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

1-inch-thick polysio board over roof sheathing at an unvented attic in Boston – okay?

Bradley ODonnell | Posted in Green Building Techniques on

I’m in Boston, climate zone 6, and looking into spraying my 2×6 rafter bays full of closed cell foam when we reroof. This would create up to approx R-30-something in the rafter bay under the existing plank sheathing (1920’s house). To prevent thermal bridging, I’d like to add 1″ of Poly-Iso foil-faced insulation over the existing roof planking/sheathing as well. This detail is shown on the GBA website as appropriate for climate zones 1,2 and 3 (unvented attics, cathedral clgs, etc..).

Though there’s likely to be little water vapor inside this roof assembly (assuming I properly tape the seams of the Poly-Iso boards), I do understand that using the zone 1-3 detail in zone 6 could be still be a bad idea as the dew point here in the winter will be well inside the rigid board and whatever water vapor does get into the assembly could condense on the old roof sheathing, potentially rotting it.

I want to avoid using the recommended zone 6 detail as the two layers of thick poly-iso board over the roof sheathing called for would necessitate paying $$ to attach a new nailing surface over that 3 1/2″ or 4″ of rigid insulation. (As I understand it, roofing shingles can be nailed through 1″ (but not through 31/2″) of rigid board to the existing roof sheathing, and therefore my desired detail doesn’t require a new nailing surface.)

Question: if I put my layer of Grace Ice and Water Shield (or similar) roof waterproofing/underlayment over the existing roof sheathing and under the new 1″ Poly Iso board (as opposed to above the Poly Iso board) would I have prudently dealt with any condensation potential? Would I have safely modified the zone 1-3 detail for zone 6, or does the risk of condensation (trapped between Poly-Iso board and Ice and Water Shield) outweigh the benefits of the additional insulation and thermal bridging prevention? The roof is 12:12 pitched, btw.


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. Keith Gustafson | | #1

    How bout more rigid and less or no spray? It is cheaper and may make up for the plywood cost. Sealing the eaves seems to be the only question or use for spray foam.

    I+W over foam is a waste of money. It does its job when on a hard surface. Just ripped a whole extremely leaky roof off that was 100 percent I+W over foam.

    Of course a 12/12 won't tend to leak

  2. GBA Editor
    Martin Holladay | | #2

    In your climate zone, the 2009 IRC has a minimum prescriptive requirement for ceiling insulation of R-49. I would try my best to acheive that.

    If you fill your rafter bays with closed-cell foam, you won't have to worry about keeping your roof sheathing above the dew point, because closed-cell foam is a vapor barrier and an air barrier. Interior moisture is not going to migrate to your roof sheathing, unless you have a bad spray-foam contractor who makes a major mistake. So you'll be fine. You can install as much or as little rigid foam on top of your roof sheathing as you want.

    I don't like your plan to install asphalt shingles directly on top of polyiso foam, however. I've never heard of this technique, and I doubt than any asphalt shingle manufacturer allows it. In 15 years, when your asphalt shingles are ready for the Dumpster, it's going to be hard to strip the old shingles off your roof without damaging or destroying the foam.

    Do yourself a favor -- spring for a layer of plywood over the foam. (And while you are at it -- buy thicker polyiso.)

    If you really want to omit the plywood, you can install purlins or furring strips instead, parallel to the ridge, and install metal roofing.

  3. Bradley ODonnell | | #3

    Thanks Martin,
    I was thinking about vapor getting in from the outside, not the inside, but of course you're correct, it would be coming from inside during the winter...
    As for shingling directly over insulation boards, I've never seen it, either, but I have less experience than many, so I figured I'd throw it out there. I had the idea myself, but then I saw it here:

    On page 5, figure 2, so I imagined it might be standard practice in some places...

    I'm in Massachusetts now, but I'm coming from California. Out there, if you were to install new plywood over foam insulation over and existing roof of 2x6 rafters and planking, the building inspector would take one look at it and make you re-engineer - and rebuild - your whole roof, based on the added weight of the new plywood. Those kinds of experiences have made me gun-shy about adding onto existing structure like that. Also, I'd have to re-detail the eaves and rakes if I built up a 3 1/2" sandwich of foam and plywood over my roof. That's where my biggest fear lies - $$ and time for new flashing, trim, barge rafters, fascia, etc...

    I'll probably just fill the rafter bays with closed-cell foam (if I can afford that). That will have to do for the cathedral parts. For the small attic areas and knee-wall voids, I'll put rigid foam against the underside of the rafters, spray down from above onto those surfaces (making sure to completely fill the bay) and then blow cellulose in to the remaining space beneath the rigid foam. That'll be approx r-36 in some spots, but beyond R-60 in others, and I'll have stopped all the air movement, if I do it right.


  4. GBA Editor
    Martin Holladay | | #4

    The BSC detail shows two layers of roof sheathing -- one below and one above the rigid foam. The font used for the detail captions is very fuzzy for some reason, but I think the caption on the upper layer of sheathing says, "3/4-inch sheathing over rigid insulation."

  5. Bradley ODonnell | | #5

    It does, indeed. I'm 0 for 2.

    I have a quote to have Icynene blown-in when I re-roof. The quoted price $4000. 5.5" of Icynene open-celled foam is going to give me R21 in the "cathedral" rafter cavity. Any idea what multiple of that price I'd be looking at for closed-cell foam of the same thickness?

    Maybe I'll blown in cellulose, put 2- 1 1/2" layers of foam, a new layer of plywood and then take the savings from not using closed-cell spray foam to fill large voids and re-detail the edges of the roof....


  6. Aj Builder, Upstate NY Zone 6a | | #6

    Bradley, Open cell spray foam is never sprayed as thin as 2.5" for an unvented roof. Talk to your spray companies that really know their work and see what they propose for you. My company sprays both open and closed cell and has been very helpful in coming up with the the insulation package for a project that they think best addresses the all.

    Closed cell has the proper perm rating in thinner lay ups, has a much higher R per inch but also costs more per R in general.

    I have had great success with open cell Icynene and unvented roof specs. Michael Chandler though has us all thinking about other set ups... Robert Riversong has us all liking cellulose.... A friend of mine has great success with double layers of taped foam exterior set ups.

  7. Keith Gustafson | | #7

    Lots of places have shingles directly over 1 inch, even 2 iso, they leak like a sieve.

    Last I checked iso was 50-60 cents a board foot, should be a lot cheaper than spray.

    3 and 3 1/2 inches are convenient thicknesses to run 2 flat or one on edge 2x4 on the edge of the roof to back up the facia

  8. Bradley ODonnell | | #8

    @ AJ Builder: I meant to write two 1 1/2" foam boards, in other words, an assembly of 5.5" of cellulose blown-in between the rafters. I'd blown the cellulose into the rafter bays by removing a run of the old roof planking. Then over that old planking, I'd lay down two 1 1/2" Poly Iso foam boards and then screw a new layer of 1/2 OSB as a nailer in place over it all. That gives a new roof profile of 3 1/2", which, as Keith G said, can be contained on the roof edges with a 2x4 on its edge.

    After I posted it, I realized that using cellulose wouldn't work in the proposed assembly because I started this discussion with an unvented attic/rafter bay in mind.

    @ AJ: could you give a rough cost multiplier difference between Icynene and closed-cell spray in?


  9. Aj Builder, Upstate NY Zone 6a | | #9

    Bradley, best to make a quick call to your area spray companies. Personally I have a love hate relationship with sprayfoam and more so with closed cell. My success and experience to date is with open cell. A really rough multiplier is 50% cost increase for closed, at twice the R per inch installed.

    Huge debate about where to use what spray foam. My guy has used open cell everywhere with success, on basement concrete walls to unvented roofs. Some say closed cell, some say don't trap moisture, you won't find roof leaks.... there are horror pics of rotten walls and roofs on the internet. If you create a building spec that traps moisture that does not dry out.... here come more pics of wildly rotten framing and sheathing. Then Joe Lstiburek shows up and says... silly humans... yaa trapped moisture... bad bad contractor. Don't do that. Do this next time.... yada yada yada....

    Go read for hours and days at Joe's site. Great info.

    Oh and... Icynene does more than just open cell...

  10. Bradley ODonnell | | #10

    Thanks for the link to Joe's site, I think I need to go modify a roof deck insulating detail I submitted last week ... quickly....

  11. Keith Gustafson | | #11

    Am I missing something...with the foam on top shouldn't the cellulose be alright unvented?

  12. GBA Editor
    Martin Holladay | | #12

    You can install dense-packed cellulose in unvented rafter bays, as long as there is a layer of rigid foam with sufficient R-value installed above the roof sheathing.

    According to the 2009 IRC (Section R806.4), it’s possible to build an unvented roof assembly with a combination of rigid foam insulation above the roof sheathing and air-permeable insulation in the rafter bays. The code requires that “rigid board or sheet insulation shall be installed directly above the structural roof sheathing as specified in Table R806.4 for condensation control.” The table calls for a minimum of R-5 foam for Climate Zones 1-3, R-10 for Climate Zone 4C, R-15 for Climate Zones 4A and 4B, R-20 for Climate Zone 5, R-25 for Climate Zone 6, R-30 for Climate Zone 7, and R-35 for Climate Zone 8.

    Brad is in climate zone 6, so that means he'll need R-25 above the roof sheathing. That means a minimum of 4 inches of polyiso, 5 inches of XPS, or 6.5 inches of EPS.

  13. Keith Gustafson | | #13

    Well, case closed, he just needs to plan another inch

    The answer is always more insulation!

  14. Bradley ODonnell | | #14

    Referring to #11, at least according to the California Residential Code, Keith should be correct that the foam can be on the inside - in the rafter bay - if it is thick enough to ensure the dew point is held within the foam. Air-permeable insulation such as cellulose could then fill the remainder of the void.

    From CRC (based on the IRC) for unvented attics, the three approved ways to have an unvented attic are to use:
    5.1. Air-impermeable insulation only. Insulation shall be applied in direct contact with the underside of the structural roof sheathing.
    5.2. Air-permeable insulation only. In addition to the air-permeable insulation installed directly below the structural sheathing, rigid board or sheet insulation with an R-value of R-4 shall be installed directly above the structural roof sheathing for condensation control.
    5.3. Air-impermeable and air-permeable insulation. The air-impermeable insulation shall be applied in direct contact with the underside of the structural roof sheathing for condensation control. The air-permeable insulation shall be installed directly under the air-impermeable insulation.

    5.2 is what Martin is referring to, no? Of course, the R-4 listed ought to give you an idea of just how pleasant it can be to live in California. 5.3 is what Keith is talking about. I assume the IRC also has this third "flash and batt" option?

    Brad (just moved to Boston from California....)

  15. Keith Gustafson | | #15

    I was talking about foam over.........mebee CA is different........seems odd

  16. GBA Editor
    Martin Holladay | | #16

    California definitely has its own residential building code, unlike that of any other state.

    Designing a roof assembly requires paying attention to two types of requirements: building codes and building science (also known as physics). If the physics don't work, it doesn't really matter whether the detail was code-compliant, because the roof assembly can rot.

  17. Bradley ODonnell | | #17

    I re-read the thread and realized that Keith was not talking about 5.3, so-called "Flash and Batt," and that I was the one to introduce that to the discussion, but what danger do you perceive if the layer of sprayed foam between the rafters and under the roof sheathing is thick enough to keep the dew point inside the foam?
    For example, assume a 2 x 12 rafter bay at an unvented cathedral ceiling, say in climate zone 6, so we need an R25 layer of ccSPF (closed cell sprayed polyurethane foam). That layer is sprayed in first, up against the underside of the roof sheathing and is perhaps 4 inches thick (approx R26). That leaves 7 1/4" inches below it to fill with cellulose of fiberglass batts (approx R25). That's an R50+ assembly with the cost of only 4" of closed cell foam, though it does not address thermal bridging through the rafters.

    Perhaps I'm overly wary of placing foam on top of the roof sheathing because it's unfamiliar to me? Option 5.3 would definitely save money in California, but perhaps here in a cold climate, it's just more sensible and cost effective to put rigid foam over the sheathing.

    Why would physics preclude option 5.3? The dew point is inside the closed-cell foam, so water ought not to condense on its inside edge and wet the cellulose/fiberglass. If it did, wouldn't also form on the inside layer edge of the rigid foam over the sheathing in option 5.2? True, then the sheathing gets wet, not the insulation, but presumably water will work through the sheathing and soak the batts/cellulose below, too.
    If we're going to place faith in keeping the dew point inside closed cell foam, does it matter if that foam is adove or below the roof deck?


  18. Bradley ODonnell | | #18

    Looks like Boston is Climate zone 5, so R20...
    I need to get up to speed quickly on building outside the bubble of California where climates and codes alike are unique.
    Thanks for all the help and responses.


  19. GBA Editor
    Martin Holladay | | #19

    If you don't want to install rigid foam on top of your roof sheathing, you don't have to. I list several options for insulating sloped roofs in my article, Creating a Conditioned Attic.

    Here's what the 2009 IRC says about your latest proposed method: it's possible to use a combination of air-impermeable and air-permeable insulation in unvented rafter bays, as long as the minimum R-value of the air-impermeable insulation that is "applied in direct contact with the underside of the structural roof sheathing" meets the requirements for condensation control shown in Table R806.4.
    The table calls for a minimum of R-25 for Climate Zone 6.

  20. Bradley ODonnell | | #20

    Thanks a lot, it's great to know there's a community of people who care enough to do this stuff, and care enough to do it right.


Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |