Install Rigid Foam On Top of Roof Sheathing.

    You wrote, "Looks like we're going to have pretty tall fascia boards!" See the illustration below for a way to keep your fascia size reasonable. [Image credit: Building Science Corp.]

    In my opinion, dense-packed insulation makes more sense than spray foam for use between studs. After all, you'll still have thermal bridging through the studs -- so you don't get much (if any) thermal benefit from the high cost of spray foam.


  5. Kevin Dickson, MSME | | #5

    Just a caveat about conventional nailbase -- the foam always shrinks, causing a 2"-3" bow in the panels. That's not a problem when attached to the structural frame.

    But the nailbase you have described doesn't sound conventional.

  6. GBA Editor
    Martin Holladay | | #6

    1. What is your source for this statement: "The foam always shrinks, causing a 2 in. to 3 in. bow in the panels"? That sounds unlikely to me.

    2. Why do you think that ventilated nailbase is unconventional? Here are some links to these products:

  7. Jodi Gunderson | | #7

    We will most likely be using GAF ThermaCal 1 nailbase panels. Martin, what are your thoughts on dense-pack cellulose in the rafter bays as well?
    Our soffit (if I can call it that) is beadboard attached to the underside of the roof rafters. The roof pitch is 8/12 and, measured on the slope is in excess of 8'. The existing Yankee gutter will be replaced with a new one of course - so perhaps that can be detailed to reduce the fascia height some (perhaps more like a box gutter). Again, thank you for your suggestions! This is very helpful.

  8. GBA Editor
    Martin Holladay | | #8

    If you plan to combine rigid foam (or nailbase) above the roof sheathing with dense-packed cellulose between the rafters, I suggest that you follow the advice given in this article: How to Build an Insulated Cathedral Ceiling.

  9. Jodi Gunderson | | #9

    I hope it's okay that I'm revisiting this topic I asked about a while back. I've read the articles a few times and still find I have questions.
    One thing that I'm unclear of is the spray foam specifics with regard to the finish ceiling. In the article, most often the spray foam type is specified while sometimes it's not. In regard to the beadboard ceiling we will install, will open-cell SPF need an air barrier between it and the ceiling boards? Or is the oc foam a sufficient air barrier? I'd much rather not have to install taped drywall on the ceilings first - mainly because of the extra weight on these already small (1 3/4" x 5 1/2") roof rafters. I understand how a t&g ceiling would leak air - but am I okay with the oc spray foam being the "air barrier"? And I won't want a vapor barrier because the assembly needs to be able to dry to the interior - right? We intend to use a latex paint on the interior beadboard.
    So from the inside out we'd have 3/4" thick beadboard strips attached to the rafters, open-cell spray foam filling (at least 5" thick) the rafter bays, the existing 1-inch-thick roof boards, a layer of GAF TigerPaw synthetic roofing felt, GAF 6 inch thick vented nailbase panel (4 1/2" of polyiso stated as having an LTTR R-value of 26.8), another layer of the synthetic roofing felt, and finally asphalt shingles.
    Does this sound like a safe and efficient assembly? Many thanks again for the advice and articles.

  10. GBA Editor
    Martin Holladay | | #10

    Five inches of open-cell spray foam is a perfectly adequate air barrier. Your roof assembly has several other potential air barriers, including the synthetic roofing underlayment and the nailbase panels, so I wouldn't worry about air leakage if I were you.

    The only comment I'll make about your plan is that the Finger Lakes Region of New York state gets cold in the winter, so you might want to use EPS nailbase instead of polyiso nailbase. (Polyiso doesn't perform wall at cold temperatures.) Here is a link to an article about this issue: In Cold Climates, R-5 Foam Beats R-6.

  11. Jodi Gunderson | | #11

    Great. I will look in to the EPS nailbase. Thanks!

  12. Jodi Gunderson | | #12

    A question regarding the airtight membrane above the roof's board sheathing (synthetic felt): should I plan to tape the seams?

  13. GBA Editor
    Martin Holladay | | #13

    You can tape the seams if you want. Lots of people like redundant air barriers -- "just in case."

    That said, it's really hard to imagine that you will have any air leaks through your roof assembly, even if you don't tape the seams of your roofing underlayment.

  14. Jodi Gunderson | | #14

    Well I seem to have come full circle on the roof insulation. I have read every article I could on the subject and sharpened my pencil till there's almost nothing left. I looked extensively at the eps nailbase offerings. Even had a rep come to the site to discuss details. I definitely agree that eps is the way to go in cold climates, but it's just proving to be too much for our roof and budget. I'm located in zone 5 - so I would need at least a 6" nailbase panel (I was planning on the 7 3/4" panel) to get the min. R20 on top. The detailing needed to make it work and look right just exceeded the practicality of it all. It doesn't help that half the roof is outboard of the building it's covering. And furring up that outboard area with 2xs on edge would add considerable weight I think.
    We will probably end up going with an unvented cathedral ceiling filling the rafter bays with closed cell spf realizing that we will end up with stripes on the roof at certain times of the year. Damn you thermal bridging!
    Still, some questions come to mind. When I was reading Joe L's article about insulating his roof on the Building Science Corporation site, he described using cc spf along with a fully adhered membrane on the old board sheathing. I took that to be a peel-and-stick ice and water shield. Several years later when he finally got around to adding rigid foam above he removed the fully adhered membrane to find the roof boards we still in great shape. This makes me wonder if less rigid foam could be added to the roof deck (say 2 - 4") with the cc spf applied underneath. I know it's been said that it's not wise to have the sheathing sandwiched between two vapor-impermeable layers. Is the problem with that have to do with bulk water intrusion? Would it be unwise for me to use an ice and water barrier on the entire roof deck if we go with an unvented roof? That was something that was suggested to me in the event we do have any ice dams. I wonder how his roof has held up with both the cc spf and 6" of polyiso sandwiching the old roof deck.
    Any more thoughts before I close the case? Thanks.
    By the way, the latest company I had give us an estimate uses Lapolla spray foam which uses a much friendlier blowing agent (with the closed-cell foam). Anyone have experience with this foam? Unfortunately the estimate for this foam was more than double the others.

  15. Expert Member
    Dana Dorsett | | #15

    With as little as 2" of ccSPF the roof deck is fully protected from interior moisture drives in a zone 5 climate even without exterior foam. WITH exterior foam (say an R13-15-ish 4" EPS nailbase panel it would be fine with a nominal inch of ccSPF on the interior, and the total stackup of ccSPF + EPS would be about R20. With interior fiber insulation below the 1" ccSPF, that total R20 is sufficient for dewpoint control for up to R30 of fiber insulation using only standard latex paint as the interior vapor retarder.

    At 1" the vapor retardency of ccSPF is 0.8 - 1.2, the boundary of class-II and class-III vapor retarders. That is sufficiently low to mitigate against moisture accumulation in the roof deck, but is vapor open enough that seasonal drying is not a problem. It is NOT a moisture trap. At 2" it would still have enough drying capacity, but more than that becomes questionable.

    A full R49 of closed cell foam between the rafters IS a moisture trap for the roof decking, since it's vapor redardency is less than 0.15 perms- almost a Class-I vapor retarder (aka "vapor barrier"). It's performance would also be dramatically less than ~R14 above the roof deck with ~R6 ccSPF under the roof deck and R30 of cellulose/rock wool/fiberglass below that, due to the thermal bridging issues. It's by far the least green too, due to the HFC245fa blowing agent.

    It's fine to use Ice & Water Shield on top of the roof deck with or without exterior foam, since roofs don't dry toward the exterior (with the exceptions being tile, slate, cedar shingles, or metal roofing mounted on purlins.) A typical #30 felt + asphalt shingle layup is only about 0.1 perm, and in zone 5 often covered with snow/rain/dew, and is never a reliable drying path (which is why venting under the roof deck is standard.)

    The more R you have above the roof deck, the less likely you are to have ice dam issues since it thermally breaks the bridging rafters. So a 4" EPS nailbase panel solution is far preferable to only 2" or 3".

  16. Jodi Gunderson | | #16

    Dana, thank you for your detailed response. I really appreciate your explanations of the different scenarios. One thing I'll mention is that the roof rafters are not on any sort of layout. They're anywhere from 20 to 26 inches apart - so using batting of any sort wouldn't be feasible. Also, I've only got 5 1/2" of rafter cavity to work with. Furring down will reduce the headroom and create problems with clearances for the stair landing and bathroom fixtures. Could the remainder of the cavity be filled with ocSPF? Or would I be better off with cellulose below the flash coat?
    When you talk about a moisture trap with more than 2" of ccSPF, where is this moisture coming from? Is this moisture coming from above due to vapor drive or outdoor seasonal humidity? Or is it a moisture trap for any potential leaks? Thanks for any more input. The nailbase is back on the table for now - though I'm not sure of the code ramifications if I put less than R20 up there.

  17. GBA Editor
    Martin Holladay | | #17

    Your question -- "where would the moisture come from?" -- is a good one. In fact, many buildings have closed-cell spray foam underneath the roof sheathing and vapor-impermeable roofing above the roof sheathing. As long as the spray foam was installed when the roof sheathing was dry, these roof assemblies will perform well until the roofing begins to leak. Of course, eventually all roofing leaks -- so these roofs are pretty good.

    If you want to install nailbase above the roof sheathing, and closed-cell spray foam under the roof sheathing, you can. This type of assembly needs to be detailed well -- with good roof flashing and attention to details -- but that's true of any roofing job.

    This assembly won't be as forgiving as a roof over a vented unconditioned attic (a type of roof that allows for easy inspection of the roof sheathing). But it won't be any more unforgiving than tens of thousands of similar roofs in the U.S.

  18. user-7061227 | | #18

    Did you apply a waterproof (ice & Water shield) membrane above the nail base or a water resistant membrane?I am think that in my similar application (closed cell below and nail base above) there is no reason at all to use resistant since moisture cannot dry out. Thinking that I should waterproof the entire roof. Thoughts please?

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |